onc brain

The longer read

PREPEC does the thing surgical technique trials rarely do: it randomises the choice and measures the outcome patients actually notice. The result is a clean positive on its own terms. Pre-pectoral placement improved BREAST-Q physical well-being of the chest at 24 months by 4.8 points (1.0 to 8.7, p=0.01), and the direction matches the mechanistic rationale, since leaving pectoralis major undisturbed avoids the muscle division, animation deformity and chest-wall discomfort that sub-pectoral placement builds in. The interval's lower bound sits just above zero, so the trial establishes that the benefit is real more convincingly than it establishes how large it is.

The secondary safety endpoint is where the read gets complicated, and it deserves more weight than its secondary status suggests, because it was prespecified with an explicit non-inferiority hypothesis that the data did not support. Unplanned loss or replacement of expander or implant reached 21.1% pre-pectoral versus 14.5% sub-pectoral, an adjusted difference of 5.7% whose interval (-2.4 to 13.8) crosses zero. A skeptic will note that the confidence interval includes no difference and refuse to call this a demonstrated harm. That reading is technically correct and practically misleading: the trial was not designed to establish superiority on this endpoint, it was designed to rule out a meaningful increase, and it failed to do so. The honest summary is that pre-pectoral placement has not been shown safe on device retention, not that it has been shown equivalent.

So the interpretive question is a trade-off, and it is one that has to be posed to the patient rather than settled by the trial. Roughly one in five women in the pre-pectoral arm required unplanned device removal or replacement within two years. Set against that, the average gain on a 0 to 100 patient-reported scale is 4.8 points. Different patients will price those differently, and a mean difference cannot tell you how the benefit distributes: it is plausible that some women gain far more than 4.8 points while others gain nothing and absorb the reoperation risk. Nothing in the source resolves that.

Two methodological features should shape confidence. First, a positioning trial cannot blind anyone, and the primary endpoint is entirely patient-reported. Knowing which technique you received, in a field where pre-pectoral has been marketed as the modern approach, is exactly the setup that inflates a subjective score. Second, longitudinal completion ranged 83-95% across timepoints and the primary estimate depends on multiple imputation, with imputed baseline values feeding the model as a covariate. Neither is a fatal flaw, and both are handled conventionally, but they mean the 4.8-point estimate carries more uncertainty than its interval alone conveys.

For a radiation oncologist the most important thing about PREPEC is what it does not report. Post-mastectomy radiotherapy is the single largest modifier of implant reconstruction outcomes, and the source gives no RT stratification, no irradiated subgroup, and no capsular contracture data. The women who most need this question answered, those proceeding to chest wall irradiation with an implant in place, cannot take an answer from these results. Whether the device-retention gap widens under radiation, and whether the well-being advantage survives it, remain open and are the questions worth following into the full publication.