JBRA Assist. Reprod. 2011;15(4):44-48
ARTIGO DE REVISÃO
doi: 10.5935/1518-0557.2011.15.4.10
1Genesis - Centre for Assistance in Human Reproduction, Brasília, Distrito Federal, Brazil
2Department of Obstetrics and Gynecology, Northeastern Ohio Universities College of Medicine and Pharmacy, United States of America
The authors of the present study declare that they have no conflict of interest regarding the attempted issues or the materials discussed in the text; this includes work relations, consultantships, honoraria and authors’ rights.
ABSTRACT
The current trends to postpone motherhood and the increase in demand for assistance in reproductive medicine highlight the need for seeking guidelines for the establishment of individualized treatment protocols. Although the ability to assess the functional ovarian reserve remains limited, the tests that are currently available provide guidance for assistance to infertile couples and ease the emotional burden of a process whose results are still lagging behind. In this paper, we will revisit literature and highlight current tests available for assessing functional ovarian reserve, intending to embase propaedeutic strategies and their interpretation in order to have reproductive success.
Keywords: Infertility, In Vitro Fertlization, Intracytoplasmic Sperm Injection, Ovarian Reserve, Assisted Reproduction, Follicle-Stimulating Hormone, Anti-Müllerian Hormone, Antral Follicle Count.
RESUMO
A tendência atual de se postergar a maternidade e o aumento das demandas pela assistência em medicina reprodutiva destacam a necessidade de se buscarem diretrizes para o estabelecimento de protocolos terapêuticos individualizados. Embora a capacidade de avaliar a reserva funcional ovariana permaneça limitada, os testes disponíveis atualmente oferecem um norte para assistência aos casais inférteis e amenizam a carga emocional de um processo de resultados ainda aquém dos almejados. Neste trabalho, revemos a literatura especializada e destacamos os testes de avaliação funcional da reserva ovariana disponíveis na atualidade, com o intuito de fornecer subsídios para o planejamento propedêutico e a interpretação dos resultados, com vistas ao sucesso reprodutivo.
Palavras-Chave: Infertilidade, Fertilização In Vitro, Injeção Intracitoplasmática de Espermatozóides, Reserva Ovariana, eprodução Assistida, Hormônio Folículo-Estimulante, Hormônio Anti-Mülleriano, Contagem de Folículos Antrais.
INTRODUCTION
Due to the modern tendency to postpone motherhood and the increasing demand for services of reproductive medicine, great attention has been given to the relationship between physiological ovarian ageing and progressive loss of follicular reserve. Follicular exhaustion is markedly accelerated from the fourth decade of life (Block, 1952), but initial decline preceeds this while women maintain regular menses between 30 and 40 years of age, representing a major challenge to specialists.
In this context, and especially when ovarian stimulation for assisted reproduction techniques (ART) is the first therapeutic choice, an ideal test for the inference of density and quality of remaining ovarian follicles after each normal menstrual cycle (ovarian reserve) is still absent (Broekmans et al, 2006; Carvalho et al, 2008; Domingues et al, 2010).
In the last two decades, great insights have been made in both serum and sonographic markers of female gonadal reserve. However, no consensus has been established on their relevance along the pre-treatment investigation of infertility. There is not strong evidence that any of the tested markers reflects the complex dynamics of ovarian follicles or allows specialists to infer quality or quantity of functional viable units. Women’s age seems to be the only confident marker of ovarian reserve reduction.
Knowledge on the currently available markers do not support real expectations on inference of the pool of follicles that would meet the continuity of ovulatory cycles and thus the perpetuation of long-term reproductive potential (Broekmans et al, 2006, Carvalho et al, 2008). Specialists deal with this complex scenario on initial counseling of infertile couples and the establishment of individualized treatment protocols.
In this article, we aim to revisit the rationale for ovarian reserve evaluation by providing a critical analysis of endocrine and sonographic tests, in addition to age as easily accessible markers, with the intent to assist physicians on their selection, interpretation, and establishment of individualized therapeutic strategies in ART cycles.
AGE AS AN OVARIAN RESERVE MARKER
Age is the single most important factor in determing the quality of ovarian reserve. It is well known that both the quantity and quality of ovarian follicles significantly decreases as a woman advances in age. The quantity of ovarian oocytes begins to decrease in utero. It is estimated that the maximum number of oocytes, approximately 6 million, occurs at twenty weeks gestation. The oocyte pool continues to decline until menopause mainly through the processes of ovulation and atresia (Maseelall & McGovern, 2008). Multiple studies have reported on the processes leading to diminished oocyte quality, including decreased discrimination in follicle selection, chromosomal abnormalities, and alterations in granulosa cell functions (Volarcik et al, 1998; Klein et al, 1996; Seifer et al, 1996; Seifer et al, 1999).
