Correspondence: Tel: ; Fax: ;

Transcription

1 RBMOnline - Vol 16. No Reproductive BioMedicine Online; on web 21 January 2008 Article Optimal timing of ultrasonographic and Doppler evaluation of uterine receptivity to implantation Herve Dechaud received his MD degree in 1996 (Obstetrics and Gynaecology). He completed his PhD at Montpellier I University (France) (tubal infertility and impairment of embryo implantation). He then attended the University of Texas, Health Science Center at San Antonio, Texas, USA as research fellow in His research interests include endometrium and embryo implantation, infertility related to endometriosis, and ovarian stimulation protocols for assisted reproduction cycles. At the present time, he is the Medical Director of the Department of Reproductive Medicine and Surgery at the University Hospital in Montpellier, France. Dr Herve Dechaud Herve Dechaud 1,3, Emmanuelle Bessueille 1, Philippe-Jean Bousquet 2, Lionel Reyftmann 1, Samir Hamamah 1, Bernard Hedon 1 1 Department of Reproductive Medicine and Biology; 2 Department of Medical Information and Biostatistics, Arnaud de Villeneuve Hospital, 371 avenue du doyen Giraud, Montpellier Cedex 5, France 3 Correspondence: Tel: ; Fax: ; h-dechaud@chu-montpellier.fr Abstract In IVF programmes, transvaginal ultrasonography is used as a non-invasive method to evaluate uterine receptivity. The aim of this study was to determine when to perform this investigation in order to optimize prediction of the likelihood of pregnancy. Over 9 months, 124 patients undergoing IVF or intracytoplasmic sperm injection were studied. The ultrasonographic evaluation included endometrial thickness, endometrial pattern, uterine artery pulsatility index, protodiastolic notch, end-diastolic blood flow, and endometrial subendometrial blood flow distribution pattern. All patients underwent ultrasonographic investigation on the days of human chorionic gonadotrophin (HCG) administration, oocyte retrieval, and embryo transfer. Statistical analysis was done using recursive-partitioning analysis. The pregnancy and implantation rates per transfer were 33 and 19.8% respectively. In terms of single parameters, women with an enddiastolic blood flow, an endometrial subendometrial blood flow and a multilayered endometrium were more likely to be pregnant than women without one or more of these signs. The most effective combination for evaluation of uterine receptivity was end-diastolic blood flow, endometrial pattern and endometrial thickness. Sensitivity and specificity of this combination were around 81%, positive predictive value was 68.2%, and negative predictive value 89.7%. The best sensitivity and specificity were obtained on the day of HCG administration: respectively 81.1 and 81.3%. Keywords: Doppler, embryo implantation, endometrium, IVF, ultrasound Introduction 368 Despite constant progress in IVF techniques and ovarian stimulation, pregnancy and embryo implantation rates remain modest (Andersen et al., 2005). The phenomenon of embryo implantation depends both on the embryo itself and on endometrial receptivity (Schwartz et al., 1997). A favourable uterine medium is necessary for embryo implantation, but endometrial receptivity can be negatively affected by ovarian stimulation (Yaron et al., 1994). Study of endometrial receptivity is especially limited in the context of IVF cycles (Friedler et al., 1996; Carbillon et al., 2001). Biopsy of the endometrium cannot be considered during the treatment, and this has encouraged the development and use of ultrasonography and Doppler-like non-invasive methods to evaluate ovarian response to the stimulation and uterine receptivity (Dickey, 1997; Fanchin, 2001; Ardaens et al., 2002; Ebrard-Charra et al., 2005). However, such indirect exploration is still a subject of debate, both in terms of methodology and its results (Kupesic et al., 2001; Puerto et al., 2003) Published by Reproductive Healthcare Ltd, Duck End Farm, Dry Drayton, Cambridge CB3 8DB, UK

2 Several studies assessing endometrial thickness and echogenicity during treatment cycles and how they are related to pregnancy have yielded conflicting findings (De Geyter et al., 2000; Bassil, 2001; Dietterich et al., 2002; Kovacs et al., 2003). The thresholds and reproducibility achieved with these techniques are hard to define. Variation may be due to the limited power of certain studies, patient characteristics or the multiplicity of ovarian stimulation protocols. Furthermore, numerous other studies on uterine blood flow have led to conflicting conclusions (Yang et al., 1999; Schild et al., 2001; Basir et al., 2002a,b; Chien et al., 2004). Despite of the absence of consensus, in the majority of IVF programmes transvaginal ultrasonography is used as a noninvasive method to evaluate uterine receptivity (Aytoz et al., 1997; Ozturk et al., 2004; Killick, 2007). The optimal timing of ultrasonography is unclear, and various moments are used: during the previous cycle, on the day of human