MD

The amniotic membrane is a multilayer tissue forming the innermost layer of the amniotic sac that surrounds the developing fetus. It is comprised of 5 layers, from the inside out: a single layer of epithelial cells, a thick basement membrane, a compact layer, a fibroblast layer, and a spongy layer that abuts the surrounding chorion (Figure 1).¹

The amniotic membrane serves several functions, including synthesis of growth factors and cytokines, regulation of pH, transport of water and solutes, and provision of a permeable barrier to amniotic macromolecules.²

Amniotic epithelial cells are derived from the pluripotent epiblast at approximately day 8 of gestation. This is well before gastrulation occurs at days 15 to 17, considered the “tipping point” when pluripotent cells differentiate into ectoderm, mesoderm, and endoderm.³ These cells express Oct-4 and Nanog, 2 molecular markers that are indicative of pluripotency.³ Two cell types have been identified in amniotic tissues that possess stem cell-like characteristics: human amniotic epithelial cells and human amniotic mesenchymal stromal cells.⁴ Both of these cell types have demonstrated the ability to differentiate into various cell lineages, including endothelial cells, adipocytes, myogenic cells, neurogenic cells, chondrocytes, tenocytes, and osteogenic cells.^5-7 These previously reported findings indicate that amniotic cells and tissue have the capability to generate mesenchymal tissues.

FDA Classification and Available Forms

The US Food and Drug Administration (FDA) classifies amnion as an allograft tissue under Human Cells, Tissues, and Cellular and Tissue-Based Products (HCT/Ps) 361. To meet criteria, the tissue needs to be minimally manipulated. It is to be for homologous use and cannot be combined with other cells or tissues. There can be no systemic effect or dependence on the metabolic activity of living cells to achieve its primary function. The tissue has to have a localized effect in vivo. Therefore, amnion allograft tissue can be commercialized, provided it is not marketed as a stem cell product or to contain viable cells.

Amniotic tissue is commercially available in several forms.

These include fresh-frozen injectable amniotic liquid that may contain viable amniotic cells and/or particulated amniotic membrane, a micronized freeze-dried (lyophilized) particulate powder that is directly applied to a wound or resuspended for injection, and a cross-linked dehydrated membrane acting as an adhesion barrier (Figure 2).

Safety

Amniotic tissue has been used for over 100 years in burn, ophthalmology, and chronic wound patients with favorable outcomes and no adverse effects reported in the literature. Unlike embryonic stem cells, which may be tumorigenic,⁸ amniotic cells do not possess any known tumorigenicity.⁹ In one study, 50 immunodeficient mice were injected with 1 to 2 million amniotic epithelial cells and observed for a maximum of 516 days with no tumorigenicity observed in any of the animals.¹⁰ In another study, amniotic epithelial cells were implanted into the forearms of healthy volunteers and no immunologic response was observed in any of the recipients.¹¹ Furthermore, viable amniotic cells were recovered via biopsy 7 weeks following transplantation, demonstrating viability of the transplanted cells.¹¹ The lack of tumorigenicity and immunologic response in hosts is due in part to the fact that amniotic cells do not express human leukocyte antigen class II antigens and only express class I antigens in small amounts.³

Advantages of Amnion Tissue

Amniotic tissue is readily available, as it is often discarded after childbirth. The use of this tissue poses no added risk to the fetus or mother, eliminating the ethical concerns associated with obtaining embryonic stem cells. Amniotic tissue is comprised of an extracellular matrix, which acts as a natural scaffold for cellular attachment and structural support for cells as well as collagen types I, III, IV, V, and VI, hyaluronic acid, and a host of growth factors.¹² In addition, it possesses antimicrobial properties, including beta-defensins.¹³

Amniotic tissue has been shown to exert an anti-inflammatory effect by inhibiting the inflammatory cascade. Specifically, it has been shown to inhibit cytokines such as tumor necrosis factor-alpha in the presence of dendritic cells,¹⁴ as well as inhibiting transforming growth factor-beta, interleukin-8, and fibroblast proliferation.¹⁵ These findings indicate that amniotic tissue has the ability to dampen the “cytokine storm” that occurs after an injury in an adult, which would lead to beneficial impacts on healing and scar formation in patients.¹⁶

Basic Science and Animal Studies

Several studies have demonstrated promising outcomes for orthopedic applications in vitro. A comparison of osteogenic potential found that amniotic fluid-derived cells were able to produce approximately 5 times more mineralized matrix than bone marrow-derived mesenchymal stem cells.¹⁷ More recently, Si and colleagues¹⁸ compared the osteogenic potential of human amniotic epithelial cells, amniotic cells, and human bone marrow-derived mesenchymal stem cells. They found that all 3 cell lines were osteogenic, though the amniotic epithelial cells had better immunomodulatory properties and marginally less osteogenic potential than the other 2 cell types. Furthermore, several in vivo animal studies have demonstrated the ability of human amniotic cells to stimulate bone growth in rats,^19,20 rabbits,²¹ and sheep.²²

 

Amniotic tissue also possesses potential for chondrogenesis. Cryopreserved human amniotic membrane cells used for in vitro human osteoarthritis tissue scaffolds did not differentiate in culture, and they integrated and repaired damaged articular cartilage.²³ Various in vitro^24,25 and animal in vivo^26,27 studies have reported similar supportive findings. Kunisaki and colleagues²⁸ used sheep amniotic fluid mesenchymal stem cells to reconstruct lamb tracheal cartilage in utero, concluding that cells obtained from the amniotic fluid possess chondrogenic capabilities. Further in utero lamb studies of cartilage artificial defects, given 7 days to settle before adding a hypocellular matrix as a scaffold, showed chondrocyte density and cell architecture was restored at the defect site after 28 days without the formation of an inflammatory response or scar tissue.²⁹

Amniotic tissue has had similar success in tendon repair studies in vivo.^9,30,31 Barboni and colleagues³² implanted amniotic epithelial cells (AECs) into artificially created sheep Achilles tendon defects in situ, inducing superior structural and mechanical recovery in the defects at a faster rate compared to controls not receiving AECs. Healing via AECs started at the healthy tissue around the borders of the defect and progressed centrally, suggesting recruitment of native progenitor cells to the lesion.³² Kueckelhaus and colleagues³³ investigated the role of amnion-derived cellular cytokine solution in the healing of transections of rat Achilles tendons, reporting improved mechanical properties of healing tendons at early time points compared to controls. Beredjiklian and colleagues³⁴ compared the healing of transected extensor tendons of pregnant ewes and of their fetus in utero, reporting a reparative form of healing with scar formation in adult subjects and regenerative form of healing without scar formation or inflammation in fetal subjects.

Amniotic tissue has properties that prevent adhesion formation around tendons following injury and reconstruction.³⁵ Ozgenel³⁶ investigated the effects of hyaluronic acid and amniotic membrane alone and in combination on the presence of adhesions and the rate of healing following chicken flexor tendon repair. The study found amniotic membrane wrapped around the repaired tendon was superior in preventing adhesion formation. Kim and colleagues³⁷ report a similar reduction in fibrosis and adhesion following application of a human amniotic membrane wrap to rabbit ulnar neurorrhaphy sites.

This barrier function of amniotic tissue has also been investigated in the prevention of surgical scarring and peridural fibrosis in animal models following spinal discectomy. A study in canine models showed a reduction of scarring following the application of cross-linked amniotic membrane compared to freeze dried amniotic membrane.³⁸ Similar reductions in scarring in rat models with the application of freeze-dried amniotic membrane compared to negative controls have been reported.³⁹

Human Studies

A randomized trial investigated the outcomes of prenatal vs postnatal repair of myelomeningocele in humans, finding a reduced need for implanted shunts and improved functional outcomes at 30 months of life in the prenatal intervention group compared to the postnatal group.⁴⁰ This study was concluded early due to the efficacy of prenatal surgery and the benefit of nervous system repair in utero in the presence of amniotic growth factors.

Vines and colleagues⁴¹ performed a 6-patient feasibility study using amnion injections to treat symptomatic knee osteoarthritis. Each patient received a single intra-articular cryopreserved amniotic suspension allograft (ASA) injection and was followed for 1 year. No adverse outcomes were reported, with the only abnormal finding being a small increase in serum immunoglobulin G and immunoglobulin E levels. Intra-articular ASA injection was found to be safe, but a large-scale trial investigating symptomatic relief was recommended.⁴¹

Most of the human studies using amnion pertain to foot and ankle surgery. Its use as a treatment for diabetic foot ulcers and recalcitrant plantar fasciitis was one of the early-recognized successes.^42-45 Zelen and colleagues⁴⁶ investigated the applications of injectable micronized dehydrated human amniotic/chorionic membrane as an alternative to surgical intervention in the treatment of refractory plantar fasciitis. This prospective, randomized trial with 45 patients showed significant improvement in plantar fasciitis symptoms at 8 weeks compared to controls (saline injections). A similar study compared the use of cryopreserved human amniotic membrane (c-hAM) injections to corticosteroid injections in plantar fasciitis patients.⁴⁷ The results indicated that c-hAM is safe and comparable to corticosteroids, with the authors noting that pain improvement was greatest in patients receiving 2 injections of c-hAM at 18 weeks.

Tendon wrapping, in which the amniotic membrane is laid over a tendon repair, has been reported with success. Amniotic membrane is superior to collagen for tendon wrapping as it actively contributes to healing while minimizing adhesions, which collagen alone cannot do.⁴⁸ The membrane serves as a protective sheath around repaired tendons with anti-inflammatory, anti-adhesive, immunomodulatory, and antimicrobial benefits. A 124-patient study demonstrated the safety of using amnion in this manner, and the authors reported a decreased rate of complication compared to previously published data.⁴⁹ Another study of 14 patients undergoing foot and ankle surgery with tendon wrapping reported clinical improvement with reduced pain and greater functional outcomes postoperatively compared to preoperative measurements.⁵⁰

 

Conclusion

Amniotic membrane-derived tissues are safe and non-tumorigenic, producing an abundance of growth factors that have shown promise as tissue scaffolds and as aids in the regeneration of human bone and soft tissues. Amnion applications in orthopedic surgery may be numerous, but development is ongoing. Given the vast array of in vitro and in vivo animal data supporting the benefits of amnion in tissue regeneration, orthopedic surgeons and researchers should place emphasis on conducting clinical studies to validate the safety and efficacy of amniotic cells in the treatment of orthopedic conditions.

Am J Orthop. 2016;45(7):E421-E425. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

The amniotic membrane is a multilayer tissue forming the innermost layer of the amniotic sac that surrounds the developing fetus. It is comprised of 5 layers, from the inside out: a single layer of epithelial cells, a thick basement membrane, a compact layer, a fibroblast layer, and a spongy layer that abuts the surrounding chorion (Figure 1).¹

The amniotic membrane serves several functions, including synthesis of growth factors and cytokines, regulation of pH, transport of water and solutes, and provision of a permeable barrier to amniotic macromolecules.²

Amniotic epithelial cells are derived from the pluripotent epiblast at approximately day 8 of gestation. This is well before gastrulation occurs at days 15 to 17, considered the “tipping point” when pluripotent cells differentiate into ectoderm, mesoderm, and endoderm.³ These cells express Oct-4 and Nanog, 2 molecular markers that are indicative of pluripotency.³ Two cell types have been identified in amniotic tissues that possess stem cell-like characteristics: human amniotic epithelial cells and human amniotic mesenchymal stromal cells.⁴ Both of these cell types have demonstrated the ability to differentiate into various cell lineages, including endothelial cells, adipocytes, myogenic cells, neurogenic cells, chondrocytes, tenocytes, and osteogenic cells.^5-7 These previously reported findings indicate that amniotic cells and tissue have the capability to generate mesenchymal tissues.

FDA Classification and Available Forms

The US Food and Drug Administration (FDA) classifies amnion as an allograft tissue under Human Cells, Tissues, and Cellular and Tissue-Based Products (HCT/Ps) 361. To meet criteria, the tissue needs to be minimally manipulated. It is to be for homologous use and cannot be combined with other cells or tissues. There can be no systemic effect or dependence on the metabolic activity of living cells to achieve its primary function. The tissue has to have a localized effect in vivo. Therefore, amnion allograft tissue can be commercialized, provided it is not marketed as a stem cell product or to contain viable cells.

Amniotic tissue is commercially available in several forms.

These include fresh-frozen injectable amniotic liquid that may contain viable amniotic cells and/or particulated amniotic membrane, a micronized freeze-dried (lyophilized) particulate powder that is directly applied to a wound or resuspended for injection, and a cross-linked dehydrated membrane acting as an adhesion barrier (Figure 2).

Safety

Amniotic tissue has been used for over 100 years in burn, ophthalmology, and chronic wound patients with favorable outcomes and no adverse effects reported in the literature. Unlike embryonic stem cells, which may be tumorigenic,⁸ amniotic cells do not possess any known tumorigenicity.⁹ In one study, 50 immunodeficient mice were injected with 1 to 2 million amniotic epithelial cells and observed for a maximum of 516 days with no tumorigenicity observed in any of the animals.¹⁰ In another study, amniotic epithelial cells were implanted into the forearms of healthy volunteers and no immunologic response was observed in any of the recipients.¹¹ Furthermore, viable amniotic cells were recovered via biopsy 7 weeks following transplantation, demonstrating viability of the transplanted cells.¹¹ The lack of tumorigenicity and immunologic response in hosts is due in part to the fact that amniotic cells do not express human leukocyte antigen class II antigens and only express class I antigens in small amounts.³

Advantages of Amnion Tissue

Amniotic tissue is readily available, as it is often discarded after childbirth. The use of this tissue poses no added risk to the fetus or mother, eliminating the ethical concerns associated with obtaining embryonic stem cells. Amniotic tissue is comprised of an extracellular matrix, which acts as a natural scaffold for cellular attachment and structural support for cells as well as collagen types I, III, IV, V, and VI, hyaluronic acid, and a host of growth factors.¹² In addition, it possesses antimicrobial properties, including beta-defensins.¹³

Amniotic tissue has been shown to exert an anti-inflammatory effect by inhibiting the inflammatory cascade. Specifically, it has been shown to inhibit cytokines such as tumor necrosis factor-alpha in the presence of dendritic cells,¹⁴ as well as inhibiting transforming growth factor-beta, interleukin-8, and fibroblast proliferation.¹⁵ These findings indicate that amniotic tissue has the ability to dampen the “cytokine storm” that occurs after an injury in an adult, which would lead to beneficial impacts on healing and scar formation in patients.¹⁶

Basic Science and Animal Studies

Several studies have demonstrated promising outcomes for orthopedic applications in vitro. A comparison of osteogenic potential found that amniotic fluid-derived cells were able to produce approximately 5 times more mineralized matrix than bone marrow-derived mesenchymal stem cells.¹⁷ More recently, Si and colleagues¹⁸ compared the osteogenic potential of human amniotic epithelial cells, amniotic cells, and human bone marrow-derived mesenchymal stem cells. They found that all 3 cell lines were osteogenic, though the amniotic epithelial cells had better immunomodulatory properties and marginally less osteogenic potential than the other 2 cell types. Furthermore, several in vivo animal studies have demonstrated the ability of human amniotic cells to stimulate bone growth in rats,^19,20 rabbits,²¹ and sheep.²²

 

Amniotic tissue also possesses potential for chondrogenesis. Cryopreserved human amniotic membrane cells used for in vitro human osteoarthritis tissue scaffolds did not differentiate in culture, and they integrated and repaired damaged articular cartilage.²³ Various in vitro^24,25 and animal in vivo^26,27 studies have reported similar supportive findings. Kunisaki and colleagues²⁸ used sheep amniotic fluid mesenchymal stem cells to reconstruct lamb tracheal cartilage in utero, concluding that cells obtained from the amniotic fluid possess chondrogenic capabilities. Further in utero lamb studies of cartilage artificial defects, given 7 days to settle before adding a hypocellular matrix as a scaffold, showed chondrocyte density and cell architecture was restored at the defect site after 28 days without the formation of an inflammatory response or scar tissue.²⁹

Amniotic tissue has had similar success in tendon repair studies in vivo.^9,30,31 Barboni and colleagues³² implanted amniotic epithelial cells (AECs) into artificially created sheep Achilles tendon defects in situ, inducing superior structural and mechanical recovery in the defects at a faster rate compared to controls not receiving AECs. Healing via AECs started at the healthy tissue around the borders of the defect and progressed centrally, suggesting recruitment of native progenitor cells to the lesion.³² Kueckelhaus and colleagues³³ investigated the role of amnion-derived cellular cytokine solution in the healing of transections of rat Achilles tendons, reporting improved mechanical properties of healing tendons at early time points compared to controls. Beredjiklian and colleagues³⁴ compared the healing of transected extensor tendons of pregnant ewes and of their fetus in utero, reporting a reparative form of healing with scar formation in adult subjects and regenerative form of healing without scar formation or inflammation in fetal subjects.

Amniotic tissue has properties that prevent adhesion formation around tendons following injury and reconstruction.³⁵ Ozgenel³⁶ investigated the effects of hyaluronic acid and amniotic membrane alone and in combination on the presence of adhesions and the rate of healing following chicken flexor tendon repair. The study found amniotic membrane wrapped around the repaired tendon was superior in preventing adhesion formation. Kim and colleagues³⁷ report a similar reduction in fibrosis and adhesion following application of a human amniotic membrane wrap to rabbit ulnar neurorrhaphy sites.

This barrier function of amniotic tissue has also been investigated in the prevention of surgical scarring and peridural fibrosis in animal models following spinal discectomy. A study in canine models showed a reduction of scarring following the application of cross-linked amniotic membrane compared to freeze dried amniotic membrane.³⁸ Similar reductions in scarring in rat models with the application of freeze-dried amniotic membrane compared to negative controls have been reported.³⁹

Human Studies

A randomized trial investigated the outcomes of prenatal vs postnatal repair of myelomeningocele in humans, finding a reduced need for implanted shunts and improved functional outcomes at 30 months of life in the prenatal intervention group compared to the postnatal group.⁴⁰ This study was concluded early due to the efficacy of prenatal surgery and the benefit of nervous system repair in utero in the presence of amniotic growth factors.

Vines and colleagues⁴¹ performed a 6-patient feasibility study using amnion injections to treat symptomatic knee osteoarthritis. Each patient received a single intra-articular cryopreserved amniotic suspension allograft (ASA) injection and was followed for 1 year. No adverse outcomes were reported, with the only abnormal finding being a small increase in serum immunoglobulin G and immunoglobulin E levels. Intra-articular ASA injection was found to be safe, but a large-scale trial investigating symptomatic relief was recommended.⁴¹

Most of the human studies using amnion pertain to foot and ankle surgery. Its use as a treatment for diabetic foot ulcers and recalcitrant plantar fasciitis was one of the early-recognized successes.^42-45 Zelen and colleagues⁴⁶ investigated the applications of injectable micronized dehydrated human amniotic/chorionic membrane as an alternative to surgical intervention in the treatment of refractory plantar fasciitis. This prospective, randomized trial with 45 patients showed significant improvement in plantar fasciitis symptoms at 8 weeks compared to controls (saline injections). A similar study compared the use of cryopreserved human amniotic membrane (c-hAM) injections to corticosteroid injections in plantar fasciitis patients.⁴⁷ The results indicated that c-hAM is safe and comparable to corticosteroids, with the authors noting that pain improvement was greatest in patients receiving 2 injections of c-hAM at 18 weeks.

Tendon wrapping, in which the amniotic membrane is laid over a tendon repair, has been reported with success. Amniotic membrane is superior to collagen for tendon wrapping as it actively contributes to healing while minimizing adhesions, which collagen alone cannot do.⁴⁸ The membrane serves as a protective sheath around repaired tendons with anti-inflammatory, anti-adhesive, immunomodulatory, and antimicrobial benefits. A 124-patient study demonstrated the safety of using amnion in this manner, and the authors reported a decreased rate of complication compared to previously published data.⁴⁹ Another study of 14 patients undergoing foot and ankle surgery with tendon wrapping reported clinical improvement with reduced pain and greater functional outcomes postoperatively compared to preoperative measurements.⁵⁰

 

Conclusion

Amniotic membrane-derived tissues are safe and non-tumorigenic, producing an abundance of growth factors that have shown promise as tissue scaffolds and as aids in the regeneration of human bone and soft tissues. Amnion applications in orthopedic surgery may be numerous, but development is ongoing. Given the vast array of in vitro and in vivo animal data supporting the benefits of amnion in tissue regeneration, orthopedic surgeons and researchers should place emphasis on conducting clinical studies to validate the safety and efficacy of amniotic cells in the treatment of orthopedic conditions.

Am J Orthop. 2016;45(7):E421-E425. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

The amniotic membrane is a multilayer tissue forming the innermost layer of the amniotic sac that surrounds the developing fetus. It is comprised of 5 layers, from the inside out: a single layer of epithelial cells, a thick basement membrane, a compact layer, a fibroblast layer, and a spongy layer that abuts the surrounding chorion (Figure 1).¹

The amniotic membrane serves several functions, including synthesis of growth factors and cytokines, regulation of pH, transport of water and solutes, and provision of a permeable barrier to amniotic macromolecules.²

Amniotic epithelial cells are derived from the pluripotent epiblast at approximately day 8 of gestation. This is well before gastrulation occurs at days 15 to 17, considered the “tipping point” when pluripotent cells differentiate into ectoderm, mesoderm, and endoderm.³ These cells express Oct-4 and Nanog, 2 molecular markers that are indicative of pluripotency.³ Two cell types have been identified in amniotic tissues that possess stem cell-like characteristics: human amniotic epithelial cells and human amniotic mesenchymal stromal cells.⁴ Both of these cell types have demonstrated the ability to differentiate into various cell lineages, including endothelial cells, adipocytes, myogenic cells, neurogenic cells, chondrocytes, tenocytes, and osteogenic cells.^5-7 These previously reported findings indicate that amniotic cells and tissue have the capability to generate mesenchymal tissues.

FDA Classification and Available Forms

The US Food and Drug Administration (FDA) classifies amnion as an allograft tissue under Human Cells, Tissues, and Cellular and Tissue-Based Products (HCT/Ps) 361. To meet criteria, the tissue needs to be minimally manipulated. It is to be for homologous use and cannot be combined with other cells or tissues. There can be no systemic effect or dependence on the metabolic activity of living cells to achieve its primary function. The tissue has to have a localized effect in vivo. Therefore, amnion allograft tissue can be commercialized, provided it is not marketed as a stem cell product or to contain viable cells.

Amniotic tissue is commercially available in several forms.

These include fresh-frozen injectable amniotic liquid that may contain viable amniotic cells and/or particulated amniotic membrane, a micronized freeze-dried (lyophilized) particulate powder that is directly applied to a wound or resuspended for injection, and a cross-linked dehydrated membrane acting as an adhesion barrier (Figure 2).

Safety

Amniotic tissue has been used for over 100 years in burn, ophthalmology, and chronic wound patients with favorable outcomes and no adverse effects reported in the literature. Unlike embryonic stem cells, which may be tumorigenic,⁸ amniotic cells do not possess any known tumorigenicity.⁹ In one study, 50 immunodeficient mice were injected with 1 to 2 million amniotic epithelial cells and observed for a maximum of 516 days with no tumorigenicity observed in any of the animals.¹⁰ In another study, amniotic epithelial cells were implanted into the forearms of healthy volunteers and no immunologic response was observed in any of the recipients.¹¹ Furthermore, viable amniotic cells were recovered via biopsy 7 weeks following transplantation, demonstrating viability of the transplanted cells.¹¹ The lack of tumorigenicity and immunologic response in hosts is due in part to the fact that amniotic cells do not express human leukocyte antigen class II antigens and only express class I antigens in small amounts.³

Advantages of Amnion Tissue

Amniotic tissue is readily available, as it is often discarded after childbirth. The use of this tissue poses no added risk to the fetus or mother, eliminating the ethical concerns associated with obtaining embryonic stem cells. Amniotic tissue is comprised of an extracellular matrix, which acts as a natural scaffold for cellular attachment and structural support for cells as well as collagen types I, III, IV, V, and VI, hyaluronic acid, and a host of growth factors.¹² In addition, it possesses antimicrobial properties, including beta-defensins.¹³

Amniotic tissue has been shown to exert an anti-inflammatory effect by inhibiting the inflammatory cascade. Specifically, it has been shown to inhibit cytokines such as tumor necrosis factor-alpha in the presence of dendritic cells,¹⁴ as well as inhibiting transforming growth factor-beta, interleukin-8, and fibroblast proliferation.¹⁵ These findings indicate that amniotic tissue has the ability to dampen the “cytokine storm” that occurs after an injury in an adult, which would lead to beneficial impacts on healing and scar formation in patients.¹⁶

Basic Science and Animal Studies

Several studies have demonstrated promising outcomes for orthopedic applications in vitro. A comparison of osteogenic potential found that amniotic fluid-derived cells were able to produce approximately 5 times more mineralized matrix than bone marrow-derived mesenchymal stem cells.¹⁷ More recently, Si and colleagues¹⁸ compared the osteogenic potential of human amniotic epithelial cells, amniotic cells, and human bone marrow-derived mesenchymal stem cells. They found that all 3 cell lines were osteogenic, though the amniotic epithelial cells had better immunomodulatory properties and marginally less osteogenic potential than the other 2 cell types. Furthermore, several in vivo animal studies have demonstrated the ability of human amniotic cells to stimulate bone growth in rats,^19,20 rabbits,²¹ and sheep.²²

 

Amniotic tissue also possesses potential for chondrogenesis. Cryopreserved human amniotic membrane cells used for in vitro human osteoarthritis tissue scaffolds did not differentiate in culture, and they integrated and repaired damaged articular cartilage.²³ Various in vitro^24,25 and animal in vivo^26,27 studies have reported similar supportive findings. Kunisaki and colleagues²⁸ used sheep amniotic fluid mesenchymal stem cells to reconstruct lamb tracheal cartilage in utero, concluding that cells obtained from the amniotic fluid possess chondrogenic capabilities. Further in utero lamb studies of cartilage artificial defects, given 7 days to settle before adding a hypocellular matrix as a scaffold, showed chondrocyte density and cell architecture was restored at the defect site after 28 days without the formation of an inflammatory response or scar tissue.²⁹

Amniotic tissue has had similar success in tendon repair studies in vivo.^9,30,31 Barboni and colleagues³² implanted amniotic epithelial cells (AECs) into artificially created sheep Achilles tendon defects in situ, inducing superior structural and mechanical recovery in the defects at a faster rate compared to controls not receiving AECs. Healing via AECs started at the healthy tissue around the borders of the defect and progressed centrally, suggesting recruitment of native progenitor cells to the lesion.³² Kueckelhaus and colleagues³³ investigated the role of amnion-derived cellular cytokine solution in the healing of transections of rat Achilles tendons, reporting improved mechanical properties of healing tendons at early time points compared to controls. Beredjiklian and colleagues³⁴ compared the healing of transected extensor tendons of pregnant ewes and of their fetus in utero, reporting a reparative form of healing with scar formation in adult subjects and regenerative form of healing without scar formation or inflammation in fetal subjects.

Amniotic tissue has properties that prevent adhesion formation around tendons following injury and reconstruction.³⁵ Ozgenel³⁶ investigated the effects of hyaluronic acid and amniotic membrane alone and in combination on the presence of adhesions and the rate of healing following chicken flexor tendon repair. The study found amniotic membrane wrapped around the repaired tendon was superior in preventing adhesion formation. Kim and colleagues³⁷ report a similar reduction in fibrosis and adhesion following application of a human amniotic membrane wrap to rabbit ulnar neurorrhaphy sites.

This barrier function of amniotic tissue has also been investigated in the prevention of surgical scarring and peridural fibrosis in animal models following spinal discectomy. A study in canine models showed a reduction of scarring following the application of cross-linked amniotic membrane compared to freeze dried amniotic membrane.³⁸ Similar reductions in scarring in rat models with the application of freeze-dried amniotic membrane compared to negative controls have been reported.³⁹

Human Studies

A randomized trial investigated the outcomes of prenatal vs postnatal repair of myelomeningocele in humans, finding a reduced need for implanted shunts and improved functional outcomes at 30 months of life in the prenatal intervention group compared to the postnatal group.⁴⁰ This study was concluded early due to the efficacy of prenatal surgery and the benefit of nervous system repair in utero in the presence of amniotic growth factors.

Vines and colleagues⁴¹ performed a 6-patient feasibility study using amnion injections to treat symptomatic knee osteoarthritis. Each patient received a single intra-articular cryopreserved amniotic suspension allograft (ASA) injection and was followed for 1 year. No adverse outcomes were reported, with the only abnormal finding being a small increase in serum immunoglobulin G and immunoglobulin E levels. Intra-articular ASA injection was found to be safe, but a large-scale trial investigating symptomatic relief was recommended.⁴¹

Most of the human studies using amnion pertain to foot and ankle surgery. Its use as a treatment for diabetic foot ulcers and recalcitrant plantar fasciitis was one of the early-recognized successes.^42-45 Zelen and colleagues⁴⁶ investigated the applications of injectable micronized dehydrated human amniotic/chorionic membrane as an alternative to surgical intervention in the treatment of refractory plantar fasciitis. This prospective, randomized trial with 45 patients showed significant improvement in plantar fasciitis symptoms at 8 weeks compared to controls (saline injections). A similar study compared the use of cryopreserved human amniotic membrane (c-hAM) injections to corticosteroid injections in plantar fasciitis patients.⁴⁷ The results indicated that c-hAM is safe and comparable to corticosteroids, with the authors noting that pain improvement was greatest in patients receiving 2 injections of c-hAM at 18 weeks.

Tendon wrapping, in which the amniotic membrane is laid over a tendon repair, has been reported with success. Amniotic membrane is superior to collagen for tendon wrapping as it actively contributes to healing while minimizing adhesions, which collagen alone cannot do.⁴⁸ The membrane serves as a protective sheath around repaired tendons with anti-inflammatory, anti-adhesive, immunomodulatory, and antimicrobial benefits. A 124-patient study demonstrated the safety of using amnion in this manner, and the authors reported a decreased rate of complication compared to previously published data.⁴⁹ Another study of 14 patients undergoing foot and ankle surgery with tendon wrapping reported clinical improvement with reduced pain and greater functional outcomes postoperatively compared to preoperative measurements.⁵⁰

 

Conclusion

Amniotic membrane-derived tissues are safe and non-tumorigenic, producing an abundance of growth factors that have shown promise as tissue scaffolds and as aids in the regeneration of human bone and soft tissues. Amnion applications in orthopedic surgery may be numerous, but development is ongoing. Given the vast array of in vitro and in vivo animal data supporting the benefits of amnion in tissue regeneration, orthopedic surgeons and researchers should place emphasis on conducting clinical studies to validate the safety and efficacy of amniotic cells in the treatment of orthopedic conditions.

Am J Orthop. 2016;45(7):E421-E425. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

Injury of the anterior cruciate ligament (ACL) is very common with over 200,000 annual injuries in the United Status.^1,2 There is a general consensus that these injuries should not be treated conservatively in patients that are younger, or who wish to remain active.^3,4 Reconstructive surgery is currently the preferred treatment in these patients, and anatomic single-bundle reconstruction with autografts is considered the gold standard.^5,6

Reconstruction of the ACL is, however, not a perfect treatment. Following single-bundle autograft reconstruction, revision rates of 3% to 8%,^6-9 contralateral injury rates of 3% to 8%,^10,11 and infection rates of 0.5% to 3%^7,12,13 have been reported. Furthermore, due to the invasive nature of graft harvesting and the surgical procedure, 10% to 25% of the patients are not satisfied following ACL reconstruction.^14,15 This can often be explained by common complaints, such as anterior knee pain (13%-43%), kneeling pain (12%-54%), quadriceps muscle atrophy (20%-30%),^16,17 and loss of range of motion (ROM) (12%-23%).^7,9,18,19 Furthermore, as a result of the invasive nature of reconstructive surgery, revisions can be difficult due to complications, such as tunnel widening, tunnel malpositioning, and preexisting hardware.^20-22 This can lead to inferior outcomes and higher rates (13%) of revision surgery compared to primary reconstruction.^23-26 Finally, reconstructive surgery does not restore native kinematics of the ACL,^27-29 which may partially explain why reconstructive surgery has not been shown to prevent osteoarthritis.^28-31

Over the past decades, there has been an increasing interest in the preservation of the ACL in an attempt to ameliorate these issues.^32-37 Ligament preservation focuses on preserving the native tissues and biology, while minimizing the surgical morbidity to the patients.

Some authors have recently reported on arthroscopic primary repair of proximal ACL tears in which the ligament is reattached onto the femoral wall using modern-day suture anchor technology.^32,38 Others have augmented this repair technique with an internal brace^39,40 or with a synthetic device.^33,41 When performing primary repair, it is believed that proprioception is maintained,^42-44 while experimental studies have suggested that primary repair also restores the native kinematics,⁴⁵ and may prevent osteoarthritis.⁴⁶ Furthermore, primary repair is a conservative approach in that no grafts need to be harvested, no tunnels need to be drilled, and revision surgery, if necessary, is more analogous to primary reconstructions.³²In patients with partial tears, some surgeons have advocated preserving the anteromedial (AM) or posterolateral (PL) bundle and performing selective single-bundle augmentation.^34,35 In addition, several authors have used remnant tensioning^36,47 or remnant preservation^37,48 in combination with reconstructive surgery in order to benefit from the biological characteristics of the remnant. These techniques lead to better proprioceptive function,^44,49,50 vascularization and ligamentization of the graft,^50-52 provide an optical guide for anatomic tunnel placement,⁵³ and decrease the incidence of tunnel widening.^54,55The feasibility and applicability of these surgical techniques mainly depends on the tear type and tissue quality of the torn ligament. In this article we (I) discuss the history of ACL preservation, (II) discuss how modern advances alter the risk-benefit ratio for ACL preservation, and (III) propose a treatment algorithm for ACL injuries that is based on tear location and tissue quality.

History of ACL Preservation

The history of the surgical treatment of ACL injuries started in 1895 when Robson⁵⁶ treated a 41-year-old male who tore both cruciate ligaments from the femoral wall. Performing primary repair with catgut ligatures, both cruciate ligaments were preserved and the patients had resolution of pain symptoms and full function at 6-year follow-up. Over the following decades, Palmer^57,58 and O’Donoghue^59,60 further popularized open primary repair for the treatment of ACL injuries, and this technique was the most commonly performed treatment in the 1970s and early 1980s.^61-65 The initial short-term results of primary repair were excellent,^61,62 but Feagin and Curl⁶⁶ were the first to note that the results deteriorated at mid-term follow-up. Despite improvements in the surgical technique of repairing the ACL, such as the usage of nonabsorbable sutures and directly tying the sutures over bone,^63,67 the results remained disappointing at longer-term follow-up.^68-70

In response to these disappointing results, surgeons sought to improve the surgical treatment by either augmenting the primary repair with a semitendinosus, a patella tendon graft or an augmentation device,^71-74 or by performing primary reconstruction.^75-77 At the end of the 1980s and early 1990s, several randomized and prospective clinical trials were performed in order to compare the outcomes of these techniques.^74,78-82 Many studies showed that results of augmented repair were more reliable when compared to primary repair, which led to the abandonment of primary repair in favor of augmented repair, and eventually primary reconstruction.⁶⁵

 

The Important Role of Tear Location in Ligament Preservation

When taking a closer look at the outcomes of primary repair and augmented repair, it seems that the results of these preservation techniques were not as disappointing as was suggested. This can be explained, in large part, by the fact that the important roles of tear location and tissue quality were not widely recognized.

Sherman and colleagues⁷⁰ reported in 1991 their mid-term results of open primary repair. Similar to others, they noted a deterioration of their results at mid-term follow-up. However, they uniquely performed an extensive subgroup analysis in order to find an explanation for this. In their study, considered a landmark paper on primary repair,^65,70 they concluded that, “poor tissue quality is typical for midsubstance tears and that a repair of these injuries will predictably fail while type I tears (proximal), with better tissue quality, show a definite trend towards better results.”⁷⁰ With these findings, they confirmed the findings of others that had recognized a trend of better outcomes with proximal tears.^64,67,83-85

A majority of the historical studies that were published before 1991 had not considered the role of tear location and tissue quality on outcomes of open primary repair. This was also true for the aforementioned randomized studies that compared primary repair with augmented repair and primary reconstruction. Because these studies randomized patients and did not take tear location into account, it can be expected that patients with midsubstance tears were included in the cohorts of primary repair and the outcomes of these studies were therefore confounded.^74,78-82 If these studies would have been aware of the role that tear location plays on primary repair outcomes, different outcomes may have been found and different conclusions on the optimal treatment for different tear types may have been drawn.⁸⁶

Open Primary ACL Repair Outcomes Stratified by Tear Location

When reviewing the literature of open primary repair outcomes stratified by tear location, it is noted that multiple studies reported excellent outcomes following primary repair of proximal ACL tears.^{73,83,84,87-90} Weaver and colleagues⁶⁴ were among the first to stratify their results by tear location, and they found that more patients with proximal tears (52 of 66; 79%) were satisfied after the procedure when compared to patients with midsubstance tears (3 of 13; 23%) at 3.5-year follow-up. They concluded that, “selection can be made with some predictability of the type of injury to the ligament as to which patients will do better.”⁶⁴ Kühne and colleagues⁸⁹ reported the outcomes of 75 patients with proximal tears treated with open primary repair and noted no failures, negative pivot shift in 88% of patients, stable or nearly stable Lachman test in 87% of patients, and 89% return to sports rate at 4-year follow-up. Raunest and colleagues⁹¹ reported a negative pivot shift and negative anterior drawer test in 84%, return to sports in 71%, and satisfaction in 75% of 51 patients that underwent open primary repair of proximal tears at 3.5-year follow-up.

Interestingly, and in contrast to the findings of Feagin and Curl,⁶⁶ no deterioration of the outcomes at mid-term follow-up was noted in patients with proximal tears. Genelin and colleagues⁸⁸ reported their results of 42 patients with proximal tears treated with open primary repair at 5- to 7-year follow-up. They found a negative pivot shift in 81%, stable or nearly stable Lachman test in 81%, and patient satisfaction in 86% of patients. Similarly, Bräm and colleagues⁸⁷ found good results at mid-term follow-up with a good-excellent Lysholm score in 79%, return to a similar level of sports in 76%, stable or nearly stable Lachman test in 91%, and anterior drawer test in 94% of patients, along with an 88% satisfaction rate and 7% failure rate in patients who underwent open primary repair of proximal tears.

On the contrary, when the outcomes of studies that performed open primary repair in mainly, or only, patients with midsubstance tears are reviewed, significantly inferior results are found. Frank and colleagues⁹² reported outcomes in 42 patients with midsubstance tears at 4-year follow-up. They reported that 56% had a stable or nearly stable anterior drawer test, 78% had a positive pivot shift, and that only 61% were satisfied with the procedure. Odensten and colleagues⁷⁸ reported outcomes of open primary repair in a subgroup of 22 patients with midsubstance tears at 1.5-year follow-up, and noted a 14% failure rate.

When reviewing the mid-term results in patients with midsubstance tears, it seems that there was more deterioration in outcomes.^69,70 Firstly, the aforementioned study by Sherman and colleagues⁷⁰ showed poor results in the patients with (type IV) midsubstance tears at mid-term follow-up. Furthermore, Kaplan and colleagues⁶⁹ reported the mid-term outcomes of 70 patients, of which 56 patients had midsubstance tears. After having reported good outcomes at short-term follow-up,^63,67 they noted that 42% of patients had >3 mm anteroposterior stability when compared to the contralateral leg, only a 62% return to sport rate, and a 17% failure rate. They concluded that, “Although … primary repair of the anterior cruciate may work in some patients, it is an unpredictable operative procedure.”

These studies showed that the outcomes of open primary repair were significantly better in patients with proximal ACL tears and sufficient tissue quality when compared to midsubstance tears. This suggests that open primary ACL repair may have been prematurely abandoned as a treatment option for patients with proximal tears.

 

Augmented ACL Repair

There were several reasons why augmented repair became the preferred treatment in the early and mid 1990s. First of all, the results of augmented repair were more consistent compared to primary repair in the aforementioned randomized and prospective studies,^74,78-82 which is not surprising given the fact that the role of tear location was not widely recognized at the time. Secondly, in the 1970s and early 1980s, patients were treated postoperatively in a cast for 6 weeks, which led to problems, such as loss of ROM, pain, and decreased function.^93,94 At the end of the1980s and 1990s, the focus shifted from prolonged joint immobilization towards early postoperative ROM.^95-97 Since many authors believed that primary repair of the ACL was not strong enough to tolerate early mobilization, an augmentation was added to the technique in order to fortify the repair and enable early ROM.⁹⁸

Interestingly, augmented repair, which is essentially a combination of primary ACL repair and ACL reconstruction, was mainly performed in the 1990s and many surgeons did recognize the role of tear location in this treatment at this point.^73,98-103 In these years, the treatment algorithm consisted of augmented ACL repair in patients with proximal tears in the acute setting and ACL reconstruction in patients with midsubstance or chronic tears. Several different augmentation techniques were used to reinforce the primary repair in these years including autograft tissues (semitendinosus tendon,^102-104 patellar tendon,¹⁰⁰ or iliotibial band [ITB]¹⁰⁵) synthetic materials (polydioxanone [PDS],^101,102,106 carbon fibre,⁷⁴ and polyester [Trevira]⁹⁷), augmentation devices (Kennedy Ligament Augmentation Device [LAD]^98-100) and extra-articular augmentations.⁷³

When reviewing the outcomes of augmented repair of the ACL, good to excellent results can be found in studies that used this technique in patients with proximal tears.^73,98-106 Kdolsky and colleagues⁹⁸ were in one of the first groups that reported their results of augmented repair in only patients with proximal tears. In 1993, they reported their mid-term outcomes (5 to 8 years) in 66 patients who underwent primary repair and augmentation with the Kennedy LAD and found that 97% of patients had stable knees (<3 mm on KT-1000 examination), 98% had a negative pivot shift, and 76% returned to previous level of sports. However, often-reported problems with the augmentation devices were found in this study with rupture of the device (12%) and decreased ROM (14%).⁹⁸ In 1995, Grøntvedt and Engebretsen¹⁰⁰ compared augmentation with the Kennedy LAD to patellar tendon augmentation in a randomized study of patients with acute proximal tears. They noted that 50% of the patients in the Kennedy LAD group had a positive pivot shift compared to 23% in the patellar tendon group. Furthermore, they found KT-1000 leg differences of <3 mm in 92% of the patellar tendon group and 54% of the Kennedy LAD group. Because the authors found significant differences between both groups at 1- and 2-year follow-up, they stopped the clinical trial.

