Quantitative assessment of subendometrial blood flow by three-dimensional-ultrasound is an important predictive factor of implantation in an in-vitro fertilization programme

R.L. Schild1,4, S. Holthaus2, J. d'Alquen2, R. Fimmers3, C. Dorn2, H. van der Ven2 and M. Hansmann1

1 Abteilung für Pränatale Diagnostik und Therapie, Frauenklinik der Universität, Sigmund Freud Strasse 25, 53105 Bonn, Germany, 2 Abteilung für Reproduktionsmedizin und Endokrinologie, Frauenklinik der Universität, and 3 Institut für Medizinische Statistik, Universität Bonn, Germany


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
This study was designed to investigate the role of three-dimensional (3D) power Doppler sonography of the (sub-) endometrial area on the first day of ovarian stimulation in predicting the outcome of an in-vitro fertilization (IVF) programme. Among the 75 cycles analysed, the overall pregnancy rate was 20% (15/75) per cycle and 23.8% (15/63) per embryo transfer. Intra-observer variability of the colour histogram was checked in 14 patients with the results demonstrating a high level of agreement. Neither endometrial measurements nor uterine blood flow were correlated with the pregnancy rate. In contrast, all 3D indices were significantly lower in conception compared with non-conception cycles (P < 0.05). Logistic regression analysis found the subendometrial flow index to be the strongest predictive factor of IVF success among the tested sonographic parameters (P = 0.04). In conclusion, quantitative assessment of spiral artery blood flow may be of predictive value for implantation in IVF cycles even before ovarian stimulation therapy is started.

Key words: IVF/pregnancy rate/quantitative assessment/sub- endometrial blood flow/3D-ultrasound


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Despite recent advances in in-vitro fertilization (IVF) techniques and ovarian stimulation regimens, the pregnancy rate has not increased accordingly and implantation rates per embryo transferred remain at a low 10–15% (Salle et al., 1998Go). Multiple factors responsible for a successful IVF outcome have been described, not the least of which are embryo quality and uterine receptivity (Coulam et al., 1994Go; Schwartz et al., 1997Go). It appears that a favourable endometrial milieu is necessary for successful implantation. What determines such a favourable endometrial milieu, however, is still controversial (Noyes et al., 1995Go). Among potential uterine predictors for implantation measurable by ultrasonography are endometrial thickness and volume, endometrial pattern as well as blood flow in the uterine and (sub)endometrial arteries. Unfortunately, neither of these parameters has been shown to be consistently predictive of pregnancy outcome (Friedler et al., 1996Go; Aytoz et al., 1997Go; Bloechle et al., 1997Go; Schild et al., 1999Go; Yuval et al., 1999Go). No study has yet explored, however, the possible interaction between overall blood supply in the (sub)endometrial area and pregnancy rate. With the advent of three-dimensional power Doppler sonography (3D-PDS) quantitative assessment of vessel density and perfusion within the (sub)endometrial area has become possible. In this study, we used the new technique of 3D-PDS to examine the relationship between spiral artery blood flow and vessel density on the first day of stimulation therapy and subsequent IVF outcome.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
For this study, a total of 96 patients with the predominant diagnosis of male factor infertility (n = 66, 68.8%), tubal occlusion (n = 22, 22.9%) or unexplained infertility (n = 8, 8.3%) was recruited at the time of a routine appointment on the first day of ovarian stimulation. All patients were scheduled to undergo ovarian stimulation and IVF therapy. Inclusion criteria for the study were sonographic evidence of a down-regulated endometrium (<5 mm), absence of ovarian cysts of >2.5 cm diameter and oestradiol concentrations of <60 pg/ml. Recruitment occurred whenever these inclusion criteria were fulfilled and the patient willing to participate. Each patient was included only once. At recruitment, seven patients each had an endometrial thickness of >5 mm, oestradiol concentrations of >60 pg/ml, and no oestradiol concentration determined respectively. These 21 patients were, therefore, excluded from further analysis. The study was approved by the local Ethics Committee of the University of Bonn, Germany, and informed consent was obtained from each patient. Prior to ovarian stimulation therapy, a standard regime of the gonadotrophin-releasing hormone (GnRH) agonist triptorelin (Decapeptyl®; Ferring, Kiel, Germany) was administered s.c. at a daily dose of 100 µg, starting in the mid-luteal phase and continuing for 12–14 days. On the first day of stimulation therapy transvaginal sonography using an electronic 7.5 MHz transducer with a 3D facility (Voluson 530D; Kretz-Technik, Zipf, Austria) was performed between 07:00 and 10:00 with the patient in the lithotomy position. When a true longitudinal view of the uterus was obtained the power Doppler mode was activated. Subsequently, the region of interest was defined by a moveable sector on the screen which was placed over the endometrium and subendometrial area. The sector had the shape of a truncated cone that was adjusted individually to minimize acquisition time while ensuring that the complete endometrial cavity from the fundus to the internal cervical os and the subendometrial area within 5 mm of the endometrial borders was included in the region of interest. Identical power Doppler settings (colour gain 44.8, pulse repetition frequency 1.9, C-PWR 3) were used in all patients. The 3D volume mode was switched on, volume acquisition time was set on fast (low resolution) to avoid artefacts and the patient was instructed to remain very still. Movements of the vaginal probe were excluded. Volume sampling took ~4–6 s during which time the area of interest was scanned in a fan-shaped pattern. Three orthogonal planes were simultaneously displayed on the screen with the perpendicular orientation of these planes being maintained throughout any translation or rotation. The data set was stored digitally on an internal disk drive for subsequent analysis and the ultrasound probe was removed. All examinations and further calculations were done by one investigator (R.L.S.). The volume was rotated into a pre-defined view with plane A (upper left) showing the longitudinal view, plane B (upper right) displaying the transverse view, and plane C (lower left) demonstrating the frontal view. A minimum-intensity projection of the vascular images (lower right plane) was created by starting the volume rendering mode. Quantitative analysis of the blood flow in the predefined area was achieved by implementing the colour histogram mode, the results of which were displayed as indices in the outer lower right panel (Figure 1Go). The vascularization index (VI) characterized vessel density, the flow index (FI) described the intensity of blood flow and the vascularization flow index (VFI) assessed both vascularization and perfusion. The indices were calculated by the built-in computer:



