THE TICK THAT BIT ME

Reactive Arthritis and the Chronic Lyme Disease Debate: Dissecting the Research 11/20/2012

As you may be aware, there is a great deal of controversy about persistence of Lyme disease symptoms after treatment—even when treatment involves long-term antibiotics. There are many theories about this, and several of them may be true, depending on the patient. The Centers for Disease Control (CDC) and the Infectious Diseases Society of America (IDSA) say that once you’ve been treated with a month of antibiotics, you no longer have a Borrelia burgdorferi infection. If you’re still experiencing symptoms, they call it Post-Lyme or antibiotic-refractory arthritis. At the other end of the spectrum, some Lyme-literate medical doctors (LLMDs) believe that Lyme is a chronic disease, and once you have it, you’ll have it for life. Long-term (read: indefinite) antibiotic treatment, they maintain, is necessary to keep the organisms from multiplying, but you’ll never fully be rid of them.

My views fall somewhere between these two extremes. (Disclaimer: I am not a medical professional.) On the one hand, I don’t think that four weeks of antibiotics, whether oral or intravenous, is really enough to kill off a Borrelia infection in a patient who’s been infected for years. (And this view is supported by Embers et al’s study of Rhesus macaques.) On the other hand, I don’t think that antibiotics-for-life is the answer either. There are just too many people who have been on antibiotics for years who don’t seem to be getting better. Plus there’s the fact that antibiotics can cause a lot of damage if you take them long-term, which significantly lowers the quality of life for people on these treatment regimens.

So when patients have been treated and they’re still experiencing symptoms, I see several possible explanations:

1) The antibiotics didn’t kill off all the bacteria, and they are still hanging around somewhere—perhaps hidden in joints, cartilage, or the brain. This, as you can imagine, is very difficult to prove, especially in living human beings.

2) The antibiotics killed off all the bacteria, but the patient was bitten by another tick and re-infected. This is highly possible if the patient’s environment, lifestyle, and preventive measures have not changed. It’s also difficult to detect when patients are only being follow-up tested with Western Blots, and not something like the C6 antibody assay, which gives you a titer so you can see if your antibodies to the bacteria suddenly increase. (A study related to reinfection was just published on Wednesday in the New England Journal of Medicine. You can read about it in the NY Times here, or read the study abstract. I’ll be doing a run-down of that study next week.)

3) The antibiotics killed off the bacteria, but the body is still making an immune response, possibly attacking its own cells, causing inflammation and continuing symptoms.

When new research comes out, I like to pay attention to see which of these explanations is supported and why. In this post, I’ll take you through a study called “Spirochete antigens persist near cartilage after murine Lyme borreliosis therapy” that was conducted by some researchers at Yale and published in the Journal of Clinical Investigation back in June. (If you want to read along, you can access the full article here.)

Some Background:

The underlying question in this study is: what causes Lyme-associated arthritis in patients who have been treated with antibiotics? Is it that infectious spirochetes are still hiding somewhere in the body, or is it possible that antibiotics “kill” all the bacteria (read: disassemble them so they can’t multiply) but leave their building blocks (referred to as antigens because the body still detects them) behind, causing inflammation.

Let me explain how killing Borrelia in the human body works. The things you need to kill Borrelia are antibodies, first and foremost. If you don’t generate an IgM response, this infection can be fatal in the first 3 to 7 days. We know that IgM has a direct bactericidal effect. In other words, this antibody can kill Borrelia on its own in the absence of complement. IgG, by comparison, is very inefficient at killing Borrelia, but we make that, too. We also need phagocytes to kill Borrelia, and in order to generate antibodies, we need B cells that work. Another thing we need is toll-like receptors (TLRs). These are important for helping antibodies bind to pathogens or parts of pathogens. If we’re deficient in TLR2, or a certain molecule in TLR called myeloid differentiation antigen 88 (abbreviated MyD88), we can have an overwhelming infection.

What they did:

In this study, the researchers used a type of mouse in which this MyD88 protein has been knocked out—i.e. the mouse is totally deficient in MyD88. For that reason, they call it a Myd88-/- mouse. (No, that thing at the end is not an emoticon.) The problem with these mice is that they die quickly of opportunistic infection (specifically, Pneumocystis carinii), so in the lab, they have to give them an antibiotic called Sulfatrim or Septra, which is actually a combination of two antibiotics—sulfamethoxazole and trimethoprim. I’ll come back to why this is important a little later.

