Melamed E

Flexor tendons are considered the strongest component of the musculotendinous unit; they generally do not rupture unless weakened by an underlying pathologic condition.¹ According to traditional teaching, when the musculotendinous unit is subjected to excessive forces, failure invariably occurs at the tendon insertion, at the musculotendinous junction, within the muscle substance, or at its origin from the bone before the tendon itself ruptures.¹

Midsubstance tears in nonrheumatoid patients are less frequent and are typically attributable to an underlying cause.² Possible pathologic conditions include, but are not limited to, osteoarthritis of the pisotriquetral joint,³ nonunion fracture of the hook of the hamate,⁴ lunate dislocation,⁵ accessory carpal bone,⁶ gouty infiltration of the flexor tendon,⁷ and tumor.⁸ In 1960, Boyes and colleagues⁹ presented a series of 80 flexor tendon ruptures in 78 patients over a 13-year period. Only 3 cases had no identifiable cause. The authors recommended using the term spontaneous for those ruptures that occur within the tendon substance without underlying or associated pathologic changes.

We describe a patient with spontaneous rupture of the flexor digitorum profundus (FDP) tendon at zone III, satisfying Boyes’ definition of the term spontaneous. The patient provided written informed consent for print and electronic publication of this case report.

Case Report

A 65-year-old, right-handed manual worker was assessed in our hand clinic 3 days after he felt a cramp in his left palm while lifting a heavy object. Shortly thereafter, he noted he could not flex his ring finger distal interphalangeal (DIP) joint. He could not recall any previous injury to his finger. No predisposing pathologic conditions or bone abnormalities were identified. Clinically, there was no tenderness, swelling, or ecchymosis evident. He had full passive range of motion (ROM) of his ring finger, and proximal interphalangeal (PIP) joint active ROM was 0/110º; however, he had no activity of the FDP of the ring finger. Preoperative radiographs were normal. The hook of the hamate was clinically and radiographically normal.

A preoperative diagnosis of FDP avulsion from the distal phalanx was made, and the operation was carried out 16 days after injury. Surgical exploration started in zone II and extended proximally into the distal palmar crease, but no stump was found in either location. Therefore, exploration was carried out to the midpalmar region, revealing the tendon rupture in zone III, in the region of the origin of the ring finger lumbrical muscle (Figure 1). The flexor digitorum superficialis tendon was intact. Macroscopically, both tendon and carpal tunnel appeared normal, with no evidence of tendon attrition; thus, the tendon was not sent for histologic examination. The ends of the ruptured FDP tendon to the ring finger were at the level of the superficial palmar arch, with the distal end appearing as though it had been cut sharply with a knife. Because of the short period of time from injury to exploration, delayed primary tendon repair was possible, along with side-to-side tenodesis to the intact ring finger flexor superficialis tendon in the palm (Figure 2). Two days after surgery, the patient started a controlled mobilization program using the Duran method.¹⁰

At final follow-up of 18 months, total active motion was 126°, which corresponds to a good outcome, according to the Strickland and Glogovac criteria.¹¹ Grip strength was 50 kg, which was 84% of grip strength on the uninjured side. The patient was back to recreational activity but had not returned to work.

Discussion

Most flexor tendon ruptures result from avulsion of the FDP tendon at its distal phalanx insertion, commonly known as Jersey finger. However, true midsubstance spontaneous ruptures are infrequent. Reports of spontaneous tendon ruptures of all types, including those of the hand, have increased in incidence in most countries.¹² Bois and colleagues,¹³ who have reviewed the literature over a 50-year period, found a total of 50 spontaneous ruptures of “normal” flexor tendon in 43 cases. The authors point to unique historical and physical examinaton findings that help differentiate spontaneous tendon ruptures from the more common FDP avulsions. Such findings include the sensation of a pop or snap, or a sudden sharp pain or cramp within the palmar region. In contrast, most avulsion ruptures cause discomfort within the region of the digit. In type I avulsion injuries of the FDP tendon, the proximal tendon stump usually retracts proximal to the digital tendon sheath, causing a tender mass in the palm.¹⁴ Flexor digitorum profundus tendon avulsions, however, are not typically associated with a snap or pop in the palm. When spontaneous ruptures of the hand occur, they typically involve the profundus tendon of the small finger, in the area of the lumbrical origin.¹³

In equivocal cases when the site of rupture is uncertain, ultrasound and magnetic resonance imaging may assist in making the diagnosis and provide important preoperative information for surgical decision-making and planning; this information may decrease postoperative morbidity by minimizing surgical dissection.

The etiology of spontaneous ruptures is incompletely understood. For any rupture of the ulnar flexor tendons, the hook of the hamate should be examined to rule out a previous fracture as a cause of tendon attrition.¹⁵ Tendon vascularization may be a cause for tendon rupture in the hand. When the blood supply of the lumbrical muscles was examined in 100 upper extremities from human cadavers using vascular injection studies,¹⁶ it was discovered that each lumbrical muscle received its arterial supply from 4 sources: the superficial palmar arch, the common palmar digital artery, the deep palmar arch, and the dorsal digital artery. There were no anastomoses between the networks supplying the lumbrical muscles and the FDP tendons within the palm, suggesting a possible watershed zone between the FDP tendon and lumbrical muscle origin. The patient described in this case had the tendon rupture in the area of potential hypovascularity at the lumbrical origin.

Important factors in the decision-making process for surgical treatment include the length of time between rupture and treatment, the site of rupture, and the condition of the ruptured tendon ends. Patients who present in the first 3 weeks of injury can be treated by primary tendon repair, provided that the ruptured tendon ends are not significantly frayed or attenuated. For patients presenting more than 3 weeks after injury, interposition tendon grafts or tendon transfers are suitable options for ruptures in zone III. Distal interphalangeal joint arthrodesis is another alternative in specific cases where reconstruction is not possible. In this case, direct end-to-end repair was possible, as well as tenodesis to the intact ring finger superficialis in order to prevent stretching of the repair.

Localizing the level of the tendon rupture clinically is difficult. When the site of the profundus tendon rupture is uncertain, and there is no tenderness in zone I or the PIP joint, the first incision should be made at the metacarpophalangeal joint level. This first incision will indicate if the rupture occurred in zone III. If the tendon is intact at that location, then the next incision should be at the level of the PIP joint.

Conclusion

We report a patient treated for spontaneous rupture of the flexor tendon in zone III. He was treated in the acute setting with direct tendon repair. It is important to consider spontaneous rupture of the tendon in patients presenting with a snap/pop and the sudden inability to flex a finger. A tendon rupture can be diagnosed as spontaneous in the absence of an underlying pathologic condition such as rheumatoid arthritis, gout, or occult carpal fractures. In the acute setting, these may be repaired primarily; however, if presenting after a few weeks, alternative surgical options, including interposition tendon grafts, tendon transfer, and DIP joint arthrodesis, should be considered.

Flexor tendons are considered the strongest component of the musculotendinous unit; they generally do not rupture unless weakened by an underlying pathologic condition.¹ According to traditional teaching, when the musculotendinous unit is subjected to excessive forces, failure invariably occurs at the tendon insertion, at the musculotendinous junction, within the muscle substance, or at its origin from the bone before the tendon itself ruptures.¹

Midsubstance tears in nonrheumatoid patients are less frequent and are typically attributable to an underlying cause.² Possible pathologic conditions include, but are not limited to, osteoarthritis of the pisotriquetral joint,³ nonunion fracture of the hook of the hamate,⁴ lunate dislocation,⁵ accessory carpal bone,⁶ gouty infiltration of the flexor tendon,⁷ and tumor.⁸ In 1960, Boyes and colleagues⁹ presented a series of 80 flexor tendon ruptures in 78 patients over a 13-year period. Only 3 cases had no identifiable cause. The authors recommended using the term spontaneous for those ruptures that occur within the tendon substance without underlying or associated pathologic changes.

