Factors predictive of clinical pregnancy in the first intrauterine insemination cycle of 306 couples with favourable female patient characteristics

Human Fertility, 2013; 16(4): 286 290 ISSN 1464-7273 print/issn 1742-8149 online DOI: 10.3109/14647273.2013.841328 ORIGINAL ARTICLE Factors predictive of clinical pregnancy in the first intrauterine insemination cycle of 306 couples with favourable female patient characteristics YUNUS AYDIN, HIKMET HASSA, TUFAN OGE & VEHBI YAVUZ TOKGOZ Department of Obstetrics and Gynecology, Eskisehir Osmangazi University School of Medicine, Eskisehir, Turkey Abstract The objective of this study was to evaluate the factors predictive of clinical pregnancy in the first superovulation/intrauterine insemination (SO/IUI) cycle of couples with favourable female characteristics. We analyzed retrospectively the first SO/IUI cycle of 306 infertile couples with mild male factor infertility and unexplained infertility. The women had a favourable prognosis in terms of ovarian reserve. Univariate logistic regression analyses identified body mass index (BMI) [odds ratio (OR) 0.9, P 0.014], sperm concentration [OR 1.007, P 0.007] and inseminating motile sperm count (IMC) [OR 1.007, P 0.032] as significant predictive factors of clinical pregnancy. Multivariate logistic regression analysis identified BMI [OR 0.87, P 0.008] and sperm concentration [OR 1.008, P 0.011] as significant factors. Pregnant and non-pregnant groups did not differ significantly in terms of the age and smoking status of the woman, duration and type of infertility, length of the stimulation, total gonadotropin dosage or antral follicle count. Of the female characteristics investigated, BMI was the most significant predictive factor of clinical pregnancy in the first SO/IUI cycle of couples with unexplained or mild male factor infertility and favourable female characteristics. In overweight women, weight loss should be advised before starting SO/IUI. Sperm concentration and IMC were significant male predictive factors for clinical pregnancy in the first SO/IUI. Keywords: Infertility; intrauterine insemination; ovarian stimulation Introduction Following the evaluation of infertile couples, superovulation/intrauterine insemination (SO/IUI) is the preferential therapeutic approach, particularly for those with unexplained infertility, mild male factor infertility, minimal mild endometriosis and cervical hostility (ESHRE Capri Workshop Group, 2009). Moreover, the acceptable pregnancy rates and relatively lower incidence of complications in addition to the lower cost of this approach lead most clinicians to direct patients to SO/IUI in routine infertility management plans. Clinical pregnancy rates following SO/IUI are generally between 5% and 20% (Stone et al., 1999). A large variety of factors may affect the result of IUI cycles, and the prognostic factors for the success of SO/IUI cycles are still controversial because most studies have comprised heterogeneous groups of patients and the analysis of multiple IUI cycles (Nuojua-Huttunen et al., 1999; Erdem et al., 2008; Merviel et al., 2010; Badawy et al., 2009). The most important determining factors reported to date are the woman s age (Nuojua- Huttunen et al., 1999), treatment indications (Nuojua- Huttunen et al., 1999; Erdem et al., 2008), mature follicle count (Erdem et al., 2008), oestradiol concentration on the day of human chorionic gonadotropin (hcg) administration (Merviel et al., 2010) and sperm parameters such as motility, morphology and concentration (Badawy et al., 2009). Preliminary information regarding predictive factors of SO/IUI outcome is primarily based on these studies and to the best of our knowledge, there are few reports which have reviewed the predictive factors of SO/IUI in couples with favourable female characteristics. In one such study, Demir et al. (2011) found that the woman s age and total number of motile sperm were predictive variables for pregnancy. Nuojua-Huttunen et al. (1999) reported that the pregnancy rate per cycle was highest in the first treatment cycle (18%) and that it remained thereafter at approximately 10% through the fourth cycle. To minimise repeating SO/IUI cycles needlessly and to decrease the Correspondence: Yunus Aydin, MD, Department of Obstetrics and Gynecology, Eskisehir Osmangazi University School of Medicine, Eskisehir 26480, Turkey. Tel: 905335168740. E-mail: aydin.yunus@yahoo.com (Received 12 November 2012 ; revised 28 March 2013 ; accepted 12 April 2013 ) 286

