A granulocytosis can kill you, but its symptoms are frustratingly broad. Some people's throats close up. Some people get diarrhea. Some people get skin infections, sores in their mouth or anus, or just a fever. Some people have it, don't know it, and get better without seeing a doctor. Some people don't see a doctor until it's too late.

Basically, agranulocytosis is a catastrophic crash in a person's immune system, which can turn a zit, a scratch, or even the bacteria that normally live in and around your body into a life-threatening infection. In one vividly described case from the 1920s, an otherwise healthy 40-year-old woman came down with a mysterious fever. Over the next nine days, under the care of baffled physicians, she sprouted "brownish papular eruptions" all over her face and body, necrotic abscesses on her neck and buttocks, and "a greyish-green dirty membrane" covering her mouth and throat with "scattered small greyish ulcers." In one cubic millimeter of blood, her doctors found 4,000,000 red blood cells but only 1,000 white blood cells. Then, after a blood transfusion, she died.

Agranulocytosis is rare and typically caused by medications: Antibiotics, gold salts (to treat arthritis), and some anti­psychotic drugs can trigger the crash. But lately, doctors have been seeing more and more cocaine users with mysterious cases of agranulocytosis linked to a mysterious cutting agent called levamisole. Levamisole was discovered in 1966 and studied for its ability to rev up the effects of chemotherapy drugs and people's immune systems. It also turned out to work wonders with intestinal worms. Levamisole is an immunomodulator, meaning it can either strengthen or weaken your immune system, depending on your genes and what other drugs you might be taking. But too many patients came down with agranulocytosis, the studies were discontinued, and the FDA withdrew its approval of the drug.

One of the last studies on levamisole use in humans was in 2001, when Iranian researchers gave the drug to a group of girls who lived in crowded, unhygienic conditions with uncontrollable lice infestations. According to the International Journal of Dermatology, a 10-day course of levamisole tablets was "completely effective": The girls took the drugs, and the drugs poisoned the lice. (The study didn't mention whether the drugs poisoned the girls.)

These days, levamisole is mostly used by farmers to deworm cows and pigs—and, for some reason, it's also used by people in the cocaine trade. The DEA first reported seeing significant amounts of levamisole-tainted cocaine in 2005, with 331 samples testing positive. Then the numbers spiked: The DEA found 6,061 tainted samples in 2008 and 7,427 in 2009. One DEA brief from 2010 reports that between October 2007 and October 2009, the percentage of seized cocaine bricks containing levamisole jumped from 2 percent to 71 percent.

Which is not only sudden, but odd. Levamisole is not like other common cutting agents—sugar, baking powder, laxatives, etc.—in three important ways:

1. It's more expensive than other cuts.

2. It makes some customers sick.

3. It's being cut into the cocaine before it hits the United States.

This last mystery is the most puzzling. Typically, smugglers like to move the purest possible product—less volume means less chance of detection—and cut their drugs once they cross into the United States.

So what's the incentive to use a relatively expensive cut of something that makes your customers sick and increases your smuggling risk? Even stranger: The cocaine trade, in both smuggling and production, has fragmented in recent years (more on that in a minute). If there's no central production, how did hundreds and hundreds of independent shops come to use the same unusual cutting agent?

Nobody seems to know, including experts I spoke with on both coasts of the United States: doctors, scholars, chemists, think-tank fellows, research scientists, federal and state public-health analysts, law enforcement agencies from the Seattle Police Department to the DEA, and even people who work in and around the drug trade. Everyone has theories, but nobody has answers.

It's a mystery.

What We Do Know

Some people are getting sick from levamisole and a few have died, but it's impossible to pin down exact numbers. In April 2008, a lab in New Mexico reported an unexplained cluster of 11 agranulocytosis cases in cocaine users. In November 2009, public health officials in Seattle announced another 10 cases. The CDC began a surveillance program in eight states.

During levamisole's early clinical trials for cancer and autoimmune disorders, around 10 percent of the patients developed agranulocytosis. If the nation's cocaine supply is so thoroughly tainted, why aren't 10 percent of cocaine users going to hospitals with unexplained infections?

"Maybe 10 percent are experiencing pressure on their neutrophils," says Dr. Phillip Coffin of the University of Washington, who has studied drug use in New York City and Seattle. (Neutrophils are the type of white blood cell wiped out by agranulocytosis.) "But only a proportion of them are getting sick enough and using enough that they come to our attention. And an even smaller proportion of those people are coming to the attention of physicians who are aware of the cocaine-levamisole problem. There are many steps in the pathway that have left such a small number of cases being reported."

