Ovarian stromal blood flow in the prediction of ovarian response during in vitro fertilization treatment

Abstract

BACKGROUND: This study evaluated the role of ovarian stromal blood flow in the prediction of the ovarian response of infertile women by comparing age of women, body mass index (BMI), basal FSH concentration, antral follicle count (AFC) and ovarian stromal blood flow indices measured by power Doppler in two-dimensional ultrasound. Patients were aged <40 years with basal FSH <10 IU/l on recruitment for IVF treatment. METHODS: All received a standard regimen of ovarian stimulation in their first IVF cycle. AFC, pulsatility index, resistance index and peak systolic blood flow velocity of ovarian stromal vessels were determined on the second day of the treatment cycle prior to ovarian stimulation. Ovarian response was represented by the number of oocytes, serum oestradiol, and the duration and dosage of gonadotrophins. RESULTS: A total of 136 women were included in the analysis. Basal FSH concentration achieved the best predictive value in relation to the number of oocytes obtained, followed by AFC and BMI. AFC was the only predictive factor of serum oestradiol concentration on the day of HCG while BMI was predictive of the gonadotrophin dosage. CONCLUSION: Ovarian stromal blood flow indices measured by power Doppler ultrasound had no predictive value for the ovarian response.

Introduction

Recruitment and development of multiple follicles in response to gonadotrophin stimulation are the key factors leading to a successful outcome of IVF/embryo transfer treatment. Poor ovarian response may be associated with poor pregnancy rates and many of these cycles are cancelled without proceeding to oocyte retrieval (Keay et al., 1997; Tarlatzis et al., 2003). On the other hand, exaggerated ovarian response leads to an increased risk of ovarian hyperstimulation syndrome (OHSS) (Aboulghar and Mansour, 2003) and the resulting high serum oestradiol (E₂) concentrations may adversely affect the outcomes of the IVF treatment (Ng et al., 2000a). Prediction of ovarian responses prior to stimulation is useful in counselling patients and may be helpful in tailoring the dosage of gonadotrophin to individual patients.

Ultrasound is essential in the modern management of couples undergoing IVF treatment because it is used to predict and monitor the ovarian response, assess endometrial receptivity, and guide the transvaginal aspiration of oocytes and subsequent transcervical transfer of embryos to the uterus. Several ultrasound parameters have been examined to predict the ovarian response to gonadotrophins, including ovarian volume (Syrop et al., 1995, 1999; Lass et al., 1997), antral follicle count (Tomás et al., 1997; Chang et al., 1998; Frattarelli et al., 2000; Ng et al., 2000b; Hsieh et al., 2001; Nahum et al., 2001; Bancsi et al., 2002) and ovarian stromal blood flow (Zaidi et al., 1996; Engmann et al., 1999; Kupesic and Kurjak, 2002; Kupesic et al., 2003; Popovic-Todorovic et al., 2003). Hormonal markers such as early follicular serum FSH (Scott and Hofmann et al., 1995; Sharara et al., 1998), serum inhibin B (Seifer et al., 1997; Tinkanen et al., 1999; Dzik et al., 2000), serum anti-Müllerian hormone (Seifer et al., 2002; van Rooij et al., 2002; Fanchin et al., 2003) are also shown to be predictive of ovarian response.

Folliculogenesis in the human ovary is a complex process regulated by a variety of endocrine and paracrine signals (McGee and Hsueh, 2000). It has been suggested that the availability of an adequate vascular supply to provide endocrine and paracrine signals may play a key role in the regulation of follicle growth (Redmer and Reynolds, 1996). It is postulated that increased ovarian stromal blood flow may lead to a greater delivery of gonadotrophins to the granulosa cells of the developing follicles. Ovarian stromal blood flow can be assessed by colour Doppler and power Doppler ultrasound. Power Doppler is better suited to the study of the ovarian stromal blood flow as it is more sensitive to lower velocities and essentially angle-independent (Rubin et al., 1994; Guerriero et al., 1999). There are, however, very few studies addressing combinations of ultrasound and hormonal markers after a standard stimulation regimen. This is especially true for ovarian stromal blood flow.

