Sentinel lymph node biopsy versus pelvic lymphadenectomy in cervical cancer: The PHENIX trial.

Abstract

LBA5501 Background: Limited data is available on survival outcomes following sentinel lymph node biopsy (SLNB) as a substitute for pelvic lymphadenectomy (PL) in cervical cancer. We aimed to prospectively compare survival outcomes between the two approaches for lym ph nod e dissectio n in cerv ix cancer. Methods: This multicenter, non-inferiority, randomized controlled trial involves patients with FIGO 2009 stage IA1 (lymphovascular invasion), IA2, IB1, or IIA1 cervical cancer, including squamous-cell carcinoma, adenocarcinoma, or adenosquamous carcinoma. Patients with radiological nodal metastases were excluded. All patients underwent SLNB first and all sentinel lymph nodes (SLNs) were examined by frozen section. Based on SLN status, patients were intraoperatively assigned to PHENIX-I (SLN-negative) or PHENIX-II (SLN-positive) cohorts and then randomized 1:1 to undergo PL or not. Side-specific PLs were performed in cases of unilateral detection and radical hysterectomies were performed for all patients. The primary endpoint was disease-free survival. Results: Between December 2015 and December 2023, 908 patients with at least one SLN detected were enrolled and randomized intraoperatively, including 838 SLN-negative patients in PHENIX-I and 70 SLN-positive patients in PHENIX-II (early terminated). Patients were assigned to undergo SLNB alone (455 cases: 420 in PHENIX-I, 35 in II) or PL (453 cases: 418 in PHENIX-I; 35 in II). The median age was 48 years (23 to 65) and the bilateral SLN detection rate was 82.6%. The clinicopathological characteristics, surgical approaches, and postoperative therapies were well-balanced between the SLNB and PL groups. SLNB demonstrated significantly shorter operative duration (P < 0.0001), less blood loss (P < 0.001), and lower morbidities (P < 0.001) compared to PL. The median follow-up time reached 52 months (1 to 104). In PHENIX-I, recurrences were observed in 16 patients in the SLNB group and 26 in the PL group. Retroperitoneal nodal recurrences were observed in 9 patients in the PL group, whereas no such recurrence was detected in the SLNB group (P < 0.05). There were 3 cancer-specific deaths in the SLNB group and 14 in the PL group. The 3-year disease-free survival rates were 96.8% for the SLNB group and 94.5% for the PL group (HR = 0.61, 95% confidence interval [CI] 0.33-1.14, P = 0.12); the 3-year cancer-specific survival rates were 100.0% for the SLNB group and 97.8% for the PL group (HR = 0.21, 95%CI 0.06-0.74, P = 0.007). Similar comparative trend was observed in PHENIX-II despite its early termination (HR = 0.47, 95%CI 0.14-1.58, P = 0.21 for disease-free survival; HR = 0.23, 95%CI 0.05-1.08, P = 0.061 for cancer-specific survival). Conclusion: When SLNB succeeds in cervical cancer, PL should be abandoned as it provides no survival benefit and increases surgical morbidity, yet may unexpectedly correlate with elevated risks of nodal recurrence and cancer-specific mortality. Clinical trial information: NCT02642471 .