Prevalence, characteristics, and risk factors of occult uterine cancer in presumed benign hysterectomy

Abstract

Background: Occult uterine cancer at the time of benign hysterectomy poses unique challenges in patient care. There is large variability and uncertainty in estimated risk of occult uterine cancer in the literature and prior research often did not differentiate/include all subtypes.

Objectives: To thoroughly examine the prevalence of occult uterine cancer in a large population-based sample of women undergoing hysterectomy for presumed benign indications and to identify associated risk factors.

Study design: Using the New York Statewide Planning and Research Cooperative System database, we identified 229,536 adult women who underwent an inpatient or outpatient hysterectomy for benign indications during the period October 1, 2003 to December 31, 2013 at civilian hospitals and ambulatory surgery centers throughout the state. Diagnosis of corpus uteri cancer within 28 days after the index hysterectomy was determined using linked state cancer registry data. We estimated the prevalence of occult uterine cancer (overall and by subtype) and developed and validated risk prediction models using a random split sample approach.

Results: Overall, 0.96% (95% confidence interval: 0.92-1.00%) of the women had occult uterine cancer, including 0.75% (95% confidence interval: 0.71-0.78%) with endometrial carcinoma and 0.22% (95% confidence interval: 0.20-0.23%) with uterine sarcoma. The prevalence of leiomyosarcoma was 0.15% (95% confidence interval: 0.13-0.17%). Seventy-one percent of the endometrial carcinomas and 58.0% of the uterine sarcomas were at localized stage. The risk for occult uterine cancer ranged from 0.10% in women aged 18-29 years to 4.40% in women aged ≥75 years; and varied from 0.14% in women undergoing hysterectomy for endometriosis to 0.62% for uterine fibroids and 8.43% for postmenopausal bleeding. The risk of occult uterine cancer was also significantly associated with race/ethnicity, obesity, comorbidity, and personal history of malignancy. Prediction models incorporating these risk factors had high negative predictive values (99.8% for endometrial carcinoma and 99.9% for uterine sarcoma) and good rule-out accuracy despite low positive predictive value.

Conclusions: In women undergoing hysterectomy for presumed benign indications, 0.96% had unexpected uterine cancer. Patient characteristics such as age, surgical indication, and medical history may help guide risk stratification.

Keywords: endometrial carcinoma; hysterectomy; leiomyosarcoma; occult uterine cancer; risk prediction; uterine sarcoma.

SPARCS = New York Statewide Planning and Research Cooperative System. ^a. To protect patient confidentiality, identifiers and dates were redacted in SPARCS data if an encounter was abortion- or HIV-related. Therefore, these encounters could not be linked to cancer registry data.

Figure 1.

Adjusted association between patient characteristics and the risk of having occult uterine cancer, development sample (n=153032) 1A. Risk for occult endometrial carcinoma 1B. Risk for occult uterine sarcoma RR, risk ratio; LCL, lower confidence limit of the 95% confidence interval; UCL, upper confidence limit of the 95% confidence interval. Notes: 1) We combined the two youngest age categories (18-29 years and 30-34 years) due to small number of patients with occult uterine cancer in these age categories. 2) Because no patients with endometriosis or benign neoplasm of ovary had occult uterine sarcoma, they were combined with other female genital disorders in the risk prediction model for occult uterine sarcoma. 3) “Other solid tumor cancer” referred to cancer of solid tumor other than breast, colon, melanoma, and urinary organs. 4) Area under receiver operating characteristic (ROC) curve was 0.87 and 0.82 for the risk prediction model for occult endometrial carcinoma and occult uterine sarcoma, respectively.

Figure 1.

Adjusted association between patient characteristics and the risk of having occult uterine cancer, development sample (n=153032) 1A. Risk for occult endometrial carcinoma 1B. Risk for occult uterine sarcoma RR, risk ratio; LCL, lower confidence limit of the 95% confidence interval; UCL, upper confidence limit of the 95% confidence interval. Notes: 1) We combined the two youngest age categories (18-29 years and 30-34 years) due to small number of patients with occult uterine cancer in these age categories. 2) Because no patients with endometriosis or benign neoplasm of ovary had occult uterine sarcoma, they were combined with other female genital disorders in the risk prediction model for occult uterine sarcoma. 3) “Other solid tumor cancer” referred to cancer of solid tumor other than breast, colon, melanoma, and urinary organs. 4) Area under receiver operating characteristic (ROC) curve was 0.87 and 0.82 for the risk prediction model for occult endometrial carcinoma and occult uterine sarcoma, respectively.