Article Prediction of pituitary down-regulation by evaluation of endometrial thickness in an IVF programme

RBMOnline - Vol 8. No 5. 2004 595-599 Reproductive BioMedicine Online; www.rbmonline.com/article/1065 on web 17 March 2004 Article Prediction of pituitary down-regulation by evaluation of endometrial thickness in an IVF programme Muammer M Dogan, MD, was born in Ankara in 1963. He graduated from Ankara University Medical Faculty in 1986 and became a specialist in obstetrics and gynaecology in 1990. Since then he has been working in the Department of Reproductive Medicine of the Zekai Tahir Burak Women s Health Research and Education Hospital, which is the largest maternity hospital in Turkey, treating 800 IVF embryo transfer cases annually. Dr Muammer M Dogan Muammer M Dogan 1, Dilek Uygur, Reyhan Neslihanoglu Alkan, Sertaç Batioglu, Leyla Mollamahmutoglu Centre for Reproductive Medicine, Zekai Tahir Burak Women s Health Research and Education Hospital, Ankara, Turkey 1 Correspondence: Turan Gunes Bul, Urdün Cad, 48 Sok, Akasya Evleri 2, Block B-21, Oran-Ankara, Turkey. Tel: +90 3124198747; Fax: +90 3124198838; e-mail: muadogan@ttnet.net.tr Abstract The aim of this retrospective study was to determine whether pituitary down-regulation after gonadotrophin-releasing hormone analogue (GnRHa) administration can be accurately predicted by transvaginal ultrasonographic measurement of endometrial thickness in the presence of menstruation. All cycles of an IVF/intracytoplasmic sperm injection programme in which a long protocol of GnRHa was used for ovarian stimulation were analysed. Overall, 209 patients underwent 223 treatment cycles. Using a serum oestradiol concentration of 50 pg/ml as a cut-off point, the sensitivity, specificity, predictive value and false positive and false negative values were calculated for prediction of pituitary down-regulation from endometrial thickness measurements. Pituitary down-regulation was achieved in 223 treatment cycles in 180 patients (80%). The best combination of the highest specificity (71.7%) and sensitivity (62.5%) is achieved with a linear appearance of the endometrium. Therefore, ultrasonographic measurement of endometrial thickness should be used in combination with serum oestradiol concentration in estimating pituitary down-regulation after GnRHa. In conclusion, the linear appearance of endometrium can be as reliable as serum oestradiol concentration in prediction of pituitary down-regulation after GnRHa. Keywords: economics, endometrium, GnRHa, IVF embryo transfer, oestradiol, ultrasound Introduction Gonadotrophin-releasing hormone analogues (GnRHa) have been successfully used in the assisted reproductive technology programmes to suppress the pituitary ovarian axis. They are used prior to and concomitantly with stimulation of follicular growth and induction of ovulation by exogenous gonadotrophins. The adequacy of pituitary down-regulation is usually assessed by measuring serum oestradiol concentrations and ultrasonographic evaluation of ovarian morphology, specifically the presence of cysts and endometrial thickness. A recent study evaluated ovarian artery resistance index. The authors reported that ovarian artery resistance index was the best Doppler predictor for pituitary suppression, with the highest specificity and positive predictive value (Dada et al., 2002). There have been studies investigating the relation of endometrial thickness with various other reproductive issues. For example, Kolibianakis et al. (2004) attempted to evaluate the predictive value of endometrial thickness in ongoing pregnancy achievement in IUI cycles and found no significant relationship. There have been attempts to simplify assisted reproduction programmes. Since repeated hormonal assays are expensive and add to the stress for patients, some IVF programmes have abandoned completely the use of hormonal assays and difficult methods for cycle monitoring when GnRHa is used (Golan et al., 1994; Roset et al., 1995). Likewise, even single hormonal assays may have higher costs compared with ultrasound scans. The cost of a single hormonal assay (~US$17) is higher than 595

