- Email: [email protected]
Insulin resistance improves in hyperandrogenic women treated with Lupron *†‡
FERTILITY AND STERILITY Copyright
©
Vol. 60, No.4, October 1993
Printed on acid-free paper in U. s. A.
1993 The American Fertility Society
Insulin resistance improves in hyperandrogenic women treated with Lupron*t*
Karen E. Elkind-Hirsch, Ph.D.§11 Cecilia T. Valdes, M.D"I** L. Russell Malinak, M.D.** Baylor College of Medicine, Houston, Texas
Objective: To examine if changes in insulin sensitivity and glucose effectiveness in women with polycystic ovarian disease (PCOD) occurred after ovarian androgen suppression with a GnRH agonist, leuprolide acetate (LA, Lupron; TAP Pharmaceuticals, Deerfield, IL) using the minimal model method. Design: Twelve patients with PCOD were tested in the untreated state (baseline) and after 6 weeks of LA treatment. Subjects were divided into two groups based on the degree of impairment of their baseline insulin sensitivity index (SI; (min- 1 )(ILU/mL- 1 ): mild insulin resistance (SI > 1) or severe insulin resistance (SI < 1). Results: In all patients, serum T was significantly decreased from elevated baseline levels to normal female concentrations after 6 weeks of LA therapy. Insulin sensitivity in PCOD patients with mild insulin resistance significantly improved from baseline after 6 weeks of LA therapy, whereas no change in SI on LA therapy was seen in PCOD women with severe insulin resistance. Glucose utilization independent of increased insulin secretion did not change as a function of LA treatment in either group. Conclusion: These findings indicate a significant improvement in SI in mildly insulin-resistant women with PCOD after suppression of ovarian function with LA treatment. Fertil Steril1993;60:634-41 Key Words: Insulin sensitivity, GnRH agonist, polycystic ovarian disease
Since the description of "the diabetes of bearded women" in 1921 by Achard and Theirs (1) the remarkable correlation between androgen levels and
Received February 1, 1993; revised and accepted July 2, 1993.
* Lupron, TAP Pharmaceuticals, Deerfield, Illinois. t Supported by the Division of Research Resources of the National Institutes of Health under grant M01RR00350, Bethesda, Maryland and in part by an Educational Grant from TAP Pharmaceuticals, Deerfield, Illinois. :I: Presented in part at the 48th Annual Meeting of the American Fertility Society, New Orleans, Louisiana, October 31 to November 5, 1992. § Reprint requests: Karen Elkind-Hirsch, Ph.D., Obstetrical and Gynecological Associates, 7550 Fannin Street, Houston, Texas 77054-1989. II Department of Medicine. 1f Recipient of an American College of Obstetrics and GynecologyjOrtho Fellowship Grant, Baylor College of Medicine, Houston, Texas from 1989 to 1990. ** Department of Obstetrics of Gynecology.
634
Elkind-Hirsch et al.
insulin resistance in hyperandrogenic states has been known. Patients with polycystic ovarian disease (peOD) and the hyperandrogen-insulin resistance-acanthosis nigricans syndrome imply an association of elevated androgens and changes in insulin sensitivity, but the mechanism by which the hyperinsulinemia seen in insulin resistance leads to ovarian hyperandrogenism in peOD remains unclear (2). A number of studies suggest elevated androgens as the primary defect in peOD. Treatment with the antiandrogen spironolactone decreases insulin resistance in peOD patients concordant with a decline in plasma androgen levels (3). Older studies using exogenous androgen treatment for patients with aplastic anemia demonstrated glucose intolerance and elevated insulin levels (4). Administration of T or T derivatives to women results in impaired glucose tolerance and hyperinsulinemia (5). Furthermore, exposing streptozotocin-treated
Insulin resistance improves with Lupron
Fertility and Sterility
female rats to androgens increases the incidence of insulin-dependent diabetes (6). Although one group of investigators (Cole C, Kitabchi AE, abstract) found that treatment ofPCOD patients with oral contraceptives (OCs, Ortho-Novum; Ortho Pharmaceutical Corporation, Raritan NJ) lowered androgen levels and improved insulin sensitivity, other studies reported no effect of OCs on either of these parameters (2). Furthermore, suppressing androgen levels with a long-acting GnRH analogue (7) or with oophorectomy (8) had no effect on insulin resistance in hyperandrogenic women. Evidence points to hyperinsulinemia as a factor in the etiology of some cases of PCOD. In vitro studies, have shown that insulin and insulinlike peptides can augment basal and LH -stimulated ovarian androstenedione and testosterone production (9). In vivo studies, however, performed with acute, high-dose insulin infusions in obese insulinresistant women with PCOD resulted in decreased, rather than increased, androgen concentrations (10). By administration of a highly potent GnRH agonist (GnRH-a), such as leuprolide acetate (LA, Lupron; TAP Pharmaceuticals, Deerfield, IL), it is possible to effectively suppress pituitary and ovarian function. Recent evidence presented by Dunaif et al. (11), using direct measurement of peripheral and hepatic insulin action, showed that normalization of plasma androgen levels for 12 weeks with a GnRH -a did not significantly improve insulin sensitivity in insulin-resistant women with PCOD. The failure to improve insulin sensitivity in this study could possibly result from sampling only a small number of PCOD patients in which obesity was a significant factor responsible for their insulin resistance. However, although the presence of obesity certainly contributes substantially to the insulin resistance observed in obese patients with PCOD, the insulin resistance observed in normal-weight women with PCOD (12) suggests contributory factors that are specifically associated with this condition. Dunaif and colleagues (Dunaif A, Green G, Finegood D, abstract) demonstrated that hyperinsulinemia and insulin resistance in PCOD is secondary to increased basal secretion and decreased clearance but not pancreatic beta cell dysfunction. However, PCOD and obesity have a synergistic deleterious effect on glucose homeostasis. Smith et al. (13) evaluated a mixed group of hyperandrogenic women and reported two sets of responses to an oral glucose tolerance test; a marked insulin response and a slight or normal insulin response. This was
Vol. 60, No.4, October 1993
the first attempt to examine individual responses rather than group mean data and the authors suggested there may be two populations of hyperandrogenic women (13). Because of the heterogeneity of PCOD, it also is possible that androgens contribute to insulin -resistance in some women. Our study was undertaken to determine whether suppression of ovarian androgen production by using LA alters insulin sensitivity and glucose effectiveness determined using the minimal model method when PCOD patients have been classified as to the severity of their pretreatment insulin resistance (separation of patients by diagnostic category).
MATERIALS AND METHODS Study Patients
Fourteen women meeting clinical and biochemical criteria for PCOD were recruited. The subjects' mean age was 26.5 and ranged from 20 to 34. All patients had oligomenorrhea or amenorrhea, hirsutism, elevated plasma levels of T, and a LHjFSH ratio ~ 2.5. Adrenal disorders and hyperprolactinemia were specifically excluded. Polycystic ovarian disease women with abnormal elevations in DHEAS levels were excluded because GnRH-a treatment does not suppress adrenal androgen production. Five of the patients were obese (body mass index [BMI] > 29 kg/m2 ). Two patients had evidence of acanthosis nigricans. None of the women had taken any medications, including hormonal contraception, for the previous 6 months. Two obese patients withdrew from the study when they elected to discontinue LA therapy after the first month. Written informed consent was obtained from each subject, and the protocol was approved by the Baylor College of Medicine and Methodist Hospital Institutional Review Boards. Description of the Study
After a complete history and physical examination, the PCOD patients were studied in the untreated state (baseline) and after 6 weeks of ovarian androgen suppression with LA; 20 ~g/kg per day subcutaneously or 3.75 mg depot monthly. Each subject underwent a tolbutamide-modified frequently sampled intravenous glucose tolerance test (lVGTT) both pretreatment and 6 weeks after LA therapy.
