diabetes drug


can, in rare cases, cause a liver meltdown that can kill. Will two new drugs in the same class called glitazone have a similar glitch?



SmithKline Beecham

(SBH) - Get Report




Eli Lilly

(LLY) - Get Report

and Japan's

Takeda Chemical

are both on the

Food and Drug Administration's

fast track for approval. They could hit the market as early as the second quarter. But it's ominously unclear whether the drugs have the same problem as Rezulin.

SmithKline and Lilly/Takeda say their clinical trials have shown that the drugs work well, have the potential to save lives and have no more effects on the liver than placebos. No big surprise.

But doctors are cautious. David Nathan, the director of the diabetes center at Boston's

Massachusetts General Hospital

, says that at least initially, "I don't see how the glitazones will avoid the stain, if you will, of liver toxicity." (Nathan hasn't been a consultant for the three U.S. companies and doesn't hold stock in them.)

Diabetes experts around the country don't really know much about the two new compounds because little data have been published or presented at scientific meetings on either drug, especially Actos. The clinical trials for Avandia and Actos are too small -- any feasible clinical trial would be -- to catch whether either drug causes the liver problems. It won't be clear until each drug has been taken by tens of thousands of patients for almost a year.

That makes for one uncertain and scary investing environment. Nevertheless, investors are betting that the two new drugs won't cause liver toxicity, pretty much because of what the companies have told them. That doesn't mean that their neck hair isn't raised. Wall Street is expecting the two new drugs eventually to sell more than a billion dollars a year worldwide.

Similar hopes for Rezulin have been dealt a severe blow. The FDA's recommendation for mucho liver-enzyme testing, borne of the safety concerns, has hampered the Warner drug. The testing is inconvenient for doctors and patients (except when it saves lives) and increases the costs associated with the drug, which makes it harder to win reimbursement from insurers. Rezulin once was supposed to be a $1 billion-plus-a-year drug; now it looks like the roughly $750 million it sold in 1998 could be its peak.

None of the companies with a new glitazone has the luxury of having anything screwed up. Eli Lilly needs a drug to plug a potentially multibillion-dollar hole in its revenue stream from the looming patent expiration of its pick-me-up



Unfortunately, Actos won't be massively profitable, even if it is huge. Carl Seiden, the pharmaceutical analyst for

J.P. Morgan

, estimates that Lilly, which will co-market with Takeda in the U.S., will get a 20% to 30% royalty on sales, though other analysts have put the figure at 40% to 50%. Most analysts are carrying estimates of only around $25 million to $30 million for Actos' sales this year, even though it could be launched in the third quarter, but some analysts are weighing in with higher estimates. Lilly won't comment on the financial terms of its Takeda deal. (Seiden rates the company a market perform, and J.P. Morgan hasn't done any recent banking for Lilly.)

SmithKline, the


of British drug companies, needs a monster, profitable blockbuster to help it stay independent amid the peer pressure among European drug concerns to merge. Avandia will be more profitable for its purveyor than Actos, but in competing with Lilly, SmithKline faces the dominant U.S. diabetes sales force. SmithKline has no diabetes presence, though it has the second-highest-rated sales force in the country, according to a 1998 survey of doctors by health consultant



Both drugs come before an FDA advisory panel of outside experts in late April. Before that, this coming Friday, the same advisory panel will consider whether Warner's Rezulin is

safe enough to be kept on the market. According to Warner, there have been up to 35 deaths and liver transplants in patients taking Rezulin that have been "associated" with the drug. On Monday, the U.K.'s drug regulatory body rejected

Glaxo Wellcome's


application to reintroduce troglitazone, the chemical name for Rezulin. That could increase the pressure on the panel. So the panel's verdict could have an enormous impact on Avandia and Actos.

All three drugs belong to the class of compounds colloquially called glitazones, which work differently than current diabetes drugs and are meant for patients who have developed the disease as adults. (These patients are said to suffer from Type II diabetes.) The drugs lower blood sugar and make patients more sensitive to insulin, which can enable them to cut down on their injections.

When the liver sustains damage, due to anything from a night of heavy drinking to a toxic drug reaction, its cells leak and enzymes seep into the blood. Patients can be tested for enzyme elevations, which can provide an early warning that something is going wrong. Patients taking Rezulin must undergo liver-function tests monthly for eight months, every other month for the next four months and periodically thereafter.

The problem is that it cannot be predicted who will experience the serious liver problems that can lead to organ failure and death. If patients show rapidly rising enzyme levels, in theory, they can be taken off the medicine and their livers normally get healthy. Or maybe not. It's not clear that the testing is regular enough for those few patients, or if the meltdown is too rapid. It may be that for those patients, the serious events are as good as undetectable. Nor is it known whether the reactions are always reversible in those unlucky few who start down the severe reaction path.

And so the glitazone class doesn't have a smooth path to blockbuster status. It may take a while, but Wall Street is betting that the positive data and the lack of liver side effects in the trials will be enough to assuage the worry of doctors who feel burned by their Rezulin experience. That remains to be seen.

"Drugs are being released so quickly that the medical community can't assimilate all of them. We need quicker, more reliable safety data," says Matthew Riddle, a diabetes specialist at

Oregon Health Sciences University

in Portland, Ore. (Riddle holds no stock in Warner, Lilly or SmithKline. He has been a paid consultant to a variety of drug companies, including SmithKline and Warner.) Until that happens, "I don't think we are through" with the phenomenon of side effects emerging with drugs only when they become widely available. "It'll be true for whatever new class comes."