Thyroid function after controlled ovarian hyperstimulation in women with and without the hyperstimulation syndrome Kris Poppe, M.D., Ph.D., a David Unuane, M.D., a Miguel D Haeseleer, M.D., a Herman Tournaye, M.D., Ph.D., b Johan Schiettecatte, b Patrick Haentjens, M.D., Ph.D., c and Brigitte Velkeniers, M.D., Ph.D. a a Department of Endocrinology, b Centre for Reproductive Medicine, and c Centre for Outcomes Research and Laboratory for Experimental Surgery, Universitair Ziekenhuis Brussel, Vrije Universiteit Brussel, Brussels, Belgium Objective: To investigate the impact of ovarian hyperstimulation syndrome (OHSS) on thyroid function in women without thyroid disorders and to compare it with that in women with uncomplicated controlled ovarian hyperstimulation (COH). Design: Retrospective analysis. Setting: Tertiary referral fertility center. Patient(s): A total of 77 women undergoing COH of whom 25 developed OHSS and 52 had no OHSS. Women with the presence of thyroid disorders were excluded, and only women pregnant after assisted reproductive technology were included. Intervention(s): Serum TSH and free T4 (ft4) levels were measured before and 2, 4, and 6 weeks after embryo transfer (ET), and thyroid peroxidase and thyroglobulin antibody levels were measured before ET to exclude thyroid autoimmunity. The diagnosis of OHSS was based on clinical, ultrasonographic, and biologic features. Main Outcome Measure(s): Thyroid function, OHSS. Result(s): Serum TSH and ft4 levels increased 2 weeks after ET in both study groups compared with prestimulation levels. In the OHSS group: TSH, 1.9 0.8 miu/l vs. 3.1 1.9 miu/l; ft4, 12.3 1.4 ng/l vs. 13.4 2.1 ng/l. In the no-ohss group: TSH, 2.1 1.1 miu/l vs. 2.6 1.9 miu/l; ft4, 13.0 1.7 ng/l vs. 13.8 1.6 ng/l. The increment was comparable between both study groups. Conclusion(s): Serum TSH levels increased significantly after COH in a comparable way in both study groups, when no thyroid disorders were present. (Fertil Steril Ò 2011;96:241 5. Ó2011 by American Society for Reproductive Medicine.) Key Words: Ovarian hyperstimulation syndrome, thyroid function, controlled ovarian hyperstimulation Controlled ovarian hyperstimulation (COH), which is an important stage in the preparation of an assisted reproductive technology (ART) procedure, is characterized by high E 2 levels, comparable to those obtained during the second trimester of spontaneous pregnancy (4000 6000 ng/l). The high E 2 levels are a strain on the hypothalamic-pituitary-thyroid axis, in addition to that of the pregnancy itself, and can therefore impair thyroid hormonal levels (1, 2). In healthy pregnant women without structural thyroid disorders, the required increase in thyroid hormone production can be met easily and thus will not lead to the development of (subclinical) hypothyroidism. In the study of Alexander et al. (3), the need for a rapid increase (4 6 weeks gestation) in T4 was identified in hypothyroid-treated women to maintain euthyroidism, and they showed that the timing of this increase was more rapid and pronounced after ART procedures. Both Muller et al. (4) and Poppe et al. (5) investigated thyroid function during the first trimester of pregnancy after ART and showed a significant impact of COH on serum TSH and ft4 levels. In addition, changes in thyroid function Received February 8, 2011; revised and accepted April 13, 2011; published online May 12, 2011. K.P. has nothing to disclose. D.U. has nothing to disclose. M.D H. has nothing to disclose. H.T. has nothing to disclose. J.S. has nothing to disclose. P.H. has nothing to disclose. B.V. has nothing to disclose. Reprint requests: Kris Poppe, M.D., Ph.D., Department of Endocrinology, Universitair Ziekenhuis Brussel, Vrije Universiteit Brussel, Brussels, Belgium (E-mail: kris.poppe@uzbrussel.be). were more pronounced and permanent in women with associated thyroid autoimmunity (TAI); therefore, in women with limited thyroid function, COH may lead to thyroid dysfunction during the first trimester of pregnancy, when the fetus is largely dependent on maternal T4 for its neuropsychologic development (6). Ovarian