Most imaging modalities that have made their way into the clinic are prognostic, at least to some extent. The novelty of PSMA PET imaging is its sensitivity. It is very sensitive, even in the early stages of disease. Further, with the maximum intensity projection views, it becomes almost impossible to miss something that is that bright on PET. 

Currently, the clearest indication for PSMA PET imaging is for patients who are in the biochemical relapse state after primary local therapy. When disease is present outside the pelvis, those patients are less likely to benefit from local therapy alone. However, individuals with completely negative imaging results, including in the prostate and prostate bed, are likely to have the highest chance of a cure with local salvage therapy with a low burden of disease and should not be denied such treatment. 

In my own practice, I tell patients with rising prostate-specific antigen levels after surgery and/or radiation that I will get a PSMA PET scan, and I let them know in advance that we hope to see nothing on the scan. I usually ask if they want to see the scan, similar to what we do with bone scans. For those who do have a small number of sites of uptake, the practice of focal therapy for “oligometastatic” disease is sometimes undertaken in the absence of level-1 data. The preference is for this to be done in the context of a clinical trial. If a trial is not available and metastases are going to be treated, then PSMA PET should be utilized since, based on the ORIOLE trial, treating only disease that is detected on conventional imaging may result in missing other sites of metastases and may lead to poorer outcomes.

In the context of PSMA-targeted radionuclide therapy for metastatic castration-resistant prostate cancer, it is not too surprising that there is a predictive nature to PSMA PET. This is supported by both prospective and retrospective data. The TheraP trial enrolled patients who were very highly PSMA positive (defined as a maximum standardized uptake value [SUVmax] of ≥20 at a site of disease and an SUVmax of >10 at sites of measurable disease ≥10 mm) with no fluorodeoxyglucose (FDG) PET–positive lesions. At this year’s symposium, Buteau et al reported a new analysis from the TheraP trial in which they evaluated a mean SUV (SUVmean) of greater than or equal to 10 on PSMA PET as a predictive biomarker for response to lutetium-177 (¹⁷⁷Lu)–PSMA-617 vs cabazitaxel (abstract 10). A PSMA PET SUVmean of greater than or equal to 10 was predictive of a higher likelihood of a favorable response to ¹⁷⁷Lu-PSMA-617 compared with cabazitaxel, while a high volume of disease on FDG PET was linked to a worse prognosis, regardless of randomly assigned treatment. 

I think that having multiple imaging modalities is always going to give you the most information. In cases where there is discordance between the FDG PET and the PSMA PET, most lesions are visible on standard cross-sectional imaging. At this point, FDG PET probably adds a little bit—but not a lot—when we take a very rigorous view of imaging; however, FDG PET does make some lesions easier to spot, and it has been shown in multiple studies that FDG PET positivity carries independent prognostic value.

With regard to patient selection based on PET imaging, I would note that this is just the beginning, and the knowledge base is still expanding. The more data we collect, the more we will be able to answer questions related to patient selection. It is not a foreign concept to treat 1 compartment rather than targeting all of the disease. For instance, if a patient has a 3-cm lymph node, along with many symptomatic bone lesions, radium-223 could be given to target what we believe to be the most important compartment (the bone), even if it ignores soft tissue metastases. Something similar could be reasonable with PSMA-targeted radionuclide therapy. A patient with 8 PSMA-positive lesions and 2 PSMA-negative lesions might not be the most optimal candidate for PSMA-targeted radioligand therapy, but, particularly in the setting of no other standard or investigational options, I think that imaging-based optimization becomes less important, and I would have no problem offering that treatment to such a patient.

Finally, on the subject of PET tracers, a new class of PSMA-targeting agents is being studied and could be of interest, including leading compound ¹⁸F-rhPSMA-7.3 (abstract 9). On first blush, one might ask: Do we need a third tracer? Although I do not have a concrete answer, if there is the potential for low urinary excretion, with less prominence of the ureters and bladder in the imaging, this might facilitate the visualization of disease in the pelvis. From my perspective, having more options tends to be preferable to the alternative. In the future, as additional diagnostics and therapeutics become available,* there will be a place for blood tests, tissue biopsy, and imaging as complementary studies. The more biomarker assessment modalities we have that will tell us something about the disease biology or prognosis, the better.

*Addendum: On March 23, 2022, subsequent to this interview, the US Food and Drug Administration (FDA) approved lutetium Lu 177 vipivotide tetraxetan (formerly referred to as ¹⁷⁷Lu-PSMA-617) for the treatment of progressive PSMA-positive metastatic castration-resistant prostate cancer. The FDA also approved a complementary gallium-68–based agent indicated for PET of PSMA-positive lesions.