Inclusion of PSMA-PET to mpMRI May Better Clinical Evaluation of Patients on Active Surveillance

Adding PSMA-PET imaging to multi-parametric MRI (mpMRI) correlated with better detection of clinically significant prostate cancer (csPC) in men who are being actively monitored, according to an interim analysis of a phase 2 study presented at the 2024 American Urological Association (AUA) Annual Meeting.

“What we saw at the interim analysis was that the addition of PSMA-PET to mpMRI improved the detection of clinically significant prostate cancer compared with mpMRI alone,” said Marcelo Bigarella, MD.

“Active surveillance is not perfect. We do see a lot of things with MRI that are not clinically significant, and the opposite is also true—sometimes we miss clinically significant prostate cancer. And we have seen that PSMA-PET has worked in higher risk patients, such as those with biochemical recurrence and advanced disease, so we wanted to examine the potential of bringing PSMA-PET to active surveillance patients with low- and intermediate-risk disease to see if it also works there. We examined the accuracy of adding PET-PSMA to MRI in the active surveillance setting,” Bigarella, a urologic oncology fellow in the Department of Urology at the University of Wisconsin School of Medicine and Public Health, said in an interview with Urology Times at the AUA meeting.

The study enrolled men on active surveillance with biopsy confirmed very low-, low-, and intermediate-risk prostate cancer. This included newly diagnosed patients who were candidates for active surveillance, patients with progression on active surveillance, and patients on active surveillance undergoing planned prostate biopsy.

The interim analysis included 57 patients with a median age at baseline of 68 years (IQR, 62-73), a median PSA of 8.1 (IQR, 5.9-10.9), a median prostate volume of 60.5 (IQR, 39.0-82.7), a median PSA density of 0.12 (IQR, 0.08 - 0.18), and a median time on active surveillance of 25 months (IQR, 13-58). Per NCCN risk group categorization, 51% of patients were very low risk, 16% were low risk, 28% were favorable intermediate, and 5% were unfavorable intermediate.

All patients received pelvic PET, whole body PET, and mpMRI, followed by MRI/TRUS targeted fusion biopsy. There were 163 lesions biopsied. Of these, 123 were determined to be PI-RADS 3-5. The mean lesion size was 11 mm (IQR, 8-14). Eighty lesions were Likert scale 3-5, and the mean SUVmax was 4.9 (IQR, 3.4-7.2).

The primary outcome measure was the detection of csPC, defined as International Society of Urological Pathologists (ISUP) Grade Group 2 or higher.

Results showed that the specificity in detecting csPC was 29% with mpMRI PI-RADS ≥3 and 68% with mpMRI PI-RADS ≥4. This went up to 87% when combining mpMRI PI-RADS ≥4 with PSMA-PET (18F-DCFPyL) Likert ≥3. The accuracy went up from 43% and 67% in the mpMRI PI-RADS ≥3 and mpMRI PI-RADS ≥4 groups to 80% when using mpMRI plus PSMA-PET. The positive predictive values were 27%, 37%, and 55%, respectively, and the negative predictive values were 93%, 88%, and 88%, respectively.

Overall, csPC was detected in 28 (49%) of 57 patients. Ten percent (n = 3) of these csPC were detected by PSMA-PET only and not seen on MRI and 14% (n = 4) of these were detected by MRI and not on PSMA-PET. Twenty-two (78%) of the patients with csPC had a positive mpMRI and PSMA-PET.

“What we saw at the interim analysis was that the addition of PSMA-PET to mpMRI improved the detection of clinically significant prostate cancer compared with mpMRI alone,” Bigarella said when presenting the data at the AUA meeting. “A predictive model using MRI PI-RADS ≥4 in addition to PSMA-PET Likert ≥3 is the one with the best accuracy…with reasonable negative predictive value of combined MRI/PSMA-PET for targeted lesions.”

REFERENCE:

Bigarella M, Lawrence E, Wiedmer A, et al. Impact of 18F-DCFPYL PSMA PET on accuracy of mpMRI in men with low and intermediate risk prostate cancer: interim analysis of a phase II diagnostic trial. Presented at: 2024 American Urological Association Annual Meeting. May 2-6, San Antonio, Texas. Abstract MP18-18.