ABSTRACT
PURPOSE: The aim of this study was to evaluate the accuracy of (11)C-choline positron emission tomography/computed tomography (PET/CT) in restaging patients affected by prostate cancer and suspected relapse due to prostate-specific antigen (PSA) increase. We also aimed to determine a PSA cutoff that is most suited to the study in terms of best compromise between sensitivity and specificity. Secondary endpoints were a comparison between (11)C-choline PET/CT and histological results, clinical findings, and radiological imaging (CT and magnetic resonance imaging). MATERIALS AND METHODS: We retrospectively evaluated 210 patients (median ± SD age 70 ± 7 years) affected by prostate cancer who underwent (11)C-choline PET/CT. RESULTS: (11)C-choline PET/CT imaging was positive in 116 (55.2%) patients and negative in 94 (44.8%). Receiver operating characteristic (ROC) analysis showed that the highest accuracy (sensitivity 76.8%, specificity 92.5%) for the whole population was achieved when the PSA level of 1.26 ng/ml level was used as the cutoff value for interpreting the results (P = 0.0001 and the area under the ROC curve AUC 0.897). For patients treated with surgery or surgery plus radiotherapy the cutoff was 0.81 ng/ml (sensitivity 73.2%, specificity 86.1%). For patients treated with radiotherapy alone, the cutoff was 2.0 ng/ml (sensitivity 81.8%, specificity 92.9%). CONCLUSION: Our results indicate that (11)C-choline PET/CT is a useful diagnostic tool in patients affected by prostate cancer and a relapsed PSA level. The highest accuracy for all patients is obtained with a PSA cutoff level of 1.26 ng/ml, above which the imaging study is performed (0.81 ng/ml for patients treated with surgery or surgery plus radiotherapy and 2.0 ng/ml for patients treated with radiotherapy alone).
