Comparison of the effects of acarbose and metformin use on ovulation rates in clomiphene citrate-resistant polycystic ovary syndrome

A.S. Sönmez1,3, L. Yasar1, K. Savan1, S. Koç1, J. Özcan1, A. Toklar1, F. Yazicioglu1, A. Akgün1 and N. Sut2

1 Süleymaniye Maternity and Women's Disease Research and Teaching Hospital, Assisted Reproductive Techniques Unit and 2 Istanbul University, Cerrahpasa Medical School, Deparment of Biostatistics, Istanbul, Turkey

3 To whom correspondence should be addressed at: Kanuni Medresesi Sok. 1/9 Eminonu, 34470, Istanbul, Turkey. Email: suhas{at}net.net.tr


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: The aim of this study was to evaluate the effects of metformin and acarbose on insulin resistance, hormone profiles and ovulation rates in patients with clomiphene citrate-resistant polycystic ovary syndrome (PCOS). METHODS: Thirty clomiphene citrate-resistant patients were selected randomly and divided into two groups. Group I was treated with 100 mg/day clomiphene citrate and 300 mg/day acarbose 100 mg/day orally, for 3 months. Group II was treated with clomiphene citrate 100 mg/day and metformin 1700 mg/day orally, for 3 months. Serum fasting insulin and glucose, FSH, LH, estradiol, progesterone, prolactin and total testosterone levels plus body mass index (BMI) were measured before and after treatment. Follicle growth was followed by transvaginal ultrasonography. RESULTS: LH:FSH ratio and total testosterone concentrations decreased (P<0.05) and ovulation rates increased in both groups. Reduction in weight and BMI was only significant in the acarbose group. CONCLUSIONS: Both treatment modalities were effective in the treatment of insulin resistance and improving ovulation rates. Increase in the number of eumenorrhoeic and normoinsulinaemic cases and decrease in the number of insulin-resistant cases were significant in both groups (P<0.05). Ovulation rate was greater in the metformin group in the second month of therapy (P<0.05). Acarbose was found to be a safe and effective agent that could be used in cases with clomiphene-resistant PCOS.

Key words: acarbose/insulin resistance/metformin/polycystic ovary


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Polycystic ovary syndrome (PCOS) is a complex disorder characterized by dysregulation between the central nervous system, pituitary and adrenal glands and ovaries. Hyperinsulinaemia is one of the diagnostic features of PCOS and patients with PCOS are found to have resistance to either endogenous or exogenous insulin (Ciaraldi et al., 1992Go; Dunaif, 1997). About one-third of obese PCO patients have impaired glucose tolerance and 7.5–10% have type II diabetes mellitus. Increased androgen secretion and insulin resistance have been reported in both obese and non-obese anovulatory women. However, insulin levels are higher and LH, sex hormone-binding globulin (SHBG) and insulin-like growth factor binding protein levels are lower in obese women with polycyctic ovaries. Results of research with insulin-sensitizing drugs are promising and beneficial effects have been observed in PCO patients who have undergone ovulation induction or IVF (Ehrmann et al., 1997Go; Laure et al., 1998Go; Nestler et al., 1998Go, 2002Go; Lilliana et al., 2001Go;Yarali et al., 2002Go).

Metformin administration has been shown to improve ovarian response to gonadotrophins in women with clomiphene citrate-resistant PCOS (Yarali et al., 2002Go), but to our knowledge, acarbose, an {alpha}-glycosidase inhibitor, has not yet been studied in clomiphene citrate-resistant PCO patients. Acarbose is widely used in the management of Type II diabetes and can lower serum insulin concentration. There is only one study showing beneficial effects of acarbose use in cases with PCOS (Lilliana et al., 2001Go). Use of acarbose was associated with a significant decrease in insulin resistance, serum LH and androgen concentrations and with a significant rise of serum SHBG concentration (Lilliana et al., 2001Go). Hyperinsulinaemia is one of the causes of hyperandrogenaemia in patients with PCOS. Patients with PCOS have an exaggerated insulin response to glucose absorbtion. {alpha}-Glycosidase inhibitors reduce absorption of monosaccharides through intestines and minimize the increment of postprandial blood glucose and insulin levels, reversibly (Coniff et al., 1995Go). Serious side-effects of acarbose are rare and gradual increments of dose may minimize the side-effects (abdominal distention, flatulence, diarrhoea) mostly seen in the gastrointestinal system. Hypoglycaemia and lactic acidosis are not the manifestations of acarbose treatment. Therefore we preferred using acarbose as the therapeutic agent for insulin and clomiphene citrate resistance.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Patient selection
Thirty cases with clomiphene citrate-resistant PCOS were enrolled in the study between December 2001 and October 2002. The following criteria were used to diagnose PCOS, based on the modified consensus of the National Institutes of Health and Child Health and Human Development. Major criteria: (i) chronic anovulation (oligomenorrhoea, cycles >35 days; or amenorrhoea, lack of menstruation for ≥6 months); (ii) hyperandrogenaemia; (iii) clinical signs of hyperandrogenaemia (hirsutism, acne and seborrhoea); (iv) other aetiologies excluded. Minor criteria: (i) insulin resistance (fasting glucose:insulin ratio <4.5); (ii) peri-menarchal onset of hirsutism and obesity [body mass index (BMI) >25 kg/m2]; (iii) elevated LH:FSH ratio (>2); (iv) intermittent anovulation associated with hyperandrogenaemia (free testosterone, dehydroepiandrosterone sulphate).