The problem might be—and probably is—larger than we know. And, because of budget crunches, last month the CDC abandoned its surveillance program in Washington State. This is worrisome not only for people who've already gotten sick and are likely to get sick again (doctors at Harborview have reported seeing the same patients multiple times for agranulocytosis), but because levamisole has a cumulative effect: The more you're exposed to it, the more likely you are to get sick, and even if you've had levamisole-tainted cocaine and not gotten sick doesn't mean you won't get sick from levamisole-tainted cocaine in the future. With the DEA reporting such a radical increase in the percentage of tainted cocaine (which more than doubled between 2008 and 2009), the number of people at risk is also increasing radically.

The Cocaine Trade

So who's lacing the world's cocaine with levamisole and why? "I honestly can't tell you," says Sanho Tree of the Institute for Policy Studies, a think tank based in Washington, D.C. An internationally recognized scholar, Tree has spent his career studying the drug trade—if anyone (outside of a drug cartel) should know where and why levamisole is being cut into the world's cocaine supply, it's him.

The ubiquity of tainted cocaine could, he says, be an unintended consequence of the drug war. Centralized drug-producing operations, like the old Medellín and Cali cartels, depended more on consistency of product and long-term business relationships. But after those cartels were infiltrated and disrupted, hundreds of small shops—many of them family-operated—jumped into the void. "We can't even count those operations, much less infiltrate and break them up," Tree says.

Recent developments in Colombia's guerilla wars have also destabilized business as usual. Right-wing paramilitary death squads—which are on U.S. lists of international terrorist organizations—have been fighting against Colombia's Marxist-Leninist guerillas for years on the government's behalf. Both the guerillas and the paramilitaries have been involved in the country's black market, but, Tree says, "guerillas deal more with peso economy"—the trade within Colombia—"and the paramilitary death squads made a play for the coastal regions. They were cutting the guerillas off from weapons and from smuggling zones. Those dynamics have been shifting over the past few years. The death squads were officially disbanded by government but have reemerged: same people, different names." (In 2007, the produce distributor Chiquita pled guilty to paying almost $2 million to these paramilitary death squads. U.S. congressman William Delahunt said Chiquita was only "the tip of the iceberg" of U.S. businesses getting tied up in paramilitary groups, which means those businesses are implicitly tied up with the cocaine trade.)

"As a result, there's much less accountability within Colombia now," Tree says. "The drug market is much more fragmented. Who are you going to complain to? Plus, you've got the meat grinder in Mexico"—where gangs ship South American product into North America. "Will any of these people even see each other again? Who knows? It's shorter-term careers these days."

Meanwhile, Peruvian shops are also stepping up production to compete with the small Colombian producers—the dynamics across the South American cocaine market are shifting rapidly and violently.

Even the old Mexican shipping networks are breaking up as turf wars make the smuggling routes less reliable and more expensive. Some gangs are making an end run around Mexico by sea—the U.S. government has begun intercepting homemade submarines, loaded with cocaine, that sail by night just beneath the surface of the water. One of the first narco-­subs was found in 2000 in Bogotá. "It had Russian blueprints and the engineers fled just before the police arrived," Tree says. "In Bogotá—8,000 feet in the Andes and nowhere near any ocean. How corrupt can you get?"

The U.S. has only intercepted around two dozen narco-subs so far. "They've got no wake, no conning towers—just a snorkel sticking up for air. By day they stay idle and throw a blue tarp over themselves so they blend in with the ocean," Tree says. "As one intelligence officer put it to me, rather frankly: 'You try finding a log floating in the Pacific Ocean.'"

With such a fragmented drug market, accountability and quality control decline. As Tree says, who are you going to complain to?

Which leaves the question of why producers and/or smugglers are cutting their cocaine with levamisole. Why that, instead of a cheaper and more benign cut?

"That," Tree says, "I don't know. This is the most interdisciplinary field in the world. The people who focus on violence and the cartels don't understand the pharmacology, and the people who understand the pharmacology don't understand the economics and shifting forces of the cartels. Nobody has a bird's-eye view of the whole thing."

Working Theories

In 2004, a controversy erupted in the horse-racing world. A string of trainers with long and distinguished reputations were accused of doping their horses after aminorex, an amphetamine-like stimulant, was detected in their animals' urine. The penalty for doping horses with aminorex is a one- to five-year suspension and a career-ending stigma. Accusations flew, the trainers protested their innocence, and scientists stepped in to investigate.

It turned out the whole thing was an accident: The horses had been injected with levamisole for deworming, which their bodies metabolized into speed. Studies in the 1970s had discovered that dogs experienced "mood elevation" after receiving doses of levamisole. And a 1998 study at Vanderbilt University showed that levamisole eased withdrawal symptoms in rats addicted to morphine.

