Determination of serum anti-Mullerian hormone levels in a low-prognosis women treated in-vitro fertilization/intracytoplasmic sperm injection: A cohort study

Abstract Background Outcome prediction of participants treated with in-vitro fertilization or intracytoplasmic sperm injection (IVF/ICSI) using anti-Mullerian hormone (AMH) concentration has been widely used. According to the patient-oriented strategies encompassing individualized oocyte number (POSEIDON) definition, low prognosis Bologna responders have changed from poor. This definition divides low prognosis into 4 groups. Objective The purpose of this study was to assess blood AMH levels in the group of women treated with IVF/ICSI who were thought to have a low prognosis. Materials and Methods A retrospective cohort study among 252 suspected low-prognosis group participants was assessed between January 2016 and December 2019 at Morula IVF, National hospital, Surabaya, Indonesia. Observed AMH serum levels and pregnancy rates were compared among 4 subgroups. Results The AMH cutoff value was 1.7 ng/mL with a sensitivity of 86.7% and a specificity of 70% for diagnosing low-prognosis women using POSEIDON criteria. There was no difference in the pregnancy rate between those groups (p > 0.05). Conclusion AMH levels may indicate a poor prognosis for women having IVF/ICSI in accordance with POSEIDON guidelines. To predict the poor prognosis in women, the cutoff value must be identified.


Introduction
Numerous predictive factors can affect whether assisted reproductive technology is successful. This factor has been well studied. Sperm and oocyte quality are the main determinants of success (1). Oocyte quality is determined by a woman's age and ovarian reserve (2,3). It has been established that parameters such as age, antral follicle count (AFC), follicle stimulating hormone (FSH), day 3 of the inhibin cycle, and random anti-Mullerian hormone (AMH) serum during the menstrual cycle are linked to follicle pool depletion degree (4).
AMH is produced primarily by pre-antral and late-antral follicles. It is one of the most widely used indicators to forecast the success of assisted reproductive technology in terms of pregnancy and cancellation rate in intracytoplasmic sperm injection (ICSI), especially for women who respond poorly to the procedure. However, the use of AMH as a single marker is still controversial. Some studies argued about AMH as a single marker, while another study found variation in the cutoff, sensitivity, and specificity (5,6).
Women treated with IVF/ICSI were classified as poor responders and normoresponders in 2011 by the European Society of Human Reproduction and Embryology. This term is well known as the "Bologna Criteria." These criteria help the clinician predict the outcome and give the best counseling to the participants. Age, prior ovarian response, and an ovarian reserve test are among the Bologna criteria for assessing the status of a woman responder treated with IVF/ICSI (7).
The patient-oriented strategies encompassing individualized oocyte number (POSEIDON) criteria have been established. Its criterion redefines "poor responder" to be a "low prognosis" based on quantitative and qualitative data on age and number of aneuploidy embryos, ovarian markers, history of ovarian response to previous therapy and the ability to take oocytes needed to obtain at least one blastocyst euploidy in each patient (8).
Therefore, the purpose of this study is to assess the AMH value in women who are thought to have a low prognosis for success with IVF/ICSI therapy using POSEIDON criteria.

Study design
This study was a retrospective cohort study (historical and prospectively). The sample for this study were participants who came to Morula

Participant
Serum AMH levels were analyzed from

Ethical considerations
The National hospital Surabaya Ethics Committee in Indonesia granted approval for this study under No. 003/MIS/DIR/2020. Participants received guarantees that their personal data would be kept private and that the research findings would be released under pseudonyms. The participants were also asked to sign consent forms.

Statistical analysis
The statistical package for the social sciences (SPSS), SPSS Inc., Chicago, Illinois, USA, version 21.0, was used to analyze the data. To compare the means across more than 2 groups, Kruskal-Wallis was employed. The percentage of pregnancies in groups was compared using chi-square. AMH levels were predicted using the area under the curve (AUC), and the ideal cut-off value was found. It was deemed significant if the p-value was less than 0.05.

Results
This study comprised 252 participants in total. Demographic data of participants based on 4 subgroups of suspected low prognosis are shown in table I, figure 1. There were significant differences in the duration of infertility, FSH, oocyte count, and post-treatment AFC (p < 0.05). However, data for age, BMI, luteinizing hormone (LH), and P4 were not significantly different (Table I). Subgroups II and IV had longer infertility than the other subgroups. FSH levels in subgroups II and IV were higher than in other subgroups. The number of oocytes and AFC after subgroups I and II therapy were higher than the other subgroups.
We divided the AMH level into 3 subgroups based on the pattern of prior research subgrouping (6). Of 252 participants, there were 31 participants with AMH levels ≤ 0.39, 93 participants with AMH levels between 0.4-2.1, and 128 participants with AMH levels > 2.1. The parameters of age, duration of infertility, FSH, oocyte count, posttreatment AFC, BMI, LH, P4, and E2 were compared in each group based on the data of AMH level. Data showed that age, duration of infertility, FSH, oocyte count, posttreatment AFC, BMI, and LH were significantly different in each subgroup, while P4 and E2 were not significantly different (Table II). A relationship between age and AMH levels is depicted in figure  2. It is evident that the logarithmic AMH level declines with increasing age (y = 313, 71e -0,131x ).
According to the POSEIDON criteria, the optimal cutoff value for determining women with a low prognosis was 1.7 ng/mL with a sensitivity of 86.7% and a specificity of 70% with an AUC of 0.887 ( Figure 3). Interestingly, there were no variations in pregnancy rates across the 4 subgroups (p > 0.05).  Figure 2. Non-linear graph AMH level and age.

Discussion
This study found an interesting thing, differences The frequencies of pregnancies in each group did not differ substantially (p > 0.05). This study found the same result as in another study (27).
However, prior studies have demonstrated that women with adequate ovarian reserve (> 5 AFC) have a higher rate of pregnancy and live babies (28). Similar to the results of this study, subgroups I and II had higher AFC than subgroups III and IV, although not significantly different. A higher amount of AFC will result in a higher oocyte yield after increasing the gonadotropin dose, which translates to higher pregnancy and lives birth rate in subsequent cycles (4,28,29).
Determination of the optimal cutoff will help the clinician determine predictions for the success of IVF/ICSI therapy. We recommend that every laboratory that performs IVF/ICSI therapy determine their own optimal AMH levels on their examinations and continue to evaluate pregnancy rates, number of cycles, and cancellation of cycles and births in participants who will undergo IVF/ICSI therapy. Proper counseling will help participants understand the conditions and predictions that are best for their success.

Conclusion
Women having IVF/ICSI may have poor prognoses based on serum AMH levels. Finding the cutoff value is crucial in order to forecast women's poor prognoses. The low-prognosis groups' pregnancy rates were the same. Further evaluation is needed on the cancellation rate of cycles and births in each subgroup based on AMH predetermined.