FERTILITY AND STERILITY
Vol. 58, No.2, August 1992
Printed on acid·free paper in U.S.A.
Copyright I!) 1992 The American Fertility Society
The administration of glucocorticoids for the prevention of ovarian hyperstimulation syndrome in in vitro fertilization: a prospective randomized study Seang-Lin Tan, M.RC.O.G., M.Med.*t:j: Adam Balen, M.RC.O.G. *§ Elsir El Hussein, M.RC.O.G.*
Stuart Campbell, F.RC.O.G.t Howard S. Jacobs, M.D.§
The Hallam Medical Centre, King's College School of Medicine and Dentistry and The Middlesex Hospital, London, United Kingdom
Objective: To determine if the administration of glucocorticoids reduced the rate of ovarian hyperstimulation syndrome (OHSS) in high-risk patients after ovarian stimulation for in vitro fertilization (IVF). Design: Prospective randomized study. Patients: Thirty-one patients who were stimulated with human menopausal gonadotropin (hMG) after pituitary desensitization by gonadotropin-releasing hormone agonist and who developed >20 follicles> 12 mm and/or had a serum estradiol (E 2) level of> 10,000 pmol/L on the day of administration of human chorionic gonadotropin (hCG). Interventions: Patients were randomly divided into two groups. Those who were randomized to receive glucocorticoids (group A) (n = 17) were administered intravenous hydrocortisone, 100 mg, immediately after ultrasound (US)-directed oocyte recovery. Prednisolone, 10 mg three times per day, was given for 5 days starting on the day of oocyte recovery followed by prednisolone 10 mg two times a day for 3 days and 10 mg/d for 2 days. Those in group B (n = 14) did not receive any glucocorticoid treatment. In both groups, luteal support was provided by intramuscular injections of gestone 100 mg/d. Results: The two groups of patients were comparable in terms of age, duration of infertility, and total dose of hMG used. All had polycystic ovaries on US examination. On the day of hCG administration, the mean number of follicles in the two groups were 26.76 ± 2.49 and 25.93 ± 1.44 and the serum E2 concentration 13,404 ± 710 and 13,915 ± 901 pmol/L, respectively. There were no significant differences in the number of oocytes collected or in the fertilization, cleavage, and implantation rates in the two groups. The pregnancy rates per initiated cycle were 41.18% and 35.71 %, respectively. Seven of the 17 patients (41.2%) who received glucocorticoids developed ovarian hyperstimulation syndrome compared with 6 of the 14 patients (42.9%) who did not receive glucocorticoids. Conclusions: Administration of glucocorticoids to high risk patients did not reduce the rate of OHSS after ovarian stimulation for IVF. Fertil Steril1992;58:378-83 Key Words: Glucocorticoids, ovarian hyperstimulation syndrome, in vitro fertilization
Received February 10, 1992; revised and accepted April 20, 1992. * The Hallam Medical Centre. t Department of Obstetrics and Gynaecology, King's College School of Medicine and Dentistry. :j: Reprint requests: Seang-Lin Tan, M.R.C.O.G., M. Med., Department of Obstetrics and Gynaecology, King's College School of Medicine and Dentistry, Denmark Hill, London SE5 8RX, United Kingdom. § Department of Reproductive Endocrinology, The Middlesex Hospital.
378
Tan et al.
Ovarian hyperstimulation and IVF
The most serious complication of gonadotropin stimulation is the ovarian hyperstimulation syndrome (ORSS). In its severe form, it is characterized by ovarian enlargement with cyst formation, ascites, pleural effusion, and electrolyte disturbances (1). Oliguria, vascular complications (2), and even fatalities have been reported (3). It was previously believed that although ORSS was a major problem in in vivo induction of ovulation, it was very uncommon during ovarian stimulation for in vitro fertilization
Fertility and Sterility
(IVF). Unfortunately, although aspiration of ovarian follicles during oocyte recovery may reduce the incidence of OHSS, it does not completely prevent its occurrence (4). Since the introduction of gonadotropin-releasing hormone agonists (GnRH-a) for pituitary desensitization before ovarian stimulation for IVF (5), the method has gained widespread popularity and is used as the predominant, if not the only, method of stimulation in many IVF programs today. Although it was initially believed that pituitary desensitization would reduce the risk of OHSS, it has not proved to be the case. In fact, recent data suggest the contrary and that the use of GnRH -a is associated with an increased incidence of OHSS (6, 7). There is, at present, no one single method that allows OHSS to be reliably predicted and prevented. It is, however, generally accepted that patients who develop many preovulatory follicles and who have a high serum estradiol (E 2 ) concentration are at greatest risk of developing OHSS. The use of prednisolone supplementation of clomiphene citrate (CC) and gonadotropin stimulation for IVF has been suggested to increase the pregnancy rate (PR) (8). More recently, several large groups, for example, at Bourn Hall Clinic in Cambridge, have administered glucocorticoids to patients considered at high risk of developing OHSS in an attempt to reduce the incidence of this condition (Brinsden P, personal communications 1990). There has, however, been no randomized study that has addressed this issue. The present study was, therefore, designed to determine if the administration of glucocorticoids to patients at high risk of developing OHSS reduces the risk of this complication. MATERIALS AND METHODS
Over the 10-month period from September 1, 1990 to June 30, 1991, patients undergoing IVF treatment at the Hallam Medical Centre who were considered to be at high risk of developing OHSS were recruited for this study that had been approved by the Institutional Review Board. Preliminary statistical analysis indicated that a sample size of 15 in each group would give a 90% power to detect a difference of 20% in the incidence of OHSS at a significance level of 5%. All patients were administered subcutaneous (SC) buserelin acetate (Suprefact; Hoechst, Hounslow, United Kingdom), 500 ~g/d from day 1 ofthe menstrual cycle to desensitize the pituitary gland. When