That study caught the attention of Dr. Mike Clark, an assistant professor of psychiatry at Harborview Medical Center, who also studies cocaine addiction in lab rats. "According to the study, levamisole acts on all three monoamine neurotransmitters," he says. "That's exactly what you'd expect from something that potentiates cocaine." In other words, levamisole may heighten cocaine's effects—or might be a stimulant all by itself. In the next few weeks, Dr. Clark will begin an experiment of his own to find out (among other things) whether levamisole, without any cocaine, can produce cocainelike effects in lab rats.

If he can demonstrate that levamisole makes cocaine more potent, we'll be a step closer to understanding what it's doing in the supply chain. Other people have other theories, including:

• Something about the chemical structure of levamisole retains the iridescent fish-scale sheen of pure cocaine, according to a chemist with ties to the cocaine trade, giving cocaine cut with levamisole the same appearance as pure cocaine.

• Levamisole is a bulking agent for crack. The process of making crack involves "washing" cocaine and filtering out impurities and cutting agents. Levamisole slips through this process, meaning you can produce more volume of crack with less pure cocaine.

• Levamisole passes the "bleach test," a simple street test used to detect impurities in cocaine. When dropped in Clorox, pure cocaine dissolves clearly. Procaine (a common cutting agent) turns reddish brown, lidocaine turns yellowish, and other impurities float to the bottom. In a lab test conducted by ­Dr. Clark, levamisole stayed clean and clear.

If levamisole can do all of these things—pass the visual test, pass the bleach test, pass the crack-purifying process, and provide a stimulant effect either on its own or in conjunction with cocaine—it explains not only why producers use it, but why so many small South American producers have independently decided to start cutting their "pure" product. "Think of it as evolution in action," Dr. Clark says. Like a mutated gene that is beneficial to a species and is passed on through the pressures of natural selection, levamisole has a variety of benefits that become, in essence, selective pressures.

Instead of the traditional smuggling model, where centralized producers ship pure product and cut it once it crosses the U.S. border, levamisole (theoretically) behaves enough like cocaine that producers can pass off cut kilos as 100 percent pure—even to the smugglers who may believe they're shipping pure product to sell to American wholesalers. This theory is supported by a couple of findings, including reports of seizures in the DEA's Microgram Bulletin. One flight from Guyana into New York's JFK airport contained 192 churros stuffed with levamisole-tainted cocaine. And DEA agents in Bogotá came across a magazine page coated in a "protective" plastic laminate that was 21.5 percent cocaine, cut with levamisole. The research and development labs that developed this relatively sophisticated smuggling technique were at the source of production. And the source of production was cutting its "pure" product with levamisole.

A source with close ties to the DEA confirmed this, saying a recent, still-classified report has revealed that Colombian cocaine producers are putting a great deal of effort into making sure they maintain access to levamisole. "More than that," the source says, "I cannot tell you right now."

The Test-Your-Own Kit

Because the official research on levamisole's effects on human beings was stopped years ago—and, apart from Dr. Clark's pending experiments with rats, there's been no official research on its effects when combined with cocaine—there's still a lot we don't know. It's possible that agranulocytosis is only one of its health hazards.

According to a 2009 article in the Journal of Analytic Toxicology, levamisole-laced cocaine might also increase the risk of cardiac problems: "Cocaine increases sympathetic activity by blocking the reuptake of norepinephrine at the postganglionic synapse. Additive, if not synergistic effects could be expected when the drugs are combined. Concerns of increased toxicity with exaggerated pressure response or development of arrhythmia could then arise when cocaine is combined with levamisole."

Heart attack and cardiac arrest are two of the common causes of death associated with cocaine overdoses—levamisole might exacerbate those risks. It's hard to say: Cocaine, according to the latest Seattle/King County drug trends report, released this June, is the most common illegal drug detected in deaths, but its manner of killing is less clear-cut than opiates. "Opiate overdose is pretty simple and straightforward," explains ­Dr. Coffin. "It's breathing. Keep them breathing and they live. Cocaine is more difficult: Is it a massive heart attack? Is it a stroke? It's not very well defined."

Currently, people who suffer cardiac problems associated with cocaine are not tested for the presence of levamisole, so we really have no idea what kind of damage this new cutting agent is inflicting on the nation's cocaine users, nor the strain on already strapped public funds—every time someone without health insurance lands in the emergency room, it costs taxpayers thousands of dollars.

