LAUSR

Dear Editor, Pyoderma gangrenosum (PG) is an ulcerative and neutrophilic dermatosis with poorly understood pathogenesis. Almost half of patients with PG have an underlying inflammatory disease [1]. The frequency of concomitant malignant disease in PG patients has been estimated at 4.5–9% [1, 2], and there have been increasing reports of PG in patients with an underlying solid malignancy [2] leading us to further characterize this association. A systematic review of PubMed, Web of Science, and the Cochrane database with the search terms “Pyoderma gangrenosum” AND “cancer” OR “malignancy” OR “tumor” OR “leukemia” OR “lymphoma” led to a total of 876 publications retrieved. Cases of PG in cancer patients that were pathergy induced (i.e., after surgical excision of the tumor), large-scale studies that did not provide patient details, and cases with concurrent solid and hematologic malignancy were excluded. A total of 186 patient cases were identified from the literature search. Eighty-two percent of patients had an underlying hematologic malignancy and 18% a solid malignancy (Table 1). In 71% of cases with solid tumors, PG presented after the diagnosis of malignancy. This was slightly decreased in patients with hematologic malignancy (54%); however, this difference was not statistically significant (Table 2). As expected, reported mortality was significantly higher in patients with hematologic malignancy versus solid malignancy (16% versus 41%, chi square test p < 0.05). Mortality was further analyzed using a univariate logistic regression model which determined that patients with leukemia and PG have a higher likelihood of death compared with PG patients with non-leukemia hematologic malignancies (p < 0.05, 95% CI = 1.01–3.34). A multivariate logistic regression did not identify any increase in mortality when adjusting for gender, size of lesion, or age. Lower extremity was the most common site of PG in both types of malignancy. Interestingly, there was a higher number of reported cases of PG of the head and neck in patients with hematologic malignancy (n = 16) compared with patients with solid malignancy (n = 1). PG was mostly treated with oral and topical corticosteroids. Most studies did not report the effect of removal of the underlying malignancy on PG; however, numerous cases resolved with active malignancy. Our systematic review has many limitations including reporting bias and differing reporting periods. However, this report still highlights the necessity of future research into the association between PG and solid malignancies. A recent meta-analysis has noted that the pooled prevalence of solid malignancies is surprisingly high at 7.4%, and our study provides more insight into the specific clinical characteristics of these patients [3]. Many patients with advanced cancer show high levels of blood neutrophilia [4, 5]. This underlying neutrophilic state may predispose these individuals to neutrophilic dermatoses such as PG. IL-1 is also highly produced by several solid tumors [6], which is a cytokine that is overexpressed in PG skin lesions [7]. * Alex G. Ortega-Loayza [email protected]