Pregnancy prediction in insemination cycles 287 cumulative expense per couple, it is important to clarify the factors determining success, particularly in the first treatment cycles of select couples. The prediction of pregnancy as a result of SO/IUI cycles is vital in improving the counselling of couples, and selecting those most suitable for this treatment modality. This may not only increase success rates but also cumulative cost effectiveness. The aim of this study was to clarify the determining factors of clinical pregnancy in the first SO/IUI cycles of infertile couples with favourable female characteristics. Materials and methods In this retrospective cohort study, the records of the first SO/IUI cycles of 306 couples with unexplained and mild male factor infertility were reviewed. Treatment cycles were administered from March 2008 to December 2011 in the Reproductive Medicine Unit of Eskisehir Osmangazi University Hospital. The local ethics committee approved the study. All couples had failed to conceive after at least 1 year of unprotected intercourse and had been evaluated by a basic infertility workup consisting of a medical history, confirmation of ovulatory cycle with a day 21 progesterone value 3 ng/ml, confirmation of bilateral tubal patency by hysterosalpingography and/or laparoscopy and a semen analysis according to the World Health Organization (WHO) criteria (WHO Reference values, 1999). Age, body mass index (BMI), smoking status, day 3 hormone levels [follicle-stimulating hormone (FSH), luteinising hormone, oestradiol], antral follicle count (AFC), infertility type and duration were recorded. Non-smoker represents women that had never smoked. Couples who had normal sperm parameters and normal tubal patency were placed in the unexplained infertility group. Male factor infertility was defined according to the WHO criteria (1999) (sperm concentration 20 million/ml, 50% progressive motile spermatozoa, normal morphology 30%). The inclusion criterion was couples with favourable female characteristics. All women in the study were 35 years old with basal FSH levels 10 IU/ml and an AFC 6. The criteria for exclusion from the analysis due to female characteristics included the diagnosis of polycystic ovary syndrome, suspected endometriosis, expected diminished ovarian reserve (basal FSH levels 10 IU/ml, AFC 6) and the presence of previous ovarian surgery. Cycles also excluded were those that did not end with IUI subsequent to hcg administration due to a poor response to stimulation (absence of a follicle larger than 10 mm on the 12th day of stimulation or 6 mm endometrial thickness at the time of hcg) or cancellation due to a hyper-response, (presence of 4 follicles larger than 17 mm on the day of hcg). Due to the inclusion and exclusion criteria, the first SO/IUI cycles of 306 couples from 1143 cycles of 454 couples only were evaluated between March 2008 and December 2011. Baseline transvaginal ultrasonography (TV-USG) was performed on day 2 3 of the menstrual cycle to evaluate basal endometrial thickness and AFC. Ovarian stimulation was initiated with 75 112.5 IU recombinant FSH (r-fsh) [Gonal-F ; Serono, Istanbul, Turkey] on day 3 of the cycle, and continued for 7 days at the same dosage. On the 7th day of stimulation, if one or more 12-mm-sized follicles were observed, ovarian stimulation was continued with the same doses of 75 112.5 IU r-fsh; otherwise, the r-fsh dosage was increased to 37.5 75 IU. In the presence of one 17 mm mature follicle with one to two 13 16 mm follicles, ovulation was triggered with 250 mcg of recombinant hcg (Ovitrelle, Serono, Istanbul, Turkey). If 4 follicles larger than 17 mm and/or a 6 mm endometrial thickness were present at the time of hcg, the cycle was cancelled. The number of follicles with diameters of 11 13 mm, 14 16 mm and 17 mm on the day of hcg administration were recorded separately. Semen specimens were collected by masturbation into sterile containers. Within 1 h after collection, the sample was analysed for volume, concentration and motility according to WHO criteria. The density gradient technique was the preferred method for semen preparation. IUI was performed with a disposable 1-ml syringe (Gynetics : Gynetics Medical Products, Achel, Belgium) 36 h after the ovulation trigger. The patients remained in a supine position for 15 20 min after insemination. Luteal phase support was not given. A pregnancy test was performed 14 days after the IUI procedure by measuring serum hcg. If the test was positive, intrauterine pregnancy was confirmed with TV-USG at approximately the 6 7th gestational week. Clinical pregnancy was defined as the presence of an embryo with cardiac activity on TV-USG. The primary outcome of our study was clinical pregnancy. SPSS 15.0 and Medcalc 11.3 were used for statistical analysis and data expressed