Elkind-Hirsch et al. Insulin resistance improves with Lupron
635
All investigations were performed in the Methodist Hospital General Clinical Research Center in Houston, Texas. Blood samples were obtained for determination of baseline serum concentrations of T, DHEAS, androstenedione (A), E 2, P, 17-hydroxyprogesterone (17-0HP), LH, FSH, PRL, morning and evening cortisol, insulin, and thyroid functions (T 3, T 4 , TSH, and T3RU). In addition, a complete blood count was determined (hematocrit of 30 was required for participation). Study Protocol
Tolbutamide-Modified IVOTT
Each subject was instructed to consume a diet containing ~300 g carbohydrate in the 3 days before each IVGTT and all studies were begun between 6 and 10 A.M. after an overnight fast. A heparin lock was placed in the forearm for administration of medications and for blood sampling. Baseline samples for insulin and glucose levels were drawn 15, 20, 25, and 30 minutes after IV placement. Serum T, E 2, P, LH, FSH, PRL, and DHEAS levels were drawn 15 minutes after IV placement before each IVGTT. Glucose as a 25% solution (0.3 g/kg) was injected over 1 minute. Beginning 20 minutes after the glucose bolus, tolbutamide as a bolus of 150 mg/m2 (300 to 500 mg) was injected over 30 seconds (Orinase Diagnostic; Upjohn, Kalamazoo, MI). Blood samples (4 mL) for measurement of glucose and insulin were drawn at 2, 3, 4, 5, 6, 8, 10, 12, 14,16,19,22,23,24,25,27,30,35,40,50,60,70,80, 90, 100, 120, 140, 160, and 180 minutes after starting the glucose injection (14). Analytical Techniques and Calculations
Serum E 2, LH, FSH, PRL, DHEAS, T, cortisol, and P concentrations were measured in duplicate with commercial assay kits (Diagnostic Products, Los Angeles, CA) in the General Clinical Research Center RIA core laboratory. The interassay coefficients of variation of all hormone assays were <12%. Androstenedione and 17-0HP measurements were performed by Nichols Reference Laboratories (San Juan Capistrano, CA). Immunoreactive insulin was measured in duplicate by RIA using a modification of the procedures described by Starr et al. (15) and performed in the Diabetes and Endocrine Research RIA core laboratory. The interassay and intra-assay coefficients of variation were <10% and <5%, respectively. Serum glucose was mea-
636
Elkind-Hirsch et al.
sured by the glucose oxidase method in the Methodist Hospital Clinical Chemistry Laboratory. The relative proportion of body fat was expressed as the BMI (kg/m 2), which is considered to be the best estimate of body fat obtained from a single measure of height and weight; it has an excellent correlation with body fat estimated from body density (16). The nonobese group was defined as subjects within 20% of their ideal body weight with a BMI < 27 kg/m 2. The mean BMI for nonobese PCOD patients was 23 ± 1.5 (±SEM) kg/m2. The obese group was defined as subjects with weight ranging between 150% and 225% of ideal body weight and BMI > 29 kg/m2 (16). The mean BMI for obese PCOD patients was 47.8 ± 2.3 kg/m2. Results of the modified IVGTTs were then fitted to the minimal model of glucose kinetics using the MINMOD program (provided by Richard Bergman), giving an insulin sensitivity index (SI X 10-4 (min- 1)(J.tU/mL -1) and glucose effectiveness index (SG/min). In analyzing the results, one limit in the measurement of insulin sensitivity using the minimal model became apparent. Extreme insulin resistance could not be easily measured because a prerequisite for the minimal model to calculate insulin sensitivity is that the endogenous secreted insulin must measurably influence the plasma glucose level (14). Thus, from a clinical standpoint, evaluating PCOD subjects with glucose tolerance reduced to a diabetic value «1) confounds the minimal-model calculated insulin sensitivity index (SI) (14). To overcome the problem of diabetic tolerance values in very insulin-resistant states, subjects were divided into two groups: mild insulin resistance (SI > 1) or severe insulin resistance (SI < 1) group based on the degree of impairment of baseline SI value All three obese PCOD patients demonstrated severe insulin resistance (severe insulin resistance group). Two PCOD patients had normal baseline insulin sensitivity (SI > 6.5) and were excluded from the study analyses. Statistical Analyses
All results are reported as mean ± SEM values. Comparisons of baseline hormone measurements between groups were performed using either unpaired t-tests or analyses of variance (ANOVA). Statistical analyses of serum hormones and IVGTT-derived measurement values were performed using subjects-by-treatments repeatedmeasures ANOV A. The treatment variable was
Insulin resistance improves with Lupron
Fertility and Sterility
either hormone concentration (T, E 2, LH, FSH, DHEAS, PRL, and insulin) or IVGTT-derivedmeasurements (Sr, SG) for the two tests. Because the observations were made repetitively on the same subject, these observations are correlated. This violates the validity of the P values for the univariate repeated measures hypothesis tests (F-tests). Therefore all the F-tests were corrected using the Greenhouse-Geisser estimate of epsilon, which is the most conservative adjustment of the F-test. The amount of adjustment is determined by the factor epsilon. The reported P values reflect this correction. For further statistical evaluation of linear relationships between basal insulin levels and Sr before and after LA treatment, comparisons of data sets were performed using the Pearson product moment correlation test. The sample size required to detect a significant change in insulin sensitivity was calculated to be four subjects, setting the P value at 0.05, the power at 95%, and the expected effect size at 2.0 X 10- 4 (min-1)(J,LU/mL- 1).
Baseline IVGTT In PCOD-MIR Patient 5000
::J
.------------~20
Si=2.08
4000
15
0
E
3000
c
2000
~
:;
E
10
CD
0
U
glucose
III
C
5
1000
-60 -30
30
60
90 120150180
(minutes)
IVGTT In Lupron-Treated PCOD-MIR Patient 5000,----:-----------r20
-
::J
SI=7.95
4000
15
E
E
3000
E 10 CD
2000
III
o
III
Vol. 60, No.4, October 1993
glucose
C
Insulin and glucose profiles obtained during a tolbutamide-modified IVGTT in two different
~ o
o
;:,
Mod-IVGTT Measurements
CI
0 0
Time
c
In all PCOD patients, baseline levels of T and E2 levels were both significantly suppressed by LA whereas DHEAS levels remained unchanged. Estradiol levels fell from a mean before treatment concentration of 58 ± 8.7 pg/mL (213 ± 32 pmol/L) to after treatment levels of 23 ± 1.9 pg/mL (85 ± 7 pmol/L) (P < 0.0025). Testosterone levels decreased from elevated baseline levels of 91 ± 12 ng/dL (3.15 ± 0.4 nmol/L) to normal female concentrations of 34 ± 9 ng/dL (1.17 ± 0.3 nmol/L) after 6 weeks of LA therapy (P < 0.0001). The adrenal androgen, DHEAS, concentrations remained unchanged and were 169.7 ± 40.7 J,Lg/dL (4.6 ± 1.1 pmol/L) before treatment and 163 ± 48.1 J,Lg/dL (4.4 ± 1.3 pmol/L) after LA therapy. Although no significant differences in T, E 2, or DHEAS were found between PCOD-mild insulin resistance and PCOD-severe insulin resistance patients, PCOD patients with severe insulin resistance tended to be more androgenized; i.e., manifested higher mean T and DHEAS levels as compared with mildly insulin-resistant patients.
.2
insulin
o
D.
