- Email: [email protected]
Management of Peristomal Pyoderma Gangrenosum
Management of Peristomal Pyoderma Gangrenosum Lisa S Poritz, MD, FACS, Marjorie A Lebo, MSN, CRNP, CWOCN, Anne D Bobb, BSN, RN, CWOCN, Christine M Ardell, BSN, RN, CWOCN, Walter A Koltun, MD, FACS Pyoderma gangrenosum (PG) occurs in about 1% to 5% of patients with inflammatory bowel disease (IBD). Peristomal pyoderma gangrenosum (PPG) is particularly difficult to manage. STUDY DESIGN: A retrospective chart review was performed on all patients with IBD in whom PPG developed from 1997 to 2007 at the Milton S Hershey Medical Center. RESULTS: Sixteen patients (11 women) were identified. Seven had Crohn’s disease (CD), seven had ulcerative colitis (UC), and two had indeterminate colitis. Six patients underwent total proctocolectomy, six patients had total abdominal colectomy (TAC), and four patients had diverting loop stomas. PPG occurred an average of 18.4 ⫾ 7.5 months after stoma creation. Twelve patients had active IBD when PPG developed. Two patients had stoma revisions and both had recurrence of the PPG with the new stoma. Medical therapy was successful in eight patients. Five patients had their stomas closed, with active PPG, and all five resolved their lesions. In four of five, surgical management was altered because of PPG (one early stoma closure, two ileal pouches without stomas, one ileal pouch with high body mass index). Of the seven and six patients treated with cyclosporine or infliximab, respectively, there were only two successes with each. CONCLUSIONS: PPG is more common in the presence of active IBD. Surgical closure of the stoma was successful in resolving PPG in all patients. Cure rate of PPG was poor with cyclosporine and only marginally better with infliximab. Medical treatment of PPG is imperfect, and the best therapy is stoma closure when possible. (J Am Coll Surg 2008;206:311–315. © 2008 by the American College of Surgeons) BACKGROUND:
Treatment of PG and PPG is not standardized, and no treatment method has had consistent success. A wide range of therapeutic interventions exist, including local and systemic medical therapy and surgical intervention. By far, the most successful method to treat PPG is stoma closure when possible. In this article, we review our experience with managing patients with PPG, and how the presence of PPG in a patient may change surgical management.
Pyoderma gangrenosum (PG) is a skin lesion that occurs in about 1% to 5% of patients with inflammatory bowel disease (IBD).1 The lesions start as pustules that break down, rapidly coalesce, and form superficial ulcers.2 The edges of the ulceration are necrotic and undermined. Despite the fact that the ulcers are superficial, they are exquisitely painful. PG exhibits pathergy, a process in which trauma, even minor, leads to development of the pustules. PG classically occurs on the lower extremities, but it can occur elsewhere on the body. Peristomal pyoderma gangrenosum (PPG) is particularly difficult to manage and often frustrating for both the patient and physician. In addition to the pain associated with the lesions, PPG complicates stomal appliance adhesion and management. Minor trauma from the stomal appliance has been suggested as a contributing factor in PPG.
METHODS A retrospective chart review was performed on all adult patients treated for PPG at the Milton S Hershey Medical Center from January 1, 1997 to April 1, 2007. PPG was diagnosed clinically by the characteristic appearance of the lesions, and successful treatment required complete resolution of the lesions. Data abstracted from the chart contained patient characteristics including diagnosis, treatment of PPG, surgical procedures, and complications. The study was approved by the Milton S Hershey Medical Center Institutional Review Board.
Competing Interests Declared: None. Received April 25, 2007; Accepted July 17, 2007. From the Division of Colon and Rectal Surgery, The Milton S Hershey Medical Center, Hershey, PA. Correspondence address: Lisa S Poritz, MD, Division of Colon and Rectal Surgery, H137, The Milton S Hershey Medical Center, 500 University Dr, Hershey, PA 17033.
© 2008 by the American College of Surgeons Published by Elsevier Inc.
