Chronic obstructive pulmonary disease (COPD) is the 4th leading cause of death worldwide and affects 1.2 million people in the UK, costing the NHS >£800 million annually. COPD patients are more susceptible to bacterial infections and both chronic and acute infections are common. COPD patients with chronic lung bacterial infection have worse quality of life, faster disease progression, more symptoms and frequent exacerbations. Acute infections are the main cause of COPD exacerbations which cause COPD patients to become acutely unwell and often result in hospitalisation especially in the winter. Antibiotics are frequently used to treat COPD exacerbations and this contributes to the development of antibiotic resistance. Therefore there is a need to develop antibiotic-independent approaches to reducing or preventing bacterial infection in COPD.

The investigators have carried out work in in animal studies and in humans showing that there is a link between high levels of glucose in the lung and bacterial lung infection. Levels of glucose in the lung are higher in COPD patients compared with people without COPD. These higher glucose levels support greater bacterial growth probably because glucose is a nutrient for bacteria. Therefore reducing airway glucose has the potential to inhibit bacterial growth in COPD patients.

In animal studies the investigators have demonstrated that the diabetic drug metformin decreases airway glucose and bacterial growth. The investigators wish to determine if metformin can achieve the same effects in COPD patients. Metformin is safe and cheap, and has been extensively used in COPD patients with diabetes with an excellent safety record. The primary aim of this study will be to determine whether metformin reduces lung glucose in a small group of non-diabetic COPD patients. If it demonstrates that metformin reduces lung glucose concentrations it will justify a larger clinical trial of metformin as a treatment for COPD.

Chronic obstructive pulmonary disease (COPD) is the 4th leading cause of death worldwide and affects 1.2 million people in the UK, costing the NHS >£800 million annually. COPD patients are more susceptible to both chronic and acute bacterial infections. Patients with chronic lung bacterial infection have worse quality of life, faster disease progression, more symptoms and frequent exacerbations. Acute infections are the main cause of acute COPD exacerbations which cause COPD patients to become acutely unwell and often result in hospitalisation especially in the winter. Bacteria are detected in 50-60% of COPD exacerbations. Antibiotics are frequently used to treat COPD exacerbations and this contributes to the development of antibiotic resistance. Therefore any intervention that prevents or reduces bacterial infection in COPD, especially if it is not an antibiotic, will have major benefits for COPD patients, the NHS and for society as a whole.

It is likely that there are many reasons why COPD patients are more susceptible to bacterial infections. From experimental work the investigators have carried out one of the reasons may be high glucose concentrations in the lung.

In healthy lungs glucose levels are kept low and this may be a mechanism that inhibits bacterial growth by depriving them of an essential nutrient. In animal studies the investigators have demonstrated that when levels of glucose in the lung are high, bacterial lung infection is more common. The investigators measured lung glucose concentrations in COPD patients and found that they are higher compared with people without COPD. COPD patients with higher levels of glucose also had more bacteria in their lungs and sputum samples from COPD patients with higher glucose concentrations supported greater bacterial growth in the laboratory.

Therefore this study was the first to link elevated glucose in the lung to bacterial infection in COPD. Therefore reducing airway glucose has the potential to inhibit bacterial growth in COPD patients.

Study Design The proposed study will be a randomised, double-blinded, placebo-controlled, cross-over study of metformin in COPD patients. The primary outcome will be sputum glucose after 3 months' treatment with metformin compared with sputum glucose in those taking placebo. In order to account for potential withdrawals 40 subjects will be recruited.

Study Procedures Potential participants will attend for a screening visit where they will have a full medical history, a physical examination and spirometry carried out to confirm the diagnosis of COPD. A blood test will also be done to measure kidney and liver function and blood glucose to exclude undiagnosed diabetes, kidney disease or liver disease.

If they fulfil the entry criteria and consent to taking part in the study they will have a baseline visit prior to being randomised.

The baseline visit will include:

1. Physical examination and measurement of vital signs
2. Completion of quality of life (St George's Respiratory Questionnaire (SGRQ)) and symptom questionnaires (COPD Assessment Test (CAT))
3. Collection of samples. The samples collected will include blood samples, nose samples collected using nasal synthetic absorption matrix (SAM) strips and induced sputum.

After baseline assessment, subjects will either be commenced on metformin (500mg twice a day after meals) or placebo for 3 months during which time the participants will have monthly visits. At these visits the same assessments and sampling as the baseline will be carried out, together with collection of data regarding exacerbations and adverse events. Following a 2 week washout period the subjects will crossover to the other study arm for another 3 months and follow the same study protocol.

Inclusion Criteria:

• Age between 40 and 75 years
• Clinical diagnosis of COPD confirmed with spirometry (Post-bronchodilator FEV1/FVC <70%).
• Smoking history >15 pack years
• Absence of infection >8 weeks prior to study entry
• No use of antibiotics and/or oral corticosteroids >8 weeks prior to study entry
• Able to understand and consent to the study procedures

Exclusion criteria:

• Diabetes including diabetes diagnosed at screening
• History of hepatic or renal impairment or diagnosed on screening bloods
• Patients already taking metformin irrespective of indication
• Known allergy or hypersensitivity to metformin
• Pregnancy or breastfeeding
• Any other significant medical condition likely to interfere with the study
• Unable to provide informed consent
• Excessive alcohol intake (>21 units/week)
• BMI < 18.5kg/m2

1 year

Adverse events were reviewed at every study visit.