View larger version (55K):
[in this window]
[in a new window]
 
Figure 1. Multiplanar image of the endometrium in the power Doppler mode on the first day of stimulation therapy. Longitudinal, transverse and frontal view in upper left, upper right and lower left plane respectively. Three dimensional projection of the vascular images and colour histogram in the lower right plane and outer lower right panel respectively.

 
VI = colour values/(total voxels – background values)

FI = weighted colour values/colour values, and

VFI = weighted colour values/(total voxels – background values).

Endometrial thickness and volume were measured as described previously (Schild et al., 1999Go). Uterine artery blood flow was assessed by colour Doppler and the mean pulsatility index (PI) and peak systolic velocity (PSV) of the right and left uterine arteries were calculated. All measurements were corrected for the insonation angle. In no case was the angle >40°.

Subsequent to the ultrasound examination, ovarian stimulation with recombinant follicle stimulating hormone (rFSH) (Puregon®; Organon, Oberschleißheim, Germany) in appropriate doses was commenced. All patients underwent serial ultrasound examinations to assess follicular growth until at least three follicles with a mean diameter of >=18 mm were seen. At this stage 10 000 IU of human chorionic gonadotrophin (HCG) (Pregnesin®; Serono, Unterschleißheim, Germany) were used to induce ovulation. Oocytes were collected by transvaginal ultrasound-directed follicular aspiration and up to three good quality embryos were transferred 48 h after oocyte retrieval. The number of blastomeres and morphological grade of each embryo were recorded. Progesterone vaginal suppositories (400 mg daily) were prescribed for 14 days as luteal support. Pregnancy was defined as the occurrence of a positive ß-HCG (>10 IU) value at day 12 after embryo transfer and a second, higher value 2 days later. Only pregnancies reaching HCG values >100 IU were considered for evaluation.

Statistical analysis
All data were analysed using the SPSS statistical package on a personal computer. Pearson's correlation coefficient, Student's two-tailed t-test and the Mann–Whitney U-test were used as appropriate. The correlation coefficient was tested for deviation from zero. Logistic regression analysis was performed with IVF outcome and the 3D indices being the independent and dependent variables respectively.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
All 96 patients underwent assessment of the blood flow in the endometrial area on the first day of ovarian stimulation. A total of 21 patients had to be excluded from final analysis since they failed to meet the entry criteria to our study.

Intra-observer variability of the 3D indices was checked in 14 patients, the results demonstrating a high level of agreement (Table IGo). Since there was no significant difference in PI and PSV values between the right and left uterine artery, the mean value was calculated and used for subsequent analysis.


