Post-VISION Real World Experience in the US: How is PSMA RLT Being Used? "Presentation" - Matthew Rettig
April 14, 2025
At the 2025 UCSF-UCLA PSMA Conference, Matthew Rettig discusses real-world PSMA radioligand therapy usage in the VA healthcare system. He highlights VA patients are older with more comorbidities than VISION trial participants, with only half having prior taxane treatment due to VA-specific pathways. Veteran survival after lutetium is only 9-10 months, likely from competing mortality risks. Dr. Rettig reviews conflicting genomic data affecting outcomes and emphasizes the need for better predictive biomarkers and earlier treatment approaches.
Biography:
Matthew Rettig, MD, Professor of Medicine and Urology, David Geffen School of Medicine at UCLA, Chief, Division of Hematology-Oncology, VA Greater Los Angeles Healthcare System, Los Angeles, CA
Biography:
Matthew Rettig, MD, Professor of Medicine and Urology, David Geffen School of Medicine at UCLA, Chief, Division of Hematology-Oncology, VA Greater Los Angeles Healthcare System, Los Angeles, CA
Read the Full Video Transcript
Matthew Rettig: Good afternoon, everybody. My name is Matt Rettig. I'm a GU medical oncologist here at UCLA, and I also work at our affiliated VA a short distance away. And I was tasked with telling you about real-world usage of PSMA radioligand therapy. And I'll be focusing on some of the data that we have from the VA.
So just as background, we have many different agents that have been approved for mCRPC, and they can be categorized in this fashion. A lot of these categorizations are not absolute, and many of these agents can fall under different categories. When we look at the median OS benefit of the various FDA-approved agents for mCRPC, we see that the median OS, in absolute terms, is relatively modest—somewhere between two and a half and four and a half, five months.
We do have some recent approvals for PARP inhibitors and receptor pathway inhibitors as doublets where the absolute improvement is greater, and there is some suggestion, based on mechanism and preclinical data, that there can be synergy. The VISION study, of course, showed a median OS benefit of about four months.
By contrast, when we use many of these same agents in mCSPC, we see a much greater absolute benefit. So instead of around four months plus or minus, we're seeing a year or more than a year. And it adheres to the general theme in oncology that drugs that work in the most advanced state of a disease tend to have a greater impact—not always, but tend to have a greater impact when brought earlier in the disease.
So let me tell you a little bit about the VA. It's the largest integrated health care system in the United States. There are over nine million veterans who actively seek care within the VA. The VA patients are older than the US population at large, but our cancer patients have a similar age to that of the general population. About a third of the 40,000 new cases of cancer in the VA system are prostate cancer, and most veterans—well over 90%—are male.
Importantly, the VA has a unified electronic medical record, so all of the clinical data gets housed at a data warehouse, and that actually can be mined for research purposes through our research program called VINCI (the VA’s Informatics and Computing Infrastructure). There are over 1.2 million veterans with prostate cancer since 2000, and all of those patients’ data is available, including PSAs. We have well over 60 million PSAs that have been well annotated. There are images, including the DICOM images of PSMA scans, MRIs, and they’re being transferred to a central cloud space where they’re de-identified and can be used for research purposes.
So amongst our veterans who did receive PSMA radioligand therapy, they have a slightly older age than that of the VISION study. They are about four, four and a half years older. They had more comorbidities: their Charlson Comorbidity Index is 2.
And interestingly, only about half of them had a prior taxane. So the VA has its own clinical pathway—it’s a little bit different than something like the NCCN, where there’s a specific therapy that is recommended for a given state of a disease, including prostate cancer. And part of that pathway has been—up until today, when we heard about PSMAfore—to utilize PSMA radioligand therapy in the pre-chemotherapy setting for patients who either refused docetaxel or were not suitable for it. And the vast majority of the patients who’ve received lutetium have had an ARPI.
And this is the usage as we see it in the VA health care system since 2022. As you remember, the VISION study led to approval in March of 2022. There was a time to ramp up the usage of it. And as you may remember, there were some manufacturing issues at the outset. But we see that overall, the usage has been increasing by quarter over the last few years.
