Cobbold Laboratories, University College and Middlesex School of Medicine, Middlesex Hospital, London, United Kingdom

Transcription

1 FERTILITY AND STERILITY Copyright 1992 The American Fertility Society Printed on acid-free paper in U.S.A. Polycystic ovaries in patients with hypogonadotropic hypogonadism: similarity of ovarian response to gonadotropin stimulation in patients with polycystic ovarian syndrome Zeev Shoham, M.D.*t Gerard S. Conway, M.R.C.P.:!: Anita Patel, D.M.U. Howard S. Jacobs, M.D.:!: II Cobbold Laboratories, University College and Middlesex School of Medicine, Middlesex Hospital, London, United Kingdom Objective: To characterize the ovarian response in patients with isolated hypogonadotropic hypogonadism with ultrasound (US) findings of polycystic ovaries (PCO). Design: Twenty-seven treatment cycles in patients with hypogonadotropic hypogonadism and US findings of normal ovaries were compared with 31 cycles in patients with hypogonadotropic hypogonadism and US-diagnosed PCO. Forty-one cycles in the hypogonadotropic hypogonadism and US-diagnosed PCO were compared with 59 cycles of patients with polycystic ovarian syndrome (PCOS) to examine pattern of response after ovulation induction. Setting: Specialist Reproductive Endocrine Unit. Patients, Participants: Twenty hypogonadotropic patients in whom 10 had US findings of PCO and 13 patients with PCOS. Main Outcome Measure: Serum estradiol (E2) concentration, number of leading follicles on US, cancellation, and pregnancy rate. Results: Hypogonadotropic patients with US-diagnosed PCO had higher baseline ovarian volume (P < 0.02) compared with patients with hypogonadotropic hypogonadism with normal ovaries. After ovarian stimulation, a higher mean serum E2 concentration (P < 0.001), endometrial thickness (P < 0.001), and increased number of leading follicles (P < ) were found in hypogonadotropic patients with US-diagnosed PCO, compared with hypogonadotropic patients with US findings of normal ovaries. Patients with PCOS had a higher serum E2 concentration (P < 0.008), although they were treated for fewer days (P < ) and with fewer ampules of gonadotropin (P < 0.001) compared with patients with hypogonadotropic hypogonadism with US-diagnosed PCO. Conclusions: We have characterized a group of hypogonadotropic patients with US findings of PCO, in which the ovarian response to ovulation induction was similar to patients with PCOS. The results have practical and theoretical implications for the etiology and treatment of patients with PCO. Fertil Steril1992;58:37-45 Key Words:, polycystic ovaries, polycystic ovarian syndrome, ultrasound, induction of ovulation Received July 29, 1991; revised and accepted March 11, * Research Fellow at the University College and Middlesex School of Medicine, Middlesex Hospital. On study leave from the Department of Obstetrics and Gynecology, Kaplan Hospital, Rehovot, Israel. t Present address: Department of Obstetrics and Gynecology, Kaplan Hospital, Rehovot 76100, Israel. :f: Cobbold Laboratories, University College and Middlesex School of Medicine, Middlesex Hospital. The etiology of polycystic ovary syndrome (peas) has been a subject of interest and controversy for almost 60 years. The disorder has been Department of Ultrasonography, Middlesex Hospital. II Reprint requests: Howard S. Jacobs, M.D., Cobbold Laboratories, Middlesex Hospital, Mortimer Street, London WIN 8AA, United Kingdom. Shoham et al. Hypogonadism with US-diagnosed pca 37

2 J IlIL - attributed to primary abnormalities in the ovaries (1), the adrenal glands (2), and to abnormal gonadotropin secretion (3). One of the first abnormal hormonal abnormalities documented in patients with PCOS was hypersecretion of luteinizing hormone (LH) (4), which is thought to playa key role in perpetuating anovulation (5). It has frequently been noted that PCOS is a heterogeneous disorder that occurs in association with several distinct pathological entities, such as congenital adrenal hyperplasia, Cushing's syndrome, hyperprolactinemia, and insulin-resistant states (6). Ovulation induction with gonadotropins in patients with PCOS, resistant to treatment with clomiphene citrate (CC), is particularly