Abstract
Despite the notion that the first baby, Louise Brown, was born in 1978 following IVF performed in a natural menstrual cycle, ovarian stimulation became the golden standard of care in clinical IVF, since the number of oocytes retrieved is directly associated with pregnancy and life birth rates (, ). The aim should be to titrate the stimulation in such a way that the optimal number of follicles develops. Too few follicles (also referred to as low response) usually means poor IVF outcome, whereas too many developing follicles induce a risk for developing OHSS and possibly reduce the chance of success with increasing number of oocytes, if stimulation is not adjusted toward the end of the follicular phase (, ).
1.1 Introduction
Despite the notion that the first baby, Louise Brown, was born in 1978 following IVF performed in a natural menstrual cycle, ovarian stimulation became the gold standard of care in clinical IVF, since the number of oocytes retrieved is directly associated with pregnancy and life birth rates (1, 2). The aim should be to titrate the stimulation in such a way that the optimal number of follicles develops. Too few follicles (also referred to as low response) usually means poor IVF outcome, whereas too many developing follicles induce a risk for developing OHSS and possibly reduce the chance of success with increasing number of oocytes, if stimulation is not adjusted toward the end of the follicular phase (3, 4).
To perform a correct individualization of treatment protocols in IVF, an accurate prediction of ovarian response should be performed, based on the most sensitive markers of ovarian reserve. In this chapter, we discuss the use of the ovarian reserve markers to categorize women based on their anticipated ovarian response. As well, we describe the therapeutic approach and adjustments for an individualized ovarian stimulation for normo- and high responders.
1.2 Evaluation of Ovarian Response
1.2.1 Markers of Ovarian Function
The first step to performing an individualized ovarian stimulation is to evaluate the ovarian response. Accurate and reliable predictors of ovarian reserve are needed to identify patients likely to have poor response, normal response, or hyperresponse to treatment and to guide physicians in selecting the optimal dose of gonadotrophins for ovarian stimulation. To predict ovarian reserve and reproductive potential, several different measures of ovarian reserve have been identified over time, including biochemical measures and ovarian imaging, with varying degrees of success.
Nowadays, the three markers used most frequently in the clinical practice are basal FSH, antral follicle count (AFC), and anti-Müllerian hormone (AMH) levels.
1.2.1.1 Early Follicular Follicle-Stimulating Hormone Levels (Basal FSH)
Early follicular phase serum FSH levels are inversely correlated with the number of follicles in the ovary as determined histologically. It is important to mention that it should always be measured with basal estradiol levels. Higher day 3 FSH positively correlates with the age of patients and negatively with the estradiol (E2) response to stimulation and the number of oocytes retrieved, though it was found that basal FSH testing is limited by wide intercycle variability, which weakens its reliability (5–7). Basal FSH may serve as a predictor of decreased ovarian response, and an abnormally high basal FSH value has a high predictive value for decreased ovarian response. However, a normal value has a low negative predictive value for poor response. In addition, basal FSH has no value in predicting OHSS.
Although basal FSH level as a predictive marker has marked shortcomings, it is clear that the pregnancy and live birth rates declined with increasing FSH and advancing age, confirming the importance of utilization of age-specific FSH levels in assessing infertile women (8). For patients with FSH receptor defect, the basal FSH levels are elevated, but other ovarian reserve markers (AFC and AMH) are normal or even high.
1.2.1.2 Antral Follicle Count (AFC)
The ovary contains three distinct populations of developing follicles: primordial follicles, early-growing follicles, and antral follicles. A small proportion of early-growing follicles develop into antral follicles larger than 2 mm. These are highly responsive to FSH and can be readily visualized using transvaginal ultrasound. AFC is typically carried out at the beginning of a cycle, counting those small antral follicles between 2 and 8 mm. However, recent evidence suggests that AFC can be obtained at any point in the cycle without compromising accuracy, and with the recent development of three-dimensional ultrasound and other improvements in ultrasound resolution, antral follicles as small as 2 mm in diameter can now be reliably counted.
For several important IVF outcomes, AFC has been associated with good predictive value, showing a linear relationship with the number of retrieved oocytes and correlation with measures of ovarian response to gonadotrophins, including cycle cancellations as a result of poor response. However, it did not predict implantation rate, pregnancy rate, or live birth rate. Presently, AFC is an easy-to-perform, noninvasive approach that immediately provides essential predictive information on ovarian responsiveness, both low and hyperresponse, with acceptable interoperator reliability (9) (Table 1.1).
| AFC | AMH |
|---|---|
|
|
| |
1.2.1.3 Anti-Müllerian Hormone (AMH)
AMH is a transforming growth factor-b superfamily member, secreted exclusively from ovarian granulosa cells by primary and secondary preantral (but not primordial) and small antral follicles (up to 5–6 mm). Through paracrine mediation, AMH contributes to control follicle development from the reserve of primordial follicles constituted early in life, and its production seems to be independent of FSH. Its serum levels positively correlate with histologically determined primordial follicle number and negatively correlate with chronologic age.
During the last few years, AMH has emerged as one of the most important clinical markers for ovarian reserve in assisted reproductive techniques (ART). It has a strong correlation with the number of follicles, it is operator independent, it can predict reproductive life-span, and it is useful as a baseline assessment preceding ovarian stimulation for individualizing the therapeutic strategy. AMH levels are associated with ovarian response, becoming an excellent predictor of poor ovarian response, excessive response, and pregnancy outcomes in IVF (10).
