Genotype-informed docosahexaenoic acid dosing based on the FADS Indel to reduce preterm and prolonged gestation: A secondary analysis of the ADORE randomized trial.

backgroundSpontaneous early preterm birth (EPB < 34 weeks) is a major contributor to newborn morbidity and mortality; it depends on ancestry, and preventive measures are limited. The insertion (I) allele of the fatty acid desaturase (FADS) Indel polymorphism (rs66698963) modulates FADS1 expression, and increases arachidonic acid synthesis, enhancing parturifacient prostaglandins. Omega-3 docosahexaenoic acid (DHA) inhibits this pathway. Whether FADS Indel genotype predicts response to DHA dose in pregnancy is unknown. METHODS AND

findingsTo determine whether FADS Indel genotype modifies the effect of DHA dose on gestational length, we conducted a post hoc secondary analysis of the Assessment of DHA on Reducing Early Preterm Birth (ADORE) trial (NCT02626299). The FADS Indel, the only genetic variant measured, was selected on mechanistic grounds and genotyped blind to outcome. The analysis included all participants consenting to genetic analysis (12-20 weeks gestation) at three US medical centers. Participants received 200 or 1,000 mg/d DHA and were genotyped for FADS Indel (D/D, I/D, I/I). Primary outcomes were EPB and preterm birth (PB, <37 weeks) analyzed using treatment-specific Bayesian binomial models. Cervical length was assessed in 574 participants. Women (n = 10,497) were screened for eligibility from June 8th 2016, to March 13th 2020, 1,032 (1,000 mg, n = 540; 200 mg, n = 492) completed the ADORE trial. A total of 761 (1,000 mg, n = 396; 200 mg, n = 365) are included in this secondary analysis. Participants carrying at least one FADS Indel I allele (I/I + I/D) receiving low-dose DHA had 4.3% EPB (9/214; 95% Credible Interval (CrI) [2.0%, 7.3%]) versus 0.8% in the D/D group (1/151; 95% CrI [0.04%, 2.7%]; Posterior Probability (PP) = 0.99), and 14.5% PB (31/214; 95% CrI [10.1%, 19.5%]) versus 8.7% in the D/D group (13/151; 95% CrI [4.7%, 13.6%]; PP = 0.96). High-dose DHA was associated with reduced EPB in the I-allele group (1.1%; 2/205; 95% CrI [0.2%, 2.8%]; PP = 0.99), and PB (9.3%; 19/205; 95% CrI [5.7%, 13.6%]; PP = 0.95), with little effect in the D/D group. Birth at ≥40 weeks was more frequent in the high-dose D/D group. Black women had higher rates of PB and EPB, have higher rates of the I allele, and benefited most from 1,000 mg/d dose. These dose effects credibly differ by genotype for EPB and prolonged gestation (genotype × dose interaction PP = 0.99 and 0.98). Both genotypes reduced PB similarly with high-dose DHA (PP = 0.56). Although genetic analysis was anticipated and genotyping was completed blind to outcome, this was a post hoc analysis and the genotype-by-dose comparison was not a prespecified trial aim. Some subgroups contained few PB and EPB events, so these estimates, particularly for PB, should be interpreted cautiously and confirmed prospectively.

conclusionsThe FADS Indel is a functional, mechanistically grounded polymorphism whose frequency varies with race/ethnicity. Preterm and early preterm birth were lower in the I-allele group receiving high-dose DHA (1,000 mg/d), while prolonged gestation was more frequent in the D/D group receiving high-dose DHA. FADS Indel genotypes predicted the direction of the response to high-dose DHA for EPB and prolonged gestation, whereas its modest PB benefit was shared by both genotypes. Analysis by retrospectively pairing high dose to I-carrier genotype and low dose to D/D was associated with the highest proportion of births at 37-40 weeks. These findings identify the FADS Indel as a potential predictive biomarker for DHA dose response and support a genotype-informed, precision-medicine strategy to optimize gestational length by reducing preterm birth in the I-allele group and prolonged gestation in the D/D group across racially/ethnically diverse populations.

trial registrationClinical trial identification number: NCT02626299, Date of registration: December 8, 2015, Date of Initial participant enrolled: June 8, 2016 https://clinicaltrials.gov/study/NCT02626299.