Renal failure and use of metformin in people with type 2 diabetes: An audit

How metformin causes lactic acidosis is unclear and is not necessarily related to the accumulation of metformin, as was previously thought, because metformin is excreted unchanged in the urine. Tissue hypoxia has been noted to be a trigger of lactic acidosis and, therefore, metformin should be discontinued when tissue hypoxia is suspected, regardless of renal function (Jones et al, 2003).

Putative risk factors for lactic acidosis with biguanide treatment include: age >60 years; decreased cardiac, hepatic or renal function; diabetic ketoacidosis; surgery; respiratory failure; ethanol intoxication; and fasting (Luft, 2001).

NICE states that metformin is contraindicated in those with renal impairment (serum creatinine levels >130 µm/l) and in those at risk of sudden deterioration of renal function (NICE, 2002). However, the British National Formulary recommends avoiding metformin in mild renal impairment, where mild renal impairment is defined as a glomerular filtration rate (GFR) of 20–50ml/min/1.73m² or serum creatinine levels in the range 150–300µm/l (British Medical Association and the Royal Pharmaceutical Society of Britain, 2007).

There are various reasons why renal impairment occurs in people with diabetes. One of the long-term complications of diabetes is diabetic nephropathy, which can progress to renal failure in some cases. The main focus of therapy in diabetic nephropathy is blood pressure. Guidelines have progressively revised the target blood pressure goal downwards and they currently stand at 130/80mmHg in people with chronic renal failure and recommend either angiotensin converting enzyme (ACE) inhibitors or angiotensin-II receptor blocker (ARB) therapy as first-line renoprotective antihypertensive therapy (British Cardiac Society et al, 2005).

Many people with type 2 diabetes are therefore prescribed ACE inhibitors. Although ACE inhibitors now have a specialised role in some forms of renal disease, they also occasionally cause impairment of renal function that may progress and become severe (at particular risk are the elderly; British Medical Association and the Royal Pharmaceutical Society of Britain, 2007). Concomitant treatment with nonsteroidal anti-inflammatory drugs (NSAIDs) increases the risk of renal damage (British Medical Association and the Royal Pharmaceutical Society of Britain, 2007). We also know that NSAIDs are popular drugs in individuals with arthritis (Longmore et al, 2004).

Unfortunately, normal aging is accompanied by a reduction in GFR by 6–10 ml/min/1.73m² per decade after the age of 40 years, so those over 70 years of age may have a GFR less than 75ml/min/1.73m² (National Kidney Association).

Study aims
This retrospective study, performed at Asplands Medical Centre in Milton Keynes (total registered population: 10026), compares current practice in primary care with NICE guidelines on the management of type 2 diabetes and the recommendations from the British National Formulary with regards to the use of metformin in people with renal failure (NICE, 2002; British Medical Association and the Royal Pharmaceutical Society of Britain, 2007). The aim is to improve treatment and routine monitoring of people with diabetes on metformin.

Method
The clinical notes of people who were on metformin when the audit commenced and who also had serum creatinine levels above 150 µm/l at any time since commencing metformin treatment were reviewed. Data on the latest serum creatinine levels (within last 6 months) were also examined.

The age, duration of diabetes, concurrent drug therapy and BMI of people with current serum creatinine levels above 130µm/l were also taken into account. Metformin therapy was considered inappropriate if the serum creatinine level was more than 150µm/l. The value of 150µm/l was used in accordance with recommendations from the British National Formulary (British Medical Association and the Royal Pharmaceutical Society of Britain, 2007) and is also used in current local practice.

Results
Individuals with type 2 diabetes comprise 2.6% (n = 262) of the total practice-registered population (10026). Of these, 223 (85.1%) have been on metformin at some time since diagnosis. Most individuals on metformin were aged 60–80 years (60.5%; n = 135; Figure 1).

