WASHINGTON -- The FDA has told Centocor Inc. that it shouldconduct a new "well-controlled" trial before Centoxin can begiven further consideration for marketing approval, accordingto a statement obtained from the agency by BioWorld onWednesday.

The FDA's position, which apparently is stronger than initiallyreported by the company, virtually eliminates the possibilitythat Centoxin will be approved in the near future.

Centocor on Wednesday announced that the FDA had decidedthat the company's clinical data could not be used to establishthe efficacy of the anti-sepsis drug. The Malvern, Pa., companysaid it hoped approval still could be obtained this year.

The FDA said it requested the additional study becausepremature unblinding of clinical trial results prior to revisionof a Phase III trial protocol could have introduced bias andtherefore invalidated the results. Compliance with the requestwill require a new Phase III trial, FDA spokeswoman MonicaRevell told BioWorld.

Centocor spokesman Richard Koenig said: "The next step is towork with the agency to determine the best way to develop theadditional data that is required. At this point we don't knowwhat that requirement is going to be, but defining that is thenext step. Then the question becomes, how do we best developthe type of trial that the agency desires."

Koenig said he had not seen the FDA statement and was unableto comment on it.

Centocor's claim of efficacy was based on an analysis of asubset of the 616 patients in the trials. The total populationhad sepsis that was presumed caused by gram-negativeinfection. The subgroup included 200 patients who were foundto have gram-negative bacterial infections in theirbloodstream.

"There was no statistically significant difference in survivalbetween treated and placebo groups 14 days after treatment,the study end point in the original protocol," the FDA said."However, prior to finishing the study, Centocor submitted arevised analytical plan to the agency clarifying and modifyingthe types of analyses to be done. The plan included anadditional primary end point -- survival over 28 days."

The result was an improved survival rate "of marginalstatistical significance, depending on the analysis used,"according to the FDA.

In September, an FDA advisory committee stated that therewas evidence for efficacy in patients with presumptive gram-negative bacteremia, particularly those in shock, but expressedconcerns about how to identify the patients likely to benefitfrom treatment. Nevertheless, it recommended approval of thedrug.

"Subsequent to this meeting, the FDA learned that certainCentocor staff were aware of the results of some interimanalyses prior to the final submission to the agency of therevised analytical plan," the FDA statement said. "Because ofthe potential for introduction of bias, results of analyses basedupon this plan cannot be accepted."

The additional "well-controlled trial" is necessary "beforefurther consideration" of the marketing application, the FDAsaid. "Populations who potentially may benefit from treatmentwith HA-1A (Centoxin) were identified in the first study, andadditional studies may focus on these groups," the agencyconcluded.

Centocor stock (NASDAQ:CNTO) fell $12.75 to close at $18.50.Analysts reached by BioWorld were not aware of the FDAstatement when the markets closed, and several held out hopethat the company would be able to repackage data fromcompassionate use or other trials.

An enormous short interest in Centocor might also have actedas a break on the stock's fall, according to Nomura ResearchInstitute analyst Jay Silverman. "Before today, you had almost7 million shares short out of 37 million outstanding shares, thelargest short interest in the OTC," Silverman said. He predictedthe stock could weaken after the shorts have covered. "It willbe completely in the traders' hands for the next few days,"Silverman said.

The company may face a cash crunch as it contemplatesadditional trials. Centocor is spending about $50 million perquarter, according to Koenig. The company had cash resourcesof $220 million in January, plus about $84 million netted fromits January offering in Tocor II, an R&D spinoff to fund smallmolecule drug development.

Tocor II units (NASDAQ:TOCRZ) also fell heavily, dropping $8.25to $17.

Rival Xoma Corp. (NASDAQ:XOMA) was down $4 to $14.50. LateTuesday, the Berkeley, Calif., company announced that the FDAhad determined that its initial clinical trial data wasinsufficient to warrant approval of its E5 sepsis treatment.However, the agency is still analyzing data from a second trial.

Centocor is carrying a large inventory of the drug, $57 millionat the end of 1991, according to Koenig. Centoxin has amaximum shelf life of 18 months. Koenig also said thatCentocor has no plans to lay off its 200-person sales force.

Centoxin sales in Europe totalled $8.3 million in 1991, andaccording to Silverman, sales on the continent have not beenbrisk. European concerns about high cost -- as much as $3,700per treatment -- have hurt sales, he said. Centocor expected1992 sales of $50 million, he said, but "now it looks closer to$15 million."

Before the announcement, Silverman had projected a $2.25 pershare loss for 1992, which included $125 million in globalCentoxin sales. "That will come down," Silverman said. EdwardKim of Robertson, Stephens changed his estimate to a $1.50 pershare loss on total revenues of $105 million in 1992, downfrom a 25 cent per share loss on $190 million in sales.

CentoRx, a cardiovascular drug that inhibits the aggregation ofblood platelets and inhibits clot formations, is the next majordrug in Centocor's pipeline. CentoRx is in Phase III trials as anagent to prevent reocclusion of coronary arteries during orfollowing angioplasty. It is undergoing Phase II trials for otherindications.

-- Steve Usdin BioWorld Washington Bureau

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