A retrospective study on IVF outcome in euthyroid patients with anti-thyroid antibodies: effects of levothyroxine, acetyl-salicylic acid and prednisolone adjuvant treatments

Autoimmune thyroid diseases are rather frequent in women in the childbearing age, affecting 5-20% of them [1]. They are characterized by the presence of anti-thyroglobulin and anti-thyroperoxidase antibodies, grouped under the definition of anti-thyroid antibodies (ATA). ATA are often detected in subjects complaining of hypo- or hyperthyroidism, but are not rarely found in patients without any sign of thyroid dysfunction [1].

Some evidence suggest that ATA could exert a negative influence on the female reproductive potential. Women with no signs of thyroid dysfunction who were ATA+ risk spontaneous miscarriage three- to five-folds more than ATA-women [2]. Moreover, some studies reported a significantly higher ATA prevalence in subfertile women complaining of pelvic endometriosis [3, 4], premature ovarian failure (POF) [5, 6], polycystic ovary (PCO) [7], or hyperprolactinemia [8].

Some authors even reported an unexpectedly high ATA prevalence in euthyroid women with a history of three or more unsuccessful IVF cycles [9], and ATA-positivity was found to be associated with a low pregnancy rate in IVF [10‐12]. On the contrary, other studies failed to detect any difference in IVF success rate between ATA+ and ATA-women [13‐15].

In our study, we analyzed the prevalence of ATA in infertile women and compared it with that observed in age-matched, fertile controls. Further, we retrospectively analyzed IVF outcome in euthyroid ATA+ women and compared it with ATA-controls. Differently from previous reports, we also studied the effect of some adjuvant medical treatments that are frequently given to ATA+ women during IVF (levothyroxine alone or associated with acetyl-salicylic acid and prednisolone), on IVF results. Levothyroxine (LT), in fact, is claimed to reduce ATA level [16] and lower the risk of miscarriage in ATA+ women [17], whereas acetyl-salicylic acid (ASA) plus prednisolone (P) was reported to improve IVF outcome in women with autoimmune disorders [18‐21]. Aim of the present study is to check the effectiveness of such adjuvant treatments as tools to improve IVF outcome in ATA+ patients.

Anti-thyroglobulin, anti-thyroperoxidase antibodies, or both, are frequently detected in subjects with hypo- or hyperthyroidism of autoimmune origin, and can also be found in patients without any sign of thyroid disease. According to some reports, ATA positivity in euthyroid women could be linked to reproductive problems, such as increased abortion rate or infertility [2].

In the population of euthyroid, infertile women that we observed, the prevalence of ATA was 10.5%, comparable to what reported in large population studies on euthyroid, fertile women between 18 and 45 years (13.8%; 95% CI 7.6-19.9) [24]. Our observation, therefore, does not support the concept of a strong association between ATA posotivity and female infertility. It must be remarked, however, that we considered clinically relevant (potentially able to affect fertility) only ATA levels above a given threshold, whereas some reports on fertile women included all levels of positivity.

We observed that some specific causes of female subfertility, such as pelvic endometriosis or reduced ovarian reserve, are more frequently associated with ATA than other infertility factors. The association between ATA and pelvic endometriosis, indeed, was already reported [3, 4]; it could be based on a common dysfunction of the immune system, as there is abundant evidence of the link between endometriosis and specific immunological features [25]. Also premature ovarian failure (POF), that is considerable an extreme form of reduced ovarian reserve, is often associated with autoimmune disorders [26], and anti-zona pellucida antibodies able to cross-react with thyroid antigens have been detected in POF patients: they could represent a link between POF syndrome and anti-thyroid autoimmunity [27]. We also observed, in agreement with other reports [8], an increased prevalence of ATA in hyperprolactinemic women. PRL is known to exert immunomodulatory effects that could lead to increased autoreactivity of the immune system [28]. Differently, we failed to confirm the previously reported observation [7] of increased PCO prevalence among ATA+ women.

