Case Reports

CD30⁺ primary cutaneous lymphoproliferative disorders (pcLPDs) are the second most common cause of cutaneous T-cell lymphoma, accounting for approximately 25% to 30% of cases.¹ These disorders comprise a spectrum that includes primary cutaneous anaplastic large-cell lymphoma (pcALCL); lymphomatoid papulosis (LyP); and borderline lesions, which share clinicopathologic features of both pcALCL and LyP. Lymphomatoid papulosis is characterized as chronic, recurrent, papular or papulonodular skin lesions that typically are multifocal and regress spontaneously within weeks to months, only leaving small scars with atrophy and/or hyperpigmentation.² Cutaneous anaplastic large-cell lymphoma typically presents as solitary or grouped nodules or tumors that may undergo spontaneous partial or complete regression in approximately 25% of cases³ but often persist if not treated. Patients may have an array of lesions comprising the spectrum of CD30 pcLPDs.⁴

There is no curative therapy for CD30⁺ pcLPDs. Although active treatment is not necessary for LyP, low-dose methotrexate (MTX)(10–50 mg weekly) or phototherapy are the preferred initial suppressive therapies for symptomatic patients with scarring, facial lesions, or multiple symptomatic lesions.⁵ Observation with expectant follow-up is an option in pcALCL, though spontaneous regression is less likely than in LyP. For single or grouped pcALCL lesions, local radiation is the first-line therapy.⁶ Multifocal pcALCL lesions also can be treated with low-dose MTX,^2,5 as in LyP, or local radiation to selected areas. Although local radiotherapy is considered a first-line treatment in pcALCL, there is limited evidence on its clinical efficacy as well as the optimal dose and technique. We report the complete response of refractory pcALCL lesions to low-dose radiation while remaining on MTX weekly without any adverse effects.

Case Report

A 51-year-old woman presented with a 3-year history of CD30⁺ pcLPD manifesting primarily as pcALCL involving the head and neck, as well as LyP involving the head, arms, and trunk (T3N0M0). For 2 years her treatment regimen included clobetasol propionate cream 0.05% as needed for new lesions and 2 courses of standard-dose localized external beam radiation for larger pcALCL tumors on the right cheek and right side of the chin (Figure 1)(total dose for each course of treatment was 20 Gy and 36 Gy, respectively, each administered over 2–3 weeks). Because new unsightly papulonodules continued to develop on the patient’s face, she subsequently required low-dose oral MTX 30 mg once weekly for suppression of new lesions and was stable on this regimen for a year. However, she experienced an increase in LyP/pcALCL activity on the face during a 2-week break from MTX when she developed a herpes zoster infection on the right side of the forehead.

On physical examination 1 month later, 5 tiny pink papules scattered on the left eyebrow, left cheek, and left side of the chin were noted. She was advised to continue applying the clobetasol cream as needed and was restarted on MTX 10 mg once weekly. However, she developed 2 additional 1-cm nodules on the left side of the chin, neck, and shoulder. Methotrexate was increased to 30 mg once weekly over 2 weeks, which was the original dose prior to interruption, but the nodules grew to 1.5-cm in diameter. Due to their clinical appearance, the nodules were believed to be early pcALCL lesions (Figure 2A). Given the cosmetically sensitive location of the nodules, palliative radiotherapy was recommended rather than observe for possible regression. Based on a prior report by Neelis et al⁷ demonstrating efficacy of low-dose radiotherapy for cutaneous T-cell lymphoma and cutaneous B-cell lymphoma, we recommended starting with low radiation doses. Our patient was treated with 400 cGy twice to the left side of the chin and left side of the neck (800 cGy total at each site) while remaining on MTX 30 mg once weekly. This treatment was well tolerated without side effects and no evidence of radiation dermatitis. On follow-up examination 1 week later, the nodules had regressed and no new lesions were present (Figure 2B).

The patient has stayed on oral MTX and occasionally develops small lesions that quickly resolve with clobetasol cream. She has been followed for 3 years after radiotherapy and all 3 previously irradiated sites have remained recurrence free. Furthermore, she has not developed any new larger nodules or tumors and her MTX dose has been decreased to 15 mg once weekly.

 

Comment

Local radiotherapy is considered a first-line treatment of pcALCL; however, there is limited evidence on its clinical efficacy as well as the optimal dose and technique. Although no standard dose exists for pcALCL, the National Comprehensive Cancer Network guidelines⁸ recommend doses of 12 to 36 Gy in mycosis fungoides/Sézary syndrome subtypes of cutaneous T-cell lymphoma, which are consistent with guidelines published by the European Society for Medical Oncology.⁹ High complete response rates have been demonstrated in pcALCL at doses of 34 to 44 Gy⁶; however, lesions tend to recur elsewhere on the skin in 36% to 41% of patients despite treatment.^2,10 Lower doses of radiation therapy would provide several advantages over higher-dose therapy if a complete response could be achieved without greatly increasing the local recurrence rate. In cases of local recurrence, low-dose radiation would more easily permit retreatment of lesions compared to higher doses of radiation. Similarly, in patients with multifocal pcALCL, lower doses of radiotherapy may allow for treatment of larger skin areas while limiting potential treatment risks. Furthermore, low-dose therapy would allow for treatments to be delivered more quickly and with less inconvenience to the patient who is likely to need multiple future treatments to other areas. Low-dose radiation has been described with a favorable efficacy profile for mycosis fungoides^7,11 but has not been studied in patients with CD30⁺ pcLPDs.

Our case is notable because the patient remained on MTX during radiation therapy. Because MTX can act as a radiosensitizer, current standard practice is to stop MTX while administering radiation therapy. High oncologic doses of MTX can occasionally produce severe skin reactions at the irradiated site^12-14; however, reports of the use of lower weekly doses in combination with radiation are sparse. Our patient had refractory disfiguring facial lesions that would immediately recur once MTX was stopped. Low-dose radiation therapy was administered while she remained on her weekly MTX doses. With this treatment, she achieved a complete response, developed no radiation dermatitis, and had no further new skin lesions appear.

Conclusion

We reported the use of low-dose radiation therapy for the treatment of localized pcALCL in a patient who remained on low-dose oral MTX. Additional studies will be necessary to more fully evaluate the efficacy of using low-dose radiation both as monotherapy and in combination with MTX for pcALCL.

CD30⁺ primary cutaneous lymphoproliferative disorders (pcLPDs) are the second most common cause of cutaneous T-cell lymphoma, accounting for approximately 25% to 30% of cases.¹ These disorders comprise a spectrum that includes primary cutaneous anaplastic large-cell lymphoma (pcALCL); lymphomatoid papulosis (LyP); and borderline lesions, which share clinicopathologic features of both pcALCL and LyP. Lymphomatoid papulosis is characterized as chronic, recurrent, papular or papulonodular skin lesions that typically are multifocal and regress spontaneously within weeks to months, only leaving small scars with atrophy and/or hyperpigmentation.² Cutaneous anaplastic large-cell lymphoma typically presents as solitary or grouped nodules or tumors that may undergo spontaneous partial or complete regression in approximately 25% of cases³ but often persist if not treated. Patients may have an array of lesions comprising the spectrum of CD30 pcLPDs.⁴

There is no curative therapy for CD30⁺ pcLPDs. Although active treatment is not necessary for LyP, low-dose methotrexate (MTX)(10–50 mg weekly) or phototherapy are the preferred initial suppressive therapies for symptomatic patients with scarring, facial lesions, or multiple symptomatic lesions.⁵ Observation with expectant follow-up is an option in pcALCL, though spontaneous regression is less likely than in LyP. For single or grouped pcALCL lesions, local radiation is the first-line therapy.⁶ Multifocal pcALCL lesions also can be treated with low-dose MTX,^2,5 as in LyP, or local radiation to selected areas. Although local radiotherapy is considered a first-line treatment in pcALCL, there is limited evidence on its clinical efficacy as well as the optimal dose and technique. We report the complete response of refractory pcALCL lesions to low-dose radiation while remaining on MTX weekly without any adverse effects.

Case Report

A 51-year-old woman presented with a 3-year history of CD30⁺ pcLPD manifesting primarily as pcALCL involving the head and neck, as well as LyP involving the head, arms, and trunk (T3N0M0). For 2 years her treatment regimen included clobetasol propionate cream 0.05% as needed for new lesions and 2 courses of standard-dose localized external beam radiation for larger pcALCL tumors on the right cheek and right side of the chin (Figure 1)(total dose for each course of treatment was 20 Gy and 36 Gy, respectively, each administered over 2–3 weeks). Because new unsightly papulonodules continued to develop on the patient’s face, she subsequently required low-dose oral MTX 30 mg once weekly for suppression of new lesions and was stable on this regimen for a year. However, she experienced an increase in LyP/pcALCL activity on the face during a 2-week break from MTX when she developed a herpes zoster infection on the right side of the forehead.

On physical examination 1 month later, 5 tiny pink papules scattered on the left eyebrow, left cheek, and left side of the chin were noted. She was advised to continue applying the clobetasol cream as needed and was restarted on MTX 10 mg once weekly. However, she developed 2 additional 1-cm nodules on the left side of the chin, neck, and shoulder. Methotrexate was increased to 30 mg once weekly over 2 weeks, which was the original dose prior to interruption, but the nodules grew to 1.5-cm in diameter. Due to their clinical appearance, the nodules were believed to be early pcALCL lesions (Figure 2A). Given the cosmetically sensitive location of the nodules, palliative radiotherapy was recommended rather than observe for possible regression. Based on a prior report by Neelis et al⁷ demonstrating efficacy of low-dose radiotherapy for cutaneous T-cell lymphoma and cutaneous B-cell lymphoma, we recommended starting with low radiation doses. Our patient was treated with 400 cGy twice to the left side of the chin and left side of the neck (800 cGy total at each site) while remaining on MTX 30 mg once weekly. This treatment was well tolerated without side effects and no evidence of radiation dermatitis. On follow-up examination 1 week later, the nodules had regressed and no new lesions were present (Figure 2B).

The patient has stayed on oral MTX and occasionally develops small lesions that quickly resolve with clobetasol cream. She has been followed for 3 years after radiotherapy and all 3 previously irradiated sites have remained recurrence free. Furthermore, she has not developed any new larger nodules or tumors and her MTX dose has been decreased to 15 mg once weekly.

 

Comment

Local radiotherapy is considered a first-line treatment of pcALCL; however, there is limited evidence on its clinical efficacy as well as the optimal dose and technique. Although no standard dose exists for pcALCL, the National Comprehensive Cancer Network guidelines⁸ recommend doses of 12 to 36 Gy in mycosis fungoides/Sézary syndrome subtypes of cutaneous T-cell lymphoma, which are consistent with guidelines published by the European Society for Medical Oncology.⁹ High complete response rates have been demonstrated in pcALCL at doses of 34 to 44 Gy⁶; however, lesions tend to recur elsewhere on the skin in 36% to 41% of patients despite treatment.^2,10 Lower doses of radiation therapy would provide several advantages over higher-dose therapy if a complete response could be achieved without greatly increasing the local recurrence rate. In cases of local recurrence, low-dose radiation would more easily permit retreatment of lesions compared to higher doses of radiation. Similarly, in patients with multifocal pcALCL, lower doses of radiotherapy may allow for treatment of larger skin areas while limiting potential treatment risks. Furthermore, low-dose therapy would allow for treatments to be delivered more quickly and with less inconvenience to the patient who is likely to need multiple future treatments to other areas. Low-dose radiation has been described with a favorable efficacy profile for mycosis fungoides^7,11 but has not been studied in patients with CD30⁺ pcLPDs.

Our case is notable because the patient remained on MTX during radiation therapy. Because MTX can act as a radiosensitizer, current standard practice is to stop MTX while administering radiation therapy. High oncologic doses of MTX can occasionally produce severe skin reactions at the irradiated site^12-14; however, reports of the use of lower weekly doses in combination with radiation are sparse. Our patient had refractory disfiguring facial lesions that would immediately recur once MTX was stopped. Low-dose radiation therapy was administered while she remained on her weekly MTX doses. With this treatment, she achieved a complete response, developed no radiation dermatitis, and had no further new skin lesions appear.

Conclusion

We reported the use of low-dose radiation therapy for the treatment of localized pcALCL in a patient who remained on low-dose oral MTX. Additional studies will be necessary to more fully evaluate the efficacy of using low-dose radiation both as monotherapy and in combination with MTX for pcALCL.

CD30⁺ primary cutaneous lymphoproliferative disorders (pcLPDs) are the second most common cause of cutaneous T-cell lymphoma, accounting for approximately 25% to 30% of cases.¹ These disorders comprise a spectrum that includes primary cutaneous anaplastic large-cell lymphoma (pcALCL); lymphomatoid papulosis (LyP); and borderline lesions, which share clinicopathologic features of both pcALCL and LyP. Lymphomatoid papulosis is characterized as chronic, recurrent, papular or papulonodular skin lesions that typically are multifocal and regress spontaneously within weeks to months, only leaving small scars with atrophy and/or hyperpigmentation.² Cutaneous anaplastic large-cell lymphoma typically presents as solitary or grouped nodules or tumors that may undergo spontaneous partial or complete regression in approximately 25% of cases³ but often persist if not treated. Patients may have an array of lesions comprising the spectrum of CD30 pcLPDs.⁴

There is no curative therapy for CD30⁺ pcLPDs. Although active treatment is not necessary for LyP, low-dose methotrexate (MTX)(10–50 mg weekly) or phototherapy are the preferred initial suppressive therapies for symptomatic patients with scarring, facial lesions, or multiple symptomatic lesions.⁵ Observation with expectant follow-up is an option in pcALCL, though spontaneous regression is less likely than in LyP. For single or grouped pcALCL lesions, local radiation is the first-line therapy.⁶ Multifocal pcALCL lesions also can be treated with low-dose MTX,^2,5 as in LyP, or local radiation to selected areas. Although local radiotherapy is considered a first-line treatment in pcALCL, there is limited evidence on its clinical efficacy as well as the optimal dose and technique. We report the complete response of refractory pcALCL lesions to low-dose radiation while remaining on MTX weekly without any adverse effects.

Case Report

A 51-year-old woman presented with a 3-year history of CD30⁺ pcLPD manifesting primarily as pcALCL involving the head and neck, as well as LyP involving the head, arms, and trunk (T3N0M0). For 2 years her treatment regimen included clobetasol propionate cream 0.05% as needed for new lesions and 2 courses of standard-dose localized external beam radiation for larger pcALCL tumors on the right cheek and right side of the chin (Figure 1)(total dose for each course of treatment was 20 Gy and 36 Gy, respectively, each administered over 2–3 weeks). Because new unsightly papulonodules continued to develop on the patient’s face, she subsequently required low-dose oral MTX 30 mg once weekly for suppression of new lesions and was stable on this regimen for a year. However, she experienced an increase in LyP/pcALCL activity on the face during a 2-week break from MTX when she developed a herpes zoster infection on the right side of the forehead.

On physical examination 1 month later, 5 tiny pink papules scattered on the left eyebrow, left cheek, and left side of the chin were noted. She was advised to continue applying the clobetasol cream as needed and was restarted on MTX 10 mg once weekly. However, she developed 2 additional 1-cm nodules on the left side of the chin, neck, and shoulder. Methotrexate was increased to 30 mg once weekly over 2 weeks, which was the original dose prior to interruption, but the nodules grew to 1.5-cm in diameter. Due to their clinical appearance, the nodules were believed to be early pcALCL lesions (Figure 2A). Given the cosmetically sensitive location of the nodules, palliative radiotherapy was recommended rather than observe for possible regression. Based on a prior report by Neelis et al⁷ demonstrating efficacy of low-dose radiotherapy for cutaneous T-cell lymphoma and cutaneous B-cell lymphoma, we recommended starting with low radiation doses. Our patient was treated with 400 cGy twice to the left side of the chin and left side of the neck (800 cGy total at each site) while remaining on MTX 30 mg once weekly. This treatment was well tolerated without side effects and no evidence of radiation dermatitis. On follow-up examination 1 week later, the nodules had regressed and no new lesions were present (Figure 2B).

The patient has stayed on oral MTX and occasionally develops small lesions that quickly resolve with clobetasol cream. She has been followed for 3 years after radiotherapy and all 3 previously irradiated sites have remained recurrence free. Furthermore, she has not developed any new larger nodules or tumors and her MTX dose has been decreased to 15 mg once weekly.

 

Comment

Local radiotherapy is considered a first-line treatment of pcALCL; however, there is limited evidence on its clinical efficacy as well as the optimal dose and technique. Although no standard dose exists for pcALCL, the National Comprehensive Cancer Network guidelines⁸ recommend doses of 12 to 36 Gy in mycosis fungoides/Sézary syndrome subtypes of cutaneous T-cell lymphoma, which are consistent with guidelines published by the European Society for Medical Oncology.⁹ High complete response rates have been demonstrated in pcALCL at doses of 34 to 44 Gy⁶; however, lesions tend to recur elsewhere on the skin in 36% to 41% of patients despite treatment.^2,10 Lower doses of radiation therapy would provide several advantages over higher-dose therapy if a complete response could be achieved without greatly increasing the local recurrence rate. In cases of local recurrence, low-dose radiation would more easily permit retreatment of lesions compared to higher doses of radiation. Similarly, in patients with multifocal pcALCL, lower doses of radiotherapy may allow for treatment of larger skin areas while limiting potential treatment risks. Furthermore, low-dose therapy would allow for treatments to be delivered more quickly and with less inconvenience to the patient who is likely to need multiple future treatments to other areas. Low-dose radiation has been described with a favorable efficacy profile for mycosis fungoides^7,11 but has not been studied in patients with CD30⁺ pcLPDs.

Our case is notable because the patient remained on MTX during radiation therapy. Because MTX can act as a radiosensitizer, current standard practice is to stop MTX while administering radiation therapy. High oncologic doses of MTX can occasionally produce severe skin reactions at the irradiated site^12-14; however, reports of the use of lower weekly doses in combination with radiation are sparse. Our patient had refractory disfiguring facial lesions that would immediately recur once MTX was stopped. Low-dose radiation therapy was administered while she remained on her weekly MTX doses. With this treatment, she achieved a complete response, developed no radiation dermatitis, and had no further new skin lesions appear.

Conclusion

We reported the use of low-dose radiation therapy for the treatment of localized pcALCL in a patient who remained on low-dose oral MTX. Additional studies will be necessary to more fully evaluate the efficacy of using low-dose radiation both as monotherapy and in combination with MTX for pcALCL.

Bullous pemphigoid (BP) is an acquired, autoimmune, subepidermal blistering disease that is more common in elderly patients.¹ An association with internal neoplasms and BP has been established; however, there is debate regarding the precise nature of the relationship.² Several gastrointestinal tract tumors have been associated with BP, including adenocarcinoma of the colon, adenosquamous cell carcinoma and adenocarcinoma of the stomach, adenocarcinoma of the rectum, and liver and bile duct malignancies.^3-5 Association with pancreatic neoplasms (eg, carcinoma of the pancreas) rarely has been reported.^5-7 We present an unusual case of a lymphoepithelial cyst of the pancreas in a patient with BP.

Case Report

A 67-year-old man presented with erythematous crusted plaques and pink scars over the chest, back, arms, and legs (Figure 1). A 1.5-cm tense bullous lesion was observed on the right knee. The patient’s medical history was notable for biopsy-proven BP of 8 months’ duration as well as diabetes mellitus and hypothyroidism. The patient was being followed by his surgeon for a 1.9-cm soft-tissue lesion in the pancreatic tail and was awaiting surgical excision at the time of the current presentation. The pancreatic lesion was discovered incidentally on magnetic resonance imaging performed following urologic concerns. At the current presentation, the patient’s medications included nifedipine, hydralazine, metoprolol, torsemide, aspirin, levothyroxine, atorvastatin, insulin lispro, and insulin glargine. He had previously been treated for BP with prednisone at a maximum dosage of 60 mg daily, clobetasol propionate cream 0.05%, and mupirocin ointment 2% without improvement. Because of substantial weight gain and poorly controlled diabetes, prednisone was discontinued.

Bullous pemphigoid had been diagnosed histopathologically by a prior dermatologist. Hematoxylin and eosin staining demonstrated a subepidermal separation with eosinophils within the perivascular infiltrate (Figure 2). Direct immunofluorescence was noted in a linear pattern at the dermoepidermal junction with IgG and C3. Bullous pemphigoid antigen antibodies 1 and 2 were obtained via enzyme-linked immunosorbent assay with a positive BP-1 antigen antibody of 19 U/mL (positive, >15 U/mL) and a normal BP-2 antigen antibody of less than 9 U/mL (reference range, <9 U/mL). The glucagon level was elevated at 245 pg/mL (reference range, ≤134 pg/mL).