Sorry, little guy. (Image via jaxmice.jax.org)

Now, an interesting thing is that this same group of researchers did a study back in 2002 using C3H mice, who don’t have the Myd88 protein knocked out—so essentially “normal” mice, in terms of their immune systems—and they showed that live Borrelia persist for 3 months after optimal treatment. These Borrelia that remained were sort of mutant Borrelia because they were missing a couple of proteins that they might need to infect other animals. The researchers knew this because the clean ticks that bit them got B. burgdorferi, but when they had those ticks bite healthy mice, those mice didn’t get infected.

Anyway, for some reason, these researchers didn’t want to use the C3H mice for this study, and they decided to use the Myd88-/- mice, who develop overwhelming infections. They also used a WT strain of mouse (which is not missing Myd88) for the sake of comparison. They infected both types of mice, and then they treated some of them with Doxycycline (through their water supply) and others with Ceftriaxone (via subcutaneous injection). Interestingly, they showed that after one day of Ceftriaxone therapy, they could kill all of the Borrelia. They also used an interesting microscopic technique which allows one to look real-time at tissue and watch an organism to see what it does. By cutting down to a mouse’s tendon, they can see what’s going on down there.

Their Findings:

“B. burgdorferi DNA can be detected in B6 Myd88-/-, but not WT, mice after treatment with Doxycycline” (p. 2).

Translation: After treating the mice with Doxycycline, the researchers couldn’t find any B. burgdorferi DNA in the normal immune system mice, but they could find the DNA in the immune-compromised mice. One of those mice had a positive culture for B. burgdorferi, and ticks that fed on that mouse also tested positive for B. burgdorferi. HOWEVER, when they took samples from the knee joints of the mice, ALL of the mice tested positive for the ospA plasmid (B. burgdorferi DNA). In addition, ear-skin samples from half the immune-compromised mice and one of the normal mice tested positive for B. burgdorferi DNA (p. 2).

“Real-time imaging of B. burgdorferi in Myd88-/- mice reveals rapid spirochete elimination after antibiotic therapy” (p. 2).

Translation: The researchers used intravital 2-photon microscopy to observe the behavior of the B. burgdorferi spirochetes in the infected mice. Specifically, they looked at the dermis (skin) and the calcaneal (achilles) tendons. They say that 24 hours after beginning treatment with Ceftriaxone, the number of spirochetes had “diminished dramatically” in both the skin and tendons. The spirochetes left behind in the skin appeared to be moving, but the ones in the tendons did not. The following day, they were not able to see any spirochetes using this imaging technique, suggesting that they had all been destroyed.

“Spirochete antigens can be detected adjacent to ear cartilage in antibiotic treated Myd88-/- mice” (p. 3).

Translation: At the end of the study, the researchers took tissue samples from the ears of all the mice. They tested these samples for B. burgdorferi using both immunofluorescence staining (looking for antibodies) and culture techniques. The mice who were “sham treated” (not given antibiotics) tested positive. The mice who were treated with Ceftriaxone had negative cultures, but some “spirochete antigens” were detected in a deep layer of skin next to the ear cartilage. By “spirochete antigens,” they mean not live spirochetes but proteins (building blocks) left over from the bacteria that can cause the immune system to react. These antigens were found at a deeper level than where the imaging had earlier been peformed (which explains why they weren’t detected using imaging). The same antigens were detected in ear tissue from mice treated with Doxycycline. The researchers conclude that because the antigens were detected, but the bacteria could not be cultured (grown), it means whatever these spirochete remains were, they were incapable of multiplying because they had been too damaged by the antibiotics.

“Live imaging reveals antigen deposits but not motile spirochetes adjacent to cartilage of Myd88-/- mice after Doxycycline treatment for B. burgdorferi infection” (p. 3).

Translation: A separate experiment was conducted in which researchers studied mice between 2 and 10 weeks after finishing a 1 month course of Doxycycline. They used a technique called xenodiagnosis, where they let clean ticks bite the mice. They could find some B. burgdorferi DNA in the ticks that fed on the mice treated with Doxycycline, but when they studied the contents of the ticks’ guts, they could not find any spirochetes. By contrast, the ticks that fed on mice not treated with antibiotics had the bacteria in their bellies. In addition to using xenodiagnosis, they did the immunofluorescence test and cultures on this group of mice, and as before the cultures were negative and the immunofluorescence found antigens near the ear cartilage. This time, they used their imaging technique to look deeper under the skin, closer to the ear cartilage. In the sham treated mice, they found that there were motile (alive) spirochetes right next to the cartilage and large “deposits of nonmotile fluorescent material” where the skin meets the cartilage. In the antibiotic-treated mice, they saw no live spirochetes, but the same deposits were present next to the cartilage. So to see if these deposits would cause infection, the researchers transplanted skin from the infected mice (both those treated with Doxycycline and the sham-treated ones) into non-infected mice. Only the skin from sham-treated mice caused infection in the new mice. This shows that the antigens adjacent to ear cartilage in the mice treated with Doxycycline were not infectious.

