The relationships between endometrial thickness, and blood flow and pregnancy rates in in-vitro fertilization

Y. Yuval1, S. Lipitz1, J. Dor2 and R. Achiron1,3

1 Diagnostic Ultrasound and 2 IVF Units, Department of Obstetrics and Gynecology, The Chaim Sheba Medical Center, Tel Hashomer, and The Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
To establish whether endometrial blood flow and thickness can predict the success rate of in-vitro fertilization, 156 cycles were evaluated. The parameters were: endometrial colour and power Doppler pulsatility index (PI), resistance index (RI), systolic/diastolic ratio (S/D) and endometrial thickness. Each patient was studied: on the day of ovum retrievalpickup, and on the day of embryo transfer. Non-endometrial parameters studied were: age, oestrestrogen and progesterone concentrations, number of oocytes, and number of embryos. Pregnancy was achieved in 31 cycles. On the day of ovum retrieval, patients who conceived had PI, RI, and S/D values of 0.997, 0.563, and 2.403, respectively. Patients who did not conceive had values of 0.994, 0.584, and 2.477 respectively. The power Doppler technique provided similar results. On the day of embryo transfer, pregnant patients had PI, RI and S/D values of 1.096, 0.590 and, 2.597 respectively, while in the non-pregnant patients the values were 1.104, 0.603 and, 2.723 respectively. Power Doppler showed similar numbers. The differences between pregnant and non-pregnant patients were not statistically significant in any of the parameters. Endometrial thickness and blood flow does not seem to correlate with pregnancy rate in IVF.

Key words: blood flow/endometrium/IVF/pregnancy-rate


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Today, the main obstacle in fertility treatment is the low rate of implantation in cycles where good embryos have been transferred to the uterus. The percentage of implantation is approximately 15% (Edwards and Craft, 1990Go; SART and ASRM, 1996Go). As yet, the reason for this low rate remains unclear.

The changes in the endometrial morphology and blood flow during the monthly cycle, can easily be identified by ultrasound examination. It has been found that the resistance to flow, which is inversely correlated to tissue perfusion, drops towards ovulation and during the luteal phase, supposedly in preparation for implantation (Achiron, 1995; Kupesic and Kurjak, 1993Go; Achiron et al., 1995Go). Both uterine arteries, as well as spiral arteries, were assessed.

In infertile women, non-uniform results were reported in different studies. The bulk of information relates to flow measurements in the uterine arteries and vessels within the myometrium or endometrium. Steer et al. (Steer et al., 1994Go) have shown that certain defects in fertility are connected with sub-optimal perfusion of the uterus in the mid-luteal phase. Similar findings to those of Steer et al. (1994) regarding uterine arteries were demonstrated by Kurjak et al. (Kurjak et al., 1991Go). Goswamy et al. showed a possible relationship between infertility and decreased uterine perfusion via the uterine arteries (Goswamy et al., 1998).

Studies that have attempted to predict implantation success by endometrial morphology have failed to show positive results. Remohi et al. found a relationship between endometrial thickness and oestradiol concentration, but neither of these, either alone or combined, could predict outcome (Remohi et al., 1997Go). No threshold was defined: endometrial thickness < 4 mm, as well as oestradiol concentration < 50 pg/ml did not preclude pregnancy. Oliveira et al. endeavoured to predict success in in-vitro fertilization (IVF) by grouping the endometrium according to three types of appearances (Oliveira et al., 1997Go). However, a positive predictive value was not established. An endometrial thickness of 7 mm minimum was found necessary for implantation success.

Sterzik et al. demonstrated that pregnant women had a significantly lower resistance in the uterine arteries on the day of ovum pickup (OPU) (Sterzik et al., 1989Go). Strohmer et al. also published similar findings (Strohmer et al., 1991Go). Steer et al. (1992) determined a threshold pulsatility index (PI) value for the uterine arteries. Other authors also showed threshold values (Favre et al., 1993Go). Zaidi et al, established a rough value: presence (good chance for implantation) or absence (poor chances for implantation) of identifiable blood flow in the region of the endometrium and sub-endometrium (Zaidi et al., 1995Go).