We describe a patient with spontaneous rupture of the flexor digitorum profundus (FDP) tendon at zone III, satisfying Boyes’ definition of the term spontaneous. The patient provided written informed consent for print and electronic publication of this case report.

Case Report

A 65-year-old, right-handed manual worker was assessed in our hand clinic 3 days after he felt a cramp in his left palm while lifting a heavy object. Shortly thereafter, he noted he could not flex his ring finger distal interphalangeal (DIP) joint. He could not recall any previous injury to his finger. No predisposing pathologic conditions or bone abnormalities were identified. Clinically, there was no tenderness, swelling, or ecchymosis evident. He had full passive range of motion (ROM) of his ring finger, and proximal interphalangeal (PIP) joint active ROM was 0/110º; however, he had no activity of the FDP of the ring finger. Preoperative radiographs were normal. The hook of the hamate was clinically and radiographically normal.

A preoperative diagnosis of FDP avulsion from the distal phalanx was made, and the operation was carried out 16 days after injury. Surgical exploration started in zone II and extended proximally into the distal palmar crease, but no stump was found in either location. Therefore, exploration was carried out to the midpalmar region, revealing the tendon rupture in zone III, in the region of the origin of the ring finger lumbrical muscle (Figure 1). The flexor digitorum superficialis tendon was intact. Macroscopically, both tendon and carpal tunnel appeared normal, with no evidence of tendon attrition; thus, the tendon was not sent for histologic examination. The ends of the ruptured FDP tendon to the ring finger were at the level of the superficial palmar arch, with the distal end appearing as though it had been cut sharply with a knife. Because of the short period of time from injury to exploration, delayed primary tendon repair was possible, along with side-to-side tenodesis to the intact ring finger flexor superficialis tendon in the palm (Figure 2). Two days after surgery, the patient started a controlled mobilization program using the Duran method.¹⁰

At final follow-up of 18 months, total active motion was 126°, which corresponds to a good outcome, according to the Strickland and Glogovac criteria.¹¹ Grip strength was 50 kg, which was 84% of grip strength on the uninjured side. The patient was back to recreational activity but had not returned to work.

Discussion

Most flexor tendon ruptures result from avulsion of the FDP tendon at its distal phalanx insertion, commonly known as Jersey finger. However, true midsubstance spontaneous ruptures are infrequent. Reports of spontaneous tendon ruptures of all types, including those of the hand, have increased in incidence in most countries.¹² Bois and colleagues,¹³ who have reviewed the literature over a 50-year period, found a total of 50 spontaneous ruptures of “normal” flexor tendon in 43 cases. The authors point to unique historical and physical examinaton findings that help differentiate spontaneous tendon ruptures from the more common FDP avulsions. Such findings include the sensation of a pop or snap, or a sudden sharp pain or cramp within the palmar region. In contrast, most avulsion ruptures cause discomfort within the region of the digit. In type I avulsion injuries of the FDP tendon, the proximal tendon stump usually retracts proximal to the digital tendon sheath, causing a tender mass in the palm.¹⁴ Flexor digitorum profundus tendon avulsions, however, are not typically associated with a snap or pop in the palm. When spontaneous ruptures of the hand occur, they typically involve the profundus tendon of the small finger, in the area of the lumbrical origin.¹³

In equivocal cases when the site of rupture is uncertain, ultrasound and magnetic resonance imaging may assist in making the diagnosis and provide important preoperative information for surgical decision-making and planning; this information may decrease postoperative morbidity by minimizing surgical dissection.

The etiology of spontaneous ruptures is incompletely understood. For any rupture of the ulnar flexor tendons, the hook of the hamate should be examined to rule out a previous fracture as a cause of tendon attrition.¹⁵ Tendon vascularization may be a cause for tendon rupture in the hand. When the blood supply of the lumbrical muscles was examined in 100 upper extremities from human cadavers using vascular injection studies,¹⁶ it was discovered that each lumbrical muscle received its arterial supply from 4 sources: the superficial palmar arch, the common palmar digital artery, the deep palmar arch, and the dorsal digital artery. There were no anastomoses between the networks supplying the lumbrical muscles and the FDP tendons within the palm, suggesting a possible watershed zone between the FDP tendon and lumbrical muscle origin. The patient described in this case had the tendon rupture in the area of potential hypovascularity at the lumbrical origin.

Important factors in the decision-making process for surgical treatment include the length of time between rupture and treatment, the site of rupture, and the condition of the ruptured tendon ends. Patients who present in the first 3 weeks of injury can be treated by primary tendon repair, provided that the ruptured tendon ends are not significantly frayed or attenuated. For patients presenting more than 3 weeks after injury, interposition tendon grafts or tendon transfers are suitable options for ruptures in zone III. Distal interphalangeal joint arthrodesis is another alternative in specific cases where reconstruction is not possible. In this case, direct end-to-end repair was possible, as well as tenodesis to the intact ring finger superficialis in order to prevent stretching of the repair.

Localizing the level of the tendon rupture clinically is difficult. When the site of the profundus tendon rupture is uncertain, and there is no tenderness in zone I or the PIP joint, the first incision should be made at the metacarpophalangeal joint level. This first incision will indicate if the rupture occurred in zone III. If the tendon is intact at that location, then the next incision should be at the level of the PIP joint.

Conclusion

We report a patient treated for spontaneous rupture of the flexor tendon in zone III. He was treated in the acute setting with direct tendon repair. It is important to consider spontaneous rupture of the tendon in patients presenting with a snap/pop and the sudden inability to flex a finger. A tendon rupture can be diagnosed as spontaneous in the absence of an underlying pathologic condition such as rheumatoid arthritis, gout, or occult carpal fractures. In the acute setting, these may be repaired primarily; however, if presenting after a few weeks, alternative surgical options, including interposition tendon grafts, tendon transfer, and DIP joint arthrodesis, should be considered.

Flexor tendons are considered the strongest component of the musculotendinous unit; they generally do not rupture unless weakened by an underlying pathologic condition.¹ According to traditional teaching, when the musculotendinous unit is subjected to excessive forces, failure invariably occurs at the tendon insertion, at the musculotendinous junction, within the muscle substance, or at its origin from the bone before the tendon itself ruptures.¹

Midsubstance tears in nonrheumatoid patients are less frequent and are typically attributable to an underlying cause.² Possible pathologic conditions include, but are not limited to, osteoarthritis of the pisotriquetral joint,³ nonunion fracture of the hook of the hamate,⁴ lunate dislocation,⁵ accessory carpal bone,⁶ gouty infiltration of the flexor tendon,⁷ and tumor.⁸ In 1960, Boyes and colleagues⁹ presented a series of 80 flexor tendon ruptures in 78 patients over a 13-year period. Only 3 cases had no identifiable cause. The authors recommended using the term spontaneous for those ruptures that occur within the tendon substance without underlying or associated pathologic changes.

We describe a patient with spontaneous rupture of the flexor digitorum profundus (FDP) tendon at zone III, satisfying Boyes’ definition of the term spontaneous. The patient provided written informed consent for print and electronic publication of this case report.

