Lyme Neuroborreliosis: Known Knowns, Known Unknowns

John J. Halperin; Randi Eikeland; John A. Branda; Rick Dersch


Brain. 2022;145(8):2635-2647. 

In This Article

CNS Disorders Without CNS Inflammation

Lyme Encephalopathy

There is likely nothing related to Lyme borreliosis that has provoked more controversy than 'Lyme encephalopathy'. First described in patients with active infection, this was originally conceptualized as the same cognitive and memory difficulty seen in patients with other systemic, non-neurological infectious and inflammatory disorders.[88] This is rarely associated with CNS Bbsl infection or inflammation—by MRI or CSF—including the absence of identifiable CSF cytokine abnormalities.[45] Recent observations suggest an association with peripheral activation of Th17-related cytokines,[45] a pattern also seen in antibiotic-resistant, presumably immune-mediated Lyme arthritis[89] and suggested in neuromyelitis optica[90] and multiple sclerosis.[91] Whether this association is in any way specific, related to any other post-treatment symptoms, or more broadly informative remains to be seen.

Unfortunately, this symptom complex has become the major construct underlying 'post-treatment Lyme disease symptoms' (PTLDS), which similarly has been broadly misconstrued as evidence of a chronic brain infection—despite overwhelming evidence to the contrary. This has created a level of fear about 'nervous system Lyme disease' that underlies much of the public concern about Lyme borreliosis.

When can Atypical Manifestations be Attributed to Lyme Neuroborreliosis?

Establishing causal relationships between other neurological disorders and LNB requires both compelling systematic data and biological plausibility. An important (Bayesian) limitation of serodiagnostic testing is that positive predictive value depends on its only being used in appropriate settings[9] (Table 3)—i.e. in individuals without plausible exposure false positives may well exceed true positives. Moreover, high background seropositivity (13+%) in some populations[85,86] makes coincidental seropositivity an important confounder. Importantly, given LNB's pathophysiology, association of CNS disorders without CNS inflammation is inherently unlikely.

Neurodegenerative Diseases. A population-based study[92] using the administrative database incorporating all Danish medical records from 1995 to 2015 compared the 2067 patients with Bbsl-specific ITAb to 20 687 age and sex-matched controls, finding no increased incidence of dementia, Alzheimer's disease, Parkinson's disease, motor neuron disease, epilepsy or Guillain-Barre syndrome among those with ITAb. A prospective 6-year study of 689 French farmers, aged 65 and over,[93] found no higher incidence of cognitive or functional decline among seropositive individuals. A study comparing 491 Dutch patients with amyotrophic lateral sclerosis (ALS) to 982 matched controls found no association between ALS and anti-Bbsl antibodies.[94] Of 414 ALS patients seen at Massachusetts General Hospital, four had positive two-tier testing. Ceftriaxone treatment in two had no effect.[95] Importantly, there is no geographic correlation between ALS and Lyme disease in the USA.[96]

This notwithstanding, rare patients with neurodegenerative disorders have been hypothesized to have the disorder as a result of Lyme borreliosis, been treated with antibiotics, and appeared to improve clinically. Of 1594 patients with dementia screened serologically for Lyme borreliosis, 38 were found to be seropositive. CSF was examined in 34/38. CSF ITAb was elevated in 20—notably in none was there a CSF pleocytosis, suggesting the ITAb reflected past, not currently active infection. Following treatment,[97] seven (0.4%) 'stabilized or improved slightly' and were felt to have Lyme dementia; the possibilities of this being an infection temporarily worsening an underlying dementia, or more likely, that this reflected prior, unrelated infection, were not discussed. Balancing these anecdotal observations against the strong epidemiological data, a relationship appears highly unlikely.

Psychiatric Disease. A study using the above-described records of all 2067 Danes with ITAb, combined with the national health database[98] found psychiatric diagnoses and hospitalizations to be no more frequent in patients with definite LNB. A second study of the same population dataset[99] (not incorporating ITAb data) concluded psychiatric disorders were more frequent in Lyme borreliosis (not LNB) generally [incidence rate ratio (IRR) 1.28, 95% confidence interval (CI) 1.20, 1.37]—but failed to point out that earlier work with the same database[100,101] found an even greater increase in psychiatric diagnoses with 'any' infection (IRR 1.62; 95% CI, 1.60–1.64) or auto-immune disorder (IRR 1.45; 95%CI, 1.39–1.52). Notably, when the Lyme borreliosis analysis was limited to the 12 616 patients administratively coded as LNB[99]—explicitly not based on ITAb—there was no increase in psychiatric diagnoses. Consistent with this, among 517 patients undergoing acute psychiatric hospitalization in an endemic area just one had a positive Lyme ELISA—but negative western blot.[102]

An often-cited study used multiple assays to compare 926 newly hospitalized psychiatric patients to 884 healthy control subjects.[103] Psychiatric patients were more likely to be positive in at least one of four assays, but two-tier testing was more frequently positive in healthy controls. Abnormal results in psychiatric patients were due entirely to unconventional tests. A second study of the same population[104] found no specific psychiatric diagnosis disproportionately represented among either seropositive or seronegative patients.

MRI Abnormalities. LNB may occasionally cause focal areas of brain inflammation, visualizable on MRI.[105,106] However, it has become commonplace to suggest an association between non-specific white matter (NSWM) MRI abnormalities and this infection. Two controlled studies, with <100 LNB patients,[107,108] found no difference in NSWM abnormalities compared to control subjects. A third, population-based study found MRI useful primarily to exclude other diagnoses.[109] No larger studies have been published but this association with a non-specific finding lacks biological plausibility. Recent US guidance recommends against including Lyme borreliosis in the differential diagnosis of such MRI findings.[9]