- Email: [email protected]
An analysis of factors influencing the establishment of a clinical pregnancy in an ultrasound-based ambulatory in vitro fertilization program
FERTILITY AND STERILITY Copyright e 1988 The American Fertility Society
Vol. 49, No. 3, March 1988 Printed in U.S.A.
An analysis of factors influencing the establishment of a clinical pregnancy in an ultrasound-based ambulatory in vitro fertilization program
Vinay Sharma, M.R.C.O.G.* Andrew Riddle, M.R.C.O.G. Bridgett A. Mason, M.B.B.S.
Julian Pampiglione, M.R.C.O.G. Stuart Campbell, F.R.C.O.G.
The Hallam Medical Centre and King's College School of Medicine and Dentistry, London, United Kingdom
In the study period (June 1984 to December 1986), 2232 cycles were stimulated in 1294 patients. Ultrasound-directed oocyte recovery (UDOR) was performed as an ambulatory procedure in 1737 (77.8%) cycles, resulting in 1375 embryos transfers (ET). Age, etiology, menstrual cycle length, number of oocytes collected, and number of embryos transferred were important determinants of the outcome. The number of attempts at in vitro fertilization did not affect the clinical pregnancy rate (CPR). In patients receiving four embryos, the CPR appeared to be highest when up to seven embryos were available for transfer. The fertilization rate in an individual cycle had a good prognostic value, the implantation rate being highest when 7 to 9 oocytes were retrieved and >60% of these were fertilized. When 10 or more oocytes were collected, the implantation rate showed a progressive decline, regardless of the fertilization rate. Furthermore, multiple pregnancies failed to occur when >12 oocytes were retrieved or more than eight embryos were available for transfer. These data suggest that, in excessively stimulated cycles, the quality of oocytes and embryos or uterine receptiveness may be suboptimal, and the transfer of more than four embryos is unlikely to increase the success rate. Fertil Steril 49:468, 1988
We have recently reported1 the results of our first 2 years of experience in a large ambulatory in vitro fertilization (IVF) program. We herein present a more detailed analysis of 2232 cycles that were stimulated between June 1984 and December 1986, relying almost completely on ultrasound for follicular monitoring and oocyte recovery. Contrary to earlier doubts, 2 our results have shown that this method is not only safe, but has achieved a clinical pregnancy rate (CPR) similar to the longer-established laparoscopy programs. 3- 5 Numerous variables, such as etiology, selection of patients, agents for stimulation, methods of monitoring follicular development, technique of oocyte recovery, and laboratory protocol, may influence the Received June 2, 1987; revised and accepted November 4, 1987. *Reprint requests: Vinay Sharma, M.R.C.O.G., Hallam Medical Centre, 77 Hallam Street, London WIN 5LR, England. 468
outcome of treatment. In this article we present the influence of some of these factors on the establishment of a clinical pregnancy. MATERIALS AND METHODS
All patients treated between June 1984 and December 1986 have been included in this analysis. Patients in all age groups were accepted in the first 6 months of our program. Because no pregnancies were obtained in patients above 40 years, we did not enroll them in the program after January 1985. Only those investigations that had been omitted prior to referral and were considered essential for an accurate diagnosis were performed before treatment. Etiologic Definitions
Patients were considered to have tubal damage if the tubes were swollen, blocked, or fixed, as proven
Sharma et al. Factors affecting outcome in an IVF program
Fertility and Sterility
by laparoscopy and hysterosalpingogram. If endometriotic deposits were identified at laparoscopy but there was no tubal damage or ovarian involvement, then the patients were considered to have mild endometriosis. If the endometriotic deposits involved the tubes and/or ovary, then a classification of severe endometriosis was made. All husbands had a routine semen analysis and a mixed antiglobulin reaction (MAR) test performed initially. The Kibrick and Trey agglutination test (TAT) was only performed in men if the MAR test was positive and in women if the postcoital or postinsemination test was negative. If either or a combination of the following were present-i.e., the sperm density was <20 X 106 /ml, sperm motility <40%, the number of abnormal sperm >60%, and/ or MAR test was positive-then the couple was diagnosed as having male factor infertility (MFI). However, if only antibodies were present in either partner, then the couple was designated as having an immune factor responsible for infertility. Any combination of abnormalities in the male or female partner was designated mixed male and female factor infertility. If no abnormalities were detected in either of the partners, then the couple was said to have unexplained infertility. Ovarian stimulation was performed in most women with a combination of clomiphene citrate (CC) and human menopausal gonadotrophin (hMG; Pergonal, Serono Laboratories, Inc., Welwyn Garden City, UK) or pure follicle-stimulating hormone (pure FSH; Metrodin, Serono Laboratories, Inc.) CC was always administered from the second to the sixth day of the menstrual cycle (100 mg orally per day). In the first treatment cycle, gonadotropin administration was commenced either on the second or the fourth day of the menstrual cycle, depending on the usual cycle length. Patients more than 36 years of age received 225 IU, and younger patients were given 150 IU gonadotrophin daily. In subsequent cycles, the daily dose was adjusted according to the folliculogenic response obtained previously. Monitoring was performed both ultrasonically (Diasonics DRF-1, Bedford, England, 3.5 MHz GPM-11 probe; Phillips SDR 1500, Hammersmith, London, UK, 3.5 and 5.0 MHz abdominal probes; International General Electric RT3000, Slough, Berkshire, UK, 3.5 and 5.0 MHz abdominal probes) and biochemically. The first ultrasound scan was performed on the second day of the patient's menstrual cycle when any abnormalities, such as ovarian cysts and fibroids, were noted. Daily ultrasound Vol. 49, No. 3, March 1988
scans were performed after the eighth day. All follicles were visualized in the transverse and the longitudinal planes, and the maximum vertical and transverse diameters measured. The arithmetic mean of these four measurements was recorded. The patients also were advised to collect 24-hour urine samples from the second to the third day, and daily after the seventh day of their menstrual cycle. Total urinary estrogen excretion was measured by the method described by Brown et al. 6 The criteria for the administration of human chorionic gonadotrophin (hCG; 5000 IU, Serono Laboratories, Inc.) are described in detail elsewhere.1 Ultrasound-directed oocyte recovery (UDOR) was performed as an ambulatory procedure 34 to 35 hours after the hCG was administered. The patients were advised to drink sufficient water in order to fill the bladder before oocyte recovery. They arrived 30 minutes before the scheduled time and usually left 2 hours after the procedure was complete. Demerol (Winthrop-Breon Laboratories, New York, NY), diazepam, and prochlorperazine were administered intramuscularly for sedation 30 minutes before the procedure in doses appropriate to the patient's body weight. During the procedure, abdominal scanning was performed in the transverse plane and the aspiration needle (gauge 16, Hallam Medical Centre IVF needle, Rocket of London Ltd., Watford, England) was introduced transcutaneously parallel to the ultrasound beam using the "free-hand" technique. Ovaries that were located lateral to the uterus were approached transvesically. Follicles in subcutaneous ovaries were aspirated by the direct transabdominal route. Ovaries located posterior to the body of the uterus in the pouch of Douglas were aspirated transvaginally, the needle being visualized with an abdominal probe. Oocytes with a fractured zona, darkened cytoplasm, or an abnormal shape were not considered to be viable. All oocytes were examined 18 hours after insemination and, if two pronuclei were visualized, fertilization was thought to have occurred. Only those embryos that had undergone at least one mitotic division after confirming fertilization (cleaved embryos) were transferred 48 to 60 hours after oocyte recovery. The technique of embryo transfer (ET) and luteal support have been described in detail elsewhere. 1 Patients were designated as being clinically pregnant when the amenorrhea continued for more than 6 weeks from the first day of the last men-
Sharma et al. Factors affecting outcome in an IVF program
469
Table 1
Age Distribution
Age groups
Jun 1984 to Dec 1984
Jan 1985 to Jun 1985
Jul1985 to Dec 1985
Jan 1986 to Jun 1986
Jul 1986 to Dec 1986
Jun 1984 to Dec 1986
Total cycles .;;25 years (%) 26-30 years(%) 31-35 years(%) 36-40 years(%) >40 years(%)
250 16 (6.4) 49 (19.6) 91 (36.4) 82 (32.8) 12 (4.8)
373 17 (4.6) 70 (18.8) 158 (42.4) 123 (33.0) 5 (1.3)
401 16 (4.0) 80 (20.0) 173 (43.1) 129 (32.2) 3 (0.7)
572 19 (3.3) 121 (21.2) 244 (42.7) 186 (32.3) 3 (0.5)
636 21 (3.3) 138 (21.7) 288 (45.3) 187 (29.4) 2 (0.3)
2232 89 (4.0) 458 (20.5) 955 (42.7) 705 (31.6) 25 (1.1)
237 11 (4.6) 48 (20.3) 106 (44.7) 71 (30.0) 1 (0.4)
371 13 (3.5) 88 (23.7) 165 (44.5) 106 (28.3)
396 12 (3.3) 96 (24.2) 180 (45.5) 107 (27.0) 1 (0.3)
1375 55 (4.0) 325 (23.7) 602 (43.8) 387 (28.1) 6 (0.4)
Proportion of patients receiving an embryo replacement 232 Total cycles 138 8 (5.8) 11 (4.7) .;;25 years (%) 33 (23.9) 60 (25.9) 26-30 years(%) 53 (38.4) 98 (42.2) 31-35 years(%) 41 (29.7) 62 (26.7) 36-40 years(%) 3 (2.2) 1 (0.4) >40 years (%)
strual period and a definite gestational sac was identified on ultrasound scanning. Those patients who had bleeding before the completion of 6 weeks of ammenorrhea, but had a plasma hCG level above 25 IU/ml, a minimum of 16 days after the oocyte recovery and 10 days after the last hCG injection, were diagnosed as having had a preclinical pregnancy. Age Distribution
All patients were divided into four age groups: under 25 years, 26 to 30 years, 31 to 35 years, and 36 to 40 years. There was no significant difference in the distribution of age groups during the study Table 2
period. Similarly, the age distribution remained uniform among the patients receiving an ET (Table 1). Etiologic Distribution
There was no alteration in the distribution of the different etiologic categories treated over the last 2 years of the study period. Similarly, the distribution of these categories has not changed among the patients receiving an ET (Table 2). Statistical Methods
All statistical comparisons on the effect of age have been made between the groups 26 to 30 years
Etiologic Distribution Jan 1985 to Jun 1985
Jul1985 to Dec 1985
Jan 1986 to Jul 1986
Jul 1986 to Dec 1986
Jun 1984 to Dec 1986
Diagnosis
Jun 1984 to Dec 1984 %
%
%
%
%
%
Total cycles Tubal damage Unexplained infertility Mild endometriosis Severe endometriosis Male factor infertility Immune factor infertility Mixed male and female
250 30.8 25.2 10.0 4.0 10.8 4.0 15.2
373 42.6 26.3 4.3 8.3 5.6 1.6 11.3
401 44.6 28.2 5.5 8.0 5.2 2.5 6.0
572 48.1 27.1 5.2 6.1 3.8 4.9 4.7
636 47.8 28.0 6.6 5.2 4.6 4.2 3.6
2232 44.5 27.2 6.0 6.3 5.4 3.6 6.9
232 47.8 25.8 4.3 6.5 2.6 1.7 11.2
237 48.1 28.3 5.5 6.8 2.5 2.5 6.3
371 50.7 26.1 4.6 6.7 2.2 4.3 5.4
396 53.3 27.0 6.1 5.8 2.0 2.5 3.3
1375 49.0 26.3 5.7 6.0 2.9 3.2 6.8
Proportion of patients receiving an embryo transfer 138 Total cycles 36.2 Tubal damage 21.7 Unexplained infertility 10.9 Mild endometriosis 2.9 Severe endometriosis 8.7 Male factor infertility 5.8 Immune factor infertility 13.8 Mixed male and female
470
Sharma et al.