as the mean SD. The following comparisons were performed between cycles with and without clinical pregnancy using the unpaired t- test: mean age, BMI, basal FSH and basal oestradiol for women; duration of infertility; sperm concentration, total motile sperm (TMS) count and inseminated motile sperm count (IMC) before sperm preparation for men; and the number of mature follicles 17 mm on the day of hcg administration. Comparisons of infertility diagnosis (primary or secondary) and aetiology (unexplained or male) between cycles with or without clinical pregnancy were performed using the chi-squared test. Univariate and backward Wald multivariate logistic regression analyses were used to determine the effect of variables as determinants of clinical pregnancy. Variables evaluated in the regression analyses were as follows: age, BMI, basal FSH level, basal oestradiol level, AFC and smoking status for females; sperm concentration, motility and morphology, TMS count and IMC for males; duration of infertility, cause of infertility and numbers of follicles with diameters of 11 13 mm, 14 16 mm and 17 mm on the day of hcg as cycle characteristics. Odds ratios (ORs) and 95% confidence intervals (95%

288 Y. Aydin et al. CIs) were estimated. A receiver-operating characteristic (ROC) curve was used to calculate the area under the ROC curve (AUC) to identify significant sperm parameters for clinical pregnancy (sperm concentration, TMS and IMC) and the cut-off and minimum values with the best sensitivity and specificity. P 0.05 was considered statistically significant. Results As previously stated, the first SO/IUI cycles of 306 couples from the 1143 cycles of 454 couples were evaluated. The clinical pregnancy and multiple pregnancy rates were 13.3% and 9%. Demographic characteristics and differences in clinical parameters with regard to clinical pregnancy are shown in Table I. BMI was significantly lower (21 vs. 26, P 0.003) and sperm concentration (81.12 10 6 vs. 51.42 10 6, P 0.003), IMC (68.92 10 6 vs. 27.34 10 6, P 0.021) and TMS (195.43 10 6 vs. 147.21 10 6, P 0.005) were significantly higher in cycles that resulted in clinical pregnancy compared to non-pregnant cycles. Univariate logistic regression analyses were used to evaluate the predictive ability of the variables for clinical pregnancy; BMI was found to be the most significant predictor, and with each unit increase of BMI, the clinical pregnancy rate decreased with an OR of 0.9 (95% CI 0.827 0.979, P 0.014). Moreover, sperm concentration [OR 1.007, 95% CI 1.002 1.013, P 0.007] and IMC [OR 1.007, 95% CI 1.005 1.013, P 0.032] Table I. Couple s characteristics of pregnant and non-pregnant group. Pregnant group (Mean SD) Non-pregnant group (Mean SD) P value Female age, years 27.25 4.23 27.85 4.16 NS Female BMI, kg/m 2 21 1.81 26 4.79 0.003 Female smoking status, n (%) Smoker 14 (44) 18 (56) Non-smoker 181 (66) 193 (34) NS Duration of 3.78 3.40 4.69 3.54 NS infertility, years Type of infertility, n (%) Primary 25 (10) 225 (90) NS Secondary 16 (29) 40 (71) Aetiology of infertility, n (%) Unexplained 18 (12) 136 (88) Male 23 (15) 129 (85) NS Basal FSH (miu/ml) 6.42 2.13 7.01 5.14 NS Antral follicle count 14.42 6.06 13.41 8.63 NS Sperm concentration 81.12 22.13 51.42 16.24 0.003 (million) IMC (million) 68.92 22.14 27.34 13.97 0.021 TMS (million) 195.43 54.65 147.21 63.21 0.005 BMI, Body mass index; IMC, inseminated motile sperm count; TMS, total motile sperm count; SD, Standard deviation; NS, nonsignificant. Table II. Univariate logistic regression analysis of the variables for the prediction of clinical pregnancy. OR 95% CI P value Female age 0.966 0.888 1.052 NS Female BMI 0.900 0.827 0.979 0.014 Female smoking status 0.918 0.382 2.208 NS (smoker versus non-smoker) Duration of infertility 0.916 0.811 1.034 NS Infertility type 0.511 0.230 1.136 NS (primary versus secondary) Cause of infertility 1.556 0.760 3.188 NS (unexplained versus male) Follicle stimulating hormone 0.872 0.746 1.018 NS Luteinising hormone 1.033 0.956 1.115 NS Antral follicle count 1.012 0.977 1.048 NS Sperm concentration 1.007 1.002 1.013 0.007 IMC 1.007 1.005 1.013 0.032 Sperm motility 1.009 0.989 1.030 NS TMS 1.001 1.000 1.003 0.059 Sperm normal morphology 1.015 0.984 1.046 NS Follicle count of 10 13 mm 0.908 0.772 1.068 NS Follicle count of 14 16 mm 1.038 0.871 1.236 NS Follicle count of 17 mm 1.014 0.705 1.457 NS BMI, Body mass index, IMC, inseminated motile sperm count, TMS, total motile sperm count, NS, non-significant; CI, confidence interval; OR: odds ratio. were significant parameters in the prediction of clinical pregnancy (Table II). A multivariate logistic regression analysis revealed