Surgical Treatment of Pyoderma Gangrenosum with Negative Pressure Wound Therapy and Skin Grafting, Including Xenografts: Personal Experience and Comprehensive Review on 161 Cases

Abstract

Significance: Pyoderma gangrenosum (PG) is a rare debilitating autoinflammatory ulcerative skin disease. No gold standard has been established for the treatment of PG. The role of surgical interventions and negative pressure wound therapy (NPWT) was discussed controversially until recently as these procedures might pose a trigger to further aggravate the condition.

Recent Advances: Recent advances confirm the paradigm change that a surgical approach of PG with split thickness skin grafting (STSG) secured by NPWT is a safe and valuable treatment if performed under adequate immunosuppression. We elaborate this on the hand of a broad literature search retrieving 101 relevant articles describing 138 patients complemented with our personal experience on 23 patients, including 2 patients treated with a porcine xenodressing.

Critical Issues: A wide range of surgical approaches have been reported, including xenografts. Treatment was finally successful in 86%, including the xenotransplant cases. Ten percent improved and failures were mainly reported without immunosuppression. Despite halting the inflammatory process, NPWT alone, without skin grafting, does not much accelerate healing time. The best surgical approach appears to be STSG fixed with NPWT as this leads to higher skin graft take. There remains the problem of the chronic nature of PG and the recurrence after tapering of immunosuppression or trauma; therefore, a sustained immunosuppressive treatment is suggested.

Future Directions: While surgical treatment is supported by the published data, the exact immunosuppression is still evolving. Due to deeper insights into pathogenesis and growing clinical reports, a broader utilization of biologic treatments and a shift from tumor necrosis factor (TNF)-alpha to interleukin (IL)-12/23 or IL-23 antibodies alone are predictable, as IL-12/23 antibodies show good clinical responses with fewer side effects. The positive results with porcine xenodressings might be due to immunological effects of the xenomaterial; they appear promising, but are preliminary and should be confirmed in a larger patient collective.

Klaus Eisendle, MD, MSc, PhD, MBA

Scope and Significance

Pyoderma gangrenosum (PG) is a rare autoinflammatory disease, characterized by uncontrolled activation of neutrophil granulocytes leading to very painful growing skin ulcerations. Treatment is complex, necessitating time-intensive wound dressings, systemic treatments, and surgical interventions. Mortality is augmented in PG patients and care is long and costly. This review focuses on the surgical and systemic management of patients with PG. A broad search of the PubMed, Medline, EBSCO Biomedical Reference Collection, and Cochrane databases was performed and a total of 101 relevant articles describing 138 patients could be retrieved and complemented with our personal experience of 23 patients.

Translational Relevance

This work is relevant to researchers studying wound healing, skin immunology, and autoinflammatory diseases, as well as scientists involved in the development and new use of biologic and immunosuppressive treatments.

Clinical Relevance

All published different surgical approaches, including xenotransplants, are critically discussed on the experience on 161 treated patients. An immunosuppressive treatment ladder with newer developments in the use of anti-interleukin (IL)-12/23 antibodies, as well as possible adjuvant approaches are proposed. This is directly relevant to all clinicians, wound specialists, and nurses caring for PG patients.

Background

PG, first described by Brunsting et al. in 1930,¹ is a rare inflammatory ulcerative skin disease belonging to the neutrophil dermatoses, which are characterized by an accumulation of neutrophils in the skin.² An incidence of 0.3–1.0/100,000 has been reported with a female predominance ranging from 55% to 59%. The mean age at diagnosis ranges from 48 to 58 years,^3–5 and the mortality of patients suffering from PG is three times higher compared with the general population.^3,6 The most common clinical presentation of PG is a pustule that progresses to a painful ulcer with violaceous undermined borders and a purulent base. Bullous, vegetative, peristomal, and extracutaneous forms of PG may also be seen.

Pathogenesis is unknown and poorly understood; however, there is strong evidence to suggest that PG has an immunologic etiology. Recently, neutrophil dermatoses have been added to the group of autoinflammatory diseases.^7,8 Successful treatment requires the reduction of the inflammation; indeed, the secretion of tumor necrosis factor (TNF) by keratinocytes and clonally expanded T cells with subsequent overexpression of IL-8, a strong chemotactic factor for neutrophils, and other cytokines, including IL-1, IL-6, IL-17, and IL-23, has been demonstrated to contribute to the genesis of PG.^9–13 Furthermore, the upregulation of matrix metalloproteinase (MMP) expression, in particular MMP 9 and 10, might also contribute to poor wound healing^14,15 and Wilkes et al. could show that the resolution of inflammation was one predictor for the eventual healing of PG.¹⁶

The clinical course of PG is unpredictable; it may stop spontaneously and stay inactive for months or years or can progress and get worse after traumas, after surgical interventions, or even without any triggers.¹⁷ In 25–50% of patients, the disease is idiopathic; in the remaining cases, an underlying systemic disease can be found,¹⁷ such as inflammatory bowel disease (9–34%),^3,18 rheumatoid arthritis (9–11%), solid neoplasms (9–33%), or hematological disorders, including monoclonal gammopathy.^{3–8,19–21} Recently, also, an association with endocrine diseases like diabetes mellitus, found in 29%, or hypothyroidism, found in 7% of patients, has been reported.³

It is essential to rule out other causes of cutaneous ulcerations, as misdiagnosis can lead to serious complications and PG remains a diagnosis by exclusion.^22,23 A helpful diagnostic criterion is the pathergy phenomenon. Trigger factors were found in 30% of patients, mostly after surgical interventions or minor trauma.^20–23 Rapid progressive ulcerations and pain are present in the majority of PG cases.²¹ Special attention must be given to patients with hypertension, especially with diabetes, as Martorell hypertensive ischemic leg ulcer has a very similar clinical presentation and a large enough wedge biopsy should be performed to exclude ischemic subcutaneous arteriolosclerosis.^24,25

Systemic treatment mainly consists of corticosteroids (CS; 0.5–2 mg/kg/day),^4,20 commonly with adjunctive immunosuppressive or immunomodulatory treatments such as cyclosporine (3–6 mg/kg daily),^5,26,27 mycophenolate mofetil (up to 2,000 mg daily),^5,26,28 methotrexate (12.5–22.5 mg/weekly),^26,29 azathioprine (25–150 mg daily),^30,31 dapsone (50–300 mg daily),^32–35 thalidomide (200–400 mg daily),^36–39 intravenous immunoglobulins,^40,41 and biologics such as TNF-alpha inhibitors infliximab,⁴² adalimumab,^43–45 and etanercept^46–48 or ustekinumab, an anti-IL 12/23 antibody.^49,50 The adverse effects of those drugs and/or the possible inadequate clinical response can limit their use in some patients, emphasizing the role of local wound management.^51,52 Local treatment consists in the application of wound dressings and topical or intralesional CS or tacrolimus.^52,53

The role of surgical interventions such as split thickness skin grafting (STSG) and negative pressure wound therapy (NPWT) is controversially discussed in the literature as these procedures might pose a trigger and further aggravate the condition, especially if performed without immunosuppression.⁵⁴ Indeed, acute postoperative PG is a feared complication after surgery and more than 200 cases have been described in the literature so far.⁵⁵ In this study, we report the largest review of surgical treatment of PG and our personal experience in the surgical management of PG under adequate immunosuppression in 23 patients, including 2 cases successfully treated with porcine xenografts (EZ-Derm^®), a new potentially promising treatment option.

