onc brain

The longer read

This series belongs to a familiar genre: a single-institution, retrospective account of a brachytherapy technique applied to elderly pts who either could not tolerate or would not accept Mohs surgery. Its 1-yr local control of 92% sits comfortably inside the 87% to 100% range the authors cite from earlier HDR surface-mold series, so the result is best read as a technical confirmation that a CT-planned catheter flap performs like other surface applicators, not as evidence that it performs better. The comparison that actually defines the field is the randomized trial of surgery vs RT in facial BCC that the authors themselves cite, where 4-yr failure was 7.5% with RT vs 0.7% with surgery. Nothing here challenges that ordering for a pt who can have surgery, and the authors do not claim it does.

The more instructive content is dosimetric. The flap places its source channels 5 mm from the skin, the CTV adds a 5 mm margin with no PTV expansion, and median tumor thickness was 5 mm with a range up to 7 mm. Against that geometry, the median V100 of 85% and D90 of 92% mean that a typical plan left part of the target below prescription, despite a stated goal of covering at least 95% of the PTV. Depth was assessed largely by palpation, with pathologic depth used only when available. For a reader deciding whether to adopt a flap technique, the open issue is the deep margin of thicker lesions, and this series cannot resolve it because recurrences were few and follow-up short.

The pattern of failure is worth more than its four events suggest. Two nasal SCCs, an eyelid BCC and a lip BCC recurred, all in the irregular, high-curvature anatomy the flap is meant to handle better than a rigid mold or an electron field. That is not proof of a technique-specific weakness, since these are also the highest-risk sites in any series, but it argues against assuming the flap has solved conformity on the nose and lip.

Follow-up is the decisive limitation for interpreting both efficacy and toxicity. A median of 15 months, with some lesions observed for only 6, is short for BCC, whose recurrences can declare late, and short for late skin effects after a hypofractionated schedule. The authors lean on Gauden et al., with 66-month median follow-up, to argue that chronic toxicity stabilizes in the first one to two years, which is plausible but borrowed rather than shown. The toxicity reporting also contains internal inconsistencies: the abstract lists the three acute events as grade 2 while the results call them grade 3, and the discussion states there were no chronic grade 2 or higher toxicities while the results report two pts with grade 2 telangiectasia at 12 months. Neither changes the overall picture of a tolerable schedule, but both lower confidence in the precision of the toxicity numbers.

What would make this wrong is a late recurrence curve that diverges from the 1-yr estimate, or deep-margin failures in thicker tumors once larger numbers accumulate. The decision it supports is narrow: for a surgery-unfit pt with a thin T1 facial lesion, 40 Gy in 8 twice-weekly HDR fractions via a flap is a reasonable, convenient option with acceptable acute effects. Whether BCC and SCC should receive different dose or margins, and how surface HDR compares with EBRT or Mohs head to head, remain open.