Newly approved DMARDs can appear less safe due to ‘channeling’ phenomenon
Key takeaways:
- Among patients with rheumatoid arthritis, observed rates of major adverse events from biologic/targeted synthetic DMARDs fell by about 25% in the years after market entry.
- This could be due to the drugs initially being prescribed for more severe disease.
Adverse events from disease-modifying antirheumatic drugs for the treatment of rheumatoid arthritis decline by as much as 25% after the drugs are established on the market, according to data published in Annals of the Rheumatic Diseases.
“Currently, there are safety concerns regarding treatment with Janus kinase inhibitors for patients with RA that limit their clinical use,” Viktor Molander, MD, PhD, a rheumatologist at Karolinska University Hospital, in Sweden, told Healio. “However, since patients with RA are at increased risk for these potential side effects — VTE, cardiovascular events and cancer — due to RA itself, it is not easy to confirm whether such observations are due to actual side effects or a consequence of RA.
“When a drug class is new on the market, it is more likely to be prescribed to those with long-lasting and difficult-to-treat RA,” he added. “We hypothesized that such potential channeling could make drugs appear more dangerous when the drug is new on the market compared to when more established.”
To investigate this effect, Molander and colleagues conducted a nationwide cohort study using seven different Swedish data sources, including registries of rheumatology patients, inpatient and outpatient hospital discharges, prescribed drugs and causes of death. The researchers used these sources to create three cohorts:
- patients with RA who initiated biologic or targeted synthetic DMARDs between January 2006 through December 2021 (n = 33,550);
- an “early bionaïve” cohort that started methotrexate within 2 years of RA diagnosis (n = 16,011); and
- a matched general population cohort (111,074).
The primary outcome was the first instance of a major adverse cardiovascular event, venous thromboembolism, cancer or infections resulting in hospitalization.
Biological and targeted synthetic DMARDs were split into five classes: TNF inhibitors, interleukin (IL)-6 inhibitors, JAK inhibitors, rituximab (Rituxan, Genentech) and abatacept (Orencia, Bristol Myers Squibb). In each class, outcomes were analyzed among patients initiating treatment fewer than 2 years since the class’s market introduction in Sweden, 2 to 5 years afterward, and greater than 5 years afterward. These risks were expressed as incidence rates and hazard ratios, calculated using Cox proportional hazards regression models and adjusted for patient characteristics.
A total of 5,862 events of interest were recorded among those who initiated DMARDs.
Overall, compared with patients who initiated treatment within 2 years of a class’s market entry, those who initiated 5 or more years afterward demonstrated lower outcome rates in an unadjusted model (HR = 0.74; 95% CI, 0.67-0.81), according to the researchers.
However, the difference practically “disappeared” (HR = 0.93; 95% CI, 0.84-1.03) in a model fully adjusted for RA disease variables, comorbidities, medications and socioeconomics, further “suggesting that channeling is behind this observation,” the researchers wrote.
Risk for fulfilling the composite outcome diminished over time. Compared with the period of 2006 to 2010, the hazard ratios were 0.87 (95% CI, 0.82-0.93) for 2011 to 2015, and 0.74 (95% CI, 0.69-0.80) for 2016 to 2021 in the fully adjusted model.
Meanwhile, the composite outcome rate “remained stable” in both the early bionaïve and general population cohorts throughout the study, the researchers wrote.
This could possibly be because patients with RA who receive biologic and targeted synthetic DMARDs today are “on average healthier compared to previously,” Molander said.
The result suggests that “the bar defining ‘acceptable’ rates of safety outcomes of new and future antirheumatic drugs should be lower than that used for past approval of existing b/tsDMARDs,” the researchers wrote.
“The observed rates of these four safety outcomes were around 25% higher during the first year after drug class market entry, compared to thereafter, which could be readily explained by changing patient characteristics, or channeling,” Molander said. “This is important to keep in mind when interpreting safety data from observational studies, especially when such results do not take all relevant patient characteristics into account.”
For more information:
Viktor Molander, MD, PhD, can be reached at viktor.molander@ki.se.
Collapse