Hum. Reprod. Advance Access originally published online on May 10, 2007
Human Reproduction 2007 22(7):1932-1941; doi:10.1093/humrep/dem104
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Does 3D ultrasound offer any advantage in the pretreatment assessment of ovarian reserve and prediction of outcome after assisted reproduction treatment?
Academic Division of Reproductive Medicine and Surgery, School of Human Development, University of Nottingham, Nottingham, UK
1 Correspondence address. Tel: +44 115 82 30700; Fax: +44 115 82 30651; E-mail: k.jayaprakasan{at}nottingham.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionAcknowledgementReferences |
|---|
BACKGROUND: This study compared the predictive value of the antral follicle count (AFC) measured from stored 3D data by three methods including a ‘2D equivalent technique’ and two 3D techniques, the ‘multiplanar view’ and ‘rendered inversion mode’, as a test of ovarian reserve and treatment outcome.
METHODS: One hundred consecutive subjects aged <40 years with basal FSH levels <12 IU/l underwent transvaginal ultrasound in the early follicular phase prior to their first cycle of assisted reproduction treatment. The relationship between the AFC made with each method and outcome was evaluated.
RESULTS: The AFC as measured by the ‘inversion mode’ method (r = 0.777) showed a higher, but statistically insignificant, correlation with the number of retrieved oocytes than did comparable measurements made with the ‘2D equivalent’ (r = 0.665) and ‘3D multiplanar’ (r = 0.687) techniques. Multiple regression analysis showed the AFC measured using any of the three methods was a significantly better predictor of the number of oocytes retrieved (P < 0.001) than age, FSH or ovarian volume. AFC was not predictive of non-conception regardless of the measurement technique.
CONCLUSIONS: For predicting ovarian respone and outcome, the AFC measured using techniques based on 3D ultrasound offers no statistically significant advantage over a measurement which is limited to information available with conventional 2D imaging.
Key words: in vitro fertilization/antral follicle count/3D ultrasound/inversion mode/ovarian reserve
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionAcknowledgementReferences |
|---|
Ovarian folliculogenesis is a lengthy developmental process which involves a prolonged period of cellular proliferation followed by rapid cellular differentiation in selected follicles that respond to stimulation by the pituitary gonadotrophins (Gougeon, 1996
). At any one time during the reproductive life of a woman, the paired ovaries will contain a heterogenous population of follicles at different developmental stages including a resting reserve of primordial follicles, pre-antral and early antral follicles (0.2–2.0 mm) which are largely gonadotrophin-independent, small antral follicles (1–5 mm) that are gonadotrophin-responsive or selectable, and a few larger antral follicles (>5 mm) that are gonadotrophin-dependent.
It is well established that the ovarian follicle population declines with increasing age. This is first evident as a woman enters her fourth decade when a reduction in the number of small antral follicles (1–5 mm) is seen (Scheffer et al., 1999). A more profound decline in the number of all follicle classes is seen in the early forties (Gougeon, 1998). Further, it is clear that a great deal of variation exists between individual women in the timing of age-related subfertility and infertility (te Velde and Pearson, 2002) which may be due in part to variation in the timing of these age-related changes in the ovarian follicle population.
Ovarian stimulation protocols in assisted reproduction treatment (ART) rely on the recruitment of a variable proportion of gonadotrophin-responsive follicles depending on the dose and duration of gonadotrophin treatment in relation to an individual woman's ovarian sensitivity to follicle-stimulating hormone (FSH) (Imani et al., 2002). Thus the estimation of the number of gonadotrophin-responsive or selectable follicles prior to treatment may assist in the design of more individualized controlled ovarian stimulation regimes. The estimation of this so-called ‘ovarian reserve’ can be made directly by counting the number of follicles measuring 2–10 mm (the antral follicle count, AFC) using ultrasound or indirectly through measurements of factors in the blood produced by these follicles (inhibin B and anti-Mullerian hormone; AMH) (Muttukrishna et al., 2005) or that are under the inhibitory endocrine control of the factors produced by small antral follicles (Bancsi et al., 2003).
Evaluation of ovarian reserve has become an essential part of the pretreatment assessment of a woman about to undergo ART (Hendriks et al., 2005). This has been most commonly achieved through recognition of the woman's age and quantification of her serum FSH level during the early follicular phase of the menstrual cycle (Templeton et al., 1996; Bancsi et al., 2003; Chuang et al., 2003). However, several studies have shown that assessment of the basal number of antral follicles, in itself assumed to be reflective of the primordial follicle population (Gougeon, 1996), is a better predictor of ovarian response as it correlates more closely with the number of oocytes retrieved (Chang et al., 1998b; Frattarelli et al., 2000) and the incidence of ovarian hyperstimulation syndrome (Ng et al., 2000). The superiority of AFCs over basal gonadotrophin measurements in the prediction of response to controlled ovarian stimulation during IVF treatment has been confirmed in a recent meta-analysis (Hendriks et al., 2005).