hyperstimulation syndrome (OHSS) is a potentially lifethreatening complication of COH, characterized by a massive enlargement of both ovaries and by increased capillary permeability (7). The exact pathophysiologic mechanism underlying OHSS remains largely unknown, but the high serum E 2 levels (R5,000 ng/l) are a characteristic biologic feature. OHSS occurs almost exclusively as a complication of COH in 1% of ART cycles. The aim of this study was to investigate the impact of OHSS on thyroid function and to compare it with that of an uncomplicated COH (no OHSS) in women pregnant after ART and without thyroid disorders. MATERIALS AND METHODS Overall Study Design The current retrospective study, conducted in collaboration with the Centre for Reproductive Medicine of the University Hospital of Brussels, Belgium, was approved by the institutional review board of our hospital. Women coming to the center for consultation were systematically screened for thyroid dysfunction (serum free T4 [ft4] and TSH levels) and for the presence of TAI by means of thyroid peroxidase and thyroglobulin antibodies. The diagnosis of an underlying thyroid disease was further evaluated from the personal history on goiter, known thyroid diseases, and/or prior use of thyroid 0015-0282/$36.00 Fertility and Sterility â Vol. 96, No. 1, July 2011 241 doi:10.1016/j.fertnstert.2011.04.039 Copyright ª2011 American Society for Reproductive Medicine, Published by Elsevier Inc.
medication. Only euthyroid women without TAI and pregnant after ART were included in this study. Thyroid function and serum hcg and E 2 levels were assessed in all women at regular intervals of 2 weeks before and 2, 4, and 6 weeks after embryo transfer (ET). Thyroid function was assessed in 2 different groups of women, and the results were subsequently compared. The first group consisted of 52 women with uncomplicated COH (women pregnant after ART from a previous study) (8). The second group consisted of 25 women who developed OHSS. The causes of infertility were comparable between both groups and especially that of women with ovulatory disorders (17% and 23%, respectively). The diagnosis of OHSS was made based upon clinical findings, transvaginal ultrasound observations, and measurements of serum E 2 values, and all degrees of OHSS severity were included, according to Golan et al. (9). Ovarian Hyperstimulation/ART Treatment All women received the same COH treatment consisting of a combination of the GnRH agonist (GnRH-a; Suprefact nasal spray; Hoechst) and 150 IU of ufsh (Menopur; Ferring Pharmaceuticals Inc.) or 150 IU of recfsh (Puregon; MSD B.V., Brussels, Belgium; Gonal-F; Merck-Serono S.A.). When the patient had at least 3 follicles with diameters of 17 mm and serum E 2 levels of 1,000 ng/l, GnRH-a and FSH were discontinued and ovulation was induced with 10,000 IU of hcg (Pregnyl; Organon). All patients had a transvaginal ultrasound-guided ovum aspiration approximately 36 hours after hcg injection under local anesthesia. With conventional IVF, each oocyte was inseminated within 3 to 4 hours after retrieval by the addition of 5,000 to 20,000 motile spermatozoa per oocyte. The intracytoplasmic sperm injection procedure was conducted as described earlier (10). After fertilization, 1 to 3 embryos were transferred depending on their morphological quality and according to the guidelines included in the Belgian law on ART (11). Pregnancy was diagnosed at least 10 days after ET by rising hcg levels of R20 IU/L in the serum on 2 occasions. Serum Assay All procedures were conducted by the RIA laboratory of the University Hospital of Brussels. Serum TSH and ft4 levels were measured using a third-generation electrochemiluminescence immunoassay (Roche Diagnostics). The reference values for TSH were 0.27 to 4.2 miu/l, and the reference values for ft4 were 9.3 to 17.0 ng/l (12 23.2 pmol/l). The quantity of thyroid peroxidase antibodies was determined with an RIA kit (BRAHMS Diagnostica), with a reference range of 0 to 34 kiu/l. Thymoglobulin antibody levels were measured with an automated competitive immunoassay (modular E170: by an automated Elecsys