Costly error in health research?

Wes Rose had already lost some use of his hands and legs, and his neurologist told him the rest of his body would eventually follow. For people with Lou Gehrig's disease, it always does.

So Rose decided to fight, joining a drug trial that required him to take two medicines morning and night: little white capsules and a powder mixed with water.

But like a dozen trials before that one, it didn't work. Here's a possible reason why:

Scientists may have been targeting the wrong disease.

That's the controversial assertion of an international team of researchers, including a husband-and-wife duo at the University of Pennsylvania. If they're right, untold millions have been spent since the mid-1990s on research with little connection to a cure for the deadly illness, formally called amyotrophic lateral sclerosis (ALS).

The drugs in Rose's trial and others were chosen because they had improved symptoms in mice with rare mutations linked to the disease.

But in May, the researchers reported a curious finding. They identified abnormal protein clumps in patients with all forms of the disease, with one glaring exception: those who had the rare mutations. So they believe that rare type of the illness, which accounts for 1 to 2 percent of cases, is likely not representative of the disease as a whole.

"I think it's a completely separate disease," said University of British Columbia neuropathologist Ian Mackenzie, lead author of the paper in May's Annals of Neurology.

The ensuing months have drawn reactions ranging from eager curiosity to bitter skepticism from those in the field, some of whom have devoted the bulk of their careers to an area whose relevance is now being called into question.

Critics argue - and the paper's authors agree - that more lab work is needed to prove that the newfound protein is a key to the disease and not just a red herring. Jeffrey D. Rothstein, a prominent ALS researcher at Johns Hopkins University, wrote a review questioning whether the new theory was mere "patho-babel."

But already, labs around the country are racing to find the genetic underpinnings of the flawed protein, whose telltale clumps appear in patients' motor neurons as compact, round bodies or longer filaments. The new finding serves as a sobering reminder that ALS, like Alzheimer's and other diseases that strike the nervous system, remains years from a solution.

The illness - whose common name comes from its most famous victim, the New York Yankees' Lou Gehrig - afflicts up to 30,000 people in the United States at any one time. Some patients live for just a few years; others, like British physicist Stephen Hawking, endure for decades.

Rose, the drug-trial participant, also is a scientist, an assistant professor of biology at Arcadia University who also does research at Fox Chase Cancer Center.

In an eerie coincidence, Rose is studying the spinal cord and how it responds to viruses - an area he got into before learning that his own nervous system was in peril, 21/2 years ago. So he is well qualified to follow the change of direction on ALS championed by Penn's Virginia M.-Y. Lee and John Q. Trojanowski, along with Mackenzie. And it doesn't faze him.

"I think that's just part of the scientific process," said Rose, 35, who finds the new research promising.

During the six-month drug trial, he was required to take creatine powder, a supplement used by athletes. He also was given a second drug but was not told which one: either the antibiotic minocycline or celecoxib - an anti-inflammatory called a COX2 inhibitor.

"I had pretty high hopes for it," Rose said. But he didn't notice any improvement. He wears braces on both legs and walks with a cane, and lately has started to lose some of the strength in his grip.

The path to that drug trial, which ended in March, began 14 years ago.

As with Alzheimer's, ALS usually occurs in patients who, like Rose, have no family history of the illness. For these patients, the cause could be environmental factors or some sort of broad genetic predisposition - a mix of seemingly benign variations in DNA, not any overt mutation.

Still, researchers agreed it was important to identify mutations in those who had them.

That's because they view a complex disease like ALS as a river, a cascade of bodily events proceeding from upstream to downstream. The disease might begin in any of several tributaries - say, a mutation or an unknown environmental trigger.

Scientists felt that by finding mutations that caused the disease for some patients, they could figure a good spot to dam the river for all patients - slowing its progression no matter the initial source.

In 1993, after years of searching, a group of 33 scientists found a key gene.

It was still fairly early in the history of linking genes to disease, and the ALS announcement was made with great fanfare - described on the front page of the New York Times, The Inquirer and other newspapers.

They found mutations in a gene that holds the recipe for superoxide dismutase - an enzyme used to neutralize harmful "free radicals."

Within a few years, researchers had engineered mice with mutations in the same gene, and have since done about 100 drug tests on the animals. About a dozen individual drugs or combinations were promising enough that they were then tried on people.

But none had much of an effect; one actually seemed to make patients sicker. Another has received FDA approval, but it seems to provide modest benefits at best.

Then along came Lee and Trojanowski, who run the Center for Neurodegenerative Disease Research at Penn's medical school.

Last year, they and colleagues identified a faulty protein that was present in patients with ALS and also in those who had a certain kind of dementia. Called TDP-43, it was a scientific unknown mentioned in just a dozen papers, none of them having to do with the nervous system.

They found clumps of that protein in "sporadic" cases of ALS - the 90-plus percent of patients whose disease has no apparent genetic cause - and in most of the inherited, "familial" cases.

But upon further scrutiny, working with British Columbia's Mackenzie and others, they saw there were no clumps of TDP-43 in patients with superoxide mutations - the very form of the disease that had received the most study. Other researchers have since replicated the results.

The various patients (and mice) have similar symptoms. Their motor neurons die, and slowly they become paralyzed. But the underlying cause may be completely different, Trojanowski said, much as pneumonia patients all have fluid in their lungs regardless of what bug has infected them.

Trojanowski and Lee, who were married in 1979 and work side by side, say they do not fault science for pursuing the mutations found in 1993. They did some research on those mutations themselves.

"The best people with the best data can go down the wrong road," Trojanowski said.

And the pair are quick to say that even if drugs had been tested on mice with a more similar form of the disease, they still might not have worked on people. Studying human diseases in an animal model is an imperfect science at best; mice, for example, have a metabolism that's different from humans'.

"Cancer is cured all the time, and it doesn't always translate," Trojanowski said. "Mice are not little people."

Despite the new protein evidence, Don Cleveland, a leading ALS researcher at the University of California, San Diego, does not feel the original thinking should be discarded.

Among the drugs that have not worked on people, none were miracle cures for the mice, either. Most improved symptoms but did not prolong the animals' lives. He feels science just hasn't found the right drugs yet.

"The mouse was trying to tell us something and we weren't listening," Cleveland said.

Lucie Bruijn, science director of the nonprofit ALS Association, which has awarded $40 million for research in the last decade, said that the old and new thinking may both play a role and that both must be studied.

"You don't know where the opportunities will come," Bruijn said. "You can't shut doors before they've been absolutely exploited."

Various labs already are trying to find exactly how TDP-43 may lead to the disease. At Penn, for example, pathologist Vivianna Van Deerlin is searching her library of patient DNA to find mutations that might be responsible, a process that relies on sophisticated statistics to sort out the harmless variations from the deadly.

It is hard to say exactly how much has been spent on the old thinking, the gene identified in 1993. Records indicate the National Institutes of Health awarded $11.5 million to study this pathway in 2006, and $12.7 million the year before. At least 400 such projects have been funded since 1993; exact dollar amounts were not available.

Whatever the total, it is small in the context of overall NIH spending. More important to patients is the years that have passed with no cure.

Yet Rose, the drug-trial patient, is philosophical. As a researcher, he knows better than most how science can proceed in fits and starts as theories are proposed, tested, modified, and sometimes discarded.

Though they may not cause the disease that Rose has, he said the mutations found in 1993 were still important.

"It was either don't study ALS or use what we have," he said. "All that work was not for naught, for sure."