Artificial-blood trials went on in face of risks
Despite evidence that a group of experimental blood substitutes nearly tripled the risk of heart attacks and caused a 30% increase in deaths, the Food and Drug Administration continued to approve some clinical trials of the products, researchers said Monday.
The agency should have known about the risks by 2000 and halted all trials, thereby preventing deaths that have occurred in the last seven years, according to a report published online by the Journal of the American Medical Assn. No such trials are proceeding in the United States now.
Had the agency reported publicly what it knew then, ongoing trials in eight other countries would probably have been blocked, the researchers said.
Dr. Jay Epstein, director of the FDA’s Office of Blood Research and Review, countered in a news conference that the agency had carefully weighed the risks and benefits of each proposed study, then blocked some and allowed others to proceed.
He added that the FDA had been “highly vigilant in its oversight” and that it also had concerns about the safety of the products, which was why “there are no ongoing studies of blood substitutes in the U.S. and no approved products.”
The FDA is convening a two-day meeting on the products beginning today.
A major blow
The latest findings, however, deal an enormous setback for the long-sought goal of developing artificial blood, which has been spearheaded by a handful of small companies.
“Two billion dollars and 4,000 patients later, we still haven’t solved the basic problem of the toxicity,” said Dr. John R. Hess of the University of Maryland, who was not involved in the study. “It’s unlikely that there are going to be any alternatives to donated blood anytime soon.”
The companies are either going to have to develop new products or find completely different ways to administer them, added Dr. William D. Hoffman of Massachusetts General Hospital, who also was not involved in the study.
“This is not the death of the field, but it may be a terminal illness for the companies, which don’t appear to have the resources to undertake these kinds of new developments,” he said.
Companies have been working for more than two decades to develop synthetic blood products because of their great promise, particularly on the battlefield and in emergency situations.
Artificial bloods need no refrigeration and can be stored for as long as a year, compared with the maximum of 42 days for whole blood. Blood types don’t need to be matched and there is no risk of infectious agents.
Different methods
Companies have generally used two approaches to producing synthetic blood. The first was to use liquid perfluorocarbons, which dissolve oxygen and are compatible with blood. Unfortunately, they don’t carry enough oxygen to be very useful.
The second and more widely used approach is to take hemoglobin -- the oxygen-carrying agent from red blood cells -- from outdated blood and remove it from the cells.
The five companies in the study -- Baxter Healthcare Corp., Biopure Corp., Hemosol Corp., Northfield Laboratories Inc., and Sangart Inc. -- all produced hemoglobin-based substitutes. Baxter and Hemosol have since abandoned the field.
Researchers knew that the bare hemoglobin could cause problems, but all of the companies “thought the way they formulated hemoglobin would solve the toxicity seen in earlier products,” Hoffman said. “They were wrong.”
The problem, the study’s authors said, is that the FDA has considered clinical trials for each product separately. It rejected some of them based on risks, but trials for other products didn’t seem to show statistically significant problems -- in part because not enough patients were enrolled.
Moreover, the FDA did not release data submitted to it until either the product was rejected or the company published its findings.
In several cases, those data were not published until five or six years after they were submitted to the FDA, and in a few cases they have never been published, according to the study.
Consistent findings
To better assess the risks, Dr. Charles Natanson of the National Institutes of Health, Dr. Sidney Wolfe of the advocacy group Public Citizen and their colleagues combined the data from 16 separate clinical trials involving 3,711 patients.
Although such studies are normally conducted on a single product for one type of application, Wolfe said, there were hints of problems that could only become statistically significant by pooling the data from all of the studies and products in what is called a meta-analysis.
They found a 30% increased risk of death and a 2.7-fold increase in heart attacks overall. Analysis of subgroups of the data showed that the risks were not restricted to a single product or a single application but were shared by all, Wolfe said.
Because of the consistency of the findings, no further trials of hemoglobin-based products should be conducted unless new animal data show that the risks have been reduced or eliminated, Dr. Lauralyn McIntyre and Dean A. Fergusson of the Ottawa Health Research Institute wrote in an editorial accompanying the report.
The FDA’s Epstein said that it was “unusual for a meta-analysis to look at different products in different use settings.”
“Our reviewers determined there were enough differences between products and their intended uses to support a careful weighing of individual clinical trial proposals, only some of which were allowed to proceed,” Epstein said.
Potential benefits
He also said that the FDA, although aware of the safety concerns, also had information on potential benefits from some of the products that was not publicly available, which was considered justification for approving the trials.
The companies involved also objected to the meta-analysis. Dr. A.G. Greenburg, vice president of medical affairs at Biopure, said: “There are vast differences among these products that make any pooling of data flawed, especially across different clinical experiences.”
But the authors of the study dismissed such objections, saying the results spoke for themselves.
“We find that results of trials across different products and in different clinical circumstances are very similar,” Wolfe said. “That argument does not hold any water at all.”
--
--
denise.gellene@latimes.com
