Anti-Müllerian hormone (AMH) is a recent addition in the armamentarium of fertility specialists. Many clinics have adopted AMH in their routine fertility assessment; however, it has to be used judiciously to tailor treatment for infertile couples. The high cost of drug regimens, discomfort to patients and the significant risk for complications associated with ovarian stimulation all justify the need for obtaining clinically relevant information before commencing treatment 1,2.
In an environment where safety and cost-effectiveness are central to assisted reproductive technology, patients have to be made aware of the possible high cost of and poor response to ovarian stimulation 3. It has been shown that the rate of decline of the ovarian reserve varies considerably between individual women, making it a challenge to design tests that estimate an individual’s remaining reproductive life span at a given age 4. In the analysis of 15 years of data from the UK (HFEA register data 2007) and multivariate analysis of US data from the Society of Assisted Reproductive Technology 5, maternal age remained the significant factor associated with reduced odds of conception and increased loss of early pregnancy.
AMH is barely detectable at birth and reaches its highest value after puberty. It then decreases progressively with age and becomes undetectable at menopause 6,7. It is synthesized in the granulosa cells of preantral and small antral follicles 8. In the ovary, AMH inhibits the recruitment of initial primordial follicles and decreases the sensitivity of preantral and small antral follicles to follicular stimulating hormone (FSH) 9,10. It has been reported that human antral follicles measuring less than 6 mm express the greatest amount of AMH and that levels decline as the antral follicles enlarge 11.
There is paucity of data on the use of AMH as an effective tool in the management of women, especially in those with low reproductive potential. It would be interesting and helpful to analyse the reproductive performance at different levels of AMH and identify, if possible, the critical level of AMH that may influence the reproductive performance of a woman.
The current trial was conducted to compare the two common modalities of treatment – intrauterine insemination (IUI) (group 1) and egg collection for IVF/ICSI (group 2) – in terms of their reproductive outcomes such as pregnancy, biochemical pregnancy, miscarriage, and clinical pregnancy rates. There is no internationally accepted cutoff value to discriminate between women with low and normal ovarian reserve. A level of 1.2 ng/ml was taken as a cutoff value in this study to discriminate between normal and diminished ovarian reserve, as it represented the lowest quartile in the population of patients attending our centre. We conducted this randomized prospective analysis of the two different modalities of treatment in women with low AMH to ascertain the mode of treatment that would be more useful for this specific group of patients.
Materials and methods
Sample size calculation
A total of 228 women were recruited into the study. An a-priori sample was calculated and the study required 80 women in each arm to detect a significant difference between the two allocated treatment groups, taking into consideration 5% significance level and a power of 80%, given an anticipated dropout rate or cancellation rate of 10%.
This study was carried out in the Dubai Gynaecology and Fertility Centre (DGFC), the Assisted Reproductive Center of the Dubai Health Authority, over a 12-month period from 2011 to 2012 after ethical approval of our local Ethical Committee was granted. We recruited 108 women into the IUI arm (group 1) and 120 women into the vaginal egg collection arm (group 2), as shown in Fig. 1. Patients were randomly assigned to one of the two groups using dark sealed envelopes prepared by an independent investigator with no clinical involvement in the trial; corresponding envelopes were opened only after the consented participants had completed the baseline assessment and it was time to allocate the interventions. The investigators, participants and sonographer were all blinded to the intervention. All patients were counselled, and those willing to participate were enrolled and underwent a baseline health assessment including a day-3 hormonal profile, which involved FSH, LH, prolactin, oestradiol, thyroid function tests and AMH, in addition to sonographic pelvic assessment performed by a single experienced sonographer.
All women attending the clinic with an AMH value of 1.2 ng/ml or less, who were found to have no other reason for their infertility (whether male or female infertility) and had no previous fertility treatment, were enrolled in the study.
Women who were not willing to participate in the trial, who had undergone previous radiotherapy, chemotherapy or ovarian surgery, who had endometriosis, proven tubal factor, single ovary, recurrent miscarriages, associated medical disease or male factor infertility and who had undergone previous fertility interventions such as IUI and IVF were excluded from the study.
