Of the 17 Kentuckians who received a blood substitute during a controversial study that didn't require consent from participants, four died.
The results, which the University of Kentucky released last week, are not scientifically significant. But the results for the entire trial — UK was one of 31 Level 1 trauma centers that participated in the study — are.
The study examined the use of a PolyHeme, a hemoglobin-based blood substitute, in trauma patients who were bleeding badly.
The idea was to see whether PolyHeme, which like blood carries oxygen to cells, saved lives because patients would be able to receive it before they got to the hospital. Blood isn't available in ambulances.
The makers of PolyHeme and its champions, including Dr. Andrew Bernard, who was the principal investigator for UK's portion of the study, say the study shows the promise of PolyHeme.
"Here's a drug we can use to treat patients when no red blood cells are available," Bernard said.
But critics have questioned the study and the benefit of hemoglobin-based blood substitutes altogether.
Since the PolyHeme trial finished in 2006, an analysis of trials of several hemoglobin-based blood substitutes published in the Journal of the American Medical Association concluded that patients who received the products had a higher risk of death and a higher risk of heart attacks.
"None of these studies have shown a statistically significant benefit," said Dr. Charles Natanson, a senior investigator at the National Institutes for Health, who led the meta-analysis. "They have only shown increased risk."
Blood substitutes were supposed to provide a safer, more available blood supply and the potential to save lives.
Like blood, substitutes such as PolyHeme can carry oxygen to cells. Unlike blood, it doesn't have to be typed, so it can be given to anyone. And it can be stored for a year without refrigeration, making it easy for ambulances to stock. Donated blood lasts 42 days.
Study was in final phase
The study that UK participated in was a Phase III trial, the final phase before a drug goes before the federal Food and Drug Administration for regulatory approval.
Northfield Laboratories Inc., the company that makes PolyHeme, plans to apply for expedited approval later this year, according to a letter to shareholders in August. PolyHeme is the Illinois company's one product, and, according to the letter, the company has just enough money to see the product through the approval process.
The phase III trial drew criticism because of the way it was conducted.
In order to participate, subjects had to be bleeding profusely and have very low pressure. Those patients are usually unconscious, so the study took place under an FDA exception that allows patients to be enrolled without their consent, as long as researchers inform the community before the study begins.
Critics have questioned why more information was not given about an earlier PolyHeme study that was stopped after 10 of the 81 patients who received PolyHeme had heart attacks within seven days and two of them died. None of the 71 patients who didn't receive PolyHeme had a heart attack.
In Lexington, Bernard did not discuss the earlier PolyHeme study during the community meetings, he said. The company had concluded that the problem was that the PolyHeme patients received too much fluid, not that the product increased the risk of a heart attack, Bernard said. In addition, he did not discuss it because of the complicated nature of the results, he said.
Information about the previous study was available in brochures, produced nationally, to educate communities about PolyHeme and the research, said Ernest "Gene" Moore, the Denver Health trauma surgeon who led the study.
Of the 720 participants in the phase III trial that included UK, those who received the drug were slightly more likely to die — 13 percent versus 10 percent. That difference is not statistically significant, researchers said.
A first analysis found that those participants who received PolyHeme were more likely to experience heart attacks than those who did not (3 percent compared with less than 1 percent.) That difference is statistically significant.
However, researchers in the study did not have clear guidelines for what constituted a heart attack, Bernard said.
Severely injured patients often have abnormal electrocardiograms or EKGs. They also have high levels of cardiac enzymes in their blood — the same enzymes that signal a heart attack, said Moore. That makes it hard to determine whether patients in the trial had a heart attack, he said.
So a group of independent cardiologists examined the data to figure out who had heart attacks.
In that analysis, 12 percent of those in the PolyHeme group "probably" had a heart attack and 8 percent in the control group. An additonal 42 percent in the PolyHeme group "possibly" had a heart attack and 44 percent in the control group.
Moore says that the study shows that PolyHeme will save lives when trauma accidents happen far from hospitals.
"Even if there is a small risk of a myocardial event, that overrides the risk of death," Moore said.
But Dean Fergusson, an expert on blood transfusion at the Ottawa Health Research Institute in Canada, doubts that there is a use for a product with the risks of PolyHeme in North America.
"The problem is it's got to be safer than the current system of stored blood. And so far they haven't demonstrated that," Fergusson said.
The statistically insignificant number of deaths in the most recent PolyHeme study, for example, add up over multiple studies.
"The trend is toward more death," he said. "Just because it wasn't statistically significant, that doesn't mean that that association isn't real. It could just be a question of not enough numbers."