The number of antral follicles can be measured using either 2D or 3D transvaginal ultrasonography with an adequate inter- and intra-observer reliability (Scheffer et al., 2002). Antral follicles can also be demonstrated and counted by using a 3D rendering technique known as the ‘inversion mode’ (Raine-Fenning and Lam, 2006). The ‘inversion mode’ is a relatively new technique in which the hypoechoic aspect of the ultrasound display is inverted to demonstrate fluid-filled areas within the 3D dataset (Lee et al., 2005). This technique offers a new method for the demonstration and objective quantification of AFCs and has the potential to facilitate automated analysis of follicular number and size in the future.
A significant correlation between the number of antral follicles measuring between 2 and 10 mm in diameter, as assessed by both conventional 2D (Hsieh et al., 2001; Ng et al., 2005) and 3D (Kupesic et al., 2003; Ng et al., 2006) transvaginal ultrasound, and the number of oocytes retrieved has been reported. However, the degree of correlation and the predictive value of AFC and basal FSH have been different in each study. This variation may reflect the measurement techniques used as the number of antral follicles varies with different measurement techniques with higher counts recorded when measurements are made from 3D datasets as opposed to conventional real-time 2D ultrasound (Scheffer et al., 2002). Contrary to this, we have recently shown lower counts when the inversion mode is used to demonstrate the antral follicles within a 3D dataset (Jayaprakasan et al., 2007).
These findings indicate that the method by which the antral follicles are counted may influence its performance as a predictive test of ovarian reserve and response. To date there are no comparative studies to address this issue. The objective of this study, therefore, was to compare the predictive value of AFCs measured from 3D data by three different methods including a ‘2D equivalent technique’, the ‘3D multiplanar view’ and the ‘inversion mode’.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionAcknowledgementReferences |
|---|
Experimental design
We recruited 112 consecutive subjects in order to get an eventual study group of 100 women in total. All subjects were under the age of 40 years with regular menstrual cycles of 21–35 days duration and an early follicular phase FSH level of <12 IU/l and were undergoing their first cycle of ART. They underwent a baseline pretreatment ultrasound assessment in the early follicular phase (Day 2–4) of the spontaneous menstrual cycle before starting the down-regulation in the luteal phase of the same cycle. Subjects were excluded if they had a history of ovarian surgery or were found to have an ovarian cyst or follicle measuring
20 mm in diameter. We continued to recruit until we had 100 subjects. The study was approved by the hospital's local ethics committee and informed consent was obtained prior to the enrollment of each subject.
Data acquisition
All subjects had a transvaginal scan performed by a single investigator (K.J.) using a Voluson Expert 730TM (General Electric Medical Systems, Zipf, Austria) and a 7.5 MHz transvaginal probe. Our technique for the acquisition of 3D data has been described in detail (Raine-Fenning et al., 2003) but briefly this included a initial 2D ultrasound assessment of the pelvis to exclude any obvious pathology before the application of a region of interest over the ovary which defined the volume to be acquired. An automated mechanical sweep of this region through 90° was then undertaken using the slow sweep mode and the resultant multiplanar display examined to ensure that the entire ovary had been captured. A single acquisition was obtained for each ovary. All subjects were scanned with their legs supported by stirrups in a modified Lloyd Davies position to limit discomfort and ensure free manipulation of the transvaginal transducer. The data were saved to the hard drive of the ultrasound machine and subsequently transferred to a personal computer via a digital video disk without any data compression.
Data measurement
All measurements were made on a personal computer using 4D view (version 5.0; General Electric Medical Systems, Zipf, Austria). The 3D dataset was initially displayed in the multiplanar view and one of three methods was used to measure the number of antral follicles, defined as those follicles measuring 2–10 mm in diameter (Swanton and Child, 2005), in each ovary. The multiplanar display is specific to 3D ultrasound and simultaneously displays the longitudinal, transverse and coronal sections of the object of interest in an orthogonal manner such that each image is at 90° the other two images (Fig. 1). The longitudinal (image plane A) and transverse (image plane B) images are obtainable with conventional 2D ultrasound but not visible simultaneously, while the coronal image (image plane C) is specific to 3D imaging, whether derived from magnetic resonance, computerized tomographic or ultrasonographic data, and depicts the image plane at 90° to the transducer and ultrasound beam.
|
In technique 1, the ‘2D equivalent’ method, the user was only able to see one image plane at a time and was restricted to making measurements in the longitudinal and transverse planes only. The user could scroll through these planes in real-time as they would in the clinical setting but they were not allowed to rotate them to produce a coronal plane. The 2D equivalent technique is so-called as the views used for measurement are comparable to those obtainable with conventional 2D ultrasound. In technique 2, the ‘3D multiplanar view’ method, AFCs were made using all three perpendicular planes, which could be rotated and viewed simultaneously to enhance the observer's spatial awareness. In technique 3, ‘the 3D rendered inversion mode’, the antral follicles were measured in the rendered view after the ovary had been defined using virtual organ computer-aided analysis (VOCAL®; GE Kretz, Zipf, Austria) and the inversion mode applied (Fig. 2). VOCAL allows the user to trace the ovarian cortex and generate a 3D model of the ovary (Raine-Fenning et al., 2003
). This model was then processed using the ‘inversion mode’ to display the follicles within the ovary and the threshold, which assigns transparency associated with fluid to opaque voxels, adjusted until the follicles were optimally visualized. The antral follicles were then counted by rotating the resultant display through 180° using an automatically calculated cine loop whose speed can be controlled by the user. The main advantage of VOCAL in this context is that it allows the user to selectively display the information within the ovary and exclude all extra-ovarian tissue, which can reduce the image quality and make accurate measurement more difficult (Fig. 3).
