Over the last quarter century, about two-dozen researcher-physicians—primarily the authors of the Infectious Diseases Society’s treatment guidelines—have dominated Lyme disease care in the United States and around the world, a dynamic that has dramatically limited physician freedom. They have done this, with the support of major medical journals and government agencies in the United States and through a series of studies, much of it adequate if limited science. Their research findings, often repeated in drumbeat fashion in multiple journals, have defined and minimized Lyme disease by relying on successes in early cases, emphasizing the risks of longer treatment, and dismissing the reality of patient suffering. They have left little room for debate, even while they acknowledge the potential limits and biases of their research.
The Myths of Lyme
Coinciding with the dawn of open-access publishing, indeed perhaps because it provided a ready platform that is seen as cheapening its value, articles that have pushed back on traditional views of Lyme disease—particularly those suggesting evidence of the pathogen’s ability to survive antibiotic treatment—have been dismissed as flawed or inconclusive. In Lyme disease, the natural process of scientific self-correction has been stymied, the battle for control less a war than a decades-old standoff. Over time, a tapestry of myths has been woven from sturdy but not incontrovertible science, creating an enduring and false likeness of Lyme disease as a problem without urgency. As a result, cases of Lyme disease have been missed, doctors have avoided caring for people with complicated presentations, and patients—many of them children—have suffered enormously and needlessly.
Among the early themes that have shaped and limited care are the following four myths:
Myth #1: Lyme is overdiagnosed. In the early 1990s, Alan Steere, the physician credited with discovering Lyme disease and an author of the first treatment guidelines, studied the records of nearly eight hundred Lyme-diagnosed patients. Fully 57 percent did not suffer from the malady, he found. As significant, nearly half of patients had tested positive in other labs and negative in Steere’s. While such findings could have been taken as a tip-off of the poor state of Lyme diagnostics, they were read as a warning sign of overtreatment, one that would produce a stream of easily replicated science.
The Steere study, published in 1993, was followed by at least ten other published articles on overdiagnosis through the rest of the 1990s. Each came to similar conclusions and, in turn, led to more studies. Overdiagnosis leads to “frequent minor adverse drug events,” said a study in Annals of Internal Medicine, such as nausea, rash, and vaginal itching. Ordering unneeded tests, it asserted, also eats up “considerable health resources,” a sum that was later put at $492 million for more than three million tests every year. By 2016, about thirty publications in the National Institutes of Health library directly or indirectly addressed Lyme disease overdiagnosis. These included several studies of purported overdiagnosis in children and in people with neurological involvement, the kind of cases careful doctors might not want to miss. Just five studies addressed potential underdiagnosis.
On the overdiagnosed side were patients who falsely tested positive—there were no doubt some—and were prescribed antibiotics for short periods. On the underdiagnosed side were patients who had falsely tested negative for a disease that, once missed, can cause lasting pain and debilitation. Science sided with the folks who should be spared unnecessary treatment, which is logical. First, do no harm. The assumptions of those studies are open to question, however. First, they rested on the premise that the risk and cost of treating appreciably outweighed not treating, a question no study asked. Second, the studies presuppose that science can prove, through standard blood sampling, who actually has Lyme disease. That brings me to the second myth.
Myth #2: Lyme disease testing is reliable. The CDC endorses what is called “two-tiered” testing, saying on its website: “Patients who have had Lyme disease for longer than 4–6 weeks, especially those with later stages of illness involving the brain or the joints, will almost always test positive.” Physicians have taken pronouncements like that, combined with prolific warnings of overdiagnosis, to mean one thing: Go by the tests. Yet dozens of scientific papers paint the Lyme test, which I’ll discuss in chapter 6, as anything but cut and dry. Because of the time needed to muster antibodies, just 30 to 40 percent of infected patients will test positive early in the disease, a Lyme fact acknowledged by the CDC and IDSA.
Testing accuracy improves later on, but only after patients are sicker, the disease more advanced. And still cases are missed. Perhaps one in four patients with disseminated disease and one in eight with neurological disease will wrongly test negative weeks to months into the disease, according to key testing studies. Then there is the question of whether published accuracy rates, which vary widely, are even reliable. Studies that have validated the tests suffer significant biases, for example proving the technology works by testing it on people who previously tested positive. What about those it failed? Wide variation in accuracy has also been found depending on which commercial test, and in what two-tiered combination, is used. One may ring positive; another negative. The only definitive symptom on which doctors are encouraged to confirm a Lyme diagnosis is the erythema migrans, or EM, rash, and it appeared, according to a CDC study of 150,000 cases, in 69.2 percent of patients. That’s three in ten cases potentially missed early on in a diagnostic milieu rife with testing pitfalls.