chorionic gonadotrophin (HCG) administration, on the day of embryo transfer (Salle et al., 1998; Schild et al., 2001; Basir et al., 2002; Dietterich et al., 2002; Puerto et al. 2003; Chien et al. 2004). The results are therefore not comparable and differ between reports. The aim of this study was to determine prospectively when to perform uterine exploration by ultrasonography and Doppler in an IVF programme, in order to optimize prediction of the likelihood of pregnancy. Materials and methods Patient inclusion criteria Over 9 months, 124 patients undergoing IVF or intracytoplasmic sperm injection (ICSI) because of male factors, tubal factors, unexplained infertility, endometriosis factors or preimplantation genetic diagnosis were found to be eligible for this study, which was approved by an internal ethics committee. Twelve patients had no embryo transfer after IVF or ICSI because of absence of oocytes (n = 1), fertilization (n = 6), or early-stage embryo development (n = 5) and were excluded from the analysis. None of the 112 remaining patients had occult ovarian failure on the basis of their day 3 FSH concentrations of <12 IU/l measured in a pretreatment cycle. Women with oligo/amenorrhoea or a history of uterine surgery, who had apparent endometrial disease (submucous myoma, synechia, polyps, uterus septum) or hydrosalpinges detected by ultrasonography, who had had three or more failed attempts at IVF and embryo transfer, or who received frozen thawed embryos or donated oocytes, were not included. All patients received no treatment other than ovarian stimulation. Ovarian stimulation and IVF All patients were stimulated with a combination of the gonadotrophin-releasing hormone (GnRH) agonist (GnRHa) (Decapeptyl LP 3; Ipsen, Paris, France) and recombinant FSH (r-fsh) (Puregon; Organon, Paris, France) to achieve multiple follicular development. Each patient received the GnRHa i.m. on day 1 of the normal cycle. Fourteen days later, pituitary desensitization was evaluated by ultrasound (no follicle >10 mm in diameter) and by serum oestradiol concentrations (<50 pg/ml). Then r-fsh was administered s.c. at a dose of IU according to the day 3 FSH concentration (measured in a pretreatment cycle), each evening over 7 days. At that point, the r-fsh dose was modified according to the degree of ovarian response, which was evaluated by daily ultrasound examination to observe follicular development and by serum oestradiol concentrations. The criteria for HCG administration (5000 IU i.m.) included an oestradiol concentration of >1500 pg/ml, and at least three follicles of >18 mm in diameter. Oocytes were retrieved 35 h after HCG administration under ultrasound guidance. Embryos were transferred 2 or 3 days later in all patients in whom embryos were obtained after IVF or ICSI. The number of embryos replaced was determined according to the following criteria: age of the patient, ovarian response to exogenous stimulation, fertilization rate, morphological appearance of the embryos and proper counselling of the couple. Micronized vaginal progesterone (600 mg a day, started on the day of oocyte retrieval and continued for up to 14 days) was used for luteal support. Assessment of embryos The embryos were evaluated by a biologist (SH) and classified as follows: grade 1: perfectly symmetrical with no fragmentation; grade 2: perfectly symmetrical with slight fragmentation (<20% fragmentation of the total embryonic volume); grade 3: uneven blastomeres with no fragmentation; grade 4: uneven blastomeres with gross fragmentation (>20% fragments). Grade 1 or 2 embryos were considered high quality. Ultrasonographic measurements On the day of HCG administration, evaluation of ovaries was performed by HD or LR, then all ultrasonographic investigations of the endometrium were performed by the same trained operator (EB) using a 7.5-MHz transvaginal transducer (EV9 4) with colour Doppler facility (Siemens Sonoline G60S). All measurements were carried out between 7:00 a.m. and 9:00 a.m. to eliminate diurnal variation (Zaidi et al., 1995). The patients and treatment staff were blind to the results of the ultrasound investigations. In the same conditions, all patients underwent ultrasonographic investigation on the day of oocyte retrieval and on the day of embryo transfer. The preimplantation ultrasonographic evaluation included all of the following parameters: endometrial thickness, endometrial pattern, uterine artery pulsatility index, protodiastolic notch, end-diastolic blood flow and endometrial subendometrial blood flow distribution pattern. Endometrial thickness was measured on static grey-scale images. The maximum anteroposterior distance between the echogenic interfaces at the endometrial myometrial junction was measured in the plane through the central longitudinal axis of the uterine body and with the callipers placed on the 369

