onc brain

The longer read

The design question here is narrower and cleaner than in most metastasis-directed therapy trials. Both arms received the same finite course of ADT and then stopped, so the comparison isolates what SBRT adds once the systemic backbone is held constant. STOMP and ORIOLE compared MDT against observation, with no hormone therapy in either arm. EXTEND tested MDT added to intermittent hormone therapy in oligometastatic prostate cancer, which is the closest precedent in logic. STOP moves that same question into disease that only PSMA PET can see, with CT and bone scan negative, which is the population now filling clinics as PSMA PET replaces conventional staging at recurrence.

The efficacy readout is biochemical progression: 21% vs 60% at 3 years, p=0.004, with the median not reached vs 27 mo. Biochemical progression after stopping ADT is a soft endpoint. It depends on how fast testosterone recovers, how often PSA is drawn and how progression is defined, and the source gives none of these. The control arm figure is still informative on its own terms: a majority of men given 8-9 months of ADT alone had progressed biochemically by 3 years, which argues that a finite hormone course without local consolidation does not hold PSMA PET-only disease for long.

Several features should temper confidence in the size of the effect. The trial enrolled 60 men and its stated success was feasibility, with 91% of eligible pts enrolled, so the efficacy comparison carries wide uncertainty and no confidence interval appears in the source. A p-value of this size in a trial this small can rest on a handful of events. The endpoint closest to what a patient feels, re-initiation of ADT, came in at p=0.06 and did not reach the conventional threshold. Re-initiation is a clinician decision, and blinding is not stated, so that endpoint is open to knowledge of arm in either direction. A QoL slide was shown but no values were captured, and toxicity is absent from the source entirely.

For a radiation oncologist the transferability gap is technique. SBRT followed institutional protocols by anatomic region, with no dose, fractionation or target volume in the source. Nodal (miN1) and distant (miM1) deposits are pooled, so the trial says nothing about whether pelvic nodal recurrences are better served by SBRT to the involved node or by elective nodal RT, and no in-field vs out-of-field failure data are given. The timing parameter is the one concrete instruction: SBRT as early as possible and within 4 months of ADT.

What this supports is the already common habit of consolidating PSMA PET-only oligometastatic HSPC with SBRT when a short ADT course is being given. What it leaves untouched is whether ADT is needed at all alongside SBRT, and whether intensified hormonal therapy would erase the difference. The result would be misleading if the PSA separation reflects delayed biochemical relapse without a matching delay in metastatic progression or castration resistance, and only a larger trial with MFS and time-off-ADT endpoints can answer that.