Clinical guidance of pyoderma gangrenosum 2022

Pyoderma gangrenosum (PG) is a rare, neutrophilic skin disease. For the purpose of accurate diagnosis and proper treatment of PG, the Japanese clinical practice guidance for PG developed by the Japanese Dermatological Association was published in 2022. In this guidance, clinical aspects, pathogenesis, current therapies, and clinical questions on PG are described from the viewpoints of current knowledge and evidence‐based medicine. Here, the English version of the Japanese clinical practice guidelines for PG is presented and is intended to be widely referred to in the clinical examination and treatment of PG.


| Disclaimer
The present treatment guidelines do not intend to restrict physicians' discretion but describe the clinical standards based on objective facts from the clinicians' viewpoints. These treatment guidelines provide current, provisional standards for treatment and do not prevent treatments based on individual cases. Therefore, the guidelines cannot be used as evidence to inquire physicians' responsibility for performing treatments different from those mentioned in it. Treatments used in Japan or abroad are mentioned, even if they are treatments not covered by health insurance (unapproved drugs). Concerning individual drugs, it is important to obtain patients' informed consent based on the drug-package inserts and the latest information.

| Future plans for updating
The present treatment guidelines were prepared after a strong request from those in the clinical forefront of treatment of PG in the absence of domestic treatment guidelines or their equivalents. Basic and clinical research concerning PG are expected to be reported in the future and many clinical trials of new treatments are being planned. Currently, treatments with high evidence levels based on RCTs are limited, and there is clearly room for improvement for the future. We hope to update the treatment guidelines 3-5 years after their implementation by considering opinions voiced from those engaged in clinical practice.

| Declaration of conflict of interest
The members of the drafting committee of these treatment guidelines have made self-declarations of conflicts of interest (COI) based on the standards concerning COI at the institutions they belong to or the Japanese Association of Medical Sciences COI management guideline.
The cost of the preparation of the guidelines was borne by each committee member. The committee members have received no compensation for their services including the preparation of the test of the guidelines and attendance at meetings. Whether the members of the treatment guidelines drafting committee or their relatives within first degree of kinship received any of the compensations mentioned below from companies involved in the diagnosis/treatment of PG in the target period (January 1, 2020 until December 31, 2021) was reported.
(1) Executive/advisor compensations, (2) capital gain, (3) patent royalties, (4) lecture fees, (5) manuscript fees, (6) clinical research funding (consigned research funds, joint research funds, trial research funds), (7) grant donations, (8) honoraria for lectures, (9) receipt of travel expenses/gifts. It was stipulated that if any of the committee members was involved in the development of a particular related drug or had COI, the member would be excluded from the assessment of the recommendation grade of the treatment concerned.

| Collection of evidence
We tried to collect previously published studies as evidence, but as studies concerning PG with high evidence levels are few, case reports were also included if they were considered important.
Recommendation grades of treatment were determined after sufficient discussion among committee members (including treatments that may be effective although their effectiveness has not been confirmed by means of an RCT), and by considering the health insurance system of Japan. b. Implementation is recommended (at least one low-quality level II, high-quality level III, or very high-quality level IV evidence indicating the efficacy must exist) c. Consideration of implementation is supported, but there is no sufficient evidence (low-quality level III-IV, multiple highquality level V, or level VI evidence recognized by the committee exists) d. Not recommended because of the absence of evidence C Recommended not to be performed (high-quality evidence indicating inefficacy or harmfulness exists)

| List of the evidence levels and recommendation grades
The diagnosis of PG is more difficult than would be expected.
The reasons of the difficulty include that (1) characteristic clinical features are not necessarily observed (disappearance of edema in the ulcer margin with time or due to the effect of treatment); (2) few diagnostic criteria have been proposed, but the clinical features mentioned as major items lack objectivity; (3) clinical features are important for the diagnosis but are difficult to become accustomed to because of their rareness; (4) there are a wide variety of diseases that present with ulcers other than PG; (5) biopsy samples show neutrophil infiltration but lack histopathological findings specific to PG, and (6) there are no useful biomarkers. A characteristic of PG is the frequent occurrence and exacerbation of symptoms or the appearance of new lesions after a mild injury, or in a minor surgical wound (pathergy). Also, the presence of underlying diseases or comorbidities (e.g., inflammatory bowel disease, rheumatoid arthritis, blood disease, aortitis syndrome) helps with the diagnosis of PG.