Several authors in the following years reported good results of augmented repair using autograft tissues. Natri and colleagues¹⁰⁵ reported the outcomes of 72 patients treated with primary repair of proximal tears augmented with the ITB at 3.5-year follow-up. They found 89% negative pivot shift rate, 93% stable or nearly stable Lachman test, 99% stable or nearly stable anterior drawer test, 79% satisfaction rate, and 91% return to previous level of sports rate. Krueger-Franke and colleagues¹⁰⁴ reported the outcomes of primary repair of proximal tears with augmentation using the semitendinosus tendon. In a retrospective study of 76 patients, they noted that 96% of patients had a negative pivot shift, 75% of patients had stable or nearly stable Lachman test, 93% were satisfied with the procedure, a mean Lysholm score of 92, a Tegner score that only decreased from 7.2 to 7.1, and KT-1000 testing with 78% <4 mm leg difference with the contralateral leg. The authors concluded that patients with femoral ruptures could be treated with augmented repair when performed in the acute setting. As this study was published in 1998, they stated that magnetic resonance imaging and arthroscopy could be helpful in identifying the tear location.

Final Abandonment of ACL Preservation

Reviewing these outcomes raises the question as to why these techniques were ultimately abandoned in the treatment algorithm of proximal ACL injuries, especially given the aforementioned advantages of ACL preservation. One of the possible answers can be found in a landmark study on ACL reconstruction and rehabilitation published by Shelbourne and colleagues¹⁰⁷ in 1991. At that time, arthrofibrosis and knee stiffness were frequently reported problems following ACL surgery, which could partially be explained by the standard conservative rehabilitation using postoperative joint immobilization.^67,70,80,88

 

Shelbourne and colleagues¹⁰⁷ aimed to assess the cause of arthrofibrosis and knee stiffness, and divided the patients into groups by number of days between injury and surgery (<7, 7 to 21 days, and >21 days between injury and surgery). Furthermore, patients within these groups underwent either a conventional or accelerated rehabilitation program. The authors not only found that patients undergoing accelerated rehabilitation had less arthrofibrosis, but they also noted that less arthrofibrosis was seen when surgery was delayed. These findings, however, contrasted with the general perception that the ACL should be repaired in the first 3 weeks postinjury to ensure optimal tissue quality with an augmented approach. As a result, the treatment of ACL injuries shifted towards ACL reconstruction after these findings. Krueger-Franke and colleagues¹⁰⁴ commented on the trend after the study of Shelbourne and colleagues:¹⁰⁷ “Less consideration has been given to the importance of the proprioceptive receptors in the tibial remnants of the torn ACL and the value of their preservation as part of a primary reconstruction.”

In addition to the trend away from an augmented repair approach due to the novel understanding of the importance of early mobilization, some discussion should focus on the technical limitations of arthroscopy at that time. While arthroscopy had been around for several decades, fluid management and arthroscopic instrumentation was slow to develop. All of the repair and augmentation techniques previously discussed had been performed via an open arthrotomy. Arthroscopic technologies of the time were not refined enough to enable surgeons to perform such complex, intra-articular techniques that would enable suturing of the ligament remnant. In this regard, arthroscopic ACL reconstruction was a much simpler technique to accomplish, and this also likely contributed to the final abandonment of the ligament preservation approach.

Role for ACL Preservation with Modern Advances

As stated in the introduction, there has been a recent resurgence of interest in preservation of the native ligament.^32-37 With the passage of time, many technologic advances have been made, which has allowed surgeons to reconsider the concept of ligament preservation.

First of all, appropriate patient selection was not applied historically, as the critical factors of tear location and tissue quality were not recognized in the era of open primary repair. In modern days, however, advances such as MRI have been developed, which can give the surgeon an idea of the status, and tear type of the ACL pre-operatively.¹⁰⁸ This may help the orthopaedic surgeon to plan the surgery and make an assessment as to whether ACL preservation is possible. Secondly, in the historic literature the postoperative regimen consisted of casting for 5 or 6 weeks,^67,70,80,88 while the focus later shifted towards early ROM.^95-97Modern day ACL rehabilitation focuses on immediate ROM to avoid the complications stiffness, pain and decreased function that plagued the outcomes when immobilization was used.^93,94 Thirdly, historically small tunnels were drilled with primary repair and sutures had to be tied over bone,^57,67 whereas currently suture anchors are available that prevent the need for tunnel drilling and enable direct suture tensioning.^32,38 Finally, and most importantly, in the historic literature patients were treated with an invasive arthrotomy technique, while modern day arthroscopic techniques readily enable the surgeon to effectively suture the remnant arthroscopically. Interestingly, in 2005, in their 20-year follow-up of primary repair surgeries, Strand and colleagues¹⁰⁹ stated, “if the same results could be accomplished by a smaller, arthroscopic procedure, primary repair might reduce the number of patients needing later reconstructions with small ‘costs’ in the way of risk and inconvenience for the patients. We therefore believe that further research and development of methods for closed (arthroscopic) repair are justified.”

Altered Risk-Benefit Ratio

Historically, the treatments of open primary repair and open ACL reconstruction were both invasive surgeries with an arthrotomy, drilling of bone tunnels, and postoperative joint immobilization for 4 to 6 weeks. However, with the modern-day advances, the risk-benefit ratio of both treatments has changed, as Strand and colleagues¹⁰⁹ had already suggested. Although ACL reconstruction can be performed arthroscopically, it remains an invasive procedure, in which tunnels are drilled, patellar tendons or hamstring tendons are harvested, and complications, such as knee pain and quadriceps atrophy, are common. The surgery of primary ACL repair, however, has benefited significantly from the modern developments.^32,38 Primary ACL repair can now be performed arthroscopically, and by using suture anchors no tunnels need to be drilled and the remnant can be tensioned directly. An additional benefit of the use of suture anchors is that revision surgery of a failed primary repair is analogous to primary reconstruction, whereas revision surgery of a failed ACL reconstruction can be problematic due to tunnel widening, tunnel malpositioning, and preexisting hardware.^20-22

 

Reviewing the differences between arthroscopic primary ACL repair and ACL reconstruction, it becomes clear that primary repair has benefited significantly from the modern advances and that the risk-benefit ratio for primary repair has been altered. This means that patients with proximal tears can be treated with a relatively straight forward, minimally invasive surgery, which has been shown to be effective in 85% to 90% of patients.^32,38

Treatment Algorithm Based on Tear Location

Since 2008, in the practice of the senior author (GSD), the surgical treatment algorithm for ACL injuries is completely based on the tear location and tissue quality of the ligament.^110,111 To describe the different tear types, we use the modified Sherman classification in which we extended his classification towards the tibial side whereas Sherman and colleagues⁷⁰ only described the femoral side of the tears (Figures A-F, Table).

In this section, we will discuss the different tear types that are seen and the corresponding treatments that can be used to treat these injuries (Table). Furthermore, we discuss current research on these topics and the reported outcomes of these techniques.

We will not provide the incidence of different tear types as the senior author’s practice is biased towards primary repair.

Type I Tears: Primary Repair

Type I tears are soft tissue avulsion type tears that can be easily treated with arthroscopic primary repair.¹⁰⁷ The length of the distal remnant has to be at least 90% and the tissue quality has to be good to excellent in order to approximate the remnant towards the femoral wall (Table).¹¹² The incidence of type I tears was 26% in the study of Sherman and colleagues,⁷⁰ although recent studies showed a lower incidence (6% to 10%) in a larger population.^32,38 Certainly, individual practices will see different percentages of type I tears based upon the mix of injury mechanisms they see most frequently. Over the last 2 years, with the recognition of the importance of tear type and tissue quality, there has been a renewed interest in arthroscopic primary ACL repair.^32,38

DiFelice and colleagues³² were the first to arthroscopically perform primary repair of the ACL in proximal tears using suture anchors. They reported the outcomes of the first 11 consecutive patients that underwent primary repair in a previously described technique.¹¹³ At mean 3.5-year follow-up, they noted only 1 failure (9%) due to re-injury; mean Lysholm score of 93.2; mean modified Cincinnati score of 91.5; pre- and postoperative Tegner score of 7.3 and 6.9, respectively; SANE score of 91.8; and subjective International Knee Documentation Committee (IKDC) score of 86.4. Of the patients with an intact repair, 9 patients had an objective IKDC rating A and 1 patient had B and all patients had KT-1000 leg differences of <3 mm with the contralateral side (three patients were not available for KT-1000 testing). The authors concluded that arthroscopic primary ACL repair could achieve short-term clinical success in a selected group of patients with proximal avulsion tears and excellent tissue quality. They further noted that mid-term outcomes are necessary given that the results of open primary repair deteriorated at longer-term follow-up in the historical literature. Recently, the senior author (GSD) has added an Internal Brace (Arthrex) to the primary repair with the goal of protecting the ligament in the first weeks to further promote healing of the ligament.^39,40,114

More recently, Achtnich and colleagues³⁸ compared the treatment of arthroscopic primary ACL repair with primary ACL reconstruction in 41 patients with type I tears at 2.3-years follow-up. Twenty-one patients consented for primary repair while 20 patients declined this procedure and underwent primary reconstruction. They noted no significant differences in Lachman test, pivot shift test, objective IKDC score, and KT-1000 scores. Although not significant, the clinical failure rate in the primary repair group (15%) was higher than the reconstruction group (0%). Interestingly, despite the higher failure rate in the repair group, the authors concluded that primary ACL repair is recommended in a carefully selected group of patients with type I tears and excellent tissue quality, which can likely be explained by the differences in the risk-benefit ratio between both procedures.

Over the last decade, the research group led by Murray^46,115,116 has performed experimental research on primary repair with a biological scaffold and reported many interesting findings that could be extrapolated to primary ACL repair. First of all, they compared bioenhanced primary repair with bioenhanced primary reconstruction in 64 Yucutan pigs and noted that there was significantly less macroscopic cartilage damage in the primary repair group at 1-year follow-up.⁴⁶ They concluded that bioenhanced ACL repair may provide a new, less invasive treatment option that reduces cartilage damage following joint injury. This may suggest that primary repair may have a lower incidence of osteoarthritis when compared to ACL reconstruction, which is interesting as osteoarthritis is very common after ACL reconstruction. Further research in this area is certainly warranted.

In another study they compared bioenhanced primary repair in juvenile, adolescent and mature Yucutan pigs and noted that functional healing depended on the level of skeletal maturity with immature animals having a more productive healing response.¹¹⁶ This indicates that primary repair might be a good treatment option in skeletally immature patients, especially since reconstruction increases the risk of premature closure of the epiphysis^117,118 and delaying treatment increases the risk of meniscus injury.¹¹⁹ Interestingly, a recent meta-analysis showed indeed that the risk of epiphysis closure was lower in primary repair when compared to ACL reconstruction and the rupture rate was also lower.¹¹⁸ Primary repair may be a good treatment option in children as the procedure has all the attributes that should be applicable to children: it is minimally morbid, tissue sparing, and it is a conservative approach that does not burn any surgical bridges for future reconstructive surgery if necessary.

Finally, the research group of Murray¹¹⁵ assessed the effect of surgical delay of primary repair following injury in Yucutan pigs and noted that better biomechanical outcomes were noted after delaying surgery for 2 weeks when compared to 6 weeks. This suggests that primary repair should preferably be performed in the acute setting, which has also been shown in historical studies since the ligament in the acute setting has optimal tissue quality and the ligament is less likely to be retracted or reabsorbed.^59,60,115

 

One Bundle Type I Tears: Single Bundle Augmented Repair

In some cases, the tear locations of the AM and PL bundle are not at the same location and Zantop and colleagues¹²⁰ reported in an arthroscopic study that this could be as frequent as in 30% of all complete tears. In some of these tears, one of the bundles can be avulsed of the femoral wall (type I tear) while the other bundle is not directly repairable (non-type I tear). In these cases, the senior author (GSD) will repair the type I tear bundle, whereas a hamstring augmentation is placed at the location of the other bundle. When reviewing the literature, a combination of primary repair of one bundle and reconstruction of the second bundle has not been described before. However, over the last decade several surgeons have performed augmentation of one bundle in the setting of partial tears.^{34,35,121-124}

Buda and colleagues³⁴ were the first to perform selective AM or PL bundle reconstruction in the setting of partial tears.³⁴ At 5-year follow-up, they reported no reruptures and only 1 patient with an IKDC C-score, although reoperation was necessary in 4 out of 47 patients (9%). Following this publication, many others reported on selective bundle reconstruction.^35,121-124 However, with partial tears, the knee is often stable and a selective augmentation technique is utilized to prevent complete rupture of the ligament. The application of this technique is essentially different from reconstruction for complete ACL tears in which the knee is unstable, there is a giving way sensation and patients have problems participating in sports.

Type II Tears: Augmented Repair

Type II tears often have good or excellent tissue quality and can be pulled up towards the femoral footprint, but are too short to be firmly attached. Sherman and colleagues⁷⁰ reported that approximately 22% patients had a type II tear, which corresponds to a tear located in the proximal part of the ligament. With this technique, multiple suture passes are used to stitch the remnant and, in addition, a smaller hamstring autograft or allograft is passed through the middle of the tibial remnant. A suture button is used proximally for the graft, and the tensioning repair sutures through the remnant are also passed through the suture button. The suture button is passed through the femoral tunnel and flipped so that the graft is proximally fixed. Then, the repair sutures of the remnant are tensioned, and the ligament is pulled towards the femoral wall as a sleeve around the graft. When the ligament is approximated to the femoral wall, the sutures are tied over the suture button. The graft is then tensioned distally to complete the augmented repair.

In the recent literature, the technique of augmentation of a primary repair using autograft tissue has not been reported. However, augmented repair using an internal brace^39,40 or augmentation devices^33,41 have been recently performed. MacKay and colleagues³⁹ reported good outcomes of arthroscopic primary repair of proximal tears using an internal brace. Eggli and colleagues³³ reported the results of the first 10 patients treated with ACL preservation using primary repair of the ligament with the addition of a dynamic screw-spring mechanism. The authors reported good preliminary results with one failure (10%) and good objective and subjective outcomes. In a next study, they reported the outcomes of 278 patients and although they reported good clinical outcomes and a revision rate of 4%, the reoperation rate for removal of the screw-spring mechanism was high (24%).⁴¹ This is not surprising when reviewing the historical literature in which high complication rates of the augmentation devices were reported.^99,100 We were unable to identify any other studies reporting surgical techniques of augmenting primary repair in the literature.

Type III Tears: Reconstruction With Remnant Tensioning

In patients with type III tears, the ligament cannot be approximated to the wall and reconstruction is necessary in order to restore knee stability. However, in these cases the ligament has sufficient length (25%-75%) and can be tensioned along or around the graft. Preservation of the ligament remnant has several (theoretical) advantages, such as better proprioceptive function,^42,49,50 vascularization and ligamentization of the graft,^50-52 an optical guide for anatomic tunnel placement,⁵³ and a decreased incidence of tunnel widening.^54,55 Furthermore, tensioning of the remnant is thought to lower the risk of cyclops lesions when compared to remnant preservation.¹²⁵ Although the difference between augmented repair and remnant tensioning seems small, the purpose of surgery is different. With augmented repair, the ligament can be approximated close to the femoral wall and the goal of surgery is to use the healing capacity that the ACL has in the proximal part of the ligament,¹²⁶ while with remnant tensioning the goal is only to benefit from some of the aforementioned advantages. Ahn and colleagues³⁶ were the first to perform this technique and stated, “Our concept is that the remnant tissue has only an additive effect.” Furthermore, with augmented repair multiple sutures are passed through the AM and PL bundle in order to sufficiently approximate the ligament to the femoral wall, while with the remnant tensioning technique generally one or a few sutures or lasso loop are passed through the proximal part to tension the ligament, prevent sagging of the remnant, and decrease the risk of cyclops lesions.^127,128

 

Several authors have recently performed remnant tensioning during ACL.^{36,47,125-127} Ahn and colleagues⁴⁷ reported excellent objective and subjective outcomes following this procedure and found that with re-arthroscopy nearly all patients had fair synovialization of the graft. Others have reported similarly good outcomes of these techniques.^125,129,130 However, studies comparing this treatment with normal ACL reconstruction and assessing outcomes, failure rates and proprioception are lacking.

Type IV Tears: Reconstruction With Remnant Preservation

Finally, in some patients the ligament is torn distally or the tissue quality is not optimal. In these patients, the remnant can be debrided to the part of good tissue quality in order to preserve the biology and minimize the risk for cyclops lesions. A standard reconstruction needs to be performed to restore the instability, but by preserving the remnant, advantages, such as proprioception,^44,49,50 graft vascularization,^50-52 an optical guide for tibial tunnel placement,⁵³ and a decreased incidence of tunnel widening^54,55 can be expected.

Lee and colleagues³⁷ presented the tibial remnant technique in which standard reconstruction was performed, and the tibial tunnel was drilled through the center of the remnant. In a later study, they compared remnant preservation with a remnant of <20% of the total ACL length with >20% of the length and found that proprioception was better with more remnant volume.⁴⁸ Similarly, Muneta and colleagues¹³¹ assessed the role of remnant length and found that remnant length is positively correlated with better stability measured on KT-1000 anteroposterior stability.

Several studies compared ACL reconstruction with remnant preservation vs conventional ACL reconstruction.^52,54,129 Takazawa and colleagues⁵² performed a retrospective study of 183 patients and found that patients in the remnant preservation group had significantly better KT-2000 stability, while they also reported a significantly lower graft rupture rate in this group (1.1% vs 7.1%) at 2-year follow-up. Hong and colleagues¹²⁹ performed a randomized clinical trial of 80 patients and did not find these differences, although there was a trend towards higher Lysholm scores in the remnant preservation group. Finally, Zhang and colleagues⁵⁴ performed a randomized clinical trial and found a lower incidence and amount of tibial tunnel widening in the preserving-remnant group when compared to the removing-remnant group. These studies show that there is likely a role for remnant preservation.

Type V Tears: Primary Repair

In some patients, the ligament is torn in the distal 10% of the ligament, which can occur as a distal avulsion tear or as a distal bony avulsion fracture.¹³² Bony avulsion fractures are most commonly seen in children whereas true distal soft tissue avulsion tears are very rare.¹³²

Treatments of these tear types include antegrade screw fixation, pullout sutures or the use of suture anchors in case of bony avulsion fractures and pullout sutures with tying over a bony bridge or ligament button in case of soft tissue avulsions. Leeberg and colleagues¹³² recently performed a systematic review of all studies reporting on treatment of distal avulsion fractures.They noted that most treatments were currently performed arthroscopically and that outcomes were generally good. Another recent biomechanical study compared antegrade screw fixation with suture anchor fixation and pullout suture fixation.¹³³ The authors noted that suture anchor fixation has slightly less displacement of the bony fragment when compared to screw fixation and pull-out sutures, and that the strength to failure was higher in the suture anchor fixation when compared to the pullout suture fixation. The outcomes of this study suggest that screw fixation and suture anchor fixation might be superior to pullout suture fixation, which might be interesting as with pullout suture fixation the ligament cannot be directly tensioned to the tibial footprint, which can lead to anteroposterior laxity.¹³² Clinical studies are necessary to assess the preferred treatment in these tear types but it seems that screw fixation is preferred in large bony avulsion fractures, while suture anchor fixation or pullout suture fixation can be used for soft tissue avulsion tears.

Complex Tears or Poor Tissue Quality: Reconstruction

If the tear is complex, multiple tears are present, or the tissue quality is poor, then preservation of the ligament is not possible, and in these cases a standard reconstruction should be performed.

Conclusion

When reviewing the literature of ACL preservation, it becomes clear that the evolution of surgical treatment of ACL injuries was biased. Preservation of the native ligament has many advantages, such as better proprioception, graft vascularization, an optical guide for tibial tunnel placement, and a decreased incidence of tunnel widening that can be expected. Furthermore, arthroscopic primary ACL repair is minimally invasive and does not burn any bridges for future reconstructions, if necessary. This is in addition to the other (theoretical) advantages of primary repair, such as restoration of native kinematics and a decreased risk of osteoarthritis. Modern advances have significantly changed the risk-benefit ratio that should make us reconsider ACL preservation approaches. Certainly, further research in this area is warranted. In this article we have presented a treatment algorithm for ACL preservation, which is based on tear location and remnant tissue quality.

Am J Orthop. 2016;45(7):E393-E405. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

Injury of the anterior cruciate ligament (ACL) is very common with over 200,000 annual injuries in the United Status.^1,2 There is a general consensus that these injuries should not be treated conservatively in patients that are younger, or who wish to remain active.^3,4 Reconstructive surgery is currently the preferred treatment in these patients, and anatomic single-bundle reconstruction with autografts is considered the gold standard.^5,6

Reconstruction of the ACL is, however, not a perfect treatment. Following single-bundle autograft reconstruction, revision rates of 3% to 8%,^6-9 contralateral injury rates of 3% to 8%,^10,11 and infection rates of 0.5% to 3%^7,12,13 have been reported. Furthermore, due to the invasive nature of graft harvesting and the surgical procedure, 10% to 25% of the patients are not satisfied following ACL reconstruction.^14,15 This can often be explained by common complaints, such as anterior knee pain (13%-43%), kneeling pain (12%-54%), quadriceps muscle atrophy (20%-30%),^16,17 and loss of range of motion (ROM) (12%-23%).^7,9,18,19 Furthermore, as a result of the invasive nature of reconstructive surgery, revisions can be difficult due to complications, such as tunnel widening, tunnel malpositioning, and preexisting hardware.^20-22 This can lead to inferior outcomes and higher rates (13%) of revision surgery compared to primary reconstruction.^23-26 Finally, reconstructive surgery does not restore native kinematics of the ACL,^27-29 which may partially explain why reconstructive surgery has not been shown to prevent osteoarthritis.^28-31

Over the past decades, there has been an increasing interest in the preservation of the ACL in an attempt to ameliorate these issues.^32-37 Ligament preservation focuses on preserving the native tissues and biology, while minimizing the surgical morbidity to the patients.

Some authors have recently reported on arthroscopic primary repair of proximal ACL tears in which the ligament is reattached onto the femoral wall using modern-day suture anchor technology.^32,38 Others have augmented this repair technique with an internal brace^39,40 or with a synthetic device.^33,41 When performing primary repair, it is believed that proprioception is maintained,^42-44 while experimental studies have suggested that primary repair also restores the native kinematics,⁴⁵ and may prevent osteoarthritis.⁴⁶ Furthermore, primary repair is a conservative approach in that no grafts need to be harvested, no tunnels need to be drilled, and revision surgery, if necessary, is more analogous to primary reconstructions.³²In patients with partial tears, some surgeons have advocated preserving the anteromedial (AM) or posterolateral (PL) bundle and performing selective single-bundle augmentation.^34,35 In addition, several authors have used remnant tensioning^36,47 or remnant preservation^37,48 in combination with reconstructive surgery in order to benefit from the biological characteristics of the remnant. These techniques lead to better proprioceptive function,^44,49,50 vascularization and ligamentization of the graft,^50-52 provide an optical guide for anatomic tunnel placement,⁵³ and decrease the incidence of tunnel widening.^54,55The feasibility and applicability of these surgical techniques mainly depends on the tear type and tissue quality of the torn ligament. In this article we (I) discuss the history of ACL preservation, (II) discuss how modern advances alter the risk-benefit ratio for ACL preservation, and (III) propose a treatment algorithm for ACL injuries that is based on tear location and tissue quality.

History of ACL Preservation

The history of the surgical treatment of ACL injuries started in 1895 when Robson⁵⁶ treated a 41-year-old male who tore both cruciate ligaments from the femoral wall. Performing primary repair with catgut ligatures, both cruciate ligaments were preserved and the patients had resolution of pain symptoms and full function at 6-year follow-up. Over the following decades, Palmer^57,58 and O’Donoghue^59,60 further popularized open primary repair for the treatment of ACL injuries, and this technique was the most commonly performed treatment in the 1970s and early 1980s.^61-65 The initial short-term results of primary repair were excellent,^61,62 but Feagin and Curl⁶⁶ were the first to note that the results deteriorated at mid-term follow-up. Despite improvements in the surgical technique of repairing the ACL, such as the usage of nonabsorbable sutures and directly tying the sutures over bone,^63,67 the results remained disappointing at longer-term follow-up.^68-70

In response to these disappointing results, surgeons sought to improve the surgical treatment by either augmenting the primary repair with a semitendinosus, a patella tendon graft or an augmentation device,^71-74 or by performing primary reconstruction.^75-77 At the end of the 1980s and early 1990s, several randomized and prospective clinical trials were performed in order to compare the outcomes of these techniques.^74,78-82 Many studies showed that results of augmented repair were more reliable when compared to primary repair, which led to the abandonment of primary repair in favor of augmented repair, and eventually primary reconstruction.⁶⁵

 

The Important Role of Tear Location in Ligament Preservation

When taking a closer look at the outcomes of primary repair and augmented repair, it seems that the results of these preservation techniques were not as disappointing as was suggested. This can be explained, in large part, by the fact that the important roles of tear location and tissue quality were not widely recognized.

Sherman and colleagues⁷⁰ reported in 1991 their mid-term results of open primary repair. Similar to others, they noted a deterioration of their results at mid-term follow-up. However, they uniquely performed an extensive subgroup analysis in order to find an explanation for this. In their study, considered a landmark paper on primary repair,^65,70 they concluded that, “poor tissue quality is typical for midsubstance tears and that a repair of these injuries will predictably fail while type I tears (proximal), with better tissue quality, show a definite trend towards better results.”⁷⁰ With these findings, they confirmed the findings of others that had recognized a trend of better outcomes with proximal tears.^64,67,83-85

A majority of the historical studies that were published before 1991 had not considered the role of tear location and tissue quality on outcomes of open primary repair. This was also true for the aforementioned randomized studies that compared primary repair with augmented repair and primary reconstruction. Because these studies randomized patients and did not take tear location into account, it can be expected that patients with midsubstance tears were included in the cohorts of primary repair and the outcomes of these studies were therefore confounded.^74,78-82 If these studies would have been aware of the role that tear location plays on primary repair outcomes, different outcomes may have been found and different conclusions on the optimal treatment for different tear types may have been drawn.⁸⁶

Open Primary ACL Repair Outcomes Stratified by Tear Location

When reviewing the literature of open primary repair outcomes stratified by tear location, it is noted that multiple studies reported excellent outcomes following primary repair of proximal ACL tears.^{73,83,84,87-90} Weaver and colleagues⁶⁴ were among the first to stratify their results by tear location, and they found that more patients with proximal tears (52 of 66; 79%) were satisfied after the procedure when compared to patients with midsubstance tears (3 of 13; 23%) at 3.5-year follow-up. They concluded that, “selection can be made with some predictability of the type of injury to the ligament as to which patients will do better.”⁶⁴ Kühne and colleagues⁸⁹ reported the outcomes of 75 patients with proximal tears treated with open primary repair and noted no failures, negative pivot shift in 88% of patients, stable or nearly stable Lachman test in 87% of patients, and 89% return to sports rate at 4-year follow-up. Raunest and colleagues⁹¹ reported a negative pivot shift and negative anterior drawer test in 84%, return to sports in 71%, and satisfaction in 75% of 51 patients that underwent open primary repair of proximal tears at 3.5-year follow-up.

Interestingly, and in contrast to the findings of Feagin and Curl,⁶⁶ no deterioration of the outcomes at mid-term follow-up was noted in patients with proximal tears. Genelin and colleagues⁸⁸ reported their results of 42 patients with proximal tears treated with open primary repair at 5- to 7-year follow-up. They found a negative pivot shift in 81%, stable or nearly stable Lachman test in 81%, and patient satisfaction in 86% of patients. Similarly, Bräm and colleagues⁸⁷ found good results at mid-term follow-up with a good-excellent Lysholm score in 79%, return to a similar level of sports in 76%, stable or nearly stable Lachman test in 91%, and anterior drawer test in 94% of patients, along with an 88% satisfaction rate and 7% failure rate in patients who underwent open primary repair of proximal tears.

On the contrary, when the outcomes of studies that performed open primary repair in mainly, or only, patients with midsubstance tears are reviewed, significantly inferior results are found. Frank and colleagues⁹² reported outcomes in 42 patients with midsubstance tears at 4-year follow-up. They reported that 56% had a stable or nearly stable anterior drawer test, 78% had a positive pivot shift, and that only 61% were satisfied with the procedure. Odensten and colleagues⁷⁸ reported outcomes of open primary repair in a subgroup of 22 patients with midsubstance tears at 1.5-year follow-up, and noted a 14% failure rate.

When reviewing the mid-term results in patients with midsubstance tears, it seems that there was more deterioration in outcomes.^69,70 Firstly, the aforementioned study by Sherman and colleagues⁷⁰ showed poor results in the patients with (type IV) midsubstance tears at mid-term follow-up. Furthermore, Kaplan and colleagues⁶⁹ reported the mid-term outcomes of 70 patients, of which 56 patients had midsubstance tears. After having reported good outcomes at short-term follow-up,^63,67 they noted that 42% of patients had >3 mm anteroposterior stability when compared to the contralateral leg, only a 62% return to sport rate, and a 17% failure rate. They concluded that, “Although … primary repair of the anterior cruciate may work in some patients, it is an unpredictable operative procedure.”

These studies showed that the outcomes of open primary repair were significantly better in patients with proximal ACL tears and sufficient tissue quality when compared to midsubstance tears. This suggests that open primary ACL repair may have been prematurely abandoned as a treatment option for patients with proximal tears.

 

Augmented ACL Repair

There were several reasons why augmented repair became the preferred treatment in the early and mid 1990s. First of all, the results of augmented repair were more consistent compared to primary repair in the aforementioned randomized and prospective studies,^74,78-82 which is not surprising given the fact that the role of tear location was not widely recognized at the time. Secondly, in the 1970s and early 1980s, patients were treated postoperatively in a cast for 6 weeks, which led to problems, such as loss of ROM, pain, and decreased function.^93,94 At the end of the1980s and 1990s, the focus shifted from prolonged joint immobilization towards early postoperative ROM.^95-97 Since many authors believed that primary repair of the ACL was not strong enough to tolerate early mobilization, an augmentation was added to the technique in order to fortify the repair and enable early ROM.⁹⁸

Interestingly, augmented repair, which is essentially a combination of primary ACL repair and ACL reconstruction, was mainly performed in the 1990s and many surgeons did recognize the role of tear location in this treatment at this point.^73,98-103 In these years, the treatment algorithm consisted of augmented ACL repair in patients with proximal tears in the acute setting and ACL reconstruction in patients with midsubstance or chronic tears. Several different augmentation techniques were used to reinforce the primary repair in these years including autograft tissues (semitendinosus tendon,^102-104 patellar tendon,¹⁰⁰ or iliotibial band [ITB]¹⁰⁵) synthetic materials (polydioxanone [PDS],^101,102,106 carbon fibre,⁷⁴ and polyester [Trevira]⁹⁷), augmentation devices (Kennedy Ligament Augmentation Device [LAD]^98-100) and extra-articular augmentations.⁷³

When reviewing the outcomes of augmented repair of the ACL, good to excellent results can be found in studies that used this technique in patients with proximal tears.^73,98-106 Kdolsky and colleagues⁹⁸ were in one of the first groups that reported their results of augmented repair in only patients with proximal tears. In 1993, they reported their mid-term outcomes (5 to 8 years) in 66 patients who underwent primary repair and augmentation with the Kennedy LAD and found that 97% of patients had stable knees (<3 mm on KT-1000 examination), 98% had a negative pivot shift, and 76% returned to previous level of sports. However, often-reported problems with the augmentation devices were found in this study with rupture of the device (12%) and decreased ROM (14%).⁹⁸ In 1995, Grøntvedt and Engebretsen¹⁰⁰ compared augmentation with the Kennedy LAD to patellar tendon augmentation in a randomized study of patients with acute proximal tears. They noted that 50% of the patients in the Kennedy LAD group had a positive pivot shift compared to 23% in the patellar tendon group. Furthermore, they found KT-1000 leg differences of <3 mm in 92% of the patellar tendon group and 54% of the Kennedy LAD group. Because the authors found significant differences between both groups at 1- and 2-year follow-up, they stopped the clinical trial.

Several authors in the following years reported good results of augmented repair using autograft tissues. Natri and colleagues¹⁰⁵ reported the outcomes of 72 patients treated with primary repair of proximal tears augmented with the ITB at 3.5-year follow-up. They found 89% negative pivot shift rate, 93% stable or nearly stable Lachman test, 99% stable or nearly stable anterior drawer test, 79% satisfaction rate, and 91% return to previous level of sports rate. Krueger-Franke and colleagues¹⁰⁴ reported the outcomes of primary repair of proximal tears with augmentation using the semitendinosus tendon. In a retrospective study of 76 patients, they noted that 96% of patients had a negative pivot shift, 75% of patients had stable or nearly stable Lachman test, 93% were satisfied with the procedure, a mean Lysholm score of 92, a Tegner score that only decreased from 7.2 to 7.1, and KT-1000 testing with 78% <4 mm leg difference with the contralateral leg. The authors concluded that patients with femoral ruptures could be treated with augmented repair when performed in the acute setting. As this study was published in 1998, they stated that magnetic resonance imaging and arthroscopy could be helpful in identifying the tear location.

Final Abandonment of ACL Preservation

Reviewing these outcomes raises the question as to why these techniques were ultimately abandoned in the treatment algorithm of proximal ACL injuries, especially given the aforementioned advantages of ACL preservation. One of the possible answers can be found in a landmark study on ACL reconstruction and rehabilitation published by Shelbourne and colleagues¹⁰⁷ in 1991. At that time, arthrofibrosis and knee stiffness were frequently reported problems following ACL surgery, which could partially be explained by the standard conservative rehabilitation using postoperative joint immobilization.^67,70,80,88

 

Shelbourne and colleagues¹⁰⁷ aimed to assess the cause of arthrofibrosis and knee stiffness, and divided the patients into groups by number of days between injury and surgery (<7, 7 to 21 days, and >21 days between injury and surgery). Furthermore, patients within these groups underwent either a conventional or accelerated rehabilitation program. The authors not only found that patients undergoing accelerated rehabilitation had less arthrofibrosis, but they also noted that less arthrofibrosis was seen when surgery was delayed. These findings, however, contrasted with the general perception that the ACL should be repaired in the first 3 weeks postinjury to ensure optimal tissue quality with an augmented approach. As a result, the treatment of ACL injuries shifted towards ACL reconstruction after these findings. Krueger-Franke and colleagues¹⁰⁴ commented on the trend after the study of Shelbourne and colleagues:¹⁰⁷ “Less consideration has been given to the importance of the proprioceptive receptors in the tibial remnants of the torn ACL and the value of their preservation as part of a primary reconstruction.”

In addition to the trend away from an augmented repair approach due to the novel understanding of the importance of early mobilization, some discussion should focus on the technical limitations of arthroscopy at that time. While arthroscopy had been around for several decades, fluid management and arthroscopic instrumentation was slow to develop. All of the repair and augmentation techniques previously discussed had been performed via an open arthrotomy. Arthroscopic technologies of the time were not refined enough to enable surgeons to perform such complex, intra-articular techniques that would enable suturing of the ligament remnant. In this regard, arthroscopic ACL reconstruction was a much simpler technique to accomplish, and this also likely contributed to the final abandonment of the ligament preservation approach.

Role for ACL Preservation with Modern Advances

As stated in the introduction, there has been a recent resurgence of interest in preservation of the native ligament.^32-37 With the passage of time, many technologic advances have been made, which has allowed surgeons to reconsider the concept of ligament preservation.

First of all, appropriate patient selection was not applied historically, as the critical factors of tear location and tissue quality were not recognized in the era of open primary repair. In modern days, however, advances such as MRI have been developed, which can give the surgeon an idea of the status, and tear type of the ACL pre-operatively.¹⁰⁸ This may help the orthopaedic surgeon to plan the surgery and make an assessment as to whether ACL preservation is possible. Secondly, in the historic literature the postoperative regimen consisted of casting for 5 or 6 weeks,^67,70,80,88 while the focus later shifted towards early ROM.^95-97Modern day ACL rehabilitation focuses on immediate ROM to avoid the complications stiffness, pain and decreased function that plagued the outcomes when immobilization was used.^93,94 Thirdly, historically small tunnels were drilled with primary repair and sutures had to be tied over bone,^57,67 whereas currently suture anchors are available that prevent the need for tunnel drilling and enable direct suture tensioning.^32,38 Finally, and most importantly, in the historic literature patients were treated with an invasive arthrotomy technique, while modern day arthroscopic techniques readily enable the surgeon to effectively suture the remnant arthroscopically. Interestingly, in 2005, in their 20-year follow-up of primary repair surgeries, Strand and colleagues¹⁰⁹ stated, “if the same results could be accomplished by a smaller, arthroscopic procedure, primary repair might reduce the number of patients needing later reconstructions with small ‘costs’ in the way of risk and inconvenience for the patients. We therefore believe that further research and development of methods for closed (arthroscopic) repair are justified.”

Altered Risk-Benefit Ratio

Historically, the treatments of open primary repair and open ACL reconstruction were both invasive surgeries with an arthrotomy, drilling of bone tunnels, and postoperative joint immobilization for 4 to 6 weeks. However, with the modern-day advances, the risk-benefit ratio of both treatments has changed, as Strand and colleagues¹⁰⁹ had already suggested. Although ACL reconstruction can be performed arthroscopically, it remains an invasive procedure, in which tunnels are drilled, patellar tendons or hamstring tendons are harvested, and complications, such as knee pain and quadriceps atrophy, are common. The surgery of primary ACL repair, however, has benefited significantly from the modern developments.^32,38 Primary ACL repair can now be performed arthroscopically, and by using suture anchors no tunnels need to be drilled and the remnant can be tensioned directly. An additional benefit of the use of suture anchors is that revision surgery of a failed primary repair is analogous to primary reconstruction, whereas revision surgery of a failed ACL reconstruction can be problematic due to tunnel widening, tunnel malpositioning, and preexisting hardware.^20-22

 

Reviewing the differences between arthroscopic primary ACL repair and ACL reconstruction, it becomes clear that primary repair has benefited significantly from the modern advances and that the risk-benefit ratio for primary repair has been altered. This means that patients with proximal tears can be treated with a relatively straight forward, minimally invasive surgery, which has been shown to be effective in 85% to 90% of patients.^32,38

Treatment Algorithm Based on Tear Location

Since 2008, in the practice of the senior author (GSD), the surgical treatment algorithm for ACL injuries is completely based on the tear location and tissue quality of the ligament.^110,111 To describe the different tear types, we use the modified Sherman classification in which we extended his classification towards the tibial side whereas Sherman and colleagues⁷⁰ only described the femoral side of the tears (Figures A-F, Table).

In this section, we will discuss the different tear types that are seen and the corresponding treatments that can be used to treat these injuries (Table). Furthermore, we discuss current research on these topics and the reported outcomes of these techniques.

We will not provide the incidence of different tear types as the senior author’s practice is biased towards primary repair.

Type I Tears: Primary Repair

Type I tears are soft tissue avulsion type tears that can be easily treated with arthroscopic primary repair.¹⁰⁷ The length of the distal remnant has to be at least 90% and the tissue quality has to be good to excellent in order to approximate the remnant towards the femoral wall (Table).¹¹² The incidence of type I tears was 26% in the study of Sherman and colleagues,⁷⁰ although recent studies showed a lower incidence (6% to 10%) in a larger population.^32,38 Certainly, individual practices will see different percentages of type I tears based upon the mix of injury mechanisms they see most frequently. Over the last 2 years, with the recognition of the importance of tear type and tissue quality, there has been a renewed interest in arthroscopic primary ACL repair.^32,38

DiFelice and colleagues³² were the first to arthroscopically perform primary repair of the ACL in proximal tears using suture anchors. They reported the outcomes of the first 11 consecutive patients that underwent primary repair in a previously described technique.¹¹³ At mean 3.5-year follow-up, they noted only 1 failure (9%) due to re-injury; mean Lysholm score of 93.2; mean modified Cincinnati score of 91.5; pre- and postoperative Tegner score of 7.3 and 6.9, respectively; SANE score of 91.8; and subjective International Knee Documentation Committee (IKDC) score of 86.4. Of the patients with an intact repair, 9 patients had an objective IKDC rating A and 1 patient had B and all patients had KT-1000 leg differences of <3 mm with the contralateral side (three patients were not available for KT-1000 testing). The authors concluded that arthroscopic primary ACL repair could achieve short-term clinical success in a selected group of patients with proximal avulsion tears and excellent tissue quality. They further noted that mid-term outcomes are necessary given that the results of open primary repair deteriorated at longer-term follow-up in the historical literature. Recently, the senior author (GSD) has added an Internal Brace (Arthrex) to the primary repair with the goal of protecting the ligament in the first weeks to further promote healing of the ligament.^39,40,114

More recently, Achtnich and colleagues³⁸ compared the treatment of arthroscopic primary ACL repair with primary ACL reconstruction in 41 patients with type I tears at 2.3-years follow-up. Twenty-one patients consented for primary repair while 20 patients declined this procedure and underwent primary reconstruction. They noted no significant differences in Lachman test, pivot shift test, objective IKDC score, and KT-1000 scores. Although not significant, the clinical failure rate in the primary repair group (15%) was higher than the reconstruction group (0%). Interestingly, despite the higher failure rate in the repair group, the authors concluded that primary ACL repair is recommended in a carefully selected group of patients with type I tears and excellent tissue quality, which can likely be explained by the differences in the risk-benefit ratio between both procedures.

Over the last decade, the research group led by Murray^46,115,116 has performed experimental research on primary repair with a biological scaffold and reported many interesting findings that could be extrapolated to primary ACL repair. First of all, they compared bioenhanced primary repair with bioenhanced primary reconstruction in 64 Yucutan pigs and noted that there was significantly less macroscopic cartilage damage in the primary repair group at 1-year follow-up.⁴⁶ They concluded that bioenhanced ACL repair may provide a new, less invasive treatment option that reduces cartilage damage following joint injury. This may suggest that primary repair may have a lower incidence of osteoarthritis when compared to ACL reconstruction, which is interesting as osteoarthritis is very common after ACL reconstruction. Further research in this area is certainly warranted.

In another study they compared bioenhanced primary repair in juvenile, adolescent and mature Yucutan pigs and noted that functional healing depended on the level of skeletal maturity with immature animals having a more productive healing response.¹¹⁶ This indicates that primary repair might be a good treatment option in skeletally immature patients, especially since reconstruction increases the risk of premature closure of the epiphysis^117,118 and delaying treatment increases the risk of meniscus injury.¹¹⁹ Interestingly, a recent meta-analysis showed indeed that the risk of epiphysis closure was lower in primary repair when compared to ACL reconstruction and the rupture rate was also lower.¹¹⁸ Primary repair may be a good treatment option in children as the procedure has all the attributes that should be applicable to children: it is minimally morbid, tissue sparing, and it is a conservative approach that does not burn any surgical bridges for future reconstructive surgery if necessary.

