(Illustration: Laura Waldusky)
On June 30, Fox 5 New York aired a highly-rated thirty-minute TV special about Lyme Disease called "Lyme and Reason," in which I was interviewed with my doctor, Yale-trained internist and preeminent Lyme specialist, Steven Phillips. You can watch our segment here.
Also interviewed was the CDC's Paul Mead, MD, Chief of Epidemiology and Surveillance Activity of the Bacterial Diseases Branch of the Centers for Disease Control and Prevention. You can watch that here.
Many patients and clinicians expressed outrage on social media regarding what they considered to be dangerous, misleading information offered by Mead, and pointed out the CDC's pattern of giving preferential treatment to outdated IDSA Lyme guidelines instead of the more scientifically-supported, patient-centered ILADS guidelines.
Interestingly, The National Guidelines Clearinghouse, a federal database that provides treatment information to health care professionals and insurance companies, has removed the IDSA Lyme treatment guidelines from its website and only lists the ILADS guidelines.
To help chronic Lyme patients who feel that the CDC has once again turned its back on them, Phillips wrote a formal rebuttal to some of Mead's statements and Fox 5 posted it on their Facebook page.
Fox invited Mead to comment further but that request remains "under review."
See Mead's statements and Phillips' responses below and watch the TV special and all of the unedited interviews here.
Please join the conversation and add your comments below. We look forward to hearing from you.
Mead: "typically people with early stages of disease get a rash..."
Phillips: I don't actually think it's accurate to say that this is typical. The initial published findings by Steere documented that 25% of patients recalled a history of rash compatible with erythema migrans (EM). Other published research has pegged the rate of prior EM in late Lyme patients at 22%. Part of the problem with some of the research that demonstrates a very high rate of EM in Lyme disease is that it's part of the CDC reporting criteria as well as being a common diagnostic criterion. So the CDC's statistics on the rates of EM in early Lyme may be inherently skewed. It may be likened to publishing a study that 95% of people in prison have committed a crime.
Mead: "well, actually 'the great imitator' is a term used to refer to syphilis which is a different disease"
Phillips: Well, actually, although syphilis was a previous illness given this nickname, Lyme disease has also carried the moniker in the medical literature. A Pubmed search of 'great imitator' and 'lyme' returned 23 results. This is because Lyme can present in so many varied ways, able to mimic a broad array of diseases.
Mead: "there are really two parts to diagnosis, there are the clinical features of the disease that is what the physician can see, a large joint, a red rash, a fever, that sort of thing, and then there is laboratory testing."
Phillips: Most of the clinical features of Lyme disease are subjective, i.e., symptoms that the physician can't see. There is robust data in the medical literature which documents that in patients diagnosed with Lyme disease based on the presence of EM or of B. burgdorferi and/or its components obtained from body tissues and/or fluids, the subjective symptoms of Lyme disease outnumber the objective signs by very significant numbers.
So to restrict clinical diagnosis to those patients who have the several objective features included in the CDC reporting criteria would mean that a great many patients would go undiagnosed.
I don't think it's justified to exclude subjective symptoms from the reporting criteria just because such symptoms can't be seen. More alarming, in my view, is that this practice reinforces a medical paradigm of not believing the patient. When do we start believing patients again?
I see lots of patients in my office who come in with the 'not-Lyme diagnosis'. When I ask the primary care physicians further about this, I'm frequently told that they don't know what the patient has, but that it's not Lyme. The exclusion of this diagnosis is most often predicated upon failure to meet CDC surveillance criteria. I find it really sad and frustrating. It's like we're speaking different languages.
Mead: "So the recommended way of diagnosing Lyme disease in the laboratory is by the use of serologic tests primarily and in general we recommend a two-step process to this where the blood is tested essentially in two steps to identify whether or not the person has evidence of infection with Borrelia burgdorferi."
"If a patient has been ill for just a few days or weeks the test may in fact be negative. However if a patient has been ill for months or years, if the test is negative, that's good evidence that their illness may be caused by something other than Borrelia burgdorferi infection."
