Initial Treatment and Outcomes of Complete Hydatidiform Mole in Women 40 Years or Older: A Multicenter Cohort Study

Objectives: To evaluate the potential associations of the type of complete hydatidiform mole (CHM) initial treatment (hysterectomy or uterine evacuation) with GTN development, need for chemotherapy, and treatment outcome in women aged ≥ 40 years. Methods: This multicentric retrospective cohort study included women ≥ 40 years with CHM, initially treated between 1990 and 2018, at four different centers. Data collected included patient demographics and clinical characteristics. The outcome variables were post-CHM GTN development, need for chemotherapy for hCG normalization, surgical complications, and time to remission. Univariate and multivariate analyses were performed using chi-square, Mann-Whitney, Fisher's exact tests, and Poisson regression. Results: 275 women with CHM aged ≥ 40 years were included in the analysis. Median patient age was significantly higher among hysterectomy patients (47 × 44 years, p = 0.01). Multivariate analysis showed that compared with uterine evacuation (244/275, 89%), hysterectomy (31/275, 11%) was associated with an 83% lower risk of GTN [RR = 0.17 95% CI = (0.04-0.71); p = 0.015] and a 92% lower risk of requiring chemotherapy [RR: 0.08 (0.01-0.64), p = 0.016]. Median time to hCG normalization did not statistically differ between treatments. No significant differences were observed between hysterectomy and uterine evacuation in terms of FIGO staging (p = 0.221) or prognostic risk score (p = 0.576). Resistance to first-line chemotherapy (17/72; 23.6%) and relapse (3/72; 4.1%) were observed only in patients undergoing initial uterine evacuation. Hysterectomy complications occurred in 45.1% (14) of the patients. Conclusions: CHM initial treatment with hysterectomy was associated with a lower risk for GTN occurrence and need for chemotherapy in women aged 40 years or older. However, shared decision-making about surgery should be tailored to each patient and their risk factors and preferences. Further, larger controlled studies are required to support our findings.