List Archives

Wed Dec 20 16:48:36 GMT 2006

'Preliminary' Findings
Of PolyHeme Death Rate
Suggest Approval Setback
By *THOMAS M. BURTON*
December 19, 2006 11:54 p.m.; Page A10

Northfield Laboratories
<http://online.wsj.com/quotes/main.html?type=djn&symbol=NFLD>Inc., seeking
to be first to market with a blood substitute, said that 46 patients died
after being giving its blood substitute, PolyHeme, in a study of trauma
patients, compared with 35 who died under standard treatment.

The findings -- of death rates 30 days after severe injuries -- appear to be
a major setback for the Evanston, Ill., company and for the overall effort
to get federal approval for a blood substitute.

Northfield said there were "discrepancies" in the data from the trauma study
and that it plans to further evaluate them. Thus, it said, it considers the
findings announced yesterday to be "preliminary." Moreover, Chief Executive
Steven A. Gould said in a conference call that the death-rate findings
weren't statistically significant.

He said the company plans to "continue to move forward toward submission" of
the data to the Food and Drug Administration.

The test results were announced after the close of regular trading. As of 4
p.m. composite trading on the Nasdaq Stock Market, the stock was at $11.42,
down $2.90, or 20%. In after-hours trading, it was quoted at $5.54, off 51%
from the 4 p.m. level.

After hearing the conference call, William D. Hoffman, chief of
cardiac-surgery critical care at Massachusetts General Hospital and former
medical director of another blood-substitute company, said, "In my opinion,
it would be irresponsible and unprecedented to grant an approval for
PolyHeme based on these data."

He said that "a 5% absolute difference in mortality in favor of the control
arm of the study is too great" for the findings to be considered
"noninferior" to standard care.

The study was designed to evaluate whether the blood substitute was either
superior to, or not worse than, standard therapy. The announcement of
preliminary data doesn't definitely answer that question, but the raw
numbers would appear to be a big problem for Northfield.

In the study, badly hemorrhaging trauma patients were randomly assigned to
PolyHeme, or to standard therapy, which is saline solution in the ambulance
and donor blood in the hospital. The company said there were a total of 722
patients originally in the study, with 349 patients ultimately receiving the
blood substitute and 363 getting standard treatment. However, it said it has
found data errors in the preliminary findings and that it plans to provide
full results as soon as possible.

Dr. Gould said that in the 586 patients for whom there were no data errors,
there were 30 deaths in the PolyHeme group and 28 in the standard-therapy
group. Dr. Gould said that, in the 586 patients, PolyHeme did achieve
statistical "noninferiority" in the study.

The results would appear to leave the FDA with a quandary: Can the agency
approve a blood substitute that may still be worse than standard therapy?

The way the study was set up, PolyHeme can be found "noninferior" to saline
and donor blood, and still not be "the same as" saline and blood. Instead,
it means that, statistically, the product hasn't been proven better or
worse.

That seemingly technical distinction makes a big difference, as seen in the
trauma study's protocol, or methodological plan, which was made available to
The Wall Street Journal. This document shows that PolyHeme could be deemed
"noninferior" and -- because of the way the term is defined and the
relatively small size of the study -- still be potentially as much as nearly
7% worse than blood and saline in terms of patients' deaths.

In the vast majority of clinical trials, a company is required to show that
its drug is better than something else, either a placebo or the standard
therapy for an illness. In the Northfield study, its blood substitute can
either be superior, inferior or -- the middle ground -- noninferior.

The design of the Northfield trauma study has drawn criticism in part
because of an earlier Northfield study. In that surgery trial, 10 of 81
patients on Northfield's PolyHeme had heart attacks, of whom two people
died. None of the 71 patients getting donor blood had a heart attack in that
earlier study. Northfield quietly shut down that trial and didn't publicly
disclose the results, saying only in a 2001 federal filing that the study
was taking too long to complete. The company has since, in interviews with
the Journal, said the heart attacks may have not been caused by the blood
substitute.

The current Northfield study in trauma patients finished enrolling the last
of its U.S. patients this summer.

List Archives

For your convenience, find the Trauma List Archive Below

Northfield Hits Blood-Test Hurdle