596 an ultrasound scan (~US$10). Moreover, hormonal assays may have to be repeated. The aim of this study was to evaluate whether the state of the hypo-oestrogenism after pituitary down-regulation can be diagnosed solely by ultrasonographic evaluation in the presence of menstruation. Materials and methods Patients All cycles of an IVF/intracytoplasmic sperm injection (ICSI) programme between January 2002 and December 2002, in which a long protocol of GnRHa was used for ovarian stimulation, were analysed. The causes of infertility were as follows: male factor (57%), tubal factor (19%), unexplained infertility (15.7%) and anovulatory infertility (4.7%). Overall, 209 patients underwent 223 treatment cycles. All patients underwent a baseline transvaginal ultrasound evaluation. When normal ovarian morphology was noted, a standard oral contraceptive pill consisting of 0.15 mg of desogestrel and 0.03 mg ethinyl oestradiol (Desolett, Organon, Oss, The Netherlands) was given for 21 days. All cycles were stimulated using the long GnRHa protocol with different GnRHa preparations according to the attending physician s preference. GnRHa was instituted in the last 4 days of oral contraceptive pill administration. After withdrawal of oral contraception, GnRHa was continued, along with gonadotrophin stimulation. GnRHa was stopped on the day of human chorionic gonadotrophin (HCG) administration. A follow-up visit was performed on day 3 of menstruation. On that morning, patients were evaluated for pituitary downregulation by measuring serum oestradiol concentration. All patients had a baseline ultrasonographic evaluation with a 5-MHz transvaginal mechanical sector probe (Combison 420; Kretztecknik, Tiefenbach, Austria). In the receiver operating characteristic (ROC) analysis, the ROC curve was 0.68 with a 95% CI of 0.59 0.77 (Figure 1) Transvaginal ultrasonography was performed by three experienced operators in all patients to assess both the ovarian morphology and the endometrium thickness. The widest wall-to-wall endometrial diameter was measured. The term linear appearance of the endometrium implies an endometrium with a diameter too small to be measured, which appears like a line. The results of oestradiol determinations were not available to the clinicians measuring the endometrium, but they were aware of the patients menstrual status. With the use of serum oestradiol concentration of 50 pg/ml as a cut-off point, the sensitivity, specificity, predictive value and false positive and false negative values were calculated for prediction of pituitary down-regulation from endometrial thickness measurements. Results If pituitary down-regulation is defined as an oestradiol concentration of 50 pg/ml, it was achieved in 180 patients. All of the patients were on day 3 of menstruation. Examples of the linear appearance and endometrial thickness are given in Table 1. The sensitivity, specificity, predictive value and negative and positive value of each endometrial thickness and oestradiol concentrations are listed in Table 2. The IVF outcomes considering the different endometrial thickness and serum oestradiol measurements are listed in Table 3. The best combination of the highest specificity (71.7%) and sensitivity (62.5%) is achieved with a linear appearance of the endometrium (Figure 2). As the endometrial thickness increases, specificity decreases very sharply (Figure 3). Oestradiol concentration of 50 pg/ml was present in 81.3% of cycles with endometrial thickness of 6 mm. Only 4.6% of patients with oestradiol >50 pg/ml had an endometrial thickness of >6 mm. The data are summarized in Table 1. Discussion A number of investigators have attempted to limit the expense of assisted reproduction by reducing monitoring. GnRHa administration is used very commonly, as it reduces the need for intensive monitoring and improves the results (Hughes et al., 1992; Tan et al., 1992). In the long protocol regimen, GnRHa therapy is commenced in either the mid-luteal phase of the previous cycle or in the early follicular phase of the therapeutic cycle. In this protocol, the effect of GnRHa in producing pituitary desensitization is usually assessed by measuring serum oestradiol concentrations. The concept that endometrial thickness can serve as a reliable marker of the patient s oestrogenic status has been advanced before. By means of transvaginal sonography, Nakamura et al. (1996) have selected an endometrial thickness of 6 mm or more to predict the occurrence of withdrawal bleeding after progesterone administration with an accuracy of 95.5%. Furthermore, endometrial thickness was superior to serum oestradiol concentration in predicting withdrawal bleeding. Thus, it reflected the bioactivity of oestrogenic state as a simple test. Shulman et al. (1989) reported that by means of abdominal ultrasonography, they were able to predict the presence of withdrawal bleeding best when the endometrial thickness Figure 1. ROC curve. Statistical analyses related to ROC curve: area = 0.686 (SE = 0.046); asymptotic signal = 0.0; asymptotic 95% CI = 0.596 0.775.

