Golimumab for the treatment of ankylosing spondylitis: a NICE single technology appraisal.

As part of the National Institute for Health and Clinical Excellence (NICE) single technology appraisal (STA) process, the Evidence Review Group (ERG) produced a report to comment on the clinical and cost effectiveness of golimumab (Simponi(®), Merck Sharp & Dohme) for the treatment of ankylosing spondylitis (AS) relative to other comparators as presented in the manufacturer's submission (MS) to NICE. The population was those with active disease who had not responded to conventional therapy. The specified comparators were conventional care and two other tumour necrosis factor alpha (TNF-α) inhibitors (adalimumab and etanercept). Outcomes to be considered were disease activity, functional capacity, disease progression, adverse effects of treatment and health-related quality of life (HR-QOL). There were no head-to-head trials comparing TNF-α inhibitors. The submission included one trial of golimumab versus placebo (the GO-RAISE trial) and additionally seven placebo-controlled randomized controlled trials (RCTs) of other TNF-α inhibitor agents (five with etanercept, and two with adalimumab). The results of these trials were generally a statistically significant improvement from each of the TNF-α inhibitors. A Bayesian mixed treatment comparison (MTC) showed there was generally overlap in the 95 % credible intervals (CrIs) between the TNF-α inhibitors. Exceptions included a greater risk of discontinuation of treatment for golimumab than for etanercept (relative risk [RR] 4.30; 95 % CrI 1.01-18.50). The cost-effectiveness analysis (CEA) compared all of these TNF-α inhibitors. Relative effectiveness was informed only by RR of response (proportion achieving at least a 50 % improvement in Bath AS Disease Activity Index [BASDAI] score; BASDAI50) from the MTC. In the base-case analysis, the incremental cost-effectiveness ratio (ICER) of golimumab versus conventional care was £26,597 and adalimumab and etanercept were extendedly dominated by golimumab. The manufacturer concluded that golimumab is a cost-effective treatment option. Generally, the ERG agreed with the MTC analyses. The main problem was that the MS used data from one trial, which included a period of cross-over. The ERG found some problems with the CEA model, mainly that it did not allow for comparison of TNF-α inhibitor sequences and did not use MTC estimates for treatment discontinuation or adverse events (AEs). The ERG could not correct the sequencing problem, but re-ran the CEA with discontinuations and AEs estimated from the MTC and using the correct trial data. The results of the ERG analysis were that golimumab was extendedly dominated by etanercept, and the preferred treatment was either conventional treatment or etanercept, depending on the ICER threshold. Uncertainty was also substantial. NICE issued guidance (technology appraisal [TA] 233), which recommended golimumab according to the indications described in TA143 for etanercept and adalimumab, i.e. as first-line therapy among the TNF-α inhibitors unless patients are intolerant to one or both alternatives. Given the factors cited by NICE for their decision, the ERG recommends that there should be greater clarity in the NICE methods guidance on handling uncertainty in CEAs as well as the incorporation of benefit from process of care.