Predictors of response in initial users of metformin and sulphonylurea derivatives: a systematic review

To provide an overview of factors predicting metformin and sulphonylurea treatment response.

A large variability between individuals in treatment response to metformin and sulphonylurea derivatives exists. Understanding which factors determine response to these drugs may pave the way for more individualized therapy.

We conducted a systematic search in the MEDLINE, Cochrane and EMBASE databases, between 2003 and 2012 for articles assessing demographic and clinical prediction factors of treatment response in initial users of metformin or sulphonylurea. A literature search of articles referenced within the studies identified was also performed. Treatment response was defined as change in HbA1c level, reaching target HbA1c levels or time to treatment change. Studies were assessed on quality, sample size and type of analysis. Results were summarized by tabulating positive, null and negative associations observed for included predictors.

A total of 10 articles (six trial reports and four cohort studies) were obtained, including three of sufficient quality. For metformin, baseline HbA1c , older age, lower BMI and shorter disease duration were found to be predictors of better treatment response in at least three studies of sufficient quality. For sulphonylurea derivatives, baseline HbA1c and shorter duration were identified as predictors of better treatment response in at least two studies of sufficient quality. Race, smoking status, lipid levels, blood pressure, kidney function and comorbidities were not significantly associated with treatment response.

Several demographic and clinical factors were identified as possible predictors of response to metformin and sulphonylurea, but the number of studies with sufficient quality was small. Generally, early treatment seems important for achieving better glycaemic outcomes.