The Use of Cellular- and/or Tissue-Based Therapy for the Management of Pyoderma Gangrenosum: A Case Series and Review of the Literature

Pyoderma gangrenosum (PG) is a rare inflammatory neutrophilic disorder that causes progressive ulcers as part of an autoimmune disease process.¹ The clinical presentation is variable, although typically it is characterized by skin lesions that progress to painful necrotizing ulcerations with blue-hued purpuric undermined edges and skin sloughing (Figure 1A). The most commonly affected site is the lower extremity, with solitary ulcerations or a disseminated presentation; however, PG may also involve the head, neck, and perianal area that are more often found in infants and children.^1-3 Other recognized variants of PG are bullous, pustular, vegetative, peristomal, extracutaneous, and postsurgical, but ulcerative PG remains the most common type.¹ 

Pyoderma gangrenosum affects individuals of all ages, with a peak incidence in young to middle-aged adults; PG has no sex predilection.^1,4,5 It has been associated with underlying diseases in up to 75% of cases.⁴ The most common underlying diseases associated with PG are inflammatory bowel disease (IBD) (34%–65%), polyarthritis (16%–19%), and hematologic disorders (12%–20%).⁴ Because PG is associated with systemic autoimmune diseases, its pathogenesis is believed to be immunologic.^1,5

The diagnosis of PG is made based on clinical judgment and by exclusion because no specific diagnostic criteria exist. Research suggests biopsy can aid in the diagnosis if neutrophilic infiltration of the dermis and skin is seen on pathologic examination.² However, a false-negative biopsy result is possible; alternatively, biopsy may potentiate an exaggerated skin injury and rapid progressive ulcerations of the wound. This characteristic phenomenon is called pathergy, which can occur after minor injury to the skin via surgery, trauma, biopsy, and even venipuncture.^1-6 

Surgical interventions such as radical debridement with full-thickness skin graft, split-thickness skin graft (STSG), and even free flaps have been reported in the literature as potential therapy for extensive PG wounds that require coverage.^6-8 Concurrent medical immunosuppressive therapy to control the inflammatory response is also vital to avoid the need for surgery, because surgical intervention can trigger pathergy and disease progression.^1,2,4-7,9

This case series and literature review provides an overview of surgical treatments and the application of cellular- and/or tissue-based products (CTPs) for PG. Literature was identified based on English-language key word search in PubMed using the key words “pyoderma gangrenosum”, “skin graft”, and “skin substitute”; in addition, the commercial names of 72 commercially available CTPs are included. 

Case 1

A 48-year-old male with a history of ulcerative colitis (UC) presented to the emergency department (ED) for a painful purpuric lesion on the right shin that showed no improvement after a 2-week course of oral trimethoprim/sulfamethoxazole and doxycycline. Prior to presentation the patient had been evaluated by his primary care physician when the wound had rapidly progressed from a small lesion the patient had unroofed. The patient had been treated with mercaptopurine (PURINETHOL; Teva Pharmaceuticals USA) and mesalamine (Asacol HD; Allergan) for UC and had not had any ulcerative lesions since diagnosis. The patient denied any gastrointestinal (GI) issues, such as bloody diarrhea or abdominal pain. A dose of vancomycin and cefepime was given by the ED physician preemptively, but it was not continued during the patient’s hospitalization. 

Physical examination revealed a 10-cm × 8-cm hyperemic shallow ulceration with violaceous borders and skin sloughing circumferentially at the edges on the right shin. Figure 1A shows the appearance of the lesion on admission. Blood tests showed a normal white blood cell (WBC) count of 8.3 K/uL (range, 4.5–11.0 K/uL), an elevated erythrocyte sedimentation rate (ESR) of 75 mm/hour (range, 0–13 mm/hour), and a C-reactive protein (CRP) level of 89.35 mg/L (range, 0-5.1 mg/L). Dermatology and gastroenterology specialists were consulted for optimization of medical management of UC, and the patient was restarted on home medications for UC as previously mentioned and prescribed 1 prednisone 50 mg tablet daily. 

