Does an epidemiological comparison support a common cellular lineage for similar subtypes of postmenopausal uterine and ovarian carcinoma? The Norwegian Women and Cancer Study

Abstract

Uterine and ovarian carcinomas have the same major histological subtypes, but whether they originate from the same cell types is a matter of ongoing debate. Uterine and ovarian endometrioid and clear cell carcinoma (ECC) and uterine and ovarian serous carcinoma (SC) may originate in the same location, or share a common lineage of differentiation. Epidemiologically, a common cellular lineage should be reflected in similar risk associations, and we explored the similarity of uterine and ovarian ECC and uterine and ovarian SC. We included 146,316 postmenopausal participants from the Norwegian Women and Cancer Study. Exposure information was taken from self-administered questionnaires, and cancer cases were identified through linkage to the Cancer Registry of Norway. Hazard ratios with 95% confidence intervals for uterine and ovarian carcinoma and their subtypes were calculated using multivariable Cox regression models, and a Wald test was used to check for heterogeneity. During 1.6 million person-years, 1,006 uterine and 601 ovarian carcinomas were identified. Parity, total menstrual lifespan, body mass index and smoking were differentially associated with total uterine and total ovarian carcinoma (p = 0.041, 0.027, <0.001 and 0.001, respectively). The corresponding associations for uterine and ovarian ECC did not differ significantly (p > 0.05). Smoking was differentially associated with uterine and ovarian SC (p = 0.021). Our epidemiological analyses do not contradict a common differentiation lineage for uterine and ovarian ECC. Uterine and ovarian SC are less likely to be of a common lineage of differentiation, based on their difference in risk associated with smoking.

Risk of high-grade serous ovarian cancer associated with pelvic inflammatory disease, parity and breast cancer

“…These unspecified (other) categories can make up as much as 30% of all ovarian cancer cases [13,31,52], and are predominantly high-grade serous [46]. Other authors have emphasized the importance of histopathological review [2] and noted that without histopathological review, estimates of risk may be less precise [59]. Strengths include the size of the study population; accurate recording of diagnoses and procedures in administrative databases rather than relying on personal report which could be subject to error and recall bias.…”