Diagnostic accuracy of cerebrospinal fluid cytochemical analysis to optimize the use of the FilmArray Meningitis/Encephalitis panel in immunocompetent and immunocompromised patients

Technical Summary: Diagnostic Accuracy of CSF Cytochemical Analysis to Optimize FilmArray Meningitis/Encephalitis Panel Use

Problem Statement
Infectious meningoencephalitis is a medical emergency requiring rapid etiological diagnosis to guide antimicrobial therapy and reduce mortality. While the FilmArray Meningitis/Encephalitis (FAME) multiplex PCR panel has improved pathogen detection speed and breadth, its indiscriminate application—particularly in cases with low pretest probability such as non-infectious headaches or subacute presentations—raises concerns regarding cost-effectiveness and diagnostic yield. Current literature suggests that molecular testing may be unnecessary when cerebrospinal fluid (CSF) lacks inflammatory markers, yet specific, validated thresholds for restricting FAME testing in both immunocompetent and immunocompromised populations remain underexplored. This study addresses the need to define CSF cytochemical criteria that can reliably predict FAME positivity, thereby optimizing test indication and reducing unnecessary molecular testing.

Methodology
The authors conducted a retrospective observational diagnostic accuracy study at two tertiary-care hospitals in Medellín, Colombia, between January 2019 and August 2023. The study cohort consisted of 478 adult patients (≥18 years) with suspected acute neuroinfection who underwent both CSF cytochemical analysis and FAME testing on the same sample.

• Index Test: CSF cytochemical parameters, including absolute and differential leukocyte counts, protein concentration, glucose levels, and opening pressure.
• Reference Standard: The FAME panel (BioFire Diagnostics), which detects 14 pathogens (7 viruses, 6 bacteria, 1 yeast). A result was considered positive if any of the 14 targets were detected.
• Stratification: Patients were categorized as immunocompetent (n=353) or immunocompromised (n=125), with the latter defined by HIV/AIDS, transplantation, chemotherapy, or high-dose steroid use.
• Analysis: The study evaluated the discriminative capacity of CSF parameters using Receiver Operating Characteristic (ROC) curve analysis. It specifically assessed the diagnostic yield of FAME when restricted to samples meeting specific cutoffs for leukocytes (<5 and <10 cells/mm³) and protein (<50 mg/dL), comparing these against the overall positivity rate.

Key Results

• Overall Positivity: The overall FAME positivity rate was 14.8% (71/478). Positivity was higher in immunocompromised patients (18.4%) compared to immunocompetent patients (13.5%). Bacteria were the most common pathogens (53.5%), followed by viruses (35.2%) and yeasts (11.2%).
• Cytochemical Differences: Patients with positive FAME results exhibited significantly higher median CSF leukocyte counts (189 vs. 3 cells/mm³), higher protein levels (167 vs. 47 mg/dL), and lower glucose levels (31 vs. 62 mg/dL) compared to FAME-negative patients.
• Diagnostic Performance: ROC analysis identified CSF protein concentration as the parameter with the highest diagnostic performance (AUC 0.793), followed by neutrophil count (AUC 0.779) and total leukocytes (AUC 0.777). Opening pressure showed the lowest discriminative value (AUC 0.666).
• Impact of Restrictive Criteria:
  • Leukocytes Only: 60% of tests were performed on samples with <10 leukocytes/mm³, yielding a low positivity rate of 5.5%.
  • Combined Criteria: When samples were restricted to those with <10 leukocytes/mm³ AND <50 mg/dL protein, 40.1% of all tests fell into this "low-risk" category. The positivity rate within this group was only 2.6% (5/192).
  • Clinical Significance: Of the 5 positive results in the low-risk group (<10 leukocytes and <50 protein), only one (0.5% of the total low-risk group) was deemed clinically significant (prompting a change in management). This case involved an immunocompetent patient with HHV-6. The other four positive results (including a yeast and bacterial case in immunocompromised patients) were considered to have limited clinical utility as they did not alter management or were confirmed by other methods.
• Optimization Potential: Applying a decision rule to restrict FAME testing to samples with >10 leukocytes/mm³ OR >50 mg/dL protein would have increased the overall diagnostic yield from 14.8% to 23% (a 55.4% relative increase) while eliminating 40.1% of unnecessary tests.

Significance and Claims
The study concludes that there is a strong association between abnormal CSF cytochemical parameters and FAME positivity, with protein concentration demonstrating the best individual diagnostic performance. The authors propose that implementing a combined clinical decision rule—restricting FAME testing to samples with >10 leukocytes/mm³ or >50 mg/dL protein—could significantly optimize resource utilization.

Specifically, the paper claims that this strategy would:

1. Increase Diagnostic Yield: By focusing testing on samples with inflammatory markers, the proportion of positive results increases by 55.4%.
2. Reduce Unnecessary Testing: Approximately 40% of molecular tests could be avoided without missing clinically significant diagnoses.
3. Extend Applicability: While previous criteria (such as the Reller criteria) were often applied primarily to immunocompetent patients, this study suggests that the combined cutoff of <10 leukocytes and <50 mg/dL protein may also be safely applied to immunocompromised patients, as the few positive results in this subgroup with normal CSF parameters were largely of limited clinical utility.

The authors acknowledge limitations, including the retrospective design, the lack of individual PCR confirmation, and the subjective assessment of "clinical significance." They emphasize that these findings should be validated through prospective studies incorporating clinical components before widespread adoption.