Continuous PSMA-PET Metrics Improve Prognostication for Patients with mCRPC
Sharpening morality risk prediction in advanced prostate cancer
Total tumor volume (TTV), measured from prostate-specific membrane antigen (PSMA) PET imaging, provides a more accurate measure of disease burden and improves prediction of overall survival compared with counting metastases in patients with advanced prostate cancer treated with lutetium 177PSMA-617 (Lu-PSMA) therapy.
A recent Radiology study found this approach could improve risk stratification by identifying patients most likely to benefit from therapy and providing a more confident assessment of disease aggressiveness.
Response to Lu-PSMA therapy varies widely among patients despite its demonstrated survival benefit.
“Despite 177Lu-PSMA-617 being a life-prolonging treatment in prostate cancer, we have observed in clinical practice, post FDA approval, that a substantial proportion of patients derive limited benefit,” said lead author Alireza Ghodsi, MD, a postdoctoral research fellow at the University of Washington (UW) and Fred Hutchison Cancer Center in Seattle.
“177Lu-PSMA-617 is increasingly being used earlier in the treatment sequence, even before chemotherapy. Improved risk stratification may help guide treatment sequencing,” added coauthor Amir Iravani, MD, a nuclear medicine physician and associate professor of radiology at UW and theranostics director at Fred Hutchinson Cancer Center.
According to Drs. Ghodsi and Iravani, the original prediction model for risk stratification used a binary classification of metastatic burden with a cutoff of 20 metastatic lesions at pretreatment PSMA PET.
“However, neither of the two nomograms used to support treatment decision-making have been validated in a real-world clinical setting,” Dr. Iravani said. “In this study, we aimed to validate a nomogram previously developed by UCLA investigators in a real-world cohort after FDA approval of Lu-PSMA, and to refine it by replacing lesion count with a more robust PSMA PET–derived metric—total tumor volume.”
Refining Risk Models with TTV
To evaluate the approach, the researchers performed a retrospective study of 168 patients with metastatic castration-resistant prostate cancer (mCRPC), all of whom underwent Lu-PSMA treatment for up to six cycles. The patients also underwent PSMA PET imaging up to three months before treatment, from which tumor volume was derived at baseline using a semi-automated segmentation approach.
“Establishing this database required significant time and effort, supported by an ACR-funded grant,” Dr. Ghodsi said. “It took approximately three years to accrue a sufficient number of patients with adequate follow-up for meaningful analysis.”
To determine the best predictor of clinical outcome, the researchers compared three different models including:
- a binary measure of tumor lesions (the existing model) which classifies tumor burden as a simple cutoff at 20 lesions.
- a continuous measure of tumor lesions, which uses the exact number of lesions as a continuous variable. While more precise than binary, it still focuses on tumor count.
- a continuous measure of TTV, which captures both number and size of lesions, reflecting overall disease burden.
Other variables included time since diagnosis, chemotherapy status, hemoglobin and whether the number of metastases was above or below the tumor burden threshold, as well as tumor mean standardized uptake value and bone and liver involvement.
The team assessed overall survival, the length of time to prostate-specific antigen (PSA) progression and PSA response, looking at whether PSA levels dropped by at least 50% after treatment.
“Our findings ultimately supported both of our initial hypotheses,” Dr. Ghodsi said. “The existing nomogram demonstrated similar discriminatory performance in our real-world patient cohort, and its predictive accuracy was further improved by incorporating total tumor volume as a measure of disease burden.”
“We believe these refined nomograms may help improve risk stratification of patients receiving Lu-PSMA and could support future clinical decision-making after additional validation,” Dr. Iravani added.
Prostate-specific membrane antigen PET images. (A) Representative examples illustrate the variability in the total tumor volume (TTV) among patients stratified by the number of metastatic lesions. (B) Case examples demonstrate TTV segmentation across a range of TTV values. Notably, as the tumor volume increased, a visible decrease in radiopharmaceutical uptake by normal organs is observed, particularly in the salivary glands, kidneys, and liver (denoted by brackets), a phenomenon known as the “sink effect.” SUV = standardized uptake value.
https://doi.org/10.1148/radiol.250649 ©RSNA 2026
Promise Tempered by Practical Limitations
The findings represent a potential advance in prognostication, though practical limitations remain. In a related commentary in Radiology, experts highlighted key barriers to clinical implementation.
“There’s no doubt that TTV, the volumetric data, theoretically is richer in information,” said Herbert Alberto Vargas, MD, from the Department of Radiology at the NYU Grossman School of Medicine in New York City. “But unfortunately, measuring TTV is not practical to do in current routine clinical practice.”
“Either in-house software needs to be developed locally, or commercially available software needs to be introduced that allows for one to calculate TTV automatically (or semiautomatically),” added Sungmin Woo, MD, PhD, of the same department.
Drs. Vargas and Woo describe a scientific limitation—the sizeable heterogeneity of PSMA expression across mCRPC disease sites—as a notable caveat to the model.
“At this stage, and especially when the patient has been heavily treated with several lines of systemic therapies, many metastatic sites may become less or even non-PSMA expressing, meaning they won’t be picked up by segmentation software and count towards the TTV,” Dr. Vargas said.
Non-PSMA expressing disease sites are associated with an even worse prognosis, highlighting a gap future models may need to address. In addition, the retrospective, single-center design may introduce selection bias and limit the generalizability of findings.
“However, the ever-present challenge beyond the study design per se is how well the findings from research studies can be replicated in real life scenarios,” Dr. Woo said. “Even though the design does provide confidence in the findings presented, further work is needed to refine and validate these models in wider populations.”
“I believe thinking of extreme examples can help establish TTV’s superiority,” Dr. Vargas said.
One such scenario highlights the advantage of TTV. “For instance, in a patient with a single large pelvic bone metastasis, this will count as one metastasis, but the prognosis will not be the same as someone who has a punctate lesion, who will be also categorized as having one metastasis.” Dr. Woo explained. “Measuring the volume can overcome such limitations.”
“Improved prognostication is critical to guide treatment selection and sequencing,” Dr. Iravani said. “Identifying patients unlikely to respond to 177Lu-PSMA therapy allows clinicians to prioritize alternative therapies and avoid exposing patients to ineffective treatment and potential toxicities.”
“Ultimately, this approach supports the principles of precision oncology—delivering the right treatment to the right patient at the right time,” Dr. Ghodsi concluded.
For More Information
Access the Radiology study, “PSMA PET/CT-derived Tumor Volume for Predicting Outcomes in Patients with Metastatic Castration-Resistant Prostate Cance Receiving 177 Lu-PSMA-617,” and the related editorial, “Turning Up the Volume: Continuous PSMA-PET Metrics Add Value for Outcome Prediction in Patients with mCRPC.”
Read previous RSNA News stories on prostate imaging: