The Drug That Could Have Been
When a young company called ImClone said it had a hot new cancer drug, its stock leapt on the bet that it would get an OK from the FDA. Now the stock has tanked, ImClone's charismatic CEO is indicted, and, Martha Stewart is in trouble, too. But the big losers are cancer victims. Here's the real story behind ImClone and why it failed to win the FDA's approval
No one listens to Sam anymore. Used to be, he would stroll into a New York society party and heads would turn. When he told tales of his plans to bring an important new cancer drug to suffering patients, friends listened raptly. "How soon would the new drug be approved?" they asked. Should they invest? Martha was one of the true believers. "Sam set up a situation," says a friend who jumped in, "where he personally got involved in people putting money into the company."
Sam, of course, would be Sam Waksal. Last week's indictment of the founder and CEO of ImClone on 13 counts of securities fraud and other charges added yet another chapter to the evolving summer scandal of seemingly boundless greed in America's corporate suites. That Stewart, the popular doyen of domesticity, has been implicated in the mess adds a nice frisson. Never mind that ordinary investors are out millions. Yet what has been lost in all the anger over corporate misfeasance is the fact that thousands of cancer sufferers for whom Waksal's drug promised a respite, and maybe even a cure, are left high and dry. When the Food and Drug Administration decided not to approve ImClone's new drug, Erbitux, it put a hole below the waterline of ImClone, and it also delayed the arrival of a potentially powerful weapon in the war on cancer, probably by at least several years. How that happened is an illuminating and cautionary tale, a story not just of fortunes lost, but of life and death.
It is also a story, a U.S. News investigation shows, of colossal failure--several, in fact. The FDA's system for bringing lifesaving drugs to market, a "fast track" process, utterly collapsed. The biotech division of the FDA that handled the Erbitux application from ImClone was manifestly not up to the task. The communication that should have taken place between the FDA and the drug maker, especially the guidance necessary to bring forth a new and useful drug, simply didn't happen. ImClone, eager for the many millions Erbitux approval would mean, tried to jump through regulatory hoops too quickly, ultimately failing to provide enough basic science to support its FDA application. Bristol-Myers Squibb, anxious to add a new cancer drug to its pipeline, failed to heed concerns raised by its own in-house advisers. In the end, in an irony that borders on the truly tragic, a study that showed that Erbitux might well be a highly effective aide in combating cancer ultimately sealed its doom.
Bright beginnings
The story of Erbitux might be said to have begun at the spring meeting of the American Society of Clinical Oncology in New Orleans in May 2000. Sam Waksal, 54, brash and charming as ever, and his more serious brother, Harlan, 49, were working the room. And they were hearing some very good news. Erbitux was being tested in clinical trials for several kinds of cancer at prestigious medical centers. One such trial was known as Study 9923. In that trial, Erbitux was combined with a proven chemotherapy drug called irinotecan being given to patients with colorectal cancer--a disease that kills 56,600 Americans annually. "Physicians involved in the study kept coming to our booth and telling us, `Your drug really works,' " says Harlan Waksal, who replaced his brother as CEO of ImClone in May. "We didn't know to what extent, but they were telling us that some patients were responding dramatically."
For the Waksals, it all made sense. Their drug appeared to inhibit tumor growth, and when used with chemotherapy, which kills tumor cells, delivered a solid one-two punch. John Mendelsohn, president of the M. D. Anderson Cancer Center in Houston, who originally had discovered the drug, had seen good effects on animals: Erbitux, when combined with chemotherapy, brought remarkable results against tumors that resisted chemotherapy alone. Now, Study 9923 showed that 20 percent of people with colorectal cancer who did not respond to chemotherapy improved when they took the combination. In the world of cancer remedies, 20 percent is a big number. Perhaps the FDA would grant Erbitux accelerated drug approval, the Waksals thought; they requested a meeting.
But a problem loomed. The trial at hand with the 20 percent results, "Study 9923," was small, with only 120 patients. And it only had one arm--that is, there was no test to see whether patients given Erbitux alone would respond.