The Physics of the Doppler shift is simple and well known due to its ubiquitous use in diagnostic clinical practice. Power Doppler is a recently introduced technology (Murphy and Rubin, 1997Go); it has a 3-fold greater sensitivity to flow than colour Doppler, and is particularly useful for minute vessels with low-velocity flow. Power Doppler is less dependent on the angle between the beam and vessel, and in areas of multiple vessels, the signal is additive, and opposite flow directions do not cancel each other's signal. The increased sensitivity is not dependent on increased energy transmitted from the machine to the tissue. Hence, safety is not compromised. Data regarding the use of this technique in predicting the success of in-vitro fertilization (IVF) are lacking.

Due to the controversy and lack of standard criteria in the literature in examining blood flow in the endometrium itself, it was decided to examine the data from the cases treated in our ward. The intention was to investigate whether success rates of IVF could be predicted by using the colour and Power Doppler technique by measuring resistance to flow in the endometrium (using currently available formulaes), and endometrial thickness.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
From September 1996 through February 1997, 192 treatment cycles that resulted in embryo transfer at the IVF Unit, The Chaim Sheba Medical Center, were available for study. Ten patients refused examination, and 17 had severe distortion of the uterine cavity (submucous myoma, congenital anomaly) and were thus excluded. All patients were examined by Y.Y.: 19 were not examined on the day of ovum retrieval, and 29 were not examined on the day of embryo transfer. Fifteen patients underwent frozen–thawed embryo cycles, only one of whom conceived. Two patients had ectopic pregnancies, and seven had chemical pregnancies. Due to the small numbers of patients in the ectopic and chemical pregnancy groups, they were excluded. The remaining 156 cycles (155 patients) comprised the two groups: patients who conceived and reached at least 13 weeks of gestation (n = 31) and those who did not (n = 125).

The patients and treatment staff were blinded to the results. Each woman was examined prior to the procedure on the day of ovum retrieval and the day of embryo transfer. Each patient received a short explanation regarding the procedure and its safety, and provided their verbal consent. The common protocol in our IVF Unit is pituitary suppression with a gonadotrophin-releasing hormone-agonist (GnRHa) and ovarian hyperstimulation with follicle stimulating hormone (FSH)-based menotrophins. Data analysis did not take into account the specific protocol, but rather the endpoints: oestradiol (E2) and progesterone (P) blood concentrations, number of oocytes aspirated and embryos transferred. All patients were examined after spontaneous emptying of the urinary bladder, lying supine with the knees slightly bent, and with a small pillow under the buttocks. Patients were examined vaginally with an ATL (HDI 9) 5 MHz transducer. Wall filter was set to 25–50 Hz, and the limit of aliasing 1 cm/sec.

At present, there is no proof that obstetrical sonography can be teratogenic. Nevertheless, the `ALARA' (as low as reasonably achievable) principle was employed in order to minimizse acoustic energy transferred to the tissue examined. In all cases Isptad (spatial peak time average intensity – derated) was <50 mW/cm2. This is well below the limits set by the Food and Drug Administration, USA. Other parameters that are known to influence acoustic output, and thus were controlled accordingly, were: Time of examination that did not exceed a total of 500 s seconds; zooming was hardly used, and for not more than 30 s seconds; 2-dimensional (2-D), colour and Doppler gains were set to the minimum, facilitating an analysable picture. A better picture with lower frame rate was achieved by setting three or more focal zones. Sector angle was as wide as possible, and narrowing was limited for a few seconds. Sample volume was always not more than 1–2 mm.