Discussion of Findings and Relevant Literature

Comprehensive review

A broad search of the PubMed, Medline, EBSCO Biomedical Reference Collection, and Cochrane databases was performed in December 2019 with the terms “PG skin graft,” “PG split thickness skin graft,” “PG skin transplantation,” “PG negative pressure wound therapy,” and “PG vacuum therapy,” and with the Medical Search Help (MeSH) Terms “PG” and “skin grafting.” Two hundred and seventeen articles published between 1967 and 2020 were identified. Reviews not reporting patients, reports without skin grafting or without NPWT, reports where NPWT or STSG were not used for treatment of PG, and double publications were excluded. The titles and abstracts were screened individually; if the articles seemed relevant or lacked adequate data in the abstract to indicate their relevance, the full-text articles were obtained and fully analyzed. Study references were cross-checked and screened manually to identify potential citations not captured by the initial database search. A total of 101 articles could be included in the analysis: number of patients, age, gender, side of PG, procedural intervention (NPWT, skin grafts, and hyperbaric oxygen therapy), immunosuppression performed, outcome, and reported recurrences were noted and are reported in Table 1.

Table 1.

All reports published in the literature so far on surgical treatment of pyoderma gangrenosum with skin graft or negative wound pressure therapy

Publication No.	Author, Year (References)	Patients	Age, Sex	Localization	Procedural Interventions	Systemic Treatment	Outcome
1	Ridenhour and Stephenson (1968)91	1	62, M	Left elbow, both lower legs	STSG	CS, topical aqueous silver nitrate	Successful graft, healed without recurrence
2	McGarity and Barnett (1977)115	1	53, F	Abdominal wall	STSG	CS	Successful graft, healed without recurrence
3	Davis et al. (1987)116	4	64, M; 62, F; 17, F; 22, F	Legs (lower) 4, elbow 1	STSG (4), HBO (4)	CS (4), in 3 CS were tapered under HBO and inactive PG grafted without	Successful graft, healed without recurrence
4	Gérard et al. (1988)117	1	41, M	Left hand	STSG	CS, plasma exchange	Successful graft, healed without recurrence
5	Dean et al. (1991)118	1	71, F	Leg (ankle)	Cultured epithelial autografts	CS	Successful, healed without recurrence
6	Límová and Mauro (1994)119	1	44, F	Leg (dorsal food)	Cultured keratinocyte autografts	CS systemic+intralesional	95% Graft take, healed without recurrence
7	Maier et al. (1995)120	1	34, F	Leg	STSG	CS, CyA	Successful graft, healed without recurrence
8	Vereecken et al. (1997)121	1	87, F	Leg (right tibia)	STSG	Lymecycline, topical benzoyl peroxide	Successful graft, healed without recurrence
9	Michel et al. (1998)122	1	68, F	Leg (calves)	STSG	CS, CyA, MMF	Successful graft, healed without recurrence
10	Havlik et al. (1998)123	1	15, F	Breast	STSG	CS	Successful graft, healed without recurrence
11	Cliff et al. (1999)92	4	66, M; 36, F; 28, F; 68, F	Legs (lower) 2, abdomen, tight left	STSG (4)	One without immunosuppression (unknown PG), CS (3)	One failed with exacerbation (Patient without immunosuppression), three healed with no recurrence
12	Sotillo-Gago et al. (1999)124	1	43, F	Breast (bilateral)	STSG without IS	No IS then CS+CyA	STSG failed without IS, healed under immunosuppression
13	Rozen et al. (2001)113	3	61, F; 73, F; 46, M	Legs (all lower)	STSG (3), rectus abdominis flap (1)	CS (3)+AZA/dapsone/Mesalamine 1; +AZA/MMF/Tacrolimus 1	Successful graft in all, one minor recurrence after 3 months managed conservatively
14	Kaddoura and Amm (2001)112	3	13, F; 15, M; 56, M	Legs (right lower leg four ulcers, left lower leg two ulcers)	STSG (3)	CS (3)	Successful in all, one small recurrence managed conservatively with systemic CS
15	de Imus et al. (2001)125	1	26, F	Leg	Allogeneic cultured human skin equivalent (Graftiskin)	CS, CyA	Successful, healed without recurrence
16	Gilmour and Stewart (2001)126	1	16, M	Leg	Cultured epidermis unsuccessful, followed by NPWT and delayed STSG	CS, CyA, MMF	Healed after 7 months, no recurrence after 10 months
17	Lifchez and Larson (2002)127	1	40, F	Breasts	STSG	CS	Successful graft, healed without recurrence
18	Gulyas and Kimble (2003)128	1	57, F	Breast	STSG	CS	Successful graft, healed without recurrence
19	Poucke et al. (2004)129	1	41, F	Breast	Integra®	CS	Healed without recurrence
20	Horner et al. (2004)130	1	21, F	Breast	NPWT, then STSG	First no IS, then CS+CyA	Failed without IS, secondary healing under immunosuppression
21	Zakhireh et al. (2004)131	3	64, F; 60, F; 30, F	Legs (all lower)	STSG (3)	CS (3) with CyA (2) one with additional MTX	Successful graft, healed without recurrence
22	Geller and Longton (2005)101	1	82, F	Leg (lower)	NPWT	CS	Halted the inflammatory process, healing after 7 months
23	Niezgoda et al. (2006)132	1	46, M	Leg (left, tibia)	NPWT, STSG, HBO	CS	Successful graft, healed without recurrence
24	Keskin et al. (2006)133	1	9, M	Legs (lower)	STSG	CS	Successful graft, healed without recurrence
25	Ghersi et al. (2007)134	1	57, F	Leg (lower left) controlled and slow-healing PG	NPWT	None, previous CS, IVIG, MMF, cyclophosphamide	Healed without recurrence after 6 weeks
26	Zutt et al. (2007)85	1	50, F	Legs (right: lower leg and dorsal foot)	NPWT and STSG	CS, AZA, CyA, dapsone, MMF, Iloprost	Successful graft, healed without recurrence
27	Hoff et al. (2008)135	1	53, M	Both lower legs	STSG	CS, topical tacrolimus	Successful graft, healed without recurrence
28	Kim et al. (2008)136	1	47, F	Legs (left calf)	STSG	CS, CyA	Successful graft, healed without recurrence
29	Toyozawa et al. (2008)137	1	69, F	Legs	STSG	CS	Successful graft, healed without recurrence
30	Descheemaeker et al. (2008)138	1	31, F	Breast	NPWT and STSG	CS	Successful graft, healed without recurrence
31	Procianoy et al. (2009)139	1	19, F	Eyelid	FTSG	CS, dapsone	Successful graft, healed without recurrence