|
|
The antral follicles were counted by two investigators (K.J. and N.H.) with each dataset measured once by both using all three measurement methods and the mean value was used for analysis. The order of measurements and the method used for each dataset was randomly determined and different. One dataset, a single ovary, was opened and the antral follicles were counted using one of the three methods. The dataset was then closed and a new dataset opened. Further measurements of antral follicles were then performed using another method, which was randomly assigned. The three methods were not applied serially in any subject, therefore, to ensure that the user could not become familiar with the dataset. The mean intra-class correlation coefficient and 95% confidence interval (CI) for measurement of the number of antral follicles using the 2D equivalent mode, the 3D multiplanar mode and the 3D rendered inversion mode were 0.962 (0.927–0.982), 0.983 (0.968–0.992) and 0.989 (0.979–0.995), respectively, between the observers indicative of good inter-observer reliability for each method. The mean intra-class correlation coefficient for ovarian volume measurement using VOCAL programme was 0.989 (0.960–0.997) between observers.
Treatment protocol
All subjects underwent IVF treatment using a standard long protocol. This involved down-regulation with gonadotrophin-releasing hormone (GnRH) agonists (500 mcg/day of Buserelin; Suprefact®, Aventis Pharma, Kent, UK or 800 mcg/day of Nafarelin; Synarel®, Pharmacia, Milton Keynes, UK) started in the mid-luteal phase of the menstrual cycle 7 days prior to the expected date of menstruation. Two weeks later, following confirmation of ovarian suppression defined as an endometrial thickness of <5 mm and no ovarian activity on ultrasound scan in association with an estradiol level below 200 pmol/l, ovarian stimulation was commenced using either recombinant FSH (Gonal-F; Serono Pharmaceuticals Ltd, Feltham, UK) or purified urinary human menopausal gonadotrophin (hMG: Menopur, Ferring Pharmaceuticals, Berks, UK). The starting dose used for stimulation was based on the subject's age (150 IU for women under 30 years of age, 225 IU for women aged between 30 and 38 years and 300 IU for women aged 38 years or more) and adjusted according to the ovarian response which was monitored daily by serial transvaginal ultrasound and serum estradiol measurements from the 6 day of stimulation. Human chorionic gonadotrophin (hCG; 6500 IU of Ovitrelle; Serono Pharmaceuticals Ltd, Feltham, UK or 10 000 IU of Pregnyl; Organon Laboratories Ltd, Cambs, UK) was administered when there were at least three follicles measuring 18 mm or more in diameter, and oocyte retrieval was performed 36 h later. ‘Subjects that did not develop at least three follicles’ measuring 18 mm or more in diameter were cancelled or converted to intrauterine insemination treatment dependent on other clinical factors including tubal patency and seminal fluid analysis. A maximum of two normally cleaved embryos were transferred into the uterus 2 days after oocyte retrieval and the level of serum hCG was measured 16 days later to determine the outcome. If the test was positive (hCG > 50 IU/l), a transvaginal ultrasound was arranged 2 weeks later to confirm the viability of the pregnancy.
Statistical analysis
Statistical Package for the Social Sciences (SPSS version 14.0) was used for statistical analysis. The primary outcome measure was the number of oocytes obtained. The correlation between AFCs made with each different method and the number of follicles measuring 10 mm or more in diameter on the day of hCG and the total number of oocytes obtained at oocyte retrieval was calculated using Pearson correlation coefficient. The difference between pairs of correlation coefficients was assessed using Fisher's z transformation with significance determined using the t-statistic. The distribution of the data was checked for normality using a normal probability plot. The variables were compared between the right and left ovaries and between the pregnant and non-pregnant groups using paired t-test and unpaired t-test, respectively. A P-value of <0.05 was considered statistically significant. Multiple linear regression analysis with least-squares regression was applied to evaluate the predictive values of age, basal FSH, ovarian volume and AFC measured using different measurement methods on the number of oocytes retrieved. Multiple logistic regression analysis was used to assess the effect of different variables on prediction of non-pregnancy. Receiver operating characteristic (ROC) curve analysis was performed to quantify the ability of AFC measured by the three methods to discriminate between poor responders and normal responders and also between pregnant and non-pregnant subjects. Sensitivity, specificity, positive likelihood ratio for an optimal cut-off value and area under the ROC curve (AUCROC) were calculated for each model. Areas under the ROC curves were compared using the MEDCALC 9.2.0 software package (Hanley and McNeil, 1983).