The objective of this prospective study was to determine the significance of ovarian stromal blood flow in the prediction of the ovarian response of infertile women undergoing the first IVF cycle using a standard regimen of ovarian stimulation by comparing age of women, BMI, basal FSH concentration, antral follicle count (AFC) and ovarian stromal blood flow indices measured by power Doppler in two-dimensional (2D) ultrasound.

Materials and methods

Consecutive women attending the Department of Obstetrics and Gynaecology at the University of Hong Kong between January 2002 and December 2002 for their first IVF/embryo transfer treatment were recruited for study when the following criteria were met: (a) non-smoker; (b) no history of ovarian surgery; (c) no history of a chocolate cyst; and (d) no steroid treatment within 6 months before the IVF treatment. Women who were aged >40 years at the time of treatment or whose basal FSH concentration on repeated testing on recruitment was ≥10 IU/l were advised against the IVF treatment according to the recruitment guidelines issued by the Hospital Authority in Hong Kong. Poor visualization of ovaries because of abdominal position, an ovarian cyst of ≥20 mm in diameter and the presence of polycystic ovaries (Adams et al., 1986) were retrospectively excluded. Every patient gave written informed consent prior to participating in the study, which was approved by the Ethics Committee, Faculty of Medicine, the University of Hong Kong. They did not receive any monetary compensation for participation in the study. The results of ovarian stromal blood flow did not affect their IVF treatment.

Indications for IVF/embryo transfer included tubal, male, endometriosis, unexplained and mixed factors. ICSI was performed for couples with severe semen abnormalities where <100 000 motile spermatozoa were recovered after sperm preparation. In case of obstructive or non-obstructive azoospermia, surgically retrieved spermatozoa from the epididymis or testis respectively were used for ICSI. The details of the long protocol of ovarian stimulation regimen, gamete handling, standard insemination and ICSI were as previously described (Ng et al., 2000b). Serum FSH level was checked on day 2–4 of the period immediately preceding the treatment cycle. Serum FSH was measured by a two-site sandwich immunoassay (Bayer Corporation, Tarrytown, NY, USA) and the inter-assay and intra-assay coefficients of variation were 2.8 and 1.7% respectively.

All women were pretreated with buserelin (Suprecur; Hoechst, Frankfurt, Germany) nasal spray 150 µg four times a day from the mid-luteal phase of the cycle preceding the treatment cycle. On the second day of the treatment cycle, transvaginal scanning was performed by E.H.Y.N. at around 8–10 a.m. using a 6.5 MHz vaginal probe (Aloka, Model SSD-5500; Aloka, Tokyo, Japan) with the same ultrasound setting after the patients had emptied the bladder. All these patients were evaluated by 2D ultrasound only. The number of antral follicles <10 mm in diameter was counted (Ng et al., 2000a) and AFC was the sum of antral follicles on both sides. The presence or absence of ovarian stromal blood flow was determined by power Doppler ultrasound. Flow velocity waveforms were obtained from stromal blood vessels away from the ovarian capsule, if present. The gate of the Doppler apparatus was positioned when the vessel with good colour signals was identified on the screen. Pulsatility index (PI), resistance index (RI) and peak systolic blood flow velocity (PSV) of stromal vessels was calculated electronically when three similar, consecutive waveforms of good quality were obtained. The intra-observer coefficient of variation was 7% for AFC, 9.6% for PI, 4.1% for RI and 16% for PSV. Ovarian stromal blood flow was evaluated at three positions at random and the one with the highest PSV was chosen. When stromal flow was shown on both ovaries, the mean value was used as there were no significant differences in PI, RI or PSV between the left and right sides. In cases of ovarian flow on one ovary only, Doppler flow indices of that side were taken. PSV was considered to be zero in patients with absent ovarian blood flow on both sides.

