Performance of patients with a ''frozen pelvis" in an in vitro fertilization program

Transcription

1 FERTILITY AND STERILITY Copyright 1987 The American Fertility Society Printed in U.8A. Performance of patients with a ''frozen pelvis" in an in vitro fertilization program David Molloy, F.R.A.C.O.G.*t Marian Martin, Ph.D.+ Andrew Speirs, F.R.A.C.O.G. * Alexander Lopata, Ph.D., M.B.B.S.+ Geoffrey Clarke, F.R.A.C.O.G.* John McBain, F.R.A.C.O.G.* Andrew Ngu, F.R.A.C.O.G. Ian H. Johnston, F.R.A.C.O.G.* The Royal Women's Hospital, Carlton, and University of Melbourne, Parkville, Australia It is now possible to identify and study the performance of different subgroups of patients in in vitro fertilization (lvf) programs. Patients with severe pelvic adhesions due to pelvic inflammatory disease (PID) or endometriosis were classed as having a frozen pelvis if,,:; 20% of total ovarian surface was visible and if the rest of the ovary was bound down with significant adhesions. IVF offers the only hope of pregnancy for these patients. Fifty-one treatment cycles in 23 such patients were matched against 51 cycles in 48 patients with adhesion-free ovaries. The study group had a significantly higher number of cancelled oocyte retrievals because of poor estradiol (E:J response. They also had a significantly lower rate of E2 rise and a lower peak value of E2 before and after the administration of human chorionic gonadotropin. These patients took longer to respond to a hyperstimulation regime, and when a response occurred they formed fewer follicles, as measured with the use of ultrasound. Lower numbers of oocytes were obtained from this group, but the fertilization rate of oocytes was the same for both groups. One pregnancy occurred in the study group and 11 in the control group. It is possible that disruption of ovarian blood supply or mechanical factors due to the pressure of significant adhesions prevent a good follicular response in patients with a frozen pelvis. Fertil Steril47:450, 1987 The use of in vitro fertilization (IVF) as a means of achieving pregnancy in patients with significant tubal disease is now an established Received December 9, 1985; revised and accepted November 4,1986. *Reproductive Biology Unit, The Royal Women's Hospital treprint requests: Dr. D. Molloy, Reproductive Biology Unit, Department of Obstetrics and Gynaecology, The Royal Women's Hospital, 132 Grattan Street, Carlton, Victoria, Australia :J:Department of Obstetrics and Gynaecology, University of Melbourne. Ultrasound Department, The Royal Women's Hospital. technique in reproductive biology.! In units with a large case load, it is possible to identify and study the performance of selected patient subgroups in an IVF program. A limiting factor in the success of IVF has been the degree of ovarian access to laparoscopy. Until ultrasound was applied as a method of guiding the needle for oocyte aspiration,2 patients with severe pelvic adhesions and minimal ovarian access had little hope of success with IVF, even though this represented their only chance of pregnancy. As more of these frozen pelvis (FP) patients had an IVF treatment cycle culminating in ultrasound-guided oocyte re- 450 Molloy et al. The "frozen pelvis" in an IVF program Fertility and Sterility