Diagnosis of PCOS was confirmed by the presence of one major and two minor criteria besides infertility. Distribution of criteria used in the diagnosis of PCOS are shown in Table I.


View this table:
[in this window]
[in a new window]
 
Table I. Distribution of clinical and laboratory criteria used in the diagnosis of polycystic ovary syndrome in patients treated with acarbose or metformin

 
Patients having menstrual cycles >35 days were defined as oligomenorrhoeic and patients who did not have menstruation for ≥6 months were defined as amennorrhoeic. Menstrual cycles with 21–35 days were accepted as normal. We defined hyperandrogenaemia as a serum total testosterone level >80 ng/dl.

Classically, the daily dose of clomiphene citrate recommended for the first treatment attempt is 50 mg, which can be increased up to a total daily dose of 150 mg, although a total daily dose of 250 mg has also been used and patients who did not respond to gradual increments of clomiphene citrate treatment are defined as clomiphene citrate resistant (Gorlitsky et al., 1978Go). In our study, cases who did not ovulate at the dose of 150 mg/day were defined as clomiphene citrate resistant. We considered anovulation in patients with a progesterone level <2.5 ng/ml on the 21st day of cycle, inadequate rise in estradiol (E2) level and absence of mature follicle on ultrasound performed in the 14th day of cycle. All women were in good health and none had taken any other medication during the study period. Diabetic and hyperprolactinaemic patients and those who smoke cigarette and drink alcohol were not included in the study. Patients were advised to avoid any restrictive diet and physical exercise during the study period.

Study design and procedures
The study was approved by the Ethical Comittee of our hospital and all the patients were given informed consent. All the patients underwent clinical examination. The body mass index was calculated using the following formula: BMI = weight (kg)/[height (m)]2. After an overnight fast, serum concentrations of insulin, glucose, LH, FSH, thyroid-stimulating hormone (TSH), thyroxine, E2, progesterone, prolactin, testosterone were measured during the early follicular phase (cycle days 2–5) of a spontaneous or progestin-induced bleeding. Patients with a fasting glucose:fasting insulin ratio <4.5 were defined as insulin resistant.

Patients were divided into two groups using a computer-generated randomization (Random Numbers Generator Pro; Segobit Software trial version) each consisting of 15 patients. Based on data from the study by Lilliana et al. (2001)Go a power analysis suggested a sample size of 14 patients per group. The parameter we used to determine the sample size was the number of cases with resumed regular menstrual rhythm. We accepted the prevalance of regular menstrual rhythm 0.60 and alternative rate 0.10 [{alpha} (Type I) error = 0.05 and {beta} (Type II) error = 0.80]. Patients in group I (acarbose group) were treated with 100 mg/day of clomiphene citrate (Klomen®, Koçak, Turkey; 50 mg tablet, twice daily) and 300 mg/day of acarbose (Glucobay®, Bayer, Turkey; 100 mg tablet, three times daily) for 3 months, patients in group II (metformin group) were treated with 100 mg/day of clomiphene citrate and 1700 mg/day metformin (Glucophage Retard®, Ilsan, Turkey; 850 mg tablet, twice daily). Serum fasting insulin and glucose, FSH, LH, TSH, thyroxine, E2, progesterone, prolactin and testosterone levels were measured and BMI were calculated in all patients after treatment. Ovulation rates were evaluated in both groups, over 3 months.

The patients underwent transvaginal ultrasonography to evaluate the endometrial thickness, uterus, ovaries and the number and diameters of the follicles on the third day of the cycle. Follicular growth and ovulation was followed by transvaginal ultrasonography. Follicular growth was followed every 2 days, from the 8th day of cycle. Serum mid-luteal progesterone concentration was measured to determine the ovulation on the 21st day of the cycle. Mid-luteal serum progesterone concentration >8 ng/ml was considered to be a sign of ovulation.