The uterus and endometrium were scanned, and the double thickness of the endometrium measured. Thereafter, a colour flow map was superimposed on the 2-D picture and Doppler studies were performed on selected areas. Intra-endometrial or the adjacent sub-endometrial region was studied. Spectral analysis was made and the lowest values for resistance to flow were selected as representative. The parameters were (Jaffe and Warsof, 1992Go): (i) resistance index (RI): the difference between maximal systolic blood flow and minimal diastolic flow divided by the peak systolic flow (S-D/S); (ii) pulsatility index (PI): the difference between maximal systolic blood flow and minimal diastolic flow divided by the mean flow throughout the cycle (S – D/mean); (iii) the ratio between peak systolic flow and lowest diastolic flow (S/D). These three parameters express the resistance to flow from the point of measurement downstream. The value increases when resistance increases, and vice versa. The diastolic flow is considered to be influenced by resistance to a greater extent than the systolic flow. To date there is no accord among investigators regarding the most appropriate parameter for any given clinical situation. All parameters were also examined by using the power Doppler system. Cases in which no blood flow could be identified received artificially extreme values in order to be incorporated in the statistical analysis: maximal (S/D was defined as 6.0, maximal RI was defined as 1.0, and maximal PI as 3.0.

Patients were divided into two groups: successful outcome, defined as pregnancy reaching 13 weeks, and failure of implantation, where no pregnancy was detected. Seven pregnancies that did not reach the end of the first trimester and two ectopics were excluded from the study. Such small numbers do not warrant a statistical analysis.

Each parameter which defines resistance to flow possesses two characteristics: (i) A continuous parameter that is expressed in a table using average and SD; (ii) a dichotic parameter – above or equal to the median in the group as a whole was defined as `high value', and under median as `low value'. Comparison of parameter averages, which define resistance to flow between the cases resulting in pregnancy and those who failed to conceive, was performed using the t-test for unmatched samples. Comparison of the pregnancy rates in the group with the `high values' of resistance to the group with `low values' of resistance was performed by using the {chi}2, or Fisher exact test in order to detect whether `high' resistance was associated with lower pregnancy rates. In an effort to establish the factors associated with the success of implantation, multivariate analysis was performed based on logistic regression. The dependent variable was pregnancy versus non-pregnancy. The independent variables were the resistance to flow and the thickness of the endometrium on the day of ovum retrieval and embryo transfer, as well as age, oestradiol and progesterone concentrations on the day of human chorionic gonadotrophin (HCG), (Chorigon; Teva Pharmaceuticals, Petah Tikva, Israel) administration, number of oocytes recovered, and number of embryos transferred.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
A total of 165 cycles (164 patients) were studied. After excluding nine ectopic and chemical pregnancies (see above), 156 cycles remained. Pregnancy was achieved in 31 cycles, while 125 were defined as failures.

Tables I–VGoGoGoGoGo depict the data comparing women who conceived to those who did not: non-endometrial parameters (Table IGo), flow indexes and endometrial thickness on the day of ovum retrieval (Table IIGo), and on the day of embryo transfer (Table IIIGo), the difference between embryo transfer value and ovum retrieval value (Table IVGo), and pregnancy rate in patients with endometrial parameters above and below the median (Table VGo).


View this table:
[in this window]
[in a new window]
 
Table I. Comparison of non-endometrial parameters
 

View this table:
[in this window]
[in a new window]
 
Table II. Comparison of flow indices and endometrial thickness on day of ovum retrieval
 

View this table:
[in this window]
[in a new window]
 
Table III. Comparison of flow indices and endometrial thickness on day of embryo transfer
 

View this table:
[in this window]
[in a new window]
 
Table IV. Embryo transfer–ovum retrieval difference in endometrial parameters
 

View this table:
[in this window]
[in a new window]
 
