The multicenter study of Centocor Inc.'s Centoxin monoclonalantibody, reported in February, was roundly criticized inletters published today inthe New England Journal of Medicine.

Centoxin neutralizes the effects of endotoxin, the cause ofgram-negative sepsis, a pernicious bacterial infection that canlead to septic shock and death.

Eight letters to the Journal's editor raised concerns aboutpossibly adverse effects of the antibody and possibleimbalances in the statistical design of the study. Theycautioned against widespread use of the agent without more-refined diagnoses of sepsis, which is a syndrome that occursin about a million cases annually in the United States andEurope.

One group from Switzerland claimed it was unable to duplicateCentocor's demonstration that the antibody's binding behaviorspecifically protects against the gram-negative bacterialendotoxin.

Most of the critics emphasized the study's lack of an overallreduction in mortality. Only the subgroup of patients laterdocumented to have had gram-negative bacteria in thebloodstream showed statistically significant results with theantibody.

"The dilemma, then," wrote Dr. Gregory Schmidt of theUniversity of Chicago, "is that clinicians can choose to givethis new therapy to all patients whose condition meets thedefinition of 'sepsis' (knowing that their outcomes are notsignificantly different whether they receive the antibody orplacebo), wait to treat only the patients whose blood culturesbecome positive (a potentially lethal delay) or attempt todevise better criteria to identify patients who will have gram-negative bacteremia (an unlikely feat)."

Scientists in the Centocor study, headed by Elizabeth Zieglerof the University of California, San Diego, said that they had nointention of demonstrating an overall reduction in mortality inall sepsis patients, but wanted only to show the antibodywould be useful in patients with bacteria in their blood as thedocumented cause of their sepsis.

They defended the specific binding of their antibody to theendotoxin and detailed further analyses of their data that showno statistical problems in the makeup of the patient groupsand no evidence suggesting the drug is toxic.

"We agree," the study authors wrote, "that better predictorsand bedside tests for endotoxemia are needed. However, untilthey are available, we believe that the very high risk-to-benefit ratio of immunotherapy with HA-1A (Centoxin)warrants its use in patients with sepsis in whom gram-negative bacteremia is suspected, including those with septicshock."

In its letter, Xoma Corp., which has sued Centocor allegingpatent infringement by Centoxin, wrote that conclusions aboutan antibody, which appears to benefit only people withdocumented blood bacteria, should not be applied to anantibody that benefits a broader range of patients.

Xoma's E5 antibody, tested in a study that has yet to bedocumented in a refereed journal, helped survival in patientswithout a positive blood culture, but who had sepsis diagnosedat infected body sites, wrote Kenneth Gorelick.

But Xoma has not shown the benefit of its E5 in increasing thesurvival of patients with gram-negative infection who werealready in shock, said Sheldon Wolff of Tufts Universitymedical school. Wolff had written an editorial thataccompanied the Centocor report in February. Centoxinsignificantly reduced the mortality rate in patients withgram-negative bacteria in their bloodstream who hadprogressed into septic shock.

Both companies cited the gag order issued in the suit, which isbeing tried in San Francisco, as reason not to comment furtheron the letters.

-- Roberta Friedman, Ph.D. Special to BioWorld

(c) 1997 American Health Consultants. All rights reserved.