Surprising New Diabetes Data
The NIH halts the ACCORD trial after finding that more people in the intensive treatment group died than in the standard care group by John Carey
There is a critical question in medicine today: How useful are the widely accepted measures of health? The most common examples are blood pressure, so-called bad cholesterol, and blood sugar in diabetics. The expectation is that lowering these metrics will result in fewer heart attacks and other bad things, and bring longer life. As a result, there is a relentless push to lower the numbers, to make them "better."
But these measures are only surrogates for disease. And in many cases, the connection between "better" numbers and better health is tenuous. In the case of cholesterol, many people won't see a health benefit from lower numbers (BusinessWeek, 1/17/08).
Now comes yet another sobering reminder that lowering a surrogate marker doesn't necessarily bring better health. On Feb. 6, the National Institutes of Health announced it was halting a key trial for diabetes. Researchers had hoped the trial, dubbed ACCORD (Action to Control Cardiovascular Risk in Diabetes), would show that more aggressive lowering of blood sugar would significantly reduce deaths. Instead, the opposite happened. More people in the intensive treatment group died than in the group getting standard care. "A thorough review of the data shows that the medical treatment strategy of intensively reducing blood sugar below current clinical guidelines causes harm in these…patients," says Dr. Elizabeth Nabel, director of the National Heart, Lung & Blood Institute.
The results fly in the face of
Benefits of Aggressive Treatment Unproven
However, the evidence that aggressive blood sugar control using drugs brings benefits has always been weak. Proponents of more intensive drug treatment cite a landmark study started back in the 1970s called the U.K. Prospective Diabetes Study (UKPDS). In that trial, researchers followed thousands of patients with type 2 diabetes for more than a decade, comparing those who merely got advice on diet and lifestyle to those who took drugs to lower their blood sugar and keep it under tighter control. The researchers claimed the more aggressive treatment worked to reduce the complications of diabetes, which include kidney failure, blindness, amputation of infected limbs, and cardiovascular disease.
But in fact, the benefits seen in the trial were tiny or nonexistent. The study's authors had to torture the data to reach their conclusions, charges Dr. Robert Ewart, associate professor of family and community medicine at Southern Illinois University School of Medicine. "The UKPDS was a particularly egregious example of data manipulation," he says. "There just isn't any evidence that tight control of type 2 diabetes improves outcomes."
Ewart has plenty of company in doubting the trial. No oral diabetes drug "has ever been shown to do anything really good for any patient," adds Dr. Nortin Hadler, professor of medicine at the
Expected Benefits Not Found
In spite of this weak evidence, drugs that successfully lowered blood sugar became blockbusters. Avandia, for instance, racked up $2.6 billion in sales in 2006. When anyone questioned the premise that aggressive treatment is better, doctors pointed to the ongoing ACCORD trial. That trial, they said, will prove the benefits of aggressive glucose control.
They were wrong. The most important finding is not necessarily that more patients in the aggressive treatment group died than in the standard treatment group. The difference was relatively small. Among the 10,251 patients, 257 people given intensive treatment died over four years, compared to 203 in the control group. That's just more than one additional death per 100 people over the four-year period.
Instead, the real significance is that the expected benefits were not found. When the trial was started, researchers hypothesized that keeping one measure of blood sugar (called HbA1c) below 6 instead of the standard 7.5 would reduce heart attacks and other cardiovascular events by at least 15%. The lack of such a benefit casts serious doubt on the more intensive use of the drugs, especially because they come with a long list of side effects, such as fluid retention (which can be life-threatening) and weight gain (which makes diabetes worse).
Focus on Clinical Outcomes
And the new finding raises red flags about using surrogate markers in general—and about whether the FDA should approve drugs just because they lower blood sugar or cholesterol. "The use of surrogate end points in the drug approval process has been problematic," wrote Dr. Bruce Psaty of the
Concludes Rosen: "We urgently need to change the regulatory pathway for drugs for the treatment of type 2 diabetes to make clinical outcomes, not surrogates, the primary end points."
What does all this mean for diabetes patients? They may not have to use so much medication in an effort to keep blood sugar levels down, thus reducing the side effects. But meanwhile, they should strive to reduce other risk factors for cardiovascular disease, such as by losing weight and keeping blood pressure in check.
Carey is a senior correspondent for BusinessWeek in