Finally, the research group of Murray¹¹⁵ assessed the effect of surgical delay of primary repair following injury in Yucutan pigs and noted that better biomechanical outcomes were noted after delaying surgery for 2 weeks when compared to 6 weeks. This suggests that primary repair should preferably be performed in the acute setting, which has also been shown in historical studies since the ligament in the acute setting has optimal tissue quality and the ligament is less likely to be retracted or reabsorbed.^59,60,115

 

One Bundle Type I Tears: Single Bundle Augmented Repair

In some cases, the tear locations of the AM and PL bundle are not at the same location and Zantop and colleagues¹²⁰ reported in an arthroscopic study that this could be as frequent as in 30% of all complete tears. In some of these tears, one of the bundles can be avulsed of the femoral wall (type I tear) while the other bundle is not directly repairable (non-type I tear). In these cases, the senior author (GSD) will repair the type I tear bundle, whereas a hamstring augmentation is placed at the location of the other bundle. When reviewing the literature, a combination of primary repair of one bundle and reconstruction of the second bundle has not been described before. However, over the last decade several surgeons have performed augmentation of one bundle in the setting of partial tears.^{34,35,121-124}

Buda and colleagues³⁴ were the first to perform selective AM or PL bundle reconstruction in the setting of partial tears.³⁴ At 5-year follow-up, they reported no reruptures and only 1 patient with an IKDC C-score, although reoperation was necessary in 4 out of 47 patients (9%). Following this publication, many others reported on selective bundle reconstruction.^35,121-124 However, with partial tears, the knee is often stable and a selective augmentation technique is utilized to prevent complete rupture of the ligament. The application of this technique is essentially different from reconstruction for complete ACL tears in which the knee is unstable, there is a giving way sensation and patients have problems participating in sports.

Type II Tears: Augmented Repair

Type II tears often have good or excellent tissue quality and can be pulled up towards the femoral footprint, but are too short to be firmly attached. Sherman and colleagues⁷⁰ reported that approximately 22% patients had a type II tear, which corresponds to a tear located in the proximal part of the ligament. With this technique, multiple suture passes are used to stitch the remnant and, in addition, a smaller hamstring autograft or allograft is passed through the middle of the tibial remnant. A suture button is used proximally for the graft, and the tensioning repair sutures through the remnant are also passed through the suture button. The suture button is passed through the femoral tunnel and flipped so that the graft is proximally fixed. Then, the repair sutures of the remnant are tensioned, and the ligament is pulled towards the femoral wall as a sleeve around the graft. When the ligament is approximated to the femoral wall, the sutures are tied over the suture button. The graft is then tensioned distally to complete the augmented repair.

In the recent literature, the technique of augmentation of a primary repair using autograft tissue has not been reported. However, augmented repair using an internal brace^39,40 or augmentation devices^33,41 have been recently performed. MacKay and colleagues³⁹ reported good outcomes of arthroscopic primary repair of proximal tears using an internal brace. Eggli and colleagues³³ reported the results of the first 10 patients treated with ACL preservation using primary repair of the ligament with the addition of a dynamic screw-spring mechanism. The authors reported good preliminary results with one failure (10%) and good objective and subjective outcomes. In a next study, they reported the outcomes of 278 patients and although they reported good clinical outcomes and a revision rate of 4%, the reoperation rate for removal of the screw-spring mechanism was high (24%).⁴¹ This is not surprising when reviewing the historical literature in which high complication rates of the augmentation devices were reported.^99,100 We were unable to identify any other studies reporting surgical techniques of augmenting primary repair in the literature.

Type III Tears: Reconstruction With Remnant Tensioning

In patients with type III tears, the ligament cannot be approximated to the wall and reconstruction is necessary in order to restore knee stability. However, in these cases the ligament has sufficient length (25%-75%) and can be tensioned along or around the graft. Preservation of the ligament remnant has several (theoretical) advantages, such as better proprioceptive function,^42,49,50 vascularization and ligamentization of the graft,^50-52 an optical guide for anatomic tunnel placement,⁵³ and a decreased incidence of tunnel widening.^54,55 Furthermore, tensioning of the remnant is thought to lower the risk of cyclops lesions when compared to remnant preservation.¹²⁵ Although the difference between augmented repair and remnant tensioning seems small, the purpose of surgery is different. With augmented repair, the ligament can be approximated close to the femoral wall and the goal of surgery is to use the healing capacity that the ACL has in the proximal part of the ligament,¹²⁶ while with remnant tensioning the goal is only to benefit from some of the aforementioned advantages. Ahn and colleagues³⁶ were the first to perform this technique and stated, “Our concept is that the remnant tissue has only an additive effect.” Furthermore, with augmented repair multiple sutures are passed through the AM and PL bundle in order to sufficiently approximate the ligament to the femoral wall, while with the remnant tensioning technique generally one or a few sutures or lasso loop are passed through the proximal part to tension the ligament, prevent sagging of the remnant, and decrease the risk of cyclops lesions.^127,128

 

Several authors have recently performed remnant tensioning during ACL.^{36,47,125-127} Ahn and colleagues⁴⁷ reported excellent objective and subjective outcomes following this procedure and found that with re-arthroscopy nearly all patients had fair synovialization of the graft. Others have reported similarly good outcomes of these techniques.^125,129,130 However, studies comparing this treatment with normal ACL reconstruction and assessing outcomes, failure rates and proprioception are lacking.

Type IV Tears: Reconstruction With Remnant Preservation

Finally, in some patients the ligament is torn distally or the tissue quality is not optimal. In these patients, the remnant can be debrided to the part of good tissue quality in order to preserve the biology and minimize the risk for cyclops lesions. A standard reconstruction needs to be performed to restore the instability, but by preserving the remnant, advantages, such as proprioception,^44,49,50 graft vascularization,^50-52 an optical guide for tibial tunnel placement,⁵³ and a decreased incidence of tunnel widening^54,55 can be expected.

Lee and colleagues³⁷ presented the tibial remnant technique in which standard reconstruction was performed, and the tibial tunnel was drilled through the center of the remnant. In a later study, they compared remnant preservation with a remnant of <20% of the total ACL length with >20% of the length and found that proprioception was better with more remnant volume.⁴⁸ Similarly, Muneta and colleagues¹³¹ assessed the role of remnant length and found that remnant length is positively correlated with better stability measured on KT-1000 anteroposterior stability.

Several studies compared ACL reconstruction with remnant preservation vs conventional ACL reconstruction.^52,54,129 Takazawa and colleagues⁵² performed a retrospective study of 183 patients and found that patients in the remnant preservation group had significantly better KT-2000 stability, while they also reported a significantly lower graft rupture rate in this group (1.1% vs 7.1%) at 2-year follow-up. Hong and colleagues¹²⁹ performed a randomized clinical trial of 80 patients and did not find these differences, although there was a trend towards higher Lysholm scores in the remnant preservation group. Finally, Zhang and colleagues⁵⁴ performed a randomized clinical trial and found a lower incidence and amount of tibial tunnel widening in the preserving-remnant group when compared to the removing-remnant group. These studies show that there is likely a role for remnant preservation.

Type V Tears: Primary Repair

In some patients, the ligament is torn in the distal 10% of the ligament, which can occur as a distal avulsion tear or as a distal bony avulsion fracture.¹³² Bony avulsion fractures are most commonly seen in children whereas true distal soft tissue avulsion tears are very rare.¹³²

Treatments of these tear types include antegrade screw fixation, pullout sutures or the use of suture anchors in case of bony avulsion fractures and pullout sutures with tying over a bony bridge or ligament button in case of soft tissue avulsions. Leeberg and colleagues¹³² recently performed a systematic review of all studies reporting on treatment of distal avulsion fractures.They noted that most treatments were currently performed arthroscopically and that outcomes were generally good. Another recent biomechanical study compared antegrade screw fixation with suture anchor fixation and pullout suture fixation.¹³³ The authors noted that suture anchor fixation has slightly less displacement of the bony fragment when compared to screw fixation and pull-out sutures, and that the strength to failure was higher in the suture anchor fixation when compared to the pullout suture fixation. The outcomes of this study suggest that screw fixation and suture anchor fixation might be superior to pullout suture fixation, which might be interesting as with pullout suture fixation the ligament cannot be directly tensioned to the tibial footprint, which can lead to anteroposterior laxity.¹³² Clinical studies are necessary to assess the preferred treatment in these tear types but it seems that screw fixation is preferred in large bony avulsion fractures, while suture anchor fixation or pullout suture fixation can be used for soft tissue avulsion tears.

Complex Tears or Poor Tissue Quality: Reconstruction

If the tear is complex, multiple tears are present, or the tissue quality is poor, then preservation of the ligament is not possible, and in these cases a standard reconstruction should be performed.

Conclusion

When reviewing the literature of ACL preservation, it becomes clear that the evolution of surgical treatment of ACL injuries was biased. Preservation of the native ligament has many advantages, such as better proprioception, graft vascularization, an optical guide for tibial tunnel placement, and a decreased incidence of tunnel widening that can be expected. Furthermore, arthroscopic primary ACL repair is minimally invasive and does not burn any bridges for future reconstructions, if necessary. This is in addition to the other (theoretical) advantages of primary repair, such as restoration of native kinematics and a decreased risk of osteoarthritis. Modern advances have significantly changed the risk-benefit ratio that should make us reconsider ACL preservation approaches. Certainly, further research in this area is warranted. In this article we have presented a treatment algorithm for ACL preservation, which is based on tear location and remnant tissue quality.

Am J Orthop. 2016;45(7):E393-E405. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

Injury of the anterior cruciate ligament (ACL) is very common with over 200,000 annual injuries in the United Status.^1,2 There is a general consensus that these injuries should not be treated conservatively in patients that are younger, or who wish to remain active.^3,4 Reconstructive surgery is currently the preferred treatment in these patients, and anatomic single-bundle reconstruction with autografts is considered the gold standard.^5,6

Reconstruction of the ACL is, however, not a perfect treatment. Following single-bundle autograft reconstruction, revision rates of 3% to 8%,^6-9 contralateral injury rates of 3% to 8%,^10,11 and infection rates of 0.5% to 3%^7,12,13 have been reported. Furthermore, due to the invasive nature of graft harvesting and the surgical procedure, 10% to 25% of the patients are not satisfied following ACL reconstruction.^14,15 This can often be explained by common complaints, such as anterior knee pain (13%-43%), kneeling pain (12%-54%), quadriceps muscle atrophy (20%-30%),^16,17 and loss of range of motion (ROM) (12%-23%).^7,9,18,19 Furthermore, as a result of the invasive nature of reconstructive surgery, revisions can be difficult due to complications, such as tunnel widening, tunnel malpositioning, and preexisting hardware.^20-22 This can lead to inferior outcomes and higher rates (13%) of revision surgery compared to primary reconstruction.^23-26 Finally, reconstructive surgery does not restore native kinematics of the ACL,^27-29 which may partially explain why reconstructive surgery has not been shown to prevent osteoarthritis.^28-31

Over the past decades, there has been an increasing interest in the preservation of the ACL in an attempt to ameliorate these issues.^32-37 Ligament preservation focuses on preserving the native tissues and biology, while minimizing the surgical morbidity to the patients.

Some authors have recently reported on arthroscopic primary repair of proximal ACL tears in which the ligament is reattached onto the femoral wall using modern-day suture anchor technology.^32,38 Others have augmented this repair technique with an internal brace^39,40 or with a synthetic device.^33,41 When performing primary repair, it is believed that proprioception is maintained,^42-44 while experimental studies have suggested that primary repair also restores the native kinematics,⁴⁵ and may prevent osteoarthritis.⁴⁶ Furthermore, primary repair is a conservative approach in that no grafts need to be harvested, no tunnels need to be drilled, and revision surgery, if necessary, is more analogous to primary reconstructions.³²In patients with partial tears, some surgeons have advocated preserving the anteromedial (AM) or posterolateral (PL) bundle and performing selective single-bundle augmentation.^34,35 In addition, several authors have used remnant tensioning^36,47 or remnant preservation^37,48 in combination with reconstructive surgery in order to benefit from the biological characteristics of the remnant. These techniques lead to better proprioceptive function,^44,49,50 vascularization and ligamentization of the graft,^50-52 provide an optical guide for anatomic tunnel placement,⁵³ and decrease the incidence of tunnel widening.^54,55The feasibility and applicability of these surgical techniques mainly depends on the tear type and tissue quality of the torn ligament. In this article we (I) discuss the history of ACL preservation, (II) discuss how modern advances alter the risk-benefit ratio for ACL preservation, and (III) propose a treatment algorithm for ACL injuries that is based on tear location and tissue quality.

History of ACL Preservation

The history of the surgical treatment of ACL injuries started in 1895 when Robson⁵⁶ treated a 41-year-old male who tore both cruciate ligaments from the femoral wall. Performing primary repair with catgut ligatures, both cruciate ligaments were preserved and the patients had resolution of pain symptoms and full function at 6-year follow-up. Over the following decades, Palmer^57,58 and O’Donoghue^59,60 further popularized open primary repair for the treatment of ACL injuries, and this technique was the most commonly performed treatment in the 1970s and early 1980s.^61-65 The initial short-term results of primary repair were excellent,^61,62 but Feagin and Curl⁶⁶ were the first to note that the results deteriorated at mid-term follow-up. Despite improvements in the surgical technique of repairing the ACL, such as the usage of nonabsorbable sutures and directly tying the sutures over bone,^63,67 the results remained disappointing at longer-term follow-up.^68-70

In response to these disappointing results, surgeons sought to improve the surgical treatment by either augmenting the primary repair with a semitendinosus, a patella tendon graft or an augmentation device,^71-74 or by performing primary reconstruction.^75-77 At the end of the 1980s and early 1990s, several randomized and prospective clinical trials were performed in order to compare the outcomes of these techniques.^74,78-82 Many studies showed that results of augmented repair were more reliable when compared to primary repair, which led to the abandonment of primary repair in favor of augmented repair, and eventually primary reconstruction.⁶⁵

 

The Important Role of Tear Location in Ligament Preservation

When taking a closer look at the outcomes of primary repair and augmented repair, it seems that the results of these preservation techniques were not as disappointing as was suggested. This can be explained, in large part, by the fact that the important roles of tear location and tissue quality were not widely recognized.

Sherman and colleagues⁷⁰ reported in 1991 their mid-term results of open primary repair. Similar to others, they noted a deterioration of their results at mid-term follow-up. However, they uniquely performed an extensive subgroup analysis in order to find an explanation for this. In their study, considered a landmark paper on primary repair,^65,70 they concluded that, “poor tissue quality is typical for midsubstance tears and that a repair of these injuries will predictably fail while type I tears (proximal), with better tissue quality, show a definite trend towards better results.”⁷⁰ With these findings, they confirmed the findings of others that had recognized a trend of better outcomes with proximal tears.^64,67,83-85

A majority of the historical studies that were published before 1991 had not considered the role of tear location and tissue quality on outcomes of open primary repair. This was also true for the aforementioned randomized studies that compared primary repair with augmented repair and primary reconstruction. Because these studies randomized patients and did not take tear location into account, it can be expected that patients with midsubstance tears were included in the cohorts of primary repair and the outcomes of these studies were therefore confounded.^74,78-82 If these studies would have been aware of the role that tear location plays on primary repair outcomes, different outcomes may have been found and different conclusions on the optimal treatment for different tear types may have been drawn.⁸⁶

Open Primary ACL Repair Outcomes Stratified by Tear Location

When reviewing the literature of open primary repair outcomes stratified by tear location, it is noted that multiple studies reported excellent outcomes following primary repair of proximal ACL tears.^{73,83,84,87-90} Weaver and colleagues⁶⁴ were among the first to stratify their results by tear location, and they found that more patients with proximal tears (52 of 66; 79%) were satisfied after the procedure when compared to patients with midsubstance tears (3 of 13; 23%) at 3.5-year follow-up. They concluded that, “selection can be made with some predictability of the type of injury to the ligament as to which patients will do better.”⁶⁴ Kühne and colleagues⁸⁹ reported the outcomes of 75 patients with proximal tears treated with open primary repair and noted no failures, negative pivot shift in 88% of patients, stable or nearly stable Lachman test in 87% of patients, and 89% return to sports rate at 4-year follow-up. Raunest and colleagues⁹¹ reported a negative pivot shift and negative anterior drawer test in 84%, return to sports in 71%, and satisfaction in 75% of 51 patients that underwent open primary repair of proximal tears at 3.5-year follow-up.

Interestingly, and in contrast to the findings of Feagin and Curl,⁶⁶ no deterioration of the outcomes at mid-term follow-up was noted in patients with proximal tears. Genelin and colleagues⁸⁸ reported their results of 42 patients with proximal tears treated with open primary repair at 5- to 7-year follow-up. They found a negative pivot shift in 81%, stable or nearly stable Lachman test in 81%, and patient satisfaction in 86% of patients. Similarly, Bräm and colleagues⁸⁷ found good results at mid-term follow-up with a good-excellent Lysholm score in 79%, return to a similar level of sports in 76%, stable or nearly stable Lachman test in 91%, and anterior drawer test in 94% of patients, along with an 88% satisfaction rate and 7% failure rate in patients who underwent open primary repair of proximal tears.

On the contrary, when the outcomes of studies that performed open primary repair in mainly, or only, patients with midsubstance tears are reviewed, significantly inferior results are found. Frank and colleagues⁹² reported outcomes in 42 patients with midsubstance tears at 4-year follow-up. They reported that 56% had a stable or nearly stable anterior drawer test, 78% had a positive pivot shift, and that only 61% were satisfied with the procedure. Odensten and colleagues⁷⁸ reported outcomes of open primary repair in a subgroup of 22 patients with midsubstance tears at 1.5-year follow-up, and noted a 14% failure rate.

When reviewing the mid-term results in patients with midsubstance tears, it seems that there was more deterioration in outcomes.^69,70 Firstly, the aforementioned study by Sherman and colleagues⁷⁰ showed poor results in the patients with (type IV) midsubstance tears at mid-term follow-up. Furthermore, Kaplan and colleagues⁶⁹ reported the mid-term outcomes of 70 patients, of which 56 patients had midsubstance tears. After having reported good outcomes at short-term follow-up,^63,67 they noted that 42% of patients had >3 mm anteroposterior stability when compared to the contralateral leg, only a 62% return to sport rate, and a 17% failure rate. They concluded that, “Although … primary repair of the anterior cruciate may work in some patients, it is an unpredictable operative procedure.”

These studies showed that the outcomes of open primary repair were significantly better in patients with proximal ACL tears and sufficient tissue quality when compared to midsubstance tears. This suggests that open primary ACL repair may have been prematurely abandoned as a treatment option for patients with proximal tears.

 

Augmented ACL Repair

There were several reasons why augmented repair became the preferred treatment in the early and mid 1990s. First of all, the results of augmented repair were more consistent compared to primary repair in the aforementioned randomized and prospective studies,^74,78-82 which is not surprising given the fact that the role of tear location was not widely recognized at the time. Secondly, in the 1970s and early 1980s, patients were treated postoperatively in a cast for 6 weeks, which led to problems, such as loss of ROM, pain, and decreased function.^93,94 At the end of the1980s and 1990s, the focus shifted from prolonged joint immobilization towards early postoperative ROM.^95-97 Since many authors believed that primary repair of the ACL was not strong enough to tolerate early mobilization, an augmentation was added to the technique in order to fortify the repair and enable early ROM.⁹⁸

Interestingly, augmented repair, which is essentially a combination of primary ACL repair and ACL reconstruction, was mainly performed in the 1990s and many surgeons did recognize the role of tear location in this treatment at this point.^73,98-103 In these years, the treatment algorithm consisted of augmented ACL repair in patients with proximal tears in the acute setting and ACL reconstruction in patients with midsubstance or chronic tears. Several different augmentation techniques were used to reinforce the primary repair in these years including autograft tissues (semitendinosus tendon,^102-104 patellar tendon,¹⁰⁰ or iliotibial band [ITB]¹⁰⁵) synthetic materials (polydioxanone [PDS],^101,102,106 carbon fibre,⁷⁴ and polyester [Trevira]⁹⁷), augmentation devices (Kennedy Ligament Augmentation Device [LAD]^98-100) and extra-articular augmentations.⁷³

When reviewing the outcomes of augmented repair of the ACL, good to excellent results can be found in studies that used this technique in patients with proximal tears.^73,98-106 Kdolsky and colleagues⁹⁸ were in one of the first groups that reported their results of augmented repair in only patients with proximal tears. In 1993, they reported their mid-term outcomes (5 to 8 years) in 66 patients who underwent primary repair and augmentation with the Kennedy LAD and found that 97% of patients had stable knees (<3 mm on KT-1000 examination), 98% had a negative pivot shift, and 76% returned to previous level of sports. However, often-reported problems with the augmentation devices were found in this study with rupture of the device (12%) and decreased ROM (14%).⁹⁸ In 1995, Grøntvedt and Engebretsen¹⁰⁰ compared augmentation with the Kennedy LAD to patellar tendon augmentation in a randomized study of patients with acute proximal tears. They noted that 50% of the patients in the Kennedy LAD group had a positive pivot shift compared to 23% in the patellar tendon group. Furthermore, they found KT-1000 leg differences of <3 mm in 92% of the patellar tendon group and 54% of the Kennedy LAD group. Because the authors found significant differences between both groups at 1- and 2-year follow-up, they stopped the clinical trial.

Several authors in the following years reported good results of augmented repair using autograft tissues. Natri and colleagues¹⁰⁵ reported the outcomes of 72 patients treated with primary repair of proximal tears augmented with the ITB at 3.5-year follow-up. They found 89% negative pivot shift rate, 93% stable or nearly stable Lachman test, 99% stable or nearly stable anterior drawer test, 79% satisfaction rate, and 91% return to previous level of sports rate. Krueger-Franke and colleagues¹⁰⁴ reported the outcomes of primary repair of proximal tears with augmentation using the semitendinosus tendon. In a retrospective study of 76 patients, they noted that 96% of patients had a negative pivot shift, 75% of patients had stable or nearly stable Lachman test, 93% were satisfied with the procedure, a mean Lysholm score of 92, a Tegner score that only decreased from 7.2 to 7.1, and KT-1000 testing with 78% <4 mm leg difference with the contralateral leg. The authors concluded that patients with femoral ruptures could be treated with augmented repair when performed in the acute setting. As this study was published in 1998, they stated that magnetic resonance imaging and arthroscopy could be helpful in identifying the tear location.

Final Abandonment of ACL Preservation

Reviewing these outcomes raises the question as to why these techniques were ultimately abandoned in the treatment algorithm of proximal ACL injuries, especially given the aforementioned advantages of ACL preservation. One of the possible answers can be found in a landmark study on ACL reconstruction and rehabilitation published by Shelbourne and colleagues¹⁰⁷ in 1991. At that time, arthrofibrosis and knee stiffness were frequently reported problems following ACL surgery, which could partially be explained by the standard conservative rehabilitation using postoperative joint immobilization.^67,70,80,88

 

Shelbourne and colleagues¹⁰⁷ aimed to assess the cause of arthrofibrosis and knee stiffness, and divided the patients into groups by number of days between injury and surgery (<7, 7 to 21 days, and >21 days between injury and surgery). Furthermore, patients within these groups underwent either a conventional or accelerated rehabilitation program. The authors not only found that patients undergoing accelerated rehabilitation had less arthrofibrosis, but they also noted that less arthrofibrosis was seen when surgery was delayed. These findings, however, contrasted with the general perception that the ACL should be repaired in the first 3 weeks postinjury to ensure optimal tissue quality with an augmented approach. As a result, the treatment of ACL injuries shifted towards ACL reconstruction after these findings. Krueger-Franke and colleagues¹⁰⁴ commented on the trend after the study of Shelbourne and colleagues:¹⁰⁷ “Less consideration has been given to the importance of the proprioceptive receptors in the tibial remnants of the torn ACL and the value of their preservation as part of a primary reconstruction.”

In addition to the trend away from an augmented repair approach due to the novel understanding of the importance of early mobilization, some discussion should focus on the technical limitations of arthroscopy at that time. While arthroscopy had been around for several decades, fluid management and arthroscopic instrumentation was slow to develop. All of the repair and augmentation techniques previously discussed had been performed via an open arthrotomy. Arthroscopic technologies of the time were not refined enough to enable surgeons to perform such complex, intra-articular techniques that would enable suturing of the ligament remnant. In this regard, arthroscopic ACL reconstruction was a much simpler technique to accomplish, and this also likely contributed to the final abandonment of the ligament preservation approach.

Role for ACL Preservation with Modern Advances

As stated in the introduction, there has been a recent resurgence of interest in preservation of the native ligament.^32-37 With the passage of time, many technologic advances have been made, which has allowed surgeons to reconsider the concept of ligament preservation.

First of all, appropriate patient selection was not applied historically, as the critical factors of tear location and tissue quality were not recognized in the era of open primary repair. In modern days, however, advances such as MRI have been developed, which can give the surgeon an idea of the status, and tear type of the ACL pre-operatively.¹⁰⁸ This may help the orthopaedic surgeon to plan the surgery and make an assessment as to whether ACL preservation is possible. Secondly, in the historic literature the postoperative regimen consisted of casting for 5 or 6 weeks,^67,70,80,88 while the focus later shifted towards early ROM.^95-97Modern day ACL rehabilitation focuses on immediate ROM to avoid the complications stiffness, pain and decreased function that plagued the outcomes when immobilization was used.^93,94 Thirdly, historically small tunnels were drilled with primary repair and sutures had to be tied over bone,^57,67 whereas currently suture anchors are available that prevent the need for tunnel drilling and enable direct suture tensioning.^32,38 Finally, and most importantly, in the historic literature patients were treated with an invasive arthrotomy technique, while modern day arthroscopic techniques readily enable the surgeon to effectively suture the remnant arthroscopically. Interestingly, in 2005, in their 20-year follow-up of primary repair surgeries, Strand and colleagues¹⁰⁹ stated, “if the same results could be accomplished by a smaller, arthroscopic procedure, primary repair might reduce the number of patients needing later reconstructions with small ‘costs’ in the way of risk and inconvenience for the patients. We therefore believe that further research and development of methods for closed (arthroscopic) repair are justified.”

Altered Risk-Benefit Ratio

Historically, the treatments of open primary repair and open ACL reconstruction were both invasive surgeries with an arthrotomy, drilling of bone tunnels, and postoperative joint immobilization for 4 to 6 weeks. However, with the modern-day advances, the risk-benefit ratio of both treatments has changed, as Strand and colleagues¹⁰⁹ had already suggested. Although ACL reconstruction can be performed arthroscopically, it remains an invasive procedure, in which tunnels are drilled, patellar tendons or hamstring tendons are harvested, and complications, such as knee pain and quadriceps atrophy, are common. The surgery of primary ACL repair, however, has benefited significantly from the modern developments.^32,38 Primary ACL repair can now be performed arthroscopically, and by using suture anchors no tunnels need to be drilled and the remnant can be tensioned directly. An additional benefit of the use of suture anchors is that revision surgery of a failed primary repair is analogous to primary reconstruction, whereas revision surgery of a failed ACL reconstruction can be problematic due to tunnel widening, tunnel malpositioning, and preexisting hardware.^20-22

 

Reviewing the differences between arthroscopic primary ACL repair and ACL reconstruction, it becomes clear that primary repair has benefited significantly from the modern advances and that the risk-benefit ratio for primary repair has been altered. This means that patients with proximal tears can be treated with a relatively straight forward, minimally invasive surgery, which has been shown to be effective in 85% to 90% of patients.^32,38

Treatment Algorithm Based on Tear Location

Since 2008, in the practice of the senior author (GSD), the surgical treatment algorithm for ACL injuries is completely based on the tear location and tissue quality of the ligament.^110,111 To describe the different tear types, we use the modified Sherman classification in which we extended his classification towards the tibial side whereas Sherman and colleagues⁷⁰ only described the femoral side of the tears (Figures A-F, Table).

In this section, we will discuss the different tear types that are seen and the corresponding treatments that can be used to treat these injuries (Table). Furthermore, we discuss current research on these topics and the reported outcomes of these techniques.

We will not provide the incidence of different tear types as the senior author’s practice is biased towards primary repair.

Type I Tears: Primary Repair

Type I tears are soft tissue avulsion type tears that can be easily treated with arthroscopic primary repair.¹⁰⁷ The length of the distal remnant has to be at least 90% and the tissue quality has to be good to excellent in order to approximate the remnant towards the femoral wall (Table).¹¹² The incidence of type I tears was 26% in the study of Sherman and colleagues,⁷⁰ although recent studies showed a lower incidence (6% to 10%) in a larger population.^32,38 Certainly, individual practices will see different percentages of type I tears based upon the mix of injury mechanisms they see most frequently. Over the last 2 years, with the recognition of the importance of tear type and tissue quality, there has been a renewed interest in arthroscopic primary ACL repair.^32,38

DiFelice and colleagues³² were the first to arthroscopically perform primary repair of the ACL in proximal tears using suture anchors. They reported the outcomes of the first 11 consecutive patients that underwent primary repair in a previously described technique.¹¹³ At mean 3.5-year follow-up, they noted only 1 failure (9%) due to re-injury; mean Lysholm score of 93.2; mean modified Cincinnati score of 91.5; pre- and postoperative Tegner score of 7.3 and 6.9, respectively; SANE score of 91.8; and subjective International Knee Documentation Committee (IKDC) score of 86.4. Of the patients with an intact repair, 9 patients had an objective IKDC rating A and 1 patient had B and all patients had KT-1000 leg differences of <3 mm with the contralateral side (three patients were not available for KT-1000 testing). The authors concluded that arthroscopic primary ACL repair could achieve short-term clinical success in a selected group of patients with proximal avulsion tears and excellent tissue quality. They further noted that mid-term outcomes are necessary given that the results of open primary repair deteriorated at longer-term follow-up in the historical literature. Recently, the senior author (GSD) has added an Internal Brace (Arthrex) to the primary repair with the goal of protecting the ligament in the first weeks to further promote healing of the ligament.^39,40,114

More recently, Achtnich and colleagues³⁸ compared the treatment of arthroscopic primary ACL repair with primary ACL reconstruction in 41 patients with type I tears at 2.3-years follow-up. Twenty-one patients consented for primary repair while 20 patients declined this procedure and underwent primary reconstruction. They noted no significant differences in Lachman test, pivot shift test, objective IKDC score, and KT-1000 scores. Although not significant, the clinical failure rate in the primary repair group (15%) was higher than the reconstruction group (0%). Interestingly, despite the higher failure rate in the repair group, the authors concluded that primary ACL repair is recommended in a carefully selected group of patients with type I tears and excellent tissue quality, which can likely be explained by the differences in the risk-benefit ratio between both procedures.

Over the last decade, the research group led by Murray^46,115,116 has performed experimental research on primary repair with a biological scaffold and reported many interesting findings that could be extrapolated to primary ACL repair. First of all, they compared bioenhanced primary repair with bioenhanced primary reconstruction in 64 Yucutan pigs and noted that there was significantly less macroscopic cartilage damage in the primary repair group at 1-year follow-up.⁴⁶ They concluded that bioenhanced ACL repair may provide a new, less invasive treatment option that reduces cartilage damage following joint injury. This may suggest that primary repair may have a lower incidence of osteoarthritis when compared to ACL reconstruction, which is interesting as osteoarthritis is very common after ACL reconstruction. Further research in this area is certainly warranted.

In another study they compared bioenhanced primary repair in juvenile, adolescent and mature Yucutan pigs and noted that functional healing depended on the level of skeletal maturity with immature animals having a more productive healing response.¹¹⁶ This indicates that primary repair might be a good treatment option in skeletally immature patients, especially since reconstruction increases the risk of premature closure of the epiphysis^117,118 and delaying treatment increases the risk of meniscus injury.¹¹⁹ Interestingly, a recent meta-analysis showed indeed that the risk of epiphysis closure was lower in primary repair when compared to ACL reconstruction and the rupture rate was also lower.¹¹⁸ Primary repair may be a good treatment option in children as the procedure has all the attributes that should be applicable to children: it is minimally morbid, tissue sparing, and it is a conservative approach that does not burn any surgical bridges for future reconstructive surgery if necessary.

Finally, the research group of Murray¹¹⁵ assessed the effect of surgical delay of primary repair following injury in Yucutan pigs and noted that better biomechanical outcomes were noted after delaying surgery for 2 weeks when compared to 6 weeks. This suggests that primary repair should preferably be performed in the acute setting, which has also been shown in historical studies since the ligament in the acute setting has optimal tissue quality and the ligament is less likely to be retracted or reabsorbed.^59,60,115

 

One Bundle Type I Tears: Single Bundle Augmented Repair

In some cases, the tear locations of the AM and PL bundle are not at the same location and Zantop and colleagues¹²⁰ reported in an arthroscopic study that this could be as frequent as in 30% of all complete tears. In some of these tears, one of the bundles can be avulsed of the femoral wall (type I tear) while the other bundle is not directly repairable (non-type I tear). In these cases, the senior author (GSD) will repair the type I tear bundle, whereas a hamstring augmentation is placed at the location of the other bundle. When reviewing the literature, a combination of primary repair of one bundle and reconstruction of the second bundle has not been described before. However, over the last decade several surgeons have performed augmentation of one bundle in the setting of partial tears.^{34,35,121-124}

Buda and colleagues³⁴ were the first to perform selective AM or PL bundle reconstruction in the setting of partial tears.³⁴ At 5-year follow-up, they reported no reruptures and only 1 patient with an IKDC C-score, although reoperation was necessary in 4 out of 47 patients (9%). Following this publication, many others reported on selective bundle reconstruction.^35,121-124 However, with partial tears, the knee is often stable and a selective augmentation technique is utilized to prevent complete rupture of the ligament. The application of this technique is essentially different from reconstruction for complete ACL tears in which the knee is unstable, there is a giving way sensation and patients have problems participating in sports.

Type II Tears: Augmented Repair

Type II tears often have good or excellent tissue quality and can be pulled up towards the femoral footprint, but are too short to be firmly attached. Sherman and colleagues⁷⁰ reported that approximately 22% patients had a type II tear, which corresponds to a tear located in the proximal part of the ligament. With this technique, multiple suture passes are used to stitch the remnant and, in addition, a smaller hamstring autograft or allograft is passed through the middle of the tibial remnant. A suture button is used proximally for the graft, and the tensioning repair sutures through the remnant are also passed through the suture button. The suture button is passed through the femoral tunnel and flipped so that the graft is proximally fixed. Then, the repair sutures of the remnant are tensioned, and the ligament is pulled towards the femoral wall as a sleeve around the graft. When the ligament is approximated to the femoral wall, the sutures are tied over the suture button. The graft is then tensioned distally to complete the augmented repair.

In the recent literature, the technique of augmentation of a primary repair using autograft tissue has not been reported. However, augmented repair using an internal brace^39,40 or augmentation devices^33,41 have been recently performed. MacKay and colleagues³⁹ reported good outcomes of arthroscopic primary repair of proximal tears using an internal brace. Eggli and colleagues³³ reported the results of the first 10 patients treated with ACL preservation using primary repair of the ligament with the addition of a dynamic screw-spring mechanism. The authors reported good preliminary results with one failure (10%) and good objective and subjective outcomes. In a next study, they reported the outcomes of 278 patients and although they reported good clinical outcomes and a revision rate of 4%, the reoperation rate for removal of the screw-spring mechanism was high (24%).⁴¹ This is not surprising when reviewing the historical literature in which high complication rates of the augmentation devices were reported.^99,100 We were unable to identify any other studies reporting surgical techniques of augmenting primary repair in the literature.

Type III Tears: Reconstruction With Remnant Tensioning

In patients with type III tears, the ligament cannot be approximated to the wall and reconstruction is necessary in order to restore knee stability. However, in these cases the ligament has sufficient length (25%-75%) and can be tensioned along or around the graft. Preservation of the ligament remnant has several (theoretical) advantages, such as better proprioceptive function,^42,49,50 vascularization and ligamentization of the graft,^50-52 an optical guide for anatomic tunnel placement,⁵³ and a decreased incidence of tunnel widening.^54,55 Furthermore, tensioning of the remnant is thought to lower the risk of cyclops lesions when compared to remnant preservation.¹²⁵ Although the difference between augmented repair and remnant tensioning seems small, the purpose of surgery is different. With augmented repair, the ligament can be approximated close to the femoral wall and the goal of surgery is to use the healing capacity that the ACL has in the proximal part of the ligament,¹²⁶ while with remnant tensioning the goal is only to benefit from some of the aforementioned advantages. Ahn and colleagues³⁶ were the first to perform this technique and stated, “Our concept is that the remnant tissue has only an additive effect.” Furthermore, with augmented repair multiple sutures are passed through the AM and PL bundle in order to sufficiently approximate the ligament to the femoral wall, while with the remnant tensioning technique generally one or a few sutures or lasso loop are passed through the proximal part to tension the ligament, prevent sagging of the remnant, and decrease the risk of cyclops lesions.^127,128

 

Several authors have recently performed remnant tensioning during ACL.^{36,47,125-127} Ahn and colleagues⁴⁷ reported excellent objective and subjective outcomes following this procedure and found that with re-arthroscopy nearly all patients had fair synovialization of the graft. Others have reported similarly good outcomes of these techniques.^125,129,130 However, studies comparing this treatment with normal ACL reconstruction and assessing outcomes, failure rates and proprioception are lacking.

Type IV Tears: Reconstruction With Remnant Preservation

Finally, in some patients the ligament is torn distally or the tissue quality is not optimal. In these patients, the remnant can be debrided to the part of good tissue quality in order to preserve the biology and minimize the risk for cyclops lesions. A standard reconstruction needs to be performed to restore the instability, but by preserving the remnant, advantages, such as proprioception,^44,49,50 graft vascularization,^50-52 an optical guide for tibial tunnel placement,⁵³ and a decreased incidence of tunnel widening^54,55 can be expected.

Lee and colleagues³⁷ presented the tibial remnant technique in which standard reconstruction was performed, and the tibial tunnel was drilled through the center of the remnant. In a later study, they compared remnant preservation with a remnant of <20% of the total ACL length with >20% of the length and found that proprioception was better with more remnant volume.⁴⁸ Similarly, Muneta and colleagues¹³¹ assessed the role of remnant length and found that remnant length is positively correlated with better stability measured on KT-1000 anteroposterior stability.

Several studies compared ACL reconstruction with remnant preservation vs conventional ACL reconstruction.^52,54,129 Takazawa and colleagues⁵² performed a retrospective study of 183 patients and found that patients in the remnant preservation group had significantly better KT-2000 stability, while they also reported a significantly lower graft rupture rate in this group (1.1% vs 7.1%) at 2-year follow-up. Hong and colleagues¹²⁹ performed a randomized clinical trial of 80 patients and did not find these differences, although there was a trend towards higher Lysholm scores in the remnant preservation group. Finally, Zhang and colleagues⁵⁴ performed a randomized clinical trial and found a lower incidence and amount of tibial tunnel widening in the preserving-remnant group when compared to the removing-remnant group. These studies show that there is likely a role for remnant preservation.

Type V Tears: Primary Repair

In some patients, the ligament is torn in the distal 10% of the ligament, which can occur as a distal avulsion tear or as a distal bony avulsion fracture.¹³² Bony avulsion fractures are most commonly seen in children whereas true distal soft tissue avulsion tears are very rare.¹³²

Treatments of these tear types include antegrade screw fixation, pullout sutures or the use of suture anchors in case of bony avulsion fractures and pullout sutures with tying over a bony bridge or ligament button in case of soft tissue avulsions. Leeberg and colleagues¹³² recently performed a systematic review of all studies reporting on treatment of distal avulsion fractures.They noted that most treatments were currently performed arthroscopically and that outcomes were generally good. Another recent biomechanical study compared antegrade screw fixation with suture anchor fixation and pullout suture fixation.¹³³ The authors noted that suture anchor fixation has slightly less displacement of the bony fragment when compared to screw fixation and pull-out sutures, and that the strength to failure was higher in the suture anchor fixation when compared to the pullout suture fixation. The outcomes of this study suggest that screw fixation and suture anchor fixation might be superior to pullout suture fixation, which might be interesting as with pullout suture fixation the ligament cannot be directly tensioned to the tibial footprint, which can lead to anteroposterior laxity.¹³² Clinical studies are necessary to assess the preferred treatment in these tear types but it seems that screw fixation is preferred in large bony avulsion fractures, while suture anchor fixation or pullout suture fixation can be used for soft tissue avulsion tears.

Complex Tears or Poor Tissue Quality: Reconstruction

If the tear is complex, multiple tears are present, or the tissue quality is poor, then preservation of the ligament is not possible, and in these cases a standard reconstruction should be performed.

Conclusion

When reviewing the literature of ACL preservation, it becomes clear that the evolution of surgical treatment of ACL injuries was biased. Preservation of the native ligament has many advantages, such as better proprioception, graft vascularization, an optical guide for tibial tunnel placement, and a decreased incidence of tunnel widening that can be expected. Furthermore, arthroscopic primary ACL repair is minimally invasive and does not burn any bridges for future reconstructions, if necessary. This is in addition to the other (theoretical) advantages of primary repair, such as restoration of native kinematics and a decreased risk of osteoarthritis. Modern advances have significantly changed the risk-benefit ratio that should make us reconsider ACL preservation approaches. Certainly, further research in this area is warranted. In this article we have presented a treatment algorithm for ACL preservation, which is based on tear location and remnant tissue quality.

Am J Orthop. 2016;45(7):E393-E405. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

In the first part of this series, “Preservation of the Anterior Cruciate Ligament: A Treatment Algorithm Based on Tear Location and Tissue Quality” we discussed the history of anterior cruciate ligament (ACL) preservation, and the historical outcomes of both open primary repair and augmented repair. We also presented our surgical treatment algorithm for ACL preservation, which is based on the tear location and tissue quality of the ligament remnant. In this article, we propose a modification of the Sherman classification¹ to identify the different tear types, and we will discuss the different surgical techniques that can be used for each one. Furthermore, we aim to provide an overview of the variations of these techniques that are seen in the literature. It is important to emphasize that these tear types and corresponding surgical techniques are to be seen as guidelines, rather than strict criteria, and that significant overlap between these tear types and surgical indications exist.

Assessment of Tear Type and Tissue Quality

The first assessment of the tear location and tissue quality is made using magnetic resonance imaging (MRI). Although MRI can give you an idea of where the tear is located, the final assessment for eligibility of each specific preservation technique is made during arthroscopy. Therefore, the routine preoperative discussion and informed consent process with the patient should encompass the gamut of surgical possibilities ranging from repair to reconstruction.

The Table shows our tear type classification, along with the corresponding preservation surgical techniques.

The location of the tear is described as the length of the distal remnant compared to the total ligament length (in percentage). The tissue quality indicates the minimum tissue quality that is generally necessary to perform a certain surgical technique. If the tissue quality is less than what is necessary for a specific ACL preservation technique, it may still be possible to perform another technique. For example, if a type II tear is found to have poor tissue quality in the upper half of the ligament, but good tissue quality in the lower half of the ligament, the remnant with poor quality is debrided and then the surgical procedure that corresponds to the length of good tissue quality can be performed (in this case remnant tensioning or remnant preservation with soft tissue graft reconstruction).

Surgical Preparation

In the operating room, the patient is placed in supine position on a standard operative table, such that the knee can be moved freely through its range of motion (ROM). The operative leg is then prepped and draped in standard fashion for knee arthroscopy. Standard knee arthroscopy equipment and implants are used, although some instruments from the standard shoulder set are also utilized. Anteromedial and anterolateral portals are created, and a general inspection of the knee is performed. By pulling the remnant ligament proximally using a broad tissue gasper, the available length of the remnant can be assessed. It is important to reduce possible anterior tibial subluxation in the sagittal plane in order to prevent “false” shortening of the distal ligament remnant. Once the length of the remnant tissue is assessed and the tissue quality is determined, the surgical preservation technique can be chosen (Table).