Statistical evaluation
Each group was evaluated for the changes in the weight, BMI, menstrual pattern, follicular growth and hormone concentrations at the end of treatment. All variables were tested for normal distribution with Kolmogorov–Smirnoff test and all were found to have normal distribution. The impact of the treatment on all variables within each group was detected by the paired t-test. Changes in the pre- and post-treatment values in acarbose and metformin groups were analysed by analysis of covariance. Mean monthly progesterone levels between both groups were compared by independent samples t-test. Monthly numbers and the ratio of cases with ovulation in each group were compared by Fisher's exact test. P<0.05S was considered significant. The statistical analysis of the data was made by SPSS for Windows 10.1.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Pre-treatment values
Distribution of clinical and laboratory criteria used in the diagnosis of PCOS was similar in both groups (Table I). In the acarbose group, there were eight cases (53.3%) with three major criteria, 11 cases (73.3%) with two major criteria, 11 cases (73.3%) with three minor criteria and 15 cases (100%) with two minor criteria. In the metformin group, there were seven cases (46.6%) with three major criteria, nine cases (60%) with two major criteria, 10 cases (66.6%) with three minor criteria, 15 cases (100%) with two minor criteria.

Mean age (±SD) was 26 (range 19–40) in the study population. Mean age was 26.13±5.08 years (range 20–40) in acarbose group and it was 26±3.92 years (range 19–36) in metformin group. Mean body weight (±SD) was 68.3±6.03 kg in group I (acarbose group) and it was 68.5±6.41 kg in group II (metformin group). There was no significant difference between the two groups.

Mean serum concentrations of FSH, LH, testosterone, E2, prolactin, fasting insulin, LH/FSH, and total testosterone concentrations in each group are shown in Table II.


View this table:
[in this window]
[in a new window]
 
Table II. Comparison of the clinical and laboratory findings before and after treatment

 
Pretreatment serum concentrations of thyroxine, TSH, fasting glucose, urea, serum glutamic oxalacetic transaminase and serum glutamic–pyruvic transaminase were in normal ranges. Blood pressures and pelvic examination findings of cases were also normal.

In the acarbose group, only two out of 15 cases (13.3%) have had normal menstrual cycles (28 and 30 days). In the acarbose group, eight cases (53.3%) were oligomenorrhoeic and five cases (33.3%) were amenorrhoeic. In the metformin group, whereas five cases (33.3%) had normal menstrual cycles (all had menstrual cyles >26 days), seven cases (46.6%) were oligomenorrhoeic and three cases (20%) were amenorrhoeic (Table II). There was no difference between the two groups with respect to percentages of menstrual irregularity (Table III).


View this table:
[in this window]
[in a new window]
 
Table III. Changes in laboratory and clinical parameters following therapy with acarbose versus metformin

 
Effects of acarbose and metformin treatment
There was a statistically significant weight reduction in the acarbose group (P<0.01). The reduction in weight was not statistically significant in the metformin group (P>0.05). In the acarbose group, mean BMI was 27.4 kg/m2 before treatment, 26.3 kg/m2 after treatment (P<0.01). Mean BMI in the metformin group decreased from 27.2 to 26.9 kg/m2 at the end of the treatment (P<0.05) (Table II). There was a statistically significant difference (P<0.05) between two groups with regard to body weight and BMI (Table II).

Normal menstrual cycle was achieved in nine of 13 (69.2%) patients in the acarbose group (two patients already had normal menstrual cycles of 28 and 30 days) (P<0.01) whereas eight of 10 (80%) patients in the metformin group achieved normal menstrual cycles (five patients already had normal menstrual cycle) (P<0.01). Shortest menstrual cycle length was 26 days in the metformin group. Both groups were similar concerning the number of cases having normal menstrual cycle at the end of the therapy (P>0.05).

E2 and prolactin concentrations did not change significantly after treatment, in both groups (Table II).

LH, LH:FSH ratio, testosterone and fasting insulin were found to be reduced and FSH was increased significantly (P<0.05), but there was no significant difference between the two groups (P>0.05) (Tables II and III). Monthly mid-luteal serum progesterone concentrations and the number of ovulated cases increased significantly in both groups (P<0.05) over the 3 month period (Table IV). Ovulation rates and mean progesterone levels were similar in groups I and II in the 1st and 3rd months of therapy. There was a significantly higher ovulation rate in the metformin group than in the acarbose group in the 2nd month of therapy (P<0.05) (Table IV).


