Objectives: The aim was to compare standardized incidence ratios (SIRs) of cancers temporally related and unrelated to active myositis in patients with myositis.
Methods: Fifty-two cancer cases were identified in 281 myositis patients. SIRs of cancers having temporal overlap with the active phase of myositis [cancers concurrent with active myositis (CAM), n = 30] and cancers not having such temporal overlap [cancers non-concurrent with active myositis (CNM), n = 22] were compared in 281 patients.
Results: Patients with CAM were older at diagnosis of myositis, had a greater tendency to be male, more frequent dysphagia and less frequent interstitial lung disease than patients with CNM. CAM SIR (95% CI) was 1.78 (1.19, 2.56) and CNM SIR 1.23 (0.75, 1.90). The peak SIR was observed in the seventh decade of life for CAM and in the third decade for CNM. When stratified by myositis-cancer intervals, CAM SIR was 9.94 (6.43, 14.67) within 1 year of myositis diagnosis, whereas no temporal relationship was found for CNM. Elevated SIRs were observed for oesophageal cancer [57.77 (11.91, 168.82)], non-Hodgkin's lymphoma [41.43 (13.45, 96.69)], adenocarcinoma of unknown primary origin [67.6 (18.42, 173.07]), lung cancer [7.27 (1.98, 18.61)] and ovarian cancer [19.15 (2.32, 69.17)] within 3 years of CAM diagnosis. The cancer stage at the time of diagnosis was more advanced in CAM than CNM (P < 0.001), with a correspondingly increased hazard ratio of mortality [4.3 (1.5, 12.7)] in patients with CAM vs CNM.
Conclusion: A significantly elevated SIR was found for CAM, whereas there was a comparable SIR for CNM relative to the general population. Multiple types of cancers showed elevated SIRs among CAM, but none among CNM. Given that cancer stages in CAM were far advanced at diagnosis, mortality risk was greater in patients with CAM.
Keywords: cancer; cancer stage; myositis; standardized incidence ratio.
© The Author 2016. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oup.com.