Laboratory hormonal assessment was carried out in our local referral laboratory using an enzyme-linked immunosorbent assay for AMH detection, which was manufactured by Beckman Coulter (California, USA) (AMH Gen II A79765). Linearity was acceptable with observed values close to the expected mean recovery (106.3%). The functional sensitivity (20% coefficient of variation) calculated from precision profile data was 0.2 ng/ml. Within-batch and between-batch imprecisions, assessed over the concentration range of 0.7–9.8 ng/ml, were 5.3–11.4 and 3.8–17.3%, respectively. A transvaginal scan was performed before the ovarian stimulation to ensure that the ovaries were quiescent. Other factors of infertility such as tubal and male factors were excluded from the study using respective tests such as HSG and semen assessment. Using the antagonist protocol for all patients, ovarian stimulation was carried out with an initial dose of 300 U of the same recombinant FSH for the first 5 days; antagonist was added on day 6, and the rFSH dose was modified according to the response depending on the treatment arm. Stimulation was continued until a maximum of three follicles were more than 16 mm size; ovulation was triggered using choriogonadotrophin α [recombinant human chorionic gonadotropin (hCG)], which was administered subcutaneously (Ovitrelle 250 mcg; Merck Serono S.p.A., Darmstadt, Italy), and serum oestradiol and progesterone levels were determined on the day of the trigger (as per unit policy). In the IUI group (group 1), insemination was performed 34–36 h after hCG trigger using a Wallace IUI catheter. In the second group,vaginal egg collection was performed under ultrasound guidance 35–36 h after hCG trigger, and embryo transfer was performed on day 2 or 3 (cleavage stage) depending on the number and grade of the embryos available using a soft Wallace ET catheter under ultrasound guidance. Pregnancy tests were scheduled 16 days after the hCG trigger in both groups, and all patients in the two arms received luteal support with progesterone suppositories (Cyclogest 400 mg/day; Actavis, Bramstaple, UK). Pregnant women continued the same treatment until 12 weeks of pregnancy. Pregnancy was defined as a serum β-hCG level of more than 5 IU/l. Biochemical pregnancy was defined as a β-hCG level of more than 5 IU/ml that has resolved spontaneously in the absence of ultrasound evidence of either intrauterine or extrauterine pregnancy. Miscarriage was defined as a loss of evident intrauterine pregnancy sac, whereas clinical pregnancies were defined as those with a detectable foetal heart beat using ultrasound at 6 weeks.
Primary outcomes were pregnancy rates, biochemical pregnancy rates, miscarriage rates and clinical pregnancy rates. Secondary outcomes were ovarian response and pregnancy outcome with respect to subanalysis of AMH levels in the two groups (≤0.5 or >0.5 ng/ml). This AMH sublevel was determined on the basis of unpublished local data from a survey covering all patients attending our unit and not necessarily participating in this study.
Collected data were analysed using SPSS, version 19 (IBM, Illinois, Chicago, USA). Frequency, percentage, mean and SD were calculated. The χ2-test was used to find the associations between variables. To find significant differences of means in the two groups, the t-test was performed. Correlation coefficient was used to find out the relationship between two variables. The Wilcoxon rank-sum test was used to find the mean significant difference in discrete variables. A P-value of 0.05 or less was considered statistically significant.
Figure 1 shows the flow diagram for patients recruited into the study. Table 1 shows demographic data of the patients; none of the variables showed any significant difference between the two groups. Tables 2 and 3 show the correlation between AMH levels and other variables; there was a negative correlation with age and FSH levels and a positive correlation with antral follicle count (Figs 2 and 3). There was no difference with respect to BMI.
Seventy-nine out of 108 patients in group 1 and 105 out of 120 patients in group 2 responded to ovarian stimulation and hence continued the study; the remaining patients were excluded because of no response in both groups.
Comparison between the treatment method and pregnancy outcomes
The pregnancy rates in both groups were not significantly different but showed a slightly higher rate in the egg collection group (Table 4). The pregnancy rate was 15 out of 79 (18.9%) in group 1, whereas of the 92 patients in group 2 who underwent embryo transfer 21 became pregnant. The pregnancy rate per embryo transfer was 22.8%, whereas the pregnancy rate per started cycle was 21 out of 105 (20%). In group 2 subanalysis, 13 patients did not undergo embryo transfer because of various reasons (such as no eggs retrieved at ovum pickup, abnormal fertilization or failure to cleave).