RESULTS Sixteen patients with PPG were identified and included in the study. Their clinical characteristics are shown in
311
ISSN 1072-7515/08/$34.00 doi:10.1016/j.jamcollsurg.2007.07.023
312
Poritz et al
Management of Peristomal Pyoderma Gangrenosum
J Am Coll Surg
Table 1. Patient Characteristics Abbreviations and Acronyms
Characteristic
CD ⫽ CyA ⫽ IBD ⫽ IPAA⫽ PG ⫽ PPG ⫽ TAC ⫽ UC ⫽
n Gender (M/F) Patients with Crohn’s disease, n Patients with ulcerative colitis, n Patients with indeterminate colitis, n Duration of disease, y Other extraintestinal manifestations, yes/no Family history of inflammatory bowel disease, yes/no
Crohn’s disease cyclosporine inflammatory bowel disease ileal pouch anal anastomosis pyoderma gangrenosum peristomal pyoderma gangrenosum total abdominal colectomy ulcerative colitis
Table 1. Eleven of 16 (69%) patients were women, which is consistent with findings in some studies,3 but lower than others that had 100% women.4,5 There were seven patients with ulcerative colitis (UC), seven with Crohn’s disease (CD), and two with indeterminate colitis. A stoma was created after total proctocolectomy in six patients, after total abdominal colectomy (TAC) in six patients, and as a diverting loop stoma in four patients. Indications for operation were intractability in 10 patients, perianal disease in 4 patients, dysplasia in 1 patient, and a dysplasia-associated lesion or mass in 1 patient. Only one patient had a history of PG before stoma creation. PPG occurred an average of 18.4 ⫾ 7.5 months (median 3 months; range 3 weeks to 72 months) after stoma creation. In 10 patients, the time between stoma creation and development of PPG was less than 6 months. Twelve (75%) patients had active intestinal disease when PPG developed. Six patients had a retained rectum after TAC as the first stage of an ileal pouch anal anastomosis (IPAA) for intractable UC, three patients with CD had active perianal disease, and three patients with CD had active small bowel disease. The remaining four patients did not have active intestinal disease. All patients were initially treated medically for PPG by a surgeon, gastroenterologist, or dermatologist. Treatment was not uniform, and most patients received multimodality therapy. Table 2 shows the treatments prescribed and the success of the individual therapies. When patients whose PPG resolved received more than one type of therapy, the successful therapy was considered to be the most recent one. Only one patient was treated with topical therapy only. This patient had intralesional injection of steroids, with resolution of PPG. Nine patients were treated with both topical and systemic therapy; the remaining six patients were treated with systemic therapy only. Three patients were admitted for PPG because of severe pain. One was treated with IV steroids and the other two were treated with IV cyclosporine (CyA). All three were converted to oral medications at the time of discharge. Medical treatment (systemic and local) was successful, with complete resolution of the lesions, in eight (50%) of
Data
16 5/11 7 7 2 9.4 ⫾ 3.0 2/14 2/14
the patients (one on oral steroids, one steroid injection, one on topical antibiotics, one on systemic steroids, two on CyA, and two on infliximab). Figure 1 shows an example of PPG before (A) and after CyA treatment (B). Both patients whose pyoderma resolved with CyA and one patient whose pyoderma resolved with infliximab had recurrence of the PPG on withdrawal of the drug and required reinstitution of the respective drugs for control. Of the patients treated with CyA or infliximab, there were only two successes in seven patients (29%) and two successes in six patients (33%), respectively. In all four patients without residual active IBD, the PPG healed successfully with medical management. Of those patients with residual active intestinal disease, only 3 of 12 (25%) healed their PPG with medical management alone. Two of the three who healed required ongoing immune modulation to remain in remission. The third patient has subsequently undergone completion proctectomy with IPAA and no longer has a stoma. Five (31%) of the patients had their stomas closed with active PPG, and in all five, the lesions resolved. In four of the five patients, surgical management of their IBD was altered because of the PPG. One patient with fairly severe PPG and CD had her stoma closed before her perianal disease was completely resolved. Her PPG had not responded to oral steroids, topical and oral antibiotics, or infliximab. Although it was believed that her persistent PPG may have been an indication of ongoing intestinal Table 2. Treatment and Results
Treatment
Steroid injection Topical antibiotics Systemic steroids Systemic antibiotics Systemic cyclosporine Infliximab Stoma closure
Receiving treatment, n
4 5 8 6 7 6 5
Treatment successful n %
1 1 1 1 2 2 5
25 20 12 17 29 33 100
Vol. 206, No. 2, February 2008
Poritz et al
Management of Peristomal Pyoderma Gangrenosum
313
Figure 1. (A) Patient with peristomal pyoderma gangrenosum after total abdominal colectomy for ulcerative colitis, before treatment for peristomal pyoderma gangrenosum. (B) Same patient after treatment with cyclosporine for peristomal pyoderma gangrenosum. Reepithelialization of the ulcer is seen. Note the cribiform pattern of the scar that formed during healing.
the operative procedure. In this patient’s case, the severity and persistence of her PPG forced us to perform IPAA without diverting loop ileostomy despite a body mass index (BMI) in excess of 53 kg/m2. She subsequently experienced recurrence of her lower extremity PG. Two other patients similarly underwent IPAA without loop ileostomy to allow the PPG to heal. One did well and has had no additional problems. The other patient experienced a pelvic abscess 2 1/2 months after creation of the IPAA without an ileostomy. She underwent abscess drainage and loop ileostomy 1 week later, and mild PPG recurred. She is suspected of having CD and is currently maintained on immunosuppressive medication (infliximab and azathioprine) to treat her CD, which has also controlled her PPG. Four (25%) patients still have stomas and PPG, and two patients had stoma revisions and both had recurrence of the PPG with the new stoma. One patient who had her stoma closed experienced extremity PG. She had a previous history of extremity PG before undergoing TAC. The only other patient who had PG at a site other than the stoma experienced PG in her midline wound after it was opened for a superficial wound infection. This same patient experienced PG in her open wound after ileostomy closure. When the wounds closed, the PG resolved. PPG developed three patients 5, 6, and 7 years after stoma creation—considerably longer than for any of the other patients (all less than 1.5 years) in our study. One patient had been doing some home remodeling and had some minor trauma around her stoma. PPG developed within a few weeks. She has responded well, but not completely, to oral antibiotics (Cipro [Merck] and Flagyl [Rhone-Poulenc Rorer]) and topical steroid cream. A pelvic abscess developed in the second patient requiring debridement of his perineum. His PPG developed in the context of recrudescence of his IBD. The third patient also had active intestinal disease.
