Incidence of malignancy in dermatomyositis vs polymyositis: A systematic review and meta-analysis.

Abstract

e22578 Background: Dermatomyositis (DM) and Polymyositis (PM) are the two major subtypes of idiopathic inflammatory myopathies (IIMs) having different cancer incidences. DM is clinically recognized to have a strong cancer link than PM, but the comparative incidence of malignancy between DM and PM including sex stratified patterns, trends of malignancies, and tumor specific risk remains unquantified. We conducted a meta-analysis to evaluate incidence of malignancies in DM and PM. Methods: We performed PRISMA based searches across three databases (PubMed, Google Scholar, and ScienceDirect) for retrospective studies which reported malignancies in adult DM and PM patients after myositis onset were included. Due to variable follow up durations, malignancies were pooled as post diagnosis events. Trend analyses were performed to determine whether the DM–PM malignancy association was consistent overall and across subgroups stratified by sex and cancer type. We applied random effects model and estimated Pooled Risk Ratios with 95% confidence intervals. Subgroup analyses evaluating sex specific risk, cancer spectrum, and trends relative to myositis cohorts was performed. We assessed statistical heterogeneity using I² statistics. Results: Twelve studies which included 10,870 patients were analyzed in this meta-analysis which met the eligibility criteria. The primary outcome, comprised of 9 studies (10,230 patients; 5,063 DM and 5,167 PM patients), was incidence of malignancy which was diagnosed after myositis onset during follow up. DM showed higher malignancy incidence than PM (RR 1.79; 95% CI 1.54–2.08; I²=21%; p<0.000001). The trend analysis showed consistent greater malignancy in DM across overall and in subgroups than PM, but some cancer types (e.g. prostate and colorectal cancer) favored PM, and some (lung and ovarian cancer) favored DM while non-Hodgkin lymphoma showed equal trend. Gender specific analysis showed male gender to have higher incidence in DM and modestly higher in PM than females. In direct subtype comparison, DM carried higher incidence in both sexes. However, tumor specific cancer incidence differed :lung cancer (RR 3.14; 95% CI 1.94–5.08; I²=0%), ovarian cancer (RR 3.86; 95% CI 1.36–10.94; I²=7%), prostate cancer (RR 0.86; 95% CI 0.27–2.75; I²=19%), non-Hodgkin lymphoma (RR 1.06; 95% CI 0.38–2.93; I²=0%) , and colorectal cancer (RR 1.57; 95% CI 0.58–4.25; I²=53%). Conclusions: It concluded that DM has higher incidence of malignancy than PM overall and across all subgroup analyses. PM patients had increased risk of malignancy than general population but lower than DM. Trend analyses showed that malignancy remained consistent in DM, but some cancer types favored DM and some PM. These results provide a clear and quantitative overview about subtype, sex specific, tumor specific, and malignancy patterns showing an overall distribution of cancer incidence in myositis.