Discovering Secrets of Autoimmune Disease

reprinted from Issue 11, Fall 2010 of Frontiers of Medicine (PDF)

Jonathan Graf, MD, and Kara Lynch, PhD, discovered that most SFGH patients testing positive for cocaine had also been exposed to levamisole, a drug used to de-worm cattle and which has begun to be used to cut cocaine. Some of these patients developed a rare rheumatic disease which may be triggered by levamisole - a development which may hold clues for preventing other types of autoimmune disease.
photo by Noah Berger

In the summer of 2009, dermatologists at San Francisco General Hospital (SFGH) saw a patient with unusual purple-black patches of dying skin. They called rheumatologist Jonathan Graf, MD, to determine if the lesions might be caused by vasculitis, a rare rheumatic disease which causes blood vessel inflammation.

"Usually vasculitis affects not just the skin, but the kidney, lung and other organs," says Graf. However, the patient's condition appeared to be limited to just the skin - still dangerous, but puzzling. Also, blood tests revealed unexplained, sky-high levels of certain antibodies. Over the next few months, five other patients came in with similar symptoms and lab results, many of them with damaged or missing skin on their ears. Graf also heard about a few other cases in other cities. "This was a rare presentation of a rare disease," says Graf. "It begs the question, what was the common exposure?"

A Common Link

It turns out that all of the patients used cocaine. At the same time, the federal government released an alert that some cocaine was being cut with levamisole, a drug primarily used to de-worm cattle. Levamisole had been tested in humans decades ago as a treatment for colon cancer and other diseases, but was discontinued because of its toxicity. Graf read up on the medical literature, finding that levamisole had caused clotting disorders of the skin, particularly the ear. The drug can reduce the number of white blood cells to dangerously low levels, leading to flu-like symptoms and persistent infection.

There was no test available for levamisole, since it was no longer prescribed. Graf teamed up with Kara Lynch, PhD, associate chief of clinical chemistry and toxicology at SFGH. Lynch bought some levamisole, consulted a scientist who had researched the drug in the 1970s, and developed a customized test for the substance using a technique called liquid chromatography tandem mass spectrometry.

In October 2009, she re-tested every urine drug screen at SFGH which had tested positive for cocaine, and found that 88 percent of those also tested positive for levamisole. "That was surprising," says Lynch.

The Investigation Continues

The rate of patients presenting with these symptoms has since slowed, but Lynch and Graf are continuing to investigate more about what might be causing an autoimmune disease in cocaine-using patients. One of the biggest questions now is why doctors haven't seen hundreds of cases, since levamisole was so prevalent among cocaine users, and why all the patients seen at SFGH have been middle-aged women.

"Common autoimmune diseases like rheumatoid arthritis and lupus tend to affect women more than men, and they tend to be women in their 30s to 50s," says Graf. "It may be that these patients who came in had an underlying predisposition, but not the full genetic requirements, to develop an autoimmune disease. Maybe with an adjuvant like levamisole, you tip them over the top, and they develop this syndrome."

"Drs. Graf and Lynch have done a remarkable bit of clinical scholarship," says John Imboden, MD, chief of the Division of Rheumatology at SFGH. "Not everybody would see cases like this and put it all together. It shows what an academic environment can do when people bring in different areas of expertise to solve a puzzle like this."

Graf, Lynch and their colleagues are now comparing patients who develop the syndrome with cocaine users who do not, learning more about patients' inflammation levels, what antibodies and enzymes they produce, how they metabolize levamisole, and whether yet another cocaine adulterant might be involved in the cases which produce the vasculitis-like symptoms.

"This research could teach us something about what causes vasculitis, because right now we don't know," says Graf. "Could we prevent it, or develop better treatments? That's how a lot of medical discovery is made - by recognizing a pattern in one place, and applying it to another situation."

Frontiers of Medicine Main Page