disease, it was decided to close her ileostomy to help heal her PPG. She remained on azathioprine and infliximab after stoma closure and has done well. One morbidly obese patient with a body mass index in excess of 53 kg/m2 underwent IPAA. She initially underwent TAC with end ileostomy for intractable disease. IPAA was not offered at the initial setting because there was major concern about the ability to make the pouch reach to her anus and the ability to create a loop ileostomy in such an obese patient. She was on high-dose prednisone so she was not a candidate for a one-stage IPAA. Typically, in obese patients, we prefer them to lose a considerable amount of weight after TAC and before completion proctectomy with IPAA to facilitate
DISCUSSION PG is a relatively rare skin lesion that is associated with IBD and hematologic malignancies. But 25% to 50% of PG can be idiopathic in origin.1,6 Typically, IBD is the underlying associated cause in only 15% to 20% of patients with PG.1 The most common site for PG is the lower extremities. In patients with intestinal stomas, PG can occur in the skin around the stoma. The first case of PPG was described by McGarity and colleagues7 in 1984. The literature on PPG is relatively sparse, with the largest series having only 15 to 20 patients. In that context, it is difficult to determine the incidence of PPG. The only study that attempted to measure the true incidence found PPG in 0.6% of patients with
314
Poritz et al
Management of Peristomal Pyoderma Gangrenosum
stomas.8 It is likely that the incidence of PPG is higher than reported because it is often underdiagnosed. A high index of suspicion and familiarity with the typical appearance of the lesions are needed for accurate and timely diagnosis. With the few number of cases reported, this expertise may not be readily available, leading to underdiagnosis and subsequent undertreatment. Diagnosis of PPG is almost always made clinically.4 The lesions have a characteristic appearance and clinical course. The lesion often begins as a discrete pustule with surrounding erythema and progresses rapidly to the typical ulcerated lesion. The border of the lesion is deep erythematous in color and overhangs the ulcer. This is from the chronic dermal inflammation that leads to thrombosis of blood vessels and necrosis of the dermis, leading to secondary ulceration and necrosis of the epidermis.1 The lesions are exquisitely painful, sometimes requiring narcotics or hospital admission for pain control, as was seen with three of our patients.1 Biopsies of the lesion are often nonspecific and reveal epidermal neutrophil infiltration, edema, and perivascular lymphocyte and mononuclear cell infiltration in the dermis.1 Because the results are nonspecific, biopsies are rarely needed to make the diagnosis of PG or PPG4; the diagnostic value of a biopsy is to rule out other causes of the lesions, such as cancer. The cause of PG and PPG is still debated. Many patients can relate the development of the skin lesions to recent trauma to the affected area, a phenomenon known as pathergy.9 Using the principle of pathergy, it has been suggested that minor trauma to the peristomal skin when the wafer is changed may initiate development of PPG.10 We had one patient in our series who had trauma to the stoma and surrounding skin and PPG developed shortly thereafter. Despite the fact that only 15% to 50% of cases of PG are associated with IBD, almost all cases of PPG that have been reported have occurred in patients with underlying IBD, including all patients reported here. There are, however, a few cases of PPG occurring around stomas created for other reasons. Sheldon and associates3 had one patient with PPG around a continent urostomy and Lyon and coworkers5 reported five patients with PPG stomas for reasons unrelated to IBD. Most studies also show that among IBD patients PPG is more common in patients with CD than UC.4,10,11 Our study was evenly divided: seven patients had CD and seven had UC. Nonperistomal PG has been reported to be equally distributed among CD and UC patients.1,5 One possible explanation for the higher incidence of PPG among CD patients is that patients with CD are more likely to have stomas, especially when they have active dis-
J Am Coll Surg
ease. Most patients with UC who have stomas have undergone total proctocolectomy and no longer have any active disease. One of the reasons our study had so many patients with UC and PPG was the fact that five of the seven UC patients experienced the PPG in the context of a retained diseased rectum after TAC for acute severe UC. It has been suggested that in patients with IBD the development of PG and PPG may reflect the severity of the underlying intestinal disease.9 Sheldon and coauthors3 showed that surgically removing active disease in CD patients with PPG decreased the average healing time of PPG from 12.4 months to 1.8 months. Proctectomy for residual rectal disease has also been successful in resolving pyoderma in some patients.10 Twelve of 16 patients in our study had active intestinal disease when pyoderma developed. Four of the 12 were patients, experienced PPG within 1 to 1.5 months after TAC for active UC, just as they were finishing their steroid taper. This suggests that the presence of active disease has a role in the development of PPG; presumably, some of the residual disease in the rectum was controlled by the steroids, and withdrawal of the steroids precipitated the PPG. Another patient in our study who had a total proctocolectomy (TPC), small bowel resection, and ileostomy experienced pyoderma 7 years (84 months) after creation of his stoma, when a large perineal abscess developed presumably from CD, despite TPC in the past. There is