3 370 outer walls of the endometrium. An endometrial thickness between 7 and 14 mm was considered as better prognosis for implantation than <7 or >14 mm (Schild et al., 2001). The endometrial pattern was assessed according to the classification proposed by Sher as follows: grade I: characterized by homogeneous echogenicity of the endometrium, and grade II: characterized by an outer peripheral layer of dense echogenicity surrounding a central sonolucent area, visualized using a transverse axis (Sher et al., 1991). On the day of HCG administration and on the day of oocyte retrieval, the endometrium has to be multilayered (a prominent outer and central hyperechogenic line and inner hypoechogenic region). On the day of embryo transfer, the echogenicity of the endometrium should be increased as compared with the day of HCG administration. The uterine artery pulsatility index (PI) was measured on a transverse scan through the uterine isthmus. Colour Doppler signals were obtained from the right and left ascending branches of the uterine arteries in 91% of cases. Once visualized, a pulsed Doppler range gate was placed over each artery to obtain velocity waveforms. After confirming that waveforms were continuous, an average of three to five cardiac cycles was selected for calculation of PI, which was determined by the ultrasound software. PI <3 was normal, and 3 was abnormal (Bied-Damon et al., 1995). The velocity waveform was analysed during three cardiac cycles. A pattern was considered to be adequate if the protodiastolic notch was absent or minimal (Ardaens et al., 1998). A severe notch or the presence of a reverse flow was abnormal. The end-diastolic blood flow was considered as a favourable factor (Salle et al., 1998). Endometrial vascularization was assessed on a longitudinal scan of the uterus. The colour gate was placed over the thickest part of the endometrium. The settings for power Doppler sonography were standardized for the highest sensitivity in the absence of apparent noise using a high-pass filter at 50 Hz, pulsed repetition frequency at 750 Hz, and moderate long persistence. To characterize the endometrialsubendometrial blood flow distribution pattern, the definition of Applebaum was adopted, summarized as follows: zone 1, vessels penetrating the outer hypoechogenic area surrounding the endometrium but not entering the hyperechogenic outer margin; zone 2, vessels penetrating the hyperechogenic outer margin of the endometrium but not entering the hypoechogenic inner area; and zone 3, vessels entering the hypoechogenic inner area (Applebaum, 1995). The vascularization was considered positive if vessels entered the hypoechogenic inner area. Statistical analysis Qualitative variables were expressed in frequency and percentage, and quantitative variables by the mean and SD. Comparisons, in particular of the success or failure of IVF, were made using the chi-squared test (or Fisher exact test if needed) for qualitative variables and Student s t-test (or the Mann Whitney U- test if needed) for quantitative variables. Various ultrasonographic parameters were studied in order to work out a predictive test of the success of IVF (i.e. achieving pregnancy). The methodology used by Haydel et al. (2000) to identify patients presenting a minor head injury and requiring computed tomography (recursive partitioning analysis) was followed. The ultrasonographic parameters studied were endometrial thickness and echogenicity, pulsatility index, presence of a notch, end-diastolic blood flow and endometrial vascularization. First, ultrasonographic parameters were selected that were discriminatory enough to be included in the test. Their distribution was analysed according to whether or not pregnancy was obtained, using the chi-squared test. For this purpose, the quantitative parameters were made qualitative (creation of classes). Those parameters for which the probability of the test was lower than the alpha threshold 0.05 were selected. The parameters were then classified in descending order, on the basis of which difference was most significant. Endometrial thickness, which cannot be dissociated from endometrial pattern, was included among the selected factors, even though there was no difference according to whether or not pregnancy occurred. In the authors clinical practice, endometrial thickness and endometrial pattern are not dissociated because endometrial pattern is relevant only if endometrial thickness is between 7 and 14 mm. From another source, endometrial thickness may be a good prognostic factor for achieving implantation whatever the endometrial pattern (Sher et al., 1991). Sensitivity, specificity and positive and negative predictive values were calculated by considering only the most significant variable. The second most significant variable was combined with the first sign. New sensitivity, specificity and predictive values were computed. This was continued until all selected signs were combined. Sensitivity, specificity, positive predictive value and negative predictive values are given with a 95% confidence interval. The