In 2016, the journal BMC Infectious Diseases published a review by a team of twenty-one European researchers of seventy-eight studies of the standard tests to diagnose Lyme disease. It was the most exhaustive review of Lyme testing ever done, analyzing studies of technologies commonly employed in the United States. Ultimately, the analysis found many test studies wanting, and all test studies plagued by at least one serious bias. The team’s article concluded, “The data in this review do not provide sufficient evidence to make inferences about the value of the tests for clinical practice.” In other words, six and a half dozen studies could not settle the issue: We don’t know for sure if the tests work in a patient-physician setting. How can we say Lyme is overdiagnosed when we cannot trust the tests?
Myth #3: Lyme disease is hard to get. In June of 2001, a study was released that prompted the New York Times to publish a page one story that said, in essence, the nation’s biggest vector-borne disease was nothing to worry about. It began: “Lyme disease is very difficult to catch, even from a deer tick in a Lyme-infested area, and can easily be stopped in its tracks with a single dose of an antibiotic, a new study shows.” The study, published in the New England Journal of Medicine (NEJM), tracked 482 people in Westchester County, New York, who were given a prophylactic dose of doxycycline after they had been bitten by ticks. Just 0.4 percent of the group that received the dose developed Lyme disease. At the same time, 3.2 percent of a control group developed a rash symptomatic of the disease. The study made two points: Lyme disease could be prevented after a tick bite. And a small number of bitten people, three in a hundred, would get Lyme disease.
There is little dispute that a smaller share of infections result among people who know they have been bitten; ticks are more likely to be removed before they infect. Beyond this, the finding that a single doxycycline dose wards off disease came with a significant caveat. Researchers used the rash alone as proof of infection, admitting that doing so “could have resulted in underestimation of the actual incidence” of the disease. Indeed, some 30 percent of Lyme patients in the CDC study did not get the rash; at least one nonrash participant in the treated group tested positive but was not added to the single treated patient who nonetheless got Lyme disease. Nor were the study’s patients followed beyond six weeks, when others might also have tested positive. Last, fifty-one people, 11 percent, dropped out of the study but were assumed not to have gotten the rash. A critique published in the journal Expert Review of Anti-Infective Therapy argued that just one more rash in the treated group would have erased the study’s
significance. The critique was included in treatment guidelines of the International Lyme and Associated Diseases Society, which favors twenty days of prophylactic treatment for tick bites based on studies that came to other conclusions. In 2004 and 2008, the single dose failed up to half the time to prevent Lyme disease in two mouse studies, while a longer-acting antibiotic injection was completely effective.
Despite the NEJM study’s limits, its single-dose prophylactic finding went on to become the standard of care for tick bites. The tough-to-get, easy-to-stop message should help allay “inflated public fear of Lyme disease,” the Times reported, while the NEJM editor, Dr. Jeffrey M. Drazen, pronounced, “This is reassuring information for people who make decisions based on evidence.”
The study’s release, along with the results of two treatment trials, was an early and clear attempt to tamp down a growing Lyme-patient insurgency, and it would not be the last. Beyond this, it was science put to the task of public relations. Its message was control—of an epidemic, with a questionable prophylactic, and of an alarmed public, with a single, oversold study that has yet to be replicated.
In the ensuing years, the Lyme disease toll would grow in the United States from about 17,000 in 2001 to 38,000 in 2015. But because of underreporting, those official figures represented just a tenth of all cases, the CDC announced in 2013. That year, Lyme disease was ranked fifth in the United States among reportable diseases even before the tenfold adjustment, especially impressive considering that 95 percent of reported cases come from just fourteen states. Do these numbers fit the description of a disease that is hard to get? Nonetheless, a physician quoted in the article, Dr. Leonard H. Sigal, who would go on to publish widely on Lyme disease overdiagnosis, said catching the disease wasn’t the problem. “The bigger epidemic,” he said, “is Lyme anxiety.”
Myth #4: Lyme disease is easy to treat—and is not a chronic disease. “Patients treated with appropriate antibiotics in the early stages of Lyme disease usually recover rapidly and completely,” says the CDC, reflecting the IDSA guidelines for care. Whether the infection survives antibiotic treatment to become chronic is the single most contentious issue in Lyme disease, one that the CDC and IDSA have long sought, unsuccessfully, to put to rest. When the Westchester County study on prophylactic treatment was released, the NEJM announced a second article on twin treatment trials, articles that were “so important,” the Times reported, that they were publicly unveiled a month ahead of publication to coincide with the onset of summer and outdoor activities. In both trials, researchers gave ninety days of intravenous and oral antibiotics to Lyme disease patients who, although previously treated, had suffered “considerable impairment of health-related quality of life.” The researchers ended the trials early, however, concluding that the treated group did no better than a group that got placebo drugs. “Prolonged and intensive treatment with antibiotics, a course of care advocated by a small group of doctors, does nothing for people with symptoms often attributed to chronic Lyme disease,” the Times reported.
Although those two trials, led by Mark Klempner of Boston University School of Medicine, were used to assert that continued antibiotic treatment is unsuccessful and dangerous, the studies were the antithesis of a slam-dunk for Lyme practitioners. Treating advanced Lyme disease had long been more art than science, as these doctors built on clinical experience and shared knowledge to treat patients no one else would. One antibiotic combination might work for patient A, another combination of drugs may help patient B. Lyme practitioners also knew that tick-borne coinfections, which the trials did not consider, often immensely complicated diagnosis and care. They saw patients improve.