When the ultrasound scanning showed no ovarian cyst and serum E₂ concentrations were below 200 pmol/l, HMG (Pergonal; Serono, Geneva, Switzerland) injections were started at 300 IU daily for the first 2 days followed by 150 IU daily. The ovarian response was monitored by serial transvaginal scanning and the HMG dosage was increased if there was no follicle ≥10 mm after 7 days of stimulation. HCG (Profasi; Serono) was given intramuscularly when the leading follicle reached 18 mm in diameter and there were at least three follicles ≥16 mm in diameter. Serum E₂ concentration was measured on the day of HCG administration. Cycles were cancelled when the follicles remained <10 mm after 14 days of stimulation. Oocyte retrieval was performed even when there was only one dominant follicle and was scheduled 36 h after the HCG injection, and any visible follicles were aspirated during the procedure.

Patients were advised to have two embryos replaced into the uterine cavity 48 h after the retrieval, but replacing three embryos was allowed. Excess good-quality embryos were frozen. All fresh embryos were cryopreserved if serum E₂ on the day of HCG injection was >20 000 pmol/l in order to reduce the risk of OHSS. The luteal phase was supported by two doses of HCG. A urine pregnancy test was done 16 days after embryo transfer. If it was positive, ultrasound examination was performed 10–14 days later to confirm intrauterine pregnancy and to determine the number of gestational sacs present. Only clinical pregnancies, defined by the presence of one or more gestational sacs or the histological confirmation of gestational product in miscarriages, were considered. Ongoing pregnancies were those pregnancies beyond 10–12 weeks of gestation, at which stage the patients were referred for antenatal care. Implantation rate was the proportion of embryos transferred that resulted in an intrauterine gestational sac.

Statistical analysis

The correlation coefficient between AFC and the number of oocytes obtained in the previous study (Ng et al., 2000b) was 0.36. Assuming that AFC and mean PSV had similar correlation coefficients, the sample size required would be 107 to give a test of significance of 0.01 and a power of 0.9 (Sigmastat; Jandel Scientific, San Rafael, CA, USA).

The primary outcome measure was the number of oocytes obtained. Secondary measures included serum E₂ level on the day of HCG and HMG duration/dosage. Continuous variables were not normally distributed and were given as median (2.5th–97.5th centiles), unless indicated otherwise. Statistical tests were carried out with the Mann–Whitney U test for continuous data and by the χ² and Fisher’s exact tests for categorical data, where appropriate. Correlation was assessed by the Spearman rank method and multiple regression analysis with the least-squares regression was applied to evaluate the predictive values of different parameters for the number of oocytes obtained, serum E₂ level on the day of HCG and HMG duration/dosage. A P value (two-tailed) of <0.05 was taken as significant.

Results

A total of 193 women underwent the first IVF cycle during the study period and 20 women were not eligible because of smoking (n = 7) or a history of ovarian surgery (n = 13). All eligible patients agreed to participate in the study and 37 women were excluded after the ultrasound examination: an ovarian cyst in 30 women and polycystic ovaries in seven women. Therefore, 136 women underwent ovarian stimulation and were included in the final analysis: 46 tubal factors; 13 endometriosis; 58 male infertility; 16 unexplained and three mixed causes. Table I summarizes the demographic data and ovarian response. One cycle did not proceed to oocyte retrieval because of absent follicular development. This patient was considered to have no oocyte obtained. Oocytes were obtained in all planned retrievals. Failed fertilization was encountered in nine cycles and in another cycle the embryo failed to cleave. Embryo transfer was postponed in three cycles because of the risk of OHSS. Embryo transfer was performed in 122 cycles and 23 clinical pregnancies resulted. The pregnancy rate was 16.9% per cycle initiated and 18.8% per transfer. Mean PI, RI and PSV of ovarian stromal vessels were comparable for different causes of infertility (data not shown).

