Safety and feasibiLIty of Metformin in patients with Impaired glucose Tolerance and a recent TIA or minor ischemic stroke (LIMIT) trial – a multicenter, randomized, open-label phase II trial

Authors

  • Heleen M. den Hertog,

    Corresponding author
    1. Department of Neurology, Erasmus MC University Medical Center, Rotterdam, The Netherlands
    • Correspondence: Heleen M den Hertog, Department of Neurology, Erasmus MC University Medical Center, PO Box 2040, 3000 CA Rotterdam, The Netherlands.

      E-mail: m.denhertog@erasmusmc.nl

    Search for more papers by this author
  • S. E. Vermeer,

    1. Department of Neurology, Rijnstate Hospital, Arnhem, The Netherlands
    Search for more papers by this author
  • A. A. M. Zandbergen,

    1. Department of Internal Medicine, Ikazia Hospital, Rotterdam, The Netherlands
    Search for more papers by this author
  • Sefanja Achterberg,

    1. Department of Neurology, Utrecht Stroke Center, Rudolf Magnus Institute of Neuroscience, University Medical Center Utrecht, Utrecht, The Netherlands
    Search for more papers by this author
  • Diederik W. J. Dippel,

    1. Department of Neurology, Erasmus MC University Medical Center, Rotterdam, The Netherlands
    Search for more papers by this author
  • Ale Algra,

    1. Department of Neurology, Utrecht Stroke Center, Rudolf Magnus Institute of Neuroscience, University Medical Center Utrecht, Utrecht, The Netherlands
    2. Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, The Netherlands
    Search for more papers by this author
  • L. J. Kappelle,

    1. Department of Neurology, Utrecht Stroke Center, Rudolf Magnus Institute of Neuroscience, University Medical Center Utrecht, Utrecht, The Netherlands
    Search for more papers by this author
  • Peter J. Koudstaal

    1. Department of Neurology, Erasmus MC University Medical Center, Rotterdam, The Netherlands
    Search for more papers by this author

  • Conflicts of interest: The authors declare no conflict of interest.

Abstract

Background and purpose

We aimed to assess the safety, feasibility, and effects on glucose metabolism of treatment with metformin in patients with TIA or minor ischemic stroke and impaired glucose tolerance.

Methods

We performed a multicenter, randomized, controlled, open-label phase II trial with blinded outcome assessment. Patients with TIA or minor ischemic stroke in the previous six months and impaired glucose tolerance (2-hour post-load glucose levels of 7.8–11.0 mmol/l) were randomized to metformin, in a daily dose of 2 g, or no metformin, for three months. Primary outcome measures were safety and feasibility of metformin, and the adjusted difference in 2-hour post-load glucose levels at three months. This trial is registered as an International Standard Randomized Controlled Trial Number 54960762.

Results

Forty patients were enrolled; 19 patients were randomly assigned metformin. Nine patients in the metformin group had side effects, mostly gastrointestinal, leading to permanent discontinuation in four patients after 3–10 weeks. Treatment with metformin was associated with a significant reduction in 2-hour post-load glucose levels of 0·97 mmol/l (95% CI 0·11–1·83) in the on-treatment analysis, but not in the intention-to-treat analysis (0·71 mmol/l; 95% CI −0·36 to 1·78).

Conclusions

Treatment with metformin in patients with TIA or minor ischemic stroke and impaired glucose tolerance is safe, but leads to minor side effects. If tolerated, it may lead to a significant reduction in post-load glucose levels. This suggests that the role of metformin as potential therapeutic agent for secondary stroke prevention should be further explored.

Introduction

Impaired glucose tolerance, an intermediate metabolic state between normal glucose tolerance and diabetes mellitus, is present in about one-third of patients with transient ischemic attack (TIA) or ischemic stroke [1-3] and is associated with a twofold risk of recurrent stroke [4-6]. The mechanisms underlying this association are not fully understood but may involve insulin resistance, endothelial dysfunction, dyslipidemia, chronic inflammation, procoagulability, and impaired fibrinolysis [7-9].

Pharmacological interventions reduce the rate of progression to type 2 diabetes by 10–60% in people with impaired glucose tolerance [10-13]. Lifestyle interventions seem to be at least as effective as drug treatment [10-13], but these are often difficult to carry out successfully, and life style advice needs to be reinforced on a regular basis.

