onc brain

The longer read

The reason this trial needed to exist is not oncologic. Hypofractionation was settled for the conserved breast years ago through START and FAST-Forward, and post-mastectomy hypofractionation without reconstruction had its own supporting data, yet many practices kept reconstructed pts on 25 fractions because the reconstruction was the perceived fragile structure. RT CHARM answers that worry directly, with a reconstruction-specific primary endpoint, across 209 academic and community sites, and with every reconstruction type and sequence allowed. That pragmatic eligibility is the main source of its external validity: this is the mix of implants, expanders, autologous flaps and delayed sequences that actually arrives in clinic. It sits next to FABREC, which addressed the same 16-fraction schedule in implant reconstruction; the added value here is the autologous and delayed cohorts and the stratification on them.

The noninferiority framing deserves a careful read. The 10% absolute margin was set against an assumed 25% control complication rate, but the control arm ran at 10.2%. Against that baseline, a 10-point margin would have tolerated roughly a doubling of complications, which is a generous threshold. The upper bound of the observed interval (6.5 points on the stated scale) is what a skeptic should anchor on, rather than the noninferiority p-value. Even so, the point estimate difference of 2.1 points is small, the completed-reconstruction analysis of 631 pts gave a consistent -2.2% with an interval reaching 7.2, and the multivariable model found arm was not a predictor. Read together, the data argue that any fractionation penalty, if real, is modest and much smaller than the effect of reconstruction timing.

That timing signal is the most decision-relevant finding for the multidisciplinary team. Delayed reconstruction carried an OR of 2.96 for complications regardless of arm, while implant-only versus autologous (OR 1.65, p=0.07) did not reach significance despite the abstract's wording. Because timing was a stratification factor rather than a randomized comparison, this is an association within a randomized cohort, confounded by why reconstruction was delayed (expander strategy, neoadjuvant course, surgeon preference). It should inform counselling, not be read as evidence that immediate reconstruction is safer.

Oncologically, 3-yr locoregional recurrence was low in both arms, 1.5% versus 2.2%, with an LRRFS HR of 1.01 and an interval spanning 0.53 to 1.96. With 18 total events the trial cannot exclude a modest LRR difference, and it was never powered to; the reassurance comes from consistency with the broader hypofractionation literature rather than from this trial alone. The acute toxicity gain is real and patient-visible, with G2 dermatitis falling from 42% to 20%.

What would make the conclusion wrong? Late reconstruction events, capsular contracture in particular, can emerge beyond 3 to 4 years, and a hypofractionation penalty on late tissue effects would surface there first. Unblinded complication assessment could bias in either direction. Finally, target volumes, internal mammary coverage, bolus and boost practice are not described in the source excerpt, so transfer to a given nodal-coverage policy assumes comparable planning. The practical read is that fear of reconstruction failure is no longer a sound reason to withhold 16-fraction PMRT, while 5-fraction schedules in this setting remain untested here.