The patient was prescribed minocycline 100 mg twice daily and niacinamide 500 mg 3 times daily. Topical treatment with clobetasol and mupirocin was continued. One month later, the patient returned with an increase in disease activity. Changes to his therapeutic regimen were deferred until after excision of the pancreatic lesion based on the decision not to start immunosuppressive therapy until the precise nature of the pancreatic lesion was determined.

The patient underwent excision of the pancreatic lesion approximately 3 months later, which proved to be a benign lymphoepithelial cyst of the pancreas. Histology of the cyst consisted of dense fibrous tissue with a squamous epithelial lining focally infiltrated by lymphocytes (Figure 3A). Immunoperoxidase staining of the cyst revealed focal linear areas of C3d staining along the basement membrane of the stratified squamous epithelium (Figure 3B).

The patient stated that his skin started to improve virtually immediately following the excision without systemic treatment for BP. On follow-up examination 3 weeks postoperatively, no bullae were observed and there was a notable decrease in erythematous crusted plaques (Figure 4).

Comment

Paraneoplastic BP has been documented; however, lymphoepithelial cysts of the pancreas in association with BP are rare. We propose that the lymphoepithelial cyst of the pancreas provided the immunologic stimulus for the development of cutaneous BP based on the observation that our patient’s condition remarkably improved with resection of the tumor.

There are fewer than 100 cases of lymphoepithelial cysts of the pancreas reported in the literature.⁸ The histologic appearance is consistent with a true cyst exhibiting a well-differentiated stratified squamous epithelium, often with keratinization, surrounded by lymphoid tissue. These tumors are most commonly seen in middle-aged men and are frequently found incidentally,^8-10 as was the case with our patient. Although histologically similar, lymphoepithelial cysts of the pancreas are considered distinct from lymphoepithelial cysts of the parotid gland or head and neck region.¹⁰ Lymphoepithelial cysts of the pancreas are unrelated to elevated glucagon levels; it is likely that our patient’s glucagon levels were associated with his history of diabetes.¹¹

 

The diagnosis of BP is characteristically confirmed by direct immunofluorescence. Although it was performed for our patient’s cutaneous lesions, it was not obtained for the lymphoepithelial cyst of the pancreas. Once the diagnosis of the lymphoepithelial cyst of the pancreas was established, as direct immunofluorescence could not be performed in formalin-fixed tissue, immunoperoxidase staining with C3d was obtained. C3 has a well-established role in activation of complement and as a marker in BP. Deposition of C3d is a result of deactivation of C3b, a cleavage product of C3. In a study of 6 autoimmune blistering disorders that included 32 patients with BP, Pfaltz et al¹² found positive immunoperoxidase staining for C3d in 31 of 32 patients, which translated to a sensitivity of 97%, a positive predictive value of 100%, and a negative predictive value of 98% among the blistering diseases being studied. Similarly, Magro and Dyrsen¹³ had positive staining of C3d in 17 of 17 (100%) patients with BP.

In theory, any process that involves deposition of C3 should be positive for C3d on immunoperoxidase staining. Other dermatologic inflammatory conditions stain positively with C3d, such as systemic lupus erythematosus, discoid lupus erythematosus, subacute cutaneous lupus erythematosus, and dermatomysositis.¹³ The staining for these diseases correlates with the site of the associated inflammatory component seen on hematoxylin and eosin staining. The staining of C3d along the basement membrane of stratified squamous epithelium in the lymphoepithelial cyst of the pancreas seen in our patient closely resembles the staining seen in cutaneous BP.

A proposed mechanism for BP in our patient would be exposure of BP-1 antigen in the pancreatic cyst leading to antibody recognition and C3 deposition along the basement membrane in the cyst, as evidenced by C3d immunoperoxidase staining. The IgG and C3 deposition along the cutaneous basement membrane would then represent a systemic response to the antigen exposure in the cyst. Thus, the lymphoepithelial cyst provided the immunologic stimulus for the development of the cutaneous BP. This theory is based on the observation of our patient’s rapid improvement without a change in his treatment regimen immediately after surgical excision of the cyst.

Despite the plausibility of our hypothesis, several questions remain regarding the validity of our assumptions. Although sensitive for C3 deposition, C3d immunoperoxidase staining is not specific for BP. If the proposed mechanism for causation is true, one might have expected that a subepithelial cleft along the basement membrane of the pancreatic cyst would be observed, which was not seen. A repeat BP antigen antibody was not obtained, which would have been helpful in determining if there was clearance of the antibody that would have correlated with the clinical resolution of the BP lesions.

 

Conclusion

Our case suggests that paraneoplastic BP is a genuine entity. Indeed, the primary tumor itself may be the immunologic stimulus in the development of BP. Recalcitrant BP should raise the question of a neoplastic process that is exposing the BP antigen. If a thorough review of systems accompanied by corroborating laboratory studies suggests a neoplastic process, the suspect lesion should be further evaluated and surgically excised if clinically indicated. Further evaluation of neoplasms with advanced staining methods may aid in establishing the causative nature of tumors in the development of BP.

Acknowledgments

We are grateful to John Stanley, MD, and Aimee Payne, MD (both from Philadelphia, Pennsylvania), for theirinsights into this case.

Bullous pemphigoid (BP) is an acquired, autoimmune, subepidermal blistering disease that is more common in elderly patients.¹ An association with internal neoplasms and BP has been established; however, there is debate regarding the precise nature of the relationship.² Several gastrointestinal tract tumors have been associated with BP, including adenocarcinoma of the colon, adenosquamous cell carcinoma and adenocarcinoma of the stomach, adenocarcinoma of the rectum, and liver and bile duct malignancies.^3-5 Association with pancreatic neoplasms (eg, carcinoma of the pancreas) rarely has been reported.^5-7 We present an unusual case of a lymphoepithelial cyst of the pancreas in a patient with BP.

Case Report

A 67-year-old man presented with erythematous crusted plaques and pink scars over the chest, back, arms, and legs (Figure 1). A 1.5-cm tense bullous lesion was observed on the right knee. The patient’s medical history was notable for biopsy-proven BP of 8 months’ duration as well as diabetes mellitus and hypothyroidism. The patient was being followed by his surgeon for a 1.9-cm soft-tissue lesion in the pancreatic tail and was awaiting surgical excision at the time of the current presentation. The pancreatic lesion was discovered incidentally on magnetic resonance imaging performed following urologic concerns. At the current presentation, the patient’s medications included nifedipine, hydralazine, metoprolol, torsemide, aspirin, levothyroxine, atorvastatin, insulin lispro, and insulin glargine. He had previously been treated for BP with prednisone at a maximum dosage of 60 mg daily, clobetasol propionate cream 0.05%, and mupirocin ointment 2% without improvement. Because of substantial weight gain and poorly controlled diabetes, prednisone was discontinued.

Bullous pemphigoid had been diagnosed histopathologically by a prior dermatologist. Hematoxylin and eosin staining demonstrated a subepidermal separation with eosinophils within the perivascular infiltrate (Figure 2). Direct immunofluorescence was noted in a linear pattern at the dermoepidermal junction with IgG and C3. Bullous pemphigoid antigen antibodies 1 and 2 were obtained via enzyme-linked immunosorbent assay with a positive BP-1 antigen antibody of 19 U/mL (positive, >15 U/mL) and a normal BP-2 antigen antibody of less than 9 U/mL (reference range, <9 U/mL). The glucagon level was elevated at 245 pg/mL (reference range, ≤134 pg/mL).

The patient was prescribed minocycline 100 mg twice daily and niacinamide 500 mg 3 times daily. Topical treatment with clobetasol and mupirocin was continued. One month later, the patient returned with an increase in disease activity. Changes to his therapeutic regimen were deferred until after excision of the pancreatic lesion based on the decision not to start immunosuppressive therapy until the precise nature of the pancreatic lesion was determined.

The patient underwent excision of the pancreatic lesion approximately 3 months later, which proved to be a benign lymphoepithelial cyst of the pancreas. Histology of the cyst consisted of dense fibrous tissue with a squamous epithelial lining focally infiltrated by lymphocytes (Figure 3A). Immunoperoxidase staining of the cyst revealed focal linear areas of C3d staining along the basement membrane of the stratified squamous epithelium (Figure 3B).

The patient stated that his skin started to improve virtually immediately following the excision without systemic treatment for BP. On follow-up examination 3 weeks postoperatively, no bullae were observed and there was a notable decrease in erythematous crusted plaques (Figure 4).

Comment

Paraneoplastic BP has been documented; however, lymphoepithelial cysts of the pancreas in association with BP are rare. We propose that the lymphoepithelial cyst of the pancreas provided the immunologic stimulus for the development of cutaneous BP based on the observation that our patient’s condition remarkably improved with resection of the tumor.

There are fewer than 100 cases of lymphoepithelial cysts of the pancreas reported in the literature.⁸ The histologic appearance is consistent with a true cyst exhibiting a well-differentiated stratified squamous epithelium, often with keratinization, surrounded by lymphoid tissue. These tumors are most commonly seen in middle-aged men and are frequently found incidentally,^8-10 as was the case with our patient. Although histologically similar, lymphoepithelial cysts of the pancreas are considered distinct from lymphoepithelial cysts of the parotid gland or head and neck region.¹⁰ Lymphoepithelial cysts of the pancreas are unrelated to elevated glucagon levels; it is likely that our patient’s glucagon levels were associated with his history of diabetes.¹¹

 

The diagnosis of BP is characteristically confirmed by direct immunofluorescence. Although it was performed for our patient’s cutaneous lesions, it was not obtained for the lymphoepithelial cyst of the pancreas. Once the diagnosis of the lymphoepithelial cyst of the pancreas was established, as direct immunofluorescence could not be performed in formalin-fixed tissue, immunoperoxidase staining with C3d was obtained. C3 has a well-established role in activation of complement and as a marker in BP. Deposition of C3d is a result of deactivation of C3b, a cleavage product of C3. In a study of 6 autoimmune blistering disorders that included 32 patients with BP, Pfaltz et al¹² found positive immunoperoxidase staining for C3d in 31 of 32 patients, which translated to a sensitivity of 97%, a positive predictive value of 100%, and a negative predictive value of 98% among the blistering diseases being studied. Similarly, Magro and Dyrsen¹³ had positive staining of C3d in 17 of 17 (100%) patients with BP.

In theory, any process that involves deposition of C3 should be positive for C3d on immunoperoxidase staining. Other dermatologic inflammatory conditions stain positively with C3d, such as systemic lupus erythematosus, discoid lupus erythematosus, subacute cutaneous lupus erythematosus, and dermatomysositis.¹³ The staining for these diseases correlates with the site of the associated inflammatory component seen on hematoxylin and eosin staining. The staining of C3d along the basement membrane of stratified squamous epithelium in the lymphoepithelial cyst of the pancreas seen in our patient closely resembles the staining seen in cutaneous BP.

A proposed mechanism for BP in our patient would be exposure of BP-1 antigen in the pancreatic cyst leading to antibody recognition and C3 deposition along the basement membrane in the cyst, as evidenced by C3d immunoperoxidase staining. The IgG and C3 deposition along the cutaneous basement membrane would then represent a systemic response to the antigen exposure in the cyst. Thus, the lymphoepithelial cyst provided the immunologic stimulus for the development of the cutaneous BP. This theory is based on the observation of our patient’s rapid improvement without a change in his treatment regimen immediately after surgical excision of the cyst.

Despite the plausibility of our hypothesis, several questions remain regarding the validity of our assumptions. Although sensitive for C3 deposition, C3d immunoperoxidase staining is not specific for BP. If the proposed mechanism for causation is true, one might have expected that a subepithelial cleft along the basement membrane of the pancreatic cyst would be observed, which was not seen. A repeat BP antigen antibody was not obtained, which would have been helpful in determining if there was clearance of the antibody that would have correlated with the clinical resolution of the BP lesions.

 

Conclusion

Our case suggests that paraneoplastic BP is a genuine entity. Indeed, the primary tumor itself may be the immunologic stimulus in the development of BP. Recalcitrant BP should raise the question of a neoplastic process that is exposing the BP antigen. If a thorough review of systems accompanied by corroborating laboratory studies suggests a neoplastic process, the suspect lesion should be further evaluated and surgically excised if clinically indicated. Further evaluation of neoplasms with advanced staining methods may aid in establishing the causative nature of tumors in the development of BP.

Acknowledgments

We are grateful to John Stanley, MD, and Aimee Payne, MD (both from Philadelphia, Pennsylvania), for theirinsights into this case.

Bullous pemphigoid (BP) is an acquired, autoimmune, subepidermal blistering disease that is more common in elderly patients.¹ An association with internal neoplasms and BP has been established; however, there is debate regarding the precise nature of the relationship.² Several gastrointestinal tract tumors have been associated with BP, including adenocarcinoma of the colon, adenosquamous cell carcinoma and adenocarcinoma of the stomach, adenocarcinoma of the rectum, and liver and bile duct malignancies.^3-5 Association with pancreatic neoplasms (eg, carcinoma of the pancreas) rarely has been reported.^5-7 We present an unusual case of a lymphoepithelial cyst of the pancreas in a patient with BP.

Case Report

A 67-year-old man presented with erythematous crusted plaques and pink scars over the chest, back, arms, and legs (Figure 1). A 1.5-cm tense bullous lesion was observed on the right knee. The patient’s medical history was notable for biopsy-proven BP of 8 months’ duration as well as diabetes mellitus and hypothyroidism. The patient was being followed by his surgeon for a 1.9-cm soft-tissue lesion in the pancreatic tail and was awaiting surgical excision at the time of the current presentation. The pancreatic lesion was discovered incidentally on magnetic resonance imaging performed following urologic concerns. At the current presentation, the patient’s medications included nifedipine, hydralazine, metoprolol, torsemide, aspirin, levothyroxine, atorvastatin, insulin lispro, and insulin glargine. He had previously been treated for BP with prednisone at a maximum dosage of 60 mg daily, clobetasol propionate cream 0.05%, and mupirocin ointment 2% without improvement. Because of substantial weight gain and poorly controlled diabetes, prednisone was discontinued.

Bullous pemphigoid had been diagnosed histopathologically by a prior dermatologist. Hematoxylin and eosin staining demonstrated a subepidermal separation with eosinophils within the perivascular infiltrate (Figure 2). Direct immunofluorescence was noted in a linear pattern at the dermoepidermal junction with IgG and C3. Bullous pemphigoid antigen antibodies 1 and 2 were obtained via enzyme-linked immunosorbent assay with a positive BP-1 antigen antibody of 19 U/mL (positive, >15 U/mL) and a normal BP-2 antigen antibody of less than 9 U/mL (reference range, <9 U/mL). The glucagon level was elevated at 245 pg/mL (reference range, ≤134 pg/mL).

The patient was prescribed minocycline 100 mg twice daily and niacinamide 500 mg 3 times daily. Topical treatment with clobetasol and mupirocin was continued. One month later, the patient returned with an increase in disease activity. Changes to his therapeutic regimen were deferred until after excision of the pancreatic lesion based on the decision not to start immunosuppressive therapy until the precise nature of the pancreatic lesion was determined.

The patient underwent excision of the pancreatic lesion approximately 3 months later, which proved to be a benign lymphoepithelial cyst of the pancreas. Histology of the cyst consisted of dense fibrous tissue with a squamous epithelial lining focally infiltrated by lymphocytes (Figure 3A). Immunoperoxidase staining of the cyst revealed focal linear areas of C3d staining along the basement membrane of the stratified squamous epithelium (Figure 3B).

The patient stated that his skin started to improve virtually immediately following the excision without systemic treatment for BP. On follow-up examination 3 weeks postoperatively, no bullae were observed and there was a notable decrease in erythematous crusted plaques (Figure 4).

Comment

Paraneoplastic BP has been documented; however, lymphoepithelial cysts of the pancreas in association with BP are rare. We propose that the lymphoepithelial cyst of the pancreas provided the immunologic stimulus for the development of cutaneous BP based on the observation that our patient’s condition remarkably improved with resection of the tumor.

There are fewer than 100 cases of lymphoepithelial cysts of the pancreas reported in the literature.⁸ The histologic appearance is consistent with a true cyst exhibiting a well-differentiated stratified squamous epithelium, often with keratinization, surrounded by lymphoid tissue. These tumors are most commonly seen in middle-aged men and are frequently found incidentally,^8-10 as was the case with our patient. Although histologically similar, lymphoepithelial cysts of the pancreas are considered distinct from lymphoepithelial cysts of the parotid gland or head and neck region.¹⁰ Lymphoepithelial cysts of the pancreas are unrelated to elevated glucagon levels; it is likely that our patient’s glucagon levels were associated with his history of diabetes.¹¹

 

The diagnosis of BP is characteristically confirmed by direct immunofluorescence. Although it was performed for our patient’s cutaneous lesions, it was not obtained for the lymphoepithelial cyst of the pancreas. Once the diagnosis of the lymphoepithelial cyst of the pancreas was established, as direct immunofluorescence could not be performed in formalin-fixed tissue, immunoperoxidase staining with C3d was obtained. C3 has a well-established role in activation of complement and as a marker in BP. Deposition of C3d is a result of deactivation of C3b, a cleavage product of C3. In a study of 6 autoimmune blistering disorders that included 32 patients with BP, Pfaltz et al¹² found positive immunoperoxidase staining for C3d in 31 of 32 patients, which translated to a sensitivity of 97%, a positive predictive value of 100%, and a negative predictive value of 98% among the blistering diseases being studied. Similarly, Magro and Dyrsen¹³ had positive staining of C3d in 17 of 17 (100%) patients with BP.

In theory, any process that involves deposition of C3 should be positive for C3d on immunoperoxidase staining. Other dermatologic inflammatory conditions stain positively with C3d, such as systemic lupus erythematosus, discoid lupus erythematosus, subacute cutaneous lupus erythematosus, and dermatomysositis.¹³ The staining for these diseases correlates with the site of the associated inflammatory component seen on hematoxylin and eosin staining. The staining of C3d along the basement membrane of stratified squamous epithelium in the lymphoepithelial cyst of the pancreas seen in our patient closely resembles the staining seen in cutaneous BP.

A proposed mechanism for BP in our patient would be exposure of BP-1 antigen in the pancreatic cyst leading to antibody recognition and C3 deposition along the basement membrane in the cyst, as evidenced by C3d immunoperoxidase staining. The IgG and C3 deposition along the cutaneous basement membrane would then represent a systemic response to the antigen exposure in the cyst. Thus, the lymphoepithelial cyst provided the immunologic stimulus for the development of the cutaneous BP. This theory is based on the observation of our patient’s rapid improvement without a change in his treatment regimen immediately after surgical excision of the cyst.

Despite the plausibility of our hypothesis, several questions remain regarding the validity of our assumptions. Although sensitive for C3 deposition, C3d immunoperoxidase staining is not specific for BP. If the proposed mechanism for causation is true, one might have expected that a subepithelial cleft along the basement membrane of the pancreatic cyst would be observed, which was not seen. A repeat BP antigen antibody was not obtained, which would have been helpful in determining if there was clearance of the antibody that would have correlated with the clinical resolution of the BP lesions.

 

Conclusion

Our case suggests that paraneoplastic BP is a genuine entity. Indeed, the primary tumor itself may be the immunologic stimulus in the development of BP. Recalcitrant BP should raise the question of a neoplastic process that is exposing the BP antigen. If a thorough review of systems accompanied by corroborating laboratory studies suggests a neoplastic process, the suspect lesion should be further evaluated and surgically excised if clinically indicated. Further evaluation of neoplasms with advanced staining methods may aid in establishing the causative nature of tumors in the development of BP.

Acknowledgments

We are grateful to John Stanley, MD, and Aimee Payne, MD (both from Philadelphia, Pennsylvania), for theirinsights into this case.

THE CASE

A 49-year-old woman visited our family medicine clinic because she’d had 3 episodes of epistaxis during the previous month. She’d already visited the emergency department, and the doctor there had treated her symptomatically and referred her to our clinic.

On physical examination, we noted a whitish mass in the patient’s right nostril that was attached to the nasal septum. The patient’s vital signs were within normal limits. She had a history of hypertension, depression, anxiety, gastroesophageal reflux disease, and post-traumatic stress disorder. Her medications included amlodipine-benazepril, atenolol-chlorthalidone, citalopram, clonazepam, prazosin, and omeprazole. The patient lived alone and denied using tobacco or illicit drugs, but she drank one to 2 glasses of brandy every day. She denied any past medical or family history of similar complaints, autoimmune disorders, or skin rashes.

A complete blood count, international normalized ratio, sedimentation rate, anti-nuclear antibody test, and an anti-neutrophil cytoplasmic antibody panel were normal.

THE DIAGNOSIS

We referred the patient to an ear, nose, and throat doctor for a nasal endoscopy and a biopsy, which showed granulation tissue. A maxillofacial computed tomography (CT) scan revealed a 1.44 cm x 0.8 cm polypoid soft tissue mass in the right nasal cavity adherent to the nasal septum with no posterior extension (FIGURE 1).

