Ortho Pharmaceutical, the Johnson & Johnson subsidiary that sponsored (but didn't sell) Suprol, did studies to find out if it caused tumors in rodents.

Documents in Food and Drug Administration files, obtained by a House Government Operations subcommittee and made public at a hearing in May, show that the studies persuaded FDA pharmacologist Conrad Chen that the drug induced tumors of the adrenal gland in rats and liver tumors in mice.

Troubled, Chen advised FDA approval of Suprol "only if the clinical benefits {to users} would outweigh this possible risk." His supervisor, Dr. David J. Richman, emphatically agreed, urging that "the findings be given adequate weight on the risk portion of {the} benefit/risk equation."

On a higher rung of the FDA ladder, Dr. John F. Palmer, director of the Division of Oncology and Radiopharmaceutical Drug Products, also saw an "increased risk."

The chorus of doubts culminated in a bureaucratic experiment: Palmer and Dr. John F. Harter, who leads the unit in charge of NSAIDs (non-steroidal anti-inflammatory drugs), drafted four alternative letters for Dr. Robert J. Temple, director of the Office of Drug Research and Review. He was to sign one of them and send it to Ortho, according to testimony at the hearing. Citing the tumor findings, three of the drafts disapproved Suprol.

Temple might have been expected to sign one of the disapproval drafts, having told subcommittee chairman Ted Weiss (D-N.Y.), at a 1983 hearing on a different drug: "It is widely stated by people who study tumorgenicity that a finding of benign tumors {in rodents} represents some carcinogenic risk for man."

That testimony concerned Zomax, a prescription pain reliever taken by millions before its withdrawal in 1983 by McNeil Pharmaceuticals, the J&J subsidiary that now wanted to sell Suprol. By then, the FDA has estimated, Zomax was associated with 2,200 reported allergic reactions, including 503 that were life-threatening and at least 15 that were fatal.

In the Suprol case, Temple, rejecting the advice of the FDA experts, signed none of the disapproval drafts. Rather, he required merely that the labeling recognize an increased incidence of "benign" liver tumors.

Although the Suprol rodent studies indicated it to be more likely than Zomax to cause tumors, the labeling neither warned against chronic use, which Temple had required the Zomax label to do, nor mentioned the studies showing adrenal tumors in rats. The prescribing instructions did offer a comforting assurance: "No evidence of carcinogenicity was found in doses as high as 40 mg/kg/day in the rat and mouse."

It was the labeling's "remarkable" omission of the possibly enhanced risk of human cancer that most upset Weiss, who said it "hedged and fudged" the issue. Temple claimed the evidence of human carcinogenicity was "more uncertain" than for Zomax.

At the May hearing, after Temple had testified, the subcommittee heard testimony that was complimentary to Ortho and devastating to the FDA from veterinarian Dr. M. Adrian Gross. He is a senior science adviser in the Environmental Protection Agency and a former associate director for preclinical investigations in the FDA.

One of his major concerns was an FDA scientist's finding of internal inconsistencies in Ortho's data on the incidence of rodent tumors. Always, Gross testified, such a finding should trigger "an immediate intensive audit," because inconsistencies have sometimes been a clue to sloppy and even fraudulent testing. Instead, Temple sent Ortho a letter asking the company to audit its own data.

Gross scorned the request for a "self-audit" as a "highly irregular procedure," one that "shifts the burden for the regulatory responsibility of the agency onto the regulated industry itself." It makes "about as much sense as self-tango or self-love or self-psychoanalysis," he said.

Gross denounced other alleged departures from sound regulation in the letter. For example, the FDA had asked Ortho "to audit merely a 'sample' of the data." Allowing such a self-audit also "signaled" -- in defiance of "the truth" and of "wisdom and propriety" -- that the FDA had "pretty much made up {its} mind that these adrenal tumors don't mean very much" and that Temple apparently had judged the adrenal as well as the liver tumors all to be benign.

"This wasn't true at all," Gross testified. He cited a "definitely malignant" liver tumor that "had metastasized, spread to the lungs . . . In this case, there was absolutely no review whatsoever of the tissue slides by the agency."

Gross has been a feared expert witness for plaintiffs in product-liability lawsuits against drug companies. But Ortho's response to the FDA's letter drew his unqualified praise.

He said that on Oct. 17, 1985 -- a "very fast" eight days after Temple's letter went out -- Ortho submitted a report that "stated exactly the significance of a great number of tumors, some malignant, some benign; and the significance was quite high."

Yet at the FDA, Gross said, a surprising development followed six days later: Temple told Palmer in a memo that Ortho's own tumor findings were not significant. -- Morton Mintz