PSMA Expression as a Theranostic Biomarker in Prostate Cancer

What the test measures

Prostate-specific membrane antigen (PSMA) is a transmembrane glycoprotein — a folate hydrolase, despite the name — that sits on the surface of prostate epithelial cells and becomes markedly overexpressed as prostate cancer dedifferentiates and progresses. That upregulation is what makes PSMA useful twice over: as an imaging target and as a delivery address for radioligand therapy. This dual role is the essence of a theranostic biomarker — the same molecular feature that lets you see the disease also lets you treat it, and the diagnostic image becomes the eligibility gate for therapy.

Two things are worth keeping straight from the outset. PSMA expression can be assessed in vivo by PET imaging, which is how it's used in routine clinical decision-making, or in situ on tissue by immunohistochemistry, which currently remains largely a research and adjunct tool. These are related but not interchangeable readouts, and conflating them causes real confusion at the multidisciplinary table.

How it's tested

PSMA PET imaging is the clinically actionable assay. A radiolabeled small-molecule ligand that binds the PSMA active site is injected, and PET-CT (or PET-MRI) then maps where that ligand concentrates. Tumor deposits that take up tracer avidly light up as "PSMA-avid" disease. The specimen, in effect, is the patient — this is whole-body, in vivo assessment, which is one of its great strengths, because it captures disease at sites a single biopsy never could.

Immunohistochemistry on formalin-fixed, paraffin-embedded (FFPE) tissue offers a complementary, cellular-resolution view of PSMA protein on the tumor membrane. It's valuable for correlative and investigational work, but it isn't the standardized gate for therapy the way PET is.

Scoring, in practice, comes down to whether disease is PSMA-avid on PET. The reader integrates uptake intensity relative to reference organs and the distribution of avid lesions across the body. Preanalytically, PET is sensitive to tracer dose, uptake time, and — importantly — to the agent used, since different PSMA ligands have somewhat different biodistribution. A clean, standardized acquisition protocol matters if the scan is going to carry therapeutic weight.

What each result state means

PSMA-positive (PSMA-avid disease): tumor sites concentrate the tracer, confirming PSMA is expressed at a level detectable in vivo. In the metastatic castration-resistant setting, that's the finding that opens the door to PSMA-targeted therapy.

PSMA-low/negative: little or no tracer uptake at known sites of disease. This is the finding that practically reshapes the conversation — when a scan comes back with disease that's radiographically evident but not PSMA-avid, the clinician has to explain that a targeted radioligand won't reliably reach those deposits, and attention shifts to other options. It's one of the more common places where imaging and clinical expectation diverge.

What it determines for treatment eligibility

The predictive value of PSMA PET was established most influentially by the VISION trial, which studied lutetium-177–PSMA-617 in men with PSMA-positive metastatic castration-resistant prostate cancer (mCRPC) who had already progressed on other therapies [1]. Enrollment required PSMA-avid disease on PET — meaning the biomarker didn't just describe the tumor, it defined the treatable population. In that trial, adding the radioligand to standard care improved outcomes in the selected group [1].

So the eligibility logic is straightforward to state and important to frame correctly: PSMA positivity on PET informs eligibility for the PSMA-targeted radioligand therapy class in appropriately selected mCRPC. It does not, by itself, direct which agent a given individual receives, nor is a positive scan a recommendation to any patient. It's a gate — a necessary molecular qualification — not a prescription. PSMA PET additionally serves a staging role, refining where disease is before those therapeutic questions even arise.

Caveats and what's evolving

The most clinically consequential caveat is intratumoral and intertumoral heterogeneity. PSMA expression isn't uniform — one metastasis can be brightly avid while another in the same patient is barely detectable, and expression can shift as tumors dedifferentiate or transform toward neuroendocrine phenotypes that downregulate PSMA entirely. A scan reported as broadly PSMA-avid may still harbor pockets of non-avid disease that the radioligand can't address, which is one reason discordant or mixed-avidity scans generate so much discussion. How next-generation eligibility frameworks should handle this — for instance, whether a threshold burden of non-avid disease ought to disqualify or merely qualify a patient — remains genuinely unsettled.

A second open question is threshold definition. What counts as "PSMA-positive" enough to predict benefit isn't a single universally fixed number; it depends on uptake intensity, the extent of avid versus non-avid disease, and the specific criteria a trial or reading protocol applied. The field is still refining these cutoffs, and the evidence suggests that a binary positive/negative label understates the biological gradient underneath. This mirrors debates seen with other predictive markers where a "low" category sits awkwardly between positive and negative — the boundaries are pragmatic, not natural.

Finally, the imaging-versus-tissue distinction deserves ongoing attention. PET reports functional expression across the whole body; IHC reports protein at cellular resolution in one sampled focus. When they disagree, neither is simply "wrong" — they're answering slightly different questions, and harmonizing them is active work.

The frontier now is whether PSMA-guided approaches move earlier — into hormone-sensitive or oligometastatic disease — where the biology of PSMA expression may differ from the heavily pretreated castration-resistant population in which it was first validated. If heterogeneity is the central problem, the next advances may come less from a better tracer and more from better ways to quantify, and act on, the fraction of a patient's disease that PSMA actually reaches.

References

1. Sartor O, de Bono J, Chi KN, et al. Lutetium-177–PSMA-617 for Metastatic Castration-Resistant Prostate Cancer (VISION trial). New England Journal of Medicine. 2021. PMID: 34161051; DOI: 10.1056/NEJMoa2107322.