The widely used oral glucose-lowering drug metformin is a biguanide that improves insulin sensitivity, enhances peripheral uptake of glucose, and decreases hepatic glucose output. It is recommended as first-line treatment in diabetes and is cheap as compared with the newer antidiabetic drugs. Metformin prevents the onset of diabetes in people with impaired glucose tolerance, with a number needed to treat of 12 [12, 13]. Furthermore, it has been shown to decrease (macro)vascular complications and to improve survival in type 2 diabetic patients [14, 15]. Whether pharmacological glucose control with metformin will reduce the risk of stroke and other cardiovascular events in patients with TIA or ischemic stroke and impaired glucose tolerance is not known. Here, we report a phase II trial aimed to assess the safety, tolerability, and feasibility of treatment with metformin in patients with TIA or minor ischemic stroke and impaired glucose tolerance, and the effect of this treatment on two-hour postload glucose levels after three-months.

Methods

LIMIT was a randomized, controlled, open-label phase II trial with blinded outcome assessment. Patients were recruited at two centers, the Erasmus University Medical Center, Rotterdam, and the University Medical Center Utrecht, Utrecht.

Participants

Patients were eligible for inclusion if they were 18 years or older; had a clinical diagnosis of TIA, amaurosis fugax, or minor ischemic stroke [defined as a modified Rankin scale (mRS) [16] score of 3 or less] within the previous six-months; and had impaired glucose tolerance (defined as two-hour postload glucose levels between 7·8 and 11·0 mmol/l) [17].

Patients were excluded if they were known to have diabetes mellitus, renal impairment (creatinin of 135 μmol/l or higher for men and 110 μmol/l or higher for women), known liver disease or disturbed liver function tests (alanine amino transferase, aspartate amino transferase, alkaline phosphatase, or γ glutamyl transferase increased to more than twice the upper limit of typical values), a history of lactic acidosis, heart failure requiring pharmacological therapy, chronic hypoxic lung disease, or in case of pregnancy or breast feeding.

The trial was approved by national and local institutional review boards, and written informed consent was obtained from all patients or their legal representatives.

Procedures

Patients were randomized to receive either open-label metformin or ‘no metformin’ for three-months. Patients allocated to metformin started with 500 mg once daily, which was slowly increased in one-month time to a daily dose of two times 1000 mg.

We randomized patients by means of sealed envelopes. Treatment allocation was based on a list, generated by computer before the start of the trial. An independent statistician, who was otherwise not involved in the study, provided the list. Irrespective of treatment allocation, patients received optimal standard care, including antithrombotic and antihypertensive agents and cholesterol lowering drugs, where appropriate [18]. In addition, a stroke nurse specialist provided general lifestyle advice including healthy diet, advice against smoking, and regular physical exercise.

At baseline, we recorded clinical features of the longest episode of focal neurological deficits in the preceding six-months, demographic data, disability score on the mRS, treatment at the time of event, vascular risk factors, and vascular history.

Follow-up visits were scheduled at one-month and three-months after inclusion. At both visits, patients completed a single questionnaire to determine the compliance and nature of any of the side-effects of the medication. We also contacted patients by telephone for recording of possible adverse events at two-weeks after inclusion. All patients underwent an oral glucose tolerance test with 75-g glucose before inclusion into the study and at three-months. Fasting glucose levels, body mass index (BMI), waist circumference, blood pressure, and lipid profile were assessed at baseline as well as at three-months.

Primary and secondary outcomes

Primary outcomes were the tolerability of metformin (assessed as number of patients still on treatment after three-months), the safety of treatment with metformin, and the adjusted difference in two-hour postload glucose levels at three-months. Secondary outcomes included adjusted differences in fasting glucose levels, BMI, waist circumference, and percentage of patients with a normal glucose tolerance at three-months.

Safety concerns and adverse events

We continuously monitored the incidence of serious adverse events by frequently checking patients and by reports from general practitioners. Serious adverse events included lactic acidosis and hypoglycemia. If a serious adverse event occurred in patients on metformin, treatment was stopped immediately.

Statistical analysis

Analyses were by intention to treat and all patients who were randomly assigned treatment were included in the prespecified analyses.

A sample of 40 was expected to provide a power of 90% to detect a difference of 10% in two-hour postload glucose levels after three-months between the two treatment groups, assuming a significance level of 0·05 and a mean glucose level of 8·9 mmol/l in the control group with a standard deviation of 0·9 mmol/l.