no consistent agreement between caregivers about the best therapy for PPG. The method by which patients are treated is often determined by the specialty and experience of the treating physician. Probably the most common initial therapy for PG is topical, whether it be steroids, antibiotics, or other wound care agents. The systemic side effects and risks of immunosuppressive medications can be avoided if these treatments are successful. Unfortunately, with PPG topical therapy, it is often difficult because the wounds are usually under the faceplate of the stoma appliance, and these medications can interfere with adherence. Options for topical therapy include local wound care, steroids (either topical or injected), cromolyn sodium, metronidazole (as 1% Metrogel [Galderma]), or tacrolimus. Systemic therapy is often the next step, when topical therapy fails or in conjunction with topical therapy. Options include high dose steroids, antibiotics, dapsone, CyA, and more recently, infliximab. In our study, only one patient responded to high-dose steroids or antibiotics (12% and 17%, respectively). Of the six and seven patients treated with CyA and infliximab, respectively, only two in each group (29% and 33%) responded. But despite re-
Vol. 206, No. 2, February 2008
Poritz et al
sponse, three of these four patients had recurrence of the PPG on withdrawal of the drug. With the success of infliximab as a therapy for CD, many have tried it for PG and PPG. Although most reports only have a few patients, Brooklyn and colleagues12 performed a randomized, double blinded placebo controlled trial. They found a remission rate or complete response of 21% by week 6. But they had a 69% beneficial clinical response. There was a relatively large nonresponse rate of 31%, as with most studies using infliximab. Other smaller studies have shown a 50% to 100% response, but the followups were short.7,9,11,13 In our study, we had six patients who received infliximab for PPG. There was successful resolution of the lesions in only two patients (33%), and one of those two patients requires ongoing infliximab therapy to maintain remission. Although these studies show a response of PPG to infliximab, it is unknown whether all of the benefit from therapy is a direct effect of the infliximab on the lesion or a humeral effect from regression of the active intestinal disease. Another approach to the treatment of PPG is to treat the active inflammation in the gastrointestinal tract, either with aggressive medical management or surgery, or to close the stoma if possible. Surgery directed at the PPG, such as stoma revision or relocation, uniformly fails, with early recurrence of the PPG at the new stoma site. Both patients in our study who underwent resiting of the stoma had recurrence of PPG at the new site. Equally disappointing results have been seen in other series, with failure rates ranging from 40% to 100%.8,3,13 Unlike resiting of the stoma, stoma closure led to resolution of the skin lesions in all patients who were candidates (five patients). Of that group, only one patient subsequently PG, and she had a history of PG before PPG developing. We altered the surgical management of four of the five patients so that their stomas could be closed as soon as possible because their PPG did not respond to medical therapy, including infliximab, in three of the patients. PPG is a rare lesion that is often, but not exclusively, associated with IBD. It is exquisitely painful, and management of such lesions is complicated by the trauma of stoma appliances. Medical management of these lesions is difficult, with no sin-
Management of Peristomal Pyoderma Gangrenosum
315
gle uniformly successful treatment or treatment regimen. The best therapy is stoma closure, when possible. Author Contributions Study conception and design: Poritz, Koltun Acquisition of data: Poritz, Lebo, Bobb, Ardell, Koltun Analysis and interpretation of data: Poritz, Lebo, Bobb, Ardell, Koltun Drafting of manuscript: Poritz Critical revision: Poritz, Lebo, Bobb, Ardell, Koltun
REFERENCES 1. Callen JP. Pyoderma gangrenosum. Lancet 1998;351:581–585. 2. Braverman IM. Skin signs of systemic disease, 3rd ed. Philadelphia, PA: WB Saunders Co; 1998:416–423. 3. Sheldon DG, Sawchuk LL, Kozarek RA, Thirlby RC. Twenty cases of peristomal pyoderma gangrenosum. Diagnostic implications and management. Arch Surg 2000;135:564–569. 4. Cairns BA, Herbst CA, Sartor BR, et al. Peristomal pyoderma gangrenosum and inflammatory bowel disease. Arch Surg 1994; 129:769–772. 5. Sapienza MS, Cohen S, Dimarino AJ. Treatment of pyoderma gangrenosum with infliximab in Crohn’s disease. Dig Dis Sci 2004;49:1454–1457. 6. Von Den Driesch P. Pyoderma gangrenosum: a report of 44 cases with follow-up. Br J Dermatol 1997;137:1000–1005. 7. McGarity WC, Robertson DB, McKeown PP, et al. Pyoderma gangrenosum at the parastomal site in patients with Crohn’s disease. Arch Surg 1984;119:1186–1188. 8. Lyon CC, Smith AJ, Beck MH, et al. Parastomal pyoderma gangrenosum: clinical features and management. J Am Acad Dermatol 2000;42:992–1002. 9. Hughes AP, Jackson JM, Callen JP. Clinical features and treatment of peristomal pyoderma gangrenosum. JAMA 2000;284: 1546–1548. 10. Tjandra JJ, Hughes LE. Parastomal pyoderma gangrenosum in inflammatory bowel disease. Dis Colon Rectum 1994;37:938– 942. 11. Yeo H, Abir F, Longo W. Management of parastomal ulcers. World J Gastroenterol 2006;12:3133–3137. 12. Brooklyn TN, Dunnill MGS, Shetty A, et al. Infliximab for the treatment of pyoderma gangrenosum: a randomized, double blind, placebo controlled trial. Gut 2006;55:505–509. 13. Kiran RP, O’Brien-Ermlich B, Achkar JP, et al. Management of peristomal pyoderma gangrenosum. Dis Colon Rectum 2005; 48:1397–1403.
