Diagnostic accuracy of 18F-PSMA-1007 PET/CT and bone scintigraphy for detecting bone metastases in prostate cancer.

OBJECTIVE

The performance of 18F-labeled prostate-specific membrane antigen (18F-PSMA-1007) positron emission tomography/computed tomography (PET/CT) and bone scintigraphy (BS) for detecting bone metastases in prostate cancer patients was compared.

MATERIALS AND

METHODS

One hundred prostate cancer patients (initial staging, n = 9; biochemical recurrence, n = 63; metastatic castration-resistant prostate cancer (mCRPC), n = 15; m0CRPC, n = 13) underwent 18F-PSMA-1007 PET/CT and BS. The patient-, region-, and lesion-based performances of 18F-PSMA-1007 PET/CT and BS for detecting bone metastases were evaluated by experienced readers using a five-point scale.

RESULTS

Bone metastases were identified in 46 (46.0%) patients, 99 (24.8%) of the 400 regions, and 528 lesions. In the analysis, the area under the receiver-operating-characteristic curve (AUC), sensitivity, specificity, and accuracy values for 18F-PSMA-1007 PET/CT were 0.9985, 100%, 94.4%, and 97.0%, respectively, whereas those for BS were 0.8507, 71.7%, 88.9%, and 81.0%, respectively. AUC, sensitivity, and accuracy were significantly higher for 18F-PSMA-1007 PET/CT than for BS (p = 0.00021, p = 0.00087, and p = 0.00018, respectively; McNemar test). 18F-PSMA-1007 PET/CT detected 527 of the 528 metastatic lesions (99.8%), whereas BS detected 253 (47.9%) (p < 0.0001). The numbers of patients with (a) no bone metastases, (b) oligometastatic disease (1-3 lesions), (c) 4-5 lesions, (d) 6-20 lesions, and (e) > 20 lesions for BS were 13, 21, 0, 8, and 4, respectively, compared with 0, 27, 3, 9, and 7, respectively, on 18F-PSMA-1007 PET/CT. In addition, 18F-PSMA-1007 PET/CT detected local recurrence (n = 51), nodal metastases (n = 37), and visceral metastases (n = 4).

CONCLUSIONS

18F-PSMA-1007 PET/CT demonstrated higher sensitivity and accuracy than BS for detecting bone metastases in prostate cancer patients.