Effect of troglitazone on endocrine and ovulatory performance in women with insulin resistance related polycystic ovary syndrome

FERTILITY AND STERILITY VOL. 71, NO. 2, FEBRUARY 1999 Copyright 1999 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A. Effect of troglitazone on endocrine and ovulatory performance in women with insulin resistance related polycystic ovary syndrome Isao Hasegawa, M.D., Haruo Murakawa, M.D., Mina Suzuki, M.D., Yasuaki Yamamoto, M.D., Takumi Kurabayashi, M.D., and Kenichi Tanaka, M.D. Department of Obstetrics and Gynecology, Niigata University School of Medicine, Niigata, Japan Objective: To investigate the effect of troglitazone, a new antidiabetic agent that improves insulin resistance, on endocrine, metabolic, and ovulatory performance in women with insulin resistance related polycystic ovary syndrome (PCOS). Design: Prospective clinical study. Setting: Infertility outpatient clinic, Niigata University Hospital, Niigata, Japan. Patient(s): Thirteen women with PCOS and insulin resistance. Intervention(s): Troglitazone (400 mg/d) was administered for 12 weeks. Main Outcome Measure(s): Insulin and other hormone (gonadotropins, androgens) levels; various parameters relating to glucose and lipid metabolism before, during, and after troglitazone administration; and ovulation rate. Result(s): The mean ( SD) fasting insulin concentration was significantly reduced, from 18.3 8.9 to 10.5 7.1 U/mL. The LH level was reduced from 9.7 3.4 to 4.8 3.9 miu/ml and the testosterone level was reduced from 0.9 0.5 to 0.5 0.3 ng/ml in accordance. Atherosclerotic lipid levels also were normalized. Before troglitazone administration, the ovulation rate during clomiphene citrate therapy was 34.9% per cycle (15/43). This increased significantly to 72.7% (8/11) during troglitazone coadministration. Further, an ovulation rate of 42.3% (11/26) was achieved with troglitazone alone. Conclusion(s): In women with PCOS and insulin resistance, the reduction of hyperinsulinemia that is produced by troglitazone improves the hyperandrogenism that characterizes PCOS, restoring ovulation. (Fertil Steril 1999;71:323 7. 1999 by American Society for Reproductive Medicine.) Key Words: PCOS, insulin resistance, troglitazone, ovulation Received June 26, 1998; revised and accepted September 15, 1998. Reprint requests: Isao Hasegawa, M.D., Department of Obstetrics and Gynecology, Niigata University School of Medicine, 1-757 Asahimachi, Niigata 951-8510, Japan (FAX: 81-025- 227-0789). 0015-0282/99/$20.00 PII S0015-0282(98)00454-3 Polycystic ovary syndrome (PCOS), characterized by chronic anovulation, elevated LH levels, and hyperandrogenism, is the most common endocrine disorder in women of reproductive age, affecting approximately 6% of this population (1). Recent reports (2, 3) strongly suggest that insulin resistance plays a pivotal role in the pathogenesis of PCOS. The elevated plasma insulin levels that occur as a result of insulin resistance increase ovarian androgen production either directly or by increasing insulin-like growth factor 1 (IGF-1) levels as a result of an insulin-mediated decrease in hepatic IGF-1 binding protein production. Insulin also suppresses hepatic sex hormone binding globulin production, resulting in elevated free androgen levels. These high androgen levels contribute to morphologic effects within the ovary that prevent normal follicular development and induce premature follicular atresia. Clomiphene citrate is the agent of first choice for inducing ovulation in women with PCOS. Administration of clomiphene citrate successfully induces ovulation in 75% of such women (4). We previously reported (5) that the 25% of women with PCOS who do not respond to clomiphene citrate by ovulating often show signs of insulin resistance. In this context, we postulated that treating insulin resistance (i.e., 323