Hormone Measurements
.§. III
~
RESULTS
~
0
1000
u 5
.2
CI
insulin
o
0
-60-300
306090120150180
Time
(minutes)
Figure 1 Insulin and glucose levels used to estimate 81 and 8 a with the MINMOD program in a moderately insulin-resistant PCOD (PCOD-mild insulin resistance) subject before (top) and after (bottom) 6 weeks of treatment with LA.
patients is shown in Figures 1 and 2. One can readily observe the distinct increases in insulin concentration after glucose bolus at the 0 time point and the tolbutamide administration at the 20-minute time point. In all patients, the stimulated insulin secretion returned to baseline by 180 minutes. Figures 1 and 2 illustrate the insulin and glucose levels used to estimate Sr and SG with the MINMOD program in individual PCOD subjects with mild insulin resistance (Fig. 1) and severe insulin resistance (Fig. 2) before (top) and after (bottom) 6 weeks of LA treatment. Figure 3 shows that mean Sr X 10-4 (min-1)(J,LU/ mL- 1) values in the PCOD-mild insulin resistance (n = 5) group were significantly increased after 6 weeks of LA treatment with no change in the
Elkind-Hirsch et al. Insulin resistance improves with Lupron
637
Baseline IVGTT in PCOD- SIR Patient
Insulin Sensitivity in Lupron-Treated peOD Women
10000 ~--------------------------r20
-
::J
Si=0.3
8000
15
E c::
>-
o E
0
.e
= ~
~
6000
Ui
c::
E
CD
en
10 4000
c::
5 2000
3
CD III
c:: 2
u .:
:::I
o
:::I III
5
4
~
CI
1
insulin
o
Baseline
0
·60·300
306090120150180
Time
Lupron
Treatment Figure 3 Mean 8 1 recorded in 10 patients with PCOD. A significant increase was observed in the PCOD-mild insulin resistance group during LA treatment compared with baseline (P < 0.006).
(minutes)
IVGTT In Lupron-Treated PCOD-SIR Patient 20
10000
Si=0.33
8000
::J ::;
15
E
E E
6000
Q.
~
c::
~ 0
0
10 CD III 0
4000
:::I III
U
c::
5 2000
:::I CI
insulin
0 -60 -30
0 0
30
Time
60
90 120150180
(minutes)
Figure 2 Insulin and glucose levels used to estimate 8 1 and 8 G with the MIN MOD program in a severely insulin-resistant PCOD (PCOD-severe insulin resistance) subject before (top) and after (bottom) 6 weeks of treatment with LA.
PCOD-severe insulin resistance (n = 5) group. The mean insulin sensitivity index was 2.6 ± 0.34 in the untreated PCOD-mild insulin resistance group, which increased to 5.3 ± 0.75 after LA treatment (P < 0.006). Polycystic ovarian disease-severe insulin resistance group mean SI was unchanged despite equivalent suppression of ovarian androgens with LA treatment (basal mean SI = 0.5 ± 0.22 versus after LA treatment mean SI = 0.69 ± 0.13; P> 0.05). A significant difference was observed between mean SI value obtained in PCOD-mild insulin resistance patients versus PCOD-severe insulin resistance patients (P < 0.001). Glucose effectiveness (SG/min) did not change as a function of the LA therapy (F = 0.45, P > 0.05).
638
Elkind-Hirsch et al.
Baseline SG estimates were 0.24 ± 0.02/min in the PCOD-mild insulin resistance group and 0.24 ± 0.05/min in the PCOD-severe insulin resistance group. After 6 weeks of LA treatment, SG estimates were 0.26 ± 0.04/min in the PCOD-mild insulin resistance group and 0.28 ± 0.06/min in the PCODsevere insulin resistance group (Fig. 4). Mean basal insulin levels were significantly different between PCOD-mild insulin resistance and PCOD-severe insulin resistance patients both before LA (16.4 ± 1.7 mUlL [117.7 ±12.2 pmol/L] versus 45 ± 11.2 mUlL [322.9 ± 80.4 pmol/L]; P < 0.035) and after LA (12 ± 1.1 mUlL [86.1 ± 7.9 pmol/L] versus 53 ± 11.6 mUlL [380.3 ± 83.2 pmol/L], respectively; P < 0.007) while there was no significant change in baseline insulin concentrations after LA treatment within each group.
Glucose Effectiveness in Lupron-Treated peOD Women
c; ~
c:: CD
~
Eiil
0.4
•
Severe IR MildlR
0.3
U
.! iii
0.2
CD
0.1
U
::>
C;
0.0
Baseline
lupron
Treatment
Figure 4 Mean 8 G recorded in 10 patients with PCOD before and after 6 weeks of LA treatment.
Insulin resistance improves with Lupron
Fertility and Sterility
Regression analyses with peripheral insulin sensitivity as the dependent variable and baseline insulin level as the independent variable were ~er formed using all patients with PCOD. A signifidmt positive correlation was observed between baseline insulin and Sr before LA therapy (r = 0.66; P < 0.02). This linear relationship was strengthened (r = 0.8, P < 0.002) when these variables were examined after LA treatment. DISCUSSION
Insulin resistance and hyperinsulinemia are common in women with the PCOS in which the androgen concentration correlates with the degree of insulin excess. In this study, GnRH-a suppression of ovarian function resulted in a significant improvement in insulin sensitivity in mildly insulin-resistant women with PCOD. This is in contrast to studies (11, 17) that demonstrated no improvement in Sr in insulin-resistant women with PCOD treated with a GnRH -a. Analysis of our patients with PCOD showed that five women had mild insulin resistance, five were severely insulin-resistant, and two had normal insulin sensitivity. Because of the heterogeneity of PCOD, it is possible that only by classifying patients with this disorder into subsets (according to the severity of their altered insulin-mediated glucose disposal) are changes in sensitivity to insulin evident with the suppression of hyperandrogenism. This approach is compatible with other investigators' suggestions (13, 18) that hyperandrogenic women need to be divided into at least two subgroups: those with insulin resistance, minimally elevated LH, and markedly elevated insulin and those with elevated LH, no insulin resistance, and normal insulin concentrations (18). Our findings suggest that in noninsulin-resistant women with PCOD, elevated LH levels would mainly act as the androgen-producing factor whereas, in insulin-resistant women with PCOD, both elevated insulin and LH levels may act synergistically in producing hyperandrogenism (18). The high insulin concentration in these insulin-resistant women may exert its steroidogenic action on the ovary through insulin -like growth factor (IGF-I). Insulin and IGF-I are closely related peptides that both have been shown to potentiate LHinduced ovarian androgen synthesis (9). Both insulin and IGF-I receptors have been found in human ovarian tissue and spill over action of insulin in states of insulin resistance such as PCOD might
Vol. 60, No.4, October 1993
induce hyperandrogenism by acting through the IGF -I receptor. Insulin -like growth factor- I and LH stimulate the ovaries, particularly theca cells, to secrete more androgen than either agent by itself (9). Thus, elevated insulin levels through IGF -1 receptors coupled with the action of LH could lead to excessive thecal secretion of androgens. Recently it was demonstrated that patients with PCOD have low serum levels of IGF-I binding proteins, which could further affect the stimulatory activity ofIGFI in the ovary (19). Insulin suppression of locally produced IGF-I binding proteins might be one of the factors in the pathogenesis of insulin-induced hyperandrogenism by increasing IGF-I receptor binding in the ovary and enhancing androgen production by thecal-interstitial and stromal cells. Given that mean GH and IGF-I concentrations were demonstrated to be significantly decreased after LA depot therapy in women with leiomyomata (20), an alternative hypothesis is that the improved insulin resistance and hyperinsulinemia in mildly insulin-resistant women with PCOD in this study was due to reduction of GH and IGF-I levels after Lupron treatment. A recent report by Ciaraldi and coworkers (21) showed functionally intact insulin receptor binding and kinase activity in adipocytes of women with PCOD. The most striking feature of insulin action in adipocytes from PCOD subjects was the marked reduction in insulin sensitivity for transport stimulation. They concluded that insulin resistance in PCOD is accompanied by the normal function of insulin receptors but involves a novel postreceptor defect in the insulin signal transduction chain between the receptor kinase and glucose transport (21). A postbinding defect in insulin action in PCOD was probably related to increasing levels of T in these women. Hormones, such as T, could induce insulin resistance in PCOD women directly by reducing the number and/or efficacy of glucose transporter proteins, in particular, the type 4 glucose transporter that appears to be responsible for the insulin -stimulated uptake of glucose in muscle and fat (22). The type 4 glucose transporters are present in high levels in tissues such as skeletal muscle and fat and mediate most insulin-stimulated glucose uptake in these tissues. A lack of a normal pool of insulin -sensitive glucose transporters could theoretically render an individual insulin resistant. In addition to stimulating insulin, glucose also
Elkind-Hirsch et al.
