Neuroendocrine and Small Cell Carcinoma of the Prostate: A Diagnosis That Rewards Suspicion

What it is, and where it sits in the classification

Most prostate cancer is acinar adenocarcinoma — glandular, androgen-driven, and graded by the familiar Gleason architecture. Neuroendocrine carcinoma is a different animal. It's a high-grade malignancy with small-cell or large-cell neuroendocrine morphology, and in the current framework it is recognized as a distinct subtype of prostatic carcinoma rather than a variant graded on the Gleason scale [1].

The entity comes in two clinical flavors, and the distinction matters. A minority arise de novo — small cell carcinoma present at diagnosis with no prior treatment. The larger and increasingly discussed group is treatment-emergent: neuroendocrine carcinoma that surfaces after a patient has been on androgen-deprivation therapy for an adenocarcinoma. This is not a coincidence of two tumors. It represents lineage plasticity — the adenocarcinoma, under the sustained pressure of androgen blockade, switches its differentiation program and re-emerges as a neuroendocrine tumor that no longer depends on the androgen receptor to survive [1]. The 5th edition treats this treatment-emergent phenotype as a clinically important entity in its own right [1].

How a pathologist recognizes it

The diagnostic sequence runs morphology, then immunohistochemistry, then molecular — in that order, and for good reason.

Morphology first. Small cell neuroendocrine carcinoma shows the classic features you'd expect from small cell carcinoma anywhere in the body: sheets of cells with scant cytoplasm, finely stippled ("salt-and-pepper") chromatin, nuclear molding, brisk mitotic activity, and often extensive necrosis. The large-cell variant shows bigger cells with more prominent nucleoli and organoid or trabecular growth, but the same high-grade, high-mitotic character. In practice, the first question a pathologist faces is a simple one: does this even look like the prostate adenocarcinoma I expected? When the answer is no — when the glands have vanished and the tumor has gone solid and blue — neuroendocrine carcinoma belongs on the differential.

Immunohistochemistry second. Morphology raises the suspicion; IHC confirms it. The neuroendocrine markers — synaptophysin, chromogranin, and INSM1 — are the workhorses here, and positivity for these supports neuroendocrine differentiation [1]. Just as informative is what disappears. The androgen receptor and PSA, reliable in ordinary acinar adenocarcinoma, are frequently lost in neuroendocrine carcinoma [1]. That loss is not a laboratory artifact. It's the immunohistochemical fingerprint of the same androgen-independence that let the tumor escape hormonal therapy in the first place.

Here's the practical pitfall, and it's a real one. If AR and PSA are both gone, how do you prove the tumor is prostatic at all rather than a small cell carcinoma that has spread in from somewhere else? This is where NKX3.1 earns its keep. NKX3.1 is often retained as a lineage marker even when AR and PSA are lost [1]. It's the anchor that ties the neuroendocrine tumor back to its prostatic origin — a genuinely useful point that trainees should commit to memory. When you see neuroendocrine morphology, absent AR/PSA, and retained NKX3.1, the story of a prostatic lineage switch becomes coherent.

Molecular last. Molecular findings support and contextualize what morphology and IHC have already suggested. The characteristic lesions are loss of RB1 and TP53, and these accompany the lineage plasticity that drives the transition from adenocarcinoma to a neuroendocrine phenotype [1]. Worth stating clearly for trainees: you don't need molecular confirmation to make this diagnosis on a routine specimen. The morphology-then-IHC sequence carries the diagnostic weight. Molecular work deepens understanding of the biology, and the evidence suggests these genomic events are central to the plasticity itself — but they follow the histology, they don't replace it.

Grading and how it reads in a report

This is where neuroendocrine carcinoma parts company with everything a pathologist does for ordinary prostate cancer. There is no Gleason score here. Gleason grading applies to acinar adenocarcinoma and its architectural patterns; neuroendocrine carcinoma is high-grade by definition and is not assigned a Gleason grade or Grade Group [1]. That's an important conceptual point for clinicians reading these reports. A neuroendocrine carcinoma diagnosis on a needle core doesn't carry a "3+4" or "4+5" — it carries a categorical statement of high-grade neuroendocrine malignancy.

In a mixed tumor — and mixed adenocarcinoma-plus-neuroendocrine tumors do occur — the report should describe both components. The adenocarcinoma portion can be Gleason-graded; the neuroendocrine portion is reported as such and cannot be folded into a Gleason number. A pathologist who blends the two into a single score has obscured the very finding the oncologist most needs to see.

The differential diagnosis worth thinking through

Because AR and PSA are so often lost, the differential widens beyond the prostate. Small cell carcinoma of the urinary bladder can extend into or abut the prostate, and metastatic small cell carcinoma of pulmonary origin can seed the region. Morphology alone won't separate these — small cell carcinoma looks much the same wherever it arises. This is precisely where retained NKX3.1 does its work, pointing toward a prostatic lineage when the more familiar prostate markers have gone silent [1]. Clinical and radiologic correlation matters too; a patient with a known adenocarcinoma treated with androgen deprivation, now presenting with a neuroendocrine tumor, tells a very different story than one with a bladder mass or a lung primary. The pathologist who reads the chart alongside the slide is the one who gets this right.

Why the distinction matters

Neuroendocrine carcinoma is aggressive, and the treatment-emergent form carries particular clinical weight because it explains a specific and frustrating scenario: a patient whose adenocarcinoma was controlled on androgen deprivation and then progressed, not because the old tumor grew, but because a new biology emerged [1]. Recognizing that shift changes the clinical conversation entirely. A tumor that has abandoned the androgen receptor won't be managed like one that still depends on it — and the pathologist's report is what flags that possibility. On therapy linkage, I'll say only what the classification supports: identifying neuroendocrine differentiation establishes a distinct diagnostic category with different biology. Specific treatment decisions belong to the treating clinician, not to the diagnostic line.

The ADT era has quietly rewritten how I read a rising-PSA-that-isn't biopsy. When a patient on long-term androgen blockade rebiopsies and the tissue comes back solid, blue, and glandless, the reflex should no longer be "poorly differentiated adenocarcinoma" — it should be a deliberate hunt for the lineage switch, synaptophysin and NKX3.1 in hand. The single most useful thing a pathologist can do is name it plainly in the diagnostic line, because that name is what tells the clinician the disease has changed underneath them.

References

1. WHO Classification of Tumours Editorial Board. Urinary and Male Genital Tumours, 5th ed. Lyon: IARC; 2022. ISBN 978-92-832-4512-4.