For all efficacy end points, baseline adjusted differences and corresponding 95% confidence intervals (CIs) between the two treatment groups were calculated with linear regression. Adjustments were made with a multivariable linear regression model that included the following factors: age, sex, time to treatment, and baseline waist.

We also did an analysis of patients who completed the three-months of treatment (on-treatment analysis), in which we included only two-hour postload glucose levels of those who used metformin in a dose of 2000 mg.

Results

Forty patients were enrolled between August 2007 and August 2008 (Fig. 1). Nineteen were randomly assigned to metformin and 21 to the control group. No patient was lost to follow-up.

Figure 1.

Enrollment and follow-up. TIA, transient ischemic attack.

The baseline characteristics of the enrolled patients are described in Table 1. There were no significant differences between the two groups. In the total study population, mean age was 62 years (standard deviation 11), 21 (53%) were men, and 18 (45%) had suffered a TIA. The main concomitant treatments at inclusion were antiplatelet drugs (88%), oral anticoagulation (12%), lipid-lowering agents (88%), angiotensin converting enzyme inhibitors (15%), and diuretics (30%). Patients were enrolled at a median of 27 days (range 1–178) after their TIA or ischemic stroke. Both groups were well matched for two-hour postload glucose levels as well as for fasting glucose levels.

Table 1. Baseline characteristics
 Metformin (n = 19)No metformin (n = 21)
  1. BMI, body mass index; SD, standard deviation.
Demographics  
Age, mean (SD)62 (12)63 (9)
Gender (male)9 (47%)12 (57%)
Qualifying event  
Transient monocular blindness01 (5%)
Transient ischemic attack10 (53%)8 (38%)
Ischemic stroke9 (47%)12 (57%)
Time from event to randomization (days), median (range)28 (5–173)27 (1–178)
Additional investigations  
Carotid stenosis ≥50%1 (5%)1 (5%)
History  
Previous stroke5 (26%)5 (24%)
Coronary artery disease1 (5%)3 (14%)
Peripheral artery disease1 (5%)1 (5%)
Risk factors  
Hypertension9 (47%)12 (57%)
Atrial fibrillation1 (5%)0
Hypercholesterolemia9 (47%)8 (38%)
Current cigarette smoking10 (53%)10 (48%)
Antiplatelet drugs16 (84%)19 (90%)
Oral anticoagulation3 (16%)2 (10%)
Lipid lowering agents17 (89%)18 (86%)
Antihypertensive drugs13 (68%)14 (67%)
Angiotensin converting enzyme inhibitors2 (15%)4 (29%)
Diuretics5 (38%)7 (50%)
Physical examination  
Systolic blood pressure, mean (SD)130 (21)131 (17)
Diastolic blood pressure, mean (SD)77 (11)77 (9)
BMI (kg/m2), mean (SD)27 (3)27 (4)
Waist (cm), mean (SD)100 (10)91 (13)
Laboratory investigations  
Fasting glucose (mmol/l), mean (SD)5·2 (0·6)5·2 (0·8)
Two-hour postload glucose (mmol/l), mean (SD)9·3 (1·0)9·3 (1·0)

Nine patients in the metformin group (50%, 95% CI 24–75%) had one or more (mostly gastrointestinal) side-effects compared with three (14%, 95% CI 3–36%) in the control group, leading to permanent discontinuation in four (22%) of them after 3–10 weeks. In three patients, metformin was decreased to a daily dose of 1000 mg. One patient died of intracerebral haemorrhage in the metformin group; no patients died in the control group (Table 2). No hypoglycemia or lactic acidosis was reported in both treatment groups.

Table 2. Serious adverse events and side-effects
 Metformin (n = 19)Relationship of event to treatmentNo metformin (n = 21)
  1. NA, not applicable.
Serious adverse events   
Death (intracerebral haemorrhage)1 (5%)Not related0
Lactic acidosis0NA0
Hypoglycemia0NA0
Side-effects   
Diarrhea3 (16%)Related0
Nausea4 (21%)Related0
Anorexia2 (11%)Related0
Other side-effects   
Headache2 (11%)Unlikely2 (11%)
Dizziness1 (5%)Unlikely0
Tiredness1 (5%)Unlikely1 (5%)

The intention-to-treat analysis showed a nonsignificant reduction in two-hour postload glucose levels in the metformin group compared with the controls (0·71 mmol/l, 95% CI −0·36 to 1·78) (Fig. 2). The on-treatment analysis showed a statistically significant reduction in two-hour postload glucose levels in patients on metformin (0·97 mmol/l, 95% CI 0·11–1·83).

