Updated from 11:24 a.m. EST

Nearly a dozen patient deaths have been recorded in an ongoing clinical trial of Celgene's ( CELG) experimental drug Revlimid, TheStreet.com has learned. If confirmed and attributable to the drug, the deaths could delay, or derail, Revlimid's chances for U.S. approval, which would force investors and analysts to reduce forecasts for the company's near-term earnings and sales growth.

Dr. Alan List of the H. Lee Moffitt Cancer Center in Tampa, Fla., and the lead investigator in Celgene's Revlimid MDS study, confirmed there were patient deaths. But he would not disclose a precise number because he is bound to a Celgene confidentiality agreement. List is a paid consultant to Celgene.

"There were some deaths in the study but very few were treatment related," says List. "Remember, we have a patient enrolled in this study who is 96, and the median age of patients is in the 70s, so the expectation for survival for one year is not high." The clinical trial in which the deaths occurred is testing Revlimid as a treatment for patients with a low-grade, low-risk form of myelodysplastic syndrome, or MDS, a cancer-like bone marrow disease.

Celgene shares fell on the news. They were recently down $1.67, or 5.7%, to $27.69 on eight times normal volume. The stock hit a three-year high of $32.58 on Nov. 11, adjusted for a 2-for-1 stock split in October.

There are 148 patients enrolled in the Revlimid study, which would put the death rate at just over 7% if the reports of 11 deaths are accurate.

Celgene's chief spokesman Brian Gill, asked to confirm or deny the patient deaths in the Revlimid study, said, "I, Brian Gill, am not aware of any deaths. I'm not saying there are or aren't any deaths." Gill declined to offer a formal Celgene response.

Gill added that a data safety monitoring board is charged with overseeing the safety of this Revlimid study in MDS patients. All serious adverse events, including patient deaths, in the study are supposed to be reported to the safety board, which is then charged with determining whether or not these adverse events are related to Revlimid.

"If the data safety monitoring board finds a correlation between deaths and the drug, then the study would be stopped," said Gill, adding that none of the ongoing Revlimid studies have been stopped.

But List acknowledged that the data safety monitoring board is being run by Celgene and several investigators involved in the study, including himself. There is no external, independent body making a determination of Revlimid's safety in the MDS study, which is not atypical for a phase II study.

Celgene has told investors that it expects to use data from this phase II study to seek marketing approval for Revlimid from the Food and Drug Administration during the first quarter of 2005. But this plan could be in jeopardy because 11 patients in this low-grade MDS study, all of whom took Revlimid, have died, according to a fund manager who learned of the deaths after speaking with a clinical investigator participating in the study. (The fund manager, who requested anonymity, has a short position in Celgene.)

Blood Cancer Is Real Prize

Celgene is betting on a quick FDA approval of Revlimid in low-risk MDS patients as the fastest way to get the drug into the hands of doctors and, of course, to begin generating revenue from the drug.

Patients diagnosed with a low-risk form of MDS represent a small commercial market opportunity. For Celgene, however, the real prize is the use of Revlimid as a treatment for multiple myeloma, a blood cancer. Many people believe that Celgene has been rushing ahead with a Revlimid filing for low-risk MDS to get the drug approved in 2005 so it can then be used off-label by doctors to treat multiple myeloma, which represents a much larger and more lucrative commercial market. Of course, Celgene won't acknowledge this because it is not allowed to promote a drug off-label.

The company is conducting pivotal phase III studies of Revlimid in multiple myeloma, with data expected next year. If positive, Celgene says it would file with the FDA for this indication in late 2005, which means the drug would not be approved until 2006.

It's not uncommon for patients to die while participating in clinical trials, including patients who die for reasons unrelated to the drug being studied. The specific causes of the deaths in the Revlimid study are not yet entirely clear and won't be until Celgene makes public data from the study. But the deaths are concerning, in any case, because the patients in this study are all diagnosed with a low-grade form of MDS in which parts of the human chromosome 5 are missing. MDS patients with so-called 5q minus deletion have a disease that progresses slowly and generally does not represent a near-term risk of death.

Scale of Risks

The merits of an experimental drug always have to be judged with a scale of risks and benefits in mind. Early data on Revlimid in patients with the 5q minus form of MDS have shown strong efficacy. After taking Revlimid, many of these patients no longer required blood transfusions or red blood cell-boosting drugs to keep the disease at bay. In addition, doctors discovered that patients no longer had cells in their body that were missing the 5q gene sequence.

Efficacy data from this current uncontrolled phase II study are expected to show similarly strong results, which Celgene believes will convince the FDA to approve the drug under an accelerated protocol -- that is, without first requiring Celgene to conduct a larger, controlled phase III study. Whether this efficacy benefit outweighs any safety and/or toxicity risk associated with the drug will be a matter for the FDA to decide.

In prior studies, for instance, Revlimid has been associated with serious cases of myelosuppression, a condition in which bone marrow activity is decreased, resulting in lower counts of red blood cells, white blood cells and platelets. Patients with myelosuppression more easily succumb to infections. Myelosuppression is a known side effect of many cancer drugs.

List agrees that MDS patients with a 5q minus deletion are generally lower risk and have a better prognosis than other MDS patients. But that doesn't mean 5q minus patients aren't sick or at risk for death; he cites statistics from medical literature that show these patients, even though they have a low-grade disease, can still have an annual mortality rate as high as 10%.

If List's historical mortality statistics are accurate, then the 7% death rate in the Revlimid study -- again, assuming that analysis is confirmed -- may not be all that unusual. But there are other published medical studies of MDS 5q minus patients that show a longer survival. A study published this year in the medical journal Leukemia by Dr. AAN Giagounidis of St. Johannes Hospital in Germany found median survival of 146 months, or more than 12 years, for this population of MDS patients. That's roughly double the median survival implied by List's 10%-per-year annual mortality statistic.

The best way to measure efficacy and safety of any drug, of course, is to conduct a large, randomized, controlled study in which some patients receive the drug and others receive a placebo. However, that is something Celgene chose not to do with Revlimid in these low-grade MDS patients. If the death rate in the Revlimid study does become a point of concern for regulators, the lack of good control data in which to assess Revlimid's safety might be the factor that forces the FDA to delay approval.

The outlook for Revlimid in low-grade MDS patients is further clouded by the fact that these patients can be treated with blood transfusions or with Vidaza, a drug marketed by Pharmion ( PHRM) that was approved earlier this year for all forms of MDS.

Celgene spokesman Gill says it is a mistake to consider 5q minus MDS patients less sick or not in urgent need of new therapies like Revlimid, and he cites a specific example: An MDS patient with the missing 5q gene identified by Celgene required weekly blood transfusions in order to keep his MDS under control. After taking Revlimid for two years, this patient is completely free of the need for transfusions and expects to live a normal, healthy life as long as he continues to take Revlimid, says Gill.

List believes Revlimid is an active drug and approvable by the FDA. "This will be a good package for possible approval, there is nothing excessive here at all."

Celgene has previously stated that it would release data from an analysis of this phase II Revlimid study in 5q minus MDS patients at a corporate event planned in conjunction with the American Society of Hematology annual meeting, which runs Dec. 4-7. But since this particular Revlimid data presentation will not be under the formal auspices of the medical meeting, the breadth and depth of the information Celgene will release is not clear.
Adam Feuerstein writes regularly for RealMoney.com. In keeping with TSC's editorial policy, he doesn't own or short individual stocks, although he owns stock in TheStreet.com. He also doesn't invest in hedge funds or other private investment partnerships. He invites you to send your feedback to adam.feuerstein@thestreet.com.