Prostate-Specific Membrane Antigen: Diagnosis & Therapy

Prostate-specific membrane antigen, better known as PSMA, has reshaped how doctors find and treat prostate cancer over the past decade. It is a protein that sits on the surface of prostate cancer cells in far greater numbers than on normal tissue, making it a nearly ideal homing beacon for both imaging scans and targeted radiation therapy. The ability to use the same molecule first to locate tumors and then to deliver treatment directly to them has given rise to one of the most successful examples of “theranostics” in modern oncology, a field that continues to expand rapidly.

What PSMA Actually Is

PSMA is a type II transmembrane glycoprotein, which means it is anchored in the cell membrane with its business end sticking out into the space around the cell. It has enzyme activity: it can break down certain folate compounds and cleave small peptides, linking it to glutamate metabolism and one-carbon metabolic pathways that cancer cells exploit for growth, energy balance, and even gene regulation.¹ Healthy prostate tissue produces some PSMA, but prostate cancer cells can express it at levels hundreds of times higher. That dramatic overexpression, combined with the fact that PSMA sits on the outside of the cell where drugs and imaging agents can reach it easily, is what makes it so useful clinically.

An important quirk is that PSMA expression can actually increase when patients receive standard hormone-blocking treatments (androgen receptor blockade). In preclinical work, tumors treated with hormone therapy showed up to a roughly twofold increase in uptake of a PSMA-targeting radiotracer compared with untreated controls.² This has practical implications: hormone therapy may inadvertently make PSMA-targeted imaging more sensitive and PSMA-targeted treatment more effective, a relationship researchers are actively exploring in combination strategies.

PSMA PET Scans for Diagnosis and Staging

The diagnostic side of PSMA theranostics relies on PET/CT scans, in which a small amount of a radioactive tracer that binds to PSMA is injected and then lights up wherever PSMA-expressing cancer cells are hiding. The most widely used tracers are gallium-68 (⁶⁸Ga) and fluorine-18 (¹⁸F) labeled compounds. These scans have proven substantially more accurate than older imaging methods for finding prostate cancer that has spread to lymph nodes or other sites.

In one study focused on detecting lymph node metastases, PSMA PET/CT achieved roughly 88% sensitivity and 97% specificity, with an overall diagnostic accuracy above 92%.³ When compared head-to-head with multiparametric MRI, PSMA PET/CT was far more sensitive for both extraprostatic extension (about 72% versus 47% for MRI) and lymph node involvement (about 92% versus 8% for MRI), and it matched histopathology findings for tumor location more closely as well.⁴ That said, the performance depends on the clinical setting. A single-center analysis of high-risk patients undergoing radical prostatectomy found a somewhat lower sensitivity of 63% for detecting positive lymph nodes, though specificity remained very high at 97% and overall accuracy was 83%.⁵ The take-home point is that PSMA PET is excellent at ruling disease in when it lights up and very good at ruling disease out when it does not, but small clusters of cancer cells can still slip below the detection threshold.

Finding Cancer That Comes Back

Where PSMA PET has arguably made the biggest clinical splash is in detecting prostate cancer recurrence. After surgery or radiation, doctors monitor a blood marker called PSA. When PSA starts rising again, the cancer has likely returned somewhere, but conventional imaging like CT and bone scans often cannot find it until PSA climbs fairly high. PSMA PET can spot recurrent disease at dramatically lower PSA levels.

A real-world study spanning seven years and two centers found that even among men with very low PSA values, about a quarter had suspicious lesions on PSMA PET, including lymph node metastases and bone metastases. The lowest PSA level at which a positive scan was recorded was just 0.02 ng/mL.⁶ A larger German multicenter analysis that pooled data from several PSMA tracers found a comparable overall detection rate of about 30% in patients with PSA at or below 0.2 ng/mL, with higher detection rates in patients whose original tumor had a more aggressive grade or who were already on hormone therapy.⁷

Finding disease this early matters because it can shift a patient from systemic therapy (treating the whole body) to targeted local treatment (radiation aimed at just the few spots where cancer has returned). That distinction can spare patients months or years of hormone therapy and its associated side effects.

How PSMA PET Changes Treatment Decisions

The scan’s practical value goes beyond just being more accurate. In an Australian prospective multicenter study, PSMA PET/CT led to a change in the planned management for more than half of all patients scanned. The effect was most striking in men with biochemical recurrence after initial treatment, where scans changed the management plan in about 62% of cases.⁸ A prospective U.S. trial confirmed this: PSMA PET prompted an intended management change in 68% of patients with recurrent disease, with almost half of those changes considered major, such as switching from watchful observation to targeted radiation or from local therapy to systemic treatment. The management change was actually carried out in about 78% of cases where follow-up data were available, and the scan also prevented roughly 150 other imaging tests that would otherwise have been ordered.⁹

From Imaging to Treatment

The theranostic concept is elegantly simple. The same small molecule that locks onto PSMA for a diagnostic scan can be attached to a different radioactive isotope, one that emits cell-killing radiation rather than just imaging photons. A patient who lights up on a PSMA PET scan is, in effect, pre-screened as likely to benefit from PSMA-targeted radioligand therapy, because the scan confirms the cancer expresses enough PSMA to be a viable target.¹⁰ This “see it, then treat it” loop is what gives the theranostic approach its name and its appeal.