Insulin resistance improves with Lupron
639
appears to be an important regulator of its own metabolism; hyperglycemia in the presence of permissive insulin levels increases peripheral glucose uptake and decreases hepatic glucose output (23). During pregnancy and OC use, both high progestational states, SI decreases whereas noninsulin-dependent glucose utilization (SG) does not change (14). In the menstrual cycle, which is associated with less dramatic elevations in progesterone, no change in SG was observed (24). The absence of change in SG with LA treatment reflects a lack of impact of androgens on noninsulin-dependent glucose utilization. Further evidence supporting a role for hyperandrogenism in modulation of insulin sensitivity comes from the recent confirmation that a marked decrease in insulin sensitivity occurs in young women affected with the mild nonclassical form of congenital adrenal hyperplasia due to 21-hydroxylase deficiency (21-0HD) (25). Using the modified frequently sampled IVGTT with minimal model analysis to determine insulin sensitivity in young women with untreated nonclassical 21-0HD, Speiser et al. (25) showed that patients with nonclassical 21-0HD are significantly less sensitive to the action of insulin than female controls of similar age and weight. By studying women with androgen excess resulting from an inborn error in adrenal steroid biosynthesis, they proposed that hyperandrogenism can be the trigger to the cycle of androgeninsulin excess (25). However, whether this was a direct effect of androgen or mediated by secondary factors, such as components of the GH-insulin-like growth factor axis, is not clear. Although some evidence suggests that hyperinsulinemia may be an etiologic factor in PCOD, the increase in insulin sensitivity with normalization of the hyperandrogenemia after GnRH -a treatment in women with PCOD suggests hyperandrogenism might also playa role in maintaining insulin resistance in PCOD. In all PCOD patients, fasting baseline insulin concentrations correlate significantly with insulin sensitivity, suggesting initial baseline insulin levels are somewhat predictive of insulin sensitivity in patients with PCOD. The ability of the LA to suppress hyperandrogenism in PCOD patients yet strengthen the correlation of baseline insulin with insulin sensitivity in PCOD women further establishes that androgens modify insulin-mediated glucose disposal without affecting basal insulin levels.
640
Acknowledgments. We express our thanks to all personnel in the Methodist Hospital General Clinical Research Center for their unselfish cooperation and to TAP Pharmaceuticals in Deerfield, Illinois for the generous gift of Lupron. REFERENCES 1. Achard C, Theirs J. Le virilisme pilaire et son association a l'insuffisance glycolytique (diabete a femmes de barbe). Bull Ann Nat! Med (Paris) 1921;86:51-3. 2. Barbieri RL, Ryan KJ. Hyperandrogenism, insulin resistance, and acanthosis nigricans syndrome: a common endocrinopathy with distinct pathophysiologic features. Am J Obstet GynecoI1983;147:90-101. 3. Shoupe D. Lobo RA. The influence of androgens on insulin resistance. Fertil Steril1984;41:385-8. 4. Woodard TS, Burghen GZ, Kitabachi AE, Wilimas JA. Glucose intolerance and insulin resistance in aplastic anemia treated with oxymetholone. J Clin Endocrinol Metab 1981; 53:905-8. 5. Landon J, Wynn V, Samols E. The effect of anabolic steroids on blood sugar and plasma insulin levels in man. Metabolism 1963; 12:924-7. 6. Paik SG, Michelis MA, Kim YT, Shin S. Induction of insulin-dependent diabetes by streptozotocin: inhibition by estrogens and potentiation by androgens. Diabetes 1982;31: 724-8. 7. Geffner ME, Kaplan SA, Bersch NB, Golde DW, Landaw EM, Chang RJ. Persistence of insulin resistance in polycystic ovarian disease after inhibition of ovarian steroid secretion. Fertil SteriI1986;45:327-33. 8. Nagamani M, Dinh TV, Kelver ME. Hyperinsulinemia in hyperthecosis of the ovaries. Am J Obstet Gynecol 1986; 154:384-7. 9. Barbieri RL, Makris A, Randall RW, Daniels G, Kistner RW, Ryan KJ. Insulin stimulates androgen accumulation in incubations of ovarian stroma obtained from women with hyperandrogenism. J Clin Endocrinol Metab 1986;62:90410. 10. Dunaif A, GrafM. Insulin administration alters gonadal steroid metabolism independent of changes in gonadotropin secretion in insulin-resistant women with polycystic ovary syndrome. J Clin Invest 1989;83:23-9. 11. Dunaif A, Green G, Futterweit W, Dobrjansky A. Suppression of hyperandrogenism does not improve peripheral or . hepatic insulin resistance in the polycystic ovary syndrome. J Clin Endocrinol Metab 1990; 70:699-704. 12. Chang RJ, Nakamura RM, Judd HL, Kaplan SA. Insulin resistance, in nonobese patients with polycystic ovarian disease. J Clin Endocrinol Metab 1983;57:356-9. 13. Smith S, Ravnikar VA, Barbieri RL. Androgen and insulin response to an oral glucose challenge in hyperandrogenic women. Fertil Steril1987;48:72-7. 14. Bergman RN. 1989 Lilly lecture toward physiological understanding of glucose tolerance. Minimal-model approach. Diabetes 1989;38:1512-27. 15. Starr JI, Horowitz DL, Rubenstein AH, Mako ME. Insulin, proinsulin and C-peptide. In: Jaffe BM, Behrman HR, editors. Methods of hormone radioimmunoassay. 2nd ed. New York: Academic Press, 1979:613-42. 16. Bray GA, Jordan HA, Sims EAH. Evaluation of the obese patient. JAMA 1979;235:1487-90.
Elkind-Hirsch et al. Insulin resistance improves with Lupron
Fertility and Sterility
17. Dale PO, Tanbo T, Djoseland 0, Jervell J, Abyholm T. Persistence of hyperinsulinemia in polycystic ovarian syndrome after ovarian suppression by GnRH -agonist. Acta Endocrinol (Copenh) 1992; 126:132-6. 18. Antitila L, Ding YQ, Ruutiainen K, Erkkola R, Irjala K, Huhtaniemi I. Clinical features and circulating gonadotropin, insulin, and androgen interactions in women with polycystic ovarian disease. Fertil SterilI991;55:1057-61. 19. Pekonen F, Laatikainen T, Buyalos R, Rutanen EM. Decreased 34K insulin-like growth factor binding protein in polycystic ovarian disease. Fertil SterilI989;51:972-5. 20. Friedman AJ, Rein MS, Pandian MR, Barbieri RL. Fasting serum growth hormone and insulin-like growth factor-I and II concentrations in women with leiomyomata uteri treated with leuprolide acetate or placebo. Fertil Steril 1990;53: 250-3.
Vol. 60, No.4, October 1993
21. Ciaraldi TP, El-Roeily A, Madar Z, Reichart D. Cellular mechanisms of insulin resistance in polycystic ovarian syndrome. J Clin Endocrinol Metab 1992; 75:577-83. 22. Moller DE, Flier JS. Insulin resistance-mechanisms, syndromes and implications. N Engl J Med 1991;325:938-48. 23. Edelman SV, Laskso M, Wallace P, Bretchtel G, Olefsky JM, Baron AD. Kinetics of insulin-mediated and non-insulin mediated glucose uptake in humans. Diabetes 1990;39: 955-64. 24. Valdes CT, Elkind-Hirsch KE. IVGTT-derived insulin sensitivity changes during the menstrual cycle. J Clin Endocrinol Metabol 1991; 72:642-6. 25. Speiser, DW, Serrat J, New MI, Gertner JM. Insulin insensitivity in adrenal hyperplasia due to nonclassical steroid 21-hydroxylase deficiency. J Clin Endocrinol Metab 1992; 75:1421-4.