In total, 43 (19.3%) individuals who have taken metformin had serum creatinine levels >130µm/l on one or more occasion since diagnosis; 25 (58.1%) of these were in the range 130–150µm/l and 18 (41.9%) had a concentration >150µm/l. Review of the clinical notes of the 18 people on metformin who had serum creatinine levels >150µm/l revealed that ten were not taking metformin and were instead using insulin or another hypoglycaemic agent. However, the remaining eight were still taking metformin. 

Table 1 shows the baseline characteristics of these eight individuals. In these individuals, the latest serum creatinine levels from within the previous 6 months revealed that three had serum creatinine levels <130µm/l, two had concentrations in the range 130–150µm/l and three >150µm/l. Table 2 shows up-to-date information on the latest GFR and serum creatinine levels, and current medications of these eight people.

Discussion
This study shows that five people (2.2%) with type 2 diabetes had serum creatinine levels >130µm/l while currently taking metformin.

Furthermore, all but one of these five people were aged >70 years and were on concurrent therapy with ACE inhibitors. The BMIs of these people varied from 22 to 32kg/m² (Table 2). Available data on latest GFR for three of these people confirmed chronic renal impairment (30, 33 and 40ml/min/1.73m²). Unfortunately, there were no data on the latest GFRs of the other two patients.

Possible explanations for the use of metformin in people with renal impairment include the following.

• Normal serum creatinine levels at time of first prescription and development of renal impairment later went unnoticed.
• The benefits of continuing metformin therapy were greater than any associated risks.
• In view of the clear contraindication to metformin in these five individuals and after discussion with treating physicians, the following management changes were decided upon.
• Individuals will be contacted to discuss replacing metformin with either insulin or another oral hypoglycaemic regimen in order to avoid the associated risks of metformin therapy in the presence of renal impairment.
• Monitoring of serum creatinine levels will be included in the routine check up of people with diabetes using metformin.

One drawback of the audit is that it does not assess whether or not serum creatinine levels were measured before starting metformin. Most individuals were on metformin for several years before the audit and these data were not available.

If the results of this audit are representative of management of people with diabetes nationally, they would imply that a large number of people with diabetes are receiving metformin inappropriately.

The danger of using nephrotoxic drugs (aminoglycosides, tetracyclines, amphotericin B, NSAIDs or ACE inhibitors) in these individuals also needs to be kept in mind (Longmore et al, 2004). ACE inhibitors, being nephrotoxic, are recommended for microalbuminurea; therefore, renal function of these individuals should be monitored closely with a view to stop metformin at recommended creatinine levels (British Medical Association and the Royal Pharmaceutical Society of Britain, 2007).

A population-based study in Tayside Scotland (population 349303) by Emslie-Smith et al (2001) showed that 24.5% of people prescribed metformin have contraindications to its use. Development of contraindications rarely results in discontinuation of metformin therapy. It also showed that despite this lactic acidosis remains rare (one episode of lactic acidosis in 4600 patient years).

Conclusion
A small percentage (2.2%) of people with diabetes audited were receiving metformin inappropriately. In the author’s opinion, renal function should be monitored regularly in people with type 2 diabetes who are taking metformin with a view to cease metformin therapy at the recommended levels of glomerular filtration rate or serum creatinine.

In view of the low incidence of lactic acidosis with metformin in people with renal failure, the benefit–risk ratio favours the use of concomitant therapy with ACE inhibitors. However, the high mortality rate associated with lactic acidosis is a very good reason to stop metformin at the first sign of renal impairment in individuals who are on concomitant ACE inhibitor therapy (especially the elderly).

In conclusion, the following contraindications to metformin are as follows (British Medical Association, Royal Pharmaceutical Society of Britain, 2007).

• Avoid metformin in mild renal impairment (serum creatinine 150–300µm/l).
• Withdraw if tissue hypoxia likely (sepsis, respiratory failure, recent myocardial infarction or hepatic impairment).
• Stop when using iodine-containing x-ray contrast medium and do not restart metformin until renal function returns to normal.
• Stop metformin 2 days before general anaesthesia and restart when renal function returns to normal.
• Avoid in pregnancy and breast feeding.