Vol. 58, No.2, August 1992
pituitary desensitization had been achieved at least 14 days later, as shown by the concentration of serum E2 < 100 pmol/L and luteinizing hormone < 5 IU/L together with the absence of ovarian follicular activity on pelvic ultrasound (US), daily intramuscular (1M) injections of human menopausal gonadotropin (hMG, Pergonal; Serono, Welwyn Garden City, United Kingdom) were commenced. The dose of buserelin acetate was decreased to 200 ~g/d and continued until, and including, the day of administration of human chorionic gonadotropin (hCG, Profasi; Serono, Welwyn Garden City, United Kingdom). Ovarian stimulation was monitored by daily US assessment of follicular growth together with estimation of daily serum E2 concentrations. On the day of hCG administration, patients who had developed >20 follicles> 12 mm and/or had a serum E2 concentration of >10,000 pmol/L were randomized into two groups. This was done by drawing serially numbered sealed envelopes, each of which contained a study group number (1 or 2) that had been allocated by reference to tables of random numbers. All patients had transvaginal US-directed oocyte recovery performed 35 hours after administration ofhCG 10,000 IU. In all cases, every follicle in both ovaries that could be seen on US scan was aspirated completely. Patients in group 1 were administered intravenous hydrocortisone 100 mg immediately after oocyte recovery and prednisolone 10 mg three times per day for 5 days followed by 10 mg twice per day for 3 days and 10 mg/d for 2 days. Patients in group 2 did not receive any treatment with glucocorticoids. Embryo transfer (ET) was performed 2 days after oocyte recovery. Luteal support was provided in all cases by daily 1M 100-mg injections of gestone. All patients were seen 5 days after ET to determine if they had developed clinical hyperstimulation, and an US scan was also performed. The classification of OHSS into mild, moderate, or severe was made according to the criteria of Schenker and Weinstein (1). The patients were then followed up every few days until they had a menstrual period or until the OHSS had subsided in those who developed the condition. The results of the two groups of patients were compared using two-sample t-test and X2 test. A P < 0.05 value was accepted as significant. RESULTS
There were 17 patients in group 1 and 14 in group 2; their indications for IVF are shown in Table 1.
Tan et aI.
Ovarian hyperstimulation and IVF
379
Table 1
Indications for IVF in the Two Groups of Patients Group 1
Group 2
6 6 5
6 (NS*) 4
Tubal damage Unexplained infertility Male factor infertility
4
Table 3
Implantation and PRs
No. of sacs Implantation rate (%) PR/initiated cycle (%) PR/ET (%)
Steroids
No steroids
0.59 ± 0.19* 21.74 41.18 43.75
0.57 ± 0.23 (NS t) 20.51 35.71 35.71
* NS, not significant.
* Values are means ± t NS, not significant.
There was no significant difference in the two groups with regard to age, duration of infertility, and number of previous IVF attempts (Table 2). The dose of hMG used and the day of heG administration were comparable in the two groups of patients (Table 2). The mean serum E2 concentration in group 1 was 13,404 ± 710 pmoljL compared with 13,915 ± 901 pmoljL in group 2. The mean number of follicles > 12 mm and mean number of oocytes collected were 26.76 ± 2.49 and 21.65 ± 2.05 in those who received glucocorticoids as compared with 25.93 ± 1.44 and 19.71 ± 1.55 in those who did not receive glucocorticoids. The mean number of viable oocytes and the
Table 2
Clinical Characteristics and Outcome of Patients Steroids
No. of patients Age (y) No. of previous IVF attempts Length of infertility (y) Total hMG (ampules) Day ofhCG administration Maximum E2 concentration (pmol/L) No. of follicles No. of oocytes collected No. of viable oocytes No. of fertilized oocytes Fertilization rate (%)
No. of cleaved embryos Cleavage rate (%)
No.ofET No. of frozen embryos
17 30.29 ± 1.26 *
No steroids 14 32.93 ± 1.07 (NS t)
1.0 ± 0.41
0.43 ± 0.17
5.77 ± 0.65
5.36 ± 2.76
28.47 ± 2.10
27.07 ± 1.06
10.47 ± 0.26
10.29 ± 0.99
13,404 ± 710 26.76 ± 2.49
13,915 ± 901 25.93 ± 1.44
21.65 ± 2.05
19.71 ± 1.55
18.47 ± 1.61
17.93 ± 1.43
10.59 ± 1.38
13.0 ± 1.65
59.03 ± 6.59
72.82 ± 6.42
7.5 ± 1.07
8.38 ± 1.53
76.13 ± 5.71 2.71 ± 0.21
67.67 ± 8.12 2.79 ± 0.11
2.69 ± 1.21
4.36 ± 1.32
* Values are means ± SE. t NS, not significant.
380
Tan et al.
Ovarian hyperstimulation and IVF
SE.
fertilization and cleavage rates were comparable in the two groups of patients (Table 2). The mean number of embryos transferred was 2.71 ± 0.21 in group 1 compared with 2.79 ± 0.11 in group 2. The implantation rate was 21.74% in group 1 compared with 20.51 % in group 2 (Table 3), and there was no significant difference in the mean number of gestational sacs (0.59 ± 0.19 versus 0.57 ± 0.23) seen in the two groups of patients. The clinical PRs per initiated cycle and per ET were 41.18% and 43.75% in those who received glucocorticoids compared with 35.71% and 35.71% in those who did not receive glucocorticoids (Table 3). These differences were not statistically significant. Table 4 shows the number of cases of OHSS in the two groups of patients. Seven of the 17 patients (41.2%) who received glucocorticoids developed OHSS compared with 6 of the 14 patients (42.9%) who did not. The difference was not statistically significant. One patient in each group developed severe OHSS. The patient in group 1 did not become pregnant, whereas the patient in group 2 achieved a twin pregnancy. DISCUSSION
The ovarian hyperstimulation syndrome is a rare but serious complication of treatment with gonad-
Table 4
Incidence of OHSS
OHSS None Mild Moderate Severe Total incidence of 0 HSS (%) Incidence of moderate or severe OHSS (%)
Steroids
No steroids
10 (5) * 4 (2) 2 (0) 1 (0) 41.2
8 (3) (NSt) 4 (1) 1 (0) 1 (1) 42.9
17.65
14.29
* Number of clinical pregnancies are shown in parentheses. t NS, not significant.