The patient underwent cauterization of the mass and was sent home with a prescription for fluticasone spray and a referral to a head and neck oncologist. Excision and biopsy of the mass revealed pyogenic granuloma (lobular capillary hemangioma) with surface ulceration (FIGURES 2 AND 3).

 

 

DISCUSSION

Pyogenic granuloma (PG) is a benign vascular tumor of the skin and mucous membranes that is not associated with an infection. Rather, it is a hyperplastic, neovascular, inflammatory response to an angiogenic stimulus. Several enhancers and inhibitors of angiogenesis have been shown to play a role in PG, including hormones, medications, and local injury. In fact, a local injury or hormonal factor is identified as a stimulus in more than half of PG patients.¹

The hormone connection. Estrogen promotes production of nerve growth factor, granulocyte-macrophage colony-stimulating factor, basic fibroblast growth factor, vascular endothelial growth factor, and transforming growth factor beta 1. Progesterone enhances inflammatory mediators as well. Although there are no direct receptors for estrogen and progesterone in the oral and nasal mucosa, some of these pro-inflammatory effects create an environment conducive to the development of PG. This is supported by several studies documenting an increased incidence of PGs with oral contraceptive use and regression of PGs after childbirth.^2-4

In more than half of patients with pyogenic granuloma, a local injury or hormonal factor is at work.

Medication may play a role. Drug-induced PG has also been described in several studies.^5,6 Offending medications include systemic and topical retinoids, capecitabine, etoposide, 5-fluorouracil, cyclosporine, docetaxel, and human immunodeficiency virus protease inhibitors.

Local injury may also be a culprit. Nasal PGs are commonly attached to the anterior septum and typically result from nasal packing, habitual picking, or nose boring.⁷ In this particular case, however, we were unable to identify the irritant.

The classic presentation

PG classically presents as a painless mass that spontaneously develops over days to weeks. The mass can be sessile or pedunculated, and is frequently hemorrhagic. Intranasal PG usually presents with epiphora.⁷ While the prevalence of intraoral PG was found to be one in 25,000 individuals³, data for nasal lesions is scarce. Most cases of PG are seen in the second and third decades of life.^1,3 In children, PG is slightly more predominant in males.^1,3 Mucosal lesions, however, have a higher incidence in females.^1,3 Granuloma gravidarum, the term used to describe mucosal PG in pregnant females, was found in 0.2% to 5% of pregnancies.^2,3,8

 

Differential Dx includes warts, squamous cell carcinoma

The differential diagnosis of PG includes Spitz nevus, glomus tumors, common warts, amelanotic melanoma, squamous cell carcinoma, basal cell carcinoma, Kaposi’s sarcoma, bacillary angiomatosis, infantile hemangioma, and angiolymphoid hyperplasia, among others.^3,5 Foreign bodies, nasal polyps, angiofibroma, meningocele, Wegener’s granulomatosis, and sarcoidosis should also be considered.

Radiologic evaluation may be beneficial—especially with nasal lesions—when looking for findings suggestive of malignancy. Both CT and magnetic resonance imaging with contrast identify PG as a soft tissue mass with lobulated contours,^9,10 but histopathologic analysis is required to confirm the diagnosis. The histopathologic appearance of PG is characterized by a polypoid lesion with circumscribed anastomosing networks of capillaries arranged in one or more lobules at the base in an edematous and fibroblastic stroma.

Alternatives to surgery include electrocautery, cryotherapy, laser therapy, and intralesional and topical agents, but the recurrence rate with these is often higher.

Treatment is determined by the location and size of the lesion

The most suitable treatment is determined by considering the location of the lesion, the characteristics of the lesion (morphology/size), its amenability to surgery, risk of scar formation, and the presence or absence of a causative irritant. Excision is often preferred because it yields a specimen for pathologic analysis. Alternative treatments include electrocautery, cryotherapy, laser therapy, and intralesional and topical agents,^3,6,7 but the recurrence rate is higher (up to 15%) with some of these modalities, when compared with excision (3.6%).³

Our patient underwent excision of the mass and was seen for an annual follow-up appointment. All of her symptoms resolved and no recurrence was noted.

THE TAKEAWAY

Although PG is a common and benign condition, it is rarely seen in the nasal cavity without an obvious history of a possible irritant. PG should be considered as a diagnosis for rapidly growing cutaneous or mucosal hemorrhagic lesions. Appropriate tissue pathology is essential to rule out malignancy and other serious conditions, such as bacillary angiomatosis and Wegener’s granulomatosis.

Treatment is usually required to avoid the frequent complications of ulceration and bleeding. Surgical treatments are preferred. The location of the lesion largely determines whether referral to a specialist is necessary.

THE CASE

A 49-year-old woman visited our family medicine clinic because she’d had 3 episodes of epistaxis during the previous month. She’d already visited the emergency department, and the doctor there had treated her symptomatically and referred her to our clinic.

On physical examination, we noted a whitish mass in the patient’s right nostril that was attached to the nasal septum. The patient’s vital signs were within normal limits. She had a history of hypertension, depression, anxiety, gastroesophageal reflux disease, and post-traumatic stress disorder. Her medications included amlodipine-benazepril, atenolol-chlorthalidone, citalopram, clonazepam, prazosin, and omeprazole. The patient lived alone and denied using tobacco or illicit drugs, but she drank one to 2 glasses of brandy every day. She denied any past medical or family history of similar complaints, autoimmune disorders, or skin rashes.

A complete blood count, international normalized ratio, sedimentation rate, anti-nuclear antibody test, and an anti-neutrophil cytoplasmic antibody panel were normal.

THE DIAGNOSIS

We referred the patient to an ear, nose, and throat doctor for a nasal endoscopy and a biopsy, which showed granulation tissue. A maxillofacial computed tomography (CT) scan revealed a 1.44 cm x 0.8 cm polypoid soft tissue mass in the right nasal cavity adherent to the nasal septum with no posterior extension (FIGURE 1).

The patient underwent cauterization of the mass and was sent home with a prescription for fluticasone spray and a referral to a head and neck oncologist. Excision and biopsy of the mass revealed pyogenic granuloma (lobular capillary hemangioma) with surface ulceration (FIGURES 2 AND 3).

 

 

DISCUSSION

Pyogenic granuloma (PG) is a benign vascular tumor of the skin and mucous membranes that is not associated with an infection. Rather, it is a hyperplastic, neovascular, inflammatory response to an angiogenic stimulus. Several enhancers and inhibitors of angiogenesis have been shown to play a role in PG, including hormones, medications, and local injury. In fact, a local injury or hormonal factor is identified as a stimulus in more than half of PG patients.¹

The hormone connection. Estrogen promotes production of nerve growth factor, granulocyte-macrophage colony-stimulating factor, basic fibroblast growth factor, vascular endothelial growth factor, and transforming growth factor beta 1. Progesterone enhances inflammatory mediators as well. Although there are no direct receptors for estrogen and progesterone in the oral and nasal mucosa, some of these pro-inflammatory effects create an environment conducive to the development of PG. This is supported by several studies documenting an increased incidence of PGs with oral contraceptive use and regression of PGs after childbirth.^2-4

In more than half of patients with pyogenic granuloma, a local injury or hormonal factor is at work.

Medication may play a role. Drug-induced PG has also been described in several studies.^5,6 Offending medications include systemic and topical retinoids, capecitabine, etoposide, 5-fluorouracil, cyclosporine, docetaxel, and human immunodeficiency virus protease inhibitors.

Local injury may also be a culprit. Nasal PGs are commonly attached to the anterior septum and typically result from nasal packing, habitual picking, or nose boring.⁷ In this particular case, however, we were unable to identify the irritant.

The classic presentation

PG classically presents as a painless mass that spontaneously develops over days to weeks. The mass can be sessile or pedunculated, and is frequently hemorrhagic. Intranasal PG usually presents with epiphora.⁷ While the prevalence of intraoral PG was found to be one in 25,000 individuals³, data for nasal lesions is scarce. Most cases of PG are seen in the second and third decades of life.^1,3 In children, PG is slightly more predominant in males.^1,3 Mucosal lesions, however, have a higher incidence in females.^1,3 Granuloma gravidarum, the term used to describe mucosal PG in pregnant females, was found in 0.2% to 5% of pregnancies.^2,3,8

 

Differential Dx includes warts, squamous cell carcinoma

The differential diagnosis of PG includes Spitz nevus, glomus tumors, common warts, amelanotic melanoma, squamous cell carcinoma, basal cell carcinoma, Kaposi’s sarcoma, bacillary angiomatosis, infantile hemangioma, and angiolymphoid hyperplasia, among others.^3,5 Foreign bodies, nasal polyps, angiofibroma, meningocele, Wegener’s granulomatosis, and sarcoidosis should also be considered.

Radiologic evaluation may be beneficial—especially with nasal lesions—when looking for findings suggestive of malignancy. Both CT and magnetic resonance imaging with contrast identify PG as a soft tissue mass with lobulated contours,^9,10 but histopathologic analysis is required to confirm the diagnosis. The histopathologic appearance of PG is characterized by a polypoid lesion with circumscribed anastomosing networks of capillaries arranged in one or more lobules at the base in an edematous and fibroblastic stroma.

Alternatives to surgery include electrocautery, cryotherapy, laser therapy, and intralesional and topical agents, but the recurrence rate with these is often higher.

Treatment is determined by the location and size of the lesion

The most suitable treatment is determined by considering the location of the lesion, the characteristics of the lesion (morphology/size), its amenability to surgery, risk of scar formation, and the presence or absence of a causative irritant. Excision is often preferred because it yields a specimen for pathologic analysis. Alternative treatments include electrocautery, cryotherapy, laser therapy, and intralesional and topical agents,^3,6,7 but the recurrence rate is higher (up to 15%) with some of these modalities, when compared with excision (3.6%).³

Our patient underwent excision of the mass and was seen for an annual follow-up appointment. All of her symptoms resolved and no recurrence was noted.

THE TAKEAWAY

Although PG is a common and benign condition, it is rarely seen in the nasal cavity without an obvious history of a possible irritant. PG should be considered as a diagnosis for rapidly growing cutaneous or mucosal hemorrhagic lesions. Appropriate tissue pathology is essential to rule out malignancy and other serious conditions, such as bacillary angiomatosis and Wegener’s granulomatosis.

Treatment is usually required to avoid the frequent complications of ulceration and bleeding. Surgical treatments are preferred. The location of the lesion largely determines whether referral to a specialist is necessary.

THE CASE

A 49-year-old woman visited our family medicine clinic because she’d had 3 episodes of epistaxis during the previous month. She’d already visited the emergency department, and the doctor there had treated her symptomatically and referred her to our clinic.

On physical examination, we noted a whitish mass in the patient’s right nostril that was attached to the nasal septum. The patient’s vital signs were within normal limits. She had a history of hypertension, depression, anxiety, gastroesophageal reflux disease, and post-traumatic stress disorder. Her medications included amlodipine-benazepril, atenolol-chlorthalidone, citalopram, clonazepam, prazosin, and omeprazole. The patient lived alone and denied using tobacco or illicit drugs, but she drank one to 2 glasses of brandy every day. She denied any past medical or family history of similar complaints, autoimmune disorders, or skin rashes.

A complete blood count, international normalized ratio, sedimentation rate, anti-nuclear antibody test, and an anti-neutrophil cytoplasmic antibody panel were normal.

THE DIAGNOSIS

We referred the patient to an ear, nose, and throat doctor for a nasal endoscopy and a biopsy, which showed granulation tissue. A maxillofacial computed tomography (CT) scan revealed a 1.44 cm x 0.8 cm polypoid soft tissue mass in the right nasal cavity adherent to the nasal septum with no posterior extension (FIGURE 1).

The patient underwent cauterization of the mass and was sent home with a prescription for fluticasone spray and a referral to a head and neck oncologist. Excision and biopsy of the mass revealed pyogenic granuloma (lobular capillary hemangioma) with surface ulceration (FIGURES 2 AND 3).

 

 

DISCUSSION

Pyogenic granuloma (PG) is a benign vascular tumor of the skin and mucous membranes that is not associated with an infection. Rather, it is a hyperplastic, neovascular, inflammatory response to an angiogenic stimulus. Several enhancers and inhibitors of angiogenesis have been shown to play a role in PG, including hormones, medications, and local injury. In fact, a local injury or hormonal factor is identified as a stimulus in more than half of PG patients.¹

The hormone connection. Estrogen promotes production of nerve growth factor, granulocyte-macrophage colony-stimulating factor, basic fibroblast growth factor, vascular endothelial growth factor, and transforming growth factor beta 1. Progesterone enhances inflammatory mediators as well. Although there are no direct receptors for estrogen and progesterone in the oral and nasal mucosa, some of these pro-inflammatory effects create an environment conducive to the development of PG. This is supported by several studies documenting an increased incidence of PGs with oral contraceptive use and regression of PGs after childbirth.^2-4

In more than half of patients with pyogenic granuloma, a local injury or hormonal factor is at work.

Medication may play a role. Drug-induced PG has also been described in several studies.^5,6 Offending medications include systemic and topical retinoids, capecitabine, etoposide, 5-fluorouracil, cyclosporine, docetaxel, and human immunodeficiency virus protease inhibitors.

Local injury may also be a culprit. Nasal PGs are commonly attached to the anterior septum and typically result from nasal packing, habitual picking, or nose boring.⁷ In this particular case, however, we were unable to identify the irritant.

The classic presentation

PG classically presents as a painless mass that spontaneously develops over days to weeks. The mass can be sessile or pedunculated, and is frequently hemorrhagic. Intranasal PG usually presents with epiphora.⁷ While the prevalence of intraoral PG was found to be one in 25,000 individuals³, data for nasal lesions is scarce. Most cases of PG are seen in the second and third decades of life.^1,3 In children, PG is slightly more predominant in males.^1,3 Mucosal lesions, however, have a higher incidence in females.^1,3 Granuloma gravidarum, the term used to describe mucosal PG in pregnant females, was found in 0.2% to 5% of pregnancies.^2,3,8

 

Differential Dx includes warts, squamous cell carcinoma

The differential diagnosis of PG includes Spitz nevus, glomus tumors, common warts, amelanotic melanoma, squamous cell carcinoma, basal cell carcinoma, Kaposi’s sarcoma, bacillary angiomatosis, infantile hemangioma, and angiolymphoid hyperplasia, among others.^3,5 Foreign bodies, nasal polyps, angiofibroma, meningocele, Wegener’s granulomatosis, and sarcoidosis should also be considered.

Radiologic evaluation may be beneficial—especially with nasal lesions—when looking for findings suggestive of malignancy. Both CT and magnetic resonance imaging with contrast identify PG as a soft tissue mass with lobulated contours,^9,10 but histopathologic analysis is required to confirm the diagnosis. The histopathologic appearance of PG is characterized by a polypoid lesion with circumscribed anastomosing networks of capillaries arranged in one or more lobules at the base in an edematous and fibroblastic stroma.

Alternatives to surgery include electrocautery, cryotherapy, laser therapy, and intralesional and topical agents, but the recurrence rate with these is often higher.

Treatment is determined by the location and size of the lesion

The most suitable treatment is determined by considering the location of the lesion, the characteristics of the lesion (morphology/size), its amenability to surgery, risk of scar formation, and the presence or absence of a causative irritant. Excision is often preferred because it yields a specimen for pathologic analysis. Alternative treatments include electrocautery, cryotherapy, laser therapy, and intralesional and topical agents,^3,6,7 but the recurrence rate is higher (up to 15%) with some of these modalities, when compared with excision (3.6%).³

Our patient underwent excision of the mass and was seen for an annual follow-up appointment. All of her symptoms resolved and no recurrence was noted.

THE TAKEAWAY

Although PG is a common and benign condition, it is rarely seen in the nasal cavity without an obvious history of a possible irritant. PG should be considered as a diagnosis for rapidly growing cutaneous or mucosal hemorrhagic lesions. Appropriate tissue pathology is essential to rule out malignancy and other serious conditions, such as bacillary angiomatosis and Wegener’s granulomatosis.

Treatment is usually required to avoid the frequent complications of ulceration and bleeding. Surgical treatments are preferred. The location of the lesion largely determines whether referral to a specialist is necessary.

CASE 1

A 13-month-old boy who was recently adopted from Ethiopia presented to a primary care physician with a 3-week history of bloody diarrhea accompanied by flatulence and bloating. Stool cultures were positive for Campylobacter and Shigella. He was prescribed azithromycin but saw only moderate improvement. He was then referred to the Infectious Diseases Department. Neonatal, pregnancy, and immunization histories were unknown and a review of systems was unremarkable. On exam, the child looked well; he weighed 9.6 kg (15th percentile), was 69.5 cm long (<3rd percentile), and his head circumference was 45 cm (10th percentile). Head and neck, cardiorespiratory, and abdominal examinations were unremarkable.

A complete blood count (CBC) showed an elevated white blood cell (WBC) count of 26 x 10⁹/L (normal: 4-10 x 10⁹/L) with predominant eosinophilia (10.4 x 10⁹/L or 40.1% of WBCs; normal: <0.45 x 10⁹/L or 0%-8%). Hemoglobin and platelets were within normal limits. Stool testing for ova and parasites showed Strongyloides stercoralis larvae. Strongyloides serology was negative and Filaria serology was equivocal.

CASE 2

A 15-year-old boy was assessed for a 3-week history of fever and eosinophilia. He had enlarged cervical lymph nodes, a new rash, and had lost 4 pounds. He denied gastrointestinal symptoms, dyspnea, headaches, or chest pain. His past medical and family histories were unremarkable and he reported no drug use or allergies. He had traveled to Cuba with his family for 15 days 3 months prior to presentation. He recalled diarrhea while traveling, which resolved spontaneously. He and his family had traveled “off the beaten track,” eating foods prepared at local establishments and swimming in local rivers. He received pre-travel immunizations.

On examination, he appeared unwell, though his vital signs were normal. He had diffuse lymphadenopathy and a petechial rash on his chest, back, upper buttocks, legs, and feet. Cardiorespiratory and abdominal examinations were unremarkable. A CBC revealed an elevated WBC count of 76.9 x 10⁹/L with predominant eosinophilia (71.5 x 10⁹/L or 92% of WBCs). Hemoglobin, platelets, electrolytes, and liver function tests were normal. The patient was referred to a tertiary care center and was admitted to the hospital. Stool testing for ova and parasites, as well as serology for parasitic infections, was negative. A bone marrow aspiration and biopsy were performed and revealed the diagnosis of acute lymphoblastic leukemia (ALL).

 

 

DISCUSSION

These 2 cases highlight how the presentation of eosinophilia can vary and how important it is to maintain a broad differential diagnosis (TABLE 1^1-4). Causes of eosinophilia are numerous and can be divided into 3 categories: primary, secondary, and idiopathic.^1,5 Hematologic malignancy, where eosinophilia is clonal, is an example of a primary etiology. Causes of secondary eosinophilia include infectious diseases, drugs (TABLE 2⁵), autoimmune disorders, and allergic conditions. Prolonged eosinophilia that is >3 x 10⁹/L is associated with end-organ damage. Dermatologic, pulmonary, gastro­intestinal, and cardiac involvement is most common.²

Eosinophilia associated with parasitic infection

In returning travelers and international adoptees, multicellular helminthic parasites are the most common causes of eosinophilia, with eosinophilia occurring during tissue migration or penetration.^1,3

Strongyloides stercoralis is found in tropical and temperate areas. Its life cycle involves both a parasitic and free-living component. It has the ability to autoinfect the host and can cause hyperinfection in those who are immunocompromised. Acute infection is often subclinical, but may include a skin rash or nonspecific intestinal symptoms.¹ Patients commonly present with asymptomatic eosinophilia. Diagnosis is most often via serology, and the first-line treatment for uncomplicated cases is ivermectin 200 µg/kg/day orally for 2 days.⁴

 

Schistosomiasis is a chronic parasitic infection of the human vascular system. It is transmitted by contact with contaminated fresh water, where cercariae penetrate the skin. High prevalence areas include Africa and Southeast Asia. Acute infection can result in Katayama fever—a febrile illness with prominent eosinophilia that occurs 4 to 7 weeks after exposure.⁴ Diagnosis is primarily clinical with appropriate epidemiology, as serology may be negative early in infection. Praziquantel is the treatment of choice, though dosing varies by species, so expert consultation should be considered.