Type I Tears: Primary Repair

In order to be a candidate for arthroscopic primary repair, sufficient tissue length and tissue quality are necessary (Figures 1A and 1B, Table).

Sufficient length is seen if the distal remnant can be approximated to the femoral wall. Sufficient tissue quality is noted if sutures can be passed through the ligament and achieve good purchase. Once the ligament is deemed suitable for repair, a malleable Passport cannula (Arthrex) is placed in the anteromedial portal to facilitate suture passage and management of the repair.

Sutures are then passed through the anteromedial bundle using the Scorpion Suture Passer (Arthrex) with a No. 2 FiberWire suture (Arthrex) (Figure 1C). Suturing is commenced at the intact distal end of the anteromedial bundle and is advanced in an alternating, interlocking Bunnell-type pattern towards the avulsed proximal end with approximately 4 mm to 5 mm between each pass. In general, 3 to 4 passes can be made before the final pass exits via the avulsed end of the ligament towards the femur (Figure 1D). The same process is then repeated for the posterolateral bundle of the ACL remnant with a No. 2 TigerWire suture (Arthrex) to optimize suture management. With each subsequent pass of the sutures, it is important to assess tissue resistance to prevent perforation of a previous stitch. Mild resistance is normal, but the suture-passing device should be repositioned when notably increased resistance is encountered. In addition, placing all of the bites in the same plane should be avoided since this can allow the sutures to “cheese cut” along the collagen fibers of the ligament remnant rather than holding firm.

After passing the sutures through both bundles, the sutures are guided outside the knee using an accessory stab incision situated just above the medial portal. Using this portal, the ligament can be retracted away from the femoral footprint for optimal visibility. The femoral footprint is then roughed using a shaver or burr, and bleeding is induced to stimulate a local healing response,² while the sutures and the ACL are protected via the portal. With the knee in flexion, an accessory inferomedial portal is then created under direct visualization using a spinal needle for localization. Care should be taken to enable the appropriate trajectory for anchor placement to be achieved.

Many different techniques can be used to provide fixation of the ACL repair to the femoral footprint; the 2 most straightforward techniques are presented here. The first technique provides fixation with knotless suture anchors,^3,4 whereas in the second technique the sutures are transosseously passed, and tied over a bone bridge, as was performed in the 1970s and 1980s.

 

Suture Anchor Fixation

With the suture anchor fixation technique, the knee is flexed in 90°, the anteromedial bundle origin within the femoral footprint is identified, and a 4.5-mm x 20-mm hole is drilled, punched, or tapped, in the case of high bone density. The FiberWire sutures are then retrieved through the accessory portal and passed through a 4.75-mm Vented BioComposite SwiveLock suture anchor (Arthrex). The suture anchor for the anteromedial bundle is then deployed into the hole within the anteromedial footprint, while tensioning the ACL remnant to the wall with a visual gap of <1 mm (Figure 1E).⁵ The procedure is then repeated using another suture anchor with TigerWire sutures for the posterolateral bundle with the knee flexed at 110° to 115°. This ensures an optimal angle of approach and avoids perforating the posterior condyle with the anchor. The drill hole and anchor are placed into the origin of the posterolateral bundle within the femoral footprint. The order of bundle tensioning and repair may be varied depending on the particulars of each case.

Once the anchors are fully deployed and flush with the femoral footprint, the handle is removed and the additional core stitches are unloaded. Occasionally, the core stitches can be passed from lateral to medial through the proximal ligament remnant and tied down with an arthroscopic knot pusher to add extra compression of the remnant to the origin. The free ends of the repair sutures are cut with an Open Ended Suture Cutter (Arthrex) so that they are flush with the notch. The repair is now complete (Figure 1F). Using a probe, the ACL remnant is tested for tension and stiffness. Finally, cycling of the knee through the full ROM confirms anatomic positioning without impingement of the graft. Manual laxity testing should reveal minimal anteroposterior translation with a firm endpoint on Lachman examination intraoperatively.

Bone Bridge Fixation

With this technique, parallel drill holes are created exiting at each bundle origin. The repair stitches can then be retrieved and tensioned proximally. One way to accomplish this is by using an ACL femoral guide (Arthrex) that is placed via the anterolateral portal and is centered on the anteromedial bundle insertion. This device guides a cannulated RetroDrill (Arthrex) to drill through the lateral femoral condyle towards the anteromedial footprint. A passing wire can then be delivered through the cannulation and used to retrieve that anteromedial bundle repair stitches. This process can then be repeated for the posterolateral bundle and the associated repair stitches. Drill holes can also be made retrograde from a low anteromedial accessory portal using a slotted pit that can be used to shuttle the repair stitches. When all the repair sutures are passed, the ligament is tensioned while being visualized arthroscopically. The knee is held at 20° of flexion and a posterior drawer force can be applied, if necessary, to reduce the tibia to its anatomic position. The suture limbs are then tensioned and can be fixated using any of a multitude of techniques, including tying over a bony bridge, tying over a 4-hole ligament button, and tying to a post.

One disadvantage of the bone bridge fixation technique, however, is the suspensory fixation that is not as stiff as tensioning and fixating with suture anchors. Despite this disadvantage, however, the senior author (GSD) has achieved excellent results with this technique at longer-term follow-up in a small group of patients. One advantage of the bone bridge fixation technique is that the procedure has lower costs than fixation with suture anchors.

One Anchor Repair Fixation

Achtnich and colleagues⁶ recently published a slightly different technique for repairing type I tears. The authors passed a No. 2 FiberWire suture through the midsubstance of both bundles of the ACL remnant to create a modified Mason-Allen stitch configuration. Subsequently, they tensioned the remnant towards the middle of the ACL footprint (between the anteromedial and posterolateral footprint) using one PushLock suture anchor (Arthrex). They hypothesized that using 1 anchor would be enough fixation for tears amenable to repair, and that doing so would minimize the invasion of the bone.

The preference of the senior author (GSD) is, however, to use 2 suture anchors for each bundle in order to more anatomically and biomechanically repair the remnant, since both bundles have different biomechanical characteristics.⁷ Similarly, the preference of the senior author is to commence the suturing as distal as possible and pass multiple sutures towards the proximal end. This ensures that the last suture pass is exited very proximally, and ensures that the proximal end is approximated towards the femoral wall. One suture passed at the midsubstance portion of the remnant might cause a different tension pattern and prevent optimal re-approximation of the most proximal part towards the femoral wall. Future studies are necessary to assess the efficacy of different suture and fixation techniques as these are currently lacking.

 

Addition of Internal Brace

Over the last few years, the senior author has added an internal brace (FiberTape, Arthrex) to the repair technique, which was first performed by MacKay and colleagues.⁸ The added internal brace protects the repair and the healing process in the first few weeks and enables early ROM.

With this technique, the previously described arthroscopic primary repair technique is performed with suturing of both bundles. However, after punching, tapping, or drilling a hole in the anteromedial origin of the femoral footprint, the anteromedial anchor is first loaded with the FiberTape in addition to the repair stitches. After placing the anteromedial suture anchor in the femoral footprint, the internal brace is fixated proximally with the suture anchor into the femoral wall.

After the normal placement of the posterolateral suture anchors and completion of the repair, the internal brace has to be fixed distally. An ACL guide is used to drill a 2.4-mm drill pin up through the tibia from the anteromedial cortex (Figure 2A) and into the anterior half of the ACL tibial insertion. This is then switched for a Straight Microsuture Lasso (Arthrex) (Figure 2B), and the nitinol wire is retrieved out of the anteromedial portal with the 2 ends of the FiberTape. The FiberTape is then shuttled along the ACL substance anteriorly and down through the tibia (Figure 2C) where it fixed with a vented suture anchor near full extension after cycling the knee.

Others, however, have advocated fixing the internal brace independently of the repaired ligament and suture anchors.⁹ With this technique, tunnels are drilled in the femur and tibia and the internal brace construct is fixed proximally using a RetroButton (Arthrex) and fixed distally in the tibial metaphysis using a suture anchor. A disadvantage of this technique is that an extra femoral tunnel needs to be drilled, which is especially important in pediatric patients with the increased risk for growth disturbances.¹⁰

One Bundle Type I Tears: Single Bundle Augmented Repair

In some cases, the anteromedial or posterolateral bundle is a type I tear with good or excellent tissue quality, whereas the other bundle is not a type I tear or has poor tissue quality (Figure 3A). In these cases, a primary repair of one bundle is performed with a hamstring reconstruction of the other bundle.

First, a No. 2 FiberWire is used to make 4 to 5 passes from distal to proximal, as previously described. Then, the remnants of the irreparable bundle are debrided (Figure 3B). Subsequently, the semitendinosus tendon is harvested in standard fashion, or soft tissue allografts can be used.

Tunnels are then drilled for the reconstruction of the second bundle. The femoral tunnel, really a socket, is drilled at the femoral origin of the irreparable bundle using a FlipCutter (Arthrex) (Figure 3B). The tibial tunnel is localized using standard ACL guides, and drilled retrograde bicortically over a guide wire up into the ACL footprint of the bundle that is going to be reconstructed. Care must be taken to avoid damaging of the remnant that is being preserved. The sutures of the repaired remnant are then pulled back through the tibial tunnel and the repair stitches are passed through the TightRope RT (Arthrex) button. Once this is completed, all sutures and the button are passed up through the tibial tunnel and into the femoral socket. The proximal button is then flipped to engage the lateral femoral cortex in standard fashion (Figure 3C). Care should be taken to not tension the repairable remnant during the passage of the sutures through the femoral socket, as too much tension can cause tears in the remnant. Once the doubled hamstring graft has been advanced up into the femoral tunnel using the cinch stitches, the remnant bundle can then be tensioned towards the femoral wall (Figure 3D), and the sutures are tied over the TightRope RT button. Finally, the knee is cycled and held near full extension, while a BioComposite interference screw (Arthrex) is placed in the tibial tunnel for distal fixation.

Type II Tears: Augmented Repair

In patients with type II tears, primary repair is not possible as the length of the remnant is too short to firmly approximate the remnant towards the femoral wall (75%-90% of native tissue length) (Figure 4A). In these patients, an augmented repair of the entire ACL is performed using hamstring autograft or soft tissue allograft.

With this technique, repair stitches are passed into the anteromedial bundle of the remnant as previously described (Figure 4B). Keeping the repair stitches anteriorly in the anteromedial bundle tends to prevent entanglement during graft passage later in the case.

In some cases, it can be noted that the remnant has scarred to the femoral wall or the femoral notch. Crain and colleagues¹¹ previously described that in 20% of the cases, the remnant was reattached to the notch or the lateral wall of the notch. Because the reattachment is not at an anatomical position, the ligament should be detached from the femoral wall to allow the repair to be placed anatomically. In these cases, the senior author leaves the bundle initially attached to the femoral wall in order to make the passage of sutures through the ligament easier, and then he detaches the ligament to restore the anatomy.

Once the repair stitches are in place, a small accessory stab incision is made just above the medial portal. The repair stitches are parked here to keep them out of harms way. Traction on the repair stitches will retract the ACL away from the lateral wall of the notch and allow work to be performed here. A small opening notchplasty is generally performed to enhance visualization and to add a bleeding surface for enhanced healing. Next, the arthroscope is placed into the medial portal, which allows the femoral guide to be placed into the lateral portal. The femoral guide is positioned to optimize the femoral tunnel location in the center of the footprint. A small incision is made laterally over the condyle and through the iliotibial band to allow access to the lateral cortex of the lateral femoral condyle. The FlipCutter is then used to back-cut the femoral socket as described above. A FiberStick (Arthrex) passing suture is then placed in the femoral tunnel and brought out through the anteromedial portal.

Next, the tibial tunnel is drilled with a tibial guide at 55° inclination. The pin is drilled up into the center of the tibial footprint and this is over-reamed with a reamer. The reaming is stopped precisely upon breaking to proximal tibial cortex so as to minimize soft tissue damage of the ACL insertion fibers that are typically pristine. Then, a grasper is passed up and through the tunnel to retrieve the repair stitches and bring them out distally for later use. At the same time, the passing suture in the femoral is also retrieved distally. The soft tissue graft is proximally prepared with a TightRope RT button, and the repair stitches are passed through the button. The passing suture from the femoral socket is then used to shuttle the draw sutures and repair stitches up through the tibia, through the ACL remnant, and out the femoral socket (Figure 4C). The TightRope RT button is then engaged on the lateral femoral cortex in standard fashion. Using the cinch stitches, the graft is delivered through the tibia, up and through the center of the ACL remnant, and into the femoral socket. The knee is then cycled and the graft is tensioned distally in standard fashion, and fixed using a BioComposite interference screw. Finally, the repair stitches can be tensioned pulling the ligament remnant up as a sleeve around the hamstring graft (Figure 4D). They are then tied over the TightRope RT button using alternating half hitches tied with a knot pusher from laterally.

 

Type III Tears: Reconstruction With Remnant Tensioning

The previously discussed techniques have the goals of preserving as much native ligament remnant as possible, approximating the ligament remnant towards the femoral wall, and promoting healing of the ligament. In some cases, however, the ligament remnant is too short for healing (Figure 5A). Although the ligament cannot be approximated to the femoral wall in these cases, there is still an argument for ACL preservation, as was discussed in the first article of this series.

If the ligament length is between 25% and 75% of the native tissue length, the senior author performs a remnant tensioning technique.

The main goal of tensioning the remnant here is to prevent fluid egress into the tibial tunnel and avoid cyclops lesions, while theoretically improving graft vascularization and proprioception. With this technique, 1 or 2 passes are made through the most proximal part of the remnant and the sutures are then parked in an accessory portal as described above. The technique is very similar to that of an augmented repair; however, the graft dimensions here tend to be significantly larger due to the diminished contribution of the remnant. No strength can be expected of this small remnant since it will not be approximated close to the wall. The tibial tunnel is then drilled in standard fashion through the center of the tibial remnant (Figure 5B). After passing the repair sutures through the TightRope RT button of the soft tissue graft, the sutures of the graft and remnant are passed through the center of the distal remnant (Figure 5C). After the passage of the soft tissue graft sutures and button through the femoral socket, the button is flipped, and the graft is delivered. The knee is then cycled and the graft is fixated distally using a BioComposite interference screw. Finally, the remnant is then tensioned as a partial sleeve around the graft (Figure 5D) and tied over the TightRope RT button using alternating half hitches tied with a knot pusher from laterally.

Type IV Tears: Reconstruction With Remnant Preservation

Finally, in some cases, the distal remnant is small or the tissue quality in the largest part of the remnant is poor, and after debriding back to good tissue quality, only 10% to 25% of the native tissue length is left (Figure 6A). In these cases, the remnant is preserved, however, tensioning of the remnant with sutures is usually not necessary for the prevention of cyclops lesions. Nonetheless, it is important to debride the parts of the remnant ligament with poor tissue quality as mop-end patterns of the remnant may increase the chance of these lesions (Figure 6B).

In this situation, any of the standard ACL reconstruction techniques can be performed with simple attention being paid to preserving what is left of the tibial insertion site. At the very least, the small insertion remnant guides the anatomic placement of the graft, and prevents egress of joint fluid into the tibial tunnel and could minimize tunnel widening.

Theoretical benefits also include improved vascularity and proprioception. The tunnels are prepared, the grafts passed, and fixation placed per the surgeon’s preference (Figure 6C). The remnant is then arthroscopically checked to confirm that there is no impingement or risk of cyclops lesions (Figure 6D).

Type V Tears: Primary Repair

Finally, in some patients a soft tissue avulsion (Figure 7A) or bony avulsion of the distal attachment of the ACL can be seen. Both injuries are relatively rare, although bony avulsions are frequently seen in children, especially those younger than 12 years old. In these cases, the same techniques and theory that are applied to proximal avulsion type tears can be used and applied to distal avulsion type tears.

However, they must be applied in an upside down manner by working from proximal to distal in this case.

First, No. 2 FiberWire sutures are passed from proximal towards the distal end of the ligament in the anteromedial bundle, and the same process is then repeated for No. 2 TigerWire sutures for the posterolateral bundle. Then both sutures are exited at the distal avulsed end at the locations of the anteromedial and posterolateral footprints (Figure 7B). A 2.4-mm ACL guide wire and a Ninitol wire are used to drill 2 tunnels from the tibia towards the tibial footprint. The repair sutures are then retrieved through both tunnels (Figure 7C) and the sutures are tied distally over a ligament button after cycling of the knee (Figure 7D). This technique is very useful for soft tissue avulsions, or when there are only small flecks of bone or when the avulsed bone is significantly comminuted. If a large bony avulsion fragment is present, this technique can also be applied with some modification, although there have been multiple other techniques described in the literature that work well in this situation including fixation with screw and washer, or with suture anchors.

 

Complex Tear or Poor Tissue Quality: Reconstruction

In some cases, the tissue quality is poor, or the ligament has complex or multiple tears. Essentially, in these cases, there is nothing to preserve and a standard reconstruction approach is performed in these cases.

Conclusion

The uniform gold standard for all ACL tear types is currently primary reconstruction. However, several disadvantages of ACL reconstruction exist, while there are multiple advantages to the concept of ACL preservation. In this surgical technique article, we have discussed our tear type classification and the recommended surgical techniques for each. With this treatment algorithm, which is based on tear location and tissue quality, an optimal and minimally invasive treatment can be chosen for each individual patient. Future studies are needed to compare and contrast these treatments with the current gold standard of ACL reconstruction.

Am J Orthop. 2016;45(7):E406-E414. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

In the first part of this series, “Preservation of the Anterior Cruciate Ligament: A Treatment Algorithm Based on Tear Location and Tissue Quality” we discussed the history of anterior cruciate ligament (ACL) preservation, and the historical outcomes of both open primary repair and augmented repair. We also presented our surgical treatment algorithm for ACL preservation, which is based on the tear location and tissue quality of the ligament remnant. In this article, we propose a modification of the Sherman classification¹ to identify the different tear types, and we will discuss the different surgical techniques that can be used for each one. Furthermore, we aim to provide an overview of the variations of these techniques that are seen in the literature. It is important to emphasize that these tear types and corresponding surgical techniques are to be seen as guidelines, rather than strict criteria, and that significant overlap between these tear types and surgical indications exist.

Assessment of Tear Type and Tissue Quality

The first assessment of the tear location and tissue quality is made using magnetic resonance imaging (MRI). Although MRI can give you an idea of where the tear is located, the final assessment for eligibility of each specific preservation technique is made during arthroscopy. Therefore, the routine preoperative discussion and informed consent process with the patient should encompass the gamut of surgical possibilities ranging from repair to reconstruction.

The Table shows our tear type classification, along with the corresponding preservation surgical techniques.

The location of the tear is described as the length of the distal remnant compared to the total ligament length (in percentage). The tissue quality indicates the minimum tissue quality that is generally necessary to perform a certain surgical technique. If the tissue quality is less than what is necessary for a specific ACL preservation technique, it may still be possible to perform another technique. For example, if a type II tear is found to have poor tissue quality in the upper half of the ligament, but good tissue quality in the lower half of the ligament, the remnant with poor quality is debrided and then the surgical procedure that corresponds to the length of good tissue quality can be performed (in this case remnant tensioning or remnant preservation with soft tissue graft reconstruction).

Surgical Preparation

In the operating room, the patient is placed in supine position on a standard operative table, such that the knee can be moved freely through its range of motion (ROM). The operative leg is then prepped and draped in standard fashion for knee arthroscopy. Standard knee arthroscopy equipment and implants are used, although some instruments from the standard shoulder set are also utilized. Anteromedial and anterolateral portals are created, and a general inspection of the knee is performed. By pulling the remnant ligament proximally using a broad tissue gasper, the available length of the remnant can be assessed. It is important to reduce possible anterior tibial subluxation in the sagittal plane in order to prevent “false” shortening of the distal ligament remnant. Once the length of the remnant tissue is assessed and the tissue quality is determined, the surgical preservation technique can be chosen (Table).

Type I Tears: Primary Repair

In order to be a candidate for arthroscopic primary repair, sufficient tissue length and tissue quality are necessary (Figures 1A and 1B, Table).

Sufficient length is seen if the distal remnant can be approximated to the femoral wall. Sufficient tissue quality is noted if sutures can be passed through the ligament and achieve good purchase. Once the ligament is deemed suitable for repair, a malleable Passport cannula (Arthrex) is placed in the anteromedial portal to facilitate suture passage and management of the repair.

Sutures are then passed through the anteromedial bundle using the Scorpion Suture Passer (Arthrex) with a No. 2 FiberWire suture (Arthrex) (Figure 1C). Suturing is commenced at the intact distal end of the anteromedial bundle and is advanced in an alternating, interlocking Bunnell-type pattern towards the avulsed proximal end with approximately 4 mm to 5 mm between each pass. In general, 3 to 4 passes can be made before the final pass exits via the avulsed end of the ligament towards the femur (Figure 1D). The same process is then repeated for the posterolateral bundle of the ACL remnant with a No. 2 TigerWire suture (Arthrex) to optimize suture management. With each subsequent pass of the sutures, it is important to assess tissue resistance to prevent perforation of a previous stitch. Mild resistance is normal, but the suture-passing device should be repositioned when notably increased resistance is encountered. In addition, placing all of the bites in the same plane should be avoided since this can allow the sutures to “cheese cut” along the collagen fibers of the ligament remnant rather than holding firm.

After passing the sutures through both bundles, the sutures are guided outside the knee using an accessory stab incision situated just above the medial portal. Using this portal, the ligament can be retracted away from the femoral footprint for optimal visibility. The femoral footprint is then roughed using a shaver or burr, and bleeding is induced to stimulate a local healing response,² while the sutures and the ACL are protected via the portal. With the knee in flexion, an accessory inferomedial portal is then created under direct visualization using a spinal needle for localization. Care should be taken to enable the appropriate trajectory for anchor placement to be achieved.

Many different techniques can be used to provide fixation of the ACL repair to the femoral footprint; the 2 most straightforward techniques are presented here. The first technique provides fixation with knotless suture anchors,^3,4 whereas in the second technique the sutures are transosseously passed, and tied over a bone bridge, as was performed in the 1970s and 1980s.

 

Suture Anchor Fixation

With the suture anchor fixation technique, the knee is flexed in 90°, the anteromedial bundle origin within the femoral footprint is identified, and a 4.5-mm x 20-mm hole is drilled, punched, or tapped, in the case of high bone density. The FiberWire sutures are then retrieved through the accessory portal and passed through a 4.75-mm Vented BioComposite SwiveLock suture anchor (Arthrex). The suture anchor for the anteromedial bundle is then deployed into the hole within the anteromedial footprint, while tensioning the ACL remnant to the wall with a visual gap of <1 mm (Figure 1E).⁵ The procedure is then repeated using another suture anchor with TigerWire sutures for the posterolateral bundle with the knee flexed at 110° to 115°. This ensures an optimal angle of approach and avoids perforating the posterior condyle with the anchor. The drill hole and anchor are placed into the origin of the posterolateral bundle within the femoral footprint. The order of bundle tensioning and repair may be varied depending on the particulars of each case.

Once the anchors are fully deployed and flush with the femoral footprint, the handle is removed and the additional core stitches are unloaded. Occasionally, the core stitches can be passed from lateral to medial through the proximal ligament remnant and tied down with an arthroscopic knot pusher to add extra compression of the remnant to the origin. The free ends of the repair sutures are cut with an Open Ended Suture Cutter (Arthrex) so that they are flush with the notch. The repair is now complete (Figure 1F). Using a probe, the ACL remnant is tested for tension and stiffness. Finally, cycling of the knee through the full ROM confirms anatomic positioning without impingement of the graft. Manual laxity testing should reveal minimal anteroposterior translation with a firm endpoint on Lachman examination intraoperatively.

Bone Bridge Fixation

With this technique, parallel drill holes are created exiting at each bundle origin. The repair stitches can then be retrieved and tensioned proximally. One way to accomplish this is by using an ACL femoral guide (Arthrex) that is placed via the anterolateral portal and is centered on the anteromedial bundle insertion. This device guides a cannulated RetroDrill (Arthrex) to drill through the lateral femoral condyle towards the anteromedial footprint. A passing wire can then be delivered through the cannulation and used to retrieve that anteromedial bundle repair stitches. This process can then be repeated for the posterolateral bundle and the associated repair stitches. Drill holes can also be made retrograde from a low anteromedial accessory portal using a slotted pit that can be used to shuttle the repair stitches. When all the repair sutures are passed, the ligament is tensioned while being visualized arthroscopically. The knee is held at 20° of flexion and a posterior drawer force can be applied, if necessary, to reduce the tibia to its anatomic position. The suture limbs are then tensioned and can be fixated using any of a multitude of techniques, including tying over a bony bridge, tying over a 4-hole ligament button, and tying to a post.

One disadvantage of the bone bridge fixation technique, however, is the suspensory fixation that is not as stiff as tensioning and fixating with suture anchors. Despite this disadvantage, however, the senior author (GSD) has achieved excellent results with this technique at longer-term follow-up in a small group of patients. One advantage of the bone bridge fixation technique is that the procedure has lower costs than fixation with suture anchors.

One Anchor Repair Fixation

Achtnich and colleagues⁶ recently published a slightly different technique for repairing type I tears. The authors passed a No. 2 FiberWire suture through the midsubstance of both bundles of the ACL remnant to create a modified Mason-Allen stitch configuration. Subsequently, they tensioned the remnant towards the middle of the ACL footprint (between the anteromedial and posterolateral footprint) using one PushLock suture anchor (Arthrex). They hypothesized that using 1 anchor would be enough fixation for tears amenable to repair, and that doing so would minimize the invasion of the bone.

The preference of the senior author (GSD) is, however, to use 2 suture anchors for each bundle in order to more anatomically and biomechanically repair the remnant, since both bundles have different biomechanical characteristics.⁷ Similarly, the preference of the senior author is to commence the suturing as distal as possible and pass multiple sutures towards the proximal end. This ensures that the last suture pass is exited very proximally, and ensures that the proximal end is approximated towards the femoral wall. One suture passed at the midsubstance portion of the remnant might cause a different tension pattern and prevent optimal re-approximation of the most proximal part towards the femoral wall. Future studies are necessary to assess the efficacy of different suture and fixation techniques as these are currently lacking.

 

Addition of Internal Brace

Over the last few years, the senior author has added an internal brace (FiberTape, Arthrex) to the repair technique, which was first performed by MacKay and colleagues.⁸ The added internal brace protects the repair and the healing process in the first few weeks and enables early ROM.

With this technique, the previously described arthroscopic primary repair technique is performed with suturing of both bundles. However, after punching, tapping, or drilling a hole in the anteromedial origin of the femoral footprint, the anteromedial anchor is first loaded with the FiberTape in addition to the repair stitches. After placing the anteromedial suture anchor in the femoral footprint, the internal brace is fixated proximally with the suture anchor into the femoral wall.

After the normal placement of the posterolateral suture anchors and completion of the repair, the internal brace has to be fixed distally. An ACL guide is used to drill a 2.4-mm drill pin up through the tibia from the anteromedial cortex (Figure 2A) and into the anterior half of the ACL tibial insertion. This is then switched for a Straight Microsuture Lasso (Arthrex) (Figure 2B), and the nitinol wire is retrieved out of the anteromedial portal with the 2 ends of the FiberTape. The FiberTape is then shuttled along the ACL substance anteriorly and down through the tibia (Figure 2C) where it fixed with a vented suture anchor near full extension after cycling the knee.

Others, however, have advocated fixing the internal brace independently of the repaired ligament and suture anchors.⁹ With this technique, tunnels are drilled in the femur and tibia and the internal brace construct is fixed proximally using a RetroButton (Arthrex) and fixed distally in the tibial metaphysis using a suture anchor. A disadvantage of this technique is that an extra femoral tunnel needs to be drilled, which is especially important in pediatric patients with the increased risk for growth disturbances.¹⁰

One Bundle Type I Tears: Single Bundle Augmented Repair

In some cases, the anteromedial or posterolateral bundle is a type I tear with good or excellent tissue quality, whereas the other bundle is not a type I tear or has poor tissue quality (Figure 3A). In these cases, a primary repair of one bundle is performed with a hamstring reconstruction of the other bundle.

First, a No. 2 FiberWire is used to make 4 to 5 passes from distal to proximal, as previously described. Then, the remnants of the irreparable bundle are debrided (Figure 3B). Subsequently, the semitendinosus tendon is harvested in standard fashion, or soft tissue allografts can be used.

Tunnels are then drilled for the reconstruction of the second bundle. The femoral tunnel, really a socket, is drilled at the femoral origin of the irreparable bundle using a FlipCutter (Arthrex) (Figure 3B). The tibial tunnel is localized using standard ACL guides, and drilled retrograde bicortically over a guide wire up into the ACL footprint of the bundle that is going to be reconstructed. Care must be taken to avoid damaging of the remnant that is being preserved. The sutures of the repaired remnant are then pulled back through the tibial tunnel and the repair stitches are passed through the TightRope RT (Arthrex) button. Once this is completed, all sutures and the button are passed up through the tibial tunnel and into the femoral socket. The proximal button is then flipped to engage the lateral femoral cortex in standard fashion (Figure 3C). Care should be taken to not tension the repairable remnant during the passage of the sutures through the femoral socket, as too much tension can cause tears in the remnant. Once the doubled hamstring graft has been advanced up into the femoral tunnel using the cinch stitches, the remnant bundle can then be tensioned towards the femoral wall (Figure 3D), and the sutures are tied over the TightRope RT button. Finally, the knee is cycled and held near full extension, while a BioComposite interference screw (Arthrex) is placed in the tibial tunnel for distal fixation.

Type II Tears: Augmented Repair

In patients with type II tears, primary repair is not possible as the length of the remnant is too short to firmly approximate the remnant towards the femoral wall (75%-90% of native tissue length) (Figure 4A). In these patients, an augmented repair of the entire ACL is performed using hamstring autograft or soft tissue allograft.

With this technique, repair stitches are passed into the anteromedial bundle of the remnant as previously described (Figure 4B). Keeping the repair stitches anteriorly in the anteromedial bundle tends to prevent entanglement during graft passage later in the case.

In some cases, it can be noted that the remnant has scarred to the femoral wall or the femoral notch. Crain and colleagues¹¹ previously described that in 20% of the cases, the remnant was reattached to the notch or the lateral wall of the notch. Because the reattachment is not at an anatomical position, the ligament should be detached from the femoral wall to allow the repair to be placed anatomically. In these cases, the senior author leaves the bundle initially attached to the femoral wall in order to make the passage of sutures through the ligament easier, and then he detaches the ligament to restore the anatomy.

Once the repair stitches are in place, a small accessory stab incision is made just above the medial portal. The repair stitches are parked here to keep them out of harms way. Traction on the repair stitches will retract the ACL away from the lateral wall of the notch and allow work to be performed here. A small opening notchplasty is generally performed to enhance visualization and to add a bleeding surface for enhanced healing. Next, the arthroscope is placed into the medial portal, which allows the femoral guide to be placed into the lateral portal. The femoral guide is positioned to optimize the femoral tunnel location in the center of the footprint. A small incision is made laterally over the condyle and through the iliotibial band to allow access to the lateral cortex of the lateral femoral condyle. The FlipCutter is then used to back-cut the femoral socket as described above. A FiberStick (Arthrex) passing suture is then placed in the femoral tunnel and brought out through the anteromedial portal.

Next, the tibial tunnel is drilled with a tibial guide at 55° inclination. The pin is drilled up into the center of the tibial footprint and this is over-reamed with a reamer. The reaming is stopped precisely upon breaking to proximal tibial cortex so as to minimize soft tissue damage of the ACL insertion fibers that are typically pristine. Then, a grasper is passed up and through the tunnel to retrieve the repair stitches and bring them out distally for later use. At the same time, the passing suture in the femoral is also retrieved distally. The soft tissue graft is proximally prepared with a TightRope RT button, and the repair stitches are passed through the button. The passing suture from the femoral socket is then used to shuttle the draw sutures and repair stitches up through the tibia, through the ACL remnant, and out the femoral socket (Figure 4C). The TightRope RT button is then engaged on the lateral femoral cortex in standard fashion. Using the cinch stitches, the graft is delivered through the tibia, up and through the center of the ACL remnant, and into the femoral socket. The knee is then cycled and the graft is tensioned distally in standard fashion, and fixed using a BioComposite interference screw. Finally, the repair stitches can be tensioned pulling the ligament remnant up as a sleeve around the hamstring graft (Figure 4D). They are then tied over the TightRope RT button using alternating half hitches tied with a knot pusher from laterally.

 

Type III Tears: Reconstruction With Remnant Tensioning

The previously discussed techniques have the goals of preserving as much native ligament remnant as possible, approximating the ligament remnant towards the femoral wall, and promoting healing of the ligament. In some cases, however, the ligament remnant is too short for healing (Figure 5A). Although the ligament cannot be approximated to the femoral wall in these cases, there is still an argument for ACL preservation, as was discussed in the first article of this series.

If the ligament length is between 25% and 75% of the native tissue length, the senior author performs a remnant tensioning technique.

The main goal of tensioning the remnant here is to prevent fluid egress into the tibial tunnel and avoid cyclops lesions, while theoretically improving graft vascularization and proprioception. With this technique, 1 or 2 passes are made through the most proximal part of the remnant and the sutures are then parked in an accessory portal as described above. The technique is very similar to that of an augmented repair; however, the graft dimensions here tend to be significantly larger due to the diminished contribution of the remnant. No strength can be expected of this small remnant since it will not be approximated close to the wall. The tibial tunnel is then drilled in standard fashion through the center of the tibial remnant (Figure 5B). After passing the repair sutures through the TightRope RT button of the soft tissue graft, the sutures of the graft and remnant are passed through the center of the distal remnant (Figure 5C). After the passage of the soft tissue graft sutures and button through the femoral socket, the button is flipped, and the graft is delivered. The knee is then cycled and the graft is fixated distally using a BioComposite interference screw. Finally, the remnant is then tensioned as a partial sleeve around the graft (Figure 5D) and tied over the TightRope RT button using alternating half hitches tied with a knot pusher from laterally.

Type IV Tears: Reconstruction With Remnant Preservation

Finally, in some cases, the distal remnant is small or the tissue quality in the largest part of the remnant is poor, and after debriding back to good tissue quality, only 10% to 25% of the native tissue length is left (Figure 6A). In these cases, the remnant is preserved, however, tensioning of the remnant with sutures is usually not necessary for the prevention of cyclops lesions. Nonetheless, it is important to debride the parts of the remnant ligament with poor tissue quality as mop-end patterns of the remnant may increase the chance of these lesions (Figure 6B).

In this situation, any of the standard ACL reconstruction techniques can be performed with simple attention being paid to preserving what is left of the tibial insertion site. At the very least, the small insertion remnant guides the anatomic placement of the graft, and prevents egress of joint fluid into the tibial tunnel and could minimize tunnel widening.

Theoretical benefits also include improved vascularity and proprioception. The tunnels are prepared, the grafts passed, and fixation placed per the surgeon’s preference (Figure 6C). The remnant is then arthroscopically checked to confirm that there is no impingement or risk of cyclops lesions (Figure 6D).

Type V Tears: Primary Repair

Finally, in some patients a soft tissue avulsion (Figure 7A) or bony avulsion of the distal attachment of the ACL can be seen. Both injuries are relatively rare, although bony avulsions are frequently seen in children, especially those younger than 12 years old. In these cases, the same techniques and theory that are applied to proximal avulsion type tears can be used and applied to distal avulsion type tears.

However, they must be applied in an upside down manner by working from proximal to distal in this case.

First, No. 2 FiberWire sutures are passed from proximal towards the distal end of the ligament in the anteromedial bundle, and the same process is then repeated for No. 2 TigerWire sutures for the posterolateral bundle. Then both sutures are exited at the distal avulsed end at the locations of the anteromedial and posterolateral footprints (Figure 7B). A 2.4-mm ACL guide wire and a Ninitol wire are used to drill 2 tunnels from the tibia towards the tibial footprint. The repair sutures are then retrieved through both tunnels (Figure 7C) and the sutures are tied distally over a ligament button after cycling of the knee (Figure 7D). This technique is very useful for soft tissue avulsions, or when there are only small flecks of bone or when the avulsed bone is significantly comminuted. If a large bony avulsion fragment is present, this technique can also be applied with some modification, although there have been multiple other techniques described in the literature that work well in this situation including fixation with screw and washer, or with suture anchors.

 

Complex Tear or Poor Tissue Quality: Reconstruction

In some cases, the tissue quality is poor, or the ligament has complex or multiple tears. Essentially, in these cases, there is nothing to preserve and a standard reconstruction approach is performed in these cases.

Conclusion

The uniform gold standard for all ACL tear types is currently primary reconstruction. However, several disadvantages of ACL reconstruction exist, while there are multiple advantages to the concept of ACL preservation. In this surgical technique article, we have discussed our tear type classification and the recommended surgical techniques for each. With this treatment algorithm, which is based on tear location and tissue quality, an optimal and minimally invasive treatment can be chosen for each individual patient. Future studies are needed to compare and contrast these treatments with the current gold standard of ACL reconstruction.

Am J Orthop. 2016;45(7):E406-E414. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

In the first part of this series, “Preservation of the Anterior Cruciate Ligament: A Treatment Algorithm Based on Tear Location and Tissue Quality” we discussed the history of anterior cruciate ligament (ACL) preservation, and the historical outcomes of both open primary repair and augmented repair. We also presented our surgical treatment algorithm for ACL preservation, which is based on the tear location and tissue quality of the ligament remnant. In this article, we propose a modification of the Sherman classification¹ to identify the different tear types, and we will discuss the different surgical techniques that can be used for each one. Furthermore, we aim to provide an overview of the variations of these techniques that are seen in the literature. It is important to emphasize that these tear types and corresponding surgical techniques are to be seen as guidelines, rather than strict criteria, and that significant overlap between these tear types and surgical indications exist.

Assessment of Tear Type and Tissue Quality

The first assessment of the tear location and tissue quality is made using magnetic resonance imaging (MRI). Although MRI can give you an idea of where the tear is located, the final assessment for eligibility of each specific preservation technique is made during arthroscopy. Therefore, the routine preoperative discussion and informed consent process with the patient should encompass the gamut of surgical possibilities ranging from repair to reconstruction.

The Table shows our tear type classification, along with the corresponding preservation surgical techniques.

The location of the tear is described as the length of the distal remnant compared to the total ligament length (in percentage). The tissue quality indicates the minimum tissue quality that is generally necessary to perform a certain surgical technique. If the tissue quality is less than what is necessary for a specific ACL preservation technique, it may still be possible to perform another technique. For example, if a type II tear is found to have poor tissue quality in the upper half of the ligament, but good tissue quality in the lower half of the ligament, the remnant with poor quality is debrided and then the surgical procedure that corresponds to the length of good tissue quality can be performed (in this case remnant tensioning or remnant preservation with soft tissue graft reconstruction).

Surgical Preparation

In the operating room, the patient is placed in supine position on a standard operative table, such that the knee can be moved freely through its range of motion (ROM). The operative leg is then prepped and draped in standard fashion for knee arthroscopy. Standard knee arthroscopy equipment and implants are used, although some instruments from the standard shoulder set are also utilized. Anteromedial and anterolateral portals are created, and a general inspection of the knee is performed. By pulling the remnant ligament proximally using a broad tissue gasper, the available length of the remnant can be assessed. It is important to reduce possible anterior tibial subluxation in the sagittal plane in order to prevent “false” shortening of the distal ligament remnant. Once the length of the remnant tissue is assessed and the tissue quality is determined, the surgical preservation technique can be chosen (Table).

Type I Tears: Primary Repair

In order to be a candidate for arthroscopic primary repair, sufficient tissue length and tissue quality are necessary (Figures 1A and 1B, Table).

Sufficient length is seen if the distal remnant can be approximated to the femoral wall. Sufficient tissue quality is noted if sutures can be passed through the ligament and achieve good purchase. Once the ligament is deemed suitable for repair, a malleable Passport cannula (Arthrex) is placed in the anteromedial portal to facilitate suture passage and management of the repair.

Sutures are then passed through the anteromedial bundle using the Scorpion Suture Passer (Arthrex) with a No. 2 FiberWire suture (Arthrex) (Figure 1C). Suturing is commenced at the intact distal end of the anteromedial bundle and is advanced in an alternating, interlocking Bunnell-type pattern towards the avulsed proximal end with approximately 4 mm to 5 mm between each pass. In general, 3 to 4 passes can be made before the final pass exits via the avulsed end of the ligament towards the femur (Figure 1D). The same process is then repeated for the posterolateral bundle of the ACL remnant with a No. 2 TigerWire suture (Arthrex) to optimize suture management. With each subsequent pass of the sutures, it is important to assess tissue resistance to prevent perforation of a previous stitch. Mild resistance is normal, but the suture-passing device should be repositioned when notably increased resistance is encountered. In addition, placing all of the bites in the same plane should be avoided since this can allow the sutures to “cheese cut” along the collagen fibers of the ligament remnant rather than holding firm.

After passing the sutures through both bundles, the sutures are guided outside the knee using an accessory stab incision situated just above the medial portal. Using this portal, the ligament can be retracted away from the femoral footprint for optimal visibility. The femoral footprint is then roughed using a shaver or burr, and bleeding is induced to stimulate a local healing response,² while the sutures and the ACL are protected via the portal. With the knee in flexion, an accessory inferomedial portal is then created under direct visualization using a spinal needle for localization. Care should be taken to enable the appropriate trajectory for anchor placement to be achieved.

Many different techniques can be used to provide fixation of the ACL repair to the femoral footprint; the 2 most straightforward techniques are presented here. The first technique provides fixation with knotless suture anchors,^3,4 whereas in the second technique the sutures are transosseously passed, and tied over a bone bridge, as was performed in the 1970s and 1980s.

 

Suture Anchor Fixation

With the suture anchor fixation technique, the knee is flexed in 90°, the anteromedial bundle origin within the femoral footprint is identified, and a 4.5-mm x 20-mm hole is drilled, punched, or tapped, in the case of high bone density. The FiberWire sutures are then retrieved through the accessory portal and passed through a 4.75-mm Vented BioComposite SwiveLock suture anchor (Arthrex). The suture anchor for the anteromedial bundle is then deployed into the hole within the anteromedial footprint, while tensioning the ACL remnant to the wall with a visual gap of <1 mm (Figure 1E).⁵ The procedure is then repeated using another suture anchor with TigerWire sutures for the posterolateral bundle with the knee flexed at 110° to 115°. This ensures an optimal angle of approach and avoids perforating the posterior condyle with the anchor. The drill hole and anchor are placed into the origin of the posterolateral bundle within the femoral footprint. The order of bundle tensioning and repair may be varied depending on the particulars of each case.

Once the anchors are fully deployed and flush with the femoral footprint, the handle is removed and the additional core stitches are unloaded. Occasionally, the core stitches can be passed from lateral to medial through the proximal ligament remnant and tied down with an arthroscopic knot pusher to add extra compression of the remnant to the origin. The free ends of the repair sutures are cut with an Open Ended Suture Cutter (Arthrex) so that they are flush with the notch. The repair is now complete (Figure 1F). Using a probe, the ACL remnant is tested for tension and stiffness. Finally, cycling of the knee through the full ROM confirms anatomic positioning without impingement of the graft. Manual laxity testing should reveal minimal anteroposterior translation with a firm endpoint on Lachman examination intraoperatively.

