Any non-glandular focus at the front predicted nodes in pT1 colorectal cancer

Design

Multicentre retrospective cohort with training and validation sets

Population

2,611 endoscopically or surgically resected pT1 colorectal cancers, six Chinese centres

Primary outcome

Lymph node metastasis

Effect

Peritumoral malignant cellular foci independently associated with node metastasis in both cohorts (P<0.05); budding and clusters were not consistently

Tumour budding and poorly differentiated clusters are both used to judge nodal risk after local excision of pT1 colorectal cancer, but they are separated by an arbitrary five-cell cutoff and reproducibility is imperfect. A six-centre Chinese study proposed a single category, peritumoral malignant cellular foci: any non-glandular focus of one or more tumour cells at the invasive front.

In 2,611 endoscopically or surgically resected pT1 cancers, neither budding nor poorly differentiated clusters was consistently independent for node metastasis across training and validation cohorts. Peritumoral malignant cellular foci was, in both, and across age, sex, site and BMI subgroups. It was also linked to more involved nodes and more strongly to apical (D3) nodes.

The abstract reports significance but not effect sizes, and the cohort is Chinese only. International guidelines still use budding grade, and this study does not change that. But the concept is simple, needs no counting threshold, and could be recorded alongside budding in reports now.

For surgeons the implication, if confirmed, is a clearer basis for recommending completion surgery after endoscopic resection.

• Continue to report tumour budding grade as guidelines require.
• Consider also recording whether any non-glandular tumour foci are present at the invasive front.
• Discuss pT1 cancers with high-risk front features at the multidisciplinary meeting before deciding on completion surgery.
• Report lymphovascular invasion, depth and margin status alongside front features.

Why it matters

It offers a cutoff-free alternative to budding counts that may predict nodes more consistently.

Don't overread it

No effect sizes are given in the abstract, and it has not been validated outside China.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.