The correlation coefficient between AFC measured using ‘2D equivalent method’ in a previous unpublished observation was 0.6. Assuming the methods specific to 3D ultrasound would improve the correlation coefficient from 0.6 to 0.75, the sample size required to give a test of significance of 0.05 between the two correlation coefficients using Fisher's z transformation was 100.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionAcknowledgementReferences |
|---|
A total of 112 subjects were recruited. Four of these subjects were excluded as they had had ovarian surgery in the past. Remaining 108 subjects underwent a pretreatment transvaginal ultrasound and eight of these had ovarian follicles or cysts measuring >20 mm in diameter. The final study group of 100 subjects was therefore obtained as planned. These subjects had a variety of causative factors for their subfertility including tubal disease (20 subjects), endometriosis (nine subjects), male factor (41 subjects), ovulatory dysfunction (four subjects), combined factors (seven subjects) and unexplained subfertility (19 subjects). The starting dose of gonadotrophin was 150 IU in 16 (16%) subjects, 225 IU in 57 (57%) subjects and 300 IU in 27 (27%) subjects. Subsequent increments or decrements in FSH were needed in nine (9%) and five (5%) subjects, respectively, based on clinical interpretation of their ovarian response. Three subjects (3%) were cancelled for inadequate ovarian response and three subjects experienced failed fertilization (3%). Overall, five (5%) subjects experienced poor ovarian response, defined by either cycle cancellation or the retrieval of less than four oocytes (Bancsi et al., 2002
). One subject developed severe ovarian hyperstimulation syndrome despite having oocyte retrieval and the cycle was subsequently cancelled to avoid exacerbation of the disease through the administration of luteal support or in the event of pregnancy. Embryo transfer was performed in 93 subjects, therefore, and resulted in 51 pregnancies (51% per cycles initiated and 54.8% per embryo transfer). Six of them were biochemical pregnancies (11.8%) as there was no evidence of a pregnancy on scan 2 weeks later at 6 weeks gestation. The clinical pregnancy rate was 45% per cycles initiated and 48.4% per embryo transfer therefore. Two pregnancies miscarried during the next 6 weeks resulting in an ongoing pregnancy rate (viable pregnancy seen on a 12 weeks scan) of 43% per cycles initiated (43/100) and 46.2% per embryo transfer (43/93). The median age of the subjects was 34 (range: 23–40) years and the median early follicular phase FSH level was 7.2 (range: 3.4–11.9) IU/l. The total AFC measured using inversion mode (median: 13; range: 2–35) was significantly lower (P < 0.05) when compared with the ‘2D equivalent’ method (median: 17; range: 4–34) and the ‘3D multiplanar’ method (median: 16.5; range: 5–37). The median number of oocytes retrieved was 10 (range: 6–35) and was similar (P > 0.05) between the right and left ovaries (5; range: 0–15 versus 4; range: 0–20, respectively). There were also no significant differences between the right and left ovaries in terms of their volume (median: 7.51 cm3; range: 2.4–16.29 versus median: 6.92 cm3; range: 0.71–17.71) or number of antral follicles as measured by the ‘2D equivalent’ method (median: 8; range: 1–17 versus median: 7; range: 1–20), the ‘3D multiplanar’ method (median: 8; range: 2–18 versus median: 7; range: 1–20), or the ‘3D rendered inversion mode’ method (median: 6.5; range: 188–15 versus median: 6; range: 1–20). The median time taken for each antral follicle measurement per subject was 73 (range: 43–118), 79 (range: 49–122) and 382 s (range: 176–457) using the ‘2D equivalent’, ‘3D multiplanar’ and ‘3D rendered inversion mode’ methods, respectively (P < 0.05).
There was a closer relationship between the number of oocytes retrieved and the number of total pretreatment antral follicles measured with ‘3D rendered inversion mode’ method (r = 0.777) than when AFCs were made with the ‘2D equivalent’ (r = 0.665) and ‘3D multiplanar’ (r = 0.687) methods when each subject was considered for analysis and the values for each ovary combined, but this was not statistically significant. The correlation between the pretreatment AFC and the number of follicles measuring 10 mm or more on the day of hCG injection, prior to oocyte retrieval, was also higher, but not statistically significant, when the follicle count was measured with the ‘3D rendered inversion mode’ (r = 0.623) as opposed to the ‘2D equivalent’ (r = 0.563) and the ‘3D multiplanar’ (r = 0.582) methods (Table 1).
|
Table 2 shows the results of the multiple linear regression analysis of age, basal FSH, mean ovarian volume and total AFC in predicting the number of oocytes retrieved. Neither age nor ovarian volume was predictive of ovarian response in terms of the number of oocytes retrieved in any of the three models. Total AFC measured with any of the three methods was the best predictor of the number of retrieved oocytes (P < 0.001) followed by basal FSH. Basal FSH levels were predictive (P < 0.05) in the models where the AFC was measured by the ‘2D equivalent’ and ‘3D multiplanar view’ methods, but not when these were made with the ‘3D rendered inversion mode’ method (P = 0.168). The best model in predicting the number of oocytes retrieved was the one in which the antral follicles were measured with the ‘3D rendered inversion mode’ method (100r2 = 61.8%) in comparison to the ‘2D equivalent’ (100r2 = 47.3%) and ‘3D multiplanar view’ (100r2 = 50.2%) methods. This may indicate the superiority of the ‘3D rendered inversion mode’ method over the other two methods in the prediction of how many oocytes may be retrieved. The AUC, the sensitivities, the specificities and the positive likelihood ratios for AFCs measured using three different methods in predicting poor ovarian response in women undergoing IVF is shown in Table 3. The AFC made with the ‘3D rendered inversion mode’ method appears to offer the best discriminative potential for poor response, expressed by the largest AUC of 0.987, although this was not statistically different from the corresponding AUC for the ‘2D equivalent’ (0.968) and ‘3D multiplanar view’ methods (0.968).
