increasing. Regardless of criteria used for the diagnosis, the incidence of GDM is increasing, which affects 2-18% of all pregnancies worldwide.^(6-8) The detection of GDM during pregnancy provides an opportunity to identify women at risk of short-term and long-term complications. Some argue that pregnancy related hyperglycaemia might be completely physiological to provide nutrient to the foetus and whether there is a need to diagnose and treat GDM. It has been shown by Crowther et al and others that diagnosis and subsequent treatment are beneficial.^(9-11) The aim of treatment is to maintain maternal blood glucose concentration within an acceptable range in a normal pregnancy. Interventions include lifestyle modification, Oral Hypoglycaemic Agents (OHAs), and insulin.

It is estimated that 70-90% of women diagnosed with GDM could achieve targeted glycaemic goals with lifestyle modification and nutrition therapy alone,^(12,13) but rest of the women often require additional treatment, which has conventionally been insulin.⁽¹⁴⁾ The disadvantages of insulin for the mother include the need to give injections, risks of hypoglycaemia, and increase in appetite and weight.⁽¹⁵⁾ Women maybe anxious about being on insulin and treatment compliance is an issue. It would be useful if there were alternative treatment options to insulin, preferably oral agents. Metformin, an oral biguanide, maybe a logical alternative to insulin for women with GDM who are unable to cope with the increasing insulin resistance of pregnancy. Metformin works primarily by decreasing hepatic glucose output, improving peripheral glucose uptake, and decreasing free fatty acid levels thus reducing insulin resistance.^(16,17) Outside pregnancy, metformin is as efficacious as insulin or a sulfonylurea in achieving glycaemic control in people with newly diagnosed type 2 diabetes and it is not associated with weight gain.⁽¹⁵⁾ Metformin crosses the placenta;^(18-20) however, there is no evidence of adverse foetal effect and it is a class B drug in pregnancy.⁽²¹⁾

There are data from over 35 years ago reporting use of metformin in women with GDM or type 2 diabetes in pregnancy in South Africa. Published cohort studies showed similar perinatal mortality and morbidity for women treated with metformin compared with insulin.^(22-24) Additional data regarding use of metformin in pregnancy comes from studies of women with polycystic ovarian syndrome, also associated with insulin resistance. Metformin has been shown to increase ovulation rates and fertility^(25-27) and continuing it through pregnancy may reduce risks of miscarriage^(28,29) and GDM. Rates of preeclampsia and perinatal loss were not increased in women taking metformin compared with the community delivery population.⁽³⁰⁾ The lack of prospective randomised data, however, creates uncertainty about the use of metformin in women with diabetes in pregnancy. In Kolkata, we carried out a prospective randomised single centre open label trial comparing metformin with insulin treatment in women with GDM. The aim of the trial is to test the hypothesis that in women with GDM. Metformin treatment compared with insulin will result in similar perinatal outcomes and be associated with improved treatment-related satisfaction.

METHODS

Ethics and Trial Design

The trial was a randomised 1:1 parallel group open labelled single centre active controlled study with an intervention period from randomisation (24 to 33 weeks of gestation) to delivery. Institutional ethics committee approved study protocol. The trial was registered in ctri.nic.in, CTRI/2014/08/004835. The study was conducted between February 2014 and July 2015.

The study design included a screening visit (24 to 32 weeks of gestation), a randomisation visit, and subsequently patients treated with metformin or insulin had several appointments with a diabetologist and an obstetrician as necessary.

Participants

Subjects were recruited from outpatient departments. Eligible participants for the study were women aged ≥18 and ≤35 years with singleton pregnancy at 24 to 34 weeks of gestation. Those who were thought to meet eligibility criteria were invited to a screening visit. Written informed consent was obtained from all participants at the randomisation visit. All the participants were counseled about diet modifications and exercises and were instructed to maintain habits during study period.

Inclusion criteria were at least one abnormal plasma glucose value at screening OGTT as per IADPSG criteria who failed to achieve glycaemic targets with trial of diet modifications and exercises for 1 week.

Exclusion Criteria included the following

1. Diagnosis of diabetes before 24 weeks of pregnancy.
2. Pre-pregnancy diagnosis of diabetes.
3. HbA1c >7.0%.
4. Renal, hepatic, or cardiac failure.
5. Diabetic ketoacidosis.
6. Pregnancy with multiple foetuses.
7. Fetal growth restriction.
8. Ruptured membranes.
9. A fetal anomaly.
10. Bad obstetric history.
11. Preeclampsia and eclampsia.
12. Chronic hypertension.
13. Other major comorbid illness.
14. Placental abnormalities or low lying placenta.
15. Oligohydramnios or polyhydramnios.
16. Gestational hypertension.
17. Unwilling to participate in the study.

