The anti-sepsis monoclonals E5 and HA-1A should remainexperimental treatments until clinical trials can show areproducible, beneficial effect, according to an opinionpublished today in the New England Journal of Medicine.

Reviewing the available clinical data on Xoma Corp.'s E5 andCentocor Inc.'s HA-1A (Centoxin), physicians writing in theJournal said that human testing has given inconsistentoutcomes and that even preclinical testing has notdemonstrated a clear-cut benefit in animal models of sepsis.

The physicians, from Massachusetts General Hospital, theNational Institutes of Health and University of TexasSouthwestern Medical Center in Dallas, said that it has been"difficult to test a therapeutic agent in this complex populationand obtain the same result twice."

An accompanying editorial suggested that the rival Xoma andCentocor products have merely paved the way "for second- andthird-generation monoclonal antibodies and for combinations ofimmunotherapies that may show much greater benefit."

While commending the two companies, which "have riskedtheir lifelines on radically new products," the Journal editorialemphasized that the trials were all analyzed primarily undercorporate direction and not initially by major investigatorsfrom academic centers.

"Surely the practice of in-house control of data should bediscouraged," continued the editorial, "and industry should seethe advantage of independent external analyses.

"In my opinion," concluded the editorial's author, Dr. Richard P.Wenzel of the University of Iowa College of Medicine in IowaCity, "both of the available anti-endotoxin monoclonalantibodies help some patients, but a second look at the datasuggests that they would be of marginal benefit in practice."

Responding to the critiques, Centocor and Elizabeth Ziegler ofThe University of California Medical Center in San Diego, whowas director of the HA-1A study published last year by theJournal, argued that the physicians' "synthesis of these data islargely subjective and differs substantially from theconclusions of the Food and Drug Administration advisorypanel that voted unanimously to recommend the licensure ofHA-1A."

In vitro analysis of HA-1A's action on bacterial endotoxin havebeen improved and now consistently show the desired action,the Centocor letter added.

Animal models, meanwhile, have limited ability to fully predictconsistent action in humans, a problem not unique to testingsepsis therapies, according to the respondents. "It certainlydoes not 'erode the stated rationale for the clinical study,' " asthe review had argued.

The FDA last week decided that Centocor's clinical data couldnot be used to establish the efficacy of its sepsis drug and thatdata from Xoma's first E5 trial did not provide sufficientevidence of efficacy for marketing approval.

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

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