Phillips: It's interesting to me that CDC currently recommends laboratory surveillance criteria for diagnosis since they've historically been recommending it for standardization purposes as below, which is still listed on their website:
"This surveillance case definition was developed for national reporting of Lyme disease; it is not intended to be used in clinical diagnosis."
http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5623a1.htm
Further, this is Dr. Mead's testimony excerpted from the State of Connecticut Department of Public Health Public Hearing on Lyme Disease January 29th, 2004:
http://www.ct.gov/ag/lib/ag/health/0129lyme.pdf
"Let me now say a few words about clinical diagnosis. The clinical diagnosis is made for the purpose of treating an individual patient and should consider the many details associated with that patient's illness. Surveillance case definitions are created for the purpose of standardization, not patient care. They exist so that health officials can reasonably compare the number and distribution of cases over space and time. Whereas physicians appropriately err on the side of over-diagnosis, thereby assuring they don't miss a case, surveillance case definitions appropriately err on the side of specificity, thereby assuring they do not inadvertently capture illnesses due to other conditions."
"...CDC has stated repeatedly that the surveillance case definition is not a substitute for sound clinical judgment. Given other compelling evidence, a physician may choose to treat a patient with Lyme Disease when their condition does not meet the case surveillance definition."
If Dr. Mead has previously stated that surveillance case definitions are not to be used for patient care, then why is Dr. Mead now recommending the opposite? I'm not aware of any groundbreaking medical literature in the past 12 years which should have radically changed the opinion of the CDC on this matter.
To the contrary, since 2004, there have been even more published medical journal articles demonstrating Lyme disease with negative Lyme antibody tests, further raising questions about the validity of CDC surveillance criteria for diagnosis. In fact, there are now over 50 medical journal articles documenting Lyme disease despite negative antibody tests. This research spans all stages of illness, including late stage disease, at which time Dr. Mead opines that Lyme antibody testing should be positive.
Further, as additional pathogenic species of borrelia are continually being discovered, the term 'Lyme disease' is being thought of more and more in a collective sense. It is logical to conclude that the Lyme ELISA would cross-react with at least some of these species. By restricting diagnosis to CDC two-tier serologic reporting criteria, a greater truth may be missed in that diagnoses of such patients with non-Lyme borrelial infections would not be possible in the absence of species-specific testing for the new species, which is largely unavailable.
Another important question regarding diagnosis relates to the use of polymerase chain reaction (PCR). PCR is a very well-established and reliable technology that has been in continuous use for more than 30 years. It replicates DNA many times over so that it can be picked up on a test. If a PCR test is positive, this means that the DNA of a microbe is present and is considered definitive proof of infection for virtually every other infectious disease known in medicine. Yet for Lyme disease, it appears to me that there may be a double standard.
I recently saw a video from the CDC in which Dr. Christina Nelson advised against using Lyme PCR to evaluate Lyme disease. She spoke of its relatively low sensitivity in vivo presumably as a motivation to avoid its use. However, if this test is positive, then active infection is confirmed, which is exactly what's needed to clarify persistent infection in Lyme disease. So it's curious that she would advise against its use, even with a less than perfect sensitivity.
It should be further noted that PCR is a very specific technology, meaning that even small differences in the targeted snippets of DNA can result in a negative test. Coupled with strain heterogeneity and highly variable gene expression in B. burgdorferi in the tick vs. the mammal, it appears that some PCR assays be may going after the wrong target. It's not clear from the video to which Lyme PCR tests Dr. Nelson is referring. Newer PCR's using multiple targets and other improved technologies may have better sensitivities.
She also expressed concern that PCR tests might be falsely positive and represent dead DNA. However published data demonstrates that injection of dead borrelial DNA into dogs did not produce positive Lyme PCR tests in as little as even a few days after inoculation. This means that dead Lyme DNA does not remain in the absence of a continual resupply by a live infection. So I'd like to know from where this concern springs? And if PCR false positivity is so problematic, then why has PCR emerged as the cornerstone for diagnosing and monitoring a multitude of infectious diseases and where are the warnings from CDC about false positive PCR for these other infectious diseases?