test thus created is considered positive if all the parameters are regarded as positive or present (in favour of pregnancy). So, if at least one of the parameters is regarded as negative or absent, the test is regarded as negative (against pregnancy). For the calculation of predictive values, the prevalence considered was that of the sample, i.e. the rate of pregnancy per transfer calculated for this study. The test giving the best sensitivity, specificity and predictive values was then selected. Additionally, in order to define the optimal timing of ultrasonography, the newly created test was performed at two other times: the day of oocyte retrieval and the day of embryo transfer. SAS version 8.1 software (SAS Institute Inc., USA) was used for the statistical analysis, with an alpha level of Results After IVF, 37 of the 112 patients included were pregnant. Patient characteristics and methods of treatment are

4 summarized in Table 1. No difference could be found between pregnant and non-pregnant women. The proportions of IVF and ICSI were similar, as were the numbers of transferred embryos. The percentage of high quality embryos transferred was not significantly different between the two groups of patients. The pregnancy rates per cycle and transfer were respectively 28.9 and 33%. The ongoing pregnancy rate per transfer was 27.7% (six first-trimester miscarriages). The multiple pregnancy rate was 21.6% (eight twin pregnancies). The implantation rate was 19.8%. Ultrasonographic and Doppler parameters of pregnant and non-pregnant women are summarized in Table 2. Uterine exploration was performed on the day of HCG administration. Women with an end-diastolic blood flow, an endometrial blood flow and a multilayered endometrium were more likely to be pregnant than women without one or more of these signs. Endometrial thickness, uterine pulsatility index and protodiastolic notch did not characterize pregnant or nonpregnant women. The study of these different parameters on the day of HCG injection according to pregnancy status enabled the calculation for each of them of the positive predictive value, i.e. the probability of a patient being pregnant if the sign is present in its favourable configuration. A study prevalence of being pregnant (33%) was used. Thus, the probability of a patient being pregnant when endometrial thickness was between 7 and 14 mm was In the same conditions, if the endometrium was multilayered, the probability was If PI was under 3, the probability was 0.34, and when enddiastolic blood flow was present, the probability rose to Finally, if a protodiastolic notch was absent, the probability was 0.40, and if endometrial blood flow was present, the probability was In agreement with the methodology described in the statistics section, and in order to reach the objective, sensitivity, specificity, positive and negative predictive values were first computed with the most significant parameter, that is with the end-diastolic blood flow. Then a second parameter, endometrial pattern, was added, and, finally, endometrial thickness was added. Results are summarized in Table 3. This methodology allowed creation of the most effective combination of three signs in evaluation of uterine receptivity. Sensitivity and specificity were around 81%, positive predictive value was 68.2%, and negative predictive value 89.7%. This combination excluded endometrial blood flow which, although significant (P = 0.01) as a single parameter, did not increase the sensitivity and specificity in combination. The same methodology was used to evaluate these parameters on the day of oocyte retrieval and on the day of embryo transfer (data not shown). The combination of the three parameters (end-diastolic blood flow, endometrial pattern and endometrial thickness) was the most efficient. Comparisons showed that the best sensitivity and specificity were obtained on the day of HCG administration: respectively 81.1 and 81.3% (Table 4). Table 1. Description of the study population according to whether or not pregnancy was achieved after IVF/ICSI. Parameter Pregnant Non-pregnant (n = 37) (n = 75) Age (years) 31.9 ± ± 5 Body mass index (kg/m 2 ) 22.1 ± ± 4.7 Smoker (%) Day 3 FSH (previous cycle) (IU/l) 6.7 ± 2 7 ± 2.3 Day 3 oestradiol (previous cycle) (pg/ml) 47.5 ± ± 42.4 IVF (%) ICSI (%) Number of transferred embryos 2.3 ± ± 0.7 High quality embryo (%) Values are expressed as mean ± SD, unless otherwise stated; ICSI = intracytoplasmic sperm injection There were no statistically significant differences between the pregnant and non-pregnant patients. 371