Soil-transmitted helminths, such as Ascaris (Ascaris lumbricoides), whipworm (Trichuris trichiura), and hookworm (Ancyclostoma duodenale and Necator americanus), can also cause eosinophilia during larval tissue migration. Following infection by ingestion or skin penetration, an acute respiratory illness, termed Löffler’s syndrome, can develop with associated eosinophilia.¹ Once the helminths reach the adult stage, eosinophilia subsides. Patients are most commonly treated with albendazole 400 mg orally for 3 days.⁴

Prolonged eosinophilia that is >3 x 10⁹/L is associated with end-organ damage.Fascioliasis is common in sheep-rearing areas. Humans are infected through ingestion of aquatic plants (eg, watercress). Parasitic migration through the duodenal wall and liver parenchyma can lead to fever, right upper quadrant pain, and eosinophilia. The incubation period is 6 to 12 weeks. Diagnosis during acute infection is by serology.⁴

Filarial infections, eg lymphatic filariasis, loiasis, and onchocerciasis, can also cause eosinophilia. The rise in eosinophils can be triggered by either the adult worms or circulating microfilariae.⁴ Treatment of fascioliasis and filarial infections varies and expert consultation is recommended.

Eosinophilia associated with primary hematologic malignancy

Eosinophilia is a rare presentation of hematologic malignancy. Acute myeloid leukemia, acute lymphoblastic leukemia (ALL), chronic myeloid leukemia, and myeloproliferative disorders have all been associated with eosinophilia. Hepatosplenomegaly, generalized lymphadenopathy, and cytopenias in other cell lines are often noted. Also, the degree of eosinophilia is often more pronounced (>5 x 10⁹/L). Patients with suspected hematologic malignancy should be urgently referred for expert consultation.⁵

 

 

A systematic approach to patients with eosinophilia

Consider the following approach in the assessment of patients with eosinophilia seen in the ambulatory care setting. Inpatients or patients being seen in developing areas may require a modified approach.

History. All patients with eosinophilia should have a thorough history taken, with particular attention paid to travel history. A travel history should make note of dates, duration and location of travel, and any relevant exposures, such as arthropod bites or swimming in freshwater. Dietary habits, such as ingestion of seafood, game, or undercooked meat can also be helpful in making a diagnosis.^3,4

Causes of secondary eosinophilia include infectious diseases, drugs, autoimmune disorders, and allergic conditions.

Physical exam. In addition to a general physical examination, the following features may be helpful in determining the etiology of eosinophilia. Wheeze is characteristic of parasites in a lung migration phase (eg, strongyloidiasis and ascariasis) or asthma. Hepatomegaly can be seen with liver flukes, visceral larva migrans, or schistosomiasis. Periorbital edema can be observed with Trichinella infection. Loa loa, a type of filarial infection, produces a transient, migratory angioedema, often localized to the wrists and large joints (termed Calabar swelling). Dermatitis of varying intensity may suggest filarial infection, schistosomiasis, or atopy. Perianal dermatitis is observed with strongyloidiasis. Cutaneous larva migrans is characterized by a linear, serpiginous rash.^3,4

Laboratory investigations. Investigation will vary depending on the patient’s history, exposures, exam findings, and degree of eosinophilia. Any patient who is unwell or has significant eosinophilia (≥3 x 10⁹/L) may warrant more urgent referral to infectious disease, travel medicine, or hematology. Basic laboratory investigations should include a CBC with differential, routine serum chemistries, and liver enzymes. In the setting of significant eosinophilia, an electrocardiogram, cardiac enzyme levels, and a chest x-ray should be obtained to screen for end-organ damage related to eosinophilia.^3-5

In patients in whom you suspect hematologic malignancy, bone marrow aspiration and biopsy are often needed to make the diagnosis.⁵

In returning travelers and international adoptees, multicellular helminthic parasites are the most common causes of eosinophilia.

Parasitic infections are most often diagnosed on stool examination for ova and parasites or by serology. Stool should be collected on 3 separate days to increase diagnostic yield. Certain species of Schistosoma can also be diagnosed on direct microscopy of urine specimens. Serologic assays are available for schistosomiasis, strongyloidiasis, Toxocara, fascioliasis, filariasis, and Trichinella. Further investigations for filiariasis, including blood films, eye exam, and skin snips will vary with filarial species, so expert consultation should be considered.^3,4

Our patients. The first patient with strongyloidiasis was treated with ivermectin 200 µg/kg/day orally for 2 days and experienced symptomatic improvement and resolution of eosinophilia. The second patient with ALL was admitted and referred to hematology and received induction chemotherapy. Treatment was well tolerated and the patient was discharged one week later, with appropriate follow-up.

THE TAKEAWAY

Eosinophilia is commonly encountered in primary care. The approach to eosinophilia and the differential diagnosis can be challenging. The correct diagnosis was reached in both cases by maintaining a broad differential diagnosis. Obtaining a travel and exposure history is fundamental, although noninfectious causes, including allergy, malignancy, and drug reaction, must always be considered.

CASE 1

A 13-month-old boy who was recently adopted from Ethiopia presented to a primary care physician with a 3-week history of bloody diarrhea accompanied by flatulence and bloating. Stool cultures were positive for Campylobacter and Shigella. He was prescribed azithromycin but saw only moderate improvement. He was then referred to the Infectious Diseases Department. Neonatal, pregnancy, and immunization histories were unknown and a review of systems was unremarkable. On exam, the child looked well; he weighed 9.6 kg (15th percentile), was 69.5 cm long (<3rd percentile), and his head circumference was 45 cm (10th percentile). Head and neck, cardiorespiratory, and abdominal examinations were unremarkable.

A complete blood count (CBC) showed an elevated white blood cell (WBC) count of 26 x 10⁹/L (normal: 4-10 x 10⁹/L) with predominant eosinophilia (10.4 x 10⁹/L or 40.1% of WBCs; normal: <0.45 x 10⁹/L or 0%-8%). Hemoglobin and platelets were within normal limits. Stool testing for ova and parasites showed Strongyloides stercoralis larvae. Strongyloides serology was negative and Filaria serology was equivocal.

CASE 2

A 15-year-old boy was assessed for a 3-week history of fever and eosinophilia. He had enlarged cervical lymph nodes, a new rash, and had lost 4 pounds. He denied gastrointestinal symptoms, dyspnea, headaches, or chest pain. His past medical and family histories were unremarkable and he reported no drug use or allergies. He had traveled to Cuba with his family for 15 days 3 months prior to presentation. He recalled diarrhea while traveling, which resolved spontaneously. He and his family had traveled “off the beaten track,” eating foods prepared at local establishments and swimming in local rivers. He received pre-travel immunizations.

On examination, he appeared unwell, though his vital signs were normal. He had diffuse lymphadenopathy and a petechial rash on his chest, back, upper buttocks, legs, and feet. Cardiorespiratory and abdominal examinations were unremarkable. A CBC revealed an elevated WBC count of 76.9 x 10⁹/L with predominant eosinophilia (71.5 x 10⁹/L or 92% of WBCs). Hemoglobin, platelets, electrolytes, and liver function tests were normal. The patient was referred to a tertiary care center and was admitted to the hospital. Stool testing for ova and parasites, as well as serology for parasitic infections, was negative. A bone marrow aspiration and biopsy were performed and revealed the diagnosis of acute lymphoblastic leukemia (ALL).

 

 

DISCUSSION

These 2 cases highlight how the presentation of eosinophilia can vary and how important it is to maintain a broad differential diagnosis (TABLE 1^1-4). Causes of eosinophilia are numerous and can be divided into 3 categories: primary, secondary, and idiopathic.^1,5 Hematologic malignancy, where eosinophilia is clonal, is an example of a primary etiology. Causes of secondary eosinophilia include infectious diseases, drugs (TABLE 2⁵), autoimmune disorders, and allergic conditions. Prolonged eosinophilia that is >3 x 10⁹/L is associated with end-organ damage. Dermatologic, pulmonary, gastro­intestinal, and cardiac involvement is most common.²

Eosinophilia associated with parasitic infection

In returning travelers and international adoptees, multicellular helminthic parasites are the most common causes of eosinophilia, with eosinophilia occurring during tissue migration or penetration.^1,3

Strongyloides stercoralis is found in tropical and temperate areas. Its life cycle involves both a parasitic and free-living component. It has the ability to autoinfect the host and can cause hyperinfection in those who are immunocompromised. Acute infection is often subclinical, but may include a skin rash or nonspecific intestinal symptoms.¹ Patients commonly present with asymptomatic eosinophilia. Diagnosis is most often via serology, and the first-line treatment for uncomplicated cases is ivermectin 200 µg/kg/day orally for 2 days.⁴

 

Schistosomiasis is a chronic parasitic infection of the human vascular system. It is transmitted by contact with contaminated fresh water, where cercariae penetrate the skin. High prevalence areas include Africa and Southeast Asia. Acute infection can result in Katayama fever—a febrile illness with prominent eosinophilia that occurs 4 to 7 weeks after exposure.⁴ Diagnosis is primarily clinical with appropriate epidemiology, as serology may be negative early in infection. Praziquantel is the treatment of choice, though dosing varies by species, so expert consultation should be considered.

Soil-transmitted helminths, such as Ascaris (Ascaris lumbricoides), whipworm (Trichuris trichiura), and hookworm (Ancyclostoma duodenale and Necator americanus), can also cause eosinophilia during larval tissue migration. Following infection by ingestion or skin penetration, an acute respiratory illness, termed Löffler’s syndrome, can develop with associated eosinophilia.¹ Once the helminths reach the adult stage, eosinophilia subsides. Patients are most commonly treated with albendazole 400 mg orally for 3 days.⁴

Prolonged eosinophilia that is >3 x 10⁹/L is associated with end-organ damage.Fascioliasis is common in sheep-rearing areas. Humans are infected through ingestion of aquatic plants (eg, watercress). Parasitic migration through the duodenal wall and liver parenchyma can lead to fever, right upper quadrant pain, and eosinophilia. The incubation period is 6 to 12 weeks. Diagnosis during acute infection is by serology.⁴

Filarial infections, eg lymphatic filariasis, loiasis, and onchocerciasis, can also cause eosinophilia. The rise in eosinophils can be triggered by either the adult worms or circulating microfilariae.⁴ Treatment of fascioliasis and filarial infections varies and expert consultation is recommended.

Eosinophilia associated with primary hematologic malignancy

Eosinophilia is a rare presentation of hematologic malignancy. Acute myeloid leukemia, acute lymphoblastic leukemia (ALL), chronic myeloid leukemia, and myeloproliferative disorders have all been associated with eosinophilia. Hepatosplenomegaly, generalized lymphadenopathy, and cytopenias in other cell lines are often noted. Also, the degree of eosinophilia is often more pronounced (>5 x 10⁹/L). Patients with suspected hematologic malignancy should be urgently referred for expert consultation.⁵

 

 

A systematic approach to patients with eosinophilia

Consider the following approach in the assessment of patients with eosinophilia seen in the ambulatory care setting. Inpatients or patients being seen in developing areas may require a modified approach.

History. All patients with eosinophilia should have a thorough history taken, with particular attention paid to travel history. A travel history should make note of dates, duration and location of travel, and any relevant exposures, such as arthropod bites or swimming in freshwater. Dietary habits, such as ingestion of seafood, game, or undercooked meat can also be helpful in making a diagnosis.^3,4

Causes of secondary eosinophilia include infectious diseases, drugs, autoimmune disorders, and allergic conditions.

Physical exam. In addition to a general physical examination, the following features may be helpful in determining the etiology of eosinophilia. Wheeze is characteristic of parasites in a lung migration phase (eg, strongyloidiasis and ascariasis) or asthma. Hepatomegaly can be seen with liver flukes, visceral larva migrans, or schistosomiasis. Periorbital edema can be observed with Trichinella infection. Loa loa, a type of filarial infection, produces a transient, migratory angioedema, often localized to the wrists and large joints (termed Calabar swelling). Dermatitis of varying intensity may suggest filarial infection, schistosomiasis, or atopy. Perianal dermatitis is observed with strongyloidiasis. Cutaneous larva migrans is characterized by a linear, serpiginous rash.^3,4

Laboratory investigations. Investigation will vary depending on the patient’s history, exposures, exam findings, and degree of eosinophilia. Any patient who is unwell or has significant eosinophilia (≥3 x 10⁹/L) may warrant more urgent referral to infectious disease, travel medicine, or hematology. Basic laboratory investigations should include a CBC with differential, routine serum chemistries, and liver enzymes. In the setting of significant eosinophilia, an electrocardiogram, cardiac enzyme levels, and a chest x-ray should be obtained to screen for end-organ damage related to eosinophilia.^3-5

In patients in whom you suspect hematologic malignancy, bone marrow aspiration and biopsy are often needed to make the diagnosis.⁵

In returning travelers and international adoptees, multicellular helminthic parasites are the most common causes of eosinophilia.

Parasitic infections are most often diagnosed on stool examination for ova and parasites or by serology. Stool should be collected on 3 separate days to increase diagnostic yield. Certain species of Schistosoma can also be diagnosed on direct microscopy of urine specimens. Serologic assays are available for schistosomiasis, strongyloidiasis, Toxocara, fascioliasis, filariasis, and Trichinella. Further investigations for filiariasis, including blood films, eye exam, and skin snips will vary with filarial species, so expert consultation should be considered.^3,4

Our patients. The first patient with strongyloidiasis was treated with ivermectin 200 µg/kg/day orally for 2 days and experienced symptomatic improvement and resolution of eosinophilia. The second patient with ALL was admitted and referred to hematology and received induction chemotherapy. Treatment was well tolerated and the patient was discharged one week later, with appropriate follow-up.

THE TAKEAWAY

Eosinophilia is commonly encountered in primary care. The approach to eosinophilia and the differential diagnosis can be challenging. The correct diagnosis was reached in both cases by maintaining a broad differential diagnosis. Obtaining a travel and exposure history is fundamental, although noninfectious causes, including allergy, malignancy, and drug reaction, must always be considered.

CASE 1

A 13-month-old boy who was recently adopted from Ethiopia presented to a primary care physician with a 3-week history of bloody diarrhea accompanied by flatulence and bloating. Stool cultures were positive for Campylobacter and Shigella. He was prescribed azithromycin but saw only moderate improvement. He was then referred to the Infectious Diseases Department. Neonatal, pregnancy, and immunization histories were unknown and a review of systems was unremarkable. On exam, the child looked well; he weighed 9.6 kg (15th percentile), was 69.5 cm long (<3rd percentile), and his head circumference was 45 cm (10th percentile). Head and neck, cardiorespiratory, and abdominal examinations were unremarkable.

A complete blood count (CBC) showed an elevated white blood cell (WBC) count of 26 x 10⁹/L (normal: 4-10 x 10⁹/L) with predominant eosinophilia (10.4 x 10⁹/L or 40.1% of WBCs; normal: <0.45 x 10⁹/L or 0%-8%). Hemoglobin and platelets were within normal limits. Stool testing for ova and parasites showed Strongyloides stercoralis larvae. Strongyloides serology was negative and Filaria serology was equivocal.

CASE 2

A 15-year-old boy was assessed for a 3-week history of fever and eosinophilia. He had enlarged cervical lymph nodes, a new rash, and had lost 4 pounds. He denied gastrointestinal symptoms, dyspnea, headaches, or chest pain. His past medical and family histories were unremarkable and he reported no drug use or allergies. He had traveled to Cuba with his family for 15 days 3 months prior to presentation. He recalled diarrhea while traveling, which resolved spontaneously. He and his family had traveled “off the beaten track,” eating foods prepared at local establishments and swimming in local rivers. He received pre-travel immunizations.

On examination, he appeared unwell, though his vital signs were normal. He had diffuse lymphadenopathy and a petechial rash on his chest, back, upper buttocks, legs, and feet. Cardiorespiratory and abdominal examinations were unremarkable. A CBC revealed an elevated WBC count of 76.9 x 10⁹/L with predominant eosinophilia (71.5 x 10⁹/L or 92% of WBCs). Hemoglobin, platelets, electrolytes, and liver function tests were normal. The patient was referred to a tertiary care center and was admitted to the hospital. Stool testing for ova and parasites, as well as serology for parasitic infections, was negative. A bone marrow aspiration and biopsy were performed and revealed the diagnosis of acute lymphoblastic leukemia (ALL).

 

 

DISCUSSION

These 2 cases highlight how the presentation of eosinophilia can vary and how important it is to maintain a broad differential diagnosis (TABLE 1^1-4). Causes of eosinophilia are numerous and can be divided into 3 categories: primary, secondary, and idiopathic.^1,5 Hematologic malignancy, where eosinophilia is clonal, is an example of a primary etiology. Causes of secondary eosinophilia include infectious diseases, drugs (TABLE 2⁵), autoimmune disorders, and allergic conditions. Prolonged eosinophilia that is >3 x 10⁹/L is associated with end-organ damage. Dermatologic, pulmonary, gastro­intestinal, and cardiac involvement is most common.²

Eosinophilia associated with parasitic infection

In returning travelers and international adoptees, multicellular helminthic parasites are the most common causes of eosinophilia, with eosinophilia occurring during tissue migration or penetration.^1,3

Strongyloides stercoralis is found in tropical and temperate areas. Its life cycle involves both a parasitic and free-living component. It has the ability to autoinfect the host and can cause hyperinfection in those who are immunocompromised. Acute infection is often subclinical, but may include a skin rash or nonspecific intestinal symptoms.¹ Patients commonly present with asymptomatic eosinophilia. Diagnosis is most often via serology, and the first-line treatment for uncomplicated cases is ivermectin 200 µg/kg/day orally for 2 days.⁴

 

Schistosomiasis is a chronic parasitic infection of the human vascular system. It is transmitted by contact with contaminated fresh water, where cercariae penetrate the skin. High prevalence areas include Africa and Southeast Asia. Acute infection can result in Katayama fever—a febrile illness with prominent eosinophilia that occurs 4 to 7 weeks after exposure.⁴ Diagnosis is primarily clinical with appropriate epidemiology, as serology may be negative early in infection. Praziquantel is the treatment of choice, though dosing varies by species, so expert consultation should be considered.

Soil-transmitted helminths, such as Ascaris (Ascaris lumbricoides), whipworm (Trichuris trichiura), and hookworm (Ancyclostoma duodenale and Necator americanus), can also cause eosinophilia during larval tissue migration. Following infection by ingestion or skin penetration, an acute respiratory illness, termed Löffler’s syndrome, can develop with associated eosinophilia.¹ Once the helminths reach the adult stage, eosinophilia subsides. Patients are most commonly treated with albendazole 400 mg orally for 3 days.⁴

Prolonged eosinophilia that is >3 x 10⁹/L is associated with end-organ damage.Fascioliasis is common in sheep-rearing areas. Humans are infected through ingestion of aquatic plants (eg, watercress). Parasitic migration through the duodenal wall and liver parenchyma can lead to fever, right upper quadrant pain, and eosinophilia. The incubation period is 6 to 12 weeks. Diagnosis during acute infection is by serology.⁴

Filarial infections, eg lymphatic filariasis, loiasis, and onchocerciasis, can also cause eosinophilia. The rise in eosinophils can be triggered by either the adult worms or circulating microfilariae.⁴ Treatment of fascioliasis and filarial infections varies and expert consultation is recommended.

Eosinophilia associated with primary hematologic malignancy

Eosinophilia is a rare presentation of hematologic malignancy. Acute myeloid leukemia, acute lymphoblastic leukemia (ALL), chronic myeloid leukemia, and myeloproliferative disorders have all been associated with eosinophilia. Hepatosplenomegaly, generalized lymphadenopathy, and cytopenias in other cell lines are often noted. Also, the degree of eosinophilia is often more pronounced (>5 x 10⁹/L). Patients with suspected hematologic malignancy should be urgently referred for expert consultation.⁵

 

 

A systematic approach to patients with eosinophilia

Consider the following approach in the assessment of patients with eosinophilia seen in the ambulatory care setting. Inpatients or patients being seen in developing areas may require a modified approach.

History. All patients with eosinophilia should have a thorough history taken, with particular attention paid to travel history. A travel history should make note of dates, duration and location of travel, and any relevant exposures, such as arthropod bites or swimming in freshwater. Dietary habits, such as ingestion of seafood, game, or undercooked meat can also be helpful in making a diagnosis.^3,4

Causes of secondary eosinophilia include infectious diseases, drugs, autoimmune disorders, and allergic conditions.

Physical exam. In addition to a general physical examination, the following features may be helpful in determining the etiology of eosinophilia. Wheeze is characteristic of parasites in a lung migration phase (eg, strongyloidiasis and ascariasis) or asthma. Hepatomegaly can be seen with liver flukes, visceral larva migrans, or schistosomiasis. Periorbital edema can be observed with Trichinella infection. Loa loa, a type of filarial infection, produces a transient, migratory angioedema, often localized to the wrists and large joints (termed Calabar swelling). Dermatitis of varying intensity may suggest filarial infection, schistosomiasis, or atopy. Perianal dermatitis is observed with strongyloidiasis. Cutaneous larva migrans is characterized by a linear, serpiginous rash.^3,4

Laboratory investigations. Investigation will vary depending on the patient’s history, exposures, exam findings, and degree of eosinophilia. Any patient who is unwell or has significant eosinophilia (≥3 x 10⁹/L) may warrant more urgent referral to infectious disease, travel medicine, or hematology. Basic laboratory investigations should include a CBC with differential, routine serum chemistries, and liver enzymes. In the setting of significant eosinophilia, an electrocardiogram, cardiac enzyme levels, and a chest x-ray should be obtained to screen for end-organ damage related to eosinophilia.^3-5

In patients in whom you suspect hematologic malignancy, bone marrow aspiration and biopsy are often needed to make the diagnosis.⁵

In returning travelers and international adoptees, multicellular helminthic parasites are the most common causes of eosinophilia.