Factors affecting outcome in an IVF program
Fertility and Sterility
and 36 to 40 years. Most patients above the age of 40 years were treated in the first 6 months of our program, which was less successful. Their numbers were also comparatively small. Also, the diagnostic conditions leading to a less favorable outcome (severe endometriosis, MFI, immune factor, mixed male and female factor infertility) were higher in patients under 25 years (under 25 years, 33. 7%; 26 to 30 years, 20.9%; 31 to 35 years, 22.3%; and 36 to 40 years, 21.4%). In the analysis of the effect of menstrual cycle length, only patients with tubal damage and unexplained infertility have been included because the other categories contained variables that could alter fertilization rates or the ovarian cycle. In subsequent attempts, some patients had their treatment regimens modified according to their previous folliculogenic response, and this could bias results. Therefore, only first attempts have been analyzed. We have introduced a new method of assessing oocyte quality and hence the prediction of the pregnancy rate based on the results obtained in the laboratory. For this analysis, patients were divided into three groups: fertilization in more than 60% of the retrieved oocytes (group A), fertilization in 31% to 60% of oocytes (group B), and fertilization in <30% of oocytes (group C). The cleavage rate, implantation rate of transferred embryos (i.e., number of gestational sacs ..;- number of embryos transferred X 100), CPR per ET, and proportion of multiple pregnancies (i.e., number of ongoing multiple pregnancies ..;- total clinical pregnancies X 100) were calculated for each of these groups. Overall, chi-square tests have been used to assess
Table 3
Total cycles % Abandoned" % Reached UDORb %Successful UDOR' %Reached ETd Overall fertilization rate• (%) Overall cleavage rate!(%) Oocytes retrieved, median (range) Viable oocytes, median (range) Oocytes fertilized, median (range) Embryos cleaved, median (range) Embryos transferred, median (range)
.;;25 yr
6 5 4 3 3
89 16.9 82.0 78.7 61.8 63.4 81.2 (2-10) (2-8) (2-6) (1-5) (2-4)
79.86, 4 df, p < 0.001. 37.32, 4 df, P < 0.001. ' X2 = 64.5, 4 df, P < 0.001 (X 2 = 229.82, 4 df, P < 0.001 when compared with total cycles or those reaching UDOR). X2
b
x2 =
RESULTS A total of 2232 cycles in 1294 patients were stimulated between June 1984 and December 1986. Influence of Age
A significantly higher proportion of younger patients (Table 3) had a satisfactory response to gonadotropin stimulation, successful egg recovery, and ET. Conversely, a significantly higher proportion of patients in the 36- to 40-year age group were abandoned as compared with those in the 26- to 30-year age group. Significant heterogeneity also was observed when those cases enjoying successful UDOR were analyzed with respect to the total number of cycles and those which reached UDOR. Hence 26- to 30-year-old patients were more likely to have oocytes collected than those who were 36 to 40 years old. The overall fertilization and cleavage rates did not show significant variation with age. However, with age, there appears to be a progressive decrease in the mean number of total oocytes collected, viable oocytes, fertilized oocytes, and cleaved embryos (Table 3). The highest number of clinical pregnancies (Table 4) with respect to the total number of cycles
Influence of Age Clinical data
a
the significance of heterogeneities in frequency distributions among groups and subgroups. Partitioned chi-square tests for linear trends were calculated whenever necessary. Individual Yate's chisquare tests were performed on separate groups to investigate sources of heterogeneity suggested by high chi-square values. 7
=
Vol. 49, No. 3, March 1988
26-30 yr
31-35 yr
36-40 yr
>40 yr
458 14.4 85.6 84.5 71.0 69.1 81.8 6 (3-9) 5 (2-8) 4 (2-6) 3 (1-5) 3 (2-4)
955 19.8 80.2 78.8 63.1 67.8 81.8 5 (2-8) 5 (2-8) 3 (1-5) 3 (1-5) 3 (2-4)
705 29.5 70.5 68.2 54.8 69.6 80.6 4 (2-6) 4 (2-6) 3 (1-5) 3 (1-5) 2 (1-3)
25 64 36 36 24 70.6 75 2 (1-3) 2 (1-3) 2 (1-3) 1 (0-2) 1 (0-2)
d x2 = • X2 = 1 X2 =
Sharma et al.
63.4, 4 df, P < 0.001. 2.1, 4 df, NS. 0.92, 4 df, NS.
Factors affecting outcome in an IVF program
471
Table 4
Influence of Age Pregnancy Rates and Outcome ~25
Variable Pregnancy rates Total clinical pregnancies" CPR per successful UDORb (%) CPR per ET' (%) Pregnancy outcome Total pregnancies % Preclinicald % Clinical• % Anembryonic 1 % Miscarried8·h
yr
31-35 yr
36-40 yr
81 20.9 24.9
122 16.2 20.2
60 12.5 15.5
89 9.0 91.0 14.6 4.5
133 8.3 91.7 9.0 3.8
70 14.3 85.7 5.7 7.1
11
15.7 20.0 14 21.4 78.6 7.1
a X2 = 25.45, 3 df, P < 0.001 when compared with total number of cycles. "X 2 = 11.26, 3 df, P < 0.05 when compared with number ofETs. b X 2 = 2.6, 3 df, NS. ' x2 = 3.82, 3 df, NS.
and to the number of ETs was obtained in women who were in the 26- to 30-year age group, with the lowest among the 36- to 40-year-old group. There is no evidence of a significant change within the age groups in the CPR per UDOR or in the CPR per ET. Age had no significant effect on the proportion of pregnancies that were either preclinical, clinical, anembryonic, or misc,arried. Influence of Etiology
There was no significant influence of etiology (Table 5) on the number of abandoned cycles. Sim-
Table 5
26-30 yr
d
x2 =
1.62, 3 df, NS.
• X2 = 3.78, 3 df, NS. t x2 = 5.62, 3 df, NS. 6
h
X2 = 0.98, 3 df, NS. Pregnancies with fetal heart movements on ultrasound scan.
ilarly, the proportion reaching UDOR was not affected significantly. The proportion of patients reaching ET was heterogenous with respect to the etiology. This was due to a proportionately greater number of patients with tubal damage, unexplained infertility, mild and severe endometriosis, and fewer cases with MFI reaching ET. The overall fertilization rate was highest among patients with tubal damage and lowest among those with MFI. Similarly, the overall cleavage rate was highest among patients with tubal damage and lowest among those with MFI. No evidence of significant variation was found in the mean number of
Influence of Etiology Clinical data
Tubal damage
Unexplained infertility
Mild endometriosis
Severe endometriosis
Male factor
Immune factor
Mixed factors
Total cycles % Abandoned" % Reached UDORb %Reached ET' Overall fertilization rated (%) Overall cleavage rate• (%) Oocytes retrieved, median (range) Viable oocytes, median (range) Oocytes fertilized, median (range) Embryos cleaved, median (range) Embryos replaced, median (range)
994 20.3 79.7 67.8 74.9 83.4 5 (2-8) 5 (2-8) 3 (2-6) 3 (1-5) 3 (2-4)
607 22.1 77.8 59.5 65.6 80.1 5 (2-8) 5 (2-8) 3 (1-5) 3 (1-5) 3 (2-4)
135 29.6 70.3 58.5 71.9 79.8 4 (2-6) 4 (2-6) 3 (1-5) 3 (2-4) 3 (2-4)
141 8.3 71.6 58.9 72.0 79.8 4 (2-6) 4 (2-6) 3 (1-5) 3 (1-4) 3 (1-5)
120 25.8 74.2 33.3 36.5 65.7 5 (2-8) 4 (2-6) 3 (1-5) 3 (1-5) 2 (1-3)
81 24.7 75.3 54.3 55.0 76.6 5 (3-7) 4 (2-6) 3 (1-5) 3 (1-5) 3 (2-4)
154 17.5 82.5 61.0 59.4 83.0 5 (2-8) 4 (2-6) 3 (1-5) 3 (1-5) 2 (1-4)
127 12.8 18.8
77 12.7 21.3
21 15.6 6 26.5 6
10 8.3 25.0
6 7.4 13.6
22 14.3 23.4
Pregnancy rates Clinical pregnancies t CPR per cycle(%) CPR per ET (%)
X2 = 12.24, 6 df, NS, checked for continuity. X2 = 12.42, 6 df, NS, checked for continuity. 'X 2 = 75.66 compared with those reaching UDOR and 59.58 compared with total cycles, each 6 df, P < 0.001. d x2 = 87.14, 6 df, P < 0.001. a
b
472
Sharma et al.
11
7.8 8 13.3 8
• X2 = 21.1, 6 df, P < 0.01, Yates corrected X2 value for tubal damage 7.1, 1 df, P < 0.01 and male factor 17.2, 1 df, P < 0.001. t x2 = 9.06, 6 df, NS. 6 p < 0.001.