32	Rajapakse et al. (2010)140	1	50, F	Breasts	NPWT	CS	Successful graft, healed without recurrence
33	Wollina et al. (2000)141	1	50, F	Both lower legs	Hyaluronic acid derivate (Hyalogran) followed by grafting of autologous cultured keratinocytes	CS, MMF	One ulcer healed, second ulcer 60%; recurred shortly thereafter
34	Rietjens et al. (2010)142	1	37, F	Breast	FTSG	CS	Successful graft
35	Schintler et al. (2010)143	1	45, F	Breasts	NPWT (instill) and free flap	CS, CyA, IVIG	Successful graft, healed without recurrence
36	Lamet et al. (2010)144	1	44, F	Legs	STSG	CS	Graft at first successful, then relapsed
37	Schoemann and Zenn (2010)145	1	50, F	Abdomen	STSG	CS	Successful graft, healed without recurrence
38	Melson et al. (2010)114	1	41, F	Eyelids	FTSG	CS	Successful, recurred after 6 months
39	Kikuchi et al. (2010)146	1	28, F	Abdomen	STSG	CS	Successful graft, healed without recurrence
40	Goshtasby et al. (2010)147	1	49, F	Breasts	Integra, STSG	CS, infliximab, MTX	Successful graft, healed without recurrence
41	Andreev (2011)148	1	30, F	Legs	STSG	CS	Successful graft, healed without recurrence
42	Hafner et al. (2006)149	1	64, F	Lower Leg	Cultured autologous epidermis	CS, CyA, AZA	Successful, but recurred within 6 weeks
43	Mandal et al. (2006)150	3	72, F; 17, M; 45, F	Leg (right knee) 2, thighs and groin 1	NPWT (3) and STSG (2) and gastrocnemius muscle flap (1)	CS (3), and CyA (1)	Successful flap (1), partial rejection at the STSG borders healed conservatively (1), improvement with NPWT healed within 6 months; no recurrences
44	Lambropoulos et al. (2011)96	1	13, F	Leg (left ankle) and left arm	NPWT, cultured autologous keratinocytes, (Epicel®)	CS, CyA	90% Graft take, no complete healing after 6 months
45	Hill et al. (2011)151	1	63, F	Leg (after knee arthroplasty)	NPWT, muscle flap with STSG, HBO	First none, CS after diagnosis	NPWT ineffective without immunosuppression, flap and graft failure secondary to sepsis, prolonged course healed eventually with HBO, no recurrence
46	Gámiz and Nielsen (2011)152	1	58, F	Both legs	NPWT and STSG	CS	Successful graft, healed without recurrence
47a	Saracino et al. (2011)153	2	Not specified (series of 26 mean age 61.5, 65% F)	Legs	Pinch graft (1), STSG (1)	CS (2)	Successful graft, no recurrence
48	Fraccalvieri et al. (2014)154	4	53, M; 59, F; 72, F; 45, F	Legs (lower) 3, breasts 1	NPWT (4)	CS (4), CyA (4), Dapsone (1), IGIV (1), cyclophosphamide (1)	Marked pain reduction, good granulation tissue, no healing reported
49	Sick et al. (2012)2	3	62, F; 70, F; 44, M	Legs (all)	STSG (3) and NPWT (1)	CS/AZA (1), CS/CyA (1), MMF/IVIG (1), topical tacrolimus and clobetasol (3)	Successful/one recurrence at a different location
50	Neiderer et al. (2012)155	1	76, M	Leg (anterior left)	NPWT, bioengineered skin substitute generated from neonatal foreskin (Apligraf®)	CS	Healed within 6 weeks
51	Kolios et al. (2012)88	1	47, F	Breasts	STSG	CS	Successful graft, healed without recurrence
52	Carrasco Cubero et al. (2012)156	1	33, M	Leg (right ankle)	STSG	CS, infliximab	Successful graft, healed without recurrence
53	Das et al. (2012)157	1	53, F	Legs, right groin	STSG	CS	Successful graft, healed without recurrence
54	Andrisani et al. (2013)158	1	76, M	Left breast	STSG	CS, infliximab	Successful graft, healed without recurrence
55	Araújo et al. (2013)159	1	50, M	Leg (left)	STSG, HBO	CS, AZA	Over 90% graft take, healed without recurrence
56	Cabalag et al. (2015)60	3	Not specified (series of 29, 59% F, mean age 71)	Legs	STSG (3) with HBO (2)	CS (3), CyA (1)	Successful graft (3), healed (1), improved (2), died due to sepsis: perforated sigmoid diverticular abscess (1)
57	Mowlds et al. (2013)160	1	61, F	Right hand	STSG and NPWT and HBO	Local acetic acid, no immunosuppression (PG diagnosed later)	Healed, second PG at the left hand
58	Lee et al. (2013)161	1	47, F	Legs (left foot)	STSG	CS	Successful graft, healed without recurrence
59a	Mooij et al. (2013)162	2	Not specified (series of 6 mean age 49, 83% F)	Legs (6)	STSG (2)	CS (2), infliximab (2)	Successful graft, healed without recurrence
60	Leonard and Khan (2013)163	1	47, F	Breast	NWPT+STSG	CS	Successful graft, healed without recurrence
61	Ratnagobal and Sinha (2013)164	1	60, F	Leg and food	STSG+HBO	Infliximab	Successful graft, recurrence after 3 months (infection), further STSG failed, primary healing with no recurrence
62	Fakhar et al. (2013)165	1	41, F	Breasts (PG after III° burn)	STSG, NPWT after breast amputation	CS	Not successful, eventually healed under prednisolone, MMF, and adalimumab
63	Aydin et al. (2015)166	1	32, F	Abdomen (after caesarian delivery)	NPWT and STSG	CS, AZA	Successful graft, healed without recurrence
64	Ye and Ye (2014)167	2	51, F; 89, F	Lower leg (1); abdomen, left tights, sacral (1)	STSG (2), primary closure sacral	CS (2), MMF (1)	>90% Graft take, healed without recurrence
65a	Stair-Buchmann et al. (2015)90	6	Not specified (series of 7: 29% F, mean age 65)	Legs (lower) 6	NPWT and STSG (4), STSG (1), NPWT (1)	CS (6), plus Dapsone (1), infliximab (1)	Three closed (NPWT and STSG 1, STSG 1), one died due to sepsis (NPWT and STSG), two lost to follow-up
66	Richmond et al. (2014)168	5	40, F; 50, M; 50, F; 30, F; 80, M	Legs (lower 4, upper 1)	Epidermal grafting with suction blister–harvesting system 5 (Cellutome)	CS (3) with CyA (1); CyA (2) with MMF (1)	All successful, three healed within 3 months, two marked ulcer size reduction
67	Tay et al. (2014)169	1	58, F	Legs (lower)	STSG	Sulfasalazine, topical betamethasone	Successful, healed within 3 months
68	Novo-Torres et al. (2015)170	2	23, F; 45, F	Leg (lower) (2)	CS+CyA+infliximab (1), no systemic therapy (1)	NWPT+STSG+Integra (2)	Successful, healed within 3 months (1), failed (1) in patient with no immunosuppressive treatment
69	Snyder et al. (2015)171	1	77, F	Leg (lower)	NWPT+allograft	CS, Dapsone, MMF	Successful
