Non-Contraceptive Progestins and Risk of Venous Thromboembolism: A Nested Case-Control Study of the Marketscan Databases

Abstract

<h3>Study Objective</h3> To evaluate for a dose-response relationship in the association between progestins used for non-contraceptive indications (e.g., endometriosis) and venous thromboembolism (VTE). <h3>Design</h3> Nested case-control study. <h3>Setting</h3> The IBM® MarketScan® Databases include routinely collected inpatient, outpatient, and pharmacy claims data among privately insured individuals in the US. <h3>Patients or Participants</h3> Records of 45 million women aged 15-49 years old were evaluated from 2009-2018. Cases with VTE (n=21,405) were identified by validated ICD-9/10 codes and excluded for recent pregnancy, recent estrogen prescription, or indications for ultra-high dose progestins (e.g., endometrial intraepithelial neoplasia). Controls (n=107,025) were individually age-matched 5:1 by risk set sampling. <h3>Interventions</h3> N/A. <h3>Measurements and Main Results</h3> Oral progestin exposures and covariates were defined by prescription claims and ICD-9/10 codes, respectively, within 1 year before the index date. Multivariate conditional logistic regression models determined odds ratios with 95% confidence intervals adjusted for VTE risk factors. The dose-response relationship was assessed by two separate exploratory analyses. The first analysis compared new and chronic use with non-use. The second analysis evaluated progestin exposure as a weighted average daily dose. Significance was set at p<0.01 to allow for multiple comparisons. Odds of VTE were significantly increased for current use of norethindrone acetate (new users aOR 4.97 [2.96-8.35], n=83; chronic users aOR 2.28 [1.48-3.51], n=118) and current use of medroxyprogesterone acetate (new users aOR 2.08 [1.47-2.96], n=194; chronic users aOR 1.76 [1.18-2.63], n=145), compared with non-use. Comparisons between new and chronic use were not significant. Odds of VTE were directly proportional to progestin dose: +22% (aOR 1.22 [1.09-1.36]) for each 2.5 mg/day increase in norethindrone acetate and +19% (aOR 1.19 [1.08-1.30]) for each 5 mg/day increase in medroxyprogesterone acetate. <h3>Conclusion</h3> Non-contraceptive progestins are associated with increased VTE risk. This study provides the first assessment and support of the dose-response relationship. Clinicians should consider using the lowest effective dose to balance risks and benefits.