Factors affecting outcome in an IVF program
Fertility and Sterility
oocytes collected, viable oocytes, fertilized oocytes, cleaved embryos, or the number of embryos transferred among the different diagnostic categories (Table 5). There was no significant effect of etiology on the total number of clinical pregnancies. However, the CPR per cycle and per ET was significantly higher in patients with mild as compared with severe endometriosis. The CPR per cycle in patients with severe endometriosis and in those with MFI was similar, although the etiology in these two groups might be different. Whereas in MFI the etiology appears to be predominantly due to a lower fertilization rate, this was not the case in severe endometriosis, where the fertilization rate was normal but the implantation rate poor. The number of pregnancies in each group is not large enough to analyze the influence of etiology on pregnancy outcome. Influence of Menstrual Cycle Length
Menstrual cycle length or etiology (Table 6) did not influence the number of cycles analyzed, number of cycles abandoned, number of successful UDORs, fertilization rate, cleavage rate, or number of clinical pregnancies. The CPR in patients with a menstrual cycle of 26 days was lower than those
Table 6
with a normal menstrual cycle length. This could not be attributed to other factors, although the effect of early hMG administration could not be excluded in this analysis. Influence of the Number of Embryos Transferred
The CPR (Table 7) progressively increased from 8.6%, when one embryo was transferred, to 29.5%, when four embryos were transferred. This trend did not appear to continue among patients receiving more than four embryos, although this group is comparatively small. An upward trend was also seen in the incidence of blighted ova per ET. The incidence of preclinical pregnancies, miscarriages, and ectopic pregnancies was not associated with the number of embryos transferred. However, a significantly higher proportion of patients conceived when four or more embryos were transferred, as compared with two or three. A similar effect also was seen in the number of ongoing multiple pregnancies. The two-ET category had a lower and the four-ET group a higher multiple pregnancy rate, with no significant effect over the three-ET category, suggesting a consistent trend on either side of the latter group. Almost half of the multiple implantations diagnosed by ultrasound failed to progress in pregnancy.
Influence of Menstrual Cycle Length Unexplained infertility
Tubal damage
Cycles• (n) Mean age (years) Mean gonadotrophin dose, daily (IU) % Abandonedb No. of UDORsc (successful) Oocytes retrieved, median (range) Viable oocytes, median (range) Oocytes fertilized, median (range) Embryos cleaved, median (range) Embryos transferred, median (range) Fertilization rated(%) Cleavage rate• (%) Clinical pregnancies' (n) CPR per cycles CPR per UDOR h CPRperET; 0.25, 1 df, NS. 0.01, 1 df, NS. c X2 = 0.37, 1 df, NS. d X2 = 0.60, 1 df, NS. • x 2 = 0.01, 1 df, NS. • X2 = b 2 =
x
Vol. 49, No. 3, March 1988
Menstrual cycle .;;26 days
Menstrual cycle ~27 days
Menstrual cycle .;;26 days
Menstrual cycle ~27 days
115 33.1 ± 4.4 150-225 20 92 (91) 5 (2-8) 4 (1-7)
166 31.8 ± 4.2 150-225 18.7 135 (129) 5 (2-8) 4 (1-7) 3 (1-5) 3 (1-5) 3 (2-4) 69.2 86.0 22 13.3 17.0 20.4
71 33.2 ± 3.8 150-225
115 33.4 ± 3.9 150-225 20.0 92 (88) 5 (2-8) 4 (1-7) 4 (1-7) 3 (1-5) 3 (2-4) 70.4 69.1 12 10.4 13.6 22.2
4 (2-6)
3 (1-5) 3 (2-4) 75.3 82.2 6 5.2 6.6
8.0
21.1
56 (53) 5 (1-9) 4 (1-7) 4 (2-6)
3 (1-5) 3 (2-4) 70.8 84.2 7 9.9 13.2 16.3
t x2 = 0.684, 1 df, NS.
P < 0.05. p < 0.05. i p < 0.01.
6
h
Sharma et al.
Factors affecting outcome in an IVF program
473
Table 7
Influence of Number of Embryos Transferred
No. of embryos transferred
No. of embryo transfers
1 2 3 4 ;;;.5
245 322 317 468 23
Clinical" pregnancy rate perET
Preclinicalb pregnancy rate perET
(n)
8.6 15.2 18.9 29.5 26.1
(21) (49) (60) (138) (6)
Incidence of' blighted ova per ET
Miscarriaged rate per ET
(n)
(n)
(n)
0.8 1.6 2.8 3.2 8.7
(2) (5) (9) (15) (2)
0.8 0.9 0.9 3.0 4.3
(2) (3) (3) (14) (1)
2.5 0.6 0.9 4.3
(8) (2) (4) (1)
Ectopice rate per ET
Multiple' implantation rate per ET
MultipleR (ongoing) pregnancy rate perET
(n)
(n)
(n)
1.6 2.2 1.6 1. 7 8.7
(4) (7) (5) (8) (2)
3.7 8.8 14.1 13.0
(12) (28) (66) (3)
1.2 (4) 3.8 (12) 7.1 (33)
Multiple implantation rate per ET: number of patients with multiple implantations 7 number of embryo transfers X 100. Multiple ongoing pregnancy rate per ET: number of patients with multiple pregnancies 7 number of embryo transfers X 100.
x2 = x2 =
54.82, 4 df, p < 0.001. 9.06, 4 df, NS (2 X 2 Yates value for one embryo, 2.42, 1 df, NS; for four embryos, 1.47, 1 df, NS). 'X 2 = 9.13, 3 df, P < 0.05 (2 X 2 Yates value for four embryos, 7.57, 1 df, p < 0.01). d x2 = 4.82, 2 df, NS. a
b
e
X2
=
1.34, 4 df, NS.
t x2 = 24.13, 2 df, P < 0.001. g X2 = 15.6, 2 df, P < 0.001 (for two embryos, X2 = 9.85, 1 df, P < 0.001; for four embryos, X2 = 12.15, 1 df, P < 0.001; for three embryos, X2 = 0.245, 1 df, NS).
receive ET, and the pregnancy rate per ET were assessed individually for patients with 1 to 5 oocytes, and thereafter for groups with 6 to 10, 11 to 15, and above 16 oocytes. The overall fertilization rate did not show a significant change with the number of oocytes collected. Similarly, the cleavage rate showed little variation in the groups noted in the preceding paragraph. However, there was a linear increase in the number of patients reaching ET when four or more
Influence of the Number of Attempts
The pregnancy rate per cycle, per UDOR, or per ET did not show any significant change with the number of attempts (Fig. 1). Influence of the Number of Oocytes Collected
The number of oocytes collected after UDOR varied between 1 and 25 (Fig. 2). The fertilization rate, cleavage rate, proportion of cycles failing to PERCENTAGE
30
Pregnancy rate per
ET
20
UDOR - -
Cycle - -
10 Cycles
1295
575
227
80
34
UDOR
1004
452
175
60
28
18
ET
759
374
151
53
23
15
Preg.
156
74
25
11
4
4
3
4
21
0 2
5
6 or above
ATTEMPT NUMBER
Figure 1 Influence of the number of attempts at IVF on the pregnancy rate. CPR per cycle: X 2 = 3.42, 5 degrees of freedom (dO, not significant (NS); CPR per UDOR: X2 = 3.16, 5 df, NS; CPR per ET: X2 = 7.18, 5 df, NS.
474
Sharma et al.
Factors affecting outcome in an IVF program
Fertility and Sterility
PERCENTAGE 100 90
80 70
60 50 40 30
20 10
0 2
3
4
6-10
5
11-15
16 or above
NUMBER OF OOCYTES COLLECTED
-
Fertilization r a t e -
" cycles reaching ET
-
Cleavage rote
~
Pregnancy rate per ET
Figure 2 Influence of the number of oocytes collected on the outcome in an IVF cycle. Fertilization rate: X2 = 9.22, 7 df, NS; cleavage rate: X2 = 0.28, 7 df, NS; percentage of cycles reaching ET: linear increase from one to five oocytes, X2 = 78.79, 7 df, P < 0.01; partitioned: X2 = 71.64, 4 df, P < 0.01; pregnancy rate per ET: linear increase from one to five oocytes, X2 = 28.34, 7 df, P < 0.001; partitioned: X2 = 19.33, 4 df, P < 0.01.
oocytes were collected as compared with one, two, and three. Similarly, the CPR per ET showed a linear rise when one to five oocytes were retrieved, was highest when five oocytes were collected, and plateaued thereafter. Influence of Proportional Fertilization in Retrieved Oocytes
As shown in Table 8, the cleavage rate in the fertilized oocytes of the three groups (A, B, and C, defined under Materials and Methods) did not show any significant variation. The implantation rate of the transferred embryos did not show a statistically significant variation between groups when up to 6 oocytes were recovered. In group C, however, when 7 or more oocytes were collected, the implantation rate of embryos was lower when compared with the other two groups. In all groups, the implantation rate was lower when more than 10 oocytes were collected. Multiple pregnancies did not occur in group B when 1, 2, 3, or above 9 oocytes were retrieved. Conversely, when up to 9 oocytes were retrieved in group A, an increasing proportion of multiple pregnancies were obtained. Multiple pregnancies did not occur when more than 12 oocytes were collected, even though the cleavage rate showed little change. Vol. 49, No.3, March 1988
Influence of the Number of Cleaved Embryos
All patients who had four embryos transferred in the study period were included in this analysis (Fig. 3). They were grouped according to the number of cleaved embryos available for transfer. The CPR appeared to be lower when more than seven embryos were available for transfer. The multiple pregnancy rate also appeared to be lower when eight or more embryos were available, although this group is comparatively small. However, these results did not reach statistical significance. Similarly, the ongoing pregnancy rate did not increase when more than four embryos were available prior toET. DISCUSSION
As shown by our data and those of others, 4•8 the age of the female partner is an important determinant in the outcome of IVF. The success rate in the 36- to 40-year age group was lower in all diagnostic categories when compared with younger patients. The effects of maternal age on fecundity and fertility are complex and multifactorial. Progressive depletion of the ovarian follicular store occurs, hence a reduction in the follicular recruitment with advancing age. The production of defective oocytes
Sharma et al. Factors affecting outcome in an IVF program
475
Table 8
Influence of Proportional Fertilization in Retrieved Oocytes Group B 31-60% fertilization
Group A :»60% fertilization
Group C
~30%
fertilization
Implant• rate
CPR per ET
Proportb multiple
Cleavage rate
Implant rate
CPR per ET
Proport multiple
Cleavage rate
Implant rate
CPR per ET
Proport multiple
%
%
%
%
%
%
%
%
%
%
%
%
82.2 87.4 85.4 83.2 78.1 75.4 77.9
3.2 8.4 10.3 11.0d 8.6d 4.7d 9.2d
6.3 22.2 28.4 27.6 23.1 37.5
17.1' 23.8 28.8' 12.5
76.9 72.3 73.5 67.9 54.3 68.0
6.7 9.2 12.8. 2.o· 5.3•
6.7 17.7 18.9 7.1 16.7
15.0 40.0
65.5 48.6 70.8 66.7 66.7
15.8 5.6
15.8 15.4
16.71
50.01
No. of oocytes
Cleavage rate
1 2,3 4,5,6 7,8,9 10, 11, 12 13, 14, 15 16 or more
p < 0.005. p < 0.005. e P < 0.005. t Small numbers.