70	Cicuto et al. (2015)172	1	62, F	Breast	NWPT+skin graft (1)	CS	Successful, healed without recurrence
71	Soncini et al. (2016)173	1	19, F	Breast bilateral	NWPT	CS	Successful, healed within 4 months
72	Leitsch et al. (2016)174	1	54, M	Hand and forearm	STSG+NPWT	CS	Successful graft, healed without recurrence
73	Zaugg et al. (2016)175	1	65, F	Leg (lower)	STSG	CS	Successful graft, healed without recurrence
74	Patel et al. (2017)176	1	41, F	Breast	First cadaveric allograft, then Integra and STSG	CS+CyA+IVIG	Failed, healed by secondary intention
75	Hradil et al. (2017)177	1	72, M	Arm	STSG	CS	Successful graft
76	Yamaguchi et al. (2018)178	2	48, M; 51, F	Leg (lower) 2	NWPT (2)	CS (2)+CyA (1)	Successful, marked improvement within 1 month
77	Soro-García et al. (2017)179	1	35, F	Parasternal	NWPT	CS	Healed within 6 months
78	Domej et al. (2017)180	1	75, F	Shoulder	NWPT	CS	Improved
79	Zelones and Nigriny (2017)104	1	37, F	Abdomen	NWPT	First no treatment, then CS, MMF, infliximab	Worsening without IS, then healed within 5 months
80	Tanini et al. (2017)181	1	39, F	Breast	NWPT	CS	Healed within 2 months
81a	Riyaz et al. (2017)182	3	Not specified series of 61 patients, mean age 49.7, 52% male, 3 surgically treated (1 man, 2 woman)	Not specified	STSG (3)	CS (3)	Healed (2), failed (1)
82	Ishikawa et al. (2015)197	1	69, M	Abdomen (peristomal)	NPWT and STSG	None	Successful graft, healed without recurrence
83	Asyyed et al. (2018)183	1	46, F	Bilateral breast	STSG	CS, CyA, IVIG	50% Graft take
84	Choi and Yoo (2018)184	1	61, F	Leg (lower)	2 × STSG without IS failed	None, then CS	STSG failed without IS, secondary healing under CS
85	Ehrl et al. (2018)185	1	NR	Whole thorax after breast augmentation	STSG+Integra	CS	Successful graft, healed without recurrence
86	Gupta et al. (2018)186	1	34, F	Scalp	STSG	None	Failed, secondary healing under immunosuppressive therapy
87	Schwaiger et al. (2018)187	5	64, F; 80, F; 46, F; 66, M; 65, M	Ankle (1), leg (lower) (4)	STSG (3), STSG+FTSG (1), NWPT (1)	CS (5)	Healed, after 2-year recurrence treated successful with STSG (1), failed with FTSG, healed with STSG (1), healed (3)
88	Yu et al. (2018)188	1	51, M	Abdomen peristomal	NWPT+STSG	None	Successful graft, healed without recurrence
89	Romanelli et al. (2018)189	1	75, F	Leg (lower)	Cadaver allograft	CS+adalimumab	Successful graft, healed without recurrence
90	Zheng et al. (2020)190	1	30, M	Leg (upper)	STSG	CS	Successful graft, healed without recurrence
91	Dillingham and Jorizzo (2019)191	1	68, M	Leg (lower)	Porcine urinary bladder matrix+NWPT+STSG	CS	Successful graft, healed without recurrence
92	Yang et al. (2019)192	1	54, M	Leg (lower)	NWPT+oxygen irrigation	None	Healed within months
93	Sousa et al. (2020)193	1	57, M	Leg (upper) vaginal and perianal	NWPT	CS+infliximab	Healed within months
94	Shen et al. (2019)194	1	32, F	Abdomen	NWPT+STSG	CS, IVIG, CyA	Successful graft, healed without recurrence
95	Borda et al. (2019)195	1	40, F	Leg (lower)	FTSG	MTX, CyA, infliximab	Wound reduction of 28% after 8 weeks with much improvement
96	Nieto et al. (2019)75	1	62, M	Leg (lower)	NWPT	GC+ustekinumab (after failure with infliximab)	Healed within 20 weeks
97	Grado Sanz et al. (2020)196	1	31, F	Leg (lower)	NWPT+punch graft	GC, CyA, hydroxychloroquine	Healed within 2 weeks
98–101	Our published experience: Deluca et al. (2015),56 Larcher et al. (2015),198 Pichler et al. (2016),199 Pichler et al. (2017)200	21	Mean age 62, 67% F (14, F; 7, M)	Legs (19), breast (2), occipital (1)	NPWT and STSG (18), STSG (1), NPWT (2)	CS (21), dapsone (10), infliximab (4), adalimumab (1), MTX (2), IVIG (2), MMF (1), hydroxychloroquine (1), Pentoxifylline (5), Iloprost (4), simvastatin (4)	Completely healed: 16 (76%), with primary healing without recurrence after first treatment cycle in 14 (67%), 2 patients healed after recurrences treated with a second cycle. Over 90% improvement: 3 (14%). Two patients still with larger ulcers (10%): one had healed, but recurred twice, 1 never improved more than 30%. Recurrences observed: 7 (33%), 4 singles, 3 with two or more
Preliminary observations in this report
This report	Unpublished data	1	48, F	Lower legs	Porcine xenograft+NPWT, then STSG+NPWT	CS, dapsone, ustekinumab (after failure of infliximab)	STSG 2 weeks after xenotransplant, then successfully healed within 3 weeks after STSG with and IS (CS, dapsone, ustekinumab)
This report	Unpublished data	1	47, F	Lower leg	STSG, porcine xenograft (EZ-Derm®)	CS, dapsone, Pentoxifylline, simvastatin	STSG failed 3 × without IS, year-long deep PG healed within 6 months
Summary	101 Articles	161 Patients	Mean age 52 95% CI 2.6, median 54, minimum 9, maximum 89, F, 115, M, 46 p < 0.001	Legs 113 (71%), breast 19 (12%), abdomen 10 (6%), other 18 (11%), of those, more than one localization 8 (5%), legs significantly more frequent than all others p < 0.001	STSG alone 59 (37%), STSG together with HBO, flap or dermal substitute 11 (7%), NPWT alone 22 (14%), NPWT with STSG 41 (26%), NPWT with flaps, punch grafts, dermal substitutes or xenografts 7 (4%), other grafts (FTSG, Pinch, suction blister, cultured cells) 21 (13%)	CS first 149 (93%) after establishment of diagnosis 156 (98%), plus a second IS 41 (26%), plus two or more IS 38 (24%). IS being CyA 31 (19%), dapsone 18 (11%), infliximab or adalimumab 18 (11%), MMF 13 (8%), IVIG 7 (4%), AZA 6 (4%), pentoxifylline 6 (4%), others used in <5 patients	139 Patients finally healed (86%), of those 17 patients healed, but had one or more recurrences (11%), 16 (10%) improved, 2 died of sepsis (1%). Healing took longer than 2 months in 22 patients (14%), of those, 17 were treated with NPWT, p < 0.001 for longer healing time with NPWT alone. 18 Partially previous, partially recurrent, or primary procedures on 15 patients failed, 12 (67%) of those without IS p < 0.05

