Objective  To investigate the value of ¹⁸F-prostate specific membrane antigen (PSMA)-1007 PET/CT in predicting the International Society of Urological Pathology (ISUP) grading in prostate cancer.

Methods  A retrospective analysis was conducted on the clinical and pathological data of 70 patients with prostate cancer who underwent ¹⁸F-PSMA-1007 PET/CT examination and were histopathologically confirmed in the First Hospital of Shanxi Medical University from January 2021 to August 2023. The mean age of the cohort was 70.7±8.4 years. In accordance with the ISUP grading system, the patients were classified into two groups: low-grade prostate cancer (ISUP 1–3) and high-grade prostate cancer (ISUP 4–5). The differences in clinicopathological data, maximum standardized uptake value (SUV_max), mean standardized uptake value (SUV_mean), and peak standardized uptake value (SUV_peak) between the two groups were analyzed and compared. The comparison between two groups of quantitative data was conducted using independent sample t-test or Mann-Whitney U test; The comparison between two groups of counting data was conducted using chi square test. The statistically significant indicators were incorporated into a binary Logistic regression analysis to develop a Logistic prediction model. The diagnostic efficacy of the model was assessed using the receiver operating characteristic (ROC) curve, and the differences in the area under the curve (AUC) between the various indicators and models were evaluated through the Delong test.

Results  Among the 70 prostate cancer patients, 26 cases belonged to the low-grade prostate cancer group and 44 cases to the high-grade prostate cancer group. The two groups exhibited statistically significant differences (Z=3.13–4.70, all P<0.05) in total prostate-specific antigen (28.13 (9.35, 88.86) ng/ml vs. 97.97 (46.72, 312.47) ng/ml), SUV_max (7.81 (5.13, 19.06) vs. 30.93 (18.01, 50.12)), SUV_mean (3.66 (2.50, 6.74) vs. 13.43 (7.75, 21.81)), and SUV_peak (6.43 (3.33, 11.20) vs. 16.20 (11.14, 31.07)). The results of the binary Logistic regression analysis indicated that SUV_max (OR=1.08, 95%CI: 1.01–1.11, P=0.012) and SUV_mean (OR=1.06, 95%CI: 1.02–1.14, P=0.013) were independent predictors of ISUP grading in prostate cancer. The results of the ROC curve analysis demonstrated that the AUCs for diagnosing high-grade prostate cancer using SUV_max, SUV_mean, and the Logistic prediction model alone were 0.71, 0.69, and 0.86, respectively. The sensitivity values were 74.0%, 86.0%, and 83.0%, respectively, while the specificity values were 68.9%, 56.2%, and 79.3%, respectively. Pairwise comparisons of AUCs between the Logistic prediction model and SUV_max and between the Logistic prediction model and SUV_mean revealed statistically significant differences (Z=1.98, 2.23; both P<0.05). By contrast, no statistically significant difference in AUCs was observed between SUV_max and SUV_mean (Z=0.58, P=0.560).

Conclusions  The Logistic prediction model based on SUV_max and SUV_mean from ¹⁸F-PSMA-1007 PET/CT has high diagnostic value for high-grade prostate cancer. Compared with individual parameters, this model provides more favorable value for predicting ISUP grading in prostate cancer.