Increased endometrial thickness is associated with improved treatment outcome for selected patients undergoing in vitro fertilization embryo transfer Xingqi Zhang, Ph.D., a Chi-Huang Chen, M.D., b Edmond Confino, M.D., a Randall Barnes, M.D., a Magdy Milad, M.D., a and Ralph R. Kazer, M.D. a a Division of Reproductive Endocrinology and Infertility, Department of Obstetrics and Gynecology, Feinberg School of Medicine of Northwestern University, Chicago, Illinois; and b Department of Obstetrics and Gynecology, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan Objective: To examine possible relationships between endometrial thickness and treatment outcome after IVF and embryo transfer, and to explore the role of potential confounding factors that may influence such relationships. Design: Retrospective study. Setting: A university-affiliated clinical IVF center. Patient(s): Patients undergoing IVF embryo transfer with their own oocytes. Intervention(s): None. Main Outcome Measure(s): Endometrial thickness was determined on the day of hcg administration, 2 days before oocyte retrieval. Clinical pregnancy was confirmed by ultrasound observation of fetal heart activity. Result(s): The study analyzed 897 IVF embryo transfer cycles. Treatment outcome (clinical pregnancy) after IVF embryo transfer was positively associated with increased endometrial thickness and peak E 2 concentrations in serum, and negatively associated with advanced age. Endometrial thickness was dependent on peak E 2 concentrations in serum, but was independent of patient age or duration of ovarian stimulation. Thin endometrium reduced PRs in relatively young patients ( 38 years old), in patients who required more than 10 days of gonadotropin stimulation, or in patients whose embryo transfers consisted of poor quality embryos. Conclusion(s): Increased endometrial thickness was associated with improved treatment outcome, but this association was dependant on patient age, duration of ovarian stimulation, and embryo quality. (Fertil Steril 2005;83:336 40. 2005 by American Society for Reproductive Medicine.) Key Words: Endometrial thickness, pregnancy rates, patient age, embryo quality Endometrial receptivity refers to a physiological state of the endometrium that is optimal for embryo implantation. Endometrial receptivity is brought about primarily by ovarian steroid hormones, and is synchronized with fertilization and embryo development (1). However, the molecular mechanism underlying endometrial receptivity is poorly understood and a reliable molecular indicator of endometrial receptivity remains elusive. Correlating molecular events in the endometrium with its receptivity is even more difficult in the human for both technical and ethical reasons. Measurement of the endometrial thickness under ultrasound is both atraumatic and simple, and has been studied as a possible indicator for endometrial receptivity and predictor for treatment outcome after IVF embryo transfer procedures (2 5). However, the findings from these studies were inconclusive largely due to limited sample sizes. The aims of this study were to examine possible relationships between endometrial thickness and treatment outcome after IVF embryo transfer with a relatively large sample size, and to determine what factors affect such relationships. Received October 6, 2003; revised and accepted September 17, 2004. Reprints requests: Xingqi Zhang, Ph.D., 675 N. Saint Clair Street, Room 14-200, Chicago, IL 60611 (FAX: 312-695-4924; E-mail: x-zhang3@ northwestern.edu). MATERIALS AND METHODS This study included 897 IVF embryo transfer cycles performed between January 1999 and June 2002 at the IVF Program of the Feinberg School of Medicine of Northwestern University. Cycles using donor oocytes or cryopreserved embryos were excluded from this study. For patients who underwent more than one IVF embryo transfer cycle during the period of the study, only the first cycle was included. Endometrial thickness was not used as a criterion for cancellation of oocyte retrieval or embryo transfer. An exempt for patient consent was approved by the Institutional Review Board of Northwestern University. Procedures for controlled ovarian hyperstimulation, oocyte retrieval, IVF, embryo culture, and embryo transfer have been described elsewhere (6). Embryo transfer was usually performed on day 3, and postponed to day 5 only if more than four good quality embryos (8-cell, little or no fragmentation) were present on day 3. To assess treatment outcome, serum -hcg levels were measured 13 days after retrieval in all patients and repeated 2 days later if the first result was positive. Clinical pregnancy was confirmed by ultrasound observation of fetal cardiac activity 2 3 weeks after positive hcg tests. 336 Fertility and Sterility Vol. 83, No. 2, February 2005 0015-0282/05/$30.00 Copyright 2005 American Society for Reproductive Medicine, Published by Elsevier Inc. doi:10.1016/j.fertnstert.2004.09.020