Case Report

A 65-year-old, right-handed manual worker was assessed in our hand clinic 3 days after he felt a cramp in his left palm while lifting a heavy object. Shortly thereafter, he noted he could not flex his ring finger distal interphalangeal (DIP) joint. He could not recall any previous injury to his finger. No predisposing pathologic conditions or bone abnormalities were identified. Clinically, there was no tenderness, swelling, or ecchymosis evident. He had full passive range of motion (ROM) of his ring finger, and proximal interphalangeal (PIP) joint active ROM was 0/110º; however, he had no activity of the FDP of the ring finger. Preoperative radiographs were normal. The hook of the hamate was clinically and radiographically normal.

A preoperative diagnosis of FDP avulsion from the distal phalanx was made, and the operation was carried out 16 days after injury. Surgical exploration started in zone II and extended proximally into the distal palmar crease, but no stump was found in either location. Therefore, exploration was carried out to the midpalmar region, revealing the tendon rupture in zone III, in the region of the origin of the ring finger lumbrical muscle (Figure 1). The flexor digitorum superficialis tendon was intact. Macroscopically, both tendon and carpal tunnel appeared normal, with no evidence of tendon attrition; thus, the tendon was not sent for histologic examination. The ends of the ruptured FDP tendon to the ring finger were at the level of the superficial palmar arch, with the distal end appearing as though it had been cut sharply with a knife. Because of the short period of time from injury to exploration, delayed primary tendon repair was possible, along with side-to-side tenodesis to the intact ring finger flexor superficialis tendon in the palm (Figure 2). Two days after surgery, the patient started a controlled mobilization program using the Duran method.¹⁰

At final follow-up of 18 months, total active motion was 126°, which corresponds to a good outcome, according to the Strickland and Glogovac criteria.¹¹ Grip strength was 50 kg, which was 84% of grip strength on the uninjured side. The patient was back to recreational activity but had not returned to work.

Discussion

Most flexor tendon ruptures result from avulsion of the FDP tendon at its distal phalanx insertion, commonly known as Jersey finger. However, true midsubstance spontaneous ruptures are infrequent. Reports of spontaneous tendon ruptures of all types, including those of the hand, have increased in incidence in most countries.¹² Bois and colleagues,¹³ who have reviewed the literature over a 50-year period, found a total of 50 spontaneous ruptures of “normal” flexor tendon in 43 cases. The authors point to unique historical and physical examinaton findings that help differentiate spontaneous tendon ruptures from the more common FDP avulsions. Such findings include the sensation of a pop or snap, or a sudden sharp pain or cramp within the palmar region. In contrast, most avulsion ruptures cause discomfort within the region of the digit. In type I avulsion injuries of the FDP tendon, the proximal tendon stump usually retracts proximal to the digital tendon sheath, causing a tender mass in the palm.¹⁴ Flexor digitorum profundus tendon avulsions, however, are not typically associated with a snap or pop in the palm. When spontaneous ruptures of the hand occur, they typically involve the profundus tendon of the small finger, in the area of the lumbrical origin.¹³

In equivocal cases when the site of rupture is uncertain, ultrasound and magnetic resonance imaging may assist in making the diagnosis and provide important preoperative information for surgical decision-making and planning; this information may decrease postoperative morbidity by minimizing surgical dissection.

The etiology of spontaneous ruptures is incompletely understood. For any rupture of the ulnar flexor tendons, the hook of the hamate should be examined to rule out a previous fracture as a cause of tendon attrition.¹⁵ Tendon vascularization may be a cause for tendon rupture in the hand. When the blood supply of the lumbrical muscles was examined in 100 upper extremities from human cadavers using vascular injection studies,¹⁶ it was discovered that each lumbrical muscle received its arterial supply from 4 sources: the superficial palmar arch, the common palmar digital artery, the deep palmar arch, and the dorsal digital artery. There were no anastomoses between the networks supplying the lumbrical muscles and the FDP tendons within the palm, suggesting a possible watershed zone between the FDP tendon and lumbrical muscle origin. The patient described in this case had the tendon rupture in the area of potential hypovascularity at the lumbrical origin.

Important factors in the decision-making process for surgical treatment include the length of time between rupture and treatment, the site of rupture, and the condition of the ruptured tendon ends. Patients who present in the first 3 weeks of injury can be treated by primary tendon repair, provided that the ruptured tendon ends are not significantly frayed or attenuated. For patients presenting more than 3 weeks after injury, interposition tendon grafts or tendon transfers are suitable options for ruptures in zone III. Distal interphalangeal joint arthrodesis is another alternative in specific cases where reconstruction is not possible. In this case, direct end-to-end repair was possible, as well as tenodesis to the intact ring finger superficialis in order to prevent stretching of the repair.

Localizing the level of the tendon rupture clinically is difficult. When the site of the profundus tendon rupture is uncertain, and there is no tenderness in zone I or the PIP joint, the first incision should be made at the metacarpophalangeal joint level. This first incision will indicate if the rupture occurred in zone III. If the tendon is intact at that location, then the next incision should be at the level of the PIP joint.

Conclusion

We report a patient treated for spontaneous rupture of the flexor tendon in zone III. He was treated in the acute setting with direct tendon repair. It is important to consider spontaneous rupture of the tendon in patients presenting with a snap/pop and the sudden inability to flex a finger. A tendon rupture can be diagnosed as spontaneous in the absence of an underlying pathologic condition such as rheumatoid arthritis, gout, or occult carpal fractures. In the acute setting, these may be repaired primarily; however, if presenting after a few weeks, alternative surgical options, including interposition tendon grafts, tendon transfer, and DIP joint arthrodesis, should be considered.

Rotator cuff tears (RCTs) are common tendon injuries that can cause chronic pain and severe functional disability. Massive RCTs do not heal spontaneously and, in many cases, result in poor clinical outcomes. Specifically, muscle atrophy and fatty infiltration correlate with poor outcomes after surgical repair.¹ Fatty infiltration of the rotator cuff is a common phenomenon that can lead to permanent structural alterations within the tendon. It has been suggested that changes in muscle fiber orientation (the pennation angle) can cause mesenchymal stem cells to migrate to the interface between muscle fibers and the region of fatty infiltration of the muscle.² Understanding the factors involved in muscle degeneration and atrophy, and in fatty infiltration, may lead to treatments that improve outcomes for patients with massive RCTs. One proposed treatment involves placing continuous mechanical traction on the ends of the torn tendon.² Findings from this research have indicated that acute tears that become chronic tears are typified by inelasticity and poor function of the muscle­–tendon unit. It is therefore important to develop a method that speeds tendon healing without causing the muscle fiber atrophy and pennation angle changes that lead to fatty atrophy, which appears to be an irreversible structural change.

On the basis of the theory that adding mesenchymal cells may improve tendon healing, investigators have studied use of transcription factors (eg, scleraxis) specific to tendogenesis in the embryonal stage.^3,4 Nevertheless, certain transcription factors are associated with formation of fibrocartilage in higher concentrations.⁴ Moreover, decalcified bone matrix increases cartilage formation when added to the tendon repair site.⁵ Cartilage formation, however, is associated with poorer functional results.⁶ Thus, there is a need for a method that facilitates faster tendon healing with higher quality tissue formation and less muscle atrophy.