5 Table 2. Ultrasonographic and Doppler parameters on the day of the human chorionic gonadotrophin administration, according to pregnancy status. Ultrasonographic parameter Pregnant Non-pregnant P-value n = 37 (%) n = 75 (%) Endometrial thickness 7 14 mm 37 (100) 72 (96) <7 or >14 mm 0 (0) 3 (4) NS Endometrial pattern Multilayered endometrium 35 (95) 49 (65) Other 2 (5) 26 (35) Uterine artery pulsatility index <3 37 (100) 73 (97) 3 0 (0) 2 (3) NS Protodiastolic notch Present 8 (22) 32 (43) Absent 29 (78) 43 (57) NS End-diastolic blood flow Present 31 (84) 28 (37) Absent 6 (16) 47 (63) Endometrial blood flow Present 25 (68) 32 (43) Absent 12 (32) 43 (57) 0.01 NS = not statistically significant. Table 3. Sensitivity, specificity, positive predictive value and negative predictive value of the ultrasonographic and Doppler parameters measured on the day of human chorionic gonadotrophin injection (recursivepartitioning analysis). Parameter Sensitivity Specificity PPV NPV TP/FP End-diastolic 83.8 ( ) 62.7 ( ) 52.5 ( ) 88.7 ( ) 31/28 blood flow End-diastolic blood 83.8 ( ) 65.3 ( ) 54.4 ( ) 89.1 ( ) 31/26 flow + endometrial thickness a End-diastolic blood 81.1 ( ) 80.0 ( ) 66.7 ( ) 89.6 ( ) 30/15 flow + endometrial pattern End-diastolic blood 81.1 ( ) 81.3 ( ) 68.2 ( ) 89.7 ( ) 30/14 flow + endometrial thickness + endometrial pattern Results are given with 95% confidence interval. PPV: positive predictive value; NPV: negative predictive value; TP: true positive; FP: false positive. a Hypothesis not selected, but reported for the reader. 372

6 Table 4. Sensitivity, specificity, positive predictive value and negative predictive value of the ultrasonographic and Doppler parameters measured on the day of human chorionic gonadotrophin (HCG) injection, the day of oocyte retrieval, and the day of embryo transfer. Ultrasonographic Sensitivity Specificity PPV NPV parameter Day of HCG administration 81.1 ( ) 81.3 ( ) 68.2 ( ) 89.7 ( ) Day of oocyte retrieval 29 (13 45) 94.1 ( ) 69.2 ( ) 74.4 ( ) Day of embryo transfer 73.9 ( ) 51 (37 65) 41.5 ( ) 80.6 ( ) Results are given with 95% confidence interval. PPV: positive predictive value; NPV: negative predictive value. Discussion There is no consensus in the literature concerning the precise time when ultrasonographic and Doppler evaluation of the uterus would optimize the prediction of pregnancy during an IVF cycle. Many authors have studied ultrasonographic parameters on the day of HCG administration or at the beginning of the luteal phase (Salle et al., 1998, 2001; Basir et al., 2002a,b; Dietterich et al., 2002; Puerto et al., 2003; Chien et al., 2004). No consensus emerges from these studies, given the heterogeneity of study populations and methodologies. The present study avoided any confounding influence of these factors because the same stimulation protocol was used for all patients, who were all studied by the same ultrasonographer according to a wellstandardized technique, and because the characteristics of these patients and the IVF data were comparable in the two groups. In addition, so far as is known, no published comparative study has considered the most convenient timing of ultrasonography in terms of prediction of embryo implantation. This comparison was prospectively made starting from three ultrasonographic and Doppler parameters, and with constant embryo quality. Analysis of the sensitivity, specificity, positive and negative predictive values shows that this evaluation of endometrial receptivity is more relevant if it is done on the day of HCG injection, rather than the day of oocyte retrieval or the day of embryo transfer. Studies published hitherto have generally evaluated one or more separately analysed parameters (endometrial thickness, endometrial pattern, uterine artery pulsatility index) on the day of HCG injection, the day of oocyte retrieval, or the day of embryo transfer (Check et al., 2000; Contart et al., 2000; Hsieh et al., 2000; Schild et al., 2000; Chien et al., 2002; Wu et al., 2003). However, neither pregnancy nor pregnancy outcome can be predicted by studying only one ultrasonographic parameter (Zhang et al., 2005). Moreover, the day of the evaluation of this parameter during an IVF cycle can also influence the result of this evaluation (Khalifa et al., 1992; Steer et al., 1994, 1995; Bied-Damon et al., 1995; Friedler et al., 1996; Dickey, 1997). In the present study, by carrying out several consecutive measurements during each IVF cycle, the variations in each ultrasonographic and Doppler parameter were observed, and the profiles of those patients who will become pregnant and those who will not were defined. In order to increase the relevance of the evaluation, it is proposed to combine three parameters (endometrial thickness, endometrial pattern, end-diastolic blood flow). This combination of several parameters reduces the potential error related to each of them. The choice of these three parameters was not made arbitrarily but rather initially on the basis of their individual evaluation. The potential impact of these parameters on whether or not pregnancy was obtained after IVF was studied. Then, in order to increase sensitivity and specificity, these parameters were combined until the test became most predictive of the chances of pregnancy (Haydel et al., 2000). This test has a sensitivity and specificity higher than 81%, and positive and