onc brain

The longer read

The RT omission evidence that shapes current practice in older women with small ER+ tumors comes from trials in which every patient took endocrine therapy. PRIME II and CALGB 9343 showed that whole-breast RT could be dropped with acceptable local control, but the low recurrence attributed to omitting RT was always partly an ET effect. TOP-1 removes that support. It studies the group Dutch guidelines already leave without ET and asks whether breast-conserving surgery alone is enough. At a median 5.1 years the answer falls short of the investigators' own bar. Local recurrence was 14.8%, and the lower confidence bound of 12.3% sits above the 10% ceiling the study was powered to exclude, a ceiling patient advocates had accepted in advance. This is a prespecified threshold failure, not a judgement made after the fact.

Several design features should raise confidence in the direction of the result. Accrual reached 792 of 800 planned across multiple centers. The analysis used cumulative incidence with death as a competing event, which is the right method in a cohort where 81 patients died; a naive Kaplan-Meier estimate would have overstated recurrence. The primary endpoint counted only histologically confirmed invasive ipsilateral events, so DCIS recurrences are excluded, and the full ipsilateral event burden is if anything higher than 14.8%. What limits the reading is the comparator. The 1.3% figure with RT comes from a national registry, not a concurrent arm, so the RT-attributable reduction cannot be estimated from this study. Selection and era differences between registry and cohort cannot account for a gap this wide unless the registry population was very different, and nothing in the source suggests that.

The central point the trial cannot settle is apportionment. RT and ET were both omitted, so the excess recurrence reflects the loss of both, and a reader cannot tell how much returns if only one is given. EUROPA, which compares RT alone with ET alone in older women, addresses that choice directly. TOP-1 therefore argues against omitting both adjuvant therapies in this population. It says nothing against the established practice of omitting RT in women who take ET.

The authors' case for keeping omission available rests on salvage and survival. Of 123 recurrences, 114 were treated surgically and 86 had repeat BCS, and only 2 of 81 deaths were breast cancer related. That survival signal is reassuring but early: low-grade ER+ disease relapses slowly, and five years is too short to show whether recurrence affects breast cancer mortality. The salvage numbers also show a cost the headline hides. 70 patients received RT at recurrence, so for a large share of those who relapsed, omission postponed radiotherapy, added a second operation, and moved both to an older age. A shared decision about omission without ET should present it as accepting roughly one-in-seven local recurrence by five years, with RT and surgery still likely if that happens, rather than avoiding treatment.

The conclusion would be wrong if longer follow-up showed that recurrences plateau after five years while competing mortality rises, which would shrink the absolute lifetime cost. That seems unlikely for ER+ disease, whose recurrence hazard tends to persist late.