Mead: "It's most valuable to order the test in people who have or who are likely to have the disease. If a person is unlikely to have the disease there's a chance that if the test comes back positive that it's more likely a false positive than a true positive"
Phillips: This statement appears to have its roots in Bayes' Theorem. Bayesian methods are one of the more controversial approaches in statistics, with the inherent limitations of being a closed system of logic. For example, who is likely to have the disease?
Everything depends on the initial assumption of probability based on clinical criteria that are under dispute. We know that the objective clinical signs described by CDC surveillance criteria are restrictive in that they do not capture the majority of Lyme diagnoses, which is echoed by CDC itself below:
"...the total number of people diagnosed with Lyme disease is roughly 10 times higher than the yearly reported number"nu
http://www.cdc.gov/me.../releases/2013/p0819-lyme-disease.html
So whereas patients having the objective clinical signs described by CDC surveillance reporting criteria are likely to have Lyme disease, it appears from CDC's own statement that these represent only the minority of patients. We'd therefore be missing the majority of diagnoses if we followed a Bayesian approach.
Mead: "CDC recommends that people rely on FDA approved tests for the diagnosis of Lyme disease"
"FDA approved tests are various forms of the two tier assay that I just mentioned, the serologic antibody testing"
Phillips: FDA approval for lab testing requires clarification as there are currently no FDA approved Lyme tests. States have Clinical Laboratory Improvement Amendments (CLIA) offices which ensure that labs adhere to certain standards.
FDA test approval is required when a test kit is sold across state lines in the US and does not, per se, indicate improved accuracy compared to CLIA approval. In the absence of a lengthy FDA approval process, tests can be cleared by FDA and given similar treatment if they are demonstrated to be roughly equivalent to a former comparator test.
The first Lyme Western blot to receive FDA clearance was the MarDx Lyme Western blot. A review of the FDA's database reveals that this test was compared to the Lyme Western Blot performed by Dr. Steere's lab at Tufts. It's not clear to me if the comparator test was ever FDA approved, but it appears from my interpretation of the data that it was not. Because most doctors don't know what this actually means, they view the lack of FDA approval or FDA clearance of a test as a bad thing. Lyme antibody assays offered by even the major universities that perform research in Lyme disease are not FDA approved. Historically, these tests have not even been cleared by FDA.
Mead: "The ILADS Guidelines advance two basic ideas: The first is that there really is not any adequate scientific information about the management of Lyme disease; and the second is that in the absence of that sort of information, healthcare providers should be free to do and treat patients in whatever way they see fit. Our concerns are that misrepresents or does not give full credit to the amount of scientific evidence there is about management of Lyme disease..."
Phillips: In response to the claim regarding the first basic idea, I think that this is a highly inaccurate statement. ILADS (International Lyme and Associated Diseases Society) is a multi-specialty medical society made up of physicians and researchers. Its members are well-published in the medical literature and the society holds large, very well-attended, CME-approved medical conferences annually both in the US and abroad.
ILADS Guidelines provide an evidence-based approach to the management of Lyme disease. They are published in the peer-reviewed medical literature and are presently the only Lyme disease treatment guidelines listed on Guidelines.gov, an agency which is under the auspices of the U.S. Department of Health and Human Services.Guidelines.gov was created by the Agency for Healthcare Research and Quality in partnership with the American Medical Association and the American Association of Health Plans, (which is now known as America's Health Insurance Plans) in order to provide physicians, other healthcare providers, and health plans with detailed information on clinical practice guidelines and to further their use. Despite this, both CDC and insurance companies do not endorse or refer to ILADS Guidelines.
The ILADS Guidelines assess the medical literature on the topic of Lyme disease and associated infections. There are many shortcomings with the state of research in the field, but the purpose of the ILADS Guidelines is not to bemoan this fact but to improve the welfare of a suffering population who may be the most disenfranchised patients in medicine.