2 trieval, it seemed that their overall performance was suboptimal, compared with patients having IVF treatment for most other indications. The purpose of this study was to analyze retrospectively the ovarian response and results ofivf treatment cycles for a group of patients with infertility due to severe pelvic adhesions. Each of the treatment cycles in the study group was carefully matched against one in a control group of patients who had IVF treatment, but not for severe pelvis adhesions, with the use of a computer to select comparatively similar clinical parameters. MATERIALS AND METHODS Twenty-three patients having 51 treatment cycles between July 1, 1982 and June 30, 1985 were included in the study group. To qualify as having an FP, the patient had to have her ovaries embed~ ded in dense adhesions that also involved the fallopian tubes, the uterus, and, usually, the bowel, to the extent that < 20% of the total ovarian surface was visible or free of adhesions. These patients, therefore, presented with hopeless access for laparoscopic oocyte retrieval. The control group consisted of 48 patients who had 51 treatment cycles in the same time period. The two groups of patients were computermatched using the criteria presented in Table 1. In a 3-year period, treatment protocols, clinical attitudes, assays, and laboratory techniques evolve and change significantly. Accordingly, the IVF cycles for each patient were matched so that they occurred within 8 weeks of each other. In our unit, protocol changes tend to occur at specific times of the year (January 1 and July 1), so matching across these dates was also avoided. Individual treatment cycles were also matched according to an ovarian hyperstimulation protocol. A comparison of the stimulation protocols in the groups is shown in Table 2. Most patients Table 1. Matching by Age and Indication for FP and Control Groups Matching criteria Age (yrs) Poor semen quality in husbanda Endometriosisa Tubal disease a Control group /8 12/10 30/30 FP group / avalue given is number of cycles in number of couples. Table 2. Matching by Ovarian Hyperstimulation Protocol for Cycles of treatment in the FP and Control Groups Type of stimulation Control FP group Statistical group cycles cycles differences Clomid 150 mg/day D Clomid 50 mg/day D hmg Clomid 100 mg/day D3, 4-7, hmg hmg (ampules/papatient) Nsa Median hmg only (ampules/ patient) Nsa Median anot significant (Wilcoxon test). received either clomiphene citrate (CC) and human menopausal gonadotropin (hmg) or hmg alone. The total hmg dosage in each patient was matched and then compared, to ensure that the study and control groups had equivalent hyperstimulation protocols. Although the main aim of the study was to examine the follicular response of the FP group, fertilization and pregnancy rates were compared as well, so it was necessary to match semen profiles in the two groups. Poor semen quality was defined as a semen analysis count of < 20 X 106/motility < 40% and abnormal forms> 70%. The following parameters of follicular response were analyzed for each group: 1. Average baseline estradiol (E2) level. It has been the practice in our unit to measure early follicular phase E2 level before the start of stimulation. To obtain an average baseline level, we calculated a mean and standard deviation of the daily E2 values for each patient up to the start of stimulation. 2. The first day of E2 rise after the start of stimulation was determined using the World Health Organization (WHO) method,3 which defines a significant increase in E2 as 1.5 times the average baseline level as calculated above. From this, two further parameters were noted: the number of days from start of stimulation until the day of E2 rise and the actual increase in E2 level above the baseline level, to establish a defined E2 response. Molloy et a1. The "frozen pelvis" in an IVF program 451

3 3. The average rate of rise in E2 from the first day of response (DR) to the day human chorionic gonadotropin (hcg) was given or the cycle cancelled (DR). The formula used for this was: (DR) - (DR) je2 value on day of hcg E2 value on Dl of response 4. The E2 value before hcg administration or the highest E2 value achieved before cancellation, if no hcg was given. This represented the maximum E2 at the end of ovarian stimulation. 5. E2 value after hcg was given. 6. Number of follicles> 10 mm in diameter, detected with the use of ultrasound, on the day of hcg or termination of the cycle. 7. The number of cycles not proceeding to oocyte pickup due to poor follicular response. In our program, this means that the E2 and ultrasound results imply that fewer than three ova can be expected at oocyte pickup. A typical response to stimulation is charted in Figure 1 to demonstrate the analysis of a patient who proceeded to hcg administration and oocyte pickup. The data relating to oocyte retrieval for both groups were analyzed for cycle day of oocyte pickup, number of oocytes retrieved, number of embryos obtained following IVF, and number of pregnancies for each group of patients 10...J 5{) ~ 05 Cycle Day Figure 1 Method of analysis of E2 response. Stimulation protocol is shown in relation to cycle day. 452 Molloy et al. The "frozen pelvis" in an IVF program RESULTS The data relating to the follicular response obtained for both groups are presented in Table 3. Patients in the FP group had a significantly poorer follicular response for most of the parameters measured. Significantly more cycles were cancelled in the FP group due to poor follicular response. Late detection of luteinizing hormone surge caused two cycle cancellations in the FP group and one cancellation in the control group. Table 4 shows the data relating to oocyte retrieval. The lower number of oocytes retrieved reflects in part the poor follicular response of the FP group, but also is due to the difficulty in obtaining oocytes at ultrasound-guided oocyte retrieval. The 3-year study period covered the development and learning phase of this technique in our program. The retrieval rate of oocytes/ follicles aspirated laparoscopically was 70% in the control group but only 52% in the ultrasoundguided FP patients. Gradual improvement in our ultrasound oocyte retrieval technique was demonstrated over the period of study. DISCUSSION In a retrospective study, the method used to choose the control group can bias the results. Use of the computer to select patients as controls with strict criteria and no prior knowledge of cycle performance is one method to eliminate any such bias. When the characteristics of the control group were compared with data from 1661 treatment cycles representing all patients treated in our program during the study period (Table 5), there were no significant differences. This suggests that the control group was a representative sample from the total group of patients and so strengthened the validity of any differences noted in the FP group study. This study clearly demonstrates that patients with severe pelvic adhesions perform poorly in an IVF program. These patients had a longer response time to a given hyperstimulation regime. When they do respond, the E2 rises more slowly than for other patients. The peak E2 at the end of stimulation and the E2 level after hcg administration are lower. Fewer follicles are detected with ultrasound scanning and fewer oocytes are retrieved, leading to fewer embryos and no clinical pregnancies in the study group. These differences are statistically significant, and the results explain our clinical impression Fertility and Sterility