Author, publication year, number of patients treated with NPWT or grafts, age, gender, localization of PG, procedures performed, systemic treatment, and outcome are reported.

^aIn the case of series, where exact data on age and gender were missing (No. 47, 59, 65, and 81), given mean age was utilized and the corresponding sex percentage.

AZA, azathioprine; CI, confidence interval; CS, corticosteroids; CyA, cyclosporine A; F, female; FTSG, full thickness skin graft; HBO, hyperbaric oxygen; IS, immunosuppression; IVIG, intravenous immunoglobulins; M, male; MMF, mycophenolate mofetil; MTX, methotrexate; NPWT, negative pressure wound therapy; NR, not reported; P, patients; PG, pyoderma gangrenosum; STSG, split thickness skin grafting.

Patient management

Patient characteristics of our experience are also summarized in Table 1 and supplemented with two additional cases not previously published. From December 2011, after the first successful treatment of a patient with chronic PG with NPWT and STSG,⁴⁶ all consecutive patients with PG at the department of dermatology, Bolzano, Italy, were treated with NPWT and STSG. The diagnosis of PG was based on the clinical presentation, the exclusion of an arterial or venous vasculopathy, skin biopsies, and laboratory investigations, including blood and differential blood count, transaminases, creatinine, urea, electrolytes, glucose, thyreoid stimulating hormone, urine analysis, antinuclear antibodies, anti-neutrophil cytoplasmic antibodies, rheumatoid factor or anti-citrulline antibodies, screening for antiphospholipid antibodies, complement C3 and C4, and serum electrophoresis.

Surgical approach

All patients were hospitalized before surgical intervention and immediately received 0.5–1 mg/kg daily (methyl)prednisolone, in addition to further immunosuppressive or immunomodulatory therapy (Table 1). In 20 of 23 patients, a foam-based NPWT (ActiV.A.C.^® therapy system; KCI Europe Holding BV, London, United Kingdom, GBR first 17 patients or Vivano System^®; PAUL HARTMANN AG, Heidenheim, Germany, latest 4 patients) was started before further surgical intervention. In all patients with V.A.C., the black polyurethane foam (V.A.C. GranuFoam) was used, except on one patient, the black polyurethane-foam was substituted by the white polyvinyl alcohol-foam (V.A.C. WhiteFoam) because of marked pain using the black foam. The negative pressure applied was −125 mmHg, which was reduced to −80 mmHg on one patient due to pain.

All interventions were performed under local tumescence anesthesia as previously described,^56,57 except in a bilateral mammary case that was conducted in general anesthesia and the xenotransplantations that did not require anesthesia. Briefly, 0.4% mepivacaine solution in ringer lactate and 0.001% epinephrine was used. In cases where more than 100 mL was required, mepivacaine was further diluted to 0.2% for donor site anesthesia.

Gentle surgical debridement of the wound bed and wound borders was performed; thereafter, 0.3 mm thick STSGs were removed with an electric Dermatome (Aesculap Power Systems, Acculan^® 3Ti Dermatome; Braun, Tuttlingen, Germany) and transferred to the recipient site and fixed by NPWT for 3–5 days. All patients also received prophylactic antibiotic therapy, while undergoing surgical management, and thrombosis prophylaxis was performed with subcutaneous low molecular weight heparin (Fig. 4, proposed management algorithm).

Figure 4.

Proposed holistic treatment algorithm for PG with basic measures, surgical procedures, possible adjuvant treatment, and a proposed immunosuppressive treatment ladder. Color images are available online.

Statistical analysis

Demographic (e.g., age) results are expressed as median, including mean for age and the 95% confidence interval (CI). Comparisons between groups were performed by the chi-square test and a p-value <0.05 was regarded as statistically significant.

Results

Up to date, including our series, surgical management of 161 patients with PG has been reported in the literature with a female predominance (females 115; 72%; p < 0.01), with a mean age of 52 years (95% CI ±2.6 years, median 54; range 9–89). Ulcers were significantly more often situated on the lower extremities (n = 113; 71%; p < 0.01 vs. all other localizations), followed by breasts (n = 19; 12%), abdomen (n = 10; 6%), and other localizations (17; 11%). Eight (5%) patients had ulcers on more than one of those sites.

Surgical treatment consisted in STSG alone (n = 59; 37%), STSG together with NPWT (n = 41; 26%), STSG together with hyperbaric oxygen, flap or dermal substitutes 11 (7%), NPWT alone (22; 14%), NPWT with flaps, punch grafts, dermal substitutes, or xenografts 7 (4%), or other forms of grafting (n = 21; 13%) being cultured cells or tissues, punch grafts, suction blister grafts, or full thickness skin grafts (FTSG).

Systemic immunosuppression

Systemic immunosuppression consisted mainly in CS (n = 156; 98%), supplemented with one (n = 41; 26%) or more immunosuppressive drugs (n = 38; 24%), being cyclosporine (n = 31; 19%), dapsone (n = 18; 11%), infliximab or adalimumab (n = 18; 11%), mycophenolate mofetil (n = 13; 8%), intravenous immunoglobulins 7 (4%), azathioprine 6 (4%), pentoxifylline (4%), and others used in <5 patients. Treatment was finally successful in 139 cases (86%), 17 of those (11%) had one or more recurrences. An improvement after the procedure was noted in 16 (10%) patients. Two patients died due to sepsis (1%).

The reports included 18 treatment failures, some of them anamnestic, on 15 patients, 12 of the treatment failures were performed without immunosuppression (67%) and failure occured significantly more often when immunosuppression was missing (p < 0.05). Failure with immunosuppression occurred with the following procedures: transplantation of cultured cells (n = 1), FTSG (n = 1), STSG without NPWT (n = 3), and one case treated with cadaveric allograft, dermal substitute (Integra^®), and STSG without NPWT (n = 1). No case of pathergy was observed, neither at the side of PG nor at the skin graft donor sites.

Personal experience

Our personal experience comprises 23 patients, with a median of hospitalization of 4 weeks (mean 8.2 weeks, 95% CI ±3.6 weeks, range 10 days to 30 weeks). The extremely painful ulcers were all situated on the lower extremities, except in one case of occipital and one case of postaugmentation mastopexy PG. The ulcer dimension ranged from 1 × 1 cm, the smallest ulcer in a patient with multiple PG, to 45 × 25 cm. The disease duration of PG in median was 5.5 months (range 1 month to 30 years).

NPWT was started within the first days of hospital admission and systemic immunosuppression was adapted or started simultaneously or a few days before applying NPWT. Immunosuppression consisted in systemic CS (n = 23), dapsone (n = 12), infliximab (n = 4), adalimumab (n = 1), ustekinumab (n = 2), methotrexate (n = 2), and mycophenolate mofetil (n = 1). Immunomodulatory adjuvant treatment consisted of intravenous immunoglobulin (n = 2), pentoxifylline (n = 6), simvastatin (n = 5), and systemic intravenous iloprost (n = 4). As basic measures, an anabolic protein-rich diet was administrated, and other diagnoses were ruled out.