Table V. Pregnancy rates for endometrial parameters above and below median values on ovum retrieval day
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
In this study, no statistically significant difference was noted between the two groups except for number of embryos (4.8 and 3.9 for pregnant and non-pregnant respectively, P < 0.05). No relationship was found between the differences in the values on the day of ovum retrieval versus the day of embryo transfer. No threshold value was found for resistance in any of the parameters that could indicate a line beyond which rates of pregnancy changed significantly. The opinion that total lack of detectable diastolic flow can predict failure of implantation was not supported. Even in the small minority of women in whom no flow signals were received at all, pregnancies were achieved. A suspected trend for the increase in the possibility for pregnancy in accordance with decreased resistance was noted in some parameters. In all patients, a trend for increased resistance, as well as an increase in endometrial thickness from the day of ovum retrieval to the day of embryo transfer was noted. Theoretical statistical analysis showed that by increasing the sample size to 200 patients in each group, while leaving the average and SD unchanged, statistical significance between the groups would be attained. Another finding which suggests that increased resistance is important can be found in the tables: all parameters, except one (power RI), are lower in the group that achieved pregnancy. No threshold value could be obtained, so that from this value forward there is an increase/decrease in the pregnancy rate. Division into medians (Table VGo) and even thirds (not shown), did not improve power of analysis. Furthermore, the difference from the day of ovum retrieval to the day of embryo transfer does not predict treatment success, with one exception: in the group that failed to conceive, the difference between the values of resistance as noted with the power S/D method were statistically significant. It appears that a significant rise in resistance to flow from the ovum retrieval to the day of embryo transfer indicates an endometrium that is less receptive. This finding, which shows no clinical application as yet, should be further investigated. From the other parameters that were examined, i.e. age of women, concentrations of oestradiol and progesterone on the day of HCG administration, and number of oocytes aspirated, none was found to forecast the likelihood of pregnancy. This is already known from the literature. The number of embryos transferred was found to be a good predictor for the possibility of pregnancy, a well known fact (SART and ASRM, 1996Go). Each embryo adds to this possibility by a factor of 1.4.

Researchers agree that there is a trend towards an increase in uterine perfusion during the spontaneous monthly cycle. Maximal resistance to flow is obtained during the end of the luteal phase and menstruation. Medium values are received in the follicular phase. The nadir of resistance is seen in the mid-luteal phase, when implantation is expected, and corpus luteum activity is maximal. So far the connecting thread between the physiological (and logical) pattern just described and pathophysiology of infertility has not been established in a way that may be applied to clinical practice.

The relationship of hormonal imbalance and infertility to suboptimal perfusion has been established (Kurjak et al., 1991Go; Steer et al., 1994Go), and it seemed reasonable to correlate perfusion to success in IVF. Uterine artery flow was found to correlate with pregnancy rates (Goswamy et al., 1988Go; Sterzik et al., 1989Go), if measured around the time of ovulation. Steer et al., (1992) claimed that a PI (33.0) threshold exists, above which implantation is unlikely. A higher threshold (PI 3.55) had been set (Favre et al., 1993Go) but found no correlation between the PI as a continuous variable and pregnancy rates was found It seemed that less resistance and a higher oestradiol concentration are interrelated, but do not predict implantation. It may be that perfusion, if excessive, is detrimental. The answer may lie in the concept of a `window` between two values of resistance, where a better chance for implantation exists.

Measurement of flow in the spiral arteries is difficult. Taylor et al. (1985) did not examine these, as the flow in the spiral arteries could not be correlated during laparotomy. Long et al. (1989) was unsuccessful in attempting to demonstrate the flow in the spiral arteries. This may have been due to the sensitivity of the apparatus in the abdominal approach. Most researchers examined uterine perfusion through the uterine arteries.

Coulman et al. (1994) made an effort to combine PI in uterine arteries with morphological endometrial parameters, but could not predict success. However, a PI >3.3 and endometrial thickness <6 mm were poor prognostic indices for implantation. A later study (Coulman et al., 1995Go) compared the PI on the day of ovum retrieval to that on the day of embryo transfer. The difference between these two measurements was not significant (similar to that of our results). Endometria in oocyte donation cycles were studied (Bustillo et al., 1995Go). The PI and RI, as well as morphological parameters, were measured. Similarly, high PI values were associated with a poor prognosis; however, their data do not contain numerical values.