4 Table 3. Analysis of Follicular Response Comparing Control and FP Groups Control group cycles (n = 51) FP group cycles (n = 51) Statistical Significance Baseline E2 levels (nmolll) No. of days to E2 rise Median (n = 50 cycles) (n = 50 cycles) E2 increase from baseline level on Dl of rise (nmol/l) Average rate of E2 rise during stimulation Peak E2 at end of stimulation (nmolll) E2 level after hcg administration (nmol/l) No. of follicles 10 mm on ultrasound scanning Median (n = 49) (n = 41) P < O.OOlb No. of cancelled cycles due to poor response 3 11 P < 0.05 c anot significant (Wilcoxon test). bwilcoxon test. cfisher's exact test. that these patients do not conform to the usual response patterns observed in hyperstimulation protocols. The higher number of cancelled cycles because of poor follicular response in the study group may have biased the analysis of the E2 response against that group. However, when the data are reanalyzed for both the study and control groups with all of the cancelled cycles excluded, it is evident that the FP patients still take longer to respond to a given protocol of hyperstimulation (P < 0.01, Wilcoxon test). The rate of rise of E2 is still lower in the FP group (P < 0.01, Wilcoxon test), but the peak E2 before hcg administration no longer shows a significant difference between the study and control groups. However, the E2 after hcg administration is still lower in the study group (P < 0.01, Wilcoxon test) and fewer follicles are seen on ultrasound scanning in the FP patients. These findings suggest that there is poorer follicular development, even in the FP patients who proceed to oocyte pickup. A possible cause of this poor follicular response may stem from the initial disease in the pelvis in general and in the ovary in particular. The majority of these patients had a history of severe pelvic inflammatory disease (PID) as the cause of their adhesions, and oophoritis as well as salpingitis may have been present. 4 Such severe infection may have compromised ovarian blood supply and, with resultant ovarian capsule scarring, may prevent optimum follicular growth. Another possible cause for this poor follicular response may be ovarian entrapment. When viewed laparoscopically, these pelvises are a frozen mass involving adhesions, the bowel, and the Molloy et al. The "frozen pelvis" in an IVF program 453