negative predictive values of 68 and 89% respectively. The methodology used did not strictly follow that of Haydel s, which considered dependence between factors. The recursive-partitioning method of Haydel was based on categorization of patients regarding the end-point assessed (in the present case, pregnancy), and tried at the outset to define significant differences in various independent parameters among these categories. In that case, after identifying the most significant independent predictor of the looked-for end-point (end-diastolic blood flow), patients positive for that factor should have been removed from further analysis. This procedure led to adding up all possible predictor factors pointing to the end-point looked for and to improved sensitivity, but decreased specificity. In the present study, the dependence between factors was not considered, improving specificity at the cost of diminished sensitivity. It appeared that the second method was more powerful, inducing a low decrease of sensitivity balanced by a high increase of specificity. By combining three signs and using this combination on various days during the IVF procedure,it was possible to determine the best day on which to predict endometrial receptivity. To be able to evaluate the chances of embryo implantation on the day of HCG injection is of considerable importance (Favre et al., 1993; Zaidi et al., 1996). It is still possible at this point to modify the IVF procedure, by cancelling the cycle, freezing the embryos, or changing ovarian stimulation parameters. Studies on the day of embryo transfer of a parameter potentially predictive of embryo implantation show that it is too late to change the decision to transfer (Puerto et al., 2003). Some reports have evaluated the chances of pregnancy on day 7 after embryo transfer, not allowing any modification (Tekay et al., 1995). Because changes are induced by ovarian stimulation, the 373

7 374 evaluation of a previous cycle cannot be superimposed to the current cycle of treatment (Yaron et al., 1994; Salle et al., 1998; Basir et al., 2002a,b). Determination of the optimal timing for prediction of uterine receptivity is important for IVF cycles, and for other assisted reproductive techniques. However, taking into account the differences between assisted reproductive techniques, the results cannot be validated without specific evaluation for each technique (Urman et al., 2005a,b). IVF lends itself well to this kind of evaluation, and offers the possibility of evaluating both embryo quality and the number of embryos to transfer. IVF only makes it possible to control the whole treatment cycle, except the phenomenon of embryo implantation (Thurin et al., 2005; Torsky et al., 2005). Prediction of the likelihood of embryo implantation, and therefore of pregnancy in IVF, is not only of value in counselling infertile couples. It also has medical and economic implications. One aim is to transfer as few embryos as possible for one patient, thereby limiting the risk of multiple pregnancy (Strandell et al., 2000). The choice of top quality embryos must be combined with better selection of patients (Strandell et al., 2000). However, in terms of the patient, the evaluation was based only on age, IVF indication and the quality of ovarian response to stimulation in the absence of objective criteria of uterine receptivity. This is why uterine evaluation by ultrasonography and Doppler is very important (Schild et al., 2000; Chien et al., 2002; Ozturk et al., 2004). This best analysis of uterine receptivity should then be integrated into a more complete algorithm, which should then be evaluated in randomized, multicentre studies (Urman et al., 2005; Killick, 2007). In the case of a guarded prognosis, assessment of uterine receptivity may induce medical teams to persuade couples not to undergo transfer, rather than to do so but with negligible chances of success, or to accept an increase in the number of embryos to be transferred. All these criteria strengthen the dialogue between medical teams and couples and lead to informed consent based on objective reasoning. Like any clinical study, the present one has its limits. This method is relatively time-consuming in daily routine. The choice of the test parameters can be criticized even though their combination reduces the risk of error associated with each one. The threshold values of each parameter are also a matter for discussion in the literature (Basir et al., 2002; Dietterich et al., 2002; Kovacs et al., 2003; Chien et al., 2004). The use of results from only one ultrasonographer reduces inter-individual variation but scarcely reflects everyday medical practice (Schild et al., 2001). Therefore, the results are validated for only one method in a given centre. However, this work can be used as a basis to prepare multicentre studies to check whether or not the methodology can be validated by different operators. One of the main difficulties in carrying out large