OBJECTIVEOlder age, dermatomyositis, and specific serum autoantibodies such as anti-TIF1-γ are associated with higher cancer risk in patients with myositis. We evaluated a vast cohort of patients with myositis for the prevalence of cancer, the association to disease features, and the performance of the recent IMACS guidelines.METHODSA retrospective cohort analysis was performed and in all cases serum autoantibodies were tested using HEp-2, immunoassays, RNA- and protein-immunoprecipitation. Myositis was defined as cancer-associated if malignancy occurred within 3 years prior to or after the onset of myositis.RESULTSNinety-five patients with IIM were followed-up for a median of 6 years (interquartile range 3-11), the majority were classified as 'high-risk' or 'intermediate-risk' of cancer based on IMACS guidelines. A diagnosis of cancer was made in 22/95 (23 %) of cases and, based on the timing of the diagnosis, 14 % patients were cancer-associated myositis, with no significant differences compared to patients without cancer. Both groups of patients with overall cancer and cancer-associated myositis had more respiratory comorbidities, anemia, and hypergammaglobulinemia, and dermatomyositis phenotype. Anti-TIF1-γ antibody positivity predicted cancer-associated myositis but not the overall cancer rate; malignancy was observed in particular in patients with isolated anti-TIF1-γ antibodies, while a lower prevalence occurred in case of additional specificities identified by immunoprecipitation.CONCLUSIONSRecent IMACS guidelines perform well in the interception of cancer, yet adjunctive history and laboratory features should be considered. Patients with anti-TIF1-γ antibodies are at risk of cancer-associated myositis, but concurrent autoantibodies negatively correlate with malignancy and warrant characterization.