that BMI and FSH levels were significant predictors in step 1 of the analysis. After 21 steps of the backward Wald multivariate regression analysis, BMI [OR 0.87, 95% CI 0.795 0.965, P 0.008] and sperm concentration [OR 1.008, 95% CI 1.002 1.014, P 0.011] were the most significant predictors of clinical pregnancy (Table III). According to the ROC curve analysis, the minimum recommended IMC, sperm concentration and TMS count for obtaining clinical pregnancy were 0.8 million, 5.8 million/ml and 2.1 million, respectively. The critical cut-off values were 51.4 million with an AUC 0.566 (95% CI, 0.508 0.623; p 0.232) for IMC, 72.8 million/ml with an AUC 0.655 (95% CI, 0.599 0.709; p 0.001) for sperm concentration and 126 million with an AUC 0.646 (95% CI, 0.589 0.699; and p 0.002) for TMS count (Table IV). Discussion In the current study, BMI, sperm concentration and IMC were found to be significant predictive factors of Table III. Multivariate logistic regression analysis of significant variables to predict clinical pregnancy. OR 95% CI P value BMI 0.876 0.795 0.965 0.008 Sperm concentration 1.008 1.002 1.014 0.011 BMI, Body mass index; CI, confidence interval; OR, odds ratio. Human Fertility

Pregnancy prediction in insemination cycles 289 Table IV. Comparison of sperm parameters as predictors of pregnancy using ROC curves. clinical pregnancy in the first SO/IUI cycle of couples with favourable female characteristics. Among these predictive factors, BMI was found to be the most important variable, and the clinical pregnancy rate decreased significantly with each BMI unit increase. Souter et al. (2011) studied the effect of BMI on 477 women in 1189 SO/IUI cycles. They concluded that obese women require higher doses of medication and produce fewer follicles for a given dose, but once the medication and response are adjusted to overcome the weight effect, the success of the treatment cycle is comparable to that of normal-weight women. However, it has also been shown that increased leptin and changes in the level of adipokines may decrease fecundity and the success of reproductive treatments by impairing oocyte maturation, embryogenesis and implantation (Brannian & Hansen, 2002). According to our clinical policy, we aimed to develop one 17 mm mature follicle with one or two 13 16 mm follicles, and the dose was adjusted in all patients. Therefore, we believe that increased BMI will most likely decrease the chance of SO/IUI, not only due to anovulation or the requirement of higher doses of gonadotropin but also due to metabolic effects secondary to obesity. Previous reports indicated that the TMS count in the initial sperm analysis was a determining factor for choosing IUI treatment or IVF. A threshold value of 5 10 10 6 TMS has been reported as a criterion for choosing IVF instead of IUI (Wainer et al., 2004). Merviel et al. (2010) and Huang and Lai (1996) stated that the best pregnancy rates in IUI cycles were obtained with a TMS count 5 10 6. In this study, no pregnancies occurred with 2.1 million motile sperm. In addition to these studies, Badawy et al. (2009) stated that IUI used for male factor infertility has little chance of success when the IMC is 5 10 6. Sakhel et al. (2005) also stated that the IMC is the most significant parameter in the success of IUI cycles and that, for optimal results, the IMC must be greater than 5 million. Our results are consistent with previous studies, and IMC was one of the most significant predictors for success; no pregnancies occurred with an IMC lower than 0.8 million. One of the most important factors predicting success in reproductive medicine is female age. However, in SO/ IUI treatment cycles, the effect of age is controversial (Khalil et al., 2001; Nuojua-Huttunen et al., 1999). In Turkey, couples generally start to seek treatment for subfertility earlier than western populations, most likely Area (Mean SE) P value Lower bound 95% CI Upper bound IMC 0.566 0.055 0.232 0.508 0.623 Sperm concentration 0.655 0.047 0.001 0.599 0.709 TMS 0.646 0.048 0.002 0.589 0.699 IMC, inseminated motile sperm count; TMS, total motile sperm count; SE, standard error; CI, confidence interval. due to the socio-demographic specifications of the Turkish population. Most of the patients available for SO/ IUI cycles were therefore young couples. However, this is one of the advantages of our study because we only evaluated the results of females younger than 35 years old, and these patients comprised most of the patient population in our clinic. In addition, we only evaluated couples with favourable female characteristics including a day 3 FSH level 10 IU/ml and an AFC 6. The current study is exceptional