Table 1. Relationships between oestradiol concentrations and endometrial thickness. Specificity = 71.7%, sensitivity = 62.5%, NPV = 88.0%, PPV = 36.6%, accuracy = 69.8%. Oestradiol Endometrial thickness Total (pg/ml) Linear appearance More than linear appearance n % n % <50 132 71.7 52 28.4 184 50 18 37.5 30 62.5 48 Total 150 82 232 P < 0.001. Table 2. Sensitivity, specificity, predictive value and negative and positive value of each endometrial thickness and oestradiol concentrations. Endometrial No. Oestradiol Oestradiol Sensitivity Specificity Predictive False False thickness of 50 >50 (%) (%) value (%) positive negative (mm) cycles pg/ml pg/ml (%) (%) Linear 139 123 16 62.5 71.7 88.0 12.0 64.2 3 178 31 8 85.0 41.8 85.9 14.0 60.0 4 203 16 9 93.3 18.6 82.7 17.2 60.0 5 213 6 4 96.6 9.3 81.6 18.3 60.0 6 223 5 5 99.4 4.6 81.3 18.6 33.3 Table 3. IVF outcomes associated with different endometrial thickness and serum oestradiol measurements Endometrial No. Oestradiol Pregnancy rate Oestradiol Pregnancy rate P-value thickness (mm) of cycles 50per embryo >50per embryo pg/ml transfer (%) pg/ml transfer (%) Linear 139 123 34.0 16 32.7 NS 3 178 31 33.6 8 34.2 NS 4 203 16 34.8 9 33.1 NS 5 213 6 32.6 4 33.4 NS 6 223 5 33.2 5 34.1 NS values were 5 mm. By using this method, patients with thin endometrium can be spared the therapeutic intervention, time, and the extra visit necessary to document a negative progesterone challenge test (Shulman et al., 1989). In addition, the occurrence of vaginal bleeding is usually due to the effect of declining steroid hormone concentrations and can be used as another marker of declining oestrogen concentrations. Barash et al. have shown that pituitary down-regulation can be predicted with a high degree of accuracy by ultrasonographic measurement of endometrial thickness. In their study, they used a cut-off value for oestradiol concentration of 55 pg/ml to confirm pituitary down-regulation (Barash et al., 1998). They evaluated the patients 17 days after the administration of GnRHa for pituitary down-regulation, irrespective of the menstrual situation. They reported that high sensitivity and high specificity was obtained with an endometrial thickness of 6 mm, resulting in a positive predictive value of 96% (Barash et al., 1998). They have suggested that whether this concept would be still valid if ultrasonographic evaluation for pituitary down-regulation was undertaken shortly after the onset of menstruation should be investigated. Sharif et al. studied the accuracy of endometrial thickness of 4 mm and/or the occurrence of vaginal bleeding in predicting pituitary desensitization at both 50 and 100 pg/ml oestradiol concentrations. They concluded that sonographic measurement of endometrial thickness combined with a history of vaginal bleeding predicts the patient s oestrogenic state after GnRHa with a high degree of accuracy (Sharif et al., 2001). 597

Figure 2. An example of the linear appearance of the endometrium (see text for details of the scan). Figure 3. Example of a thick endometrium. 598 Some IVF centres report use of a history of vaginal bleeding as the sole indicator of pituitary desensitization (Wikland et al., 1994; Hurst et al., 1997). The present study has used the combination of menstrual bleeding history and endometrial thickness to predict pituitary desensitization. Many IVF programmes perform ultrasonographic examination before starting ovulation induction, to detect ovarian cysts. An important problem is that the optimum level for oestradiol concentration that predicts pituitary desensitization is not defined precisely. In the present study, a cut-off value for oestradiol concentration of 50 pg/ml has been used to confirm pituitary down-regulation. This value has been chosen by many IVF groups (Felberbaum et al., 1997; Speroff et al., 1999). For this cut-off value, a linear appearance of endometrium had a negative predictive value of 88%, specificity of 71.7%, and sensitivity of 62.5%. Above this thickness, as the sensitivity increases, the specificity decreases very sharply. Except for the linear appearance of the endometrium, the specificity of the measurement of the endometrial thickness is too low to be used as the sole test in predicting pituitary down-regulation in the presence of menstruation. Based on the results of this study, it is concluded that the ultrasonographic appearance of menstrual endometrium is variable in both echogenicity and measurements. The amount of blood present affects the appearance and the echo texture. In conclusion, in the presence of menstruation, only the linear appearance of endometrium can serve as a reliable marker of the pituitary down-regulation.

