onc brain

The longer read

The case for a brachytherapy boost in prostate cancer has rested largely on ASCENDE-RT, which compared an LDR boost with dose-escalated EBRT and favoured the boost on biochemical control. That trial gave ADT and enrolled a population weighted toward higher-risk disease. GETUG P05 asks a narrower question that matters for daily practice: in intermediate-risk disease, with no ADT, does a boost add anything over 80 Gy external beam? On the primary endpoint the answer is no. Five-year bPFS of 88.0 vs 88.2% is about as null as a randomised comparison gets, and the absence of a signal in either GG2 or GG3 makes it harder to argue that the effect is hiding in a less favourable subgroup.

The obvious objection is follow-up. A median of 61.7 months is reasonable for a trial in intermediate-risk disease, but biochemical failure after dose-escalated RT accrues slowly, and the boost advantage in prior work became clearer with longer observation. A gap could still open. Still, with curves essentially superimposed at five years, any late divergence would have to come from a small number of events in a population where most pts are already controlled. The KM slide also prints a log-rank p of 0.9900 against a headline of 0.90, which needs to be cleared up in the full report but does not change the direction of the result.

The toxicity data are where the trial moves practice most. The boost reduced acute GI toxicity, including all acute GI grade 2 and 3 events, but the late picture runs the other way. Late GU toxicity of any grade was 82.2 vs 63.6% (p=0.0006), late GU grade 2 nearly doubled (34.1 vs 17.9%), and late global grade 3 was 10.4 vs 6.6%. Late GU morbidity matters more to pts over the long run than transient bowel symptoms, so for a reader weighing a boost in intermediate-risk disease, the trade is unfavourable when control is equal.

Several design features limit how far this travels. LDR and HDR boosts are pooled, and the source gives no separate outcomes, so a reader with an HDR-only programme cannot tell whether the late GU excess belongs to one technique. The control arm is 80 Gy in 2 Gy fractions over eight weeks, which many centres have replaced with moderate hypofractionation or SBRT. The result therefore shows that a boost does not beat conventional dose escalation, not that it fails against modern regimens, which were never tested. The design also excluded ADT, so the trial cannot address whether a boost interacts with hormonal therapy as it may have in ASCENDE-RT.

The practical read is that in intermediate-risk disease treated with RT alone, dose-escalated EBRT reaches the same biochemical control with less late GU morbidity, and a boost needs a reason beyond bPFS. What remains open is whether high-risk features, ADT, or longer follow-up recover a boost benefit, and how a boost compares with ultrahypofractionated external beam.