|
|
There was no significant difference in age, basal FSH and mean ovarian volume between subjects that subsequently became pregnant and those that did not. However, the total AFC measured using any of the three methods was significantly lower in the non-pregnant group compared to the pregnant group (Table 4). Multiple logistic regression analysis showed that all of the variables tested, including the AFC measured by any of the three methods, were poor predictors of non-conception (Table 5). The AUC for AFC in predicting non-conception in women undergoing IVF was low for all three methods (Table 6). However, the AUC for follicle counts measured by the ‘3D rendered inversion mode’ (0.641) was higher than the corresponding AUC for the ‘2D equivalent’ (0.621) and ‘3D multiplanar view’ (0.630) methods, although the difference was not statistically significant. Figures 4–6 show the total AFC, calculated by combining the follicle number of both ovaries, measured with the three different methods and the treatment outcome. These figures illustrate that the pregnancies were scattered over a wide range of follicle counts but that the pregnancy rate was lower when the total AFC was less than eight.
|
|
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionAcknowledgementReferences |
|---|
This is the first study to compare the predictive value of AFC measurement made using both 2D and 3D ultrasound in determining the outcome of response to ovarian stimulation as measured by the number of follicles that develop, the number of oocytes retrieved and the pregnancy rate following ART. It is also the first study to evaluate the relationship between AFCs made from stored 3D data using the inversion mode and outcome. The data in this study indicate that a pretreatment AFC measured using methods specific to 3D ultrasound offers minimal additional information from that derived from conventional 2D ultrasound in the prediction of the number of follicles measuring 10 mm or more that will be evident on the day of hCG, the actual number of oocytes that will be retrieved thereafter, and the incidence of non-conception. Furthermore, measurements made with the inversion mode take significantly longer than those made with the 2D equivalent and 3D multiplanar view techniques. While the total AFC measured using any of the three methods is a better predictor of ovarian response than age, basal FSH levels and mean ovarian volume, none of these variables are predictive of non-conception.
AFCs measured using the ‘3D rendered inversion mode’ were significantly lower (P < 0.05), when compared with the ‘2D equivalent’ and ‘3D multiplanar view’ methods. This may reflect an overestimation by the ‘2D equivalent’ and ‘3D multiplanar view’ methods or underestimation by the ‘3D rendered inversion mode’ method. The better statistical correlation with the number of oocytes retrieved and total AFC by the inversion mode may be due to such an underestimation of the absolute number. Application of the ‘inversion mode’ to an ovarian volume defined using VOCAL allows an immediate objective demonstration of the antral follicles within an ovary with a high inter-observer reliability (Jayaprakasan et al., 2007). Inversion rendering of the fluid-filled areas from hypoechoic to echogenic images improves binocular vision allowing a better perception of the image (Timor-Tritsch et al., 2005). It also provides a simultaneous and complete view of all of the follicles within the volume rather than in a single plane at one time and thereby attenuates the difficulties that arise with conventional 2D and 3D image planes, which require the observer to scroll through the images. It also facilitates the identification of individual follicles which can be difficult when there are increased numbers. The use of the 3D cine function to evaluate the ‘inverted’ follicles from different angles by rotating the images through 180° at a desired speed controllable by the user further enhances the accuracy of AFC measurement.
Contrary to the work of Scheffer et al. (2002), who noted a higher number of antral follicles when counts were made with 3D ultrasound rather than 2D ultrasound, this study has shown a comparable performance between the ‘2D equivalent’ and the ‘3D multiplanar view’ measurement methods in terms of the mean AFC. Scheffer et al. used different ultrasound machines for the 2D and 3D ultrasound examination which may have contributed to the variation in the follicle count if there were differences in the image quality which has been shown to be an important parameter in determining the AFC (Jayaprakasan et al., 2007). In the present study, we overcame this problem by using ultrasound images generated from the same machine using the same probe.