difficult because there is a high frequency of multifollicular development that is responsible for the high rate of hyperstimulation and multiple pregnancies (7). The availability of high-resolution ultrasound (US) scanning has offered a noninvasive means of defining the typical polycystic morphology of ovaries in women with reproductive disturbances. The first US observations of enlarged polycystic ovaries (PCO) in patients with Stein-Leventhal syndrome were described by Kratochwil et al. (8) and Zemlyn (9) and more recently were confirmed by Swanson et al. (10), Parisi et al. (11), Adams et al. (12), and Ardaens et al. (13). Patients with hypogonadotropic hypogonadism have a clinical syndrome that results from impairment of neuroendocrine function and presents with amenorrhea and small ovaries with arrested folliculogenesis. In such patients after induction of ovulation with gonadotropins, usually only one or two leading follicles develop. During the last several years, we have noticed that in some patients with hypo gonadotropic hypogonadism a pretreatment pelvic US examination shows a typical polycystic morphology, whereas the endocrine milieu is completely different from patients with PCOS. Therefore, we decided to investigate whether, after induction of ovulation with gonadotropins, this particular group of patients behaved differently from hypogonadotropic patients with US findings of normal ovaries and whether they behave in a similar way to patients with PCOS, with respect to multifollicular development, cancellation, and pregnancy rate (PR). MATERIALS AND METHODS Patients This study involved 20 infertile women with isolated hypo gonadotropic hypogonadism and 13 women with PCOS attending the Infertility Clinic of the Middlesex Hospital, for induction of ovulation between January 1990 and January Twenty patients with hypo gonadotropic hypogonadism, mean (±SD) age of 29.3 ± 4.5 years, were divided into two groups according to the results of the baseline sonographic appearance of their ovaries. There were 10 patients with US findings of normal ovaries and 10 patients with US-diagnosed PCO. The diagnostic criteria for hypogonadotropic hypogonadism included amenorrhea with a history of failure to undergo spontaneous puberty before the age of 18 years, low serum gonadotropin (LH 0.9 ± 0.2 and follicle-stimulating hormone [FSH] 0.6 ± 0.2 lull) and serum estradiol (E2) concentrations (50.3 ± 14.7 pmoljl). At the time of the study, all patients had normal basal thyroid and adrenal function and normal serum prolactin (PRL) (202 ± 137 mull) concentrations. The mean serum testosterone (T) concentration was 0.93 ± 0.7 nmol/l. An expanding pituitary lesion was excluded by x ray of the sella (4 patients) or computed tomography (16 patients). All patients were in good health and without chronic illness. A summary of the clinical and baseline hormonal data appear in Table 1, divided according to the ultrasonographic findings of the patients' ovaries. The mean age of the 13 patients with PCOS was 27.8 ± 5.2 years (Table 1). All fulfilled the classical criteria of PCOS. Seven of them were amenorrheic, and the remainder had oligomenorrhea. All had elevated basal serum LH concentrations (18.4 ± 6.5 lull), whereas serum FSH concentrations were normal (4.7 ± 0.9 lull); the mean LH:FSH ratio was 3.9 ± 0.9. All except one had elevated serum T concentrations (mean 3.6 ± 1.3 nmoljl for the whole group). Nine patients had presented with hirsutism or acne, and 7 had had amenorrhea for at least 6 months. The body mass index (BMI = weightl height2) of 8 was above the normal range, with a mean of 28.4 ± 5.3 for the whole group. All patients had the ultrasonographic appearance of PCO. Treatment with CC up to 100 mgld for 5 consecutive days had previously failed to induce ovulation in all patients. Drugs Used for Ovulation Induction Patients with PCOS were treated with either human menopausal gonadotropin (hmg, Pergonal, 75 IU of FSH and 75 IU of LH per ampule; Serono Laboratories Ltd., Welwyn Garden City, Herts, United Kingdom) or purified FSH (Metrodin, 75 IU 38 Shoham et al. Hypogonadism with US-diagnosed pea Fertility and Sterility

