Objective: To review diagnostic accuracy of positron emission tomography/computed tomography (PET/CT) for colorectal liver metastases.
Background: Colorectal liver metastases can be treated with surgical resection; however, recurrence is seen in 58% of patients. PET/CT may better detect extra-hepatic disease before surgery to more accurately identify eligible candidates for surgery and, through better selection, improve patient prognosis.
Methods: We conducted a comprehensive systematic review on adults with colorectal liver metastases who received PET/CT and CT scans to detect metastases. The gold standard to confirm the diagnosis was histology. Study selection, quality assessment, and data extraction were completed independently by 2 investigators. Pooling of results was not feasible because of heterogeneity. A qualitative summary of results is presented.
Results: From 1083 citations, we identified 6 studies (440 patients) for the review. For extra-hepatic lesions (3 studies; 178 patients), PET/CT was more sensitive than CT, but specificities were similar (PET/CT sensitivity [SN] = 75%–89% and specificity [SP] = 95%–96% vs. CT SN = 58%–64% and SP = 87%–97%). For hepatic lesions (5 studies; 316 patients), PET/CT had higher SN and SP than CT (PET/CT SN = 91%–100% and SP = 75%–100%; CT SN = 78%–94% and SP = 25%–98%). For local recurrence (3 studies; 206 patients), PET/CT also had better accuracy than CT with SN = 93% to 100% and SP = 97% to 98% versus SN = 0 %–100% and SP = 97%–98%.
Conclusion: Based on this systematic review, we conclude that PET/CT has a higher accuracy for detection of extra-hepatic and hepatic colorectal metastatic disease than CT alone. However, the results are based on a small number of studies and should be interpreted cautiously.