Eighteen European centers assembled 179 patients whose endoscopically resected T1 colorectal cancers showed deep submucosal invasion and no other high-risk feature on the pathology report. Among the 124 who went on to surgery, 16 had lymph node metastasis, 12.9 percent. Almost all of that nodal disease sat in tumors invading 2000 micrometers or deeper.
- Design: retrospective series across 18 European centers, patients treated 2009 through 2022.
- Population: 179 patients, en bloc endoscopic resection of T1 colorectal cancer, isolated submucosal invasion deeper than 1000 micrometers.
- Arms: 124 underwent additional surgery, 55 went to surveillance.
- Node positivity: 16 of 124 surgical specimens, 12.9 percent (95% CI 7.7 to 20.4).
- Depth split: 1 of 50 below 2000 micrometers (2.0 percent) against 15 of 74 at or above it (20.3 percent).
- Surveillance arm: 2 of 55 recurred locally or distantly, 3.6 percent (95% CI 0.4 to 12.6).
A Cohort Built by Subtraction
Eligibility ran the other way from most T1 colorectal series. Investigators kept only the patients whose reports carried submucosal invasion beyond 1000 micrometers and nothing else, excluding anyone whose specimen also showed one of the accompanying high-risk findings that ordinarily drive the decision. That isolates the variable the paper set out to test, and it also removes any internal comparison: the series carries no arm with lymphovascular invasion, poor differentiation or high-grade tumor budding to measure the depth criterion against. The abstract does not list which features had to be absent, and it reports invasion as a measured depth in micrometers rather than by Kikuchi or Haggitt level.
Recruitment spanned 2009 to 2022 and reached beyond France into Spain, Italy, Belgium and Portugal, with contributing units at Erasmus Hospital in Brussels, University Hospitals Leuven, Hospital Universitario de La Princesa in Madrid, Tor Vergata in Rome and Centro Hospitalar Universitario de Sao Joao in Porto.
Two Thousand Micrometers Split the Risk
One of the 50 surgical patients whose invasion measured under 2000 micrometers had a positive node, 2.0 percent. Among the 74 at or beyond that depth, 15 did, 20.3 percent. Exploratory multivariable analysis returned two associations, depth at or above 2000 micrometers (odds ratio 3.59, 95 percent CI 1.14 to 13.71) and colonic versus rectal location (OR 3.63, 95 percent CI 1.07 to 16.77). The intervals are wide enough on both to leave the effect size unsettled, and the authors label the analysis exploratory.
Surveillance Patients Were Never Staged Surgically
The surveillance group produced two recurrences, local or distant, across 55 patients, 3.6 percent with a confidence interval running from 0.4 to 12.6. That figure and the 12.9 percent nodal rate are not directly comparable, because the two groups were separated by clinical judgment rather than randomization and the surveillance patients were never staged surgically. Barthet and colleagues put the conclusion in terms that stop short of a recommendation.
Isolated deep submucosal invasion in T1 CRC was associated with a non-negligible rate of LNM in this real-world cohort. In exploratory analyses, SID ≥2000 µm was associated with LNM.
Who Assembled the Series
First author Lucas Barthet and senior author Mathieu Pioche both work in the gastroenterology and endoscopy unit at Hopital Edouard Herriot in Lyon, which supplied the largest block of the 33-name author list. Pioche is a co-editor of Endoscopy, disclosed in the paper's conflict of interest statement; the remaining authors declare none. Thieme published the article online ahead of print on August 21 and serves it to subscribers only, blocking automated access outright, so the figures here come from the structured abstract and the author affiliations held by PubMed and Crossref. The full report stayed out of reach.
Why This Matters to the APO|APE Reader
Two pathologists sit on the author list, Valerie Hervieu and Tanguy Fenouil of the pathology unit at Hopital Edouard Herriot, and the finding lands on their side of the workflow. A synoptic report that records submucosal invasion as a yes or no against a single threshold cannot support a decision made at 2000 micrometers, so the measured depth has to survive into the report as a number. Services already recording it that way can re-interrogate their own T1 archive against this cohort without touching a slide. The rest would be back at the microscope with a micrometer scale, one case at a time.
