Women in the age group ≥40 years now represent about 12.5% of women attending in vitro fertilization (IVF) clinics (1). For the older woman desiring pregnancy, chronologic age is the best single predictor for clinical pregnancy. However, within this age group, the selection of those with the best prognosis for pregnancy and live birth has proved difficult (2). Although ovarian reserve tests, both static and dynamic, may predict ovarian response to controlled ovarian hyperstimulation (COH) with varying degrees of success, they have been found to be of limited use for the prediction of pregnancy 3, 4, 5, 6.

In their systematic and meta-analytic reviews, Broekmans et al. (2) concluded that the ovarian reserve tests known to date have only very modest predictive properties for pregnancy and are therefore far from suitable for relevant clinical use. Moreover, Mol et al. (7), while evaluating whether the expected benefits based on probabilities and value judgments of outcomes would justify testing for ovarian reserve before a first-cycle IVF, found that, for realistic ranges of the success rate after IVF and for distress ranges as were measured, treatment of all couples without testing was found to generate less distress than testing for ovarian reserve. They therefore concluded that, where current test accuracy and preference inventory among patients and physicians were used, testing for ovarian reserve did not seem to be useful for current IVF programs.

Ovarian stimulation characteristics during an IVF cycle are considered the most reliable sign of decrease ovarian reserve (8). Moreover, in older patients undergoing IVF, high estradiol (E2) to follicle or E2 to oocyte ratios on day of human chorionic gonadotropin (hCG) administration were associated with poor IVF outcome (9).

Following the aforementioned thorough systematic review of tests predicting IVF outcome, Broekmans et al. (2) asked whether the a priori identification of actual poor responders in the first IVF cycle has any prognostic value for their chances of conception in the course of a series of IVF cycles. Thus, our preliminary study of women in the 40 to 45 years age group undergoing IVF for the first time examined whether information obtained from the first IVF cycle has any prognostic value for the achievement of pregnancy within the first three treatment cycles.

Basal and IVF parameters immediately before and during each woman's first ever IVF cycle were carefully recorded for 97 women aged 40 to 45 years. All but three reached the stage of ovum retrieval after ovarian stimulation with either a midluteal gonadotropin-releasing hormone (GnRH) long protocol or flexible multidose GnRH antagonist protocol. Gonadotropins were administered in doses adjusted according to follicular development and serum E2 concentrations performed every 2 to 3 days. Clinical pregnancy was defined as fetal heart activity and visualization of a gestational sac on transvaginal ultrasound.

Ovarian stimulation characteristics, number of oocytes retrieved, and number of embryos transferred in the first IVF cycle attempt as well as the outcome of up to three consecutive cycles were recorded for each woman. A comparison of the data from the first treatment cycle was made for those who achieved a clinical pregnancy within three treatment cycles as compared with those who did not.

Results are presented as mean ± standard deviation. Differences in variables between the patients who conceived (pregnant group) and those who did not (nonpregnant group) were statistically analyzed with nonparametric Wilcoxon signed rank test and Student's t-test, as appropriate. P<.05 was considered statistically significant.

Of the 97 patients aged 40 to 45 years who started treatment with IVF, 25 achieved a clinical pregnancy in one of their first three consecutive treatment cycles. Table 1 shows the clinical characteristics of the first IVF cycle in women who conceived in one of the three consecutive IVF cycle attempts compared with those who did not.

Table 1. Comparison of basal and IVF stimulation characteristics in the first treatment cycle according to whether a clinical pregnancy was achieved in the first three consecutive cycle attempts.

		Pregnant group (n = 25)	Nonpregnant group (n = 72)	P value
	Age (years)	41.2 ± 1.3	42.2 ± 1.8	<.001
	Gravidity	2.1 ± 2.3	1.4 ± 1.5	NS
	Parity	0.7 ± 0.7	0.9 ± 1.0	NS
	Day-3 FSH (IU/L)	6.0 ± 2.7	7.3 ± 3.3	NS
	Days of stimulation	9.5 ± 2.2	10.1 ± 2.4	NS
	Gonadotropins required (IU)	3300 ± 1275	4025 ± 1725	NS
	Estradiol on day of hCG administration (pg/mL)	1628 ± 1213	1365 ± 1056	NS
	Progesterone on day of hCG administration (ng/mL)	0.7 ± 4.8	0.8 ± 0.6	NS
	Number of follicles of >14 mm on day of hCG administration	7.5 ± 4.8	6.7 ± 4.7	NS
	Number of oocytes recovered	7.8 ± 5.5	5.3 ± 4.2	<.02
	Number of embryos transferred	2.6 ± 0.9	2.2 ± 1.1	NS

Note: NS, not statistically significant.

