BACKGROUND : Almost one-third of colorectal cancer (CRC) patients experience recurrence after resection; nevertheless, follow-up strategies remain controversial. We sought to systematically assess and compare the accuracy of carcinoembryonic antigen (CEA), imaging [positron emission tomography (PET) and computed tomography (CT) scans], and circulating tumor DNA (CtDNA) as surveillance strategies.

PATIENTS AND METHODS : PubMed, Medline, Embase, Scopus, Cochrane, Web of Science, and CINAHL were systematically searched. The Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) was used to assess methodological quality. We performed a bivariate random-effects meta-analysis and reported pooled sensitivity, specificity, and diagnostic odds ratio (DOR) values for each surveillance strategy.

RESULTS : Thirty studies were included in the analysis. PET scans had the highest sensitivity to detect recurrence (0.95; 95%CI 0.91-0.97), followed by CT scans (0.77; 95%CI 0.67-0.85). CtDNA positivity had the highest specificity to detect recurrence (0.95; 95%CI 0.91-0.97), followed by increased CEA levels (0.88; 95%CI 0.82-0.92). Furthermore, PET scans had the highest DOR to detect recurrence (DOR 120.7; 95%CI 48.9-297.9) followed by CtDNA (DOR 37.6; 95%CI 20.8-68.0).

CONCLUSION : PET scans had the highest sensitivity and DOR to detect recurrence, while CtDNA had the highest specificity and second highest DOR. Combinations of traditional cross-sectional/functional imaging and newer platforms such as CtDNA may result in optimized surveillance of patients following resection of CRC.