Endometrial thickness was defined as the maximal distance between the echogenic interfaces of the myometrium and the endometrium measured in the plane through the central longitudinal axis of the uterus. The measurement recorded on the day of hcg injection was used for the analysis in this study. Peak serum E 2 concentration was defined as the value measured on the day of hcg injection. A step-wise multiple logistical regression analysis was used to assess the impact of age, endometrial thickness, peak E 2 level, and days of stimulation on treatment outcome. The possible dependence of endometrial thickness on patient age, peak serum E 2 concentration, or the number of days of ovarian stimulation before hcg injection was then assessed using a stepwise multiple linear regression analysis. Finally, study cycles were divided into two groups according to patient age at the beginning of the IVF cycle, days of gonadotropin stimulation, the quality of embryos for transfer, or the day of transfer. Each group was subdivided into three endometrial thickness groups ( 9 mm, 9 14 mm, and 14 mm), as in earlier studies (5, 7). A 2 test was used to compare differences in pregnancy rates (PR) between the three endometrial thickness groups separately for young ( 38 years old) or old patients ( 38 years old), short or long duration of gonadotropin stimulation ( 11 days or 11 days), transfers consisting of zero or at least one good quality embryo (8-cell stage without fragmentation on day 3, or blastocyst stage with well-defined inner cell mass and trophectoderm on day 5), and transfers performed on day 3 or day 5. The level of statistical significance was defined as P.05. In addition, the relationship between endometrial thickness and treatment outcome in each of the subgroups was also analyzed by a receiver operator characteristic (ROC) curve, using computer software from Analyse-It Ltd (Leeds, UK). RESULTS The age of the patients in this study ranged from 23 to 44 years, averaging 35.6 years. The overall PR per transfer was 43% (Table 1). Other demographic data in this study, including pregnancy history, length of infertility, and diagnosis were comparable to the latest national summary data published by the Centers for Disease Control. TABLE 1 Summary of demographic data. Total number of IVF embryo 897 transfer cycles Age (years) 35.6 / 3.7 Months of infertility 30 / 24 Diagnosis (%)* Male factor 26 Endometriosis/uterine 11 factor Ovarian factor 21 Tubal factor 15 Unexplained/other 36 No. of embryos transferred 2.6 / 0.9 Clinical pregnancy rates (%) 43.1 *The sum is greater than 100 because some patients have more than one diagnosis. Association of Cycle Characteristics with Treatment Outcome The relationships between treatment outcome (clinical pregnancy) and patient age, endometrial thickness, peak E 2 levels, and number of days of gonadotropin stimulation are presented in Table 2. Stepwise multiple logistical regression analyses showed that patient age is negatively correlated with treatment outcome, and that increasing endometrial thickness and peak serum E 2 levels are associated with improved PRs. Treatment outcome is independent of the number of days of stimulation. Factors Influencing Endometrial Thickness and its Effect on Treatment Outcome Endometrial thickness is positively correlated with peak E 2 concentrations in serum (Fig. 1), but is independent of patient age and the number of days of ovarian stimulation. When the IVF embryo transfer cycles were divided according to patient age, cycles with thin endometrium had lower PRs than the medium or thick groups only for the TABLE 2 Multiple logistical regression analyses. Independent variables Patient age Endometrial thickness Peak E 2 level Days of stimulation R 0.053 0.062 0.121 0.004 P 0.007 0.011 0.020 0.621 R Regression coefficient; R is considered statistically significant if P.05. Fertility and Sterility 337
FIGURE 1 peak estradiol concentration in serum. Correlation coefficient 0.06, P.04. FIGURE 3 treatment outcome: effect of duration ovarian stimulation. Value on top of each histogram represents number of clinical pregnancies/number IVF embryo transfer cycles. a vs. b, and a vs. c : P.001. relatively young patients (Fig. 2). Pregnancy rates were lower in older patients regardless of their endometrial thickness. Thin endometrium led to reduced PRs if the number of days of ovarian stimulation was 11 days (Fig. 3). Similarly, PRs were negatively affected by thin endometrium if the embryos available for transfer were of poor quality, or if embryo transfers were performed on day 3 (Table 3). The areas under the ROC curve of endometrial thickness vs. treatment outcome ranged from 0.51 to 0.54 (95% confidence interval, 0.45 0.58) for each of the subgroups by patient age, duration of stimulation, embryo quality, and day of embryo transfer. This suggests that the data analyzed in this study cannot establish a true endometrial thickness cutoff value below which pregnancy would not occur. FIGURE 2 treatment outcome: effect of patient