Fertility and Sterility
otropins. The precise etiologic and pathophysiological factors underlying this condition remain uncertain, although recent evidence has pointed to the existence of an ovarian renin-angiotension system that may be a prime mover in its generation (9). There are also some recent data that suggest that there is a reduction in the colloid osmotic gradient (plasma colloid osmotic pressure-interstitial colloid osmotic pressure) in women who have ovarian stimulation for IVF that might be mediated by the exaggerated levels of serum E 2 , causing an increased capillary permeability to plasma proteins (10). Friedman et al. (11) described the first case of severe OHSS after ovarian stimulation for IVF. When GnRH -a was initially incorporated into ovarian stimulation regimens, it was suggested that their use may reduce the incidence of OHSS. This was based on the observation that OHSS is rare in patients with hypo gonadotrophic hypogonadism, and it was hoped that by converting the patient to a hypogonadal state, OHSS could be prevented (12), perhaps by a direct action of GnRH -a on the ovary to inhibit steroidogenesis and gonadotropin receptor formation (13). Unfortunately, although the use of GnRH-a, especially in the long protocol in which pituitary desensitization is achieved before hMG is administered (14), has resulted in significant practical advantages for IVF programs (15), it does not reduce the incidence of OHSS. In fact, several studies have recently suggested that there may be an increased incidence of OHSS when GnRH -a are used (6,7). In our own unit, we have recently found that the prevalence of severe OHSS was significantly higher when GnRH-a and hMG were used compared with hMG and CC (1.1 % versus 0.2%, P < 0.05) (16). Because there is no definitive treatment of OHSS and management is symptomatic rather than curative, prevention is very important. Unfortunately, there is at present no one single method that allows OHSS to be reliably prevented. Although OHSS is generally associated with high preovulatory estrogen (E) levels (4, 6), OHSS may sometimes occur in the presence of normal or even low E levels. On the other hand, patients with high E levels do not always develop OHSS. Similarly, a large number of ovarian follicles detected on pelvic US on the day of hCG administration is associated with an increased risk of OHSS (17), but again this is not an invariable finding. Nevertheless, it is generally accepted that the combination of a large number of ovarian follicles with high preovulatory serum E2 concentrations does place the patient at a high risk of devel-
Vol. 58, No.2, August 1992
oping the syndrome. For example, the incidence of moderate or severe OHSS in the present study of 16.13% (5 of 31 patients) was much higher than the normal incidence observed in our clinic, which was 1.15% for all patients undergoing IVF (16). Recently, several large centers have reported that treatment with glucocorticoids may be effective in the prevention of OHSS. Because the effects of these steroids have not been systematically evaluated for this indication, we undertook the present prospective randomized study to address this issue. The two groups of patients were comparable in terms of age, number of previous IVF attempts, and number of years of infertility. They received a similar number of ampules ofhMG, developed a comparable number of follicles, and had a similar levels of serum E2 on the day of hCG administration. The administration of glucocorticoids had no effect on the outcome of the treatment cycles. The incidence of OHSS was comparable in both groups, irrespective of whether all degrees of OHSS (41.2% versus 42.9%) or only moderate and severe OHSS were considered (23.53% versus 21.43%). Interestingly, the administration of glucocorticoids did not affect the implantation rates observed in the two groups. This result differs from the report of Kemeter and Feichtinger (8), although ovarian stimulation in that study involved the use of CC and hMG, whereas in our study GnRH -a and hMG were used instead. At present, there are only a few strategies for the management of patients who are found to have high preovulatory levels of serum E2 and many preovulatory and medium sized follices on the day of hCG administration. First, hCG can be withheld, the patient advised to avoid coitus (if she has nontubal infertility), the cycle abandoned, and a new cycle commenced lor 2 months later. This would appear to be the safest option in terms of preventing OHSS. However, given the difficulties of treatment as well as the emotional and financial commitment invested by the patient, many patients and clinicians are reluctant to embark on this course of action. Moreover, this policy will result in cancellation of a high proportion of cycles that will not progress to clinical hyperstimulation. Second, hCG could be withheld, the patient advised to avoid coitus if she has nontubal infertility, and the GnRH-a continued until pituitary desensitization is achieved, at which point hMG is re-commenced at a lower dose (6). There are two problems with this approach. First, it may take anywhere between 24 to 56 days between discontinuing hMG and re-commencing follicular
Tan et al.
Ovarian hyperstimuiation and IVF
381
stimulation with a lower dose of gonadotropin (6). Second, repeat stimulation with a lower dose of hMG may still lead to excessive stimulation or, on the other hand, it may lead to inadequate response of the ovaries (18). A third strategy is to proceed with oocyte collection and IVF but to avoid ET in the primary treatment cycle. Instead, all the preembryos are electively cryopreserved for transfer in a subsequent cycle (19, 20). The rationale for this approach is that OHSS is more common in conception cycles, which may be associated with a more severe and prolonged course. The problem with this approach is that although the incidence and severity of OHSS may be reduced, the possibility of its occurrence is not eliminated. Moreover, there is recent evidence to suggest that frozen-thawed embryos resulting from GnRH-a/hMG-treated cycles have a greatly reduced pregnancy potential compared with embryos from hMG-stimulated cycles, even in centers with established cryopreservation facilities (21). There is, therefore, still no ideal method of preventing OHSS. Because polycystic ovaries are a definite high-risk factor (MacDougall MJ, Tan SL, Jacobs HS, unpublished observations, 1992), it would be prudent to assess the ovarian morphology by US scan at the commencement of treatment. If polycystic ovaries are seen, the dose of gonadotropins used should be initially low and cautiously increased if necessary. If, despite these precautions, many preovulatory follicles and a high serum E2 level are noted on the day of presumed heG administration, the choice of which ofthe three strategies described above will have to be carefully considered for the individual patient. In practice, the decision will be influenced to a large extent by the results of cryopreservation in the particular IVF program. If a decision is made to proceed with oocyte recovery, every follicle, including small and medium sized ones, should be emptied. This is because these small and medium sized follicles contribute significantly to the risk of developing OHSS, more so than the number of large follicles (17). Similarly, luteal support should be with progesterone only, and heG injections should be avoided. In this study, these two precautions were undertaken, which probably accounts for the relatively low incidence of moderate to severe OHSS in both groups of patients (2 of 31, [6.45%] incidence of severe and 5 of 31, 16.13% incidence of moderate to severe OHSS). In contrast, in the report by Forman et al. (6) in which heG for luteal support was used in some of the patients developing OHSS, it was predicted that the probability
382
Tan et al.
