Treatment Trends Among Patients with Rheumatoid Arthritis Who Discontinued Initial TNFi

In a large cohort of commercially insured patients with RA, switching to another TNFi was more common than initiating a non-TNFi. However, patients who swapped to a non-TNFi remained on treatment longer than patients who received a second TNFi. These results were published in the journal Arthritis Care & Research.

Investigators extracted claims data from the Truven Health MarketScan Research database, which comprises the healthcare claims of commercially insured individuals in the United States. Patients with RA were identified using the relevant diagnostic codes. Patients who switched from an initial TNFi to a new drug between 2008 and 2015 were eligible for inclusion. Baseline and clinical characteristics were compared between patients who swapped to another TNFi and patients who switched to a non-TNFi. Six-month healthcare costs were calculated using individual claims for each drug. Survival time—defined as the time between initiation and discontinuation—was computed for each drug. Cox proportional hazards models were used to identify predictors of drug survival.

The study cohort comprised of 10,442 patients, among whom 63.5% cycled to a new TNFi and 36.5% swapped to a non-TNFi. Mean follow-up time was 2.91 years. Patients who swapped to a non-TNFi agent were significantly older and had more comorbidities than patients who cycled to another TNFi (both P <.001). Etanercept and adalimumab were the most commonly used initial TNFi, accounting for 43.6% and 31.7% of initial prescriptions, respectively. These 2 drugs were also the most common agents in patients who cycled to another TNFi. Among patients who switched to a non-TNFi, abatacept was the most prevalent second-line option. Treatment continuation rate was lower in patients who initiated another TNFi compared with patients who swapped to a non-TNFi (33.4% vs 39.6%; P <.001). The number of patients who switched to a third agent was also higher in the TNFi vs non-TNFi group (52.5% vs 45.9%; P <.001). Estimated mean drug survival time was 489 days in patients who started another TNFi and 605 days in patients who started a non-TNFi. In Cox analyses, patients who swapped to a non-TNFi were at significantly decreased risk of switching to a third agent (hazard ratio [HR], 0.82-0.91; 95% confidence interval [CI], 0.82-0.91; P <.001). Greater comorbidity predicted higher likelihood of switching to a third treatment option (HR, 1.16; 95% CI, 1.05-1.29; P =.005). Mean healthcare costs were generally lower for patients who cycled to another TNFi than for patients who started a drug with another mechanism of action. 

Study limitations include that the MarketScan database does not contain information on disease activity, disease severity, or treatment response.

While patients with RA more often swapped from one TNFi to another, switching to a non-TNFi was associated with greater treatment persistence. However, TNFi cycling was more cost-effective. Further analysis incorporating patient-specific characteristics is necessary to confirm these findings. 

Disclosure: Several study authors declared affiliations with the pharmaceutical industry. Please see the original reference for a full list of authors’ disclosures.

Reference

Karpes Matusevich AR, Duan Z, Zhao H, et al. Treatment sequences after discontinuing a tumor necrosis factor inhibitor in patients with rheumatoid arthritis. A comparison of cycling versus swapping strategies [published online June 17, 2020]. Arthritis Care Res (Hoboken). doi:10.1002/acr.24358