Survival in patients with rheumatoid arthritis and recently diagnosed early-stage colorectal, lung, or prostate cancer receiving tumour necrosis factor inhibitors: a retrospective cohort study.

backgroundTumor necrosis factor (TNF) inhibitors could impair tumoural immunity in patients with rheumatoid arthritis and cancer. We aimed to investigate the association between survival and TNF inhibitor treatment during the first 3 years after a diagnosis of colorectal, lung, or prostate cancer in patients with rheumatoid arthritis.

methodsIn this cohort study, we conducted a secondary data analysis of the Surveillance, Epidemiology, and End Results Medicare-linked dataset. We included patients aged 66 years and older with rheumatoid arthritis diagnosed with colorectal, lung, or prostate cancer between Jan 1, 2008, and Dec 31, 2019, using ICD-O-3 site and histology codes. We limited the cohort to patients who had early-stage cancer (localised or regional). We only included patients who received TNF inhibitors, conventional synthetic disease-modifying antirheumatic drugs (DMARDs), or no DMARDs in the first year after cancer diagnosis. The primary outcomes were 5-year overall survival and cancer-specific survival. Exposures were use of TNF inhibitors, conventional synthetic DMARDs, or no DMARDs within 3 years after cancer diagnosis. Other covariates included demographics and comorbidities. We conducted landmark analyses at years 1, 2, and 3, with Cox regression adjusted by propensity scores. People with lived experience of rheumatoid arthritis and cancer were not involved in the design or conduct of this study.

findingsWe identified three cohorts of patients diagnosed with early-stage colorectal cancer (n=514), lung cancer (n=864), or prostate cancer (n=603) between Jan 1, 2008, and Dec 31, 2019. In the colorectal cancer cohort, the mean age was 76·1 years (SD 6·4), 385 (75%) of 514 patients were female, 129 (25%) were male, and 405 (79%) were White and non-Hispanic. In the lung cancer cohort, the mean age was 74·8 years (SD 5·9), 632 (73%) of 864 patients were female, 232 (27%) were male, and 743 (86%) were White and non-Hispanic. In the prostate cancer cohort, the mean age was 73·1 years (SD 5·1), 603 (100%) patients were male, and 492 (82%) were White and non-Hispanic. 80 (16%) of 514 patients with colorectal cancer, 102 (12%) of 864 patients with lung cancer, and 120 (20%) of 603 patients with prostate cancer received TNF inhibitors with or without conventional synthetic DMARDs at any time during the first year after cancer diagnosis. No significant deleterious association was observed for overall survival or cancer-specific survival for any of the cancers at any of the three landmark points. Hazard ratios and 95% CIs for overall survival for year 1 comparing TNF inhibitors with conventional synthetic DMARDs in the three cohorts were 0·72 (0·43-1·21) for colorectal cancer, 0·70 (0·49-1·00) for lung cancer, and 0·80 (0·44-1·44) for prostate cancer. Patients who received glucocorticoids in the first year had significantly worse overall survival and cancer-specific survival than those who did not in the multivariable Cox proportional hazards models for the three cancers.

interpretationTreatment with TNF inhibitors during the first 3 years after diagnosis of colorectal, lung, or prostate cancer was not associated with poorer survival compared with those who received conventional synthetic DMARDs or those who did not receive any DMARDs. However, these findings might not be generalisable to other populations or types of cancer.

fundingNational Institutes of Health (NIH)/National Institute of Arthritis and Musculoskeletal and Skin Diseases and NIH/National Cancer Institute through MD Andeson's Cancer Center Support Grant and Komen.