View this table:
[in this window]
[in a new window]
 
Table IV. Consecutive mid-luteal serum progesterone concentrations and percentage of ovulation in acarbose and metformin groups

 
None of the patients treated either with acarbose or metformin discontinued the treatment because of side-effects. Minimal gastrointestinal side-effects (abdominal distention and diarrhoea) occured in nine cases in the acarbose group and seven cases in the metformin group. There was no difference between the groups with respect to side-effects.


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Acarbose, an inhibitor of the enzyme {alpha}-glycosidase, is a drug in the treatment of non-insulin-dependent diabetes mellitus (NIDDM) with minimal side-effects and high clinical efficiency as an anti-hyperglycaemic agent (Coniff et al., 1995Go).

Acarbose has not been studied sufficiently in the treatment of patients with PCOS. Lilliana et al. (2001)Go had impressive results concerning acarbose in the treatment of PCOS patients. In that study, 30 hyperinsulinaemic PCOS patients with normal weight and a control group of 15 patients were included. The clinical improvement was relevant to the significant decline of insulin after the glycose loading, the decrease in LH and total testosterone and androsteneidone concentrations and the increase in SHBG concentrations. PCOS patients treated with placebo showed no clinical, metabolic or hormonal change.

Nestler et al. (2002)Go showed that PCOS patients with obesity had minimum response to the induction of ovulation by clomiphene citrate and in these cases metformin significantly induced not only clomiphene citrate-induced ovulation but also spontaneous ovulation. By contrast, some other studies have shown the inefficacy of metformin in PCOS patients (Ehrmann et al., 1997Go). In the study by Nestler et al. (1998)Go, clomiphene citrate was used as the first choice of the monotherapy and metformin was combined if ovulation was achieved by clomipene citrate alone. In PCOS patients with infertility, metformin use as an alternative to clomiphene citrate therapy is still speculative since more studies are required.

Toeller (1991) used acarbose in a 300 mg daily dose and showed reliable metabolic and endocrinological results. This influenced us to use acarbose, 300 mg daily, in our study. Before the treatment, mean body weight was estimated as 68.3 kg whereas it was 65.6 kg after the patients' use of acarbose. This result was found to be significant (P<0.01). On the other hand, the mean body weight was 68.5 kg before the metformine treatment and only 67.7 kg after treatment (not insignificant). The effect of mean body weight in terms of BMI was as follows: in the acarbose group, mean BMI was 27.4 kg/m2 before treatment and 26.3 kg/m2 after treatment. In the metformin group, mean BMI was 27.2 and 26.9 kg/m2, before and after treatment respectively. The decrease of BMI with acarbose use was found to be significant (P<0.01) whereas with the use of metformin it was not. Nevertheless there was a significant difference between groups (P<0.05). In a similar study, 61 women with PCOS were treated with metformin and no significant change in BMI was observed (Nestler et al., 2002Go). In a study by Lilliana et al. (2001)Go, acarbose was used in 30 PCOS patients with normal BMI and no significant change in BMI was observed in this group. Whereas we evaluated the impact of acarbose use on the ovulation rates in CC resistant cases with PCOS in the latter, study population was consisting particularly hyperinsulinaemic and non-obese cases with PCO and so there may be differences between our results and theirs, e.g. in BMI.

There is no study concerning the role of acarbose on ovulation rates in PCOS patients. In the study by Lilliana et al. (2001)Go, no methods were used to determine ovulation rates during acarbose treatment (e.g. serum progesterone concentration, ultrasonography). On the other hand, there have been many studies showing that metformin induces spontaneous ovulation. In the study by Moghetti et al. (2000)Go, 32 patients were followed for 26 months. The patients achieving menstrual cycles ‘normal’ or ‘better than pretreatment’ was considered to be a ‘positive’ response and 54% of patients using metformin had a positive outcome in terms of having ‘normal menstrual cycles’.

Acarbose has a lower absorption rate than metformin through the intestine (0.5–1.7%) and this feature limits the adverse effects seen in a dose-related manner such as abdominal distention, flatulence and diarrhoea. These adverse effects can be minimized by gradually increasing the dose of acarbose. Acarbose does not cause hypoglycaemia or lactic acidosis. The most frequent side-effects of metformin are nausea, vomiting, fatigue, diarrhoea, especially in patients with chronic renal failure. Therefore patients should be investigated for renal functions during metformin use. Metformin also reduces B12 absorption through the intestine (Campbell and Gerich, 1990Go).

We observed significant improvement in the number of normoinsulinaemic, normotestosteronaemic and ovulating patients with the outcome of regular menstrual cycles in both groups. Ovulation rate in metformin group was greater than in acarbose group in second month of therapy. Insulin resistance was overcome in most of the patients and LH:FSH ratio was lowered to <2, but the difference between the two groups was not statistically significant.

