659: Genetic variation and recurrent prematurity among women receiving 17-alpha hydroxyprogesterone caproate for the prevention of recurrent preterm birth

Abstract

17-alpha-hydroxyprogesterone caproate (17P) reduces recurrent preterm birth (PTB) in some, but not all, women. We hypothesized that genetic polymorphisms affect variable response to 17P for the prevention of recurrent PTB. Prospective cohort of women with ≥1 prior singleton SPTB 16-36 weeks, recruited prospectively from a PTB prevention clinic over a 3 year period at a single tertiary referral center. This planned nested cohort analysis included the subset of women who received weekly 17P. Women were classified as a 17P responder or non-responder based on the difference in delivery GA between their 17P-treated and 17P untreated pregnancy. Non-responders were those delivering <3 weeks of the GA of their earliest prior PTB, and responders delivered ≥ 3 weeks later with 17P. Women were genotyped for single nucleotide polymorphisms (SNPs) from candidate genes in immunomodulatory, steroid hormone, and collagen pathways using the Illumina Infinium Global Screening Array. Logistic regression models evaluated the association between genotype (additive inheritance), clinical variables, and 17P responder status, using Plink v. 1.9. We adjusted models a priori for black race and 3rd trimester vaginal bleeding. 175 women met inclusion criteria; 27 (15%) were non-responders. The GA of each participant’s earliest PTB differed by responder status (median 26 wks for responders vs. 35 wks for non-responders, p<0.001), and responders achieved a median +11 wks longer with 17P, versus +1 wk for non-responders (p<0.001). Non-responders were more likely to have vaginal bleeding in the third trimester (25.9% vs. 10.8%, p=0.03). Similar proportions of responders and non-responders were black, had multiple previous PTB, smoked, and had a short cervix <20mm. All SNPs met quality control filters. We evaluated 18 SNPs in 9 genes (PGR, CYP3A5, IL6ST, NOS2, FLT1, COL5A1, MMP9, COL9A1, ESR2) and did not find any significant associations between maternal genotype and 17P response status (Table). Maternal genotype was not associated with non-response to 17P in this high-risk cohort of women with at least one prior SPTB. Pregnancy outcomes in women receiving 17P remain difficult to predict using clinical and genetic data.