Ovarian hyperstimulation and IVF
of patients who have 15 or more follicles at least 12 mm in diameter associated with plasma E2 of 2,000 pg/mL or more, developing severe OHSS would be 27%. It therefore appears that at the present time there is still no ideal strategy for preventing the development of OHSS, which remains an enigma in need of a solution. Nevertheless, our study suggests that glucocorticoids have no role to play in the prevention of the development of this condition. REFERENCES 1. Schenker JG, Weinstein D. Ovarian hyperstimulation syndrome: a current survey. Fertil Steril 1978;30:255-68. 2. Kaaja R, Siegberg R, Titinen A, Koskimies A. Severe ovarian hyperstimulation and deep vein thrombosis. Lancet 1989;2: 1043. 3. Moses M, Bogowsky H, Anteby E, Lunenfied B, Rabau E, Serr D, et al. Thromboembolic phenomena after ovarian stimulation with human menopausal gonadotropins. Lancet 1965;2:1213. 4. Golan A, Ron-EI R, Herman A, Weinraub Z, Soffer Y, Caspi E. Ovarian hyperstimulation following D-Trp-6 luteinizing hormone-releasing hormone microcapsules and menotropin for in vitro fertilization. Fertil Steril 1988;50:912-6. 5. Porter RN, Smith W, Craft IL, Abdulwahid NA, Jacobs HS. Induction of ovulation for in vitro fertilization using buserelin and gonadotropins. Lancet 1984;2:1284-5. 6. Forman RG, Frydman R, Egan D, Ross C, Barlow DH. Severe ovarian hyperstimulation syndrome using agonists of gonadotropin-releasing hormone for in vitro fertilization: a European series and a proposal for prevention. Fertil Steril1990;53: 502-9. 7. Tanbo T, Dale PO, Kjekshus E, Haug E, Abyholm T. Stimulation with human menopausal gonadotropin versus folliclestimulating hormone after pituitary suppression in polycystic ovarian syndrome. Fertil Steril 1990;53:798-803. 8. Kemeter P, Feichtinger W. Prednisolone supplementation to clomid and/or gonadotropin stimulation for in-vitro fertilization-a prospective randomized trial. Hum Reprod 1986;1:441-4. 9. Ong ACM, Rennie DP, Homburg R, Lachelin GCL, Jacobs HS, Slater JDH. The pathogenesis of ovarian hyperstimulation syndrome: a possible role for renin. Clin Endocrinol (Oxf) 1991;34:43-9. 10. Tollan A, Holst N, Forsdahl F, Fadnes HO, Oian P, Malthau JM. Transcapillary fluid dynamics during ovarian stimulation for in vitro fertilization. Am J Obstet Gynecol 1990;162:554-8. 11. Friedman RG, Schmidt GE, Chang FE, Kim MH. Severe ovarian hyperstimulation following follicular aspiration. Am J Obstet GynecoI1984;150:436-7. 12. Salat-Baroux J, Alvarez S, Antoine JM, Tibi C, Cornet D, Mandelbaum J, et al. Comparison between long and short protocols of LHRH agonist in the treatment of polycystic ovary disease by in vitro fertilization. Hum Reprod 1988;3: 535-9. 13. Hsueh AJW, Jones PBC. Extrapituitary actions of gonadotropin-releasing hormone. Endocr Rev 1981;2:437-61. 14. Tan SL, Kingsland C, Campbell S, Mills C, Bradfield J, Alexander N, et al. The long protocol of administration of go-
Fertility and Sterility
nadotropin-releasing hormone agonist is superior to the short protocol for ovarian stimulation for in vitro fertilization. Fertil Steril 1992;57:810-4. 15. Tan SL, Balen A, Hussein EI, Mills C, Campbell S, Yovich J, et al. A prospective randomized study of the optimum timing of human chorionic gonadotropin administration after pituitary desensitization in in vitro fertilization. Fertil Steril. 1992;57:1259-64. 16. MacDougall MJ, Tan SL, Jacobs HS. In vitro fertilization and the ovarian hyperstimulation syndrome. Hum Reprod. 1992;7:597-600. 17. Blackstein J, Shalev J, Saadon T, Kukia EE, Rabinovici J, Pariete C, et al. Ovarian hyperstimulation syndrome: prediction by number and size of preovulatory ovarian follicles. Fertil Steril1987;47:597-602.
Vol. 58, No.2, August 1992
18. Amso NN, Shaw RW, Ahuja KK, Morris N. Prevention of ovarian hyperstimulation syndrome. Fertil Steril 1991;55: 220-1. 19. Amso NN, Ahuja KK, Morris N, Shaw RW. The management of predicted ovarian hyperstimulation involving gonadotropin-releasing hormone analog with effective cryopreservation of all pre-embryos. Fertil Steril1990;53:1087-90. 20. Salat-Baroux J, Alvarez S, Antoine JM, Cornet D, Tibi Ch, Plachot M, et al. Treatment of hyperstimulation during invitro fertilization. Hum Reprod 1990;5:36-9. 21. Testart J, Forman R, Belaisch-Allart J, Volante M, Hazout A, Strubb N, et al. Embryo quality and uterine receptivity in in vitro fertilization cycles with and without agonists of gonadotropin releasing hormone. Hum Reprod 1989;4:198201.
Tan et al.
