Background/Purpose: There is a dearth of knowledge around the safety of biologic disease modifying antirheumatic drugs (bDMARDS) in immune checkpoint inhibitor (ICI)-treated rheumatoid arthritis (RA) patients with cancer. The goal of this study was to quantify the association between bDMARD use after ICI initiation and overall survival (OS) in RA patients with metastatic non-small cell lung cancer (mNSCLC), melanoma and urothelial cancer.

Methods: We used a curated Medicare claims dataset that consists of a 100% sample of patients with RA. We included patients ≥ 66 years of age with both a diagnosis of RA and of mNSCLC, melanoma or urothelial cancer, who initiated nivolumab, pembrolizumab or atezolizumab 2015-2019 and who received an intravenous (IV) bDMARD at some time during the year prior to ICI initiation. ICI and DMARD use were identified using J and C codes. We limited the study to IV bDMARDs because the timing of administration can be accurately ascertained. Patients with mNSCLC, melanoma and urothelial cancers were required to have two claims associated with the relevant ICD-9-CM or ICD-10-CM codes. RA was defined as having two claims associated with an ICD-9-CM or ICD-10-CM diagnosis code for RA. Patients who took a specific bDMARD after ICI initiation were compared to patients who took the same bDMARD in the year prior, but not in the year after ICI initiation. Kaplan Meier (KM) curves and cox proportional hazard models (CPHM) (un-adjusted, age and ICI adjusted) were created to measure OS from first ICI initiation. Patients were followed through 12/31/2019 and were censored at time of death or last recorded visit in the database.

Results: A total of 6350 ICI-treated RA patients with mNSCLC, melanoma, and urothelial cancer were identified. Of these, only 466 (7.3%) received an IV bDMARD in the year prior to ICI initiation. 166/466 (35.6 %) took the same DMARD in the 1 year after ICI initiation. Overall median follow up was 186.5 days (IQR 76, 355). Patients who took a bDMARD prior to ICI only were more likely to be female than those who continued/initiated that bDMARD after ICI initiation (63.3% vs 51.2%, p= 0.011) (Table1). Figure 1 displays the KM curves comparing patients who continued/initiated a TNFi after ICI initiation to those who took a TNFi only prior to ICI initiation (1A, log rank p-value 0.02) and a similar analysis for Rituximab (1B; p-value 0.06). In CPHM for mNSCLC, patients who took an IV TNFi only prior to ICI initiation had worse OS than those who continued/initiated a TNFi after ICI initiation (Table 2). Adjusted models for other cancers and other IV DMARDS were not statistically significant.

Conclusion: Few RA patients with mNSCLC, melanoma, and urothelial cancer use IV bDMARDS in the year prior to ICI initiation, even among those who were previously receiving them before their cancer diagnosis. Patients who continued/initiated an IV TNFi after ICI initiation appear to have better OS than patients who take a TNFi only prior to ICI initiation. However, this may reflect unmeasured differences between these two groups, such as the use of concomitant chemotherapy, number of prior cancer treatments or comorbidities, which we will address in further analyses.

« Back to ACR Convergence 2023

ACR Meeting Abstracts - https://acrabstracts.org/abstract/biologic-disease-modifying-antirheumatic-drug-use-in-immune-checkpoint-inhibitor-treated-cancer-patients-with-rheumatoid-arthritis-utilization-and-overall-survival/