Long versus short‐course treatment with metformin plus clomiphene citrate for ovulation induction in women with polycystic ovary syndrome
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
To determine the effectiveness of short‐course (less than weeks) metformin plus CC versus long‐course (four weeks or more) metformin plus CC with regard to ovulation and achievement of pregnancy in infertile women with PCOS.
Background
Polycystic ovary syndrome (PCOS) is the most common endocrinopathy among reproductive‐age women and affects approximately 5 to 10% of this population group (Hull 1987; Polson 1988). It is also the most common cause of anovulatory infertility.
Clomiphene citrate (CC) is currently the first‐line, most widely used oral medication to induce ovulation in women with PCOS (Kim 2000). However, only 70 to 85% of women with PCOS respond to clomiphene citrate, with a pregnancy rate of only 30 to 40% (Franks 1995; Lobo 1982). This could be attributed to the anti‐estrogenic effect of clomiphene citrate on cervical mucous and endometrium (Randall 1991; Nakamura 1997).
Apart from infertility women with PCOS often have other endocrine disorders, including a reported 50 to 100% of obese women with PCOS with insulin resistance and hyperinsulinemia and as high as 22% in women who were lean (Dale 1992). Hyperinsulinemia appears to lead to hyperandrogenism, which may adversely affect follicular development and ovulation (Barbieri 1986; Nestler 1998a). Metformin is a medication that has an insulin‐sensitizing effect. Metformin could also ameliorate hyperandrogenism in PCOS women (Nestler 1998b; Pirwany 1999) and thus possibly correct the endocrinopathy. Several studies have demonstrated that treatment with metformin before administration of clomiphene citrate in women with PCOS may significantly increase ovulation and pregnancy rates (Vandermolen 2001; Velazquez 1994). The most recent research synthesis revealed that metformin was 50% better than placebo for increasing ovulation in infertile PCOS patients and that metformin plus clomiphene citrate may be three to four‐fold superior to clomiphene citrate alone for producing ovulation and achievement of pregnancy (Kashyap 2004). The same study, however, showed that metformin alone had no confirmed benefit over placebo for achievement of pregnancy.
Although it has become clear that the combination of metformin and clomiphene citrate is more effective in pregnancy attainment than clomiphene citrate (CC) alone (Kashyap 2004; Lord 2004) the optimal duration of metformin pretreatment before CC administration in women with PCOS is unknown. Previous studies usually used 4 to 12 weeks of metformin before beginning clomiphene citrate (Vandermolen 2001; Velazquez 1994) but many women found such duration of metformin pretreatment inconvenient. Long‐term use of metformin may also be associated with several adverse effects such as lactic acidosis and gastrointestinal disturbances (Lord 2004). One study recently revealed that ultra‐short (12 days) metformin pretreatment before administration of CC significantly increased ovulation and pregnancy rates as compared to CC alone (Hwu 2005). It is useful, therefore, to determine whether short‐course (less than four weeks) metformin treatment in conjunction with CC is as effective as the conventional long‐course (at least four 4 weeks) metformin plus CC with regard to ovulation and achievement of pregnancy in infertile women diagnosed with polycystic ovary syndrome.
Objectives
To determine the effectiveness of short‐course (less than weeks) metformin plus CC versus long‐course (four weeks or more) metformin plus CC with regard to ovulation and achievement of pregnancy in infertile women with PCOS.
Methods
Criteria for considering studies for this review
Types of studies
All published, unpublished and ongoing randomized controlled trials comparing short‐course (less than four weeks) metformin plus CC versus long‐course (four weeks or more) metformin plus CC to achieve ovulation or pregnancy in infertile women with PCOS. Studies reported only in abstract form will be included in the 'studies awaiting assessment' category and will be included in the analysis when published as full reports.
Types of participants
Women of reproductive age (between 15 and 40 years) with anovulatory infertility attributed to PCOS.
Anovulation is defined as a lack of evidence of serum progesterone within the luteal range for the reference laboratory, menstrual cycles that are less frequent than every 35 days or fewer than six periods per year.