The rate of biochemical pregnancy was significantly higher in the IUI group [4/15 (26.7%)] compared with that in the egg collection group [1/21(4.8%)] (P<0.05), whereas the miscarriage rate had no statistical significance in either group [2/11 (18.18%) and 2/20 (10%) in groups 1 and 2, respectively].
Anti-Müllerian hormone concentration and ovarian response
Subgroup analysis was performed in both groups; the groups were divided into two on the basis of their AMH levels being equal to or less than 0.5 ng/ml and more than 0.5 ng/ml.
In group 1 (IUI), 40 out of 79 women had an AMH level of less than or equal to 0.5 ng/ml and the remaining 39 had an AMH level of more than 0.5 ng/ml. The mean number of developed follicles was 2 and 3, respectively, which was not statistically significant, whereas the pregnancy rates were 5 and 33.3%, respectively; the difference was statistically significant (P<0.05), which might be related to and may reflect the quality of developed oocytes in addition to their number. Similarly, the clinical pregnancy rates were significantly different in both subgroups (0/15 and 9/15, respectively) (Tables 5 and 6).
In group 2 (egg collection), the mean number of follicles developed was 3.9, the mean number of eggs collected was 3.7, and the average number of mature oocytes retrieved was 3.9 versus 2.9 (21/105 and 84/105) in the pregnant and nonpregnant population, respectively. The average number of mature eggs was significantly higher (P<0.05) in patients who became pregnant compared with those who did not (Table 7). Normal fertilization was observed in 95/105 women (95.2%), whereas 5/105 women (4.7%) failed to fertilize, 2/105 (1.9%) had an abnormal fertilization, 1/105 (0.9%) failed to cleave and no eggs could be retrieved in 2/105 (1.9%) women. The mean number of embryos transferred was 2.09, and no embryos were available for cryopreservation in any of the patients.
In the subgroup analysis of women with an AMH level of less than or equal to 0.5 in group 2 (Table 8), significantly lower basal FSH levels, fewer follicles, fewer retrieved oocytes and mature oocytes, lower total dose of gonadotropins and lower oestradiol levels on the day of hCG trigger were seen when compared with women with an AMH level greater than 0.5 ng/ml. There was a highly significant difference (P<0.01) in pregnancy and clinical pregnancy rates between the two subgroups in group 2, with more favourable outcome for those with AMH levels greater than 0.5 ng/ml. There was a significant association (P<0.05) between AMH and embryo transfer rate in group 2; 77.3% of patients with an AMH level less than or equal to 0.5 ng/ml underwent embryo transfer compared with 95% of patients with an AMH level greater than 0.5 ng/ml. There was a statistically significant correlation (r=0.54) between antral follicle count and follicles more than 16 mm size, number of eggs retrieved and number of mature eggs (Table 9) in both subgroups. In group 2, it was also noted that the pregnancy rates were higher in patients with progesterone level lower than 1.5 ng/ml on the day of hCG trigger, which was a highly significant result (Table 10).
Our data show that women with a low level of AMH had a significantly lower biochemical pregnancy rate in the egg collection group compared with those in the IUI group. Women showed a higher trend of achieving a successful pregnancy (the pregnancy and clinical pregnancy rates) and a lower trend of miscarriage, although both did not reach the level of statistical significance, if they undergo egg collection rather than intrauterine insemination.
Our findings were comparable to the result of a recent study conducted by Nicopoullos and Abdalla 12 who studied 248 women who underwent IUI and 807 women who underwent egg collection and suggested that vaginal egg collection may represent the best chance for a successful outcome in poor responders and that abandoning and conversion to IUI does not improve the outcome.
