Background Individuals with rheumatoid arthritis (RA) receiving anti-tumour necrosis factor-α (anti-TNF) therapy may have an inadequate response (IR) to first-line treatment. Among those patients for whom treatment fails, little is known about the associated costs of treatment or the subsequent proportion with an IR to a second-line anti-TNF treatment.
Objectives This study quantified the RA-related healthcare cost of treating patients who respond inadequately to anti-TNF therapy using an adaptation of a published administrative claims-based algorithm to identify IR. We examined the probability of IR to first- and second-line anti-TNF biologics, and assessed RA-related healthcare costs among patients with an IR to first- or second-line treatments.
Methods This was a retrospective, observational cohort study based on administrative claims data. Adult (aged ≥18 years) patients with RA initiating anti-TNF therapy between 1 January 2009 and 31 December 2013 were included in the analysis. Patients were required to have continuous insurance enrolment for 12 months before initiating anti-TNF therapy, and no treatment with any biologics during this period. Patients who experienced an IR to first-line anti-TNF therapy and were treated with second-line anti-TNF therapy were analysed as a sub-sample. Variable-length follow-up was measured for both the first- and second-line treatments, beginning with initiation of an anti-TNF and continuing until one of the following events: use of a biologic that is different from the initiated anti-TNF, disenrolment from health insurance or reaching the study end date of 31 December 2013. IR was defined as a composite of discontinuation or switch of the initiated anti-TNF, anti-TNF dose escalation, initiation of a new non-biologic DMARD, new/increased oral glucocorticoid (OGC) use/dose or receiving ≥2 injections of glucocorticoid (IGC). Kaplan–Meier analysis was used to calculate probability of IR over time. RA-related healthcare utilization and costs were expressed in per-patient per-month (PPPM) units and corresponded to medical claims with an International Classification of Diseases, Ninth Revision, Clinical Modification diagnosis code for RA, the initiated anti-TNF and non-biologic DMARDs.
Results A total of 26,785 patients met our inclusion criteria for the first-line analysis; 4638 for the second-line analysis. Mean follow-up time was 525 days for first line and 383 days for second line; Kaplan–Meier-estimated probabilities of IR were 67% at 6 months, 82% at 12 months and 91% at 24 months for first line; 77%, 89% and 95% at 6, 12 and 24 months for second line. In both lines, the primary reason for IR was discontinuation of initiated anti-TNF (52–61% of patients), followed by: 2+ IGC (30–33%); switch of initiated anti-TNF (23–34%); initiation of a new non-biologic DMARD (14–27%); anti-TNF dose escalation (15–19%); new/increased OGC use/dose (13–15%). Average RA-related PPPM costs were $3883 (SD $10,697) for patients with an IR to first-line therapy and $6831 (SD $32,849) for patients with an IR to second-line therapy.
Conclusions In this retrospective study using real-world data, a high proportion of patients with RA experienced an event indicative of IR to first- or second-line anti-TNF treatment. The RA-related healthcare costs of treating patients with IR were substantial.
Disclosure of Interest E. Thomson Grant/research support from: Bristol-Myers Squibb, Employee of: Truven Health Analytics, A. Nadkarni Shareholder of: Bristol-Myers Squibb, Employee of: Bristol-Myers Squibb, M. Brouillette Grant/research support from: Bristol-Myers Squibb, Employee of: Truven Health Analytics, S. Johnston Consultant for: Bristol-Myers Squibb, Employee of: Truven Health Analytics