3 Table 1 Clinical and Baseline Hormonal and US Data of the Three Patient Groups with normal ovaries with US-diagnosed PCO PC OS No. of cycles Age (y) BMI (kg/m2) LH (IU/L) FSH (IU/L) LH:FSH ratio T (nmol/l) PRL (mu/l) Hirsutism Amenorrhea/oligomenorrhea E2 (pmol/l) Endometrial thickness (mm) Uterine cross-sectional area (mm2) Ovarian volume (ml) ± 5.1* 22.3 ± ± ± ± ± ± 139 None 10/0 45 ± ± ± ± ± ± 2.2t 0.9 ± 0.2:1: 0.7 ± 0.1:1: 1.5 ± 0.3:1: 0.9 ± 0.7:1: 191 ± 141 None 10/0 56 ± 14:1: 3.8 ± 0.9:1: 15.7 ± ± 2.34:1: ± ± 5.3t 18.4 ± 6.5:1: 4.7 ± 0.9:1: 3.9 ± 0.9:1: 3.6 ± 1.3:1: 260 ± patients 7/6 195 ± 51:1: 6.4 ± 0.9:1: 26.9 ± 1L ± 5.9:1: * Values are means ± SD. t p < :I: P < P < of FSH with <1.0 IU of LH per ampule; Serono Laboratories Ltd.) Patients with hypogonadotropic hypogonadism were stimulated using hmg. Human chorionic gonadotropin (hcg, Gonadotraphon LH; Paines and Byrne Ltd., Greenford, United Kingdom) was used to trigger ovulation. Treatment Protocol Patients were treated with one of two protocols according to their response. After a baseline US examination, treatment was started in all patients with one ampule per day for 7 days. If necessary, the dose was increased by one ampule every 7 days until there was evidence of active follicular development (follicles 2 10 mm in diameter) and an increase in the thickness of the endometrium, based on ultrasonographic examination. Further dosages of the drug were administered according to each individual's response. Human chorionic gonadotropin, 10,000 U, was injected when not more than three leading follicles 2 16 mm (maximum diameter) were seen on US examination. If there were more than three leading follicles, hcg was not injected, and no further treatment was administered during that cycle. However, in the subsequent cycles the patients were treated with the low-dose protocol. The low-dose protocol (7) started with one ampule per day for the first 7 days. The dose was increased by 37.5 IU (half of 1 ampule) every 7 days, according to the criteria given above. Monitoring of the cycle, hcg dosage, criteria for hcg administration or withholding were the same as for the conventional protocol. No additional drugs were given during the luteal phase. Monitoring of Patients Baseline hormone concentrations were measured before treatment. Serum E 2, gonadotropin, and progesterone (P) concentrations were measured during the induced cycles on serial blood samples taken on the same day as the US scans. Ultrasound monitoring included ovarian morphology and volume, uterine longitudinal, and the anteroposterior diameters, and maximum endometrial thickness. Monitoring was performed by transabdominal pelvic ultrasonography using a 3.5-mHz transducer, or where necessary, by transvaginal US giving a 7.5-mHz transducer, both attached to a Diasonics SPA 1000 sector scanner (Diasonics Incorporated, Millpitas, CA). Ovarian volume was calculated according to the formula 4/3 7r (! diameter), where the diameter is considered as the mean of the length, width, and depth of the ovary, in the absence of a dominant follicle (14). Where possible the mean ofthe volume of both ovaries was recorded. The ovarian morphology was defined as polycystic if there were 10 or more cysts, 2 to 8 mm in diameter, arranged around a dense stroma (12). In hypogonadotropic patients, this appearance became more prominent during treatment. When scanning transabdominally using the fullbladder technique, the uterus was assessed in both longitudinal and transverse sections (the bladder Shoham et al. Hypogonadism with US-diagnosed pea 39

4 : 2 _sns a was considered to be optimally full when the long axis of the uterus was demonstrated at 45 at least to the vertical, so that the entire length [funduscervix] was clearly visualized. The uterus was scanned transversely to assess the obliquity of its lie within the pelvis. Once this was assessed, the transducer was turned by 90 0, and by appropriate manipulation of the probe, a true longitudinal section of the uterus was obtained with the entire length of the endometrial cavity being seen. Uterine dimensions were measured in the saggital plane on an electronically "frozen" US image and expressed as the cross-sectional diameter of the