Ovarian hyperstimulation and IVF
383
Vol. 58, No.2, August 1992
Printed on acid·free paper in U.S.A.
Copyright I!) 1992 The American Fertility Society
The administration of glucocorticoids for the prevention of ovarian hyperstimulation syndrome in in vitro fertilization: a prospective randomized study Seang-Lin Tan, M.RC.O.G., M.Med.*t:j: Adam Balen, M.RC.O.G. *§ Elsir El Hussein, M.RC.O.G.*
Stuart Campbell, F.RC.O.G.t Howard S. Jacobs, M.D.§
The Hallam Medical Centre, King's College School of Medicine and Dentistry and The Middlesex Hospital, London, United Kingdom
Objective: To determine if the administration of glucocorticoids reduced the rate of ovarian hyperstimulation syndrome (OHSS) in high-risk patients after ovarian stimulation for in vitro fertilization (IVF). Design: Prospective randomized study. Patients: Thirty-one patients who were stimulated with human menopausal gonadotropin (hMG) after pituitary desensitization by gonadotropin-releasing hormone agonist and who developed >20 follicles> 12 mm and/or had a serum estradiol (E 2) level of> 10,000 pmol/L on the day of administration of human chorionic gonadotropin (hCG). Interventions: Patients were randomly divided into two groups. Those who were randomized to receive glucocorticoids (group A) (n = 17) were administered intravenous hydrocortisone, 100 mg, immediately after ultrasound (US)-directed oocyte recovery. Prednisolone, 10 mg three times per day, was given for 5 days starting on the day of oocyte recovery followed by prednisolone 10 mg two times a day for 3 days and 10 mg/d for 2 days. Those in group B (n = 14) did not receive any glucocorticoid treatment. In both groups, luteal support was provided by intramuscular injections of gestone 100 mg/d. Results: The two groups of patients were comparable in terms of age, duration of infertility, and total dose of hMG used. All had polycystic ovaries on US examination. On the day of hCG administration, the mean number of follicles in the two groups were 26.76 ± 2.49 and 25.93 ± 1.44 and the serum E2 concentration 13,404 ± 710 and 13,915 ± 901 pmol/L, respectively. There were no significant differences in the number of oocytes collected or in the fertilization, cleavage, and implantation rates in the two groups. The pregnancy rates per initiated cycle were 41.18% and 35.71 %, respectively. Seven of the 17 patients (41.2%) who received glucocorticoids developed ovarian hyperstimulation syndrome compared with 6 of the 14 patients (42.9%) who did not receive glucocorticoids. Conclusions: Administration of glucocorticoids to high risk patients did not reduce the rate of OHSS after ovarian stimulation for IVF. Fertil Steril1992;58:378-83 Key Words: Glucocorticoids, ovarian hyperstimulation syndrome, in vitro fertilization
Received February 10, 1992; revised and accepted April 20, 1992. * The Hallam Medical Centre. t Department of Obstetrics and Gynaecology, King's College School of Medicine and Dentistry. :j: Reprint requests: Seang-Lin Tan, M.R.C.O.G., M. Med., Department of Obstetrics and Gynaecology, King's College School of Medicine and Dentistry, Denmark Hill, London SE5 8RX, United Kingdom. § Department of Reproductive Endocrinology, The Middlesex Hospital.
378
Tan et al.
Ovarian hyperstimulation and IVF
The most serious complication of gonadotropin stimulation is the ovarian hyperstimulation syndrome (ORSS). In its severe form, it is characterized by ovarian enlargement with cyst formation, ascites, pleural effusion, and electrolyte disturbances (1). Oliguria, vascular complications (2), and even fatalities have been reported (3). It was previously believed that although ORSS was a major problem in in vivo induction of ovulation, it was very uncommon during ovarian stimulation for in vitro fertilization
Fertility and Sterility
(IVF). Unfortunately, although aspiration of ovarian follicles during oocyte recovery may reduce the incidence of OHSS, it does not completely prevent its occurrence (4). Since the introduction of gonadotropin-releasing hormone agonists (GnRH-a) for pituitary desensitization before ovarian stimulation for IVF (5), the method has gained widespread popularity and is used as the predominant, if not the only, method of stimulation in many IVF programs today. Although it was initially believed that pituitary desensitization would reduce the risk of OHSS, it has not proved to be the case. In fact, recent data suggest the contrary and that the use of GnRH -a is associated with an increased incidence of OHSS (6, 7). There is, at present, no one single method that allows OHSS to be reliably predicted and prevented. It is, however, generally accepted that patients who develop many preovulatory follicles and who have a high serum estradiol (E 2 ) concentration are at greatest risk of developing OHSS. The use of prednisolone supplementation of clomiphene citrate (CC) and gonadotropin stimulation for IVF has been suggested to increase the pregnancy rate (PR) (8). More recently, several large groups, for example, at Bourn Hall Clinic in Cambridge, have administered glucocorticoids to patients considered at high risk of developing OHSS in an attempt to reduce the incidence of this condition (Brinsden P, personal communications 1990). There has, however, been no randomized study that has addressed this issue. The present study was, therefore, designed to determine if the administration of glucocorticoids to patients at high risk of developing OHSS reduces the risk of this complication. MATERIALS AND METHODS
Over the 10-month period from September 1, 1990 to June 30, 1991, patients undergoing IVF treatment at the Hallam Medical Centre who were considered to be at high risk of developing OHSS were recruited for this study that had been approved by the Institutional Review Board. Preliminary statistical analysis indicated that a sample size of 15 in each group would give a 90% power to detect a difference of 20% in the incidence of OHSS at a significance level of 5%. All patients were administered subcutaneous (SC) buserelin acetate (Suprefact; Hoechst, Hounslow, United Kingdom), 500 ~g/d from day 1 ofthe menstrual cycle to desensitize the pituitary gland. When
