This article is more than 5 years old. Information may no longer be current.
Should clinical approach to type 2 diabetes treatment change?
More focus on insulin resistance, less focus on treatments producing overwork of beta cell output may now be proper.
Eight years after the results of the United Kingdom Prospective Diabetes Study were presented, we have a partial answer for one of the central findings of the study, in which type 2 diabetes was seen as a progressive disorder in terms of loss of beta-cell function. Worse, none of the treatments tested in that study either ameliorated or averted the inexorable, continuing loss of beta-cell function post-diagnosis. Surprisingly, not lifestyle modification, nor insulin, nor metformin, nor sulfonylureas significantly modified the progressive loss of function in the 5,102 patients enrolled in that study.
However, no thiazolidinedione was included in the UKPDS, as this class of oral antidiabetic agents was discovered long after the study had begun. Preliminary findings from other short-term studies with TZDs have indicated an impressive ability to assert durable diabetes control seemingly superior to other agents lacking insulin-sensitizing properties. This is of some consequence since the failure of other oral agents to improve insulin sensitivity may explain their inability to mitigate the loss of beta-cell function as seen in the UKPDS. For these reasons, a test of glycemic durability, ADOPT — or, A Diabetes Outcome Progression Trial — was designed to assess the potential benefits (or lack thereof) of the TZD rosiglitazone as initial therapy in drug-naive, recently diagnosed patients with type 2 diabetes.
|
|
To perform such an assessment, an active comparator trial was designed using two different agents in standard practice for initial therapy in type 2 diabetes, specifically metformin and the sulfonylurea glyburide. About 4,360 recently diagnosed (<2 years diabetes duration) drug-naive patients were randomized to one of three different monotherapy treatments including metformin 500 mg daily, glyburide 2.5 mg daily and rosiglitazone 4 mg daily. Diabetes medications were administered in a doubled-blind manner and were titrated to assure adequate glycemic control. The primary endpoint was monotherapy failure defined as an on-treatment fasting glucose of more than 180 mg/dL that is reproducible. Secondary endpoints included monotherapy failure at 140 mg/dL, rather than 180 mg/dL, as well as the proportion of patients remaining at HbA1c goals of <7%. Measures of beta cell function were also included, as were measures of patient safety. Endpoints were assessed at four years after randomization.
Rosiglitazone superiority
Rosiglitazone was substantially superior to metformin in maintaining glucose control below 140 mg/dL or 180 mg/dL (by RR=0.36 or 0.32, respectively). This effect was even more striking compared with glyburide (RR=0.62 or 0.63, respectively). Fasting glucose levels remained nearly constant during the four-year treatment interval, as did HbA1c levels. The time to 50% failure of glucose control (HbA1c >7%) was 33 months with glyburide, 45 months with metformin and was estimated at 60 months for rosiglitazone. This result implies longer and more frequent periods of poor control when glycemic control fails sooner on certain agents. It also has the potential to escalate costs in terms of added medications, visits and the resulting complications of more prolonged poor control.
These very large improvements in constancy of glucose control with rosiglitazone may be the result of its unique mechanism of action, which proceeds largely by improving insulin resistance and thereby reducing insulin secretion. Clearly, measures of beta-cell function improved with rosiglitazone by much more than with metformin or glyburide. Thus, it may be that improving insulin resistance has a disease-modifying influence on beta-cell decline and loss of beta-cell function, as seen in the UKPDS. It is therefore possible that we may be seeing for the first time that a therapeutic agent can modify the course of diabetes, at least type 2 diabetes.
Confirmation of such an hypothesis may be obtained from the findings of the DREAM trial, which recently showed that rosiglitazone, more than any other antidiabetic agent, seemed to prevent progression of impaired glucose tolerance or impaired fasting glucose to overt diabetes. Thus, it may be that the decline in beta-cell function — which begins long before type 2 diabetes begins and continues long after its diagnosis — is fundamentally modifiable by an agent overcoming insulin resistance, quite aside from the agent’s glucose lowering abilities, which appear to be no different than other antidiabetic agents with lesser apparent ability to stabilize beta-cell function.
The adverse effect profile of rosiglitazone in ADOPT was similar to that seen elsewhere. Patients gained more weight with rosiglitazone than with glyburide (2.5 kg) or metformin (6.9 kg). Despite this, BP was lower with rosiglitazone as were microalbumin-creatinine ratios. Furthermore, hospitalizations and cardiac events with rosiglitazone were no different than with metformin, including episodes of congestive heart failure. On the other hand, hypoglycemia was substantial with glyburide and far less with rosiglitazone or with metformin. Of note, total fractures were slightly increased with rosiglitazone, especially in women (average age 56.3 years at randomization). These fractures involved not the hip or spine but the bones of the extremities.
In light of these findings, one must wonder whether or not the clinical approach to management of type 2 diabetes should change with more of a focus on insulin resistance and less focus on treatments producing overwork of beta cell output. Whether or not improved insulin sensitivity is the potential mechanism that accounts for the seeming beneficial effect of rosiglitazone is open to question. What is not questionable, however, is the apparent stability of glycemic control before and after diabetes diagnosis with this agent. This finding may therefore be of fundamental significance for diabetes management and its implementation may be similar to present efforts toward prolongation of the honeymoon period in patients with type 1 diabetes.
Alan J. Garber, MD, PhD, is Professor, Departments of Medicine, Biochemistry and Molecular Biology and Cellular and Molecular Biology at Baylor College of Medicine; and he is a member of Today in Cardiology’s Editorial Board. Garber is also Chief Medical Editor of Endocrine Today, another SLACK Incorporated publication.