One other group has examined the predictive value of AFCs made with 2D and 3D ultrasound but as two separate studies. In their initial study, which used 2D ultrasound, they showed basal FSH was a better predictor of the total number of oocytes retrieved than the AFC (Ng et al., 2005). In a subsequent study, of similar design but based on 3D ultrasound, the same group found AFCs made using the multiplanar view were more predictive than basal FSH levels (Ng et al., 2006). While their results suggest that 3D is superior to 2D therefore, albeit in different subject groups, they only compared AFC and FSH levels and the comparison between the two measurement techniques was not the objective of either study. Our study was specifically designed to investigate the technique used for counting antral follicles, was appropriately blinded, and has used multiple regression analysis to compare the significance of several factors known to be predictive of the number of oocytes retrieved. Our results demonstrate that the AFC, irrespective of the measurement method used, has the highest predictive value for determining the total number of oocytes retrieved at oocyte collection followed by basal FSH levels. However, the basal FSH level was not a significant predictor when compared with AFCs measured using the ‘3D rendered inversion mode’ method in contrast to measurements made with the ‘2D equivalent’ and ‘3D multiplanar view’ methods. Through the application of VOCAL, which was used to define the ovarian cortex and eliminate extra-ovarian tissue prior to assessment of the number of antral follicles with the inversion mode, we were also able to calculate ovarian volume which has been recommended as an integral part of the pretreatment assessment of women undergoing ART (Lass et al., 1997). However, ovarian volume was less predictive of number of oocytes retrieved than the AFC and basal FSH level in our subjects.
The majority of subjects in our study had fewer number of oocytes obtained than the total AFC (Fig. 4–6). The number of oocytes obtained was higher than the AFC in only four (4%), two (2%) and five (5%) subjects when measurements were made with the ‘2D equivalent’, ‘3D multiplanar view’ and ‘3D rendered inversion mode’ methods, respectively. This may reflect either an incorrect measurement of the total number of antral follicles or a variation in the cohort of follicles available for recruitment between cycles or after down-regulation. While biological variation in the number of follicles between cycles has been demonstrated (Hansen et al., 2003), the effects of down-regulation on the follicle count are conflicting (Jarvela et al., 2003; Yu Ng et al., 2004). The degree of recruitment and subsequent follicular development also depend on the dose of FSH used for ovarian stimulation, the subject's ovarian sensitivity to FSH and their individual ‘FSH threshold’ (Imani et al., 2002). Based on the linear regression analysis, the predicted number of oocytes retrieved was 50, 53 and 72% of the total AFC made with the ‘2D equivalent’, ‘3D multiplanar view’ and ‘3D rendered inversion mode’ methods, respectively. This observation has potential clinical implications and could be used to identify women at risk of a poor or exaggerated response to ovarian stimulation and possibly to ascertain the suitability of women who wish to act as oocyte donors. We are currently examining these factors in a series of studies designed to investigate factors that may affect the AFC and are undertaking a randomized prospective study based on a fixed dose of ovarian stimulation.
AFCs made with any of the three methods do not appear to be predictive of non-conception. Pregnancies were seen only in those subjects who had at least four oocytes retrieved. This would equate to a total AFC of between six and eight at best below which is indicative of an impaired ovarian reserve. ROC curve analysis indicated that the best antral follicle cut-off value to discriminate between conception and non-conception cycles was eight and at this level the specificity of follicle counts measured by any of the methods to predict non-conception is high but limited by a relatively low sensitivity. However, for a specificity of 96.1%, the sensitivity and positive likelihood ratio were higher with the ‘3D rendered inversion mode’ method than the ‘2D equivalent’ and ‘3D multiplanar view’ methods. Logistic regression analysis showed all the predictive variables tested in the study including the AFCs measured using any of the three methods were not predictive of non-conception following ART. Because of the limitation in predicting non-conception, ovarian reserve assessment prior to IVF is controversial (Broekmans et al., 2006). However, women who are predicted to have normal ovarian response can have acceptable chance of conception through IVF even at an advanced reproductive age (Klinkert et al., 2005). AFCs measured with any of the methods used in this study were significantly higher in subjects that became pregnant. Women with higher AFCs are more likely to have higher numbers of oocytes retrieved and therefore more embryos cleave which may allow the embryologist a wider choice in their selection of embryos for transfer (Chang et al., 1998a).
Follicular recruitment is dependent on the starting dose of FSH used for ovarian stimulation and, but to a lesser degree, the total dose of FSH which is determined by the duration of stimulation and the need for any variation in the daily dose. This study included subjects with different starting doses of FSH and did not account for subsequent increments or reduction in the daily regimen. However, while the gonadotrophin dose may have influenced the recruitment and development of follicles and thus the number of oocytes retrieved, a recent randomized trial failed to demonstrate any significant difference in the number of retrieved oocytes between two different dose regimens (150 IU versus 300 IU) among predicted poor responders based on AFC (Khalaf et al., 2002). Many ART units have variable starting doses and modify the daily FSH dose based on the ultrasonographic or endocrine response to stimulation and so this study is therefore reflective of true clinical practice. It is only possible to overcome this limitation by performing a study using a standard fixed dose of gonadotrophin for ovarian stimulation, which is not altered through the treatment cycle.
|
Conclusion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionAcknowledgementReferences |
|---|
Stored 3D data can be used for the assessment of ovarian reserve through the calculation of ovarian volume and assessment of the total number of antral follicles. AFCs measured using techniques based on 3D imaging appear to offer a small advantage over methods limited to information available with conventional 2D imaging in predicting ovarian response in an ART programme. AFCs are a significantly better predictor of the total number of oocytes retrieved at oocyte collection than age, basal FSH and ovarian volume. The highest correlation of AFC with the number of follicles that develop during ovarian stimulation and the total number of oocytes retrieved is seen when the AFC is made with the ‘inversion mode’, which is based on 3D rendering, but this rendering technique takes significantly longer than all other methods. AFCs made with any of the three techniques appear to be a good predictor of poor ovarian response, with a high sensitivity and specificity, at an optimal cut-off value of six or seven follicles. Regardless of measurement technique, AFCs do not appear to predict non-conception.