Chitosan, a linear polysaccharide, is associated with scarless healing of soft tissues and prevention of adhesion formation both intraperitoneally and during tendon healing after surgery.^7,8 Chitosan tends to precipitate in physiologic pH, thereby mitigating its potency. Fortunately, a chitosan solution that does not precipitate in physiologic conditions was recently developed.⁹ The solution’s lack of precipitation, coupled with its in situ gelling, allows it to adhere to the repair site long enough to take effect. These characteristics could allow for intimate contact between gel and tendon, facilitating guided-tissue regeneration and preventing adhesion of the rotator cuff to surrounding tissue. By contrast, other biological agents (eg, platelet-rich plasma) are administered as fluid rather than gel and are therefore more susceptible to diffusing from the repair site, mitigating their effects. Thus, chitosan gel is fairly unique among agents.

In the study reported here, we histologically investigated whether a chitosan gel would help improve healing of rotator cuff tendon (acute supraspinatus) tears in a rat model.

Materials and Methods

Supraspinatus Surgical Model

Forty Wistar rats, each weighing between 300 and 400 g, were used in this study. All procedures were approved by the Institutional Animal Care and Use Committee at Rabin Medical Center in Petah Tikva, Israel. The rats were anesthetized with ketamine 90 mg/kg and xylazine 10 mg/kg, both administered intramuscularly, and anesthesia was prolonged as needed with 2% isoflurane, administered by nose cone. The skin was incised 5 cm along the upper back following the midline of the spine. The resulting skin flaps were retracted and the scapula exposed. Careful blunt dissection allowed visualization of the rotator cuff and the trans-scapular arch. A full-thickness incision of the supraspinatus tendon was then made 2 mm distal to the arch. This procedure was performed on both shoulders. For the right supraspinatus tendon, a bioabsorbable chitosan–hydrochloric acid solution (>70% de-acetylated chitosan, molecular weight of 600 kDa; Heppe Medical Chitosan GmbH, Halle, Germany) was sterilely applied to the ends of the tendon (total volume, 0.5 mL) and automatically gelled in situ by heating to about 37°C (rat’s internal body temperature). The tendon ends were subsequently approximated with a single 4-0 Prolene suture (Ethicon, Somerville, New Jersey). The left shoulder (tendon repaired with suture only) served as a control.

The rats were housed for a maximum of 12 weeks after surgery. They were sacrificed (in groups of 5 each) 2 hours, 3 days, 1 week, 2 weeks, 4 weeks, 6 weeks, 8 weeks, and 12 weeks after surgery. After each rat was sacrificed, both shoulder girdles were harvested, and the sutures were removed from the supraspinatus tendons.

Histologic Analysis

After routine fixation with 4% formalin for 48 hours and decalcification with 10% ethylenediaminetetraacetic acid (EDTA) for 3 weeks, the specimens were sectioned with a microtome blade. Care was taken to ensure the plane of the microtome blade was parallel with the longitudinal plane of the supraspinatus muscle and tendon to allow for evaluation of pennation angle. Hematoxylin-eosin staining and Masson trichrome staining were subsequently performed.

A variety of histologic measurements were obtained with use of ImageJ software (US National Institutes of Health). Percentage of fibrous tissue was determined by examining the slides at low magnification fields (×25) at the tendon healing site. Three such fields were evaluated per specimen. The fibrous tissue was circled manually, and percentage of tissue area was assessed and compared with total region of interest. Cellularity was carefully outlined and measured as percentage of total tendon area occupied by cells. Fatty atrophy was defined as either present or absent. Muscle fiber diameter was defined as average diameter of 10 muscle fibers measured within 2 mm of the tendon laceration site. Inflammatory cell collections were defined as either large (>100 µm in diameter) or small (<100 µm in diameter) and were dichotomized to either present or absent. Pennation angle was defined as average angle between muscle fibers and longitudinal axis of supraspinatus muscle and tendon unit. Ten fibers proximal to and within 2 mm of the laceration site were randomly selected, measured, and averaged.

Statistical Analysis

Statistical analysis was performed with Analyse-it 2.20 for Microsoft Excel 2010 (Analyse-it Software, Leeds, United Kingdom). Data were initially analyzed with the Kolmogorov-Smirnov test to assess for normality of distribution. The t test was used to compare continuous variables when the data were normally distributed and the Mann-Whitney test when the data were not normally distributed.

Results

All tendons (both groups) healed within 12 weeks. Generally, the tissue formed at the repair site exhibited a mixture of tenocyte-like cells (fibrotic tissue) and granulation tissue without clear orientation. As noted in Figure 1, the tendons treated with chitosan had more fibrotic tissue (overall mean, 21.5%) relative to the control group (mean, 12.3%), and the difference was significant (P = .003). The most notable differences were found at time points later than 1 week after surgery. In addition, amount of cellularity (Figure 2) was higher in chitosan-treated tendon and control tendon than in the normal, uninjured adjacent tendon at all time points (P < .001). Chitosan-treated tendons had significantly higher cellularity than untreated control tendons from 1 to 2 weeks (P < .001), and control tendons were significantly hypercellular compared with chitosan-treated tendons from 4 to 8 weeks (P < .001), but both groups exhibited similar cellularity by 12 weeks (P > .05). Fatty atrophy was found at significantly higher rates in control rats than in chitosan-treated rats (P = .001; Table). Furthermore, as noted in Figure 3, muscle fiber diameter decreased in both groups after injury (P < .001).

Figure 4 shows that the amount of inflammatory collections was significantly smaller in the chitosan-treated group than in the control group over the course of the study (P = .01). In addition, pennation angle steadily decreased in the control group throughout the study period, whereas it transiently decreased in the chitosan-treated group (until 2 weeks) before returning to its immediate postoperative level by 12 weeks (Figure 5). Overall, the chitosan-treated group maintained a higher pennation angle than the control group did (P < .001).

Discussion

RCTs affect more than 40% of patients over age 60 years and are a common cause of debilitating pain, reduced shoulder function, and weakness.¹⁰ Thirty thousand to 75,000 rotator cuff repairs are performed annually in the United States.¹¹ Although the best treatment for this disorder remains a topic of debate, arthroscopic and (when necessary) open surgical repair is the accepted gold standard for the treatment of tears that do not improve with conservative management. Despite advances in the surgical treatment of these tears, the surgical failure rates are high (range, 20%-90%), with failures attributed to factors beyond patient age, tear size and chronicity, muscle atrophy and degeneration, tendon quality, repair technique, and postoperative rehabilitation.^12,13 Repair strategies that biologically enhance the patient’s intrinsic healing potential are needed.

In tendon repair, choice of repair material (eg, graft) is crucial in determining the success of tissue engineering approaches. The ideal scaffold is biocompatible and does not elicit a host inflammatory response. The selected scaffold in its composition and fabricated form must be capable of holding and supporting cells. In addition, the scaffold should be biodegradable, serving as a temporary support for such cells and mechanically augmenting the repaired tendon while allowing for eventual replacement by matrix components. Moreover, the scaffold should have high porosity and a large surface area. Furthermore, the material should mimic the native tendon extracellular matrix (ECM) architecture to allow cells to be distributed throughout the scaffold and to facilitate diffusion of nutrients and factors that promote cellular proliferation and ECM production.

