Small steps, large problems: advances in the management of diabetes mellitus

Abstract

The evidence of a diabetes epidemic is as ubiquitous as the disease: the prevalence in U.S. adults approaches 8%, with 800,000 new cases each year (1); new classes of medications (2) and new forms of old drugs, such as insulin (3), have been developed to treat type 1 and type 2 diabetes; and clinical trials abound. In the current issue of the Journal, Gerich reviews new insulin formulations and their clinical use (3), and Metchick and colleagues discuss the treatment of diabetes in hospitalized patients and propose management strategies (4). The articles are well timed because more patients are being treated with insulin, and management of patients with diabetes in the hospital setting has become more complex with shorter inpatient stays. Insulin use has increased, largely owing to more type 2 diabetic patients being treated with insulin, a trend that is likely to continue with the recognition that beta-cell function in type 2 diabetes wanes over time (5) and that insulin therapy is necessary to achieve the near-normal glucose levels required to prevent long-term complications (6). The number of tools to manage diabetes has increased in the past two decades. New self-monitoring devices, algorithms to achieve near-normal glycemia, methods of measuring levels of chronic glycemia, and medications have all contributed to modern management. The oldest of the glucose-lowering medications, insulin, has also been transformed, with recombinant insulin and analogs engineered to provide a broader spectrum of activity profiles. As reviewed by Gerich, the new insulins are associated with some advantages. In particular, rapid-acting analogs provide increased convenience for patients on frequent treatment regimens. Unfortunately, such insulins have not eliminated the need for frequent injection regimens in type 1 diabetes, which are necessary to achieve the near-normal glycemia for the prevention or delay of long-term complications (7). Moreover, although the very rapid-acting analogs decrease hypoglycemia, the relative benefit compared with more conventional rapid-acting insulins is modest (8). Inhaled insulins, which are also rapid-acting, have not been demonstrated to normalize glucose levels, and long-term safety remains a concern that must be addressed before these formulations can be approved (9). The very long-acting, peakless insulin analogs can provide a consistent basal insulin supply to type 1 or type 2 diabetic patients. In type 1 diabetic patients, however, multiple daily injection regimens with glargine supplying the basal insulin have not been shown to provide longterm control comparable with the Diabetes Control and Complications Trial goals (7). In fact, the lack of a peak may obligate type 1 diabetic patients to add a prelunch injection of rapid-acting insulin, which may not be necessary in regimens with morning intermediate-acting insulin. The absence of a peak may also increase fasting glucose levels, compared with regimens using bedtime intermediate-acting insulin, owing to the dawn rise in insulin requirements (10). In type 2 diabetes, the very long-acting analog is not demonstrably superior to regimens using intermediate-acting insulin with regard to overall glycemia, providing only a modest decrease in episodes of hypoglycemia (11). Gerich unfortunately does not distinguish between insulin therapy for patients with type 1 and type 2 diabetes. Whereas type 1 diabetic patients require carefully orchestrated insulin regimens to provide physiologic insulin replacement to match meals, ambient glucose levels, and physical activity, patients with type 2 diabetes require supplementation, but without the need for frequent injections. Near-normal glycemia has been achieved in type 2 diabetes with relatively simple injection regimens comprising one or two intermediate-acting or long-acting insulin injections per day, taking advantage of the patients’ ability to secrete insulin in response to meals (6,12,13). The major issue for insulin treatment in type 2 diabetes is how much insulin to give, not how often. In addition, the risk of severe hypoglycemia with insulin therapy is substantially lower in type 2 diabetes than in type 1 diabetes. Not surprisingly, the benefits provided by the new, very rapid-acting insulins have not affected type 2 diabetes patients as much as they have patients with type 1 diabetes. Whether the very long-acting insulin analogs are beneficial in type 2 diabetes and whether they are a significant improvement over older formulations is still to be established. Metchick and colleagues point out the disruption of self-care and glucose control that often occurs with hospitalized diabetic patients (4). The premise of inpatient diabetes care in the past has been to maintain “safe” levels of glycemia, avoiding the dehydration and catabolic state of severe hyperglycemia, and the acute consequences and danger of hypoglycemia. Plasma glucose levels between 100 and 250 mg/dL were generally consistent with these goals. Metchick et al. note two clinical studies of acutely ill patients in which insulin and glucose administration with Am J Med. 2002;113:339 –340. From the Diabetes Center, Massachusetts General Hospital, Boston, Massachusetts.