Risk of epithelial ovarian cancer Type I and II after hysterectomy, salpingectomy and tubal ligation-A nationwide case-control study

The proposed different origins and pathways to of the dualistic model of epithelial ovarian cancer (EOC) may affect and alter the potential risk reduction related to hysterectomy, salpingectomy and tubal ligation. The aim of our study was to analyze associations between hysterectomy, salpingectomy or tubal ligation and risk reduction of EOC Type I and II. In this nationwide register-based case-control study, women diagnosed with EOC, Fallopian tube or primary peritoneal cancer between 2008 and 2014 were included. Cases were classified into Type I and II according to histology and predefined criteria. The exposure variables: hysterectomy, salpingectomy and tubal ligation were identified from national registries. Conditional logistic regression analyses were performed to evaluate associations between Type I and II EOC and the exposure variables. Among 4669 registered cases, 4040 were eligible and assessed for subtyping resulting in 1033 Type I and 3007 Type II. Ten controls were randomly assigned to each case from the register of population. In regression analyses, women with previous salpingectomy had a significantly lower risk of EOC Type II (odds ratio [OR] 0.62; 95% confidence interval [95%CI] 0.45-0.85) but not Type I (OR 1.16; 95%CI 0.75-1.78). Hysterectomy was associated with a reduced risk of both EOC Type I (OR 0.71; 95%CI 0.52-0.99) and Type II (OR 0.81; 95%CI 0.68-0.96). Similar estimates were obtained for tubal ligation, although without statistical significance. The association between salpingectomy and reduced risk of EOC Type II supports the proposed theory of high-grade serous cancer originating from the tubal fimbriae.

Automated summary

The study found that salpingectomy is associated with a reduced risk of EOC Type II, while hysterectomy is associated with a reduced risk of both EOC Type I and Type II. Tubal ligation also shows a trend towards reduced risk, although without statistical significance.