TABLE 1 Endocrinologic background of 13 women with insulin resistance related polycystic ovary syndrome. Case no. Age (y) BMI (kg/m 2 ) Fasting insulin level ( U/mL) ACU for insulin ( U/mL/min) LH level (miu/ml) FSH level (miu/ml) PRL level T level A4 level DHEAS level ( g/ml) 1 26 32.5 17.7 8,921 7.3 7.5 7.3 0.8 3.36 1,298 2 37 41.3 27.0 19,515 8.2 6.4 8.7 1.4 2.56 1,439 3 30 27.9 16.4 15,051 9.6 5.3 11.3 1.5 3.85 4,563 4 30 32.8 19.4 10,394 11.0 10.5 3.2 0.3 3.25 3,576 5 32 26.2 12.4 11,627 10.7 7.2 6.0 0.9 4.69 2,694 6 27 20.4 10.8 8,821 10.3 6.5 8.3 0.2 3.23 3,906 7 29 33.4 17.4 8,031 11.3 9.2 7.2 0.5 3.12 1,770 8 33 29.4 21.8 10,038 1.2 5.9 7.4 0.6 3.41 2,810 9 26 24.0 19.0 10,506 4.7 4.8 11.8 0.4 3.11 2,970 10 31 29.1 12.5 9,083 10.4 7.7 7.1 1.5 2.49 2,254 11 29 20.7 10.3 12,227 15.7 9.8 11.6 0.7 3.95 1,857 12 28 27.2 22.3 18,605 8.7 6.8 16.4 1.3 9.98 4,016 13 25 22.9 10.3 10,632 12.1 9.2 4.9 0.5 2.78 4,705 Mean SD 29.8 3.2 28.7 5.9 16.7 5.2 11,804 3,691 9.7 3.4 7.6 1.6 8.3 3.6 0.9 0.5 3.89 2.02 2,907 1,216 Note: ACU area under the curve in 2-hour oral glucose tolerance test; A4 androstenedione; BMI body mass index; PRL prolactin; T testosterone. reducing hyperinsulinemia) may improve the endocrinologic abnormalities associated with PCOS and thus facilitate ovulation. Various methods, such as diet modification or administration of the biguanide agent metformin (6), have been used to treat insulin resistance. Troglitazone is a newly developed antidiabetic agent that improves insulin sensitivity (7). Dunaif et al. (8) first used troglitazone in obese women with PCOS and found a marked improvement in hyperandrogenism. However, they did not report details of changes in ovulatory performance. We undertook the present study to investigate the effect of troglitazone on ovulatory performance and endocrine and metabolic function in women with insulin resistance related PCOS. MATERIALS AND METHODS The subjects of this study were 13 women with PCOS and insulin resistance. Their mean ( SD) age was 29.8 3.2 years (range, 25 33 years) and their mean ( SD) body mass index was 28.7 5.9 kg/m 2 (range, 20.4 41.3 kg/m 2 ). Polycystic ovary syndrome was diagnosed when the following three criteria were fulfilled: [1] presence of chronic ovulatory disorders such as oligomenorrhea, anovulatory cycles, or secondary amenorrhea; [2] presence of hyperandrogenism (elevated testosterone, androstenedione, or DHEAS levels with or without elevated LH levels); and [3] presence of polycystic ovaries on transvaginal ultrasound examination. All the women had a fasting insulin level of 10 U/mL and an accumulated insulin level (area under the curve for insulin during a 2-hour, 75-g oral glucose tolerance test) of 8,000 U/mL/min. Troglitazone (one 200-mg tablet twice daily; Sankyo Co. Ltd., Tokyo, Japan) therapy was administered orally beginning on the second day of progestin-induced withdrawal bleeding and was continued for 12 weeks. During troglitazone administration, the following endocrine and metabolic parameters were analyzed every 4 weeks. Serum hormones, including FSH, LH, prolactin, testosterone, androstenedione, DHEAS, and 17-hydroxyprogesterone, were measured by RIA; glycohemoglobin was measured by liquid chromatography; and serum lipids, including total cholesterol, phospholipids, triglycerides, -lipoprotein, free fatty acids, highdensity lipoprotein cholesterol, and