Nina Niu Sanford: Our work on clinical trial endpoints and inclusion criteria in GI oncology trials

Differences persisted when stratifying by trial Phase. We also found differences in inclusion criteria, with more systemic trials being limited to smaller patient subsets vs. most RT trials were for all comers, at diagnosis.

All RT trials (except 1, RTOG 1112) have been negative. Our argument is that RT will not improve OS in unselected patients. Thus, future trials in GI cancers should consider the following:

1) Incorporate new patient-centered endpoints. Arguably, RT does not improve OS in most neoadjuvant or adjuvant settings. However, patients also value QOL – sometimes as much, or even more than OS. We believe that RT can improve QOL……but QOL has historically been poorly captured and often using long, burdensome instruments. Future trials should include more specific QOL endpoints such as time toxicity, financial toxicity, and QOL should continue to be measured at progression and until death.

2) Refine trial inclusion criteria. Most RT trials in GI cancers have broad inclusion criteria. This makes no sense for a local modality such as RT, since the predominant form of disease progression for locally advanced GI cancers is distant.

Thus, RT trials need to more selective – the way surgeons select patients for a Whipple, or med oncs choose patients for targeted therapy. Our final paragraph summarizes it all.