Given the importance of glycosaminoglycans (GAGs) in supporting the reticular structure of the matrix, use of GAGs or GAG-analogues as components of a tendon tissue scaffold for enhancing repair is well documented.¹⁴ One such candidate is chitosan, a partially de-acetylated derivative of chitin found in arthropod exoskeletons. Structurally, chitosan shares some characteristics with various GAGs and hyaluronic acid.¹⁵ More specifically, chitosan is a linear polysaccharide composed of glucosamine and N-acetyl glucosamine units linked by β-glycosidic bonds. Investigators have studied the properties of chitosan, including its biocompatibility, biodegradability, antibacterial activity, mucoadhesivity, and wound healing.^16,17

One of the most promising features of chitosan is that it can be processed into porous structures for use in cell transplantation and tissue regeneration.^18,19 Porous chitosan structures can be formed by freezing and lyophilizing chitosan-acetic acid solutions; chondrogenic cell adhesion and proliferation onto these structures have been reported.^20,21 This chitosan scaffolding method has also been used to test different composites with collagens, gelatins, GAGs, and hyaluronic acid, all of which have also been proposed as useful 3-dimensional materials for tissue repair.²²

In the present study, we used chitosan matrix in RCT repair. We hypothesized that chitosan matrix could enhance rotator cuff repair the same way it enhances repair in epidermal tissues.¹⁶ Histologic findings demonstrated that the percentage of fibrous tissue was significantly higher in the chitosan-treated group than in the control group. This improved fibroblastic response may be attributed to the ability of chitosan to enhance cell migration and serve as a scaffold for repair. Other studies have indicated that chitin, of which chitosan is the primary derivative, accelerated the healing of skin and subcutaneous tissues by increased cell migration.²³ Moreover, Okamoto and colleagues²⁴ reported that chitin implants stimulated abundant angiogenesis through the same mechanism.

Inadequate initial strength of a repair may lead to a recurrent cuff tear or a disability of rotator cuff function in the early healing stages. In our study, the chitosan matrix tended to be absorbed by 6 weeks after surgery. Its adherence to and ultimate absorption at the repair site may be challenged by the flow of irrigation fluid through the subacromial space in the setting of arthroscopic surgery. However, because the chitosan remains in a more robust gel form, it is better able to resist being washed from the repair site. For augmentation, it may be possible to apply a biocompatible patch over the gel to further protect it from being dislodged. In addition, histologic findings showed that the fibrous repair tissue gradually increased until reaching a peak 8 weeks after surgery—an indication that the absorption rate of the chitosan scaffold lags behind full recovery of the repair tissue. Given this relationship, further studies are needed to determine the mechanical strength of the repair between 6 and 8 weeks, which is important for avoiding recurrent tears.

This study had a few limitations. First, as with any animal model, the anatomy and function of the rat shoulder differ from those of the human shoulder. The acromial arch differs in quadruped animals, with less coverage of the supraspinatus and more of the subscapularis.²⁵ These anatomical differences could yield altered stress mechanics that could affect tendon repair. Furthermore, rats and humans differ in their RCT healing rates. Thus, the pathophysiology of muscle atrophy and fat infiltration in rats may slightly differ from that in humans. In addition, no mechanical testing was performed to compare chitosan-treated and untreated rotator cuff repairs, and such testing is needed to clarify the biomechanical importance of augmentation. Furthermore, no immunohistochemical analysis was performed for collagen. In the repair of rotator cuff tendons, surgeons must consider not only the number of cells but also the production of ECM. Although not directly confirmed in this study, chitosan induced fibrous tissue proliferation that mirrored production of a large amount of collagen fibers. Last, we used an open RTC model. As an arthroscopic model was not used, no definitive conclusions can be drawn regarding use of chitosan in arthroscopy.

Conclusion

Use of chitosan as an acellular matrix improved formation of healing fibrous tissue, increased the number of cells, and prevented fatty atrophy and inflammatory aggregates inside repair sites while facilitating recovery of the natural pennation angle of the tissue. These results demonstrate that chitosan can enhance tendon healing in the setting of acute RCT. Further research, including biomechanical testing of repaired tendons, is needed to further delineate the utility of chitosan in regenerating irreparable RCTs.

Rotator cuff tears (RCTs) are common tendon injuries that can cause chronic pain and severe functional disability. Massive RCTs do not heal spontaneously and, in many cases, result in poor clinical outcomes. Specifically, muscle atrophy and fatty infiltration correlate with poor outcomes after surgical repair.¹ Fatty infiltration of the rotator cuff is a common phenomenon that can lead to permanent structural alterations within the tendon. It has been suggested that changes in muscle fiber orientation (the pennation angle) can cause mesenchymal stem cells to migrate to the interface between muscle fibers and the region of fatty infiltration of the muscle.² Understanding the factors involved in muscle degeneration and atrophy, and in fatty infiltration, may lead to treatments that improve outcomes for patients with massive RCTs. One proposed treatment involves placing continuous mechanical traction on the ends of the torn tendon.² Findings from this research have indicated that acute tears that become chronic tears are typified by inelasticity and poor function of the muscle­–tendon unit. It is therefore important to develop a method that speeds tendon healing without causing the muscle fiber atrophy and pennation angle changes that lead to fatty atrophy, which appears to be an irreversible structural change.

On the basis of the theory that adding mesenchymal cells may improve tendon healing, investigators have studied use of transcription factors (eg, scleraxis) specific to tendogenesis in the embryonal stage.^3,4 Nevertheless, certain transcription factors are associated with formation of fibrocartilage in higher concentrations.⁴ Moreover, decalcified bone matrix increases cartilage formation when added to the tendon repair site.⁵ Cartilage formation, however, is associated with poorer functional results.⁶ Thus, there is a need for a method that facilitates faster tendon healing with higher quality tissue formation and less muscle atrophy.

Chitosan, a linear polysaccharide, is associated with scarless healing of soft tissues and prevention of adhesion formation both intraperitoneally and during tendon healing after surgery.^7,8 Chitosan tends to precipitate in physiologic pH, thereby mitigating its potency. Fortunately, a chitosan solution that does not precipitate in physiologic conditions was recently developed.⁹ The solution’s lack of precipitation, coupled with its in situ gelling, allows it to adhere to the repair site long enough to take effect. These characteristics could allow for intimate contact between gel and tendon, facilitating guided-tissue regeneration and preventing adhesion of the rotator cuff to surrounding tissue. By contrast, other biological agents (eg, platelet-rich plasma) are administered as fluid rather than gel and are therefore more susceptible to diffusing from the repair site, mitigating their effects. Thus, chitosan gel is fairly unique among agents.

In the study reported here, we histologically investigated whether a chitosan gel would help improve healing of rotator cuff tendon (acute supraspinatus) tears in a rat model.

Materials and Methods

Supraspinatus Surgical Model

Forty Wistar rats, each weighing between 300 and 400 g, were used in this study. All procedures were approved by the Institutional Animal Care and Use Committee at Rabin Medical Center in Petah Tikva, Israel. The rats were anesthetized with ketamine 90 mg/kg and xylazine 10 mg/kg, both administered intramuscularly, and anesthesia was prolonged as needed with 2% isoflurane, administered by nose cone. The skin was incised 5 cm along the upper back following the midline of the spine. The resulting skin flaps were retracted and the scapula exposed. Careful blunt dissection allowed visualization of the rotator cuff and the trans-scapular arch. A full-thickness incision of the supraspinatus tendon was then made 2 mm distal to the arch. This procedure was performed on both shoulders. For the right supraspinatus tendon, a bioabsorbable chitosan–hydrochloric acid solution (>70% de-acetylated chitosan, molecular weight of 600 kDa; Heppe Medical Chitosan GmbH, Halle, Germany) was sterilely applied to the ends of the tendon (total volume, 0.5 mL) and automatically gelled in situ by heating to about 37°C (rat’s internal body temperature). The tendon ends were subsequently approximated with a single 4-0 Prolene suture (Ethicon, Somerville, New Jersey). The left shoulder (tendon repaired with suture only) served as a control.