low-density lipoprotein cholesterol were measured by enzymatic analysis. In addition, liver enzymes, such as aspartate aminotransferase, alanine aminotransferase, and -glutamyl transpeptidase, were measured. On the 12th day of each cycle, measurement of follicular size by transvaginal ultrasound examination was begun. When a follicle was 18 mm in diameter and its rupture was confirmed, troglitazone therapy was discontinued until day 2 of the next cycle. In this case, serum progesterone was measured and ovulation was considered to have been established by progesterone levels of 5 ng/ml. If no mature follicle was observed by the 25th day of each cycle, withdrawal bleeding was induced with a progestin. In this case, clomiphene citrate (100 mg/d for 5 days) was added in the subsequent cycle at the woman s request. This study was approved by our institutional review board, and informed consent was obtained from all patients. RESULTS Table 1 shows the endocrine and metabolic characteristics of the 13 study participants. Nine of the women were obese, 324 Hasegawa et al. Effect of troglitazone on PCOS Vol. 71, No. 2, February 1999

TABLE 2 Endocrine and metabolic changes before and during troglitazone administration. Parameter Week 0 Week 4 Week 8 Week 12 LH level (miu/ml) 9.7 3.4 4.7 2.9* 7.0 3.6 4.8 3.9* FSH level (miu/ml) 7.6 1.6 5.3 1.9* 6.6 2.1 5.9 1.8 PRL level 8.3 3.6 11.2 7.2 8.0 3.7 9.8 6.1 T level 0.9 0.5 0.4 0.2* 0.6 0.3 0.5 0.3 A4 level 3.9 2.0 2.6 0.9 3.8 2.6 2.7 1.4 DHEAS level ( g/ml) 2.9 1.2 1.8 0.9 2.0 0.9 2.2 0.8 17-OHP level 0.77 0.30 1.7 1.6 1.3 0.9 1.6 2.0 Fasting insulin level ( U/mL) 18.3 8.9 10.9 8.6 13.7 7.8 10.5 7.1 Hb-A1c level (%) 5.1 0.4 5.0 0.3 5.0 0.4 4.8 0.3 Fructosamine level ( mol/l) 226 13 216 8 224 16 227 15 Total cholesterol level (mg/dl) 207 39 192 26 191 25 189 30 Phospholipid level (mg/dl) 231 38 209 23 213 21 209 15 Triglyceride level (mg/dl) 136 70 124 72 110 61 94 51 -Lipoprotein level (mg/dl) 450 147 385 99 363 114 340 117 Free fatty acid level ( Eq/L) 519 313 360 181 276 250 500 310 HDL cholesterol level (mg/dl) 54 14 54 16 59 17 59 18 LDL cholesterol level (mg/dl) 154 106 114 24 111 24 110 32 Note: A4 androstenedione; Hb-A1c glycohemoglobin; HDL high-density lipoprotein; LDL low-density lipoprotein; 17-OHP 17-hydroxyprogesterone; PRL prolactin; T testosterone. * P.05 compared with week 0 (before administration). P.01 compared with week 0 (before administration). with a body mass index exceeding 26 kg/m 2. All the women had an elevated insulin area under the curve ( 8,000 U/ ml/min) and hyperandrogenism (elevation of at least one androgen level). Ten of the women had elevated LH levels (LH 7 miu/ml and LH/FSH ratio 1). Table 2 shows the changes in the endocrine and metabolic parameters that occurred during troglitazone administration. Fasting plasma insulin levels were significantly reduced during troglitazone administration, and glycohemoglobin and fructosamine levels, which reflect mean blood glucose levels, were reduced in accordance with the decrease in fasting insulin levels, suggesting that the pharmacologic effect of troglitazone was adequate. Serum levels of gonadotropins (especially LH) and androgens (testosterone, androstenedione, and DHEAS) also were reduced significantly by troglitazone. A gradual decrease in total cholesterol, low-density lipoprotein cholesterol, and triglyceride levels and a significant reduction in -lipoprotein and free fatty acid levels was observed. Before the present study, the 13 women had undergone a collective total of 43 cycles of clomiphene citrate therapy in an attempt to induce ovulation (Table 