Abstract
The essential cause of the existence of the menopause is that the ovary contains a finite number of follicles: these are progressively lost with time until insufficient remain to support menstrual cyclicity. Ovarian follicles are both the source of the female gamete, and the key site of reproductive hormone production. Depletion of follicle numbers therefore results in both loss of fertility and gonadal estrogen production, and thus differs substantially from the situation in the male where the two functions of the gonad are anatomically and functionally more independent, and loss of one does not necessitate loss of the other.
The essential cause of the existence of the menopause is that the ovary contains a finite number of follicles: these are progressively lost with time until insufficient remain to support menstrual cyclicity. Ovarian follicles are both the source of the female gamete, and the key site of reproductive hormone production. Depletion of follicle numbers therefore results in both loss of fertility and gonadal estrogen production, and thus differs substantially from the situation in the male where the two functions of the gonad are anatomically and functionally more independent, and loss of one does not necessitate loss of the other.
Establishment and Loss of the Ovarian Reserve
Ovarian follicles are formed during fetal life, with primordial follicles first seen in the ovary from about 18 weeks of gestation. Prior to this, the female germ cells have been specified, migrated to the gonadal ridge where they have proliferated before exiting mitosis and enter meiosis only to arrest at diplotene of meiosis. During this process they reorganize their interactions with surrounding somatic cells to form primordial follicles. This process is completed during later pregnancy and indeed some newly formed follicles start growing immediately so that during later fetal life and throughout childhood the ovary contains follicles at a range of stages of development up to small antral sizes. Subsequent development to ovulatory stages does not, of course, occur until after puberty when there is sufficient gonadotropic stimulation to support folliculogenesis through to completion of growth and maturation.
Ovaries from different women contain a very wide range of numbers of follicles. This is demonstrated in histologic studies, which have also led to the development of models showing the decline in the primordial follicle pool with age (Figure 3.1) [1]. These studies show that there is a range of at least 50-fold in the number of follicles in the ovaries of women of the same age. The various models that have demonstrated this have tended to present their data on a logarithmic scale which promotes the view that the decline in follicle number accelerates with age. While this may be true when presented as a percentage of the number of follicles present, a very different perspective is gained when the number of follicles is presented on a linear scale (Figure 3.2). This highlights how the vast majority of follicles are lost during the early years of life, even before reproductive maturity is achieved. These data can also be used to calculate the number of follicles being lost to either growth or atresia per month in women with a larger or smaller follicle complement. This analysis indicates that in the ovaries of the ‘average’ woman, with expected age of menopause of 51 years, some 900 follicles will start to grow each month at the peak of this activity (which is at age 14 years), declining to 200 per month at age 35. In women with a high number of follicles, this number is approximately 1900 follicles a month at age 35, whereas in women at the lower end of the still normal range (i.e. with expected age of menopause 42 years) it is approximately 26 follicles a month at the same age, thus fewer than 1 per day. Many of these early growing follicles will then be lost across all stages of folliculogenesis. This has implications for understanding the range of normality across the major milestones of reproductive life. In this context it has been proposed that women become essentially sterile some 10 years before the menopause and subfertile a further 10 years before that [2]. If one considers the range of age at menopause, it will rapidly become clear that the age at which a women may become subfertile will also vary considerably, and this also highlights that the reproductive lifespan, i.e. the interval between puberty and this proposed time of subfertility, will be relatively brief in women destined to go through a menopause in their early forties.
Figure 3.1 The number of non-growing follicles in women from conception to the menopause. The figure shows the data set (n = 325), the model (thicker line), the 95 per cent confidence interval for the model (dotted lines), and the 95 per cent prediction limits of the NGF population (outer solid lines).
Figure 3.2 Rates of non-growing follicle (NGF) recruitment towards maturation. Each sub-figure describes the absolute number of NGFs recruited per month, for ages from birth to 55 years, based on population decline predicted by the model shown in Figure 2.1. The top panel denotes recruitment for “average” women with menopause aged 51; maximum recruitment of 880 follicles per month occurs at 14 years 2 months, falling to 221 per month at age 35. The lower panels denote recruitment for women who will have an early or late menopause, left and right respectively (42 and 58 years). These indicate maximum recruitment of 104 follicles vs 7,520 follicles per month, at 14 years 2 months, falling to 26 and 1,900 per month at age 35.
