Design considerations for randomized comparisons of neoadjuvant–adjuvant versus adjuvant-only cancer immunotherapy when tumor measurement schedules do not align (SWOG S1801)

Abstract

Background

In 2022, SWOG S1801 was the first trial to demonstrate that single-agent anti-PD-1 checkpoint inhibition used as neoadjuvant–adjuvant therapy leads to significantly improved outcomes compared to adjuvant-only therapy. Endpoints in trials comparing neoadjuvant–adjuvant to adjuvant strategies need special consideration to ensure that event measurement timing is appropriately accounted for in analyses to avoid biased comparisons artificially favoring one arm over another.

Methods

The S1801 trial is used a case study to evaluate the issues involved in selecting endpoints for trials comparing neoadjuvant–adjuvant versus adjuvant-only strategies.

Results

Definitions and timing of measurement of events is provided. Trial scenarios when recurrence-free versus event-free survival should be used are provided.

Conclusions

In randomized trials comparing neoadjuvant–adjuvant to adjuvant-only strategies, event-free survival endpoints measured from randomization are required for unbiased comparison of the arms. The time at which events can be measured on each arm needs to be carefully considered. If measurement of events occurs at different times on the randomized arms, modified definitions of event-free survival must be used to avoid bias.