We conducted a retrospective chart review of adult patients with diagnosed Anaplasma phagocytophilum infection (International Classification of Diseases, 9th Revision, codes A77.49 or A79.82) who were admitted to Mayo Clinic Rochester or Mayo Clinic Health System sites in Minnesota and Wisconsin during November 1, 2014–November 29, 2024. The study was approved by the Mayo Clinic Institutional Review Board (approval no. 18-007901).

We manually reviewed a total of 101 patient charts. We included patients >18 years of age if they had a positive A. phagocytophilum PCR performed on EDTA-anticoagulated whole blood and predominantly had CNS symptoms when care was sought that prompted LP and CSF analysis.

We performed molecular detection by using the automated MagNA Pure 96 system (Roche Diagnostics, https://www.roche.com) for DNA extraction, followed by amplification of a conserved region of the groEL heat shock protein operon gene. We identified organisms via melting curve analysis by using the LightCycler 480 Instrument II (Roche Diagnostics). That assay uses fluorescence resonance energy transfer probes to detect and differentiate A. phagocytophilum, Ehrlichia chaffeensis, E. muris eauclairensis, and E. ewingii/canis (1).

All patients underwent standard CSF testing including a meningitis/encephalitis panel per institutional protocol. We defined CNS symptoms as severe headache, confusion or altered mental status, central focal neurologic deficits, or signs of meningitis (including neck stiffness, photophobia, or positive meningeal signs).

Of 101 hospitalized patients with anaplasmosis diagnoses, 10 (10%) met inclusion criteria. Among those, 2 patients were immunocompromised and 4 required intensive care unit (ICU) admission. Six patients had clinical features consistent with meningitis, 3 with meningoencephalitis, and 1 with encephalitis (Table 1). All 10 patients underwent extensive infectious evaluation for alternative bacterial, viral, endemic fungal, and tickborne etiologies (Appendix Table), including testing for other Ixodes scapularis tick–transmitted pathogens. All of those results were negative, thereby excluding co-infection.

In all cases, LP was performed before initiation of doxycycline therapy. Five (50%) patient samples demonstrated pleocytosis, and the other 5 had CSF cell counts within reference ranges. Of the 5 patients with pleocytosis, 4 had mild lymphocytic pleocytosis (5–50 cells/μL [reference range 0–5 cells/μL]) and 1 had neutrophilic pleocytosis attributable to traumatic LP. CSF protein and glucose levels were within reference ranges in all patients except for patient 8, who had a traumatic LP.

Among the 10 patients described in this study, 6 had clinical features consistent with meningitis, 3 with meningoencephalitis, and 1 with encephalitis. Notable neurologic findings included akathisia and left hemiparesis in 2 patients who also exhibited meningitic symptoms. Two patients were immunocompromised, and 4 required ICU admission. All patients demonstrated clinical improvement within 24–72 hours of initiating doxycycline therapy and achieved full recovery. Total duration of doxycycline therapy ranged from 10 to 28 days; most patients received 10–14 days of doxycycline therapy.

CSF profiles were variable and differed substantially from those typically observed in bacterial meningitis or meningoencephalitis (Table 1). One patient with severe meningitic symptoms requiring ICU-level care had CSF findings within reference ranges. Another patient experienced a traumatic LP resulting in neutrophilic pleocytosis and low protein levels. Because traumatic LP more commonly leads to elevated CSF protein concentrations, the low protein observed in this case was unexpected and might reflect laboratory measurement variability without clear clinical significance. In all remaining patients, mild lymphocytic pleocytosis was observed with protein and glucose levels within reference ranges. Total CSF cell counts generally were 0–50 cells/µL.

Those findings suggest that CNS symptom severity does not correlate with the degree of CSF pleocytosis or inflammation. Neurologic manifestations of anaplasmosis therefore might be mediated by systemic inflammatory responses rather than direct CNS invasion. Although human data are lacking, a canine study of meningoencephalomyelitis with concurrent anaplasmosis failed to detect pathogen nucleic acid in brain tissue or CSF, supporting the hypothesis that direct CNS invasion is unlikely (17). That interpretation is further supported by the limited penetration of doxycycline across the blood–brain barrier, suggesting that clinical improvement likely results from attenuation of systemic and neuroinflammatory processes rather than direct antimicrobial activity within the CNS. Doxycycline penetrates the blood–brain barrier at ≈15%–30% of serum levels. However, CNS penetration increases when inflammation is present. Penetration might be higher in cases of Anaplasma meningitis (18), but the rapid symptomatic improvement noted after doxycycline administration suggests that the primary benefit of the drug likely stems from attenuation of systemic processes rather than direct CNS activity.

