The efficacy of metformin and clomiphene citrate combination compared with clomiphene citrate alone for ovulation induction in infertile patients with PCOS.

CONTEXT: Low ovulatory and pregnancy rates with clomiphene citrate (CC) in anovulatory polycystic ovarian syndrome (PCOS). AIM: To find out the ovulatory and pregnancy rates in infertile PCOS subjects who receive CC alone and a combination of metformin and CC. SETTING AND
DESIGN: A prospective controlled clinical trial conducted in the outpatient department from August 2003 to August 2005.
MATERIALS AND METHODS: Twenty-four infertile PCOS women received CC alone at incremental doses of 50 mg up to 150 mg for three cycles and then at a dose of 150 mg for another three cycles (control group). The study group (16 PCOS) received the same dose of CC along with 1500mg of metformin. Ovulation was monitored by transvaginal sonography up to six cycles or till pregnancy occurred. STATISTICAL ANALYSIS: This was carried out using software SSPS, version 10. Fisher's exact test was used to calculate the ovulatory rates. Nine subjects of the control group who failed to conceive with CC had opted for CC and metformin and their ovulatory rate was calculated using statistical software, namely SPSS 15.0, Stata 8.0, MedCalc 9.0.1 and Systat 11.0 using Fischer's exact test.
RESULTS: The metformin and clomiphene combination resulted in a significantly higher rate of ovulation (P = 0.0016). The pregnancy rate was 8% with CC and 24% with metformin and CC. The CC failure group also ovulated at a similar rate as that of the study group.
CONCLUSIONS: The ovulatory rate and the pregnancy rate with the metformin-CC combination was found to be higher when compared with CC alone. Metformin increased the ovulatory rate in CC failures, also implying increased sensitivity to CC.

RCT Entities:   Population Interventions Outcomes

INTRODUCTION

Polycystic ovarian syndrome (PCOS) is one of the most common causes of anovulatory infertility. The initial therapy for anovulation is clomiphene citrate (CC) but, unfortunately, ovulation is not achieved in almost 40% of PCOS.[1] Further, when the patient is subjected to more than six cycles of CC, there is a possibility of increased risk of ovarian cancer.[2]

Metformin is found to increase ovulation and thereby pregnancy rates by improving insulin resistance in PCOS.[34] The hyperinsulinemia in PCOS accounts for decreased response to CC. This study aimed to assess the ovulatory rate by transvaginal sonography in infertile PCOS subjects receiving CC alone and in those who received the CC and metformin combination. An attempt is also made to find out the response of CC failures to the metformin and CC combination.

MATERIALS AND METHODS

This was a controlled clinical trial conducted in the Department of Obstetrics and Gynaecology from August 2003 to August 2005. Women reporting for infertility treatment were diagnosed to have PCOS by the Rotterdam criteria (2003).[5] After explaining about the drugs used to induce ovulation and the necessity of regular frequent visits for follicular monitoring, they were grouped into the control group to receive CC alone (24) and the study group to receive the CC and metformin combination (16). The male factor was normal in both groups and the tubal factor was excluded by performing hysterosalphingogram or laparoscopy chromotubation. Cases with recent pelvic inflammatory disease were excluded. An informed consent was taken from all the patients and the study was approved by the Institute Scientific Counsel and Ethics committee.

The study group received 500 mg three times a day of metformin continuously (the same dose) from the first cycle for 6 months or till pregnancy was confirmed. CC was started at a dose of 50 mg from day 2 of the menstrual cycle till day 6 for a total of 5 days. The dose of CC was increased to 100 mg in the second cycle and 150 mg during the third cycle and CC was given at a dose of 150 mg for the remaining three cycles. Follicular growth was monitored by transvaginal ultrasound with a 5 MHz frequency probe UAGVO 14A of Toshiba ECCO CEECX, SSA340A, Minato Ku, Tokyo, Japan. from the 10th day of the menstrual cycle till the follicle reaches 18–20 mm. At this time, the patient was given 10,000 IU of HCG intramuscularly and was advised to have coitus after 36 h. Confirmation of ovulation was made by noting the signs of ovulation by transvaginal sonography after 48 h of administration of HCG.

The control group received only CC at the same dosage as that of the study group and was monitored similarly. Monitoring was performed for a maximum of six cycles or till pregnancy occurred.

There were nine patients who did not conceive with six cycles of CC alone and these patients were given CC + metformin for an additional six cycles (or till pregnancy) at their request for further treatment. These are designated as CC failure group/subgroup.

