Sentinel lymph node biopsy versus complete lymphadenectomy in high grade endometrial cancer: a systematic review and meta-analysis of performance characteristics

Abstract

Objectives: Sentinel lymph node (SLN) biopsy is widely accepted as the standard of care for surgical staging in low grade endometrial cancer, but its value in high grade endometrial cancer remains unclear. The aim of this systematic review and meta-analysis was to evaluate the performance characteristics of SLN biopsy compared with pelvic lymphadenectomy in endometrial cancer patients with high-grade histology (PROSPERO ID: CRD42020160280). Methods: We systematically searched MEDLINE, EMBASE, the Cochrane Central Register of Controlled Trials, the Cochrane Database of Systematic Reviews, ClinicalTrials.gov, Ovid MEDLINE Epub Ahead of Print and In-Process & Other Non-Indexed Citations (January 1, 2000 to February 4, 2020) for prospective observational studies evaluating SLN biopsy in clinical stage I high grade endometrial cancer patients (grade 3 endometrioid, serous, clear cell, carcinosarcoma, mixed, undifferentiated/dedifferentiated, high grade not otherwise specified). Studies in which patients underwent cervical injection of indocyanine green for SLN detection, and at least a bilateral pelvic lymphadenectomy as a reference standard, were eligible for inclusion. If data was not reported specifically for patients with high grade histology, authors were contacted for aggregate data. We pooled detection rates and measures of diagnostic accuracy using a generalized linear mixed-effects model with a logit link, and assessed risk of bias using the QUADAS-II tool. Results: We identified 20 eligible studies, of which the authors for 8 provided data on 427 high grade patients specifically. The study-level median age was 66.9 years (range 44-82.8) and the study-level median BMI was 28.8 kg/m2 (range 18.9-45.2). The pooled detection rates were 91% per patient (95% CI 85-95; I2=62%) and 64% bilaterally (95% CI 53-73, I 2=72%). The overall node-positivity rate was 26% (95% CI 20-34; I 2=47%). Of 87 node positive patients, SLN biopsy correctly identified 80, yielding a pooled sensitivity of 92% per patient (95% CI 84-96; I2=0%) (Figure 1), false negative rate of 8% (95% CI 4-16; I2=0%), and negative predictive value of 97% (95% CI 95-99; I 2=0%). Conclusions: SLN biopsy accurately detects lymph node metastases in high grade endometrial cancer, with a false negative rate comparable to that observed in melanoma and breast cancer. Diagnostic accuracy would be even higher if a strict SLN algorithm was followed in all included studies. SLN biopsy can replace complete lymphadenectomy in patients with high grade endometrial cancer. Sentinel lymph node (SLN) biopsy is widely accepted as the standard of care for surgical staging in low grade endometrial cancer, but its value in high grade endometrial cancer remains unclear. The aim of this systematic review and meta-analysis was to evaluate the performance characteristics of SLN biopsy compared with pelvic lymphadenectomy in endometrial cancer patients with high-grade histology (PROSPERO ID: CRD42020160280). We systematically searched MEDLINE, EMBASE, the Cochrane Central Register of Controlled Trials, the Cochrane Database of Systematic Reviews, ClinicalTrials.gov, Ovid MEDLINE Epub Ahead of Print and In-Process & Other Non-Indexed Citations (January 1, 2000 to February 4, 2020) for prospective observational studies evaluating SLN biopsy in clinical stage I high grade endometrial cancer patients (grade 3 endometrioid, serous, clear cell, carcinosarcoma, mixed, undifferentiated/dedifferentiated, high grade not otherwise specified). Studies in which patients underwent cervical injection of indocyanine green for SLN detection, and at least a bilateral pelvic lymphadenectomy as a reference standard, were eligible for inclusion. If data was not reported specifically for patients with high grade histology, authors were contacted for aggregate data. We pooled detection rates and measures of diagnostic accuracy using a generalized linear mixed-effects model with a logit link, and assessed risk of bias using the QUADAS-II tool. We identified 20 eligible studies, of which the authors for 8 provided data on 427 high grade patients specifically. The study-level median age was 66.9 years (range 44-82.8) and the study-level median BMI was 28.8 kg/m2 (range 18.9-45.2). The pooled detection rates were 91% per patient (95% CI 85-95; I2=62%) and 64% bilaterally (95% CI 53-73, I 2=72%). The overall node-positivity rate was 26% (95% CI 20-34; I 2=47%). Of 87 node positive patients, SLN biopsy correctly identified 80, yielding a pooled sensitivity of 92% per patient (95% CI 84-96; I2=0%) (Figure 1), false negative rate of 8% (95% CI 4-16; I2=0%), and negative predictive value of 97% (95% CI 95-99; I 2=0%). SLN biopsy accurately detects lymph node metastases in high grade endometrial cancer, with a false negative rate comparable to that observed in melanoma and breast cancer. Diagnostic accuracy would be even higher if a strict SLN algorithm was followed in all included studies. SLN biopsy can replace complete lymphadenectomy in patients with high grade endometrial cancer.