**Background:** Bacterial meningitis (BM) triggers a host inflammatory response that includes oxidative/nitrosative stress, which is thought to contribute to neuronal damage. Protein oxidation biomarkers, such as ortho-tyrosine (o-Tyr) formed from phenylalanine (Phe) by hydroxyl radicals, can be measured in cerebrospinal fluid (CSF). While oxidative stress is generally considered harmful, its relationship with clinical outcomes in pediatric BM is not fully understood. This study aimed to investigate the association between CSF protein oxidation biomarker levels at admission and the course and outcomes of BM in children.
**Methods:** This was a sequential pediatric study using data and CSF samples from a prospective single-center study conducted at the Pediatric Hospital of Luanda, Angola (2005–2008). CSF samples from 79 children with confirmed BM were analyzed. Biomarkers o-Tyr, 3-chlorotyrosine (3Cl-Tyr), and 3-nitrotyrosine (3NO₂-Tyr) were measured by liquid chromatography-tandem mass spectrometry (LC-MS/MS) and normalized to their precursors (Phe or para-tyrosine, p-Tyr). Matrix metalloproteinase (MMP)-8, MMP-9, and tissue inhibitor of metalloproteinase (TIMP)-1 were also measured. Patient characteristics, course of illness (seizures, secondary fever), and outcomes (neurological sequelae, ataxia, hearing impairment, Glasgow Outcome Scale [GOS]) were recorded at Day 7 and one month after discharge. Statistical analyses included Spearman's rank correlation, Mann-Whitney U-test, and odds ratios (OR) with 95% confidence intervals (CI) for biomarker levels above vs. below the median.
**Key Results:** The median age was 12 months (IQR 7–42), and 35% (26/79) died by Day 7. The median o-Tyr/Phe ratio was 0.002 (IQR 0.001–0.013). A lower o-Tyr/Phe ratio was significantly associated with male sex (p=0.020), longer preadmission symptoms (>3 days, p=0.047), and seizures prior to admission (p=0.013) or at admission (p=0.001). A higher o-Tyr/Phe ratio correlated with higher CSF leukocyte count (rho 0.384, p=0.00047) and higher admission temperature (rho 0.229, p=0.004). A lower o-Tyr/Phe ratio correlated with higher CSF TIMP-1 (rho -0.503, p<0.0001) and higher 3NO₂-Tyr/p-Tyr (rho -0.351, p=0.002).
Children with an o-Tyr/Phe ratio below the median (0.0015) had significantly increased odds of: seizures on admission (OR 4.35; 95% CI 1.67–11.33), below-median CSF leukocyte count (OR 3.71; 95% CI 1.47–9.42), above-median TIMP-1 (OR 8.70; 95% CI 3.16–23.99), above-median 3NO₂-Tyr/p-Tyr (OR 5.25; 95% CI 2.01–13.70), secondary fever after Day 7 (OR 3.34; 95% CI 1.08–10.39), and focal seizures during illness (OR 2.86; 95% CI 1.14–7.16).
Regarding outcomes, a lower o-Tyr/Phe ratio was associated with worse outcomes: any neurological sequelae on Day 7 (OR 8.55; 95% CI 2.27–32.22) and one month after discharge (OR 8.00; 95% CI 1.51–42.45), ataxia on Day 7 (OR 8.55; 95% CI 2.27–32.22) and one month after discharge (OR 5.83; 95% CI 1.12–30.40), and GOS below 5 on Day 7 (OR 2.85; 95% CI 1.14–7.14) and one month after discharge (OR 5.23; 95% CI 1.66–16.52). A higher o-Tyr/Phe ratio correlated with a higher GOS at one month (rho 0.271, p=0.036). The 3NO₂-Tyr/p-Tyr ratio showed associations with slower recovery (longer duration with GCS <15, rho 0.224, p=0.047) and more severe disease (lower GCS, OR 4.90; 95% CI 1.44–16.66), but no direct association with outcome. The 3Cl-Tyr/p-Tyr ratio showed no significant associations with outcomes.
**Clinical Implications:** This study provides novel evidence that a higher CSF o-Tyr/Phe ratio at admission, indicating greater protein oxidation, is associated with a better clinical course and fewer neurological sequelae in pediatric BM. This challenges the conventional view that oxidative stress is uniformly detrimental and suggests that the host's oxidative response may have a protective role in certain contexts. The findings highlight the complexity of the inflammatory response in BM and suggest that therapeutic strategies aimed at broadly suppressing oxidative stress may be counterproductive. Instead, a more nuanced approach targeting specific pathways may be needed. The study also underscores the importance of early diagnosis and treatment, as the median duration of illness before admission was 4 days, and many children presented with advanced disease. Limitations include the small sample size (79 of 723 original patients), loss to follow-up, and the single-center design in a resource-limited setting. Further research with larger cohorts and serial biomarker measurements is warranted to confirm these findings and explore underlying mechanisms.