multicentre studies is ensuring that the same methodology is used throughout the study and in the various study centres. In the field of IVF, this methodological and clinical difficulty is further compounded by marked biological variability. In conclusion, determining the most convenient day and method for assessment of uterine receptivity constitutes the first step towards improved clinical practice in IVF treatment. The second step is the subject of work in progress, aiming to reduce subjectivity in deciding when to transfer embryos, by putting forward objective clinical arguments regarding the likelihood of successful IVF. References Andersen AN, Gianaroli L, Felberbaum R et al Assisted reproductive technology in Europe, Results generated from European registers by ESHRE. Human Reproduction 20, Applebaum M 1995 The uterine biophysical profile. Ultrasound Obstetrics Gynecology 5, Ardaens Y, Gougeon A, Lefebvre C et al [Contribution of ovarian and uterine color Doppler in medically assisted reproduction techniques (ART)]. Gynecologie Obstetrique Fertilite 30, Ardaens Y, Leroy M, Lefebvre C et al [Color echo-doppler and endometrial receptivity in in vitro fertilization]. Contraception Fertilité Sexualité 26, Aytoz A, Ubaldi F, Tournaye H et al The predictive value of uterine artery blood flow measurements for uterine receptivity in an intracytoplasmic sperm injection program. Fertility and Sterility 68, Basir GS, Lam TP, O W, Chau MT et al. 2002a Cycle-to-cycle variation in utero ovarian hemodynamic indices in ovarian stimulation and natural cycles of the same women and its effect on the outcome of assisted reproduction treatment. Fertility and Sterility 78, Basir GS, O WS, So WW et al. 2002b Evaluation of cycle-to-cycle variation of endometrial responsiveness using transvaginal sonography in women undergoing assisted reproduction. Ultrasound Obstetrics Gynecology 19, Bassil S 2001 Changes in endometrial thickness, width, length and pattern in predicting pregnancy outcome during ovarian stimulation in in vitro fertilization. Ultrasound Obstetrics Gynecology 18, Bied-Damon V, Salle B, Payan F et al [Study of uterine blood supply by pulsed doppler before embryo transfer]. Contraception Fertilité Sexualité 23, Carbillon L, Perrot N, Uzan M, Uzan S 2001 Doppler ultrasonography and implantation: a critical review. Fetal Diagnosis Therapy 16, Check JH, Dietterich C, Lurie D 2000 Non-homogeneous hyperechogenic pattern 3 days after embryo transfer is associated with lower pregnancy rates. Human Reproduction 15, Chien LW, Lee WS, Au HK, Tzeng CR 2004 Assessment of changes in utero ovarian arterial impedance during the peri-implantation period by Doppler sonography in women undergoing assisted reproduction. Ultrasound Obstetrics Gynecology 23, Chien LW, Au HK, Chen PL et al Assessment of uterine receptivity by the endometrial subendometrial blood flow distribution pattern in women undergoing in vitro fertilization embryo transfer. Fertility and Sterility 78, Contart P, Baruffi RL, Coelho J et al Power Doppler endometrial evaluation as a method for the prognosis of embryo implantation in an ICSI program. Journal of Assisted Reproduction Genetics 17, De Geyter C, Schmitter M, De Geyter M et al Prospective evaluation of the ultrasound appearance of the endometrium in a cohort of 1,186 infertile women. Fertility and Sterility 73, Dickey RP 1997 Doppler ultrasound investigation of uterine and ovarian blood flow in infertility and early pregnancy. Human Reproduction Update 3, Dietterich C, Check JH, Choe JK et al Increased endometrial thickness on the day of human chorionic gonadotropin injection does not adversely affect pregnancy or implantation rates following in vitro fertilization embryo transfer. Fertility and Sterility 77, Ebrard-Charra S, Reyftmann L, Hedon B, Dechaud H 2005 [Ultrasonographical predictive factors of ovarian response to

8 stimulation prior to in vitro fertilization]. Gynecologie Obstetrique Fertilité 33, Fanchin R 2001 Assessing uterine receptivity in 2001: ultrasonographic glances at the new millennium. Annals of the New York Academy of Sciences 943, Favre R, Bettahar K, Grange G et al Predictive value of transvaginal uterine Doppler assessment in an in vitro fertilization program. Ultrasound Obstetrics Gynecology 3, Friedler S, Schenker JG, Herman A, Lewin A 1996 The role of ultrasonography in the evaluation of endometrial receptivity following assisted reproductive treatments: a critical review. Human Reproduction Update 2, Haydel MJ, Preston CA, Mills TJ et al Indications for computed tomography in patients with minor head injury. New England Journal of Medicine 343, Hsieh YY, Chang FC, Tsai HD 2000 Doppler evaluation of the uterine and spiral arteries from different sampling sites and phases of the menstrual cycle during controlled ovarian hyperstimulation. Ultrasound Obstetrics Gynecology 16, Khalifa E, Brzyski RG, Oehninger S et al Sonographic appearance of the endometrium: the predictive value for the