BACKGROUND The timely diagnosis of bacterial meningitis is of utmost importance due to the need to institute antibiotic treatment as early as possible. Moreover, the differentiation from other causes of meningitis/encephalitis is critical because of differences in management such as the need for antiviral or immunosuppressive treatments. Considering our previously reported association between free membrane phospholipids in cerebrospinal fluid (CSF) and CNS involvement in neuroinfections we evaluated phosphatidylcholine PC ae C44:6, an integral constituent of cell membranes, as diagnostic biomarker for bacterial meningitis. METHODS We used tandem mass spectrometry to measure concentrations of PC ae C44:6 in cell-free CSF samples (n = 221) from patients with acute bacterial meningitis, neuroborreliosis, viral meningitis/encephalitis (herpes simplex virus, varicella zoster virus, enteroviruses), autoimmune neuroinflammation (anti-NMDA-receptor autoimmune encephalitis, multiple sclerosis), facial nerve and segmental herpes zoster (shingles), and noninflammatory CNS disorders (Bell's palsy, Tourette syndrome, normal pressure hydrocephalus). RESULTS PC ae C44:6 concentrations were significantly higher in bacterial meningitis than in all other diagnostic groups, and were higher in patients with a classic bacterial meningitis pathogen (e.g. Streptococcus pneumoniae, Neisseria meningitidis, Staphylococcus aureus) than in those with less virulent or opportunistic pathogens as causative agents (P = 0.026). PC ae C44:6 concentrations were only moderately associated with CSF cell count (Spearman's ρ = 0.45; P = 0.009), indicating that they do not merely reflect neuroinflammation. In receiver operating characteristic curve analysis, PC ae C44:6 equaled CSF cell count in the ability to distinguish bacterial meningitis from viral meningitis/encephalitis and autoimmune CNS disorders (AUC 0.93 both), but had higher sensitivity (91% vs. 41%) and negative predictive value (98% vs. 89%). A diagnostic algorithm comprising cell count, lactate and PC ae C44:6 had a sensitivity of 97% (specificity 87%) and negative predictive value of 99% (positive predictive value 61%) and correctly diagnosed three of four bacterial meningitis samples that were misclassified by cell count and lactate due to low values not suggestive of bacterial meningitis. CONCLUSIONS Increased CSF PC ae C44:6 concentrations in bacterial meningitis likely reflect ongoing CNS cell membrane stress or damage and have potential as additional, sensitive biomarker to diagnose bacterial meningitis in patients with less pronounced neuroinflammation.

中文翻译：

---

脑脊液中的磷脂酰胆碱PC ae C44：6是细菌性脑膜炎的敏感生物标志物。

背景技术细菌性脑膜炎的及时诊断是最重要的，因为需要尽早进行抗生素治疗。此外，由于管理上的差异（例如需要抗病毒或免疫抑制治疗），与其他原因引起的脑膜炎/脑炎的区别至关重要。考虑到我们先前报道的脑脊髓液（CSF）中游离膜磷脂与中枢神经系统参与神经感染之间的关联，我们评估了磷脂酰胆碱PC ae C44：6（细胞膜的组成部分）作为细菌性脑膜炎的诊断生物标志物。方法我们采用串联质谱法测量了急性细菌性脑膜炎，神经性硼脑病，病毒性脑膜炎/脑炎（单纯疱疹病毒，水痘带状疱疹病毒，肠病毒），自身免疫性神经炎症（抗NMDA受体自身免疫性脑炎，多发性硬化症），面神经和节段性带状疱疹（带状疱疹）以及非炎性CNS疾病（贝尔氏麻痹，图雷特）综合征，正常压力脑积水）。结果在细菌性脑膜炎中，PC ae C44：6的浓度显着高于所有其他诊断组，在经典细菌性脑膜炎病原体（例如，肺炎链球菌，脑膜炎奈瑟氏菌，金黄色葡萄球菌）的患者中，PC ae C44：6的浓度高于那些毒性或机会性较低的患者病原体作为病原体（P = 0.026）。PC ae C44：6浓度仅与CSF细胞计数适度相关（Spearmanρ= 0.45； P = 0.009），表明它们不仅反映了神经炎症。在接收器工作特征曲线分析中，PC ae C44：6在区分细菌性脑膜炎与病毒性脑膜炎/脑炎和自身免疫性中枢神经系统疾病（AUC均为0.93）的能力上与CSF细胞计数相等，但敏感性更高（91％比41％）。和阴性预测值（98％比89％）。包含细胞计数，乳酸和PC AE C44：6的诊断算法的敏感性为97％（特异性87％），阴性预测值为99％（阳性预测值为61％），并且正确诊断了四个细菌性脑膜炎样本中的三个由于细胞计数和乳酸值较低，因此分类错误，不能提示细菌性脑膜炎。结论CSF PC ae C44增加：