Cervical cancer is a malignant disease that poses a serious threat to women’s health. For early-stage cervical cancer, surgery remains the primary treatment, with pelvic lymph node dissection (PLND) serving as a crucial component. However, this procedure has inherent limitations. Since the majority of early-stage patients do not have lymph node metastases, routine PLND may lead to overtreatment and is associated with various complications such as lymphoedema, venous thrombosis and lymphocyst formation. Furthermore, PLND may disrupt the anatomical basis for immune surveillance. In the era of precision medicine, there is a growing imperative to re-evaluate this classic surgical approach.
Sentinel lymph node biopsy (SLNB) offers a more precise and minimally invasive alternative. By using tracers to identify the first lymph node(s) draining the tumour, SLNB allows patients with negative sentinel nodes to avoid systematic lymphadenectomy, thereby reducing surgical trauma. Over the past two decades, numerous studies, including large prospective multicentre trials like AGO and SENTICOL I, have validated the diagnostic value of SLNB in cervical cancer.1 2 Meta-analyses confirm its high sensitivity and negative predictive value for detecting lymph node metastasis.3 The SENTICOL II study further demonstrated that replacing PLND with SLNB reduces surgical complications and improves patients' quality of life.4
Nevertheless, a critical question persists: Can SLNB safely replace PLND in terms of long-term oncological outcomes? This requires evidence from prospective randomised controlled trials with survival as the primary endpoint. It was in this context that the multicentre PHENIX trial, led by Chinese researchers, was initiated.
The PHENIX trial is a multicentre, randomised controlled, non-inferiority phase III study.5 All patients underwent SLNB first, followed by immediate frozen section analysis. Patients with negative SLNs were randomised in a 1:1 ratio to either the ‘biopsy-only group’ (omitting PLND) or the ‘dissection group’ (receiving PLND). The primary endpoint was 3-year disease-free survival (DFS). Based on historical data, the 3-year DFS for the standard PLND group was assumed to be 94%, with a non-inferiority margin set at 5 percentage points. Secondary endpoints included retroperitoneal lymph node recurrence rate, cancer-specific survival and surgical complications.
Prior research indicated that the sensitivity of intraoperative frozen section analysis might be insufficient to guide surgical decisions. To address this, the PHENIX team conducted a systematic review and meta-analysis before trial initiation, finding that latitudinal sectioning protocol significantly improved the accuracy of frozen sections.6 This protocol was incorporated into the study’s standard operating procedures, laying a solid foundation for trial quality.
After a median follow-up of 62.8 months, the 3-year DFS was 96.9% in the ‘biopsy-only group’ and 94.6% in the ‘dissection group’. The between-group difference was −2.3 percentage points, with the upper limit of the 95% CI (−5.0 to 0.5) below the prespecified non-inferiority margin (p<0.001), confirming that SLNB is non-inferior to systematic PLND regarding oncological outcomes. Moreover, the biopsy group had fewer surgical complications.
Secondary findings from the trial prompted deeper reflection. A notable observation was that all nine cases of retroperitoneal lymph node recurrence occurred in the ‘dissection group’, with none in the ‘biopsy-only group’. Subgroup analysis revealed that among patients undergoing laparoscopic surgery, DFS was significantly better in the ‘biopsy-only group’ (97.3%) compared with the ‘dissection group’ (93.6%), a difference not observed in patients who underwent open surgery.
This potentially redirects the discussion sparked by the LACC trial7—which suggested that laparoscopic surgery might increase the risk of recurrence in cervical cancer—toward a new dimension: Could the potential risks of laparoscopic surgery be associated with the concomitant performance of systematic lymphadenectomy? To explain these findings, the research team proposed two hypotheses. The first is the ‘immunological hypothesis’: Cervical cancer exhibits high immunogenicity, and an intact lymphatic network serves as a crucial barrier for immune surveillance. Removing a large number of tumour-free lymph nodes may weaken regional immune defenses, potentially facilitating the escape of minimal residual disease. The second is the ‘two-step iatrogenic dissemination hypothesis’: During laparoscopic surgery, hysterectomy (if done first) may contaminate instruments with tumour cells, and subsequent lymphadenectomy could disseminate these cells into the newly created retroperitoneal surgical field. Omitting the lymphadenectomy step interrupts this potential pathway.
The PHENIX trial also has limitations. The relatively high rate of adjuvant therapy (approximately 50%) may have attenuated the differences between groups. The predominant use of a single tracer like methylene blue and the lack of mandatory pathological ultrastaging are also noted. However, these choices were based on pragmatic considerations: the adjuvant therapy standard aligned with contemporary clinical practice and the STARS trial,8 and the use of low-cost, widely available tracers and flexible pathology protocols enhances the generalisability and applicability of the conclusions, particularly in many developing countries.
PHENIX is not lonely on the road it walks. The recently published SENTIX study by Cibula et al is an international multicentre single-arm trial assessing the safety of omitting systematic PLND after negative SLNB in early-stage cervical cancer.9 It reported a 2-year recurrence rate of 6.1% in 594 enrolled patients, meeting the prespecified endpoint of non-inferiority compared with a historical PLND cohort recurrence rate of 7%.
Another notable ongoing study is SENTICOL III, an international multicentre, randomised trial designed to validate the clinical value of SLNB in early-stage cervical cancer.10 This trial is taking place in France with the collaboration of the INCA-labelled French Cooperative Group ARCAGY-GINECO, the ENGOT network (European Network of Gynaecological Oncology Trial groups), GCIG (Gynecologic Cancer InterGroup) and CCRN (Cervix Cancer Research Network) accredited centres. It plans to enrol 950 patients, with co-primary endpoints of 3-year DFS and health-related quality of life. Patient enrolment has now been completed, and the outcomes are anticipated to conclude in 2026.
In summary, the PHENIX trial, with its high-level evidence, formally establishes SLNB as a new standard in the surgical management of early-stage cervical cancer. This large-scale, internationally significant study led by Chinese researchers will benefit countless patients worldwide. Furthermore, its unique design, rigorous execution and interpretation of unexpected findings provide valuable intellectual inspiration for the field of surgical oncology.