In the present study, we did not find any correlation between low AMH and BMI levels, irrespective of the woman’s age. In contrast to our study, Freeman et al. 13 has recently reported in a large cross-sectional study that AMH levels in obese women were 65% lower than those in nonobese women of late reproductive years. The results of this study were different, probably because of the difference in the study population. He studied women in the late reproductive age group, whereas our study included women of all age groups. However, another study investigating the relationship between BMI and ovarian reserve markers, including AMH, failed to demonstrate a similar association 14.
In the subgroup analysis of group 1 (IUI), we noted significantly higher pregnancy and clinical pregnancy rates in women with an AMH level of more than 0.5 ng/ml compared with those with levels equal to or below 0.5 ng/ml. We could not find any comparable studies. In group 2, a significantly higher number of eggs could be collected from women with an AMH level greater than 0.5 ng/ml, who also demonstrated a higher percentage of mature oocytes and a higher percentage of successful embryo transfer and pregnancy, when compared with women with an AMH level less than 0.5 ng/ml, as was seen in the study by Li et al.15. A similar observation was made in a study by Abdallah and Nicopoullous 16, who studied 3916 cycles, all of which were stratified into different groups of AMH levels with respect to the number of cycles, mean number of eggs, patients achieving embryo transfer and pregnancy rates.
In the statistical analysis of group 2, the total requirement of gonadotropins was found to be significantly higher in women with AMH level less than 0.5 ng/ml when compared with women who had an AMH level greater than 0.5 ng/ml, which was similar to the result of a recent publication by Li et al.15.
There was no significant difference in gonadotropin dosage in women who underwent embryo transfer and those who had cancelled cycles, in contrast to a recent study in which the gonadotropin requirements were lower in the cancelled cycles 17.
In another point of subanalysis, we noticed no pregnancy in women with a low AMH level who underwent egg collection and had a progesterone level greater than 1.5 ng/ml on the day of hCG trigger. We could argue that cryopreservation of the available embryos in these women may further improve their chances of pregnancy, although further studies are required to confirm the finding. Similar results were observed in a study by Papanicolaou in women with antagonist cycles on a day 3 transfer 13. It was argued that when the progesterone level was high, performing transfer on day 3 exposed the embryo to an advanced endometrium and to progesterone-induced uterine growth factors ‘out of phase’, leading to the establishment of an asynchronous embryoendometrium cross-dialog 18.
Low level of AMH defines the poor responders but it does not warrant withholding treatment as surprising results are not unheard of in women with a low level of AMH. Lower levels may be associated with lower pregnancy rates but the factors that might improve their chances of a successful pregnancy are still a matter of important clinical query. Low levels of AMH should not be used to deny treatment but to predict prognosis and guide the choice of treatment. The results of our study may aid in the decision making by advising couples to choose egg collection for IVF/ICSI. The results of the subgroup analysis of AMH value (less than or more than 0.5 ng/ml) may partially address the clinical query, as this cutoff can be a good counselling tool to help couples make informed choices by providing a pragmatic approach to women with a low AMH level; especially, if they have a level of more than 0.5 ng/ml, they should choose egg collection for IVF/ICSI rather than IUI.
The evidence in this study provides a background for women to make informed choices about their mode of treatment. Electing a procedure such as egg collection over IUI may improve their chances of a successful pregnancy, although larger studies with a higher number of cycles are warranted. Observing the well-demonstrated correlation between AMH levels and various ovarian responses, a cutoff of 0.5 ng/ml can be used in clinical practice to maximize the chances of a successful pregnancy and minimize the risk of cycle or fresh embryo transfer cancellation. Embryos could be cryopreserved to improve the success rates if the progesterone values are higher. This approach is likely to benefit subfertile patients enrolled in assisted conception programs psychologically, physically and financially.
This study was funded by local DGFC funds. The authors thank all participants as well as nursing and medical staff at DGFC for their valuable contributions at various stages in this study. In addition, they thank the Statistical Department in Ajman University for performing the statistical work in this study.
Conflicts of interest
There are no conflicts of interest.
1. Macklon NS, Stouffer RL, Giudice LC, Fauser BC.The science behind 25 years of ovarian stimulation for in vitro fertilization.Endocr Rev2006;27:170–207.
