Lyme Disease Antibiotic Treatment Research

For early Lyme disease, a short course of oral antibiotics, such as doxycycline or amoxicillin, cures the majority of cases. In more complicated cases, Lyme disease can usually be successfully treated with three to four weeks of antibiotic therapy.

After being treated for Lyme disease, a minority of patients may still report non-specific symptoms, including persistent pain, joint and muscle aches, fatigue, impaired cognitive function, or unexplained numbness. These patients often show no evidence of active infection and may be diagnosed with post-treatment Lyme disease syndrome (PTLDS). In patients with PTLDS, studies have shown that more antibiotic therapy is not beneficial and the risks outweigh the benefits.

Clinical Studies

NIAID has funded three placebo-controlled clinical trials to learn more about the efficacy of prolonged antibiotic therapy for treating PLDS. The published results were subjected to rigorous statistical, editorial, and scientific peer review.

These trials were designed to ensure that several key parameters were addressed:

  • The susceptibility of Borrelia burgdorferi, the bacterium that causes Lyme disease, to specific antibiotics
  • The ability of antibiotics to cross the blood-brain barrier, access the central nervous system, and persist at effective levels throughout the course of therapy
  • The ability of antibiotics to kill bacteria living both outside and inside mammalian cells
  • The safety and welfare of patients enrolled in the trials

The first clinical trial, which included two studies conducted at multiple research sites, provided no evidence that extended antibiotic treatment is beneficial (New Engl J Med 345: 85-92, 2001). In those studies, physicians examined long-term antibiotic therapy in patients with a well-documented history of previous Lyme disease, but who reported persistent pain, fatigue, impaired cognitive function, or unexplained numbness. Patients were treated with 30 days of an intravenous antibiotic followed by 60 days of treatment with an oral antibiotic.

These studies did reinforce the evidence that patients reporting PTLDS symptoms have a severe impairment in overall physical health and quality of life. However, prolonged antibiotic therapy showed no benefit when compared with groups who received placebo.

In another study, published in 2003, researchers examined the effect of 28 days of intravenous antibiotic compared with placebo in 55 patients reporting persistent, severe fatigue at least six months following treatment for laboratory-diagnosed Lyme disease. Patients were assessed for improvements in self-reported fatigue and cognitive function (Neurology 60: 1923-30, 2003).

In that study, people receiving antibiotics did report a greater improvement in fatigue than those on placebo. However, no benefit to cognitive function was observed. In addition, six of the study participants had serious adverse events associated with intravenous antibiotic use, and four patients required hospitalization. Overall, the study authors concluded that additional antibiotic therapy for PTLDS was not supported by the evidence.

More recently, a study supported by the National Institute of Neurological Disorders and Stroke (NINDS), also part of the National Institutes of Health, again showed that long-term antibiotic use for Lyme disease is not an effective strategy for cognitive improvement (Neurology 70(13): 992-1003, 2008). Researchers studied 37 individuals with well-documented Lyme disease, at least three weeks of prior intravenous antibiotics, a current positive Lyme test (IgG Western blot), and self-reported memory impairment. Patients received either 10 weeks of intravenous ceftriaxone or intravenous placebo, and were examined for clinical improvement using objective memory tests. In a complicated statistical model, the ceftriaxone group showed a slightly greater improvement at 12 weeks, but at 24 weeks both the ceftriaxone and the placebo groups had improved similarly from baseline. In addition, adverse effects attributed to intravenous ceftriaxone occurred in 26 percent of patients. The authors concluded that because of the limited duration of the cognitive improvement and the risks involved, 10 weeks of intravenous ceftriaxone was not an effective strategy for cognitive improvement in these patients, and more durable and safer treatment strategies are still needed.

For more information, please see the Chronic Lyme Disease section.

Animal Models

Animal models have provided considerable information on the transmission and pathogenesis of Lyme disease, as well as on the mechanisms involved in the development of protective immunity. Studies of the effects of antibiotic therapy in animals infected with B. burgdorferi have been conducted most often with mice but also with hamsters, gerbils, dogs, and non-human primates. (See Wormser and Schwartz, Clin Microbiol Rev 22(3): 387-395, 2009 for a recent review.)