| Diagnostic criteria
At present, there are no standardized diagnostic criteria for PG. The diagnostic criteria proposed to date are listed in Reference 1 and are described in Japanese as reference material (Table 1). In 1997, von den Driesch et al. defined PG as a condition that fulfills both of two major criteria: (1) the occurrence of primary sterile, chronic ulceration(s), typically with violaceous undermined borders; and (2) the exclusion of relevant differential diagnoses (e.g., pyoderma, ulcerations based on arterial or venous vessel diseases, and ulcerations based on a classic leukocytoclastic vasculitis). Furthermore, more than two minor criteria are required out of the following three items; (1) the histology from the border of the ulceration: neutrophil-rich infiltration of the dermis with signs for vasculitis and deposits of Ig (Ig) and/or complement factors on the vessels; (2) the presence of a relevant associated disease (e.g., chronic autoimmune bowel disease, chronic autoimmune arthritis, paraproteinemia, or haematoproliferative disease); and (3) response to treatment with systemic immunosuppressive therapy, but little or no response to conventional ulcer therapy. 6 The diagnostic criteria for PG proposed by Su et al. 7 require fulfillment of both of the two major criteria, (i.e., rapid progression of a painful, necrolytic cutaneous ulcer with an irregular, violaceous, and undermined border, and exclusion of other causes of cutaneous ulceration) as well as at least two of the following minor criteria: a history suggestive of pathergy or clinical finding of cribriform scarring; systemic diseases associated with pyoderma gangrenosum; histopathological findings (sterile dermal neutrophilia, ±mixed inflammation, ±lymphocytic vasculitis); and a treatment response (rapid response to systemic steroid treatment). 7,8 In contrast, according to a survey among German dermatology experts reported in 2014, the items, i.e., (1) rapid progression, (2) erythematous-iolaceous undermined borders, (3) exclusion of relevant differential diagnosis, and (4) rapid response to systemic immunosuppressive therapy, were most important for the diagnosis of PG. 9 In 2018, diagnostic criteria for ulcerative type PG were proposed by the Delphi consensus method. 10  In a clinical study conducted in Japan, the complete cure rate of target PG ulcers (pyoderma gangrenosum area reduction 100: PGAR100) was used as the primary evaluation criterion for the course of treatment. 14 In addition, objective assessment by evaluators using indices including changes in the ulcer area and severity of erythema of the ulcer margin, subjective evaluation of the dermatology life quality index, and pain by the patients themselves were also made.

| EPIDEMI OLOGY OF PYODERMA G ANG RENOSUM
Although the exact number is unknown, the annual incidence of PG is reported to be about 3-10 per million people annually. [15][16][17][18] The prevalence is reported to be 5.8 per 100 000 people. 19 The susceptible age range is 20-60 years, but the occurrence in older people is not rare. Females are slightly more often affected. 19 In

| Clinical characteristics
Pyoderma gangrenosum is a rapidly and efferently expanding, proliferative/gangrenous undermined ulcer with a slightly elevated, dike-like margin that is further surrounded by edema. It begins with sterile small pustules, small papules, or bloody blisters but is rapidly ulcerated and spreads to the surrounding area. During progression of ulceration, purulent fur attaches to the surface and bacteria is often detected secondarily. Edema surrounding the ulcer disappears in the chronic phase. PG frequently affects the extensor aspect of the lower leg but may also occur in the head and neck region, trunk, vulva, and fingers. Since the disease has few characteristic histopathological findings, it is often suspected primarily due to its clinical characteristics.

| Subtypes of pyoderma gangrenosum
Pyoderma gangrenosum is classified into five major subtypes: ulcerative, pustular, bullous, and vegetative (superficial) and peri-stomal types. The ulcerative subtype, which is the most common, is also called classic ulcerative PG. The disease predominantly affects the lower leg. In the pustular type, solitary pustules accompanied by a red halo occur in the head, trunk, and limbs. The bullous type often accompanies blood disease and frequently affects the dorsum of the hand, forearm, and lower leg. The vegetative type is an upwardly elevated rather than downwardly undermined lesion and is often regarded as synonymous with the superficial type because of the relative shallowness of ulceration. Of the superficial type lesions, those that histologically show pseudo-cancerous growth of the epidermis and those that show neutrophil and lymphocyte infiltration around giant cells in the dermis are called superficial granulomatous pyoderma, and it is often mitigated by external treatment. Multiple types occur simultaneously in many patients. Peri-stomal (parastomal) PG occurs around the stoma. It has been increasingly recognized recently, also by non-dermatologists and there are occasional consultations from nurses involved in stoma care. According to a report from a single facility, a rapidly enlarging type was observed in only one of the 11 patients, and PG was mostly the slowly progressive type. 22 Along with the above classification of five subtypes of PG, there are terms, such as postoperative PG and drug-induced PG as the cause of PG, and some studies include them in addition to the above five subtypes.

| Postoperative pyoderma gangrenosum
Concerning factors for the development and exacerbation of PG, while those of pathergy include slight stimuli, such as needle insertion for drip infusion or blood collection, PG induced by more invasive surgery or treatment is referred to as postoperative (postsurgical) PG. The surgery may be for anything including malignant neoplasm, hernia, and appendicitis, and even PG after a cesarean section is included. According to a report of 18 cases of postoperative PG from a single facility, it occurred most frequently after breast reconstruction, followed by the occurrence in the thoracic to abdominal region after surgery for malignant or benign diseases. Such PG more frequently affected females and even patients without an underlying disease. 23

| Rare sites
The head, face, ear, fingers, vulva, and mouth (oral mucosa and tongue) are relatively rare sites of PG. [32][33][34][35] Wang et al. 36 reported that PG avoids the nipple and areolar region, palm, and planta, which lack pilosebaceous gland tissue, but there are cases of PG affecting the planta. They proposed the pathological hypothesis that PG is the destruction of hair follicles by autoreactive T cells targeting hair follicle adnexa from the observation and that reactivation of PG is rare in areas after healing with scar formation. They also proposed that T cells rather than neutrophils infiltrate around blood vessels and hair follicles in early lesions.

