Subscribe to eTOC

New Study Reports Short-term Benefits But Not Sustained Improvement in Cognition With Prolonged Antibiotic Use for Lyme Disease Encephalopathy


  • ✓ Patients taking IV ceftriaxone for Lyme disease encephalopathy had significantly improved cognition than did controls and those on placebo at 12 weeks, but the effect was not sustained at week 24.

The latest placebo-controlled, double-masked trial to test the safety and efficacy of IV antibiotic treatment in patients with post-treatment Lyme encephalopathy found short-term but not sustainable benefit for cognitive function. The study, published online in advance of print on Oct. 10 in Neurology, adds to a growing body of evidence suggesting no benefit for prolonged antibiotic use — beyond four weeks — for Lyme disease. Despite the findings, some physicians say the jury is out on how to translate the findings into clinical practice.

Lead investigator Brian Fallon, MD, associate professor of clinical psychiatry at the Columbia University Medical Center and director of the Lyme and Tick-Borne Diseases Research Center at the Columbia University Medical Center, said: “Because IV ceftriaxone therapy is associated with considerable risks and because the improvement in cognition was not sustained, we concluded that 10 weeks of IV ceftriaxone therapy followed by 14 weeks of no antibiotic is not an effective strategy for sustained cognitive improvement.”

But patients with greater severity at the start of the study had moderate improvement in pain and physical functioning — a finding, Dr. Fallon said, that needs to be evaluated further in future studies.


The current study examined neurocognitive impairment in patients with persistent joint pain, fatigue, and memory loss after Lyme disease diagnosis. To be eligible for the study, participants between the ages of 18 and 65 years old had to have had a history of physician-documented erythema migrans or another objective manifestation of Lyme disease as defined by the Centers for Disease Control and Prevention; current positive IgG Western blot on serologic testing; at least three weeks of past IV ceftriaxone treatment for Lyme disease, completed at least four months before study entry; and objective memory impairment that started after the onset of Lyme disease. Dr. Fallon stressed that the investigators in the current study, unlike other studies, specifically recruited patients with cognitive impairments (memory complaints).

He noted that they excluded patients who had significant pre-Lyme psychiatric morbidity as that might have confounded assessments of cognitive impairment. But they did assess patients in the study for psychiatric morbidity, using the Beck Depression Inventory and the Zung Anxiety Scale, and there were no differences at baseline in their scores on these issues.

Twenty-three patients were assigned to 10 weeks of double-masked treatment with IV ceftriaxone (2 g/d) and 14 patients to IV placebo, and then 14 weeks off all antibiotics. Eighteen healthy individuals served as controls, receiving neither ceftriaxone nor placebo.

All were tested at baseline, and then at weeks 12 and 24. Five patients withdrew during the first 12 weeks: three on the drug and two on placebo. Three additional patients remained in the study despite adverse events that required early termination of the medication, but “each of these patients continued in a masked fashion through to the 12 and 24 week evaluations.” No patients withdrew between week 12 and week 24.

To determine if the two-time neuropsychologic testing led to improved performance due to practice, the group of healthy controls also took the same tests twice.


Among findings, there was a significant difference in cognitive improvement for the three groups (placebo, ceftriaxone, controls) over the six-month interval (p=0.04), favoring the antibiotic at week 12. The cognitive improvement between baseline and week 12 was better in the experimental group than in the controls (p<0.01) and the placebo-treated group (p=0.053). However, the effect was not sustained at week 24.

The study authors wrote that the inability of the group on ceftriaxone to sustain the acute-phase improvement in cognition during the antibiotic-free interval “resulted in a loss of the differential treatment effect among the three groups at week 24.”

“It should be noted that for the specific between-group comparisons at week 12, the drug-placebo difference fell just above the margin of significance at the p=0.053 level so that this finding of a drug-placebo difference should be considered to have a slightly increased risk for having occurred by chance and would require confirmation from a subsequent study,” Dr. Fallon said.

Dr. Fallon pointed out that the group receiving ceftriaxone who had the most severe symptoms of pain and impaired physical functioning at baseline had greater improvement at week 12, and the effects were sustained at week 24. Adverse events from the study medication or the PICC (peripherally inserted central catheter) IV line used for extended antibiotic therapy occurred in six of the 23 (26.1 percent) given ceftriaxone, and one of the 14 (7.1 percent) given placebo.

The study “does not tell us the mechanism by which IV ceftriaxone may have been beneficial, as ceftriaxone has both antimicrobial-specific properties and non-antimicrobial properties that may be beneficial,” Dr. Fallon said. “For example, it is possible that short-term improvement in cognition may have resulted from an up-regulation of glutamate transporters in the CNS induced by ceftriaxone and a consequent reduction of extracellular neurotoxicity.”


John J. Halperin, MD, a neurologist at the Atlantic Neuroscience Institute and Overlook Hospital in Summit, NJ, pointed out that Dr. Fallon's study, in abstract form, was included in the review performed by the AAN for its 2007 guidelines on Lyme disease treatment (Neurology 2007;69:91–102). Dr. Halperin, lead author of the Lyme treatment guidelines, noted that AAN guidelines normally do not quote unpublished studies, but made an exception for the Fallon study “because we knew it would be viewed as important.”

“The published paper added a lot of technical detail to the previously presented abstract,” he said, “but, if anything, made its very limited claims of success even weaker.”

Dr. Fallon's findings lacked statistical significance, Dr. Halperin noted. At 12 weeks, the improvement in patients who had received ceftriaxone appeared slightly greater than in those on placebo but did not quite reach statistical significance (p=0.053). He added that the p-value did not quite reach 0.05 until they incorporated a second control group: healthy individuals who neither had Lyme disease nor received any type of treatment.