What is very clear is that B. burgdorferi has been isolated alive from both animals and humans despite administration of antibiotics that are deemed curative by IDSA and CDC. More alarmingly, this organism has also been isolated alive from humans after antibiotic therapies measured in many months to years, when the administered therapies are far in excess of what is declared curative by these same agencies.
Much of the medical literature in which B. burgdoferi was isolated from patients despite "appropriate" antibiotic therapy was penned by IDSA Guidelines authors. These medical journal articles were either not referenced in the IDSA Guidelines, or when they were, they did not specifically refer to the aspects that document persistent borrelial infection despite antibiotic therapy. Despite this, CDC gives preference to IDSA Guidelines and does not endorse ILADS Guidelines.
In response to the claim regarding the second basic idea, ILADS Guidelines provide a heuristic algorithm for optimal treatment of an extremely heterogeneous group of very ill patients given the best available published medical literature on the topic. Inherent to this, there must be a place for sound clinical judgment, which may be the freedom to which Dr. Mead is referring. I fear the day when individualized clinical judgment is replaced by medical guidelines of any kind. You simply can't rigidly standardize Lyme disease treatment due to strain heterogeneity, co-infections, and differences in immune system types among patients.
Mead: "...and perhaps more importantly, we are concerned about patients who are being treated with unproven therapies; with therapies that are sometimes harmful. There are case reports of patients who died as the result of long term therapies for Lyme disease when in fact there wasn't even necessarily evidence that they were infected."
Phillips: I think one thing that we can all agree upon is that we want to minimize risks to patients from treatments and maximize benefits, which is why ILADS has physician training programs.
Therapies for most serious diseases likewise have potentially serious side effects, but the risk of fatality from long-term antibiotic therapy is quite low. For example, far more fatalities have been caused by Lyme disease than by its treatment.
The risk of fatal outcomes in the treatment of inflammatory diseases with immunosuppressive agents and cancer with chemotherapy is far higher than that with antibiotic therapy, but the difference in those diseases is that they are well accepted by CDC as legitimate so that the risk is deemed worthy. I would further counter that in the rare case reports of patients who died during therapy for Lyme disease, there was indeed evidence of infection, albeit not meeting strict CDC surveillance reporting criteria in all cases.
Although it's correct that some therapies being used to treat chronic Lyme patients are of unproven efficacy, the high rates of treatment failures of even early, and most certainly late Lyme, are quite well-documented in the medical literature. It has been demonstrated in study after study without equivocation that short-term antibiotics are not effective in a great many cases. This, coupled with the documentation of microbial persistence despite such short term antibiotic therapies, makes the case for longer and better treatments. It's not uncommon in medicine for innovative therapies to be used in patients before well-designed trials document their utility. I see it all the time in other diseases.
Mead: "the reports of cases of Lyme being mistaken for other diseases are really quite rare"
Phillips: There are a great many cases in the published medical literature where Lyme disease has been mistaken for other diseases. Of course, in these cases the diagnosis of Lyme was eventually made, hence the ability to have the published reports. The natural question which follows is how many patients never get appropriately diagnosed with Lyme, never get treated, and remain in the category of 'other disease'.
In my clinical experience of treating over 20,000 patients, I see patients come in with diagnoses of fibromyalgia, chronic fatigue syndrome, various inflammatory disorders, cardiac symptoms, and MS on a regular basis. When I find evidence for Lyme and/or other zoonotic infection such as bartonella, the large majority of these patients markedly improve with antibiotic therapy.
Mead: "...patients who have been told they have Lyme and are then treated for Lyme, when in fact they have other things: pituitary adenomas, other types of malignancy, and some of those patients were seriously harmed by being given a diagnosis and long term therapy for Lyme disease when in fact their underlying condition went untreated."
Phillips: Medical mistakes will occur as long as the practice of medicine exists. Of that I'm certain. It's crucial for physicians in any area of medicine to pursue the correct diagnoses. But again, if we examine the published literature, there are relatively few reports of Lyme being diagnosed instead of another serious diagnosis which was missed.
