Radical hysterectomy or chemoradiotherapy for clinically early-stage cervical cancer with suspicious lymph nodes on imaging: a retrospective cohort study

Objective: The optimal treatment of clinically early-stage cervical cancer with suspicious lymph nodes on pretreatment imaging is unclear. Therefore, we aimed to compare surgery (i.e., radical hysterectomy and pelvic lymphadenectomy±adjuvant therapy) with primary chemoradiotherapy as treatment strategies in this patient group regarding recurrence-free, overall survival and toxicity. Methods: Women diagnosed between 2009–2017 with the International Federation of Gynecology and Obstetrics (2009) stage IA–IIA and suspicious nodes based on radiologic assessment of pretreatment imaging were retrospectively selected from the Netherlands Cancer Registry. Cox proportional hazard was used to estimate survival and logistic regression for toxicity. Inverse probability weighting was used to correct for confounding. Grade ≥2 surgery-related (≤30 days) and grade ≥3 chemotherapy or radiotherapy-related (≤6 months) toxicity were collected. Missing data were imputed. Results: Of 330 patients included, 131 (40%) received surgery (followed by adjuvant therapy in 54%) and 199 (60%) chemoradiotherapy. Pathological nodal status was known in 100% of the surgery group and 32% (n=63) of the chemoradiotherapy group, of whom 43% (56/131) and 89% (56/63), respectively, had metastases. After adjustment for confounders, the recurrence-free survival (hazard ratio [HR]=0.67; 95% confidence interval [CI]=0.34–1.31) and overall survival (HR=0.75; 95% CI=0.38–1.47) were not significantly different between both groups, while surgery was associated with more toxicity (odds ratio=2.82; 95% CI=1.42– 5.60), mainly surgery-related. Conclusion: In patients with clinically early-stage cervical cancer and suspicious nodes on imaging, surgery and primary chemoradiotherapy yielded comparable results in terms of survival, whereas surgery might be associated with more (surgery-related) short-term toxicity.