“Spirochete antigens can be detected in joints of antibiotic-treated C3H Myd88-/- mice” (p. 4).

Translation: Here, the researchers decided to look at the knee joints of the mice to see if the same antigen deposits exist in antibiotic-treated mice. They looked at the knees of mice that had been infected for 4 months (which is a long time considering mice only live for about a year). When they treated these mice with Ceftriaxone, intravital microscopy showed that the spirochetes died off, but debris was left behind. They are pretty sure the spirochetes died off because the cultures were negative.

“Tissues from antibiotic-treated mice contain immunogenic and inflammatory B. burgdorferi antigens” (p. 4).

Translation: Finally, they wanted to test whether the deposits left behind in the knees of the mice actually contained B. burgdorferi antigens. They did this by immunizing new mice with knee tissue from the infected mice. They found that both the tissue from sham-treated and antibiotic-treated mice caused an IgG immune response to several B. burgdorferi proteins in the new mice.

Problems with the study:

1. It’s in mice. If you follow the research on B. burgdorferi, you’ll see that many of the studies are done in mice. That’s because it’s much less expensive to study disease in mice than in other animals. However, if we really want to learn about arthritis and B. burgdorferi in the human body, it would be better to do a study like this in Rhesus monkeys, which are much more similar to humans. Hopefully, this study will make it possible for some researchers to try to replicate this work in a primate model so that we can learn more.

2. The use of a “lab” strain of B. burgdorferi. They used a 297 strain of Borrelia burgdorferi which is “stable”–in other words, it’s not changing. It’s old and predictable. The problem is that most Borrelia in the wild are likely mutating and changing. They could even be developing resistance to antibiotics. After all, these organisms have been around for thousands of years; they are masters of adaptation. It would be much more interesting to do a study like this using a “wild” strain of B. burgdorferi, as this would more closely mimic the average patient’s experience.

3. The ambiguous blot analysis. The researchers used their own immunoblot to look at the antigens in the patellas (knees) of the mice. It’s not clear to me why they didn’t just use a Western Blot (since that’s the test used on us humans). Another odd thing they did was use a dilution of 1:1,000 for the blot, which is ten times the dilution used for other blots. Perhaps a lower dilution showed too many similarities between the sham group and the antibiotic group? In any case, for a study like this, I would expect more justification for these unusual choices.

4. The study’s “lack of heart” (and brain). I’m referring to the fact that the researchers failed to examine the effect of B. burgdorferi in the hearts of the mice. We know that, in addition to knee problems, patients with B. burgdorferi infections are at risk for a variety of heart problems, including myocarditis. The researchers were so bent on showing that antibiotics could kill the bacteria near joints, but what does that matter if the heart is still infected? If they truly believe that 24 hours of Ceftriaxone in mice eliminates B. burgdorferi, they missed a golden opportunity to show it by neglecting to examine the hearts–and the brains, for that matter. Not only did they ignore the heart, but they wasted valuable word space in their discussion section attacking the research design of Embers et al’s 2012 study of B. burgdorferi infection in Rhesus macaques. Now, the infamous monkey study is far from perfect, but they did do one thing right, which was to look at the heart tissue of the monkeys post-mortem–and guess what they found? In 3 out of 12 monkeys who were treated with antibiotics (that’s 25%), B. burgdorferi RNA could be detected in the heart. That’s despite the fact that in all 12 of the treated animals, C6 antibody titers decreased steadily over the course of treatment.