| Pyoderma gangrenosum in pregnant women
During pregnancy, neutrophils are likely to be activated due to neutrophilia and a marked elevation of the blood G-CSF concentration associated with pregnancy, 37 and, therefore, the body is considered to be in a preliminary stage for induction of PG by external stimuli. In later pregnancy, tilting of the blood cytokine balance to Th1 may also contribute to the development of PG. 38 Induction of PG after Cesarean section has often been reported even without an underlying disease. 39 A case in which PG was induced after surgery for appendicitis and was induced again after 10 years, following a cesarean section on the suture line has also been reported. 40

| Pyoderma gangrenosum in children
According to a past report, children account for only 4%-5% of all PG patients. 41 A search of the literature for cases of PG in patients aged 18 years and younger, found that underlying diseases were observed in 93 of the 170 cases. Of the underlying diseases, inflammatory bowel disease (18 cases of Crohn's disease, 17 cases of ulcerative colitis) was the most frequent, followed by blood disease, angiitis, and immunodeficiency, in descending order. 42 No sex difference was observed. PG most frequently affected the lower limb but was also observed in the perineal region, head, mouth, back, peristomal region, upper limb, and thoracic region. Treatment was most frequently oral administration of corticosteroids alone, accounting for 42%.

| Pathergy
Pathergy is an excessive reaction to external stimulation/stress. Reaction to needle puncture observed in Behçet's disease as well as PG is also considered pathergy. Activated neutrophils may migrate to the damaged site and form a pustule/abscess there. Histological examination of biopsy specimens reveals infiltration of lymphocytes in addition to neutrophils. If epidermal keratinocytes are damaged, alarmin is released and inflammatory cytokines, such as interleukin-1β (IL-1β), IL-6 and TNFα, are produced. Moreover, pathergy can be induced by events such as activation of dendritic cells, production of chemokines that stimulate neutrophil migration, activation of T cells, and shift to the dominance of Th1 cytokines. 43 Pathergy is reported to be observed in 20%-30% of PG patients, 44 and is useful for the diagnosis of PG. On the other hand, of the 126 PG patients in Japan and abroad who underwent surgery (found by a literature search), postoperative exacerbation or reactivation of PG was observed in 21 (16.7%) patients, which was a high percentage. 45 In addition, in a study of 166 PG patients at a single facility, postoperative exacerbation or reactivation was observed in 25 (15.1%) patients, a comparable frequency. 46

| Clinical laboratory findings
In a period of high disease activity, C-reactive protein (CRP) and erythrocyte sedimentation rate, which are indices of acute inflammation, are elevated and leukocytosis and/or neutrophilia are present in peripheral blood. Moreover, although rarely, an abnormal increase in while blood cells may occur, a condition called a neutrophilic leukemoid reaction. 47 Also, anti-neutrophil cytoplasmic antibody (ANCA), (e.g., myeloperoxidase (MPO)-ANCA and proteinase 3 (PR3)-ANCA), may be detected. 48

| HIS TO PAT HOL OG I C AL PROFILE OF PYODERMA G ANG RENOSUM
The histopathological profile of PG is non-specific and varies widely according to the lesion type, stage (early or chronic), and sampling site (ulcer margin or bed). Therefore, ultimately, a histopathological examination is performed to eliminate other diseases, as in the clinical diagnosis. [49][50][51][52] A histopathological examination is particularly useful as a method for elimination of a malignant tumor, vasculitis, and infection. 7,52,53 For the above-mentioned reasons, it is difficult to determine the evidence level regarding the usefulness or specificity of the histopathological diagnosis of PG.
Basic findings common to all subtypes of PG including the classic (ulcerative) type are biphasic changes consisting of (1) necrotizing suppurative inflammation showing lytic changes in collagen fibers, bleeding, nuclear breakdown along with neutrophil infiltration in the center of the lesion, and (2) Sweet-syndrome-like vascular reaction accompanied by neutrophil infiltration in the margin. 50,51 The former causes destruction of existing tissues, including hair follicles and sweat glands, 52 often leading to ulceration, and such a course is particularly notable in the classic type. The latter finding is of lymphocyte infiltration in the vascular wall and around the vessels surrounded by neutrophils, usually without fibrinoid degeneration. 54,55 Although features of vasculitis accompanied by leukocyte nuclear breakdown are occasionally observed, it must be noted that in PG, vascular destruction is secondary thrombotic destruction due to neutrophilic inflammation and not primary vasculitis. 51,52 Also, in the early phase, edema of dermal papillae and neutrophilic interface dermatitis resembling dermatitis herpetiformis are occasionally observed. 51 In the pustular and bullous types, sterile, necrotic pustular change in hair follicles is the primary lesion. 49 52 There is also the view that PG starts with such folliculitis, i.e., a large number of neutrophils spread from around hair follicles densely into the dermis due to rupture of the infundibular hair follicles filled with neutrophils. Another opinion is that aggregates of neutrophils are formed subcorneally and intradermally by migration through the epidermis. 57,58 In the bullous type, depending on the sampling site, subcorneal bullae, subepidermal edema accompanied by intraepidermal bullae, and subepidermal bullae are observed; subepidermal and intraepidermal neutrophil aggregation is also a characteristic finding. 49,52 The vegetative type of PG is characterized by pseudoepitheliomatous hyperplasia and intradermal abscess accompanied by palisading granuloma consisting of histiocytes and foreign body giant cells. 59,60 There appears to be no mention of specific findings for the peristomal type. 61 If PG is complicated by Crohn's disease, granuloma may develop in areas where dense neutrophil infiltration and collagen fiber lysis/degeneration are observed. 50,62 Regarding a histopathological differential diagnosis, PG is suspected if (1) leukocytoclastic vasculitis can be excluded when an undermining ulcer is accompanied by dense neutrophil infiltration; (2) fungal, bacterial, and mycobacterial infection can be excluded by PAS, Grocott, Gram, or Ziehl-Neelsen staining or in some cases, by tissue culturing; or (3) a malignant neoplasm can be excluded by examination of atypia, along with clinical information. 50,51 Early lesions of PG are often indistinguishable from lesions of Sweet syndrome, but in Sweet syndrome, the folliculotropic changes observed in PG are rare, and lytic changes in dermal collagen fibers or necrotic changes in blood vessels are not observed, even in areas that show dense neutrophil infiltration. 51 By the direct immunofluorescence test, deposition of Ig (Ig) M, C3, and fibrin in the vascular wall is observed in a majority of patients; deposition of IgG or IgA is rare. 63 Deposition of Ig and complement is not specific to PG, and this test is performed for the exclusion of autoimmune blistering diseases, lupus erythematosus, and vasculitis. Lastly, if a skin biopsy is performed, it is recommended that spindle-shaped tissue samples are collected from areas encompassing the marginal erythema, which is the active lesion in a relatively early stage with vascular and follicular changes, including an ulcer bed. 52