Interventions

Study participants were randomly assigned to receive 1) Metformin (500 mg to 2000 mg in divided dosages as per glycaemic status. Patients also received supplemental insulin if required. 2) Insulin (Regular, neutral protamin hagedorn or both) was started at randomisation visit according to fasting and 2-hour postprandial plasma glucose values. Patients were frequently followed up as deemed suitable and monitored for glycaemic targets, any drug-related adverse events, hypoglycaemia, adherence to diet modifications along with foetal and maternal wellbeing. Dosages of metformin or insulin were titrated to achieve glycaemic targets, fasting plasma glucose ≤95 mg/dL and 2-hour postprandial plasma glucose ≤120 mg/dL. Additional insulin supplement was given in metformin group if glycaemic targets were not achieved with lifestyle modification and maximum metformin dosage of 2 grams per day.

Outcomes

Primary endpoint was to test whether there was a difference between neonatal birth weights in the metformin and insulin groups. Secondary endpoints were gestational hypertension, maternal weight gain, preeclampsia, mode of delivery, gestational weeks at delivery, induction of delivery, shoulder dystocia, neonatal hyperbilirubinaemia, neonatal hypoglycaemia, Apgar score at 5 min after delivery, neonate transfer to intensive care unit, respiratory distress, preterm birth (<37 wk. of gestation) and satisfaction with the allotted treatment modality.

Assessments

All assessments were performed at the randomisation visit and at the study end visits if not stated otherwise.

OGTT

A sample for fasting plasma glucose was obtained. Glucose (75 gm) in water was then given orally and plasma glucose was obtained at 60 and 120 minutes after glucose administration. The baseline investigation was performed at screening visit.

Laboratory

Glucose was measured by glucose oxidase method. HbA_1C was analysed by high performance liquid chromatography. Baseline biochemistry was performed at screening visit.

Other Assessments

A standard electronic scale was used to measure weight. Foetal ultrasound was carried out in Department of Radiology by experienced radiologist for all the subjects at baseline and thereafter as required. Subjects were evaluated for both maternal and fetal wellbeing by obstetrician at randomisation visit and thereafter as per obstetrician advice.

Safety

Adverse events were collected and recorded at all contacts with subjects after randomisation visit until the last contact, i.e. 7 days after delivery. A clinical examination was performed at the randomisation and followup visits. All biochemistry needed for assessment of eligibility was remeasured at delivery. Self-monitored blood glucose values and blood pressure were obtained at each visit.

Randomisation

A simple random sampling was performed creating a list including numbers 1-76 and intervention/active control treatment in a 1:1 ratio. An independent statistician created the list by computer software. Study subjects were allocated a randomisation number consecutively at the randomisation visit. They were then provided the corresponding medication. No blocking or stratification was performed.

Blinding

Study participants, study personnel, and investigators were not blinded to the allocated study intervention. All baseline data collection and its recording in the database were made before the randomisation code was broken.

Statistics

Descriptive statistical analysis has been carried out in the present study. Data is presented as means with SDs for continuous measurements as medians with IQRs or counts with percentages for categorical measurements. Significance is assessed at 5% level of significance. Normality of data tested by Shapiro-Wilk test. Unpaired t-test or Wilcoxon Mann-Whitney test has been used to find the significance of study parameters between two groups of patients according to the distribution of data. Chi-square/Fisher exact test has been used to find the significance of study parameters on categorical scale between two groups.

RESULTS

Participant flow

Participant flow is shown in flow chart (Figure 1). In study metformin arm (n=38), 3 patients withdrew consent, 2 patients lost to follow up, 1 patient violated protocol, and one patient discontinued metformin and continued on insulin treatment to maintain normoglycaemia because of gastrointestinal side effects. In the insulin group (n=38), 2 patients withdrew consent, 1 patient lost to follow up and 2 patients had violation of protocol. For the analyses, all subjects were referred to their original treatment groups.

Baseline data

The baseline maternal characteristics are seen in Table I. In this study, there were no differences in baseline maternal characteristics between the metformin and insulin groups.

Outcomes

As shown in Table II, there were no statistically significant differences in maternal data between metformin and insulin groups in this study except for postprandial plasma glucose and treatment satisfaction.

Neonatal data are presented in Table III. In this study, there were no statistically significant differences in birth weight (grams). There were no other significant differences in terms of neonatal complications between the groups. In study, there was one stillbirth at gestational week 38 in the insulin group. The likely cause of which was an umbilical cord complication. Data of this pregnancy were included in the results of maternal and neonatal data until birth.

Need for additional insulin among patients on metformin

In this study, 25% of the patients in the metformin group (8/32) needed additional insulin to maintain adequate glucose control. Maternal baseline characteristics such as age, parity, OGTT at randomisation, and HbA_1C did not differ significantly in the metformin subgroups (Metformin only and metformin with additional insulin).