To assess if PSMA PET quantitative parameters are associated with pathologic ISUP grade group (GG) and upgrading/downgrading. PCa patients undergoing radical prostatectomy with or without pelvic lymph node dissection staged with preoperative PSMA PET at seven referral centres worldwide were evaluated. PSMA PET parameters which included SUVmax, PSMAvolume, and total PSMA accumulation (PSMAtotal) were collected. Multivariable logistic regression evaluated the association between PSMA PET quantified parameters and surgical ISUP GG. Decision-tree analysis was performed to identify discriminative thresholds for all three parameters related to the five ISUP GGs The ROC-derived AUC was used to determine whether the inclusion of PSMA quantified parameters improved the ability of multivariable models to predict ISUP GG ≥ 4. A total of 605 patients were included. Overall, 2%, 37%, 37%, 10% and 13% patients had pathologic ISUP GG1, 2, 3, 4, and 5, respectively. At multivariable analyses, all three parameters SUVmax, PSMAvolume and PSMAtotal were associated with GG ≥ 4 at surgical pathology after accounting for PSA and clinical T stage based on DRE, hospital and radioligand (all p < 0.05). Addition of all three parameters significantly improved the discrimination of clinical models in predicting GG ≥ 4 from 68% (95%CI 63 - 74) to 74% (95%CI 69 - 79) for SUVmax, 72% (95%CI 67 - 76) for PSMAvolume, 74% (70 - 79) for PSMAtotal and 75% (95%CI 71 - 80) when all parameters were included (all p < 0.05). Decision-tree analysis resulted in thresholds that discriminate between GG (SUVmax 0-6.5, 6.5-15, 15-28, > 28, PSMAvol 0-2, 2-9, 9-20 and > 20 and PSMAtotal 0-12, 12-98 and > 98). PSMAvolume was significantly associated with GG upgrading (OR 1.03 95%CI 1.01 - 1.05). In patients with biopsy GG1-3, PSMAvolume ≥ 2 was significantly associated with higher odds for upgrading to ISUP GG ≥ 4, compared to PSMAvolume < 2 (OR 6.36, 95%CI 1.47 - 27.6). Quantitative PSMA PET parameters are associated with surgical ISUP GG and upgrading. We propose clinically relevant thresholds of these parameters which can improve in PCa risk stratification in daily clinical practice.
Read full abstract