Thyroxine treatment modified in infertile women according to thyroxine‐releasing hormone testing: 5 year follow‐up of 283 women referred after exclusion of absolute causes of infertility

Human Reproduction Vol.18, No.4 pp. 707±714, 2003 DOI: 10.1093/humrep/deg142 Thyroxine treatment modi®ed in infertile women according to thyroxine-releasing hormone testing: 5 year follow-up of 283 women referred after exclusion of absolute causes of infertility W.Raber1,3, P.Nowotny1, E.Vytiska-Binstorfer2 and H.Vierhapper1 1 Department of Medicine III, Division of Endocrinology and Metabolism and 2Department of Gynecology, University of Vienna, Vienna, Austria 3 To whom correspondence should be addressed at: Department of Medicine III, Division of Endocrinology & Metabolism, University of Vienna, AKH, Waehringer Guertel 18±20, 1090 Wien, Austria. E-mail: BACKGROUND: Mild hypothyroidism may contribute to disturbed reproductive function. We hypothesized that frequent thyroxine-releasing hormone (TRH) testing to ®ne-tune thyroxine (T4) therapy instituted upon every TRH-induced thyroid-stimulating hormone (TSH) rise above the mean of a healthy population (i.e. 15 mIU/l) would improve fecundity compared with historical data. METHODS: In a cohort of 283 infertile women followed over 5 years, we assessed (i) pregnancy, abortion and delivery rates, (ii) thyroid function over time in women who conceived compared with those who did not, and (iii) various thyroid parameters with respect to fertility. RESULTS: Overall conception rate of 37% was higher (P < 0.05) than previously reported and independent of thyroid function prior to T4 therapy, thyroxine dose or elevated thyroid autoantibodies. Never achieving basal TSH <2.5 IU/l or TRH-stimulated TSH <20 mIU/l with T4 therapy resulted in lower conception rates (P < 0.05). Median time to conception was 6 months, but 18 months in women who declined TRH testing (P < 0.02). Overall abortion rate was 9%. Only ®rst trimester miscarriages occurred. CONCLUSIONS: Based on the presented protocol, high pregnancy and parturition rates were observed. Whether this is due to early T4 therapy remains to be determined. Abortions appeared to be associated with higher TSH but not with elevated thyroid antibodies. Key words: cohort study/infertility/mild hypothyroidism/thyroxine therapy/TRH test Introduction Approximately 10±15% of all married couples in the USA are childless (Davajan and Israel, 1998). Examination of parish registries between 1550 and 1850 in Cambridge, UK, from an era that predated contemporary methods of sterilization and contraception, revealed that 8% of married women were sterile throughout their reproductive life (Trussel and Wilson, 1985). Population-based infertility data of women with subclinical hypothyroidism are not available. Hyperprolactinaemia due to increased hypothalamic thyroxine-releasing hormone (TRH) secretion was suggested 17 years ago as a cause of infertility in hypothyroidism (Thomas and Reid, 1986). However, pulsatile secretion of LH, FSH and prolactin is not altered in infertile women with overt hypothyroidism (Tomasi et al., 1997) or with a TRH-stimulated thyroid-stimulating hormone (TSH) >20 mIU/l in the presence of a normal basal TSH (Bals-Pratsch et al., 1997). In addition, basal and TRH-stimulated serum prolactin concentrations in mild and severe hypothyroidism are not different from euthyroid subjects (Bigos et al., 1978). This corresponds to recent epidemiological and clinical observations ã European Society of Human Reproduction and Embryology of a large number of patients demonstrating that hypothyroidism is associated with only minor menstrual disturbancies and minimal changes in the serum prolactin concentrations (Vanderpump et al., 1998; Krassas et al., 1999; Raber et al., 2003). Thyroid hormones, on the other hand, play a role in the modulation of the LH- and FSH-mediated control of granulosa cell function. There are experimental data of both stimulatory (Maruo et al., 1987; Wakim et al., 1995) and inhibitory effects (Channing et al., 1976; Wakim et al., 1993; Cecconi et al., 1999) of thyroid hormones on mammalian granulosa cell gonadotropin-induced steroidogenesis. These controversial (stimulatory or inhibitory) effects of thyroid hormones may be due to the different responsiveness to tri-iodothyronine (T3) of granulosa cells isolated from follicles at different stages of antral development, with small and medium follicles displaying a higher number of T3 binding sites than large antral follicles (Maruo et al., 1993). They may, on the other hand, also be due to the different species studied, and/or due to different culture conditions used (Cecconi et al., 1999). In fact, 707 W.Raber et al. stimulatory effects of T3 on mammalian granulosa cell steroidogenesis have been suggested to depend upon the presence of insulin in the culture medium (Channing et al., 1976; Wakim et al., 1993; Cecconi et al., 1999). As obtained from patients undergoing therapeutic abortions at 7±8 weeks gestation, thyroxine (T4) and T3 in ®rst trimester placentas were ampli®ers of differentiated trophoblast function (Maruo et al., 1991). In addition, data from clinical studies have demonstrated that thyroid hormone replacement therapy increased the success rate of ovulation induction by clomiphene citrate in women with subclinical hypothyroidism (Maruo et al., 1993). Taken together, hypothyroidism may, even at an early stage, have an important impact on conception. Once pregnancy has occurred, thyroid hormones contribute to the stability of the feto-placental unit, protecting from early loss of the conceptus (Maruo et al., 1992; La Marca et al., 1998). In that context it is worth mentioning that until recently it was believed that fetal human tissues during early phases of development are exposed to only trace amounts of thyroid hormones (Burrow et al., 1994). However, it was lately demonstrated that amniotic ¯uid, coelomic ¯uid and fetal blood T4 concentrations of ®rst trimester human pregnancies (as early as the 5th week of gestation) are in the same range as those available to adult tissues and depend ultimately on the circulating maternal T4 serum concentrations (Calvo et al., 2002). In the present study, a cohort of 283 infertile women was followed over 5 years under routine daily outpatient care. After exclusion of absolute causes of sterility (such as bilateral tubal obstruction and azoospermia of the male partner) all women with primary or secondary infertility were included. Due to the lack of prospective data on the in¯uence of thyroid hormone replacement on fertility of women with mild thyroid impairment, and on the optimal TSH threshold for therapy, patients with a TRH-stimulated TSH response to >15 mIU/l [the mean of a healthy euthyroid population (Keller, 1986; Vierhapper, 1997)] were treated with T4. The aims were: (i) to compare pregnancy rates, abortion rates, and proportions of delivery of healthy babies between women with different thyroid function at ®rst visit, (ii) to compare thyroid function over time in patients who conceived with those who did not, and (iii) to identify possible fact (...truncated)