To tolerate the NPWT, especially in the first night after application, adequate pain medication was given, including opioids. Despite the initial painful application of NPWT, pain was markedly on the decrease within the first 24 h of NPWT. Nineteen patients were treated simultaneously with NPWT and STSG, one of them with a previous porcine xenograft described below. Two patients were not grafted because PG first showed fast intrinsic healing with good epithelialization under NPWT and the mammary case was grafted without applying NPWT. Thirteen of the 19 patients grafted had a complete graft take and healed within 2 weeks after STSG. Six patients had a 75–90% graft survival and then they further improved or healed with follow-up.

Four patients treated contemporarily with NPWT and STSG showed a reactivation in the grafted area in 1, 2, 3, and 11 months after intervention and one patient had two recurrences after 4 and 9 months, but all in much minor dimensions. Of those, one patient had a traumatic abrasion of the grafted area and one patient stopped immunosuppressive CS treatment against medical advice a week before reactivation, which might have facilitated the relapse. All patients were successfully retreated. One further patient developed a second PG under infliximab and glucocorticoids (GC) at her left breast after 6 months and was successfully treated and healed within 4 weeks with NPWT followed by STSG secured by NPWT under GC, dapsone, and ustekinumab (Fig. 1).

Figure 1.

Forty-four-year old woman who developed a second PG on her left breast successfully treated with NPWT and STSG. (a) First day of presentation with painful erythematous-violaceous plaque on her left medial breast. (b) Two days later with painful PG ulcer before application of NPWT, (c) stabilized PG after NPWT before STSG. (d) PG after STSG secured 5 days by NPWT, (e) healed ulcer 3 weeks after STSG (f) and at 1-year follow-up. NPWT, negative pressure wound therapy; PG, pyoderma gangrenosum; STSG, split thickness skin grafting. Color images are available online.

The outcome of the patients treated with NPWT alone initially showed good intrinsic healing, but the time for complete wound closure nearly took 1 year for one patient and the second patient showed an epithelialization of 20% in 4 months and is still in follow-up. The median time of disease-free follow-up in the patients without relapse was 9 months (range 3–24 months). No reactivation of PG was observed at the skin donor sites. and donor sites healed without complications. The patient with postmastopexy PG underwent esthetic corrective surgery under 50 mg prednisolone after 6 months of the original PG without complications, especially without reactivation of PG.

Preliminary results in the treatment of pyoderma gangrenosum with porcine xenograft

Two previous unpublished patients were treated with porcine Xenografts (EZ Derm^®). The first patient was a 46-year-old woman who had previously failed three STSG performed without immunosuppression in a peripheral hospital. PG was known for 10 years and the current ulcer had been active and growing for 2 years. On hospital admission, she presented with a deep 12 × 17 cm PG ulcer on her left leg with already partially exposed tendons. She was treated by xenograft transplantation fixed by Vivano-NPWT for 5 days with antibiotic and thrombotic prophylaxis under immunosuppression with 0.5 mg/kg methylprednisolone, 100 mg dapsone and adjuvant 400 mg pentoxifylline, and 20 mg simvastatin.

Pain was immediately better the first day after xenograft application and the patient could be discharged after 2 weeks. One month after application of the porcine xenograft, the wound bed was covered completely by granulation tissue. Although the patient refused a definitive closure with STSG, a complete healing by second intention could be achieved after 6 months (Fig. 2). Two years after the first xenotransplant, the patient developed a smaller local recurrence after trauma, which again improved after grafting with EZ-Derm.

Figure 2.

Pyoderma gangrenosum successfully treated with NPWT and a porcine xenograft (EZ-Derm^®). (a) PG before application of NPWT and the porcine xenograft, (b) stabilized PG 2 weeks after xenograft application, (c) complete granulation of PG ulcer and closure of the free tendon 1 month after xenograft transplantation, (d) follow-up at 2 months, (e) 3 months (f), 4 months, and (g) 5 months. Wound closed at 6 months. Color images are available online.

The second patient was a 48-year-old woman with already known smaller recurrent cribriform PG ulcers on her legs that were under control with low-dose GC (4 mg methylprednisolone) and infliximab. She developed a large very painful ulcer a few days after a simple trauma, which did not improve with local treatment. She was admitted in our department for immunosuppressive and local treatment. The first application of NPWT failed due to pain, so a porcine xenograft was transplanted and fixed by NPWT for 5 days.

Pain disappeared within the first hours after xenotransplantation. The xenograft was removed with the first NPWT change and showed a markedly improved ulcer, which was covered by simple NPWT that was well tolerated without pain for a further 5 days. Thereafter, an STSG fixed by NPWT could be performed and the ulcer completely healed within 3 weeks after STSG. Her immunosuppressive treatment was changed from infliximab standard dosing (5 mg/kg) to subcutaneous ustekinumab 90 mg in month 1, 2, and every 2 months thereafter, beginning the day of the STSG; further 100 mg dapsone was added and oral methylprednisolone was increased to 36 mg and tapered thereafter (Fig. 3).

Figure 3.

Pyoderma gangrenosum successfully treated with NPWT and STSG. (a) PG before application of NPWT, (b) stabilized PG after NPWT before STSG. (c) PG after STSG (d) at 8-week follow-up, (e) at 3 months and (f) 6 months follow up. Color images are available online.

Discussion

Definitive guidelines for the treatment of PG are lacking due to the rarity of the disease. Therapy in general consists of a combination of systemic anti-inflammatory or immunosuppressive treatment as well as local wound dressings and the treatment of underlying associated diseases.^4,58

Systemic treatment

Systemic CS form the backbone of the immunosuppressive treatment. The common use of CS was confirmed in the literature search with a use in over 90% of cases (Table 1). Accordingly, all our personal treated patients received CS at some point of diagnosis. As a second step, we prefer dapsone as anti-inflammatory steroid-sparing agent over cyclosporine, especially in elderly patients, as it does not affect renal function or blood pressure and for its well-known additional antimicrobial properties.⁵⁹ Indeed, mortality rates from 11% to 16% have been reported for PG and were mostly related to sepsis,^5,60 and septic complications were the death cause in the two deceased patients from the literature review. Furthermore, the highest rate of adverse events per therapeutic attempt has been described to occur with cyclosporine (40%), followed by cyclophosphamide (38%) and azathioprine (32%) in a large study on 52 patients by Herberger et al.⁶¹

As third step, if deemed necessary, we commonly use the anti-TNF-alpha antibody infliximab for its known rapid action. It is the biologic most used in the treatment of PG and the sole agent with a positive randomized, double-blinded, placebo-controlled trial for PG showing improvement in 20 of 29 patients.^42,62,63 A recent retrospective study found the following complete response rates for TNF-alpha inhibitors: infliximab 64% (n = 33), adalimumab 57% (n = 28), and etanercept 71% (n = 7). These results were all superior to CS (49%; n = 78).^61,64

As second-line biologic treatment. we suggest the use of ustekinumab after the failure of infliximab. It is known to cause fewer side effects and has shown a higher efficacy than infliximab, and is recently commonly used for the treatment of PG after the failure of TNF-alpha inhibitors.⁶¹ So Herberger et al.⁶¹ reported the successful use in six of nine patients with PG and de Risi-Pugliese et al.⁶⁵ treated four patients successfully after the failure of infliximab. This could be confirmed in our two cases that failed infliximab, where ustekinumab (90 mg subcutaneous) was successfully used. Up to date, together with our cases, 31 patients have been treated with ustekinumab for PG, 24 (77%) healed, 3 (10%) improved, and 4 patients (13%) failed the treatment. Fourteen of the 31 patients had previously received a TNF-alpha blocker, mainly infliximab, and all of them achieved a complete response under ustekinumab (Table 2).^61,65–78

Table 2.

