reFLASH: a proof-of-concept meta-analytic framework to validate surrogate endpoints in the absence of randomized clinical trial data.

What was studied

This study tested a new way to check stand-in measures for trial outcomes. Such checks normally need patient data from many randomised trials. The authors used data from 12 follicular lymphoma trials with 3996 patients. They split each trial's arms apart and paired them again into 12 artificial trials. They used weighting to make both arms look alike. They compared complete response at 30 months with progression-free survival.

What they found

Treatment effects varied widely between the artificial trials. Still, the link between both measures stayed strong. Both models gave a value of 0.769 for this link. This met the preset limits. A second pairing method gave nearly the same value. Leave-one-out checks generally kept the value above 0.70. The link was lower than in the original trials.

What it means for patients

The authors say artificial trials from single-arm data recovered a high link between both measures. They say the idea offers a practical path when randomised trials are scarce.

Why it is not certain yet

The authors say the method is not a substitute for randomisation. It is open to leftover confounding and to pairing choices. The abstract calls it a proof of concept. The link was lower than in the benchmark built from real trials.

Source

reFLASH: a proof-of-concept meta-analytic framework to validate surrogate endpoints in the absence of randomized clinical trial data.

We restate facts from the source record in our own words. The original wording belongs to its publisher.