In conclusion, both acarbose and metformin improved hormonal parameters and effected ovulation in clomiphene citrate-resistant PCOS patients. However, decrease in BMI was greater in the acarbose group and the difference between the two groups was significant. Metformin administration has already been shown to improve ovarian response to gonadotrophin in women with clomiphene citrate-resistant PCOS. To our knowledge, acarbose, which has been widely used in the treatment of type II diabetes mellitus, has not been studied in clomiphene citrate-resistant PCOS patients and this is the first study to evaluate the clinical and endocrinological effects of acarbose on clomiphene citrate-resistant PCOS patients. We evaluated the effects of acarbose on ovulation rates; previously no data on ovulation rates were available (Lilliana et al., 2001Go). Acarbose seems to be an effective and safe agent to be used in patients with clomiphene citrate-resistant PCOS, but further long-term, double-blind, controlled, large-scale studies are needed to determine whether acarbose can be used with another drug or alone in the treatment of such patients.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Campbell PJ and Gerich JE (1990) Impact of obesity on insulin action in volunteers with normal glucose tolerance; demonstration of a threshold for the adverse effect of obesity. Clin Endocrinol Metab 70, 1114–1118.

Ciaraldi TP, el Roeiy A, Madar Z, Reichart D, Olefsky JM and Yen SC (1992) Cellular mechanisms of insulin resistance in polycystic ovary syndrome. J Clin Endocrinol Metab 75, 577–583.[Abstract]

Coniff RF, Seton TB and Shapiro JA (1995) Reduction of glycosylated hemoglobin and postprandial hyperglicemia by acarbose in patients with NIDDM. Diabetes Care 18, 817–820.[Abstract]

Dunaif A (1997) Insulin resistance and policystic ovary syndrome. Endocr Rev 18, 774.[Abstract/Free Full Text]

Ehrmann DA, Cavaghan MK, Imperial J, Sturis J, Rosenfield RL and Polonsky KS (1997) Effects of metformin on insulin secretion, insulin action, and ovarian steroidogenesis in women with polycystic ovary syndrome. J Clin Endocrinol Metab 82, 524–530.[Abstract/Free Full Text]

Gorlitsky GA, Kase NG and Speroff L (1978) Ovulation and pregnancy rates with clomiphene citrate. Obstet Gynecol 51, 265–269.[Abstract]

Laure CM, Koivunen RM, Ruokonen A and Martikainen HK (1998) Metformin therapy improves the menstrual pattern with minimal endocrine and metabolic effects in women with polycystic ovary syndrome. Fertil Steril 69, 691–696.[CrossRef][ISI][Medline]

Lilliana C, Calogero AE, Farina M, De Leo V, La Marca A and Cianci A (2001) Clinical, endocrine and metabolic effects of acarbose, an alpha-glucosidase inhibitor, in PCOS patients with increased insulin response and normal glucose tolerance. Hum Reprod 10, 2066–2072.[CrossRef]

Moghetti P, Castello R, Negri C, Tosi F, Perrone F and Caputo M (2000) Metformin; effects of clinical features, endocrin and metabolic profiles and insulin sensitivity in polycystic ovary syndrome: a randomized double-blind, placebo controlled, 6-month trial, followed by open, long term clinical evaluation. J Clin Endocrinol Metab 85, 139–146.[Abstract/Free Full Text]

Nestler JE, Jakubowicz DJ, Evans WS and Renato P (1998) Effects of metformin on spontaneous and clomiphen-induced ovulation in the polycystic ovary syndrome. N Engl J Med 25, 1876–1880.[CrossRef]

Nestler JE, Stowall D, Akhter N, Iuorno JM and Jackubowicz DJ (2002) Strategies for the use of insulin-sensitizing drugs to treat infertility in women with polycystic ovary syndrome. Fertil Steril 77, 209–213.[CrossRef][ISI][Medline]

Toeller M (1991) Modulation of intestinal glucose absorption: postponement of glucose absorption by alpha-glucosidase inhibitors. In Mogensen CE and Standdl E (eds) Pharmacology of Diabetes. Walter de Gruyter, Berlin.

Yarali H, Yildiz OB, Demirol A, Zeyneloglu HB, Yigit N, Bükülmez O and Koray Z (2002) Co-administration of metformin during rFSH treatment in patients with clomiphene citrate resistant polycystic ovarian syndrome: a prospective randomized trial. Hum Reprod 17, 289–294.[Abstract/Free Full Text]

Submitted on November 6, 2003; resubmitted on March 10, 2004; accepted on September 29, 2004.