• Implantation rate: number of gestational sacs + number of embryos transferred X 100. b Proportion multiple pregnancies: number of ongoing multiple pregnancies + total clinical pregnancies X 100.
c
d
occurs probably as a result of a predisposition to abnormal chromosomal segregation in oocytes ovulated in later reproductive years. 9 •10 A reduced competence of organs to sustain a gestation as a result of sclerosis and stenosis of uterine blood vessels 11 and abnormal uterine microvillous architecture12 may further contribute to the lower implantation and high spontaneous abortion rates in older women. In the MFI group, significantly fewer patients had successful fertilization, cleavage, and ET, PERCENTAGE
hence the CPR cycle was lowest in this group. However, once fertilization and cleavage were achieved successfully, the pregnancy rate per ET was comparable to other diagnostic categories. Interestingly, the pregnancy rate per ET was significantly higher in the MFI group when the female partner was younger than 30 years of age (unpublished data, P < 0.001). This could be the result of enhanced fertility and fecundity of a normal female partner in a stimulated cycle. It has been reported that the oocyte recovery rate n(): number of pregnancies
n: number of ET's
60
n(S) 50
n(13) n(6) n(32)
40 n(13)
n(6) 30
n(9)
20
10
0
4
5
6
8
7
9 or above
NUMBER OF CLEAVED EMBRYOS
lllll!
Multiple pregnancies (ll of clinical pregnancies excluding ectopics)
~
CPR per ET (ll)
Figure 3 Influence of the number of cleaved embryos available prior to the ET on the CPR per ET and the incidence of multiple pregnancies in patients receiving four embryos. Multiple pregnancies: X2 = 0.74, 3 df, NS (sets 7 and 8 grouped together). CPR per ET: X2 = 616, 5 df, NS.
476
Sharma et al. Factors affecting outcome in an IVF program
Fertility and Sterility
and fertilization rate is adversely affected by endometriosis compared with that in patients with tubal damage or unexplained infertility. 13 This effect has not been confirmed by our data and those of others. 4·14·15 Although the clinical and laboratory parameters, including the fertilization rate and the mean number of embryos transferred, were similar, the CPR was significantly higher (P < 0.025) in patients with mild endometriosis when compared with those with severe endometriosis, suggesting that the outcome of IVF may be related to the severity of this disease. In patients with extensive involvement of the genital tract, the oocyte and, consequently, the embryo quality may be impaired and/or uterine receptiveness altered. Interestingly, the CPR per ET appears to be lower among patients with tubal damage and severe endometriosis as compared with those with normal pelvic organs (e.g., female partners in MFI and mild endometriosis). Furthermore, the CPR per cycle appears to be similar in patients with severe endometriosis and in those with MFI. Whereas in the group with MFI this appears to be largely due to fewer patients reaching ET, in patients with severe endometriosis, this may be related to lower uterine receptivity. It is possible that pelvic disease adversely affects embryo viability and implantation. The CPR was significantly lower in patients with cycles of 26 days or fewer. Patients with short cycles appeared to have a more rapid follicular growth and estrogen rise, and therefore reached an earlier oocyte recovery than those with longer cycles. This may result in transfer of embryos to an inadequately primed and discordant uterine environment, with a lower conception rate. However, the effect of the modified treatment regimen in women with short cycles cannot be excluded in this analysis. An overall linear rise in the pregnancy rate with the number of embryos transferred up to four embryos3·4·6·8·16·17 is now well established. Our results are similar to the World Collaborative ReportY The incidence of anembryonic pregnancies also showed an upward trend with the number of embryos transferred, but that of preclinical pregnancies and miscarriage in embryonic pregnancies did not show any variation. The likelihood of a multiple pregnancy also increased when more than one embryo was transferred. Interestingly, in early pregnancy, twice as many patients showed multiple implantations (as judged by the number of gestational sacs) compared with ongoing multiple pregnancies. Excess implantations that did not conVol. 49, No. 3, March 1988
tinue development might represent the "helper" embryos postulated to induce decidualization and/ or provide the luteotropic stimulus, thus aiding the dominant embryo with the greatest potential to continue its growth. The number of attempts at IVF failed to show a beneficial or deleterious effect on the success rate. It appears that embryonic and uterine factors alone determine the chances of success in an individual cycle and the CPR in an individual cycle remains fairly constant. Similar results have been reported recently .18 These data justify the continuation of treatment as long as cleaving embryos are obtained. Hyperstimulation regimens are employed to encourage recruitment and growth of multiple follicles, retrieval of multiple oocytes, and transfer of multiple embryos. However, the proportion of patients reaching an ET did not significantly improve when more than 5 oocytes were collected. The CPR per ET also showed a linear rise only when up to 5 oocytes were collected. Thereafter, the success rate appears to reach a plateau. Assuming that the proportion of follicles punctured and oocytes retrieved remains fairly constant, there appears to be little additional benefit when more than 5 oocytes are available for retrieval. It also appears that success depends upon the proportion of oocytes that fertilize. The best results were obtained when 7 to 9 oocytes were collected and more than 60% oocytes fertilized. Therefore, retrieval of a large number of oocytes alone fails to confer a higher chance of success. The fertilization rate in a particular cycle appears to have a greater prognostic value. The low fertilization rate often associated with a high retrieval may reflect a deficiency in the quality of oocytes and embryos when patients are stimulated excessively. Uterine receptiveness also may be severely compromised in these patients. Similarly, in the lower fertilization groups B and C, there was a striking absence of multiple implantations when more than 10 oocytes were collected, and even in group A when more than 12 oocytes were retrieved. In summary, the success of implantation was highest when 7 to 9 oocytes were retrieved and more than 60% of these fertilized, and lower with higher oocyte retrievals, even when the fertilization rate was above 60%. The presence of spare embryos at transfer also fails to positively influence the success rate. Hence, stimulation beyond optimum not only increases the risks of hyperstimulation to the patient, but also may adversely affect the success rate. Sharma et al.
Factors affecting outcome in an IVF program
477
From our data, if optimal factors are operating, a CPR of 37% per ET can be expected. An understanding of the effect of these variables can help in assessing the prognosis in an individual cycle. Acknowledgment. We would like to acknowledge the contribution of the nursing and laboratory staff of Hallam Medical Center in this study.
REFERENCES 1. Riddle A, Sharma V, Mason B, Ford NT, Pampiglione JS, Parsons J, Campbell S: Two years experience of ultrasound directed oocyte recovery. Fertil Steril 48:454, 1987 2. Marrs RP: Does the method of oocyte collection have a major influence on in vitro fertilization? Fertil Steril 46:193, 1986 3. Edwards RG, Steptoe PC: Current status of in-vitro fertilisation and implantation of human embryos. Lancet 2:1263, 1983 4. Jones HW, Acosta AA, Andrews MC, Garcia JE, Jones GS, Mayer J, McDowell JS, Rosenwaks Z, Sandow BA, Veeck LL, Wilkes CA: Three years of in vitro fertilization at Norfolk. Fertil Steril 42:826, 1984 5. Rogers P, Molloy D, Healey D, McBain J, Howlett D, Bourne H, Thomas A, Wood C, Johnston I, Trounson A: Cross over trial of superovulation protocols from two major in-vitro fertilization centers. Fertil Steril 46:424, 1986 6. Brown JB, McLeod FC, McNaughton MC, Smith MB, Smyth A: A rapid method for estimating estrogens in urine using a semiautomatic extractor. J Endocrinol 42:5, 1968
478
Sharma et al. Factors affecting outcome in an IVF program
7. Seigal S: Non Parametric Statistics for the Behavioral Sciences. New York, McGraw-Hill, 1956 8. Edwards RG, Fishel SB, Cohen J, Fehilly CB, Purdy JM, Slater JM, Steptoe PC, Webster JM: Factors influencing the success of in vitro fertilization for alleviating human infertility. J In Vitro Fert Embryo Transfer 1:3, 1984 9. Faddy MJ, Jones EC, Edwards RG: An analytical model for ovarian follicle dynamics. J Exp Zool197:173, 1976 10. Henderson SA, Edwards RG: Chiasma frequency and maternal age in mammals. Nature 218:22, 1968 11. Naeye RL: Maternal age, obstetric complications and the outcome of pregnancy. Obstet Gynecol 61:210, 1983 12. Smith AF: Ultrastructure of the uterine luminal epithelium at the time of implantation in ageing mice. J Reprod Fertil 42:183, 1975 13. Wardle PG, Mitchell JD, McLaughlin EA, Ray BD, McDermott A, Hull MG: Endometriosis and ovulatory disorder: reduced fertilization in vitro compared with tubal and unexplained infertility. Lancet 2:236, 1985 14. Chillik CF, Acosta AA, Garcia JE, Perera S, Van Vern JFHM, Rosenwaks Z, Jones HW: The role of in vitro fertilization in infertile patients with endometriosis. Fertil Steril 44:56, 1985 15. Matson PL, Yovich JL: The treatment of infertility associated with endometriosis by in vitro fertilization. Fertil Steril 46:432, 1986 16. Craft I, Porter R, Green S, Tucker M, Smith B, Twigg H, Ahuja K, Whittingham D: Success of fertility, embryo number and in vitro fertilization. Lancet 1:732, 1984 17. Seppala M: The World Collaborative Report on in vitro fertilization and embryo replacement: current state of the art in January 1984. Ann NY Acad Sci 442:558, 1985 18. Guzick DS, Wilkes C, Jones HW: Cumulative pregnancy rates for in vitro fertilization. Fertil Steril 46:663, 1986
Fertility and Sterility
Vol. 49, No. 3, March 1988 Printed in U.S.A.