maximum length (from fundus to cervix). The anteroposterior diameter was measured at 90 0 to this plane at the level of the fundusjbody junction. The uterine cross-sectional area was defined as the product of the maximum length and anteroposterior diameter of the uterus, as measured in the saggital plane (15). The endometrium was assessed in real time, but once again measurements were made on the electronically "frozen" US image. The thickness was assessed to be the distance from the hyperechogenic interface of the endometrium and the myometrium to the opposite interface and included the stronger midline echo (endometrial interface). This linear measurement of the endometrium, therefore, represents twice the endometrial thickness. The highest value of endometrial thickness in the plane through the central longitudinal axis of the uterine body was recorded. In the midluteal phase, a mature corpus luteum (CL) was detected according to the following criteria: an irregular structure with a small central echo-free area, surrounded by a thick layer of an echogenic wall giving the appearance of a doughnut. Ovulation was documented by the presence of a CL 7 ± 2 days after hcg administration, with serum P concentration of >25 nmoljl and menstrual bleeding 13 ± 2 days after hcg administration. Frequency of monitoring was limited to once a week until there was evidence of active follicular development. From the day of first evidence of follicular development and increase in endometrial thickness, monitoring was done every 2 to 3 days to determine the correct timing for hcg administration. Endocrine Assessment Serum E2 was determined by radioimmunoassay (RIA) using a commercial kit (MD ; Baxter Healthcare Ltd., Compton, Newbury, United Kingdom). This was a direct RIA (i.e., without extraction) with a solid-phase separation. The sensitivity was 20 pmoljl and the intra-assay and interassay coefficients of variation (CVs) were 5% and 18%, respectively. For LH and FSH measurements, we used a double-antibody RIA (Chelsea method; Endocrine Department, Hammersmith Hospital, London, United Kingdom), using Medical Research Council 68/40 and 69/104 standards, respectively. The sensitivity of the LH assay was 0.3 U /L and 0.2 U /L for the FSH. The intra-assay and interassay CVs were 3.0% and 10.3%, respectively, for LH and 3.0% and 10.0%, respectively, for FSH. Serum Twas measured by an extracted assay (St. Thomas's Hospital Method, Department of Chemical Pathology, London, United Kingdom). The sensitivity was 0.5 nmoljl and the intra-assay and interassay CVs were 4.0% and 6.0%, respectively. Serum PRL concentrations were measured by RIA using an antihuman PRL antiserum raised in rabbit and a second antibody polyethylene glycol separation system (North East Thames Region Immunoassay Unit, London, United Kingdom). A purified preparation of human pituitary PRL was labeled with 1251 and also used as a reference standard (IRP 83/562). Intra-assay and interassay CVs were 4.3% and 6.5%, respectively. Progesterone was measured by RIA using a commercial kit (IDS Ltd., Usworth Hall, Tyne and Wear, United Kingdom). The working range of the assay was 0.5 to 100 nmoljl and the intra-assay and interassay CVs were 6.9% and 8.4%, respectively. Samples obtained in individual cycles were measured in a single assay. Statistical Analyses Results are presented as means ± SD for normally distributed data, and the statistical analysis applied to this group was Student's t-test for paired data. Results are expressed as medians with ranges when the data were not normally distributed, and differences were then analyzed by a nonparametric statistical test (Mann-Whitney U-test). RESULTS Table 1 shows the characteristics of the 33 patients participating in the study. At baseline the results from patients with hypogonadotropic hypogonadism and US findings of normal ovaries were similar to those from patients with hypogonadotropic hypogonadism and US-diagnosed PCO except 40 Shoham et al. Hypogonadism with US-diagnosed pea Fertility and Sterility