View this table:
[in this window]
[in a new window]
 
Table I. Results of intra-observer variability testing
 
Overall pregnancy rate was 20% (15/75) per cycle and 23.8% (15/63) per embryo transfer. In 14 of the 15 pregnancies fetal heart activity was demonstrable on ultrasound, one pregnancy was diagnosed as biochemical. In this study, the outcome of IVF/embryo transfer was not dependent on patient's age, body mass index, cause of infertility, number of previous IVF attempts, duration or total dose of rFSH administered, serum oestradiol concentration, number of oocytes harvested or fertilized, sperm concentration and motility, the mean number of embryos transferred, or the embryo score (Table IIGo). Likewise, both endometrial measurements and uterine artery blood flow indices showed no association with IVF outcome (Table IIIGo).


View this table:
[in this window]
[in a new window]
 
Table II. Clinical details of the 75 patients included in the study. Values are given as mean ± SD
 

View this table:
[in this window]
[in a new window]
 
Table III. Endometrial measurements and uterine blood flow at stimulation induction. Values are given as mean ± SD
 
In contrast, all 3D indices were significantly lower in conception compared with non-conception cycles (Table IVGo, Figures 2–4GoGoGo). Among the tested 3D indices, logistic regression analysis found the subendometrial FI the strongest predicting factor of IVF success (P = 0.04). Figure 5Go displays receiver operator curves for the individual 3D indices.


View this table:
[in this window]
[in a new window]
 
Table IV. Three dimensional subendometrial indices. Values are given as mean ± SD
 


View larger version (28K):
[in this window]
[in a new window]
 
Figure 2. Values of the subendometrial vascularization index (VI) in non-conception and conception cycles. The boxes represent 50% of values (25th and 75th centile) while the vertical lines lead to the outer ranges without considering outliers. The median value is shown by the horizontal bar in the box.

 


View larger version (26K):
[in this window]
[in a new window]
 
Figure 3. Values of the subendometrial flow index (FI) in non-conception and conception cycles.

 


View larger version (23K):
[in this window]
[in a new window]
 
Figure 4. Values of the subendometrial vascularization flow index (VFI) in non-conception and conception cycles.

 


View larger version (13K):
[in this window]
[in a new window]
 
Figure 5. Receiver operator curves for the individual three dimensional subendometrial indices.

 
Of all cycles analysed, five cycles (6.7%) demonstrated poor ovarian response to stimulation therapy and further treatment was discontinued. In seven cycles (9.3%) no fertilization occurred. A total of 54 cycles (72%), however, were characterized by transfer of at least two good quality embryos (grade 3 or 4 according to the classification of Steer et al., 1992). In this subgroup, pregnancy rate was 24.1% (13/54) with all tested 3D indices being significantly lower in conception compared with non-conception cycles (Table VGo).


View this table:
[in this window]
[in a new window]
 
Table V. Subgroup of patients with at least two good quality embryos (grade 3 or 4 according to Steer et al., 1992) transferred. Values are given as mean ± SD of the three dimensional subendometrial indices
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Although previous studies have reported on the use of transvaginal sonography to determine intrauterine blood flow, this is, to the best of our knowledge, the first study to relate quantitative assessment of intrauterine blood flow on the first day of ovarian stimulation to IVF outcome.

Achiron et al. (1995) evaluated the endometrial blood flow response to hormone replacement therapy in women with premature ovarian failure who planned to enter an oocyte donation programme. The authors examined 18 women with ovarian failure in the study group and 12 volunteers with normal menstrual cycles in the control group. With the exception of the early follicular phase no significant difference was revealed between the two groups. Endometrial blood flow increased during the proliferative stage of the cycle because of reduced downstream impedance manifested by increased diastolic flow. From the point of maximum oestradiol concentration, rising diastolic impedance resulted in decreased flow to the endometrium in the secretory phase (Achiron et al., 1995Go).

Zaidi et al. (1995) examined endometrial thickness, morphology, vascular penetration and velocimetry on the day of HCG administration in an IVF programme. Between conception and non-conception cycles there was no significant difference with regard to endometrial thickness, endometrial pattern, subendometrial PSV or subendometrial PI. Failure of implantation was associated with absent subendometrial blood flow. This was deemed a useful predictor of adverse IVF outcome (Zaidi et al., 1995Go).

A new concept of evaluating uterine receptivity by means of a uterine score including assessment of endometrial blood flow within zone 3 was introduced by Applebaum (1995). With absent endometrial blood flow, despite maximum values for the other parameters, no conception was achieved (Applebaum, 1995Go).

