EULAR 2016 | Daily Highlights
RELATIVE EFFICACY OF ADALIMUMAB VERSUS SECUKINUMAB IN ACTIVE ANKYLOSING SPONDYLITIS: A MATCHING-ADJUSTED INDIRECT COMPARISONAbstract: OP0115
Authors: K. A. Betts1, M. Mittal2, J. Song1, M. Skup2, A. Joshi2,*
Co Authors: 1Analysis Group, Inc., Boston, MA, 2AbbVie Inc., North Chicago, IL, United States
Two phase 3 clinical trials, MEASURE 1 and 2, demonstrated the efficacy of secukinumab for active ankylosing spondylitis (AS). However, there is limited data comparing the efficacy of secukinumab with established AS treatments.
The objective of this study was to compare relative efficacy between adalimumab and secukinumab, estimate the number needed to treat (NNT), and evaluate the incremental cost per responder (CPR) for the treatment of active AS.
An indirect comparison was conducted using individual patient data from the adalimumab clinical trial (ATLAS)1 and published data from the secukinumab trials (MEASURE 1 and 2).2 Only dosages approved by the European Medicines Agency (adalimumab 40mg and secukinumab 150mg) were included in the analysis. To adjust for the cross-trial differences, individual patients in ATLAS were re-weighted to match the mean baseline characteristics in the MEASURE trials.3,4 Specifically, age, weight, gender, duration of AS, baseline methotrexate use, baseline sulfasalazine use, presence of positive HLA-B27, BASDAI score and total back pain score were included in the baseline matching. The ASAS20, ASAS40 and ASAS 5/6 response rates relative to placebo at the end of the trial period (adalimumab: 12 weeks, secukinumab: 16 weeks) were compared. NNTs to achieve one additional ASAS20, ASAS40 and ASAS 5/6 responder and the corresponding incremental CPRs were also calculated. Drug costs in terms of 2015 Euros were obtained from the Lauer Taxe.5
After matching, mean baseline characteristics were balanced across the ATLAS and MEASURE 1 and 2 populations. Compared to secukinumab (n=197), adalimumab (n=204) demonstrated similar efficacy in ASAS20/40 and ASAS 5/6. The mean differences (95% confidence interval) between adalimumab and secukinumab in relative ASAS20, ASAS40 and ASAS 5/6 response rates were 2.3% (-12.8%, 17.5%), -0.4% (-14.1%, 13.4%), and -3.1% (-17.3%, 11.2%), respectively. The NNTs to achieve one additional responder were also comparable between adalimumab and secukinumab: 2.9 vs 3.1 in ASAS20, 3.7 vs 3.7 in ASAS40, and 3.1 vs 2.8 in ASAS 5/6. Incremental CPRs relative to placebo were lower for adalimumab compared to secukinumab across all efficacy measures (Figure 1).
In the absence of head to head trials comparing adalimumab and secukinumab, the current indirect comparison reduced cross-trial differences and provides timely and reliable evidence for physicians and payers. After adjusting for cross-trial differences in baseline characteristics, adalimumab demonstrated similar efficacy as secukinumab in terms of ASAS20, ASAS40 and ASAS 5/6 responses and was associated with lower incremental CPR compared with secukinumab over the trial period among patients with active AS.
- van der Heijde D, et al. Arthritis Rheum. 2006;54:2136–46.
- Baeten D, et al. N Engl J Med. 2015;373:2534–48.
- Signorovitch JE, et al. Pharmacoeconomics. 2010;28:935–45.
- Signorovitch JE, et al. Value Health. 2012;15:940–7.
- http://www.lauer-fischer.de. Accessed December 2015.
Disclosure of Interest
K. Betts Consultant for: AbbVie, M. Mittal Shareholder of: AbbVie, Employee of: AbbVie, J. Song Consultant for: AbbVie, M. Skup Shareholder of: AbbVie, Employee of: AbbVie, A. Joshi Shareholder of: AbbVie, Employee of: AbbVie
While TNF inhibitors have been successfully used for the treatment of AS for more than a decade, secukinumab, an IL-17-blocking agent, has recently been shown to also be effective in AS. It remains to be determined which patients will mostly benefit from one class of agents in comparison to the other. In the absence of head-to-head trials (TNF inhibitor versus secukinumab), indirect cross-trial comparisons can by biased by differences in patient characteristics. Adjustment for such biases are usually impossible, as individual patient data are rarely available publicly for all trials. However, many researchers have the opportunity to access individual patient data for trials of one treatment, but only aggregate data for trials of comparator treatments. A new statistical method – called Matching-Adjusted Indirect Comparison (MAIC), recently proposed by J.E. Signorovitch – leverages all available data in this setting by adjusting average patient characteristics in trials with individual patient data to match those reported for trials without individual patient data. Treatment outcomes, including continuous, categorical and censored time-to-event outcomes, can then be compared across balanced trial populations - a major limitation being, however, that significantly lower patient numbers are available for comparison. The two abstracts present comparisons of secukinumab versus adalimumab from the respective perspective of Novartis and Abbvie (each pharmaceutic company having the individual patient data for their own drug, but only the published aggregate data of the comparator treatment). Have the two comparisons of the same randomized trials of secukinumab and adalimumab found the same result? Of course, not… While the study sponsored by Novartis demonstrates a significantly higher improvement in clinical outcomes upon treatment with secukinumab, the study sponsored by Abbvie finds a similar efficacy of adalimumab and secukinumab. We might want to wait for the results of comparisons from observational analyses in AS registries to help us guide the choice of treatment. If this last sentence is interpreted as a call for including all Swiss AS patients on secukinumab or ustekinumab into SCQM we might have the results in a near future.
PD Dr. Adrian Ciurea
> zurück zur Übersicht