onc brain

About · curated by Nick Boehling, MD · @nb2276
Caveats dominate

ctDNA and Local Regrowth/Distant Mets in Nonoperative Rectal Cancer

ForMSS stage I-III rectal ca, cCR/nCR after NAT, on nonoperative management

TL;DRctDNA sensitivity for local regrowth only 41% (12/29), specificity 94%; distant mets sensitivity 74%, specificity 97%.

Why it mattersRadiation oncology

The number that gates watch-and-wait practice is 41% sensitivity for local regrowth (12/29 samples): a negative ctDNA cannot license lengthening MRI or endoscopy intervals in an organ-preservation protocol. Specificity 94% means a positive result is worth acting on, and ctDNA+ at regrowth tracked ypT3-4 in 6/8 vs 3/14 (p=0.01).

Monday clinic

In a stage I-III MSS rectal pt in watch-and-wait after TNT or CRT, a positive ctDNA supports intensified assessment for regrowth or distant disease; a negative result does not support relaxing endoscopic or MRI surveillance intervals.

ctDNA and Local Regrowth/Distant Mets in Nonoperative Rectal Cancer
OutcomeSensitivitySpecificityAccuracy
For local regrowth12 / 29 (41%)480 / 509 (94%)492 / 538 (91%)
For distant metastasis31 / 42 (74%)611 / 627 (97%)642 / 669 (96%)
11 details

Single-institution cohort from MD Anderson's INTERCEPT program, 2020-2024, N=110, with serial tumor-informed ctDNA during nonoperative management. Median follow-up 25 months (IQR 18-37).

Microsatellite stable stage I-III rectal adenocarcinoma achieving cCR or near-cCR after neoadjuvant therapy and managed nonoperatively. Median age 56; baseline cT3 in 72 and cT4 in 16; 69 received TNT, 41 CRT or chemotherapy alone.

Local regrowth and distant metastasis by longitudinal ctDNA status, by first post-NAT ctDNA (within 180 days), and per-sample accuracy for an event within ±90 days of each draw. Salvage-surgery pathology by ctDNA status at regrowth.

Twenty-three pts (21%) had local regrowth and 12 (11%) distant metastasis. Ever-positive ctDNA separated both curves (log rank p=0.0002 for regrowth, p<.0001 for metastasis). The per-sample table carries the operating characteristics.

MSS stage I-III rectal cancer in cCR/nCR on nonoperative management with tumor-informed ctDNA surveillance
Does not represent MSI-high disease, pts taken to upfront TME, or ctDNA assays other than the tumor-informed exome platform used here.

Per-sample analysis pools 669 draws from 110 pts without accounting for repeated measures, so the confidence around 41% is softer than the denominator suggests. The first-post-NAT comparison rests on n=12 evaluable pts as reported in the source. No comparison against MRI or endoscopy, the tests ctDNA would have to beat.

The asymmetry between local (41%) and distant (74%) sensitivity is the informative result: intraluminal regrowth from a small residual burden sheds too little DNA to be caught reliably, while metastatic disease does. That biology argues for ctDNA as a distant-recurrence tool layered onto, not substituted for, luminal surveillance.

Retrospective single-center cohort; per-sample analysis treats 669 draws from 110 pts as independent. No head-to-head against MRI/endoscopy surveillance.

  • Does ctDNA add anything over MRI plus endoscopy in NOM surveillance?
  • Can draw timing or a lower assay threshold raise local regrowth sensitivity?
  • Does ctDNA-triggered restaging improve salvage outcomes?
📚 Sources · 🐦 1 tweet

The longer read

The interesting result here is not that ctDNA is prognostic, which is by now the expected finding in colorectal cancer, but that its performance splits so cleanly by site of failure. Sensitivity of 41% for local regrowth against 74% for distant metastasis, in the same patients, on the same assay, at the same draws, is close to a controlled comparison of what tumor-informed ctDNA can and cannot see. The most parsimonious reading is a shedding-burden effect: a small intraluminal regrowth in an irradiated rectum releases too little circulating DNA to clear the assay's threshold, while metastatic deposits do. If that is right, no amount of assay optimization at the current sensitivity floor rescues local surveillance, and the finding generalizes beyond this platform.

That matters because the decision this test is being marketed into is precisely the local one. Watch-and-wait is sustained by intensive proctoscopy and MRI at short intervals, and the burden of that schedule is what a blood test is meant to relieve. A test that misses more than half of local regrowths cannot relieve it. The 94% specificity points the other way: a positive result carries real information and is worth acting on, so the honest framing is a rule-in adjunct with essentially no rule-out capacity for the endpoint clinicians most want ruled out.

Two features of the analysis should move confidence downward, and one upward. Downward: the headline operating characteristics come from a per-sample analysis pooling 669 draws across 110 patients, which treats correlated observations as independent. Patients contributing many draws dominate; the interval around 41% is wider than the denominator implies. Also downward: the first post-NAT timepoint analysis, the one closest to a clinically usable single-test strategy, is reported in the source with only 12 evaluable patients, which is a signal to note rather than a result to build on. Upward: the salvage pathology finding, ypT3-4 in 6 of 8 ctDNA-positive versus 3 of 14 ctDNA-negative regrowths, is internally coherent with the shedding hypothesis and was not a foregone conclusion. It suggests the misses are not random noise but are concentrated in low-volume regrowth, which is the subset salvage surgery handles best.

What it would take for this to be wrong: if the missed regrowths were driven by draw timing rather than biology, tightening the schedule around clinical suspicion could lift sensitivity substantially, and the ±90-day event window used here would be the artifact. That is testable and this design does not test it. Likewise, the cohort is microsatellite stable and single-institution, and INTERCEPT patients are surveilled more intensively than most, which if anything biases toward detecting the regrowths ctDNA missed and makes the 41% a relatively favorable estimate rather than a pessimistic one.

For a clinician running an organ-preservation program, this does not change the surveillance schedule. It does argue for reading a positive result as a prompt to look harder, both luminally and systemically, and for resisting the inference that a string of negative results has earned the patient longer intervals between scopes.