Main findings
In this study, we systematically investigated the diagnostic performance of FSE in SLNB for cervical cancer, based on the available data over nearly twenty years. To our knowledge, this is the first systematic review and meta-analysis on this topic. The sensitivity of FSE in SLNB, which had been surrounded by controversies, was found to be dissatisfactory in pooled analysis (0.77, 95%CI 0.66–0.85). This data is very close to that previously reported in breast cancer.54However, a more important finding of this study is that the sectioning protocol of FSE had great impact on diagnostic accuracy, which also generated a high heterogeneity (I 2=69.73%). Significantly improved sensitivities (pooled: 0.86, 95%CI: 0.79–0.91) and low heterogeneity (I 2=0) were observed among the studies using L-protocol, whereas an even lower sensitivity (0.59, 0.46–0.72) was pooled under E-protocol. This difference was so remarkable that we expected it would hardly be denied or reversed in future studies. Thus, our study provides strong evidence supporting L-protocol as the standard for intraoperative pathological examination of SLNs.
PLND shares the same incision with hysterectomy, lowering the priority of a two-step surgery strategy.29 In addition, it is still inconclusive whether PLND can be completely replaced by SLNB in early-stage cervical cancer. For these reasons, intraoperative pathological diagnosis remains an important element in SLNB for cervical cancer. Despite wide recognition, the benefits of SLNB in cervical cancer have long been confined by the lack of precise intraoperative pathology.18Similar dilemmas are also encountered in other malignancies such as gastric cancer.55 The major contribution of this meta-analysis is that, for the first time, we identified a simple method to achieve more precise intraoperative SLN assessment, which enables immediate decision-making for individualized treatments.7