To date five studies have been examining the flow in the spiral arteries. Kupesic and Kurjak (1993) studied the spiral arteries near ovulation. They compared spontaneous to hyperstimulated cycles in healthy women treated for male infertility. All were examined vaginally. A medium degree of resistance was found before ovulation (PI 1.32) and a nadir one day after ovulation (average PI 0.72). During the cycles with ovulation induction higher values were obtained prior to (average PI 1.32) as well as after (average PI 1.09) ovulation. The authors concluded that the treatment causes a disturbance in uterine perfusion, and that possibly better treatment may improve flow.

In this study, uterine arteries were also examined. Here too, the values showed the same trend as the spiral arteries. The second study (Achiron et al., 1995Go) examined the change in perfusion of the uterine endometrium in menopausal women due to premature ovarian failure. This research was part of a treatment programme for egg donation. Ultrasonographic examination was performed vaginally, a colour flow map was obtained for all those examined, and healthy women were the controls. Examination of perfusion was performed only 1 week prior to treatment and not during the cycles of women not treated. The researchers found a high resistance to flow in the phase prior to treatment (that manifested as absent end diastolic flow) and at the beginning of treatment (early follicular phase; average RI 0.85). Towards ovulation, a significant decrease in resistance was observed (average RI 0.57) with a slight increase after ovulation (average RI 0.67). The medium resistance remained throughout the luteal phase.

Similar values were obtained from healthy women, except for the early follicular phase, in which healthy women had lower average scores (RI 0.68 compared to 0.85). In the third study (Zaidi et al., 1995Go) the blood flow in the sub-endometrial region, and in the endometrium itself was studied. No correlation was found between resistance and pregnancy rates in IVF cycles, except for the fact that the absence of flow predicts poor chances for pregnancy.

Recently, a fourth study was published (Brown et al. 1997Go). Only 10 patients were studied and the results were inconclusive. The last study (Battaglia et al., 1997Go) showed that the PI values on the day of ovum retrieval were significantly lower in patients who became pregnant, compared to those who did not. This is the only study in the literature with contradictory results. The difference can be attributed to: (i) the small number of patients; (ii) patients with no colour flow detected were excluded from the study, thereby creating a selection bias for those with better perfusion of the endometrium; the number of cases excluded was not specified; (iii) their results require further verification, as they are not supported by others. Zaidi et al. (1995) found a PI of 1.3, Kupesic and Kurjak (1993) a result of 1.3, and our results were 0.9 and higher. The article of Battaglia et al. (1997) shows that the PI in pregnant patients was 0.6–0.7. Ours is the largest study that has aimed at examining the relationship between blood flow in the endometrium itself and rates of success in IVF cycles.

In summary, endometrial perfusion is, apparently, an important factor in the process of implantation. At present, it is not possible to predict the chances of implantation based on measurements of resistance to flow. As this procedure is known to reduce implantation rates (SART and ASRM, 1996Go), women with poor perfusion cannot be advised to freeze embryos and to postpone transfer to later cycles. Most probably, better prediction of the success rate in IVF will be obtained after larger studies that integrate a group of parameters including, among others, sonographic and endocrinological data.


    Notes
 
3 To whom correspondence should be addressed at: Department of Obstetrics and Gynecology, The Chaim Sheba Medical Center, Tel Hashomer 52621, Israel. Tel: 972-3-5302890 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]

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]

Brown, J.M., Schwartz, L.B., Olive, D. et al. (1997) Evaluation of Doppler ultrasonography as a means of monitoring in vitro fertilization and embryo transfer cycles: preliminary results and findings. J. Ultrasound Med., 16, 411–416.[Abstract]

Bustillo, M., Krysa, L.W. and Coulman, C.B. (1995) Uterine receptivity in an oocyte donation programme. Hum. Reprod., 10, 442–445.[Abstract]

Coulman, 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]