Because of its size and to aid in slowing down progression, the cutaneous lesion was treated with CTP. Sharp debridement of the edges of sloughed skin was performed with a scalpel under local anesthesia. Fetal bovine dermis (FBD) (PriMatrix; Integra LifeSciences) was applied and sutured in place with simple interrupted 3-0 chromic gut sutures. Biopsy of the wound edge was also done at the time of debridement, revealing skin with ulceration, severe acute inflammation and abscess, and absence of giant cells. The FBD was bolstered to the wound bed with nonadherent cellulose acetate mesh (ADAPTIC TOUCH Non-Adhering Silicone Dressing; 3M) and wrapped with a multilayered zinc oxide compression boot. 

The dressing was removed on postoperative day 4, at which time the wound appeared less hyperemic and exhibited successful engraftment of FBD (Figure 1B). A new multilayer dressing was placed, and the patient was discharged on a prednisone taper. Steroid treatment duration and type were dictated by dermatology and gastroenterology specialists. 

By the follow-up appointment with the dermatology specialist 1 week later, the wound had contracted around the edges and granulation tissue had formed at the base. The borders were clean and without purpura (Figure 1C). Triamcinolone acetonide 80 mg (Kenalog 20 mg/mL) was injected at the periphery of the wound. The patient was also evaluated at the authors’ clinic on the same day at which time the dressing was reapplied. At 1-month follow-up, the wound bed was almost completely reepithelialized, with small areas of ulceration at the edges (Figure 1D). 

Case 2

A 31-year-old female with a history of UC presented to the ED with worsening right foot pain for 1 month. She had been evaluated in the ED multiple times previously with similar issues, but laboratory values and radiologic images demonstrated no signs of infection and the patient was discharged home on oral antibiotics. By the time of this most recent visit the patient’s pain had progressed to the point that she could not ambulate without difficulty; at that time these authors were consulted in the ED. The patient had been on mesalamine when she was diagnosed with acute flare of UC 2 years previously, but at the time she was not undergoing treatment with corticosteroids or immunomodulators. The patient also denied any GI issues. 

Physical examination revealed marked edema of the right medial forefoot to mid shin and ruptured bullae with overlying eschar that was exquisitely tender to palpation (Figure 2A). Pedal pulses were palpable bilaterally. Intravenous (IV) vancomycin and ceftriaxone was started for a presumed working differential diagnosis of PG and cellulitis. Blood tests showed a normal WBC count of 8.3 K/uL as well as an elevated ESR and CRP level of 78 mm/hr and 58.3 mg/L, respectively. Magnetic resonance imaging of the foot demonstrated osteomyelitis of the navicular bone with devitalized soft tissue. The dermatology and gastroenterology specialists who were consulted did not believe the wound was PG and deferred treatment or workup. 

The patient underwent wide hydrosurgical debridement; FBD was used to cover the wound and was sutured in place with 4-0 chromic gut (Figure 2B, 2C). The authors’ protocol for use of tangential hydrosurgical debridement includes cutting the skin edges back to a 45° angle to healthy bleeding tissue under ×2.5 loupe magnification. The base of the wound was then debrided with tangential hydrosurgery until punctate bleeding was observed under ×2.5 loupe magnification. Skin biopsy was also done at the time of debridement, which showed skin with underlying abscess that was nonspecific with neutrophilic infiltration and hemorrhage. The wound was covered with a petrolatum-based gauze and topical bacitracin and bolstered to the wound bed with moist gauze and compressive wrap. Wound culture of the drainage grew no organisms after 5 days. The patient was discharged on 6 weeks of oral antibiotics given the underlying abscess. By 3-week follow-up, the wound exhibited adequate graft take (Figure 2D). Fetal bovine dermis that was not fully engrafted to the wound was excised, and the base of the wound was pulse irrigated with saline. Split-thickness skin graft was harvested from the contralateral hip and secured to the wound with nonadherent mesh and negative pressure wound therapy (NPWT). On postoperative day 4, NPWT was discontinued and the patient was discharged with a multilayer compression dressing. The STSG appeared to exhibit 100% take and complete healing after 1 month (Figure 2E). 