Parasitic infections are most often diagnosed on stool examination for ova and parasites or by serology. Stool should be collected on 3 separate days to increase diagnostic yield. Certain species of Schistosoma can also be diagnosed on direct microscopy of urine specimens. Serologic assays are available for schistosomiasis, strongyloidiasis, Toxocara, fascioliasis, filariasis, and Trichinella. Further investigations for filiariasis, including blood films, eye exam, and skin snips will vary with filarial species, so expert consultation should be considered.^3,4

Our patients. The first patient with strongyloidiasis was treated with ivermectin 200 µg/kg/day orally for 2 days and experienced symptomatic improvement and resolution of eosinophilia. The second patient with ALL was admitted and referred to hematology and received induction chemotherapy. Treatment was well tolerated and the patient was discharged one week later, with appropriate follow-up.

THE TAKEAWAY

Eosinophilia is commonly encountered in primary care. The approach to eosinophilia and the differential diagnosis can be challenging. The correct diagnosis was reached in both cases by maintaining a broad differential diagnosis. Obtaining a travel and exposure history is fundamental, although noninfectious causes, including allergy, malignancy, and drug reaction, must always be considered.

Angiomyomatous hamartoma (AMH) of the lymph node is an extremely uncommon vascular disorder of unknown etiology, first described by Chan and colleagues in 1992.^1-3 Angiomyomatous hamartoma particularly involves inguinal and femoral lymph nodes, with few cases reported in the cervical, popliteal, and submandibular lymph nodes.¹ Angiomyomatous hamartoma can occasionally be associated with edema of the ipsilateral limb. To the authors’ knowledge, to date only 18 cases of AMH have been reported.⁴

Case Presentation

A 40-year-old white man started to have a left inguinal and scrotal pain along with left thigh swelling at age 22 while serving in the U.S. Army.

A left hydrocelectomy was performed overseas; however, the patient’s symptoms persisted. He was evaluated in the U.S. at age 36, and the physical examination revealed a left medial thigh nonpitting edema, an enlarged left inguinal lymph node, bilateral axillary lymphadenopathy, and hepatosplenomegaly.

A complete blood count, basic metabolic panel, liver function tests, hepatitis workup, autoimmune profile, iron studies,

alpha-fetoprotein, and ceruloplasmin and copper levels were within normal limits.

An abdominal Doppler ultrasound did not show any evidence of portal hypertension. A thoraco-abdomino-pelvic computed tomography (CT) scan showed bilateral axillary, subcarinal (Figure 1), mesenteric and retroperitoneal (Figure 2), and left inguinal (Figure 3) lymphadenopathy. Excisional biopsy of a 3.5 x 2.5 x 1.5 cm left inguinal lymph node was performed, and histopathology showed extensive smooth muscle and vascular proliferation replacing most of the lymph node (Figure 4), a finding consistent with AMH. A trichrome staining (Figure 5) and immunohistochemical study for smooth muscle actin (Figure 6) were performed and supported the diagnosis. Due to persistent pain in the scrotal area, the patient underwent a left spermatic cord denervation. Currently, the patient has persistent left thigh swelling. His condition remains stable with a regular follow-up CT scan showing unchanged lymphadenopathy.

Discussion

Angiomyomatous hamartoma is a rare, primary vascular tumor of the lymph nodes occurring almost exclusively in the inguinal and femoral lymph nodes and occasionally associated with edema of the ipsilateral limb.¹ A few cases with popliteal and cervical lymph node involvement have been reported.¹ There are no prior reports of cases with either generalized adenopathy or hepatosplenomegaly.

The histopathogenesis of AMH remains unclear. Chan and colleagues first reported this distinct clinicopathologic entity in 1992 as a primary vascular tumor of the lymph node.^1-3 The hamartomatous nature of the disease was postulated by the authors on the basis of a disorganized growth pattern of smooth muscle cells and blood vessels noted on pathology.^2,3 The AMH could represent a localized malformation in a congenitally damaged lymphatic vessel system.⁵ Other hypothesis suggests lymphedema as a possible etiology of AMH through continuous stimulation of lymphatic vessels, which triggers vasoproliferation and eventually the vascular transformation of the lymph nodes.⁵

Differential diagnoses of AMH include nodal lymphangiomyomatosis, which is most prevalent in women, particularly presenting with thoracic and intra-abdominal lymph nodes and plumper HMB45 (human melanoma black 45) -positive tumor cells⁶; leiomyomato

sis, which tends to have an intra-abdominal localization with compact bundles of smooth muscle cells and insignificant vascular component; and angiomyolipoma, which typically presents with retroperitoneal nodes and HMB45-positive cells.

Treatment is either conservative or surgical, depending on clinical judgment. This is only the 19th case of AMH reported so far in the literature and the fifth reported case in which the patient presented with ipsilateral lymphedema of the limb. Importantly, it is the first reported case with generalized (axillary, subcarinal, mesenteric, inguinal and retroperitoneal) lymphadenopathy and unexplained hepatosplenomegaly.

 

Conclusion

Angiomyomatous hamartoma of the lymph nodes is an exceedingly rare diagnosis but should be considered when evaluating patients with lymphatic tumors. This patient remains relatively asymptomatic and on observation at this time and seems to have more extensive disease than prior reports in the literature.

Angiomyomatous hamartoma (AMH) of the lymph node is an extremely uncommon vascular disorder of unknown etiology, first described by Chan and colleagues in 1992.^1-3 Angiomyomatous hamartoma particularly involves inguinal and femoral lymph nodes, with few cases reported in the cervical, popliteal, and submandibular lymph nodes.¹ Angiomyomatous hamartoma can occasionally be associated with edema of the ipsilateral limb. To the authors’ knowledge, to date only 18 cases of AMH have been reported.⁴

Case Presentation

A 40-year-old white man started to have a left inguinal and scrotal pain along with left thigh swelling at age 22 while serving in the U.S. Army.

A left hydrocelectomy was performed overseas; however, the patient’s symptoms persisted. He was evaluated in the U.S. at age 36, and the physical examination revealed a left medial thigh nonpitting edema, an enlarged left inguinal lymph node, bilateral axillary lymphadenopathy, and hepatosplenomegaly.

A complete blood count, basic metabolic panel, liver function tests, hepatitis workup, autoimmune profile, iron studies,

alpha-fetoprotein, and ceruloplasmin and copper levels were within normal limits.

An abdominal Doppler ultrasound did not show any evidence of portal hypertension. A thoraco-abdomino-pelvic computed tomography (CT) scan showed bilateral axillary, subcarinal (Figure 1), mesenteric and retroperitoneal (Figure 2), and left inguinal (Figure 3) lymphadenopathy. Excisional biopsy of a 3.5 x 2.5 x 1.5 cm left inguinal lymph node was performed, and histopathology showed extensive smooth muscle and vascular proliferation replacing most of the lymph node (Figure 4), a finding consistent with AMH. A trichrome staining (Figure 5) and immunohistochemical study for smooth muscle actin (Figure 6) were performed and supported the diagnosis. Due to persistent pain in the scrotal area, the patient underwent a left spermatic cord denervation. Currently, the patient has persistent left thigh swelling. His condition remains stable with a regular follow-up CT scan showing unchanged lymphadenopathy.

Discussion

Angiomyomatous hamartoma is a rare, primary vascular tumor of the lymph nodes occurring almost exclusively in the inguinal and femoral lymph nodes and occasionally associated with edema of the ipsilateral limb.¹ A few cases with popliteal and cervical lymph node involvement have been reported.¹ There are no prior reports of cases with either generalized adenopathy or hepatosplenomegaly.

The histopathogenesis of AMH remains unclear. Chan and colleagues first reported this distinct clinicopathologic entity in 1992 as a primary vascular tumor of the lymph node.^1-3 The hamartomatous nature of the disease was postulated by the authors on the basis of a disorganized growth pattern of smooth muscle cells and blood vessels noted on pathology.^2,3 The AMH could represent a localized malformation in a congenitally damaged lymphatic vessel system.⁵ Other hypothesis suggests lymphedema as a possible etiology of AMH through continuous stimulation of lymphatic vessels, which triggers vasoproliferation and eventually the vascular transformation of the lymph nodes.⁵

Differential diagnoses of AMH include nodal lymphangiomyomatosis, which is most prevalent in women, particularly presenting with thoracic and intra-abdominal lymph nodes and plumper HMB45 (human melanoma black 45) -positive tumor cells⁶; leiomyomato

sis, which tends to have an intra-abdominal localization with compact bundles of smooth muscle cells and insignificant vascular component; and angiomyolipoma, which typically presents with retroperitoneal nodes and HMB45-positive cells.

Treatment is either conservative or surgical, depending on clinical judgment. This is only the 19th case of AMH reported so far in the literature and the fifth reported case in which the patient presented with ipsilateral lymphedema of the limb. Importantly, it is the first reported case with generalized (axillary, subcarinal, mesenteric, inguinal and retroperitoneal) lymphadenopathy and unexplained hepatosplenomegaly.

 

Conclusion

Angiomyomatous hamartoma of the lymph nodes is an exceedingly rare diagnosis but should be considered when evaluating patients with lymphatic tumors. This patient remains relatively asymptomatic and on observation at this time and seems to have more extensive disease than prior reports in the literature.

Angiomyomatous hamartoma (AMH) of the lymph node is an extremely uncommon vascular disorder of unknown etiology, first described by Chan and colleagues in 1992.^1-3 Angiomyomatous hamartoma particularly involves inguinal and femoral lymph nodes, with few cases reported in the cervical, popliteal, and submandibular lymph nodes.¹ Angiomyomatous hamartoma can occasionally be associated with edema of the ipsilateral limb. To the authors’ knowledge, to date only 18 cases of AMH have been reported.⁴

Case Presentation

A 40-year-old white man started to have a left inguinal and scrotal pain along with left thigh swelling at age 22 while serving in the U.S. Army.

A left hydrocelectomy was performed overseas; however, the patient’s symptoms persisted. He was evaluated in the U.S. at age 36, and the physical examination revealed a left medial thigh nonpitting edema, an enlarged left inguinal lymph node, bilateral axillary lymphadenopathy, and hepatosplenomegaly.

A complete blood count, basic metabolic panel, liver function tests, hepatitis workup, autoimmune profile, iron studies,

alpha-fetoprotein, and ceruloplasmin and copper levels were within normal limits.

An abdominal Doppler ultrasound did not show any evidence of portal hypertension. A thoraco-abdomino-pelvic computed tomography (CT) scan showed bilateral axillary, subcarinal (Figure 1), mesenteric and retroperitoneal (Figure 2), and left inguinal (Figure 3) lymphadenopathy. Excisional biopsy of a 3.5 x 2.5 x 1.5 cm left inguinal lymph node was performed, and histopathology showed extensive smooth muscle and vascular proliferation replacing most of the lymph node (Figure 4), a finding consistent with AMH. A trichrome staining (Figure 5) and immunohistochemical study for smooth muscle actin (Figure 6) were performed and supported the diagnosis. Due to persistent pain in the scrotal area, the patient underwent a left spermatic cord denervation. Currently, the patient has persistent left thigh swelling. His condition remains stable with a regular follow-up CT scan showing unchanged lymphadenopathy.

Discussion

Angiomyomatous hamartoma is a rare, primary vascular tumor of the lymph nodes occurring almost exclusively in the inguinal and femoral lymph nodes and occasionally associated with edema of the ipsilateral limb.¹ A few cases with popliteal and cervical lymph node involvement have been reported.¹ There are no prior reports of cases with either generalized adenopathy or hepatosplenomegaly.

The histopathogenesis of AMH remains unclear. Chan and colleagues first reported this distinct clinicopathologic entity in 1992 as a primary vascular tumor of the lymph node.^1-3 The hamartomatous nature of the disease was postulated by the authors on the basis of a disorganized growth pattern of smooth muscle cells and blood vessels noted on pathology.^2,3 The AMH could represent a localized malformation in a congenitally damaged lymphatic vessel system.⁵ Other hypothesis suggests lymphedema as a possible etiology of AMH through continuous stimulation of lymphatic vessels, which triggers vasoproliferation and eventually the vascular transformation of the lymph nodes.⁵

Differential diagnoses of AMH include nodal lymphangiomyomatosis, which is most prevalent in women, particularly presenting with thoracic and intra-abdominal lymph nodes and plumper HMB45 (human melanoma black 45) -positive tumor cells⁶; leiomyomato

sis, which tends to have an intra-abdominal localization with compact bundles of smooth muscle cells and insignificant vascular component; and angiomyolipoma, which typically presents with retroperitoneal nodes and HMB45-positive cells.

Treatment is either conservative or surgical, depending on clinical judgment. This is only the 19th case of AMH reported so far in the literature and the fifth reported case in which the patient presented with ipsilateral lymphedema of the limb. Importantly, it is the first reported case with generalized (axillary, subcarinal, mesenteric, inguinal and retroperitoneal) lymphadenopathy and unexplained hepatosplenomegaly.

 

Conclusion

Angiomyomatous hamartoma of the lymph nodes is an exceedingly rare diagnosis but should be considered when evaluating patients with lymphatic tumors. This patient remains relatively asymptomatic and on observation at this time and seems to have more extensive disease than prior reports in the literature.

Septic arthritis predominantly involves the weight-bearing joints of the hip and knee, which account for nearly 60% of cases.¹ In contrast, the shoulder joint is involved in 10% to 15% of cases, though this number may be higher among intravenous (IV) drug users.² The most common causative organisms are the Staphylococcus species, followed closely by β-hemolytic streptococci, with these 2 groups accounting for more than 90% of all cases.³ The Streptococcus viridans group belongs to normal oral flora residing predominantly on the surface of teeth. Although well known for its ability to colonize heart valves and frequently cause bacterial endocarditis, this group has rarely been associated with septic arthritis. Furthermore, Streptococcus mitis, a subgroup of S viridans, has been implicated even less commonly.

In this article, we report a case of glenohumeral joint septic arthritis caused by S mitis. To our knowledge, such a case has not been previously reported in the English literature. Given the low virulence of this orally based bacterium, treating physicians must maintain clinical suspicion for the organism in the setting of persistent joint effusion and pain in association with periodontal disease or trauma. The patient provided written informed consent for print and electronic publication of this case report.

Case Report

A right-hand-dominant 54-year-old man presented to Dr. Gruson with complaints of persistent right shoulder pain associated with worsening range of motion (ROM). Three weeks earlier, the patient reported being assaulted and noted progressive swelling about the right shoulder. He denied fevers, chills, or prior shoulder problems. Although his past medical history was remarkable for hepatitis C and diabetes, he was not taking any diabetic medications at that time. A review of systems was remarkable for poor dental hygiene, and the patient was missing several teeth, which he said had been knocked out during the assault. Physical examination revealed diffuse tenderness about the right shoulder and severe pain with all passive movement. The shoulder was pseudoparalyzed. There were no subcutaneous collections, wounds, or ecchymosis about the shoulder. Mild calor was noted on the right shoulder relative to the left. Radiographs of the right shoulder showed no acute osseous abnormalities.

Magnetic resonance imaging (MRI), which was urgently obtained to assess the integrity of the rotator cuff and the location of the effusion, showed a large subacromial and glenohumeral joint effusion as well as diffuse muscular edema (Figures 1A-1C).

At follow-up, the patient reported having lost 10 pounds since his assault, as well as new-onset fevers and chills. C-reactive protein (CRP) level was 5.2 mg/dL (reference, <0.9 mg/dL), and erythrocyte sedimentation rate (ESR) was 48 mm/h (reference, <21 mm/h). White blood cell count was normal. Fluoroscopy-guided aspiration of the glenohumeral joint, performed under sterile conditions, yielded only 4 cc of hematoma. Gram stain was negative; though there was no growth on the primary plates, broth cultures grew S mitis. Repeat bloodwork demonstrated persistently increased CRP level (6.4 mg/dL) and ESR (55 mm/h).

In light of the elevated infection findings of the laboratory tests and the positive culture, urgent arthroscopic irrigation and débridement of the right shoulder were indicated. Given the organism identified, transesophageal echocardiography was performed; there were no valvular vegetations. Creation of the posterior glenohumeral portal resulted in egress of turbid fluid, which was sent for culture. The subacromial space and the glenohumeral joint were thoroughly lavaged and the copious hemorrhagic synovitis débrided (Figures 2A, 2B).

Chondral surfaces appeared grossly intact. All cultures from the surgery ultimately yielded S mitis. A peripherally inserted central catheter line was started, as was a 4-week course of IV ceftriaxone, as recommended by an infectious disease consultant. At postoperative visits in the orthopedic clinic, a new-onset right axillary abscess consisting of purulent material and organized hematoma was drained. After the ceftriaxone regimen was completed, a 4-week course of oral amoxicillin was started.

The 8-week course of antibiotics normalized the patient’s ESR to 13 mm/h. Follow-up MRI showed improvement in the soft-tissue edema. Clinically, the patient reported minimal shoulder pain. He was undergoing physical therapy to regain strength and ROM.

Discussion

Staphylococcus aureus is the leading causative organism of septic arthritis, accounting for more than 60% of all cases.⁴ Conversely, the Streptococcus viridans group is rarely implicated in septic arthritis, accounting for <1% of cases.⁴S viridans is part of the commensal oral flora and has low virulence. This heterogeneous group is subdivided into S mitis, S salivarius, S anginosus, S mutans, and S bovis. The S mitis group is further subdivided into S sanguinis (formerly known as S sanguis) and S mitis. Infection by an organism of the S viridans group usually occurs on a previously injured focus, and the organism is a causative agent of bacterial endocarditis.⁵ Reported cases of septic arthritis caused by S viridans have predominantly involved the knee joint—with severe osteoarthritis, poor dental hygiene, and prior IV drug use identified as risk factors.^5-7The shoulder joint is seldom involved in septic arthritis; estimated incidence is under 8%.⁸ Although overall incidence may rise in an increasingly elderly patient population, incidence of shoulder infection remains low.^2,9

 

The main routes for developing septic arthritis include direct inoculation secondary to penetrating trauma or hematologic spread.¹⁰ Coatsworth and colleagues¹¹ reported on iatrogenic S mitis septic arthritis of a shoulder arthroplasty during ultrasonography-guided aspiration by a technician who was not wearing a mask. Our institutional policy is to perform joint aspiration under strictly sterile conditions, which were adhered to in the present case. We surmise our patient developed transient bacteremia from the loss of several teeth, particularly given his poor dentition. Yombi and colleagues⁵ documented 2 cases of septic arthritis caused by Streptococcus gordonii, a relative of S sanguinis. One involved a previously replaced knee, and the other a native knee joint. Other cases of S viridans group septic arthritis have involved the knee,^6,7,12,13 the sternoclavicular joint,^14-16 and the acromioclavicular joint.¹⁷S sanguinis^6,7,12,15,16 and S gordonii⁵ have been implicated in most cases, and an unspeciated S viridans in others.^13,14,17 Concomitant periodontal disease has been reported in most cases as well,^6,7,12,15 including our patient’s case. In the English-language literature, we found no other reports of S mitis as the causative agent of acute septic glenohumeral joint arthritis from hematogenous spread.

There should be no delay in diagnosing septic arthritis, and infected material should be removed from the joint. In animal models, complete joint destruction occurred only 5 weeks after inoculation with Staphylococcus aureus.¹⁰ Garofalo and colleagues¹⁸ reported a trend toward improved functional outcomes after earlier operative treatment. The choice of open surgical drainage vs repeat needle aspiration seems to be of little consequence, as both have good long-term outcomes, but open surgical drainage seems to result in better long-term functional ROM.^2,9 However, results of a recent study suggested surgical treatment is not always superior to medical treatment for septic arthritis in native joints.¹⁹ In some cases involving S viridans species, treatment consisted of a combination of IV antibiotics and onetime or repeat aspiration;^6,12-15 treatment in the remaining cases was surgical débridement.^5,7,16,17 Given that S viridans is associated with bacterial endocarditis, echocardiography is essential if this organism is to be identified. Medical management and antibiotic treatment should be initiated after consultation with medical and infectious disease specialists.¹⁹We have reported a case of septic shoulder caused by S mitis, a low-virulence organism seldom associated with joint infection. The patient’s infection likely resulted from hematogenous spread from the oral cavity (dentition was poor). Urgent aspiration of the joint and baseline infection laboratory tests are recommended. MRI of the shoulder may show an effusion. Urgent arthroscopic irrigation and débridement can yield good clinical outcomes.