A similar concept was pursued by Salle et al. (1998) who proposed a uterine score calculated in the secretory phase of the menstrual cycle preceding IVF. Among other parameters, the presence or absence of colour in the subendometrial region was determined. The vascularization was considered as positive if more than three vessels penetrating the outer hypoechogenic area surrounding the endometrium could be seen. None of the individual ultrasonographic or Doppler parameters tested were of sufficient accuracy to predict uterine receptivity, whereas the uterine score seemed to be a useful predictor of implantation (Salle et al., 1998Go).

Bourne et al. (1996) defined subendometrial blood flow arbitrarily as waveforms emanating from an area <=3.0 mm from the apparent basal layer of the endometrium. In their study of vascular and morphological changes in the human uterus after a positve self-test for the urinary luteinizing hormone (LH) surge, the authors reached the conclusion that the PSV and the time-averaged maximum velocity (TAMXV) were the most sensitive markers of physiological function for small vessels, e.g. the subendometrial spiral arteries, whereas an index of impedance appeared to be more appropriate for large vessels, e.g. the uterine arteries (Bourne et al., 1996Go).

Conflicting results of uterine artery blood flow assessment at the beginning of ovarian stimulation have been described. In the study of Bloechle et al. (1997) conception cycles were characterized by significantly lower PI and resistance index (RI) values in the ascendent uterine artery at the beginning of ovarian stimulation. In contrast, our data did not reveal any correlation between uterine artery Doppler findings at stimulation induction and subsequent pregnancy rate.

Battaglia et al. (1997) studied thromboxane production and colour Doppler changes in women undergoing different ovarian stimulation protocols. Thromboxane concentrations were measured on blood samples taken at the beginning of ovarian stimulation and on the day of oocyte retrieval. Thromboxane production was also evaluated on endometrial cells sampled and cultured on the day of ovum retrieval. Significantly lower endometrial culture cell thromboxane values were observed in conception cycles while plasma thromboxane concentrations were not significantly different between conception and non-conception cycles. The Doppler PI values of uterine and spiral arteries on the day of oocyte retrieval were significantly lower in women achieving pregnancy. In this group the presence and endometrial penetration of spiral arteries appeared to be more intense although this subjective impression could not be quantified (Battaglia et al., 1997Go). In our opinion, this obvious lack of quantitative assessment cannot provide firm evidence that endometrial vascularity is indeed increased in conception cycles.

Recently, Yuval et al. (1999) examined whether endometrial blood flow and thickness could predict the success of IVF. In their study of 156 cycles, the authors assessed the endometrial PI and RI, the systolic/diastolic ratio and endometrial thickness on the day of oocyte retrieval and on the day of embryo transfer. Neither endometrial thickness nor endometrial blood flow appeared to be correlated with pregnancy rate (Yuval et al., 1999Go).

In our study, no significant difference was found between the pregnant and non-pregnant group with regard to serum oestradiol concentrations, endometrial thickness and volume. However, significantly lower 3D indices were observed in conception cycles compared with non-conception cycles, suggesting that a lesser degree of intrauterine vascularization and perfusion at the beginning of ovarian stimulation indicated a more favourable endometrial milieu. These results were rather surprising, as the opposite might have been expected. Possibly, a better functional down-regulation of the endometrium following GnRH agonist application for 12–14 days, as expressed by lower 3D indices, increases the chances of successful implantation. This hypothesis, however, will need to be tested in further studies.

Previous work on the microvasculature in the endometrium throughout the normal menstrual cycle has demonstrated that the capillary basal lamina is loosely formed and discontinuous in the early proliferative phase. Progressive but heterogeneous differentiation of the endothelial cells occurs towards the mid-secretory phase of the cycle. It seems plausible that the co-ordinated growth and development of the endometrium is supported by extensive physical contacts connecting its component cell types (Roberts et al., 1992Go). These findings demonstrate the variable nature of the endometrial vessel system and suggest that specific alterations may be necessary in preparation for embryo implantation. In a review on the structure and function of endometrial blood vessels (Rogers, 1996Go), it was confirmed that endometrial blood vessels form a vascular bed with unusual properties in that these vessels undergo constant cycles of growth and regression during the reproductive life of the woman (Rogers, 1996Go).

In conclusion, at the initiation of ovarian stimulation following GnRH down-regulation, significant differences in (sub-) endometrial blood flow and vessel density can be found between subsequently pregnant or non-pregnant women. These findings may allow earlier prediction of a non-receptive endometrium, at a time when further stimulation therapy could be postponed to a cycle with a more favourable endometrial milieu. Further studies are needed to evaluate whether this proposed management will result in a higher implantation rate.