for two main reasons; only couples with favourable female characteristics were analysed, and only the first IUI cycles were evaluated because previous reports have indicated that couples have the greatest chance of clinical pregnancy in their first cycle (Nuojua-Huttunen et al., 1999). Reporting the results of specific patient groups, as in our study, will provide clinicians with more information on how to increase the success of SO/IUI treatment. The retrospective nature of our study and limited number of patients are the most important limitations. Prospective studies performed in specific groups of couples will make it easier for clinicians to determine the most effective infertility treatment in terms of pregnancy success and lower cost. In conclusion, BMI in women was the most significant factor predicting clinical pregnancy in the first SO/IUI cycles of couples with unexplained and mild male factor infertility with favourable female characteristics including age 35 years, basal FSH 10 IU/ml and an antral follicle count 6. As a result, we advise weight loss before starting SO/IUI treatment in women with an elevated BMI. Sperm concentration and IMC were the most important male predictive factors for clinical pregnancy in IUI cycles. Although this was a retrospective analysis of couples with unexplained and/or mild male factor infertility with favourable female characteristics, the information could be used while counselling such couples, particularly in their first SO/IUI cycles. Acknowledgements None Declaration of Interest : The authors report no declarations of interest. The authors alone are responsible for the content and writing of the paper. References Badawy, A., Elnashar, A., & Eltotongy, M. (2009). Effect of sperm morphology and number on success of intrauterine insemination. Fertility and Sterility, 91, 777 781. Brannian, J.D. & Hansen, K.A. (2002). Leptin and ovarian folliculogenesis: implications for ovulation induction and ART outcomes. Seminars in Reproductive Medicine, 20, 103 112. Demir, B., Dilbaz, B., Cinar, O., Karadag, B., Tasci, Y., Kocak, M., et al. (2011). Factors affecting pregnancy outcome of intrauterine insemination cycles in couples with favourable female characteristics. Journal of Obstetrics and Gynaecology, 31, 420 423.

290 Y. Aydin et al. Erdem, A., Erdem, M., Atmaca, S., Korucuoglu, U., & Karabacak, O. (2008). Factors affecting live birth rate in intrauterine insemination cycles with recombinant gonadotrophin stimulation. Reproduction Biomedicine Online, 17, 199 206. ESHRE Capri Workshop Group. (2009). Intrauterine insemination. Human Reproduction Update, 15, 265 277. Huang, H.Y. & Lai, Y.M. (1996). The impact of the total motile sperm count on the success of intrauterine insemination with husband s spermatozoa. Journal of Assisted Reproduction and Genetics, 13, 56 63. Khalil, M.R., Rasmussen, P.E., Erb, K., Laursen, S.B., Rex, S., & Westergaard, L.G. (2001). Homologous intrauterine insemination. An evaluation of prognostic factors based on a review of 2473 cycles. Acta Obstetrica et Gynecologica Scandinavica, 80, 74 81. Nuojua-Huttunen, S., Tomas, C., Bloigu, R., Tuomivaara, L., & Martikainen, H. (1999). Intrauterine insemination treatment in subfertility: an analysis of factors affecting outcome. Human Reproduction, 14, 698 703. Merviel, P., Heraud, M.H., Grenier, N., Lourdel, E., Sanguinet, P., & Copin, H. (2010). Predictive factors for pregnancy after intrauterine insemination.iui.: an analysis of 1038 cycles and a review of the literature. Fertility and Sterility, 935, 79 88. Sakhel, K., Schwarck, S., Ashraf, M., & Abuzeid, M. (2005). Semen parameters as determinants of success in 1662 cycles of intrauterine insemination after controlled ovarian hyperstimulation. Fertility and Sterility, 84, 248 249. Souter, I., Baltagi, L.M., Kuleta, D., Meeker, J.D., & Petrozza, J.C. (2011). Women, weight, and fertility: the effect of body mass index on the outcome of superovulation/intrauterine insemination cycles. Fertility and Sterility, 95, 1042 1047. Stone, B.A., Vargyas, J.M., Ringler, G.E., Stein, A.L., & Marrs, R.P. (1999). Determinants of the outcome of intrauterine insemination: analysis of outcomes of 9963 consecutive cycles. American Journal of Obstetrics and Gynecology, 180, 1522 1534. Wainer, R., Albert, M., Dorion, A., Bailly, M., Berg è re, M., Lombroso, R., et al. (2004). Influence of the number of motile spermatozoa inseminated and of their morphology on the success of intrauterine insemination. Human Reproduction, 19, 2060 2065. World Health Organization. (1999). Reference Values of Semen Variables. 4th ed. Cambridge: Cambridge University Press. Human Fertility