OBJECTIVE: To assess the accuracy of 18F choline PET/magnetic resonance (MR) when compared with the reference standard MRI-guided template mapping (TPM) biopsies for evaluation of biochemical relapse of prostate cancer in patients with clinically significant and insignificant disease. MATERIAL AND METHODS: A total of 26 prostate lobes in 13 patients with rise in prostate-specific antigen (PSA), age range 56-79; median 66 years, who had 18F choline PET/MR and underwent TPM. Whole body PET/MR images were correlated with maximum cancer core length (MCCL) and total cancer core length (TCCL) obtained via TPM. RESULTS: The mean age of the patients was 66 years (median: 66, range: 56-79 years), and the mean PSA level was 5.17 ng/ml (median: 4.40 ng/ml, range: 1.0-11.1 ng/ml; Roche Modular method). The mean maximum standard uptake value (SUVmax ) was 3.18 (median: 2.3, range: 0.6-12.3). The mean MCCL was 1.5 mm (range: 1-9 mm), mean TCCL was 3.1 mm (range: 1-28 mm). We evaluated the correlation (P value) between MCCL, TCCL, and SUVmax on 18F choline PET/MR with univariate analysis. P value for MCCL was 0.078 and for TCCL 0.093. Overall, the sensitivity, specificity, positive predictive value, negative predictive value, and area under the curve were 77, 70, 58, 85, and 0.621 (95% confidence interval: 2.16-4.19). CONCLUSION: 18F choline PET/MR allows detection of clinically significant and insignificant prostate cancer. Multiple previous treatments can give false positive results. False negative results with 18F choline PET/MR can be because of very small volume (≤2 mm) disease.
Haroon et al. (Tue,) studied this question.