5 for mean ovarian volume, which was higher in the group with US-diagnosed peo (1.74 ± 0.94 versus 4.06 ± 2.34 ml, respectively, P < 0.02) (Fig. 1). Pretreatment results in patients with peos differed from those in patients with hypogonadotropic hypogonadism and US-diagnosed peo, the former having a higher BMI (28.4 ± 5.3 versus 22.4 ± 2.2, respectively, P < 0.003), higher mean serum LH and FSH concentrations (P < ) with higher LH:FSH ratio (3.9 ± 0.9 versus 1.5 ± 0.3, respectively, P < ), higher mean serum T concentration (3.6 ± 1.3 versus 0.9 ± 0.7 nmoljl, respectively, P < ), higher mean serum E2 concentration (195 ± 51 versus 56 ± 14 pmoljl, respectively, P < ), with thicker endometrium (6.4 ± 0.9 versus 3.8 ± 0.9, respectively, P < ), higher uterine cross-sectional area (26.9 ± 11.1 versus 15.7 ± 6.0 mm 2, respectively, P < 0.02), and a higher mean ovarian volume (14.8 ± 5.9 versus 4.06 ± 2.34 ml, respectively, P < ) (Fig. 1). No significant difference was noted in age and mean serum PRL concentrations. The ovarian response to induction of ovulation was studied by comparing first the results in 31 treatment cycles in patients with hypogonadotropic hypogonadism and US-diagnosed peo with the results in 27 treatment cycles in patients with hypogonadotropic hypogonadism and US findings of normal ovaries (Table 2). No significant difference was found in the number of ampules used and the number of days needed to achieve a degree of stimulation appropriate for the administration of heg. However, significantly higher mean serum LH and FSH concentrations were found in patients with hypogonadotropic hypogonadism and US findings of normal ovaries compared with those found in the hypogonadotropic hypogonadism with US-diagnosed peo (2.3 ± 0.6 versus 1.8 ± 0.6 U jl, respectively, P < for the LH and 11.2 ± 5.1 versus 7.7 ± 2.9 UjL, respectively, P < for the FSH). Despite this finding, however, in the patients with hypogonadotropic hypogonadism with US-diagnosed peo a significantly higher number of follicles ;0:: 14 mm in diameter (4 [range 1 to 13] versus 1 [range 1 to 3], respectively, P < ) developed during ovulation induction, with higher mean serum E2 concentrations (1,612 [range 183 to 3,694] versus 905 [range 118 to 2,060] pmoljl, respectively, P < 0.001) and higher endometrial thickness (9.0 ± 2.0 versus 7.7 ± 1.1 mm, respectively, P < 0.001). The ovhrian response to ovulation induction was then compared in 31 treatment cycles in patients Figure 1 A baseline US image of the ovaries, in three patients, using 7.5-mHz transvaginal transducer. (A), A small normal ovary of 2.6 ml found in patient with hypogonadotropic hypogonadism. (B), An ovary of 6.3 ml found in patient with hypogonadotropic hypogonadism with US-diagnosed peo. In this patient, the main feature of peo can be easily observed; approximately 10 microcysts are peripherally arranged around a dense core of stroma. (C), A classical feature of peo, 11.2 ml in volume found in patient with peos. The ovaries on (B) and (C) are almost identical except the difference in the ovarian volume. with hypogonadotropic hypogonadism and US-diagnosed peo with 32 treatment cycles in patients with peos, both treated with the conventional protocol, and 10 cycles versus 27 cycles, respectively, treated using the low-dose protocol. When the conventional protocol was used (Table 3), patients with peos required significantly fewer ampules (P < 0.001) and fewer days of treatment (P < ) compared with hypogonadotropic patients with US-diagnosed peo. Moreover, ovarian stimulation resulted in higher mean serum E2 concentration on day of heg administration (P < 0.008). However, no significant difference was found in the number offollicles ;0:: 14 mm between the peos and the hypogonadotropic patients with US-diagnosed peo (Fig. 2) (4 [range 1 to 18] versus 4 [range 1 to Shoham et al. Hypogonadism with US-diagnosed peg 41