An analysis of factors influencing the establishment of a clinical pregnancy in an ultrasound-based ambulatory in vitro fertilization program
Vinay Sharma, M.R.C.O.G.* Andrew Riddle, M.R.C.O.G. Bridgett A. Mason, M.B.B.S.
Julian Pampiglione, M.R.C.O.G. Stuart Campbell, F.R.C.O.G.
The Hallam Medical Centre and King's College School of Medicine and Dentistry, London, United Kingdom
In the study period (June 1984 to December 1986), 2232 cycles were stimulated in 1294 patients. Ultrasound-directed oocyte recovery (UDOR) was performed as an ambulatory procedure in 1737 (77.8%) cycles, resulting in 1375 embryos transfers (ET). Age, etiology, menstrual cycle length, number of oocytes collected, and number of embryos transferred were important determinants of the outcome. The number of attempts at in vitro fertilization did not affect the clinical pregnancy rate (CPR). In patients receiving four embryos, the CPR appeared to be highest when up to seven embryos were available for transfer. The fertilization rate in an individual cycle had a good prognostic value, the implantation rate being highest when 7 to 9 oocytes were retrieved and >60% of these were fertilized. When 10 or more oocytes were collected, the implantation rate showed a progressive decline, regardless of the fertilization rate. Furthermore, multiple pregnancies failed to occur when >12 oocytes were retrieved or more than eight embryos were available for transfer. These data suggest that, in excessively stimulated cycles, the quality of oocytes and embryos or uterine receptiveness may be suboptimal, and the transfer of more than four embryos is unlikely to increase the success rate. Fertil Steril 49:468, 1988
We have recently reported1 the results of our first 2 years of experience in a large ambulatory in vitro fertilization (IVF) program. We herein present a more detailed analysis of 2232 cycles that were stimulated between June 1984 and December 1986, relying almost completely on ultrasound for follicular monitoring and oocyte recovery. Contrary to earlier doubts, 2 our results have shown that this method is not only safe, but has achieved a clinical pregnancy rate (CPR) similar to the longer-established laparoscopy programs. 3- 5 Numerous variables, such as etiology, selection of patients, agents for stimulation, methods of monitoring follicular development, technique of oocyte recovery, and laboratory protocol, may influence the Received June 2, 1987; revised and accepted November 4, 1987. *Reprint requests: Vinay Sharma, M.R.C.O.G., Hallam Medical Centre, 77 Hallam Street, London WIN 5LR, England. 468
outcome of treatment. In this article we present the influence of some of these factors on the establishment of a clinical pregnancy. MATERIALS AND METHODS
All patients treated between June 1984 and December 1986 have been included in this analysis. Patients in all age groups were accepted in the first 6 months of our program. Because no pregnancies were obtained in patients above 40 years, we did not enroll them in the program after January 1985. Only those investigations that had been omitted prior to referral and were considered essential for an accurate diagnosis were performed before treatment. Etiologic Definitions
Patients were considered to have tubal damage if the tubes were swollen, blocked, or fixed, as proven
Sharma et al. Factors affecting outcome in an IVF program
Fertility and Sterility
by laparoscopy and hysterosalpingogram. If endometriotic deposits were identified at laparoscopy but there was no tubal damage or ovarian involvement, then the patients were considered to have mild endometriosis. If the endometriotic deposits involved the tubes and/or ovary, then a classification of severe endometriosis was made. All husbands had a routine semen analysis and a mixed antiglobulin reaction (MAR) test performed initially. The Kibrick and Trey agglutination test (TAT) was only performed in men if the MAR test was positive and in women if the postcoital or postinsemination test was negative. If either or a combination of the following were present-i.e., the sperm density was <20 X 106 /ml, sperm motility <40%, the number of abnormal sperm >60%, and/ or MAR test was positive-then the couple was diagnosed as having male factor infertility (MFI). However, if only antibodies were present in either partner, then the couple was designated as having an immune factor responsible for infertility. Any combination of abnormalities in the male or female partner was designated mixed male and female factor infertility. If no abnormalities were detected in either of the partners, then the couple was said to have unexplained infertility. Ovarian stimulation was performed in most women with a combination of clomiphene citrate (CC) and human menopausal gonadotrophin (hMG; Pergonal, Serono Laboratories, Inc., Welwyn Garden City, UK) or pure follicle-stimulating hormone (pure FSH; Metrodin, Serono Laboratories, Inc.) CC was always administered from the second to the sixth day of the menstrual cycle (100 mg orally per day). In the first treatment cycle, gonadotropin administration was commenced either on the second or the fourth day of the menstrual cycle, depending on the usual cycle length. Patients more than 36 years of age received 225 IU, and younger patients were given 150 IU gonadotrophin daily. In subsequent cycles, the daily dose was adjusted according to the folliculogenic response obtained previously. Monitoring was performed both ultrasonically (Diasonics DRF-1, Bedford, England, 3.5 MHz GPM-11 probe; Phillips SDR 1500, Hammersmith, London, UK, 3.5 and 5.0 MHz abdominal probes; International General Electric RT3000, Slough, Berkshire, UK, 3.5 and 5.0 MHz abdominal probes) and biochemically. The first ultrasound scan was performed on the second day of the patient's menstrual cycle when any abnormalities, such as ovarian cysts and fibroids, were noted. Daily ultrasound Vol. 49, No. 3, March 1988
scans were performed after the eighth day. All follicles were visualized in the transverse and the longitudinal planes, and the maximum vertical and transverse diameters measured. The arithmetic mean of these four measurements was recorded. The patients also were advised to collect 24-hour urine samples from the second to the third day, and daily after the seventh day of their menstrual cycle. Total urinary estrogen excretion was measured by the method described by Brown et al. 6 The criteria for the administration of human chorionic gonadotrophin (hCG; 5000 IU, Serono Laboratories, Inc.) are described in detail elsewhere.1 Ultrasound-directed oocyte recovery (UDOR) was performed as an ambulatory procedure 34 to 35 hours after the hCG was administered. The patients were advised to drink sufficient water in order to fill the bladder before oocyte recovery. They arrived 30 minutes before the scheduled time and usually left 2 hours after the procedure was complete. Demerol (Winthrop-Breon Laboratories, New York, NY), diazepam, and prochlorperazine were administered intramuscularly for sedation 30 minutes before the procedure in doses appropriate to the patient's body weight. During the procedure, abdominal scanning was performed in the transverse plane and the aspiration needle (gauge 16, Hallam Medical Centre IVF needle, Rocket of London Ltd., Watford, England) was introduced transcutaneously parallel to the ultrasound beam using the "free-hand" technique. Ovaries that were located lateral to the uterus were approached transvesically. Follicles in subcutaneous ovaries were aspirated by the direct transabdominal route. Ovaries located posterior to the body of the uterus in the pouch of Douglas were aspirated transvaginally, the needle being visualized with an abdominal probe. Oocytes with a fractured zona, darkened cytoplasm, or an abnormal shape were not considered to be viable. All oocytes were examined 18 hours after insemination and, if two pronuclei were visualized, fertilization was thought to have occurred. Only those embryos that had undergone at least one mitotic division after confirming fertilization (cleaved embryos) were transferred 48 to 60 hours after oocyte recovery. The technique of embryo transfer (ET) and luteal support have been described in detail elsewhere. 1 Patients were designated as being clinically pregnant when the amenorrhea continued for more than 6 weeks from the first day of the last men-
Sharma et al. Factors affecting outcome in an IVF program
469
Table 1
Age Distribution
Age groups
Jun 1984 to Dec 1984
Jan 1985 to Jun 1985
Jul1985 to Dec 1985
Jan 1986 to Jun 1986
Jul 1986 to Dec 1986
Jun 1984 to Dec 1986
Total cycles .;;25 years (%) 26-30 years(%) 31-35 years(%) 36-40 years(%) >40 years(%)
250 16 (6.4) 49 (19.6) 91 (36.4) 82 (32.8) 12 (4.8)
373 17 (4.6) 70 (18.8) 158 (42.4) 123 (33.0) 5 (1.3)
401 16 (4.0) 80 (20.0) 173 (43.1) 129 (32.2) 3 (0.7)
572 19 (3.3) 121 (21.2) 244 (42.7) 186 (32.3) 3 (0.5)
636 21 (3.3) 138 (21.7) 288 (45.3) 187 (29.4) 2 (0.3)
2232 89 (4.0) 458 (20.5) 955 (42.7) 705 (31.6) 25 (1.1)
237 11 (4.6) 48 (20.3) 106 (44.7) 71 (30.0) 1 (0.4)
371 13 (3.5) 88 (23.7) 165 (44.5) 106 (28.3)
396 12 (3.3) 96 (24.2) 180 (45.5) 107 (27.0) 1 (0.3)
1375 55 (4.0) 325 (23.7) 602 (43.8) 387 (28.1) 6 (0.4)
Proportion of patients receiving an embryo replacement 232 Total cycles 138 8 (5.8) 11 (4.7) .;;25 years (%) 33 (23.9) 60 (25.9) 26-30 years(%) 53 (38.4) 98 (42.2) 31-35 years(%) 41 (29.7) 62 (26.7) 36-40 years(%) 3 (2.2) 1 (0.4) >40 years (%)
strual period and a definite gestational sac was identified on ultrasound scanning. Those patients who had bleeding before the completion of 6 weeks of ammenorrhea, but had a plasma hCG level above 25 IU/ml, a minimum of 16 days after the oocyte recovery and 10 days after the last hCG injection, were diagnosed as having had a preclinical pregnancy. Age Distribution
All patients were divided into four age groups: under 25 years, 26 to 30 years, 31 to 35 years, and 36 to 40 years. There was no significant difference in the distribution of age groups during the study Table 2
period. Similarly, the age distribution remained uniform among the patients receiving an ET (Table 1). Etiologic Distribution
There was no alteration in the distribution of the different etiologic categories treated over the last 2 years of the study period. Similarly, the distribution of these categories has not changed among the patients receiving an ET (Table 2). Statistical Methods
All statistical comparisons on the effect of age have been made between the groups 26 to 30 years
Etiologic Distribution Jan 1985 to Jun 1985
Jul1985 to Dec 1985
Jan 1986 to Jul 1986
Jul 1986 to Dec 1986
Jun 1984 to Dec 1986
Diagnosis
Jun 1984 to Dec 1984 %
%
%
%
%
%
Total cycles Tubal damage Unexplained infertility Mild endometriosis Severe endometriosis Male factor infertility Immune factor infertility Mixed male and female
250 30.8 25.2 10.0 4.0 10.8 4.0 15.2
373 42.6 26.3 4.3 8.3 5.6 1.6 11.3
401 44.6 28.2 5.5 8.0 5.2 2.5 6.0
572 48.1 27.1 5.2 6.1 3.8 4.9 4.7
636 47.8 28.0 6.6 5.2 4.6 4.2 3.6
2232 44.5 27.2 6.0 6.3 5.4 3.6 6.9
232 47.8 25.8 4.3 6.5 2.6 1.7 11.2
237 48.1 28.3 5.5 6.8 2.5 2.5 6.3
371 50.7 26.1 4.6 6.7 2.2 4.3 5.4
396 53.3 27.0 6.1 5.8 2.0 2.5 3.3
1375 49.0 26.3 5.7 6.0 2.9 3.2 6.8
Proportion of patients receiving an embryo transfer 138 Total cycles 36.2 Tubal damage 21.7 Unexplained infertility 10.9 Mild endometriosis 2.9 Severe endometriosis 8.7 Male factor infertility 5.8 Immune factor infertility 13.8 Mixed male and female
470
Sharma et al.
