Introduction
Over the past decades, sentinel lymph node biopsy (SLNB) has become an attractive surgical procedure in many malignancies.1-5The benefits of SLNB mainly include a reduction in surgical morbidity and time cost, achieved by replacing lymphadenectomies in selected patients,6,7 and revelation of aberrant drainage regions that are probably omitted during routine lymphadenectomies.8 Furthermore, by pathological ultrastaging of sentinel lymph nodes (SLN), the micro-metastatic burden in lymphatic system can be evaluated conveniently, enabling more precise individualized treatments.9-11 With these benefits and high accuracy.12 SLNB is currently recommended as an alternative of pelvic lymph node dissection (PLND) in early-stage cervical cancer.13,14
However, in clinical practice, only a few authors have attempted to perform SLNB alone without a further PLND.6,7,15 A recent international survey by the Gynecologic Cancer Intergroup showed great divergence regarding the SLNB strategy,16reflecting worldwide mistrust on intraoperative decisions made based on SLNB. Indeed, previous studies had shown that the accuracy of SLNB largely relied on postoperative ultrastaging,17,18which is time consuming and unavailable for intraoperative decision-making. As a result, many gynecologists choose to directly replace PLND with SLNB in radical surgery and wait for final pathology. However, nodal metastasis has been included as IIIC stage in the latest International Federation of Gynecology and Obstetrics (FIGO) staging system,19 so it is becoming increasingly important to acquire the lymphatic status before deciding treatment modality. The European Society of Gynaecological Oncology/European Society for Radiotherapy and Oncology/European Society of Pathology guidelines (2018) therefore recommended submitting SLNs for intraoperative assessment to immediately triage patients towards radical surgery or definitive chemoradiotherapy.14 Consequently, accurate intraoperative pathology of SLNs is urgently required.
Frozen section examination (FSE) is the most common method for intraoperative SLN assessment.20 Compared with other methods, FSE has a natural superiority of almost 100% specificity.21-23However, the sensitivity of FSE varies considerably between the published studies.24-35 Some authors had cautioned the high false-negative rate of FSE in SLNB,24-27,29 whereas others provided satisfying results.30,32-35 The reason for such a discrepancy remains unclear and may be associated with the heterogeneity among these studies, including the differences in methodologies, patients’ characteristics, volume of metastases, as well as pathologists’ experiences. However, few studies had concerned the impacts of these factors on the sensitivity of FSE.29,33 It spontaneously interests us whether there exists an optimal protocol by which the FSE can yield the highest sensitivity for intraoperative decision-making.
To our knowledge, there are several ongoing international multicenter trials targeting the validation of SLNB in early-stage cervical cancer.36,37 All these trials were designed with an intraoperative randomization or assignment depended on the results of FSE. Therefore, it is of great importance to validate the accuracy of FSE first. In this systematic review and meta-analysis, we analyzed the available data on this issue, in order to determine the diagnostic performance and optimal protocol of FSE in SLNB for cervical cancer and provide evidence for ongoing and future studies.