Fecal Microbiota and Volatile Metabolome Changes Predict Late-Onset Meningitis in Preterm Infants
Overview
This study demonstrates that alterations in fecal microbiota composition can accurately predict late-onset meningitis (LOM) in preterm infants 1–3 days before clinical diagnosis. Volatile metabolome changes showed moderate association with LOM, while no single discriminative metabolites were identified. These findings suggest potential for noninvasive early biomarkers to improve LOM diagnosis.
Background
Late-onset neonatal meningitis (LOM) is a severe infection in preterm infants with high mortality and long-term neurological sequelae. Diagnosis is challenging due to difficulties in cerebrospinal fluid (CSF) sampling and delayed culture results. Prior studies indicate that gut microbiota and metabolome alterations precede other neonatal infections like late-onset sepsis (LOS). Given the frequent co-occurrence of LOS and LOM, this study investigates whether fecal microbiota and volatile metabolome changes precede LOM, potentially serving as early diagnostic biomarkers.
Data Highlights
Parameter
Value
Number of infants included
1397
LOM cases
21 (1.5%)
LOM with concomitant LOS
19 (90%)
Random forest AUC for microbiota prediction (1–3 days pre-LOM)
0.88 (n=147)
GC-IMS pattern recognition AUC (3 days pre-LOM)
0.70–0.76 (P < .05, n=92)
GC-TOF-MS discriminative metabolites
None identified (n=66)
Key Findings
Fecal microbiota composition differs significantly in preterm infants prior to LOM diagnosis compared to controls.
A random forest model using six microbiota features predicted LOM with high accuracy (AUC 0.88) 1–3 days before diagnosis.
Volatile metabolome analysis by GC-IMS showed moderate discrimination of pre-LOM samples (AUC 0.70–0.76).
GC-TOF-MS failed to identify individual discriminative metabolites associated with LOM.
90% of LOM cases had concomitant late-onset sepsis, supporting gut origin of infection.
Microbiota and metabolome alterations precede clinical onset, suggesting potential for early noninvasive biomarkers.
Clinical Implications
Early detection of LOM in preterm infants may be improved by monitoring fecal microbiota composition, enabling timely intervention before clinical deterioration. Noninvasive fecal biomarker analysis could complement or reduce reliance on invasive CSF sampling, especially in unstable patients. Moderate volatile metabolome changes may provide additional diagnostic information but require further validation.
Conclusion
Alterations in fecal microbiota composition precede late-onset meningitis in preterm infants and can accurately predict disease onset days before clinical diagnosis. These findings highlight the potential of fecal microbiota profiling as a noninvasive early diagnostic tool for LOM.
References
Original Study 2024 -- Changes in Fecal Microbiota and Volatile Metabolome Patterns May Indicate Late-Onset Meningitis in Preterm Infants
by Nina M Frerichs, Nancy Deianova, Sofia el Manouni el Hassani, Animesh Acharjee, Mohammed Nabil Quraishi, Willem P de Boode, Veerle Cossey, Christian V Hulzebos, Anton H van Kaam, Boris W Kramer, Esther d’Haens, Wouter J de Jonge, Daniel C Vijlbrief, Mirjam M van Weissenbruch, Emma Daulton, Alfian N Wicaksono, James A Covington, Marc A Benninga, Nanne K H de Boer, Johannes B van Goudoever, Hendrik J Niemarkt, Tim G J de Meij