The main limitation of AMH is the presence of significant biological intra- and intercycle variation, as we will discuss below, and the assay methods. Due to dissimilarity in the antibodies and assay sensitivities, in addition to interlaboratory variations, a considerable difference has been found between different assays, particularly for low AMH values, making their interpretation complicated. Furthermore, the presence of different AMH isoforms not detectable with most of the assays reduces the accuracy of the results. The highly sensitive, fully automated AMH assays that have been available since 2014 have replaced the older ELISA assays, thereby both providing faster results and improving interobserver reliability. However, no reliable converting factor has been identified; therefore, the cut points developed and reported for one commercial AMH assay are not generalizable to others (6, 10) (Table 1.1).
1.2.2 Variability of the Ovarian Reserve Test
Ovarian reserve tests have three main goals (Summary 1.1): [1] counseling IVF patients based on ovarian response prediction and the probability of live birth; [2] employing predicted ovarian response to optimize ovarian stimulation and minimize safety risks; and [3] assessing current and future fertility potential to allow women to decide when and how to proceed with family planning, fertility treatment, or fertility preservation. The ideal ovarian reserve test should be convenient; be reproducible; display little, if any, intracycle and intercycle variability; demonstrate high specificity to minimize the risk of wrongly diagnosing women as having diminished ovarian reserve; and accurately identify those at greatest risk of developing ovarian hyperstimulation prior to fertility treatment. An ovarian reserve measure without limitations has not yet been discovered, although both AFC and AMH have good predictive value and clearly have an added value together with female age and basal FSH for predicting ovarian response in IVF (6). Unfortunately, still today (anno 2019), there is a lack of studies combining the BMI to the aforementioned ovarian reserve markers, in order to have a more objective assessment of the patient’s response.
Summary 1.1 Evaluation of Ovarian Reserve
Ovarian Reserve Marker Goals
1. To counsel patients based on ovarian response prediction and live birth rate
2. To optimize ovarian stimulation and minimize safety risks
3. To assess current and future fertility potential for family planning
The Ideal Ovarian Reserve Test
Convenient for the patient
Reproducible
Little intracycle and intercycle variability
Accurate to identify patients who will have poor response or hyperresponse
AMH has been considered an ovarian reserve marker that can be measured independently of the cycle phase with minimal fluctuations in the menstrual cycle. Initially, those fluctuations were associated with the analysis, as analytical variability. It is true that different platforms will deliver different results depending on which molecular form of AMH is being measured, sample storage, freezing of samples, the assay protocols, and manual or automated methods used. However, recent studies have revealed inter- and intracycle variations that cannot be explained only by analytical variability, underlying the presence of a biologic AMH dynamic that is not yet fully understood. During the natural cycle, serum AMH levels seem to be higher during the follicular phase than the luteal phase, and recently, our group described a 20 percent intraindividual AMH variability during the ovarian cycle, using a fully automated AMH assay. As well, we observed 28 percent of short-term intercycle variations, probably caused by a biological variation in the number of AMH-producing antral follicles, since those follicles, due to their size, produce a significant amount of AMH. Similarly, when the intercycle variation of AFC was examined, cycle-to-cycle measurements revealed only moderate agreement in any range of counts due to the variable size of the growing follicle cohort among separate cycles. These observations may explain the intercycle variations of the ovarian response for the same patient when the same stimulation protocol is used. Hence, fluctuations in the same woman, intercycle and intracycle during natural cycles, question whether a single AMH or AFC measurement is enough for decision-making in our daily practice (10, 11).
1.2.3 The Best Moment to Measure AMH
Nowadays, it seems clear that AMH can fluctuate during the menstrual cycle with not only intracycle but also short-term intercycle variations that cannot be explained by the AMH assaying. As well, this happens with AFC. This variability should be considered carefully before making any decision in assisted reproductive technologies.
Hence, when should AMH and AFC be used?
Based on their variations shown during the whole cycle, it would be useful during our daily work to standardize the moment for the evaluation, as the results for our patients will be more homogeneous. As a routine, we perform the AFC during the first days of the cycle before starting the ovarian stimulation, so we visualize by ultrasound the follicles that would be expected to respond to the medication. This information is combined with the basal AMH levels on days 1–3 of the cycle, measured with automated assay (Elecsys, Roche). With both ovarian reserve markers and AMH and AFC during the early follicular phase, combined with the patient’s age and basal FSH, more accurate information is obtained for the ovarian stimulation prognosis, and the dosage of medication for each patient is adjusted based on the results.
1.3 Ovarian Response Classification: The Cutoff Values
An important factor when using ovarian reserve markers as predictors of ovarian response is to establish acceptable cutoff levels, values that can distinguish with sufficient accuracy women who are likely to have normal responses from those likely to have abnormal responses to ovarian stimulation. AMH and AFC values reported in literature are very variable (cutoff levels of AMH values for poor ovarian response have been reported between 0.1 and 2.97 ng/mL, which is within the range of normal values for AMH in healthy women), thus creating difficulties for clinicians in selecting the best cutoff values based on evidence (12).
The lack of a uniform definition of a poor response makes it difficult to compare studies and challenging to develop or assess any protocol to improve the outcome. In 2011, a consensus was reached to standardize the definition of poor ovarian response (POR) in a reproducible manner: the Bologna Criteria. These include at least two of the following three features: (1) advanced maternal age or any other risk factor for poor ovarian response (POR); (2) a previous POR; and (3) an abnormal ovarian reserve test (ORT). Two episodes of POR after maximal stimulation are sufficient to define a patient as a poor responder in the absence of advanced maternal age or abnormal ORT. A low or poor ovarian response was considered to be the retrieval of three or fewer oocytes (Summary 1.2) (13).
