Physical examination indicated that the patient's height is 5 feet 9 inches; weight, 178 lb; calculated body mass index, 26.3; waist circumference, 38 inches; blood pressure, 132/76 mm Hg; and heart rate, 76 beats per minute. Other physical examination findings are normal, excluding some funduscopic changes consistent with hypertension. 

When comparing baseline laboratory test results with current results (Table 2), his HbA_1c decreased from 7.8% to 7.4%. Dave's fasting glucose and mean glucose levels also decreased (187 mg/dL to 132 mg/dL and 177 mg/dL to 166 mg/dL, respectively). 

Liver function studies, electrolytes, serum urea nitrogen/creatinine ratio, and urine for microalbumin were all normal at baseline and at presentation. Lipid profile, taken at diagnosis, revealed a total cholesterol of 154 mg/dL; low-density lipoprotein cholesterol, 68 mg/dL; high-density lipoprotein cholesterol, 58 mg/dL; and triglycerides, 89 mg/dL. His calculated non–high-density lipoprotein cholesterol was 96 mg/dL. 

Given Dave's profile at baseline and then 2 months later at presentation, you must consider how to move forward in terms of treating this patient. 

Jay Shubrook, DO: This patient is not at goal at 2 months. This is a young patient who is relatively new to diabetes, so the target HbA_1c should be 6.5%. This person can be protected for a number of years, so if he can get his HbA_1c below 6.5% without hypoglycemia, that would be acceptable. 

Frank Lavernia, MD: I agree that he is not at goal. If he can avoid hypoglycemia and weight gain and keep his HbA_1c levels between 6.5% and 7.0%, that would be a satisfactory target. 

Jeffrey S. Freeman, DO: I agree also. This patient was diagnosed as having diabetes only 2 months ago, but he has probably had diabetes for 5 to 7 years. His treatment has only just begun, so there is ample time to try different strategies and make therapeutic change. The patient's history of coronary artery disease may preclude the ability to drive his HbA_1c levels to 6.5% or lower. 

Frank Svec, MD, PhD: Both metformin and exenatide are good options. I think exenatide is going to be an exciting possibility because of its effect on weight and its other actions. However, exenatide is still a fairly new agent and more data are needed before using it for first-line therapy. On the other hand, metformin is a good initial therapy because of its effect in overweight patients, as shown in the UKPDS [United Kingdom Prospective Diabetes Study].¹ Metformin also reduces the risk of cardiovascular outcomes. Currently, my first choice would be metformin because we have outcome data to support its use as a first-line therapy. 

Dr Lavernia: I agree with Dr Svec that metformin should be the first choice. The patient's HbA_1c levels do not dictate that he should begin with dual therapy. Metformin as monotherapy could be effective in reducing his HbA_1c levels. 

Dr Shubrook: You could pick any agent and he would improve. His improvement would be even better if the patient included lifestyle changes also. I choose one agent for each percentage point that the patient is beyond the goal. So, at 7.4%, one agent would be adequate, at 8.4%, two agents are needed, and at 9.4%, I would add insulin or insulin replacement therapy. Metformin is definitely my first choice of therapy. 

Dr Shubrook: This is where the physician needs to individualize treatment decisions. I would ask the patient to check his fasting and postprandial blood glucose levels. I would want to determine whether there is a difference between them. I would also want to discuss his goals with him, find out what his insurance will cover, and encourage him to decide what he wants to do. If he has dyslipidemia or fatty liver disease, I would choose a TZD [thiazolidinedione]; if he has postprandial hyperglycemia, I would choose a secretagogue; and if weight is an issue, I would choose a DPP-4 [dipeptidyl peptidase-4] inhibitor or exenatide. 

Dr Lavernia: My concern is whether this patient will agree to inject himself. I would want to see whether he can tolerate the medications, so I would start him on free samples of the agents. If he loses weight, he will most likely be satisfied with the medication and be willing to continue using it. If he develops nausea and can't tolerate the medication, I will move on to something else. Cost must also be considered, as well as weight gain and hypoglycemia, which are possible with a sulfonylurea. Basal insulin also has a cost attached, particularly if taken twice a day. Analog insulin, which is the glargine, is more expensive and effective. 

Dr Svec: I am strongly opposed to using a sulfonylurea. The UKPDS trial¹ showed that the combination of metformin and a sulfonylurea leads to increased complications. No one initially believed that result, but a meta-analysis found that the sulfonylurea/metformin combination increased cardiovascular complications. Sulfonylureas are inexpensive and have been available for 50 years, but there is no study showing that they reduce complications. Other medications, such as TZDs, DPP-4 inhibitors, and exenatide are expensive. Exenatide is a very good choice because of its effect on weight and because it has a favorable effect on lipids. 

Dr Freeman: This patient is overweight but not obese. He lost 2 pounds in 2 months with some lifestyle modifications. He also has a history of cardiac disease. I would not rule out a TZD/metformin combination in this particular patient because they have not been shown to cause myocardial infarction. There are data showing that TZDs may improve the proinflammatory environment. 

Dr Svec: TZDs do not cause heart attacks, but they also do not reduce those events. The outcome data must be considered, but findings are limited. 

Dr Freeman: The PROactive trial² compared a TZD to placebo, and there are data from the RECORD trial^3,4 that showed some measures of improvement in CVD risk or neutrality, but these studies were conducted in high-risk patients. 

Dr Shubrook: The American Diabetes Association has recommended a higher HbA_1c target for someone with a limited life expectancy, other complicating factors, or if the patient is a child. If this patient were aged 86 years, I would focus more on this patient's blood pressure than his glucose. I would not give metformin to an 86-year-old patient. Many hospitals do not use metformin in patients older than 75 years because of the higher incidence of renal disease associated with it, which is harder to monitor in older patients. 

Dr Lavernia: For this older patient, it is important to monitor the estimated GFR [glomerular filtration rate]. If the GFR is far below 50 mL per minute, I would be cautious about administering metformin. If you are considering using metformin for the first time with an older patient, assess the estimated GFR. 

Dr Svec: For an 86-year-old patient, I would be satisfied with an HbA_1c of 8%. I would consider prescribing acarbose because it is a mild agent and does not cause hypoglycemia. The medication can be taken as needed and it helps relieve constipation in elderly patients. 

Dr Freeman: Acarbose actually led to a 25% reduction in myocardial infarction in one of the leading diabetes prevention trials.⁵ 

Dr Svec: This could cause more problems. Metformin is associated with renal problems. Thiazolidinediones are associated with edema. I would not use any agent that poses a risk of hypoglycemia in this patient. I wouldn't want him to have HbA_1c levels of 11% to 13%, but if he can achieve an HbA_1c of 8%, I would be satisfied. I do not think there is going to be an advantage of achieving an HbA_1c of 7% as opposed to 8% for this patient. 

Dr Shubrook: I believe DPP-4 inhibitors would be an ideal agent for patients with a limited lifespan and for those in a nursing home. Diabetes is difficult to manage in the elderly, and an agent that is glucose-dependent that could be implemented with minimal monitoring would be ideal. Patients in nursing homes should not be administered glyburide because of the risk of hypoglycemia and the difficulty of monitoring these patients by nursing staff. 

Dr Freeman: In a study comparing sitagliptin to placebo specifically in elderly patients (aged ≥65 years), it was observed that there were no safety issues with sitagliptin compared to placebo.⁶