The inverse relationship between fecundity and age is long established. Menken et al reported that when compared to women aged 20 to 24 years, fertility is reduced by 14% in women 30 to 34 years and by 31% in women age 35 to 39 (Menken et al, 1986). Recent published data demonstrated significantly elevated pregnancy rates among women aged less than 35 years, even if follicle-stimulating hormone (FSH) levels were elevated, when compared to older patients with normal FSH levels (Luna et al, 2007).
Van Rooij reported that age was the significant factor associated with a poor response and embryo quality in a cohort of women aged 41 years and older undergoing in vitro fertilization (Van Rooij, 2003). Other studies suggest a similar relationship between advanced maternal age and response to assisted reproduction technologies (Gomes et al, 2009; Carvalho et al, 2010a).
According to the Centers for Disease Control and Prevention’s 2007 ART Success Rates Report, the percentages of ART cycles resulting in pregnancy and live births continuously declined beginning in the mid 30’s (CDC/ ASRM/SART, 2009). Data from Red Latinoamericana de Reproducción Asistida demonstrated a significant reduction of pregnancy rates per cycle with follicular aspiration: 38% in women 30 to 34 years; 31% in women 35 to 39 years; and 16% in older patients (Zegers-Hochschild et al, 2008). Furthermore, studies have shown that women of advanced maternal age unable to achieve pregnancy through in vitro fertlization (IVF) are able to conceive using donor oocytes from younger women (Steiner & Paulson, 2006).
Fertility is most favorable during the mid twenties. While there is no known cutoff value in age to predict fecundity, the literature suggests optimal fertility occurs before age 30. As we revisit serum markers in the evaluation of ovarian reserve, it is imperative to keep age at the forefront of our assessment as serum tests have greater predictability in older women.
BASAL ENDOCRINE MARKERS
1. Estradiol
Basal estradiol (E2) is still used in some services for the initial counselling of infertile couples and as pre-ART approach, but, considering the low predictive accuracy and the lack of standardization of a cut-off with high sensitivity and specificity, its adoption as a determinant for inclusion in therapeutic programs should be avoided (Broekmans et al, 2006).While a higher rate of cancelled ART cycles has been demonstrated in patients whose levels of E2 were < 20 pg / mL or ≥ 80 pg / mL (Fratarelli et al, 2000), the bulk of literature is flawed in supporting its clinical applicability as a marker of ovarian reserve. Even studies that favor its use for this purpose have been unable to correlate it to the follicular development or demonstrate its ability to predict the occurrence of pregnancy (Licciardi et al, 1995; Smotrych et al, 1995; Fratarelli et al, 2000). Fiçicioglu et al have also showed no significant differences on serum levels of E2 between good and poor responders’ to ART stimulation (Fiçicioglu et al, 2006).
2. Follicle-stimulating hormone
Basal FSH is the most studied and worldwidely used endocrine ovarian reserve marker (Broekmans et al, 2006; Silberstein et al, 2006). However there is great variability of in serum FSH levels within and between menstrual cycles, and the postulated interference of external factors such as smoking (Lambalk & de Koning, 1998).According with Broekmans et al, the accuracy of FSH in predicting poor response to ART stimulation would depend on the identification of very high serum thresholds, which are infrequent (Broekmans et al, 2006). Ashrafi et al showed a significant reduction in the number of retrieved oocytes and increased cancellation rates in IVF cycles of women with FSH levels ≥ 15 IU/mL (Ashrafi et al, 2005). Klinkert et al also suggested lower rates of pregnancy between women with levels ≥ 15 IU/L and those with lower levels, but with no statistical significance (Klinkert et al, 2005).In fact, it was recently demonstrated that pregnancy rates in women aged < 35 years with elevated basal FSH were higher than those of older women with normal levels of the hormone (Luna et al, 2007), reinforcing age as a main ovarian reserve marker. FSH is an easily accessible and low-cost marker, and could be useful in the pre-treatment evaluation of infertile women in specific groups, such as those carrying anovulatory cycles (van Montfrans et al, 2000) or endometriosis (Carvalho et al, 2010b), or in patients over 35 years of age (Watt et al, 2000, Luna et al, 2007).