Vol. 58, No.2, August 1992
pituitary desensitization had been achieved at least 14 days later, as shown by the concentration of serum E2 < 100 pmol/L and luteinizing hormone < 5 IU/L together with the absence of ovarian follicular activity on pelvic ultrasound (US), daily intramuscular (1M) injections of human menopausal gonadotropin (hMG, Pergonal; Serono, Welwyn Garden City, United Kingdom) were commenced. The dose of buserelin acetate was decreased to 200 ~g/d and continued until, and including, the day of administration of human chorionic gonadotropin (hCG, Profasi; Serono, Welwyn Garden City, United Kingdom). Ovarian stimulation was monitored by daily US assessment of follicular growth together with estimation of daily serum E2 concentrations. On the day of hCG administration, patients who had developed >20 follicles> 12 mm and/or had a serum E2 concentration of >10,000 pmol/L were randomized into two groups. This was done by drawing serially numbered sealed envelopes, each of which contained a study group number (1 or 2) that had been allocated by reference to tables of random numbers. All patients had transvaginal US-directed oocyte recovery performed 35 hours after administration ofhCG 10,000 IU. In all cases, every follicle in both ovaries that could be seen on US scan was aspirated completely. Patients in group 1 were administered intravenous hydrocortisone 100 mg immediately after oocyte recovery and prednisolone 10 mg three times per day for 5 days followed by 10 mg twice per day for 3 days and 10 mg/d for 2 days. Patients in group 2 did not receive any treatment with glucocorticoids. Embryo transfer (ET) was performed 2 days after oocyte recovery. Luteal support was provided in all cases by daily 1M 100-mg injections of gestone. All patients were seen 5 days after ET to determine if they had developed clinical hyperstimulation, and an US scan was also performed. The classification of OHSS into mild, moderate, or severe was made according to the criteria of Schenker and Weinstein (1). The patients were then followed up every few days until they had a menstrual period or until the OHSS had subsided in those who developed the condition. The results of the two groups of patients were compared using two-sample t-test and X2 test. A P < 0.05 value was accepted as significant. RESULTS
There were 17 patients in group 1 and 14 in group 2; their indications for IVF are shown in Table 1.
Tan et aI.
Ovarian hyperstimulation and IVF
379
Table 1
Indications for IVF in the Two Groups of Patients Group 1
Group 2
6 6 5
6 (NS*) 4
Tubal damage Unexplained infertility Male factor infertility
4
Table 3
Implantation and PRs
No. of sacs Implantation rate (%) PR/initiated cycle (%) PR/ET (%)
Steroids
No steroids
0.59 ± 0.19* 21.74 41.18 43.75
0.57 ± 0.23 (NS t) 20.51 35.71 35.71
* NS, not significant.
* Values are means ± t NS, not significant.
There was no significant difference in the two groups with regard to age, duration of infertility, and number of previous IVF attempts (Table 2). The dose of hMG used and the day of heG administration were comparable in the two groups of patients (Table 2). The mean serum E2 concentration in group 1 was 13,404 ± 710 pmoljL compared with 13,915 ± 901 pmoljL in group 2. The mean number of follicles > 12 mm and mean number of oocytes collected were 26.76 ± 2.49 and 21.65 ± 2.05 in those who received glucocorticoids as compared with 25.93 ± 1.44 and 19.71 ± 1.55 in those who did not receive glucocorticoids. The mean number of viable oocytes and the
Table 2
Clinical Characteristics and Outcome of Patients Steroids
No. of patients Age (y) No. of previous IVF attempts Length of infertility (y) Total hMG (ampules) Day ofhCG administration Maximum E2 concentration (pmol/L) No. of follicles No. of oocytes collected No. of viable oocytes No. of fertilized oocytes Fertilization rate (%)
No. of cleaved embryos Cleavage rate (%)
No.ofET No. of frozen embryos
17 30.29 ± 1.26 *
No steroids 14 32.93 ± 1.07 (NS t)
1.0 ± 0.41
0.43 ± 0.17
5.77 ± 0.65
5.36 ± 2.76
28.47 ± 2.10
27.07 ± 1.06
10.47 ± 0.26
10.29 ± 0.99
13,404 ± 710 26.76 ± 2.49
13,915 ± 901 25.93 ± 1.44
21.65 ± 2.05
19.71 ± 1.55
18.47 ± 1.61
17.93 ± 1.43
10.59 ± 1.38
13.0 ± 1.65
59.03 ± 6.59
72.82 ± 6.42
7.5 ± 1.07
8.38 ± 1.53
76.13 ± 5.71 2.71 ± 0.21
67.67 ± 8.12 2.79 ± 0.11
2.69 ± 1.21
4.36 ± 1.32
* Values are means ± SE. t NS, not significant.
380
Tan et al.
Ovarian hyperstimulation and IVF
SE.
fertilization and cleavage rates were comparable in the two groups of patients (Table 2). The mean number of embryos transferred was 2.71 ± 0.21 in group 1 compared with 2.79 ± 0.11 in group 2. The implantation rate was 21.74% in group 1 compared with 20.51 % in group 2 (Table 3), and there was no significant difference in the mean number of gestational sacs (0.59 ± 0.19 versus 0.57 ± 0.23) seen in the two groups of patients. The clinical PRs per initiated cycle and per ET were 41.18% and 43.75% in those who received glucocorticoids compared with 35.71% and 35.71% in those who did not receive glucocorticoids (Table 3). These differences were not statistically significant. Table 4 shows the number of cases of OHSS in the two groups of patients. Seven of the 17 patients (41.2%) who received glucocorticoids developed OHSS compared with 6 of the 14 patients (42.9%) who did not. The difference was not statistically significant. One patient in each group developed severe OHSS. The patient in group 1 did not become pregnant, whereas the patient in group 2 achieved a twin pregnancy. DISCUSSION
The ovarian hyperstimulation syndrome is a rare but serious complication of treatment with gonad-
Table 4
Incidence of OHSS
OHSS None Mild Moderate Severe Total incidence of 0 HSS (%) Incidence of moderate or severe OHSS (%)
Steroids
No steroids
10 (5) * 4 (2) 2 (0) 1 (0) 41.2
8 (3) (NSt) 4 (1) 1 (0) 1 (1) 42.9
17.65
14.29
* Number of clinical pregnancies are shown in parentheses. t NS, not significant.