Elkind-Hirsch et al. Insulin resistance improves with Lupron
641
©
Vol. 60, No.4, October 1993
Printed on acid-free paper in U. s. A.
1993 The American Fertility Society
Insulin resistance improves in hyperandrogenic women treated with Lupron*t*
Karen E. Elkind-Hirsch, Ph.D.§11 Cecilia T. Valdes, M.D"I** L. Russell Malinak, M.D.** Baylor College of Medicine, Houston, Texas
Objective: To examine if changes in insulin sensitivity and glucose effectiveness in women with polycystic ovarian disease (PCOD) occurred after ovarian androgen suppression with a GnRH agonist, leuprolide acetate (LA, Lupron; TAP Pharmaceuticals, Deerfield, IL) using the minimal model method. Design: Twelve patients with PCOD were tested in the untreated state (baseline) and after 6 weeks of LA treatment. Subjects were divided into two groups based on the degree of impairment of their baseline insulin sensitivity index (SI; (min- 1 )(ILU/mL- 1 ): mild insulin resistance (SI > 1) or severe insulin resistance (SI < 1). Results: In all patients, serum T was significantly decreased from elevated baseline levels to normal female concentrations after 6 weeks of LA therapy. Insulin sensitivity in PCOD patients with mild insulin resistance significantly improved from baseline after 6 weeks of LA therapy, whereas no change in SI on LA therapy was seen in PCOD women with severe insulin resistance. Glucose utilization independent of increased insulin secretion did not change as a function of LA treatment in either group. Conclusion: These findings indicate a significant improvement in SI in mildly insulin-resistant women with PCOD after suppression of ovarian function with LA treatment. Fertil Steril1993;60:634-41 Key Words: Insulin sensitivity, GnRH agonist, polycystic ovarian disease
Since the description of "the diabetes of bearded women" in 1921 by Achard and Theirs (1) the remarkable correlation between androgen levels and
Received February 1, 1993; revised and accepted July 2, 1993.
* Lupron, TAP Pharmaceuticals, Deerfield, Illinois. t Supported by the Division of Research Resources of the National Institutes of Health under grant M01RR00350, Bethesda, Maryland and in part by an Educational Grant from TAP Pharmaceuticals, Deerfield, Illinois. :I: Presented in part at the 48th Annual Meeting of the American Fertility Society, New Orleans, Louisiana, October 31 to November 5, 1992. § Reprint requests: Karen Elkind-Hirsch, Ph.D., Obstetrical and Gynecological Associates, 7550 Fannin Street, Houston, Texas 77054-1989. II Department of Medicine. 1f Recipient of an American College of Obstetrics and GynecologyjOrtho Fellowship Grant, Baylor College of Medicine, Houston, Texas from 1989 to 1990. ** Department of Obstetrics of Gynecology.
634
Elkind-Hirsch et al.
insulin resistance in hyperandrogenic states has been known. Patients with polycystic ovarian disease (peOD) and the hyperandrogen-insulin resistance-acanthosis nigricans syndrome imply an association of elevated androgens and changes in insulin sensitivity, but the mechanism by which the hyperinsulinemia seen in insulin resistance leads to ovarian hyperandrogenism in peOD remains unclear (2). A number of studies suggest elevated androgens as the primary defect in peOD. Treatment with the antiandrogen spironolactone decreases insulin resistance in peOD patients concordant with a decline in plasma androgen levels (3). Older studies using exogenous androgen treatment for patients with aplastic anemia demonstrated glucose intolerance and elevated insulin levels (4). Administration of T or T derivatives to women results in impaired glucose tolerance and hyperinsulinemia (5). Furthermore, exposing streptozotocin-treated
Insulin resistance improves with Lupron
Fertility and Sterility
female rats to androgens increases the incidence of insulin-dependent diabetes (6). Although one group of investigators (Cole C, Kitabchi AE, abstract) found that treatment ofPCOD patients with oral contraceptives (OCs, Ortho-Novum; Ortho Pharmaceutical Corporation, Raritan NJ) lowered androgen levels and improved insulin sensitivity, other studies reported no effect of OCs on either of these parameters (2). Furthermore, suppressing androgen levels with a long-acting GnRH analogue (7) or with oophorectomy (8) had no effect on insulin resistance in hyperandrogenic women. Evidence points to hyperinsulinemia as a factor in the etiology of some cases of PCOD. In vitro studies, have shown that insulin and insulinlike peptides can augment basal and LH -stimulated ovarian androstenedione and testosterone production (9). In vivo studies, however, performed with acute, high-dose insulin infusions in obese insulinresistant women with PCOD resulted in decreased, rather than increased, androgen concentrations (10). By administration of a highly potent GnRH agonist (GnRH-a), such as leuprolide acetate (LA, Lupron; TAP Pharmaceuticals, Deerfield, IL), it is possible to effectively suppress pituitary and ovarian function. Recent evidence presented by Dunaif et al. (11), using direct measurement of peripheral and hepatic insulin action, showed that normalization of plasma androgen levels for 12 weeks with a GnRH -a did not significantly improve insulin sensitivity in insulin-resistant women with PCOD. The failure to improve insulin sensitivity in this study could possibly result from sampling only a small number of PCOD patients in which obesity was a significant factor responsible for their insulin resistance. However, although the presence of obesity certainly contributes substantially to the insulin resistance observed in obese patients with PCOD, the insulin resistance observed in normal-weight women with PCOD (12) suggests contributory factors that are specifically associated with this condition. Dunaif and colleagues (Dunaif A, Green G, Finegood D, abstract) demonstrated that hyperinsulinemia and insulin resistance in PCOD is secondary to increased basal secretion and decreased clearance but not pancreatic beta cell dysfunction. However, PCOD and obesity have a synergistic deleterious effect on glucose homeostasis. Smith et al. (13) evaluated a mixed group of hyperandrogenic women and reported two sets of responses to an oral glucose tolerance test; a marked insulin response and a slight or normal insulin response. This was
Vol. 60, No.4, October 1993
the first attempt to examine individual responses rather than group mean data and the authors suggested there may be two populations of hyperandrogenic women (13). Because of the heterogeneity of PCOD, it also is possible that androgens contribute to insulin -resistance in some women. Our study was undertaken to determine whether suppression of ovarian androgen production by using LA alters insulin sensitivity and glucose effectiveness determined using the minimal model method when PCOD patients have been classified as to the severity of their pretreatment insulin resistance (separation of patients by diagnostic category).
MATERIALS AND METHODS Study Patients
Fourteen women meeting clinical and biochemical criteria for PCOD were recruited. The subjects' mean age was 26.5 and ranged from 20 to 34. All patients had oligomenorrhea or amenorrhea, hirsutism, elevated plasma levels of T, and a LHjFSH ratio ~ 2.5. Adrenal disorders and hyperprolactinemia were specifically excluded. Polycystic ovarian disease women with abnormal elevations in DHEAS levels were excluded because GnRH-a treatment does not suppress adrenal androgen production. Five of the patients were obese (body mass index [BMI] > 29 kg/m2 ). Two patients had evidence of acanthosis nigricans. None of the women had taken any medications, including hormonal contraception, for the previous 6 months. Two obese patients withdrew from the study when they elected to discontinue LA therapy after the first month. Written informed consent was obtained from each subject, and the protocol was approved by the Baylor College of Medicine and Methodist Hospital Institutional Review Boards. Description of the Study
After a complete history and physical examination, the PCOD patients were studied in the untreated state (baseline) and after 6 weeks of ovarian androgen suppression with LA; 20 ~g/kg per day subcutaneously or 3.75 mg depot monthly. Each subject underwent a tolbutamide-modified frequently sampled intravenous glucose tolerance test (lVGTT) both pretreatment and 6 weeks after LA therapy.