Measurement of the Ovarian Reserve
From these considerations, the size of the remaining follicle pool is clearly a major determinant of age at menopause and indeed of time to menopause, although the rate of loss is also an essential consideration. The size of the ovarian reserve is starting to become clinically assessable. Analysis of the rate of loss cannot be determined from single measurements, although it can be extrapolated with repeated measures over time. Data are starting to emerge as to the value of measuring the ovarian reserve in the context of predicting the menopause, although this field remains in its infancy, due to the essential requirement for long-term follow-up. It is therefore not possible to predict ‘how long a woman will be fertile for’.
Before addressing how one might measure the ovarian reserve, it is important to discuss what is meant by the term, as unfortunately it is used to mean two separate although related aspects of ovarian biology. Most commonly in the clinical literature, and generally in the context of studies involving assisted reproduction, the ovarian reserve is used to mean the number of follicles that can be recruited to grow by administration of supraphysiological doses of FSH, i.e. as administered during ovarian stimulation for IVF. This is a very valuable measure as it will predict (to some extent) the number of oocytes that will be obtained after ovarian stimulation. It can be used to identify women either at risk of over response and therefore ovarian hyperstimulation syndrome, or conversely those whose response is less than their age or other markers of ovarian function (notably FSH) would have otherwise predicted. The follicles identified through this usage are already at an advanced, antral stage of gonadotropin-dependent growth and will have been in the growth phase for many weeks already: they constitute what can be termed the functional ovarian reserve. The second usage of the term ovarian reserve is used to mean the size of the primordial follicle pool. This is a more accurate biological usage and thus while ultimately more correct, its value in clinical practice is limited as the primordial follicle pool can only be determined at present by histologic analysis, and not in vivo. The size of the two follicle pools is related under normal circumstances, although the relationship between the two may well vary in different physiological and pathological states, for example in adolescence versus later adulthood, and in women with disorders such as hypothalamic amenorrhea and polycystic ovary syndrome, as well as in a range of systemic illnesses. It may also be partially suppressed during hormonal contraceptive use.
The assessment of the ovarian reserve has long been a goal in reproductive medicine, particularly in assisted reproduction, to optimize the prediction of the response of an individual woman. It can be used to improve the safety and effectiveness of ovarian stimulation regimes, and in the development of new regimes. As the ovarian reserve declines with age, then age is itself a measure of the ovarian reserve. It also includes an aspect of the quality of the oocytes within that reserve, reflected clinically in the increasing risk of non-conception, miscarriage and chromosomally abnormal conceptions with age. It does not, however, allow much in the way of individualization and therefore a range of biochemical and biophysical tests have been explored over the years. It has long been recognized that serum FSH increases with age and a high FSH is one of the diagnostic tests of the menopause, i.e. loss of the ovarian reserve. The biological function of FSH, however, is to regulate antral stages, follicle growth and selection such that only a single follicle emerges as dominant and mono-ovulation occurs, and the early stages of follicle growth are gonadotropin independent. Follicle-stimulating hormone remains a useful screening test in that a high FSH predicts a poor ovarian response at IVF, but the accuracy of this prediction is poor and the marked cycle-to-cycle variation within a single woman (as well as variation through the menstrual cycle) has led to a search for more robust indices. Measuring a woman’s FSH would be of little value in predicting her fertile life, although a high value (>10 IU/L, or more worrying >25 IU/L) would indicate that it is short, and that the menopause may be imminent. Estradiol is even less use as its production largely reflects the function of the single preovulatory follicle of that particular menstrual cycle and not any measure of the ovarian reserve. Inhibin B was identified as a product of the granulosa cells of smaller follicles and indeed is of predictive value in assisted reproduction. It also declines prior to the menopause although this decline is relatively late. The key physiological role of inhibin B is the negative regulation of FSH secretion particularly in the early follicular phase, so the two hormones are functionally interrelated. The identification of anti-Müllerian hormone (AMH) as a product of smaller preantral as well as early antral follicles has led to dramatic development in our ability to clinically assess the ovarian reserve and it has become of routine use in many IVF clinics around the world [3]. AMH is produced by granulosa cells of the follicles as soon as they start to grow, although not by primordial follicles (Figure 3.3). While the concentration of AMH in blood reflects the size of the true as well as the functional ovarian reserve, the relationship with the true ovarian reserve is therefore indirect. As it is produced by smaller and therefore less gonadotropin-dependent follicles, its concentration through the menstrual cycle is much less variable than that of the aforementioned reproductive hormones. While there is some variability, this is generally regarded as not clinically significant as relatively few women will be misclassified; this is a very substantial practical clinical advantage particularly where transvaginal ultrasound (to measure antral follicle count) is not immediately available.