The fuzziness surrounding cocaine's destructive qualities makes harm-reduction strategies more difficult for cocaine than for opiates. Nobody doubts that cocaine is destructive. "It's toxic to heart-muscle cells," Dr. Coffin says. "Even in its purest form, it's among the worst recreational drugs for the cardiovascular system." But its spectrum of harmful qualities, some of which are exacerbated by levamisole, makes it tricky to pinpoint good maintenance programs for chronic addicts. Opiate addicts, Dr. Coffin says, can live on methadone or other controlled dosing mechanisms their entire lives with no medical harm besides constipation and loss of libido. But cocaine- and amphetamine-maintenance programs haven't shown any conclusive results, despite attempts in Colombia to prescribe coca tablets and tea to addicts.

One thing that can be done: develop an inexpensive field-test kit to try to detect levamisole. Dr. Clark has invented such a kit and—in association with The Stranger, a few folks in the local harm-reduction community, and the People's Harm Reduction Alliance (PHRA), which runs the U-District needle exchange—hopes to begin distributing kits in a few weeks. Unfortunately, kits are technically drug paraphernalia under Washington State law, not only because the kits will contain cocaine residue, but because it is illegal for any person to possess something used to "process, prepare, test, analyze, pack, repack, store, contain, conceal, inject, ingest, inhale, or otherwise introduce into the human body a controlled substance." It's a perfect example of how drug prohibition laws make drugs more dangerous—an unregulated market for cocaine, with no quality control, has encouraged the use of levamisole as a cutting agent. And U.S. drug laws make it illegal for users to test their cocaine for poison—if users could, they might stop buying from dealers who sell tainted cocaine, putting economic pressure on the market to be less dangerous. It's a classically self-defeating chain of policies, but some antidrug warriors defend it on the grounds that since drugs are illegal, users get what they deserve. And if cocaine is perceived as more dangerous, perhaps fewer people will use it.

This, of course, is a cruel, stupid, and expensive way to deal with the problem. As Dr. Clark put it: "The idea of letting addicts die to make drugs scarier is reprehensible."

It's not quite the same in the heroin world: Because of the public outcry about the health risks of sharing needles, hypodermic syringes have a special exemption. Crack users need similar exemptions.

"If you read the paraphernalia laws, cocaine is both the forgotten drug and in some ways the most hated drug," says Shiloh Murphy, director of PHRA, an independent nonprofit that isn't affiliated with public-health-funded needle exchanges. He gestures behind him to a tower of cardboard boxes full of hypodermic needles. "A young person just starting to inject knows he shouldn't share his syringes," he says. "But a 20-year crack veteran doesn't realize that every time he smokes and burns his lips and passes on his stem, he could be transferring the same diseases—it's open sores to open sores."

To combat this problem, Murphy has begun a controversial program to distribute crack stems, rubber crack "condoms," and fresh steel wool to users. (Steel wool, which is used as a filter in crack pipes, weakens and flakes off after repeated use, sending red-hot chunks of metal into users' throats and lungs, which leads to infections and abscesses.) Murphy got the idea for his crack program one afternoon two years ago, when he was approached by an angry crack user.

"I was sitting at the table, handing out flyers and things," Murphy says, "and a man said to me: 'You're a real motherfucker, you know that? You're sitting here with all these syringes and talking about health. I use crack and my friends are dying of HIV and hepatitis C and there's nothing on this table for us. I guess crack users are always just left to die.' I said, 'You're right. I'm sorry. Tell me what you need.' It was an enlightening moment for me."

"All we have to do," he says, "is save one person from getting HIV, and we've become economically worth it." PHRA's annual budget is around $385,000. Its budget for the crack program is currently $6,000. The lifetime cost for the state to take care of an uninsured person with HIV, he says, is half a million dollars. "We've saved the state thousands and thousands of dollars."

Now Murphy will be at the forefront of our combined attempt to distribute Dr. Clark's levamisole kits to cocaine users. The kits will contain instructions for use, a fact sheet about levamisole and agranulocytosis, and a survey on a prestamped postcard about where and when the cocaine was purchased, whether it's powder or rock cocaine, whether it tested positive for levamisole, and a few other research questions. Hopefully, that data will help us—me, Dr. Clark, PHRA, and a local harm-reduction organization called DanceSafe—develop a better understanding of how levamisole-tainted cocaine is distributed through the city and whether some neighborhoods face greater health risks than others. (Is the cocaine you can buy on the street in Georgetown, for example, more or less tainted than the cocaine at some millionaire's house party in Bellevue?)

As for its illegality: After a meeting with ACLU lawyer Alison Holcomb, Stranger publisher Tim Keck, and me, Seattle city attorney Pete Holmes and King County prosecutor Dan Satterberg decided to allow us to distribute levamisole test kits—and collect data about whether people are finding levamisole—without prosecution. They notified new Seattle police chief John Diaz, who supports the program. recommended

The levamisole test kits will be available in a few weeks—watch for updates in The Stranger and on Slog. This piece is the first in an investigative series.

This story has been updated since its original publication.