3. Inhibin-B
The studies of Franchimont et al and Fowler et al pointed to the composition of follicular fluid as an index of maturation and suggested a possible association between concentrations of inhibin and oocyte quality (Franchimont et al, 1990; Fowler et al, 1995).Tinkanen et al investigated infertile women between 24 and 40 years of age, and found significant negative correlation between serum inhibin-B and FSH. Also, the authors found significant positive correlation between inhibin B and antral follicle count at basal evaluation (Tinkanen et al, 2001). In agreement with them, Seifer et al demonstrated the presence of higher estrogenic response and number of oocytes obtained after stimulation for ART when women had serum inhibin-B ≥ 45 pg/mL, while cancellations were three times more frequent among patients with lower levels (Seifer et al, 1997).Other studies, however, were unable to repeat favorable results on the use of inhibin-B as a marker of ovarian reserve (Corson et al, 1999; Scheffer et al, 2007). The systematic review of Broekmans et al was categorical in stating that the high rate of false positives in routine determination of basal inhibin-B should lead to unnecessary exclusion of patients from IVF programs. Moreover, they noted that even using very low levels, the accuracy of inhibin-B in predicting poor response was only modest when compared the other available markers (Broekmans et al, 2006).
4. Anti-müllerian hormone
In the last decade, anti-müllerian hormone (AMH) has been used as predictor ovarian follicular reserve, since it signals the size of the pool of growing follicles, which in turn would be correlated to the stock of primordial follicles. In other words, it is considered to be an accurate marker, able to estimate the amount and activity of recruitable follicles at early stages of maturation, being, thus, more reliable for prediction of ovarian reserve (van Rooij et al, 2002; Fanchin et al, 2003, Muttukrishna et al, 2005, Scheffer et al, 2007).Since AMH reduced variability in serum has been demonstrated along menstrual cycles, the marker has gained credibility and consistency when compared with FSH, inhibin-B and E2 (Fanchin et al, 2005; La Marca et al, 2006; Elgindy et al, 2008). Elgindy et al performed enzyme immunoassay and found median AMH values of 1.4 ± 1.1 ng / mL, 1.43 ± 1.08 ng / mL and 1.35 ± 1.02 ng / mL in follicular, ovulatory and mid-luteal phases, respectively, confirming results from previous studies and realizing that there was no gonadotrophic influence on its production (Hehenkamp et al, 2006; La Marca et al, 2006). Likewise, Tsepelidis et al found stable levels, even higher, averaging 2.4 ± 1.1 ng / mL along menstrual cycle in normo-ovulatory women (Tsepelidis et al, 2007). Reproducibility of AMH in subsequent cycles was demonstrated by Fanchin et al, who studied its behavior in infertile women aged between 20 and 40 years. The authors demonstrated that variations in AMH serum levels in consecutive cycles were lower than those found with FSH, inhibin-B and E2; moreover, there was a positive correlation with antral follicle count (AFC), but this marker was more susceptible to variations (Fanchin et al, 2005).In ART cycles, studies have associated lower levels of AMH to poor response in gonadotropin stimulated cycles (van Rooij et al, 2002; Tremellen et al, 2005; Fiçicioglu et al, 2006; Gleicher et al, 2010). Muttukrishna et al prospectively evaluated women over 38 years of age with a previous poor response and supposed low ovarian reserve based on elevated basal FSH; reduced levels of AMH and inhibin-B were clearly demonstrated among women whose cycles were canceled, but AMH were considered to be a superior marker among the three markers (Muttukrishna et al, 2004).La Marca et al confirmed expectations on AMH as a predictor of poor response, demonstrating a sensitivity of 80% and specificity of 93% in ART cycles, when considered as a threshold of 0.75 ng/mL (La Marca et al, 2007). Similar conclusions have been reached with higher cut off levels (Gnoth et al, 2008), maintaining a lack of standardization of a cut-off with satisfactory sensitivity and specificity for routine use.Also, an association between AMH levels and the occurrence of pregnancy after ART has been studied. Values above 2.7 ng / mL were associated with higher rates of implantation and pregnancy, but with no significance (Silberstein et al, 2006). Recent data from the study of Gnoth et al defended AMH as a marker of response to stimulus but was unable to find its value in predicting pregnancy (Gnoth et al, 2008). Arabzadeh et al have recently demonstrated positive correlation between serum basal AMH levels and the number of oocytes retrieved and the percentage of mature oocytes, but significance was also not reached in correlation between the marker and implantation rate (Arabzaedh et al, 2010).Given the scarcity of studies and lack of standardization of a cut-off with high sensitivity and specificity, it is still early to determine the real significance of AMH in the prediction of ovarian reserve and response. However, based on the favorable results presented and the significant correlation with important variables such as AFC, number of oocyte yielded and their maturity (van Rooij et al, 2002; Fiçicioglu et al, 2006; La Marca et al, 2007), we believe in the promising future of this marker.
