Adenocarcinoma, NOS (Colorectal): The Default Diagnosis That Still Demands Rigor

What It Is and Where It Sits in the Classification

When a pathologist signs out a colorectal carcinoma without further qualification, the diagnosis is almost always adenocarcinoma, not otherwise specified (NOS). In the WHO 5th edition classification of digestive system tumours, this is a recognized subtype — the default category of colorectal carcinoma against which every named variant (mucinous, signet-ring, medullary, and the rest) is measured [1]. It is by a wide margin the most common malignant epithelial tumour of the large bowel, and most of what we intuitively picture when we say "colon cancer" is this entity.

Calling it the "default" carries a risk, though. The word invites a mental shortcut — as if NOS meant unremarkable or requiring less thought. The opposite is closer to the truth. Because this category is defined largely by exclusion (it is glandular adenocarcinoma that doesn't meet criteria for a special subtype), the meaningful clinical information often lives not in the H&E slide alone but in the molecular characterization that follows it. The morphology tells you what it is; the ancillary work tells you how it will behave and what therapeutic classes the patient may be eligible for.

How a Pathologist Recognizes Colorectal Adenocarcinoma, NOS

The diagnostic sequence runs morphology → immunohistochemistry → molecular, and that order matters because each step earns its place only when the prior one leaves a question unresolved.

Morphology. The two defining features are glandular differentiation and desmoplastic invasion [1]. On the slide, you see malignant epithelium arranging itself into glands — lumina lined by atypical columnar cells with the coarse chromatin and prominent nucleoli of colorectal-type malignancy — set within a reactive, fibrotic (desmoplastic) stroma that betrays true invasion. That desmoplastic response is what separates invasive carcinoma from an intramucosal or in-situ process, and it's worth saying plainly that the assessment of genuine stromal invasion is where inexperienced eyes most often stumble. Dirty necrosis within gland lumina is a familiar accompanying feature. For the great majority of cases, this is enough: the H&E is diagnostic on its own.

Immunohistochemistry. IHC is confirmatory rather than routinely necessary, and it earns its keep in two scenarios — a metastasis of unknown primary, or a poorly differentiated tumour whose lineage isn't obvious. The classic colorectal signature is CK20-positive, CDX2-positive, SATB2-positive, and CK7-negative [1]. That said, treat the panel as a pattern, not a checklist. None of these markers is perfectly sensitive or specific. CK7 is described as typically negative, but it can be focally positive, particularly in right-sided and high-grade tumours, so a little CK7 staining does not exclude a colorectal origin. CDX2 and SATB2 both mark caudal-type differentiation but are not unique to the colon — CDX2 appears in upper-GI and pancreaticobiliary adenocarcinomas as well. SATB2 tends to be most useful precisely when CDX2 or CK20 falter, for instance in poorly differentiated or MSI-high tumours where the conventional colorectal markers may weaken. The strength of the signature lies in the combination, read against clinical and radiologic context.

Molecular. This is where the sign-out stops being purely descriptive and starts being actionable. Two broad pathways underlie colorectal adenocarcinoma [1]. The chromosomal-instability (CIN) pathway, driven by the classic APC → KRAS → TP53 sequence, accounts for the large majority of sporadic cases. The alternative is the microsatellite-instability (MSI) / CpG-island methylator (CIMP) route, associated with the serrated pathway of tumourigenesis [1]. Determining which pathway a given tumour travelled — above all, whether it is mismatch-repair deficient — reframes prognosis and eligibility for particular therapeutic classes in a way morphology alone cannot.

Grading and Classification

Grading colorectal adenocarcinoma, NOS is deliberately simple. The WHO 5th edition uses a two-tier system based on gland formation [1]. Low-grade tumours (well to moderately differentiated) form glands in at least 50% of the tumour; high-grade tumours (poorly differentiated) form glands in less than 50% [1]. Collapsing the older four-tier scheme into two tiers improves reproducibility, which is no small thing — inter-observer agreement on grade has always been softer than we'd like, and a binary threshold at least gives us a defensible line.

Here is the single most important caveat in this entire entity, and it deserves emphasis: MSI-high tumours frequently look high-grade on morphology yet behave better than their appearance predicts [1]. They tend toward poor differentiation, mucin production, and a brisk lymphocytic response — features that would, on morphology alone, prompt a grim assessment. Grading by gland formation still applies, but MMR status reframes prognosis independent of what the glands are doing. In my experience this is the pitfall most worth internalizing early: the tumour that looks the worst under the microscope is sometimes the one carrying the more favourable biology.

Tumour budding — the presence of isolated single cells or small clusters at the invasive front — is increasingly recognized within the WHO framework as an adverse prognostic parameter, particularly relevant in the assessment of early-stage disease. It's an emerging area, and standardized counting has taken time to settle, but it's worth watching as a supplement to conventional grade.

Why the Distinction Matters Clinically

The reason we push every "default" adenocarcinoma through molecular characterization comes down to MMR/MSI status, which does two things at once.

First, prognosis. As noted, MSI-high tumours often present with alarming morphology but a comparatively favourable natural history. A pathologist who grades on glands alone and stops there gives an incomplete picture; the MMR result recalibrates the clinical expectation.

Second — and this is the practical downstream consequence clinicians rely on — identifying MSI-high or MMR-deficient tumours is the entry point to Lynch syndrome screening. A deficient result flags the possibility of an inherited predisposition, triggering the reflex workflow (further molecular testing and, where appropriate, germline evaluation and genetic counselling) that can protect not just the patient but their relatives. In this sense the humble MMR immunostain does double duty: it informs the individual tumour's biology and it opens a door onto hereditary risk. MMR/MSI status also determines eligibility for particular systemic therapy classes, though the specifics of any individual's treatment sit with the treating oncology team, not the report.

So the "default" diagnosis turns out to demand as much discipline as any named subtype — arguably more, precisely because its ordinariness tempts us to relax. The gland-forming morphology and desmoplastic invasion get you to the diagnosis; the two-tier grade adds a layer of prognostic texture; but it's the molecular step, the MMR status above all, where the clinically actionable information actually emerges. An adenocarcinoma, NOS worked up only to the H&E is a diagnosis half-finished.

References

1. WHO Classification of Tumours Editorial Board. Digestive System Tumours, 5th ed. Lyon: IARC; 2019. ISBN 978-92-832-4499-8.