Neuroendocrine Neoplasms of the Colon: Reading the Two Axes

A biopsy that forces the question

A core biopsy arrives from a right colon mass. Under the scope you see organoid nests and trabeculae, uniform cells with stippled "salt-and-pepper" chromatin, unmistakably neuroendocrine. Then the Ki-67 comes back at 28%. So where does this tumor sit? Is it a high-grade well-differentiated NET, or is it a neuroendocrine carcinoma? That question isn't academic. The two answers point to different biology, different molecular expectations, and a different clinical trajectory, and the proliferation index alone won't settle it.

This is the central puzzle of colorectal neuroendocrine neoplasms (NENs), and it rewards a careful walkthrough.

What it is, and where it sits

In the WHO 5th edition scheme, neuroendocrine neoplasms of the digestive system are recognized entities, distinct from conventional adenocarcinoma [1]. The category has an internal architecture built on two independent axes, and confusing them is the most common conceptual error I see in trainees.

The first axis is differentiation, a qualitative, morphology-driven judgment. Well-differentiated tumors are called neuroendocrine tumors (NETs); poorly differentiated ones are neuroendocrine carcinomas (NECs) [1]. The second axis is grade, a quantitative measure of proliferation. That's the Ki-67 and mitotic count.

Here's the point that resolves the opening biopsy: a NET can be high-grade (G3) and still be a NET. It is not automatically a NEC because Ki-67 crosses some threshold. NEC is a diagnosis rooted in poorly differentiated morphology, not in the proliferation number alone [1]. Sitting alongside these is the mixed neuroendocrine, non-neuroendocrine neoplasm (MiNEN), in which a neuroendocrine component coexists with a non-neuroendocrine component, typically adenocarcinoma, each comprising at least 30% of the tumor [1].

How a pathologist recognizes it

The diagnostic order matters, and it runs morphology → immunohistochemistry → molecular.

Morphology first. Neuroendocrine differentiation announces itself architecturally: organoid nests, trabeculae, rosettes, and that characteristic finely granular chromatin. Well-differentiated NETs show this organoid pattern with relatively monotonous cytology. NECs look different, sheets of poorly differentiated cells, often with the small-cell or large-cell appearance familiar from pulmonary NEC, brisk mitoses, and necrosis [1].

Immunohistochemistry confirms. The neuroendocrine phenotype is verified with the established markers synaptophysin and chromogranin [2]. INSM1, a nuclear neuroendocrine transcription factor, is increasingly used as an adjunct, though its diagnostic performance specifically in gastrointestinal neuroendocrine neoplasms is not yet established in validated accuracy data [1]. A Ki-67 stain is then run to establish the proliferation index, and mitoses are counted. IHC confirms lineage; it does not, by itself, assign differentiation, that's still a morphologic call.

Molecular last, and confirmatory. Poorly differentiated colorectal NECs share the molecular profile of high-grade neuroendocrine carcinomas elsewhere, with recurrent TP53 and RB1 alterations, the same co-loss that characterizes high-grade neuroendocrine carcinoma at other sites [1,2]. That signature can support a NEC diagnosis when the morphologic distinction from a G3 NET is difficult, precisely the situation posed by our 28% case.

Grading and classification

Well-differentiated NETs are graded G1, G2, or G3 using Ki-67 and mitotic count [1]. NEC is, by definition, a high-grade tumor and is not subdivided along the same NET grading axis [1]. MiNEN is classified by the proportions of its components, each of which must reach the 30% threshold to qualify [1]. The NET grading axis parallels what's used for gastroenteropancreatic and pancreatic NETs, a shared framework that makes the colorectal grading intuitive if you already work with pancreatic cases [1].

Why the distinction matters clinically

This is where getting the classification right earns its keep.

Differentiation and grade carry different clinical meaning, and treating them as interchangeable can misdirect care. A well-differentiated G3 NET and a poorly differentiated NEC may share a similar Ki-67, yet they are biologically distinct diseases, different natural histories, different molecular underpinnings, and eligibility for different therapeutic classes rather than any single regimen [1]. Misfiling a G3 NET as a NEC, or the reverse, can therefore steer a patient toward an entirely inappropriate line of management. That's not a subtle error; it's a categorical one.

The MiNEN designation guards against a related trap in both directions. Under-calling is the more insidious problem: a small nest of scattered neuroendocrine cells within an adenocarcinoma does not make a MiNEN. Unless the neuroendocrine component reaches that 30% bar, the tumor stays an adenocarcinoma with neuroendocrine differentiation [1]. Over-call it and you've relabeled a common cancer as a rare mixed entity. Under-recognize a genuine high-grade neuroendocrine component that does meet the threshold, and you've missed a clinically consequential population of tumor cells that behaves aggressively. The 30% rule exists precisely to impose discipline here, and I'd argue it's the single most misapplied number in this part of the classification.

The grading axis within NETs also carries a prognostic gradient, the ordering from G1 through G3 reflects increasing proliferative activity and, correspondingly, more aggressive behavior [1]. That gradient is the reason we count carefully rather than eyeball a rough estimate; the difference between a G1 and a G2 hinges on the index, and the index can move management.

A note on where this is heading

Return to the 28% biopsy. The morphology, organoid, monotonous, no sheeting or necrosis, argues for a G3 NET, and if TP53 and RB1 come back intact, that reading holds. The molecular data increasingly do this work at the margins, helping separate the well-differentiated high-grade tumor from true NEC when the H&E alone leaves room for doubt [1]. It's a evolving space, and the boundary between G3 NET and NEC is one to watch as the molecular evidence accumulates.

This article is educational and describes how these diagnoses are made; it is not treatment advice.

References

1. WHO Classification of Tumours Editorial Board. Digestive System Tumours, 5th ed. Lyon: IARC; 2019. ISBN 978-92-832-4499-8.
2. Beltran H, Prandi D, Mosquera JM, et al. Divergent clonal evolution of castration-resistant neuroendocrine prostate cancer. Nat Med. 2016;22(3):298, 305. PMID: 26855148.