| D IFFERENTIAL D IAG NOS IS OF PYODERMA G ANG RENOSUM
Diseases that present with lower leg ulcers should be considered in the differential diagnosis. Differentiation is necessary from (1) ulcers due to circulatory disorders; (2) ulcers due to thrombosis in antiphospholipid antibody syndrome and livedo vasculopathy; (3) ulcers due to malignant neoplasms, such as squamous cell carcinoma, basal cell carcinoma, and malignant lymphoma; (4) ulcers due to vasculitis, such as polyarteritis nodosa, granulomatosis with polyangiitis, rheumatoid vasculitis, and granulomatous vasculitis; and (5) ulcers due to infections, such as deep mycosis, non-tuberculous mycobacteriosis, and necrotizing fasciitis. [64][65][66] Biopsy is essential in the course, but histological exclusion of vasculitis, in particular, may be difficult. Features resembling vasculitis may be observed with infiltration of neutrophils and mononuclear cells around blood vessels in the dermis and fibrin deposition in the vascular wall, but it is often a condition secondary to a skin ulcer (pseudo-vasculitis) rather than genuine vasculitis. The diagnosis may still be difficult, even by skin biopsy, because the findings vary with the site and time of biopsy and biopsy from multiple sites may be required. In addition, ulcerated lesions resembling PG may be observed in granulomatosis with polyangiitis (Wegener's granulomatosis). 67 Histopathological examination may present with findings of granulomatosis with polyangiitis in some cases, but findings of PG in others, and the latter is occasionally called malignant pyoderma. In the differential diagnosis from primary vasculitis, primary vasculitis is considered more likely if ANCA is positive in peripheral blood in addition to skin biopsy findings of necrotizing vasculitis.

| PATHOLOGY OF PYODERMA G ANG RENOSUM
Although the cause of PG is unclear, sterile inflammation due to neutrophils plays the central role in its pathology as is clear from its histological profile and negative results of various cultures. In addition to activation and functional impairment of neutrophils, proliferation of T cell clones at the site of the lesion and Th1/Th17-predominant cytokine imbalance have also been reported, and both natural immunity and acquired immunity play important roles. 68 Various inflammatory cytokines, including the neutrophil chemotactic factors IL-8, G-CSF, and GROα, have been suggested to be involved in inflammatory reactions at the lesions.
Pyoderma gangrenosum is classified mainly into the ulcerative, pustular, bullous, superficial, and peristomal types, and the pathology may vary among these types. In 2000, Oka et al. 69 reported that PG-like ulcers developed on skin grafts from humans in mice administered IL-8, which is a neutrophil chemotactic factor, and that IL-8 production was increased in fibroblasts of PG patients. 70  is also enhanced, and as infiltration of cells of the natural immune system, such as macrophages and neutrophils, is induced by IL-17E, the role of IL-17E in the pathology of PG is highlighted. 78 On the other hand, it is reported that by comparison of neutrophilic dermatoses,t IL-17E production is increased in Sweet syndrome but low in PG. 79 Furthermore, TNFα is considered important in the pathology of PG, because the disease responds to anti-TNFα preparations, and because TNFα induces the production of IL-8 and growth-related oncogeneα. It was also reported that the IL-23 expression was enhanced at the lesions of PG. 80 In addition, PG patients with janus kinase (JAK)2 gene mutation have been reported, 81 and JAK inhibitors are effective for the treatment of PG. 82 Taken together, the following pathological condition of PG has been proposed: 83