Basal FSH concentration was negatively correlated with AFC (r = –0.329; P < 0.001), the number of eggs aspirated (r = –0.362; P < 0.001) and serum E₂ on the day of HCG (r = –0.254; P = 0.003), but was positively correlated with HMG dosage (r = 0.231; P = 0.007) and HMG duration (r = 0.202; P = 0.02). AFC was negatively correlated with age of women (r = –0.278; P < 0.001), basal FSH concentration (r = –0.329; P < 0.001), HMG dosage (r = –0.319; P < 0.001) and HMG duration (r = –0.314; P < 0.001) but positively correlated with mean RI (r = 0.239; P = 0.03), the number of eggs aspirated (r = 0.363; P < 0.001) and serum E₂ on HCG day (r = 0.336; P < 0.001). Mean RI was positively correlated with AFC (r = 0.239; P = 0.03).

Age of women, BMI, basal FSH concentration, AFC and mean PI/RI/PSV of ovarian stromal vessels were entered in a stepwise fashion in the multiple regression analysis using the number of oocytes obtained as the dependent variable, with a constant included in the equation. Basal FSH concentration had the largest R² change, which was followed by AFC and BMI (Table II). When these parameters were entered in a stepwise fashion in the multiple regression analysis using serum E₂ concentration on the day of HCG as the dependent variable, AFC was the only predictive factor (Table III). When these parameters were entered in a stepwise fashion in the multiple regression analysis using the dosage of HMG used as the dependent variable, BMI was the only predictive parameter (Table IV). The duration of HMG could not be predicted by these factors.

Women with absent (n = 6) or unilateral (n = 35) ovarian stromal blood flow were classified as Group A while those with bilateral ovarian stromal blood flow were classified as Group B. There were no significant differences in age of women, primary/secondary infertility, cause of infertility, duration of infertility, BMI, basal FSH concentration, AFC, HMG dosage/duration, serum E₂ concentration on the day of HCG and the number of oocytes between Group A and Group B (Table V). Women in Group B had significantly higher mean PSV than those in Group A (median 12.4 and 9.4 cm/s respectively, P = 0.01, Mann–Whitney test).

Women with mean RI ≤0.56 had significantly lower AFC only when compared with those with mean RI >0.56 (median 7.5 versus 9.0, P = 0.009, Mann–Whitney test). No significant differences were shown in all the above parameters between women with mean PSV <10 and ≥10 cm/s (data not shown). Twenty-nine (21.3%, 29/136) patients required an increase in the daily HMG dose after 7 days of stimulation. Women receiving increased HMG dosage had similar mean PI, RI and PSV of ovarian stromal vessels when compared with those on the usual dosage (data not shown).

Discussion

Many studies on the prediction of ovarian response are retrospective in nature and patients received different stimulation regimens and different starting doses of gonadotrophins. Moreover, various factors affecting ovarian response, such as age of women, BMI and AFC, were not considered at the same time. In order to avoid these confounding factors, we recruited women undergoing their first IVF cycle and receiving the same starting dose of gonadotrophin after a standard protocol of pituitary down-regulation for ovarian stimulation. All patients receiving ovarian stimulation were offered oocyte collection, unless there was absent follicular development after 14 days of stimulation. A multiple regression analysis was also applied to compare the values of these factors in the prediction of ovarian response. Ovarian response was represented by the number of oocytes aspirated. Serum E₂ on the day of HCG and HMG dosage/duration were considered secondary outcome measures because we have found that these parameters are also important in the counselling of patients undergoing IVF treatment. It is important to highlight that we examined a group of infertile women having a relatively normal ovarian reserve, as suggested by age <40 years at the time of treatment and basal FSH concentration <10 IU/l. Those patients who did not satisfy these criteria could have self-funded treatment cycles in a private unit.