Treatment of PG and PPG is not standardized, and no treatment method has had consistent success. A wide range of therapeutic interventions exist, including local and systemic medical therapy and surgical intervention. By far, the most successful method to treat PPG is stoma closure when possible. In this article, we review our experience with managing patients with PPG, and how the presence of PPG in a patient may change surgical management.
Pyoderma gangrenosum (PG) is a skin lesion that occurs in about 1% to 5% of patients with inflammatory bowel disease (IBD).1 The lesions start as pustules that break down, rapidly coalesce, and form superficial ulcers.2 The edges of the ulceration are necrotic and undermined. Despite the fact that the ulcers are superficial, they are exquisitely painful. PG exhibits pathergy, a process in which trauma, even minor, leads to development of the pustules. PG classically occurs on the lower extremities, but it can occur elsewhere on the body. Peristomal pyoderma gangrenosum (PPG) is particularly difficult to manage and often frustrating for both the patient and physician. In addition to the pain associated with the lesions, PPG complicates stomal appliance adhesion and management. Minor trauma from the stomal appliance has been suggested as a contributing factor in PPG.
METHODS A retrospective chart review was performed on all adult patients treated for PPG at the Milton S Hershey Medical Center from January 1, 1997 to April 1, 2007. PPG was diagnosed clinically by the characteristic appearance of the lesions, and successful treatment required complete resolution of the lesions. Data abstracted from the chart contained patient characteristics including diagnosis, treatment of PPG, surgical procedures, and complications. The study was approved by the Milton S Hershey Medical Center Institutional Review Board.
Competing Interests Declared: None. Received April 25, 2007; Accepted July 17, 2007. From the Division of Colon and Rectal Surgery, The Milton S Hershey Medical Center, Hershey, PA. Correspondence address: Lisa S Poritz, MD, Division of Colon and Rectal Surgery, H137, The Milton S Hershey Medical Center, 500 University Dr, Hershey, PA 17033.
© 2008 by the American College of Surgeons Published by Elsevier Inc.
RESULTS Sixteen patients with PPG were identified and included in the study. Their clinical characteristics are shown in
311
ISSN 1072-7515/08/$34.00 doi:10.1016/j.jamcollsurg.2007.07.023
312
Poritz et al
Management of Peristomal Pyoderma Gangrenosum
J Am Coll Surg
Table 1. Patient Characteristics Abbreviations and Acronyms
Characteristic
CD ⫽ CyA ⫽ IBD ⫽ IPAA⫽ PG ⫽ PPG ⫽ TAC ⫽ UC ⫽
n Gender (M/F) Patients with Crohn’s disease, n Patients with ulcerative colitis, n Patients with indeterminate colitis, n Duration of disease, y Other extraintestinal manifestations, yes/no Family history of inflammatory bowel disease, yes/no
Crohn’s disease cyclosporine inflammatory bowel disease ileal pouch anal anastomosis pyoderma gangrenosum peristomal pyoderma gangrenosum total abdominal colectomy ulcerative colitis
Table 1. Eleven of 16 (69%) patients were women, which is consistent with findings in some studies,3 but lower than others that had 100% women.4,5 There were seven patients with ulcerative colitis (UC), seven with Crohn’s disease (CD), and two with indeterminate colitis. A stoma was created after total proctocolectomy in six patients, after total abdominal colectomy (TAC) in six patients, and as a diverting loop stoma in four patients. Indications for operation were intractability in 10 patients, perianal disease in 4 patients, dysplasia in 1 patient, and a dysplasia-associated lesion or mass in 1 patient. Only one patient had a history of PG before stoma creation. PPG occurred an average of 18.4 ⫾ 7.5 months (median 3 months; range 3 weeks to 72 months) after stoma creation. In 10 patients, the time between stoma creation and development of PPG was less than 6 months. Twelve (75%) patients had active intestinal disease when PPG developed. Six patients had a retained rectum after TAC as the first stage of an ileal pouch anal anastomosis (IPAA) for intractable UC, three patients with CD had active perianal disease, and three patients with CD had active small bowel disease. The remaining four patients did not have active intestinal disease. All patients were initially treated medically for PPG by a surgeon, gastroenterologist, or dermatologist. Treatment was not uniform, and most patients received multimodality therapy. Table 2 shows the treatments prescribed and the success of the individual therapies. When patients whose PPG resolved received more than one type of therapy, the successful therapy was considered to be the most recent one. Only one patient was treated with topical therapy only. This patient had intralesional injection of steroids, with resolution of PPG. Nine patients were treated with both topical and systemic therapy; the remaining six patients were treated with systemic therapy only. Three patients were admitted for PPG because of severe pain. One was treated with IV steroids and the other two were treated with IV cyclosporine (CyA). All three were converted to oral medications at the time of discharge. Medical treatment (systemic and local) was successful, with complete resolution of the lesions, in eight (50%) of