immunochemistry analyzer; Roche Diagnostics), with a reference range of 0 to 115 IU/mL. Serum hcg and E 2 levels were measured by an automated Elecsys immunochemistry analyzer (Roche Diagnostics). The intra-assay and interassay coefficients of variation were, respectively, <5% and <7% for hcg and <5% and <10% for E 2. Normal serum E 2 values in women during the first trimester of pregnancy are approximately 800 to 4,500 ng/l. Statistical Analysis Serum TSH, ft4, and E 2 values from the total cohort of women passed the Kolmogoroff-Smirnoff test for normal distribution and are expressed as means SD. The serum hcg values are expressed as median (range) values. The paired Student s t test was used to compare the differences between serum TSH and ft4 values before ET and 2 weeks after ET. For comparison of the differences in thyroid function between patients with and without OHSS before and 2 weeks after ET, the unpaired Student s t test was used. A 1-way (single-group) repeated-measures ANOVA was performed to evaluate the time effect on serum TSH and ft4 values determined at 4 time points: 2 weeks before ETand at weeks 2, 4, and 6 after ET. A 2-way (between-groups) repeated-measures ANOVA was used to assess the impact of OHSS vs. no OHSS on thyroid function, as measured by serum TSH and ft4 values the first weeks after ET. The correlations between serum E 2 level and serum TSH and ft4 levels were performed by means of a Spearman s correlation test. All data analyses were performed using SPSS version 12.0 (SPSS, Inc.). Results were considered significant when P<.05. RESULTS Baseline Characteristics Table 1 shows the clinical and biochemical baseline characteristics of all women included in this study (n ¼ 77) and stratified according to OHSS status. In the entire study group, the mean values SD for serum TSH and ft4 levels were 2.0 1.0 miu/l and 12.8 1.6 ng/l, respectively. There was no significant difference between the OHSS and no-ohss group for baseline serum TSH levels (2.1 1.1 miu/l and 1.9 0.9 miu/l, respectively; P¼not significant) and for serum ft4 levels (12.3 1.4 ng/l and 13.0 1.7 ng/l, respectively; P¼not significant). In the entire study group, the mean age was 31 5 years, comparable between both groups. Thyroid Function Two weeks after ET Serum TSH and ft4 values significantly increased in both study groups 2 weeks after ET compared with their respective baseline values. In the OHSS group, the serum TSH values increased from 1.9 0.9 miu/l to 3.1 1.9 miu/l (P<.001), and serum ft4 values increased from 12.3 1.4 ng/l to 13.4 2.1 ng/l (P¼.001). In the no-ohss group, the serum TSH values increased from 2.1 1.1 miu/l to 2.6 1.9 miu/l 2 weeks after ET (P¼.009), and serum ft4 values from 13.0 1.7 ng/l to 13.8 1.6 ng/l (P<.001). There were no significant differences between the 2 groups regarding the serum TSH and ft4 values (P¼.240 and P¼.370, respectively) 2 weeks after ET. In 4 patients in the OHSS group (4/25 [16%]), the serum TSH values rose to supranormal levels 2 weeks after ET (4.3, 4.5, 4.6, and 5.0 miu/l, respectively), and the same was observed in 2 patients (2/52 [4%]) in the no-ohss group (7.2 and 12.2 miu/l, respectively). The 4 patients in the OHSS group and the 2 patients in the no-ohss group with serum TSH values >4.2 miu/l 2 weeks after ET had serum TSH values <2.5 miu/l at the next time points measured. During the first weeks of pregnancy Figure 1 shows the pattern of change in serum TSH and ft4 values during the first weeks of pregnancy, stratified according to OHSS status. In both groups, there was a statistically significant change over time for TSH and ft4 levels over the different time points (P<.001 for TSH and P¼.004 for ft4). When analyzed according to OHSS vs. no-ohss status, both the serum TSH and ft4 curves during this time period remained strictly comparable (P¼.930 for TSH and P¼.090 for ft4). The exact data (mean SD) are given at the bottom of Figure 1. TABLE 1 Baseline characteristics of all women stratified according to OHSS status. All women OHSS No OHSS No. of patients (%) 77 (100) 25 (32) 52 (68) Age (y) 31 5 30 5 31 5 TSH (miu/l) a 2.0 1.0 1.9 0.9 2.1 1.1 ft4 (ng/l) b 12.8 1.6 12.3 1.4 13.0 1.7 Note: Data presented as mean SD unless otherwise noted. a Reference range for TSH is 0.27 to 4.2 miu/l. b Reference range for ft4 is 9.3 to 17 ng/l. 242 Poppe et al. Thyroid function and ovarian hyperstimulation Vol. 96, No. 1, July 2011