The statistical analysis was performed with the software SPSS; version 10.0. The ovulatory rates were calculated by Fisher's exact test. The ovulatory rates of the subgroup (CC failure group) was calculated using statistical software, namely SPSS 15.0, Stata 8.0, MedCalc 9.0.1 and Systat 11.0 using Fischer's exact test.

The ovulatory rates are calculated for the study group of 16 patients and the control group of 24 patients [Table 1] and for the CC failure group/subgroup [Table 2] and also for the total cases who received CC + metformin (16 + 9 = 25) and for the control group (24) for comparison [Table 3].

Table 1

Comparison of ovulatory rates (CC alone vs CC + metformin)

CC dose (mg)	Control group (n = 24)			Study group (n = 16)			P-value
	Ovulatory cycles	Total cycles	% ovulated	Ovulatory cycles	Total cycles	% ovulated
50	8	24	33.3	7	16	43.8	0.505 NS
100	6	24	25.0	9	16	56.3	0.046 S
150 (1)	10	24	41.6	12	16*	75.0	0.038 S
150 (2)	8	24	33.3	10	14*	71.4	0.044 S
150 (3)	12	23*	52.1	11	13*	84.6	0.075 NS
150 (4)	12	23*	52.1	9	12*	75.0	0.282 NS
Total	56	142	39.4	58	87	66.6

Patients who became pregnant were excluded; NS= not significant; S= significant

Table 2A

Comparison of ovulation before and after addition of metformin in the same patient

Cycle	CC dose (mg)	Control group (24)			Subgroup (9) (CC failure group)			P-value⌖
		Ovulatory cycles	Total cycles	% ovulated	Ovulatory cycles	Total cycles	% ovulated
1	50	2	9	22	5	9	55	0.375
2	100	1	9	11	5	9	55	0.219
3	150 (1)	1	9	11	7*	9	77	0.031 S
4	150 (2)	1	9	11	6*	8	75	0.002 S
5	150 (3)	6	9	66	7	8	87	0.05 S
6	150 (4)	6	9	66	7	7	100	0.018
Total		16	54	29	36	50	72

Patients who became pregnant were excluded

P-value shown in this table is calculated by the 2/2 contigency tables shown in Table 3B

Table 3

Comparison of ovulatory rates in each cycle for total cycles

CC dose (mg)	Control group (n = 24)			Study group (n = 16 + 9 = 25)			P-value
	Total cycles	Ovulatory cycles	% ovulated	Total cycles	Ovulatory cycles	% ovulated
50	24	8	33.3	25	13	52	0.251 NS
100	24	6	−25	25	14	56	0.042 S
150 (1)	24	−10	41.6	25	19	76	0.021 S
150 (2)	24	8	33.3	−22*	16	−72.7	0.046 S
150 (3)	−23*	12	52.1	−21*	18	−85.7	0.024 S
150 (4)	−23*	12	52.1	−19*	16	−84.2	0.047 S
Total	−142	56	39.4	137	96	70	0.00016 S

Patients conceived were excluded; P-value <0.05 is considered significant

RESULTS

The age group of the subjects varied between 20 and 38 years. Only 12.5% were more than 31 years of age. There were two subjects who were more than 35 years of age in the study group and both of them were 38 years old. There was only one patient who was more than 35 years in the control group and she was also 38 years of age. Hirsutism was the most common clinical finding in all subjects of the study group and 75% of the control group showed hirsutism of various grades. Amennorrhea was more common than oligomenorrhea. Obesity was present in 37.5%. There were three patients with hypothyroidism and one patient with non-insulin dependent diabetes mellitus, which was well controlled before enrollment in this study. There is no statistical significance among the clinical features between the study group and the control group except for hirsutism, which was more in the study group [Table 41].

The ovulatory rates in both the groups are shown in Table 1. From this, it can be inferred that the combination of metformin and clomiphene resulted in higher ovulatory rates, which are statistically significant at a dose of 100 and 150 mg of CC (P = 0.046 and 0.038, respectively). Subjecting to a second course of 150 mg resulted in an almost similar ovulatory rate (71.4%). In the control group, the highest ovulatory rate was 52.1% and this was achieved only with the third and fourth courses of 150 mg of CC. Pregnancy resulted with 150 mg of CC in both the groups. In the study group, pregnancy was achieved earlier. i.e. during the first course of 150 mg of CC, whereas in the control group, pregnancy occurred only after receiving two courses of 150 mg of CC [Table 2]. Additional third and fourth courses of CC + metformin improved the ovulatory rate by 10% and pregnancy was achieved in two cases. Thus, it can be inferred that in subjects receiving CC alone, an ovulatory rate of 52% can be achieved only with 150 mg and pregnancy occurs at a later period, i.e. during the third and subsequent doses of 150 mg, most probably after a period of sensitization to that particular dose. Addition of metformin improved the ovulatory rate significantly in the first course of 150 mg itself (implying increase in sensitivity to CC) and also the pregnancy rate, although not statistically significant.

