Attempts to navigate the tempestuous waters of diabetes have once again been dashed against the rocks of clinical reality. The trial failed to find any evidence of clinical benefit for either valsartan of nateglinide in patients with impaired glucose tolerance.

Results of NAVIGATOR (Nateglinide and Valsartan in Impaired Glucose Tolerance Outcomes Research) were presented today in Atlanta at the ACC and published simultaneously in the New England Journal of Medicine.

In a 2 x 2 factorial design, 9306 patients with impaired glucose tolerance and with cardiovascular disease or at high risk for CV disease were randomized to valsartan or placebo and nateglinide (a short-acting insulin secretagogue) or placebo. Patients were followed for a mean of 5 years for diabetes incidence and 6.5 years for vital status. There were 3 coprimary outcomes:

• the development of diabetes
• extended composite outcome (death from CV causes, nonfatal MI, nonfatal stroke, hospitalization for HF, arterial revascularization, or hospitalization for unstable angina
• core composite outcome (above except it excluded unstable angina and revascularization)

In the valsartan versus placebo portion of the trial, the incidence of diabetes was significantly reduced in the valsartan group (33.1% in the valsartan group versus 36.8% in the placebo group, HR 0.86; CI .80=.92, p<0.001). However, there were no significant differences in either the extended composite outcome (14.5% vs 14.8%, p=-.43) or the core composite outcome (8.1% in both groups)

The investigators noted several comparisons with other trials, while acknowledging the limitations of indirect comparisons across trials. In one comparison, they noted that lifestyle modifications in two trials achieved a much larger reduction in the incidence of diabetes than the reduction associated with valsartan in NAVIGATOR. By contrast, they observed that the reduction in diabetes by valsartan in NAVIGATOR was somewhat larger than the effect of ramipril in the DREAM study.

…pharmacologically preventing the “development of diabetes” is not a real goal. It is a surrogate for a surrogate of a surrogate.

In the nateglinide portion of the trial, there were no significant differences in any of the primary outcomes, including the incidence of diabetes (36% for nateglinide versus 34% for placebo, HR 1.07, CI 1.00-1.15, p=0.05), the core composite outcome (7.9% vs 8.3%), or the extended composite outcome (14.2% vs 15.2%).

An accompanying editorial by David Nathan criticized the decision to report the two arms of the trial in separate papers because this “belies one of the stated goals of NAVIGATOR– specifically, to study the combined effects of the two study drugs.”

Nathan concluded: “The prevention of diabetes remains a critical public health priority, but for now we should steer away from these two drugs and use effective lifestyle interventions and, in selected persons, metformin to combat the epidemic.”

James Stein sent the following detailed comment to CardioBrief:

Studies that test the strategy of “preventing diabetes” by administering an anti-glycemic medication really represents a “sleight of hand” by the pharmaceutical sponsors, who are looking to expand their market for blood sugar-lowering medication. After all, what is the difference between a pre-diabetic patient on a diabetes drug and a diabetic patient on a diabetes drug? Nothing, except catching the disease at an earlier stage, with an intervention of unknown long-term benefit and well known short-term risks and expense. You prevent diabetes, defined by a number, by lowering the number — not a really impressive trick. Now, if treating early prevents micro- and macro-vascular complications of diabetes, there might be a benefit to starting them early. But pharmacologically preventing the “development of diabetes” is not a real goal. It is a surrogate for a surrogate of a surrogate.

To this end, I congratulate the NAVIGATOR investigators for looking at preventing CVD events, but I am not at all surprised it was a negative study in regard to nateglinide. Secretagogues like nateglinide do not reduce CVD events even in individuals with diabetes, and there are mechanistic reasons to suspect harm, related to hypoglycemia and perhaps the effect of inhibiting the K-ATP channel in the myocardium.

Also, I think the authors’ conclusion that the drug “appears to be safe” is overstated. Although well-powered for efficacy, it was not big enough or long enough to exclude safety concerns, and it certainly can’t address safety in people with established CVD, in whom hypoglycemia has been shown to be predictive of mortality and CVD events, and in whom K-ATP channel inhibition could be very harmful. It is noteworthy that nateglinide binds more avidly to the cardiac K-ATP receptor (SUR2A) than the pancreatic K-ATP (SUR 1) and does so with an affinity similar to tolbutamide, which increased CHD events back in the 1980′s University Group Diabetes Study.

Finally, to say nateglinide “did not reduce the incidence of diabetes” is completely misleading– it increased it by 7% with a p=0.05. To me that is not safe, either. The explanation of rebound is an unconvincing hypothesis; “special pleading” for a drug without a niche. I suspect long-term beta cell failure was the reason.

In the end, our focus on drugs for individuals with glucose intolerance is a big distraction from proven interventions for diabetes and CVD prevention – diet and exercise. And for those who do fail, we have two good drugs: acarbose, which in patients with pre-diabetes reduced CVD events and progression to diabetes and metformin, which modestly reduced development of diabetes, but at least seems to reduce CVD events among diabetic patients. This disease is all about lifestyle, and if they fail lifestyle, acarbose may be the best answer. It may cause gas – but I suspect most people would put up with it if they knew it reduced diabetes, HTN, and CVD by 36-49% over 3.3 years.

Regarding the valsartan arm – it was a small effect for a class of dugs that we commonly use in these sorts of patients only. It is of interest, but not a game-changer.