FIGURE 1 Pattern of change over time for the mean TSH and free T4 (ft4) serum values collected at 4 time points: before ET (time 0) and at weeks 2, 4, and 6 after ET. P¼not significant for patients with ovarian hyperstimulation syndrome (OHSS) (solid red lines) vs. no OHSS (dotted blue lines). Serum hcg and E 2 Levels After ET Serum hcg levels were comparable between both study groups at all measured time points, as shown in Table 2. Serum E 2 values were at all time points significantly higher in the OHSS group as compared TABLE 2 Serum hcg and E 2 levels at different time points after successful ART in women with and without OHSS and OHSS. OHSS No OHSS hcg (IU/L) Week 2 416 (250 6,166) 204 (101 2,370) Week 4 28,868 (3,402 135,154) 27,752 (3,907 81,898) Week 6 90,051 (42,172 232,970) 97,973 (21,569 144,680) E 2 (ng/l) Week 2 3,786 1,346 a 1,307 1,129 Week 4 5,158 2,038 a 1,913 1,700 Week 6 3,592 944 a 1,872 1,537 Note: Data presented as median (range) or mean SD. a P<.001 compared with no OHSS. with those in the no-ohss group (P<.001). The peak value of serum E 2 in the OHSS group was measured 4 weeks after ET. Correlations Between TSH and ft4 and E 2 A significant positive correlation between TSH and E 2 (r ¼.53; P¼.018) and a negative correlation between ft4 and E 2 (r ¼.52; P¼.016) was observed 6 weeks after ET in all study patients. DISCUSSION This study showed that COH had an important impact on thyroid function and that despite higher E 2 levels in patients with OHSS, a higher prevalence of (subclinical) hypothyroidism did not occur, when no underlying thyroid disease was present. The initial increase in serum TSH levels was probably a pituitary feedback response to the ft4 levels that had begun to decrease (lower ft4 because of an E 2 -induced increase in T4-binding globulin and total T4). Total T4 and T4-binding globulin, however, were not measured in our study, and the observed increase in ft4 levels might have been the consequence of the time point at which we measured thyroid function (i.e., soon after the administration of 10,000 U of hcg for ovulation induction, which is known to stimulate the thyroid gland) (12). Whether the use of the ft4 assay could have played a role in the higher serum ft4 levels we observed is difficult to assess. The controversy on the use of ft4 during pregnancy is ongoing, although we recently showed that according to the equilibrium dialysis Fertility and Sterility â 243
technique (a reference measurement procedure), the ft4 assay used in this study was more reliable than others (13). At the later time points, serum TSH levels progressively decreased, probably because of the thyrotrophic action of the increasing serum hcg levels, aiming to maintain normal ft4 levels during the first trimester of pregnancy (2). In another study on thyroid function and COH, this finding was not reported, probably because thyroid function was only measured at 1 time point after COH, before high serum hcg levels could have blunted serum TSH (4). At the 6-week time point, serum TSH and ft4 levels were correlated (positively and negatively, respectively) with the E 2 levels in both study groups, which was also the case in the study by Muller et al. (4). At all time points measured, the E 2 levels were significantly higher in the OHSS group compared with those in the no-ohss group, but serum TSH and ft4 levels in both groups remained comparable over all time points measured. The development of subclinical hypothyroidism (defined as serum TSH >4.2 miu/l) that was observed in patients of both study groups at 2 weeks after ET was reversible because at the later time points, serum TSH levels were within the reference range and were even <2.5 miu/l, the currently advised upper limit during the first trimester of pregnancy (14). Our data prove that the thyroid