Sentinel lymph node mapping and intraoperative assessment in a prospective, international, multicentre, observational trial of patients with cervical cancer: The SENTIX trial

Abstract

BackgroundSENTIX (ENGOT-CX2/CEEGOG-CX1) is an international, multicentre, prospective observational trial evaluating sentinel lymph node (SLN) biopsy without pelvic lymph node dissection in patients with early-stage cervical cancer. We report the final preplanned analysis of the secondary end-points: SLN mapping and outcomes of intraoperative SLN pathology. MethodsForty-seven sites (18 countries) with experience of SLN biopsy participated in SENTIX. We preregistered patients with stage IA1/lymphovascular space invasion-positive to IB2 (4 cm or smaller or 2 cm or smaller for fertility-sparing treatment) cervical cancer without suspicious lymph nodes on imaging before surgery. SLN frozen section assessment and pathological ultrastaging were mandatory. Patients were registered postoperatively if SLN were bilaterally detected in the pelvis, and frozen sections were negative. Trial registration: ClinicalTrials.gov (NCT02494063). ResultsWe analysed data for 395 preregistered patients. Bilateral detection was achieved in 91% (355/395), and it was unaffected by tumour size, tumour stage or body mass index, but it was lower in older patients, in patients who underwent open surgery, and in sites with fewer cases. No SLN were found outside the seven anatomical pelvic regions. Most SLN and positive SLN were localised below the common iliac artery bifurcation. Single positive SLN above the iliac bifurcation were found in 2% of cases. Frozen sections failed to detect 54% of positive lymph nodes (pN1), including 28% of cases with macrometastases and 90% with micrometastases. InterpretationSLN biopsy can achieve high bilateral SLN detection in patients with tumours of 4 cm or smaller. At experienced centres, all SLN were found in the pelvis, and most were located below the iliac vessel bifurcation. SLN frozen section assessment is an unreliable tool for intraoperative triage because it only detects about half of N1 cases.