3). Ovulation occurred in 15 of these cycles, for an ovulation success rate of 34.9%. During troglitazone administration, 37 cycles (troglitazone only, 26 cycles; troglitazone with clomiphene citrate, 11 cycles) were monitored in the 13 women (Table 2). In the cycles with combined troglitazone and clomiphene citrate, ovulation was confirmed in 8 of 11 cycles (72.7%). This ovulation rate was significantly higher (P.01) than that achieved in previous clomiphene citrate only cycles. Further, in the troglitazone-only cycles, the ovulation rate was as high as 42.3% (11/26). Patients who ovulated while receiving troglitazone therapy tended to have higher androstenedione and DHEAS levels than those who did not ovulate. No pregnancy occurred in this series. None of the 13 women had any side effects, such as liver dysfunction or hypoglycemia. DISCUSSION The most characteristic feature of PCOS is chronic hyperandrogenic anovulation. Initially, this hyperandrogenism was thought to be caused by elevated LH levels resulting from dysregulation of the hypothalamic-pituitary axis. The theca interna, which possesses LH receptors, undergoes hyperplastic changes in response to elevated LH levels. These result in a hyperandrogenic environment within the follicle and lead to follicular atresia. Hyperinsulinemia has recently attracted attention because it also can generate hyperandrogenic anovulation (2, 3). Hyperinsulinemia is considered to be a compensatory reaction to inefficient insulin activity (i.e., insulin resistance). An elevated insulin level itself, or the coexisting elevation in the IGF-1 level, causes hyperplasia of the theca interna, resulting in ovarian hyperandrogenism and premature follicular atresia. There are two hypotheses concerning the relation between elevated LH levels and hyperinsulinemia. One is that ovarian FERTILITY & STERILITY 325

TABLE 3 Occurrence of ovulation before and during troglitazone administration. Case no. Before troglitazone 1st cycle 2nd cycle 3rd cycle Regimen Ovulation Regimen Ovulation Regimen Ovulation Regimen Ovulation 1 CC 1/2 Tr ( ) Tr ( ) NO NO 2 CC 0/4 Tr ( ) Tr ( ) NO NO 3 CC 4/5 Tr ( ) Tr ( ) Tr ( ) 4 CC 0/3 Tr ( ) Tr ( ) Tr ( ) 5 CC 3/4 Tr ( ) Tr ( ) Tr CC ( ) 6 CC 0/4 Tr ( ) Tr ( ) Tr CC ( ) 7 CC 1/4 Tr ( ) Tr ( ) Tr CC ( ) 8 CC 1/4 Tr ( ) Tr ( ) Tr CC ( ) 9 CC 1/2 Tr ( ) Tr ( ) Tr CC ( ) 10 CC 1/4 Tr ( ) Tr ( ) Tr CC ( ) 11 CC 0/2 Tr ( ) Tr ( ) Tr CC ( ) 12 CC 2/3 Tr ( ) Tr CC ( ) Tr CC ( ) 13 CC 1/2 Tr ( ) Tr CC ( ) Tr CC ( ) Note: CC clomiphene citrate, NO not observed, Tr troglitazone, ( ) not ovulated, ( ) ovulated. Fractions indicate (no. of ovulatory cycles/total no. of treated cycles) before administration of troglitazone. hyperandrogenism and subsequent chronic anovulation are caused either by hypothalamic dysregulation characterized by elevated LH levels (2) or hyperinsulinemia, and the other is that a dual defect involving both these hormones induces PCOS (3). Insulin receptors have been identified in human pituitary tissue (9), and there is a possibility that insulin might augment the release of LH (10). Cara and Rosenfield (11) also reported that insulin and IGF-1 potentiate LHinduced androgen synthesis by theca interna cells. Therefore, attention should be paid to both these hormones in the management of PCOS. In the present study, troglitazone administration significantly suppressed insulin levels, as confirmed by reductions in serum glycohemoglobin and fructosamine levels. In accordance with this reduction, androgen levels (testosterone, androstenedione, or DHEAS) and LH levels decreased. These results agree with the previously