Elkind-Hirsch et al. Insulin resistance improves with Lupron
635
All investigations were performed in the Methodist Hospital General Clinical Research Center in Houston, Texas. Blood samples were obtained for determination of baseline serum concentrations of T, DHEAS, androstenedione (A), E 2, P, 17-hydroxyprogesterone (17-0HP), LH, FSH, PRL, morning and evening cortisol, insulin, and thyroid functions (T 3, T 4 , TSH, and T3RU). In addition, a complete blood count was determined (hematocrit of 30 was required for participation). Study Protocol
Tolbutamide-Modified IVOTT
Each subject was instructed to consume a diet containing ~300 g carbohydrate in the 3 days before each IVGTT and all studies were begun between 6 and 10 A.M. after an overnight fast. A heparin lock was placed in the forearm for administration of medications and for blood sampling. Baseline samples for insulin and glucose levels were drawn 15, 20, 25, and 30 minutes after IV placement. Serum T, E 2, P, LH, FSH, PRL, and DHEAS levels were drawn 15 minutes after IV placement before each IVGTT. Glucose as a 25% solution (0.3 g/kg) was injected over 1 minute. Beginning 20 minutes after the glucose bolus, tolbutamide as a bolus of 150 mg/m2 (300 to 500 mg) was injected over 30 seconds (Orinase Diagnostic; Upjohn, Kalamazoo, MI). Blood samples (4 mL) for measurement of glucose and insulin were drawn at 2, 3, 4, 5, 6, 8, 10, 12, 14,16,19,22,23,24,25,27,30,35,40,50,60,70,80, 90, 100, 120, 140, 160, and 180 minutes after starting the glucose injection (14). Analytical Techniques and Calculations
Serum E 2, LH, FSH, PRL, DHEAS, T, cortisol, and P concentrations were measured in duplicate with commercial assay kits (Diagnostic Products, Los Angeles, CA) in the General Clinical Research Center RIA core laboratory. The interassay coefficients of variation of all hormone assays were <12%. Androstenedione and 17-0HP measurements were performed by Nichols Reference Laboratories (San Juan Capistrano, CA). Immunoreactive insulin was measured in duplicate by RIA using a modification of the procedures described by Starr et al. (15) and performed in the Diabetes and Endocrine Research RIA core laboratory. The interassay and intra-assay coefficients of variation were <10% and <5%, respectively. Serum glucose was mea-
636
Elkind-Hirsch et al.
sured by the glucose oxidase method in the Methodist Hospital Clinical Chemistry Laboratory. The relative proportion of body fat was expressed as the BMI (kg/m 2), which is considered to be the best estimate of body fat obtained from a single measure of height and weight; it has an excellent correlation with body fat estimated from body density (16). The nonobese group was defined as subjects within 20% of their ideal body weight with a BMI < 27 kg/m 2. The mean BMI for nonobese PCOD patients was 23 ± 1.5 (±SEM) kg/m2. The obese group was defined as subjects with weight ranging between 150% and 225% of ideal body weight and BMI > 29 kg/m2 (16). The mean BMI for obese PCOD patients was 47.8 ± 2.3 kg/m2. Results of the modified IVGTTs were then fitted to the minimal model of glucose kinetics using the MINMOD program (provided by Richard Bergman), giving an insulin sensitivity index (SI X 10-4 (min- 1)(J.tU/mL -1) and glucose effectiveness index (SG/min). In analyzing the results, one limit in the measurement of insulin sensitivity using the minimal model became apparent. Extreme insulin resistance could not be easily measured because a prerequisite for the minimal model to calculate insulin sensitivity is that the endogenous secreted insulin must measurably influence the plasma glucose level (14). Thus, from a clinical standpoint, evaluating PCOD subjects with glucose tolerance reduced to a diabetic value «1) confounds the minimal-model calculated insulin sensitivity index (SI) (14). To overcome the problem of diabetic tolerance values in very insulin-resistant states, subjects were divided into two groups: mild insulin resistance (SI > 1) or severe insulin resistance (SI < 1) group based on the degree of impairment of baseline SI value All three obese PCOD patients demonstrated severe insulin resistance (severe insulin resistance group). Two PCOD patients had normal baseline insulin sensitivity (SI > 6.5) and were excluded from the study analyses. Statistical Analyses
All results are reported as mean ± SEM values. Comparisons of baseline hormone measurements between groups were performed using either unpaired t-tests or analyses of variance (ANOVA). Statistical analyses of serum hormones and IVGTT-derived measurement values were performed using subjects-by-treatments repeatedmeasures ANOV A. The treatment variable was
Insulin resistance improves with Lupron
Fertility and Sterility
either hormone concentration (T, E 2, LH, FSH, DHEAS, PRL, and insulin) or IVGTT-derivedmeasurements (Sr, SG) for the two tests. Because the observations were made repetitively on the same subject, these observations are correlated. This violates the validity of the P values for the univariate repeated measures hypothesis tests (F-tests). Therefore all the F-tests were corrected using the Greenhouse-Geisser estimate of epsilon, which is the most conservative adjustment of the F-test. The amount of adjustment is determined by the factor epsilon. The reported P values reflect this correction. For further statistical evaluation of linear relationships between basal insulin levels and Sr before and after LA treatment, comparisons of data sets were performed using the Pearson product moment correlation test. The sample size required to detect a significant change in insulin sensitivity was calculated to be four subjects, setting the P value at 0.05, the power at 95%, and the expected effect size at 2.0 X 10- 4 (min-1)(J,LU/mL- 1).
Baseline IVGTT In PCOD-MIR Patient 5000
::J
.------------~20
Si=2.08
4000
15
0
E
3000
c
2000
~
:;
E
10
CD
0
U
glucose
III
C
5
1000
-60 -30
30
60
90 120150180
(minutes)
IVGTT In Lupron-Treated PCOD-MIR Patient 5000,----:-----------r20
-
::J
SI=7.95
4000
15
E
E
3000
E 10 CD
2000
III
o
III
Vol. 60, No.4, October 1993
glucose
C
Insulin and glucose profiles obtained during a tolbutamide-modified IVGTT in two different
~ o
o
;:,
Mod-IVGTT Measurements
CI
0 0
Time
c
In all PCOD patients, baseline levels of T and E2 levels were both significantly suppressed by LA whereas DHEAS levels remained unchanged. Estradiol levels fell from a mean before treatment concentration of 58 ± 8.7 pg/mL (213 ± 32 pmol/L) to after treatment levels of 23 ± 1.9 pg/mL (85 ± 7 pmol/L) (P < 0.0025). Testosterone levels decreased from elevated baseline levels of 91 ± 12 ng/dL (3.15 ± 0.4 nmol/L) to normal female concentrations of 34 ± 9 ng/dL (1.17 ± 0.3 nmol/L) after 6 weeks of LA therapy (P < 0.0001). The adrenal androgen, DHEAS, concentrations remained unchanged and were 169.7 ± 40.7 J,Lg/dL (4.6 ± 1.1 pmol/L) before treatment and 163 ± 48.1 J,Lg/dL (4.4 ± 1.3 pmol/L) after LA therapy. Although no significant differences in T, E 2, or DHEAS were found between PCOD-mild insulin resistance and PCOD-severe insulin resistance patients, PCOD patients with severe insulin resistance tended to be more androgenized; i.e., manifested higher mean T and DHEAS levels as compared with mildly insulin-resistant patients.
.2
insulin
o
D.