All reports in the literature so far reporting treatment of pyoderma gangrenosum with ustekinumab

Publication, Year (Reference)	Patients No.	Complete Response No.	Partial Response No.	Failure No.	Failed Anti TNF-Alpha No.	Previous Anti TNF-Alpha Molecule
Guenova et al. (2011)66	1	1	0	0	0
Fahmy et al. (2012)67	1	1	0	0	1	Infxmb, Ada
Goldminz et al. (2012)49	1	1	0	0	1	Ada, Goli, Infxmb
Greb et al. (2016)68,a
Cosgarea et al. (2016)69	1	0	0	1	0
Acquitter et al. (2015)70	1	1	0	0	1	Etn
Barbosa et al. (2016)71	2	1	1	0	0
Benzaquen et al. (2017)72	1	1	0	0	1	Ada
Low and Mar (2018)73	3	1	2	0
Nunes et al. (2019)74	1	1	0	0	1	Infxmb
Nieto et al. (2019)75	1	1	0	0	1	Infxmb
Herberger et al. (2019)61	9	6	0	3	NR	NR
García Cámara et al. (2019)76	1	1	0	0	1	Infxmb, Ada
Vallerand and Hardin (2019)77	1	1	0	0	0
Piqueras-García et al. (2019)78	1	1	0	0	1	Infxmb
De Risi-Pugliese et al. (2019)65	4	4	0	0	4	All Infxmb
This report	2	2	0	0	2	All Infxmb
Summary	n = 31	n = 24 (77.4%)	n = 3 (9.7%)	n = 4 (12.9%)	n = 14 (45.2%)	Infxmb 12, Ada 4, Goli 1, Etn 1

Patient numbers treated per report, outcome, and previous failure of anti-TNF alpha treatment are reported.

^aSame patient as Goldminz, but higher ustekinumab dose.

Ada, adalimumab; Etn, etanercept; Goli, golimumab; Infxmb, infliximab; No., number of patients, TNF, tumor necrosis factor.

We do not recommend the use of anti-IL-17 or anti IL-17 receptor antibodies as there are a few concerning reports of PG induced by this class of biologics^79–81 and of their well-known risk to induce or aggravate inflammatory bowel disease, known to be associated with PG.⁸²

For PG on the legs, we further used pentoxifylline as adjuvant treatment in six patients for its known positive rheologic, immunomodulatory, and anti-TNF-alpha activities. Indeed, it is advised by a Cochrane review for leg ulcers and successfully used in other diseases with neutrophil dysfunction like oral aphthous ulcers or Behcet's disease.⁸³

Another drug with anti-inflammatory activity and positive effects on wound healing demonstrated by Evangelista et al. in 2014 in a blinded randomized clinical trial is simvastatin.⁸⁴ After publication of this trial, it was used in five patients. In difficult cases, the adjuvant use of prostaglandins (iloprost) the day after STSG was used in four patients to facilitate skin grafting. The first to use iloprost in PG and STSG was Zutt et al.⁸⁵ Its positive effects on inflammation, ischemia, and venous leg ulcers are known.⁸⁶ Our systemic and chirurgic treatment approach is summarized in Fig. 4.

Surgical treatment

Surgical treatment of active PG ulcers was controversially discussed due to the pathergy phenomenon, which presents with new PG lesions or the rapid progression of existing ulcers after tissue traumatization or surgery.^87–89 On the other hand, PG ulcers without skin grafting show prolonged healing (up to several years), also in patients under adequate immunosuppression, which, as a side effect, also defers proper wound healing. Moreover, open wounds are more prone to secondary infections, which potentially represent a further pathergy trigger. Importantly, sepsis is a feared complication of PG in immunosuppressed patients.^60,90

In addition, long-term systemic immunosuppressive therapy is associated with a high rate of complications.⁶⁰ Therefore, an urgent clinical need exists to close open wounds in these patients as fast as possible. The first documented report of a successful skin autograft in PG was in 1968 by Ridenhour and Stephenson⁹¹ in a patient treated with a combination of local aqueous silver nitrate and systemic CS. Cliff et al. described the use of STSG in the management of PG in the first case series of four consecutive patients. They conclude that STSG may have a role in the management of this disease under immunosuppressive therapy.⁹² Subsequently, more authors described the role of surgical management in PG and up to date, including our series, 161 patients have been successfully managed surgically with skin grafts or NPWT combined with an effective immunosuppressive therapy (Table 1).

The simultaneous immunosuppression is of uppermost importance because STSG without immunosuppression exacerbates or triggers PG and new lesions might even occur at the donor site.^93–95 Six of seven patients have been successfully treated with cultured human cells or tissue, but one took longer than 6 months to heal and two recurred shortly thereafter. These results are not superior if compared to classical STSG alone or in combination with NPWT, and do not justify this costly and time-consuming procedure; indeed, the last attempt published with this method was 2011.⁹⁶ Similar results are achieved with dermal substitutes (Integra n = 7 and Apligraf^® n = 1), which were used in eight procedures (alone n = 1, with STSG n = 3, with NPWT and STSG n = 2, and with NPWT n = 2), two of those failed, one without immunosuppression (Table 1).

Aggressive surgical debridement alone without immediate skin grafting should not be performed as it leads to pathergy or aggravation of PG in up to 20% of cases even in patients with immunosuppression.⁶⁰

A valuable option for wound conditioning is NPWT. It has become an important tool for the management of complex wounds. Despite its use in daily practice, the evidence base demonstrating effectiveness of NPWT to heal chronic wounds remains poor due to the paucity of randomized clinical trials.⁹⁷ Better evidence exists for the NPWT after skin graft placement in difficult wound beds.⁹⁸ The percentage of overall skin graft take and maximum graft take is higher if grafts are secured by NPWT and significantly fewer repeated grafts are required if NPWT is used.^99,100 This is also confirmed by our personal experience and the cases from the literature. No single failure was described in patients where the STSG was secured by NPWT, whereas STSG without fixation failed in 14 procedures, in 4 with and in 10 without immunosuppression.