5 reproductive organs. It is often impossible to see the ovaries at all, and it may be that the mechanical pressures of bowel and adhesions around the ovaries prevent adequate follicular growth. If the optimum follicular diameter of 18 mm is to be reached, then space must exist or be created for them to develop. This may not be the case in a severely FP. The presence of severe endometriosis as a cause for FP may also compromise follicular growth. Active endometriosis has been associated with infertility and anovulation. 5 The findings presented in this study may affect clinical IVF practice in several ways. IVF programs with limited resources or at the initial stages of development may choose to defer treatment in these patients, because of their poor success rates. Patients with this indication for IVF need sympathetic treatment, because this represents their only real hope of pregnancy. They may require extra counselling and should have a realistic appraisal of their poor chances of success. More aggressive hyperstimulation protocols may be routinely used in these patients. Our FP patients are now often started earlier on higher doses ofhmg, which may be continued for longer. This has resulted in some measure of increased follicular response, compared with previous cycles of treatment, and two continuing pregnancies since June A complication of this approach is the risk of hyperstimulation syndrome, which has occurred in three of our patients undergoing ultrasound-guided oocyte retrieval. This may be related to incomplete follicular aspiration, especially of smaller follicles, with this technique. Table 4. Analysis of Data Related to Oocyte Retrieval, Comparing Control and FP Groups Day ofopu Median No. of oocytes retrieved/patient Fertilization rate (%) Pregnancy number Pregnancy rate (%) Deliveries anot significant. bwilcoxon test. CChi-square test. dfisher's exact tst. Control group cycles (n = 47) D FP group cycles (n = 38) o Statistical significance Nsa,c P < 0.05 d P < 0.05 d Table 5. Comparison of Characteristics Between the Control Group and All Patient Cycles: July 1,1982 to June 30,1985 Median age (yrs) Seminal factors (%) Endometriosis (%) Tubal disease (%) Cancellation rate (%) Oocytesfpatient Fertilization rate (%) Percentage of patients who had a 4 EfT (%) Pregnancy rateflaparoscopy (%) Deliveries Control group cycles (n = 51) All patient cycles (n = 1661) NS, chi-squared test In our experience, expertly performed ultrasound-guided oocyte retrieval may enhance the chances of pregnancy in this group. Other units 6 using this technique for all groups of patients have shown oocyte aspiration efficiencies of up to 90% per follicle aspirated. Our study clearly shows that fewer follicles will develop in FP patients, therefore optimum aspiration is essential. The poor number of ova retrieved here will have contributed to the lower pregnancy rate in the study. Ovarian relocation surgery may be indicated in this group. This would improve laparoscopic access and could enhance follicular development by removing the mechanical factors that may be preventing it. Some success with this has been noted in our program. FP patients represent a significant challenge for any IVF unit. Careful assessment and a test treatment cycle may be a prelude to more aggressive management in an attempt to improve the poor chance of pregnancy that seems to be the fate of these patients despite the rapid developments in the IVF technique. Acknowledgments. We thank Mrs. Anne McCartin and Ms. Lyn Parker for their typing assistance, Ms. Agnes Soo for her assistance with the computer matching, and Drs. Hugh Robinson and Lachlan de Crespigny for their excellent work in providing ultrasound services. REFERENCES 1. Edwards RG: In vitro fertilization and embryo replacement. In Proceedings of III World Congress of In Vitro Fertilization and Embryo Transfer, Opening Lecture on In Vitro Fertilization and Embryo Transfer, Edited by M Seppala, RG Edwards. Ann NY Acad Sci 442:1, Molloy et al. The "frozen pelvis" in an IVF program Fertility and Sterility

6 2. Hamberger L, Wikland M: Clinical experience with ultrasound guided follicle aspiration. In Recent Progress in Human In Vitro Fertilization, Edited by W Feichtinger, P Kemeter. Cofese, Palermo, 1984, p World Health Organization, Task Force on Methods for The Determination of the Fertile Period, Special Program of Research, Development and Research Training in Human Reproduction: Temporal relationships between ovulation and defined changes in the concentration of plasma estradiol-17f3 luteinizing hormone, follicle-stimulating hormone and progesterone. I. Probit analysis. Am J Obstet Gynecol 128:383, Hager WD, Eschenbach DA, Spence MR, Sweet RL: Criteria for diagnosis and grading of salpingitis. Obstet Gynecol 61:113, Fowles MR, Malinak LR, Bury R, Poindexter A: Endometriosis and anovulation: a coexisting problem in the infertile female. Am J Obstet Gynecol 125:412, Feichtinger W, Kemeter P: In vitro fertilization and embryo transfer: an outpatient/office procedure. In Recent Progress in Human In Vitro Fertilization, Edited by W Feichtinger, P Kemeter. Cofese, Palermo, 1984, p 285 Molloy et al. The "frozen pelvis" in an IVF program 455