The rats were housed for a maximum of 12 weeks after surgery. They were sacrificed (in groups of 5 each) 2 hours, 3 days, 1 week, 2 weeks, 4 weeks, 6 weeks, 8 weeks, and 12 weeks after surgery. After each rat was sacrificed, both shoulder girdles were harvested, and the sutures were removed from the supraspinatus tendons.

Histologic Analysis

After routine fixation with 4% formalin for 48 hours and decalcification with 10% ethylenediaminetetraacetic acid (EDTA) for 3 weeks, the specimens were sectioned with a microtome blade. Care was taken to ensure the plane of the microtome blade was parallel with the longitudinal plane of the supraspinatus muscle and tendon to allow for evaluation of pennation angle. Hematoxylin-eosin staining and Masson trichrome staining were subsequently performed.

A variety of histologic measurements were obtained with use of ImageJ software (US National Institutes of Health). Percentage of fibrous tissue was determined by examining the slides at low magnification fields (×25) at the tendon healing site. Three such fields were evaluated per specimen. The fibrous tissue was circled manually, and percentage of tissue area was assessed and compared with total region of interest. Cellularity was carefully outlined and measured as percentage of total tendon area occupied by cells. Fatty atrophy was defined as either present or absent. Muscle fiber diameter was defined as average diameter of 10 muscle fibers measured within 2 mm of the tendon laceration site. Inflammatory cell collections were defined as either large (>100 µm in diameter) or small (<100 µm in diameter) and were dichotomized to either present or absent. Pennation angle was defined as average angle between muscle fibers and longitudinal axis of supraspinatus muscle and tendon unit. Ten fibers proximal to and within 2 mm of the laceration site were randomly selected, measured, and averaged.

Statistical Analysis

Statistical analysis was performed with Analyse-it 2.20 for Microsoft Excel 2010 (Analyse-it Software, Leeds, United Kingdom). Data were initially analyzed with the Kolmogorov-Smirnov test to assess for normality of distribution. The t test was used to compare continuous variables when the data were normally distributed and the Mann-Whitney test when the data were not normally distributed.

Results

All tendons (both groups) healed within 12 weeks. Generally, the tissue formed at the repair site exhibited a mixture of tenocyte-like cells (fibrotic tissue) and granulation tissue without clear orientation. As noted in Figure 1, the tendons treated with chitosan had more fibrotic tissue (overall mean, 21.5%) relative to the control group (mean, 12.3%), and the difference was significant (P = .003). The most notable differences were found at time points later than 1 week after surgery. In addition, amount of cellularity (Figure 2) was higher in chitosan-treated tendon and control tendon than in the normal, uninjured adjacent tendon at all time points (P < .001). Chitosan-treated tendons had significantly higher cellularity than untreated control tendons from 1 to 2 weeks (P < .001), and control tendons were significantly hypercellular compared with chitosan-treated tendons from 4 to 8 weeks (P < .001), but both groups exhibited similar cellularity by 12 weeks (P > .05). Fatty atrophy was found at significantly higher rates in control rats than in chitosan-treated rats (P = .001; Table). Furthermore, as noted in Figure 3, muscle fiber diameter decreased in both groups after injury (P < .001).

Figure 4 shows that the amount of inflammatory collections was significantly smaller in the chitosan-treated group than in the control group over the course of the study (P = .01). In addition, pennation angle steadily decreased in the control group throughout the study period, whereas it transiently decreased in the chitosan-treated group (until 2 weeks) before returning to its immediate postoperative level by 12 weeks (Figure 5). Overall, the chitosan-treated group maintained a higher pennation angle than the control group did (P < .001).

Discussion

RCTs affect more than 40% of patients over age 60 years and are a common cause of debilitating pain, reduced shoulder function, and weakness.¹⁰ Thirty thousand to 75,000 rotator cuff repairs are performed annually in the United States.¹¹ Although the best treatment for this disorder remains a topic of debate, arthroscopic and (when necessary) open surgical repair is the accepted gold standard for the treatment of tears that do not improve with conservative management. Despite advances in the surgical treatment of these tears, the surgical failure rates are high (range, 20%-90%), with failures attributed to factors beyond patient age, tear size and chronicity, muscle atrophy and degeneration, tendon quality, repair technique, and postoperative rehabilitation.^12,13 Repair strategies that biologically enhance the patient’s intrinsic healing potential are needed.

In tendon repair, choice of repair material (eg, graft) is crucial in determining the success of tissue engineering approaches. The ideal scaffold is biocompatible and does not elicit a host inflammatory response. The selected scaffold in its composition and fabricated form must be capable of holding and supporting cells. In addition, the scaffold should be biodegradable, serving as a temporary support for such cells and mechanically augmenting the repaired tendon while allowing for eventual replacement by matrix components. Moreover, the scaffold should have high porosity and a large surface area. Furthermore, the material should mimic the native tendon extracellular matrix (ECM) architecture to allow cells to be distributed throughout the scaffold and to facilitate diffusion of nutrients and factors that promote cellular proliferation and ECM production.

Given the importance of glycosaminoglycans (GAGs) in supporting the reticular structure of the matrix, use of GAGs or GAG-analogues as components of a tendon tissue scaffold for enhancing repair is well documented.¹⁴ One such candidate is chitosan, a partially de-acetylated derivative of chitin found in arthropod exoskeletons. Structurally, chitosan shares some characteristics with various GAGs and hyaluronic acid.¹⁵ More specifically, chitosan is a linear polysaccharide composed of glucosamine and N-acetyl glucosamine units linked by β-glycosidic bonds. Investigators have studied the properties of chitosan, including its biocompatibility, biodegradability, antibacterial activity, mucoadhesivity, and wound healing.^16,17

One of the most promising features of chitosan is that it can be processed into porous structures for use in cell transplantation and tissue regeneration.^18,19 Porous chitosan structures can be formed by freezing and lyophilizing chitosan-acetic acid solutions; chondrogenic cell adhesion and proliferation onto these structures have been reported.^20,21 This chitosan scaffolding method has also been used to test different composites with collagens, gelatins, GAGs, and hyaluronic acid, all of which have also been proposed as useful 3-dimensional materials for tissue repair.²²

In the present study, we used chitosan matrix in RCT repair. We hypothesized that chitosan matrix could enhance rotator cuff repair the same way it enhances repair in epidermal tissues.¹⁶ Histologic findings demonstrated that the percentage of fibrous tissue was significantly higher in the chitosan-treated group than in the control group. This improved fibroblastic response may be attributed to the ability of chitosan to enhance cell migration and serve as a scaffold for repair. Other studies have indicated that chitin, of which chitosan is the primary derivative, accelerated the healing of skin and subcutaneous tissues by increased cell migration.²³ Moreover, Okamoto and colleagues²⁴ reported that chitin implants stimulated abundant angiogenesis through the same mechanism.

Inadequate initial strength of a repair may lead to a recurrent cuff tear or a disability of rotator cuff function in the early healing stages. In our study, the chitosan matrix tended to be absorbed by 6 weeks after surgery. Its adherence to and ultimate absorption at the repair site may be challenged by the flow of irrigation fluid through the subacromial space in the setting of arthroscopic surgery. However, because the chitosan remains in a more robust gel form, it is better able to resist being washed from the repair site. For augmentation, it may be possible to apply a biocompatible patch over the gel to further protect it from being dislodged. In addition, histologic findings showed that the fibrous repair tissue gradually increased until reaching a peak 8 weeks after surgery—an indication that the absorption rate of the chitosan scaffold lags behind full recovery of the repair tissue. Given this relationship, further studies are needed to determine the mechanical strength of the repair between 6 and 8 weeks, which is important for avoiding recurrent tears.