2. Fauser BC, Devroey P, Yen SS, Gosden R, Crowley WF Jr, Baird DT, Bouchard P.Minimal ovarian stimulation for IVF: appraisal of potential benefits and drawbacks.Hum Reprod1999;14:2681–2686.
3. Streuli I, Fraisse T, Chapron C, Bijaoui G, Bischof P, de Ziegler D.Clinical uses of anti-Müllerian hormone assays: pitfalls and promises.Fertil Steril2009;91:226–230.
4. Te Velde ER, Pearson PL.The variability of female reproductive aging.Hum Reprod Update2002;8:141–154.
5. Baker VL, Luke B, Brown MB, Alvero R, Frattarelli JL, Usadi R, et al..Multivariate analysis of factors affecting probability of pregnancy and live birth with in vitro fertilization: an analysis of the Society for Assisted Reproductive Technology Clinic Outcomes Reporting System.Fertil Steril2009;94:1410–1416.
6. Hudson PL, Dougas I, Donahoe PK, Cate RL, Epstein J, Pepinsky RB, MacLaughlin DT.An immunoassay to detect human Müllerian inhibiting substance in males and females during normal development.J Clin Endocrinol Metab1990;70:16–22.
7. de Vet A, Laven JS, de Jong FH, Themmen AP, Fauser BC.Anti-Müllerian hormone serum levels: a putative marker for ovarian aging.Fertil Steril2002;77:357–362.
8. Modi D, Bhartiya D, Puri C.Developmental expression and cellular distribution of Müllerian inhibiting substance in the primate ovary.Reproduction2006;132:443–453.
9. Durlinger ALL, Gruijters MJG, Kramer P, Karels B, Kumar TR, Matzuk MM, et al..Anti-Müllerian hormone attenuates the effects of FSH on follicle development in the mouse ovary.Endocrinology2001;142:4891–4899.
10. Durlinger AL, Gruijters MJ, Kramer P, Karels B, Kumar TR, Matzuk MM, et al..Anti-Müllerian hormone inhibits initiation of primordial follicle growth in the mouse ovary.Endocrinology2002;143:1076–1084.
11. Weenen C, Laven JS, Von Bergh AR, Cranfield M, Groome NP, Visser JA, et al..Anti-Müllerian hormone expression pattern in the human ovary: potential implications for initial and cyclic follicle recruitment.Mol Hum Reprod2004;10:77–83.
12. Nicopoullos JD, Abdalla H.Poor response cycles: when should we cancel? Comparison of outcome between egg collection, intrauterine insemination conversion, and follow-up cycles after abandonment.Fertil Steril2011;95:68–71.
13. Freeman EW, Gracia CR, Sammel MD, Lin H, Lim LC, Strauss JF III.Association of anti-Müllerian hormone levels with obesity in late reproductive age women.Fertil Steril2007;87:101–106.
14. Buyuk E, Seifer DB, Illions E, Grazi RV, Lieman H.Elevated body mass index is associated with lower serum anti-Müllerian hormone levels is infertile women with diminished ovarian reserve but not with normal ovarian reserve.Fertil Steril2011;95:2364–2367.
15. Li HW, Yeung WS, Lau EY, Ho PC, Ng EH.Evaluating the performance of serum anti-Müllerian hormone concentration in predicting the live birth rate of controlled ovarian stimulation and intrauterine insemination.Fertil Steril2010;94:2177–2181.
16. Abdallah H, Nicopoullous J.Anti-Müllerian hormone in assisted reproduction: counseling tool not a rationing tool2011.Stockholm, Sweden:27th Annual meeting of ESHRE;P464.
17. Nardo LG, Gelbaya TA, Wilkinson H, Roberts SA, Yates A, Pemberton P, Laing I.Circulating basal anti-Müllerian hormone levels as predictor of ovarian response in women undergoing ovarian stimulation for in vitro fertilization.Fertil Steril2009;92:1586–1593.
18. Papanikolaou EG, Kolibianakis EM, Pozzobon C, Tank P, Tournaye H, Bourgain C, et al..Progesterone rise on the day of human chorionic gonadotropin administration impairs pregnancy outcome in day 3 single-embryo transfer, while has no effect on day 5 single blastocyst transfer.Fertil Steril2009;91:949–952.