5. The use of septra/Sulfatrim. This one is a doozy. Earlier, I mentioned that the researchers added the antibiotic sulfamethoxazole-trimethoprim (Sulfatrim) to the mice’s drinking water “to reduce opportunistic infection”(p. 7). They claim that this drug has “no effect on B. burgdorferi infection or disease”(p. 7). I’m guessing they think that because of this 1996 study done in Austria. In that study, several species of Borrelia were evaluated to see whether they were susceptible to amoxicillin, azithromycin, cefotaxime, ceftriaxone, doxycycline, penicillin G sodium, roxithromycin, and trimethoprim-sulfamethoxazole (Sulfatrim) in vitro. The researchers used 30 different strains of Borrelia, but only 4 of those were Borrelia burgdorferi, and they were European B. burgdorferi at that. They found that B. burgdorferi was resistant to trimethoprim-sulfamethoxazole. Now, even though both ceftriaxone and trimethoprim-sulfamethoxazole were studied, the Austrian researchers didn’t investigate what would happen if you used both of these drugs on the Borrelia at the same time (which is what was done in the Yale study). In fact, after scouring PubMed, I was unable to find any synergistic Borrelia studies using ceftriaxone and trimethoprim-sulfamethoxazole. I did, however, find this study, also from Austria, published in 1997. They found that “trimethoprim was more active against Borrelia burgdorferi than against a sensitive strain of Escherichia coli, but sulfamethoxazole was not active against Borrelia burgdorferi.” In other words, one of the drugs that makes up Sulfatrim kills Borrelia, and the other doesn’t. The question is, if you add Ceftriaxone, does Sulfatrim start killing the Borrelia? To actually know whether or not Sulfatrim has an effect on B. burgdorferi when combined with Ceftriaxone, our Yale researchers would have had to do a synergy study, to see what would happen if you took the 50% kill rate of Doxycycline and Ceftriaxone, and add Sulfatrim. But they didn’t show that adding Sulfatrim didn’t affect the kill. So when they’re saying that “Ceftriaxone rapidly [within 24 hours] reduces pathogen burden in the skin,”(p. 5) they’re not taking into account that the Sulfatrim in the drinking water is also probably helping kill off spirochetes. It’s ironic how critical they are of other studies when their own study isn’t exactly “clean.”

Doxycycline (Image via en.wikipedia.org)

6. The way Doxycycline was administered. Doxycycline was given to the mice in their drinking water. This means that the amount of Doxycycline in each mouse’s system depended on how much water it drank. The researchers said their reason for doing the Doxy in the water instead of force-feeding it to the mice twice a day was that when it was given twice daily, serum drug levels fell too low and they were not able to kill all the bacteria (p. 5). When I read this, I thought to myself, “Well, let’s see, how many humans do I know who are taking their Doxycycline through their drinking water? Oh, that’s right. None.” So here we have a study of Borrelia burgdorferi infection in mice in which the researchers choose not to give oral antibiotics because they believe not that the drug doesn’t work, but that the drug delivery system doesn’t work because it can’t get a high enough level of the drug into the blood stream. Yet, these are the same doctors who are saying that one month of oral Doxycycline should be enough to treat the same infection in humans. Curious, isn’t it? They even admit that one mouse may have stayed sick “due to a drinking pattern that led to inconsistent Doxycycline levels” (p. 5). So I guess either that mouse just wasn’t as thirsty as all the other mice, or he was eating his food and drinking his water in one sitting, and the food interfered with the drug absorption (as it can in people!).

(via valdo.com)

7. They didn’t treat the arthritis. Okay, I get that the researchers were having lots of fun with their innovative real-time imaging technique. They tried to accomplish a lot with this study, and it already appears that they may have spread themselves a little thin. However, it bothers me that they spent no time examining ways to treat the arthritis caused by the deposits left behind by B. burgdorferi. In the world outside the laboratory, it doesn’t so much matter to people whether their arthritis is caused by live spirochetes or dead ones. They want to know what’s going to make them feel better. The study’s authors suggest that more antibiotics likely won’t work, but they don’t explore any alternatives, like steroids, for treating Lyme arthritis.

Some interesting (and some unexpected) implications:

1. Cool pictures. Intravital microscopy, the real-time imaging technology used in this study, is pretty nifty, and could be used in better-designed studies to find out a lot of useful information. The researchers in this study could even see some of the spirochetes changing into spherical forms, but they didn’t really investigate or discuss this in detail, beyond saying they don’t think those forms are bacterial cysts. It might be useful to have an entire study dedicated to investigating that.

Image sequences of Bb914 (Borrelia burgdorferi) changing from an elongated spirochete to a spherical form. (Image via ncbi.nlm.nih.gov; J Clin Invest. 2012 July 2; 122(7): 2652–2660.)