ULTRASONOGRAPHIC MARKERS
1. Antral Follicle Count
Pre-therapeutic ultrasonographic AFC has been shown to be an excellent predictor of ovarian reserve and response to gonadotropins, with significant superiority in relation to other markers (Bancsi et al, 2002; Scheffer et al, 2003; Verhagen et al, 2008). According with Muttukrishna et al, AFC should allow identification of 89% of poor responders, in spite of low specificity (Muttukrishna et al, 2005), and other authors admit its importance as a screening test for infertile couples (Broekmans et al, 2006). Data from the study of Maseelall et al demonstrated that chance of live birth is lower with an AFC of 10 or less; authors also suggested that patients should be advised about the increased risks of miscarriage, cycle cancellation, use of more gonadotropins, fewer oocytes yielded, and lower clinical pregnancy rates if compared with women with an AFC of 11 or greater (Maseelall et al, 2009).Recent studies have shown significant correlations between AFC and the concentrations of classically used serum markers (Haadsma et al, 2007) and AMH (Fanchin et al, 2003, Visser & Themmen, 2005), as well as significant differences between women with normal response to stimulation and poor responders. In a recent study, Elgindy et al assessed AFC (follicles with up to 10 mm in diameter) and found 10.1 ± 3.0 and 5.7 ± 1.0 antral follicles for those two groups, respectively (Elgindy et al, 2008). Special attention has been given to initial antral follicles when considering ovarian reserve evaluation. Klinkert et al demonstrated that the frequency of normal response to stimulation was significantly elevated in patients with AFC ≥ 5 units with up to 5 mm in diameter (Klinkert et al, 2005) and Haadsma et al demonstrated significant correlation between follicles with up to 6 mm in diameter and all ovarian reserve endocrine tests, in contrast to that observed with larger follicles, which were consistent only with ovarian volume and serum levels of inhibin-B. Also, the authors postulated that smaller follicles should be significantly reduced in number with the advance of age, while the pool of larger follicles should virtually remain unchanged until the middle of fifth decade of life (Haadsma et al, 2007).In summary, results of literature seem to converge to recognition of the importance of AFC as a predictor of ovarian response. Although the test should not be used to exclude women from IVF programs, it should a very useful instrument for couples counseling and determining individualized treatment protocols in IVF cycles. Since endovaginal ultrasound evaluation is normally required in pre-ART evaluation, we believe that it should be definitely included as a routine marker of gonadal reserve.
2. Ovarian Volume
Ovarian volume (OV) has also been associated with fertility potential. In a study evaluating women undergoing ART cycles, between 23 and 46 years of age, Syrop et al demonstrated a clear reduction in the number of yielded oocytes and pregnancy rates, with decreased ovarian volumes (Syrop et al, 1999). However, ten well-designed studies on ovarian volume as a marker of gonadal reserve were systematically revised by Broekmans et al, with little clinical applicability in the prediction of poor response or pregnancy (Broekmans et al, 2006).Recently, Bowen et al demonstrated significant correlation between the decrease in ovarian measures, increased age and elevated serum FSH (Bowen et al, 2007), which was not demonstrated by Elgindy et al, whose data showed no difference in OV between young women with normal and poor response in ART cycles (means of 4.1 ± 0.66 cm3 and 3.36 ± 0.71 cm3, respectively) (Elgindy et al, 2008). McIlven et al reached a similar conclusion when evaluating women at high risk for ART cancellation (McIlven et al, 2007).
FINAL CONSIDERATIONS
Since single ovarian reserve markers have unsatisfactory sensitivity and specificity, a combination of of markers should be a more efficient strategy in clinical practice.
Although a recent meta-analysis has concluded for the equivalent potential of multivariate models and AFC alone for predicting poor response in IVF cycles (Verhagen et al, 2008), combinations have not been sufficiently evaluated and there are few data comparing AMH and AFC. Further studies evaluating this and other combinations will be necessary to discard the use of multivariate models, which, in our view, are still the best strategy to address conclusions on ovarian reserve. For now, none of the currently available tests of ovarian reserve can reliably predict response or pregnancy in ART. They shall be used as complementary tools in counseling process, which must be based on age and individuality. This rationale is further supported by a review of ovarian testing by Sun et al (Sun etal, 2008). The authors state that interpretation of ovarian reserve testing should be based on women’s age and that in women younger than 35 any single test suggesting decreased reserve should be supported by additional testing.
Menken J, Trussel J, Larsen U. Age and Infertility. Science. 1986;233:1389-94.
Steiner AZ, Paulson RJ. Oocyte Donation. Clin Onstet Gyncol. 2006;49:44-54.