Thursday, April 14 2005

Bad Outcomes Related to, But Not Caused by, Misdosing of Clotbusters in Heart Attack Patients
Duke University Medical Center News Office

DURHAM, N.C. – The death or clinical complication in heart attack patients given potent drugs to re-open clogged arteries is more likely to be due to individual patient characteristics than to modest misdosing of the drugs, researchers from the Duke Clinical Research Institute (DCRI) have determined. Thus, they said, physicians should not be overly worried about the possibility of minor errors in dosing -- a concern that the Duke physicians said likely contributes to insufficient prescribing of the drugs to treat heart attacks.

These findings are important, the researchers said, because approximately one in three patients rushed to emergency rooms with heart attack symptoms do not receive these potentially life-saving drugs. One reason cited for this under-usage is that many physicians may be overly – albeit sometimes justifiably -- concerned about potential adverse events due to incorrect dosing, the researchers said.

The drugs, known as fibrinolytic agents, have been proven effective in clinical trials in breaking apart blood clots in coronary arteries, restoring blood flow to heart and saving heart muscle. The doses of some of the drugs in this class are determined by a calculation that includes body weight, which must often be estimated in emergency situations. The amount of drug administered is crucial, the researchers said, since too much can cause unwanted bleeding events, while too little will not dissolve the blood clot.

"We have found that there does not appear to be a cause-and-effect relationship between small dosing errors and worse outcomes for patients," Granger continued. "These findings should be somewhat reassuring to busy physicians in emergency rooms who can assume that as long as the drugs are given carefully to the right patients, small misdosings should not lead to bad outcomes. These are a class of drugs that can save lives, but tend to be underused."

For their study, the Duke team analyzed the results of the 16,949-patient clinical trial concluded in 1998 known as ASSENT-2 (Assessment of the Safety and Efficacy of a New Thrombolytic). The trial compared the effectiveness of tissue plasminogen activator (t-PA) and TNK, a newer genetically altered version of t-PA. t-PA works by dissolving blood clots and is most effective when administered within hours of heart attack symptoms.

Using the data collected from ASSENT-2, the researchers analyzed to what extent incorrect dosing – whether over or under – was responsible for death, stroke or major bleeding within 30 days of treatment.

When the researchers adjusted the data for these confounding factors, they found that the association was greatly reduced. The patient-specific factors that appeared to bestow the greatest risks for adverse events included advanced age, female gender, being African-American, shorter stature, lower body weight, and lower blood pressure.

Specifically, the 30-day mortality rate for those who received a t-PA overdose was 9.8 percent, and 19.5 percent for those receiving an underdose. Those receiving a correct dose had a mortality rate of 5.4 percent. Conversely, those who received an overdose of placebo had a 10 percent mortality rate, and 23.5 percent rate for underdose. Those who received the correct dose of placebo also had a 5.4 percent mortality rate.

"These findings are counterintuitive – when given the wrong dose of placebo, there was much worse outcome," Granger said. "This is clearly a remarkable finding, as well as a reason to pause – whenever we see a relationship between a treatment and an outcome, we must be careful in ascribing cause and effect. Other factors may also be involved, and additional studies are needed to better understand what these factors may be."

Granger cited as a similar example the issue of hormone replacement therapy (HRT) for post-menopausal women. Earlier observational studies of thousands of women indicated an association between HRT and a lowered risk of heart attack.

"However, when the randomized, scientific studies were done the opposite was proven to be true, that there was in fact an increased risk of blood clots and stroke," Granger said. "Another example is vitamin E. At first, earlier studies appeared to show that vitamin E was associated with better outcomes, but when the proper studies were conducted, it turned out not to be the case."

While Granger does not believe that the results of this analysis will change clinical practice, he does believe that hospitals and physicians should not withhold beneficial drugs because of concern over misdosing, as long as the drugs are given carefully and errors minimized. Less concern about minor misdosing could lead to an increased usage of the drugs. Additionally, he said that the patients at the highest risk for adverse outcomes should be closely monitored while they are receiving fibrinolytic therapy.

Mehta believes that the results of this study could also have an impact on medical malpractice litigation, which is often driven by adverse outcomes.

"This study finds that adverse outcomes following modest errors in fibrinolytic dosing are often times not caused by the error, but by clinical factors that may contribute to higher likelihood of the error occurring," Mehta said.

The ASSENT-2 trial was funded by Boehringer Ingelheim and Genentech, Inc. Granger's analysis of the ASSENT-2 data was supported by the DCRI.

Other members of the team, from the DCRI, were John Alexander, M.D., Karen Pieper, Jyotsna Garg, and Robert Califf, M.D. Other colleagues were Frans Van de Werf, M.D., University Hospital Lueven, Belgium, and Paul Armstrong, M.D., University of Alberta, Edmonton.

John Alexander, M.D.

Karen Pieper

Rob Califf, M.D.