outcome of in-vitro fertilization in stimulated cycles. Human Reproduction 7, Killick S 2007 Ultrasound and the receptivity of the endometrium. Reproductive BioMedicine Online 15, Kovacs P, Matyas S, Boda K, Kaali SG 2003 The effect of endometrial thickness on IVF/ICSI outcome. Human Reproduction 18, Kupesic S, Bekavac I, Bjelos D, Kurjak A 2001 Assessment of endometrial receptivity by transvaginal color Doppler and three-dimensional power Doppler ultrasonography in patients undergoing in vitro fertilization procedures. Journal of Ultrasound Medicine 20, Ozturk O, Bhattacharya S, Saridogan E et al Role of utero ovarian vascular impedance: predictor of ongoing pregnancy in an IVF embryo transfer programme. Reproductive BioMedicine Online 9, Puerto B, Creus M, Carmona F et al Ultrasonography as a predictor of embryo implantation after in vitro fertilization: a controlled study. Fertility and Sterility 79, Salle B, Bied-Damon V, Benchaib M et al Preliminary report of an ultrasonography and colour Doppler uterine score to predict uterine receptivity in an in-vitro fertilization programme. Human Reproduction 13, Schild RL, Knobloch C, Dorn C et al Endometrial receptivity in an in vitro fertilization program as assessed by spiral artery blood flow, endometrial thickness, endometrial volume, and uterine artery blood flow. Fertility and Sterility 75, Schild RL, Holthaus S, d Alquen J et al Quantitative assessment of subendometrial blood flow by three-dimensional-ultrasound is an important predictive factor of implantation in an in-vitro fertilization programme. Human Reproduction 15, Schwartz LB, Chiu AS, Courtney M et al The embryo versus endometrium controversy revisited as it relates to predicting pregnancy outcome in in-vitro fertilization embryo transfer cycles. Human Reproduction 12, Sher G, Herbert C, Maassarani G, Jacobs MH 1991 Assessment of the late proliferative phase endometrium by ultrasonography in patients undergoing in-vitro fertilization and embryo transfer (IVF/ ET). Human Reproduction 6, Steer CV, Tan SL, Dillon D et al Vaginal color Doppler assessment of uterine artery impedance correlates with immunohistochemical markers of endometrial receptivity required for the implantation of an embryo. Fertility and Sterility 63, Steer CV, Tan SL, Mason BA, Campbell S 1994 Midluteal-phase vaginal color Doppler assessment of uterine artery impedance in a subfertile population. Fertility and Sterility 61, Strandell A, Bergh C, Lundin K 2000 Selection of patients suitable for one-embryo transfer may reduce the rate of multiple births by half without impairment of overall birth rates. Human Reproduction 15, Tekay A, Martikainen H, Jouppila P 1995 Blood flow changes in uterine and ovarian vasculature, and predictive value of transvaginal pulsed colour Doppler ultrasonography in an in-vitro fertilization programme. Human Reproduction 10, Thurin A, Hardarson T, Hausken J et al Predictors of ongoing implantation in IVF in a good prognosis group of patients. Human Reproduction 20, Torsky SP, Amato P, Cisneros PL et al Algorithm to predict assisted reproductive technology pregnancy outcome reveals minimal embryo synergy. Fertility and Sterility 83, Urman B, Yakin K, Balaban B 2005a Recurrent implantation failure in assisted reproduction: how to counsel and manage. A. General considerations and treatment options that may benefit the couple. Reproductive BioMedicine Online 11, Urman B, Yakin K, Balaban B 2005b Recurrent implantation failure in assisted reproduction: how to counsel and manage. B. Treatment options that have not been proven to benefit the couple. Reproductive BioMedicine Online 11, Wu HM, Chiang CH, Huang HY et al Detection of the subendometrial vascularization flow index by three-dimensional ultrasound may be useful for predicting the pregnancy rate for patients undergoing in vitro fertilization embryo transfer. Fertility and Sterility 79, Yang JH, Wu MY, Chen CD et al Association of endometrial blood flow as determined by a modified colour Doppler technique with subsequent outcome of in-vitro fertilization. Human Reproduction 14, Yaron Y, Botchan A, Amit A et al Endometrial receptivity in the light of modern assisted reproductive technologies. Fertility and Sterility 62, Zaidi J, Pittrof R, Shaker A et al Assessment of uterine artery blood flow on the day of human chorionic gonadotropin administration by transvaginal color Doppler ultrasound in an in vitro fertilization program. Fertility and Sterility 65, Zaidi J, Jurkovic D, Campbell S et al Description of circadian rhythm in uterine artery blood flow during the peri-ovulatory period. Human Reproduction 10, Zhang X, Chen CH, Confino E et al Increased endometrial thickness is associated with improved treatment outcome for selected patients undergoing in vitro fertilization embryo transfer. Fertility and Sterility 83, Declaration: The authors report no financial or commercial conflicts of interest. Received 24 July 2007; refereed 17 August 2007; accepted 7 November