Optimizing metagenomic next-generation sequencing in CNS infections: a diagnostic model based on CSF parameters

Area of Science:

• Neuroscience
• Infectious Diseases
• Molecular Diagnostics

Background:

• Central nervous system (CNS) infections pose diagnostic challenges, often requiring advanced molecular techniques like metagenomic next-generation sequencing (mNGS).
• Routine cerebrospinal fluid (CSF) biochemical parameters are readily available but their association with mNGS results in CNS infections is not fully elucidated.
• Optimizing the use of mNGS, a sensitive but costly diagnostic tool, is crucial, particularly in resource-constrained settings.

Purpose of the Study:

• To evaluate the correlation between standard CSF biochemical markers and mNGS findings in patients with suspected CNS infections.
• To develop and validate a predictive model using CSF parameters to guide mNGS testing strategies.
• To enhance the diagnostic yield and cost-effectiveness of CNS infection diagnostics.

Main Methods:

• Retrospective analysis of 110 patients with suspected CNS infections, comparing CSF analysis and mNGS results.
• Logistic regression was employed to identify independent predictors of mNGS positivity, focusing on CSF cell count and protein concentration.
• A predictive nomogram was developed and validated internally (10-fold cross-validation, bootstrap) and externally (40 patients).

Main Results:

• mNGS detected infections in 56.36% of cases, significantly higher than CSF culture (6.36%).
• Patients with mNGS-positive results exhibited significantly higher CSF cell counts, protein levels, and turbidity, along with lower glucose levels compared to mNGS-negative cases.
• The predictive model incorporating CSF cell count and protein demonstrated good performance, with an Area Under the Curve (AUC) of 0.782 in the derivation cohort and 0.763 in the external validation cohort.

Conclusions:

• CSF cell count and protein concentration are robust predictors of mNGS positivity in suspected CNS infections.
• The developed predictive model shows consistent diagnostic performance and clinical utility.
• This model can assist clinicians in making informed decisions regarding mNGS testing, thereby optimizing diagnostic workflows, especially in resource-limited healthcare settings.