Bone Bridge Fixation

With this technique, parallel drill holes are created exiting at each bundle origin. The repair stitches can then be retrieved and tensioned proximally. One way to accomplish this is by using an ACL femoral guide (Arthrex) that is placed via the anterolateral portal and is centered on the anteromedial bundle insertion. This device guides a cannulated RetroDrill (Arthrex) to drill through the lateral femoral condyle towards the anteromedial footprint. A passing wire can then be delivered through the cannulation and used to retrieve that anteromedial bundle repair stitches. This process can then be repeated for the posterolateral bundle and the associated repair stitches. Drill holes can also be made retrograde from a low anteromedial accessory portal using a slotted pit that can be used to shuttle the repair stitches. When all the repair sutures are passed, the ligament is tensioned while being visualized arthroscopically. The knee is held at 20° of flexion and a posterior drawer force can be applied, if necessary, to reduce the tibia to its anatomic position. The suture limbs are then tensioned and can be fixated using any of a multitude of techniques, including tying over a bony bridge, tying over a 4-hole ligament button, and tying to a post.

One disadvantage of the bone bridge fixation technique, however, is the suspensory fixation that is not as stiff as tensioning and fixating with suture anchors. Despite this disadvantage, however, the senior author (GSD) has achieved excellent results with this technique at longer-term follow-up in a small group of patients. One advantage of the bone bridge fixation technique is that the procedure has lower costs than fixation with suture anchors.

One Anchor Repair Fixation

Achtnich and colleagues⁶ recently published a slightly different technique for repairing type I tears. The authors passed a No. 2 FiberWire suture through the midsubstance of both bundles of the ACL remnant to create a modified Mason-Allen stitch configuration. Subsequently, they tensioned the remnant towards the middle of the ACL footprint (between the anteromedial and posterolateral footprint) using one PushLock suture anchor (Arthrex). They hypothesized that using 1 anchor would be enough fixation for tears amenable to repair, and that doing so would minimize the invasion of the bone.

The preference of the senior author (GSD) is, however, to use 2 suture anchors for each bundle in order to more anatomically and biomechanically repair the remnant, since both bundles have different biomechanical characteristics.⁷ Similarly, the preference of the senior author is to commence the suturing as distal as possible and pass multiple sutures towards the proximal end. This ensures that the last suture pass is exited very proximally, and ensures that the proximal end is approximated towards the femoral wall. One suture passed at the midsubstance portion of the remnant might cause a different tension pattern and prevent optimal re-approximation of the most proximal part towards the femoral wall. Future studies are necessary to assess the efficacy of different suture and fixation techniques as these are currently lacking.

 

Addition of Internal Brace

Over the last few years, the senior author has added an internal brace (FiberTape, Arthrex) to the repair technique, which was first performed by MacKay and colleagues.⁸ The added internal brace protects the repair and the healing process in the first few weeks and enables early ROM.

With this technique, the previously described arthroscopic primary repair technique is performed with suturing of both bundles. However, after punching, tapping, or drilling a hole in the anteromedial origin of the femoral footprint, the anteromedial anchor is first loaded with the FiberTape in addition to the repair stitches. After placing the anteromedial suture anchor in the femoral footprint, the internal brace is fixated proximally with the suture anchor into the femoral wall.

After the normal placement of the posterolateral suture anchors and completion of the repair, the internal brace has to be fixed distally. An ACL guide is used to drill a 2.4-mm drill pin up through the tibia from the anteromedial cortex (Figure 2A) and into the anterior half of the ACL tibial insertion. This is then switched for a Straight Microsuture Lasso (Arthrex) (Figure 2B), and the nitinol wire is retrieved out of the anteromedial portal with the 2 ends of the FiberTape. The FiberTape is then shuttled along the ACL substance anteriorly and down through the tibia (Figure 2C) where it fixed with a vented suture anchor near full extension after cycling the knee.

Others, however, have advocated fixing the internal brace independently of the repaired ligament and suture anchors.⁹ With this technique, tunnels are drilled in the femur and tibia and the internal brace construct is fixed proximally using a RetroButton (Arthrex) and fixed distally in the tibial metaphysis using a suture anchor. A disadvantage of this technique is that an extra femoral tunnel needs to be drilled, which is especially important in pediatric patients with the increased risk for growth disturbances.¹⁰

One Bundle Type I Tears: Single Bundle Augmented Repair

In some cases, the anteromedial or posterolateral bundle is a type I tear with good or excellent tissue quality, whereas the other bundle is not a type I tear or has poor tissue quality (Figure 3A). In these cases, a primary repair of one bundle is performed with a hamstring reconstruction of the other bundle.

First, a No. 2 FiberWire is used to make 4 to 5 passes from distal to proximal, as previously described. Then, the remnants of the irreparable bundle are debrided (Figure 3B). Subsequently, the semitendinosus tendon is harvested in standard fashion, or soft tissue allografts can be used.

Tunnels are then drilled for the reconstruction of the second bundle. The femoral tunnel, really a socket, is drilled at the femoral origin of the irreparable bundle using a FlipCutter (Arthrex) (Figure 3B). The tibial tunnel is localized using standard ACL guides, and drilled retrograde bicortically over a guide wire up into the ACL footprint of the bundle that is going to be reconstructed. Care must be taken to avoid damaging of the remnant that is being preserved. The sutures of the repaired remnant are then pulled back through the tibial tunnel and the repair stitches are passed through the TightRope RT (Arthrex) button. Once this is completed, all sutures and the button are passed up through the tibial tunnel and into the femoral socket. The proximal button is then flipped to engage the lateral femoral cortex in standard fashion (Figure 3C). Care should be taken to not tension the repairable remnant during the passage of the sutures through the femoral socket, as too much tension can cause tears in the remnant. Once the doubled hamstring graft has been advanced up into the femoral tunnel using the cinch stitches, the remnant bundle can then be tensioned towards the femoral wall (Figure 3D), and the sutures are tied over the TightRope RT button. Finally, the knee is cycled and held near full extension, while a BioComposite interference screw (Arthrex) is placed in the tibial tunnel for distal fixation.

Type II Tears: Augmented Repair

In patients with type II tears, primary repair is not possible as the length of the remnant is too short to firmly approximate the remnant towards the femoral wall (75%-90% of native tissue length) (Figure 4A). In these patients, an augmented repair of the entire ACL is performed using hamstring autograft or soft tissue allograft.

With this technique, repair stitches are passed into the anteromedial bundle of the remnant as previously described (Figure 4B). Keeping the repair stitches anteriorly in the anteromedial bundle tends to prevent entanglement during graft passage later in the case.

In some cases, it can be noted that the remnant has scarred to the femoral wall or the femoral notch. Crain and colleagues¹¹ previously described that in 20% of the cases, the remnant was reattached to the notch or the lateral wall of the notch. Because the reattachment is not at an anatomical position, the ligament should be detached from the femoral wall to allow the repair to be placed anatomically. In these cases, the senior author leaves the bundle initially attached to the femoral wall in order to make the passage of sutures through the ligament easier, and then he detaches the ligament to restore the anatomy.

Once the repair stitches are in place, a small accessory stab incision is made just above the medial portal. The repair stitches are parked here to keep them out of harms way. Traction on the repair stitches will retract the ACL away from the lateral wall of the notch and allow work to be performed here. A small opening notchplasty is generally performed to enhance visualization and to add a bleeding surface for enhanced healing. Next, the arthroscope is placed into the medial portal, which allows the femoral guide to be placed into the lateral portal. The femoral guide is positioned to optimize the femoral tunnel location in the center of the footprint. A small incision is made laterally over the condyle and through the iliotibial band to allow access to the lateral cortex of the lateral femoral condyle. The FlipCutter is then used to back-cut the femoral socket as described above. A FiberStick (Arthrex) passing suture is then placed in the femoral tunnel and brought out through the anteromedial portal.

Next, the tibial tunnel is drilled with a tibial guide at 55° inclination. The pin is drilled up into the center of the tibial footprint and this is over-reamed with a reamer. The reaming is stopped precisely upon breaking to proximal tibial cortex so as to minimize soft tissue damage of the ACL insertion fibers that are typically pristine. Then, a grasper is passed up and through the tunnel to retrieve the repair stitches and bring them out distally for later use. At the same time, the passing suture in the femoral is also retrieved distally. The soft tissue graft is proximally prepared with a TightRope RT button, and the repair stitches are passed through the button. The passing suture from the femoral socket is then used to shuttle the draw sutures and repair stitches up through the tibia, through the ACL remnant, and out the femoral socket (Figure 4C). The TightRope RT button is then engaged on the lateral femoral cortex in standard fashion. Using the cinch stitches, the graft is delivered through the tibia, up and through the center of the ACL remnant, and into the femoral socket. The knee is then cycled and the graft is tensioned distally in standard fashion, and fixed using a BioComposite interference screw. Finally, the repair stitches can be tensioned pulling the ligament remnant up as a sleeve around the hamstring graft (Figure 4D). They are then tied over the TightRope RT button using alternating half hitches tied with a knot pusher from laterally.

 

Type III Tears: Reconstruction With Remnant Tensioning

The previously discussed techniques have the goals of preserving as much native ligament remnant as possible, approximating the ligament remnant towards the femoral wall, and promoting healing of the ligament. In some cases, however, the ligament remnant is too short for healing (Figure 5A). Although the ligament cannot be approximated to the femoral wall in these cases, there is still an argument for ACL preservation, as was discussed in the first article of this series.

If the ligament length is between 25% and 75% of the native tissue length, the senior author performs a remnant tensioning technique.

The main goal of tensioning the remnant here is to prevent fluid egress into the tibial tunnel and avoid cyclops lesions, while theoretically improving graft vascularization and proprioception. With this technique, 1 or 2 passes are made through the most proximal part of the remnant and the sutures are then parked in an accessory portal as described above. The technique is very similar to that of an augmented repair; however, the graft dimensions here tend to be significantly larger due to the diminished contribution of the remnant. No strength can be expected of this small remnant since it will not be approximated close to the wall. The tibial tunnel is then drilled in standard fashion through the center of the tibial remnant (Figure 5B). After passing the repair sutures through the TightRope RT button of the soft tissue graft, the sutures of the graft and remnant are passed through the center of the distal remnant (Figure 5C). After the passage of the soft tissue graft sutures and button through the femoral socket, the button is flipped, and the graft is delivered. The knee is then cycled and the graft is fixated distally using a BioComposite interference screw. Finally, the remnant is then tensioned as a partial sleeve around the graft (Figure 5D) and tied over the TightRope RT button using alternating half hitches tied with a knot pusher from laterally.

Type IV Tears: Reconstruction With Remnant Preservation

Finally, in some cases, the distal remnant is small or the tissue quality in the largest part of the remnant is poor, and after debriding back to good tissue quality, only 10% to 25% of the native tissue length is left (Figure 6A). In these cases, the remnant is preserved, however, tensioning of the remnant with sutures is usually not necessary for the prevention of cyclops lesions. Nonetheless, it is important to debride the parts of the remnant ligament with poor tissue quality as mop-end patterns of the remnant may increase the chance of these lesions (Figure 6B).

In this situation, any of the standard ACL reconstruction techniques can be performed with simple attention being paid to preserving what is left of the tibial insertion site. At the very least, the small insertion remnant guides the anatomic placement of the graft, and prevents egress of joint fluid into the tibial tunnel and could minimize tunnel widening.

Theoretical benefits also include improved vascularity and proprioception. The tunnels are prepared, the grafts passed, and fixation placed per the surgeon’s preference (Figure 6C). The remnant is then arthroscopically checked to confirm that there is no impingement or risk of cyclops lesions (Figure 6D).

Type V Tears: Primary Repair

Finally, in some patients a soft tissue avulsion (Figure 7A) or bony avulsion of the distal attachment of the ACL can be seen. Both injuries are relatively rare, although bony avulsions are frequently seen in children, especially those younger than 12 years old. In these cases, the same techniques and theory that are applied to proximal avulsion type tears can be used and applied to distal avulsion type tears.

However, they must be applied in an upside down manner by working from proximal to distal in this case.

First, No. 2 FiberWire sutures are passed from proximal towards the distal end of the ligament in the anteromedial bundle, and the same process is then repeated for No. 2 TigerWire sutures for the posterolateral bundle. Then both sutures are exited at the distal avulsed end at the locations of the anteromedial and posterolateral footprints (Figure 7B). A 2.4-mm ACL guide wire and a Ninitol wire are used to drill 2 tunnels from the tibia towards the tibial footprint. The repair sutures are then retrieved through both tunnels (Figure 7C) and the sutures are tied distally over a ligament button after cycling of the knee (Figure 7D). This technique is very useful for soft tissue avulsions, or when there are only small flecks of bone or when the avulsed bone is significantly comminuted. If a large bony avulsion fragment is present, this technique can also be applied with some modification, although there have been multiple other techniques described in the literature that work well in this situation including fixation with screw and washer, or with suture anchors.

 

Complex Tear or Poor Tissue Quality: Reconstruction

In some cases, the tissue quality is poor, or the ligament has complex or multiple tears. Essentially, in these cases, there is nothing to preserve and a standard reconstruction approach is performed in these cases.

Conclusion

The uniform gold standard for all ACL tear types is currently primary reconstruction. However, several disadvantages of ACL reconstruction exist, while there are multiple advantages to the concept of ACL preservation. In this surgical technique article, we have discussed our tear type classification and the recommended surgical techniques for each. With this treatment algorithm, which is based on tear location and tissue quality, an optimal and minimally invasive treatment can be chosen for each individual patient. Future studies are needed to compare and contrast these treatments with the current gold standard of ACL reconstruction.

Am J Orthop. 2016;45(7):E406-E414. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

Musculoskeletal allografts are becoming increasingly accepted as a viable alternative to autografts in a variety of orthopedic procedures. A 2006 American Orthopaedic Society for Sports Medicine (AOSSM) survey indicated that 86% of the participating 365 orthopedic surgeons use allografts in their practice.¹ Although the overwhelming majority of orthopedic surgeons use allografts, they share common concerns, including safety, tissue integrity, and biologic incorporation. It is essential for the orthopedic surgeon to understand the current standards of tissue banking, risks and benefits related to the use of allografts, and common indications for safe use in clinical practice. This article reviews the current status of musculoskeletal allografts, including tissue procurement and processing, infections, complications, and specific uses tailored to ligament reconstruction.

Donor Bank, Processing, Sterilization, and Regulation

In the United States, the American Association of Tissue Banks (AATB) is responsible for establishing the standards for more than 100 accredited tissue banks. These tissue banks recover tissue from approximately 30,000 donors annually and account for an estimated 90% of the available musculoskeletal allografts used in the United States. While not all tissue banks are accredited by the AATB, all are required to register with the Food and Drug Administration (FDA), which allows for unannounced inspections of any facility. Facilities are required to abide by the FDA-implemented Current Good Tissue Practices (CGTP), which encompasses regulations on all donor tissue collected after May 2005 to help prevent the transmission of communicable diseases. The FDA released an updated draft in January 2009 that emphasizes safe practices and regulations spanning from environmental control to specific equipment.²

The safety of a transplanted allograft tissue begins within the tissue bank. Donor screening and testing is the first step in reducing the risk of transmission. Screening consists of collecting medical and social history from the family and any healthcare resources to assess the eligibility of the donor. If prior blood donations or autopsy information is available, that information is scrutinized. Donor tissue undergoes nucleic acid testing (NAT), which is required by both the AATB and FDA. All donor tissue must be screened for both types of human immunodeficiency virus (HIV), hepatitis B virus (HBV), hepatitis C virus (HCV), treponema pallidum, and human transmissible spongiform encephalopathies.³ NAT of donor tissue effectively reduces the risk of viral transmission. Additionally, routine preprocessing swabs for bacterial and fungal cultures are performed, although the sensitivity of these cultures ranges from 78% to 92%.⁴

After donor screening and testing, allograft tissues are usually obtained under aseptic conditions, though this is not FDA-required.⁵ Once procured, the tissue undergoes sterilization. Currently, there is no standard method ubiquitous to all tissue banks, nor does the FDA require a specific method. Rather, the FDA and AATB require tissue banks to validate their sterilization process and provide supporting data. The goal of sterilization is to inactivate viruses and eradicate bacteria while maintaining the biological and mechanical properties of the tissue. The AATB requires a Sterility Assurance Level (SAL) of 10^-6, meaning there is no more than one in a million chance that a nonviral viable microbe exists on or within the tissue. Sterilization techniques may include both radiation and a variety of chemical reagents. Gamma irradiation is a commonly used method of sterilizing soft tissue allografts, although some studies indicate that it is detrimental to tissue biology.⁶ Newer methods of sterilization are being tested, one of which includes carbon dioxide in combination with antioxidants and irradiation. Bui and colleagues⁷ directly compared the biomechanical and histological properties of allograft tissue after either the standard 25 kGy gamma irradiation or supercritical carbon dioxide techniques. Although there is no histological difference, the samples treated with supercritical carbon dioxide had less biomechanical damage.⁷ Finally, the terminally sterilized allograft tissue is frozen to temperatures between -40°C and -80°C.⁵

Infections

One major concern of allografts is the risk of disease transmission. While numerous studies have investigated the incidence of bacterial infection following transplantation of allograft tissue, there are challenges associated with differentiating common postoperative infections from ones directly associated with the transmission of bacteria within the graft. There is a wide array of reported incidences of infection in the literature, from the Tomford and colleagues⁸ 1981 study that reported a 6.9% rateto the 2001 study by Munting and colleagues,⁹ who reported 0% in their series. Multiple confounding variables exist, such as possible contamination during handling of an otherwise noncontaminated or properly sterilized allograft with inappropriate inclusion of all postoperative infections. In contrast, recognizing viral transmission has been somewhat easier, although reporting of these incidences has been variable in the past. In either case, there is no accredited reporting system for infections related to allografts.

 

Bacterial Transmission

Clostridium species. Clostridium species are commonly found among intestinal flora. There is a general consensus that between 24 to 48 hours after death intestinal flora transmigrates into the surrounding tissue and blood. Therefore, a commonly accepted recommendation is that cadaveric tissue needs to be excised prior to 24 hours postmortem.¹⁰

In 2001, a 23-year-old man underwent reconstructive knee surgery with a femoral condyle allograft. A few days after surgery, he became septic and ultimately died from the infection. Clostridium sordellii was cultured from the tissue. Several days later, a 17-year-old boy underwent reconstructive knee surgery with a fresh femoral condyle and frozen meniscus from the same donor. Twenty-four hours after surgery, he developed a fever and was readmitted a week later for presumed infection and treated effectively with penicillin and ampicillin/sulbactam. Tissue from the same cadaveric donor had been transplanted into 7 other patients without reports of infection. In a 2002 Centers for Disease Control and Prevention (CDC) update report,¹¹ there were 26 total bacterial cases from allografts and 13 cases were attributed to Clostridium. Malinin and colleagues¹⁰ reviewed 795 consecutive cadaveric donors and found that 64 (8.1%) had positive cultures for Clostridia. Of all the positive cultures for Clostridia, 81.3% had positive blood cultures, 57.8% had positive bone marrow aspirate cultures, and 46.9% had positive tissue cultures. They concluded that multiple cultures are required for cadaveric tissue donors in order to reach a higher sensitivity for Clostridial contamination, and these should be done routinely to guide the sterilization process.

Strep species. In 2003, a 17-year-old boy underwent anterior cruciate ligament (ACL) reconstruction with a patellar tendon allograft.¹² About 1 week later, he was admitted for signs of infection and received intravenous antibiotics. He required surgical debridement, and intraoperative cultures grew Group A Streptococcus (GAS) that was also identified in the postmortem donor cultures. The tissues underwent processing in an antimicrobial solution and postprocessing cultures were negative for bacteria, but they were not sterilized. Tissues from this donor had been implanted in 5 other patients without report of infection. Following this event, recommendations have been made for prompt rejection of tissue with cultures positive for GAS, unless a sterilizing procedure is used.

Other bacteria. According to the 2002 CDC update, 11 of the 26 cases of bacterial infection reported to the agency were a combination of gram-negative bacilli, polymicrobial flora, or culture negative.¹¹

Viral Transmission

The most effective way to prevent transmission of a viral disease from allografts is thorough donor screening. Since the AATB implemented NAT in 2005 for HIV and HCV, there have been no reported cases of transmission.³ Even prior to this, regular blood screening along with social questionnaires completed by donors or donor families eliminated high-risk donors and significantly decreased the rate of transmission.

Human Immunodeficiency Virus. The first reported case of HIV transmission via implantation of allograft was in 1988. Further investigation revealed that there were 8 transmissions between 1984 and 1986, when routine screening of donors had not yet been implemented. The last reported case of HIV transmission occurred in 1996 with an untested donor.¹³Hepatitis C Virus. There are several reported cases of HCV transmission that occurred where the donors initially tested negative for HCV. In one case, 40 allografts from the same donor were transplanted over a period of nearly 2 years. This resulted in at least 8 patients being infected with HCV.¹⁴ Another case of HCV transmission was reported in 2005 after a patient developed acute HCV 6 weeks after transplantation of a patellar tendon allograft. Further investigation revealed that there had been 3 additional cases over a year from the same donor. Researchers determined that if the initial case had been reported, at least 3 transmissions could have been prevented.¹⁵Human T-cell Lymphotropic Virus (HTLV).The first reported transmission of HTLV was in 1991. This was reported in an asymptomatic patient who received a femoral head allograft from a donor who had been previously infected via a blood transfusion.¹⁶Zika virus. With recent outbreaks of the Zika virus, the FDA recently released recommendations regarding the screening and deferral of donors, mainly for blood transfusion. Orthopedists should take into consideration the potential for transmission through allografts. The FDA states that all potential donors should be screened for Zika virus using questionnaires and whole blood tests. Symptomatic donors are deferred at least 4 weeks following resolution of symptoms. While this is a recent recommendation from the FDA, orthopedists must be cognizant of the potential harms from this unfamiliar and evolving situation.¹⁷

Graft Specifics

Anterior Cruciate Ligament

ACL reconstruction is one of the most commonly performed surgeries by orthopedic surgeons, with an estimated 200,000 reconstructions per year.¹⁸Despite the popularity of this surgery, controversies remain regarding the optimal graft for reconstruction.^19,20 One would provide adequate strength, be readily available, not elicit an immunologic response from the host, rapidly incorporate, elicit low morbidity, and vascularize early. Current options include both autografts and allografts. Common autograft options include patellar bone-tendon-bone (PBTB), hamstrings tendon, quadriceps tendon, and iliotibial band. PBTB autograft remains a common choice among orthopedic surgeons, as it allows early incorporation of the graft into bone and eliminates immune rejection. However, donor site morbidity, including anterior knee pain, weakness of knee extension, joint stiffness, increased postoperative pain, and iatrogenic patella fractures, have been reported in the literature.²¹ Commonly used allograft options include donor bone-patellar tendon-bone, quadriceps tendon, Achilles tendon, anterior and posterior tibialis tendons, hamstring tendons, and iliotibial band. Allografts provide the advantage of avoiding donor site morbidity, being readily available, allowing for shorter operative times, and providing lower postoperative pain compared to autografts, although they carry the risk of disease transmission, rejection, and slower incorporation into bone.^22-27

 

Autograft donor site morbidities. One of the general disadvantages of autografts is the donor site morbidity associated with harvesting the grafts. In specific, PBTB grafts allow for bony blocks on both ends of the graft to incorporate into the host bone. However, this technique comes with the risk of disrupting the extensor mechanism.^28,29 Milankov and colleagues³⁰ published a retrospective review of over 2000 ACLs using autologous PBTB graft. They noted a 0.45% incidence of patella fracture and 0.18% patellar tendon rupture.³⁰ Others have reported that intraoperative repair of the patellar tendon after tendon harvesting can increase infrapatellar fibrosis, thus increasing the risk for stiffness.^31-33

Hamstring autografts include the semitendinosus and the gracilis tendons. The harvesting process is technically demanding and can be complicated by inadvertent amputation of the tendons, making the graft unsuitable for reconstructive purposes.³⁴ Additionally, several reports have identified persistent numbness and hyperesthesia following hamstring harvesting due to iatrogenic injury to the prepatellar branches of the saphenous nerve.^35,36A comprehensive review by Slone and colleagues³⁷ reported comparable functional outcomes with quadriceps tendon autograft compared to PBTB; however, this comes with the risk of postoperative hematoma formation and the potential for thigh compartment syndrome.

Biology and Biomechanics of Allografts

One of the major disadvantages of allografts is the reduced ability to incorporate into the host tissue. Several in vitro and animal studies have suggested that allografts incorporate in the host slower than autografts.^24,26,38 Early studies by Jackson and colleagues²⁴ on goat models demonstrated that allografts and autografts have similar structural and biological properties initially, but allografts display significantly slower incorporation into the host tissue at 6 months. Histologically, allografts demonstrated lower revascularization, a smaller cross-sectional area, and a prolonged inflammatory response at 6 months postoperatively.^24,39,40 Muramatsu and colleagues⁴¹ further showed through the use of magnetic resonance imaging a slower rate of revascularization of allografts over 2 years post-reconstruction.

Given the delayed biologic incorporation of allografts, studies have identified a lower strength-to-failure rate in the early postoperative period compared to autografts. An animal model study by Nikolaou and colleagues³⁸ showed that the strength of allografts was lower for up to 2 years following surgery. Additional biomechanical studies demonstrated that allografts were nearly 75% structurally weaker compared to autografts at 1 year following surgery.⁴²

Acknowledging these limitations, one should use caution when choosing to use an allograft or starting aggressive early rehabilitation after an allograft reconstruction, especially in athletes and young patients.

Clinical Outcomes

Although in vitro studies demonstrate inferior strength and delayed incorporation of allografts in the early postoperative period, there is still controversy surrounding the clinical and functional outcomes. Numerous studies have identified allografts as a viable option for ACL reconstruction, with similar reported patient satisfaction scores compared to autografts.^43,44

The MOON Consortium recently published a prospective study of nearly 2500 subjects looking to identify risk factors for failure of ACL reconstruction. The study found that allografts had an odds ratio for failure 5.2 times that of PBTB autografts, correlating this factor to an increased re-tear rate of 6.9% in the allograft group compared to 3.2% in the PBTB group (P < .01).⁴⁵ The elevated risk is more prevalent in younger patients, especially athletic teenagers. This issue has been reiterated in multiple studies.^45-50A meta-analysis by Hu and colleagues²³ identified 9 studies, either randomized control trials or prospective cohort studies, that looked at clinical outcomes between the different graft choices. They showed there was no significant difference between graft options in terms of instrumental laxity (P = .59), Lachman test (P = .41), pivot shift test (P = .88), and multiple functional outcome scores, including the International Knee Documentation Committee (IKDC), Lysholm, and Tegner scores.^23,51-59Processing and sterilization techniques are thought to play a role in allograft failure. Guo and other researchers have demonstrated a significantly higher rate of failure for patients who received gamma-irradiated allografts compared to fresh frozen allografts.^23,58-64 With improved sterilization techniques and a strict selection process of donors, gamma radiation has fallen out of favor to protect the biological characteristics of the tissue graft.^5,65,66Several factors need to be considered when selecting between allograft or autograft tissue for ligamentous reconstruction. The selection must be balanced between the surgeon’s experience, patient and surgeon preferences, age of the patient, level of physical activity, primary or revision surgical setting, multiligamentous failure, geographical availability of donor grafts, and economical factors.

Medial Patellofemoral Ligament Reconstruction

Another relatively recent application for allografts has been described for the reconstruction of the medial patellofemoral ligament (MPFL) in recurrent lateral patellar dislocations.^67-74

Typically, MPFL reconstructions make use of autografts, including quadriceps tendon, patellar tendon, and hamstring ligaments. However, allografts have the potential to limit postoperative donor site morbidity and to allow a faster rehabilitation.^75,76 Allografts include semitendinosus, gracilis, anterior tibialis, posterior tibialis, and quadriceps tendons.

Calvo Rodríguez and colleagues⁷⁶ performed a retrospective review in 2015 comparing allografts to autografts for MPFL reconstruction with respect to postoperative knee function and re-dislocation rates. Among the collective 28 patients, there was no difference in overall functional scores or dislocation rates between the grafts. Although this was a retrospective review and had a small number of subjects, the findings identify allografts as a reliable graft option for MPFL reconstruction.⁷⁶While there has been a surge of interest in techniques for MPFL reconstruction, there is limited research available regarding the superiority of allografts compared to autografts. For this specific application, it seems that clinical outcomes correlate more to adequate stabilization of the patellofemoral joint than to the type of graft used.^77,78 Future research should be dedicated to prospective randomized control trials to delineate any disadvantages to using allografts for MPFL reconstruction.

 

Discussion

Musculoskeletal allografts are gaining popularity for ligamentous reconstruction as their safety and efficacy continue to improve. With the great majority of tissue banks being accredited by the AATB and specific regulations such as NAT screening becoming common practice, infection rates and transmission of diseases have become incredibly rare. However, a thorough consideration needs to be taken into account when choosing between autograft and allograft on a case-by-case basis (Table).

Although the incidence of donor site complications is low with autografts, there are inherent risks, such as harvest site hyperesthesia, persistent numbness, cosmetic dissatisfaction, pain, weakness, functional implications, and unsuitability of the harvested graft. While it may appear that allografts may obviate donor site morbidity, one must consider the reduced potential for the donor tissue to incorporate into the host. Several studies have suggested that incorporation into the host tissue is inferior and slower for allografts. With this knowledge, factors such as clinical outcomes, future expectations, rehabilitation protocol, and individual patient characteristics all need to be considered when selecting the source of the tissue to be transplanted. Given that there is a growing need for availability of allografts, a well-rounded understanding of the biologic and physiologic aspects of the transplanted tissues is imperative. Future research will need to focus on improving the rate and quality of the biological incorporation of the transplanted graft into the host while eliminating the risk of disease transmission and infection.

Am J Orthop. 2016;45(7):446-453. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

Musculoskeletal allografts are becoming increasingly accepted as a viable alternative to autografts in a variety of orthopedic procedures. A 2006 American Orthopaedic Society for Sports Medicine (AOSSM) survey indicated that 86% of the participating 365 orthopedic surgeons use allografts in their practice.¹ Although the overwhelming majority of orthopedic surgeons use allografts, they share common concerns, including safety, tissue integrity, and biologic incorporation. It is essential for the orthopedic surgeon to understand the current standards of tissue banking, risks and benefits related to the use of allografts, and common indications for safe use in clinical practice. This article reviews the current status of musculoskeletal allografts, including tissue procurement and processing, infections, complications, and specific uses tailored to ligament reconstruction.

Donor Bank, Processing, Sterilization, and Regulation

In the United States, the American Association of Tissue Banks (AATB) is responsible for establishing the standards for more than 100 accredited tissue banks. These tissue banks recover tissue from approximately 30,000 donors annually and account for an estimated 90% of the available musculoskeletal allografts used in the United States. While not all tissue banks are accredited by the AATB, all are required to register with the Food and Drug Administration (FDA), which allows for unannounced inspections of any facility. Facilities are required to abide by the FDA-implemented Current Good Tissue Practices (CGTP), which encompasses regulations on all donor tissue collected after May 2005 to help prevent the transmission of communicable diseases. The FDA released an updated draft in January 2009 that emphasizes safe practices and regulations spanning from environmental control to specific equipment.²

The safety of a transplanted allograft tissue begins within the tissue bank. Donor screening and testing is the first step in reducing the risk of transmission. Screening consists of collecting medical and social history from the family and any healthcare resources to assess the eligibility of the donor. If prior blood donations or autopsy information is available, that information is scrutinized. Donor tissue undergoes nucleic acid testing (NAT), which is required by both the AATB and FDA. All donor tissue must be screened for both types of human immunodeficiency virus (HIV), hepatitis B virus (HBV), hepatitis C virus (HCV), treponema pallidum, and human transmissible spongiform encephalopathies.³ NAT of donor tissue effectively reduces the risk of viral transmission. Additionally, routine preprocessing swabs for bacterial and fungal cultures are performed, although the sensitivity of these cultures ranges from 78% to 92%.⁴

After donor screening and testing, allograft tissues are usually obtained under aseptic conditions, though this is not FDA-required.⁵ Once procured, the tissue undergoes sterilization. Currently, there is no standard method ubiquitous to all tissue banks, nor does the FDA require a specific method. Rather, the FDA and AATB require tissue banks to validate their sterilization process and provide supporting data. The goal of sterilization is to inactivate viruses and eradicate bacteria while maintaining the biological and mechanical properties of the tissue. The AATB requires a Sterility Assurance Level (SAL) of 10^-6, meaning there is no more than one in a million chance that a nonviral viable microbe exists on or within the tissue. Sterilization techniques may include both radiation and a variety of chemical reagents. Gamma irradiation is a commonly used method of sterilizing soft tissue allografts, although some studies indicate that it is detrimental to tissue biology.⁶ Newer methods of sterilization are being tested, one of which includes carbon dioxide in combination with antioxidants and irradiation. Bui and colleagues⁷ directly compared the biomechanical and histological properties of allograft tissue after either the standard 25 kGy gamma irradiation or supercritical carbon dioxide techniques. Although there is no histological difference, the samples treated with supercritical carbon dioxide had less biomechanical damage.⁷ Finally, the terminally sterilized allograft tissue is frozen to temperatures between -40°C and -80°C.⁵

Infections

One major concern of allografts is the risk of disease transmission. While numerous studies have investigated the incidence of bacterial infection following transplantation of allograft tissue, there are challenges associated with differentiating common postoperative infections from ones directly associated with the transmission of bacteria within the graft. There is a wide array of reported incidences of infection in the literature, from the Tomford and colleagues⁸ 1981 study that reported a 6.9% rateto the 2001 study by Munting and colleagues,⁹ who reported 0% in their series. Multiple confounding variables exist, such as possible contamination during handling of an otherwise noncontaminated or properly sterilized allograft with inappropriate inclusion of all postoperative infections. In contrast, recognizing viral transmission has been somewhat easier, although reporting of these incidences has been variable in the past. In either case, there is no accredited reporting system for infections related to allografts.

 

Bacterial Transmission

Clostridium species. Clostridium species are commonly found among intestinal flora. There is a general consensus that between 24 to 48 hours after death intestinal flora transmigrates into the surrounding tissue and blood. Therefore, a commonly accepted recommendation is that cadaveric tissue needs to be excised prior to 24 hours postmortem.¹⁰

In 2001, a 23-year-old man underwent reconstructive knee surgery with a femoral condyle allograft. A few days after surgery, he became septic and ultimately died from the infection. Clostridium sordellii was cultured from the tissue. Several days later, a 17-year-old boy underwent reconstructive knee surgery with a fresh femoral condyle and frozen meniscus from the same donor. Twenty-four hours after surgery, he developed a fever and was readmitted a week later for presumed infection and treated effectively with penicillin and ampicillin/sulbactam. Tissue from the same cadaveric donor had been transplanted into 7 other patients without reports of infection. In a 2002 Centers for Disease Control and Prevention (CDC) update report,¹¹ there were 26 total bacterial cases from allografts and 13 cases were attributed to Clostridium. Malinin and colleagues¹⁰ reviewed 795 consecutive cadaveric donors and found that 64 (8.1%) had positive cultures for Clostridia. Of all the positive cultures for Clostridia, 81.3% had positive blood cultures, 57.8% had positive bone marrow aspirate cultures, and 46.9% had positive tissue cultures. They concluded that multiple cultures are required for cadaveric tissue donors in order to reach a higher sensitivity for Clostridial contamination, and these should be done routinely to guide the sterilization process.

Strep species. In 2003, a 17-year-old boy underwent anterior cruciate ligament (ACL) reconstruction with a patellar tendon allograft.¹² About 1 week later, he was admitted for signs of infection and received intravenous antibiotics. He required surgical debridement, and intraoperative cultures grew Group A Streptococcus (GAS) that was also identified in the postmortem donor cultures. The tissues underwent processing in an antimicrobial solution and postprocessing cultures were negative for bacteria, but they were not sterilized. Tissues from this donor had been implanted in 5 other patients without report of infection. Following this event, recommendations have been made for prompt rejection of tissue with cultures positive for GAS, unless a sterilizing procedure is used.

Other bacteria. According to the 2002 CDC update, 11 of the 26 cases of bacterial infection reported to the agency were a combination of gram-negative bacilli, polymicrobial flora, or culture negative.¹¹

Viral Transmission

The most effective way to prevent transmission of a viral disease from allografts is thorough donor screening. Since the AATB implemented NAT in 2005 for HIV and HCV, there have been no reported cases of transmission.³ Even prior to this, regular blood screening along with social questionnaires completed by donors or donor families eliminated high-risk donors and significantly decreased the rate of transmission.

Human Immunodeficiency Virus. The first reported case of HIV transmission via implantation of allograft was in 1988. Further investigation revealed that there were 8 transmissions between 1984 and 1986, when routine screening of donors had not yet been implemented. The last reported case of HIV transmission occurred in 1996 with an untested donor.¹³Hepatitis C Virus. There are several reported cases of HCV transmission that occurred where the donors initially tested negative for HCV. In one case, 40 allografts from the same donor were transplanted over a period of nearly 2 years. This resulted in at least 8 patients being infected with HCV.¹⁴ Another case of HCV transmission was reported in 2005 after a patient developed acute HCV 6 weeks after transplantation of a patellar tendon allograft. Further investigation revealed that there had been 3 additional cases over a year from the same donor. Researchers determined that if the initial case had been reported, at least 3 transmissions could have been prevented.¹⁵Human T-cell Lymphotropic Virus (HTLV).The first reported transmission of HTLV was in 1991. This was reported in an asymptomatic patient who received a femoral head allograft from a donor who had been previously infected via a blood transfusion.¹⁶Zika virus. With recent outbreaks of the Zika virus, the FDA recently released recommendations regarding the screening and deferral of donors, mainly for blood transfusion. Orthopedists should take into consideration the potential for transmission through allografts. The FDA states that all potential donors should be screened for Zika virus using questionnaires and whole blood tests. Symptomatic donors are deferred at least 4 weeks following resolution of symptoms. While this is a recent recommendation from the FDA, orthopedists must be cognizant of the potential harms from this unfamiliar and evolving situation.¹⁷

Graft Specifics

Anterior Cruciate Ligament

ACL reconstruction is one of the most commonly performed surgeries by orthopedic surgeons, with an estimated 200,000 reconstructions per year.¹⁸Despite the popularity of this surgery, controversies remain regarding the optimal graft for reconstruction.^19,20 One would provide adequate strength, be readily available, not elicit an immunologic response from the host, rapidly incorporate, elicit low morbidity, and vascularize early. Current options include both autografts and allografts. Common autograft options include patellar bone-tendon-bone (PBTB), hamstrings tendon, quadriceps tendon, and iliotibial band. PBTB autograft remains a common choice among orthopedic surgeons, as it allows early incorporation of the graft into bone and eliminates immune rejection. However, donor site morbidity, including anterior knee pain, weakness of knee extension, joint stiffness, increased postoperative pain, and iatrogenic patella fractures, have been reported in the literature.²¹ Commonly used allograft options include donor bone-patellar tendon-bone, quadriceps tendon, Achilles tendon, anterior and posterior tibialis tendons, hamstring tendons, and iliotibial band. Allografts provide the advantage of avoiding donor site morbidity, being readily available, allowing for shorter operative times, and providing lower postoperative pain compared to autografts, although they carry the risk of disease transmission, rejection, and slower incorporation into bone.^22-27

 

Autograft donor site morbidities. One of the general disadvantages of autografts is the donor site morbidity associated with harvesting the grafts. In specific, PBTB grafts allow for bony blocks on both ends of the graft to incorporate into the host bone. However, this technique comes with the risk of disrupting the extensor mechanism.^28,29 Milankov and colleagues³⁰ published a retrospective review of over 2000 ACLs using autologous PBTB graft. They noted a 0.45% incidence of patella fracture and 0.18% patellar tendon rupture.³⁰ Others have reported that intraoperative repair of the patellar tendon after tendon harvesting can increase infrapatellar fibrosis, thus increasing the risk for stiffness.^31-33

Hamstring autografts include the semitendinosus and the gracilis tendons. The harvesting process is technically demanding and can be complicated by inadvertent amputation of the tendons, making the graft unsuitable for reconstructive purposes.³⁴ Additionally, several reports have identified persistent numbness and hyperesthesia following hamstring harvesting due to iatrogenic injury to the prepatellar branches of the saphenous nerve.^35,36A comprehensive review by Slone and colleagues³⁷ reported comparable functional outcomes with quadriceps tendon autograft compared to PBTB; however, this comes with the risk of postoperative hematoma formation and the potential for thigh compartment syndrome.

Biology and Biomechanics of Allografts

One of the major disadvantages of allografts is the reduced ability to incorporate into the host tissue. Several in vitro and animal studies have suggested that allografts incorporate in the host slower than autografts.^24,26,38 Early studies by Jackson and colleagues²⁴ on goat models demonstrated that allografts and autografts have similar structural and biological properties initially, but allografts display significantly slower incorporation into the host tissue at 6 months. Histologically, allografts demonstrated lower revascularization, a smaller cross-sectional area, and a prolonged inflammatory response at 6 months postoperatively.^24,39,40 Muramatsu and colleagues⁴¹ further showed through the use of magnetic resonance imaging a slower rate of revascularization of allografts over 2 years post-reconstruction.

Given the delayed biologic incorporation of allografts, studies have identified a lower strength-to-failure rate in the early postoperative period compared to autografts. An animal model study by Nikolaou and colleagues³⁸ showed that the strength of allografts was lower for up to 2 years following surgery. Additional biomechanical studies demonstrated that allografts were nearly 75% structurally weaker compared to autografts at 1 year following surgery.⁴²

Acknowledging these limitations, one should use caution when choosing to use an allograft or starting aggressive early rehabilitation after an allograft reconstruction, especially in athletes and young patients.

Clinical Outcomes

Although in vitro studies demonstrate inferior strength and delayed incorporation of allografts in the early postoperative period, there is still controversy surrounding the clinical and functional outcomes. Numerous studies have identified allografts as a viable option for ACL reconstruction, with similar reported patient satisfaction scores compared to autografts.^43,44

The MOON Consortium recently published a prospective study of nearly 2500 subjects looking to identify risk factors for failure of ACL reconstruction. The study found that allografts had an odds ratio for failure 5.2 times that of PBTB autografts, correlating this factor to an increased re-tear rate of 6.9% in the allograft group compared to 3.2% in the PBTB group (P < .01).⁴⁵ The elevated risk is more prevalent in younger patients, especially athletic teenagers. This issue has been reiterated in multiple studies.^45-50A meta-analysis by Hu and colleagues²³ identified 9 studies, either randomized control trials or prospective cohort studies, that looked at clinical outcomes between the different graft choices. They showed there was no significant difference between graft options in terms of instrumental laxity (P = .59), Lachman test (P = .41), pivot shift test (P = .88), and multiple functional outcome scores, including the International Knee Documentation Committee (IKDC), Lysholm, and Tegner scores.^23,51-59Processing and sterilization techniques are thought to play a role in allograft failure. Guo and other researchers have demonstrated a significantly higher rate of failure for patients who received gamma-irradiated allografts compared to fresh frozen allografts.^23,58-64 With improved sterilization techniques and a strict selection process of donors, gamma radiation has fallen out of favor to protect the biological characteristics of the tissue graft.^5,65,66Several factors need to be considered when selecting between allograft or autograft tissue for ligamentous reconstruction. The selection must be balanced between the surgeon’s experience, patient and surgeon preferences, age of the patient, level of physical activity, primary or revision surgical setting, multiligamentous failure, geographical availability of donor grafts, and economical factors.