|
|
Acknowledgement |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionAcknowledgementReferences |
|---|
The authors would like to thank Sarah Armstrong, Trent Research and Development Support Unit, the University of Nottingham for her assistance with the statistical analysis.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionAcknowledgementReferences |
|---|
Bancsi LF, Broekmans FJ, Eijkemans MJ, de Jong FH, Habbema JD, te Velde ER. Predictors of poor ovarian response in in vitro fertilization: a prospective study comparing basal markers of ovarian reserve. Fertil Steril (2002) 77:328–336.[CrossRef][Web of Science][Medline]
Bancsi LF, Broekmans FJ, Mol BW, Habbema JD, te Velde ER. Performance of basal follicle-stimulating hormone in the prediction of poor ovarian response and failure to become pregnant after in vitro fertilization: a meta-analysis. Fertil Steril (2003) 79:1091–1100.[CrossRef][Web of Science][Medline]
Broekmans FJ, Kwee J, Hendriks DJ, Mol BW, Lambalk CB. A systematic review of tests predicting ovarian reserve and IVF outcome. Hum Reprod Update (2006) 12:685–718.
Chang MY, Chiang CH, Chiu TH, Hsieh TT, Soong YK. The antral follicle count predicts the outcome of pregnancy in a controlled ovarian hyperstimulation/intrauterine insemination program. J Assist Reprod Genet (1998a) 15:12–17.[CrossRef][Web of Science][Medline]
Chang MY, Chiang CH, Hsieh TT, Soong YK, Hsu KH. Use of the antral follicle count to predict the outcome of assisted reproductive technologies. Fertil Steril (1998b) 69:505–510.[CrossRef][Web of Science][Medline]
Chuang CC, Chen CD, Chao KH, Chen SU, Ho HN, Yang YS. Age is a better predictor of pregnancy potential than basal follicle-stimulating hormone levels in women undergoing in vitro fertilization. Fertil Steril (2003) 79:63–68.[CrossRef][Web of Science][Medline]
Frattarelli JL, Lauria-Costab DF, Miller BT, Bergh PA, Scott RT. Basal antral follicle number and mean ovarian diameter predict cycle cancellation and ovarian responsiveness in assisted reproductive technology cycles. Fertil Steril (2000) 74:512–517.[CrossRef][Web of Science][Medline]
Gougeon A. Regulation of ovarian follicular development in primates: facts and hypotheses. Endocr Rev (1996) 17:121–155.
Gougeon A. Ovarian follicular growth in humans: ovarian ageing and population of growing follicles. Maturitas (1998) 30:137–142.[CrossRef][Web of Science][Medline]
Hanley JA, McNeil BJ. A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology (1983) 148:839–843.
Hansen KR, Morris JL, Thyer AC, Soules MR. Reproductive aging and variability in the ovarian antral follicle count: application in the clinical setting. Fertil Steril (2003) 80:577–583.[CrossRef][Web of Science][Medline]
Hendriks DJ, Mol BW, Bancsi LF, Te Velde ER, Broekmans FJ. Antral follicle count in the prediction of poor ovarian response and pregnancy after in vitro fertilization: a meta-analysis and comparison with basal follicle-stimulating hormone level. Fertil Steril (2005) 83:291–301.[CrossRef][Web of Science][Medline]
Hsieh YY, Chang CC, Tsai HD. Antral follicle counting in predicting the retrieved oocyte number after ovarian hyperstimulation. J Assist Reprod Genet (2001) 18:320–324.[CrossRef][Web of Science][Medline]
Imani B, Eijkemans MJ, Faessen GH, Bouchard P, Giudice LC, Fauser BC. Prediction of the individual follicle-stimulating hormone threshold for gonadotropin induction of ovulation in normogonadotropic anovulatory infertility: an approach to increase safety and efficiency. Fertil Steril (2002) 77:83–90.[Web of Science][Medline]
Jarvela IY, Sladkevicius P, Kelly S, Ojha K, Campbell S, Nargund G. Effect of pituitary down-regulation on the ovary before in vitro fertilization as measured using three-dimensional power Doppler ultrasound. Fertil Steril (2003) 79:1129–1135.[CrossRef][Web of Science][Medline]
Jayaprakasan K, Walker KF, Clewes JS, Johnson IR, Raine-Fenning NJ. The interobserver reliability of off-line antral follicle counts made from stored three-dimensional ultrasound data: a comparative study of different measurement techniques. Ultrasound Obstet Gynecol (2007) 29:335–341.[CrossRef][Medline]
Khalaf Y, El-Toukhy T, Taylor A, Braude P. Increasing the gonadotrophin dose in the course of an in vitro fertilization cycle does not rectify an initial poor response. Eur J Obstet Gynecol Reprod Biol (2002) 103:146–149.[CrossRef][Web of Science][Medline]
Klinkert ER, Broekmans FJ, Looman CW, Habbema JD, te Velde ER. The antral follicle count is a better marker than basal follicle-stimulating hormone for the selection of older patients with acceptable pregnancy prospects after in vitro fertilization. Fertil Steril (2005) 83:811–814.[CrossRef][Web of Science][Medline]
Kupesic S, Kurjak A, Bjelos D, Vujisic S. Three-dimensional ultrasonographic ovarian measurements and in vitro fertilization outcome are related to age. Fertil Steril (2003) 79:190–197.[CrossRef][Web of Science][Medline]
Lass A, Skull J, McVeigh E, Margara R, Winston RM. Measurement of ovarian volume by transvaginal sonography before ovulation induction with human menopausal gonadotrophin for in-vitro fertilization can predict poor response. Hum Reprod (1997) 12:294–297.
