A study describes an additivity model that can be used to predict the likelihood of success or failure of combination therapies in phase III trials
Most of approved drug combinations in oncology have an additive, but not a synergistic clinical efficacy, according to a recently published study that analysed progression-free survival (PFS) data from phase III trials by a model of drug additivity (Nat Cancer. 2023 Nov 16).
The method developed by the team of the UNC Lineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill, Chapel Hill, US, used PFS distributions of individual drugs to generate a joint distribution, with correlation as supported by experimental data. Considering PFS as a clinical efficacy metric, the researchers defined clinical drug additivity as the sum of PFS benefits in individual patients. To assess whether the additivity model could predict the ability of combination therapies to significantly improve PFS in phase III trials, approved combinations over a 25-year interval (1995 to 2020) were firstly examined, followed by both positive and negative trials published in a 5-year interval (2014-2018). For each approved combination, a hazard ratio was calculated for the expected combination therapy PFS versus the trial’s control arm (Cox proportional hazards model; ɑ = 0.05).
Among the 37 examined drug combinations satisfying the study inclusion criteria, dose-matched monotherapy data show that 95% of combination therapies exhibited PFS distributions that were equal to the sum of their parts or less, with no unexpected deviations from the efficacy profiles of single agents. “These findings do not mean that the approved drug combinations are not clinically effective, but they show that a more-than-additive effect is not necessarily equal to a clinical superiority – for example, additive effects of highly active drugs is better than synergy among weaker drugs”, says Dr Marco Donia, National Center for Cancer Immune Therapy, Department of Oncology, Copenhagen University Hospital, Herlev, Denmark. “While the study does not specifically identify combinations with more than additive or less than additive effects based on the mechanism of action, it highlights that combinations involving a new drug without monotherapy approval and those with marginal improvement over the control arm tend to result in less than additive effects.”
The methodological framework provided by Hwangbo and colleagues demonstrated that the additivity model can be used to predict the likelihood of success or failure of drug combinations in phase III trials, with success measured as statistically-significant improvement of PFS to the control arm using monotherapies. Also, the results reported imply that sequential treatment can be as good as concurrent combinations. “Assuming combinations with additive efficacy, where drug A and drug B improve PFS by a given value regardless of being administered in first or second line, a strategy where drug A is administered first and drug B is administered at progression - sequencing, rather than combination - may achieve similar efficacy in the long run, which is what counts the most for patients”, explains Donia. “In this context, if drug A and drug B induce toxicities synergistically, using drug A plus drug B in combination may be a disadvantage for quality of life and not improve overall survival, even while they improve time to first progression.”
As Donia concludes, the methodology proposed appears adaptable to identify synergistic versus additive toxicities of combinations: “Having established that most combination in oncology have additive but not synergistic effects, it will be critical to understand whether toxicity of the individual combinations are additive or synergistic to inform treatment guidelines and guide future studies of combination versus sequencing.”