Medial Patellofemoral Ligament Reconstruction

Another relatively recent application for allografts has been described for the reconstruction of the medial patellofemoral ligament (MPFL) in recurrent lateral patellar dislocations.^67-74

Typically, MPFL reconstructions make use of autografts, including quadriceps tendon, patellar tendon, and hamstring ligaments. However, allografts have the potential to limit postoperative donor site morbidity and to allow a faster rehabilitation.^75,76 Allografts include semitendinosus, gracilis, anterior tibialis, posterior tibialis, and quadriceps tendons.

Calvo Rodríguez and colleagues⁷⁶ performed a retrospective review in 2015 comparing allografts to autografts for MPFL reconstruction with respect to postoperative knee function and re-dislocation rates. Among the collective 28 patients, there was no difference in overall functional scores or dislocation rates between the grafts. Although this was a retrospective review and had a small number of subjects, the findings identify allografts as a reliable graft option for MPFL reconstruction.⁷⁶While there has been a surge of interest in techniques for MPFL reconstruction, there is limited research available regarding the superiority of allografts compared to autografts. For this specific application, it seems that clinical outcomes correlate more to adequate stabilization of the patellofemoral joint than to the type of graft used.^77,78 Future research should be dedicated to prospective randomized control trials to delineate any disadvantages to using allografts for MPFL reconstruction.

 

Discussion

Musculoskeletal allografts are gaining popularity for ligamentous reconstruction as their safety and efficacy continue to improve. With the great majority of tissue banks being accredited by the AATB and specific regulations such as NAT screening becoming common practice, infection rates and transmission of diseases have become incredibly rare. However, a thorough consideration needs to be taken into account when choosing between autograft and allograft on a case-by-case basis (Table).

Although the incidence of donor site complications is low with autografts, there are inherent risks, such as harvest site hyperesthesia, persistent numbness, cosmetic dissatisfaction, pain, weakness, functional implications, and unsuitability of the harvested graft. While it may appear that allografts may obviate donor site morbidity, one must consider the reduced potential for the donor tissue to incorporate into the host. Several studies have suggested that incorporation into the host tissue is inferior and slower for allografts. With this knowledge, factors such as clinical outcomes, future expectations, rehabilitation protocol, and individual patient characteristics all need to be considered when selecting the source of the tissue to be transplanted. Given that there is a growing need for availability of allografts, a well-rounded understanding of the biologic and physiologic aspects of the transplanted tissues is imperative. Future research will need to focus on improving the rate and quality of the biological incorporation of the transplanted graft into the host while eliminating the risk of disease transmission and infection.

Am J Orthop. 2016;45(7):446-453. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

Musculoskeletal allografts are becoming increasingly accepted as a viable alternative to autografts in a variety of orthopedic procedures. A 2006 American Orthopaedic Society for Sports Medicine (AOSSM) survey indicated that 86% of the participating 365 orthopedic surgeons use allografts in their practice.¹ Although the overwhelming majority of orthopedic surgeons use allografts, they share common concerns, including safety, tissue integrity, and biologic incorporation. It is essential for the orthopedic surgeon to understand the current standards of tissue banking, risks and benefits related to the use of allografts, and common indications for safe use in clinical practice. This article reviews the current status of musculoskeletal allografts, including tissue procurement and processing, infections, complications, and specific uses tailored to ligament reconstruction.

Donor Bank, Processing, Sterilization, and Regulation

In the United States, the American Association of Tissue Banks (AATB) is responsible for establishing the standards for more than 100 accredited tissue banks. These tissue banks recover tissue from approximately 30,000 donors annually and account for an estimated 90% of the available musculoskeletal allografts used in the United States. While not all tissue banks are accredited by the AATB, all are required to register with the Food and Drug Administration (FDA), which allows for unannounced inspections of any facility. Facilities are required to abide by the FDA-implemented Current Good Tissue Practices (CGTP), which encompasses regulations on all donor tissue collected after May 2005 to help prevent the transmission of communicable diseases. The FDA released an updated draft in January 2009 that emphasizes safe practices and regulations spanning from environmental control to specific equipment.²

The safety of a transplanted allograft tissue begins within the tissue bank. Donor screening and testing is the first step in reducing the risk of transmission. Screening consists of collecting medical and social history from the family and any healthcare resources to assess the eligibility of the donor. If prior blood donations or autopsy information is available, that information is scrutinized. Donor tissue undergoes nucleic acid testing (NAT), which is required by both the AATB and FDA. All donor tissue must be screened for both types of human immunodeficiency virus (HIV), hepatitis B virus (HBV), hepatitis C virus (HCV), treponema pallidum, and human transmissible spongiform encephalopathies.³ NAT of donor tissue effectively reduces the risk of viral transmission. Additionally, routine preprocessing swabs for bacterial and fungal cultures are performed, although the sensitivity of these cultures ranges from 78% to 92%.⁴

After donor screening and testing, allograft tissues are usually obtained under aseptic conditions, though this is not FDA-required.⁵ Once procured, the tissue undergoes sterilization. Currently, there is no standard method ubiquitous to all tissue banks, nor does the FDA require a specific method. Rather, the FDA and AATB require tissue banks to validate their sterilization process and provide supporting data. The goal of sterilization is to inactivate viruses and eradicate bacteria while maintaining the biological and mechanical properties of the tissue. The AATB requires a Sterility Assurance Level (SAL) of 10^-6, meaning there is no more than one in a million chance that a nonviral viable microbe exists on or within the tissue. Sterilization techniques may include both radiation and a variety of chemical reagents. Gamma irradiation is a commonly used method of sterilizing soft tissue allografts, although some studies indicate that it is detrimental to tissue biology.⁶ Newer methods of sterilization are being tested, one of which includes carbon dioxide in combination with antioxidants and irradiation. Bui and colleagues⁷ directly compared the biomechanical and histological properties of allograft tissue after either the standard 25 kGy gamma irradiation or supercritical carbon dioxide techniques. Although there is no histological difference, the samples treated with supercritical carbon dioxide had less biomechanical damage.⁷ Finally, the terminally sterilized allograft tissue is frozen to temperatures between -40°C and -80°C.⁵

Infections

One major concern of allografts is the risk of disease transmission. While numerous studies have investigated the incidence of bacterial infection following transplantation of allograft tissue, there are challenges associated with differentiating common postoperative infections from ones directly associated with the transmission of bacteria within the graft. There is a wide array of reported incidences of infection in the literature, from the Tomford and colleagues⁸ 1981 study that reported a 6.9% rateto the 2001 study by Munting and colleagues,⁹ who reported 0% in their series. Multiple confounding variables exist, such as possible contamination during handling of an otherwise noncontaminated or properly sterilized allograft with inappropriate inclusion of all postoperative infections. In contrast, recognizing viral transmission has been somewhat easier, although reporting of these incidences has been variable in the past. In either case, there is no accredited reporting system for infections related to allografts.

 

Bacterial Transmission

Clostridium species. Clostridium species are commonly found among intestinal flora. There is a general consensus that between 24 to 48 hours after death intestinal flora transmigrates into the surrounding tissue and blood. Therefore, a commonly accepted recommendation is that cadaveric tissue needs to be excised prior to 24 hours postmortem.¹⁰

In 2001, a 23-year-old man underwent reconstructive knee surgery with a femoral condyle allograft. A few days after surgery, he became septic and ultimately died from the infection. Clostridium sordellii was cultured from the tissue. Several days later, a 17-year-old boy underwent reconstructive knee surgery with a fresh femoral condyle and frozen meniscus from the same donor. Twenty-four hours after surgery, he developed a fever and was readmitted a week later for presumed infection and treated effectively with penicillin and ampicillin/sulbactam. Tissue from the same cadaveric donor had been transplanted into 7 other patients without reports of infection. In a 2002 Centers for Disease Control and Prevention (CDC) update report,¹¹ there were 26 total bacterial cases from allografts and 13 cases were attributed to Clostridium. Malinin and colleagues¹⁰ reviewed 795 consecutive cadaveric donors and found that 64 (8.1%) had positive cultures for Clostridia. Of all the positive cultures for Clostridia, 81.3% had positive blood cultures, 57.8% had positive bone marrow aspirate cultures, and 46.9% had positive tissue cultures. They concluded that multiple cultures are required for cadaveric tissue donors in order to reach a higher sensitivity for Clostridial contamination, and these should be done routinely to guide the sterilization process.

Strep species. In 2003, a 17-year-old boy underwent anterior cruciate ligament (ACL) reconstruction with a patellar tendon allograft.¹² About 1 week later, he was admitted for signs of infection and received intravenous antibiotics. He required surgical debridement, and intraoperative cultures grew Group A Streptococcus (GAS) that was also identified in the postmortem donor cultures. The tissues underwent processing in an antimicrobial solution and postprocessing cultures were negative for bacteria, but they were not sterilized. Tissues from this donor had been implanted in 5 other patients without report of infection. Following this event, recommendations have been made for prompt rejection of tissue with cultures positive for GAS, unless a sterilizing procedure is used.

Other bacteria. According to the 2002 CDC update, 11 of the 26 cases of bacterial infection reported to the agency were a combination of gram-negative bacilli, polymicrobial flora, or culture negative.¹¹

Viral Transmission

The most effective way to prevent transmission of a viral disease from allografts is thorough donor screening. Since the AATB implemented NAT in 2005 for HIV and HCV, there have been no reported cases of transmission.³ Even prior to this, regular blood screening along with social questionnaires completed by donors or donor families eliminated high-risk donors and significantly decreased the rate of transmission.

Human Immunodeficiency Virus. The first reported case of HIV transmission via implantation of allograft was in 1988. Further investigation revealed that there were 8 transmissions between 1984 and 1986, when routine screening of donors had not yet been implemented. The last reported case of HIV transmission occurred in 1996 with an untested donor.¹³Hepatitis C Virus. There are several reported cases of HCV transmission that occurred where the donors initially tested negative for HCV. In one case, 40 allografts from the same donor were transplanted over a period of nearly 2 years. This resulted in at least 8 patients being infected with HCV.¹⁴ Another case of HCV transmission was reported in 2005 after a patient developed acute HCV 6 weeks after transplantation of a patellar tendon allograft. Further investigation revealed that there had been 3 additional cases over a year from the same donor. Researchers determined that if the initial case had been reported, at least 3 transmissions could have been prevented.¹⁵Human T-cell Lymphotropic Virus (HTLV).The first reported transmission of HTLV was in 1991. This was reported in an asymptomatic patient who received a femoral head allograft from a donor who had been previously infected via a blood transfusion.¹⁶Zika virus. With recent outbreaks of the Zika virus, the FDA recently released recommendations regarding the screening and deferral of donors, mainly for blood transfusion. Orthopedists should take into consideration the potential for transmission through allografts. The FDA states that all potential donors should be screened for Zika virus using questionnaires and whole blood tests. Symptomatic donors are deferred at least 4 weeks following resolution of symptoms. While this is a recent recommendation from the FDA, orthopedists must be cognizant of the potential harms from this unfamiliar and evolving situation.¹⁷

Graft Specifics

Anterior Cruciate Ligament

ACL reconstruction is one of the most commonly performed surgeries by orthopedic surgeons, with an estimated 200,000 reconstructions per year.¹⁸Despite the popularity of this surgery, controversies remain regarding the optimal graft for reconstruction.^19,20 One would provide adequate strength, be readily available, not elicit an immunologic response from the host, rapidly incorporate, elicit low morbidity, and vascularize early. Current options include both autografts and allografts. Common autograft options include patellar bone-tendon-bone (PBTB), hamstrings tendon, quadriceps tendon, and iliotibial band. PBTB autograft remains a common choice among orthopedic surgeons, as it allows early incorporation of the graft into bone and eliminates immune rejection. However, donor site morbidity, including anterior knee pain, weakness of knee extension, joint stiffness, increased postoperative pain, and iatrogenic patella fractures, have been reported in the literature.²¹ Commonly used allograft options include donor bone-patellar tendon-bone, quadriceps tendon, Achilles tendon, anterior and posterior tibialis tendons, hamstring tendons, and iliotibial band. Allografts provide the advantage of avoiding donor site morbidity, being readily available, allowing for shorter operative times, and providing lower postoperative pain compared to autografts, although they carry the risk of disease transmission, rejection, and slower incorporation into bone.^22-27

 

Autograft donor site morbidities. One of the general disadvantages of autografts is the donor site morbidity associated with harvesting the grafts. In specific, PBTB grafts allow for bony blocks on both ends of the graft to incorporate into the host bone. However, this technique comes with the risk of disrupting the extensor mechanism.^28,29 Milankov and colleagues³⁰ published a retrospective review of over 2000 ACLs using autologous PBTB graft. They noted a 0.45% incidence of patella fracture and 0.18% patellar tendon rupture.³⁰ Others have reported that intraoperative repair of the patellar tendon after tendon harvesting can increase infrapatellar fibrosis, thus increasing the risk for stiffness.^31-33

Hamstring autografts include the semitendinosus and the gracilis tendons. The harvesting process is technically demanding and can be complicated by inadvertent amputation of the tendons, making the graft unsuitable for reconstructive purposes.³⁴ Additionally, several reports have identified persistent numbness and hyperesthesia following hamstring harvesting due to iatrogenic injury to the prepatellar branches of the saphenous nerve.^35,36A comprehensive review by Slone and colleagues³⁷ reported comparable functional outcomes with quadriceps tendon autograft compared to PBTB; however, this comes with the risk of postoperative hematoma formation and the potential for thigh compartment syndrome.

Biology and Biomechanics of Allografts

One of the major disadvantages of allografts is the reduced ability to incorporate into the host tissue. Several in vitro and animal studies have suggested that allografts incorporate in the host slower than autografts.^24,26,38 Early studies by Jackson and colleagues²⁴ on goat models demonstrated that allografts and autografts have similar structural and biological properties initially, but allografts display significantly slower incorporation into the host tissue at 6 months. Histologically, allografts demonstrated lower revascularization, a smaller cross-sectional area, and a prolonged inflammatory response at 6 months postoperatively.^24,39,40 Muramatsu and colleagues⁴¹ further showed through the use of magnetic resonance imaging a slower rate of revascularization of allografts over 2 years post-reconstruction.

Given the delayed biologic incorporation of allografts, studies have identified a lower strength-to-failure rate in the early postoperative period compared to autografts. An animal model study by Nikolaou and colleagues³⁸ showed that the strength of allografts was lower for up to 2 years following surgery. Additional biomechanical studies demonstrated that allografts were nearly 75% structurally weaker compared to autografts at 1 year following surgery.⁴²

Acknowledging these limitations, one should use caution when choosing to use an allograft or starting aggressive early rehabilitation after an allograft reconstruction, especially in athletes and young patients.

Clinical Outcomes

Although in vitro studies demonstrate inferior strength and delayed incorporation of allografts in the early postoperative period, there is still controversy surrounding the clinical and functional outcomes. Numerous studies have identified allografts as a viable option for ACL reconstruction, with similar reported patient satisfaction scores compared to autografts.^43,44

The MOON Consortium recently published a prospective study of nearly 2500 subjects looking to identify risk factors for failure of ACL reconstruction. The study found that allografts had an odds ratio for failure 5.2 times that of PBTB autografts, correlating this factor to an increased re-tear rate of 6.9% in the allograft group compared to 3.2% in the PBTB group (P < .01).⁴⁵ The elevated risk is more prevalent in younger patients, especially athletic teenagers. This issue has been reiterated in multiple studies.^45-50A meta-analysis by Hu and colleagues²³ identified 9 studies, either randomized control trials or prospective cohort studies, that looked at clinical outcomes between the different graft choices. They showed there was no significant difference between graft options in terms of instrumental laxity (P = .59), Lachman test (P = .41), pivot shift test (P = .88), and multiple functional outcome scores, including the International Knee Documentation Committee (IKDC), Lysholm, and Tegner scores.^23,51-59Processing and sterilization techniques are thought to play a role in allograft failure. Guo and other researchers have demonstrated a significantly higher rate of failure for patients who received gamma-irradiated allografts compared to fresh frozen allografts.^23,58-64 With improved sterilization techniques and a strict selection process of donors, gamma radiation has fallen out of favor to protect the biological characteristics of the tissue graft.^5,65,66Several factors need to be considered when selecting between allograft or autograft tissue for ligamentous reconstruction. The selection must be balanced between the surgeon’s experience, patient and surgeon preferences, age of the patient, level of physical activity, primary or revision surgical setting, multiligamentous failure, geographical availability of donor grafts, and economical factors.

Medial Patellofemoral Ligament Reconstruction

Another relatively recent application for allografts has been described for the reconstruction of the medial patellofemoral ligament (MPFL) in recurrent lateral patellar dislocations.^67-74

Typically, MPFL reconstructions make use of autografts, including quadriceps tendon, patellar tendon, and hamstring ligaments. However, allografts have the potential to limit postoperative donor site morbidity and to allow a faster rehabilitation.^75,76 Allografts include semitendinosus, gracilis, anterior tibialis, posterior tibialis, and quadriceps tendons.

Calvo Rodríguez and colleagues⁷⁶ performed a retrospective review in 2015 comparing allografts to autografts for MPFL reconstruction with respect to postoperative knee function and re-dislocation rates. Among the collective 28 patients, there was no difference in overall functional scores or dislocation rates between the grafts. Although this was a retrospective review and had a small number of subjects, the findings identify allografts as a reliable graft option for MPFL reconstruction.⁷⁶While there has been a surge of interest in techniques for MPFL reconstruction, there is limited research available regarding the superiority of allografts compared to autografts. For this specific application, it seems that clinical outcomes correlate more to adequate stabilization of the patellofemoral joint than to the type of graft used.^77,78 Future research should be dedicated to prospective randomized control trials to delineate any disadvantages to using allografts for MPFL reconstruction.

 

Discussion

Musculoskeletal allografts are gaining popularity for ligamentous reconstruction as their safety and efficacy continue to improve. With the great majority of tissue banks being accredited by the AATB and specific regulations such as NAT screening becoming common practice, infection rates and transmission of diseases have become incredibly rare. However, a thorough consideration needs to be taken into account when choosing between autograft and allograft on a case-by-case basis (Table).

Although the incidence of donor site complications is low with autografts, there are inherent risks, such as harvest site hyperesthesia, persistent numbness, cosmetic dissatisfaction, pain, weakness, functional implications, and unsuitability of the harvested graft. While it may appear that allografts may obviate donor site morbidity, one must consider the reduced potential for the donor tissue to incorporate into the host. Several studies have suggested that incorporation into the host tissue is inferior and slower for allografts. With this knowledge, factors such as clinical outcomes, future expectations, rehabilitation protocol, and individual patient characteristics all need to be considered when selecting the source of the tissue to be transplanted. Given that there is a growing need for availability of allografts, a well-rounded understanding of the biologic and physiologic aspects of the transplanted tissues is imperative. Future research will need to focus on improving the rate and quality of the biological incorporation of the transplanted graft into the host while eliminating the risk of disease transmission and infection.

Am J Orthop. 2016;45(7):446-453. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

In 1975, Skolnick and colleagues¹ introduced unicompartmental knee arthroplasty (UKA) for patients with isolated unicompartmental osteoarthritis (OA). They reported a study of 14 UKA procedures, of which 12 were at the medial and 2 at the lateral side. Forty years since this procedure was introduced, UKA is used in 8% to 12% of all knee arthroplasties.^2-6 A minority of these procedures are performed at the lateral side (5%-10%).^6-8

The considerable anatomical and kinematical differences between compartments^9-14 make it impossible to directly compare outcomes of medial and lateral UKA. For example, a greater degree of femoral roll and more posterior translation at the lateral side in flexion^9,10,13 can contribute to different pattern and volume differences of cartilage wear.¹⁵ Because of these differences, and because of implant design factors and lower surgical volume, lateral UKA is considered a technically more challenging surgery compared to medial UKA.^12,16,17

Since isolated lateral compartment OA is relatively scarce, current literature on lateral UKA is limited, and most studies combine medial and lateral outcomes to report UKA outcomes and failure modes.^3,4,18-20 However, as the UKA has grown in popularity over the last decade,^2,21-25 the number of reports about the lateral UKA also has increased. Recent studies reported excellent short-term survivorship results of the lateral UKA (96%-99%)^26,27 and smaller lateral UKA studies reported the 10-year survivorship with varying outcomes from good (84%)^14,28-30 to excellent (94%-100%).^8,31,32 Indeed, a recent systematic review showed survivorship of lateral UKA at 5, 10, and 15 years of 93%, 91%, and 89%, respectively.³³Because of the differences between the medial and lateral compartment, it is important to know the failure modes of lateral UKA in order to improve clinical outcomes and revision rates. We performed a systematic review of cohort studies and registry-based studies that reported lateral UKA failure to assess the causes of lateral UKA failure. In addition, we compared the failure modes in cohort studies with those found in registry-based studies.

Patients and Methods

Search Strategy and Criteria

Databases of PubMed, Embase, and Cochrane (Cochrane Central Register of Clinical Trials) were searched with the terms “knee, arthroplasty, replacement,” “unicompartmental,” “unicondylar,” “partial,” “UKA,” “UKR,” “UCA,” “UCR,” “PKA,” “PKR,” “PCA,” “prosthesis failure,” “reoperation,” “survivorship,” and “treatment failure.” After removal of duplicates, 2 authors (JPvdL and HAZ) scanned the articles for their title and abstract to assess eligibility for the study.

The full text of these eligible articles was further viewed and useful studies were selected using the inclusion and exclusion criteria. The references of these articles were scanned for additional studies and national registries (Figure).

Inclusion criteria were: (I) English language articles describing studies in humans published in the last 25 years, (II) retrospective and prospective studies, (III) featured lateral UKA, (IV) OA was indication for surgery, and (V) included failure modes data. The exclusion criteria were studies that featured: (I) only a specific group of failure (eg, bearing dislocations only), (II) previous surgery in ipsilateral knee (high tibial osteotomy, medial UKA), (III) acute concurrent knee diagnoses (acute anterior cruciate ligament rupture, acute meniscal tear), (IV) combined reporting of medial and lateral UKA, or (V) multiple studies with the same patient database.

Data Collection

All studies that reported modes of failure were used in this study and these failure modes were noted in a datasheet in Microsoft Excel 2011 (Microsoft).

The data of failures of lateral UKA are presented in Table 1 and are divided in cohort studies and registry-based studies. The final failure mode rates were presented in percentages (Table 2).

Statistical Analysis

For this systematic review, statistical analysis was performed with IBM SPSS Statistics 22 (SPSS Inc.). We performed chi square tests and Fisher’s exact tests to assess a difference between cohort studies and registry-based studies with the null hypothesis of no difference between both groups. A difference was considered significant when P < .05.

Results

Through the search of the databases, 1294 studies were identified and 26 handpicked studies were added. Initially, based on the title and abstract, 184 of these studies were found eligible.

After reviewing the full text of these articles, 25 studies (23 cohort studies and 2 registry-based studies) met the inclusion criteria and were included for the analysis of lateral UKA failure (Figure).

A total of 366 lateral UKA failures were included. The most common failure modes were progression of OA (29%), aseptic loosening (23%), and bearing dislocation (10%). Infection (6%), instability (6%), unexplained pain (6%), and fractures (4%) were less common causes of failure of lateral UKA (Table 2).

One hundred fifty-five of these failures were reported in the cohort studies. The most common modes of failure were OA progression (36%), bearing dislocation (17%) and aseptic loosening (16%). Less common were infection (10%), fractures (5%), pain (5%), and other causes (6%). In registry-based studies, with 211 lateral UKA failures, the most common modes of failure were aseptic loosening (28%), OA progression (24%), other causes (12%), instability (10%), pain (7%), bearing dislocation (5%), and polyethylene wear (4%) (Table 2).

When pooling cohort and registry-based studies, progression of OA was significantly more common than aseptic loosening (29% vs 23%, respectively; P < .01). It was also significantly more common in the cohort studies (36% vs 16%, respectively; P < .01) but no significant difference was found between progression of OA and aseptic loosening in registry-based studies (24% and 28%, respectively; P = .16) (Table 2).

When comparing cohort with registry-based studies, progression of OA was higher in cohort studies (36% vs. 24%; P < .01). Other failures modes that were more common in cohort studies compared with registry-based studies were bearing dislocation (17% vs 5%, respectively; P < .01) and infections (10% vs 3%, P < .01). Failure modes that were more common in registry-based studies than cohort studies were aseptic loosening (28% vs 16%, respectively; P < .01), other causes (12% vs 6%, respectively, P = .02), and instability (10% vs 1%, respectively, P < .01) (Table 2).

 

Discussion

In this systematic review, the most common failure modes in lateral UKA review were OA progression (29%), aseptic loosening (23%), and bearing dislocation (10%). Progression of OA and bearing dislocation were the most common modes of failure in cohort studies (36% and 17%, respectively), while aseptic loosening and OA progression were the most common failure modes in registry-based studies (28% and 24%, respectively).

As mentioned above, there are differences in anatomy and kinematics between the medial and lateral compartment. When the lateral UKA failure modes are compared with studies reporting medial UKA failure modes, differences in failure modes are seen.³⁴ Siddiqui and Ahmad³⁵ performed a systematic review of outcomes after UKA revision and presented a table with the failure modes of included studies. Unfortunately they did not report the ratio of medial and lateral UKA. However, when assuming an average percentage of 90% to 95% of medial UKA,^6,7,36 the main failure mode in their review in 17 out of 21 studies was aseptic loosening. Indeed, a recent systematic review on medial UKA failure modes showed that aseptic loosening is the most common cause of failure following this procedure.³⁴ Similarly, a search through registry-based studies^6,7 and large cohort studies^37-40 that only reported medial UKA failures showed that the majority of these studies^7,37-39 also reported aseptic loosening as the main cause of failure in medial UKA. When comparing the results of our systematic review of lateral UKA failures with the results of these studies of medial UKA failures, it seems that OA progression seems to play a more dominant role in failures of lateral UKA, while aseptic loosening seems to be more common in medial UKA.

Differences in anatomy and kinematics of the medial and lateral compartment can explain this. Malalignment of the joint is an important factor in the etiology of OA^41,42 and biomechanical studies showed that this malalignment can cause decreased viability and further degenerative changes of cartilage of the knee.⁴³ Hernigou and Deschamps⁴⁴ showed that the alignment of the knee after medial UKA is an important factor in postoperative joint changes. They found that overcorrection of varus deformity during medial UKA surgery, measured by the hip-knee-ankle (HKA) angle, was associated with increased OA at the lateral condyle and less tibial wear of the medial UKA. Undercorrection of the varus caused an increase in tibial wear of polyethylene. Chatellard and colleagues⁴⁵ found the same results in the correction of varus, measured by HKA. In addition, they found that when the prosthetic (medial) joint space was smaller than healthy (lateral) joint space, this was correlated with lower prosthesis survival. A smaller joint space at the healthy side was correlated with OA progression at the lateral compartment and tibial component wear.

These studies explain the mechanism of progression of OA and aseptic loosening. Harrington⁴⁶ assessed the load in patients with valgus and varus deformity. Patients with a valgus deformity have high mechanical load on the lateral condyle during the static phase, but during the dynamic phase, a major part of this load shifts to the medial condyle. In the patients with varus deformity, the mechanical load was noted on the medial condyle during both the static and dynamic phase. Ohdera and colleagues⁴⁷ advised, based on this biomechanical study and their own experiences, to correct the knee during lateral UKA to a slight valgus angle (5°-7°) to prevent OA progression at the medial side. van der List and colleagues⁴⁸ similarly showed that undercorrection of 3° to 7° was correlated with better functional outcomes when compared to more neutral alignment. Moreover, Khamaisy and colleagues⁴⁹ recently showed that overcorrection during UKA surgery is more common in lateral than medial UKA.

These studies are important to understanding why OA progression is more common as a failure mode in lateral UKA. The shift of mechanical load from the lateral to medial epicondyle during the dynamic phase also could explain why aseptic loosening is less common in lateral UKA. As Hernigou and Deschamps⁴⁴ and Chatellard and colleagues⁴⁵ stated, undercorrection of varus deformity in medial UKA is associated with higher mechanical load on the medial prosthesis side and smaller joint space width. These factors are correlated with mechanical failure of medial UKA. We think this process can be applied to lateral UKA, with the addition that the mechanical load is higher on the healthy medial compartment during the dynamic phase. This causes more forces on the healthy (medial) side in lateral UKA, and in medial UKA more forces on the prosthesis (medial) side, which results in more OA progression in lateral UKA and more aseptic loosening in medial UKA. This finding is consistent with the results of our review of more OA progression and less aseptic loosening in lateral UKA. This study also suggests that medial and lateral UKA should not be reported together in studies that present survivorship, failure modes, or clinical outcomes.

A large discrepancy was seen in bearing dislocation between cohort studies (17%) and registry-based studies (5%). When we take a closer look to the bearing dislocation failures in the cohort studies, most of the failures were reported in only 2 cohort studies.^50,51 In a study by Pandit and colleagues,⁵⁰ 3 different prosthesis designs were used in 3 different time periods. In the first series of lateral UKA (1983-1991), 6 out of 51 (12%) bearings dislocated. In the second series (1998-2004), a modified technique was used and 3 out of 65 (5%) bearings dislocated. In the third series (2004-2008), a modified technique and a domed tibial component was used and only 1 out of 68 bearings dislocated (1%). In a study published in 1996, Gunther and colleagues⁵¹ also used surgical techniques and implants that were modified over the course of the study period. Because of these modified techniques, different implant designs, and year of publication, bearing dislocation most likely plays a smaller role than the 17% reported in the cohort studies. This discrepancy is a good example of the important role for the registries and registry-based studies in reporting failure modes and survivorship, especially in lateral UKA due to the low surgical frequency. Pabinger and colleagues⁵² recently performed a systematic review of cohort studies and registry-based studies in which they stated that the reliability in non-registry-based studies should be questioned and they considered registry-based studies superior in reporting UKA outcomes and revision rates. Furthermore, given the differences in anatomic and kinematic differences between the medial and lateral compartment and different failure modes between medial and lateral UKA, it would be better if future studies presented the medial and lateral failures separately. As stated above, most large cohort studies and especially annual registries currently do not report modes of failure of medial and lateral UKA separately.^3,4,18-20

There are limitations in this study. First, this systematic review is not a full meta-analysis but a pooled analysis of collected study series and retrospective studies. Therefore, we cannot exclude sampling bias, confounders, and selection bias from the literature. We included all studies reporting failure modes of lateral UKA and excluded all case reports. We made a conscious choice about including all lateral UKA failures because this is the first systematic review of lateral UKA failure modes. Another limitation is that the follow-up period of the studies differed (Table 1) and we did not correct for the follow-up period. As stated in the example of bearing dislocations, some of these studies reported old or different techniques, while other, more recently published studies used more modified techniques^11,29,53-56 Unfortunately, most studies did not report the time of arthroplasty survival and therefore we could not correct for the follow-up period.

In conclusion, progression of OA is the most common failure mode in lateral UKA, followed by aseptic loosening. Anatomic and kinematic factors such as alignment, mechanical forces during dynamic phase, and correction of valgus seem to play important roles in failure modes of lateral UKA. In the future, failure modes of medial and lateral UKA should be reported separately.

Am J Orthop. 2016;45(7):432-438, 462. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

In 1975, Skolnick and colleagues¹ introduced unicompartmental knee arthroplasty (UKA) for patients with isolated unicompartmental osteoarthritis (OA). They reported a study of 14 UKA procedures, of which 12 were at the medial and 2 at the lateral side. Forty years since this procedure was introduced, UKA is used in 8% to 12% of all knee arthroplasties.^2-6 A minority of these procedures are performed at the lateral side (5%-10%).^6-8

The considerable anatomical and kinematical differences between compartments^9-14 make it impossible to directly compare outcomes of medial and lateral UKA. For example, a greater degree of femoral roll and more posterior translation at the lateral side in flexion^9,10,13 can contribute to different pattern and volume differences of cartilage wear.¹⁵ Because of these differences, and because of implant design factors and lower surgical volume, lateral UKA is considered a technically more challenging surgery compared to medial UKA.^12,16,17

Since isolated lateral compartment OA is relatively scarce, current literature on lateral UKA is limited, and most studies combine medial and lateral outcomes to report UKA outcomes and failure modes.^3,4,18-20 However, as the UKA has grown in popularity over the last decade,^2,21-25 the number of reports about the lateral UKA also has increased. Recent studies reported excellent short-term survivorship results of the lateral UKA (96%-99%)^26,27 and smaller lateral UKA studies reported the 10-year survivorship with varying outcomes from good (84%)^14,28-30 to excellent (94%-100%).^8,31,32 Indeed, a recent systematic review showed survivorship of lateral UKA at 5, 10, and 15 years of 93%, 91%, and 89%, respectively.³³Because of the differences between the medial and lateral compartment, it is important to know the failure modes of lateral UKA in order to improve clinical outcomes and revision rates. We performed a systematic review of cohort studies and registry-based studies that reported lateral UKA failure to assess the causes of lateral UKA failure. In addition, we compared the failure modes in cohort studies with those found in registry-based studies.

Patients and Methods

Search Strategy and Criteria

Databases of PubMed, Embase, and Cochrane (Cochrane Central Register of Clinical Trials) were searched with the terms “knee, arthroplasty, replacement,” “unicompartmental,” “unicondylar,” “partial,” “UKA,” “UKR,” “UCA,” “UCR,” “PKA,” “PKR,” “PCA,” “prosthesis failure,” “reoperation,” “survivorship,” and “treatment failure.” After removal of duplicates, 2 authors (JPvdL and HAZ) scanned the articles for their title and abstract to assess eligibility for the study.

The full text of these eligible articles was further viewed and useful studies were selected using the inclusion and exclusion criteria. The references of these articles were scanned for additional studies and national registries (Figure).

Inclusion criteria were: (I) English language articles describing studies in humans published in the last 25 years, (II) retrospective and prospective studies, (III) featured lateral UKA, (IV) OA was indication for surgery, and (V) included failure modes data. The exclusion criteria were studies that featured: (I) only a specific group of failure (eg, bearing dislocations only), (II) previous surgery in ipsilateral knee (high tibial osteotomy, medial UKA), (III) acute concurrent knee diagnoses (acute anterior cruciate ligament rupture, acute meniscal tear), (IV) combined reporting of medial and lateral UKA, or (V) multiple studies with the same patient database.

Data Collection

All studies that reported modes of failure were used in this study and these failure modes were noted in a datasheet in Microsoft Excel 2011 (Microsoft).

The data of failures of lateral UKA are presented in Table 1 and are divided in cohort studies and registry-based studies. The final failure mode rates were presented in percentages (Table 2).

Statistical Analysis

For this systematic review, statistical analysis was performed with IBM SPSS Statistics 22 (SPSS Inc.). We performed chi square tests and Fisher’s exact tests to assess a difference between cohort studies and registry-based studies with the null hypothesis of no difference between both groups. A difference was considered significant when P < .05.

Results

Through the search of the databases, 1294 studies were identified and 26 handpicked studies were added. Initially, based on the title and abstract, 184 of these studies were found eligible.

After reviewing the full text of these articles, 25 studies (23 cohort studies and 2 registry-based studies) met the inclusion criteria and were included for the analysis of lateral UKA failure (Figure).

A total of 366 lateral UKA failures were included. The most common failure modes were progression of OA (29%), aseptic loosening (23%), and bearing dislocation (10%). Infection (6%), instability (6%), unexplained pain (6%), and fractures (4%) were less common causes of failure of lateral UKA (Table 2).

One hundred fifty-five of these failures were reported in the cohort studies. The most common modes of failure were OA progression (36%), bearing dislocation (17%) and aseptic loosening (16%). Less common were infection (10%), fractures (5%), pain (5%), and other causes (6%). In registry-based studies, with 211 lateral UKA failures, the most common modes of failure were aseptic loosening (28%), OA progression (24%), other causes (12%), instability (10%), pain (7%), bearing dislocation (5%), and polyethylene wear (4%) (Table 2).

When pooling cohort and registry-based studies, progression of OA was significantly more common than aseptic loosening (29% vs 23%, respectively; P < .01). It was also significantly more common in the cohort studies (36% vs 16%, respectively; P < .01) but no significant difference was found between progression of OA and aseptic loosening in registry-based studies (24% and 28%, respectively; P = .16) (Table 2).

When comparing cohort with registry-based studies, progression of OA was higher in cohort studies (36% vs. 24%; P < .01). Other failures modes that were more common in cohort studies compared with registry-based studies were bearing dislocation (17% vs 5%, respectively; P < .01) and infections (10% vs 3%, P < .01). Failure modes that were more common in registry-based studies than cohort studies were aseptic loosening (28% vs 16%, respectively; P < .01), other causes (12% vs 6%, respectively, P = .02), and instability (10% vs 1%, respectively, P < .01) (Table 2).

 

Discussion

In this systematic review, the most common failure modes in lateral UKA review were OA progression (29%), aseptic loosening (23%), and bearing dislocation (10%). Progression of OA and bearing dislocation were the most common modes of failure in cohort studies (36% and 17%, respectively), while aseptic loosening and OA progression were the most common failure modes in registry-based studies (28% and 24%, respectively).

As mentioned above, there are differences in anatomy and kinematics between the medial and lateral compartment. When the lateral UKA failure modes are compared with studies reporting medial UKA failure modes, differences in failure modes are seen.³⁴ Siddiqui and Ahmad³⁵ performed a systematic review of outcomes after UKA revision and presented a table with the failure modes of included studies. Unfortunately they did not report the ratio of medial and lateral UKA. However, when assuming an average percentage of 90% to 95% of medial UKA,^6,7,36 the main failure mode in their review in 17 out of 21 studies was aseptic loosening. Indeed, a recent systematic review on medial UKA failure modes showed that aseptic loosening is the most common cause of failure following this procedure.³⁴ Similarly, a search through registry-based studies^6,7 and large cohort studies^37-40 that only reported medial UKA failures showed that the majority of these studies^7,37-39 also reported aseptic loosening as the main cause of failure in medial UKA. When comparing the results of our systematic review of lateral UKA failures with the results of these studies of medial UKA failures, it seems that OA progression seems to play a more dominant role in failures of lateral UKA, while aseptic loosening seems to be more common in medial UKA.

Differences in anatomy and kinematics of the medial and lateral compartment can explain this. Malalignment of the joint is an important factor in the etiology of OA^41,42 and biomechanical studies showed that this malalignment can cause decreased viability and further degenerative changes of cartilage of the knee.⁴³ Hernigou and Deschamps⁴⁴ showed that the alignment of the knee after medial UKA is an important factor in postoperative joint changes. They found that overcorrection of varus deformity during medial UKA surgery, measured by the hip-knee-ankle (HKA) angle, was associated with increased OA at the lateral condyle and less tibial wear of the medial UKA. Undercorrection of the varus caused an increase in tibial wear of polyethylene. Chatellard and colleagues⁴⁵ found the same results in the correction of varus, measured by HKA. In addition, they found that when the prosthetic (medial) joint space was smaller than healthy (lateral) joint space, this was correlated with lower prosthesis survival. A smaller joint space at the healthy side was correlated with OA progression at the lateral compartment and tibial component wear.

These studies explain the mechanism of progression of OA and aseptic loosening. Harrington⁴⁶ assessed the load in patients with valgus and varus deformity. Patients with a valgus deformity have high mechanical load on the lateral condyle during the static phase, but during the dynamic phase, a major part of this load shifts to the medial condyle. In the patients with varus deformity, the mechanical load was noted on the medial condyle during both the static and dynamic phase. Ohdera and colleagues⁴⁷ advised, based on this biomechanical study and their own experiences, to correct the knee during lateral UKA to a slight valgus angle (5°-7°) to prevent OA progression at the medial side. van der List and colleagues⁴⁸ similarly showed that undercorrection of 3° to 7° was correlated with better functional outcomes when compared to more neutral alignment. Moreover, Khamaisy and colleagues⁴⁹ recently showed that overcorrection during UKA surgery is more common in lateral than medial UKA.

These studies are important to understanding why OA progression is more common as a failure mode in lateral UKA. The shift of mechanical load from the lateral to medial epicondyle during the dynamic phase also could explain why aseptic loosening is less common in lateral UKA. As Hernigou and Deschamps⁴⁴ and Chatellard and colleagues⁴⁵ stated, undercorrection of varus deformity in medial UKA is associated with higher mechanical load on the medial prosthesis side and smaller joint space width. These factors are correlated with mechanical failure of medial UKA. We think this process can be applied to lateral UKA, with the addition that the mechanical load is higher on the healthy medial compartment during the dynamic phase. This causes more forces on the healthy (medial) side in lateral UKA, and in medial UKA more forces on the prosthesis (medial) side, which results in more OA progression in lateral UKA and more aseptic loosening in medial UKA. This finding is consistent with the results of our review of more OA progression and less aseptic loosening in lateral UKA. This study also suggests that medial and lateral UKA should not be reported together in studies that present survivorship, failure modes, or clinical outcomes.

A large discrepancy was seen in bearing dislocation between cohort studies (17%) and registry-based studies (5%). When we take a closer look to the bearing dislocation failures in the cohort studies, most of the failures were reported in only 2 cohort studies.^50,51 In a study by Pandit and colleagues,⁵⁰ 3 different prosthesis designs were used in 3 different time periods. In the first series of lateral UKA (1983-1991), 6 out of 51 (12%) bearings dislocated. In the second series (1998-2004), a modified technique was used and 3 out of 65 (5%) bearings dislocated. In the third series (2004-2008), a modified technique and a domed tibial component was used and only 1 out of 68 bearings dislocated (1%). In a study published in 1996, Gunther and colleagues⁵¹ also used surgical techniques and implants that were modified over the course of the study period. Because of these modified techniques, different implant designs, and year of publication, bearing dislocation most likely plays a smaller role than the 17% reported in the cohort studies. This discrepancy is a good example of the important role for the registries and registry-based studies in reporting failure modes and survivorship, especially in lateral UKA due to the low surgical frequency. Pabinger and colleagues⁵² recently performed a systematic review of cohort studies and registry-based studies in which they stated that the reliability in non-registry-based studies should be questioned and they considered registry-based studies superior in reporting UKA outcomes and revision rates. Furthermore, given the differences in anatomic and kinematic differences between the medial and lateral compartment and different failure modes between medial and lateral UKA, it would be better if future studies presented the medial and lateral failures separately. As stated above, most large cohort studies and especially annual registries currently do not report modes of failure of medial and lateral UKA separately.^3,4,18-20

There are limitations in this study. First, this systematic review is not a full meta-analysis but a pooled analysis of collected study series and retrospective studies. Therefore, we cannot exclude sampling bias, confounders, and selection bias from the literature. We included all studies reporting failure modes of lateral UKA and excluded all case reports. We made a conscious choice about including all lateral UKA failures because this is the first systematic review of lateral UKA failure modes. Another limitation is that the follow-up period of the studies differed (Table 1) and we did not correct for the follow-up period. As stated in the example of bearing dislocations, some of these studies reported old or different techniques, while other, more recently published studies used more modified techniques^11,29,53-56 Unfortunately, most studies did not report the time of arthroplasty survival and therefore we could not correct for the follow-up period.