2. Rethinking oral Doxycycline. It’s been my belief for a while now that oral antibiotics are just not as effective at killing Borrelia as IV antibiotics like Ceftriaxone. I’m not expecting everyone to agree with me on this, but let me tell you why I think so. While Doxycycline is by far the best choice among oral antibiotics for killing Borrelia (as it’s better at crossing the blood-brain barrier than many other drugs), there is an inherent disadvantage to all oral drugs because they have to be delivered through our digestive system. As a patient who took Doxy for a month, I can tell you that no matter how responsible and organized you are, it is very difficult to eat meals at the same time every day, to space the doses 12 hours apart, and to avoid ingesting things like milk that interfere with drug absorption. Reading Bockenstedt et al’s article made me further question the effectiveness of oral Doxy, as the researchers decided that oral Doxy twice per day would not be enough to keep serum drug levels consistent. Instead, they opted to deliver the drug through the mice’s water supply, which poses other problems with consistency. In any case, if it’s not good enough for lab mice, I don’t see how it’s good enough for humans.

3. A new drug combo? The study shows that in the presence of trimethoprim, you can rapidly kill Borrelia with Ceftriaxone. That means we should be doing more studies on how this works and whether it is safe for humans. There is always concern with killing off bacteria too rapidly because macrophages need time to clear the debris (which we think causes arthritis). However, this drug combination seems worth examining in other laboratory studies.

4. Location, location, location. The study doesn’t show that cartilage can be infected with Borrelia burgdorferi, but it does show that deposits are left over near cartilage after the bacteria have been disassembled by antibiotics. If the infection is in cartilage, that’s bad news, because there is no blood flow to cartilage, so it’s very difficult to eradicate an infection there. We need more studies that examine how this bacterium acts around cartilage.

5. Chronic Lyme? Contrary to what Yale alum and journalist Carole Bass would have you believe (Thanks to Becki from Bloody Lymey for opening my eyes to that one.), this study neither proves nor disproves the existence of Chronic Lyme disease, so despite the agenda that may underlie this study, patients need not see it as a threat. The study authors themselves admit in their Discussion section that they’re not quite sure what all their data mean: “The significance of B. burgdorferi DNA in xenodiagnostic ticks and in mouse tissues after antibiotic therapy is unclear” (p. 5). One possibility is that “Some B. burgdorferi DNA could remain intact if it is sequestered in cellular debris such as the GFP deposits.” They’re saying they think that the B. burgdorferi DNA they detected is just remnants of dead spirochetes that were preserved because they were stuck in the debris left behind by the antibiotics. However, they also admit to another possibility: “Alternatively, spirochete DNA could represent a minor subpopulation of B. burgdorferi that is not killed by the antibiotic treatment.” It’s a one-sentence admission in a 9-page paper, but it’s there–and it means that despite what these researchers think is going on, they still can’t say with 100% certainty that the antibiotics completely eradicated the infection.

What the study does show is that there are deposits in mouse tissue that the researchers insinuate are dead organisms (they have Borrelia antigens, are immunogenic, and don’t appear to be infectious). Because they don’t examine all of the tissue–including the cartilage, the heart, and the brain–it’s difficult to say whether they have completely eliminated the bacteria with antibiotics. What Embers et al showed in their primate model is that there seems to be persistence of spirochetes following 4 weeks of IV Ceftriaxone treatment and 8 weeks of Doxycycline. Until somebody does another study in Rhesus monkeys and proves that they’re wrong, that study stands.

6. Treating the arthritis. We know that reactive arthritis caused by Borrelia infections is a real phenomenon, and this study suggests that the cause is the debris left behind by spirochetes following antibiotic treatment. However, what patients and doctors alike need is access to information about how best to treat this unique form of arthritis. I’ve heard anecdotes from patients and doctors about the helpfulness of steroids like prednisone during or following antibiotic treatment, but there really isn’t enough research being done on this. It would be nice if researchers on both sides of the Chronic Lyme debate would pool their resources for the sake of better patient care.

I hope you enjoyed this installment of Tick-Lit Tuesday. It’s good to be back.

What has been your experience with Lyme or Tick-borne Relapsing Fever and reactive arthritis? What questions would you like to see addressed in future research?

Related articles

• Ceftriaxone (Rocephin): Is your doctor following directions?
• Arthralgias, myalgias, and herx—oh my! (Symptom vocabulary for TBIDs)
• New Infection, Not Relapse, Brings Back Symptoms of Lyme Disease, Study Finds (nytimes.com)