Lutetium-177 PSMA Therapy

The most established PSMA-targeted treatment is ¹⁷⁷Lu-PSMA-617, now marketed as Pluvicto. It uses lutetium-177, a beta-emitting radionuclide. Beta particles travel a few millimeters in tissue, enough to kill a cluster of cancer cells while limiting damage to surrounding structures. The landmark VISION trial established this therapy in men with metastatic castration-resistant prostate cancer who had already been through chemotherapy and newer hormone agents. Adding ¹⁷⁷Lu-PSMA-617 to standard care roughly doubled the time before the cancer progressed on imaging (a median of about 8.7 months versus 3.4 months) and extended overall survival by about four months (a median of 15.3 months versus 11.3 months).¹¹

Subsequent work tried to move the therapy earlier in the disease course. The PSMAfore trial enrolled men with metastatic castration-resistant disease who had not yet received chemotherapy and whose cancer had progressed on one type of hormone therapy. Instead of simply switching to another hormone agent, half the patients received ¹⁷⁷Lu-PSMA-617. The radioligand therapy arm had a median progression-free survival of about 11.6 months compared with roughly 5.6 months for the hormone-switch arm, and the rate of serious side effects was actually lower in the radioligand group.¹² These results are pushing PSMA radioligand therapy into earlier lines of treatment, before chemotherapy becomes necessary.

Alpha Particle Therapy With Actinium-225

Lutetium-177’s beta particles are effective, but they scatter their energy over a relatively wide path. Alpha particles, by contrast, are much heavier and dump a large amount of energy over an extremely short distance, which makes them potent single-cell killers. Actinium-225 (²²⁵Ac) linked to the same PSMA-617 molecule is being studied for patients whose disease has progressed even after lutetium therapy. In one early series of 26 such patients, about two-thirds achieved a PSA decline of 50% or more, a meaningful biochemical response in a population with very few remaining options. Median overall survival was about 7.7 months.¹³ The biology behind this advantage is intuitive: research has shown that PSMA-mediated internalization of the tracer into the cell strongly enhances the killing power of short-range alpha radiation, while having little additional effect on longer-range beta radiation, which does its damage whether or not it gets pulled inside the cell.¹⁴ The main challenge with actinium-225 is supply: it is difficult and expensive to produce, and scaling up for widespread clinical use remains an ongoing hurdle.

Side Effects and Salivary Gland Toxicity

PSMA is not exclusively a cancer protein. Healthy tissues including the salivary glands, lacrimal (tear) glands, and kidneys also express it at low levels. Because the radioligand homes to any PSMA it finds, these organs can receive some radiation exposure. The most common side effects of lutetium-177 therapy are fatigue, dry mouth, nausea, and blood count drops. A large analysis of the FDA’s adverse event reporting system found that blood-related problems, including anemia, low platelet counts, and broader bone marrow suppression, were the most frequently flagged serious events. Part of this may stem from the cancer itself: prostate cancer commonly infiltrates the bone marrow, blurring the line between normal tissue and tumor. When the radioligand irradiates tumors embedded in bone marrow, nearby healthy marrow cells get caught in the crossfire.¹⁵

Dry mouth is the signature nuisance side effect. Researchers have tried to protect the salivary glands by cooling them with ice packs during treatment, hoping that reduced blood flow would lower tracer delivery. During diagnostic scans with gallium-68, cooling one side of the face did modestly reduce PSMA uptake in the cooled parotid gland by about 14-15%.¹⁶ But when the same approach was tested during actual lutetium-177 therapy, it failed: imaging at 4 and 24 hours after treatment showed no significant difference in radiation uptake between the cooled and uncooled glands.¹⁷ The therapeutic radionuclide circulates for much longer than a diagnostic tracer, so any brief reduction from cooling gets washed out over hours.

A more creative approach involves injecting botulinum toxin directly into the salivary glands to suppress their metabolic activity. In one case report, a single injection into one parotid gland reduced PSMA uptake in that gland by about 64% on a follow-up scan 45 days later.¹⁸ This is a provocative proof of concept, but a single patient is far from proof of efficacy. Larger trials are needed before botulinum toxin injection could become standard practice.