Lee W, Goncalves LF, Espinoza J, Romero R. Inversion mode: a new volume analysis tool for 3-dimensional ultrasonography. J Ultrasound Med (2005) 24:201–207.
Muttukrishna S, McGarrigle H, Wakim R, Khadum I, Ranieri DM, Serhal P. Antral follicle count, anti-mullerian hormone and inhibin B: predictors of ovarian response in assisted reproductive technology? Bjog (2005) 112:1384–1390.[Web of Science][Medline]
Ng EH, Tang OS, Ho PC. The significance of the number of antral follicles prior to stimulation in predicting ovarian responses in an IVF programme. Hum Reprod (2000) 15:1937–1942.
Ng EH, Tang OS, Chan CC, Ho PC. Ovarian stromal blood flow in the prediction of ovarian response during in vitro fertilization treatment. Hum Reprod (2005) 20:3147–3151.
Ng EH, Tang OS, Chan CC, Ho PC. Ovarian stromal vascularity is not predictive of ovarian response and pregnancy. Reprod Biomed Online (2006) 12:43–49.[Web of Science][Medline]
Raine-Fenning NJ, Campbell BK, Clewes JS, Johnson IR. The interobserver reliability of ovarian volume measurement is improved with three-dimensional ultrasound, but dependent upon technique. Ultrasound Med Biol (2003) 29:1685–1690.[CrossRef][Web of Science][Medline]
Raine-Fenning NJ, Lam PM. Assessment of ovarian reserve using the inversion mode. Ultrasound Obstet Gynecol (2006) 27:104–106.[CrossRef][Web of Science][Medline]
Scheffer GJ, Broekmans FJ, Dorland M, Habbema JD, Looman CW, te Velde ER. Antral follicle counts by transvaginal ultrasonography are related to age in women with proven natural fertility. Fertil Steril (1999) 72:845–851.[CrossRef][Web of Science][Medline]
Scheffer GJ, Broekmans FJ, Bancsi LF, Habbema JD, Looman CW, Te Velde ER. Quantitative transvaginal two- and three-dimensional sonography of the ovaries: reproducibility of antral follicle counts. Ultrasound Obstet Gynecol (2002) 20:270–275.[CrossRef][Web of Science][Medline]
Swanton A, Child T. Reproduction and ovarian ageing. J Br Menopause Soc (2005) 11:126–131.
te Velde ER, Pearson PL. The variability of female reproductive ageing. Hum Reprod Update (2002) 8:141–154.
Templeton A, Morris JK, Parslow W. Factors that affect outcome of in-vitro fertilisation treatment. Lancet (1996) 348:1402–1406.[CrossRef][Web of Science][Medline]
Timor-Tritsch IE, Monteagudo A, Tsymbal T, Strok I. Three-dimensional inversion rendering: a new sonographic technique and its use in gynecology. J Ultrasound Med (2005) 24:681–688.
Yu Ng EH, Chi Wai Chan C, Tang OS, Shu Biu Yeung W, Chung Ho P. Effect of pituitary downregulation on antral follicle count, ovarian volume and stromal blood flow measured by three-dimensional ultrasound with power Doppler prior to ovarian stimulation. Hum Reprod (2004) 19:2811–2815.
Submitted on December 16, 2006; resubmitted on March 20, 2007; accepted on March 29, 2007.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  I. Y. Jarvela, A. Ruokonen, and A. Tekay Effect of rising hCG levels on the human corpus luteum during early pregnancy Hum. Reprod., December 1, 2008; 23(12): 2775 - 2781. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Jayaprakasan, B.K. Campbell, J.F. Hopkisson, J.S. Clewes, I.R. Johnson, and N.J. Raine-Fenning Effect of pituitary desensitization on the early growing follicular cohort estimated using anti-Mullerian hormone Hum. Reprod., November 1, 2008; 23(11): 2577 - 2583. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Jayaprakasan, J.F. Hopkisson, B.K. Campbell, J. Clewes, I.R. Johnson, and N.J. Raine-Fenning Quantification of the effect of pituitary down-regulation on 3D ultrasound predictors of ovarian response Hum. Reprod., July 1, 2008; 23(7): 1538 - 1544. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||