| UNDERLYING DIS E A S E S/ COMPLIC ATIONS OF PYODERMA G ANG RENOSUM
Pyoderma gangrenosum often affects patients with an underlying disease, which is frequently an inflammatory bowel disease, such as ulcerative colitis and Crohn's disease, or rheumatoid arthritis. 5,20,87 In the case of a refractory ulcer of the lower leg complicating rheumatoid arthritis, exclusion of rheumatoid vasculitis by histopathological examination is required. Ulcerated skin lesions often reflect the activity of such complications, and exacerbation of the activity of bowel disease following skin lesions is a common experience. PG is also known to complicate blood diseases (myelodysplastic syndrome, malignant lymphoma, myelocytic leukemia). Racial differences in the frequency of complications have not been confirmed. 88 Table 3 shows details of 473 cases encountered at dermatology facilities in Japan. Among the underlying diseases/complications, ulcerative colitis (23.5%) was observed most frequently, followed by rheumatoid arthritis (9.7%), blood disease (8.7%), and Crohn's disease (7.2%). 20 Also, in a survey of 103 cases in the United States, PG was complicated by inflammatory bowel disease (34.0%), blood disease (20.3%), and rheumatoid arthritis (9.7%). Racial differences in the frequency of complications have not been confirmed. 89 Although PG is defined as neutrophilic dermatosis, activated neutrophils may cause extracutaneous symptoms. Arthralgia, joint symptoms (e.g., sternocostoclavicular arthritis), sterile abscess (lung, liver, kidney, spleen), eye symptoms (scleritis, corneal ulcer), and neutrophilic myositis are known as such conditions. In addition, there are instances of induction of neutrophilic dermatosis secondary to intestinal surgery, and the condition is called bowelassociated dermatosis-arthritis syndrome or bowel-bypass syndrome. It is considered to be caused by some immune reaction due to excessive bacterial proliferation in the intestine, involvement of immune complexes, or reaction to surgical stress. 5,89 There is also a pathological condition called aseptic abscess syndrome, which is an autoinflammatory disease that complicates Crohn's disease and produces sterile abscesses in the spleen and skin accompanied by abdominal pain and fever. 90 Palmoplantar pustulosis, psoriasis, and erythema nodosum have been reported as cutaneous complications of PG. 91,92 Folliculitis may be the initial symptom of PG. If severe acne or suppurative hidradenitis is concurrent, complication by an autoinflammatory disease should be suspected.

| Autoinflammatory disease and pyoderma gangrenosum
Autoinflammatory disease is a concept proposed in 1999, as a group of diseases that do not fall in the conventional category of immunological diseases. 93 It is a group of diseases with gene mu-

| Clinical matters
Three primary manifestations of PAPA syndrome are pyogenic arthritis, PG, and acne. These symptoms rarely occur simultaneously. Periodic fever, which is observed in other autoinflammatory diseases, does not occur. Pyogenic arthritis has onset in early childhood in many patients.
It is triggered by trauma, and neutrophil infiltration is observed by arthrocentesis, but cultures are negative and sterile. The disease often affects large joints and is recurrent. Joint destruction/contracture occurs in the advanced stage. PG develops from around puberty; it primarily affects the leg and is recurrent. Acne becomes prominent during or after puberty, is difficult to treat, and is likely to cause scarring. The disease has a relatively favorable prognosis, but because it is complicated by various chronic disorders, including hematological disorders (e.g., splenomegaly, hemolytic anemia, and thrombocytopenia), inflammatory diseases, (e.g., inflammatory bowel disease, uveitis, glomerulonephritis), and diabetes mellitus, caution is needed.

| TRE ATMENT FOR PYODERMA G ANG RENOSUM
The basic treatment for ulcerative lesions of the lower leg, including PG, is rest. lesions will not heal readily if the lower legs are not relieved from weight bearing. Therefore, treatment for PG also often requires hospitalization for rest. Hospitalization for detailed evaluation, treatment, and pain control is recommended in PG as well as other diseases regardless of the site, size, or number of ulcers. Topical treatment is the same as that is generally used for ulcers: maintaining a wet condition, use of an appropriate external preparation and dressing material, protection of the skin, and bandage for alleviation of edema. If necrotic tissue is adhered to the skin surface, it may be removed carefully with scissors, but excessive debridement and direct fixation of the adhesive surface of the dressing material to the skin should be avoided.
Checking for contact dermatitis due to previously used external preparations or disinfectant is also necessary.
Some treatment algorithms have been proposed. 96,101,102 There are not many major differences, and mild or early PG may be treated with external preparations, but oral drug therapy or the use of biological preparations should be considered for moderate, severe, or progressing PG. In addition, Maverakis et al. 96

| CQ3: Does biopsy exacerbate ulcers?
Comment: Theoretically, since PG is associated with the problem of pathergy and occurs in an ischemic condition due to secondary vasculitis, there is the possibility that biopsy exacerbates ulcers. Regarding the exacerbation of lesions due to blood flow disturbance, there are data of biopsy of lower leg ulcers instead of PG. Of the 866 patients studied, 614 (70.9%) had vascular ulcers, and exacerbation of the lesions after biopsy was observed in five (0.6%), all of whom had vasculitis. 116 Therefore, exacerbation of ulcers after biopsy of PG is considered to be due mostly to pathergy, but its frequency is unexpectedly low, reportedly being 2.9% (6/209). 46 According to that report, the risk of the recurrence of PG was higher after surgery than after minor invasion such as biopsy or needle puncture. Pompeo 117 observed that clinicians hesitate to perform biopsy in PG because of the concern over exacerbation of the lesion due to latent pathergy. But, based on his experience, Pompeo recommends initiation of corticosteroids therapy from before the biopsy and its continuation until elimination of PG.
This method is worth consideration if there is strong concern over pathergy. The judgment of whether biopsy should be performed is made case by case, and it should be performed strictly for the diagnosis by exclusion, with sufficiently explaining the possibility of exacerbation to the patient in advance.