Nine subjects of the control group who failed to conceive with six cycles of CC had opted for CC + metformin therapy, extending their treatment to a further six cycles. (CC failure in this study is considered when there is no conception after receiving six cycles, although there is ovulation.) The ovulatory rate before and after addition of metformin at the same dose of CC as per protocol was compared and is shown in Table 2A as subgroup analysis. It is seen that the percentage of ovulatory cycles was high in the same patient with CC combined with metformin than CC alone. On using Mcnemers χ2 test, there was no statistical significance as the sample size in the subgroup is small. Hence, using the 2/2 contingency table for all doses, the ovulatory rate before and after addition of metformin was calculated in the same patient [Table 2B]. The statistical software, namely SPSS 15.0, Stata 8.0, MedCalc 9.0.1 and Systat 11.0 were used for the analysis of the data. CC + metformin gave a significant outcome of ovulation Cycle 3, Cycle 4 and Cycle 6 when compared with CC alone.

Table 2B

Effect of treatment of CC + metformin on ovulation in comparison with CC alone

Cycle	Group	Total	Not ovulated	Ovulated	P-value
1	CC + metformin	9	4	5	0.335
	CC alone	9	7	2
2	CC + metformin	9	4	5	0.131
	CC alone	9	8	1
3	CC + metformin	9	2	7	0.041*
	CC alone	9	6	1
4	CC + metformin	8	2	6	0.007**
	CC alone	8	8	0
5	CC + metformin	8	1	7	0.588
	CC alone	8	3	5
6	CC + metformin	7	0	7	0.070+
	CC alone	9	4	3

CC + metformin has given a significant outcome of ovulation in Cycle 3, Cycle 4 and Cycle 6 when compared with CC alone

The pregnancy rates in both the groups are shown in Table 5. The pregnancy rate was 25% in the study group when compared with 8% in the control group. In the CC failure group, two patients achieved pregnancy out of the nine treated (22.2%), which is also similar to the study group. Although there is no significant statistical difference, pregnancy rates were high with the CC and metformin combination. This may be due to the small sample size. Thus, it can be inferred that the ovulatory and pregnancy rates are better in PCOS when metformin is a zdded to CC.

Table 5

Pregnancy rates

Drug regimen	Total no. of patients	No. of pregnancies	Percentage	P-value*
CC alone (control group)	24	2	8	0.194
CC and metformin (study group)	16	4	25
CC failures (subgroup)	−9	−2	−22.2	0.291

P-value <0.05 is considered significant

Side effects

Of the 25 subjects who received metformin along with CC, 80% complained of loss of appetite and 24% had nausea and vomiting, but none of them discontinued therapy. There were no cases of hyperstimulation either in the controls or in the study group and subgroup.

DISCUSSION

Anovulation or oligoovulation is one of the most important diagnostic criteria for PCOS.[5] Franks,[1] in his study, found 75% to have had oligomenorrhea and the rest amenorrhea. In the present study, 56.3% in the study group and 67% in the control group presented with oligomenorrhea (< 6 cycles/year). The occurrence of hirsutism of 85% in the present study is almost similar to that of Adams[6] study. We encountered non-obese PCOS more than obese, i.e. only 37.5% of women had a body mass index of >25. In the study by Dunaif,[7] 65% of PCOS were obese.

Ovulatory rate with CC alone

Gyseler[8] reported that 50% of the patients had ovulated with 50 mg CC and 74% with 100 mg, whereas Batnkan[9] reported only a 28.2% ovulatory rate with 100 mg. We have found a 33.3% ovulatory rate with 50 mg and 25% with 100 mg, which is in agreement with that of Batukan's. The maximum ovulatory rate of 52% was attained with 150 mg/ three cycles in the present study, whereas Lobo[10] reported only 32% with a dose of 200 mg.

The anovulatory rate at a dose of 150 mg in the study of Branigan et al.[11] was 28% and 20% in Lobo's[10] study. The present study showed a 48% anovulatory rate with 150 mg/ four cycles. Batukan[8] reported a rate as high as 72% anovulation with 100 mg CC in three cycles.

