The impact of levothyroxine therapy on pregnancy and neonatal outcomes in euthyroid pregnant women with thyroid autoimmunity: A systematic review, meta-analysis and trial sequential analysis

Background: At present, only one systematic review has investigated the effect of levothyroxine (LT4) in the treatment of euthyroid pregnant women with thyroid autoimmunity, but some problems [such as merging different types of research for meta-analysis, lacking neonatal outcomes, and so on] exist in this study, satisfactory results can not be provided. So, this systematic review was performed to investigate the effect of LT4 in euthyroid pregnant women with thyroid autoimmunity, in the hope of providing more comprehensive evidence for clinical use. Methods: Medline (Ovid), Embase (Ovid), and Cochrane Central Register of Controlled Trials were electronically searched from database inception to March 2022. We included cohort studies and RCTs that evaluated the impact of LT4 therapy on pregnancy and neonatal outcomes in euthyroid pregnant women with thyroid autoimmunity. Meta-analyses of different types of studies were performed separately, and meta-analyses were further performed by only including researches with low and moderate risk of bias. We used the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) approach to evaluate the quality of evidence, and used TSA to test the sufficiency of the evidence. Results: Finally, 2,901 euthyroid pregnant women with thyroid autoimmunity in six RCTs and five cohort studies were included. In all outcomes, no statistically significant differences were found between LT4 group and control group, including miscarriage [RR = 0.85, 95%CI (0.69,1.05), p = 0.14, I (2) = 1%], preterm birth [RR = 0.80, 95%CI (0.59,1.08), p = 0.14, I2 = 0%], preeclampsia [RR = 0.68, 95%CI (0.12, 3.91), p = 0.66, I (2) = 0%], placenta abruption [Peto’ OR = 0.14, 95%CI (0.00, 6.94), p = 0.32, I (2) = 0%], birth weight [MD = -36.00, 95%CI (-170.41, 98.41), p = 0.60, I (2) = 0%], gestational age at delivery [MD = -0.10, 95%CI (-0.61, 0.41), p = 0.70, I (2) = 0%] and neonatal admission [RR = 1.33, 95%CI (0.21, 8.58), p = 0.76, I (2) = 0%]. The results for all outcomes were insufficient and inconclusive as demonstrated by TSA. The GRADE assessments showed that the quality of evidence of 4 outcomes (miscarriage, preterm birth, birth weight and gestational age at delivery) were moderate, and 3 outcomes (preeclampsia, placenta abruption and neonatal admission) were low or very low. Conclusion: For pregnancy and neonatal outcomes in euthyroid pregnant women with thyroid autoimmunity, we did not find benefit of LT4 treatment in this study. Systematic Review Registration: https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022346745, identifier CRD42022346745.