NIAID-supported studies have shown that B. burgdorferi can be detected in mice for at least three months after treatment with therapeutic doses of various antibiotics (ceftriaxone, doxycycline, or azithromycin). In these studies, surviving bacteria could not be transmitted to healthy mice, and some lacked genes associated with infectivity. By six months, antibiotic-treated mice no longer tested positive for the presence of B. burgdorferi, even when their immune systems were suppressed. Nine months after antibiotic treatment, low levels of Borrelia DNA still could be detected in some—but not all—of the mice (J Infect Dis 186: 1430, 2002). These findings indicate that noninfectious B. burgdorferi can persist for a limited period of time after antibiotic therapy. The implications of these findings in terms of persistent infection and the nature of PTLDS in humans still need to be evaluated.

In a 2008 study, mice were treated with the antibiotic ceftriaxone for one month following either early or later stages of infection with B. burgdorferi. After completing antibiotic treatment, tissues from the mice were analyzed using a variety of scientific methods. B. burgdorferi could not be grown in cell cultures for any of the mice treated with antibiotic. In a small number of the treated mice, however, bacteria were detected in their tissues. Further, when Ixodes scapularis ticks fed on some of the antibiotic-treated mice, the ticks were able to transmit the bacteria to mice with weakened immune systems who were not previously infected with B. burgdorferi.Highly sensitive laboratory tests were able to detect the presence of B. burgdorferi in these mice. However, the bacteria did not grow in cell cultures (Antimicrob Agents Chemother 52: 1728-1736 , 2008). The implications of these findings in terms of persistent infection and the nature of PTLDS in humans are yet to be fully understood.

Several recent studies suggest that B. burgdorferi may persist in animals after antibiotic therapy. In one study, NIAID-supported scientists found that remnants of B. burgdorferi remained in mice after antibiotic treatment (J Clin Invest 122(7):2652-60, 2012). Another team of NIAID-supported investigators found that intact B. burgdorferi persist in nonhuman primates after antibiotic treatment. It was not possible to culture these bacteria and it is not clear whether they are infectious. (PLoS One 7(1): e29914, 2012). More recent work by Hodzic et al. (PLoS One 9(1):e86907, 2014) replicated the earlier finding of persisting DNA but non-cultivatable B. burgdorferi after antibiotic treatment using a mouse model. Additional research is needed and continues to be supported by NIAID to learn more about persistent infection in animal models and its potential implication for human disease.

NIAID, in collaboration with NINDS, supported comprehensive studies on non-human primates looking at the neuropathology associated with post-treatment Lyme disease syndrome (Ann Neurol 56: 361, 2004). A major goal of these studies was to optimize the Rhesus monkey model of Lyme disease as well as to determine the pathogenesis of the disease with a focus on its neurological manifestations.

The results of these studies have expanded the knowledge of those factors that contribute to the pathology associated with B. burgdorferi infection of the central nervous system. Among the findings are the following:

  • Non-human primates are easily infected with low levels of B. burgdorferi through skin injection. In healthy non-human primates, the infection is transient and rapidly cleared as the animals developed antibodies to fight off infection. However, in non-human primates with suppressed immune systems, infection persists and involves the central and peripheral nervous systems, as well as organs, such as the heart, bladder, skin, and skeletal muscle (Clin Diagnostic Lab Immunol 8: 225, 2001; Ann Neurol. 50: 330, 2001; Infect Immun 71: 7087, 2003).
  • The presence of the bacteria’s genetic material was particularly prevalent in the central nervous systems of infected non-human primates compared to other tissues (Proc Natl Acad Sci 100: 1953, 2003; Microbial Path 39: 27, 2005).
  • The pattern of infection is dependent upon the strain of Borrelia used to induce infection; B. garinii induces a different type of disease in mice and non-human primates than does B. burgdorferi (Ann Neurol 54: S49, 2004).
Content last reviewed on December 16, 2014