A Comparative Observational Study Evaluating the Incidence Rate of Endometrial Cancer in Women aged 50 Years and Over Who Use Low dose Vaginal Estrogen Unopposed by a Progestogen: A Post-authorization Safety Study in the United States and Sweden (Low dose vaginal estrogen and endometrial cancer)

Study type

• Study topic:
  Disease /health condition

  Human medicinal product
• Study type:
  Non-interventional study
• Scope of the study:
  Assessment of risk minimisation measure implementation or effectiveness
• Data collection methods:
  Secondary use of data

Non-interventional study

• Non-interventional study design:
  Cohort

Study drug and medical condition

• Medicinal product name:
  PREMARIN

  ESTRING

  VAGIFEM
• Medicinal product name, other:
  Oestring, Yuvafem, Imvexxy
• Anatomical Therapeutic Chemical (ATC) code:
  (G03CA03) estradiol

  estradiol

  (G03CA57) conjugated estrogens

  conjugated estrogens
• Medical condition to be studied:
  Endometrial cancer metastatic

  Endometrial cancer recurrent

  Endometrial cancer stage 0

  Endometrial cancer stage I

  Endometrial cancer stage II

  Endometrial cancer stage III

  Endometrial cancer stage IV

  Endometrial cancer

Population studied

• Short description of the study population:
  The study includes women aged >50 years with an intact uterus and no prior use of vaginal estrogens. The exposed group of LDVE will include new users who had ≥1 dispensing of LDVE (daily doses ≤10 mcg/day estradiol or ≤0.3 mg/day conjugated estrogen) unopposed by a progestogen, regardless of prior use of other hormone therapy between 01 January 2007 and 31 December 2020 (or most recent available date) in the HIRD, and between 01 July 2007 to 31 December 2019 (or most recent available date) in the Swedish National Registers (the exposure ascertainment period). Similarly, two additional exposure groups will be created for MDVE (daily doses >10 mcg – 25 mcg/day estradiol or >0.3 mg – 0.45 mg/day conjugated estrogen) and, in the HIRD only, HDVC (daily doses >0.45 mg/day conjugated estrogen). These two additional groups will also include women with ≥1 dispensing of MDVE or HDVC, respectively, regardless of prior use of hormone therapy. The difference in study periods between the HIRD and Swedish databases is due to the difference in lag time for data availability ; between the two data sources. The comparator groups are: (1) women with ≥1 gynecological visit on or after age 50 with no prior use of vaginal estrogen at the time of their gynecological visit (i.e., non-users) and (2) new users of E+P HT with no prior use of hormone therapy. Women with less than 12 months of continuous health plan enrollment in the HIRD or less than 24 months of medical history in Sweden prior to the index date, without a uterus, or with a history of endometrial cancer will be excluded.
• Age groups:
  • Adults (46 to < 65 years)
  • Adults (65 to < 75 years)
  • Adults (75 to < 85 years)
  • Adults (85 years and over)
• Estimated number of subjects:
  152941

Study design details

• Study design:
  This is a non-interventional, retrospective cohort study among women aged ≥50 years using a US ; healthcare claims data source (the HIRD) and longitudinal data collected from five Swedish ; National Registers. The three exposed groups are new users of vaginal estrogen grouped into 3 ; dose categories.
• Main study objective:
  Estimate and compare IR of endometrial cancer in postmenopausal women aged >50 years with a uterus regardless of prior hormone therapy (any estrogen, progestin, E+P HT or opposed estrogen hormone therapy) in the following groups: a. New users of LDVE vs. non-users of vaginal estrogen b. New users of MDVE vs. non-users of vaginal estrogen c. New users of HDVC vs. non-users of vaginal estrogen
• Setting:
  The study population consists of women in the HIRD and Sweden who are at least 50 years of age. For subjects using a vaginal estrogen or E+P HT, cohort qualification status will be assessed for the first study drug dispensing (index date) during the exposure ascertainment period (see Table 1) based on the inclusion and exclusion criteria described below. For the purpose of the primary objective, non-users of vaginal estrogen, subjects with at least one gynecological visit after age 50 and without any prior or current use of vaginal estrogens at the time of their gynecological visit (index date for non-users) will be included in this population regardless of their prior use of hormone therapy. Subjects may have multiple qualifying gynecological visits. For the purpose of the secondary ; objective, E+P HT new users with no prior use of hormone therapy will be included.
• Outcomes:
  The endpoint is the occurrence of endometrial cancer. In the United States, the cases of endometrial cancer will be identified using a validated algorithm in the United States and using the Swedish Cancer Register in Sweden.
• Data analysis plan:
  Propensity score matching will be applied to each cohort pair (comparison) with a 1:4 exposed:comparator matching ratio. Propensity scores will be estimated separately for each comparison, using logistic regression and a set of baseline covariates. Descriptive statistics and all additional analyses will be performed on the PS-matched cohorts. Baseline characteristics will be compared across the study cohorts to assess comparability. The IR of endometrial cancer and 95% confidence intervals (CIs) will be computed for each study cohort. Adjusted hazard ratios (HRs) will be calculated using Cox proportional hazards models to compare the hazard of endometrial cancer between 1) unopposed LDVE and non-users, 2) unopposed MDVE and non-users, 3) unopposed HDVC with non-users (US only).