IL-6Ri vs csIM in Polymyalgia Rheumatica
Key Takeaways
- Among older adults with PMR in matched US Medicare claims cohorts, IL-6Ri initiation was associated with significantly more year-1 GC discontinuation than csIM initiation.
- Minimal GC use, defined as prednisone-equivalent dosing of 2 mg/day or less or stopping GCs, was also achieved significantly more often with IL-6Ri therapy than with csIM therapy.
- In the methotrexate subgroup, minimal GC use favored IL-6Ri therapy, while year-1 GC discontinuation versus methotrexate was not significantly different overall.
- Primary hospitalized infections were higher in year 1 with IL-6Ri therapy, whereas year-2 rates and other adverse-event patterns were broadly similar.
In the study, investigators used US Medicare fee-for-service medical and Part D claims data to build a retrospective comparative cohort of patients older than 50 years with polymyalgia rheumatica (PMR), identified by 1 inpatient or 2 outpatient claims at least 30 days apart, who were already receiving glucocorticoids (GCs), had no prior interleukin-6 receptor inhibitor (IL-6Ri) exposure, and started tocilizumab or sarilumab versus methotrexate, leflunomide, or azathioprine. Direct matching was followed by 1:1 propensity score matching, yielding 415 matched treatment pairs, including 187 conventional synthetic immunomodulator (csIM)-naive pairs and 228 csIM-experienced pairs. Effectiveness was followed through year 1 and safety through year 2, with primary endpoints of time to GC discontinuation and time to minimal GC use, defined as a prednisone-equivalent dose of 2 mg/day or less or stopping GCs.
Year-1 steroid-sparing outcomes favored IL-6Ri therapy over csIM therapy in the pooled matched analysis. The adjusted hazard ratio for time to GC discontinuation was 1.28 (95% CI 1.02 to 1.60; p=0.031), and the adjusted hazard ratio for time to minimal GC use was 1.28 (95% CI 1.03 to 1.58; p=0.025). Both the csIM-naive and csIM-experienced matched cohorts showed the same direction of benefit. In the methotrexate subgroup, the hazard ratio for time to minimal GC use was 1.41 (95% CI 1.05 to 1.89; p=0.022), while GC discontinuation versus methotrexate did not reach significance overall.
Safety findings were more mixed. In hospitalised infection outcomes in polymyalgia rheumatica, primary incidence rates were 12.2 versus 5.7 per 100 patient-years in year 1 and 5.8 versus 6.4 per 100 patient-years in year 2 for IL-6Ri versus csIM therapy. Drug-induced liver injury, gastrointestinal perforation, and broader adverse-event categories were otherwise low or not significantly different over 2 years, while hospitalized infections in any diagnosis position were more frequent during year 1 with IL-6Ri. Glucocorticoid dose reduction over time also generally trended in the same direction as the primary effectiveness findings.
Interpretation remains bounded by a nonrandomized, claims-based comparison in which GC discontinuation and minimal GC use were surrogate treatment outcomes derived from prescription claims rather than direct remission measures. Residual confounding by disease severity remains plausible because IL-6Ri use for PMR during most of the study period was off-label. The authors also noted a small methotrexate subgroup in csIM-experienced disease, differing prior exposure patterns, possible coding-related confounding from seronegative rheumatoid arthritis, and generalizability centered on a Medicare-age population in whom IL-6Ri exposure mainly reflected tocilizumab. Findings were directionally similar when patients with prior giant cell arteritis were excluded.
Clinician Questions
Which patients with polymyalgia rheumatica were represented in the Medicare comparison of IL-6 receptor inhibitors and csIM therapy?
Adults older than 50 years in US Medicare fee-for-service and Part D data were included if PMR was identified by at least 1 inpatient or 2 outpatient claims 30 days apart, patients were already receiving glucocorticoids, had no prior IL-6 receptor inhibitor exposure, and had an index prednisone-equivalent dose of 25 mg/day or less. The analysis was separated into csIM-naive and csIM-experienced cohorts. Key exclusions included seropositive rheumatoid arthritis, other inflammatory arthritis or connective tissue disease, organ transplant, multiple sclerosis, active malignancy treatment, and same-day IL-6Ri plus csIM initiation.
What events ended effectiveness follow-up in the PMR claims analysis of IL-6 receptor inhibitors?
Effectiveness follow-up lasted up to 1 year after treatment initiation and ended at the earliest of loss of Medicare coverage, a treatment gap longer than 60 days, switching to or adding another PMR treatment, death, or day 365.
Why was the methotrexate subgroup harder to interpret in csIM-experienced polymyalgia rheumatica?
In csIM-experienced PMR, the matched methotrexate subgroup included 41 pairs, and prior drug exposure was uneven across comparison groups: most IL-6Ri initiators had previously received methotrexate, whereas most methotrexate initiators had previously received leflunomide. The authors said those differences, along with possible selection bias and coding-related confounding, could make subgroup estimates less stable.
How far do these IL-6 receptor inhibitor findings in polymyalgia rheumatica extend beyond Medicare-age patients?
The analysis mainly reflects a Medicare-age PMR population, with most patients aged 65 years or older, and IL-6Ri exposure in the dataset was largely tocilizumab. According to the authors, younger patients with PMR were underrepresented, so the findings are most directly applicable to older US claims-based populations rather than to all age groups.
Recommended Reading
- For more on steroid-sparing vasculitis treatment: Upadacitinib in Giant Cell Arteritis: Phase 3 Trial Results