discussed relation between insulin, androgens, and LH. The recovery of ovulatory function may have been the consequence of both general and local (intrafollicular) reductions in androgen levels. There are many methods of inducing ovulation in women with PCOS. Although clomiphene citrate is the agent of first choice, approximately 25% of women with PCOS are resistant to this agent, particularly those with insulin resistance (5). Human menopausal gonadotropin or the recently introduced pure FSH are more effective in inducing ovulation in PCOS but are associated with serious problems such as ovarian hyperstimulation syndrome and multiple gestation. Other methods of inducing ovulation in patients with PCOS that have been reported include pulsatile infusions of hmg or GnRH and laparoscopic ovarian cautery with the use of a laser. Both these methods are more expensive and less convenient than clomiphene citrate for ordinary clinical use. The object of these pharmacologic methods of inducing ovulation is strictly to increase FSH levels; no consideration is given to improving hyperinsulinemia and the resulting hyperandrogenism. In this study, when troglitazone was administered, the reduction of androgen and LH levels that followed the normalization of insulin levels allowed successful ovulation (42.3% ovulation rate). The addition of clomiphene citrate to troglitazone therapy might ensure improvement in both the hyperinsulinemia and the hypothalamic-pituitary disorder, facilitating ovulation. When the two agents were used together in this study, the ovulation rate was 72.7%. However, no pregnancy occurred. It is possible that sufficient luteal support may have been required. In a previous study (6), the investigators attempted to reduce hyperinsulinemia in women with PCOS with the use of the biguanide agent metformin. This agent decreases gluconeogenesis resulting from hepatic glucose production. The investigators reported that elevated insulin levels increased ovarian 17 -hydroxylase activity, resulting in hyperandrogenism. The decrease in insulin secretion caused by metformin resulted in an improvement in hyperandrogenism through a reduction in ovarian 17 -hydroxylase activity. Troglitazone exerts its effect on insulin resistance through the activation of tyrosine phosphorylation by tyrosine kinase on the -subunit of the insulin receptor (12). After improvement of insulin resistance with troglitazone, the blood glucose level can be controlled with less insulin, and this results in a decrease in the blood insulin level (13). Dunaif et al. (8) also used troglitazone to reduce insulin levels. They reported 326 Hasegawa et al. Effect of troglitazone on PCOS Vol. 71, No. 2, February 1999

that 400 mg daily, the same dosage as that used in the present study, was optimal for this purpose. Patients with PCOS often have dyslipidemia, characterized by increased total cholesterol, low-density lipoprotein cholesterol, and triglyceride concentrations and a decreased high-density lipoprotein cholesterol concentration. This dyslipidemia is associated with insulin resistance rather than with obesity (14). Meirow et al. (15) also reported that in insulin-resistant women with PCOS, the degree of dyslipidemia correlates with the concentrations of insulin and DHEAS. In the present study, -lipoprotein and free fatty acid levels were decreased significantly by the administration of troglitazone. In addition, although total cholesterol and low-density lipoprotein cholesterol levels did not change significantly, gradual reductions were observed. It is possible that the administration of troglitazone for longer