onc brain

The longer read

TNT is no longer the question in high-risk rectal cancer; the question is which TNT. RAPIDO, PRODIGE-23 and STELLAR established that moving systemic therapy before surgery improves distant control, and they did it with three different architectures: short-course RT followed by consolidation chemotherapy, induction chemotherapy before long-course chemoradiation, and short-course RT plus chemotherapy tested against conventional chemoradiation. TNTCRT tests a fourth, an uninterrupted doublet that brackets long-course chemoradiation on both sides with oxaliplatin running concurrently with RT rather than only before or after it. That framing sets how much weight the result deserves. This is a regimen contribution inside an accepted paradigm, not a challenge to it, and the authors say as much when they position the regimen as a standard option rather than the standard.

The internal pattern of the result is more informative than its headline. DFS and MFS move together while locoregional failure sits at 6.03% and 6.19%, essentially identical and low for a cohort that was 47.82% cT4 and 75.98% cN2. Read plainly, intensification bought distant control and nothing local, which is what one should expect: long-course chemoradiation with a fluoropyrimidine already operates near its local ceiling here, and the residual failure mode in high-risk LARC is metastatic. For a radiation oncologist the practical consequence is that nothing in this trial licenses a change to the RT prescription, and the source text never reports dose, fractionation, technique or target volume, so the RT component cannot be audited against local practice at all.

Two things should temper confidence. First, the comparison is not simply more chemotherapy against less. The control arm's oxaliplatin was scheduled after surgery, and postoperative chemotherapy in rectal cancer is delivered unreliably; if completion in the control arm was poor, part of the DFS separation reflects therapy planned rather than therapy given in a better order. The source does not report control-arm adjuvant delivery, so the mechanistic claim stays open. Second, accrual ran from June 2017 to December 2023, a window in which TNT went from investigational to routine. A control arm defined in 2017 was contemporary at the start and progressively less so at the end, a hazard shared by every long-accruing trial testing a strategy the field is adopting underneath it.

Overall survival is the usual objection to trials like this, and here it is unresolved rather than negative: 90.2% against 87.5% at three years, P = .167. With absolute survival that high and follow-up at a median of 51 months, an OS separation was never plausible on this event count, so DFS carries the result and should be read as such rather than as a surrogate that failed.

The finding likeliest to change behavior is not the primary endpoint. pCR rose from 9.80% to 26.37%, and the preceding phase 2 on the same backbone reported a 36.2% complete response rate. That magnitude puts intensified long-course TNT into the organ-preservation conversation, which is where rectal practice is actually moving. The caution is that pCR is a resection-specimen endpoint measured in the roughly 87% and 90% of pts who reached total mesorectal excision, and this trial neither randomized nor protocolized watch-and-wait. What it establishes is that the response rate is available; whether it converts into durable nonoperative management is a different trial. The comparative reassurance worth carrying forward is that this intensification did not import RAPIDO's locoregional recurrence signal.