Three months postoperatively, however, the patient had an acute UC flare and was prescribed hydrocortisone for 2 weeks; she declined to continue immunosuppressant therapy when her GI symptoms resolved. She was later examined in the authors’ clinic for swelling over the medial malleolus of the STSG site. The fluid was aspirated and sent for culture, which was found to be negative for any organisms. The patient presented to the ED the follow week with purple discoloration at the site of aspiration and intense pain. The patient reported subjective fevers at home and indicated she was taking oral antibiotics prescribed by her primary care physician for a presumed cough. On physical examination, the 20% of the graft at its center appeared to have a violaceous plaque with hemorrhagic oozing. Blood tests showed an elevated ESR and CRP level with a normal WBC count. At the time of presentation, the patient was experiencing infrequent episodes of diarrhea and bright red blood per rectum, but she continued to decline immunosuppressive therapy. The patient returned to the operating room for debridement of the necrotic center of the wound and copious irrigation with hydrogen peroxide. She was advised to soak her foot daily in hydrogen peroxide and cover her wound with petrolatum-based gauze. The patient was lost to follow-up after discharge.

Case 3

A 35-year-old male with a history of sickle cell anemia presented to an outpatient vascular surgery clinic with an ulcer of the right lower extremity (Figure 3A). The patient was treated with local wound care and compression therapy for possible venous stasis ulcer. At 2-week follow-up, the wound appeared to be enlarging with erythema and induration around the borders, measuring 6 cm × 7 cm and 3 cm × 2 cm (Figure 3B). The patient was sent to the ED and was found to have a normal WBC count with an elevated ESR (150 mm/hour) and CRP level (51.3 mg/L). He was again presumed to have worsening of the venous stasis ulcer and was admitted for IV antibiotics and daily dressing changes with SILVADENE Cream (Monarch Pharmaceuticals). 

After discharge, the patient was evaluated in the authors’ clinic for a second opinion. At that time, the decision was made to start the patient on prednisone owing to high suspicion for PG. However, his wounds had coalesced and enlarged to a size of 11 cm × 8 cm (Figure 3C), and he was readmitted to the hospital for IV antibiotics and local wound care. The dermatology specialist who was consulted for PG recommended nonsurgical management with outpatient follow-up. The medical team also adamantly refused to allow the patient to undergo CTP placement. The patient was discharged after 10 days of treatment with IV antibiotics, steroids, and local wound care. He was evaluated in the dermatology clinic 2 days after discharge; at that time he was on a steroid taper and received intralesional Kenalog 20 mg/mL injections. Weekly follow-up was continued for 3 months, and the patient eventually underwent NPWT to manage excessive drainage from the wound (Figure 3D). Surgical management with debridement and placement of CTPs was offered, but the patient declined and continued NPWT for 4 months. At the end of NPWT treatment, the wound was approximately one-half its original size (Figure 3E). 

Because of the variable presentation and course of healing or pathergy, many patients with PG are often misdiagnosed and thus experience delayed treatment. Past medical history of autoimmune diseases, including IBD, can offer contextual clues for diagnosing PG, which often occurs during acute flares with GI symptoms. Pyoderma gangrenosum has a prevalence of 0.5% in patients with IBD, with UC more common than in Crohn disease.¹ The presence of an inflammatory nidus in patients with autoimmune disease and elevated inflammatory mediators in lesions of PG suggest a pathological inflammatory process that contributes to its development.^1,4,5 However, 25% to 50% of patients with PG either have no apparent underlying disorder or develop PG after surgical trauma. A high level of suspicion is needed to promptly start medical treatment and to avoid multiple debridements that have the potential to worsen lesions.^4,6,8,10,11 The patients in this study (cases 1 and 2) did not present with any GI symptoms, which suggests that PG lesions can occur even in quiescent stages of IBD.

Diagnosis is based on clinical judgement, and many physicians believe that biopsy is not mandatory for diagnosis of PG.² Biopsy results are often nondiagnostic or nonspecific, and a misdiagnosis rate of approximately 10% was reported in one study.¹² Histopathologic studies often show edema, neutrophilic inflammation, necrosis, and abscess formation resulting from infiltration of acute inflammatory cells.^2,13,14 Although biopsy has been deferred to avoid pathergy, the literature shows that a full-thickness biopsy at the periphery of the lesion is warranted to exclude other diseases, including vasculitis, malignancy, and drug-induced tissue injury.^2,12,14 For case 1, biopsy was done after the patient had been started on systemic steroids. No pathergic response occurred. Similarly, no pathergic response occurred in case 2, in which immunosuppression was not used. In the authors’ experience, the most notable biopsy findings include the relative sterility of the wound compared with the massive neutrophil response, and the lack of a systemic leukocytosis. 