The only statistically significant differences during the first treatment cycle between those who conceived and those who did not are in mean age (41.2 versus 42.2 years, respectively; P<.001) and mean number of oocytes recovered (7.8 versus 5.3, respectively; P<.02). Notably, there were no differences between the groups regarding day-3 follicle-stimulating hormone (FSH) concentrations, duration or amount of gonadotropin stimulation, serum E2 and progesterone levels on the day of hCG administration, or number of embryos transferred.

After the relative failure of a large battery of ovarian reserve tests to accurately predict clinical pregnancy resulting from IVF treatment (2), this study examined whether parameters obtained during the first IVF cycle of patients aged 40 to 45 years can improve the predictive value. The current data do not support this suggestion.

The well-documented prognosticators of age and number of oocytes recovered remained the only statistically significant differences between women who achieved a clinical pregnancy and those who did not. As age is obviously known and number of oocytes recovered is closely paralleled by the number of larger follicles obtained on ovarian stimulation, the stimulation characteristics and endocrinologic parameters obtained during the first cycle of IVF seem to offer little more in terms of prediction of clinical pregnancy than age, cycle day-3 FSH, inhibin B, or antimüllerian hormone serum levels, antral follicle count, or any combination of these.

Further research is thus needed to find reliable indicators for patients in the advanced maternal age group who can benefit from treatment with IVF and those who should be refused IVF. It has been suggested that, in view of the unreliability of tests that can predict pregnancy in this group, entering the first cycle of IVF without any prior testing is the preferable strategy today (2), but our current findings demonstrate that this strategy is also unhelpful as far as prediction of pregnancy is concerned and merely demonstrates our helplessness.

OBJECTIVE: To determine whether underweight and overweight affect controlled ovarian hyperstimulation and IVF outcomes in PCOS women.

DESIGN: Retrospective cohort analysis.

MATERIALS AND METHODS: A retrospective analysis involving 57 patients with PCOS diagnosis according to 2003 Rotterdam criteria. Patients were divided in two groups: group “lean PCOS” with BMI < 18.5 (n = 16) and group “overweight PCOS” with BMI > 25, (n = 41). Maternal and controlled ovarian hyperstimulation variables and IVF outcomes were analyzed: FSH, LH and E2 on day 3, dose of r-FSH, E2 level the day of hCG, ovarian responsiveness to gonadotropins defined by r-FSH dose per retrieved oocyte, number of oocytes, number of embryos and clinical pregnancy rate.

RESULTS: No difference was observed concerning the age between the two groups (31.8 years in group “lean PCOS” and 31.7 years in group “overweight PCOS”, p = 0.98). No difference was seen in basal hormonal status on day 3 (FSH = 5.8 IU/L versus 6.1 IU/L, p = 0.67 and LH = 7.4 IU/L versus 7.3 IU/L, p = 0.88 in “lean PCOS” and “overweight PCOS” respectively). Higher doses of gonadotropins were used for overweight PCOS (1790 IU versus 1042 IU, p < 0.001). No significant difference was observed concerning the number of retrieved oocytes but a trend to a higher number of retrieved oocytes in the group “lean PCOS” was noticed (14.8 versus 10.8, p = 0.08). Ovarian responsiveness to gonadotrophins was significantly higher in lean PCOS (93.5 r-FSH IU per retrieved oocyte versus 208 r-FSH IU per retrieved oocyte, p = 0.016) with higher E2 peak the day of hCG (3312.3 pg/mL versus 2208.9 pg/mL, p = 0.02). A higher number of embryos was obtained in the group “lean PCOS” (8.1 versus 4.5, p < 0.001). No significant difference was found in term of implantation rate (50% in group “lean PCOS” and 34.7% in group “overweight PCOS”, p = 0.22). No significant difference was observed concerning the pregnancy rate even if a trend to a higher pregnancy rate was noticed in group “lean PCOS” (31.3% per cycle versus 12.1%, p > 0.05).

CONCLUSIONS: In PCOS women, weight is an important parameter of r-FSH dose requirement. Lean PCOS have a greater ovarian responsiveness to gonadotropins and a milder stimulation is recommended. Overweight PCOS require more r-FSH dose in relation to a lower ovarian sensitivity.