6 abel tj 2 jj L isaac Jli Ii Table 2 Comparing Results Between the Hypogonadotropic Hypogonadism Patients With Normal Ovaries With the Hypogonadotropic Hypogonadism Patients With US-Diagnosed PCO* Parameter with normal ovaries with US-diagnosed PCO Probability value No. of cycles No. of ampules No. of days LH (IU/L) FSH (IU/L) E2 (pmol/l) Endometrial thickness (mm) Follicle"" 14 mm (10 to 48) 17.7 ± ± ± (118 to 2,060) 7.7 ± (1 to 3) (8 to 70) 16.9 ± ± ± 2.9 1,612 (183 to 3,694) 9.3 ± (1 to 13) * Values recorded on day of hcg administration are medians with ranges in parentheses or means ± SD. t NS, not significant. 13], respectively) or in the US-determined endometrial thickness. Cancellation and PRs in both groups were similar (23% versus 29% and 6.2% versus 12.9%, respectively). Furthermore, when patients in both groups who produced more than three leading follicles at the time of hcg administration, were subsequently treated with the low-dose protocol (Table 3), no significant differences were noted comparing the number of ampules used, number of days needed for stimulation, number of leading follicles on day of hcg administration, cancellation rate, and PRo The only significant difference between the two groups, those with hypogonadotropic hypogonadism and US-diagnosed PCO compared with PCOS, was found comparing serum E2 concentration (888 pmol/l [range 388 to 1,698] versus 2711 pmol/l [range 550 to 9,093], respectively, P < 0.001) and serum LH concentration (1.8 ± 0.3 lull versus 7.3 ± 4.1 lull, respectively, P < ). Surprisingly, there was not a single case of miscarriage in either group using the two different protocols. DISCUSSION The concept of this study originated from the observation that by using high-frequency US we obtain sufficiently high resolution of ovarian morphology to identify a group of patients with hypogonadotropic hypogonadism but with an ovarian appearance similar to that found in PCO (i.e., a necklace of 10 or more small cysts surrounding a dense stroma). After induction of ovulation, we have found that ovaries of these patients behaved similarly to those of patients with PCOS with regard to the development of multiple follicles, even though their baseline endocrine milieu is different; moreover, these ovaries respond differently from those in patients with hypogonadotropic hypogonadism and the US findings of small normal ovaries. We used the patients with hypogonadotropic hypogonadism and the US findings of normal ovaries as the reference model to characterize the disturbed response of the patients with hypogonadotropic hypogonadism with US-diagnosed PCO. Before treatment, the only significant difference, comparing the two groups, was the finding of a higher mean ovarian volume in patients with USdiagnosed PCO. Comparing the response of these women's ovaries after induction of ovulation, the striking difference was the higher number of follicles developed in the US-diagnosed PCO group, leading to higher serum E2 concentration and an increase in endometrial thickness, despite serum LH and FSH concentrations being significantly lower during treatment. These data led us to characterize a new group of hypogonadotropic patients according to the pretreatment ultrasonographic appearance, which is similar to that found in PCOS patients. In these patients, the ovarian response after induction of ovulation is different from that usually seen in hypogonadotropic patients. The difference occurs even though these patients have low serum LH, FSH, and androgen concentrations throughout life. To explore further the concept that patients with hypogonadotropic hypogonadism and US-diagnosed PCO respond to ovulation induction in a similar way to that of patients with PCOS, we compared the ovarian performance of our newly characterized 42 Shoham et al. Hypogonadism with US-diagnosed pea Fertility and Sterility

7 group with that of patients with pcas. After previous observations ofthe superiority ofthe low-dose gonadotropin protocol over the conventional protocol for ovulation induction in patients who developed more than three leading follicles at the time ofhcg administration (7, 16), the same criteria for using either the conventional or the low-dose protocol were applied to both groups of patients, and the ovarian response to stimulation was almost the same. In general, patients with pcas have more sensitive ovaries, together with thecal hyperplasia, which provides large amounts of androstenedione (A) and T that act as substrates to the aromatase enzyme complex in the granulosa cells for E2 production (17). In the hypogonadotropic hypogonadism group, because there was only a minimum