In conclusion, progression of OA is the most common failure mode in lateral UKA, followed by aseptic loosening. Anatomic and kinematic factors such as alignment, mechanical forces during dynamic phase, and correction of valgus seem to play important roles in failure modes of lateral UKA. In the future, failure modes of medial and lateral UKA should be reported separately.

Am J Orthop. 2016;45(7):432-438, 462. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

In 1975, Skolnick and colleagues¹ introduced unicompartmental knee arthroplasty (UKA) for patients with isolated unicompartmental osteoarthritis (OA). They reported a study of 14 UKA procedures, of which 12 were at the medial and 2 at the lateral side. Forty years since this procedure was introduced, UKA is used in 8% to 12% of all knee arthroplasties.^2-6 A minority of these procedures are performed at the lateral side (5%-10%).^6-8

The considerable anatomical and kinematical differences between compartments^9-14 make it impossible to directly compare outcomes of medial and lateral UKA. For example, a greater degree of femoral roll and more posterior translation at the lateral side in flexion^9,10,13 can contribute to different pattern and volume differences of cartilage wear.¹⁵ Because of these differences, and because of implant design factors and lower surgical volume, lateral UKA is considered a technically more challenging surgery compared to medial UKA.^12,16,17

Since isolated lateral compartment OA is relatively scarce, current literature on lateral UKA is limited, and most studies combine medial and lateral outcomes to report UKA outcomes and failure modes.^3,4,18-20 However, as the UKA has grown in popularity over the last decade,^2,21-25 the number of reports about the lateral UKA also has increased. Recent studies reported excellent short-term survivorship results of the lateral UKA (96%-99%)^26,27 and smaller lateral UKA studies reported the 10-year survivorship with varying outcomes from good (84%)^14,28-30 to excellent (94%-100%).^8,31,32 Indeed, a recent systematic review showed survivorship of lateral UKA at 5, 10, and 15 years of 93%, 91%, and 89%, respectively.³³Because of the differences between the medial and lateral compartment, it is important to know the failure modes of lateral UKA in order to improve clinical outcomes and revision rates. We performed a systematic review of cohort studies and registry-based studies that reported lateral UKA failure to assess the causes of lateral UKA failure. In addition, we compared the failure modes in cohort studies with those found in registry-based studies.

Patients and Methods

Search Strategy and Criteria

Databases of PubMed, Embase, and Cochrane (Cochrane Central Register of Clinical Trials) were searched with the terms “knee, arthroplasty, replacement,” “unicompartmental,” “unicondylar,” “partial,” “UKA,” “UKR,” “UCA,” “UCR,” “PKA,” “PKR,” “PCA,” “prosthesis failure,” “reoperation,” “survivorship,” and “treatment failure.” After removal of duplicates, 2 authors (JPvdL and HAZ) scanned the articles for their title and abstract to assess eligibility for the study.

The full text of these eligible articles was further viewed and useful studies were selected using the inclusion and exclusion criteria. The references of these articles were scanned for additional studies and national registries (Figure).

Inclusion criteria were: (I) English language articles describing studies in humans published in the last 25 years, (II) retrospective and prospective studies, (III) featured lateral UKA, (IV) OA was indication for surgery, and (V) included failure modes data. The exclusion criteria were studies that featured: (I) only a specific group of failure (eg, bearing dislocations only), (II) previous surgery in ipsilateral knee (high tibial osteotomy, medial UKA), (III) acute concurrent knee diagnoses (acute anterior cruciate ligament rupture, acute meniscal tear), (IV) combined reporting of medial and lateral UKA, or (V) multiple studies with the same patient database.

Data Collection

All studies that reported modes of failure were used in this study and these failure modes were noted in a datasheet in Microsoft Excel 2011 (Microsoft).

The data of failures of lateral UKA are presented in Table 1 and are divided in cohort studies and registry-based studies. The final failure mode rates were presented in percentages (Table 2).

Statistical Analysis

For this systematic review, statistical analysis was performed with IBM SPSS Statistics 22 (SPSS Inc.). We performed chi square tests and Fisher’s exact tests to assess a difference between cohort studies and registry-based studies with the null hypothesis of no difference between both groups. A difference was considered significant when P < .05.

Results

Through the search of the databases, 1294 studies were identified and 26 handpicked studies were added. Initially, based on the title and abstract, 184 of these studies were found eligible.

After reviewing the full text of these articles, 25 studies (23 cohort studies and 2 registry-based studies) met the inclusion criteria and were included for the analysis of lateral UKA failure (Figure).

A total of 366 lateral UKA failures were included. The most common failure modes were progression of OA (29%), aseptic loosening (23%), and bearing dislocation (10%). Infection (6%), instability (6%), unexplained pain (6%), and fractures (4%) were less common causes of failure of lateral UKA (Table 2).

One hundred fifty-five of these failures were reported in the cohort studies. The most common modes of failure were OA progression (36%), bearing dislocation (17%) and aseptic loosening (16%). Less common were infection (10%), fractures (5%), pain (5%), and other causes (6%). In registry-based studies, with 211 lateral UKA failures, the most common modes of failure were aseptic loosening (28%), OA progression (24%), other causes (12%), instability (10%), pain (7%), bearing dislocation (5%), and polyethylene wear (4%) (Table 2).

When pooling cohort and registry-based studies, progression of OA was significantly more common than aseptic loosening (29% vs 23%, respectively; P < .01). It was also significantly more common in the cohort studies (36% vs 16%, respectively; P < .01) but no significant difference was found between progression of OA and aseptic loosening in registry-based studies (24% and 28%, respectively; P = .16) (Table 2).

When comparing cohort with registry-based studies, progression of OA was higher in cohort studies (36% vs. 24%; P < .01). Other failures modes that were more common in cohort studies compared with registry-based studies were bearing dislocation (17% vs 5%, respectively; P < .01) and infections (10% vs 3%, P < .01). Failure modes that were more common in registry-based studies than cohort studies were aseptic loosening (28% vs 16%, respectively; P < .01), other causes (12% vs 6%, respectively, P = .02), and instability (10% vs 1%, respectively, P < .01) (Table 2).

 

Discussion

In this systematic review, the most common failure modes in lateral UKA review were OA progression (29%), aseptic loosening (23%), and bearing dislocation (10%). Progression of OA and bearing dislocation were the most common modes of failure in cohort studies (36% and 17%, respectively), while aseptic loosening and OA progression were the most common failure modes in registry-based studies (28% and 24%, respectively).

As mentioned above, there are differences in anatomy and kinematics between the medial and lateral compartment. When the lateral UKA failure modes are compared with studies reporting medial UKA failure modes, differences in failure modes are seen.³⁴ Siddiqui and Ahmad³⁵ performed a systematic review of outcomes after UKA revision and presented a table with the failure modes of included studies. Unfortunately they did not report the ratio of medial and lateral UKA. However, when assuming an average percentage of 90% to 95% of medial UKA,^6,7,36 the main failure mode in their review in 17 out of 21 studies was aseptic loosening. Indeed, a recent systematic review on medial UKA failure modes showed that aseptic loosening is the most common cause of failure following this procedure.³⁴ Similarly, a search through registry-based studies^6,7 and large cohort studies^37-40 that only reported medial UKA failures showed that the majority of these studies^7,37-39 also reported aseptic loosening as the main cause of failure in medial UKA. When comparing the results of our systematic review of lateral UKA failures with the results of these studies of medial UKA failures, it seems that OA progression seems to play a more dominant role in failures of lateral UKA, while aseptic loosening seems to be more common in medial UKA.

Differences in anatomy and kinematics of the medial and lateral compartment can explain this. Malalignment of the joint is an important factor in the etiology of OA^41,42 and biomechanical studies showed that this malalignment can cause decreased viability and further degenerative changes of cartilage of the knee.⁴³ Hernigou and Deschamps⁴⁴ showed that the alignment of the knee after medial UKA is an important factor in postoperative joint changes. They found that overcorrection of varus deformity during medial UKA surgery, measured by the hip-knee-ankle (HKA) angle, was associated with increased OA at the lateral condyle and less tibial wear of the medial UKA. Undercorrection of the varus caused an increase in tibial wear of polyethylene. Chatellard and colleagues⁴⁵ found the same results in the correction of varus, measured by HKA. In addition, they found that when the prosthetic (medial) joint space was smaller than healthy (lateral) joint space, this was correlated with lower prosthesis survival. A smaller joint space at the healthy side was correlated with OA progression at the lateral compartment and tibial component wear.

These studies explain the mechanism of progression of OA and aseptic loosening. Harrington⁴⁶ assessed the load in patients with valgus and varus deformity. Patients with a valgus deformity have high mechanical load on the lateral condyle during the static phase, but during the dynamic phase, a major part of this load shifts to the medial condyle. In the patients with varus deformity, the mechanical load was noted on the medial condyle during both the static and dynamic phase. Ohdera and colleagues⁴⁷ advised, based on this biomechanical study and their own experiences, to correct the knee during lateral UKA to a slight valgus angle (5°-7°) to prevent OA progression at the medial side. van der List and colleagues⁴⁸ similarly showed that undercorrection of 3° to 7° was correlated with better functional outcomes when compared to more neutral alignment. Moreover, Khamaisy and colleagues⁴⁹ recently showed that overcorrection during UKA surgery is more common in lateral than medial UKA.

These studies are important to understanding why OA progression is more common as a failure mode in lateral UKA. The shift of mechanical load from the lateral to medial epicondyle during the dynamic phase also could explain why aseptic loosening is less common in lateral UKA. As Hernigou and Deschamps⁴⁴ and Chatellard and colleagues⁴⁵ stated, undercorrection of varus deformity in medial UKA is associated with higher mechanical load on the medial prosthesis side and smaller joint space width. These factors are correlated with mechanical failure of medial UKA. We think this process can be applied to lateral UKA, with the addition that the mechanical load is higher on the healthy medial compartment during the dynamic phase. This causes more forces on the healthy (medial) side in lateral UKA, and in medial UKA more forces on the prosthesis (medial) side, which results in more OA progression in lateral UKA and more aseptic loosening in medial UKA. This finding is consistent with the results of our review of more OA progression and less aseptic loosening in lateral UKA. This study also suggests that medial and lateral UKA should not be reported together in studies that present survivorship, failure modes, or clinical outcomes.

A large discrepancy was seen in bearing dislocation between cohort studies (17%) and registry-based studies (5%). When we take a closer look to the bearing dislocation failures in the cohort studies, most of the failures were reported in only 2 cohort studies.^50,51 In a study by Pandit and colleagues,⁵⁰ 3 different prosthesis designs were used in 3 different time periods. In the first series of lateral UKA (1983-1991), 6 out of 51 (12%) bearings dislocated. In the second series (1998-2004), a modified technique was used and 3 out of 65 (5%) bearings dislocated. In the third series (2004-2008), a modified technique and a domed tibial component was used and only 1 out of 68 bearings dislocated (1%). In a study published in 1996, Gunther and colleagues⁵¹ also used surgical techniques and implants that were modified over the course of the study period. Because of these modified techniques, different implant designs, and year of publication, bearing dislocation most likely plays a smaller role than the 17% reported in the cohort studies. This discrepancy is a good example of the important role for the registries and registry-based studies in reporting failure modes and survivorship, especially in lateral UKA due to the low surgical frequency. Pabinger and colleagues⁵² recently performed a systematic review of cohort studies and registry-based studies in which they stated that the reliability in non-registry-based studies should be questioned and they considered registry-based studies superior in reporting UKA outcomes and revision rates. Furthermore, given the differences in anatomic and kinematic differences between the medial and lateral compartment and different failure modes between medial and lateral UKA, it would be better if future studies presented the medial and lateral failures separately. As stated above, most large cohort studies and especially annual registries currently do not report modes of failure of medial and lateral UKA separately.^3,4,18-20

There are limitations in this study. First, this systematic review is not a full meta-analysis but a pooled analysis of collected study series and retrospective studies. Therefore, we cannot exclude sampling bias, confounders, and selection bias from the literature. We included all studies reporting failure modes of lateral UKA and excluded all case reports. We made a conscious choice about including all lateral UKA failures because this is the first systematic review of lateral UKA failure modes. Another limitation is that the follow-up period of the studies differed (Table 1) and we did not correct for the follow-up period. As stated in the example of bearing dislocations, some of these studies reported old or different techniques, while other, more recently published studies used more modified techniques^11,29,53-56 Unfortunately, most studies did not report the time of arthroplasty survival and therefore we could not correct for the follow-up period.

In conclusion, progression of OA is the most common failure mode in lateral UKA, followed by aseptic loosening. Anatomic and kinematic factors such as alignment, mechanical forces during dynamic phase, and correction of valgus seem to play important roles in failure modes of lateral UKA. In the future, failure modes of medial and lateral UKA should be reported separately.

Am J Orthop. 2016;45(7):432-438, 462. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

The annual number of total shoulder arthroplasties (TSAs) is rising with the growing elderly population and development of new technologies such as reverse shoulder arthroplasty.¹ In 2008, 47,000 shoulder arthroplasties were performed in the US compared with 19,000 in 1998.¹ As of 2011, there were 53,000 shoulder arthroplasties performed annually.² Pain control after shoulder procedures, particularly TSA, is challenging. ³

Several modalities exist to manage pain after shoulder arthroplasty. The interscalene brachial plexus nerve block is considered the “gold standard” for shoulder analgesia. A new approach is the periarticular injection method, in which the surgeon administers a local anesthetic intraoperatively. Liposomal bupivacaine (Exparel, Pacira Pharmaceuticals, Inc.) is a nonopioid anesthetic that has been shown to improve pain control, shorten hospital stays, and decrease costs for total knee and hip arthroplasty compared with nerve blocks.^4-6 Patients who were treated with liposomal bupivacaine consumed less opioid medication than a placebo group.⁷

Our purpose was to compare intraoperative local liposomal bupivacaine injection with preoperative single-shot interscalene nerve block (ISNB) in terms of pain control, opioid use, and length of hospital stay (LOS) after shoulder arthroplasty. We hypothesized that patients in the liposomal bupivacaine group would have lower pain scores, less opioid use, and shorter LOS compared with patients in the ISNB group.

Methods

A retrospective cohort analysis was conducted with 58 patients who underwent shoulder arthroplasty by 1 surgeon at our academically affiliated community hospital from January 2012 through January 2015. ISNBs were the standard at the beginning of the study period and were used until Exparel became available on the hospital formulary in 2013. We began using Exparel for all shoulder arthroplasties in November 2013. No other changes were made in the perioperative management of our arthroplasty patients during this period. Patients who underwent TSA, reverse TSA, or hemiarthroplasty of the shoulder were included. Patients who underwent revision TSA were excluded. Twenty-one patients received ISNBs and 37 received liposomal bupivacaine injections. This study was approved by our Institutional Review Board.

Baseline data for each patient were age, sex, body mass index, and the American Society of Anesthesiologists (ASA) Physical Status Classification. The primary outcome measure was the numeric rating scale (NRS) pain score at 4 post-operative time intervals. The NRS pain score has a range of 0 to 10, with 10 representing severe pain. Data were gathered from nursing and physical therapy notes in patient charts. The postoperative time intervals were 0 to 1 hour, 8 to 14 hours, 18 to 24 hours, and 27 to 36 hours. Available NRS scores for these time intervals were averaged. Patients were included if they had pain scores for at least 3 of the postoperative time intervals documented in their charts. Secondary outcome measures were LOS and opioid consumption during hospital admission. Intravenous acetaminophen use was also measured in both groups. All data on opioids were converted to oral morphine equivalents using the method described by Schneider and colleagues.⁸

A board-certified, fellowship-trained anesthesiologist, experienced in regional anesthesia, administered the single-shot ISNB before surgery. The block was administered under ultrasound guidance using a 44-mm, 22-gauge needle with the patient in the supine position. No indwelling catheter was used. The medication consisted of 30 mL of 5% ropivacaine (5 mg/mL). The surgeon injected liposomal bupivacaine (266 mg diluted into 40 mL of injectable saline) near the end of the procedure throughout the pericapsular area and multiple layers of the wound, per manufacturer guidelines.⁹ A 60-mL syringe with a 20-gauge needle was used. All operations were performed by 1 board-certified, fellowship-trained surgeon using a standard deltopectoral approach with the same surgical equipment. The same postoperative pain protocol was used for all patients, including intravenous acetaminophen and patient-controlled analgesia. Additional oral pain medication was provided as needed for all patients. Physical therapy protocols were identical between groups.

Statistical Analysis

Mean patient ages in the 2 treatment groups were compared using the Student t test. Sex distribution and the ASA scores were compared using a χ² test and a Fisher exact test, respectively. Arthroplasty types were compared using a Fisher exact test. The medians and interquartile ranges of the NRS scores at each time point measured were tabulated by treatment group, and at each time point the difference between groups was tested using nonparametric rank sum tests.

We tested the longitudinal trajectory of NRS scores over time, accounting for repeated measurements in the same patients using linear mixed model analysis. Treatment group, time period as a categorical variable, and the interaction between treatment and time period were included as fixed effects, and patient identification number was included as the random effect. An initial omnibus test was performed for all treatment and treatment-by-time interaction effects. Subsequently, the treatment-by-time interaction was tested for each of the time periods. The association of day of discharge (as a categorical variable) with treatment was tested using the Fisher exact test. All analyses were conducted using Stata, version 13, software (StataCorp LP). P values <.05 were considered significant.

 

Sample Size Analysis

We calculated the minimum detectable effect size with 80% power at an alpha level of 0.05 for the nonparametric rank sum test in terms of the proportion of every possible pair of patients treated with the 2 treatments, where the patient treated with liposomal bupivacaine has a lower pain score than the patient treated with ISNB. For pain score at 18 to 24 hours, the sample sizes of 33 patients treated with liposomal bupivacaine and 20 treated with ISNB, the minimum detectable effect size is 73%.

Results

Fifty-eight patient charts (21 in the ISNB group and 37 in the liposomal bupivacaine group) were reviewed for the study. Patient sex distribution, mean age, mean body mass index, and mean baseline ASA scores were not statistically different (Table 1).

In the ISNB group, 5 patients had hemiarthroplasty, 12 had TSA, and 4 had reverse TSA. In the liposomal bupivacaine group, 1 patient had hemiarthroplasty, 23 had TSA, and 13 had reverse TSA. Frequency of procedure types was significantly different between groups (P = .039), with the liposomal bupivacaine group undergoing fewer hemiarthroplasties.

The primary outcome measure, NRS pain score, showed no significant differences between groups at 0 to 1 hour after surgery (P = .99) or 8 to 14 hours after surgery (P = .208).

At 18 to 24 hours after surgery, the liposomal bupivacaine group had a lower mean NRS score than the ISNB group (P = .001). This was statistically significant when taking repeated measures of variance into account (Figure 1). Mean NRS score was also lower for the liposomal bupivacaine group at 27 to 36 hours after surgery (P = .029).

This was a significant difference when repeated measures of variance was considered (Table 2).

There was no difference in the amount of intravenous acetaminophen given during the hospital stay between groups. There was no significant difference in opioid consumption on postoperative day 1 in the hospital (P = .59) (Figure 2). However, there were significant differences between groups on postoperative days 2 and 3.

On postoperative day 2, the ISNB group required significantly more opioids (mean, 112 mg morphine equivalents) than the liposomal bupivacaine group (mean, 37 mg morphine equivalents) (P = .001). The ISNB group also required significantly more opioids (mean, 25 mg morphine equivalents) on postoperative day 3 than the liposomal bupivacaine group (mean, 5 mg) (P = .002).

Sixteen of 37 patients in the liposomal bupivacaine group and 2 of 21 in the ISNB group were discharged on the day after surgery (P = .010) (Table 3).

The mean LOS was 46 ± 20 hours for the liposomal bupivacaine group and 57 ± 14 hours for the ISNB group (P = .012).

There were no major cardiac or respiratory events in either group. No long-term paresthesias or neuropathies were noted. There were no readmissions for either group.

Discussion

Postoperative pain control after shoulder arthroplasty can be challenging, and several modalities have been tried in various combinations to minimize pain and decrease adverse effects of opioid medications. The most common method for pain relief after shoulder arthroplasty is the ISNB. Several studies of ISNBs have shown improved pain control after shoulder arthroplasty with associated decreased opioid consumption and related side effects.¹⁰ Patient rehabilitation and satisfaction have improved with the increasing use of peripheral nerve blocks.¹¹

Despite the well-established benefits of ISNBs, several limitations exist. Although the superior portion of the shoulder is well covered by an ISNB, the inferior portion of the brachial plexus can remain uncovered or only partially covered.¹² Complications of ISNBs include hemidiaphragmatic paresis, rebound pain 24 hours after surgery,¹³ chronic neurologic complications,¹⁴ and substantial respiratory and cardiovascular events.¹⁵ Nerve blocks also require additional time and resources in the perioperative period, including an anesthesiologist with specialized training, assistants, and ultrasonography or nerve stimulation equipment contraindicated in patients taking blood thinners.¹⁶

Periarticular injections of local anesthetics have also shown promise in reducing pain after arthroplasty.⁴ Benefits include an enhanced safety profile because local injection avoids the concurrent blockade of the phrenic nerve and recurrent laryngeal nerve and has not been associated with the risk of peripheral neuropathies. Further, local injection is a simple technique that can be performed during surgery without additional personnel or expertise. A limitation of this approach is the relatively short duration of effectiveness of the local anesthetic and uncertainty regarding the best agent and the ideal volume of injection.⁶ Liposomal bupivacaine is a new agent (approved by the US Food and Drug Administration in 2011¹⁷) with a sustained release over 72 to 96 hours.¹⁸ The most common adverse effects of liposomal bupivacaine are nausea, vomiting, constipation, pyrexia, dizziness, and headache.¹⁹ Chondrotoxicity and granulomatous inflammation are more serious, yet rare, complications of liposomal bupivacaine.²⁰

We found that liposomal bupivacaine injections were associated with lower pain scores compared with ISNB at 18 to 24 hours after surgery. This correlated with less opioid consumption in the liposomal bupivacaine group than in the ISNB group on the second postoperative day. These differences in pain values are consistent with the known pharmacokinetics of liposomal bupivacaine.¹⁸ Peak plasma levels normally occur approximately 24 hours after injection, leaving the early postoperative period relatively uncovered by anesthetic agent. This finding of relatively poor pain control early after surgery has also been noted in patients undergoing knee arthroplasty.⁵ On the basis of the findings of this study, we have added standard bupivacaine injections to our separate liposomal bupivacaine injection to cover early postoperative pain. Opioid consumption was significantly lower in the liposomal bupivacaine group than in the ISNB group on postoperative days 2 and 3. We did not measure adverse events related to opioid consumption, so we cannot comment on whether the decreased opioid consumption was associated with the rate of adverse events. However, other studies^21,22 have established this relationship.

We found the liposomal bupivacaine group to have earlier discharges to home. Sixteen of 37 patients in the liposomal bupivacaine group compared with 2 of 21 patients in the ISNB group were discharged on the day after surgery. A mean reduction in LOS of 18 hours for the liposomal bupivacaine group was statistically significant (P = .012). This reduction in LOS has important implications for hospitals and value analysis committees considering whether to keep a new, more expensive local anesthetic on formulary. Savings from reduced LOS and improvements in patient satisfaction may justify the expense (approximately $300 per 266-mg vial) of Exparel.

From a societal cost perspective, liposomal bupivacaine is more economical compared with ISNB, which adds approximately $1500 to the cost of anesthesia per patient.²³ Eliminating the costs associated with ISNB administration in shoulder arthroplasties could result in substantial savings to our healthcare system. More research examining time savings and exact costs of each procedure is needed to determine the true cost effectiveness of each approach.

Limitations of our study include the retrospective design, relatively small numbers of patients in each group, missing data for some patients at various time points, variation in the types of procedures in each group, and lack of long-term outcome measures. It is important to note that we did not confirm the success of the nerve block after administration. However, this study reflects the effectiveness of each of the modalities in actual clinical conditions (as opposed to a controlled experimental setting). The actual effectiveness of a nerve block varies, even when performed by an experienced anesthesiologist with ultrasound guidance. Furthermore, immediate postoperative pain scores in the nerve block group are consistent with those of prior research reporting pain values ranging from 4 to 5 and a mean duration of effect ranging from 9 to 14 hours.^23,24 Additionally, the patients, surgeon, and nursing team were not blinded to the treatment group. Although we did note a significant difference in the types of procedures between groups, this finding is related to the greater number of hemiarthroplasties performed in the ISNB group (N = 5) compared with the liposomal group (N = 1). Because of this variation and the decreased invasiveness of hemiarthroplasties, the bias is against the liposomal group. Finally, our primary outcome variable was pain, which is a subjective, self-reported measure. However, our opioid consumption data and LOS data corroborate the improved pain scores in the liposomal bupivacaine group.

Limiting the study to a single surgeon may limit external validity. Another limitation is the lack of data on adverse events related to opioid medication use. There was no additional experimental group to determine whether less expensive local anesthetics injected locally would perform similarly to liposomal bupivacaine. In total knee arthroplasty, periarticular injections of liposomal bupivacaine were not as effective as less expensive periarticular injections.²⁵ It is unclear which agents (and in what doses or combinations) should be used for periarticular injections. Finally, we acknowledge that our retrospective study design cannot account for all potential factors affecting discharge time.

This is the first comparative study of liposomal bupivacaine and ISNB in TSA. The study design allowed us to control for variables such as surgical technique, postoperative protocols (including use and type of sling), and use of other pain modalities such as patient-controlled analgesia and intravenous acetaminophen that are likely to affect postoperative pain and LOS. This study provides preliminary data that confirm relative equipoise between liposomal bupivacaine and ISNB, which is needed for the ethical conduct of a randomized controlled trial. Such a trial would allow for a more robust comparison, and this retrospective study provides appropriate pilot data on which to base this design and the clinical information needed to counsel patients during enrollment.

Our results suggest that liposomal bupivacaine may provide superior or similar pain relief compared with ISNB after shoulder arthroplasty. Additionally, the use of liposomal bupivacaine was associated with decreased opioid consumption and earlier discharge to home compared with ISNB. These findings have important implications for pain control after TSA because pain represents a major concern for patients and providers after surgery. In addition to clinical improvements, use of liposomal bupivacaine may save time and eliminate costs associated with administering nerve blocks. Local injection may also be used in patients who are contraindicated for ISNB such as those with obesity, pulmonary disease, or peripheral neuropathy. Although we cannot definitively suggest that liposomal bupivacaine is superior to the current gold standard ISNB for pain control after shoulder arthroplasty, our results suggest a relative clinical equipoise between these modalities. Larger analytical studies, including randomized trials, should be performed to explore the potential benefits of liposomal bupivacaine injections for pain control after shoulder arthroplasty.

Am J Orthop. 2016;45(7):424-430. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

The annual number of total shoulder arthroplasties (TSAs) is rising with the growing elderly population and development of new technologies such as reverse shoulder arthroplasty.¹ In 2008, 47,000 shoulder arthroplasties were performed in the US compared with 19,000 in 1998.¹ As of 2011, there were 53,000 shoulder arthroplasties performed annually.² Pain control after shoulder procedures, particularly TSA, is challenging. ³

Several modalities exist to manage pain after shoulder arthroplasty. The interscalene brachial plexus nerve block is considered the “gold standard” for shoulder analgesia. A new approach is the periarticular injection method, in which the surgeon administers a local anesthetic intraoperatively. Liposomal bupivacaine (Exparel, Pacira Pharmaceuticals, Inc.) is a nonopioid anesthetic that has been shown to improve pain control, shorten hospital stays, and decrease costs for total knee and hip arthroplasty compared with nerve blocks.^4-6 Patients who were treated with liposomal bupivacaine consumed less opioid medication than a placebo group.⁷

Our purpose was to compare intraoperative local liposomal bupivacaine injection with preoperative single-shot interscalene nerve block (ISNB) in terms of pain control, opioid use, and length of hospital stay (LOS) after shoulder arthroplasty. We hypothesized that patients in the liposomal bupivacaine group would have lower pain scores, less opioid use, and shorter LOS compared with patients in the ISNB group.

Methods

A retrospective cohort analysis was conducted with 58 patients who underwent shoulder arthroplasty by 1 surgeon at our academically affiliated community hospital from January 2012 through January 2015. ISNBs were the standard at the beginning of the study period and were used until Exparel became available on the hospital formulary in 2013. We began using Exparel for all shoulder arthroplasties in November 2013. No other changes were made in the perioperative management of our arthroplasty patients during this period. Patients who underwent TSA, reverse TSA, or hemiarthroplasty of the shoulder were included. Patients who underwent revision TSA were excluded. Twenty-one patients received ISNBs and 37 received liposomal bupivacaine injections. This study was approved by our Institutional Review Board.

Baseline data for each patient were age, sex, body mass index, and the American Society of Anesthesiologists (ASA) Physical Status Classification. The primary outcome measure was the numeric rating scale (NRS) pain score at 4 post-operative time intervals. The NRS pain score has a range of 0 to 10, with 10 representing severe pain. Data were gathered from nursing and physical therapy notes in patient charts. The postoperative time intervals were 0 to 1 hour, 8 to 14 hours, 18 to 24 hours, and 27 to 36 hours. Available NRS scores for these time intervals were averaged. Patients were included if they had pain scores for at least 3 of the postoperative time intervals documented in their charts. Secondary outcome measures were LOS and opioid consumption during hospital admission. Intravenous acetaminophen use was also measured in both groups. All data on opioids were converted to oral morphine equivalents using the method described by Schneider and colleagues.⁸

A board-certified, fellowship-trained anesthesiologist, experienced in regional anesthesia, administered the single-shot ISNB before surgery. The block was administered under ultrasound guidance using a 44-mm, 22-gauge needle with the patient in the supine position. No indwelling catheter was used. The medication consisted of 30 mL of 5% ropivacaine (5 mg/mL). The surgeon injected liposomal bupivacaine (266 mg diluted into 40 mL of injectable saline) near the end of the procedure throughout the pericapsular area and multiple layers of the wound, per manufacturer guidelines.⁹ A 60-mL syringe with a 20-gauge needle was used. All operations were performed by 1 board-certified, fellowship-trained surgeon using a standard deltopectoral approach with the same surgical equipment. The same postoperative pain protocol was used for all patients, including intravenous acetaminophen and patient-controlled analgesia. Additional oral pain medication was provided as needed for all patients. Physical therapy protocols were identical between groups.

Statistical Analysis

Mean patient ages in the 2 treatment groups were compared using the Student t test. Sex distribution and the ASA scores were compared using a χ² test and a Fisher exact test, respectively. Arthroplasty types were compared using a Fisher exact test. The medians and interquartile ranges of the NRS scores at each time point measured were tabulated by treatment group, and at each time point the difference between groups was tested using nonparametric rank sum tests.

We tested the longitudinal trajectory of NRS scores over time, accounting for repeated measurements in the same patients using linear mixed model analysis. Treatment group, time period as a categorical variable, and the interaction between treatment and time period were included as fixed effects, and patient identification number was included as the random effect. An initial omnibus test was performed for all treatment and treatment-by-time interaction effects. Subsequently, the treatment-by-time interaction was tested for each of the time periods. The association of day of discharge (as a categorical variable) with treatment was tested using the Fisher exact test. All analyses were conducted using Stata, version 13, software (StataCorp LP). P values <.05 were considered significant.

 

Sample Size Analysis

We calculated the minimum detectable effect size with 80% power at an alpha level of 0.05 for the nonparametric rank sum test in terms of the proportion of every possible pair of patients treated with the 2 treatments, where the patient treated with liposomal bupivacaine has a lower pain score than the patient treated with ISNB. For pain score at 18 to 24 hours, the sample sizes of 33 patients treated with liposomal bupivacaine and 20 treated with ISNB, the minimum detectable effect size is 73%.

Results

Fifty-eight patient charts (21 in the ISNB group and 37 in the liposomal bupivacaine group) were reviewed for the study. Patient sex distribution, mean age, mean body mass index, and mean baseline ASA scores were not statistically different (Table 1).

In the ISNB group, 5 patients had hemiarthroplasty, 12 had TSA, and 4 had reverse TSA. In the liposomal bupivacaine group, 1 patient had hemiarthroplasty, 23 had TSA, and 13 had reverse TSA. Frequency of procedure types was significantly different between groups (P = .039), with the liposomal bupivacaine group undergoing fewer hemiarthroplasties.

The primary outcome measure, NRS pain score, showed no significant differences between groups at 0 to 1 hour after surgery (P = .99) or 8 to 14 hours after surgery (P = .208).

At 18 to 24 hours after surgery, the liposomal bupivacaine group had a lower mean NRS score than the ISNB group (P = .001). This was statistically significant when taking repeated measures of variance into account (Figure 1). Mean NRS score was also lower for the liposomal bupivacaine group at 27 to 36 hours after surgery (P = .029).

This was a significant difference when repeated measures of variance was considered (Table 2).

There was no difference in the amount of intravenous acetaminophen given during the hospital stay between groups. There was no significant difference in opioid consumption on postoperative day 1 in the hospital (P = .59) (Figure 2). However, there were significant differences between groups on postoperative days 2 and 3.

On postoperative day 2, the ISNB group required significantly more opioids (mean, 112 mg morphine equivalents) than the liposomal bupivacaine group (mean, 37 mg morphine equivalents) (P = .001). The ISNB group also required significantly more opioids (mean, 25 mg morphine equivalents) on postoperative day 3 than the liposomal bupivacaine group (mean, 5 mg) (P = .002).

Sixteen of 37 patients in the liposomal bupivacaine group and 2 of 21 in the ISNB group were discharged on the day after surgery (P = .010) (Table 3).

The mean LOS was 46 ± 20 hours for the liposomal bupivacaine group and 57 ± 14 hours for the ISNB group (P = .012).

There were no major cardiac or respiratory events in either group. No long-term paresthesias or neuropathies were noted. There were no readmissions for either group.

Discussion

Postoperative pain control after shoulder arthroplasty can be challenging, and several modalities have been tried in various combinations to minimize pain and decrease adverse effects of opioid medications. The most common method for pain relief after shoulder arthroplasty is the ISNB. Several studies of ISNBs have shown improved pain control after shoulder arthroplasty with associated decreased opioid consumption and related side effects.¹⁰ Patient rehabilitation and satisfaction have improved with the increasing use of peripheral nerve blocks.¹¹

Despite the well-established benefits of ISNBs, several limitations exist. Although the superior portion of the shoulder is well covered by an ISNB, the inferior portion of the brachial plexus can remain uncovered or only partially covered.¹² Complications of ISNBs include hemidiaphragmatic paresis, rebound pain 24 hours after surgery,¹³ chronic neurologic complications,¹⁴ and substantial respiratory and cardiovascular events.¹⁵ Nerve blocks also require additional time and resources in the perioperative period, including an anesthesiologist with specialized training, assistants, and ultrasonography or nerve stimulation equipment contraindicated in patients taking blood thinners.¹⁶

Periarticular injections of local anesthetics have also shown promise in reducing pain after arthroplasty.⁴ Benefits include an enhanced safety profile because local injection avoids the concurrent blockade of the phrenic nerve and recurrent laryngeal nerve and has not been associated with the risk of peripheral neuropathies. Further, local injection is a simple technique that can be performed during surgery without additional personnel or expertise. A limitation of this approach is the relatively short duration of effectiveness of the local anesthetic and uncertainty regarding the best agent and the ideal volume of injection.⁶ Liposomal bupivacaine is a new agent (approved by the US Food and Drug Administration in 2011¹⁷) with a sustained release over 72 to 96 hours.¹⁸ The most common adverse effects of liposomal bupivacaine are nausea, vomiting, constipation, pyrexia, dizziness, and headache.¹⁹ Chondrotoxicity and granulomatous inflammation are more serious, yet rare, complications of liposomal bupivacaine.²⁰

We found that liposomal bupivacaine injections were associated with lower pain scores compared with ISNB at 18 to 24 hours after surgery. This correlated with less opioid consumption in the liposomal bupivacaine group than in the ISNB group on the second postoperative day. These differences in pain values are consistent with the known pharmacokinetics of liposomal bupivacaine.¹⁸ Peak plasma levels normally occur approximately 24 hours after injection, leaving the early postoperative period relatively uncovered by anesthetic agent. This finding of relatively poor pain control early after surgery has also been noted in patients undergoing knee arthroplasty.⁵ On the basis of the findings of this study, we have added standard bupivacaine injections to our separate liposomal bupivacaine injection to cover early postoperative pain. Opioid consumption was significantly lower in the liposomal bupivacaine group than in the ISNB group on postoperative days 2 and 3. We did not measure adverse events related to opioid consumption, so we cannot comment on whether the decreased opioid consumption was associated with the rate of adverse events. However, other studies^21,22 have established this relationship.

We found the liposomal bupivacaine group to have earlier discharges to home. Sixteen of 37 patients in the liposomal bupivacaine group compared with 2 of 21 patients in the ISNB group were discharged on the day after surgery. A mean reduction in LOS of 18 hours for the liposomal bupivacaine group was statistically significant (P = .012). This reduction in LOS has important implications for hospitals and value analysis committees considering whether to keep a new, more expensive local anesthetic on formulary. Savings from reduced LOS and improvements in patient satisfaction may justify the expense (approximately $300 per 266-mg vial) of Exparel.

From a societal cost perspective, liposomal bupivacaine is more economical compared with ISNB, which adds approximately $1500 to the cost of anesthesia per patient.²³ Eliminating the costs associated with ISNB administration in shoulder arthroplasties could result in substantial savings to our healthcare system. More research examining time savings and exact costs of each procedure is needed to determine the true cost effectiveness of each approach.

Limitations of our study include the retrospective design, relatively small numbers of patients in each group, missing data for some patients at various time points, variation in the types of procedures in each group, and lack of long-term outcome measures. It is important to note that we did not confirm the success of the nerve block after administration. However, this study reflects the effectiveness of each of the modalities in actual clinical conditions (as opposed to a controlled experimental setting). The actual effectiveness of a nerve block varies, even when performed by an experienced anesthesiologist with ultrasound guidance. Furthermore, immediate postoperative pain scores in the nerve block group are consistent with those of prior research reporting pain values ranging from 4 to 5 and a mean duration of effect ranging from 9 to 14 hours.^23,24 Additionally, the patients, surgeon, and nursing team were not blinded to the treatment group. Although we did note a significant difference in the types of procedures between groups, this finding is related to the greater number of hemiarthroplasties performed in the ISNB group (N = 5) compared with the liposomal group (N = 1). Because of this variation and the decreased invasiveness of hemiarthroplasties, the bias is against the liposomal group. Finally, our primary outcome variable was pain, which is a subjective, self-reported measure. However, our opioid consumption data and LOS data corroborate the improved pain scores in the liposomal bupivacaine group.

Limiting the study to a single surgeon may limit external validity. Another limitation is the lack of data on adverse events related to opioid medication use. There was no additional experimental group to determine whether less expensive local anesthetics injected locally would perform similarly to liposomal bupivacaine. In total knee arthroplasty, periarticular injections of liposomal bupivacaine were not as effective as less expensive periarticular injections.²⁵ It is unclear which agents (and in what doses or combinations) should be used for periarticular injections. Finally, we acknowledge that our retrospective study design cannot account for all potential factors affecting discharge time.

This is the first comparative study of liposomal bupivacaine and ISNB in TSA. The study design allowed us to control for variables such as surgical technique, postoperative protocols (including use and type of sling), and use of other pain modalities such as patient-controlled analgesia and intravenous acetaminophen that are likely to affect postoperative pain and LOS. This study provides preliminary data that confirm relative equipoise between liposomal bupivacaine and ISNB, which is needed for the ethical conduct of a randomized controlled trial. Such a trial would allow for a more robust comparison, and this retrospective study provides appropriate pilot data on which to base this design and the clinical information needed to counsel patients during enrollment.

Our results suggest that liposomal bupivacaine may provide superior or similar pain relief compared with ISNB after shoulder arthroplasty. Additionally, the use of liposomal bupivacaine was associated with decreased opioid consumption and earlier discharge to home compared with ISNB. These findings have important implications for pain control after TSA because pain represents a major concern for patients and providers after surgery. In addition to clinical improvements, use of liposomal bupivacaine may save time and eliminate costs associated with administering nerve blocks. Local injection may also be used in patients who are contraindicated for ISNB such as those with obesity, pulmonary disease, or peripheral neuropathy. Although we cannot definitively suggest that liposomal bupivacaine is superior to the current gold standard ISNB for pain control after shoulder arthroplasty, our results suggest a relative clinical equipoise between these modalities. Larger analytical studies, including randomized trials, should be performed to explore the potential benefits of liposomal bupivacaine injections for pain control after shoulder arthroplasty.

Am J Orthop. 2016;45(7):424-430. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

The annual number of total shoulder arthroplasties (TSAs) is rising with the growing elderly population and development of new technologies such as reverse shoulder arthroplasty.¹ In 2008, 47,000 shoulder arthroplasties were performed in the US compared with 19,000 in 1998.¹ As of 2011, there were 53,000 shoulder arthroplasties performed annually.² Pain control after shoulder procedures, particularly TSA, is challenging. ³

Several modalities exist to manage pain after shoulder arthroplasty. The interscalene brachial plexus nerve block is considered the “gold standard” for shoulder analgesia. A new approach is the periarticular injection method, in which the surgeon administers a local anesthetic intraoperatively. Liposomal bupivacaine (Exparel, Pacira Pharmaceuticals, Inc.) is a nonopioid anesthetic that has been shown to improve pain control, shorten hospital stays, and decrease costs for total knee and hip arthroplasty compared with nerve blocks.^4-6 Patients who were treated with liposomal bupivacaine consumed less opioid medication than a placebo group.⁷

Our purpose was to compare intraoperative local liposomal bupivacaine injection with preoperative single-shot interscalene nerve block (ISNB) in terms of pain control, opioid use, and length of hospital stay (LOS) after shoulder arthroplasty. We hypothesized that patients in the liposomal bupivacaine group would have lower pain scores, less opioid use, and shorter LOS compared with patients in the ISNB group.

Methods

A retrospective cohort analysis was conducted with 58 patients who underwent shoulder arthroplasty by 1 surgeon at our academically affiliated community hospital from January 2012 through January 2015. ISNBs were the standard at the beginning of the study period and were used until Exparel became available on the hospital formulary in 2013. We began using Exparel for all shoulder arthroplasties in November 2013. No other changes were made in the perioperative management of our arthroplasty patients during this period. Patients who underwent TSA, reverse TSA, or hemiarthroplasty of the shoulder were included. Patients who underwent revision TSA were excluded. Twenty-one patients received ISNBs and 37 received liposomal bupivacaine injections. This study was approved by our Institutional Review Board.