Am J Orthop. 2016;45(6):E343-E346. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

Septic arthritis predominantly involves the weight-bearing joints of the hip and knee, which account for nearly 60% of cases.¹ In contrast, the shoulder joint is involved in 10% to 15% of cases, though this number may be higher among intravenous (IV) drug users.² The most common causative organisms are the Staphylococcus species, followed closely by β-hemolytic streptococci, with these 2 groups accounting for more than 90% of all cases.³ The Streptococcus viridans group belongs to normal oral flora residing predominantly on the surface of teeth. Although well known for its ability to colonize heart valves and frequently cause bacterial endocarditis, this group has rarely been associated with septic arthritis. Furthermore, Streptococcus mitis, a subgroup of S viridans, has been implicated even less commonly.

In this article, we report a case of glenohumeral joint septic arthritis caused by S mitis. To our knowledge, such a case has not been previously reported in the English literature. Given the low virulence of this orally based bacterium, treating physicians must maintain clinical suspicion for the organism in the setting of persistent joint effusion and pain in association with periodontal disease or trauma. The patient provided written informed consent for print and electronic publication of this case report.

Case Report

A right-hand-dominant 54-year-old man presented to Dr. Gruson with complaints of persistent right shoulder pain associated with worsening range of motion (ROM). Three weeks earlier, the patient reported being assaulted and noted progressive swelling about the right shoulder. He denied fevers, chills, or prior shoulder problems. Although his past medical history was remarkable for hepatitis C and diabetes, he was not taking any diabetic medications at that time. A review of systems was remarkable for poor dental hygiene, and the patient was missing several teeth, which he said had been knocked out during the assault. Physical examination revealed diffuse tenderness about the right shoulder and severe pain with all passive movement. The shoulder was pseudoparalyzed. There were no subcutaneous collections, wounds, or ecchymosis about the shoulder. Mild calor was noted on the right shoulder relative to the left. Radiographs of the right shoulder showed no acute osseous abnormalities.

Magnetic resonance imaging (MRI), which was urgently obtained to assess the integrity of the rotator cuff and the location of the effusion, showed a large subacromial and glenohumeral joint effusion as well as diffuse muscular edema (Figures 1A-1C).

At follow-up, the patient reported having lost 10 pounds since his assault, as well as new-onset fevers and chills. C-reactive protein (CRP) level was 5.2 mg/dL (reference, <0.9 mg/dL), and erythrocyte sedimentation rate (ESR) was 48 mm/h (reference, <21 mm/h). White blood cell count was normal. Fluoroscopy-guided aspiration of the glenohumeral joint, performed under sterile conditions, yielded only 4 cc of hematoma. Gram stain was negative; though there was no growth on the primary plates, broth cultures grew S mitis. Repeat bloodwork demonstrated persistently increased CRP level (6.4 mg/dL) and ESR (55 mm/h).

In light of the elevated infection findings of the laboratory tests and the positive culture, urgent arthroscopic irrigation and débridement of the right shoulder were indicated. Given the organism identified, transesophageal echocardiography was performed; there were no valvular vegetations. Creation of the posterior glenohumeral portal resulted in egress of turbid fluid, which was sent for culture. The subacromial space and the glenohumeral joint were thoroughly lavaged and the copious hemorrhagic synovitis débrided (Figures 2A, 2B).

Chondral surfaces appeared grossly intact. All cultures from the surgery ultimately yielded S mitis. A peripherally inserted central catheter line was started, as was a 4-week course of IV ceftriaxone, as recommended by an infectious disease consultant. At postoperative visits in the orthopedic clinic, a new-onset right axillary abscess consisting of purulent material and organized hematoma was drained. After the ceftriaxone regimen was completed, a 4-week course of oral amoxicillin was started.

The 8-week course of antibiotics normalized the patient’s ESR to 13 mm/h. Follow-up MRI showed improvement in the soft-tissue edema. Clinically, the patient reported minimal shoulder pain. He was undergoing physical therapy to regain strength and ROM.

Discussion

Staphylococcus aureus is the leading causative organism of septic arthritis, accounting for more than 60% of all cases.⁴ Conversely, the Streptococcus viridans group is rarely implicated in septic arthritis, accounting for <1% of cases.⁴S viridans is part of the commensal oral flora and has low virulence. This heterogeneous group is subdivided into S mitis, S salivarius, S anginosus, S mutans, and S bovis. The S mitis group is further subdivided into S sanguinis (formerly known as S sanguis) and S mitis. Infection by an organism of the S viridans group usually occurs on a previously injured focus, and the organism is a causative agent of bacterial endocarditis.⁵ Reported cases of septic arthritis caused by S viridans have predominantly involved the knee joint—with severe osteoarthritis, poor dental hygiene, and prior IV drug use identified as risk factors.^5-7The shoulder joint is seldom involved in septic arthritis; estimated incidence is under 8%.⁸ Although overall incidence may rise in an increasingly elderly patient population, incidence of shoulder infection remains low.^2,9

 

The main routes for developing septic arthritis include direct inoculation secondary to penetrating trauma or hematologic spread.¹⁰ Coatsworth and colleagues¹¹ reported on iatrogenic S mitis septic arthritis of a shoulder arthroplasty during ultrasonography-guided aspiration by a technician who was not wearing a mask. Our institutional policy is to perform joint aspiration under strictly sterile conditions, which were adhered to in the present case. We surmise our patient developed transient bacteremia from the loss of several teeth, particularly given his poor dentition. Yombi and colleagues⁵ documented 2 cases of septic arthritis caused by Streptococcus gordonii, a relative of S sanguinis. One involved a previously replaced knee, and the other a native knee joint. Other cases of S viridans group septic arthritis have involved the knee,^6,7,12,13 the sternoclavicular joint,^14-16 and the acromioclavicular joint.¹⁷S sanguinis^6,7,12,15,16 and S gordonii⁵ have been implicated in most cases, and an unspeciated S viridans in others.^13,14,17 Concomitant periodontal disease has been reported in most cases as well,^6,7,12,15 including our patient’s case. In the English-language literature, we found no other reports of S mitis as the causative agent of acute septic glenohumeral joint arthritis from hematogenous spread.

There should be no delay in diagnosing septic arthritis, and infected material should be removed from the joint. In animal models, complete joint destruction occurred only 5 weeks after inoculation with Staphylococcus aureus.¹⁰ Garofalo and colleagues¹⁸ reported a trend toward improved functional outcomes after earlier operative treatment. The choice of open surgical drainage vs repeat needle aspiration seems to be of little consequence, as both have good long-term outcomes, but open surgical drainage seems to result in better long-term functional ROM.^2,9 However, results of a recent study suggested surgical treatment is not always superior to medical treatment for septic arthritis in native joints.¹⁹ In some cases involving S viridans species, treatment consisted of a combination of IV antibiotics and onetime or repeat aspiration;^6,12-15 treatment in the remaining cases was surgical débridement.^5,7,16,17 Given that S viridans is associated with bacterial endocarditis, echocardiography is essential if this organism is to be identified. Medical management and antibiotic treatment should be initiated after consultation with medical and infectious disease specialists.¹⁹We have reported a case of septic shoulder caused by S mitis, a low-virulence organism seldom associated with joint infection. The patient’s infection likely resulted from hematogenous spread from the oral cavity (dentition was poor). Urgent aspiration of the joint and baseline infection laboratory tests are recommended. MRI of the shoulder may show an effusion. Urgent arthroscopic irrigation and débridement can yield good clinical outcomes.

Am J Orthop. 2016;45(6):E343-E346. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

Septic arthritis predominantly involves the weight-bearing joints of the hip and knee, which account for nearly 60% of cases.¹ In contrast, the shoulder joint is involved in 10% to 15% of cases, though this number may be higher among intravenous (IV) drug users.² The most common causative organisms are the Staphylococcus species, followed closely by β-hemolytic streptococci, with these 2 groups accounting for more than 90% of all cases.³ The Streptococcus viridans group belongs to normal oral flora residing predominantly on the surface of teeth. Although well known for its ability to colonize heart valves and frequently cause bacterial endocarditis, this group has rarely been associated with septic arthritis. Furthermore, Streptococcus mitis, a subgroup of S viridans, has been implicated even less commonly.

In this article, we report a case of glenohumeral joint septic arthritis caused by S mitis. To our knowledge, such a case has not been previously reported in the English literature. Given the low virulence of this orally based bacterium, treating physicians must maintain clinical suspicion for the organism in the setting of persistent joint effusion and pain in association with periodontal disease or trauma. The patient provided written informed consent for print and electronic publication of this case report.

Case Report

A right-hand-dominant 54-year-old man presented to Dr. Gruson with complaints of persistent right shoulder pain associated with worsening range of motion (ROM). Three weeks earlier, the patient reported being assaulted and noted progressive swelling about the right shoulder. He denied fevers, chills, or prior shoulder problems. Although his past medical history was remarkable for hepatitis C and diabetes, he was not taking any diabetic medications at that time. A review of systems was remarkable for poor dental hygiene, and the patient was missing several teeth, which he said had been knocked out during the assault. Physical examination revealed diffuse tenderness about the right shoulder and severe pain with all passive movement. The shoulder was pseudoparalyzed. There were no subcutaneous collections, wounds, or ecchymosis about the shoulder. Mild calor was noted on the right shoulder relative to the left. Radiographs of the right shoulder showed no acute osseous abnormalities.

Magnetic resonance imaging (MRI), which was urgently obtained to assess the integrity of the rotator cuff and the location of the effusion, showed a large subacromial and glenohumeral joint effusion as well as diffuse muscular edema (Figures 1A-1C).

At follow-up, the patient reported having lost 10 pounds since his assault, as well as new-onset fevers and chills. C-reactive protein (CRP) level was 5.2 mg/dL (reference, <0.9 mg/dL), and erythrocyte sedimentation rate (ESR) was 48 mm/h (reference, <21 mm/h). White blood cell count was normal. Fluoroscopy-guided aspiration of the glenohumeral joint, performed under sterile conditions, yielded only 4 cc of hematoma. Gram stain was negative; though there was no growth on the primary plates, broth cultures grew S mitis. Repeat bloodwork demonstrated persistently increased CRP level (6.4 mg/dL) and ESR (55 mm/h).

In light of the elevated infection findings of the laboratory tests and the positive culture, urgent arthroscopic irrigation and débridement of the right shoulder were indicated. Given the organism identified, transesophageal echocardiography was performed; there were no valvular vegetations. Creation of the posterior glenohumeral portal resulted in egress of turbid fluid, which was sent for culture. The subacromial space and the glenohumeral joint were thoroughly lavaged and the copious hemorrhagic synovitis débrided (Figures 2A, 2B).

Chondral surfaces appeared grossly intact. All cultures from the surgery ultimately yielded S mitis. A peripherally inserted central catheter line was started, as was a 4-week course of IV ceftriaxone, as recommended by an infectious disease consultant. At postoperative visits in the orthopedic clinic, a new-onset right axillary abscess consisting of purulent material and organized hematoma was drained. After the ceftriaxone regimen was completed, a 4-week course of oral amoxicillin was started.

The 8-week course of antibiotics normalized the patient’s ESR to 13 mm/h. Follow-up MRI showed improvement in the soft-tissue edema. Clinically, the patient reported minimal shoulder pain. He was undergoing physical therapy to regain strength and ROM.

Discussion

Staphylococcus aureus is the leading causative organism of septic arthritis, accounting for more than 60% of all cases.⁴ Conversely, the Streptococcus viridans group is rarely implicated in septic arthritis, accounting for <1% of cases.⁴S viridans is part of the commensal oral flora and has low virulence. This heterogeneous group is subdivided into S mitis, S salivarius, S anginosus, S mutans, and S bovis. The S mitis group is further subdivided into S sanguinis (formerly known as S sanguis) and S mitis. Infection by an organism of the S viridans group usually occurs on a previously injured focus, and the organism is a causative agent of bacterial endocarditis.⁵ Reported cases of septic arthritis caused by S viridans have predominantly involved the knee joint—with severe osteoarthritis, poor dental hygiene, and prior IV drug use identified as risk factors.^5-7The shoulder joint is seldom involved in septic arthritis; estimated incidence is under 8%.⁸ Although overall incidence may rise in an increasingly elderly patient population, incidence of shoulder infection remains low.^2,9

 

The main routes for developing septic arthritis include direct inoculation secondary to penetrating trauma or hematologic spread.¹⁰ Coatsworth and colleagues¹¹ reported on iatrogenic S mitis septic arthritis of a shoulder arthroplasty during ultrasonography-guided aspiration by a technician who was not wearing a mask. Our institutional policy is to perform joint aspiration under strictly sterile conditions, which were adhered to in the present case. We surmise our patient developed transient bacteremia from the loss of several teeth, particularly given his poor dentition. Yombi and colleagues⁵ documented 2 cases of septic arthritis caused by Streptococcus gordonii, a relative of S sanguinis. One involved a previously replaced knee, and the other a native knee joint. Other cases of S viridans group septic arthritis have involved the knee,^6,7,12,13 the sternoclavicular joint,^14-16 and the acromioclavicular joint.¹⁷S sanguinis^6,7,12,15,16 and S gordonii⁵ have been implicated in most cases, and an unspeciated S viridans in others.^13,14,17 Concomitant periodontal disease has been reported in most cases as well,^6,7,12,15 including our patient’s case. In the English-language literature, we found no other reports of S mitis as the causative agent of acute septic glenohumeral joint arthritis from hematogenous spread.

There should be no delay in diagnosing septic arthritis, and infected material should be removed from the joint. In animal models, complete joint destruction occurred only 5 weeks after inoculation with Staphylococcus aureus.¹⁰ Garofalo and colleagues¹⁸ reported a trend toward improved functional outcomes after earlier operative treatment. The choice of open surgical drainage vs repeat needle aspiration seems to be of little consequence, as both have good long-term outcomes, but open surgical drainage seems to result in better long-term functional ROM.^2,9 However, results of a recent study suggested surgical treatment is not always superior to medical treatment for septic arthritis in native joints.¹⁹ In some cases involving S viridans species, treatment consisted of a combination of IV antibiotics and onetime or repeat aspiration;^6,12-15 treatment in the remaining cases was surgical débridement.^5,7,16,17 Given that S viridans is associated with bacterial endocarditis, echocardiography is essential if this organism is to be identified. Medical management and antibiotic treatment should be initiated after consultation with medical and infectious disease specialists.¹⁹We have reported a case of septic shoulder caused by S mitis, a low-virulence organism seldom associated with joint infection. The patient’s infection likely resulted from hematogenous spread from the oral cavity (dentition was poor). Urgent aspiration of the joint and baseline infection laboratory tests are recommended. MRI of the shoulder may show an effusion. Urgent arthroscopic irrigation and débridement can yield good clinical outcomes.

Am J Orthop. 2016;45(6):E343-E346. Copyright Frontline Medical Communications Inc. 2016. All rights reserved.

Bullous pemphigoid (BP) is an autoimmune subepidermal blistering disease.¹ The majority of BP cases are idiopathic and occur in patients older than 60 years. The disease is characterized by the development of circulating IgG autoantibodies reacting with the BP180 antigen of the basement membrane zone.¹ Psoriasis vulgaris (PV) is a common, chronic, immune-mediated disease affecting approximately 2% of the world’s population including children and adults.² Both entities may coexist with internal disorders such as hypertension, diabetes mellitus, coronary heart disease, congestive heart failure, hyperlipidemia, and cerebrovascular accident. It has been postulated that BP more often coexists with neurological disorders, such as stroke and Parkinson disease,³ whereas PV usually is associated with cardiovascular disorders and diabetes mellitus.² We report the case of a 35-year-old man with chronic PV and metabolic syndrome who developed BP that was successfully treated with methotrexate (MTX).

Case Report

A 35-year-old man with a 15-year history of PV, class 3 obesity (body mass index, 69.2), and thrombosis of the left leg was referred to the dermatology department due to a sudden extensive erythematous and bullous eruption located on the trunk, arms, and legs with involvement of the oral mucosa that had started 4 weeks prior. The skin lesions were accompanied by severe pruritus. On admission to the hospital, the patient presented with stable psoriatic plaques located on the trunk, arms, and proximal part of the lower legs with a psoriasis area severity index score of 11.8 (Figure 1A). He also had disseminated tense blisters and erosions partially arranged in an annular pattern located on the border of the psoriatic plaques as well as on an erythematous base or within unaffected skin (Figure 1B). Additionally, a few small erosions were present on the oral mucosa.

The patient’s father had a history of PV, but there was no family history of obesity or autoimmune blistering disorders. On physical examination, central obesity was noted with a waist circumference of 180 cm and a body mass index of 69.2; his blood pressure was 220/150 mm Hg. Laboratory tests revealed leukocytosis (20.06×10⁹/L [reference range, 4.5–11.0×10⁹/L]) with neutrophilia (16.2×10⁹/L [reference range, 1.6–7.6×10⁹/L]; 80.9% [reference range, 40.0%–70.0%]), eosinophilia (1.01×10⁹/L [reference range, 0–0.5×10⁹/L]), elevated C-reactive protein levels (49.4 mg/L [reference range, 0.0–9.0 mg/L]), elevated erythrocyte sedimentation rate (35 mm/h [reference range, 0–12 mm/h]), elevated γ-glutamyltransferase (66 U/L [reference range, 0–55 U/L]), decreased high-density lipoprotein levels (38 mg/dL [reference range, ≥40 mg/dL]), elevated fasting plasma glucose (116 mg/dL or 6.4 mmol/L [reference range, 70–99 mg/dL or 3.9–5.5 mmol/L]), elevated total IgE (1540 µg/L [reference range, 0–1000 µg/L]), elevated D-dimer (3.21 µg/mL [reference range, <0.5 µg/mL]), and low free triiodothyronine levels (130 pg/dL [reference range, 171–371 pg/dL]). The total protein level was 6.5 g/dL (reference range, 6.0–8.0 g/dL) and albumin level was 3.2 g/dL (reference range, 4.02–4.76 g/dL). A chest radiograph showed no abnormalities.

Based on the physical examination and laboratory testing, it was determined that the patient fulfilled 4 of 5 criteria for metabolic syndrome described by the International Diabetes Federation in 2006 (Table).⁴ Direct immunofluorescence performed on normal-appearing perilesional skin demonstrated linear IgG and C3 deposits along the basement membrane zone. Indirect immunofluorescence detected circulating IgG autoantibodies at a titer of 1:80. Serum studies using biochip mosaics⁵ revealed the reactivity of circulating IgG antibodies to the epidermal side of salt-split skin and with antigen dots of tetrameric BP180-NC16a, which prompted the diagnosis of BP (Figure 2).

Oral treatment with MTX 12.5 mg once weekly with clobetasol propionate cream applied to affected skin was initiated for 4 weeks. The PV resolved completely and blister formation stopped. A few weeks later BP reappeared, even though the patient was still taking MTX. The treatment failure may have been related to the patient’s class 3 obesity; therefore, the dose was increased to 20 mg once weekly for 8 weeks, which led to rapid healing of BP erosions. The patient was monitored for 2 months with no symptoms of recurrence.