| CQ4: Are histopathological findings in the ulcer or its margin useful for the diagnosis of PG?
Comment: In the classic (ulcerative) type, relatively useful findings are obtained in the center of the lesion in the pre-ulcerated stage, but, as the lesion is ulcerated, which means that tissue destruction is advanced, only non-specific findings remain in many cases. Therefore, it is recommended that tissue is collected by fusiform excision encompassing the ulcer margin, in which specific findings are more likely to remain. The ulcer floor is included, partly because the possibility of a malignant tumor cannot be excluded.

| CQ5: Is the immunofluorescence technique useful for the diagnosis of PG?
Comment: On the direct immunofluorescence method, deposition of IgG and IgA is observed rarely in the vascular wall in the lesions of PG, but deposition of IgM, C3, and fibrin is noted in most cases.
However, it is not a finding specific to PG, and immunofluorescence findings do not suffice for a definitive diagnosis. This method is significant for the exclusion of autoimmune bullous disease, lupus erythematosus, and vasculitis.

| CQ6: Does peristomal PG occur more often in patients with inflammatory bowel disease?
Comment: Peristomal PG, which is a relatively rare type, has been reported mostly in patients with inflammatory bowel disease and is often diagnosed as abscess, due to wound infection or at the suture site, contact dermatitis, due to stimulation by stool or urine leakage, and progression of Crohn's disease as an underlying disease. 22,61,115,118   12.11 | CQ11: Is external application of corticosteroids useful?

| Recommendation grade: B
Comment: Sixty documents could be retrieved by search of PubMed with "glucocorticoid," "topical," and "pyoderma gangrenosum" as search words. None of them was the report of an RCT concerning topical use of steroids. One was a report of an open-label study of peri-stomal PG, 122 one was a report of a prospective cohort study, 13 and two were case series studies about peri-stomal PG. 123,124 Although there were many case reports, the effectiveness of external steroid therapy was reported mostly by reports of its use for peri-stomal PG or localized PG, but there were also reports about superficial granulomatous PG and vegetative PG. 125 According to a report of a questionnaire survey about which treatment was used for PG, external therapy alone was given to about 11%; external therapy with some systemic therapy was administered in about 66% of the cases. The primary treatment for PG was mostly systemic therapy, but external therapy appears to be performed in combination with systemic therapy. 126 Since external steroid therapy is a treatment expected to be effective for relatively localized refractory PG, such as peri-stomal PG, superficial granulomatous PG, and vegetative PG, the recommendation grade was rated as B.

| Recommendation grade: B
Comment: 73 documents could be retrieved by a search of PubMed using "tacrolimus", "topical", and "PG" as search words. They included no RCT, one non-RCT, two cohort studies, and many case reports or case series reports. 127 In the cohort studies, case reports, and case series reports, relatively localized PG, such as peri-stomal PG and superficial granulomatous PG, was treated by external application of tacrolimus ointment, which was effective in most cases. 13,128,129 Topical tacrolimus therapy is a treatment expected to be effective against localized refractory PG, and its recommendation grade was rated as B. 12.14 | CQ14: Are anti-ulcer agents (e.g., Isodine sugar, Geben cream, Prostaglandin ointment, Actosin ointment) useful?

| Recommendation grade: C2
Comment: Three reports were retrieved by search using "povidoneiodine" and "PG" as search words. Among them, one case report stated that povidone-iodine sugar was effective as a topical treatment used in combination with systemic administration of methylprednisolone. 130 Five reports were retrieved by search using "sulfadiazine silver" and "PG" as search words. In one report, Geben cream (sulfadiazine silver) was used for the treatment of PG, but kidney disorder was caused by its use in a large dose (200 g/day). 131 One report was retrieved by search using "prostaglandin" and "PG" as search words, and it concerned a case of drug-induced development of new vascular lesions. No reference could be found by search using "bucladesine" and "PG", "dibutyril cyclic AMP" and "PG," or "cyclic AMP" and "PG" as search words. Search for reports of the use of Actosin ointment for PG could not be made. No report could be retrieved by searches using "prostaglandin E1", "topical" and "PG" or "alprostadil", "topical" and "PG" as search words. As a result, use external anti-ulcer agents for the treatment of PG is not recommended because of the lack of sufficient evidence. However, their use for chronic ulcers after the acute phase of PG, similarly to the treatment of usual refractory ulcers, has been found to be tolerated.