age. Value on top of each histogram represents number of clinical pregnancies/number IVF embryo transfer cycles. a vs. b, a vs. c, and b vs. c : P.001. DISCUSSION Many characteristics of the human endometrium, including thickness, structure, and vascularization, can be studied noninvasively by ultrasound. Among these characteristics, endometrial thickness is probably the easiest and most reproducible measurement of endometrial development (8, 9). Conflicting reports exist concerning possible relationships between endometrial thickness and treatment outcome after IVF embryo transfer presumably because of the multiple confounding factors that may affect such a relationship (2 5). This study examined the possible association of endometrial thickness with treatment outcome, as well as factors that may affect endometrial thickness and its relationship with treatment outcome. Most studies in the literature compared the mean endometrial thickness between IVF embryo transfer cycles that resulted in a pregnancy and those that did not result in pregnancy, and very few of the reported studies analyzed more than 200 cycles (2). In a recent study of 1,186 IVF cycles using an ovarian stimulation protocol similar to the current study, the odds ratio for achieving a pregnancy was reduced if the endometrium was thin (4). This is supported by the findings of the current study. Studies that failed to find an association between endometrial thickness and treatment outcome either had a much smaller sample size (e.g., 10, 11) or used different ovarian stimulation protocols, such as the clomiphene citrate hmg ovarian stimulation protocol (e.g., 12, 13). An endometrial thickness 14 mm was found to be detrimental for successful pregnancy after IVF (3). However, data from a more recent study (5) and those from this study do not support this finding. 338 Zhang et al. Endometrial thickness and IVF outcome Vol. 83, No. 2, February 2005
TABLE 3 treatment outcome: effect of embryo quality or day of embryo transfer. Endometrial thickness (mm) Number of good-quality embryos in each transfer Day of embryo transfer At least one None Day 3 Day 5 9 41/93 (44) 6/40 (15) a 36/134 (27) a 39/98 (40) 9 14 219/469 (47) 36/118 (31) b 132/357 (37) b 91/156 (58) 14 62/117 (53) 23/60 (38) c 52/114 (46) c 32/58 (55) Treatment outcome is presented as clinical pregnancies/embryo transfers (%); within the same column, a vs. b, and a vs. c : P.05 ( 2 test). The possible associations of PRs with ovarian responses were analyzed in more detail in an earlier study (6). The present study, with a substantially larger sample size, confirmed our previous findings in that PRs are positively associated with peak E 2 levels but are independent of the length of ovarian stimulation (Table 2), the number of large follicles, or the number of oocytes retrieved (data not shown). An increase in endometrial thickness has been associated with younger patients and patients who have high peak serum E 2 levels (7). However, it is difficult to discern from that study whether age and E 2 levels can independently affect the endometrial thickness. If a correlation between endometrial thickness and patient age exits, the association of increased endometrial thickness with improved PRs could well be attributed to an age effect. Using a stepwise regression analysis, the present study found that endometrial thickness was dependent on serum E 2 levels, but not on patient age (Fig. 1). Because endometrial thickness is independent of patient age, it is probably not a proxy for any age-related effect on treatment outcome. The length of ovarian stimulation may presumably affect the endometrial thickness because a longer period of ovarian stimulation may allow additional time for the endometrium to develop. However, this assumption was not supported by the stepwise multiple regression analysis that failed to demonstrate a statistically significant relationship between the number of days of ovarian stimulation and endometrial thickness. Maternal age is a well-known determinant of fertility and treatment outcome after IVF embryo transfer (14 16). When the possible impact of endometrial thickness on treatment outcome was examined separately according to patient age, endometrial thickness was found to be positively associated with treatment outcome only for younger patients ( 38 years old). However, PRs were low in the older patients group regardless of their endometrial thickness. This finding suggests that the poor prognostic factors associated with advanced reproductive age cannot be compensated for by optimal endometrial development. Conversely, high quality embryos may be able to compensate for poor endometrial development. Although the duration of ovarian stimulation did not appear to influence endometrial thickness, the length of ovarian stimulation was a confounding factor influencing the impact of endometrial thickness on treatment outcome. Thus, the current results indicate that a thin endometrial strip predicts poor treatment