periods ( 12 weeks) may result in significant reductions in the levels of these atherogenic lipids. In conclusion, troglitazone is a promising treatment that may improve both ovulatory function and endocrine and metabolic conditions in insulin-resistant women with PCOS, because insulin resistance often is the cause of such disorders. Further, as shown by Dunaif (16), women with PCOS are at increased risk for the development of non insulindependent diabetes mellitus, particularly the early onset type. In view of these findings, the administration of troglitazone to these women appears to be justified. References 1. Franks S. Polycystic ovary syndrome. N Engl J Med 1995;333:853 61. 2. Meirow D, Yossepowitch O, Rosler A, Brzezinski A, Schenker JG, Laufer N, et al. Insulin resistant and non-resistant polycystic ovary syndrome represent two clinical and endocrinological subgroups. Hum Reprod 1995;10:1951 6. 3. Poretsky L, Piper B. Insulin resistance, hypersecretion of LH and a dual-defect hypothesis for the pathogenesis of polycystic ovary syndrome. Obstet Gynecol 1994;84:613 21. 4. Gysler M, March CM, Mishell DR, Bailey EJ. A decade s experience with an individualized clomiphene treatment regimen including its effect on the postcoital test. Fertil Steril 1982;37:161 7. 5. Murakawa H, Hasegawa I, Kurabayashi T, Tanaka K. Insulin resistance and ovulatory responses to clomiphene citrate in women with polycystic ovary syndrome. J Reprod Med. In press. 6. Nestler JE, Daniela J, Jakubowicz MD. Decreases in ovarian cytochrome P450c17 activity and serum free testosterone after reduction of insulin secretion in polycystic ovary syndrome. N Engl J Med 1996;335:617 23. 7. Nolan JJ, Ludvik B, Beerdsen P, Joyce M, Olefsky J. Improvement in glucose tolerance and insulin resistance in obese subjects treated with troglitazone. N Engl J Med 1994;331:1188 93. 8. Dunaif A, Scott D, Finegood D, Quintana B, Whitcomb R. The insulinsensitizing agent troglitazone improves metabolic and reproductive abnormalities in the polycystic ovary syndrome. J Clin Endocrinol Metab 1996;81:3299 306. 9. Unger JW, Livingston JN, Moss AM. Insulin receptors in the central nervous system: localization, signalling mechanisms and functional aspects. Prog Neurobiol 1991;36:343 62. 10. Adashi EY, Hsueh AJW, Yen SSC. Insulin enhancement of luteinizing hormone and follicle-stimulating hormone release by cultured pituitary cells. Endocrinology 1981;108:1441 9. 11. Cara JF, Rosenfield RL. Insulin-like growth factor I and insulin potentiate luteinizing hormone-induced androgen production by rat ovarian thecal-interstitial cells. Endocrinology 1988;123:733 9. 12. Kellerer M, Kroder G, Tippmer S, Berti L, Kiehn R, Mosthaf L, et al. Troglitazone prevents glucose-induced insulin resistance of insulin receptor in rat-1 fibroblasts. Diabetes 1994;43:447 53. 13. Iwamoto Y, Kuzuya T, Matsuda A, Awata T, Kumakura S, Inooka G, et al. Effect of new oral antidiabetic agent CS-045 on glucose tolerance and insulin secretion in patients with NIDDM. Diabetes Care 1991;14: 1083 6. 14. Robinson S, Henderson AD, Gelding SV, Kiddy D, Niththyananthan R, Bush A, et al. Dyslipidemia is associated with insulin resistance in women with polycystic ovaries. Clin Endocrinol (Oxf) 1996;44:277 84. 15. Meirow D, Raz I, Yossepowitch O, Brzezinski A, Rosler A, Schenker JG, et al. Dyslipidemia in polycystic ovarian syndrome: different groups, different aetiologies? Hum Reprod 1996;11:1848 53. 16. Dunaif A. Hyperandrogenic anovulation (PCOS): a unique disorder of insulin action associated with an increased risk of NIDDM. Am J Med 1995;98:33 9. FERTILITY & STERILITY 327