It was July 2003 that I set out on a journey to Cape Girardeau, Missouri, to visit Edwin J. Masters, the doctor involved in hand-to-hand combat with the Centers for Disease Control and Prevention over the existence of Lyme disease in the Southern United States. Working with a few intrepid colleagues, Dr. Masters managed to generate powerful evidence for Southern Lyme, though his evidence was continually undermined.
For two days straight I sat with Dr. Masters in his oversized basement, reviewing document after document showing how data had been cast to shed doubt on the disease. Ed Masters' story sheds light not only on Lyme disease but also the dangers we all face when medicine is politicized and studies skewed. His great persistence finally led to recognition of Masters' disease, the Lyme of the south.
The heroic Dr. Masters died on June 21 2009. In his honor, I'll spend several days retelling his sprawling, riveting, and most important tale.
For two weeks in July 1991, CDC researchers occupied a space in Masters' office, interviewing his Lyme patients and reviewing their charts. Based on their interviews, the CDC scientists went to the areas the patients said they had been bitten and collected ticks. Masters provided the CDC with blood and biopsy samples taken from the patients during their illness. Finally, the CDC took the ticks, the samples, and the data back to Fort Collins to conduct their analysis and write their report.
It was almost two years later in May of 1993 that the CDC sent a final draft to Masters for input, and he was alarmed. The first thing that stood out was the CDC assertion that, according to "unpublished data," Missouri rashes differed from real Lyme rashes on the basis of "coloration, degree of homogeneity, sharpness of borders, and shape." The CDC also contended that EM lesions in Missouri were smaller, on average, than those observed on Lyme disease patients in Wisconsin, based on a 12-patient study still in press.
Masters was flabbergasted. After all, he had shown his rash photos throughout the world, and the most expert dermatologists on the planet, from Sweden and Germany to Long Island, had said he had a "ringer." As to lesion size, why, he wondered, had the CDC deferred to this small Wisconsin study over the work of dermatologist Bernard Berger, an internationally recognized Lyme disease expert who had done the seminal studies on the rash? Berger's widely cited report on 196 patients showed that the average rash size of confirmed Lyme cases and Missouri Lyme cases were exactly the same.
Could some of the confusion be traced to careless error? Comparing his patient charts to CDC data, Masters found inexplicable mistakes. One patient with a rash stretching across his back was reported with a lesion just a quarter inch in diameter. Two patients with obvious bull's-eye rashes were listed as having "no central clearing." And a patient whose chart contained a photograph of a rash across his abdomen had been reported as having no rash at all.
"I called Fort Collins and a group of them were around the speaker phone. I went ballistic," Masters recalls. "I said, you are telling me that my rashes are visibly distinct from real Lyme, which means that you can tell by looking, that these are not real rashes? Then why are we doing the friggin' study? Hey if you can tell by looking, I demand that you hold a press conference to teach us dumb yokels out here in the boonies how to do it. I'm an author on this paper --and you say you have unpublished data? I have never seen it, I have never even heard of it, and if this data exists it is one of the most important keys to this puzzle. I want to see the data now."
The CDC team was virtually silent. Then, three days later, Masters received a new version of the article, with the material he'd challenged removed. But as Masters and his colleague, Denny Donnell, read through what would be several more drafts of the CDC manuscript, they knew they could never sign on. As Masters saw it, the CDC had "skewed everything, and had literally tossed out data, to make what we found in Missouri look like a rash-only illness, and as different from Lyme disease as could be."
Most misleading, he felt, were the arbitrary stop and start dates the CDC had imposed on the study after collection of data was complete. In most such studies, each patient is studied for the same amount of time as all the other participants. Start-dates are based on the start of illness for each individual patient and end-dates are determined by adding a consistent amount of time --the same for each and every patient-- onto that. But in the Missouri study, the CDC decided to cut all patients off on the same date, no matter when the illness had begun. Thus, some patients were followed for a couple of years, others for a couple of weeks. The disturbing part was that the stop dates placed patients with the most objective signs of illness, including carditis (inflammation of the heart) and arthritis, outside the study period and thus, beyond the scope of the report. To wit: Even though patients in the study had developed serious, late-stage signs considered classic for Lyme disease, the CDC paper referred, without qualification, to the "absence of documented early neurologic, cardiac, and arthritic complications." (An analogy would be the situation in which scientists studying HIV infection cut their study off after two weeks and so conclude the virus is not a cause of AIDS.)
Masters complained passionately about the cut-off dates to Phillip R. Lee, MD, Assistant Secretary for Health, who he hoped might intervene. One of the patients, he wrote to Lee, was a previously healthy young man who had developed carditis 14 days after appearance of his rash, but five days after the arbitrary end of the study period. There was no reason to cut him off, said Masters, because it took another year before the manuscript was ready to submit. Similarly, three cases of arthritis were excluded from the "study period," and from the report. One patient's rash occurred before the CDC start date, "and what a shame," said Masters, "because he also had a positive Western blot and documented joint swelling in his knee." Two other patients developed arthritis after the stop-point. "One of those arthritis patients had symptoms emerge 58 days after the onset of the rash, just 34 days after the study's arbitrarily determined end. Such omissions were "absurd," said Masters, given that onset of Lyme carditis occurs, on average, 4.8 weeks after the rash and Lyme arthritis may not develop for months.
There was more: The CDC had dismissed laboratory work, a wide swath of it, that indicated some kind of borreliosis in play. For instance, the report did not put much stock in the finding, by CDC scientists, of "motile spirochetes" in nearly five percent of the lone star nymphal ticks observed by darkfield microscopy. Nor did the CDC report that when spirochetes from the ticks were inoculated into mice, they later cultured spirochetes from the animals' ears. "Having seen these spirochetes myself, I am not comfortable with the CDC's position that Missouri is the hole in the donut," Masters wrote to Lee, "and that somehow Missouri is a magical, ‘Lyme-free' zone and that these observed Missouri spirochetes have nothing to do with human disease."
Finally, and this was the last straw to Masters, the CDC rejected dozens of positive blood tests performed at its own lab. While it was true that the CDC used more specific tests as years went on --methodologies unavailable when the study began-- Masters could not understand how so many positive and equivocal results over so many samples were not considered suggestive of another, similar organism, even if not Borrelia burgdorferi itself. Instead, positive results where no Lyme disease was possible --as in Missouri-- were just the sort of incentive the CDC had needed to tighten the valve. The CDC kept refining and re-testing because, as the results from Missouri showed, the bar for diagnosis had been too low.
In a nutshell, the CDC insisted the illness in Missouri, whatever it was, had nothing to to with Lyme disease, while Masters insisted the evidence had been left on the cutting room floor. When it came to Masters' insistence, Duane Gubler, head of the Vector-Borne Disease Division out in Fort Collins, was especially clear: When you biopsied a Lyme rash in the northeast, you cultured Borrelia burgdorferi -not so with rashes from Missouri. If he couldn't culture Borrelia burgdorferi from the rashes, then it wasn't causing the illness. If the Missouri rashes weren't caused by B. burdorferi, then no amount of other evidence could convince Gubler to call the outbreak Lyme disease. Whatever data Masters felt had been left out, it could mean little compared to that.
Adapted from Cure Unknown, Inside the Lyme Epidemic.(St. Martins Press, 2008)