Since 1997, the FDA has had in place a "fast track" process for drugs, intended to treat a serious or life-threatening condition for which there is no other cure--and this process allows the FDA to accept data that are less robust than its normal gold standard. Still, not everyone at the FDA thought that Study 9923 met the FDA's basic standards. In an internal meeting before sitting down with ImClone, Susan Jerian, the primary FDA medical reviewer, made notes that reflected her view: Study 9923 did not "meet accelerated approval criteria." She did not think the small, one-armed study would provide enough data for the FDA to review. ImClone, however, did not know this.
A key meeting
On Aug. 11, 2000, team ImClone arrived at the FDA's Woodmont One building off a busy, mall-bedecked highway in Rockville, Md., and took seats along one side of a long conference table. They included Harlan Waksal; three other members of ImClone's regulatory and clinical team, one of whom was Michael Needle, who was in charge of the clinical trials; and five consultants, practicing oncologists involved in the trial. Among them was Leonard Saltz, a world-famous oncologist at Memorial Sloan-Kettering Cancer Center in New York and the developer of a chemotherapy regimen for colon cancer that bears his name. Saltz was the lead author of the 9923 study. Seated opposite them were Jerian, FDA reviewer Richard Steffen, and their supervisor, Patricia Keegan, from the FDA's Center for Biologics Evaluation and Research, along with eight other agency officials. Keegan's division, known as CBER, handles biotechnology products, where some believe the future of cures lies, but it had limited experience with drug approvals.
Keegan was impressed with Study 9923 and its 20 percent response rate. She asked how the FDA could be sure that patients would not have responded to continued chemotherapy treatment without the addition of Erbitux. The company said it had "pictures"--that is, it could provide scans to show how tumors grew under chemotherapy alone and then shrank when Erbitux was added. If that were the case, said Keegan, the study design was "probably acceptable" for fast-track consideration. ImClone went home happy.
That was Keegan's call. But Jerian says she was not convinced that the study was good enough for FDA purposes. Richard Pazdur, the director of oncology review in the FDA's more experienced Center for Drug Evaluation and Research, says that relying on Study 9923 put the drug in jeopardy of being rejected. For starters, the Waksals presented backup animal studies to show that the drug worked best in combination. That was "a very risky venture," Pazdur said, because human studies can bring different results. ImClone's human study, 9923, did not even test whether Erbitux worked alone. That was important, because chemotherapy is toxic; perhaps patients could take Erbitux alone and avoid chemotherapy. The company needed to do a randomized trial looking at the combination of chemotherapy with Erbitux versus Erbitux alone, Pazdur says. "That would be the correct way of developing this drug."
By January, Keegan's team had come to the same conclusion. So one week after informing the company that Erbitux had been accepted for fast-track review, the FDA sent a letter requesting the additional randomized study. The company was floored. On January 26, a conference call was held with Keegan, her reviewers, and ImClone. According to minutes recorded from the call that were obtained by U.S. News, ImClone protested. It would be "unethical" to conduct a large study of Erbitux alone, the company said, because it believed the drug would not work for sick patients.
In addition, the company said it had left the Aug. 11, 2000, meeting with the understanding that only Study 9923 would be needed. "As there was no discussion regarding the need to supply data for [Erbitux alone] . . . ImClone has not pursued" that kind of trial, the company said. It had planned to study Erbitux alone later, in follow-up trials. FDA members now say that they raised the issue of whether Erbitux would work by itself at the August meeting, but meeting minutes do not reflect that discussion.
By the end of the call, the two sides had come to a compromise: ImClone would conduct a small study, in 20 to 50 patients, using Erbitux alone. Pazdur is now critical of that decision. "In order for that to work, you had to have a 0 percent [effectiveness] almost in the single agent Erbitux study," he says. "So in essence they were betting against their own drug to get that combination approved."