 

Comment

Psoriasis Comorbidities

The correlation between PV and cardiovascular disorders such as myocardial infarction, cerebrovascular accident, and pulmonary embolism has been well established and is widely accepted.² It also has been documented that the risk for metabolic syndrome with components such as diabetes mellitus, hypertension, lipid abnormalities, obesity, and arteriosclerosis is notably increased in PV patients.⁶ Moreover, associated internal disorders are responsible for a 3- to 4-year reduction in life expectancy in patients with moderate to severe PV.⁷

Correlation of PV and BP

Psoriasis also may coexist with autoimmune disorders such as rheumatoid arthritis, lupus erythematosus, and blistering disorders.⁸ There are more than 60 known cases reporting PV in association with various types of subepidermal blistering diseases, including pemphigus vulgaris, epidermolysis bullosa acquisita, anti-p200 pemphigoid, and BP.^8,9 The pathogenetic relationship between BP and PV remains obscure. In most published cases, PV preceded BP by 5 to 30 years, possibly ascribable to patients being diagnosed with PV at a younger age.⁹ In general, patients with BP and PV are younger than patients with BP only, with a mean age of 62 years.⁹ Because our patient was in his mid-30s when he developed BP, in such cases physicians should take under consideration any triggering factors (eg, drugs). Physical examination and detailed laboratory findings allowed us to make the patient aware of the potential for development of metabolic syndrome. This condition in combination with PV could be a predisposing factor for BP development. According to more recent research, PV is considered a generalized inflammatory process rather than a disorder limited to the skin and joints.¹⁰ The chronic inflammatory process in psoriatic skin results in exposure of autoantigens, leading to an immune response and the production of BP antibodies. The neutrophil elastase enzyme present in psoriatic lesions also may take part in dermoepidermal junction degradation and blister formation of BP.¹¹ According to other observations, some antipsoriatic therapies (eg, psoralen plus UVA, UVB, dithranol, coal tar) could be associated with development of BP.¹² Moreover, it was shown that psoralen plus UVA therapy, which is widely used in PV treatment, alters the cytokine profile from helper T cells T_H1 to T_H2.¹² T_H2-dependent cytokines predominate the sera and erosions in BP patients and seem to be notably relevant to the pathophysiology of the disease.¹³ The history of our patient’s psoriatic treatment included only topical corticosteroids, keratolytic agents, and occasionally dithranol and coal tar; however, UV phototherapy or any other systemic therapies had never been utilized. Three previously reported cases of patients with PV and BP also revealed no history of UV phototherapy,^8,9 which suggests that mechanisms responsible for coexistence of PV and BP are more complex. It has been proven that proinflammatory cytokines secreted by T_H1 and T_H17 cells, in particular tumor necrosis factor α, IL-17, IL-22, and IL-23, play an important role in the development of psoriatic lesions.¹⁰ On the other hand, these cytokines are known to contribute to vascular inflammation, leading to development of arteriosclerosis, as well as to regulate adipogenesis and obesity.^14,15 Arakawa et al¹⁶ reported increased expression of IL-17 in lesional skin in BP. They concluded that IL-17 may contribute to the recruitment of eosinophils and neutrophils and tissue damage in BP. Therefore, it is highly likely that IL-17 might be a common factor underlying the coexistence of BP with PV and metabolic syndrome. More such reports are required for better understanding this association.

BP Treatment

Selecting a therapy for BP with coexistent PV is challenging, especially in patients with extreme obesity and metabolic syndrome. It is well established that obesity correlates with a higher incidence of PV and more severe disease. On the other hand, obesity also influences response to therapy. Systemic corticosteroids are contraindicated in psoriasis patients because of severe side effects, such as rebound phenomenon of psoriatic lesions and risk for development of generalized pustular PV. Although systemic corticosteroids are effective in BP, high-dose therapy may potentially be life-threatening, particularly in these obese patients with conditions such as hypertension and diabetes mellitus, among others,¹ as was observed in our case. Taking into consideration the above mentioned conditions and our experience on such cases, the current patient had received MTX (12.5 mg once weekly) and clobetasol propionate cream, which led to the rapid healing of the psoriatic plaques, whereas BP was more resistant to this therapy. This response may be explained by our patient’s class 3 obesity (body mass index, 69.2). Therefore, the dose of MTX was increased to 20 mg once weekly and was successful. The decision to use MTX was supported by evidence that this medicine may reduce the risk for arteriosclerosis and cardiovascular disorders.¹⁷

There are some alternative therapeutic options for patients with coexisting BP and PV, such as cyclosporine,¹⁸ combination low-dose cyclosporine and low-dose systemic corticosteroids,¹⁹ dapsone,²⁰ azathioprine,²¹ mycophenolate mofetil,²² and acitretin.²³ It also has been shown that biologics (eg, ustekinumab) may be a successful solution in patients with PV and antilaminin-γ1 pemphigoid.²⁴ However, these alternative therapeutic regimens could not be considered in our patient because of serious coexisting internal disorders.

 

Conclusion

We present a case of concomitant BP and PV in a patient with metabolic syndrome. Although the pathogenic role of this unique coexistence is not fully understood, MTX proved suitable and effective in this single case. Further studies should be performed to elucidate the pathogenic relationship and therapeutic solutions for cases with coexisting PV, BP, and metabolic syndrome.

Bullous pemphigoid (BP) is an autoimmune subepidermal blistering disease.¹ The majority of BP cases are idiopathic and occur in patients older than 60 years. The disease is characterized by the development of circulating IgG autoantibodies reacting with the BP180 antigen of the basement membrane zone.¹ Psoriasis vulgaris (PV) is a common, chronic, immune-mediated disease affecting approximately 2% of the world’s population including children and adults.² Both entities may coexist with internal disorders such as hypertension, diabetes mellitus, coronary heart disease, congestive heart failure, hyperlipidemia, and cerebrovascular accident. It has been postulated that BP more often coexists with neurological disorders, such as stroke and Parkinson disease,³ whereas PV usually is associated with cardiovascular disorders and diabetes mellitus.² We report the case of a 35-year-old man with chronic PV and metabolic syndrome who developed BP that was successfully treated with methotrexate (MTX).

Case Report

A 35-year-old man with a 15-year history of PV, class 3 obesity (body mass index, 69.2), and thrombosis of the left leg was referred to the dermatology department due to a sudden extensive erythematous and bullous eruption located on the trunk, arms, and legs with involvement of the oral mucosa that had started 4 weeks prior. The skin lesions were accompanied by severe pruritus. On admission to the hospital, the patient presented with stable psoriatic plaques located on the trunk, arms, and proximal part of the lower legs with a psoriasis area severity index score of 11.8 (Figure 1A). He also had disseminated tense blisters and erosions partially arranged in an annular pattern located on the border of the psoriatic plaques as well as on an erythematous base or within unaffected skin (Figure 1B). Additionally, a few small erosions were present on the oral mucosa.

The patient’s father had a history of PV, but there was no family history of obesity or autoimmune blistering disorders. On physical examination, central obesity was noted with a waist circumference of 180 cm and a body mass index of 69.2; his blood pressure was 220/150 mm Hg. Laboratory tests revealed leukocytosis (20.06×10⁹/L [reference range, 4.5–11.0×10⁹/L]) with neutrophilia (16.2×10⁹/L [reference range, 1.6–7.6×10⁹/L]; 80.9% [reference range, 40.0%–70.0%]), eosinophilia (1.01×10⁹/L [reference range, 0–0.5×10⁹/L]), elevated C-reactive protein levels (49.4 mg/L [reference range, 0.0–9.0 mg/L]), elevated erythrocyte sedimentation rate (35 mm/h [reference range, 0–12 mm/h]), elevated γ-glutamyltransferase (66 U/L [reference range, 0–55 U/L]), decreased high-density lipoprotein levels (38 mg/dL [reference range, ≥40 mg/dL]), elevated fasting plasma glucose (116 mg/dL or 6.4 mmol/L [reference range, 70–99 mg/dL or 3.9–5.5 mmol/L]), elevated total IgE (1540 µg/L [reference range, 0–1000 µg/L]), elevated D-dimer (3.21 µg/mL [reference range, <0.5 µg/mL]), and low free triiodothyronine levels (130 pg/dL [reference range, 171–371 pg/dL]). The total protein level was 6.5 g/dL (reference range, 6.0–8.0 g/dL) and albumin level was 3.2 g/dL (reference range, 4.02–4.76 g/dL). A chest radiograph showed no abnormalities.

Based on the physical examination and laboratory testing, it was determined that the patient fulfilled 4 of 5 criteria for metabolic syndrome described by the International Diabetes Federation in 2006 (Table).⁴ Direct immunofluorescence performed on normal-appearing perilesional skin demonstrated linear IgG and C3 deposits along the basement membrane zone. Indirect immunofluorescence detected circulating IgG autoantibodies at a titer of 1:80. Serum studies using biochip mosaics⁵ revealed the reactivity of circulating IgG antibodies to the epidermal side of salt-split skin and with antigen dots of tetrameric BP180-NC16a, which prompted the diagnosis of BP (Figure 2).

Oral treatment with MTX 12.5 mg once weekly with clobetasol propionate cream applied to affected skin was initiated for 4 weeks. The PV resolved completely and blister formation stopped. A few weeks later BP reappeared, even though the patient was still taking MTX. The treatment failure may have been related to the patient’s class 3 obesity; therefore, the dose was increased to 20 mg once weekly for 8 weeks, which led to rapid healing of BP erosions. The patient was monitored for 2 months with no symptoms of recurrence.

 

Comment

Psoriasis Comorbidities

The correlation between PV and cardiovascular disorders such as myocardial infarction, cerebrovascular accident, and pulmonary embolism has been well established and is widely accepted.² It also has been documented that the risk for metabolic syndrome with components such as diabetes mellitus, hypertension, lipid abnormalities, obesity, and arteriosclerosis is notably increased in PV patients.⁶ Moreover, associated internal disorders are responsible for a 3- to 4-year reduction in life expectancy in patients with moderate to severe PV.⁷

Correlation of PV and BP

Psoriasis also may coexist with autoimmune disorders such as rheumatoid arthritis, lupus erythematosus, and blistering disorders.⁸ There are more than 60 known cases reporting PV in association with various types of subepidermal blistering diseases, including pemphigus vulgaris, epidermolysis bullosa acquisita, anti-p200 pemphigoid, and BP.^8,9 The pathogenetic relationship between BP and PV remains obscure. In most published cases, PV preceded BP by 5 to 30 years, possibly ascribable to patients being diagnosed with PV at a younger age.⁹ In general, patients with BP and PV are younger than patients with BP only, with a mean age of 62 years.⁹ Because our patient was in his mid-30s when he developed BP, in such cases physicians should take under consideration any triggering factors (eg, drugs). Physical examination and detailed laboratory findings allowed us to make the patient aware of the potential for development of metabolic syndrome. This condition in combination with PV could be a predisposing factor for BP development. According to more recent research, PV is considered a generalized inflammatory process rather than a disorder limited to the skin and joints.¹⁰ The chronic inflammatory process in psoriatic skin results in exposure of autoantigens, leading to an immune response and the production of BP antibodies. The neutrophil elastase enzyme present in psoriatic lesions also may take part in dermoepidermal junction degradation and blister formation of BP.¹¹ According to other observations, some antipsoriatic therapies (eg, psoralen plus UVA, UVB, dithranol, coal tar) could be associated with development of BP.¹² Moreover, it was shown that psoralen plus UVA therapy, which is widely used in PV treatment, alters the cytokine profile from helper T cells T_H1 to T_H2.¹² T_H2-dependent cytokines predominate the sera and erosions in BP patients and seem to be notably relevant to the pathophysiology of the disease.¹³ The history of our patient’s psoriatic treatment included only topical corticosteroids, keratolytic agents, and occasionally dithranol and coal tar; however, UV phototherapy or any other systemic therapies had never been utilized. Three previously reported cases of patients with PV and BP also revealed no history of UV phototherapy,^8,9 which suggests that mechanisms responsible for coexistence of PV and BP are more complex. It has been proven that proinflammatory cytokines secreted by T_H1 and T_H17 cells, in particular tumor necrosis factor α, IL-17, IL-22, and IL-23, play an important role in the development of psoriatic lesions.¹⁰ On the other hand, these cytokines are known to contribute to vascular inflammation, leading to development of arteriosclerosis, as well as to regulate adipogenesis and obesity.^14,15 Arakawa et al¹⁶ reported increased expression of IL-17 in lesional skin in BP. They concluded that IL-17 may contribute to the recruitment of eosinophils and neutrophils and tissue damage in BP. Therefore, it is highly likely that IL-17 might be a common factor underlying the coexistence of BP with PV and metabolic syndrome. More such reports are required for better understanding this association.

BP Treatment

Selecting a therapy for BP with coexistent PV is challenging, especially in patients with extreme obesity and metabolic syndrome. It is well established that obesity correlates with a higher incidence of PV and more severe disease. On the other hand, obesity also influences response to therapy. Systemic corticosteroids are contraindicated in psoriasis patients because of severe side effects, such as rebound phenomenon of psoriatic lesions and risk for development of generalized pustular PV. Although systemic corticosteroids are effective in BP, high-dose therapy may potentially be life-threatening, particularly in these obese patients with conditions such as hypertension and diabetes mellitus, among others,¹ as was observed in our case. Taking into consideration the above mentioned conditions and our experience on such cases, the current patient had received MTX (12.5 mg once weekly) and clobetasol propionate cream, which led to the rapid healing of the psoriatic plaques, whereas BP was more resistant to this therapy. This response may be explained by our patient’s class 3 obesity (body mass index, 69.2). Therefore, the dose of MTX was increased to 20 mg once weekly and was successful. The decision to use MTX was supported by evidence that this medicine may reduce the risk for arteriosclerosis and cardiovascular disorders.¹⁷

There are some alternative therapeutic options for patients with coexisting BP and PV, such as cyclosporine,¹⁸ combination low-dose cyclosporine and low-dose systemic corticosteroids,¹⁹ dapsone,²⁰ azathioprine,²¹ mycophenolate mofetil,²² and acitretin.²³ It also has been shown that biologics (eg, ustekinumab) may be a successful solution in patients with PV and antilaminin-γ1 pemphigoid.²⁴ However, these alternative therapeutic regimens could not be considered in our patient because of serious coexisting internal disorders.

 

Conclusion

We present a case of concomitant BP and PV in a patient with metabolic syndrome. Although the pathogenic role of this unique coexistence is not fully understood, MTX proved suitable and effective in this single case. Further studies should be performed to elucidate the pathogenic relationship and therapeutic solutions for cases with coexisting PV, BP, and metabolic syndrome.

Bullous pemphigoid (BP) is an autoimmune subepidermal blistering disease.¹ The majority of BP cases are idiopathic and occur in patients older than 60 years. The disease is characterized by the development of circulating IgG autoantibodies reacting with the BP180 antigen of the basement membrane zone.¹ Psoriasis vulgaris (PV) is a common, chronic, immune-mediated disease affecting approximately 2% of the world’s population including children and adults.² Both entities may coexist with internal disorders such as hypertension, diabetes mellitus, coronary heart disease, congestive heart failure, hyperlipidemia, and cerebrovascular accident. It has been postulated that BP more often coexists with neurological disorders, such as stroke and Parkinson disease,³ whereas PV usually is associated with cardiovascular disorders and diabetes mellitus.² We report the case of a 35-year-old man with chronic PV and metabolic syndrome who developed BP that was successfully treated with methotrexate (MTX).

Case Report

A 35-year-old man with a 15-year history of PV, class 3 obesity (body mass index, 69.2), and thrombosis of the left leg was referred to the dermatology department due to a sudden extensive erythematous and bullous eruption located on the trunk, arms, and legs with involvement of the oral mucosa that had started 4 weeks prior. The skin lesions were accompanied by severe pruritus. On admission to the hospital, the patient presented with stable psoriatic plaques located on the trunk, arms, and proximal part of the lower legs with a psoriasis area severity index score of 11.8 (Figure 1A). He also had disseminated tense blisters and erosions partially arranged in an annular pattern located on the border of the psoriatic plaques as well as on an erythematous base or within unaffected skin (Figure 1B). Additionally, a few small erosions were present on the oral mucosa.

The patient’s father had a history of PV, but there was no family history of obesity or autoimmune blistering disorders. On physical examination, central obesity was noted with a waist circumference of 180 cm and a body mass index of 69.2; his blood pressure was 220/150 mm Hg. Laboratory tests revealed leukocytosis (20.06×10⁹/L [reference range, 4.5–11.0×10⁹/L]) with neutrophilia (16.2×10⁹/L [reference range, 1.6–7.6×10⁹/L]; 80.9% [reference range, 40.0%–70.0%]), eosinophilia (1.01×10⁹/L [reference range, 0–0.5×10⁹/L]), elevated C-reactive protein levels (49.4 mg/L [reference range, 0.0–9.0 mg/L]), elevated erythrocyte sedimentation rate (35 mm/h [reference range, 0–12 mm/h]), elevated γ-glutamyltransferase (66 U/L [reference range, 0–55 U/L]), decreased high-density lipoprotein levels (38 mg/dL [reference range, ≥40 mg/dL]), elevated fasting plasma glucose (116 mg/dL or 6.4 mmol/L [reference range, 70–99 mg/dL or 3.9–5.5 mmol/L]), elevated total IgE (1540 µg/L [reference range, 0–1000 µg/L]), elevated D-dimer (3.21 µg/mL [reference range, <0.5 µg/mL]), and low free triiodothyronine levels (130 pg/dL [reference range, 171–371 pg/dL]). The total protein level was 6.5 g/dL (reference range, 6.0–8.0 g/dL) and albumin level was 3.2 g/dL (reference range, 4.02–4.76 g/dL). A chest radiograph showed no abnormalities.

Based on the physical examination and laboratory testing, it was determined that the patient fulfilled 4 of 5 criteria for metabolic syndrome described by the International Diabetes Federation in 2006 (Table).⁴ Direct immunofluorescence performed on normal-appearing perilesional skin demonstrated linear IgG and C3 deposits along the basement membrane zone. Indirect immunofluorescence detected circulating IgG autoantibodies at a titer of 1:80. Serum studies using biochip mosaics⁵ revealed the reactivity of circulating IgG antibodies to the epidermal side of salt-split skin and with antigen dots of tetrameric BP180-NC16a, which prompted the diagnosis of BP (Figure 2).

Oral treatment with MTX 12.5 mg once weekly with clobetasol propionate cream applied to affected skin was initiated for 4 weeks. The PV resolved completely and blister formation stopped. A few weeks later BP reappeared, even though the patient was still taking MTX. The treatment failure may have been related to the patient’s class 3 obesity; therefore, the dose was increased to 20 mg once weekly for 8 weeks, which led to rapid healing of BP erosions. The patient was monitored for 2 months with no symptoms of recurrence.

 

Comment

Psoriasis Comorbidities

The correlation between PV and cardiovascular disorders such as myocardial infarction, cerebrovascular accident, and pulmonary embolism has been well established and is widely accepted.² It also has been documented that the risk for metabolic syndrome with components such as diabetes mellitus, hypertension, lipid abnormalities, obesity, and arteriosclerosis is notably increased in PV patients.⁶ Moreover, associated internal disorders are responsible for a 3- to 4-year reduction in life expectancy in patients with moderate to severe PV.⁷

Correlation of PV and BP

Psoriasis also may coexist with autoimmune disorders such as rheumatoid arthritis, lupus erythematosus, and blistering disorders.⁸ There are more than 60 known cases reporting PV in association with various types of subepidermal blistering diseases, including pemphigus vulgaris, epidermolysis bullosa acquisita, anti-p200 pemphigoid, and BP.^8,9 The pathogenetic relationship between BP and PV remains obscure. In most published cases, PV preceded BP by 5 to 30 years, possibly ascribable to patients being diagnosed with PV at a younger age.⁹ In general, patients with BP and PV are younger than patients with BP only, with a mean age of 62 years.⁹ Because our patient was in his mid-30s when he developed BP, in such cases physicians should take under consideration any triggering factors (eg, drugs). Physical examination and detailed laboratory findings allowed us to make the patient aware of the potential for development of metabolic syndrome. This condition in combination with PV could be a predisposing factor for BP development. According to more recent research, PV is considered a generalized inflammatory process rather than a disorder limited to the skin and joints.¹⁰ The chronic inflammatory process in psoriatic skin results in exposure of autoantigens, leading to an immune response and the production of BP antibodies. The neutrophil elastase enzyme present in psoriatic lesions also may take part in dermoepidermal junction degradation and blister formation of BP.¹¹ According to other observations, some antipsoriatic therapies (eg, psoralen plus UVA, UVB, dithranol, coal tar) could be associated with development of BP.¹² Moreover, it was shown that psoralen plus UVA therapy, which is widely used in PV treatment, alters the cytokine profile from helper T cells T_H1 to T_H2.¹² T_H2-dependent cytokines predominate the sera and erosions in BP patients and seem to be notably relevant to the pathophysiology of the disease.¹³ The history of our patient’s psoriatic treatment included only topical corticosteroids, keratolytic agents, and occasionally dithranol and coal tar; however, UV phototherapy or any other systemic therapies had never been utilized. Three previously reported cases of patients with PV and BP also revealed no history of UV phototherapy,^8,9 which suggests that mechanisms responsible for coexistence of PV and BP are more complex. It has been proven that proinflammatory cytokines secreted by T_H1 and T_H17 cells, in particular tumor necrosis factor α, IL-17, IL-22, and IL-23, play an important role in the development of psoriatic lesions.¹⁰ On the other hand, these cytokines are known to contribute to vascular inflammation, leading to development of arteriosclerosis, as well as to regulate adipogenesis and obesity.^14,15 Arakawa et al¹⁶ reported increased expression of IL-17 in lesional skin in BP. They concluded that IL-17 may contribute to the recruitment of eosinophils and neutrophils and tissue damage in BP. Therefore, it is highly likely that IL-17 might be a common factor underlying the coexistence of BP with PV and metabolic syndrome. More such reports are required for better understanding this association.

