onc brain

The longer read

The value of this analysis is not that SABR controls oligometastatic disease, which the randomised literature already established for selected populations, but that it puts a number on what in-field control looks like when the technique leaves the trial and enters routine practice at 57 institutions. 88.6% local control at three years is close enough to what single-institution SABR series report that the main read is reassurance: the effect does not appear to be a product of trial-grade patient selection or trial-grade planning. That matters because the oligometastatic literature has been built on small, highly curated cohorts, and the standard skeptical position has been that community delivery would erode the signal. On this evidence it does not erode much.

What should move a reader's confidence in the other direction is the design. This is a registry, dose and fractionation were institution-chosen, and there is no comparator arm, so every association reported here is an observation about which lesions did well, not a demonstration that a planning decision caused them to do well. The minimum PTV dose finding is the clearest example. It is entirely plausible on radiobiological grounds, and it is also exactly what you would expect to see if the lesions that received poor minimum coverage were the ones abutting bowel, cord, or previously irradiated tissue. Those constraints are also markers of anatomically difficult, biologically unfavourable, or re-treated disease. Without the dose thresholds and the site-stratified analysis, which the source does not give, a reader cannot tell the causal story from the confounded one, and should not adopt a new minimum-dose constraint on the strength of this.

The same caution applies harder to the de novo versus repeat OMD comparison, which the authors attribute to higher delivered dose. Repeat oligometastatic disease differs from de novo disease in ways that have nothing to do with dose: it has already demonstrated the ability to progress through a prior local therapy, and it is being treated in a patient whose disease has declared a more aggressive trajectory. Attributing that gap to dose is the least likely explanation on the list, and a registry has no way to adjudicate between them.

The colorectal result is the finding most likely to change something. Worst local control at 19.6% at three years while receiving the highest median dose per fraction is a genuinely informative combination, because it rules out the easy explanation. If CRC mets were simply being undertreated, the dose data would show it and it does not. That leaves intrinsic radioresistance, and it argues that pushing dose further within conventional SABR schedules may have limited headroom, which is why the authors reach for combination strategies. For a radiation oncologist counselling a patient with a colorectal oligomet, the practical consequence is a different expectation of durability than the pooled 88.6% figure implies, and a lower threshold for considering trial enrolment or systemic intensification alongside local therapy.

What this analysis leaves untouched is whether any of this local control translates into survival, since the reported endpoint is in-field progression and the registry carries no comparator. Local control is a necessary condition for the oligometastatic hypothesis to work, not a sufficient one, and a 3.5-percentage-point difference in three-year in-field failure between two primaries says nothing about whether either patient lives longer for having been treated.