Factors affecting outcome in an IVF program
Fertility and Sterility
and 36 to 40 years. Most patients above the age of 40 years were treated in the first 6 months of our program, which was less successful. Their numbers were also comparatively small. Also, the diagnostic conditions leading to a less favorable outcome (severe endometriosis, MFI, immune factor, mixed male and female factor infertility) were higher in patients under 25 years (under 25 years, 33. 7%; 26 to 30 years, 20.9%; 31 to 35 years, 22.3%; and 36 to 40 years, 21.4%). In the analysis of the effect of menstrual cycle length, only patients with tubal damage and unexplained infertility have been included because the other categories contained variables that could alter fertilization rates or the ovarian cycle. In subsequent attempts, some patients had their treatment regimens modified according to their previous folliculogenic response, and this could bias results. Therefore, only first attempts have been analyzed. We have introduced a new method of assessing oocyte quality and hence the prediction of the pregnancy rate based on the results obtained in the laboratory. For this analysis, patients were divided into three groups: fertilization in more than 60% of the retrieved oocytes (group A), fertilization in 31% to 60% of oocytes (group B), and fertilization in <30% of oocytes (group C). The cleavage rate, implantation rate of transferred embryos (i.e., number of gestational sacs ..;- number of embryos transferred X 100), CPR per ET, and proportion of multiple pregnancies (i.e., number of ongoing multiple pregnancies ..;- total clinical pregnancies X 100) were calculated for each of these groups. Overall, chi-square tests have been used to assess
Table 3
Total cycles % Abandoned" % Reached UDORb %Successful UDOR' %Reached ETd Overall fertilization rate• (%) Overall cleavage rate!(%) Oocytes retrieved, median (range) Viable oocytes, median (range) Oocytes fertilized, median (range) Embryos cleaved, median (range) Embryos transferred, median (range)
.;;25 yr
6 5 4 3 3
89 16.9 82.0 78.7 61.8 63.4 81.2 (2-10) (2-8) (2-6) (1-5) (2-4)
79.86, 4 df, p < 0.001. 37.32, 4 df, P < 0.001. ' X2 = 64.5, 4 df, P < 0.001 (X 2 = 229.82, 4 df, P < 0.001 when compared with total cycles or those reaching UDOR). X2
b
x2 =
RESULTS A total of 2232 cycles in 1294 patients were stimulated between June 1984 and December 1986. Influence of Age
A significantly higher proportion of younger patients (Table 3) had a satisfactory response to gonadotropin stimulation, successful egg recovery, and ET. Conversely, a significantly higher proportion of patients in the 36- to 40-year age group were abandoned as compared with those in the 26- to 30-year age group. Significant heterogeneity also was observed when those cases enjoying successful UDOR were analyzed with respect to the total number of cycles and those which reached UDOR. Hence 26- to 30-year-old patients were more likely to have oocytes collected than those who were 36 to 40 years old. The overall fertilization and cleavage rates did not show significant variation with age. However, with age, there appears to be a progressive decrease in the mean number of total oocytes collected, viable oocytes, fertilized oocytes, and cleaved embryos (Table 3). The highest number of clinical pregnancies (Table 4) with respect to the total number of cycles
Influence of Age Clinical data
a
the significance of heterogeneities in frequency distributions among groups and subgroups. Partitioned chi-square tests for linear trends were calculated whenever necessary. Individual Yate's chisquare tests were performed on separate groups to investigate sources of heterogeneity suggested by high chi-square values. 7
=
Vol. 49, No. 3, March 1988
26-30 yr
31-35 yr
36-40 yr
>40 yr
458 14.4 85.6 84.5 71.0 69.1 81.8 6 (3-9) 5 (2-8) 4 (2-6) 3 (1-5) 3 (2-4)
955 19.8 80.2 78.8 63.1 67.8 81.8 5 (2-8) 5 (2-8) 3 (1-5) 3 (1-5) 3 (2-4)
705 29.5 70.5 68.2 54.8 69.6 80.6 4 (2-6) 4 (2-6) 3 (1-5) 3 (1-5) 2 (1-3)
25 64 36 36 24 70.6 75 2 (1-3) 2 (1-3) 2 (1-3) 1 (0-2) 1 (0-2)
d x2 = • X2 = 1 X2 =
Sharma et al.
63.4, 4 df, P < 0.001. 2.1, 4 df, NS. 0.92, 4 df, NS.
Factors affecting outcome in an IVF program
471
Table 4
Influence of Age Pregnancy Rates and Outcome ~25
Variable Pregnancy rates Total clinical pregnancies" CPR per successful UDORb (%) CPR per ET' (%) Pregnancy outcome Total pregnancies % Preclinicald % Clinical• % Anembryonic 1 % Miscarried8·h
yr
31-35 yr
36-40 yr
81 20.9 24.9
122 16.2 20.2
60 12.5 15.5
89 9.0 91.0 14.6 4.5
133 8.3 91.7 9.0 3.8
70 14.3 85.7 5.7 7.1
11
15.7 20.0 14 21.4 78.6 7.1
a X2 = 25.45, 3 df, P < 0.001 when compared with total number of cycles. "X 2 = 11.26, 3 df, P < 0.05 when compared with number ofETs. b X 2 = 2.6, 3 df, NS. ' x2 = 3.82, 3 df, NS.
and to the number of ETs was obtained in women who were in the 26- to 30-year age group, with the lowest among the 36- to 40-year-old group. There is no evidence of a significant change within the age groups in the CPR per UDOR or in the CPR per ET. Age had no significant effect on the proportion of pregnancies that were either preclinical, clinical, anembryonic, or misc,arried. Influence of Etiology
There was no significant influence of etiology (Table 5) on the number of abandoned cycles. Sim-
Table 5
26-30 yr
d
x2 =
1.62, 3 df, NS.
• X2 = 3.78, 3 df, NS. t x2 = 5.62, 3 df, NS. 6
h
X2 = 0.98, 3 df, NS. Pregnancies with fetal heart movements on ultrasound scan.
ilarly, the proportion reaching UDOR was not affected significantly. The proportion of patients reaching ET was heterogenous with respect to the etiology. This was due to a proportionately greater number of patients with tubal damage, unexplained infertility, mild and severe endometriosis, and fewer cases with MFI reaching ET. The overall fertilization rate was highest among patients with tubal damage and lowest among those with MFI. Similarly, the overall cleavage rate was highest among patients with tubal damage and lowest among those with MFI. No evidence of significant variation was found in the mean number of
Influence of Etiology Clinical data
Tubal damage
Unexplained infertility
Mild endometriosis
Severe endometriosis
Male factor
Immune factor
Mixed factors
Total cycles % Abandoned" % Reached UDORb %Reached ET' Overall fertilization rated (%) Overall cleavage rate• (%) Oocytes retrieved, median (range) Viable oocytes, median (range) Oocytes fertilized, median (range) Embryos cleaved, median (range) Embryos replaced, median (range)
994 20.3 79.7 67.8 74.9 83.4 5 (2-8) 5 (2-8) 3 (2-6) 3 (1-5) 3 (2-4)
607 22.1 77.8 59.5 65.6 80.1 5 (2-8) 5 (2-8) 3 (1-5) 3 (1-5) 3 (2-4)
135 29.6 70.3 58.5 71.9 79.8 4 (2-6) 4 (2-6) 3 (1-5) 3 (2-4) 3 (2-4)
141 8.3 71.6 58.9 72.0 79.8 4 (2-6) 4 (2-6) 3 (1-5) 3 (1-4) 3 (1-5)
120 25.8 74.2 33.3 36.5 65.7 5 (2-8) 4 (2-6) 3 (1-5) 3 (1-5) 2 (1-3)
81 24.7 75.3 54.3 55.0 76.6 5 (3-7) 4 (2-6) 3 (1-5) 3 (1-5) 3 (2-4)
154 17.5 82.5 61.0 59.4 83.0 5 (2-8) 4 (2-6) 3 (1-5) 3 (1-5) 2 (1-4)
127 12.8 18.8
77 12.7 21.3
21 15.6 6 26.5 6
10 8.3 25.0
6 7.4 13.6
22 14.3 23.4
Pregnancy rates Clinical pregnancies t CPR per cycle(%) CPR per ET (%)
X2 = 12.24, 6 df, NS, checked for continuity. X2 = 12.42, 6 df, NS, checked for continuity. 'X 2 = 75.66 compared with those reaching UDOR and 59.58 compared with total cycles, each 6 df, P < 0.001. d x2 = 87.14, 6 df, P < 0.001. a
b
472
Sharma et al.
11
7.8 8 13.3 8
• X2 = 21.1, 6 df, P < 0.01, Yates corrected X2 value for tubal damage 7.1, 1 df, P < 0.01 and male factor 17.2, 1 df, P < 0.001. t x2 = 9.06, 6 df, NS. 6 p < 0.001.
