Sentinel lymph node biopsy in endometrial cancer: meta-analysis of 26 studies.

OBJECTIVE The validity of the sentinel lymph node (SLN) procedure for the assessment of nodal status in patients with endometrial cancer is unclear. We aimed to assess the diagnostic performance of this procedure. METHODS We searched the PubMed and Embase databases for studies published before June 1, 2011. Eligible studies had a sample size of at least 10 patients, and reported the detection rate and/or sensitivity of the SLN biopsy. RESULTS We identified 26 eligible studies, which included 1101 SLN procedures. The overall weighted-mean number of harvested SLNs was 2.6. The detection rate and the sensitivity were 78% (95% confidence interval [CI]=73%-84%) and 93% (95% CI=87%-100%), respectively. Significant between-study heterogeneity was observed in the analysis of the detection rate (I-squared statistic, 80%). The use of pericervical injection was correlated with the increase of the detection rate (P=0.031). The hysteroscopic injection technique was associated with the decrease of the detection rate (P=0.045) and the subserosal injection technique was associated with the decrease of the sensitivity (P=0.049), if they were not combined with other injection techniques. For the detection rate, significant small-study effects were noted (P<0.001). CONCLUSIONS Although SLN biopsy has shown good diagnostic performance in endometrial cancer, such performance should be interpreted with caution because of significant small study effects. Current evidence is not yet sufficient to establish the true performance of SLN biopsy in endometrial cancer.