Prediction of early pregnancy maternal thyroid impairment in women affected with unexplained recurrent miscarriage

Human Reproduction, Vol.26, No.6 pp. 1324– 1330, 2011 Advanced Access publication on March 23, 2011 doi:10.1093/humrep/der069 ORIGINAL ARTICLE Early pregnancy Prediction of early pregnancy maternal thyroid impairment in women affected with unexplained recurrent miscarriage 1 Ignazio Silone 100, Operative Unit of Spontaneous Recurrent Abortion, Rome, Italy 2Unit of Obstetric and Gynecology, University of Rome TorVergata, Fatebenefratelli Hospital “San Giovanni Calibita”, Rome, Italy 3SeSMIT-Fatebenefratelli Association for Research, Rome, Italy 4 Casa di Cura San Raffaele-Cassino, Epidemiology and Biostatistics, Rome, Italy 5Unit of Obstetric and Gynecology II, San Giovanni Addolorata Hospital, Rome, Italy 6Department of Internal Medicine, University of Rome TorVergata 7Section of Reproductive Endocrinology, Fatebenefratelli Hospital “San Giovanni Calibita”, Isola Tiberina, 00186 Rome, Italy *Correspondence address. E-mail: Submitted on May 17, 2010; resubmitted on February 8, 2011; accepted on February 18, 2011 background: Proper maternal thyroid function is necessary for a successful pregnancy. In order to identify women who may experience miscarriage due to transient impairment of the pituitary– thyroid axis in early pregnancy, we aimed to investigate the ratio between basal and peak thyroid stimulating hormone (TSH) [following stimulus with thyrotrophin-releasing hormone (TRH)] in euthyroid women with unexplained recurrent miscarriage (RM). methods: We have established a ‘iTSHa index’ (TSH increase after TRH adjusted for the levels of basal TSH), determining TSH serum levels at time 0 and 20 min after TRH stimulus in 463 consecutive women attending two antenatal care units for two or more miscarriages occurring within the first 10 weeks of pregnancy. results: The mean basal TSH serum levels were higher (P , 0.001) in RM women [2.1 mIU/ml; 95% confidence interval (CI): 2.0 –2.2] compared with the controls (1.3 mIU/ml; 95% CI: 1.2– 1.4). Establishing serum TSH at an individual level, a large overlap was observed and the receiver operating characteristic curves did not allow us to find an optimal cut-off point with an adequate sensitivity/specificity ratio. Therefore, we suggest a novel statistical model, the ‘iTSHa index’ (available on www.afar.it/tsh-trh-miscarriage), that is capable of identifying women with RM due to transient thyroid function impairment of the early pregnancy, in particular when baseline serum TSH is less than 1.5 mIU/ml, i.e. well below the conventional upper cut-off indicated as ‘safe’ in those who want to conceive. conclusions: A transient impairment of thyroid function in early pregnancy may cause an inadequate adaptation to the increased thyroid requirement and may be implicated in RM. The evaluation of the proposed iTSHa index, if validated in a larger cohort of patients, may provide information useful to identifying a subset of healthy women, without evidence of thyroid dysfunction or autoimmunity and a TSH in the low-normal reference range, who may be at risk of RM. Key words: TSH / iTSHa / recurrent miscarriage / thyroid / pregnancy Introduction The relevance of suitable maternal thyroid function in early pregnancy is well established (Glinoer, 1993, 1997). In particular, during this period, there is a physiological increase in maternal thyroid hormone requirements and women with hypothyroidism may experience several adverse outcomes including miscarriage, impaired neuropsychological development of the fetus, premature birth and increased fetal mortality (Haddow et al., 1999; Casey et al., 2005). For this reason, consensus guidelines for the investigation and medical treatment of recurrent miscarriage (RM) (Jauniaux et al., 2006), and in particular for management of thyroid function, must be followed during pregnancy and in the post-partum period (Abalovich et al., 2007). The thyroid function of all women considering & The Author 2011. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please email: Alessandro Dal Lago 1, Elena Vaquero 2, Patrizio Pasqualetti 3,4, Natalia Lazzarin 2, Caterina De Carolis 5, Roberto Perricone 6, and Costanzo Moretti 6,7,* 1325 Thyroid function and recurrent miscarriage Table I Characteristics of women with RM and control group. Women with primary RM (n 5 307) Women with secondary RM (n 5 156) Control group (n 5 101) ........................................................................................ Age (years) mean (SD) 35.1 (3.4) 36.6 (4.0) 35.6 (3.9) BMI (kg/m2) mean (SD) 24.6 (1.6) 25.2 (1.9) 23.4 (1.5) Episodes of miscarriage: median (min–max) 3 (2–15) 3 (2–5) Week of miscarriage: mean (SD) 13 (3) 10 (3) Thyroid-stimulating hormone (mUI/ml)a 2.1 (1.1) 2.2 (1.1) 1.3 (0.5) Free thyroxine (pg/ ml)b 10.2 (4.0) 9.9 (3.6) 11.9 (1.4) Free triiodothyronine (pg/ml)c 2.9 (0.7) 3.2 (0.9) 2.8 (0.3) TPO-Ab (IU/ml)d ,115 ,115 ,115 e TG-Ab (IU/ml) ,65 ,65 ,65 aPL Negative Negative Negative HOMA-IRf ,2.7 ,2.7 ,2.7 a mUI/ml (0.20 – 3.5). pg/ml (8.5– 20.0). c pg/ml (2 –5). d Anti thyreoperoxidase antibodies: negative , 115 IU/ml. e Anti thyreoglobulin antibodies: negative , 65 IU/ml. f Homeostasis Model Assessment ([Insulin (mU/mL) × Glucose (mmol/l)]/22.5): negative , 2.7. b of unexplained RM and to assess their diagnostic value in identifying women who may develop a transient impairment of thyroid function, thus increasing their risk of pregnancy disruption. Materials and Methods Patients The study was conducted in two centers and included 463 women with a history of unexplained recurrent spontaneous abortion and 101 healthy controls of similar age (35.6 + 3.8 years for patients versus 35.6 + 3.9 years for controls) and BMI (24.8 + 1.7 versus 23.4 + 1.5 kg/m2), who had had prior normal pregnancies (Table I). The study design aimed at re-evaluating two groups of women matching for age, BMI and HOMA (Homeostatic Model Assessment) index, in a ratio of 4.6:1 between cases and controls. The protocol received institutional Ethics Committee approval and both patients and controls gave their informed consent after the scope of the study was explained to them. The 463 non-pregnant women attended the Antenatal Outpatient Clinic of the University of Tor Vergata Hospital in Rome and the Outpatient Clinic of the Ambulatory Care Unit of the Rome C district, had a history of two or more consecutive miscarriages. The majority of our sample (85%) was constituted by women with three or more episodes of miscarriage and a smaller proportion (15%) had had only two episodes of miscarriage but shared identical clinical characteristics. In the studied population, 307 women were affected by primary RM, 156 by secondary RM (Table I). All women evaluated in this retrospective study, affected with idiopathic RM, represented the 26% of the whole population attending our centers (Fig. 1 (...truncated)