outcome only if the length of ovarian stimulation is greater than the median duration. It is possible that a shorter period of gonadotropin administration reflects better ovarian function and therefore better oocyte quality, which in turn can compensate for less optimal endometrial development. When examined separately by the day of embryo transfer, treatment outcome is negatively affected by thin endometrium only for day 3 transfers but not for day 5 transfers. Furthermore, thin endometrium is associated with reduced PRs only for transfers in the absence of good quality embryos, but not for transfers with at least one good quality embryo. Taken together, our findings suggest that the requirement for optimal endometrial environment is less stringent when high quality embryos are available for transfer. Alternatively, the additional 2-day delay may provide an inadequate endometrium with additional time to develop. In conclusion, these data suggest that endometrial thickness has predictive value for treatment outcome in younger patients, in patients who require prolonged gonadotropin stimulation, or in those who have poor quality embryos. However, this study failed to establish a cutoff value for endometrial thickness below which pregnancy would not occur. With the exception of the age parameter, these results, taken together, suggest that inadequate endometrial thickness is primarily a problem in patients with relatively poor oocyte/embryo quality. The absence of an association be- Fertility and Sterility 339
tween endometrial thickness and PRs in the older patients might be due to the dominant impact of age per se on PRs. Acknowledgments: The authors wish to acknowledge the diligent technical assistance of the Embryology Laboratory staff: Karen Horan, Kathleen Kallmann, Yelena Pasman, Melanie Scott, and Gabriel Levine. REFERENCES 1. Psychoyos A. Hormonal control of ovoimplantation. Vitamins & Hormones 1973;31:201 56. 2. Freidler S, Schenker JG, Herman A, Lewin A. The role of ultrasonography in the evaluation of endometrial receptivity following assisted reproductive treatments: a critical review. Hum Reprod Update 1996; 2:323 35. 3. Weissman A, Gotileb L, Casper RF. The detrimental effect of increased endometrial thickness on implantation and pregnancy rates and outcome in an in vitro fertilization program. Fertil Steril 1999;71:147 9. 4. De Geyter C, Schmitter M, De Geyter M, Nieschlag E, Holzgreve W, Schneider HP. Prospective evaluation of the ultrasound appearance of the endometrium in a cohort of 1,186 infertile women. Fertil Steril 2000;73:106 13. 5. Dietterich C, Check JH, Choe JK, Nazari A, Lurie D. Increased endometrial thickness on the day of human chorionic gonadotropin (hcg) injection does not adversely effect pregnancy or implantation rates following in vitro fertilization-embryo transfer (IVF-ET). Fertil Steril 2002;77:781 6. 6. Chen C-H, Zhang X, Barnes R, Confino E, Milad M, Puscheck E, et al. Relationship between peak serum estradiol levels and treatment outcome in in vitro fertilization cycles after embryo transfer on day 3 or day 5. Fertil Steril 2003;80:75 9. 7. Noyes N, Liu HC, Sultan K, Schattman G, Rosenwaks Z. Endometrial thickness appears to be a significant factor in embryo implantation in in-vitro fertilization. Hum Reprod 1995;10:919 22. 8. Delisle MF, Villeneuve M, Boulvain M. Measurement of endometrial thickness with transvaginal ultrasonography: is it reproducible? J Ultrasound Med 1998;17:481 4. 9. Spandorfer SD, Arrendondo-Soberon F, Loret de Mola JR, Feinberg RF. Reliability of intraobserver and interobserver sonographic endometrial stripe thickness measurements. Fertil Steril 1998;70: 152 4. 10. Coulam CB, Bustillo M, Soensken DM, Britten S. Ultrasonographic predictors of implantation after assisted reproduction. Fertil Steril 1994; 62:1004 10. 11. Serafini P, Batzofin J, Nelson J, Olive D. Sonographic uterine predictors of pregnancy in women undergoing ovulation induction for assisted reproductive treatments. Fertil Steril 1994;62:815 22. 12. Oliveira JB, Baruffi RL, Mauri AL, Petersen CG, Campos MS, Franco JG Jr. Endometrial ultrasonography as a predictor of pregnancy in an in-vitro fertilization programme. Hum Reprod 1993;8:1312 5. 13. Strohmer H, Obruca A, Radner KM, Feichtinger W. Relationship of the individual uterine size and the endometrial thickness in stimulated cycles. Fertil Steril 1994;61:972 5. 14. Van Kooij RJ, Looman CW, Habbema JD, Dorland M, te Velde ER. Age-dependent decrease in embryo implantation rate after IVF. Fertil Steril 1996;66:769 75. 15. Yaron Y, Botchan A, Amit A, Kogosowski A, Yovel I, Lessing JB. Endometrial receptivity: the age-related decline in pregnancy rates and the effect of ovarian function. Fertil Steril 1993;60:312 8. 16. Dew JE, Don RA, Hughes GJ, Johnson TC, Steigrad SJ. The influence of advanced age on the outcome of assisted reproduction. J Assist Reprod Genet 1998;15:210 4. 340 Zhang et al. Endometrial thickness and IVF outcome Vol. 83, No. 2, February 2005