And in terms of survival, our veterans’ median survival after lutetium is only about 9 to 10 months, as compared to the VISION patients. Our veterans, as I mentioned, have more comorbidities. They’re more likely to suffer toxicity and have competing causes for mortality, which can be quite significant, even if they have mCRPC.
When we looked at a subset of patients who had tumor suppressor gene alterations, namely TP53, RB, and PTEN, we saw that there’s clearly a difference in the outcomes of these patients. Now, these are well-known prognostic markers, and this is really not telling us whether or not there is differential benefit of PSMA radioligand therapy in these patients. And here are some of the other factors, mainly the Charlson Comorbidity Index, in a multivariate Cox proportional hazards analysis that predict for outcome.
Some other studies have correlated genomics with outcome. Ohio State showed, interestingly, that tumor suppressor genes do not have an impact on survival in patients getting radioligand therapy. Similarly, for patients with homologous recombination deficiency, there’s no difference amongst those patients who have it versus those that don’t. But cell cycle gene amplification was clearly associated with a poor outcome from radioligand therapy.
Another study from the University of Washington, UCSF, and University of Colorado did show tumor suppressor gene alterations—one or more—is associated with poor outcomes, as were DNA damage repair gene alterations. And finally, a study from the Frankfurt Metastatic Cancer Database also demonstrated that BRCA alterations are associated with poor outcomes, but not tumor suppressor genes.
So you see that the data are kind of all over the place. It would be interesting if tumor suppressor gene–altered patients do as well as those that do not have those mutations, since it’s such a poor prognostic factor. And I think if that were the case—if that were demonstrated in a prospective study—that would suggest that a tumor suppressor gene alteration might benefit, in particular, from PSMA radioligand therapy.
So finally, some take-home messages. One, we’re seeing that usage of PSMA RLT is increasing with time. There are genomic factors that may be prognostic. The studies vary, as I just showed you, and we need better genomic predictive markers that should be incorporated into prospective clinical trials whenever feasible.
And we need novel approaches to improve survival in patients with mCRPC. We need newer approaches—ADCs, immunotherapies, leveraging what we know about the tumor microenvironment, and potentially coming up with synergistic agents such as PARPIs and ARPIs that have been used and recently approved.
And again, the concept of bringing the drugs that are effective for mCRPC to the mCSPC space can result in improved outcomes, and we look forward to the results of some of the studies, such as the PSMAddition trial, that have tested and finished accrual for patients with mCSPC testing the role of PSMA radioligand therapy.
So just a few acknowledgments. I really want to focus on a colleague, Martin Schoen, who really did a lot of the data extraction and analysis for the VA data. So thank you very much.
Matthew Rettig: Good afternoon, everybody. My name is Matt Rettig. I'm a GU medical oncologist here at UCLA, and I also work at our affiliated VA a short distance away. And I was tasked with telling you about real-world usage of PSMA radioligand therapy. And I'll be focusing on some of the data that we have from the VA.
So just as background, we have many different agents that have been approved for mCRPC, and they can be categorized in this fashion. A lot of these categorizations are not absolute, and many of these agents can fall under different categories. When we look at the median OS benefit of the various FDA-approved agents for mCRPC, we see that the median OS, in absolute terms, is relatively modest—somewhere between two and a half and four and a half, five months.
We do have some recent approvals for PARP inhibitors and receptor pathway inhibitors as doublets where the absolute improvement is greater, and there is some suggestion, based on mechanism and preclinical data, that there can be synergy. The VISION study, of course, showed a median OS benefit of about four months.
By contrast, when we use many of these same agents in mCSPC, we see a much greater absolute benefit. So instead of around four months plus or minus, we're seeing a year or more than a year. And it adheres to the general theme in oncology that drugs that work in the most advanced state of a disease tend to have a greater impact—not always, but tend to have a greater impact when brought earlier in the disease.
So let me tell you a little bit about the VA. It's the largest integrated health care system in the United States. There are over nine million veterans who actively seek care within the VA. The VA patients are older than the US population at large, but our cancer patients have a similar age to that of the general population. About a third of the 40,000 new cases of cancer in the VA system are prostate cancer, and most veterans—well over 90%—are male.