Coulman, C.B., Stern, J.J., Soenksen, D.M. et al. (1995) Comparison of pulsatility indices on the day of oocyte retrieval and embryo transfer. Hum. Reprod., 10, 82–84.[Abstract]

Edwards, R.G. and Craft, I. (1990) Development of assisted conception. Br. Med. Bull., 46, 565–579.[Abstract]

Favre, R., Bettahar, K., Grange, G. et al. (1993) Predictive value of transvaginal uterine Doppler assessment in an in vitro fertilization program. Ultrasound Obstet. Gynecol., 3, 350–353.[ISI][Medline]

Goswamy, R. K., Williams, G. and Steptoe, P.C. (1988) Decreased uterine perfusion: a cause of infertility. Hum. Reprod., 3, 955–959.[Abstract]

Jaffe, R. and Warsof, S.L. (1992) Color Doppler Imaging in Obstetrics and Gynecology. McGraw-Hill, New York.

Kupesic, S. and Kurjak, A. (1993) Uterine and ovarian perfusion during the periovulatory period assessed by transvaginal color Doppler. Fertil. Steril., 60, 439–443.[ISI][Medline]

Kurjak, A., Kupesic-Urek, S., Schulman, H. and Zalud, I. (1991) Transvaginal color flow Doppler in the assessment of ovarian and uterine blood flow in infertile women. Fertil. Steril., 56, 870–873.[ISI][Medline]

Long, M.G., Boultbee, J.E., Hanson, M.E. and Begent, R.H.J. (1989) Doppler time velocity waveform studies of the uterine artery and uterus. Br. J. Obstet. Gynecol., 96, 588–593.[ISI][Medline]

Murphy, K.J. and Rubin, J.M. (1997) Power Doppler: it's a good thing. Semin. Ultrasound, CT, and MRI. 18, 13–21.

Oliveira, J.B.A., Baruffi, R.L., Mauri, A.L. et al. (1997) Endometrial ultrasonography as a predictor of pregnancy in an in-vitro fertilization programme after ovarian stimulation and gonadotrophin-releasing hormone and gonadotrophins. Hum. Reprod., 12, 2515–2518.[Abstract]

Remohi, J., Ardiles, G., Garcia-Velasco, J.A. et al. (1997) Endometrial thickness and oestradiol concentrations as predictors of outcome in oocyte donation. Hum. Reprod., 12, 2271–2276.[Abstract]

SART and ASRM (1996) Assisted reproductive technology in the United States and Canada: 1994 results generated from the American Society for Reproductive Medicine/Society for Assisted Reproductive Technology Registry. Fertil. Steril., 66, 697–705.[ISI][Medline]

Steer, C.V., Campbell, S., Tan, S.L. et al. (1992) The use of transvaginal color flow imaging after in vitro fertilization to identify optimum uterine conditions before embryo transfer. Fertil. Steril., 57, 372–376.[ISI][Medline]

Steer, C.V., Tan, S.L., Mason, B.A. and Campbell, S. (1994) Midluteal-phase vaginal color Doppler assessment of uterine artery impedance in a subfertile population. Fertil. Steril., 61, 53–58.[ISI][Medline]

Sterzik, K., Grab, D., Sasse, V. et al. (1989) Doppler sonographic findings and their correlation with implantation in an in vitro fertilization program. Fertil. Steril., 52, 825–828.[ISI][Medline]

Strohmer, H., Herczeg, C., Plöckinger, B. et al. (1991) Prognostic appraisal of success and failure of an in vitro fertilization program by transvaginal Doppler ultrasound at the time of ovulation induction. Ultrasound Obstet. Gynecol., 1, 272–274.[ISI][Medline]

Taylor, K.J.W., Burns, P.N., Wells, P.N.T. et al. (1985) Ultrasound Doppler flow of the ovarian and uterine arteries. Br. J. Obstet. Gynaecol., 92, 240–246.[ISI][Medline]

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

Submitted on February 20, 1998; accepted on December 22, 1998.