BP Treatment

Selecting a therapy for BP with coexistent PV is challenging, especially in patients with extreme obesity and metabolic syndrome. It is well established that obesity correlates with a higher incidence of PV and more severe disease. On the other hand, obesity also influences response to therapy. Systemic corticosteroids are contraindicated in psoriasis patients because of severe side effects, such as rebound phenomenon of psoriatic lesions and risk for development of generalized pustular PV. Although systemic corticosteroids are effective in BP, high-dose therapy may potentially be life-threatening, particularly in these obese patients with conditions such as hypertension and diabetes mellitus, among others,¹ as was observed in our case. Taking into consideration the above mentioned conditions and our experience on such cases, the current patient had received MTX (12.5 mg once weekly) and clobetasol propionate cream, which led to the rapid healing of the psoriatic plaques, whereas BP was more resistant to this therapy. This response may be explained by our patient’s class 3 obesity (body mass index, 69.2). Therefore, the dose of MTX was increased to 20 mg once weekly and was successful. The decision to use MTX was supported by evidence that this medicine may reduce the risk for arteriosclerosis and cardiovascular disorders.¹⁷

There are some alternative therapeutic options for patients with coexisting BP and PV, such as cyclosporine,¹⁸ combination low-dose cyclosporine and low-dose systemic corticosteroids,¹⁹ dapsone,²⁰ azathioprine,²¹ mycophenolate mofetil,²² and acitretin.²³ It also has been shown that biologics (eg, ustekinumab) may be a successful solution in patients with PV and antilaminin-γ1 pemphigoid.²⁴ However, these alternative therapeutic regimens could not be considered in our patient because of serious coexisting internal disorders.

 

Conclusion

We present a case of concomitant BP and PV in a patient with metabolic syndrome. Although the pathogenic role of this unique coexistence is not fully understood, MTX proved suitable and effective in this single case. Further studies should be performed to elucidate the pathogenic relationship and therapeutic solutions for cases with coexisting PV, BP, and metabolic syndrome.

Metastatic Crohn disease (MCD) is defined by the presence of cutaneous noncaseating granulomatous lesions that are noncontiguous with the gastrointestinal (GI) tract or fistulae.¹ The clinical presentation of MCD is so variable that its diagnosis requires a high index of suspicion.^1,2 In particular, the presence of erythematous tender nodules on the legs is easily mistaken for erythema nodosum (EN). Skin biopsy has an important role in confirming the diagnosis, as histopathological examination would reveal a noncaseating granuloma similar to those in the involved GI tract.² Herein, we report a case of MCD on the right leg that was clinically reminiscent of unilateral EN.

Case Report

A 21-year-old woman presented to the dermatology department with 2 painful erythematous nodules on the lower right leg of 2 weeks’ duration. She also reported abdominal pain, diarrhea, and bloody stool. She had been diagnosed with Crohn disease (CD) 6 years prior that had been well controlled with systemic low-dose steroids (5–15 mg/d), metronidazole (750 mg/d), and intermittent mesalamine and antidiarrheal drugs. However, she had not taken her medication for several weeks on her own authority. Subsequently, the patient developed skin lesions, which were characterized by ill-defined erythematous nodules with tenderness on the right lower leg along with GI symptoms (Figure 1). Laboratory studies revealed anemia (hemoglobin, 9.9 g/dL [reference range, 12.0–16.0 g/dL]) and an elevated C-reactive protein level (4.3 mg/dL [reference range, 0–0.3 mg/dL]). Other routine laboratory findings were normal.

Histopathologically, a skin biopsy from the right ankle showed vague, ill-defined, noncaseating granulomas scattered in the deep dermis and lobules of the subcutis (Figure 2). The granulomas were composed of epithelioid cells and Langerhans-type giant cells. Lymphocytes and neutrophils also were present, but eosinophils were absent. Immunohistochemical staining revealed that the infiltrating cells were mostly CD4⁺ helper/inducer T cells intermixed with CD8⁺ suppressor/cytotoxic T cells. The CD4:CD8 ratio was approximately 2:1. Counts of CD20⁺ B cells were low. Epithelioid cells and giant cells were positive for CD68.

A colonoscopy was performed to evaluate the aggravation of CD. Multiple longitudinal ulcers were observed in the ileocecal valve area and from the transverse colon to the sigmoid colon (Figure 3A). Histopathologic findings from the colon showed mucosal ulceration and noncaseating granulomas with heavy infiltration of lymphocytes and plasma cells (Figure 3B). Staining for infectious microorganisms (eg, Ziehl-Neelsen, periodic acid–Schiff, Gram) was negative. A polymerase chain reaction performed on sections cut from the paraffin block of the skin biopsy was negative for Mycobacterium tuberculosis DNA.

Based on the clinical and histopathologic findings, the patient was diagnosed with MCD that was clinically reminiscent of unilateral EN. Four weeks after the initiation of therapy with systemic corticosteroids (25 mg/d), oral metronidazole (750 mg/d), and mesalamine (1200 mg/d) for CD, the skin lesions were completely resolved and the patient’s GI symptoms improved simultaneously.

Comment

Crohn disease is a chronic inflammatory granulomatous disease of the GI tract that often is associated with reactive cutaneous lesions including EN, pyoderma gangrenosum, necrotizing vasculitis, and epidermolysis bullosa acquisita. Of these, EN is the most common to appear in CD patients and has been reported to occur in 1% to 15% of patients.^3-5 In particular, skin lesions on the leg presenting as tender erythematous nodules and patches are often diagnosed as EN, which is relatively common. In our case, we initially suspected EN due to the rare presentation of MCD and lack of specific clinical features; however, the skin biopsy revealed noncaseating granulomas in the mid to deep dermis and subcutis consistent with MCD.

Metastatic Crohn disease is a rare disease entity and is characterized by the presence of noncaseating granulomas of the skin at sites separated from the GI tract by normal tissue.¹ Although its pathogenesis is unclear, it has been suggested that immune complexes deposited in the skin could be responsible for the granulomatous reactions.⁴ A T lymphocyte–mediated type IV hypersensitivity reaction also could be responsible.^6,7 Because antimicrobial therapy can be curative for infection-related MCD, special histologic stains and/or tissue cultures can help to exclude an infectious etiology.⁸

Clinical presentations of MCD vary greatly, with observations such as single or multiple erythematous swellings, papules, plaques, nodules, abscesses, and ulcers.^1,2 The relationship between these clinical presentations and the intestinal activity of CD still is unknown; in some cases, however, the metastatic granulomatous lesions and the bowel disease show comparable severity.^2,9,10 In a review of the literature, MCD was generally reported to present in the genital area in children. In adults, lesions most frequently present in the genital area, followed by ulcers on the arms and legs.^1,2 These variations in clinical features and location resemble benign or infectious disease and can lead to delays in diagnosis.

 

Histopathologically, MCD lesions usually are ill-defined noncaseating granulomas with numerous multinucleated giant cells and lymphomononuclear cells located mostly in the dermis and occasionally extending into the subcutis. The cutaneous granulomata are similar to those present in the affected GI tract. Lymphocytes and plasma cells also are commonly present and eosinophils can be prominent.^1,2,11 In some cases of MCD, granulomatous vasculitis of small- to medium-sized vessels can be found and is associated with dermal and subcutaneous granulomatous inflammation.^8,11,12 Misago and Narisawa¹³ suggested that granulomatous vasculitis and panniculitis associated with CD is considered to be a rare subtype of MCD. Few cases of MCD presenting as granulomatous panniculitis have been described in the literature.^14-16 Our patient presented with lesions that clinically resembled EN; however, the biopsy was more consistent with MCD. The Table summarizes the distinguishing clinical and histopathological features of MCD in our case and classic EN.

Although some authors believe that MCD is not related to CD activity, others assert that MCD lesions may parallel GI activity.^1,2 Our patient was treated with systemic corticosteroids, oral metronidazole, and mesalamine to control the GI symptoms associated with CD. Four weeks after treatment, the GI symptoms and skin lesions improved simultaneously without any additional dermatologic treatment. We believe that MCD has the potential to serve as an early marker of the recurrence of CD and can help with the early diagnosis of CD aggravation, though an association between MCD and CD activity has not been confirmed.

Conclusion

We reported a case of MCD that was clinically reminiscent of unilateral EN and associated with GI disease activity. Physicians should be aware of the possibility of skin manifestations in CD, especially when erythematous nodular lesions are present on the leg.

Metastatic Crohn disease (MCD) is defined by the presence of cutaneous noncaseating granulomatous lesions that are noncontiguous with the gastrointestinal (GI) tract or fistulae.¹ The clinical presentation of MCD is so variable that its diagnosis requires a high index of suspicion.^1,2 In particular, the presence of erythematous tender nodules on the legs is easily mistaken for erythema nodosum (EN). Skin biopsy has an important role in confirming the diagnosis, as histopathological examination would reveal a noncaseating granuloma similar to those in the involved GI tract.² Herein, we report a case of MCD on the right leg that was clinically reminiscent of unilateral EN.

Case Report

A 21-year-old woman presented to the dermatology department with 2 painful erythematous nodules on the lower right leg of 2 weeks’ duration. She also reported abdominal pain, diarrhea, and bloody stool. She had been diagnosed with Crohn disease (CD) 6 years prior that had been well controlled with systemic low-dose steroids (5–15 mg/d), metronidazole (750 mg/d), and intermittent mesalamine and antidiarrheal drugs. However, she had not taken her medication for several weeks on her own authority. Subsequently, the patient developed skin lesions, which were characterized by ill-defined erythematous nodules with tenderness on the right lower leg along with GI symptoms (Figure 1). Laboratory studies revealed anemia (hemoglobin, 9.9 g/dL [reference range, 12.0–16.0 g/dL]) and an elevated C-reactive protein level (4.3 mg/dL [reference range, 0–0.3 mg/dL]). Other routine laboratory findings were normal.

Histopathologically, a skin biopsy from the right ankle showed vague, ill-defined, noncaseating granulomas scattered in the deep dermis and lobules of the subcutis (Figure 2). The granulomas were composed of epithelioid cells and Langerhans-type giant cells. Lymphocytes and neutrophils also were present, but eosinophils were absent. Immunohistochemical staining revealed that the infiltrating cells were mostly CD4⁺ helper/inducer T cells intermixed with CD8⁺ suppressor/cytotoxic T cells. The CD4:CD8 ratio was approximately 2:1. Counts of CD20⁺ B cells were low. Epithelioid cells and giant cells were positive for CD68.

A colonoscopy was performed to evaluate the aggravation of CD. Multiple longitudinal ulcers were observed in the ileocecal valve area and from the transverse colon to the sigmoid colon (Figure 3A). Histopathologic findings from the colon showed mucosal ulceration and noncaseating granulomas with heavy infiltration of lymphocytes and plasma cells (Figure 3B). Staining for infectious microorganisms (eg, Ziehl-Neelsen, periodic acid–Schiff, Gram) was negative. A polymerase chain reaction performed on sections cut from the paraffin block of the skin biopsy was negative for Mycobacterium tuberculosis DNA.

Based on the clinical and histopathologic findings, the patient was diagnosed with MCD that was clinically reminiscent of unilateral EN. Four weeks after the initiation of therapy with systemic corticosteroids (25 mg/d), oral metronidazole (750 mg/d), and mesalamine (1200 mg/d) for CD, the skin lesions were completely resolved and the patient’s GI symptoms improved simultaneously.

Comment

Crohn disease is a chronic inflammatory granulomatous disease of the GI tract that often is associated with reactive cutaneous lesions including EN, pyoderma gangrenosum, necrotizing vasculitis, and epidermolysis bullosa acquisita. Of these, EN is the most common to appear in CD patients and has been reported to occur in 1% to 15% of patients.^3-5 In particular, skin lesions on the leg presenting as tender erythematous nodules and patches are often diagnosed as EN, which is relatively common. In our case, we initially suspected EN due to the rare presentation of MCD and lack of specific clinical features; however, the skin biopsy revealed noncaseating granulomas in the mid to deep dermis and subcutis consistent with MCD.

Metastatic Crohn disease is a rare disease entity and is characterized by the presence of noncaseating granulomas of the skin at sites separated from the GI tract by normal tissue.¹ Although its pathogenesis is unclear, it has been suggested that immune complexes deposited in the skin could be responsible for the granulomatous reactions.⁴ A T lymphocyte–mediated type IV hypersensitivity reaction also could be responsible.^6,7 Because antimicrobial therapy can be curative for infection-related MCD, special histologic stains and/or tissue cultures can help to exclude an infectious etiology.⁸

Clinical presentations of MCD vary greatly, with observations such as single or multiple erythematous swellings, papules, plaques, nodules, abscesses, and ulcers.^1,2 The relationship between these clinical presentations and the intestinal activity of CD still is unknown; in some cases, however, the metastatic granulomatous lesions and the bowel disease show comparable severity.^2,9,10 In a review of the literature, MCD was generally reported to present in the genital area in children. In adults, lesions most frequently present in the genital area, followed by ulcers on the arms and legs.^1,2 These variations in clinical features and location resemble benign or infectious disease and can lead to delays in diagnosis.

 

Histopathologically, MCD lesions usually are ill-defined noncaseating granulomas with numerous multinucleated giant cells and lymphomononuclear cells located mostly in the dermis and occasionally extending into the subcutis. The cutaneous granulomata are similar to those present in the affected GI tract. Lymphocytes and plasma cells also are commonly present and eosinophils can be prominent.^1,2,11 In some cases of MCD, granulomatous vasculitis of small- to medium-sized vessels can be found and is associated with dermal and subcutaneous granulomatous inflammation.^8,11,12 Misago and Narisawa¹³ suggested that granulomatous vasculitis and panniculitis associated with CD is considered to be a rare subtype of MCD. Few cases of MCD presenting as granulomatous panniculitis have been described in the literature.^14-16 Our patient presented with lesions that clinically resembled EN; however, the biopsy was more consistent with MCD. The Table summarizes the distinguishing clinical and histopathological features of MCD in our case and classic EN.

Although some authors believe that MCD is not related to CD activity, others assert that MCD lesions may parallel GI activity.^1,2 Our patient was treated with systemic corticosteroids, oral metronidazole, and mesalamine to control the GI symptoms associated with CD. Four weeks after treatment, the GI symptoms and skin lesions improved simultaneously without any additional dermatologic treatment. We believe that MCD has the potential to serve as an early marker of the recurrence of CD and can help with the early diagnosis of CD aggravation, though an association between MCD and CD activity has not been confirmed.

Conclusion

We reported a case of MCD that was clinically reminiscent of unilateral EN and associated with GI disease activity. Physicians should be aware of the possibility of skin manifestations in CD, especially when erythematous nodular lesions are present on the leg.

Metastatic Crohn disease (MCD) is defined by the presence of cutaneous noncaseating granulomatous lesions that are noncontiguous with the gastrointestinal (GI) tract or fistulae.¹ The clinical presentation of MCD is so variable that its diagnosis requires a high index of suspicion.^1,2 In particular, the presence of erythematous tender nodules on the legs is easily mistaken for erythema nodosum (EN). Skin biopsy has an important role in confirming the diagnosis, as histopathological examination would reveal a noncaseating granuloma similar to those in the involved GI tract.² Herein, we report a case of MCD on the right leg that was clinically reminiscent of unilateral EN.

Case Report

A 21-year-old woman presented to the dermatology department with 2 painful erythematous nodules on the lower right leg of 2 weeks’ duration. She also reported abdominal pain, diarrhea, and bloody stool. She had been diagnosed with Crohn disease (CD) 6 years prior that had been well controlled with systemic low-dose steroids (5–15 mg/d), metronidazole (750 mg/d), and intermittent mesalamine and antidiarrheal drugs. However, she had not taken her medication for several weeks on her own authority. Subsequently, the patient developed skin lesions, which were characterized by ill-defined erythematous nodules with tenderness on the right lower leg along with GI symptoms (Figure 1). Laboratory studies revealed anemia (hemoglobin, 9.9 g/dL [reference range, 12.0–16.0 g/dL]) and an elevated C-reactive protein level (4.3 mg/dL [reference range, 0–0.3 mg/dL]). Other routine laboratory findings were normal.

Histopathologically, a skin biopsy from the right ankle showed vague, ill-defined, noncaseating granulomas scattered in the deep dermis and lobules of the subcutis (Figure 2). The granulomas were composed of epithelioid cells and Langerhans-type giant cells. Lymphocytes and neutrophils also were present, but eosinophils were absent. Immunohistochemical staining revealed that the infiltrating cells were mostly CD4⁺ helper/inducer T cells intermixed with CD8⁺ suppressor/cytotoxic T cells. The CD4:CD8 ratio was approximately 2:1. Counts of CD20⁺ B cells were low. Epithelioid cells and giant cells were positive for CD68.

A colonoscopy was performed to evaluate the aggravation of CD. Multiple longitudinal ulcers were observed in the ileocecal valve area and from the transverse colon to the sigmoid colon (Figure 3A). Histopathologic findings from the colon showed mucosal ulceration and noncaseating granulomas with heavy infiltration of lymphocytes and plasma cells (Figure 3B). Staining for infectious microorganisms (eg, Ziehl-Neelsen, periodic acid–Schiff, Gram) was negative. A polymerase chain reaction performed on sections cut from the paraffin block of the skin biopsy was negative for Mycobacterium tuberculosis DNA.

Based on the clinical and histopathologic findings, the patient was diagnosed with MCD that was clinically reminiscent of unilateral EN. Four weeks after the initiation of therapy with systemic corticosteroids (25 mg/d), oral metronidazole (750 mg/d), and mesalamine (1200 mg/d) for CD, the skin lesions were completely resolved and the patient’s GI symptoms improved simultaneously.

Comment

Crohn disease is a chronic inflammatory granulomatous disease of the GI tract that often is associated with reactive cutaneous lesions including EN, pyoderma gangrenosum, necrotizing vasculitis, and epidermolysis bullosa acquisita. Of these, EN is the most common to appear in CD patients and has been reported to occur in 1% to 15% of patients.^3-5 In particular, skin lesions on the leg presenting as tender erythematous nodules and patches are often diagnosed as EN, which is relatively common. In our case, we initially suspected EN due to the rare presentation of MCD and lack of specific clinical features; however, the skin biopsy revealed noncaseating granulomas in the mid to deep dermis and subcutis consistent with MCD.

Metastatic Crohn disease is a rare disease entity and is characterized by the presence of noncaseating granulomas of the skin at sites separated from the GI tract by normal tissue.¹ Although its pathogenesis is unclear, it has been suggested that immune complexes deposited in the skin could be responsible for the granulomatous reactions.⁴ A T lymphocyte–mediated type IV hypersensitivity reaction also could be responsible.^6,7 Because antimicrobial therapy can be curative for infection-related MCD, special histologic stains and/or tissue cultures can help to exclude an infectious etiology.⁸

Clinical presentations of MCD vary greatly, with observations such as single or multiple erythematous swellings, papules, plaques, nodules, abscesses, and ulcers.^1,2 The relationship between these clinical presentations and the intestinal activity of CD still is unknown; in some cases, however, the metastatic granulomatous lesions and the bowel disease show comparable severity.^2,9,10 In a review of the literature, MCD was generally reported to present in the genital area in children. In adults, lesions most frequently present in the genital area, followed by ulcers on the arms and legs.^1,2 These variations in clinical features and location resemble benign or infectious disease and can lead to delays in diagnosis.

 

Histopathologically, MCD lesions usually are ill-defined noncaseating granulomas with numerous multinucleated giant cells and lymphomononuclear cells located mostly in the dermis and occasionally extending into the subcutis. The cutaneous granulomata are similar to those present in the affected GI tract. Lymphocytes and plasma cells also are commonly present and eosinophils can be prominent.^1,2,11 In some cases of MCD, granulomatous vasculitis of small- to medium-sized vessels can be found and is associated with dermal and subcutaneous granulomatous inflammation.^8,11,12 Misago and Narisawa¹³ suggested that granulomatous vasculitis and panniculitis associated with CD is considered to be a rare subtype of MCD. Few cases of MCD presenting as granulomatous panniculitis have been described in the literature.^14-16 Our patient presented with lesions that clinically resembled EN; however, the biopsy was more consistent with MCD. The Table summarizes the distinguishing clinical and histopathological features of MCD in our case and classic EN.

Although some authors believe that MCD is not related to CD activity, others assert that MCD lesions may parallel GI activity.^1,2 Our patient was treated with systemic corticosteroids, oral metronidazole, and mesalamine to control the GI symptoms associated with CD. Four weeks after treatment, the GI symptoms and skin lesions improved simultaneously without any additional dermatologic treatment. We believe that MCD has the potential to serve as an early marker of the recurrence of CD and can help with the early diagnosis of CD aggravation, though an association between MCD and CD activity has not been confirmed.

Conclusion

We reported a case of MCD that was clinically reminiscent of unilateral EN and associated with GI disease activity. Physicians should be aware of the possibility of skin manifestations in CD, especially when erythematous nodular lesions are present on the leg.