Infertility is defined as the inability to get pregnant after one year of unprotected sexual intercourse.
PCOS is defined according to the European Society of Human Reproduction and Embryology (ESHRE) and the American Society for Reproductive Medicine (ASRM) criteria (ESHRE/ASRM 2003). Two of the following three manifestations are required for diagnosis of PCOS: (1) oligo or anovulation (menstrual cycles less frequent than every 35 days or fewer than six periods per year); (2) clinical or biochemical signs of hyperandrogenism, or both (clinical hirsutism or acne; or biochemically elevated testosterone, dehydroepiandrosterone, or androstenedione levels); and (3) polycystic ovary (ultrasound scanning shows enlarged ovary with peripheral cystic structures surrounded by an increased stromal mass).
Exclusion criteria
(1) Women with hyperprolactinemia (greater than three times the upper limit of normal of the reporting laboratory's reference range), congenital adrenal hyperplasia (CAH) and Cushing's syndrome will be excluded since these conditions preclude the diagnosis of PCOS.
(2) Women diagnosed with hypogonadotropic hypogonadism (WHO group 1 anovulation) and ovarian failure (WHO group 3 anovulation) will also be excluded from this review.
The study and control groups should have no other infertility diagnosis and preferably couples would have a documented normal semen analysis.
Types of interventions
The comparison between short‐course (less than four weeks) metformin plus CC versus long‐course (four weeks or more) metformin plus CC will be looked for.
Types of outcome measures
Primary outcome
Pregnancy rate (per woman). Pregnancy will be defined as positive urinary human choriotrophic gonadotrophin CHECK(hCG) or serum B‐hCG or evidence of pregnancy by ultrasound evaluation.
Secondary outcomes
1. Ovulation rate (per woman). Ovulation will be defined as a midluteal phase serum progesterone level greater than 3 ng/mL or in the luteal range for the reference laboratory or evidence of ovulation documented by ultrasound evaluation.
2. Miscarriage rate (per pregnancy). Miscarriage will be defined as the involuntary loss of pregnancy before 20 weeks of gestation.
3. Incidence of adverse effects (per woman). This will be defined according to the definition of reporting authors.
Search methods for identification of studies
This review will follow the Cochrane Menstrual Disorders and Subfertility Group search strategy. The literature search will aim to locate randomized controlled trials reported in all languages.
(1) We will search the Cochrane Menstrual Disorders and Subfertility Group Trials Register.
(2) We will search the Cochrane Central Register of Controlled Trials (The Cochrane Library 2006, Issue 1) for the keywords: polycystic ovary syndrome (PCOS), metformin.
(3) We will search the listed electronic databases for studies in all languages using the following terms.
MEDLINE (1966 to April 2006)
1) randomized controlled trial.pt
2) controlled clinical trial.pt
3) randomized controlled trial.pt
4) random allocation
5) double‐blind method/
6) single‐blind method/
7) or/1‐6
8) clinical trial.pt
9) exp clinical trials/
10) (clin$ adj25 trial$).tw
11) ((singl$ or doubl$ or treb$ or tripl$) adj25 (blind$ or mask$)).tw
12) metformin/
13) metformin$.tw
14) random$.tw
15) research design/
16) or/8‐15
17) animal/ not (human/ and animal/)
18) 7 or 16
19) 18 not 17
20) polycystic adj5 ovar$.tw
21) PCOS.tw
22) PCO.tw
23) Polycystic Ovary Syndrome/
24) or/20‐23
25) anovul$.tw
26) infertil$.tw
27) subfertil$.tw
28) INFERTILITY/
29) or 25‐28
30) 24 and 19
31) 7 or 19
31) 30 and 31
EMBASE (1980 to April 2006)
1) randomized controlled trial/
2) random allocation/
3) double‐blind method/
4) single‐blind method/
5) or/1‐4
6) exp clinical trials/
7) (clin$ adj25 trial$).tw
8) ((singl$ or doubl$ or treb$ or trip$) adj25 (blind$ or mask$)).tw
9) metformin/
10) metformin$.tw
11) random$.tw
12) research design/
13) or/6‐12
14) animal/ not(human/ and animal/)
15) 5 or 13
16) 15 not 14
17) polycystic adj5 ovar$.tw
18) PCOS.tw
19) PCO.tw
20) Polycystic Ovary Syndrome
21) or/17‐20
22) anovu$.tw
23) infertil$.tw
24) subfertil$.tw
25) INFERTILITY
26) or/22‐25
27) 21 and 26
28) 5 or 16
29) 27 and 28
(4) We will check references of relevant reviews and randomized controlled trials (RCTs).
