PURPOSE

The incidence rates of metastatic hormone-sensitive prostate cancer (mHSPC) have increased rapidly. Androgen deprivation therapy (ADT) with AR signaling inhibitors (ARSIs) has been determined survival benefit for mHSPC patients in several randomized trials. However, patients do not respond uniformly. Whole-body tumor-burden schemes guided by prostate-specific membrane antigen (PSMA) PET/CT have been proven to be a useful predictive tool for PSMA-targeted radioligand therapy (RLT), while the value for hormone therapy was unclear. We hypothesized a visual whole-body tumor-burden classification based on PSMA PET/CT can enable selective patient stratification and prognostic evaluation for hormone treatment.

MATERIALS AND METHODS

Patients diagnosed with de novo mHSPC through pathological test and [F]F-PSMA PET/CT between February 2022 and December 2023 who received ADT alone or ADT plus first-generation antiandrogens or ADT plus second-generation/novel ARSIs in our hospital were included. Only can hormone treatments (ADT or ADT plus first-generation antiandrogens or ADT plus novel ARSIs) be adopted at least six months after initial diagnosis. Prostate Cancer Multidisciplinary Team (MDT) of our hospital proposed a newly visual whole-body tumor-burden scheme based on PSMA PET/CT (MDT scheme: high vs. low): MDT high group (fulfilling any one of the three following criteria): (I) the number of metastatic lesions is more than 10 (diffused involvement of single bone is counted as 4 lesions) and PSMA uptake levels of 80% lesions are higher than that of parotid glands, (II) the presence of visceral metastases, (III) at least 4 bone metastases (≥ 1 beyond the vertebral bodies or the pelvis). In addition, other three tumor-burden classification methods (PSMA-CHAARTED, PSMA-LATITUDE, revised-vPSG schemes) were also assessed in this study. A series of other parameters including SUV-derived features of PSMA PET/CT and potential clinical and pathological factors were evaluated. SUV-derived features were determined for measurable locations and included: SUVmax, SUVpeak, SUVmean and tumor volume (TV) of prostatic primary lesions, the highest SUVmax of all lesions in whole body (wbSUVmax), primary-tumor SUVmax ratio backgrounds (including blood pool of liver/ spleen/ mediastinum/parotid glands), wbSUVmax ratio backgrounds above. Serum prostate-specific antigen (PSA) lower than 0.2 ng/ml after six-month hormone treatment was set as the primary endpoint for prediction of PSA response. PSA99 (PSA reduction ≥ 99%) was the second endpoint for survival analysis. All parameters above including the four tumor-burden classification schemes were evaluated for the predictive and prognostic value according to the endpoints using logistic and Cox proportional hazards regression analysis, respectively. All P values < 0.05 were considered significant.

RESULTS

A total of 165 patients were included. The average age was 69.30 ± 8.12 years. In univariate logistic regression analysis, MDT, PSMA-CHAARTED, revised-vPSG tumor-burden classifications, type of hormone treatment and primary-tumor TV were significantly related to PSA response, and PSMA-LATITUDE scheme wasn't relevant to PSA response. The results of further multivariate logistic regression revealed MDT scheme (MDT high group vs. low group: OR = 5.34, 95%CI: 2.40-11.87, P < 0.001), type of hormone treatment (ADT + second-generation ARSIs vs. ADT alone or ADT + first-generation antiandrogens: OR = 0.21, 95%CI: 0.08-0.53, P = 0.001), and primary-tumor TV (≥ 12.49 cm vs. < 12.49 cm: OR = 2.93, 95%CI: 1.25-6.89, P = 0.014) were proven to be independent significant predictors for mHSPC patients. Subgroups analysis of patients treated with ADT + second-generation ARSIs (N = 133) showed MDT, PSMA-CHAARTED, PSMA-LATITUDE, revised-vPSG tumor-burden classifications and primary-tumor TV were significantly associated with PSA response through univariate analysis, and in multivariate regression MDT scheme (MDT high group vs. low group: OR = 5.73, 95%CI: 2.47-13.30, P < 0.001) and primary-tumor TV (≥ 12.49 cm vs. < 12.49 cm: OR = 2.75, 95%CI: 1.09-6.96, P = 0.032) were independent significant predictors for PSA response of novel ARSIs. The AUC of three-predictors model of 165 mHSPC patients was 0.740 (95% CI: 0.664-0.816, P < 0.001). The AUC of two-predictors model of 133 mHSPC patients treated with ADT + novel ARSIs was 0.751 (95% CI: 0.666-0.836, P < 0.001). Univariate survival analysis revealed that patients treated with ADT + second-generation ARSIs were prone to obtaining PSA remission in shorter period (P < 0.001). In multivariate Cox regression, MDT scheme (MDT high group vs. low group: HR = 0.52, 95%CI: 0.36-0.74, P < 0.001) and type of hormone treatment (ADT + second-generation ARSIs vs. ADT alone or ADT + first-generation antiandrogens: HR = 3.34, 95%CI: 2.02-5.54, P < 0.001) were demonstrated to be independent significant prognostic indicators of patients with mHSPC. The survival analysis of patients treated with ADT + second-generation ARSIs (N = 133), patients with low burden of MDT (P < 0.001) or PSMA-CHAARTED (P = 0.029) or PSMA-LATITUDE (P = 0.009) could achieve PSA 99% reduction faster, and the low and high group patients based on revised-vPGS weren't displayed the significant difference in PSA99. Through the multivariate Cox survival regression, only MDT scheme (MDT high group vs. low group: HR = 0.47, 95%CI: 0.33-0.69, P < 0.001) was selected as the independent significant prognostic biomarker for mHSPC patients treated with ADT + second-generation ARSIs.

CONCLUSION

The visual whole-body tumor-burden classification based on PSMA PET/CT should be an effective stratification strategy for mHSPC patients treated with ADT + ARSIs, especially with ADT + second-generation ARSIs. PSA99 (PSA reduction ≥ 99%) could be a superior endpoint for patients with HSPC. This classification scheme could be a promising method for stratifying mHSPC patients. These findings need to be confirmed and validated through a longer follow-up, prospective and clinical data.