The etiology and pathogenesis of PG are still debated, and as a result, there is no consensus regarding treatment options. The literature indicates that preventing propagation of ulceration requires recognizing underlying immune disease and starting systemic steroids to reduce inflammation.^1,2,5-8 Historically, surgical intervention was discouraged because grafts had poor take and because of the risk of new ulcerations at donor sites, aggravation of the wound, or the onset of pathergy.^15,16 With advances in immunosuppressive therapies, however, surgical interventions have become a safe option because pathergy can be prevented and the inflammatory response dampened. Niezgoda et al,¹⁷ who are proponents of surgical debridement, concluded in their case study that excisional debridement of all necrotic tissue is essential to early treatment and the prevention of pathergy. A subsequent case report indicated that wounds that did not heal with steroid therapy alone demonstrated complete healing after debridement and primary repair.¹⁸ For case 2 reported herein, hydrosurgical debridement was performed; it is not known whether such debridement has the potential to cause the same tissue damage or pathergy as sharp debridement. When performing tangential hydrosurgery, the device was set to take 0.015 inch per pass, which affords greater control of amount of tissue removed from the wound compared with sharp debridement. Overall, surgery has been shown to accelerate wound healing and allow quick tapering of steroid dosing, thereby reducing the potential complication of immunosuppression and its side effects.^17-19

The type of surgical therapy performed is dependent on the extent and location of the PG lesions. In limited case reports, large soft tissue defects involving joints or exposed tendons or bones have been successfully managed with full-thickness skin grafts, local muscle flaps, and microvascular tissue transfer.^11,13,20 The authors of those case reports postulated that transferring well-vascularized tissue to the infected wound bed can aid in enhancing the immunologic milieu and can deliver oxygen necessary for healing. Few reported flap necrosis or recurrence at the wound periphery, which were able to be treated with only medical immunosuppresion.^11,13,20 Likewise, STSG has been used extensively for large wounds that continued to progress despite adequate medical therapy alone. Most case series and reports on surgical treatment propose the use of STSG for early closure and improving pain control by covering raw open surfaces of PG ulcers.^19,21,22 Use of cadaveric allograft has also resulted in successful engraftment of PG ulcers and has the benefit of avoiding any potential pathergy that could occur when harvesting autografts.⁹ Although rare, donor site pathergy even in the setting of adequate immunosuppression has been reported.^21,23 Early failures with STSG were reported by Rozen et al²⁴ when performed on the recipient site at the time of initial debridement, likely owing to a persistent inflammatory process and a poorly prepared recipient site. Rozen et al²⁴ proposed serial skin allografts. Other physicians have also used NPWT to prepare the wound bed before placing autografts.^19,21-23

Given the concern for potential pathergy, CTPs have also been used to limit donor site morbidity and to prepare a robust wound bed for effective graft fixation in complex wounds. Cellular and tissue-based products have been used to manage tissue defects by restoring normal barrier function while stimulating wound repair responses by providing matrix elements and growth factors.^25,26 Application of CTPs has been well-documented in the literature for use on burns and chronic wounds, including ulcers caused by diabetes mellitus and venous insufficiency.^25,27 An overview of commercial CTPs used to aid in accelerating PG wound closure is presented herein.

Apligraf graft skin (Organogensis) is a bilayered living cellular tissue composed of bovine collagen gel with neonatal fibroblasts for the dermis and keratinocytes for the epidermal layer.²⁶ A case report by Neiderer et al²⁸ details a patient with rheumatoid arthritis who presented with a small anterior leg ulcer that was presumed to be the result of chronic venous stasis. The wound had increased in size with only local wound care, but it also continued to increase in size after 9 months of treatment with prednisone and topical tacrolimus after a delayed diagnosis of PG. Bilayered living cellular tissue was applied every 2 weeks for a total of 5 applications with NPWT. Wound size decreased after 12 weeks, at which time it was possible to taper the prednisone. Full epithelization was observed after 16 weeks of treatment.²⁸ Similarly, Duchini et al²⁹ and de Imus et al³⁰ noted poor healing with use of only systemic corticosteroids and triamcinolone acetonide injected into the lesion. In both case reports, bilayered living cellular tissue was placed to decrease wound contracture and hasten healing.^29,30 de Imus et al³⁰ observed significantly flattened wound edges at the graft site as well as the formation of healthy granulation tissue after 1 week, with complete reepithelialization after 6 weeks. In all 3 case reports, quick tapering of immunosuppressants and improvement in pain control and cosmesis occurred.^28-30