Data
16 5/11 7 7 2 9.4 ⫾ 3.0 2/14 2/14
the patients (one on oral steroids, one steroid injection, one on topical antibiotics, one on systemic steroids, two on CyA, and two on infliximab). Figure 1 shows an example of PPG before (A) and after CyA treatment (B). Both patients whose pyoderma resolved with CyA and one patient whose pyoderma resolved with infliximab had recurrence of the PPG on withdrawal of the drug and required reinstitution of the respective drugs for control. Of the patients treated with CyA or infliximab, there were only two successes in seven patients (29%) and two successes in six patients (33%), respectively. In all four patients without residual active IBD, the PPG healed successfully with medical management. Of those patients with residual active intestinal disease, only 3 of 12 (25%) healed their PPG with medical management alone. Two of the three who healed required ongoing immune modulation to remain in remission. The third patient has subsequently undergone completion proctectomy with IPAA and no longer has a stoma. Five (31%) of the patients had their stomas closed with active PPG, and in all five, the lesions resolved. In four of the five patients, surgical management of their IBD was altered because of the PPG. One patient with fairly severe PPG and CD had her stoma closed before her perianal disease was completely resolved. Her PPG had not responded to oral steroids, topical and oral antibiotics, or infliximab. Although it was believed that her persistent PPG may have been an indication of ongoing intestinal Table 2. Treatment and Results
Treatment
Steroid injection Topical antibiotics Systemic steroids Systemic antibiotics Systemic cyclosporine Infliximab Stoma closure
Receiving treatment, n
4 5 8 6 7 6 5
Treatment successful n %
1 1 1 1 2 2 5
25 20 12 17 29 33 100
Vol. 206, No. 2, February 2008
Poritz et al
Management of Peristomal Pyoderma Gangrenosum
313
Figure 1. (A) Patient with peristomal pyoderma gangrenosum after total abdominal colectomy for ulcerative colitis, before treatment for peristomal pyoderma gangrenosum. (B) Same patient after treatment with cyclosporine for peristomal pyoderma gangrenosum. Reepithelialization of the ulcer is seen. Note the cribiform pattern of the scar that formed during healing.
the operative procedure. In this patient’s case, the severity and persistence of her PPG forced us to perform IPAA without diverting loop ileostomy despite a body mass index (BMI) in excess of 53 kg/m2. She subsequently experienced recurrence of her lower extremity PG. Two other patients similarly underwent IPAA without loop ileostomy to allow the PPG to heal. One did well and has had no additional problems. The other patient experienced a pelvic abscess 2 1/2 months after creation of the IPAA without an ileostomy. She underwent abscess drainage and loop ileostomy 1 week later, and mild PPG recurred. She is suspected of having CD and is currently maintained on immunosuppressive medication (infliximab and azathioprine) to treat her CD, which has also controlled her PPG. Four (25%) patients still have stomas and PPG, and two patients had stoma revisions and both had recurrence of the PPG with the new stoma. One patient who had her stoma closed experienced extremity PG. She had a previous history of extremity PG before undergoing TAC. The only other patient who had PG at a site other than the stoma experienced PG in her midline wound after it was opened for a superficial wound infection. This same patient experienced PG in her open wound after ileostomy closure. When the wounds closed, the PG resolved. PPG developed three patients 5, 6, and 7 years after stoma creation—considerably longer than for any of the other patients (all less than 1.5 years) in our study. One patient had been doing some home remodeling and had some minor trauma around her stoma. PPG developed within a few weeks. She has responded well, but not completely, to oral antibiotics (Cipro [Merck] and Flagyl [Rhone-Poulenc Rorer]) and topical steroid cream. A pelvic abscess developed in the second patient requiring debridement of his perineum. His PPG developed in the context of recrudescence of his IBD. The third patient also had active intestinal disease.
