onc brain

The longer read

The question here is narrower than the headline suggests. TNT is already standard in pMMR LARC, and earlier phase II work had suggested that immunotherapy layered onto it raises complete response. What was missing was a controlled comparison at scale, and this trial supplies one: the same short-course RT, the same CAPOX, with sintilimab as the only randomized variable. That design isolates the drug cleanly. It does not test the radiotherapy, so the result should not be read as an endorsement of short-course over long-course chemoradiation as the partner for a checkpoint inhibitor. Trials such as UNION and TORCH explored short-course RT with checkpoint blockade, but neither held everything else constant against an immunotherapy-free TNT arm, which is the gap this one fills.

Confidence in the direction of the effect should be high. Overall CR was 51.9% vs 26.8% on denominators of 289 and 291, which is close to the whole randomized population, so the comparison is not built on a selected subset. The pCR figures need more care. They are computed among operated pts only, 211 in the iTNT arm and 247 in the TNT arm. Fewer iTNT pts reaching surgery is what one would expect if more of them were diverted to watch-and-wait after a clinical complete response, although the source does not give the per-arm split. Removing the best responders from the surgical denominator would, if anything, pull the iTNT pCR rate down, so the pCR gap is unlikely to be flattered by that selection. The softer component is the clinical one. A cCR is an assessor-dependent call, the source does not say whether assessment was blinded, and a clinician who knows a pt received immunotherapy may be more willing to defer surgery. That is where a skeptic should push.

The population is high risk: T3-4 in 94.9%, node-positive in 92.9%, MRF+ in 48.8%, EMVI in 45.4%, and lower tumors in 71.3%. This is the group in which sphincter loss is most at stake and organ preservation is hardest to achieve, so the result applies directly to the low rectal tumor facing a permanent stoma. It says less about the early, low-risk tumors that some organ-preservation programs select.

The cost is measurable. Grade 3-4 toxicity was 27.7% vs 19.9%, with grade 3-5 immune-related events in 3.4%. Treatment delivery did not suffer: every pt completed SCRT and chemotherapy completion was 92.9% vs 94.1%. Surgical morbidity and late effects are not reported, and the grade 3-5 category leaves open whether any immune-related event was fatal.

What the trial has not shown is the endpoint it was powered for in its phase 3 portion, 3-year DFS. Complete response is a response endpoint, and the experience with short-course TNT in RAPIDO is a reminder that early response gains do not guarantee durable locoregional control. The result would be undermined if regrowths in the watch-and-wait cohort erode the clinical component of the CR advantage, or if DFS curves fail to separate in a pMMR population where checkpoint blockade has historically done little on its own. Until those data arrive, this supports discussing the regimen as an organ-preservation option, not treating it as a new standard.