Factors affecting outcome in an IVF program
Fertility and Sterility
oocytes collected, viable oocytes, fertilized oocytes, cleaved embryos, or the number of embryos transferred among the different diagnostic categories (Table 5). There was no significant effect of etiology on the total number of clinical pregnancies. However, the CPR per cycle and per ET was significantly higher in patients with mild as compared with severe endometriosis. The CPR per cycle in patients with severe endometriosis and in those with MFI was similar, although the etiology in these two groups might be different. Whereas in MFI the etiology appears to be predominantly due to a lower fertilization rate, this was not the case in severe endometriosis, where the fertilization rate was normal but the implantation rate poor. The number of pregnancies in each group is not large enough to analyze the influence of etiology on pregnancy outcome. Influence of Menstrual Cycle Length
Menstrual cycle length or etiology (Table 6) did not influence the number of cycles analyzed, number of cycles abandoned, number of successful UDORs, fertilization rate, cleavage rate, or number of clinical pregnancies. The CPR in patients with a menstrual cycle of 26 days was lower than those
Table 6
with a normal menstrual cycle length. This could not be attributed to other factors, although the effect of early hMG administration could not be excluded in this analysis. Influence of the Number of Embryos Transferred
The CPR (Table 7) progressively increased from 8.6%, when one embryo was transferred, to 29.5%, when four embryos were transferred. This trend did not appear to continue among patients receiving more than four embryos, although this group is comparatively small. An upward trend was also seen in the incidence of blighted ova per ET. The incidence of preclinical pregnancies, miscarriages, and ectopic pregnancies was not associated with the number of embryos transferred. However, a significantly higher proportion of patients conceived when four or more embryos were transferred, as compared with two or three. A similar effect also was seen in the number of ongoing multiple pregnancies. The two-ET category had a lower and the four-ET group a higher multiple pregnancy rate, with no significant effect over the three-ET category, suggesting a consistent trend on either side of the latter group. Almost half of the multiple implantations diagnosed by ultrasound failed to progress in pregnancy.
Influence of Menstrual Cycle Length Unexplained infertility
Tubal damage
Cycles• (n) Mean age (years) Mean gonadotrophin dose, daily (IU) % Abandonedb No. of UDORsc (successful) Oocytes retrieved, median (range) Viable oocytes, median (range) Oocytes fertilized, median (range) Embryos cleaved, median (range) Embryos transferred, median (range) Fertilization rated(%) Cleavage rate• (%) Clinical pregnancies' (n) CPR per cycles CPR per UDOR h CPRperET; 0.25, 1 df, NS. 0.01, 1 df, NS. c X2 = 0.37, 1 df, NS. d X2 = 0.60, 1 df, NS. • x 2 = 0.01, 1 df, NS. • X2 = b 2 =
x
Vol. 49, No. 3, March 1988
Menstrual cycle .;;26 days
Menstrual cycle ~27 days
Menstrual cycle .;;26 days
Menstrual cycle ~27 days
115 33.1 ± 4.4 150-225 20 92 (91) 5 (2-8) 4 (1-7)
166 31.8 ± 4.2 150-225 18.7 135 (129) 5 (2-8) 4 (1-7) 3 (1-5) 3 (1-5) 3 (2-4) 69.2 86.0 22 13.3 17.0 20.4
71 33.2 ± 3.8 150-225
115 33.4 ± 3.9 150-225 20.0 92 (88) 5 (2-8) 4 (1-7) 4 (1-7) 3 (1-5) 3 (2-4) 70.4 69.1 12 10.4 13.6 22.2
4 (2-6)
3 (1-5) 3 (2-4) 75.3 82.2 6 5.2 6.6
8.0
21.1
56 (53) 5 (1-9) 4 (1-7) 4 (2-6)
3 (1-5) 3 (2-4) 70.8 84.2 7 9.9 13.2 16.3
t x2 = 0.684, 1 df, NS.
P < 0.05. p < 0.05. i p < 0.01.
6
h
Sharma et al.
Factors affecting outcome in an IVF program
473
Table 7
Influence of Number of Embryos Transferred
No. of embryos transferred
No. of embryo transfers
1 2 3 4 ;;;.5
245 322 317 468 23
Clinical" pregnancy rate perET
Preclinicalb pregnancy rate perET
(n)
8.6 15.2 18.9 29.5 26.1
(21) (49) (60) (138) (6)
Incidence of' blighted ova per ET
Miscarriaged rate per ET
(n)
(n)
(n)
0.8 1.6 2.8 3.2 8.7
(2) (5) (9) (15) (2)
0.8 0.9 0.9 3.0 4.3
(2) (3) (3) (14) (1)
2.5 0.6 0.9 4.3
(8) (2) (4) (1)
Ectopice rate per ET
Multiple' implantation rate per ET
MultipleR (ongoing) pregnancy rate perET
(n)
(n)
(n)
1.6 2.2 1.6 1. 7 8.7
(4) (7) (5) (8) (2)
3.7 8.8 14.1 13.0
(12) (28) (66) (3)
1.2 (4) 3.8 (12) 7.1 (33)
Multiple implantation rate per ET: number of patients with multiple implantations 7 number of embryo transfers X 100. Multiple ongoing pregnancy rate per ET: number of patients with multiple pregnancies 7 number of embryo transfers X 100.
x2 = x2 =
54.82, 4 df, p < 0.001. 9.06, 4 df, NS (2 X 2 Yates value for one embryo, 2.42, 1 df, NS; for four embryos, 1.47, 1 df, NS). 'X 2 = 9.13, 3 df, P < 0.05 (2 X 2 Yates value for four embryos, 7.57, 1 df, p < 0.01). d x2 = 4.82, 2 df, NS. a
b
e
X2
=
1.34, 4 df, NS.
t x2 = 24.13, 2 df, P < 0.001. g X2 = 15.6, 2 df, P < 0.001 (for two embryos, X2 = 9.85, 1 df, P < 0.001; for four embryos, X2 = 12.15, 1 df, P < 0.001; for three embryos, X2 = 0.245, 1 df, NS).
receive ET, and the pregnancy rate per ET were assessed individually for patients with 1 to 5 oocytes, and thereafter for groups with 6 to 10, 11 to 15, and above 16 oocytes. The overall fertilization rate did not show a significant change with the number of oocytes collected. Similarly, the cleavage rate showed little variation in the groups noted in the preceding paragraph. However, there was a linear increase in the number of patients reaching ET when four or more
Influence of the Number of Attempts
The pregnancy rate per cycle, per UDOR, or per ET did not show any significant change with the number of attempts (Fig. 1). Influence of the Number of Oocytes Collected
The number of oocytes collected after UDOR varied between 1 and 25 (Fig. 2). The fertilization rate, cleavage rate, proportion of cycles failing to PERCENTAGE
30
Pregnancy rate per
ET
20
UDOR - -
Cycle - -
10 Cycles
1295
575
227
80
34
UDOR
1004
452
175
60
28
18
ET
759
374
151
53
23
15
Preg.
156
74
25
11
4
4
3
4
21
0 2
5
6 or above
ATTEMPT NUMBER
Figure 1 Influence of the number of attempts at IVF on the pregnancy rate. CPR per cycle: X 2 = 3.42, 5 degrees of freedom (dO, not significant (NS); CPR per UDOR: X2 = 3.16, 5 df, NS; CPR per ET: X2 = 7.18, 5 df, NS.
474
Sharma et al.
Factors affecting outcome in an IVF program
Fertility and Sterility
PERCENTAGE 100 90
80 70
60 50 40 30
20 10
0 2
3
4
6-10
5
11-15
16 or above
NUMBER OF OOCYTES COLLECTED
-
Fertilization r a t e -
" cycles reaching ET
-
Cleavage rote
~
Pregnancy rate per ET
Figure 2 Influence of the number of oocytes collected on the outcome in an IVF cycle. Fertilization rate: X2 = 9.22, 7 df, NS; cleavage rate: X2 = 0.28, 7 df, NS; percentage of cycles reaching ET: linear increase from one to five oocytes, X2 = 78.79, 7 df, P < 0.01; partitioned: X2 = 71.64, 4 df, P < 0.01; pregnancy rate per ET: linear increase from one to five oocytes, X2 = 28.34, 7 df, P < 0.001; partitioned: X2 = 19.33, 4 df, P < 0.01.
oocytes were collected as compared with one, two, and three. Similarly, the CPR per ET showed a linear rise when one to five oocytes were retrieved, was highest when five oocytes were collected, and plateaued thereafter. Influence of Proportional Fertilization in Retrieved Oocytes
As shown in Table 8, the cleavage rate in the fertilized oocytes of the three groups (A, B, and C, defined under Materials and Methods) did not show any significant variation. The implantation rate of the transferred embryos did not show a statistically significant variation between groups when up to 6 oocytes were recovered. In group C, however, when 7 or more oocytes were collected, the implantation rate of embryos was lower when compared with the other two groups. In all groups, the implantation rate was lower when more than 10 oocytes were collected. Multiple pregnancies did not occur in group B when 1, 2, 3, or above 9 oocytes were retrieved. Conversely, when up to 9 oocytes were retrieved in group A, an increasing proportion of multiple pregnancies were obtained. Multiple pregnancies did not occur when more than 12 oocytes were collected, even though the cleavage rate showed little change. Vol. 49, No.3, March 1988
Influence of the Number of Cleaved Embryos
All patients who had four embryos transferred in the study period were included in this analysis (Fig. 3). They were grouped according to the number of cleaved embryos available for transfer. The CPR appeared to be lower when more than seven embryos were available for transfer. The multiple pregnancy rate also appeared to be lower when eight or more embryos were available, although this group is comparatively small. However, these results did not reach statistical significance. Similarly, the ongoing pregnancy rate did not increase when more than four embryos were available prior toET. DISCUSSION
As shown by our data and those of others, 4•8 the age of the female partner is an important determinant in the outcome of IVF. The success rate in the 36- to 40-year age group was lower in all diagnostic categories when compared with younger patients. The effects of maternal age on fecundity and fertility are complex and multifactorial. Progressive depletion of the ovarian follicular store occurs, hence a reduction in the follicular recruitment with advancing age. The production of defective oocytes
Sharma et al. Factors affecting outcome in an IVF program
475
Table 8
Influence of Proportional Fertilization in Retrieved Oocytes Group B 31-60% fertilization
Group A :»60% fertilization
Group C
~30%
fertilization
Implant• rate
CPR per ET
Proportb multiple
Cleavage rate
Implant rate
CPR per ET
Proport multiple
Cleavage rate
Implant rate
CPR per ET
Proport multiple
%
%
%
%
%
%
%
%
%
%
%
%
82.2 87.4 85.4 83.2 78.1 75.4 77.9
3.2 8.4 10.3 11.0d 8.6d 4.7d 9.2d
6.3 22.2 28.4 27.6 23.1 37.5
17.1' 23.8 28.8' 12.5
76.9 72.3 73.5 67.9 54.3 68.0
6.7 9.2 12.8. 2.o· 5.3•
6.7 17.7 18.9 7.1 16.7
15.0 40.0
65.5 48.6 70.8 66.7 66.7
15.8 5.6
15.8 15.4
16.71
50.01
No. of oocytes
Cleavage rate
1 2,3 4,5,6 7,8,9 10, 11, 12 13, 14, 15 16 or more
p < 0.005. p < 0.005. e P < 0.005. t Small numbers.