| Recommendation grade: A
Comment: Vegetative type PG is likely to be mitigated by external preparations alone, and small ulcers may be cured by external therapy alone. 13 However, systemic administration of corticosteroids, immunosuppressant, or colchicine is usually necessary. 101,102,132 The administration of prednisolone is started at 20 mg/day or 0.5 mg/kg and may be increased to 1 mg/kg for severe cases. Cyclosporine (3-5 mg/kg) is used in combination for poorly responsive cases and cases that show a tendency of reactivation with dose reduction, although it is used off-label. An RCT to compare the effectiveness of prednisolone and cyclosporine (STOP GAP trail) was carried out by dividing the participants into prednisolone (0.75 mg/kg; 75 mg/day at a maximum) and cyclosporine (4 mg/kg; 400 mg/day at a maximum) groups, but no significant difference was observed in the reduction rate of the ulcer area until after 6 weeks, or the cure rate after 6 months. In addition, no significant difference was observed in the recurrence rate between the two groups. 133 12.16 | CQ16: Is oral immunosuppressant therapy useful?

| Recommendation grade: A (cyclosporine), C1-C2 (other immunosuppressants)
Comment: Cyclosporine is a drug with the clearest rapid effect on PG, along with corticosteroids. It began to be used in the 1980s and has been reported to be effective in a large number of cases.
Multiple cohort studies have also been reported. 134 The STOP GAP trial, an RCT carried out in the UK, examined the rapid effects of these two drugs in the initial treatment. 133,135 The results were that cyclosporine was as effective as corticosteroids and was superior in cost-effectiveness. Generally, corticosteroids were the first choice for the treatment of PG, and cyclosporine was the second choice, before this trial. However, these results suggest that cyclosporine can be a choice comparable to corticosteroids. There have been multiple systematic reviews on the treatment for PG, and corticosteroids have been used most frequently, followed by cyclosporine. [136][137][138][139][140] Since these two drugs have their characteristic profiles, it is recommended to select either drug on a case-by-case basis. Although these drugs are effective for the initial treatment for PG, it takes a very long time (6 months) before cure; the development of more effective treatments is awaited. The recommendation grade of cyclosporine is rated as A.
Tacrolimus with a similar action mechanism as cyclosporine, is a This condition is often treated with a combination of corticosteroids and chemotherapy as a treatment for the current disease. 144 The risk of inpatient hospital death is highest among these complications. 145 For these reasons, treatment aiming at controlling the underlying disease is important in PG patients with underlying diseases. If the effectiveness of corticosteroids or cyclosporine is low, the concomitant use of other immunosuppressants or the introduction of biological preparations is recommended. On the other hand, for PG without an underlying disease and PG secondary to breast reconstruction or orthopedic surgery, corticosteroids or cyclosporine, alone or in combination, are recommended as the first choice. 146,147 12.17 | CQ17: Is dapsone useful?

| Recommendation grade: C1
Comment: Dapsone has the action of inhibiting neutrophil migration and suppressing neutrophil bactericidal action and is used for the treatment of PG primarily consisting of neutrophil infiltration. 148 There is a retrospective series of 27 cases of PG (mean age: 61.2 years) treated with dapsone. 149 Complete remission was attained in four (16%), 22 attained partial remission (81%), and one showed no response. An adverse drug reaction was observed in nine (33%) patients, and the administration was suspended in one of them.

| Recommendation grade: C1
Comment: Colchicine prevents microtubule formation and inhibits mitosis by binding with tubulin, which is a major protein of microtubules, and inhibits its polymerization. It produces the inflammation-controlling effect by inhibiting neutrophil migration and phagocytosis and reducing the expression of factors of neutrophil adhesion to blood vessels and cytokines, a mechanism different from that of steroid or immunosuppressants. Colchicine has been reported to be effective as a treatment for PG [150][151][152][153] and has been used at doses of 0.6-1 mg/day. Colchicine is more often reported to be used as an addition to steroid therapy rather than alone. For these reasons, colchicine may be evaluated as an additional treatment of oral steroid therapy for PG.

| Recommendation grade: C1
Comment: Minocycline is a tetracycline antibiotic with a wide spectrum of bacteriostatic action. In addition to the antibacterial action, it is known to inhibit neutrophil migration, suppress the inflammatory cytokine production, and inhibit granuloma formation. There are case reports that PG responded to minocycline used in combination with steroid or immunosuppressants. [154][155][156] It is effective at a dose of 100 mg/day by concomitant use with steroid or immunosuppressants. There is also a report of four cases that recovered in a few weeks of minocycline administration at 200-300 mg/day. 157 For these reasons, minocycline may be evaluated as an addition of systemic steroid or immunosuppressant therapy for PG.

| Recommendation grade: C1
Comment: Potassium iodide has a history of about two centuries as a drug and has been used empirically in the field of dermatology for the treatment of disorders including erythema nodosum and Sweet syndrome as well as PG, without clarification of its pharmacological action or action mechanism. However, since the 1980s, it has been reported to suppress active oxygen by neutrophils and to suppress neutrophil migration, 158,159 providing the grounds for its clinical effects.
Concerning the usefulness of potassium iodide for PG, there are only case reports, and no case series study or RCT has been conducted. [160][161][162] In the case reports, potassium iodide was admin- From the action mechanism of apremilast, PDE4 inhibitors may be effective.