Observing IVF outcome, we noticed that it tended to be noticeably poorer in ATA+ than ATA-women. ATA+ patients who did not take any adjuvant treatment responded significantly worse to ovarian stimulation than ATA-controls, their gonadotropin dose/retrieved oocyte ratio (a reliable marker of ovarian responsiveness to stimulation) being significantly higher. A normal responsiveness to ovarian stimulation in ATA+ women was previously reported [29], but that study included only 16 ATA+ patients and unfortunately did not report their basal FSH level, nor their antral follicle count, two variables that deeply influence ovarian response to hormones.

We observed that the presence of ATA could have affected also the oocyte and consequently the embryo deriving from it; in fact, we recorded a lower fertilization rate, pregnancy and implantation rates in untreated ATA+ patients vs. ATA-controls. Interestingly enough, Bussen found a higher prevalence of ATA in infertile women with a history of three or more unsuccessful IVF attempts [9], and other studies reported poorer IVF results in ATA+ patients [10‐12]. Other authors, however, failed to confirm the existence of an association between ATA and IVF failure [13‐15]. The inconsistency in the reported results could be linked to several factors, like different criteria of patient inclusion in the study (e.g. only those with unexplained infertility or all infertile patients), different ATA level threshold needed to include a patient, different criteria to chose the control group, etc.

We observed a significantly higher abortion rate after IVF in ATA+ patients vs. ATA-controls. A similar finding was already reported by others [10, 14, 15, 30] and it is not surprising since an increased abortion rate occurs in ATA+ women even after spontaneous conception [2].

ATA+ women who despite being euthyroid received a low dose of LT during IVF showed a slightly higher ovarian responsiveness to gonadotropins, but IVF outcome was still significantly poorer in comparison to ATA- patients. Some doctors use to give a low dose of LT to ATA-positive euthyroid women because this treatment was reported to reduce ATA level [16] and to lower the risk of miscarriage in recurrent aborters who conceived spontaneously [17]. In our study, the abortion rate after IVF was not influenced by LT, and IVF results were not relevantly improved; this could be due to the high estradiol levels reached during IVF (approximately ten folds higher than in the physiological cycle) that could have induced high levels of thyroxine binding globulin, in turn able to bind and partially inactivate LT.

The adjuvant treatment with acetylsalicylic acid (ASA) combined with prednisolone (P) is advised by some doctors to patients with autoantibodies undergoing IVF; in fact, it was reported to improve IVF outcome in women with autoimmune disorders [18‐21] or with recurrent abortions [31]. Beneficial effects of this adjuvant treatment on IVF outcome, however, were not confirmed when patients without autoantibodies were studied [22, 32]. Our retrospective analysis shows that women treated with LT+ASA+P showed a higher ovarian responsiveness to gonadotropins and higher pregnancy, implantation and ongoing pregnancy rates than other ATA+ patients. Their IVF results were comparable to those observed in ATA- subjects. LT+ASA+P treatment, however, was unable to decrease the abortion rate, that remained significantly higher than the one observed in women without ATA. The advantage of giving LT+ASA+P to euthyroid ATA+ women could depend on the combination of the immunomodulatory activity of LT and P with a better ovarian and endometrial blood perfusion due to ASA [33]. Prednisolone could also lower the local inflammatory reaction to the ET catheter and facilitate implantation through improving the tolerance of the immune system toward the transferred embryos [34].

In conclusion, this retrospective study on euthyroid ATA+ women shows that: a) ATA-positivity in infertile women is as frequent as in euthyroid fertile women of comparable age; only in case of pelvic endometriosis, reduced ovarian reserve, or hyperprolactinemia, ATA prevalence is significantly higher than normal; b) ATA+ patients have significantly worse IVF outcome when compared to ATA-women; c) the adjuvant treatment with LT gives a slight improvement of ovarian responsiveness to stimulation, but has no benefit on IVF results; d) the adjuvant treatment with LT+ASA+P significantly improves IVF outcome; e) no adjuvant treatment decreases the abortion rate after IVF.

The beneficial effect of LT+ASA+P adjuvant treatment on IVF outcome observed herein in euthyroid, ATA+ patients, together with its low potential risk [22], suggests its possible clinical, targeted application in this specific subgroup of IVF patients. However, our findings need to be validated in prospective randomized investigations before the application of LT+ASA+P adjuvant treatment is encouraged.