When the treated patients were compared to this larger group of untreated ones, the statistics became a little more favorable: p=0.04 marginally significant. He noted that the apparent improvement in patients on the drug — as compared with the expanded control group — was likely attributable not to any substantial treatment effect but to the dilution of the improvement in the placebo-treated group affected by including these other controls. Importantly, when patients were reassessed at 24 weeks, even this limited apparent improvement was no longer evident.

“The only emphasized conclusions were post-hoc and not statistically significant,” Dr. Halperin said. Post-hoc analyses are exploratory rather than hypothesis-driven, requiring independent confirmation. They are problematic, he said, because the approach consists of positing multiple questions until an interesting observation appears. Statistical tests must then be adjusted for the actual number of comparisons performed. As an example, while an exploratory study predicted that the Fallon study would demonstrate memory improvement in response to treatment, a primary endpoint of this study, this did not turn out to be statistically significant. “The study clearly showed that 10 weeks of antibiotics was harmful and offered no sustained benefit, just as previous studies did,” he said.

Physicians should accept the study as “yet more compelling data that prolonged antibiotic treatment does not help these patients and exposes them to substantial risk,” Dr. Halperin said. “It may be worth emphasizing that every time patients receive unnecessary antibiotics, we are needlessly increasing the potential for MRSA [Methicillin-resistant Staphylococcus aureus] and other antibiotic-resistant bacteria.”

Daniel Cameron, MD, president-elect of the International Lyme and Associated Disease Society (ILADS) and chief author of the ILADS guidelines that recommend extended use of antibiotics for post-Lyme syndrome, disagreed. He said the Fallon trial “adds additional evidence to support prolonged therapy not only for cognitive impairment, but for the other severe symptoms of chronic Lyme disease, including severe pain, fatigue, and other physical dysfunction.” However, he added: “The trial sample size of 37 subjects is too small to draw definitive conclusions.”

Indeed, the study authors wrote that the primary limitations of the study are restrictive inclusion criteria, small sample size, and lack of post-treatment neurologic exam, or lumbar puncture. Dr. Fallon emphasized that the patients in his study made up a strictly defined sample, and that the results from the study may not be generalizable to post-treatment Lyme patients without cognitive impairment or to seronegative patients with persistent symptoms.

“Because of the small sample size in this study, the results of this study are best considered as a stimulus for further investigation into the heterogeneous nature of post-treatment Lyme symptoms,” he said.


Dr. Cameron noted that the issue of post-treatment Lyme disease is still up for debate. “Controversies arise when there is little evidence,” he said, and it is unlikely that the small size of the current randomized controlled trials coupled with concerns over the generalizability of the results will resolve the debate about post-Lyme treatment. Physicians should review the evidence of this and previous trials using evidence-based medicine, he added.

The finding of cognitive improvement at 12 weeks should be studied further in trials of patients with symptoms much earlier in their illness, Dr. Cameron continued. Inasmuch as Lyme disease affects an estimated 200,000 people in the U.S. every year, he said, the small sample size was disappointing, and larger trials are needed as well as expanded inclusion criteria.

Future trials “may need to include symptomatic treatment and rehabilitation along with antibiotics for Lyme disease patients with a poor quality of life, delayed treatment, and years of illness,” he added.

Dr. Halperin recommended a different mindset. “These patients have no evidence of a persisting infection. The symptom complex overlaps with many others. It is a real disorder that disrupts people's well-being and we neither understand its pathophysiology nor know how best to treat it. We need to go back to basics, stop exposing people to unnecessary physical and psychological risk, and determine how best to help them.”

When patients become convinced that they have a persistent infection that only the ‘Lyme-literate doctors’ understand, he said, they become angry with conventional medicine, lose confidence in their well-intentioned ‘regular doctors,’ and often have other medical problems that go undiagnosed. “It's a real problem!”


Lyme encephalopathy is one manifestation of neuroborreliosis, or nervous system Lyme disease, in which patients experience confusion and have difficulty with memory and concentration.

Lyme encephalopathy is described differently by different people, said John Halperin, MD, a neurologist at the Atlantic Neuroscience Institute and Overlook Hospital in Summit, NJ. Originally, it referred to altered status in cognitive function and memory in people with active Lyme disease, most of whom had systemic but not nervous system Lyme infection. Only rarely do these patients actually have infection in the substance of the brain itself, referred to as encephalitis.

“The diagnosis of Lyme encephalopathy should be limited, in my opinion, to people with definite Lyme disease; objectively demonstrable alteration in memory or cognition, not explained by disorders of mood or affect; and no evidence of actual brain infection,” said Dr. Halperin, lead author of the 2007 AAN Lyme disease treatment guidelines.


Neurology Today has covered a series of articles in the past few months on Lyme disease:

  • “Guidelines on Trial: AAN Subpoenaed as Part of Investigation into Treatment Parameters for Lyme Disease” (Oct. 16, page 1);
  • “New Criteria Are Proposed for Diagnosing Neuroborreliosis” (Oct. 16, page 13);
  • “In New Guidelines, AAN Panel Finds No Benefit in Long-Term Antibiotics for Lyme Symptoms” (June 19, page 1).


• Fallon BA, Keilp JG, Sackheim HA, et al. A randomized, placebo-controlled trial of repeated IV antibiotic therapy for Lyme encephalopathy. Neurology 2007; E-pub 2007 Oct. 10.
    • Halperin JJ. Editorial: Prolonged Lyme disease treatment. Neurology 2007;E-pub 2007 Oct. 10.