On March 27, Harlan Waksal, Leonard Saltz, and John Mendelsohn marched back into the FDA's Woodmont One building for a second face-to-face meeting. They argued over the second study of Erbitux, according to meeting minutes, with the FDA telling the company it wanted the study to be part of the official application and the company complaining that this was a late request by the agency. They then talked about other elements of the application, and ImClone began enrolling patients for the second study on April 1.
Pressure mounts
On the afternoon of Saturday, May 12, 2001, Leonard Saltz stood on stage in a huge convention hall at the American Society of Clinical Oncology conference in San Francisco. His face, beaming, was captured on several large projection screens. Saltz had big news: Erbitux, he said, used in conjunction with chemotherapy shrank tumors in 22.5 percent of colon cancer patients for whom other treatments had failed. The results were "knock-your-socks-off exciting," Saltz told the crowd. In the room were not just cancer physicians but also scores of investors looking for the next blockbuster. Wall Street analysts coming to the meeting expecting good news got what they were looking for.
Pressure from cancer patients to get Erbitux was also mounting. A 60 Minutes segment highlighted the drug's remarkable efficacy, but showed that most patients couldn't get the treatment. On June 20, in Washington, the House Committee on Government Reform held a hearing. "We tried to get Abigail into narrowly defined clinical trials but she did not qualify," said Frank Burroughs of his daughter, who had died two weeks earlier from head and neck cancer. "We worked very hard to acquire the drugs on a compassionate-use basis, but got nowhere." Compassionate-use programs allow patients not eligible for a clinical trial to use the drug, but ImClone did not have one. Sam Waksal told the committee that supply of the drug was limited; committee members told the FDA to hurry up and get the drug approved.
Meanwhile, investors were seeing dollar signs. ImClone's stock price jumped from $27 in April to $53 in July. At the same time, pharmaceutical giant Bristol-Myers Squibb was talking to ImClone about buying a 20 percent stake in ImClone and the rights to market the drug. (The deal would eventually earn the Waksals $111 million.) Bristol-Myers Squibb was desperate to find a new drug, especially in the area of oncology. It was doing well selling to cancer specialists, but with its blockbuster cancer drug Taxol recently off patent, the company needed a new product.
Yet Bristol-Myers medical experts were cautious. "On the whole this remains a very high risk opportunity," Bristol-Myers Senior Vice President Laurie Smaldone wrote in a June E-mail to senior executive Peter Ringrose. She noted that the Erbitux-only trial had not yet been completed and that "there is no agreement that we could find that is reassuring regarding activity level needed for approval." (Indeed, there never was such agreement.) A Bristol-Myers analysis also found the response rate in Study 9923 to be 13 percent instead of ImClone's stated 22.5 percent. But on September 4, Bristol-Myers senior executives recommended a "go." Smaldone now says the company viewed the FDA's consideration of a combination study for a fast-track approval as a "validator." The $2 billion investment deal between Bristol-Myers and ImClone was a public shot in the arm for Sam Waksal's claims. "We don't have any worries about the FDA," Waksal told a group of investors on September 19. Tony Russo, an investor-relations expert at Noonan Russo Presence, says the deal with Bristol-Myers made ImClone suddenly seem "like a very safe bet." And in October, the investment bank UBS Warburg predicted that Erbitux sales could reach over $1 billion by 2006, assuming it was approved by the FDA in the first half of 2002.
The FDA struggles
The deal spurred patient advocates to press ImClone. In November, in a series of conference calls with advocacy groups, ImClone said it might begin a program within 90 days: It would allow 30 patients per month to gain access to Erbitux outside of clinical trials. Frank Burroughs, cofounder of the Abigail Alliance for Better Access to Developmental Drugs, says that seemed low. Still, he said, "I thought it was better to get one started, and that 30 patients is better than none."