    Notes
 
4 To whom correspondence should be addressed Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Achiron, R., Levran, D., Sivan, E. et al. (1995) Endometrial blood flow response to hormone replacement therapy in women with premature ovarian failure: a transvaginal Doppler study. Fertil. Steril., 63, 550–554.[ISI][Medline]

Applebaum, M. (1995) The uterine biophysical profile. Ultrasound Obstet. Gynecol., 5, 67–68.[ISI][Medline]

Aytoz, A., Ubaldi, F., Tournaye, H. et al. (1997) The predictive value of uterine artery blood flow measurements for uterine receptivity in an intracytoplasmic sperm injection program. Fertil. Steril., 68, 935–937.[ISI][Medline]

Battaglia, C., Artini, P.G., Giulini, S. et al. (1997) Colour Doppler changes and thromboxane production after ovarian stimulation with gonadotrophin-releasing hormone agonist. Hum. Reprod., 12, 2477–2482.[Abstract]

Bloechle, M., Schreiner, Th., Küchler, I. et al. (1997) Colour Doppler assessment of ascendent uterine artery perfusion in an in-vitro fertilization-embryo transfer programme after pituitary desensitization and ovarian stimulation with human recombinant follicle stimulating hormone. Hum. Reprod., 12, 1772–1777.[Abstract]

Bourne, T.H., Hagström, H.-G., Granberg, S. et al. (1996) Ultrasound studies of vascular and morphological changes in the human uterus after a positive self-test for the urinary luteinizing hormone surge. Hum. Reprod., 11, 369–375.[Abstract]

Coulam, C.B., Bustillo, M., Soenksen, D.M. and Britten, S. (1994) Ultrasonographic predictors of implantation after assisted reproduction. Fertil. Steril., 62, 1004–1010.[ISI][Medline]

Friedler, S., Schenker, J.G., Herman, A. and Lewin, A. (1996) The role of ultrasonography in the evaluation of endometrial receptivity following assisted reproductive treatments: a critical review. Hum. Reprod. Update, 2, 323–335.[Abstract/Free Full Text]

Noyes, N., Liu, H.-C., Sultan, K. et al. (1995) Endometrial thickness appears to be a significant factor in embryo implantation in in-vitro fertilization. Hum. Reprod., 10, 919–922.[Abstract]

Roberts, D.K., Parmley, T.H., Walker, N.J. and Horbelt, D.V. (1992) Ultrastructure of the microvasculature in the human endometrium throughout the normal menstrual cycle. Am. J. Obstet. Gynecol., 166, 1393–1406.[ISI][Medline]

Rogers, P.A.W. (1996) Structure and function of endometrial blood vessels. Hum. Reprod. Update, 2, 57–62.[Abstract/Free Full Text]

Salle, B., Bied-Damon, V., Benchaib, M. et al. (1998) Preliminary report of an ultrasonography and colour Doppler uterine score to predict uterine receptivity in an in-vitro fertilization programme. Hum. Reprod., 13, 1669–1673.[Abstract]

Schild, R.L., Indefrei, D., Eschweiler, S. et al. (1999) Three-dimensional endometrial volume calculation and pregnancy rate in an in-vitro fertilization programme. Hum. Reprod., 14, 1255–1258.[Abstract/Free Full Text]

Schwartz, L.B., Chiu, A.S., Courtney, M. et al. (1997) The embryo versus endometrium controversy revisited as it relates to predicting pregnancy outcome in in-vitro fertilization-embryo transfer cycles. Hum. Reprod., 12, 45–50.[ISI][Medline]

Steer, C.V., Mills, C.L., Tan, S.L. et al. (1992) The cumulative embryo score: a predictive scoring technique to select the optimal number of embryos to transfer in an in-vitro fertilization and embryo transfer programme. Hum. Reprod., 7, 117–119.[Abstract]

Yuval, Y., Lipitz, S., Dor, J. and Achiron, R. (1999) The relationship between endometrial thickness, and blood flow and pregnancy rates in in-vitro fertilization. Hum. Reprod., 14, 1067–1071.[Abstract/Free Full Text]

Zaidi, J., Campbell, S., Pittrof, R. and Tan, S.L. (1995) Endometrial thickness, morphology, vascular penetration and velocimetry in predicting implantation in an in vitro fertilization program. Ultrasound Obstet. Gynecol., 6, 191–198.[ISI][Medline]

Submitted on June 9, 1999; accepted on October 7, 1999.