The hyperinflammatory potential of anaplasmosis is further evidenced by its association with secondary hemophagocytic lymphohistiocytosis and markedly elevated serum ferritin levels during acute infection (19). In our study, 2 patients with sepsis who required ICU admission and in whom anaplasmosis was suspected at admission received an initial 200-mg intravenous dose of doxycycline to increase CNS concentrations because of its limited penetration across the blood–brain barrier. However, that practice is not recommended by current guidelines (20), and whether it provides clinical benefit remains unclear. Patient 1 initially had stroke-like symptoms, including hemiparesis and urinary retention, in addition to symptoms of meningitis. Although neuroimaging in this patient was unrevealing and symptoms resolved with doxycycline therapy, anaplasmosis-associated stroke has been reported in other cases and is thought to result from endothelial injury (8,15,21,22).

As observed previously, our findings suggest that CNS manifestations of anaplasmosis are common but seldom reported (7). Although encephalitis has been described as a relatively frequent CNS manifestation (23), many reported cases lacked CSF evaluation, raising the possibility of encephalopathy rather than true encephalitis. As of April 2025, only 1 prior case from our institution had documented encephalitis supported by CSF findings with rapid response to doxycycline therapy (13). Further studies are needed to elucidate the pathophysiologic mechanisms underlying CNS involvement in anaplasmosis.

Diagnosis of anaplasmosis relies on clinical evaluation within an appropriate epidemiologic context. Most patients exhibit fever, pancytopenia (particularly thrombocytopenia), elevated transaminases, and gastrointestinal symptoms (5). In endemic regions, recognition of that clinical constellation often prompts empiric doxycycline therapy, resulting in rapid symptom resolution. In all our cohort patients, CNS symptoms improved within 24–72 hours of initiating doxycycline. We hypothesize that many patients improve before LP can be performed, which might explain the relatively low frequency of CSF evaluation in anaplasmosis.

Although bacterial culture remains the most sensitive diagnostic modality for anaplasmosis (24), routine clinical use is impractical because specialized cell culture systems are needed and the incubation period can be up to 4 weeks (24). Consequently, blood-based molecular diagnostic tests commonly are used (25). Serologic testing is frequently negative early in the course of disease and cannot reliably exclude acute infection (24,25). We identified only 1 reported case in which Anaplasma spp. nucleic acid was detected in CSF (26).

Approximately 30% of patients with anaplasmosis require hospitalization with encephalopathy identified as a risk factor (1). In our cohort, 4 patients with CNS symptoms required ICU admission; however, all were clinically stable for transfer to a general medical ward within 48 hours, and only 1 required vasopressor support. ICU admission remains uncommon in anaplasmosis, as demonstrated in a multicenter study in France (27). The higher ICU admission rate in our cohort might reflect recognition that anaplasmosis in Europe is generally a milder disease than in the United States (28). Genetic differences between Anaplasma spp. from North America and Europe might partially explain those clinical variations (28). Differences in strain pathogenicity, reporting practices, and underdiagnosis also might contribute to the observed disparity (28).

Our literature search identified 9 reported cases of anaplasmosis with CNS manifestations in which patients underwent LP and CSF examination (Table 2). Four of those patients had clinical signs and symptoms of encephalitis, 2 of meningoencephalitis, and 1 of meningitis. The other 2 patients who underwent LP had headache and CSF findings within reference ranges; they ultimately had trigeminal neuralgia and stroke diagnosed in addition to anaplasmosis. One patient with encephalitis also had concomitant ischemic stroke. Other rare neurologic manifestations, including opsoclonus-myoclonus-ataxia syndrome and hypometric saccades, have been described (12). Most (7/9) of the patients had CSF protein and glucose levels within reference ranges, and all 9 reported case-patients demonstrated lymphocytic pleocytosis, except for 1 patient with meningoencephalitis who had an equal percentage of neutrophils and lymphocytes in the CSF (Table 2). Although not routinely performed, metagenomic next-generation sequencing can detect Anaplasma spp. bacteria in CSF in select cases of meningitis or encephalitis (26,29).

In conclusion, patients with anaplasmosis who develop meningitis, encephalitis, or meningoencephalitis typically demonstrate minimal CSF abnormalities, characterized by absent or mild pleocytosis with protein and glucose levels within reference ranges. LP and CSF analysis might be of limited use for managing anaplasmosis-associated CNS symptoms and might be more valuable for excluding alternative infectious etiologies that can also cause neurologic symptoms after a tick bite, particularly neuroborreliosis, ehrlichiosis, and Powassan virus infection.