Liver metastases from prostate cancer at 11C-Choline PET/CT: a multicenter, retrospective analysis

During our daily clinical practice using 11C-Choline PET/CT for restaging patients affected by relapsing prostate cancer (rPCa) we noticed an unusual but significant occurrence of hypodense hepatic lesions with a different tracer uptake. Thus, we decided to evaluate the possible correlation between rPCa and these lesions as possible hepatic metastases.

We retrospectively enrolled 542 patients diagnosed with rPCa in biochemical relapse after a radical treatment (surgery and/or radiotherapy). Among these, patients with a second tumor or other benign hepatic diseases were excluded. All patients underwent 11C-Choline PET/CT during the standard restaging workup of their disease. We analyzed CT images to evaluate the presence of hypodense lesions and PET images to identify the relative tracer uptake. In accordance to the subsequent oncological history, five clinical scenarios were recognized [Table 1]: normal low dose CT (ldCT) and normal tracer distribution (Group A); evidence of previously unknown hepatic round hypodense areas at ldCT with normal rim uptake (Group B); evidence of previously known hepatic round hypodense areas at ldCT stable over time and with normal rim uptake (Group C); evidence of previously known hepatic round hypodense areas at ldCT, in a previous PET/CT scan, with or without rim uptake and significantly changing over time in terms of size and/or uptake (Group D); evidence of hepatic round hypodense areas at ldCT with or without rim uptake confirmed as prostate liver metastases by histopathology, triple phase ceCT, ce-ultra sound (CEUS) and clinical/biochemical evaluation (Group E). We evaluated the correlation with PSA level at time of scan, rim SUVmax and association with local relapse or non-hepatic metastases (lymph nodes, bone, other parenchyma).

Five hundred and forty-two consecutive patients were retrospectively enrolled. In 140 of the 542 patients more than one 11C-choline PET/CT had been performed. A total of 742 11C-Choline PET/CT scans were analyzed. Of the 542 patients enrolled, 456 (84.1%) had a normal appearance of the liver both at ldCT and PET (Group A). 19/542 (3,5%) belonged to Group B, 13/542 (2.4%) to Group C, 37/542 (6.8%) to Group D and 18/542 (3.3%) to Group E. Mean SUVmax of the rim was: 4.5 for Group B; 4.2 for Group C; 4.8 for Group D; 5.9 for Group E. Mean PSA level was 5.27 for Group A, 7.9 for Group B, 10.04 for Group C, 10.01 for Group D, 9.36 for Group E. Presence of positive findings at 11C-Choline PET/CT in any further anatomical area (local relapse, lymph node, bone, other extra hepatic sites) correlated with an higher PSA (p = 0.0285). In both the univariate and multivariate binary logistic regression analyses. PSA, SUVmax of the rim, local relapse, positive nodes were not associated to liver mets (Groups D-E) (p > 0.05). On the contrary, a significant correlation was found between the presence of liver metG (group D-E) and bone lesions (p= 0.00193).

Our results indicate that liver metastases in relapsing prostate cancer may occur frequently. The real incidence evaluation needs more investigations. In this case and despite technical limitations, Choline PET/CT shows alterations of tracer distribution within the liver that could eventually be mistaken for simple cysts but can be suspected when associated to high trigger PSA, concomitant bone lesions or modification over time. In this clinical setting an accurate analysis of liver tracer distribution (increased or decreased uptake) by the nuclear medicine physician is, therefore, mandatory.