You are an internist treating an overweight 56-year-old woman who was diagnosed as having type 2 diabetes mellitus 8 years ago. Her glycosylated hemoglobin (HbA1c) level has not been well controlled (typically at 8.3%) during the past 12 months, despite the fact that she is receiving metformin therapy and you have been providing her with repeated counseling regarding weight and exercise. Her low-density lipoprotein cholesterol (LDL-C) and blood pressure (99 mg/dL and <135/80 mm Hg, respectively) are managed with statin therapy and an angiotensin-converting enzyme (ACE) inhibitor combined with a thiazide diuretic. You are aware that your patient's blood glucose level is too high, but on the basis of previous discussions, you are also aware that she is reluctant to take more medications and that she is opposed to insulin, which may increase her struggles with her weight. Your patient is very astute, and she has asked you about the effect of the medication on complications related to diabetes, including stroke and myocardial infarction.
A colleague has been talking to you about exenatide, a once-weekly injectable glucagon-like peptide 1 receptor agonist, lauding its better control of HbA1c in patients with inadequately controlled type 2 diabetes. He mentions the drug has essentially no risk of hypoglycemia and has beneficial effects on weight compared with dipeptidyl peptidase-4 (DPP-4) inhibitors, sulfonylureas, thiazolidinediones, and insulin. Wondering if your patient would agree to an injectable drug in light of the expected advantages, you are optimistic that if improved glucose control with exenatide is also associated with a reduction in macrovascular complications of diabetes, then shared decision making may lead her to try the drug. You, therefore, conduct a literature search to determine what evidence is available that addresses this issue.
You are aware that research that addresses antidiabetic therapies is voluminous and complex because investigators have tested many treatments, dosages, and combinations of treatments against each other. For a rapid overview, you start with a resource that will summarize existing evidence at a topic-level (see Chapter 5, Finding Current Best Evidence). You opt for UpToDate, type “exenatide,” and find the chapter title, “Glucagon-like peptide-1 (GLP-1) based therapies for the treatment of type 2 diabetes mellitus.” Among several studies that partially inform your question, the chapter cites a Cochrane systematic review on GLP-1 that included 17 randomized clinical trials (RCTs) comparing different regimens.1 Accessing the full review, you browse through included studies for a head-to-head-comparison between exenatide plus metformin vs metformin alone that reports patient-important outcomes, such as cardiovascular events. Failing to find such a trial, you identify another one that is rated as having a low risk of bias and that compared the glucose-lowering effect of exenatide when added to metformin vs sitagliptin and pioglitazone with HbA1c being the end point of interest.2
What is a Surrogate Outcome?
Ideally, clinicians making treatment decisions should refer to methodologically strong RCTs examining the effect of therapy on patient-important outcomes, such as stroke, myocardial infarction, health-related quality of life (HRQL), and death. Often, however, conducting these trials requires such a large sample size or extended patient follow-up that researchers or drug companies look for alternatives. Substituting laboratory or physiologic measures that are associated with patient-important outcomes (surrogate end points) permits researchers to conduct smaller and shorter trials, thus offering a seemingly efficient solution to the dilemma.3