Skip to main content

Cancer Experts Say ImClone's Tests Were Flawed

Doctors who reviewed the leaked trial design of cancer drug Erbitux say it needs new trials.

Updated from 12:16 p.m. EST

Billionaire financier Carl Icahn's interest in

ImClone Systems


is an interesting sidelight to the really important issue facing the beleaguered biotech company: What's it going to take to get the cancer drug Erbitux approved by the Food and Drug Administration?

ImClone and the FDA will sit down together on Feb. 26 to discuss details of the agency's refuse-to-file letter for Erbitux, issued at the end of last year.

This meeting will be closely watched because it should determine how quickly ImClone can get Erbitux back on track. Will it be able to salvage its completed clinical trial? Or will it have to start from scratch?

Three independent oncologists, working from newly obtained information, believe the data ImClone has already collected on Erbitux are not sufficient for approval, according to the Feb. 15 edition of the

Cancer Letter

newsletter. Another oncologist who was contacted by

and had reviewed the FDA's letter to the company agreed. This doctor, who asked to remain anonymous, says he believes ImClone will have to conduct new trials or rely on one by European partner

Merck KGaA


Poor Design

The doctors found the trial design put together by ImClone to test Erbitux was faulty, raising serious doubts about the eligibility of patients in the trial as well as the drug's reported results.


Scroll to Continue

TheStreet Recommends

Cancer Letter

obtained a confidential copy of the clinical trial protocol, or plan, used by ImClone to conduct its Erbitux study. It then asked the oncologists -- all experienced cancer drug researchers -- for their assessment, all of whom said the protocol shows the study was flawed from the very beginning.

"Overall, this is a protocol that asks the wrong questions, and then is not tightly written and efficient," says Dr. Otis Brawley, a professor of oncology at Emory University Winship Cancer Institute and a member of the FDA Oncologic Drugs Advisory Committee, in the

Cancer Letter

. "The protocol generates far more questions than it could ever answer. It is a blueprint for the production of vague findings."

The need for new Erbitux trials will not surprise ImClone critics. After all, the Dec. 28 refuse-to-file letter sent to the company from the FDA states the agency's belief that additional studies will be needed because the existing trial was "not adequate and well-controlled." And ImClone's severely depressed stock price to a large extent reflects this.

ImClone closed Friday up 84 cents, or 4.8%, to $18.44 a share, buoyed on news that Icahn filed for permission to purchase $500 million in ImClone common stock. Icahn and ImClone CEO Sam Waksal are tennis buddies and occasional business partners, so the billionaire financier's interest in ImClone was being viewed as a signal of support for the biotech company.

But ImClone executives have publicly insisted -- with some caveats -- that the biotech's completed trial was well-run and sufficient for approval. CEO Sam Waksal admits to "screwing up" but says the mistakes are fixable from data already compiled by the company and its researchers.

Congress, the Justice Department and the

Securities and Exchange Commission

are investigating ImClone to determine whether the company misled investors about Erbitux. The new revelations about the apparent inadequacy of the Erbitux clinical trial will likely be weighed as evidence in these inquiries.


The Erbitux clinical trial conducted by ImClone tested the hypothesis that colon cancer patients who had already failed irinotecan, an existing chemotherapy treatment, would respond to a cocktail of Erbitux plus irinotecan. ImClone enrolled 120 "irinotecan refractory" patients in this trial and reported a 22.5% response rate, a very good result considering how sick these patients were.

But the trial protocol obtained by the

Cancer Letter

suggests that the entry criteria for patients were poorly defined. In other words, the definition of "refractory" was so broad that some patients could have actually responded to irinotecan alone, and not the combination of Erbitux and irinotecan.

"Indeed, the fact that this drug received a lot of hype and positive press means a physician hoping to do the best for his or her patient might use the vagueness of the inclusion and exclusion criteria to their advantage, enrolling patients the authors of the trial would have excluded." Brawley tells the

Cancer Letter


ImClone has maintained that it put together an independent review committee of oncologists and radiologists -- known in medical parlance as the IRAC -- to verify both the patients' initial refractoriness to irinotecan as well as their response to the combination therapy. In fact, the company says the main reason for the FDA rejection was because the company did not provide enough documentation from the IRAC's findings.

But Dr. Howard Ozer, director of the Oklahoma University Cancer Center, says the formation of the IRAC panel was not part of the original trial protocol and wasn't actually done until after the trial was started.

ImClone "was trying to put some kind of overall quality control on the protocol, but this is a retrospective manipulation, and that's inappropriate methodology in clinical trial," Ozer says in the

Cancer Letter


"A retrospective analysis introduces a bias. In this case, it's equivalent to introducing a sliding scale grading system after you've given a test to a class and found that only 20% passed, and now you go back to fix it by assigning a sliding scale and letting 50% or 70% -- whatever your target is -- pass the test," he adds.

ImClone also seems to have changed the type and number of patients enrolled in the trial midstream. The original plan called for the trial to enroll 49 colon cancer patients whose tumors were growing, as well as an equal number of patients whose tumors were stable.

But somewhere during the course of the trial, the study was changed to boost the number of patients with growing tumors to 120, according to the

Cancer Letter

. ImClone has never discussed publicly the results of the 49 patients with stable disease. That suggests the data submitted to the FDA consisted of only a subset of patients in the trial -- which would be a big red flag.

Drug Works, Supporters Say

ImClone supporters shrug off these latest revelations, believing that there is enough anecdotal evidence to show Erbitux works.

"I still believe that there is a 99% chance that the FDA decides to approve Erbitux in a few months," says Hemant Shah, an independent drug analyst with HKS & Co.

Shah adds that every oncologist with experience with Erbitux believes strongly it's an approvable cancer drug, despite FDA concerns and problems with the clinical trials. And that includes, most prominently, Dr. Leonard Saltz of Memorial Sloan-Kettering Cancer Institute, one of the world's most renowned oncologists and Erbitux's lead investigator.

"My father told me to have faith in great men, and Dr. Saltz is a great man," says Shah.