machinery correctly counteracted the strain induced by the E 2 levels increase in the circumstance of COH and OHSS, as long as no concomitant structural thyroid disorders (absence of TAI, adequate LT4 substitution after thyroidectomy, and no iodine deficiency) were present before ART. The reason why some women developed OHSS and others did not remains unknown. In both groups, the indications for COH were comparable, with an equal number of patients undergoing the procedure for polycystic ovaries. To avoid a bias owing to the use of different COH protocols, all patients received the same type of COH. It is therefore noteworthy that in the study by Davis (15), the impact of different types of COH procedures on thyroid function was investigated in women treated with LT4. No difference was observed in the dose of LT4 increment between women pregnant after SC injections (32%) or after oral ovarian hyperstimulation (30%). It should however be noted that postconception serum TSH levels were higher after COH compared with those obtained in spontaneous pregnancies (3.8 vs. 2.2 miu/l) (15). Data on the impact of COH on thyroid function in infertile women with concomitant TAI showed an evolution toward a significantly higher serum TSH level and lower ft4 level compared with those in women without TAI, highlighting the inability of the thyroid to correctly counteract the strain imposed by COH (5, 16). Only in 1 case report have changes in thyroid function been reported in a woman with Hashimoto disease developing OHSS (17). This showed that OHSS led to severe hypothyroidism, despite LT4 substitution and serum TSH <4.2 miu/l before COH. These observations highlighted that in the presence of TAI, the dosage of LT4 should be adapted before COH, but the extent to which the serum TSH level should be lowered to decrease pregnancy morbidity or mortality remains to be investigated in prospective randomized clinical trials. According to recent data (18), however, there could be a continuous relationship between the risk of miscarriage and serum TSH levels; thus, even within the normal range of TSH, women with a slightly higher TSH level have a higher risk of miscarriage compared with those with a lower value of TSH. It remains important to highlight that women without TAI and a serum TSH <2.5 miu/l still have a risk for miscarriage like the general population, highlighting the fact that it is a multifactorial problem related to factors such as age and chromosomal abnormalities (8, 19). In addition, Negro et al. (20) was the only interventional study with LT4 in assisted pregnancies not showing a beneficial effect in terms of miscarriage reduction; the results could, however, have been masked by a lack of power. Based on the present and previous studies on the impact of COH on thyroid function, it is strongly advised to measure thyroid function and TAI before starting COH. Follow-up of thyroid function is warranted after COH to keep the serum TSH levels <2.5 miu/l during pregnancy; when serum TSH is >2.5 miu/l before COH and TAI is associated, LT4 should be given before COH (14, 16, 17). As has recently been shown in a study by Abalovich et al. (21), even in spontaneous pregnancies, almost 50% of women taking LT4 for reasons of thyroidectomy and/or TAI and with serum TSH levels between 1.2 and 2.5 miu/l need to increase their LT4 to keep the level <2.5 miu/l during pregnancy. When TAI is absent, thyroid function tests should be checked after COH before therapeutic decisions are made. The limitation of the current study was its retrospective design. The period of thyroid function follow-up was limited to 6 weeks after ET, and the OHSS group was not subdivided according to its severity. In conclusion, serum TSH levels increased significantly immediately after COH in both the OHSS and no-ohss groups. In women without associated thyroid disorders, the presence of OHSS had no additional impact on thyroid function compared with those in the no-ohss group. During the follow-up period, serum TSH levels normalized in both study groups. REFERENCES 1. Braude P, Rowell P. Assisted reproduction I: general principles. BMJ 2003;327:799 801. 