Predicting Who Will Respond

Not every patient benefits equally from PSMA radioligand therapy, and the PSMA PET scan itself offers clues about who is most likely to respond. Researchers measure how intensely tumors light up on the scan (SUV values) and how much total disease volume is visible (PSMA tumor volume). A meta-analysis found that for each unit increase in mean tumor uptake intensity, the risk of death dropped by about 12% and the odds of achieving a meaningful PSA decline more than doubled. Conversely, larger total tumor volumes predicted worse outcomes, with each doubling of tumor volume associated with roughly a 16% increase in mortality risk.¹⁹ An earlier bicentric analysis similarly identified PSMA tumor volume and overall tumor tracer load as independent negative predictors of survival.²⁰

In plain terms, the ideal patient is one whose individual tumor spots glow brightly on the scan (lots of PSMA per cell, meaning good drug delivery) but whose total disease burden is not enormous (fewer cells to kill). The VISION post-hoc analysis reinforced this: higher SUV values on the pretreatment scan correlated with better PSA response rates in the treatment arm, though no single imaging parameter reliably predicted who would gain a survival advantage versus standard care alone.²¹ This is an area of active research: the goal is to move from one-size-fits-all dosing toward genuinely personalized treatment guided by what the scan reveals.

Personalized Dosing and Dosimetry

Currently, most patients receive a fixed amount of lutetium-177 at each cycle, regardless of their body size, tumor burden, or kidney function. Yet studies consistently find large differences from one patient to the next in how much radiation dose their tumors and organs actually absorb. One dosimetry study explicitly noted these wide inter-individual variations and called for patient-specific treatment planning rather than a uniform protocol.²² A more recent assessment found that the doses organs at risk actually received were within acceptable ranges and suggested that increasing administered activity may be feasible without unacceptable toxicity in many patients.²³ Moving toward dose optimization, where the amount of radioligand is tailored based on individual imaging and dosimetry rather than a flat dose, could improve both efficacy and safety.

When PSMA-Targeted Therapy Stops Working

Some prostate cancers lose their PSMA expression as they evolve under treatment pressure, rendering PSMA-targeted approaches ineffective. One well-described escape route is neuroendocrine differentiation, in which the cancer transforms into a subtype that no longer looks or behaves like typical prostate cancer. This neuroendocrine state is characterized by a loss of PSMA, making the tumor invisible on PSMA PET and unreachable by PSMA radioligand therapy.²⁴ Research on circulating tumor cells has shown that the relationship between androgen receptor signaling and PSMA expression is complex: some molecular patterns associated with advanced resistance, such as loss of a particular androgen receptor splice variant, may also correlate with reduced PSMA levels. This means clinicians need to maintain a high index of suspicion for neuroendocrine transformation in patients whose PSMA scans become unexpectedly dim despite rising PSA, and they should have alternative diagnostic tools and treatment pathways ready.

PSMA Beyond Prostate Cancer

PSMA is not limited to the prostate. It turns up in the blood vessels that feed many types of solid tumors, an expression pattern known as tumor-associated neovasculature. In renal cell carcinoma, for example, PSMA staining has been found in the neovasculature of the majority of kidney tumors, particularly the clear cell subtype, though the protein was not present in the tumor cells themselves.²⁵ A broader survey found moderate to strong PSMA staining in the neovasculature of over 80% of bladder, ovarian, and pancreatic carcinoma specimens, and in a sizable fraction of glioblastoma, melanoma, and non-Hodgkin lymphoma samples.²⁶ Because the PSMA in these tumors sits on blood vessels rather than on cancer cells themselves, the therapeutic strategy would differ from prostate cancer: the goal would be to starve tumors by destroying their blood supply rather than killing individual cancer cells. Early-phase clinical trials exploring this concept are underway.

Immunotherapies Targeting PSMA

Radioligand therapy is not the only way to weaponize PSMA as a target. Two classes of immune-based therapy are in clinical development: bispecific antibodies that grab both PSMA on the tumor and CD3 on a T cell, physically dragging immune cells to the cancer, and CAR-T cells engineered to recognize PSMA. Both approaches have shown some antitumor activity in early trials, though challenges remain, including cytokine release syndrome (a systemic inflammatory reaction), “on-target, off-tumor” toxicity from PSMA in healthy tissues, and difficulty sustaining immune responses inside the immunosuppressive environment that prostate tumors create.²⁷

One bispecific antibody in development, CC-1, has shown some intriguing early results. In a first-in-human trial involving 28 heavily pretreated men with metastatic castration-resistant prostate cancer, nearly all patients who received the target dose saw rapid drops in PSA, with reductions of up to 62% from baseline. The most common side effect was cytokine release syndrome, seen in 86% of patients but capped at moderate severity.²⁸ Whether these PSA declines translate into longer survival remains to be seen, but the signal is encouraging enough that larger trials are planned. If immune-based PSMA therapies mature alongside radioligand approaches, clinicians may eventually be able to sequence or combine them, attacking the same target through entirely different biological mechanisms.