Fertility and Sterility
otropins. The precise etiologic and pathophysiological factors underlying this condition remain uncertain, although recent evidence has pointed to the existence of an ovarian renin-angiotension system that may be a prime mover in its generation (9). There are also some recent data that suggest that there is a reduction in the colloid osmotic gradient (plasma colloid osmotic pressure-interstitial colloid osmotic pressure) in women who have ovarian stimulation for IVF that might be mediated by the exaggerated levels of serum E 2 , causing an increased capillary permeability to plasma proteins (10). Friedman et al. (11) described the first case of severe OHSS after ovarian stimulation for IVF. When GnRH -a was initially incorporated into ovarian stimulation regimens, it was suggested that their use may reduce the incidence of OHSS. This was based on the observation that OHSS is rare in patients with hypo gonadotrophic hypogonadism, and it was hoped that by converting the patient to a hypogonadal state, OHSS could be prevented (12), perhaps by a direct action of GnRH -a on the ovary to inhibit steroidogenesis and gonadotropin receptor formation (13). Unfortunately, although the use of GnRH-a, especially in the long protocol in which pituitary desensitization is achieved before hMG is administered (14), has resulted in significant practical advantages for IVF programs (15), it does not reduce the incidence of OHSS. In fact, several studies have recently suggested that there may be an increased incidence of OHSS when GnRH -a are used (6,7). In our own unit, we have recently found that the prevalence of severe OHSS was significantly higher when GnRH-a and hMG were used compared with hMG and CC (1.1 % versus 0.2%, P < 0.05) (16). Because there is no definitive treatment of OHSS and management is symptomatic rather than curative, prevention is very important. Unfortunately, there is at present no one single method that allows OHSS to be reliably prevented. Although OHSS is generally associated with high preovulatory estrogen (E) levels (4, 6), OHSS may sometimes occur in the presence of normal or even low E levels. On the other hand, patients with high E levels do not always develop OHSS. Similarly, a large number of ovarian follicles detected on pelvic US on the day of hCG administration is associated with an increased risk of OHSS (17), but again this is not an invariable finding. Nevertheless, it is generally accepted that the combination of a large number of ovarian follicles with high preovulatory serum E2 concentrations does place the patient at a high risk of devel-
Vol. 58, No.2, August 1992
oping the syndrome. For example, the incidence of moderate or severe OHSS in the present study of 16.13% (5 of 31 patients) was much higher than the normal incidence observed in our clinic, which was 1.15% for all patients undergoing IVF (16). Recently, several large centers have reported that treatment with glucocorticoids may be effective in the prevention of OHSS. Because the effects of these steroids have not been systematically evaluated for this indication, we undertook the present prospective randomized study to address this issue. The two groups of patients were comparable in terms of age, number of previous IVF attempts, and number of years of infertility. They received a similar number of ampules ofhMG, developed a comparable number of follicles, and had a similar levels of serum E2 on the day of hCG administration. The administration of glucocorticoids had no effect on the outcome of the treatment cycles. The incidence of OHSS was comparable in both groups, irrespective of whether all degrees of OHSS (41.2% versus 42.9%) or only moderate and severe OHSS were considered (23.53% versus 21.43%). Interestingly, the administration of glucocorticoids did not affect the implantation rates observed in the two groups. This result differs from the report of Kemeter and Feichtinger (8), although ovarian stimulation in that study involved the use of CC and hMG, whereas in our study GnRH -a and hMG were used instead. At present, there are only a few strategies for the management of patients who are found to have high preovulatory levels of serum E2 and many preovulatory and medium sized follices on the day of hCG administration. First, hCG can be withheld, the patient advised to avoid coitus (if she has nontubal infertility), the cycle abandoned, and a new cycle commenced lor 2 months later. This would appear to be the safest option in terms of preventing OHSS. However, given the difficulties of treatment as well as the emotional and financial commitment invested by the patient, many patients and clinicians are reluctant to embark on this course of action. Moreover, this policy will result in cancellation of a high proportion of cycles that will not progress to clinical hyperstimulation. Second, hCG could be withheld, the patient advised to avoid coitus if she has nontubal infertility, and the GnRH-a continued until pituitary desensitization is achieved, at which point hMG is re-commenced at a lower dose (6). There are two problems with this approach. First, it may take anywhere between 24 to 56 days between discontinuing hMG and re-commencing follicular
Tan et al.
Ovarian hyperstimuiation and IVF
381
stimulation with a lower dose of gonadotropin (6). Second, repeat stimulation with a lower dose of hMG may still lead to excessive stimulation or, on the other hand, it may lead to inadequate response of the ovaries (18). A third strategy is to proceed with oocyte collection and IVF but to avoid ET in the primary treatment cycle. Instead, all the preembryos are electively cryopreserved for transfer in a subsequent cycle (19, 20). The rationale for this approach is that OHSS is more common in conception cycles, which may be associated with a more severe and prolonged course. The problem with this approach is that although the incidence and severity of OHSS may be reduced, the possibility of its occurrence is not eliminated. Moreover, there is recent evidence to suggest that frozen-thawed embryos resulting from GnRH-a/hMG-treated cycles have a greatly reduced pregnancy potential compared with embryos from hMG-stimulated cycles, even in centers with established cryopreservation facilities (21). There is, therefore, still no ideal method of preventing OHSS. Because polycystic ovaries are a definite high-risk factor (MacDougall MJ, Tan SL, Jacobs HS, unpublished observations, 1992), it would be prudent to assess the ovarian morphology by US scan at the commencement of treatment. If polycystic ovaries are seen, the dose of gonadotropins used should be initially low and cautiously increased if necessary. If, despite these precautions, many preovulatory follicles and a high serum E2 level are noted on the day of presumed heG administration, the choice of which ofthe three strategies described above will have to be carefully considered for the individual patient. In practice, the decision will be influenced to a large extent by the results of cryopreservation in the particular IVF program. If a decision is made to proceed with oocyte recovery, every follicle, including small and medium sized ones, should be emptied. This is because these small and medium sized follicles contribute significantly to the risk of developing OHSS, more so than the number of large follicles (17). Similarly, luteal support should be with progesterone only, and heG injections should be avoided. In this study, these two precautions were undertaken, which probably accounts for the relatively low incidence of moderate to severe OHSS in both groups of patients (2 of 31, [6.45%] incidence of severe and 5 of 31, 16.13% incidence of moderate to severe OHSS). In contrast, in the report by Forman et al. (6) in which heG for luteal support was used in some of the patients developing OHSS, it was predicted that the probability
382
Tan et al.
