For a man with a rising prostate-specific antigen (PSA) after definitive treatment, the first fork in the road is now an imaging result.

“There are really two scenarios in there,” said Dr. Michael Morris, Prostate Cancer Section Head in the Genitourinary Oncology Service at Memorial Sloan Kettering Cancer Center.

“One is they get a Prostate-specific membrane antigen PSMA PET, and it’s negative. The other is that it’s a PSMA PET, and it’s positive.”

“Prostate cancer, the appropriate treatment for it is dependent on your understanding of where the disease is,” he told SurvivorNet. “It really changes the treatment plan for many patients.”

A University of Toronto meta-analysis of 34 studies, including 3,680 men with biochemical recurrence, found PSMA PET was positive in 68.2% and changed management in a pooled 56.4%. Most of these studies measure intended management changes, not whether those changes improved survival.

Track One: PSMA PET-Negative, High-Risk Biochemical Recurrence

For the patient with a pure biochemical relapse, Dr. Morris points to one trial. “That’s the only trial that has established the role of the sort of patient who is just a pure biochemical relapse with no other evidence of disease. So that’s the EMBARK trial.”

EMBARK randomized 1,068 men with high-risk biochemical recurrence (PSA doubling time of 9 months or less) to enzalutamide plus leuprolide, leuprolide alone, or enzalutamide monotherapy. Metastasis-free survival improved with the combination (HR 0.42) and with monotherapy (HR 0.63) versus leuprolide alone.

“The best results in terms of cancer outcomes are with dual therapy, ADT and enza, but enza is approved as monotherapy in that context as a castration sparing approach,” Dr. Morris said.

The final overall survival analysis, published in the “New England Journal of Medicine” in October 2025, confirms that ranking. Eight-year OS was 78.9% with the combination vs 69.5% with leuprolide alone (HR 0.60; 95% CI 0.44–0.80). Monotherapy reached 73.1%, which was not significantly different from leuprolide alone (HR 0.83; 95% CI 0.63–1.10). The FDA label allows enzalutamide with or without a GnRH analog.

One caveat shapes how this applies in the PSMA era. EMBARK defined nonmetastatic disease by conventional imaging, not PSMA PET. Many EMBARK patients would likely have had PSMA-visible disease. A PSMA-negative patient is therefore an extrapolation from the trial population, not the population itself.

Track Two: PSMA PET-Positive Disease

When the scan lights up, the next step is molecular.

“Genomically profile the patient, see what those results are, and that’s when you get back to the decision-making of triplet therapy versus doublet therapy,” Dr. Morris said. “And if triplet therapy, then are we talking about a PARP inhibitor? Are we talking about capivasertib… or Pluvicto or chemotherapy?”

All four options are now on the table in hormone-sensitive disease. Niraparib plus abiraterone is approved for BRCA2-mutated mCSPC.

Capivasertib plus abiraterone is approved for PTEN-deficient disease, identified by an FDA-authorized test.

Pluvicto plus an ARPI is approved for PSMA-positive disease.

Chemotherapy-based triplets are also established. In ARASENS, published in the New England Journal of Medicine, adding darolutamide to ADT and docetaxel significantly improved overall survival versus placebo plus ADT and docetaxel in mHSPC, along with key secondary endpoints.

The same caveat applies in reverse. The trials behind these triplets enrolled patients with metastases on conventional imaging; PSMAddition required this explicitly. Disease seen only on PSMA PET sits between EMBARK and the mHSPC trials, and no phase 3 trial has defined its treatment.

The Next Biomarker: ctDNA

“It really does seem that ctDNA is a very promising biomarker for likely to benefit from theranostics more so than other biomarkers,” Dr. Morris said. “And a decline in ctDNA to unmeasurable levels is highly prognostic for a good clinical outcome.”

The strongest data come from TheraP, a randomized phase 2 trial of Lu-PSMA-617 versus cabazitaxel in docetaxel-pretreated mCRPC.

In an exploratory analysis, men with ctDNA fraction below 2% had significantly higher odds of PSA response to Lu-PSMA than to cabazitaxel according to data published in the Journal of Clinical Oncology. Higher ctDNA fraction predicted shorter PFS on Lu-PSMA but not on cabazitaxel (interaction p=0.035). The predictive value was additive to PSMA SUVmean rather than a replacement for it.

A 2026 review concluded that ctDNA evidence in mCRPC remains predominantly prognostic rather than definitively predictive, according to research published in “Frontiers in Molecular Medicine.”

Why Earlier Is the Direction of Travel

“In prostate cancer, we have found that every time we move therapy earlier, the clinical gains that we achieve are more significant,” Dr. Morris said.

The pattern is broadly true for ARPIs, as EMBARK’s combination-arm survival benefit in biochemical recurrence shows.

However, moving treatment earlier in the sequence isn’t always the answer. Enzalutamide monotherapy in EMBARK did not significantly improve OS, as data published in the New England Journal of Medicine confirm.

Overall survival for Pluvicto in hormone-sensitive disease remains immature (HR 0.80; 95% CI 0.63–1.01).

Remaining Questions Physicians Are Still Trying to Answer

The open questions are how to treat PSMA-only disease, whether ctDNA can be validated as a selection tool in hormone-sensitive disease, and whether earlier radioligand therapy translates into longer survival.