As part of an extensive literature review on STSG, Eisendle et al²³ report on 2 patients who underwent application of EZ Derm (Mölnlycke Health Care). This aldehyde cross-linked porcine xenograft (PX) is an acellular dermal matrix commonly used in patients with burns.^23,31 Unlike STSG, use of PX did not require anesthesia and its placement was performed under local tumescence.²³ One patient refused closure by STSG after complete granulation tissue was formed and was instead treated with PX and NPWT. Complete wound healing by secondary intention was achieved in 6 months. Another patient underwent STSG after PG ulcer healing 3 weeks after application of PX. The authors of that study were in favor of the material because it was less expensive than other skin substitutes and provided similar benefits in reduction of pain and infection.²³ 

Another commonly used tissue-based product is Integra (Integra LifeSciences), a dermal regenerative template (DRT) with an outer layer made of thin silicone film that acts as the epidermis and an inner layer of a matrix of cross-linked fibers. In one case report, 2 patients with pretibial PG wounds were treated with DRT placed over periosteal tissue with a NPWT dressing.³² The dressings were changed every 5 days to 7 days, and within 21 days there was retraction of wound edges, quality coverage of the defect, and no violaceous borders. In addition, prior to application of DRT the wound bed was prepared with NPWT for 10 to 15 days. After DRT placement, the silicone lamina was removed and covered with STSG. The authors found no recurrence of PG at either 1-year or 2-year follow-up.³² 

Complications of PG after bilateral mastopexies have also been reported; these were misdiagnosed as necrotizing soft tissue infections and worsened with serial debridement.^14,33 Once systemic steroids were begun after diagnosis of PG, the wounds were deemed too large to heal by secondary intention and it was believed poor cosmesis would result. The authors also noted that STSG alone might not have been successful as the first surgical therapy and thus, DRT was applied 2 to 4 weeks after STSG.^14,33 In a report by Gottlieb and Furman,³⁴ 21 of 111 patients had lower extremity ulcers of primarily immunopathic origin (including PG) that were managed with DRT and STSG. Time to complete wound epithelialization ranged from 2 months to 18 months. Of the 21 patients, 14 experienced complete wound healing and 4 experienced a decrease in wound size by one-third to two-thirds. Recurrences seemed to coincide with flare-up of autoimmune disease. The authors reported that when DRT was applied to the wound, the inflammatory pathway halted and no microscopic inflammatory cell infiltrates were observed in the wound matrix. They also noted that the chondroitin matrix had an appearance similar to that of normal tissue and implied that DRT could be “invisible” to or recognized as “self” by platelets and inflammatory leukocytes, thus enabling suppression of inflammation and pathergy.³⁴ 

In the present series and in the 3 cases reported in this study, PriMatrix (Integra LifeSciences) was used because of its potential for engraftment, general resilience to bacterial colonization, and the need for only 1 application. Fetal bovine dermis is an acellular dermal matrix that is designed to provide an ideal environment to support cellular repopulation and revascularization necessary for wound healing and tissue generation. It is uniquely rich in type III collagen, which is necessary for developing and healing tissues.^35,36 At the authors' institution, FBD has been used on numerous chronic leg and foot ulcers, with positive outcomes in patients with ulcers caused by diabetes mellitus or venous stasis ulcers.³⁷ In that representative case, gentle debridement of the wound edges was done to excise all necrotic tissue, and only 1 application of FBD was necessary. The patient’s pain level decreased during the healing process, and nearly complete reepithelialization had occurred at 1-month follow-up. Prednisone was also tapered and discontinued within 20 days. Since the collagen matrix of FBD is not degraded or reabsorbed, but rather incorporates into the wound bed and is revascularized, successful healing outcomes have occurred with less frequent applications compared with other tissue-based products. A major limitation of these bioengineered skin substitutes is cost. Evaluating the total cost associated with different therapies, including the price of the therapy itself, hospital stay, number of applications, and time to healing may aid clinicians in choosing between products. 