amount of LH during the stimulation period, the theca cells would be expected to produce smaller amount of A and T than in the pcas group (18). It should also be noted that patients with pcas have an increased ovarian volume, part of which is because of greater ovarian stroma, which is able to produce androgens. These facts may contribute to the difference in serum E2 concentrations be- S 20 :z: 18 '0 18 ~ peo : 0 N.S. E 12 E 0 ;t I 8 ~ B ~ g esebe I : HH with normal HH with polycyatlc Polycyatlc Ovary ovari on US ovarl on US Syndrome Figure 2 The dotogram presents the number of follicles ~ 14 mm in diameter, developed in the three different groups of patients after induction of ovulation with either the conventional protocol (e) or the low-dose protocol (0). No significant difference was noted comparing multifollicular development in patients with hypogonadotropic hypogonadism PCO on US with PCOS patients. tween the hypo gonadotropic patients with US-diagnosed pca and patients with pcas found in the present study. The observation that there was Table 3 Results of Hypogonadotropic Hypogonadism Patients With US-Diagnosed PCO Compared With the PC OS Group in the Two Different Treatment Protocols * Parameter with US-diagnosed PCO J;>COS Probability value Conventional protocol No. of cycles No. of ampules No. of days LH (U/L) FSH (U/L) E z (pmol/l) Endometrial thickness (mm) Follicle ~ 14 mm Cancellation rate Pregnancy rate (%) (8 to 70) 16.9 ± ± ± 2.9 1,612 (183 to 3,694) 9.3 ± (1 to 13) (5 to 39) 11.6 ± ± ± 1.7 3,465 (300 to 16,340) 9.1 ± (1 to 18) Low-dose protocol No. of cycles No. of ampules No. of days LH (U/L) FSH (U/L) Ez (pmol/l) Endometrial thickness (mm) Follicle ~ 14 mm Cancellation rate (%) Pregnancy rate (%) (14 to 53) 19 ± ± ± (388 to 1,698) 8.3 ± (1 to 10) (9 to 49) 18 ± ± ± 7.5 2,711 (550 to 9,093) 9.6 ± (1 to 11) * Values recorded on day of hcg administration are medians with ranges in parentheses or means ± SD. t NS, not significant. Shoham et al. Hypogonadism with US-diagnosed pea 43

8 no significant difference between the number of follicles that developed in patients with hypogonadotropic hypogonadism and US-diagnosed peo and patients with peos who were treated with either the conventional protocol or the low-dose protocol shows that the excessive development of follicles usually seen in ordinary gonadotropin stimulation for peos is the same in patients with hypogonadotropic hypogonadism and US-diagnosed peo. The high cancellation and the lower PRs noted in patients with peos are well documented in the literature (19). This observation was confirmed in our study in hypogonadotropic patients with USdiagnosed peo as well as in patients with peos, both treated with the conventional protocol. Using the low-dose protocol, a significant difference was noted in serum LH and E2 concentrations comparing patients with hypogonadotropic hypogonadism and US-diagnosed peo and the peos patients. However, in both groups, a smaller number of leading follicles developed. These results are probably achieved by reproducing almost the precise gonadotropin dosages required (7, 20). Although the etiology of peo and peos is still obscure, there is increasing evidence that insulin and polypeptide growth factors, which modulate the action of gonadotropin within the ovary, have an important role in the physiology of the ovary (21). The results of this study, together with the possibility of identifying a group of hypogonadotropic patients with peo, reinforces the hypothesis previously suggested by our group (1) that the primary disorder in patients with peo is within the ovary itself and that the clinical and endocrine abnormalities associated with the syndrome is secondary to the ovarian disturbance. This hypothesis is also consistent with the observation of Hague et al. (22) that peo are inherited; the finding of unilateral peo (23) in some patients is also consistent with this idea. The presence of other endocrine disturbances associated with polycystic ovarian morphology or the appearance of hormonal and metabolic disturbances does not rule out the possibility that such derangements are secondary to a primary ovarian abnormality in those patients