(5) Personal communication with manufacturers and content experts in the field will be conducted in an attempt to search grey literature.
Data collection and analysis
Selection of studies
This review will be undertaken by four review authors. The search strategy described previously will be employed to obtain titles and, where possible, abstracts of studies that are potentially relevant to the review. The titles and abstracts will be screened by the first review author who will discard studies that are clearly ineligible but the aim is to be overly inclusive rather than risk losing relevant studies at this stage.
The first review author will then obtain copies of the full text articles. After removing all information that could identify the authors, the publishers or the results of the study, the methods section will be sent to the first and the second review authors. Both authors will independently assess whether the studies meet the pre‐stated inclusion criteria. Any disagreement will be resolved by discussion and final arbitration by the third author. Further information will be sought from the study authors where papers contain insufficient information to make a decision about eligibility. The fourth review author will be responsible for planning of analysis, data analysis, data interpretation and data presentation.
Quality assessment
The quality of all studies that are deemed eligible for the review will then be assessed independently by two review authors with disagreement resolved by discussion or, if necessary, by the third review author. The quality of allocation concealment will be graded as adequate (A), unclear (B), or inadequate (C) following the detailed description of these categories provided by the Cochrane Menstrual Disorders and Subfertility Review Group, as follows.
1. Assessment of methodological quali ty
Each item will be rated
Clearly yes: rate A
Not sure: rate B (seek details from authors)*
Clearly no: rate C
Section i: internal validity
1) Was the assigned treatment adequately concealed prior to allocation?
2) Were the outcomes of patients who withdrew or were excluded after allocation described and included in an 'intention‐to‐treat' analysis?
3) Were the outcomes assessors blind to assignment status?
4) Were the treatment and control groups comparable at entry?
5) Were the subjects blind to assignment status following allocation?
6) Were the treatment providers blind to assignment status?
7) Were the care programs, other than the trial options, identical?
8) Were the withdrawals < 10% of study population?
Section ii: external validity
9) Were the inclusion and exclusion criteria for entry clearly defined?
10) Were the outcome measures used clearly defined?
11) Were the accuracy, precision and observer variation of the outcome measures adequate?
12) Was the timing of the outcome measures appropriate?
* May want to adapt a decision rule such as 'if what was done is not clearly stated, assume that it is not done appropriately' or 'assign partial credit'.
2. Allocation score
Was the assigned treatment adequately concealed prior to allocation?
Clearly yes: score A
‐ Some form of centralized randomization scheme, such as having to provide participant details by phone to receive treatment group allocation
‐ A scheme controlled by a pharmacy
‐ In a pharmaceutical study, sequential administration of pre‐numbered or coded containers to enrolled participants
‐ An on‐site computer system given that allocations are in a locked unreadable file which can be accessed only after inputting participant details
‐ Assignment envelopes provided that they are sequentially, sealed and opaque
‐ Other combinations which appear to provide assurance of adequate concealment
Unclear: score B
‐ Assignment envelopes, without description of adequate safeguards
‐ Use of a 'list' or 'table'
‐ Flip of a coin
‐ A trial in which the description suggests adequate concealment but other features are suspicious, for example markedly unequal controls and trial groups
‐ Stated as random but unable to obtain further details
Clearly no: score C
‐ Alteration
‐ Case record numbers, dates of birth, day of week or any other such approach
‐ Any allocation procedure transparent before assignment, such as an open list random numbers
It is intended that this allocation score grading be used in investigation of any heterogeneity and in sensitivity analysis. Other aspects of study quality including the extent of blinding (if appropriate), whether groups were comparable at baseline, the extent of loses to follow up, non‐compliance, whether the outcome assessment was standardized and whether an intention‐to‐treat analysis was undertaken will provide a context for discussing the reliability of the results.