disease, it was decided to close her ileostomy to help heal her PPG. She remained on azathioprine and infliximab after stoma closure and has done well. One morbidly obese patient with a body mass index in excess of 53 kg/m2 underwent IPAA. She initially underwent TAC with end ileostomy for intractable disease. IPAA was not offered at the initial setting because there was major concern about the ability to make the pouch reach to her anus and the ability to create a loop ileostomy in such an obese patient. She was on high-dose prednisone so she was not a candidate for a one-stage IPAA. Typically, in obese patients, we prefer them to lose a considerable amount of weight after TAC and before completion proctectomy with IPAA to facilitate
DISCUSSION PG is a relatively rare skin lesion that is associated with IBD and hematologic malignancies. But 25% to 50% of PG can be idiopathic in origin.1,6 Typically, IBD is the underlying associated cause in only 15% to 20% of patients with PG.1 The most common site for PG is the lower extremities. In patients with intestinal stomas, PG can occur in the skin around the stoma. The first case of PPG was described by McGarity and colleagues7 in 1984. The literature on PPG is relatively sparse, with the largest series having only 15 to 20 patients. In that context, it is difficult to determine the incidence of PPG. The only study that attempted to measure the true incidence found PPG in 0.6% of patients with
314
Poritz et al
Management of Peristomal Pyoderma Gangrenosum
stomas.8 It is likely that the incidence of PPG is higher than reported because it is often underdiagnosed. A high index of suspicion and familiarity with the typical appearance of the lesions are needed for accurate and timely diagnosis. With the few number of cases reported, this expertise may not be readily available, leading to underdiagnosis and subsequent undertreatment. Diagnosis of PPG is almost always made clinically.4 The lesions have a characteristic appearance and clinical course. The lesion often begins as a discrete pustule with surrounding erythema and progresses rapidly to the typical ulcerated lesion. The border of the lesion is deep erythematous in color and overhangs the ulcer. This is from the chronic dermal inflammation that leads to thrombosis of blood vessels and necrosis of the dermis, leading to secondary ulceration and necrosis of the epidermis.1 The lesions are exquisitely painful, sometimes requiring narcotics or hospital admission for pain control, as was seen with three of our patients.1 Biopsies of the lesion are often nonspecific and reveal epidermal neutrophil infiltration, edema, and perivascular lymphocyte and mononuclear cell infiltration in the dermis.1 Because the results are nonspecific, biopsies are rarely needed to make the diagnosis of PG or PPG4; the diagnostic value of a biopsy is to rule out other causes of the lesions, such as cancer. The cause of PG and PPG is still debated. Many patients can relate the development of the skin lesions to recent trauma to the affected area, a phenomenon known as pathergy.9 Using the principle of pathergy, it has been suggested that minor trauma to the peristomal skin when the wafer is changed may initiate development of PPG.10 We had one patient in our series who had trauma to the stoma and surrounding skin and PPG developed shortly thereafter. Despite the fact that only 15% to 50% of cases of PG are associated with IBD, almost all cases of PPG that have been reported have occurred in patients with underlying IBD, including all patients reported here. There are, however, a few cases of PPG occurring around stomas created for other reasons. Sheldon and associates3 had one patient with PPG around a continent urostomy and Lyon and coworkers5 reported five patients with PPG stomas for reasons unrelated to IBD. Most studies also show that among IBD patients PPG is more common in patients with CD than UC.4,10,11 Our study was evenly divided: seven patients had CD and seven had UC. Nonperistomal PG has been reported to be equally distributed among CD and UC patients.1,5 One possible explanation for the higher incidence of PPG among CD patients is that patients with CD are more likely to have stomas, especially when they have active dis-
J Am Coll Surg
ease. Most patients with UC who have stomas have undergone total proctocolectomy and no longer have any active disease. One of the reasons our study had so many patients with UC and PPG was the fact that five of the seven UC patients experienced the PPG in the context of a retained diseased rectum after TAC for acute severe UC. It has been suggested that in patients with IBD the development of PG and PPG may reflect the severity of the underlying intestinal disease.9 Sheldon and coauthors3 showed that surgically removing active disease in CD patients with PPG decreased the average healing time of PPG from 12.4 months to 1.8 months. Proctectomy for residual rectal disease has also been successful in resolving pyoderma in some patients.10 Twelve of 16 patients in our study had active intestinal disease when pyoderma developed. Four of the 12 were patients, experienced PPG within 1 to 1.5 months after TAC for active UC, just as they were finishing their steroid taper. This suggests that the presence of active disease has a role in the development of PPG; presumably, some of the residual disease in the rectum was controlled by the steroids, and withdrawal of the steroids precipitated the PPG. Another patient in our study who had a total proctocolectomy (TPC), small bowel resection, and ileostomy experienced pyoderma 7 years (84 months) after creation of his stoma, when a large perineal abscess developed presumably from CD, despite TPC in the past. There is