Hormone Measurements
.§. III
~
RESULTS
~
0
1000
u 5
.2
CI
insulin
o
0
-60-300
306090120150180
Time
(minutes)
Figure 1 Insulin and glucose levels used to estimate 81 and 8 a with the MINMOD program in a moderately insulin-resistant PCOD (PCOD-mild insulin resistance) subject before (top) and after (bottom) 6 weeks of treatment with LA.
patients is shown in Figures 1 and 2. One can readily observe the distinct increases in insulin concentration after glucose bolus at the 0 time point and the tolbutamide administration at the 20-minute time point. In all patients, the stimulated insulin secretion returned to baseline by 180 minutes. Figures 1 and 2 illustrate the insulin and glucose levels used to estimate Sr and SG with the MINMOD program in individual PCOD subjects with mild insulin resistance (Fig. 1) and severe insulin resistance (Fig. 2) before (top) and after (bottom) 6 weeks of LA treatment. Figure 3 shows that mean Sr X 10-4 (min-1)(J,LU/ mL- 1) values in the PCOD-mild insulin resistance (n = 5) group were significantly increased after 6 weeks of LA treatment with no change in the
Elkind-Hirsch et al. Insulin resistance improves with Lupron
637
Baseline IVGTT in PCOD- SIR Patient
Insulin Sensitivity in Lupron-Treated peOD Women
10000 ~--------------------------r20
-
::J
Si=0.3
8000
15
E c::
>-
o E
0
.e
= ~
~
6000
Ui
c::
E
CD
en
10 4000
c::
5 2000
3
CD III
c:: 2
u .:
:::I
o
:::I III
5
4
~
CI
1
insulin
o
Baseline
0
·60·300
306090120150180
Time
Lupron
Treatment Figure 3 Mean 8 1 recorded in 10 patients with PCOD. A significant increase was observed in the PCOD-mild insulin resistance group during LA treatment compared with baseline (P < 0.006).
(minutes)
IVGTT In Lupron-Treated PCOD-SIR Patient 20
10000
Si=0.33
8000
::J ::;
15
E
E E
6000
Q.
~
c::
~ 0
0
10 CD III 0
4000
:::I III
U
c::
5 2000
:::I CI
insulin
0 -60 -30
0 0
30
Time
60
90 120150180
(minutes)
Figure 2 Insulin and glucose levels used to estimate 8 1 and 8 G with the MIN MOD program in a severely insulin-resistant PCOD (PCOD-severe insulin resistance) subject before (top) and after (bottom) 6 weeks of treatment with LA.
PCOD-severe insulin resistance (n = 5) group. The mean insulin sensitivity index was 2.6 ± 0.34 in the untreated PCOD-mild insulin resistance group, which increased to 5.3 ± 0.75 after LA treatment (P < 0.006). Polycystic ovarian disease-severe insulin resistance group mean SI was unchanged despite equivalent suppression of ovarian androgens with LA treatment (basal mean SI = 0.5 ± 0.22 versus after LA treatment mean SI = 0.69 ± 0.13; P> 0.05). A significant difference was observed between mean SI value obtained in PCOD-mild insulin resistance patients versus PCOD-severe insulin resistance patients (P < 0.001). Glucose effectiveness (SG/min) did not change as a function of the LA therapy (F = 0.45, P > 0.05).
638
Elkind-Hirsch et al.
Baseline SG estimates were 0.24 ± 0.02/min in the PCOD-mild insulin resistance group and 0.24 ± 0.05/min in the PCOD-severe insulin resistance group. After 6 weeks of LA treatment, SG estimates were 0.26 ± 0.04/min in the PCOD-mild insulin resistance group and 0.28 ± 0.06/min in the PCODsevere insulin resistance group (Fig. 4). Mean basal insulin levels were significantly different between PCOD-mild insulin resistance and PCOD-severe insulin resistance patients both before LA (16.4 ± 1.7 mUlL [117.7 ±12.2 pmol/L] versus 45 ± 11.2 mUlL [322.9 ± 80.4 pmol/L]; P < 0.035) and after LA (12 ± 1.1 mUlL [86.1 ± 7.9 pmol/L] versus 53 ± 11.6 mUlL [380.3 ± 83.2 pmol/L], respectively; P < 0.007) while there was no significant change in baseline insulin concentrations after LA treatment within each group.
Glucose Effectiveness in Lupron-Treated peOD Women
c; ~
c:: CD
~
Eiil
0.4
•
Severe IR MildlR
0.3
U
.! iii
0.2
CD
0.1
U
::>
C;
0.0
Baseline
lupron
Treatment
Figure 4 Mean 8 G recorded in 10 patients with PCOD before and after 6 weeks of LA treatment.
Insulin resistance improves with Lupron
Fertility and Sterility
Regression analyses with peripheral insulin sensitivity as the dependent variable and baseline insulin level as the independent variable were ~er formed using all patients with PCOD. A signifidmt positive correlation was observed between baseline insulin and Sr before LA therapy (r = 0.66; P < 0.02). This linear relationship was strengthened (r = 0.8, P < 0.002) when these variables were examined after LA treatment. DISCUSSION
Insulin resistance and hyperinsulinemia are common in women with the PCOS in which the androgen concentration correlates with the degree of insulin excess. In this study, GnRH-a suppression of ovarian function resulted in a significant improvement in insulin sensitivity in mildly insulin-resistant women with PCOD. This is in contrast to studies (11, 17) that demonstrated no improvement in Sr in insulin-resistant women with PCOD treated with a GnRH -a. Analysis of our patients with PCOD showed that five women had mild insulin resistance, five were severely insulin-resistant, and two had normal insulin sensitivity. Because of the heterogeneity of PCOD, it is possible that only by classifying patients with this disorder into subsets (according to the severity of their altered insulin-mediated glucose disposal) are changes in sensitivity to insulin evident with the suppression of hyperandrogenism. This approach is compatible with other investigators' suggestions (13, 18) that hyperandrogenic women need to be divided into at least two subgroups: those with insulin resistance, minimally elevated LH, and markedly elevated insulin and those with elevated LH, no insulin resistance, and normal insulin concentrations (18). Our findings suggest that in noninsulin-resistant women with PCOD, elevated LH levels would mainly act as the androgen-producing factor whereas, in insulin-resistant women with PCOD, both elevated insulin and LH levels may act synergistically in producing hyperandrogenism (18). The high insulin concentration in these insulin-resistant women may exert its steroidogenic action on the ovary through insulin -like growth factor (IGF-I). Insulin and IGF-I are closely related peptides that both have been shown to potentiate LHinduced ovarian androgen synthesis (9). Both insulin and IGF-I receptors have been found in human ovarian tissue and spill over action of insulin in states of insulin resistance such as PCOD might
Vol. 60, No.4, October 1993
induce hyperandrogenism by acting through the IGF -I receptor. Insulin -like growth factor- I and LH stimulate the ovaries, particularly theca cells, to secrete more androgen than either agent by itself (9). Thus, elevated insulin levels through IGF -1 receptors coupled with the action of LH could lead to excessive thecal secretion of androgens. Recently it was demonstrated that patients with PCOD have low serum levels of IGF-I binding proteins, which could further affect the stimulatory activity ofIGFI in the ovary (19). Insulin suppression of locally produced IGF-I binding proteins might be one of the factors in the pathogenesis of insulin-induced hyperandrogenism by increasing IGF-I receptor binding in the ovary and enhancing androgen production by thecal-interstitial and stromal cells. Given that mean GH and IGF-I concentrations were demonstrated to be significantly decreased after LA depot therapy in women with leiomyomata (20), an alternative hypothesis is that the improved insulin resistance and hyperinsulinemia in mildly insulin-resistant women with PCOD in this study was due to reduction of GH and IGF-I levels after Lupron treatment. A recent report by Ciaraldi and coworkers (21) showed functionally intact insulin receptor binding and kinase activity in adipocytes of women with PCOD. The most striking feature of insulin action in adipocytes from PCOD subjects was the marked reduction in insulin sensitivity for transport stimulation. They concluded that insulin resistance in PCOD is accompanied by the normal function of insulin receptors but involves a novel postreceptor defect in the insulin signal transduction chain between the receptor kinase and glucose transport (21). A postbinding defect in insulin action in PCOD was probably related to increasing levels of T in these women. Hormones, such as T, could induce insulin resistance in PCOD women directly by reducing the number and/or efficacy of glucose transporter proteins, in particular, the type 4 glucose transporter that appears to be responsible for the insulin -stimulated uptake of glucose in muscle and fat (22). The type 4 glucose transporters are present in high levels in tissues such as skeletal muscle and fat and mediate most insulin-stimulated glucose uptake in these tissues. A lack of a normal pool of insulin -sensitive glucose transporters could theoretically render an individual insulin resistant. In addition to stimulating insulin, glucose also
Elkind-Hirsch et al.