This study had a few limitations. First, as with any animal model, the anatomy and function of the rat shoulder differ from those of the human shoulder. The acromial arch differs in quadruped animals, with less coverage of the supraspinatus and more of the subscapularis.²⁵ These anatomical differences could yield altered stress mechanics that could affect tendon repair. Furthermore, rats and humans differ in their RCT healing rates. Thus, the pathophysiology of muscle atrophy and fat infiltration in rats may slightly differ from that in humans. In addition, no mechanical testing was performed to compare chitosan-treated and untreated rotator cuff repairs, and such testing is needed to clarify the biomechanical importance of augmentation. Furthermore, no immunohistochemical analysis was performed for collagen. In the repair of rotator cuff tendons, surgeons must consider not only the number of cells but also the production of ECM. Although not directly confirmed in this study, chitosan induced fibrous tissue proliferation that mirrored production of a large amount of collagen fibers. Last, we used an open RTC model. As an arthroscopic model was not used, no definitive conclusions can be drawn regarding use of chitosan in arthroscopy.

Conclusion

Use of chitosan as an acellular matrix improved formation of healing fibrous tissue, increased the number of cells, and prevented fatty atrophy and inflammatory aggregates inside repair sites while facilitating recovery of the natural pennation angle of the tissue. These results demonstrate that chitosan can enhance tendon healing in the setting of acute RCT. Further research, including biomechanical testing of repaired tendons, is needed to further delineate the utility of chitosan in regenerating irreparable RCTs.

Rotator cuff tears (RCTs) are common tendon injuries that can cause chronic pain and severe functional disability. Massive RCTs do not heal spontaneously and, in many cases, result in poor clinical outcomes. Specifically, muscle atrophy and fatty infiltration correlate with poor outcomes after surgical repair.¹ Fatty infiltration of the rotator cuff is a common phenomenon that can lead to permanent structural alterations within the tendon. It has been suggested that changes in muscle fiber orientation (the pennation angle) can cause mesenchymal stem cells to migrate to the interface between muscle fibers and the region of fatty infiltration of the muscle.² Understanding the factors involved in muscle degeneration and atrophy, and in fatty infiltration, may lead to treatments that improve outcomes for patients with massive RCTs. One proposed treatment involves placing continuous mechanical traction on the ends of the torn tendon.² Findings from this research have indicated that acute tears that become chronic tears are typified by inelasticity and poor function of the muscle­–tendon unit. It is therefore important to develop a method that speeds tendon healing without causing the muscle fiber atrophy and pennation angle changes that lead to fatty atrophy, which appears to be an irreversible structural change.

On the basis of the theory that adding mesenchymal cells may improve tendon healing, investigators have studied use of transcription factors (eg, scleraxis) specific to tendogenesis in the embryonal stage.^3,4 Nevertheless, certain transcription factors are associated with formation of fibrocartilage in higher concentrations.⁴ Moreover, decalcified bone matrix increases cartilage formation when added to the tendon repair site.⁵ Cartilage formation, however, is associated with poorer functional results.⁶ Thus, there is a need for a method that facilitates faster tendon healing with higher quality tissue formation and less muscle atrophy.

Chitosan, a linear polysaccharide, is associated with scarless healing of soft tissues and prevention of adhesion formation both intraperitoneally and during tendon healing after surgery.^7,8 Chitosan tends to precipitate in physiologic pH, thereby mitigating its potency. Fortunately, a chitosan solution that does not precipitate in physiologic conditions was recently developed.⁹ The solution’s lack of precipitation, coupled with its in situ gelling, allows it to adhere to the repair site long enough to take effect. These characteristics could allow for intimate contact between gel and tendon, facilitating guided-tissue regeneration and preventing adhesion of the rotator cuff to surrounding tissue. By contrast, other biological agents (eg, platelet-rich plasma) are administered as fluid rather than gel and are therefore more susceptible to diffusing from the repair site, mitigating their effects. Thus, chitosan gel is fairly unique among agents.

In the study reported here, we histologically investigated whether a chitosan gel would help improve healing of rotator cuff tendon (acute supraspinatus) tears in a rat model.

Materials and Methods

Supraspinatus Surgical Model

Forty Wistar rats, each weighing between 300 and 400 g, were used in this study. All procedures were approved by the Institutional Animal Care and Use Committee at Rabin Medical Center in Petah Tikva, Israel. The rats were anesthetized with ketamine 90 mg/kg and xylazine 10 mg/kg, both administered intramuscularly, and anesthesia was prolonged as needed with 2% isoflurane, administered by nose cone. The skin was incised 5 cm along the upper back following the midline of the spine. The resulting skin flaps were retracted and the scapula exposed. Careful blunt dissection allowed visualization of the rotator cuff and the trans-scapular arch. A full-thickness incision of the supraspinatus tendon was then made 2 mm distal to the arch. This procedure was performed on both shoulders. For the right supraspinatus tendon, a bioabsorbable chitosan–hydrochloric acid solution (>70% de-acetylated chitosan, molecular weight of 600 kDa; Heppe Medical Chitosan GmbH, Halle, Germany) was sterilely applied to the ends of the tendon (total volume, 0.5 mL) and automatically gelled in situ by heating to about 37°C (rat’s internal body temperature). The tendon ends were subsequently approximated with a single 4-0 Prolene suture (Ethicon, Somerville, New Jersey). The left shoulder (tendon repaired with suture only) served as a control.

The rats were housed for a maximum of 12 weeks after surgery. They were sacrificed (in groups of 5 each) 2 hours, 3 days, 1 week, 2 weeks, 4 weeks, 6 weeks, 8 weeks, and 12 weeks after surgery. After each rat was sacrificed, both shoulder girdles were harvested, and the sutures were removed from the supraspinatus tendons.

Histologic Analysis

After routine fixation with 4% formalin for 48 hours and decalcification with 10% ethylenediaminetetraacetic acid (EDTA) for 3 weeks, the specimens were sectioned with a microtome blade. Care was taken to ensure the plane of the microtome blade was parallel with the longitudinal plane of the supraspinatus muscle and tendon to allow for evaluation of pennation angle. Hematoxylin-eosin staining and Masson trichrome staining were subsequently performed.

A variety of histologic measurements were obtained with use of ImageJ software (US National Institutes of Health). Percentage of fibrous tissue was determined by examining the slides at low magnification fields (×25) at the tendon healing site. Three such fields were evaluated per specimen. The fibrous tissue was circled manually, and percentage of tissue area was assessed and compared with total region of interest. Cellularity was carefully outlined and measured as percentage of total tendon area occupied by cells. Fatty atrophy was defined as either present or absent. Muscle fiber diameter was defined as average diameter of 10 muscle fibers measured within 2 mm of the tendon laceration site. Inflammatory cell collections were defined as either large (>100 µm in diameter) or small (<100 µm in diameter) and were dichotomized to either present or absent. Pennation angle was defined as average angle between muscle fibers and longitudinal axis of supraspinatus muscle and tendon unit. Ten fibers proximal to and within 2 mm of the laceration site were randomly selected, measured, and averaged.

