Assessing contribution of effect in oncology combination therapies: lessons learned to inform and optimize future registrational trial designs.

Combination therapies can be used to significantly improve clinical outcomes in patients undergoing treatment for cancer; however, they also present challenges when establishing the contribution of effect (COE) for each component. Although factorial trials remain the most rigorous approach for determining COE, they are often impractical, such as in settings with rare cancers, aggressive disease, evolving standards of care, or monotherapy components with limited activity. As more combination regimens are co-developed from the outset, alternative approaches to registrational trial designs are needed to balance scientific rigor with feasibility and patient acceptability. Thoughtfully designed alternative approaches can support an understanding of COE when grounded in prior evidence and mechanistic understanding. Four major categories of evidence emerge to support designing combination trials without factorial arms: (1) data from trials of the same combination in other tumor types; (2) data from trials of agents with similar mechanisms of action; (3) data from early-phase or parallel trials of the combination within the same disease setting; and (4) strong mechanistic or clinical evidence that a monotherapy component lacks activity and/or imposes harm, supporting omission of inactive arms. Early regulatory engagement, prespecified decision rules, and transparent justification of design choices are essential to ensure that combination regimens provide meaningful improvement in efficacy over monotherapies while minimizing unnecessary toxicity. Applying these principles to combination therapy development can maximize patient benefit while maintaining scientific rigor.