Acknowledgements We thank Eyüp Ekici for collection of ultrasonographic photos and Kenan Köse for statistical analysis. References Barash A, Weissman A, Manor M et al. 1998 Prospective evaluation of endometrial thickness as a predictor of pituitary downregulation after gonadotropin releasing hormone analogue administration in an in vitro fertilization program. Fertility and Sterility 69, 496 499. Dada T, Salha O, Allgar V et al. 2002 Utero ovarian blood flow characteristics of pituitary desensitization. Human Reproduction 17, 523. Felberbaum R, Rabe T, Diedrich K 1997 Gonadotropin-releasing hormone: agonists and antagonists. In: Rabe T, Diedrich K, Runnebaum B (eds) Manual on Assisted Reproduction. Springer, Berlin and Heidelberg, pp. 127 153. Golan A, Herman A, Soffer Y et al. 1994 Ultrasonic control without hormone determination for ovulation induction in in-vitro fertilization/embryo transfer with gonadotropin-releasing hormone analogue and human menopausal gonadotrophin. Human Reproduction 9, 1631 1633. Hughes EG, Federkow DM, Daya S et al. 1992 The routine use of gonadotropin-releasing hormone agonists prior to in vitro fertilization and gamete intrafallopian transfer: a meta-analysis of randomized controlled trials. Fertility and Sterility 58, 888 896. Hurst BS, Tucker KE, Awoniniyi CA et al. 1997 Preprogrammed, unmonitored ovarian stimulation reduces expense without compromising the outcome of assisted reproduction. Fertility and Sterility 68, 282 286. Kolibianakis EM, KA Zikopoulos, HM Fatemi et al. 2004 Endometrial thickness cannot predict ongoing pregnancy achievement in cycles stimulated with clomiphene citrate for intrauterine insemination Reproductive BioMedicine Online 8, 115 118. Nakamura S, Douchi T, Oki T et al. 1996 Relationship between sonographic endometrial thickness and progestin induced withdrawal bleeding. Obstetrics and Gynecology 87, 722 725. Roset J, Verhoeff A, van Heusden AM et al. 1995 Minimal monitoring of ovarian hyperstimulation: a useful simplification of the clinical phase of in vitro fertilization treatment. Fertility and Sterility 64, 552 556. Sharif K, Afnan M, Yassen I et al. 2001 Endometrial thickness and vaginal bleeding as predictor of pituitary desensitization using long protocol GnRH-agonists in IVF. Middle East Fertility Society Journal 6, 116 122. Shulman A, Shulman N, Weissenglass L et al. 1989 Ultrasonic assessment of the endometrium as a predictor of estrogen status in amenorrhoeic patients. Human Reproduction 4, 616 619. Speroff L, Glass RH, Kase NG 1999 Induction of ovulation. In: Speroff L, Glass RH, Kase NG (eds) Clinical Gynecologic Endocrinology and Infertility. Lippincott Williams and Wilkins, Philadelphia, pp. 1097 1132. Tan SL, Balen A, Hussein EL et al. 1992 A prospective randomized study of the optimum timing of human chorionic gonadotropin administration after pituitary desensitization in in vitro fertilization. Fertility and Sterility 57, 1259 1264. Wikland M, Borg J, Hamberger L et al. 1994 Simplification of IVF: minimal monitoring and the use of subcutaneous highly purified FSH administration for ovulation induction. Human Reproduction 9, 1430 1436. Received 10 July 2003; refereed 30 July 2003; accepted 23 February 2004. 599