Abstract:
Objective To systematically review the clinical diagnostic value of
18FDG-PET/PET-CT in patients with suspected local recurrent colorectal carcinoma.Methods Computer-based retrieval was conducted on Pubmed, Embase, SCI, CB and Medion (from their establishment to 07/2011) to search published English articles about local recurrent colorectal carcinoma diagnosed by
18FDG-PET/PET-CT. Two reviewers independently selected the studies, extracted the data and accessed the quality according to the inclusion and exclusion criteria. A funnel plot was drawn using deeks method of stata10.1 software and the pooled sensitivity, specificity, positive likelihood ratio (LR+), negative likelihood ratio (LR-) and diagnostic odds ratio (DOR) were calculated. The summary receiving operating characteristic (SROC) was drawn and the areas under the curve (AUC) and Q value were measured in the form of Meta-analysis by MetaDiSc1.4 software.Results Twenty-two studies entered this Meta-analysis. The total area under the curve (AUC) and Q value were 0.978 and 0.933 respectively. The results of Meta-analysis showed sensitivity was 0.944 (95%
CI:0.918-0.964, I
2=0), DOR was 208.670 (95%
CI:109.56-397.44, I
2=23.5%), LR- was 0.084 (95%
CI:0.060-0.119, I
2=0). Good homogeneity, and heterogeneity existed in specificity and LR+. The pooled specificity and LR+ obtained by incorporating homogeneous studies directly accroding to the forest plot or meta-regression analysis, were 0.986 (95%
CI:0.975-0.993, I
2=15%) and 34.903 (95%
CI:21.321-57.126, I
2=0) or 0.981 (95%
CI:0.968-0.989, I
2=5.5%) and 27.937 (95%
CI:15.953-48.926, I
2=25.5%) respectively.Conclusion
18FDG PET/PET-CT has a high sensitivity and specificity for local recurrence diagnosis in colorectal carcinoma.