Figure 2.

(a) and (b) Changes in two-hour postload glucose levels and fasting glucose levels from baseline (mean ± standard deviation) in the intention-to-treat analysis.

Patients on metformin had lower fasting glucose levels at three-months from the start of treatment than the control group (0·43 mmol/l, 95% CI −0·13 to 1·03) (Fig. 2), but this was not statistically significant.

There was a nonsignificant trend to a reduction in BMI (0·65 kg/m2, 95% CI −0·25 to 1·54) in the metformin group. No differences were found in waist circumference between treatment groups (0·32 cm, 95% CI −3·20 to 3·88). No significant changes were observed for blood pressure, cholesterol, or triglycerides in the course of the study.

At three-months, nine of the patients on metformin reverted to normal glucose tolerance (50%) compared with 10 (48%) in the control group.

Discussion

We found that treatment with metformin in patients with recent TIA or ischemic stroke and impaired glucose tolerance is safe but often leads to side-effects. Metformin significantly reduced two-hour postload glucose levels in the on-treatment analysis.

Another phase II trial has found that pioglitazone, a thiazolidinedione, was effective in improving insulin sensitivity in patients with recent TIA or ischemic stroke and impaired glucose tolerance [19]. However, the potential cardiovascular benefits of pioglitazone must be considered against the increased risk of fluid retention, heart failure, loss of bone mineral density, and bone fractures [20-23].

Potential limitations of our study are the lack of a placebo group and its open design. However, outcome assessment was blinded for treatment allocation, and this design assessed real-life treatment strategies.

The incidence of side-effects of treatment with metformin was higher in our study than reported in previous studies in individuals with impaired glucose tolerance and no vascular history [12-15]. However, patients with TIA or ischemic stroke are on average older, and the majority uses other drugs such as statins, antihypertensives, and antithrombotic drugs, which may increase the risk of side-effects. Slower increase in dose of metformin might reduce the incidence of side-effects in these patients. Moreover, as most side-effects disappear after six-weeks of treatment, better information and support on this might have resulted into less discontinuation of metformin treatment.

We did not observe significant effects of metformin on fasting glucose levels, BMI, waist circumference, blood pressure, or serum lipids. Other studies in patients with manifest diabetes mellitus suggest that metformin treatment is associated with a decrease in fasting glucose levels, blood pressure, body weight, and triglycerides and an increase in high-density lipoprotein cholesterol levels [13-15]. These effects, however, are small compared with the effect on two-hour postload glucose levels and might not have been detected in our study because of the small sample size and short follow-up.

There is a growing recognition that patients with impaired glucose tolerance should be treated more aggressively, as these patients do not only have an increased risk of developing diabetes but also of cardiovascular events. Treatment for impaired glucose tolerance in patients with TIA or ischemic stroke might have a great impact, given its high prevalence and associated risk of recurrent stroke in these patients [1-6].

There is no clear evidence that tight glycemic control reduces the risk of stroke in patients with diabetes or impaired glucose tolerance. In the UK Prospective Diabetes Study, however, metformin therapy was associated with less cardiovascular events in newly diagnosed diabetics [14]. Furthermore, a recent large randomized placebo-controlled trial found that metformin reduces macrovascular complications when added to insulin treatment in type 2 diabetes [15]. A recent meta-analysis on glucose-lowering pharmacological interventions in patients with impaired glucose tolerance found no beneficial effects on all-cause mortality or death due to major cardiovascular events, with the possible exception of stroke [24]. None of the novel antidiabetic agents has been proven to reduce this risk of cardiovascular events [11, 25]. Lifestyle interventions might be as least as effective and lead to fewer and less serious side-effects than drug treatment [10, 13]. However, these interventions are often difficult to achieve, in particular, in patients with functional and cognitive impairment. Furthermore, their effect may not last and advice on smoking, diet, and exercise needs to be reinforced on a regular basis.

Our findings suggest that metformin treatment is safe in patients with TIA or ischemic stroke and impaired glucose tolerance and probably leads to improved glucose tolerance. Hence, metformin might be a suitable treatment in such patients. However, more information about strategies to reduce the incidence of side-effects is required before a phase III study on the efficacy and safety of metformin therapy in reducing the incidence of stroke and other vascular complications in patients with TIA or ischemic stroke and impaired glucose tolerance can be initiated.

Ancillary