no consistent agreement between caregivers about the best therapy for PPG. The method by which patients are treated is often determined by the specialty and experience of the treating physician. Probably the most common initial therapy for PG is topical, whether it be steroids, antibiotics, or other wound care agents. The systemic side effects and risks of immunosuppressive medications can be avoided if these treatments are successful. Unfortunately, with PPG topical therapy, it is often difficult because the wounds are usually under the faceplate of the stoma appliance, and these medications can interfere with adherence. Options for topical therapy include local wound care, steroids (either topical or injected), cromolyn sodium, metronidazole (as 1% Metrogel [Galderma]), or tacrolimus. Systemic therapy is often the next step, when topical therapy fails or in conjunction with topical therapy. Options include high dose steroids, antibiotics, dapsone, CyA, and more recently, infliximab. In our study, only one patient responded to high-dose steroids or antibiotics (12% and 17%, respectively). Of the six and seven patients treated with CyA and infliximab, respectively, only two in each group (29% and 33%) responded. But despite re-
Vol. 206, No. 2, February 2008
Poritz et al
sponse, three of these four patients had recurrence of the PPG on withdrawal of the drug. With the success of infliximab as a therapy for CD, many have tried it for PG and PPG. Although most reports only have a few patients, Brooklyn and colleagues12 performed a randomized, double blinded placebo controlled trial. They found a remission rate or complete response of 21% by week 6. But they had a 69% beneficial clinical response. There was a relatively large nonresponse rate of 31%, as with most studies using infliximab. Other smaller studies have shown a 50% to 100% response, but the followups were short.7,9,11,13 In our study, we had six patients who received infliximab for PPG. There was successful resolution of the lesions in only two patients (33%), and one of those two patients requires ongoing infliximab therapy to maintain remission. Although these studies show a response of PPG to infliximab, it is unknown whether all of the benefit from therapy is a direct effect of the infliximab on the lesion or a humeral effect from regression of the active intestinal disease. Another approach to the treatment of PPG is to treat the active inflammation in the gastrointestinal tract, either with aggressive medical management or surgery, or to close the stoma if possible. Surgery directed at the PPG, such as stoma revision or relocation, uniformly fails, with early recurrence of the PPG at the new stoma site. Both patients in our study who underwent resiting of the stoma had recurrence of PPG at the new site. Equally disappointing results have been seen in other series, with failure rates ranging from 40% to 100%.8,3,13 Unlike resiting of the stoma, stoma closure led to resolution of the skin lesions in all patients who were candidates (five patients). Of that group, only one patient subsequently PG, and she had a history of PG before PPG developing. We altered the surgical management of four of the five patients so that their stomas could be closed as soon as possible because their PPG did not respond to medical therapy, including infliximab, in three of the patients. PPG is a rare lesion that is often, but not exclusively, associated with IBD. It is exquisitely painful, and management of such lesions is complicated by the trauma of stoma appliances. Medical management of these lesions is difficult, with no sin-
Management of Peristomal Pyoderma Gangrenosum
315
gle uniformly successful treatment or treatment regimen. The best therapy is stoma closure, when possible. Author Contributions Study conception and design: Poritz, Koltun Acquisition of data: Poritz, Lebo, Bobb, Ardell, Koltun Analysis and interpretation of data: Poritz, Lebo, Bobb, Ardell, Koltun Drafting of manuscript: Poritz Critical revision: Poritz, Lebo, Bobb, Ardell, Koltun
REFERENCES 1. Callen JP. Pyoderma gangrenosum. Lancet 1998;351:581–585. 2. Braverman IM. Skin signs of systemic disease, 3rd ed. Philadelphia, PA: WB Saunders Co; 1998:416–423. 3. Sheldon DG, Sawchuk LL, Kozarek RA, Thirlby RC. Twenty cases of peristomal pyoderma gangrenosum. Diagnostic implications and management. Arch Surg 2000;135:564–569. 4. Cairns BA, Herbst CA, Sartor BR, et al. Peristomal pyoderma gangrenosum and inflammatory bowel disease. Arch Surg 1994; 129:769–772. 5. Sapienza MS, Cohen S, Dimarino AJ. Treatment of pyoderma gangrenosum with infliximab in Crohn’s disease. Dig Dis Sci 2004;49:1454–1457. 6. Von Den Driesch P. Pyoderma gangrenosum: a report of 44 cases with follow-up. Br J Dermatol 1997;137:1000–1005. 7. McGarity WC, Robertson DB, McKeown PP, et al. Pyoderma gangrenosum at the parastomal site in patients with Crohn’s disease. Arch Surg 1984;119:1186–1188. 8. Lyon CC, Smith AJ, Beck MH, et al. Parastomal pyoderma gangrenosum: clinical features and management. J Am Acad Dermatol 2000;42:992–1002. 9. Hughes AP, Jackson JM, Callen JP. Clinical features and treatment of peristomal pyoderma gangrenosum. JAMA 2000;284: 1546–1548. 10. Tjandra JJ, Hughes LE. Parastomal pyoderma gangrenosum in inflammatory bowel disease. Dis Colon Rectum 1994;37:938– 942. 11. Yeo H, Abir F, Longo W. Management of parastomal ulcers. World J Gastroenterol 2006;12:3133–3137. 12. Brooklyn TN, Dunnill MGS, Shetty A, et al. Infliximab for the treatment of pyoderma gangrenosum: a randomized, double blind, placebo controlled trial. Gut 2006;55:505–509. 13. Kiran RP, O’Brien-Ermlich B, Achkar JP, et al. Management of peristomal pyoderma gangrenosum. Dis Colon Rectum 2005; 48:1397–1403.