PORTFOLIO MANAGEMENT TO INFORM CANCER CLINICAL TRIALS RESEARCH INVESTMENT DECISIONS IN THE US: A PROOF OF CONCEPT EVALUATION

Purpose: A recent report from the Institute of Medicine (IOM) emphasized the need for new approaches to rigorously and systematically evaluate and prioritize public investments in cancer clinical trials.  Our objective was to evaluate the potential application of a portfolio management framework to inform the selection and prioritization of trial proposals within SWOG, a large cancer clinical trials cooperative group.

Methods:  We previously identified characteristics of cooperative group-sponsored trials that were associated with insufficient accrual (defined as <50% target) and developed a prediction model for accrual feasibility based on these factors.  We also developed methods to estimate the societal return of SWOG's trial proposals (defined as the net benefits from reduced uncertainty for the treatment decision being studied, given full accrual, minus the trial's expected costs) using minimal modeling value of information analyses that were feasible for use in real-time prospective evaluation and acceptable to SWOG members.  We demonstrate the application of these methods within a proof-of-concept portfolio evaluation using a sample of 9 randomized phase II and III clinical trial proposals reviewed by SWOG between 2008-2013, of which 5 were approved and funded and 4 were not. 

Results: As originally designed, the predicted risk of insufficient accrual ranged from 2 to 28% and the expected net benefits of sampling (with successful accrual) ranged from -$3.7 billion to +$4.8 billion across the 9 trial proposals included in this retrospective evaluation (Figure).  The overall risk-adjusted return of SWOG's sample portfolio of approved trials was valued at $2.7 billion.  We illustrate how a portfolio management framework can provide context in the approval decisions of individual prospective trial proposals, make explicit the trade offs between the expected return and accrual feasibility with different trial designs and planned sample sizes, and inform longer-term strategic planning and prioritization.

Conclusions: A portfolio management framework is a feasible and potentially useful response to the IOM's call for more rigorous and systematic approaches to select and prioritize trials within the cancer clinical trials cooperative groups.  Future work is needed to assess the role of this framework in prioritization decisions.

Figure: Accrual feasibility and expected net benefits for a SWOG's sample portfolio. Trial proposals that were approved and funded are shown in green; those that were not funded are shown in red.