Metal Concentrations in Blood and Cerebrospinal Fluid of Patients With Arthroplasty Implants.

IMPORTANCE

Arthroprosthetic metal exposure has been linked to neurological dysfunction, but evidence of central nervous system exposure is largely missing.

OBJECTIVE

To compare levels of all arthroplasty-relevant metals in cerebrospinal fluid (CSF), whole blood, and serum of patients with and without large joint replacement.

DESIGN, SETTING, AND

PARTICIPANTS

The NeuroWear pilot study was a single-site hospital-based cross-sectional study conducted between April 2018 and November 2019. Laboratory personnel and investigators were blinded to group allocation. Patients presenting for elective surgery under spinal anesthesia and patients scheduled for lumbar puncture at a university medical center were eligible for inclusion. Patients aged 18 years and older with at least 1 large joint replacement in situ (103 case participants) and age- and sex-matched arthroplasty-naive patients (108 control participants) were recruited. One case and 6 control participants were excluded. Each case participant was matched with 1 control participant to achieve equal self-reported sex and a minimal age difference. The data analyses were performed between May 2023 and February 2024.

EXPOSURE: Presence of large joint replacement.

MAIN OUTCOMES AND MEASURES: CSF, whole blood, and serum levels of aluminum, cobalt, chromium, molybdenum, nickel, niobium, tantalum, titanium, vanadium, and zirconium were quantified.

RESULTS

A total of 204 patients (118 [58%] women and 86 [42%] men; median [range] age 69.4 [21.3-93.1] years) were included for metal analyses. In CSF, median (range) cobalt levels were significantly higher in the implant group (0.03 [0.01-0.64] μg/L) compared with the control group (0.02 [0.01-0.19] μg/L). Cobalt levels in CSF were significantly correlated with cobalt levels in serum (r = 0.72; 95% CI, 0.53-0.85) and whole blood (r = 0.82; 95% CI, 0.62-0.92). Significantly higher whole blood median (range) levels of cobalt (implant: 0.27 [0.07-24.10] μg/L; control: 0.16 [0.08-0.99] μg/L), chromium (implant: 0.47 [0.24-4.76] μg/L; control: 0.42 [0.21-1.52] μg/L), titanium (implant: 8.05 [1.14-37.20] μg/L; control: 7.15 [1.80-20.70] μg/L), niobium (implant: 0.02 [0.01-1.14] μg/L; control: 0.01 [0.01-0.11] μg/L), and zirconium (implant: 0.05 [0.01-39.90] μg/L; control: 0.03 [0.01-1.95] μg/L) were detected. Patients with an implant were found to have significantly higher median (range) CSF metal levels for titanium (implant: 0.75 [0.12-1.40] μg/L; control: 0.57 [0.13-1.10] μg/L), niobium (implant: 0.02 [0.01-0.16] μg/L; control: 0.01 [0.01-0.03] μg/L), and zirconium (implant: 0.05 [0.01-0.44] μg/L; control: 0.04 [0.01-0.28] μg/L) if those metal levels were elevated in serum. Patients with cobalt-chromium-molybdenum implant components exhibited significantly higher median (range) chromium levels in CSF than control participants (implant: 0.31 [0.02-2.05] μg/L; control: 0.23 [0.02-1.10] μg/L).

CONCLUSIONS AND RELEVANCE: In this cross-sectional study, chronic exposure to arthroplasty implants was associated with corresponding metal accumulation in blood and CSF. Arthroprosthetic cobalt, chromium, titanium, niobium, and zirconium seemed to cross neural barriers and accumulate in CSF. Correlation analyses suggested cobalt-specific transport mechanisms across neural barriers. In view of the neurotoxic effects of cobalt, subsequent studies are needed to determine whether CSF metal concentrations correlate with objective measures of neurotoxic effects and whether this may be of relevance, particularly in patients with new-onset or rapid deterioration of neurological conditions following arthroplasty.