Mead: "Why there seems to be so much controversy, I'm not really entirely clear."
"There is a lot of misinformation about Lyme disease, about basic facts about it that are spread a great deal on the internet."
"I fear that sometimes if people have heard something enough times, they come to really believe that it must be true."
Phillips: I very much believe that Dr. Mead is earnest in his statements. But then again, I'm conflicted because such statements appear to me to be myopic at best. How could this intelligent physician working at the highest levels of CDC not realize why there is so much controversy?
Bias blind spot in medical terminology refers to recognizing the impact of biases on the judgment of others, while failing to see the impact of biases on one's own judgment. Published studies demonstrate that bias blind spot is pervasive in humans, begins in childhood, and is greater in individuals with greater cognitive ability. Confirmation bias in medical terminology refers to the tendency to interpret new evidence as confirmation of one's existing beliefs or theories. Published evidence documents that confirmation bias is very common and can lead to implacable decision making.
I think that it's all too easy for intelligent individuals in large groups to fall prey to these biases.
Mead: "There have been a number of studies, including one just published recently in the New England Journal from the Netherlands, which have looked at longer courses of antibiotic therapy for these patients, and of course what they find is that when patients get that therapy that they improve. The problem is that the patients who got the placebo also improved at the same rate. So I think it's a little bit misleading to suggest that there's been no science done on this problem."
"We can't just dismiss out of hand the numerous studies that have been done."
"We certainly do recognize that there are patients who've had Lyme disease and who have been treated and who will have persistent subjective symptoms, as I mentioned fatigue, some difficulty with their sleep, and with thinking, and with muscle aches and pains.
Those patients we refer to has having 'post-treatment Lyme disease syndrome'. So we recognize that that condition exists. The fundamental question is what is the cause of that condition? Is it a persistent infection? Or is it a complication of prior infection? And what is the best treatment for it. Is it long-term antibiotics or not? On the first question, we don't know the answer. It's possible that it's either one of those. We do have data on the second question. As I mentioned, a number of placebo-controlled trials which have looked at patients with 'post-treatment Lyme disease syndrome' and given them various courses of prolonged antibiotic and the bottom line of those studies is that overall they do not seem to benefit those patients in the long run."
Phillips: The gold standard in microbiology for diagnosing an infectious disease has always been to culture the organism alive. Despite notorious difficulties in culturing B. burgdorferi, in about 30 studies this organism has been cultured alive from patients despite at least standard antibiotic therapy, and in many cases after antibiotics far in excess of what is deemed curative by IDSA and CDC.
If the pathogen that causes a disease is still present in conjunction with symptoms compatible with that infection, it would appear to me that these 'fundamental questions about the cause of long term symptoms' should have been answered a very long time ago. To add insult to injury, recent studies from Tulane, Johns Hopkins, and Northeastern University all demonstrate that we can't even kill B. burgdorferi in the test tube with the currently recommended antibiotics.
I cannot in good conscience use the term 'post-treatment Lyme disease syndrome' given the wealth of published information that this organism persists. It is, I believe, by its very definition, an illogical construct. What are the chances that a second disease of mysterious etiology but with the same symptoms as the first disease, would come and replace the first disease when there is published evidence that the pathogen which causes the first disease persists despite both short and long-term antibiotics?
There are numerous chronic bacterial infections which require long-term combination antibiotic therapies: Tuberculosis, leprosy, coxiella endcocarditis, brucellosis, Whipple's. Why should Lyme be different?
By referring to patients with persistent symptoms of Lyme disease after a short course of antibiotics as 'post-treatment Lyme disease syndrome' this may produce a de facto fait accompli, in that such patients, when desperately searching for answers on the CDC website, may feel that antibiotics can't possibly help them.
This may only delay their care further and increase the likelihood of subsequent antibiotic treatment failure. Because semantics guide patient care, I believe that the term 'post-treatment Lyme disease syndrome' is harmful.