The first successful treatment of PG with NPWT without skin grafting was reported in 2005 by Geller and Longton.¹⁰¹ Since then, 22 patients have been reported with NPWT alone for the treatment of PG. As with STSG, it is important to perform NPWT in inactive PG or under adequate immunosuppression as NPWT alone does not halt the progression of active PG.^102–104 The combination of NPWT with skin grafting has been reported in 22 cases and our experience adds further 19 cases of PG under immunosuppression with NPWT and STSG. We confirm the safe use of NPWT in the treatment of PG with a drastic improvement of pain perception within the first day. Indeed, NPWT seems to halt the inflammatory process probably by modifying the inflammatory micromilieu by decreasing proinflammatory cytokines. Granulation tissue usually forms within few days.

Despite halting the inflammatory process, NPWT, without skin grafting, seems not to accelerate the healing time; indeed, the time to complete healing in the reported cases ranged from several months to years.^{92–95,97–100} A longer healing time than 2 months was reported in 22 patients; of those, 17 were treated with NPWT alone, this was significantly more often if compared to all other treatment methods with grafting (p < 0.001). This applied also for our two cases not grafted after successful stopping of the underlying inflammation by immunosuppression and NPWT, where healing took several months. In contrast, combined NPWT with STSG resulted in healing within 1 month after skin grafting (mean 4 weeks); therefore, a combination of NPWT with STSG is highly recommended to accelerate wound healing.

Preliminary observational study on the use of porcine xenografts in PG

In some patients, we are faced with temporary or persistent contraindications for autologous STSG. These can be wounds with exposed tendons or bones or atrophic skin at the donor sites due to long-term treatment with systemic CS.¹⁰⁵ To preserve the positive effect of wound bed preparation with NPWT under immunosuppression, an alternative way of durable wound coverage in these patients would be desirable. Skin substitutes, initially developed for coverage of partial thickness burns,^106,107 are used in different types of acute or chronic wounds.^108,109 They protect the wound bed from infection and water loss and can create an environment suitable for epidermal growth.¹¹⁰

In one of our patients with a PG on the lower leg, the ulcer had an exposed tendon; despite adequate wound bed preparation with previous NPWT, we applied a temporary wound coverage with a porcine xenograft (EZ Derm). This is an aldehyde crosslinked porcine dermis, which is applied directly on the prepared wound bed. In difference to cadaver allografts, no graft vascularization occurs. The porcine xenotransplant adheres tightly to the wound bed with accordingly positive effects on pain, confirmed also in our patients, wound environment, and risk of infection.¹¹¹ We prefer the porcine xenograft to other dermal substitutes because of our positive experience in burn patients and patient with difficult–to-treat wounds,¹⁰⁵ and the significantly reduced costs compared to other dermal substitutes like Integra or Apligraf. We further speculate that the good clinical effects observed by the xenograft might be a direct immunological effect of the xenomaterial itself, shifting the local immune response from an autoinflammatory granulocyte-mediated status to a T cell-mediated orientation. Indeed, after 5 days of contact mediated by NPWT, good “healthy” granulation tissue had been formed in both cases.

Future Directions

This largest comprehensive review confirms the paradigm change that surgical treatment of PG with STSG secured by NPWT is a safe treatment, if performed under adequate immunosuppression; indeed, it could be considered the first-line treatment option in larger PG ulcers as it markedly improves healing time. Due to deeper insights into pathogenesis and growing clinical reports, a broader utilization of biologic treatments and a shift from TNF-alpha to IL-12/23 or IL-23 antibodies alone is predictable, as the later shows good clinical responses with fewer side effects. The positive results with porcine xenotransplants might be due to immunological effects of the xenomaterial and are promising, but preliminary, and should be confirmed in a larger patient collective.

Summary

According to the literature and personal experience, wound preparation with NPWT and surgical treatment with gentle debridement and STSG under adequate immunosuppressive therapy is a very valuable and safe treatment of PG. This treatment accelerates wound healing, shortens the administration of pain medication and high-dose immunosuppressive treatment, and reduces the risk for infection. STSG can be safely performed in local tumescence anesthesia without complications or reactivation of PG in the postsurgical period, neither in the transplanted area nor on the donor site.

There remains the problem of the chronic nature of PG and the recurrence after tapering of immunosuppression or trauma,^112–114 as we observed in our case series with a recurrence rate of 28%. Therefore, a sustained immunosuppressive or immunomodulatory treatment with slow tapering after complete wound healing is suggested.⁶⁰ All recurrences reported, however, were minor and successfully revisited by surgery, NPWT, or more often, even conservatively.

Take-Home Messages

• PG is a rare but debilitating autoinflammatory skin disease leading to large very painful skin ulcerations mainly in the lower extremities.

• Mortality of patients with PG is elevated mainly to sepsis and treatment of PG ulcers is time-consuming, involving local wound care, including NPWT, systemic immunosuppression, and surgical interventions.

• PG itself is a feared complication of surgical interventions as trauma might aggravate the condition, but surgery, especially STSG performed with adequate immunosuppression, is a safe treatment method for PG.

• Without surgical intervention and systemic immunosuppression, ulcers take months to years for healing, if healing at all.

• The best surgical approach appears to be STSG secured by NPWT as this leads to higher skin graft take, no failure of this method has been reported in the treatment of PG.

• Despite halting the inflammatory process, NPWT alone, without skin grafting, does not accelerate healing time and should be followed by STSG if possible.

• In case of exposed tendons or dermatoporosis, where STSG is not an adequate option, dermal substitutes, especially porcine xenografts, might be a good treatment option.

• Immunosuppression should be performed with systemic glucocorticoids as backbone, but other steroid-sparing agents should be added. We propose the following ladder: first add dapsone and then the TNF-alpha antibody infliximab. If this should fail, the anti-IL-12/23 antibody ustekinumab should be tried in adequate dosing, as it has been shown to work when infliximab fails.

• We predict a broader utilization of biologic treatment and a shift from TNF-alpha to IL-12/23 or IL-23 antibodies alone, as the later shows good clinical responses with fewer side effects.

Abbreviations and Acronyms

CI

confidence interval

CS

corticosteroids

FTSG

full thickness skin graft

IL

interleukin

MMP

matrix metalloproteinase

NPWT

negative pressure wound therapy

PG

pyoderma gangrenosum

STSG

split thickness skin grafting

TNF

tumor necrosis factor

Author Disclosure and Ghostwriting

No competing financial interests exist. The content of this article was expressly written by the authors listed. No ghostwriters were used to write this article.

About the Authors

Klaus Eisendle, MD, MSc, PhD, MBA, is the president of the College of Health Care Professions Claudiana and Department Head of the Academic Teaching Department of Dermatology, Venereology and Allergology in the Central Teaching Hospital of Bolzano, the capital of South Tyrol province in Italy. As dermatologic surgeon, he has a special interest in wound care and wound healing. Tobias Thuile, MD, is a specialist in dermatology and wound management in Bolzano Central Hospital. Jenny Deluca, MD, is a specialist in dermatology, with focalization on immunology. Maria Pichler, MD, is a specialist in dermatology and made her thesis on PG.