Active surveillance (AS) is a management option for lower-risk prostate cancer (PC) patients that seeks to avoid or delay the toxicities of treatment while maximizing longevity. This conservative strategy involves monitoring patients with low-risk PC (defined as any cancer ≤cT2a, grade group [GG] 1, PSA <10 ng/mL) until intervention is warranted.¹ Patients who choose AS require routine follow-up, and key components include serial PSA, digital rectal examination, multiparametric (mp) MRI, and repeat prostate biopsy, tailored to patient and tumor factors. However, AS includes several dissatisfiers, including repeated biopsies, which pose a risk of pain, infection, and psychological stress.^2-4 There is a need for improved noninvasive approaches to avoid misclassification and to detect progression while on AS.

MRI is used on clinical AS protocols to improve baseline risk stratification and detection upon the progression of clinically significant PC (csPC). Nonetheless, patients with negative/equivocal MRI findings still undergo prostate biopsy for false-negative concerns. Next-generation prostate-specific membrane antigen (PSMA) positron emission tomography (PET) imaging is shifting the landscape of diagnosis and management of PC, particularly in the setting of the initial staging of high-risk patients and recurrence.¹⁸ F-DCFPyL-PSMA PET/MRI accurately identifies csPC and may have an incremental benefit to the MRI alone in the AS setting.^5,6 Notwithstanding the potential benefits of this imaging modality, there is no study to date assessing the role of PSMA PET in the AS setting.

The goal of the current institutional review board–approved prospective single-arm phase 2 imaging trial is to determine whether the combination of PET PSMA and mpMRI was superior to mpMRI alone in detecting csPC (GG ≥2) in AS patients undergoing repeat prostate biopsy, as well as to determine the negative predictive value (NPV) and the optimal predictive model for MRI and PET scoring.

To test the diagnostic accuracy of ¹⁸F-DCFPyL PSMA PET with prostate mpMRI, we used a new dedicated PET/MRI scanner (GE Signa). MRI is interpreted blindly by genitourinary radiologists, and PSMA PET by nuclear medicine physicians. Our planned accrual is 100 men on AS, with biopsy-proven low- or intermediate-risk PC, being considered for transrectal biopsy based on clinical risk. Patients undergo PET/MRI targeted and systematic prostate biopsy. We determined sensitivity, specificity, predictive values (NPV and positive predictive value), and accuracy for detecting csPC for mpMRI, PSMA PET, and PSMA PET + mpMRI. PSMA PET + mpMRI was defined as positive when mpMRI Prostate Imaging Reporting and Data System (PI-RADS) ≥ 3 in addition to PSMA Likert ≥ 3 (scale: Likert 1: highly negative, Likert 5: highly positive), and csPC was defined as any International Society of Urological Pathology GG ≥ 2 malignancy.

For this interim analysis, 57 completed the entire study protocol. The mean age of these patients was 68 ± 6 years, with a median PSA of 8.1 (5.9-10.9) ng/mL, and a PSA density of 0.12 (0.08-0.18) ng/mL². These men were followed on AS for an average of 25 (13-58) months, and the majority (67%) had very low- or low-risk disease (PSA <10 ng/mL, GG1, cT1–cT2a). Fifty-six (98%) had a positive MRI (PI-RADS ≥3); a positive PSMA PET (Likert ≥3) was reported in 57 (100%). Including all PI-RADS ≥3 and Likert ≥3 lesions, 163 areas were marked for targeted biopsy, which included 55 (33%) PI-RADS ≥ 3 and Likert < 3 (MRI positive, PET negative), 40 (24.5%) having PI-RADS < 3 and Likert ≥ 3 (MRI negative, PET negative; Figure), and 68 (42%) with PI-RADS ≥ 3 and Likert ≥ 3 (MRI positive, PET positive). Out of the 163 lesions targeted for biopsy, histology was benign in 99 (60.4%), GG1 in 36 (22%), and GG ≥ 2 in 36 (22%).

The addition of PSMA PET increased the accuracy from 43% (mpMRI alone) to 69% (MRI PI-RADS ≥3 + PET Likert ≥3) or at the optimal predictive model 80% (MRI PI-RADS ≥4 and PET Likert ≥3). A high NPV of 93% (MRI alone) and 91% (mpMRI + PSMA PET) is maintained. Three out of the 28 patients (10%) identified with csPC were detected only by PSMA PET, who would have otherwise continued AS based solely on MRI-targeted biopsies. Maximum standardized uptake value of csPC increased with GG; GG2, GG3, and GG5 were 5.5, 8.6, and 11.1, respectively.

In conclusion, for this low-risk cohort on AS, the addition of PSMA PET to mpMRI improved accuracy and the NPV, with csPC being detected at a higher rate. Pending the results of the entire cohort, PSMA PET use in AS protocols has an additive ability to identify csPC and may avoid inappropriately classifying patients with csPC on AS based on mpMRI alone. Additionally, clinical predictive models incorporating the use of PSMA PET might be able to identify whether biopsy can be safely omitted in men with clinical suspicion but negative combined imaging, given the high NPV of the mpMRI + PSMA PET combination. This approach could reduce the number of repeat biopsies, avoiding the clinical burden associated with them in AS patients. Although the current cost might limit its widespread use, selecting patients based on clinical or imaging variables who would benefit from the addition of the PSMA PET could help its application to AS protocols.