In response to the second question, it's true that there have been a number of studies addressing the response to antibiotics vs placebo of patients with persistent symptoms after Lyme disease, but I think that it's very important to specify what that number is and what those studies showed. There have only been 3 NIH-funded randomized placebo controlled studies looking at this problem since 2001. New Lyme infections as estimated by CDC are over 329,000 cases per year, which is more than new diagnoses of invasive breast cancer and HIV combined, yet the NIH has only funded 3 studies on this topic, and two of those 3 studies have demonstrated responses to antibiotics, albeit imperfect responses, as explored in the following discussion.
The first study was by Klempner. This study evaluated antibiotic vs placebo. The study was terminated early due to the determined likelihood that a beneficial effect would not be found. When this was critically analyzed with biostatistical methods, an article was published which I believe demonstrates that Klempner's study was so poorly designed and analyzed that in order for a treatment effect to have been observed, the antibiotic treated patients would have had to improve to a level of health which was a full standard deviation better than the average health of the general population. It's a reasonable hope for antibiotics to return a patient to a somewhat normal life; it's not a reasonable hope that they would improve that patient's health status to better than average.
The second study was by Krupp. It showed a reduction in fatigue in patients treated with antibiotics and not with placebo. There was no improvement in cognitive functioning. A biostatistical analysis demonstrated that fatigue was the only outcome of the study for which it was properly designed.
The third study was by Fallon. This was, in my opinion, the best designed of the 3 trials. It demonstrated improved cognition in antibiotic treated patients and not placebo, but these patients relapsed when antibiotics were discontinued. There were also benefits to antibiotic treated patients in fatigue and body pain by subgroup analysis.
Another study was performed by Cameron, which demonstrated benefits to antibiotics and not placebo across quality of life assessments. However, this study may have had statistical issues with baseline randomization.
The PLEASE study from The Netherlands to which Dr. Mead is referring is, in my opinion, a study looking for a question to answer but failing to find one. When studies are designed, they must be designed thoughtfully to answer important questions. An important question which requires further study is whether longer and innovative antibiotic treatment regimens are superior to placebo in patients with chronic Lyme symptoms after a previous short course of antibiotics. The PLEASE study did not have a true placebo group in that all patients were treated with antibiotics. Further, the patient population was heterogeneous in that although most patients had been treated previously with antibiotics, some had not been previously treated.
The placebo aspect of this study was such that after 2 weeks of IV antibiotics, the patients then received further oral antibiotics vs placebo. There was no benefit to oral antibiotics piggybacked directly onto the IV antibiotics. Again, very little useful information, if anything, in my opinion, can be gleaned from a study using a heterogeneous patient population in which all of them received some form of antibiotic.
Mead: "We are very concerned about patients who are ill, both those patients with 'post-treatment Lyme disease syndrome as well as those patients who may have been misdiagnosed with it."
Phillips: The only way for real progress to be made is to consider a problem from all sides with, to paraphrase from Buddhism, the proverbial 'new eyes'. Abraham Lincoln famously countered confirmation bias by forming his cabinet with those that publicly disagreed with him. Why isn't the CDC doing something similar? Why isn't the IDSA?
IDSA physicians, assuming that they follow IDSA Guidelines, are not able to see the beneficial effect of following ILADS Guidelines because they don't treat in that way. However, ILADS physicians see the other side of the equation every day in the form of innumerable treatment failures after short term and limited range antibiotic treatment options.
I think it's very compelling that a study funded by CDC and published in 2015 demonstrated that the majority of US physicians polled treat Lyme disease with antibiotics for longer than 4 weeks, which is in excess of what IDSA Guidelines recommend.
If someone of a CDC/IDSA centric view were to have confirmation bias, they may interpret this data as a wake-up call to better educate physicians on the proper way to treat Lyme disease.
However, a more neutral approach might be to ask oneself if there is a reason that most physicians aren't following IDSA Guidelines. Perhaps it's because they don't work well in curing patients.