Insulin resistance improves with Lupron
639
appears to be an important regulator of its own metabolism; hyperglycemia in the presence of permissive insulin levels increases peripheral glucose uptake and decreases hepatic glucose output (23). During pregnancy and OC use, both high progestational states, SI decreases whereas noninsulin-dependent glucose utilization (SG) does not change (14). In the menstrual cycle, which is associated with less dramatic elevations in progesterone, no change in SG was observed (24). The absence of change in SG with LA treatment reflects a lack of impact of androgens on noninsulin-dependent glucose utilization. Further evidence supporting a role for hyperandrogenism in modulation of insulin sensitivity comes from the recent confirmation that a marked decrease in insulin sensitivity occurs in young women affected with the mild nonclassical form of congenital adrenal hyperplasia due to 21-hydroxylase deficiency (21-0HD) (25). Using the modified frequently sampled IVGTT with minimal model analysis to determine insulin sensitivity in young women with untreated nonclassical 21-0HD, Speiser et al. (25) showed that patients with nonclassical 21-0HD are significantly less sensitive to the action of insulin than female controls of similar age and weight. By studying women with androgen excess resulting from an inborn error in adrenal steroid biosynthesis, they proposed that hyperandrogenism can be the trigger to the cycle of androgeninsulin excess (25). However, whether this was a direct effect of androgen or mediated by secondary factors, such as components of the GH-insulin-like growth factor axis, is not clear. Although some evidence suggests that hyperinsulinemia may be an etiologic factor in PCOD, the increase in insulin sensitivity with normalization of the hyperandrogenemia after GnRH -a treatment in women with PCOD suggests hyperandrogenism might also playa role in maintaining insulin resistance in PCOD. In all PCOD patients, fasting baseline insulin concentrations correlate significantly with insulin sensitivity, suggesting initial baseline insulin levels are somewhat predictive of insulin sensitivity in patients with PCOD. The ability of the LA to suppress hyperandrogenism in PCOD patients yet strengthen the correlation of baseline insulin with insulin sensitivity in PCOD women further establishes that androgens modify insulin-mediated glucose disposal without affecting basal insulin levels.
640
Acknowledgments. We express our thanks to all personnel in the Methodist Hospital General Clinical Research Center for their unselfish cooperation and to TAP Pharmaceuticals in Deerfield, Illinois for the generous gift of Lupron. REFERENCES 1. Achard C, Theirs J. Le virilisme pilaire et son association a l'insuffisance glycolytique (diabete a femmes de barbe). Bull Ann Nat! Med (Paris) 1921;86:51-3. 2. Barbieri RL, Ryan KJ. Hyperandrogenism, insulin resistance, and acanthosis nigricans syndrome: a common endocrinopathy with distinct pathophysiologic features. Am J Obstet GynecoI1983;147:90-101. 3. Shoupe D. Lobo RA. The influence of androgens on insulin resistance. Fertil Steril1984;41:385-8. 4. Woodard TS, Burghen GZ, Kitabachi AE, Wilimas JA. Glucose intolerance and insulin resistance in aplastic anemia treated with oxymetholone. J Clin Endocrinol Metab 1981; 53:905-8. 5. Landon J, Wynn V, Samols E. The effect of anabolic steroids on blood sugar and plasma insulin levels in man. Metabolism 1963; 12:924-7. 6. Paik SG, Michelis MA, Kim YT, Shin S. Induction of insulin-dependent diabetes by streptozotocin: inhibition by estrogens and potentiation by androgens. Diabetes 1982;31: 724-8. 7. Geffner ME, Kaplan SA, Bersch NB, Golde DW, Landaw EM, Chang RJ. Persistence of insulin resistance in polycystic ovarian disease after inhibition of ovarian steroid secretion. Fertil SteriI1986;45:327-33. 8. Nagamani M, Dinh TV, Kelver ME. Hyperinsulinemia in hyperthecosis of the ovaries. Am J Obstet Gynecol 1986; 154:384-7. 9. Barbieri RL, Makris A, Randall RW, Daniels G, Kistner RW, Ryan KJ. Insulin stimulates androgen accumulation in incubations of ovarian stroma obtained from women with hyperandrogenism. J Clin Endocrinol Metab 1986;62:90410. 10. Dunaif A, GrafM. Insulin administration alters gonadal steroid metabolism independent of changes in gonadotropin secretion in insulin-resistant women with polycystic ovary syndrome. J Clin Invest 1989;83:23-9. 11. Dunaif A, Green G, Futterweit W, Dobrjansky A. Suppression of hyperandrogenism does not improve peripheral or . hepatic insulin resistance in the polycystic ovary syndrome. J Clin Endocrinol Metab 1990; 70:699-704. 12. Chang RJ, Nakamura RM, Judd HL, Kaplan SA. Insulin resistance, in nonobese patients with polycystic ovarian disease. J Clin Endocrinol Metab 1983;57:356-9. 13. Smith S, Ravnikar VA, Barbieri RL. Androgen and insulin response to an oral glucose challenge in hyperandrogenic women. Fertil Steril1987;48:72-7. 14. Bergman RN. 1989 Lilly lecture toward physiological understanding of glucose tolerance. Minimal-model approach. Diabetes 1989;38:1512-27. 15. Starr JI, Horowitz DL, Rubenstein AH, Mako ME. Insulin, proinsulin and C-peptide. In: Jaffe BM, Behrman HR, editors. Methods of hormone radioimmunoassay. 2nd ed. New York: Academic Press, 1979:613-42. 16. Bray GA, Jordan HA, Sims EAH. Evaluation of the obese patient. JAMA 1979;235:1487-90.
Elkind-Hirsch et al. Insulin resistance improves with Lupron
Fertility and Sterility
17. Dale PO, Tanbo T, Djoseland 0, Jervell J, Abyholm T. Persistence of hyperinsulinemia in polycystic ovarian syndrome after ovarian suppression by GnRH -agonist. Acta Endocrinol (Copenh) 1992; 126:132-6. 18. Antitila L, Ding YQ, Ruutiainen K, Erkkola R, Irjala K, Huhtaniemi I. Clinical features and circulating gonadotropin, insulin, and androgen interactions in women with polycystic ovarian disease. Fertil SterilI991;55:1057-61. 19. Pekonen F, Laatikainen T, Buyalos R, Rutanen EM. Decreased 34K insulin-like growth factor binding protein in polycystic ovarian disease. Fertil SterilI989;51:972-5. 20. Friedman AJ, Rein MS, Pandian MR, Barbieri RL. Fasting serum growth hormone and insulin-like growth factor-I and II concentrations in women with leiomyomata uteri treated with leuprolide acetate or placebo. Fertil Steril 1990;53: 250-3.
Vol. 60, No.4, October 1993
21. Ciaraldi TP, El-Roeily A, Madar Z, Reichart D. Cellular mechanisms of insulin resistance in polycystic ovarian syndrome. J Clin Endocrinol Metab 1992; 75:577-83. 22. Moller DE, Flier JS. Insulin resistance-mechanisms, syndromes and implications. N Engl J Med 1991;325:938-48. 23. Edelman SV, Laskso M, Wallace P, Bretchtel G, Olefsky JM, Baron AD. Kinetics of insulin-mediated and non-insulin mediated glucose uptake in humans. Diabetes 1990;39: 955-64. 24. Valdes CT, Elkind-Hirsch KE. IVGTT-derived insulin sensitivity changes during the menstrual cycle. J Clin Endocrinol Metabol 1991; 72:642-6. 25. Speiser, DW, Serrat J, New MI, Gertner JM. Insulin insensitivity in adrenal hyperplasia due to nonclassical steroid 21-hydroxylase deficiency. J Clin Endocrinol Metab 1992; 75:1421-4.
Elkind-Hirsch et al. Insulin resistance improves with Lupron
641