Statistical Analysis

Statistical analysis was performed with Analyse-it 2.20 for Microsoft Excel 2010 (Analyse-it Software, Leeds, United Kingdom). Data were initially analyzed with the Kolmogorov-Smirnov test to assess for normality of distribution. The t test was used to compare continuous variables when the data were normally distributed and the Mann-Whitney test when the data were not normally distributed.

Results

All tendons (both groups) healed within 12 weeks. Generally, the tissue formed at the repair site exhibited a mixture of tenocyte-like cells (fibrotic tissue) and granulation tissue without clear orientation. As noted in Figure 1, the tendons treated with chitosan had more fibrotic tissue (overall mean, 21.5%) relative to the control group (mean, 12.3%), and the difference was significant (P = .003). The most notable differences were found at time points later than 1 week after surgery. In addition, amount of cellularity (Figure 2) was higher in chitosan-treated tendon and control tendon than in the normal, uninjured adjacent tendon at all time points (P < .001). Chitosan-treated tendons had significantly higher cellularity than untreated control tendons from 1 to 2 weeks (P < .001), and control tendons were significantly hypercellular compared with chitosan-treated tendons from 4 to 8 weeks (P < .001), but both groups exhibited similar cellularity by 12 weeks (P > .05). Fatty atrophy was found at significantly higher rates in control rats than in chitosan-treated rats (P = .001; Table). Furthermore, as noted in Figure 3, muscle fiber diameter decreased in both groups after injury (P < .001).

Figure 4 shows that the amount of inflammatory collections was significantly smaller in the chitosan-treated group than in the control group over the course of the study (P = .01). In addition, pennation angle steadily decreased in the control group throughout the study period, whereas it transiently decreased in the chitosan-treated group (until 2 weeks) before returning to its immediate postoperative level by 12 weeks (Figure 5). Overall, the chitosan-treated group maintained a higher pennation angle than the control group did (P < .001).

Discussion

RCTs affect more than 40% of patients over age 60 years and are a common cause of debilitating pain, reduced shoulder function, and weakness.¹⁰ Thirty thousand to 75,000 rotator cuff repairs are performed annually in the United States.¹¹ Although the best treatment for this disorder remains a topic of debate, arthroscopic and (when necessary) open surgical repair is the accepted gold standard for the treatment of tears that do not improve with conservative management. Despite advances in the surgical treatment of these tears, the surgical failure rates are high (range, 20%-90%), with failures attributed to factors beyond patient age, tear size and chronicity, muscle atrophy and degeneration, tendon quality, repair technique, and postoperative rehabilitation.^12,13 Repair strategies that biologically enhance the patient’s intrinsic healing potential are needed.

In tendon repair, choice of repair material (eg, graft) is crucial in determining the success of tissue engineering approaches. The ideal scaffold is biocompatible and does not elicit a host inflammatory response. The selected scaffold in its composition and fabricated form must be capable of holding and supporting cells. In addition, the scaffold should be biodegradable, serving as a temporary support for such cells and mechanically augmenting the repaired tendon while allowing for eventual replacement by matrix components. Moreover, the scaffold should have high porosity and a large surface area. Furthermore, the material should mimic the native tendon extracellular matrix (ECM) architecture to allow cells to be distributed throughout the scaffold and to facilitate diffusion of nutrients and factors that promote cellular proliferation and ECM production.

Given the importance of glycosaminoglycans (GAGs) in supporting the reticular structure of the matrix, use of GAGs or GAG-analogues as components of a tendon tissue scaffold for enhancing repair is well documented.¹⁴ One such candidate is chitosan, a partially de-acetylated derivative of chitin found in arthropod exoskeletons. Structurally, chitosan shares some characteristics with various GAGs and hyaluronic acid.¹⁵ More specifically, chitosan is a linear polysaccharide composed of glucosamine and N-acetyl glucosamine units linked by β-glycosidic bonds. Investigators have studied the properties of chitosan, including its biocompatibility, biodegradability, antibacterial activity, mucoadhesivity, and wound healing.^16,17

One of the most promising features of chitosan is that it can be processed into porous structures for use in cell transplantation and tissue regeneration.^18,19 Porous chitosan structures can be formed by freezing and lyophilizing chitosan-acetic acid solutions; chondrogenic cell adhesion and proliferation onto these structures have been reported.^20,21 This chitosan scaffolding method has also been used to test different composites with collagens, gelatins, GAGs, and hyaluronic acid, all of which have also been proposed as useful 3-dimensional materials for tissue repair.²²

In the present study, we used chitosan matrix in RCT repair. We hypothesized that chitosan matrix could enhance rotator cuff repair the same way it enhances repair in epidermal tissues.¹⁶ Histologic findings demonstrated that the percentage of fibrous tissue was significantly higher in the chitosan-treated group than in the control group. This improved fibroblastic response may be attributed to the ability of chitosan to enhance cell migration and serve as a scaffold for repair. Other studies have indicated that chitin, of which chitosan is the primary derivative, accelerated the healing of skin and subcutaneous tissues by increased cell migration.²³ Moreover, Okamoto and colleagues²⁴ reported that chitin implants stimulated abundant angiogenesis through the same mechanism.

Inadequate initial strength of a repair may lead to a recurrent cuff tear or a disability of rotator cuff function in the early healing stages. In our study, the chitosan matrix tended to be absorbed by 6 weeks after surgery. Its adherence to and ultimate absorption at the repair site may be challenged by the flow of irrigation fluid through the subacromial space in the setting of arthroscopic surgery. However, because the chitosan remains in a more robust gel form, it is better able to resist being washed from the repair site. For augmentation, it may be possible to apply a biocompatible patch over the gel to further protect it from being dislodged. In addition, histologic findings showed that the fibrous repair tissue gradually increased until reaching a peak 8 weeks after surgery—an indication that the absorption rate of the chitosan scaffold lags behind full recovery of the repair tissue. Given this relationship, further studies are needed to determine the mechanical strength of the repair between 6 and 8 weeks, which is important for avoiding recurrent tears.

This study had a few limitations. First, as with any animal model, the anatomy and function of the rat shoulder differ from those of the human shoulder. The acromial arch differs in quadruped animals, with less coverage of the supraspinatus and more of the subscapularis.²⁵ These anatomical differences could yield altered stress mechanics that could affect tendon repair. Furthermore, rats and humans differ in their RCT healing rates. Thus, the pathophysiology of muscle atrophy and fat infiltration in rats may slightly differ from that in humans. In addition, no mechanical testing was performed to compare chitosan-treated and untreated rotator cuff repairs, and such testing is needed to clarify the biomechanical importance of augmentation. Furthermore, no immunohistochemical analysis was performed for collagen. In the repair of rotator cuff tendons, surgeons must consider not only the number of cells but also the production of ECM. Although not directly confirmed in this study, chitosan induced fibrous tissue proliferation that mirrored production of a large amount of collagen fibers. Last, we used an open RTC model. As an arthroscopic model was not used, no definitive conclusions can be drawn regarding use of chitosan in arthroscopy.

Conclusion

Use of chitosan as an acellular matrix improved formation of healing fibrous tissue, increased the number of cells, and prevented fatty atrophy and inflammatory aggregates inside repair sites while facilitating recovery of the natural pennation angle of the tissue. These results demonstrate that chitosan can enhance tendon healing in the setting of acute RCT. Further research, including biomechanical testing of repaired tendons, is needed to further delineate the utility of chitosan in regenerating irreparable RCTs.