B-313 Aberrant Lithium Results After HDL Assay on Roche Cobas c503 Instrument

Abstract

Abstract Background Lithium is prescribed clinically to manage mania in patients with bipolar disorder. However, blood levels higher than 1.5 mmol/L (12 h post administration) indicate toxicity accompanied by severe nausea and vomiting, severe hand tremors, confusion, and vision changes. If neglected, lithium toxicity can lead to coma, brain damage, or even death. Diligent monitoring of blood lithium levels periodically is critical to preventing lithium toxicity. Case Presentation: A 37 y/o female with schizophrenia and bipolar disorder on lithium since 2019 visited the outpatient clinic for her routing biannual lithium check. The patient’s plasma lithium levels were detected at 3.9 mmol/L (previously 0.4 mmol/L; Therapeutic range: 0.60–1.20 mmol/L; Toxic: >2.00 mmol/L). The cause for elevated lithium level was unclear; EKG and vital sign were unremarkable and patient was asymptomatic for toxicity. Nevertheless, patient was then managed with IV infusions and admitted to emergency room for follow up. A redrawn lithium sample ∼22 h later was 0.3 mmol/L, suggesting no toxicity. We then investigated the discrepancy in lithium levels reported by the laboratory. Retesting of the original sample detected lithium at 0.5 mmol/L, a value significantly less than the previously reported 3.9 mmol/L. Our laboratory had recently introduced the Roche Cobas® pro integrated solutions into the lab and chemistry testing (including Lithium) are performed on the Roche Cobas c503 analyzers (an arm of the Roche pro integrated solutions). Our goal was to investigate the observed phenomenon and find potential solutions, especially as it had significant implications to patient care. Method We reviewed all high lithium levels recently tested our new instrument within a week of this incident. Seven samples (lithium values >1.2 mmol/L) were identified and lithium measurements repeated. We also investigated and ruled out the possibility of carry over contamination by rerunning a low lithium sample before and after a high lithium sample. Results Retesting all seven samples resulted in lower lithium level with an average % change of 365%. Further investigations and discussions with manufacturers alluded to interference after HDL testing on the same instrument. To investigate this hypothesis, we reran two previously reported low samples (0.3 and 0.6 mmol/L) immediately after running the HDL cholesterol assay on the Roche Cobas c503 analyzer. Retesting resulted in lithium levels of 7.32 and 1.36 mmol/L (2313.3% and 126.7% change) respectively. Conclusion Our investigation suggests that running the HDL cholesterol assay prior to the lithium assay on the Roche Cobas c503 analyzer results in positive interference. As oppose to other Roche chemistry analyzers (such as the Roche c501), the Roche Cobas c503 analyzer has no wash step between the HDL cholesterol and the lithium assays. Detergents in buffers and reactive byproducts (like peroxide) of the HDL cholesterol assay, may leave behind salts in the instrument probe that can interfere with the lithium assay. We are currently working on adding a wash step between the HDL cholesterol and lithium assays. In the interim, we are repeating lithium testing on high lithium samples.