But the FDA was struggling. Two new reviewers, George Mills and Lee Pai-Scherf, had arrived on the scene. They saw that some patients' CT scans were missing, that patients who weren't eligible were in the trial, and patients were being given varying doses of chemotherapy--all skewing the results. James Abbruzzese, the chairman of the gastrointestinal medical oncology department at M. D. Anderson, says he warned the company about this problem. "The writing [in the protocol] was not precise," he says. He says he had an E-mail exchange with ImClone over the issue and that after he raised objections, the E-mail "just stopped." M. D. Anderson declined to participate in the trial. One investigator who did work on the trial saw another flaw: Because of media attention, patients were "calling from all over the country begging for it." Perhaps, this investigator said, doctors felt compelled to relax eligibility rules so desperate patients could get the drug.
By November 30, before the results of the Erbitux-only study were even submitted, the FDA's reviewers had decided to recommend what the FDA calls a "refusal to file" letter. Keegan would later concur. By not accepting the drug's application, its approval would be stalled, maybe for years. At this point, however, ImClone knew only that the FDA saw some problems in its study, problems its vice president for regulatory affairs, Lily Lee, had promised to remedy. So on December 4, ImClone turned in the results of its Erbitux-only study. Now it was the FDA's turn to be floored. The study showed the drug shrank tumors in six of the 57 patients, more than 10 percent. Ten percent is noteworthy--thus the study suddenly threw into question whether the Erbitux-chemotherapy combination was needed. Maybe Erbitux worked well enough by itself!
Keegan had had enough. She had trusted the company, believing that it would show that Erbitux worked only with chemotherapy. "I felt misled personally," she said. ImClone, however, didn't see the problem. Why not approve Erbitux for combination use, and continue studies to see if could also be used alone? Still, the company appreciated the FDA's position at this point: The agency felt it had been fooled.
Breakdown
On December 4, ImClone's Vice President Lee hopped a train from New Jersey to make a surprise visit to FDA headquarters. She was worried about the application. The FDA, however, was noncommittal. All options were open, reviewers Mills and Pai-Scherf told Lee: ImClone could submit additional information, or the company could withdraw it. The FDA could accept the application or refuse to accept it. Mills presented the options but was careful not to put extra weight on any one of them. Although both reviewers say that Lee asked whether the FDA would refuse the application, she was not given an answer.
Eight days later, the agency teleconferenced with ImClone and Bristol-Myers Squibb. It was an ominous discussion, but still the agency did not reveal that it had made a decision. By December 20, however, when ImClone called to ask about the status of Erbitux, the agency told the company not to make contact again. "We were informed at the time that a decision had been made and that it would be coming sometime in the next week," Harlan Waksal told a congressional committee. Bristol-Myers, frantic, asked a consultant lawyer to find out what was happening at the FDA. On December 24, Sam Turner of Bennett, Turner & Coleman called the FDA's Pazdur, who told him that the refusal-to-file letter was coming. Sam Waksal called friends at the agency but could not find someone to intervene. On December 28, at 2:55 p.m., the FDA faxed the refusal-to-file letter to ImClone. In the two days before the letter arrived, Waksal's daughter Aliza, his father, and Martha Stewart unloaded ImClone stock worth, respectively, $2.5 million, $8 million, and $227,000. Sam Waksal's indictment charges that he had unsuccessfully tried to dump $4.9 million worth himself and that he had tipped others off. He has pleaded not guilty; a plea bargain is still possible, sources say. Stewart, meanwhile, is under investigation for insider trading and making false statements about her stock sale.
As word of the FDA's refusal spread, investors dumped too. Within weeks, ImClone had lost nearly $1.5 billion of market value. In a December 31 conference call, Sam Waksal tried to downplay the seriousness of the problems, but on Jan. 4, 2002, The Cancer Letter, a Washington, D.C., newsletter for oncologists, published the FDA's letter--and investors realized they'd bought a lemon. Indeed, investors saw for the first time the many hurdles the FDA had set for Erbitux's approval--and how sloppy ImClone's evidence really was. They felt they'd been conned. But not all were had. Lance Willsey, a physician who at the time helped run a hedge fund, had decided early on to short-sell the stock, and the fund profited when ImClone tanked. He just did not believe the FDA would approve the drug based on a combination treatment. David Hines, president and research director at Avalon Research Group, says one warning sign was the Waksal brothers' promotional style. "When I hear management make overstatements, I get interested," he said. He put out his first sell recommendation in April 2001.