In the present study, we showed that basal FSH concentration achieved the best predictive value in relation to the number of oocytes obtained, followed by AFC and BMI. AFC was the only predictive factor of serum E₂ concentration on the day of HCG while BMI was predictive of the HMG dosage used. With a similar study design, our previous study (Ng et al., 2000b) demonstrated that AFC achieved the best predictive value in the number of oocytes obtained, followed by basal FSH concentration, BMI and age of women (Ng et al., 2000b). Taking the results of both studies together, we can postulate that the predictive value of basal FSH concentration may be better than that of AFC in some patients.

Mean ovarian stromal PSV prior to pituitary down-regulation was demonstrated to be significantly correlated with the number of follicles, after controlling for patients’ age (Zaidi et al., 1996). Those with normal ovarian responses (more than six follicles at retrieval) had significantly higher velocity than poor responders (10.2 ± 5.8 versus 5.2 ± 4.2 cm/s). Other Doppler flow indices were not useful. Similarly, ovarian stromal PSV after pituitary down-regulation was the most important single independent predictor of the number of oocytes obtained in patients with normal basal FSH concentration, when compared with age of women, basal FSH concentration, E₂ concentration or FSH:LH ratio (Engmann et al., 1999). Bassil et al., (1997) reported that women with RI of ovarian blood flow >0.56 had a significantly longer stimulation and a significantly lower mean number of oocytes retrieved. Neither BMI nor AFC was included in these three studies. Both Zaidi et al., (1996) and Engmann et al., (1999) recruited patients with polycystic ovaries and used different starting doses of gonadotrophins.

Popovic-Todorovic et al. (2003) also evaluated ovarian stromal blood by 2D power Doppler ultrasound but a semiquantitative score was allocated to each ovary according to the number and area of the power Doppler signals. Total Doppler score was the sum of scores for each ovary: score 1 for poor flow, score 2 for moderate flow and score 3 for good flow. The number of oocytes was predicted by AFC, total Doppler score, serum testosterone concentration and smoking status. The scoring for power Doppler signals was subjective. Using 3D ultrasound with power Doppler, Kupesic and Kurjak (2002) demonstrated that AFC achieved the best predictive value for favourable IVF outcome, followed by ovarian stromal flow index, E₂ on HCG administration day, total ovarian volume, total ovarian stromal area and age of women. Subsequently, increasing patient age was found to be associated with poor ovarian response, as represented by smaller ovarian volume, lower AFC and poor stromal vascularity (Kupesic et al., 2003).

Ovarian stromal blood flow was not determined in the previous study (Ng et al., 2000b) as Power Doppler ultrasound was not available at that time. The results of the present study indicated that ovarian stromal blood flow indices measured by power Doppler ultrasound had no predictive value for the number of oocytes obtained, serum E₂ concentration and HMG dosage/duration. The ovarian response of women with absent ovarian stromal blood flow (n = 6) was not compared with that of women with stromal blood flow present because of the small number of subjects. These results may be explained by our previous studies assessing ovarian stromal blood flow in fertile Chinese women (Ng et al., 2003, 2004). We could not find any effect of age of these women on mean PSV of ovarian stromal blood vessels determined by 2D colour Doppler ultrasound. In combination with 3D ultrasound, power Doppler provides a unique tool with which to examine the ovarian stromal blood supply as a whole as opposed to analysis of small individual stromal vessels in 2D planes. We have recently shown that ovarian stromal vascularity was significantly lower in fertile Chinese women aged ≥41 years and the rate of decline of total ovarian vascularity index was only 0.18% per year (Ng et al., 2004). These data strongly suggest that reduction in ovarian stromal blood flow with increasing age is a relatively late phenomenon and that ovarian stromal blood flow is unlikely to be an early marker of ovarian reserve tests.

In summary, ovarian stromal blood flow indices measured by 2D power Doppler ultrasound had no predictive value for the ovarian response, including the number of oocytes obtained, serum E₂ concentration and HMG dosage/duration.

This study was funded by the Hong Kong Research Grant Council (HKU 7280/01M).

References