who have functional hypothalamic-pituitaryaxis. In conclusion, this study comparing the ovarian response after induction of ovulation in hypogonadotropic hypogonadism patients with US findings of small normal ovaries and those with US-diagnosed peo, gave us an opportunity to characterize a new group of hypogonadotropic patients with ultrasonographic morphology of peo who respond in a similar way to patients with classical peos, regarding multifollicular development and the cancellation and PRs. REFERENCES 1. Jacobs HS. Polycystic ovaries and polycystic ovary syndrome. Gynecol Endocrinol 1978;1: Oake RJ, Davies SJ, MacLachlan MS, Thomas JP. Plasma testosterone in adrenal and ovarian vein blood of hirsute women. Q J Med 1974;43: Yen SS, Vela P, Rankin J. Inappropriate secretion of follicle-stimulating hormone and luteinizing hormone in polycystic ovarian disease. J Clin Endocrinol Metab 1970;30: McArthur JW, Ingersall FM, Worcester J. The urinary excretion of interstitial cell and follicle stimulating hormone activity by women with diseases of the reproductive system. J Clin EndocrinoI1985;18: Yen SS. The polycystic ovary syndrome. Clin Endocrinol (Oxf) 1980;12: McKenna T J. Pathogenesis and treatment of polycystic ovary syndrome. N Engl J Med 1988;318: Shoham Z, Patel A, Jacobs HS. Polycystic ovarian syndrome: safety and effectiveness of stepwise and low-dose administration of purified follicle-stimulating hormone. Fertil Steril 1991;55: Kratochwil A, Urban G, Friedrich F. Ultrasound tomography of the ovaries. Ann Chir GynaecoI1972;61: Zemlyn S. Comparison of pelvic ultrasonography and pneumography for ovarian size. J Clin Ultrasound 1974;2: Swanson M, Sauerbrei EE, Cooperberg PL. Medical implications of ultrasonically detected polycystic ovaries. J Clin Ultrasound 1981;9: Parisi L, Tramonti M, Casciano S, Zurli A, Gazzarini O. The role of ultrasound in the study of polycystic ovarian disease. J Clin Ultrasound 1982;10: Adams J, Franks S, Polson DW, Mason HD, Abdulwahid N, Tucker M, et al. Multifollicular ovaries: clinical and endocrine features and response to pulsatile gonadotropin releasing hormone. Lancet 1985;2: Ardaens Y, Robert Y, Lemaitre L, Fossati P, Dewailly D. Polycystic ovarian disease: contribution of vaginal endosonography and reassessment of ultrasonic diagnosis. Fertil Steril 1991;55: Campbell S, Goswamy R, Goessens L, Whitehead M. Realtime ultrasonography for determination of ovarian morphology and volume. Lancet 1982;1: Polson DW, Franks S, Reed MJ, Cheng RW, Adams J, James VH. The distribution of oestradiol in plasma in relation to uterine cross-sectional area in women with polycystic or multifollicular ovaries. Clin Endocrinol (Oxf) 1987;26: Polson DW, Mason HD, Kiddy DS, Winston RM, Margara R, Franks S. Low-dose follicle-stimulating hormone in the treatment of polycystic ovary syndrome: a comparison of pulsatile subcutaneous with daily intramuscular therapy. Br J Obstet Gynaecol 1989;96: Shoham et al. Hypogonadism with US-diagnosed pca Fertility and Sterility

9 17. Franks S. Polycystic ovary syndrome: a changing perspective. Clin Endocrinol (Oxf) 1989;31: Couzinet B, Lestrat N, Brailly S, Forest M, Schaison G. Stimulation of ovarian follicular maturation with pure follicle-stimulating hormone in women with gonadotropin deficiency. J Clin Endocrinol Metab 1988;66: Lunenfeld B, Bloom S, Blankstein J. Resultants de l'induction de l'ovulation par les menotropins classiques (hmg-hcg). In: Buvat J, Bringer J, editors. Induction et stimulation de l'ovulation. Paris: Doin Publishers, 1986: Brown JB. Pituitary control of ovarian function-concepts derived from gonadotropin therapy. Aust N Z J Obstet GynaecoI1978;18: Conway GS, Jacobs HS, Holly JM, Wass JAH. Effects of luteinizing hormone, insulin, insuline-like growth factor-i and insulin-like growth factor small binding protein 1 in the polycystic ovary syndrome. Clin Endocrinol (Oxf) 1990;33: Hague WM, Adams J, Reeders ST, Peto TE, Jacobs HS. Familial polycystic ovaries: a genetic disease? Clin Endocrinol (Oxf) 1988;29: Polson DW, Adams J, Steer PJ, Franks S. Unilateral polycystic ovary. Case report. Br J Obstet GynaecoI1986;93: Shoham et al. Hypogonadism with US-diagnosed pca 45