Baseline data for each patient were age, sex, body mass index, and the American Society of Anesthesiologists (ASA) Physical Status Classification. The primary outcome measure was the numeric rating scale (NRS) pain score at 4 post-operative time intervals. The NRS pain score has a range of 0 to 10, with 10 representing severe pain. Data were gathered from nursing and physical therapy notes in patient charts. The postoperative time intervals were 0 to 1 hour, 8 to 14 hours, 18 to 24 hours, and 27 to 36 hours. Available NRS scores for these time intervals were averaged. Patients were included if they had pain scores for at least 3 of the postoperative time intervals documented in their charts. Secondary outcome measures were LOS and opioid consumption during hospital admission. Intravenous acetaminophen use was also measured in both groups. All data on opioids were converted to oral morphine equivalents using the method described by Schneider and colleagues.⁸

A board-certified, fellowship-trained anesthesiologist, experienced in regional anesthesia, administered the single-shot ISNB before surgery. The block was administered under ultrasound guidance using a 44-mm, 22-gauge needle with the patient in the supine position. No indwelling catheter was used. The medication consisted of 30 mL of 5% ropivacaine (5 mg/mL). The surgeon injected liposomal bupivacaine (266 mg diluted into 40 mL of injectable saline) near the end of the procedure throughout the pericapsular area and multiple layers of the wound, per manufacturer guidelines.⁹ A 60-mL syringe with a 20-gauge needle was used. All operations were performed by 1 board-certified, fellowship-trained surgeon using a standard deltopectoral approach with the same surgical equipment. The same postoperative pain protocol was used for all patients, including intravenous acetaminophen and patient-controlled analgesia. Additional oral pain medication was provided as needed for all patients. Physical therapy protocols were identical between groups.

Statistical Analysis

Mean patient ages in the 2 treatment groups were compared using the Student t test. Sex distribution and the ASA scores were compared using a χ² test and a Fisher exact test, respectively. Arthroplasty types were compared using a Fisher exact test. The medians and interquartile ranges of the NRS scores at each time point measured were tabulated by treatment group, and at each time point the difference between groups was tested using nonparametric rank sum tests.

We tested the longitudinal trajectory of NRS scores over time, accounting for repeated measurements in the same patients using linear mixed model analysis. Treatment group, time period as a categorical variable, and the interaction between treatment and time period were included as fixed effects, and patient identification number was included as the random effect. An initial omnibus test was performed for all treatment and treatment-by-time interaction effects. Subsequently, the treatment-by-time interaction was tested for each of the time periods. The association of day of discharge (as a categorical variable) with treatment was tested using the Fisher exact test. All analyses were conducted using Stata, version 13, software (StataCorp LP). P values <.05 were considered significant.

 

Sample Size Analysis

We calculated the minimum detectable effect size with 80% power at an alpha level of 0.05 for the nonparametric rank sum test in terms of the proportion of every possible pair of patients treated with the 2 treatments, where the patient treated with liposomal bupivacaine has a lower pain score than the patient treated with ISNB. For pain score at 18 to 24 hours, the sample sizes of 33 patients treated with liposomal bupivacaine and 20 treated with ISNB, the minimum detectable effect size is 73%.

Results

Fifty-eight patient charts (21 in the ISNB group and 37 in the liposomal bupivacaine group) were reviewed for the study. Patient sex distribution, mean age, mean body mass index, and mean baseline ASA scores were not statistically different (Table 1).

In the ISNB group, 5 patients had hemiarthroplasty, 12 had TSA, and 4 had reverse TSA. In the liposomal bupivacaine group, 1 patient had hemiarthroplasty, 23 had TSA, and 13 had reverse TSA. Frequency of procedure types was significantly different between groups (P = .039), with the liposomal bupivacaine group undergoing fewer hemiarthroplasties.

The primary outcome measure, NRS pain score, showed no significant differences between groups at 0 to 1 hour after surgery (P = .99) or 8 to 14 hours after surgery (P = .208).

At 18 to 24 hours after surgery, the liposomal bupivacaine group had a lower mean NRS score than the ISNB group (P = .001). This was statistically significant when taking repeated measures of variance into account (Figure 1). Mean NRS score was also lower for the liposomal bupivacaine group at 27 to 36 hours after surgery (P = .029).

This was a significant difference when repeated measures of variance was considered (Table 2).

There was no difference in the amount of intravenous acetaminophen given during the hospital stay between groups. There was no significant difference in opioid consumption on postoperative day 1 in the hospital (P = .59) (Figure 2). However, there were significant differences between groups on postoperative days 2 and 3.

On postoperative day 2, the ISNB group required significantly more opioids (mean, 112 mg morphine equivalents) than the liposomal bupivacaine group (mean, 37 mg morphine equivalents) (P = .001). The ISNB group also required significantly more opioids (mean, 25 mg morphine equivalents) on postoperative day 3 than the liposomal bupivacaine group (mean, 5 mg) (P = .002).

Sixteen of 37 patients in the liposomal bupivacaine group and 2 of 21 in the ISNB group were discharged on the day after surgery (P = .010) (Table 3).

The mean LOS was 46 ± 20 hours for the liposomal bupivacaine group and 57 ± 14 hours for the ISNB group (P = .012).

There were no major cardiac or respiratory events in either group. No long-term paresthesias or neuropathies were noted. There were no readmissions for either group.

Discussion

Postoperative pain control after shoulder arthroplasty can be challenging, and several modalities have been tried in various combinations to minimize pain and decrease adverse effects of opioid medications. The most common method for pain relief after shoulder arthroplasty is the ISNB. Several studies of ISNBs have shown improved pain control after shoulder arthroplasty with associated decreased opioid consumption and related side effects.¹⁰ Patient rehabilitation and satisfaction have improved with the increasing use of peripheral nerve blocks.¹¹

Despite the well-established benefits of ISNBs, several limitations exist. Although the superior portion of the shoulder is well covered by an ISNB, the inferior portion of the brachial plexus can remain uncovered or only partially covered.¹² Complications of ISNBs include hemidiaphragmatic paresis, rebound pain 24 hours after surgery,¹³ chronic neurologic complications,¹⁴ and substantial respiratory and cardiovascular events.¹⁵ Nerve blocks also require additional time and resources in the perioperative period, including an anesthesiologist with specialized training, assistants, and ultrasonography or nerve stimulation equipment contraindicated in patients taking blood thinners.¹⁶

Periarticular injections of local anesthetics have also shown promise in reducing pain after arthroplasty.⁴ Benefits include an enhanced safety profile because local injection avoids the concurrent blockade of the phrenic nerve and recurrent laryngeal nerve and has not been associated with the risk of peripheral neuropathies. Further, local injection is a simple technique that can be performed during surgery without additional personnel or expertise. A limitation of this approach is the relatively short duration of effectiveness of the local anesthetic and uncertainty regarding the best agent and the ideal volume of injection.⁶ Liposomal bupivacaine is a new agent (approved by the US Food and Drug Administration in 2011¹⁷) with a sustained release over 72 to 96 hours.¹⁸ The most common adverse effects of liposomal bupivacaine are nausea, vomiting, constipation, pyrexia, dizziness, and headache.¹⁹ Chondrotoxicity and granulomatous inflammation are more serious, yet rare, complications of liposomal bupivacaine.²⁰

We found that liposomal bupivacaine injections were associated with lower pain scores compared with ISNB at 18 to 24 hours after surgery. This correlated with less opioid consumption in the liposomal bupivacaine group than in the ISNB group on the second postoperative day. These differences in pain values are consistent with the known pharmacokinetics of liposomal bupivacaine.¹⁸ Peak plasma levels normally occur approximately 24 hours after injection, leaving the early postoperative period relatively uncovered by anesthetic agent. This finding of relatively poor pain control early after surgery has also been noted in patients undergoing knee arthroplasty.⁵ On the basis of the findings of this study, we have added standard bupivacaine injections to our separate liposomal bupivacaine injection to cover early postoperative pain. Opioid consumption was significantly lower in the liposomal bupivacaine group than in the ISNB group on postoperative days 2 and 3. We did not measure adverse events related to opioid consumption, so we cannot comment on whether the decreased opioid consumption was associated with the rate of adverse events. However, other studies^21,22 have established this relationship.

We found the liposomal bupivacaine group to have earlier discharges to home. Sixteen of 37 patients in the liposomal bupivacaine group compared with 2 of 21 patients in the ISNB group were discharged on the day after surgery. A mean reduction in LOS of 18 hours for the liposomal bupivacaine group was statistically significant (P = .012). This reduction in LOS has important implications for hospitals and value analysis committees considering whether to keep a new, more expensive local anesthetic on formulary. Savings from reduced LOS and improvements in patient satisfaction may justify the expense (approximately $300 per 266-mg vial) of Exparel.

From a societal cost perspective, liposomal bupivacaine is more economical compared with ISNB, which adds approximately $1500 to the cost of anesthesia per patient.²³ Eliminating the costs associated with ISNB administration in shoulder arthroplasties could result in substantial savings to our healthcare system. More research examining time savings and exact costs of each procedure is needed to determine the true cost effectiveness of each approach.

Limitations of our study include the retrospective design, relatively small numbers of patients in each group, missing data for some patients at various time points, variation in the types of procedures in each group, and lack of long-term outcome measures. It is important to note that we did not confirm the success of the nerve block after administration. However, this study reflects the effectiveness of each of the modalities in actual clinical conditions (as opposed to a controlled experimental setting). The actual effectiveness of a nerve block varies, even when performed by an experienced anesthesiologist with ultrasound guidance. Furthermore, immediate postoperative pain scores in the nerve block group are consistent with those of prior research reporting pain values ranging from 4 to 5 and a mean duration of effect ranging from 9 to 14 hours.^23,24 Additionally, the patients, surgeon, and nursing team were not blinded to the treatment group. Although we did note a significant difference in the types of procedures between groups, this finding is related to the greater number of hemiarthroplasties performed in the ISNB group (N = 5) compared with the liposomal group (N = 1). Because of this variation and the decreased invasiveness of hemiarthroplasties, the bias is against the liposomal group. Finally, our primary outcome variable was pain, which is a subjective, self-reported measure. However, our opioid consumption data and LOS data corroborate the improved pain scores in the liposomal bupivacaine group.

Limiting the study to a single surgeon may limit external validity. Another limitation is the lack of data on adverse events related to opioid medication use. There was no additional experimental group to determine whether less expensive local anesthetics injected locally would perform similarly to liposomal bupivacaine. In total knee arthroplasty, periarticular injections of liposomal bupivacaine were not as effective as less expensive periarticular injections.²⁵ It is unclear which agents (and in what doses or combinations) should be used for periarticular injections. Finally, we acknowledge that our retrospective study design cannot account for all potential factors affecting discharge time.

This is the first comparative study of liposomal bupivacaine and ISNB in TSA. The study design allowed us to control for variables such as surgical technique, postoperative protocols (including use and type of sling), and use of other pain modalities such as patient-controlled analgesia and intravenous acetaminophen that are likely to affect postoperative pain and LOS. This study provides preliminary data that confirm relative equipoise between liposomal bupivacaine and ISNB, which is needed for the ethical conduct of a randomized controlled trial. Such a trial would allow for a more robust comparison, and this retrospective study provides appropriate pilot data on which to base this design and the clinical information needed to counsel patients during enrollment.

Our results suggest that liposomal bupivacaine may provide superior or similar pain relief compared with ISNB after shoulder arthroplasty. Additionally, the use of liposomal bupivacaine was associated with decreased opioid consumption and earlier discharge to home compared with ISNB. These findings have important implications for pain control after TSA because pain represents a major concern for patients and providers after surgery. In addition to clinical improvements, use of liposomal bupivacaine may save time and eliminate costs associated with administering nerve blocks. Local injection may also be used in patients who are contraindicated for ISNB such as those with obesity, pulmonary disease, or peripheral neuropathy. Although we cannot definitively suggest that liposomal bupivacaine is superior to the current gold standard ISNB for pain control after shoulder arthroplasty, our results suggest a relative clinical equipoise between these modalities. Larger analytical studies, including randomized trials, should be performed to explore the potential benefits of liposomal bupivacaine injections for pain control after shoulder arthroplasty.

Am J Orthop. 2016;45(7):424-430. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

Restoring native kinematics of the knee has been a primary goal of anterior cruciate ligament (ACL) procedures. Double-bundle ACL reconstruction, compared to single-bundle, has been hypothesized to more effectively re-establish rotational stability by re-creating the anatomic ACL, but has not yet proven to result in better clinical outcomes.¹

In 1879, Dr. Paul Segond described a “fibrous, pearly band” at the lateral aspect of the knee that avulsed off the anterolateral proximal tibia during many ACL injuries.² The role of the lateral tissues in knee stability and their relationship with ACL pathology has attracted noteworthy attention in recent time. There have been multiple studies presenting an anatomical description of a structure at the anterolateral portion of the knee with definitive femoral, meniscal, and tibial attachments, which helps control internal rotational forces.^3-7 Claes and colleagues⁴ later found that band of tissue to be the anterolateral ligament (ALL) and determined its injury to be pathognomonic with ACL ruptures.

The ALL is a vital static stabilizer of the tibio-femoral joint, especially during internal tibial rotation.^8-10 In their report on ALL and ACL reconstruction, Helito and colleagues¹¹ acknowledge the necessity of accurate assessment of the lateral structures through imaging to determine the presence of extra-articular injury. Musculoskeletal diagnostic ultrasound has been established as an appropriate means to identify the ALL.¹²

Ultrasound can accurately determine the exact anatomic location of the origin and insertion of the ALL. Reconstruction of the ALL could yield better patient outcomes for those who experience concurrent ACL/ALL injury. Here we present an innovative technique for an ultrasound-guided percutaneous method for reconstruction of the ALL and report on a patient who had underwent ALL reconstruction.

Surgical Indications

All patients undergo an ultrasound evaluation preoperatively to determine if the ALL is intact or injured. Our experience has shown that when ultrasound evaluation reveals an intact ALL, the pivot shift has never been a grade III.

Our indications for a combined ACL and ALL reconstruction are a positive ultrasound diagnosis of an ALL tear, and a grade III pivot shift test in an ACL-deficient knee or a grade I-II pivot shift test in an ACL-intact knee (Table 1). The ACL cannot be left insufficient if the patient is to have a successful ALL reconstruction.

Surgical Technique

For a demonstration of this technique, see the video that accompanies this article.

The pivot shift test is conducted under anesthesia to determine whether an ALL reconstruction is required. The patient is placed in a supine position with the knee flexed at 30^o, at neutral rotation, and without any varus or valgus stress.

The knee is prepped and draped under sterile conditions. Under ultrasound guidance, the origin and insertion of the ALL are identified and marked with an 18-gauge spinal needle (Figure 1).

A No. 15 blade is used to make a small incision centered on each spinal needle. The spinal needle is replaced with a 2.4-mm drill pin (Figure 2).

A 90^o hemostat is used to establish a plane under the iliotibial (IT) band between the 2 incisions to pass a looped FiberWire suture (Arthrex) for passage of the graft and FiberTape (Arthrex). The FiberTape acts as an internal brace. A socket 22 mm in length is drilled using a 5.0-mm cannulated reamer over each drill pin.

A 4.5-mm semitendinosus graft was prepared with a collagen-coated FiberTape attached to a 5.5-mm BioComposite Vented SwiveLock anchor (Arthrex) and fixed into the femoral socket (Figure 3).

The graft and FiberTape are then passed under the IT band to the distal incision. Using the length of the BioComposite SwiveLock anchor as a guide, a mark is made on the graft after tensioning the construct in line with the leg, distal to the tibial drill pin (Table 2, Figure 4).

The second 5.5-mm BioComposite SwiveLock anchor is attached to the FiberTape and graft at the mark.

The rest of the graft is resected. The BioComposite SwiveLock anchor, graft, and FiberTape are fixed into the tibial socket, completing the reconstruction (Figure 5). Passive range of motion should then be checked to ensure the construct is not overtensioned.

Rehabilitation

Rehabilitation following an ALL procedure is similar to traditional ACL rehabilitation with an added emphasis on minimizing rotational torque of the tibia in the early stages.

Our protocol allows for early weight-bearing and minimal use of assistive devices (ie, immobilizer brace and crutches) because an internal brace is performed on every ALL reconstruction. The protocol emphasizes full range of motion and linear power with a progression to lateral and rotational activities. This enables the client to begin rehabilitation within 1 week and regain normal daily function quickly. Return to heavy lifting, physical activity, and sports is delayed until after 6 months to allow for the graft maturity and integration, which takes quite a while, as grafts are weakest after 6 weeks.¹³ When patients return to sports and activity, a brace is used for up to 1 year postoperatively to limit shearing forces inherent in pivoting and cutting.

 

Case Report

In January 2013, a 17-year-old male soccer player suffered an ACL rupture of his right knee. Later that spring, he had an ACL reconstruction with an allograft. Twelve months postoperatively, the patient returned, saying that he felt much better; however, anytime he tried to plant his foot and rotate over that fixed foot, his knee felt unstable. The physical examination revealed both negative Lachman and anterior drawer tests but a I+ pivot shift test. A magnetic resonance imaging (MRI) examination revealed an intact ACL graft. A diagnostic ultrasound evaluation revealed a distal ALL injury. After discussing the risks, benefits, and goals with the patient, we opted for a diagnostic arthroscopy and a percutaneous, ultrasound-guided reconstruction of the ALL.

Postoperatively, the patient did very well. One week after surgery, he returned, saying he felt completely stable and demonstrated by repeating the rotation of his knee. The patient continued to have no issues until he returned 13 months post-ALL surgery, complaining of a recent injury that had caused the return of his feelings of instability. An MRI evaluation showed an intact ACL graft and the possibility of a ruptured ALL. Fifteen months after the initial ALL reconstruction, we proceeded with surgery. At arthroscopy, the patient was found to have a pivot shift of I+ and an intact ACL graft. The ALL was reconstructed again using an allograft, internal brace, and bone marrow concentrate. At 13 months post-ALL reconstruction revision, the patient had no complaints.

Discussion

Reconstruction of the ALL is aimed to restore anatomic rotational kinematics. Sonnery-Cottet and colleagues¹⁴ have reported promising initial results in their 2-year follow-up study of combined ACL and ALL reconstruction outcomes. This surgical technique includes use of an internal brace, which negates the necessity for external support devices and allows for earlier mobilization of the joint. A reconstruction of the ALL, performed concurrently with the ACL, does not add recovery time, but could prevent postsurgical complications and improve rehabilitation by eliminating rotational instability that presents in some ACL-reconstructed patients.

Sonnery-Cottet and colleagues¹⁵ state that their arthroscopic identification of the ALL can help to cultivate a “less invasive and more anatomic” reconstruction. The use of musculoskeletal ultrasound allows our technique to utilize a completely noninvasive imaging tool that allows proper establishment of ALL anatomy prior to the procedure. The entirety of the ALL is easily identifiable,^4,12 which has proven to be shortcoming of MRI evaluation.^15-17 Accurate preoperative assessment of the lateral structures is necessary in ACL-deficient individuals.^11,15 Sonography also provides a means of accurate guidance and socket creation, without generating large incisions.

If the ALL is responsible for internal rotatory stability as asserted, the structure should exhibit biomechanical properties during movement. In their study on the function of the ligament, Parsons and colleagues⁹ established the inverse relationship between the ALL and ACL during internal rotation. As their cadaveric knees were subjected to an internal rotatory force through increasing angles of flexion, the contribution of the ALL towards stability significantly increased while the ACL declined. Helito and colleagues⁸ and Zens and colleagues¹⁰ have demonstrated length changes of the ligament through varying degrees of flexion and internal rotation. Their reports indicate greater tension during knee movements, coinciding with the description of increasing ALL stability contribution by Parsons and colleagues.⁹ Kennedy and colleagues⁷ conducted a pull-to-failure test on the ALL. The average failure load was 175 N with a stiffness of 20 N/mm, illustrating the structure is a candidate for most traditional soft tissue grafts. The biomechanical evidence of the structural properties of the ALL confirms its importance in knee function and the necessity for its reconstruction.

With the understanding that ACL contributes to rotatory stability to some extent, the notion begs the question of how a centrally located ligament is able to prevent excessive rotation in a structure with a large relative radius. Biomechanically, with such a small moment arm, the ACL would experience tremendous stress when a rotatory force is applied. The same torque applied to a more superficial structure, with a greater moment, would sustain a large reduction in the applied force. The concept of a wheel and an axle should be considered. The equation is F₁ × R₁ = F₂ × R₂. We measured on a cadaveric knee the distance from the center of rotation to the ACL and the ALL, finding the radii were 5 mm and 30 mm, respectively. Taking these measurements, we would then expect the force experienced on the axle (ACL) to be 6 times greater than what would be experienced on the periphery of the wheel (ALL). The ALL (wheel) has a significant biomechanical advantage over the ACL (axle) in controlling and enduring internal rotatory forces of the knee. This would imply that if the ALL were damaged and not re-established, the ACL would experience a 6 times greater force trying to control internal rotation, which would result in a significantly increased chance of failure and rupture.

While there is a degree of dissent on the presence of the ALL, a number of studies have classified the tissue as an independent ligamentous structure.^3-7 While there is disagreement on the precise location of the femoral attachment, there is a consensus on the location of the tibial and meniscal attachments. Claes and colleagues⁴ originally outlined the femoral attachment as anterior and distal to the origin of the fibular collateral ligament (FCL), which is the description this technique follows. Since Claes and colleagues’⁴ report, many have investigated the ligament’s femoral origin with delineations ranging from posterior and proximal^3,5,7 to anterior and distal.^6,16-18

The accurate, noninvasive nature of the musculoskeletal ultrasound prior to any incisions being made makes this technique innovative and superior to other open surgical techniques or those that require fluoroscopy.

This is the greatest advantage of the procedure (Table 3). Not only does the use of ultrasound make this specific operation exceptional, but its practice is widely applicable. To date, this is the only ultrasound-guided reconstruction of any kind and can serve as a template for not only ALL procedures, but many other procedures as well.

 

Conclusion

The ALL has been determined to play an integral role in the rotational stability of the knee. In the setting of instability and insufficiency, reconstruction will lead to better patient outcomes for concurrent ACL/ALL injuries and postsurgical rotatory instability following ACL procedures. This innovative technique utilizes ultrasound to ascertain the precise anatomical attachments of the ALL prior to the operation. The novel nature of this ultrasound-guided reconstruction has the potential to be applicable in many other surgical procedures.

Restoring native kinematics of the knee has been a primary goal of anterior cruciate ligament (ACL) procedures. Double-bundle ACL reconstruction, compared to single-bundle, has been hypothesized to more effectively re-establish rotational stability by re-creating the anatomic ACL, but has not yet proven to result in better clinical outcomes.¹

In 1879, Dr. Paul Segond described a “fibrous, pearly band” at the lateral aspect of the knee that avulsed off the anterolateral proximal tibia during many ACL injuries.² The role of the lateral tissues in knee stability and their relationship with ACL pathology has attracted noteworthy attention in recent time. There have been multiple studies presenting an anatomical description of a structure at the anterolateral portion of the knee with definitive femoral, meniscal, and tibial attachments, which helps control internal rotational forces.^3-7 Claes and colleagues⁴ later found that band of tissue to be the anterolateral ligament (ALL) and determined its injury to be pathognomonic with ACL ruptures.

The ALL is a vital static stabilizer of the tibio-femoral joint, especially during internal tibial rotation.^8-10 In their report on ALL and ACL reconstruction, Helito and colleagues¹¹ acknowledge the necessity of accurate assessment of the lateral structures through imaging to determine the presence of extra-articular injury. Musculoskeletal diagnostic ultrasound has been established as an appropriate means to identify the ALL.¹²

Ultrasound can accurately determine the exact anatomic location of the origin and insertion of the ALL. Reconstruction of the ALL could yield better patient outcomes for those who experience concurrent ACL/ALL injury. Here we present an innovative technique for an ultrasound-guided percutaneous method for reconstruction of the ALL and report on a patient who had underwent ALL reconstruction.

Surgical Indications

All patients undergo an ultrasound evaluation preoperatively to determine if the ALL is intact or injured. Our experience has shown that when ultrasound evaluation reveals an intact ALL, the pivot shift has never been a grade III.

Our indications for a combined ACL and ALL reconstruction are a positive ultrasound diagnosis of an ALL tear, and a grade III pivot shift test in an ACL-deficient knee or a grade I-II pivot shift test in an ACL-intact knee (Table 1). The ACL cannot be left insufficient if the patient is to have a successful ALL reconstruction.

Surgical Technique

For a demonstration of this technique, see the video that accompanies this article.

The pivot shift test is conducted under anesthesia to determine whether an ALL reconstruction is required. The patient is placed in a supine position with the knee flexed at 30^o, at neutral rotation, and without any varus or valgus stress.

The knee is prepped and draped under sterile conditions. Under ultrasound guidance, the origin and insertion of the ALL are identified and marked with an 18-gauge spinal needle (Figure 1).

A No. 15 blade is used to make a small incision centered on each spinal needle. The spinal needle is replaced with a 2.4-mm drill pin (Figure 2).

A 90^o hemostat is used to establish a plane under the iliotibial (IT) band between the 2 incisions to pass a looped FiberWire suture (Arthrex) for passage of the graft and FiberTape (Arthrex). The FiberTape acts as an internal brace. A socket 22 mm in length is drilled using a 5.0-mm cannulated reamer over each drill pin.

A 4.5-mm semitendinosus graft was prepared with a collagen-coated FiberTape attached to a 5.5-mm BioComposite Vented SwiveLock anchor (Arthrex) and fixed into the femoral socket (Figure 3).

The graft and FiberTape are then passed under the IT band to the distal incision. Using the length of the BioComposite SwiveLock anchor as a guide, a mark is made on the graft after tensioning the construct in line with the leg, distal to the tibial drill pin (Table 2, Figure 4).

The second 5.5-mm BioComposite SwiveLock anchor is attached to the FiberTape and graft at the mark.

The rest of the graft is resected. The BioComposite SwiveLock anchor, graft, and FiberTape are fixed into the tibial socket, completing the reconstruction (Figure 5). Passive range of motion should then be checked to ensure the construct is not overtensioned.

Rehabilitation

Rehabilitation following an ALL procedure is similar to traditional ACL rehabilitation with an added emphasis on minimizing rotational torque of the tibia in the early stages.

Our protocol allows for early weight-bearing and minimal use of assistive devices (ie, immobilizer brace and crutches) because an internal brace is performed on every ALL reconstruction. The protocol emphasizes full range of motion and linear power with a progression to lateral and rotational activities. This enables the client to begin rehabilitation within 1 week and regain normal daily function quickly. Return to heavy lifting, physical activity, and sports is delayed until after 6 months to allow for the graft maturity and integration, which takes quite a while, as grafts are weakest after 6 weeks.¹³ When patients return to sports and activity, a brace is used for up to 1 year postoperatively to limit shearing forces inherent in pivoting and cutting.

 

Case Report

In January 2013, a 17-year-old male soccer player suffered an ACL rupture of his right knee. Later that spring, he had an ACL reconstruction with an allograft. Twelve months postoperatively, the patient returned, saying that he felt much better; however, anytime he tried to plant his foot and rotate over that fixed foot, his knee felt unstable. The physical examination revealed both negative Lachman and anterior drawer tests but a I+ pivot shift test. A magnetic resonance imaging (MRI) examination revealed an intact ACL graft. A diagnostic ultrasound evaluation revealed a distal ALL injury. After discussing the risks, benefits, and goals with the patient, we opted for a diagnostic arthroscopy and a percutaneous, ultrasound-guided reconstruction of the ALL.

Postoperatively, the patient did very well. One week after surgery, he returned, saying he felt completely stable and demonstrated by repeating the rotation of his knee. The patient continued to have no issues until he returned 13 months post-ALL surgery, complaining of a recent injury that had caused the return of his feelings of instability. An MRI evaluation showed an intact ACL graft and the possibility of a ruptured ALL. Fifteen months after the initial ALL reconstruction, we proceeded with surgery. At arthroscopy, the patient was found to have a pivot shift of I+ and an intact ACL graft. The ALL was reconstructed again using an allograft, internal brace, and bone marrow concentrate. At 13 months post-ALL reconstruction revision, the patient had no complaints.

Discussion

Reconstruction of the ALL is aimed to restore anatomic rotational kinematics. Sonnery-Cottet and colleagues¹⁴ have reported promising initial results in their 2-year follow-up study of combined ACL and ALL reconstruction outcomes. This surgical technique includes use of an internal brace, which negates the necessity for external support devices and allows for earlier mobilization of the joint. A reconstruction of the ALL, performed concurrently with the ACL, does not add recovery time, but could prevent postsurgical complications and improve rehabilitation by eliminating rotational instability that presents in some ACL-reconstructed patients.

Sonnery-Cottet and colleagues¹⁵ state that their arthroscopic identification of the ALL can help to cultivate a “less invasive and more anatomic” reconstruction. The use of musculoskeletal ultrasound allows our technique to utilize a completely noninvasive imaging tool that allows proper establishment of ALL anatomy prior to the procedure. The entirety of the ALL is easily identifiable,^4,12 which has proven to be shortcoming of MRI evaluation.^15-17 Accurate preoperative assessment of the lateral structures is necessary in ACL-deficient individuals.^11,15 Sonography also provides a means of accurate guidance and socket creation, without generating large incisions.

If the ALL is responsible for internal rotatory stability as asserted, the structure should exhibit biomechanical properties during movement. In their study on the function of the ligament, Parsons and colleagues⁹ established the inverse relationship between the ALL and ACL during internal rotation. As their cadaveric knees were subjected to an internal rotatory force through increasing angles of flexion, the contribution of the ALL towards stability significantly increased while the ACL declined. Helito and colleagues⁸ and Zens and colleagues¹⁰ have demonstrated length changes of the ligament through varying degrees of flexion and internal rotation. Their reports indicate greater tension during knee movements, coinciding with the description of increasing ALL stability contribution by Parsons and colleagues.⁹ Kennedy and colleagues⁷ conducted a pull-to-failure test on the ALL. The average failure load was 175 N with a stiffness of 20 N/mm, illustrating the structure is a candidate for most traditional soft tissue grafts. The biomechanical evidence of the structural properties of the ALL confirms its importance in knee function and the necessity for its reconstruction.

With the understanding that ACL contributes to rotatory stability to some extent, the notion begs the question of how a centrally located ligament is able to prevent excessive rotation in a structure with a large relative radius. Biomechanically, with such a small moment arm, the ACL would experience tremendous stress when a rotatory force is applied. The same torque applied to a more superficial structure, with a greater moment, would sustain a large reduction in the applied force. The concept of a wheel and an axle should be considered. The equation is F₁ × R₁ = F₂ × R₂. We measured on a cadaveric knee the distance from the center of rotation to the ACL and the ALL, finding the radii were 5 mm and 30 mm, respectively. Taking these measurements, we would then expect the force experienced on the axle (ACL) to be 6 times greater than what would be experienced on the periphery of the wheel (ALL). The ALL (wheel) has a significant biomechanical advantage over the ACL (axle) in controlling and enduring internal rotatory forces of the knee. This would imply that if the ALL were damaged and not re-established, the ACL would experience a 6 times greater force trying to control internal rotation, which would result in a significantly increased chance of failure and rupture.

While there is a degree of dissent on the presence of the ALL, a number of studies have classified the tissue as an independent ligamentous structure.^3-7 While there is disagreement on the precise location of the femoral attachment, there is a consensus on the location of the tibial and meniscal attachments. Claes and colleagues⁴ originally outlined the femoral attachment as anterior and distal to the origin of the fibular collateral ligament (FCL), which is the description this technique follows. Since Claes and colleagues’⁴ report, many have investigated the ligament’s femoral origin with delineations ranging from posterior and proximal^3,5,7 to anterior and distal.^6,16-18

The accurate, noninvasive nature of the musculoskeletal ultrasound prior to any incisions being made makes this technique innovative and superior to other open surgical techniques or those that require fluoroscopy.

This is the greatest advantage of the procedure (Table 3). Not only does the use of ultrasound make this specific operation exceptional, but its practice is widely applicable. To date, this is the only ultrasound-guided reconstruction of any kind and can serve as a template for not only ALL procedures, but many other procedures as well.

 

Conclusion

The ALL has been determined to play an integral role in the rotational stability of the knee. In the setting of instability and insufficiency, reconstruction will lead to better patient outcomes for concurrent ACL/ALL injuries and postsurgical rotatory instability following ACL procedures. This innovative technique utilizes ultrasound to ascertain the precise anatomical attachments of the ALL prior to the operation. The novel nature of this ultrasound-guided reconstruction has the potential to be applicable in many other surgical procedures.

Restoring native kinematics of the knee has been a primary goal of anterior cruciate ligament (ACL) procedures. Double-bundle ACL reconstruction, compared to single-bundle, has been hypothesized to more effectively re-establish rotational stability by re-creating the anatomic ACL, but has not yet proven to result in better clinical outcomes.¹

In 1879, Dr. Paul Segond described a “fibrous, pearly band” at the lateral aspect of the knee that avulsed off the anterolateral proximal tibia during many ACL injuries.² The role of the lateral tissues in knee stability and their relationship with ACL pathology has attracted noteworthy attention in recent time. There have been multiple studies presenting an anatomical description of a structure at the anterolateral portion of the knee with definitive femoral, meniscal, and tibial attachments, which helps control internal rotational forces.^3-7 Claes and colleagues⁴ later found that band of tissue to be the anterolateral ligament (ALL) and determined its injury to be pathognomonic with ACL ruptures.

The ALL is a vital static stabilizer of the tibio-femoral joint, especially during internal tibial rotation.^8-10 In their report on ALL and ACL reconstruction, Helito and colleagues¹¹ acknowledge the necessity of accurate assessment of the lateral structures through imaging to determine the presence of extra-articular injury. Musculoskeletal diagnostic ultrasound has been established as an appropriate means to identify the ALL.¹²

Ultrasound can accurately determine the exact anatomic location of the origin and insertion of the ALL. Reconstruction of the ALL could yield better patient outcomes for those who experience concurrent ACL/ALL injury. Here we present an innovative technique for an ultrasound-guided percutaneous method for reconstruction of the ALL and report on a patient who had underwent ALL reconstruction.

Surgical Indications

All patients undergo an ultrasound evaluation preoperatively to determine if the ALL is intact or injured. Our experience has shown that when ultrasound evaluation reveals an intact ALL, the pivot shift has never been a grade III.

Our indications for a combined ACL and ALL reconstruction are a positive ultrasound diagnosis of an ALL tear, and a grade III pivot shift test in an ACL-deficient knee or a grade I-II pivot shift test in an ACL-intact knee (Table 1). The ACL cannot be left insufficient if the patient is to have a successful ALL reconstruction.

Surgical Technique

For a demonstration of this technique, see the video that accompanies this article.

The pivot shift test is conducted under anesthesia to determine whether an ALL reconstruction is required. The patient is placed in a supine position with the knee flexed at 30^o, at neutral rotation, and without any varus or valgus stress.

The knee is prepped and draped under sterile conditions. Under ultrasound guidance, the origin and insertion of the ALL are identified and marked with an 18-gauge spinal needle (Figure 1).

A No. 15 blade is used to make a small incision centered on each spinal needle. The spinal needle is replaced with a 2.4-mm drill pin (Figure 2).

A 90^o hemostat is used to establish a plane under the iliotibial (IT) band between the 2 incisions to pass a looped FiberWire suture (Arthrex) for passage of the graft and FiberTape (Arthrex). The FiberTape acts as an internal brace. A socket 22 mm in length is drilled using a 5.0-mm cannulated reamer over each drill pin.

A 4.5-mm semitendinosus graft was prepared with a collagen-coated FiberTape attached to a 5.5-mm BioComposite Vented SwiveLock anchor (Arthrex) and fixed into the femoral socket (Figure 3).

The graft and FiberTape are then passed under the IT band to the distal incision. Using the length of the BioComposite SwiveLock anchor as a guide, a mark is made on the graft after tensioning the construct in line with the leg, distal to the tibial drill pin (Table 2, Figure 4).

The second 5.5-mm BioComposite SwiveLock anchor is attached to the FiberTape and graft at the mark.

The rest of the graft is resected. The BioComposite SwiveLock anchor, graft, and FiberTape are fixed into the tibial socket, completing the reconstruction (Figure 5). Passive range of motion should then be checked to ensure the construct is not overtensioned.

Rehabilitation

Rehabilitation following an ALL procedure is similar to traditional ACL rehabilitation with an added emphasis on minimizing rotational torque of the tibia in the early stages.

Our protocol allows for early weight-bearing and minimal use of assistive devices (ie, immobilizer brace and crutches) because an internal brace is performed on every ALL reconstruction. The protocol emphasizes full range of motion and linear power with a progression to lateral and rotational activities. This enables the client to begin rehabilitation within 1 week and regain normal daily function quickly. Return to heavy lifting, physical activity, and sports is delayed until after 6 months to allow for the graft maturity and integration, which takes quite a while, as grafts are weakest after 6 weeks.¹³ When patients return to sports and activity, a brace is used for up to 1 year postoperatively to limit shearing forces inherent in pivoting and cutting.

 

Case Report

In January 2013, a 17-year-old male soccer player suffered an ACL rupture of his right knee. Later that spring, he had an ACL reconstruction with an allograft. Twelve months postoperatively, the patient returned, saying that he felt much better; however, anytime he tried to plant his foot and rotate over that fixed foot, his knee felt unstable. The physical examination revealed both negative Lachman and anterior drawer tests but a I+ pivot shift test. A magnetic resonance imaging (MRI) examination revealed an intact ACL graft. A diagnostic ultrasound evaluation revealed a distal ALL injury. After discussing the risks, benefits, and goals with the patient, we opted for a diagnostic arthroscopy and a percutaneous, ultrasound-guided reconstruction of the ALL.

Postoperatively, the patient did very well. One week after surgery, he returned, saying he felt completely stable and demonstrated by repeating the rotation of his knee. The patient continued to have no issues until he returned 13 months post-ALL surgery, complaining of a recent injury that had caused the return of his feelings of instability. An MRI evaluation showed an intact ACL graft and the possibility of a ruptured ALL. Fifteen months after the initial ALL reconstruction, we proceeded with surgery. At arthroscopy, the patient was found to have a pivot shift of I+ and an intact ACL graft. The ALL was reconstructed again using an allograft, internal brace, and bone marrow concentrate. At 13 months post-ALL reconstruction revision, the patient had no complaints.

Discussion

Reconstruction of the ALL is aimed to restore anatomic rotational kinematics. Sonnery-Cottet and colleagues¹⁴ have reported promising initial results in their 2-year follow-up study of combined ACL and ALL reconstruction outcomes. This surgical technique includes use of an internal brace, which negates the necessity for external support devices and allows for earlier mobilization of the joint. A reconstruction of the ALL, performed concurrently with the ACL, does not add recovery time, but could prevent postsurgical complications and improve rehabilitation by eliminating rotational instability that presents in some ACL-reconstructed patients.

Sonnery-Cottet and colleagues¹⁵ state that their arthroscopic identification of the ALL can help to cultivate a “less invasive and more anatomic” reconstruction. The use of musculoskeletal ultrasound allows our technique to utilize a completely noninvasive imaging tool that allows proper establishment of ALL anatomy prior to the procedure. The entirety of the ALL is easily identifiable,^4,12 which has proven to be shortcoming of MRI evaluation.^15-17 Accurate preoperative assessment of the lateral structures is necessary in ACL-deficient individuals.^11,15 Sonography also provides a means of accurate guidance and socket creation, without generating large incisions.

If the ALL is responsible for internal rotatory stability as asserted, the structure should exhibit biomechanical properties during movement. In their study on the function of the ligament, Parsons and colleagues⁹ established the inverse relationship between the ALL and ACL during internal rotation. As their cadaveric knees were subjected to an internal rotatory force through increasing angles of flexion, the contribution of the ALL towards stability significantly increased while the ACL declined. Helito and colleagues⁸ and Zens and colleagues¹⁰ have demonstrated length changes of the ligament through varying degrees of flexion and internal rotation. Their reports indicate greater tension during knee movements, coinciding with the description of increasing ALL stability contribution by Parsons and colleagues.⁹ Kennedy and colleagues⁷ conducted a pull-to-failure test on the ALL. The average failure load was 175 N with a stiffness of 20 N/mm, illustrating the structure is a candidate for most traditional soft tissue grafts. The biomechanical evidence of the structural properties of the ALL confirms its importance in knee function and the necessity for its reconstruction.

With the understanding that ACL contributes to rotatory stability to some extent, the notion begs the question of how a centrally located ligament is able to prevent excessive rotation in a structure with a large relative radius. Biomechanically, with such a small moment arm, the ACL would experience tremendous stress when a rotatory force is applied. The same torque applied to a more superficial structure, with a greater moment, would sustain a large reduction in the applied force. The concept of a wheel and an axle should be considered. The equation is F₁ × R₁ = F₂ × R₂. We measured on a cadaveric knee the distance from the center of rotation to the ACL and the ALL, finding the radii were 5 mm and 30 mm, respectively. Taking these measurements, we would then expect the force experienced on the axle (ACL) to be 6 times greater than what would be experienced on the periphery of the wheel (ALL). The ALL (wheel) has a significant biomechanical advantage over the ACL (axle) in controlling and enduring internal rotatory forces of the knee. This would imply that if the ALL were damaged and not re-established, the ACL would experience a 6 times greater force trying to control internal rotation, which would result in a significantly increased chance of failure and rupture.

While there is a degree of dissent on the presence of the ALL, a number of studies have classified the tissue as an independent ligamentous structure.^3-7 While there is disagreement on the precise location of the femoral attachment, there is a consensus on the location of the tibial and meniscal attachments. Claes and colleagues⁴ originally outlined the femoral attachment as anterior and distal to the origin of the fibular collateral ligament (FCL), which is the description this technique follows. Since Claes and colleagues’⁴ report, many have investigated the ligament’s femoral origin with delineations ranging from posterior and proximal^3,5,7 to anterior and distal.^6,16-18

The accurate, noninvasive nature of the musculoskeletal ultrasound prior to any incisions being made makes this technique innovative and superior to other open surgical techniques or those that require fluoroscopy.

This is the greatest advantage of the procedure (Table 3). Not only does the use of ultrasound make this specific operation exceptional, but its practice is widely applicable. To date, this is the only ultrasound-guided reconstruction of any kind and can serve as a template for not only ALL procedures, but many other procedures as well.

 

Conclusion

The ALL has been determined to play an integral role in the rotational stability of the knee. In the setting of instability and insufficiency, reconstruction will lead to better patient outcomes for concurrent ACL/ALL injuries and postsurgical rotatory instability following ACL procedures. This innovative technique utilizes ultrasound to ascertain the precise anatomical attachments of the ALL prior to the operation. The novel nature of this ultrasound-guided reconstruction has the potential to be applicable in many other surgical procedures.