Still waiting
As a congressional investigation sifts through the debris, several things become clear. Certainly, congressional sources say, ImClone should have run better studies and not hyped the drug. But the FDA bears a big share of the blame. It had put the drug's review in the hands of its biologics division, whose experience in dealing with companies was limited. Pazdur, director of oncology in the FDA's old-school drug division, says when his reviews are not going well the company is urged to quietly withdraw its application--thus avoiding volatile market reactions to bumps in the regulatory road. Pazdur's division routinely meets with new drug applicants early in the process to design trials the FDA will accept. His division had such protocol assessments for 120 new drugs; the division that handled Erbitux had done one. On June 27, U.S. Reps. Billy Tauzin of Louisiana and James Greenwood of Pennsylvania, both Republicans investigating the ImClone affair, demanded that high-level FDA officials meet with their committee to discuss how to fix problems with fast-track approval.
Cancer patients are still hopeful that Erbitux can be approved. Despite the plunge in its stock price, ImClone is flush. It has raised $400 million in cash through stock sales and the Bristol-Myers deal, which it vows to use to build an Erbitux plant in New Jersey. In February, ImClone and Bristol-Myers met with the FDA, and now they are pushing ahead with new clinical trials to get Erbitux approved. That could take a couple of years, however. Meanwhile, advocates say they are frustrated with the stalled progress of the compassionate-use program. In January, the company was saying it needed to meet with the FDA in February before it could consider starting the program. In June, Abigail Alliance founder Burroughs says, he had more conversations with Harlan Waksal but got nowhere. By July, ImClone was promising it would apply to start the program in the fall.
Meanwhile, only patients like Rita Essigmann, who is in a clinical trial, are able to get Erbitux. Every Wednesday, the 81-year-old colon cancer patient travels from her home on Cape Cod to the Dana-Farber Cancer Institute in Boston. After a checkup with her doctor she sits in a recliner in the hospital's "infusion room" and receives her dose. She says it seems to be helping. "I have no symptoms and I feel better," she says. "I've been told that I look well, too."
THE COMPANIES
Harlan Waksal (above left), now CEO of ImClone, talks with his clinical trials chief, Michael Needle. The likelihood of ImClone's drug Erbitux being approved caught the attention of Bristol-Myers Squibb, which struck a $2 billion deal with ImClone. But Bristol-Myers Senior Vice President Laurie Smaldone (right) expressed her doubts. At far right: friends Sam Waksal and Martha Stewart at Lincoln Center in New York.
The rise and fall of a hot prospect
ImClone's stock price hinged on the chance that its cancer drug, Erbitux, would get FDA approval. Investors made and lost millions by taking that bet.
[Chart is not available.]
[Chart labels]
March 6, 2000
Stock price peaks at $84.56 on the buzz that ImClone has a new drug
Aug. 11, 2000
ImClone's first meeting with the FDA
May 12, 2001
ImClone's study makes big news at conference
Sept. 19, 2001
Bristol-Myers Squibb announces deal with ImClone
Dec. 28, 2001
FDA faxes refusal letter to ImClone
0
10
20
30
40
50
60
70
80
Oct. '99
Jan. '00
Jan. '01
Jan. '02
THE REGULATORS
Richard Pazdur, an FDA division chief, says the Erbitux matter could have been handled better. Patricia Keegan (center), supervisor of the Erbitux application review team, tried to work with ImClone officials. But Susan Jerian, on Keegan's team, remained skeptical that ImClone's studies would pass muster.
THE PATIENT
Rita Essigmann, an 81-year-old colon cancer patient, travels each week to Boston where she is enrolled in a cancer trial. Patients not enrolled in such trials, however, cannot get Erbitux.
With Pamela Sherrid and Katherine Hobson
This story appears in the August 19, 2002 print edition of U.S. News & World Report.