| CQ24: Is systemic administration of antibiotics useful for wound infection in PG?
12 Devices Ag. 14 PGAR100 of target PG ulcers, which was the primary endpoint, was achieved in 12 of 22 patients with active ulcerative PG (p < 0.001 at week 26 of administration). The ulcer area also decreased by 64% (at week 26 of administration). However, no specific adverse event due to the administration of adalimumab for PG has been reported. As a result, adalimumab was approved first in Japan as a drug for PG covered by health insurance in 2020. In the above clinical trial, it was also shown that the response rate to adalimumab was higher in the group that used an oral steroid at baseline than in the group that did not. The guidance for the use of adalimumab mentions suitable candidates, such as those showing resistance to preexisting drugs (e.g., oral corticosteroids and immunosuppressants), those with complications (e.g., inflammatory bowel disease and rheumatoid arthritis), and those in which dose reduction of corticosteroids is difficult. 1 The clinical trial of adalimumab performed in Japan was an open-label study, but as its use for PG is already covered by health insurance, the committee rated its recommendation grade as A, similarly to infliximab described next.
Concerning other anti-TNF drugs, a small-scale, short-term, randomized placebo-controlled trial of infliximab was performed in the UK. 166 The clinical improvement rate 2 weeks after beginning the administration was 6% in the placebo group (n = 17), but was high at 46% in the infliximab group (5 mg/kg, n = 13; p = 0.025). Of the 29 subjects who were administered infliximab after 6 weeks as an openlabel study, 69% (n = 20) showed clinical improvements, and 21% (n = 6) attained remission. The effect of etanercept was examined, although retrospectively, in 11 PG ulcer lesions of seven patients in the USA. 167 Of the 11 skin ulcers in seven patients, eight (73%) were reported to be completely cured after a mean period of 12.5 weeks.
Cases of PG treated with certolizumab pegol (PEGylated anti-TNFα antibody) have also been reported. 168 Other than anti-TNF antibodies/inhibitors, there have been case reports that PG responded to ustekinumab (IL-12/23 p40 antibody), 169  12.26 | CQ26: Are molecularly targeted drugs other than biological preparations useful?

| Recommendation grade: C1
Comment: Reports of the use of Janus kinase (JAK) inhibitors, which have been shown to be effective for the treatment of inflammatory skin diseases such as psoriasis, atopic dermatitis and alopecia areata, are increasing in PG. Orfaly et al. 177 reported that ulcers were completely epithelialized in four of the 10 PG patients by tofacitinib, which is a JAK inhibitor (JAK3/1 inhibitor); that ulcers were also improved in the remaining six patients; and that the treatment was not discontinued due to adverse reactions in any patient. There are reports of successful treatment with baricitinib (JAK1/2 inhibitor). 82 There are no reports concerning the therapeutic effect of external delgocitinib (JAK1/2 inhibitor) or tyrosine kinase 2 (Tyk2) inhibitor on PG (as of July 2021).
In addition, there is a case report that PG responded to apremilast (PDE 4 inhibitor) 164 and alefacept (lymphocyte function associated antigen (LFA) 3 inhibitor). 178 The reports of the effects of these molecularly targeted drugs on PG are based on case reports and no randomized or active comparator-controlled trial has been performed. Also, the use of these molecularly targeted drugs for PG is not covered by health insurance (as of January 2022). Regarding the effects and safety of GMA therapy against PG, there are only case reports and reports and reviews of case series studies; no multicenter joint research, including RCTs, has been made. 185 patients, and responses including partial remission were obtained in 43 (88%). However, no treatment effect was observed in six patients (12%). The mean time until the initial response was 3.5 weeks, and the mean duration of the final administration period was 5.9 months.

12.27
When the high-dose of intravenous Ig therapy was divided into three groups (i.e., <2 g/kg, 2 g/kg, and >2 g/kg), the mean time before the appearance of the initial effect was 4.5, 3.3, and 1.8 weeks respectively, and the response was obtained earlier at a higher dose. However, the final administration period was 3, 7.7, and 1.8 months, respectively.
Concerning adverse reactions, vomiting was observed in six (12%) patients, and headache was observed in two (4%).
There is a paper that summarized 45 PG patients until 2017. 193 Systemic steroid therapy was performed concomitantly in 39 (87%) patients. High-dose intravenous Ig therapy produced remission in 23 patients, whereas remission was achieved in 22. The remission rate by high-dose intravenous Ig therapy was 4.1 times higher in patients with only one skin ulcer compared with those with two or more skin ulcers.

| Recommendation grade: C1
Comment: Recommendation of negative-pressure wound therapy for PG is controversial and the risk of pathergy is considered high, particularly, if it is performed without immunosuppressive therapy. 194 Pichler et al. 195 analyzed the effectiveness of split-thickness skin grafting, negative-pressure wound therapy, and their combination in skin ulcers of 115 PG patients, consisting of their own cases and cases in the literature. They reported that the wound healed in 84 patients (73%) and that none developed pathergy. 195 In this report, the majority of patients (93%) received some systemic immunosuppressive therapy or anti-TNFα antibody administration. 195 From these observations, a combination of split-thickness skin grafting and negative-pressure wound therapy with concomitant systemic immunosuppressive therapy or anti-TNFα antibody administration is weakly recommended.

CO N FLI C T O F I NTE R E S T S TATE M E NT
None declared.