The authors of this case report chose to use FBD for case 2 as a preliminary step to prepare the wound bed. Because of the location of the wound at the ankle joint and concern for contraction, the patient underwent STSG for secondary closure. This patient required 2 separate hospitalizations—1 for administering FBD and another for STSG 3 weeks later (Table). At the time of STSG, it was necessary to excise devitalized borders of the wound that did not take FBD. By 3 months, however, full wound healing was achieved. In contrast, the patient in case 1 had areas of ulcerations within the wound at 3-month follow-up. Unfortunately, the patient in case 2 was not receptive to steroid therapy and experienced a large recurrence of PG in the center of her healed STSG site. 

Negative pressure wound therapy and hyperbaric oxygen therapy have commonly been used as adjunct tools to aid healing.^17,23 Without immunosuppression, NPWT alone was seen to cause increased pain and significant skin maceration under the adhesive.³⁸ However, Eisendle et al²³ found that with adequate steroid treatment the percentage of overall skin graft take in PG lesions was higher if grafts were secured by NPWT. The rationale for this finding is that NPWT applies mechanical stress to the wound surface that enhances cellular proliferation and removes inflammatory factors that aggravate wound healing.^22,23 Negative pressure wound therapy alone (without STSG) was not shown to accelerate wound healing and in fact was associated with a longer time to healing (> 2 months) than NPWT in combination with STSG (≤ 1 month).²³ A few patients could not tolerate NPWT at −125 mm Hg and were given increased analgesia or continued to use NPWT at a lower pressure.²³ Likewise, in the patient reported herein as case 3, the wound did not reduce in size with only steroids, gentle debridement, and local wound care for 3 months. With NPWT, the wound was reduced in size by 50% at 4-month follow-up; by the same follow-up time, the other 2 patients had experienced almost complete healing. In cases 1 and 2, only a multilayered compression dressing that allowed FBD to adhere to the wound bed and achieved successful take of the skin substitute was used. Additional studies are necessary to determine the role of NPWT in the management of pathergy in complex PG wounds.

At the authors' institution, steroid therapy management for patients with PG is shared between the gastroenterology, dermatology, and vascular surgery departments as well as the general medicine department. This has resulted in disagreements concerning surgical therapies, different protocols, and routes of steroid administration, which in part has spurred the authors to review their experience with CTPs; currently, they are working as an institution to create a more standardized approach concerning the medical and surgical treatment of patients with PG. 

The diagnosis of PG should be considered when a patient presents with or develops painful violaceous skin ulceration that rapidly manifests within days after minor trauma or surgery. With prompt, accurate diagnosis immunosuppressive therapy can be started without delay. Per the literature, the optimal timing of surgical intervention has been variable; some physicians propose completing immunosuppression therapy and operating at the quiescent phase, while others support debridement of necrotic tissue in addition to use of steroids to allow prompt healing and pain control. In cases of wide destruction and extensive wound defects, surgical intervention appears to be necessary to decrease length of hospital stay, decrease pain, and achieve sufficient wound healing. To date, the authors of this case study have found debridement and placement of FBD and CTPs in conjunction with the initiation of immunosuppressive therapy to be effective.

Tissue-based products can be offered as an alternative dressing to provide protection of the wound, decrease long-term immunosuppressive treatment, and potentially reduce the risk of pathergy at the donor site. To the authors' knowledge, this is the first report of using FBD on PG lesions. This case report demonstrates that simultaneous medical therapy and coverage of the wound with FBD coupled with debridement and multilayered compression dressing can be a successful strategy to heal PG lesions. The result after a single application was durable and restored the defect to the level of the surrounding skin, allowing discontinuation of steroid therapy in 20 days. This method of FBD engraftment can also be an effective tool for preparing the wound bed prior to STSG.

Authors: Mabel Ching-Yee Chan, MD; Crystal James, MD; Munir Patel, MD; Scott Ellis, MD; and John C. Lantis II, MD

Affiliations: Icahn School of Medicine Mount Sinai Morningside and West Hospitals, New York City, NY

Correspondence: Mabel Ching-Yee Chan, MD, Mount Sinai Roosevelt: Mount Sinai West Medical Center, Vascular Surgery, 425 West 59th Street, FL 7, New York, NY 10019; mabel.chan2@mountsinai.org

Disclosure: The authors disclose no additional conflicts of interest.