The pregnancy rate with CC when used alone was 8% in the present study, which is high when compared with 4% reported by Batukan,[9] but less than that reported by Fleming (17.4%).[12]

Ovulatory rate with the CC and metformin combination

In 1998, Nestler[13] found a 90% ovulatory rate with the CC and metformin combination when compared with 8% with CC alone. In the present study, during the first 2 months of metformin and CC combination, the ovulatory rate was less than 60% but, by the end of 6 months, 75–85% were ovulating, i.e. the overall ovulatory rate was 70% for 6 months when compared with 39% with CC. Batukan[9] reported the ovulatory rate to be only 57.9% whereas Vandermolen[5] reported it to be 75%, which is in agreement with the present study.

The pregnancy rate with this combination was as high as 56%, as reported by Malkawi,[14] and as low as 11%, in a review by Kocak.[12] A pregnancy rate of 55% was found in the study of Vondermolen,[15] which was high when compared with our study (24%). Costello's[16] review showed a pregnancy rate of 35%.

Average duration of treatment for conception

The mean duration with the CC and metformin combination for conception is 3.8 months in the present study, which is in agreement with that of Batukan[9] (4 months time). With CC alone, the mean duration is 6 months in the present study.

Pregnancy outcome

There were no twin pregnancies encountered in the present study with CC alone, whereas Gysler[8] reported seven twin pregnancies out of 193 patients who received only CC. We encountered one twin gestation in the CC and metformin combination among 25 patients who received the same. Batukan[9] reported one twin gestation out of 16 PCOS who received CC and metformin.

Gysler[8] reported a 12.9% abortion rate with CC alone. Batukan[9] found an abortion rate of 20% and Palomba[17] 15% when CC and metformin was used for ovulation induction. We have not encountered any loss of early gestation and there were no anomalies encountered. In 2002, Jakubowicz,[18] in a retrospective study, concluded that metformin administration reduced first trimester pregnancy loss in women with PCOS. Glucek,[19] in the same year, conducted a prospective study on 72 women with PCOS who conceived on metformin and reported 17% of first trimester losses as compared with 62% first trimester losses when metformin was not continued. No congenital anomalies were reported in either of these studies.

CC failures

Various authors report on failure to ovulate at a particular dose/cycles[111520] or failure to ovulate and conceive as CC resistance. Vandermolen (2001) defined CC resistance as failure to ovulate while receiving 150 mg of CC. In this study, failure to conceive despite ovulation is taken as CC failures.

We had the opportunity to treat nine of the CC failures of our control group with the CC and metformin combination. The ovulatory rate increased from 29% to 72% on addition of metformin in these cases. This is almost similar to the study group, which received CC + metformin initially itself. The pregnancy rate in these CC failure cases is also similar to that of the study group. Batukan[9] in 2002 added metformin to patients who did not conceive on CC and found that the pregnancy rate improved from 4.2% to 65.2%.

Vandermolen studied the effect of CC and metformin in 12 CC-resistant PCOS and compared it with placebo and CC in 15 CC-resistant PCOS. The ovulatory rate with the metformin and CC combination was 75% and the pregnancy rate was 55% as against the 27% ovulatory rate and the 7% pregnancy rate in the CC and placebo group.[15]

CONCLUSIONS

The ovulatory rate with the metformin and CC combination was found to be significantly higher when compared with CC given alone (P = 0.0016) in infertile PCOS subjects.

The ovulatory rate was found to be increasing when CC of 150 mg was given in combination with metformin for three cycles.

The pregnancy rate was high with the CC and metformin combination when compared with CC alone and even in CC-resistance PCOS. Hence, this combination should be initiated early in the course of treatment.

Table 4

Characteristics of PCOS

Characteristic	Controls (%)	Cases (%)
I. Age in years:
20–25	12 (50)	6 (37.5)
26–30	10 (41.7)	7 (43.8)
>31	2 (8.3)	3 (18.8)
II. Clinical features:
Oligomennorhea	16 (67.0)	9 (56.3)
Amenorrhea	8 (33.0)	7 (43.8)
Hirsutism	18 (75.0)	16 (100)
Acne	9 (37.5)	3 (18.8)
BMI < 25	15 (62.5)	10 (62.5)
BMI > 25	9 (37.5)	6 (37.5)
Hypothyroidism	1 (4.0)	2 (12.5)
Diabetes	1 (4.0)	-