Importantly, the VA has a unified electronic medical record, so all of the clinical data gets housed at a data warehouse, and that actually can be mined for research purposes through our research program called VINCI (the VA’s Informatics and Computing Infrastructure). There are over 1.2 million veterans with prostate cancer since 2000, and all of those patients’ data is available, including PSAs. We have well over 60 million PSAs that have been well annotated. There are images, including the DICOM images of PSMA scans, MRIs, and they’re being transferred to a central cloud space where they’re de-identified and can be used for research purposes.
So amongst our veterans who did receive PSMA radioligand therapy, they have a slightly older age than that of the VISION study. They are about four, four and a half years older. They had more comorbidities: their Charlson Comorbidity Index is 2.
And interestingly, only about half of them had a prior taxane. So the VA has its own clinical pathway—it’s a little bit different than something like the NCCN, where there’s a specific therapy that is recommended for a given state of a disease, including prostate cancer. And part of that pathway has been—up until today, when we heard about PSMAfore—to utilize PSMA radioligand therapy in the pre-chemotherapy setting for patients who either refused docetaxel or were not suitable for it. And the vast majority of the patients who’ve received lutetium have had an ARPI.
And this is the usage as we see it in the VA health care system since 2022. As you remember, the VISION study led to approval in March of 2022. There was a time to ramp up the usage of it. And as you may remember, there were some manufacturing issues at the outset. But we see that overall, the usage has been increasing by quarter over the last few years.
And in terms of survival, our veterans’ median survival after lutetium is only about 9 to 10 months, as compared to the VISION patients. Our veterans, as I mentioned, have more comorbidities. They’re more likely to suffer toxicity and have competing causes for mortality, which can be quite significant, even if they have mCRPC.
When we looked at a subset of patients who had tumor suppressor gene alterations, namely TP53, RB, and PTEN, we saw that there’s clearly a difference in the outcomes of these patients. Now, these are well-known prognostic markers, and this is really not telling us whether or not there is differential benefit of PSMA radioligand therapy in these patients. And here are some of the other factors, mainly the Charlson Comorbidity Index, in a multivariate Cox proportional hazards analysis that predict for outcome.
Some other studies have correlated genomics with outcome. Ohio State showed, interestingly, that tumor suppressor genes do not have an impact on survival in patients getting radioligand therapy. Similarly, for patients with homologous recombination deficiency, there’s no difference amongst those patients who have it versus those that don’t. But cell cycle gene amplification was clearly associated with a poor outcome from radioligand therapy.
Another study from the University of Washington, UCSF, and University of Colorado did show tumor suppressor gene alterations—one or more—is associated with poor outcomes, as were DNA damage repair gene alterations. And finally, a study from the Frankfurt Metastatic Cancer Database also demonstrated that BRCA alterations are associated with poor outcomes, but not tumor suppressor genes.
So you see that the data are kind of all over the place. It would be interesting if tumor suppressor gene–altered patients do as well as those that do not have those mutations, since it’s such a poor prognostic factor. And I think if that were the case—if that were demonstrated in a prospective study—that would suggest that a tumor suppressor gene alteration might benefit, in particular, from PSMA radioligand therapy.
So finally, some take-home messages. One, we’re seeing that usage of PSMA RLT is increasing with time. There are genomic factors that may be prognostic. The studies vary, as I just showed you, and we need better genomic predictive markers that should be incorporated into prospective clinical trials whenever feasible.
And we need novel approaches to improve survival in patients with mCRPC. We need newer approaches—ADCs, immunotherapies, leveraging what we know about the tumor microenvironment, and potentially coming up with synergistic agents such as PARPIs and ARPIs that have been used and recently approved.
And again, the concept of bringing the drugs that are effective for mCRPC to the mCSPC space can result in improved outcomes, and we look forward to the results of some of the studies, such as the PSMAddition trial, that have tested and finished accrual for patients with mCSPC testing the role of PSMA radioligand therapy.
So just a few acknowledgments. I really want to focus on a colleague, Martin Schoen, who really did a lot of the data extraction and analysis for the VA data. So thank you very much.