• Implantation rate: number of gestational sacs + number of embryos transferred X 100. b Proportion multiple pregnancies: number of ongoing multiple pregnancies + total clinical pregnancies X 100.
c
d
occurs probably as a result of a predisposition to abnormal chromosomal segregation in oocytes ovulated in later reproductive years. 9 •10 A reduced competence of organs to sustain a gestation as a result of sclerosis and stenosis of uterine blood vessels 11 and abnormal uterine microvillous architecture12 may further contribute to the lower implantation and high spontaneous abortion rates in older women. In the MFI group, significantly fewer patients had successful fertilization, cleavage, and ET, PERCENTAGE
hence the CPR cycle was lowest in this group. However, once fertilization and cleavage were achieved successfully, the pregnancy rate per ET was comparable to other diagnostic categories. Interestingly, the pregnancy rate per ET was significantly higher in the MFI group when the female partner was younger than 30 years of age (unpublished data, P < 0.001). This could be the result of enhanced fertility and fecundity of a normal female partner in a stimulated cycle. It has been reported that the oocyte recovery rate n(): number of pregnancies
n: number of ET's
60
n(S) 50
n(13) n(6) n(32)
40 n(13)
n(6) 30
n(9)
20
10
0
4
5
6
8
7
9 or above
NUMBER OF CLEAVED EMBRYOS
lllll!
Multiple pregnancies (ll of clinical pregnancies excluding ectopics)
~
CPR per ET (ll)
Figure 3 Influence of the number of cleaved embryos available prior to the ET on the CPR per ET and the incidence of multiple pregnancies in patients receiving four embryos. Multiple pregnancies: X2 = 0.74, 3 df, NS (sets 7 and 8 grouped together). CPR per ET: X2 = 616, 5 df, NS.
476
Sharma et al. Factors affecting outcome in an IVF program
Fertility and Sterility
and fertilization rate is adversely affected by endometriosis compared with that in patients with tubal damage or unexplained infertility. 13 This effect has not been confirmed by our data and those of others. 4·14·15 Although the clinical and laboratory parameters, including the fertilization rate and the mean number of embryos transferred, were similar, the CPR was significantly higher (P < 0.025) in patients with mild endometriosis when compared with those with severe endometriosis, suggesting that the outcome of IVF may be related to the severity of this disease. In patients with extensive involvement of the genital tract, the oocyte and, consequently, the embryo quality may be impaired and/or uterine receptiveness altered. Interestingly, the CPR per ET appears to be lower among patients with tubal damage and severe endometriosis as compared with those with normal pelvic organs (e.g., female partners in MFI and mild endometriosis). Furthermore, the CPR per cycle appears to be similar in patients with severe endometriosis and in those with MFI. Whereas in the group with MFI this appears to be largely due to fewer patients reaching ET, in patients with severe endometriosis, this may be related to lower uterine receptivity. It is possible that pelvic disease adversely affects embryo viability and implantation. The CPR was significantly lower in patients with cycles of 26 days or fewer. Patients with short cycles appeared to have a more rapid follicular growth and estrogen rise, and therefore reached an earlier oocyte recovery than those with longer cycles. This may result in transfer of embryos to an inadequately primed and discordant uterine environment, with a lower conception rate. However, the effect of the modified treatment regimen in women with short cycles cannot be excluded in this analysis. An overall linear rise in the pregnancy rate with the number of embryos transferred up to four embryos3·4·6·8·16·17 is now well established. Our results are similar to the World Collaborative ReportY The incidence of anembryonic pregnancies also showed an upward trend with the number of embryos transferred, but that of preclinical pregnancies and miscarriage in embryonic pregnancies did not show any variation. The likelihood of a multiple pregnancy also increased when more than one embryo was transferred. Interestingly, in early pregnancy, twice as many patients showed multiple implantations (as judged by the number of gestational sacs) compared with ongoing multiple pregnancies. Excess implantations that did not conVol. 49, No. 3, March 1988
tinue development might represent the "helper" embryos postulated to induce decidualization and/ or provide the luteotropic stimulus, thus aiding the dominant embryo with the greatest potential to continue its growth. The number of attempts at IVF failed to show a beneficial or deleterious effect on the success rate. It appears that embryonic and uterine factors alone determine the chances of success in an individual cycle and the CPR in an individual cycle remains fairly constant. Similar results have been reported recently .18 These data justify the continuation of treatment as long as cleaving embryos are obtained. Hyperstimulation regimens are employed to encourage recruitment and growth of multiple follicles, retrieval of multiple oocytes, and transfer of multiple embryos. However, the proportion of patients reaching an ET did not significantly improve when more than 5 oocytes were collected. The CPR per ET also showed a linear rise only when up to 5 oocytes were collected. Thereafter, the success rate appears to reach a plateau. Assuming that the proportion of follicles punctured and oocytes retrieved remains fairly constant, there appears to be little additional benefit when more than 5 oocytes are available for retrieval. It also appears that success depends upon the proportion of oocytes that fertilize. The best results were obtained when 7 to 9 oocytes were collected and more than 60% oocytes fertilized. Therefore, retrieval of a large number of oocytes alone fails to confer a higher chance of success. The fertilization rate in a particular cycle appears to have a greater prognostic value. The low fertilization rate often associated with a high retrieval may reflect a deficiency in the quality of oocytes and embryos when patients are stimulated excessively. Uterine receptiveness also may be severely compromised in these patients. Similarly, in the lower fertilization groups B and C, there was a striking absence of multiple implantations when more than 10 oocytes were collected, and even in group A when more than 12 oocytes were retrieved. In summary, the success of implantation was highest when 7 to 9 oocytes were retrieved and more than 60% of these fertilized, and lower with higher oocyte retrievals, even when the fertilization rate was above 60%. The presence of spare embryos at transfer also fails to positively influence the success rate. Hence, stimulation beyond optimum not only increases the risks of hyperstimulation to the patient, but also may adversely affect the success rate. Sharma et al.
Factors affecting outcome in an IVF program
477
From our data, if optimal factors are operating, a CPR of 37% per ET can be expected. An understanding of the effect of these variables can help in assessing the prognosis in an individual cycle. Acknowledgment. We would like to acknowledge the contribution of the nursing and laboratory staff of Hallam Medical Center in this study.
REFERENCES 1. Riddle A, Sharma V, Mason B, Ford NT, Pampiglione JS, Parsons J, Campbell S: Two years experience of ultrasound directed oocyte recovery. Fertil Steril 48:454, 1987 2. Marrs RP: Does the method of oocyte collection have a major influence on in vitro fertilization? Fertil Steril 46:193, 1986 3. Edwards RG, Steptoe PC: Current status of in-vitro fertilisation and implantation of human embryos. Lancet 2:1263, 1983 4. Jones HW, Acosta AA, Andrews MC, Garcia JE, Jones GS, Mayer J, McDowell JS, Rosenwaks Z, Sandow BA, Veeck LL, Wilkes CA: Three years of in vitro fertilization at Norfolk. Fertil Steril 42:826, 1984 5. Rogers P, Molloy D, Healey D, McBain J, Howlett D, Bourne H, Thomas A, Wood C, Johnston I, Trounson A: Cross over trial of superovulation protocols from two major in-vitro fertilization centers. Fertil Steril 46:424, 1986 6. Brown JB, McLeod FC, McNaughton MC, Smith MB, Smyth A: A rapid method for estimating estrogens in urine using a semiautomatic extractor. J Endocrinol 42:5, 1968
478
Sharma et al. Factors affecting outcome in an IVF program
7. Seigal S: Non Parametric Statistics for the Behavioral Sciences. New York, McGraw-Hill, 1956 8. Edwards RG, Fishel SB, Cohen J, Fehilly CB, Purdy JM, Slater JM, Steptoe PC, Webster JM: Factors influencing the success of in vitro fertilization for alleviating human infertility. J In Vitro Fert Embryo Transfer 1:3, 1984 9. Faddy MJ, Jones EC, Edwards RG: An analytical model for ovarian follicle dynamics. J Exp Zool197:173, 1976 10. Henderson SA, Edwards RG: Chiasma frequency and maternal age in mammals. Nature 218:22, 1968 11. Naeye RL: Maternal age, obstetric complications and the outcome of pregnancy. Obstet Gynecol 61:210, 1983 12. Smith AF: Ultrastructure of the uterine luminal epithelium at the time of implantation in ageing mice. J Reprod Fertil 42:183, 1975 13. Wardle PG, Mitchell JD, McLaughlin EA, Ray BD, McDermott A, Hull MG: Endometriosis and ovulatory disorder: reduced fertilization in vitro compared with tubal and unexplained infertility. Lancet 2:236, 1985 14. Chillik CF, Acosta AA, Garcia JE, Perera S, Van Vern JFHM, Rosenwaks Z, Jones HW: The role of in vitro fertilization in infertile patients with endometriosis. Fertil Steril 44:56, 1985 15. Matson PL, Yovich JL: The treatment of infertility associated with endometriosis by in vitro fertilization. Fertil Steril 46:432, 1986 16. Craft I, Porter R, Green S, Tucker M, Smith B, Twigg H, Ahuja K, Whittingham D: Success of fertility, embryo number and in vitro fertilization. Lancet 1:732, 1984 17. Seppala M: The World Collaborative Report on in vitro fertilization and embryo replacement: current state of the art in January 1984. Ann NY Acad Sci 442:558, 1985 18. Guzick DS, Wilkes C, Jones HW: Cumulative pregnancy rates for in vitro fertilization. Fertil Steril 46:663, 1986
Fertility and Sterility