2. Krassas GE, Poppe K, Glinoer D. Thyroid function and human reproductive health. Endocr Rev 2010;31:702 55. 3. Alexander EK, Marqusee E, Lawrence J, Jarolim P, Fischer GA, Larsen PR. Timing and magnitude of increases in levothyroxine requirements during pregnancy in women with hypothyroidism. N Engl J Med 2004;351:241 9. 4. Muller AF, Verhoeff A, Mantel MJ, De Jong FH, Berghout A. Decrease of free thyroxine levels after controlled ovarian hyperstimulation. J Clin Endocrinol Metab 2000;85:545 8. 5. Poppe K, Glinoer D, Tournaye H, Schiettecatte J, Devroey P, Van Steirteghem A, et al. Impact of ovarian hyperstimulation on thyroid function in women with and without thyroid autoimmunity. J Clin Endocrinol Metab 2004;89:3808 12. 6. Haddow JE, Palomaki GE, Allan WC, Williams JR, Knight GJ, Gagnon J, et al. Maternal thyroid deficiency during pregnancy and subsequent neuropsychological development of the child. N Eng J Med 1999;341:549 55. 7. Delvigne A. Symposium: update on prediction and management of OHSS. Epidemiology of OHSS. Reprod Biomed Online 2009;19:8 13. 8. Poppe K, Glinoer D, Tournaye H, Schiettecatte J, Haentjens P, Velkeniers B. Thyroid function after controlled ovarian hyperstimulation in women free of thyroid disease. Fertil Steril 2005;83:1753 7. 9. Golan A, Weissman A. Symposium: update on prediction and management of OHSS. A modern classification of OHSS. Reprod Biomed Online 2009;19:28 32. 10. Van Steirteghem A, Liu J, Joris H, Nagy Z, Janssenswillen C, Tournaye H, et al. Higher success rate by intracytoplasmatic sperm injection than by subzonal insemination. Report of a second series of 300 consecutive treatment cycles. Hum Reprod 1993;8:1055 60. 11. Van Landuyt L, Verheyen G, Tournaye H, Camus M, Devroey P, Van Steirteghem A. New Belgian embryo transfer policy leads to sharp decrease in multiple pregnancy rate. Reprod Biomed Online 2006;13: 765 71. 244 Poppe et al. Thyroid function and ovarian hyperstimulation Vol. 96, No. 1, July 2011
12. Noppen M, Velkeniers B, Buydens P, Devroey P, Van Steirteghem A, Vanhaelst L. Hyperthyroidism after gonadotrophic ovarian stimulation. Br Med J (Clin Res Ed) 1986;16:171 2. 13. Anckaert E, Poppe K, Van Uytfanghe K, Schiettecatte J, Foulon W, Thienpont LM. FT4 immunoassays may display a pattern during pregnancy similar to the equilibrium dialysis ID-LC/tandem MS candidate reference measurement procedure in spite of susceptibility towards binding protein alterations. Clin Chim Acta 2010;6:1348 53. 14. Abalovich M, Amino N, Barbour LA, Cobin RH, De Groot LJ, Glinoer D, et al. Management of thyroid dysfunction during pregnancy and postpartum: an Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab 2007;92:S1 47. 15. Davis LB. The effect of infertility medication on thyroid function in hypothyroid women who conceive. Thyroid 2007;17:773 7. 16. Stuckey BG, Yeap D, Turner SR. Thyroxine replacement during super-ovulation for in vitro fertilization: a potential gap in management? Fertil Steril 2010;93:2414.e1 3. 17. Poppe K, Glinoer D, Tournaye H, Devroey P, Velkeniers B. Impact of the ovarian hyperstimulation syndrome on thyroid function. Thyroid 2008;18: 801 2. 18. Benhadi N, Wiersinga WM, Reitsma JB, Vrijkotte TG, Bonsel GJ. Higher maternal TSH levels in pregnancy are associated with increased risk for miscarriage, fetal or neonatal death. Eur J Endocrinol 2009;160:985 91. 19. Toth B, Jeschke U, Rogenhofer N, Scholz C, W urfel W, Thaler CJ, et al. Recurrent miscarriage: current concepts in diagnosis and treatment. J Reprod Immunol 2010;85:25 32. 20. Negro R, Mangieri T, Coppola L, Presicce G, Casavola EC, Gismondi R, et al. Levothyroxine treatment in thyroid peroxidase antibody-positive women undergoing assisted reproduction technologies: a prospective study. Hum Reprod 2005;20: 529 33. 21. Abalovich M, Alcaraz G, Kleiman-Rubinsztein J, Pavlove MM, Cornelio C, Levalle O, et al. The relationship of preconception thyrotropin levels to requirements for increasing the levothyroxine dose during pregnancy in women with primary hypothyroidism. Thyroid 2010;20:1175 8. Fertility and Sterility â 245