Data extraction
The following information is to be extracted from the studies included in the review
General information
(a) Title
(b) Publication status
(c) Authors
(d) Contact address
(e) Country
(f) Resource
(g) Publication year
(h) Publication language
(i) Duplication of publishing
Trial Characteristics
(a) Randomization
(b) Allocation concealment
(c) Trial design: multi‐center or single center; single phase or crossover design
(d) Blinding
(e) Number of patients randomized, excluded and analyzed
(f) Source of funding
Baseline characteristics of the studied groups
(a) Definition and duration of pre‐existing infertility
(b) Age of the patients
(c) Body mass index (BMI) of the patients
(d) Investigative work up
(e) Other causes of infertility
(f) Previous administered treatment(s)
Intervention
(a) Type of intervention
(b) Duration of treatment with metformin
(c) Dose regimen
Outcomes
(a) Outcomes reported
(b) How are outcomes defined?
(c) How are outcomes measured?
(d) Timing of outcome measurement?
All data will be extracted independently by two of the authors using data extraction forms designed according to Cochrane guidelines. The form designed will be pilot testing with a sample of the studies to ensure that it is understandable, easy to complete and comprehensive. Additional information will be sought on trial methodology, actual trial data or both from the authors of the trials which appear to meet eligible criteria but have aspects of methodology that are unclear or data in an unsuitable form for meta‐analysis. Differences of opinion between the two review authors will be resolved by the third review author. Reasons for excluding any trial will be provided.
Statistical analysis
Data will be entered onto Review Manager and checked for accuracy by performing double data entry. Data analysis and graphical displays will be facilitated using RevMan 4.2 software distributed by The Cochrane Collaboration.
All dichotomous outcomes and results will be expressed as odds ratio with 95% confidence intervals. In order to perform meta‐analysis using dichotomous data, we will extract the number in each of the two categories in each of the intervention groups (the numbers needed to be filled in the 2 x 2 table).
The data extracted from different trials will then be assessed for heterogeneity by using several methods as follows.
1. Inspection of individual 95% confidence interval (CI) in the forest plots.
2. Using the Cochrane Q statistic (which is approximate to the chi‐square test on K‐1 degrees of freedom, where K is the number of studies in the meta‐analysis). A P value of less than 0.10 is used to indicate significant heterogeneity.
3. Calculating the I2 statistic (Higgins 2003). I2 is the percentage of variability in effect estimate due to heterogeneity rather than sampling error. A value of I2 greater than 50% is usually taken to indicate heterogeneity.
If studies are clinically and statistically homogeneous, meta‐analysis will be conducted using a fixed‐effect model.
If significant heterogeneity is found, which means that the data are not comparable beyond chance effects, several strategies will be attempted in order to determine the factors that could be responsible for such heterogeneity. These strategies include pre‐specified subgroup analysis and meta‐regression. If the studies demonstrating heterogeneous results are found to be comparable, we will undertake statistical synthesis of the results using a random‐effect model.
Publication bias will be investigated by evaluating the funnel plots. Asymmetry of the funnel plots may indicate publication bias.
We are planning to perform subgroup analysis according to the body mass index (BMI) of study participants. The pre‐specified subgroups in this review include studies with participants who have a mean baseline BMI > 30 Kg/m2 and those with a mean baseline BMI < 30 Kg/m2.
Sensitivity analysis will also be performed in order to test the robustness of the review conclusions by taking into account key decisions and assumptions that were made in the process of conducting the review. These approaches include:
1. repeating the analysis with exclusion of the trials most susceptible to bias, based on the quality assessment (such as the trials with inadequate allocation concealment, high levels of postrandomization losses or exclusions);
2. repeating the analysis with exclusion of trials using the following filters: publication language and country.
Timeline
Completion of the review is expected within two years of publication of the protocol in The Cochrane Library.