Ovarian hyperstimulation and IVF
of patients who have 15 or more follicles at least 12 mm in diameter associated with plasma E2 of 2,000 pg/mL or more, developing severe OHSS would be 27%. It therefore appears that at the present time there is still no ideal strategy for preventing the development of OHSS, which remains an enigma in need of a solution. Nevertheless, our study suggests that glucocorticoids have no role to play in the prevention of the development of this condition. REFERENCES 1. Schenker JG, Weinstein D. Ovarian hyperstimulation syndrome: a current survey. Fertil Steril 1978;30:255-68. 2. Kaaja R, Siegberg R, Titinen A, Koskimies A. Severe ovarian hyperstimulation and deep vein thrombosis. Lancet 1989;2: 1043. 3. Moses M, Bogowsky H, Anteby E, Lunenfied B, Rabau E, Serr D, et al. Thromboembolic phenomena after ovarian stimulation with human menopausal gonadotropins. Lancet 1965;2:1213. 4. Golan A, Ron-EI R, Herman A, Weinraub Z, Soffer Y, Caspi E. Ovarian hyperstimulation following D-Trp-6 luteinizing hormone-releasing hormone microcapsules and menotropin for in vitro fertilization. Fertil Steril 1988;50:912-6. 5. Porter RN, Smith W, Craft IL, Abdulwahid NA, Jacobs HS. Induction of ovulation for in vitro fertilization using buserelin and gonadotropins. Lancet 1984;2:1284-5. 6. Forman RG, Frydman R, Egan D, Ross C, Barlow DH. Severe ovarian hyperstimulation syndrome using agonists of gonadotropin-releasing hormone for in vitro fertilization: a European series and a proposal for prevention. Fertil Steril1990;53: 502-9. 7. Tanbo T, Dale PO, Kjekshus E, Haug E, Abyholm T. Stimulation with human menopausal gonadotropin versus folliclestimulating hormone after pituitary suppression in polycystic ovarian syndrome. Fertil Steril 1990;53:798-803. 8. Kemeter P, Feichtinger W. Prednisolone supplementation to clomid and/or gonadotropin stimulation for in-vitro fertilization-a prospective randomized trial. Hum Reprod 1986;1:441-4. 9. Ong ACM, Rennie DP, Homburg R, Lachelin GCL, Jacobs HS, Slater JDH. The pathogenesis of ovarian hyperstimulation syndrome: a possible role for renin. Clin Endocrinol (Oxf) 1991;34:43-9. 10. Tollan A, Holst N, Forsdahl F, Fadnes HO, Oian P, Malthau JM. Transcapillary fluid dynamics during ovarian stimulation for in vitro fertilization. Am J Obstet Gynecol 1990;162:554-8. 11. Friedman RG, Schmidt GE, Chang FE, Kim MH. Severe ovarian hyperstimulation following follicular aspiration. Am J Obstet GynecoI1984;150:436-7. 12. Salat-Baroux J, Alvarez S, Antoine JM, Tibi C, Cornet D, Mandelbaum J, et al. Comparison between long and short protocols of LHRH agonist in the treatment of polycystic ovary disease by in vitro fertilization. Hum Reprod 1988;3: 535-9. 13. Hsueh AJW, Jones PBC. Extrapituitary actions of gonadotropin-releasing hormone. Endocr Rev 1981;2:437-61. 14. Tan SL, Kingsland C, Campbell S, Mills C, Bradfield J, Alexander N, et al. The long protocol of administration of go-
Fertility and Sterility
nadotropin-releasing hormone agonist is superior to the short protocol for ovarian stimulation for in vitro fertilization. Fertil Steril 1992;57:810-4. 15. Tan SL, Balen A, Hussein EI, Mills C, Campbell S, Yovich J, et al. A prospective randomized study of the optimum timing of human chorionic gonadotropin administration after pituitary desensitization in in vitro fertilization. Fertil Steril. 1992;57:1259-64. 16. MacDougall MJ, Tan SL, Jacobs HS. In vitro fertilization and the ovarian hyperstimulation syndrome. Hum Reprod. 1992;7:597-600. 17. Blackstein J, Shalev J, Saadon T, Kukia EE, Rabinovici J, Pariete C, et al. Ovarian hyperstimulation syndrome: prediction by number and size of preovulatory ovarian follicles. Fertil Steril1987;47:597-602.
Vol. 58, No.2, August 1992
18. Amso NN, Shaw RW, Ahuja KK, Morris N. Prevention of ovarian hyperstimulation syndrome. Fertil Steril 1991;55: 220-1. 19. Amso NN, Ahuja KK, Morris N, Shaw RW. The management of predicted ovarian hyperstimulation involving gonadotropin-releasing hormone analog with effective cryopreservation of all pre-embryos. Fertil Steril1990;53:1087-90. 20. Salat-Baroux J, Alvarez S, Antoine JM, Cornet D, Tibi Ch, Plachot M, et al. Treatment of hyperstimulation during invitro fertilization. Hum Reprod 1990;5:36-9. 21. Testart J, Forman R, Belaisch-Allart J, Volante M, Hazout A, Strubb N, et al. Embryo quality and uterine receptivity in in vitro fertilization cycles with and without agonists of gonadotropin releasing hormone. Hum Reprod 1989;4:198201.
Tan et al.
Ovarian hyperstimulation and IVF
383
