Hundreds of inconsistencies in the nomenclature of pyoderma gangrenosum: The need to harmonise terminology for phenotypes

Pyoderma gangrenosum (PG) is a rare, ulcerative neutrophilic dermatosis mainly affecting the lower extremities.1, 2 Patients present with differing phenotypes that lead to diagnostic and therapeutic challenges.3 Phenotypic descriptions of PG in the literature are ambiguous, with terminology such as ‘widespread’, ‘disseminated’, ‘extensive’, ‘multifocal’ and ‘generalised’ being inconsistently applied to varied clinical scenarios. In this study, we evaluate the heterogeneous nomenclature and clinical features of PG in case studies and propose a framework that can be utilised to streamline the categorisation of this condition. To identify previously published cases, we searched studies published from 1947 to 2022 in PubMed using the following string: ‘pyoderma gangrenosum case’. The search yielded 2714 total results, which were categorised into single-ulcer and multi-ulcer (3+ ulcers in separate locations) case reports. Studies that used the terms ‘widespread’, ‘disseminated’, ‘multifocal’, ‘extensive’, or ‘generalised’ were included. Cases with ulcer recurrence in the same or different anatomical area were excluded. For studies that did not state the number of ulcers, eligibility was determined based on patient photos and descriptions. We defined the distinct anatomical locations as lower extremities, upper extremities, trunk, head/neck and genitals. Chi-squared tests were used to compare the two categories of single versus multi-ulcer patients with statistical significance established as p ≤ 0.05. For the multi-ulcer data, the category of ‘multifocal’ was analysed independently due to a markedly lower average ulcer count (Figure 1). ‘Extensive’ and ‘widespread’ were combined, and ‘generalised’ and ‘disseminated’ were combined due to similarities in average number of ulcers. Four hundred twenty-nine single-ulcer cases and 204 multi-ulcer cases were identified, of which 22 studies used the descriptor ‘extensive’, 23 as ‘widespread’, 8 as ‘multifocal’, 7 as ‘disseminated’ and 6 as ‘generalised’. One hundred forty-six cases had no description used and were excluded. The clinical presentation and systemic involvement were distinct between single-ulcer and multi-ulcer groups. Patients with skin ulcers in more than three anatomical regions were younger at presentation than patients with a single ulcer with a mean age of 40.4 (SD ±20.8) versus 51.7 (SD ±17.7, p < 0.0001). Multi-ulcer patients were also found to have more severe features of disease such as extracutaneous manifestations and hematologic malignancy (Table 1). These findings build on the literature suggesting there is an association between PG and haematologic malignancies.4 In patients with a single ulcer, the most common anatomical locations were the lower extremities (48.7%) and trunk including the buttocks (31.5%). In contrast, patients with multi-ulcers had lesions mostly distributed in the lower extremities (85.5%), upper extremities (69.5%) and trunk including the buttocks (64.7%) (Figure 1). Multi-ulcer patients also had more lesions found on the upper extremities (p < 0.0001) and head/neck (p < 0.0001) compared with single-ulcer patients. Of the patients, 0.9% with a single ulcer had concomitant erosion of the oral mucosa and none had lung involvement, in contrast to 9.4% affecting the oral mucosa and 3.8% affecting the lungs of extensive/widespread patients, 7.7% affecting the oral mucosa and 7.7% affecting the lungs of generalised/disseminated patients, and no cases of either among multifocal patients (Table 1, p < 0.05). In comparison with single-ulcer patients, multi-ulcer patients had significantly more cases (p < 0.05) of extracutaneous manifestations including involvement of PG in the oral mucosa, lung, spleen, liver, pancreas, bowels, kidney and bone (Table 1). Together, these findings highlight epidemiologic and clinical differences between single-ulcer PG and multi-ulcer PG described as ‘extensive’, ‘widespread’, ‘multifocal’, ‘disseminated’ and ‘generalised’. While the literature has largely used this terminology interchangeably, patients among these groups differ in age of onset, number and distribution of ulcers, extracutaneous manifestations, and underlying comorbidities. For example, despite both falling into the category of ‘multi-ulcer’, the patients who presented with PG described as ‘generalised/disseminated’ had an average ulcer count of 9.5, whereas patients with ‘multifocal’ PG only had an average ulcer count of 3.6. The implications of phenotypical presentations on underlying disease pathogenesis, response to therapy and prognosis is still yet to be elucidated.5 As such, future directions will require the lexicon for describing PG to be standardised to characterise patients who present with multiple skin ulcers. Future work should establish consensus around categorisation of multi-ulcer patients based on comorbidities, presence of extracutaneous manifestations, number of lesions and anatomic location of ulcers. A standardised nomenclature will enable consistent pathophysiologic and epidemiologic evaluation of variations of this rare disorder. This article has no funding source. Dr. Mostaghimi reports no conflicts of interest relevant to this manuscript. He has received consulting fees from Pfizer, hims, Digital Diagnostics, Concert, Lilly, Abbvie, Equillium, Boehringer Ingelheim, LEO and ACOM. Dr. Ortega-Loayza has served on advisory boards for Janssen, BMS and Boehringer Ingelheim, and as a consultant for Genentech and Guidepoint. He has also received research grants from Eli Lilly Company, OHSU School of Medicine Gerlinger research award and Medical Research Foundation of Oregon.