In 2012, a team led by Allison DeLong, a Brown University statistical researcher, analyzed the methodology and findings of Klempner’s two treatment trials and two others by Lauren Krupp of Stony Brook University in 2003 and Brian Fallon of Columbia University in 2008. Like Klempner, Krupp and Fallon had reported mixed or fleeting patient improvement from additional rounds of antibiotics in studies that largely discounted their effectiveness. The review, published in Contemporary Clinical Trials, found otherwise. One of Klempner’s two trials used “unrealistic” measures, DeLong found, deeming patients to have improved only when they surpassed the norm for the US population by essentially one standard deviation (placing them in the 85th percentile for health). Sample sizes were also likely too small to detect meaningful improvement, the analysis asserted. In fact, all four studies suffered statistical and analytical flaws that underestimated their positive effects on patients—in particular, in Krupp and Fallon, respectively, on fatigue and cognitive functioning. “Our careful examination of the trials suggests that, for some patients with Lyme disease, retreatment can, in fact, be beneficial,” the review concluded. Fallon himself revisited his and Krupp’s trial results in 2012, concluding that demonstrable benefits had been overshadowed by side effects that perhaps could be ameliorated.
Cheering on Failure
What was striking when the Klempner study was released was the enthusiasm with which researchers touted trials in which no solution was found, in which their efforts to heal had failed. “If this had been a study looking at tuberculosis,” Tom Grier, a Lyme disease patient, wrote in 2001, “would the study be stopped early and then be rushed to press to tell us that the patients did not feel any better after three months of antibiotics?” This is the reality of Lyme disease treatment research, then and in the nearly two decades since Klempner’s research. What little treatment research there is has been used to reinforce the dogma of the one side that claims to know what works (short-course antibiotics) and what doesn’t (anything longer). Researchers whose findings challenge prevailing science, as I’ll discuss here and in other chapters, are marginalized to small journals, they have told me, or made to revise their conclusions. Some do not get published at all when the results suggest that Lyme disease may be bigger, more widespread, and more persistent than the prevailing side believes. These researchers struggle to get funding for research from government agencies whose review boards are dominated by scientists with accepted and traditional views of Lyme disease.
The United States may not export much in the twenty-first century, but its made-in-America vision of Lyme disease sells well in many countries around the world. Lyme is exaggerated. Lyme can be dealt with. Lyme is no big deal. That has been the theme of paper after paper, in particular in the New England Journal of Medicine, where Klempner, the lead author on the treatment studies, has served as an associate editor for ten years. In 2007, the journal ran an “appraisal” of chronic Lyme disease, in which it reviewed the evidence against the use of antibiotics for lingering symptoms of Lyme disease, repeating, as other articles and commentaries have, the primacy of the Klempner findings. The timing of the review’s publication was interesting, arriving after Connecticut Attorney General Richard Blumenthal had opened an antitrust investigation of the Infectious Diseases Society in November of 2006. The probe centered on potential conflicts of interest involving the physicians and researchers who wrote the guidelines that dictate Lyme disease care. Seven months after the appraisal was published, Blumenthal, who went on to become a US senator, delivered this stunning conclusion: “The IDSA’s 2006 Lyme disease guideline panel undercut its credibility by allowing individuals with financial interests—in drug companies, Lyme disease diagnostic tests, patents and consulting arrangements with insurance companies—to exclude divergent medical evidence and opinion.”
The New England Journal article was seen as an attempt to ward off the Blumenthal probe, with the nation’s foremost medical journal employed in the service of pushback and public relations. The tone of the appraisal was indignant and some of the content had little to do with science. “Physicians and laypeople who believe in the existence of chronic Lyme disease have formed societies, created charitable foundations, started numerous support groups (even in locations in which B. burgdorferi infection is not endemic), and developed their own management gu
idelines,” the appraisal stated. “Scientists who challenge the notion of chronic Lyme disease have been criticized severely.”
Yet such criticism was nothing compared to what doctors who practiced outside the guidelines faced. At the time, several Lyme physicians in New York State—including those who had formed their own medical society and developed their own guidelines—were being investigated by the state’s physician licensing board for prescribing antibiotics outside of the IDSA’s treatment guidelines. These doctors told me the reviews had often lasted for years and cost them huge sums in legal fees. One physician had been suspended and another was cited for lapses in record keeping, insignificant infractions that were enough to scare other doctors away from treating difficult cases of Lyme disease. Similar actions were playing out in other states and a few countries, too. To guideline adherents, this rogue behavior was simply not acceptable. “Chronic Lyme disease, which is equated with chronic B. burgdorferi infection,” the appraisal concluded, “is a misnomer, and the use of prolonged, dangerous, and expensive antibiotic treatments for it is not warranted.”
Lyme Page 8