Diagnostic and Monitoring Value of β-2 Transferrin and Transferrin for Intracranial Infection After Neurosurgery.

Abstract

After neurosurgery, intracranial infection is a common complication with high rates of clinical impairment and death. Traditional diagnostic approaches are time-consuming. Early and correct diagnosis improves infection control, treatment success, and survival. Novel markers are used to diagnose and classify post-neurosurgical meningitis (PNM) to overcome the difficulties of diagnosing postoperative intracranial infections and avoid the drawbacks of existing diagnostic measures. The objective was to investigate the diagnostic value of β-2 transferrin (β-2TF) and transferrin (TF) in the cerebrospinal fluid (CSF) for the identification of intracranial infection after neurosurgery. Owing to their symptoms and laboratory results, 168 patients with suspected intracranial infection after neurosurgery were divided into 3 groups: post-neurosurgical bacterial meningitis (PNBM; n = 61), post-neurosurgical aseptic meningitis (PNAM; n = 45), and non-PNM (n = 62). We measured lactate (LA), β-2TF, and TF levels in the CSF. CSF LA levels were significantly higher in the PNM, PNBM, and PNAM groups compared with the non-PNM group ( P < .05). The CSF β-2TF level in PNM, PNBM, and PNAM were statistically higher than those in non-PNMs ( P < .05). CSF TF levels in the PNBM group were statistically higher than those in the PNAM and non-PNM groups ( P < .05). The PNBM and non-PNM receiver operating curve (ROC) analysis indicates that the cutoff values for the combination (LA, β-2TF, TF) was 0.349, and the area under the curve (AUC) was 0.945 ( P < .0001), with 92.86% sensitivity and 92.98% specificity. The PNAM and non-PNM ROC analysis indicates that the cutoff values for the combination (LA, β-2TF, TF) was 0.346, and the AUC was 0.962 ( P < .0001), with 89.29% sensitivity and 90.24% specificity. The PNM and non-PNM ROC analysis indicates that the cutoff values for the combination (LA, β-2TF, TF) was 0.609, and the AUC was 0.941 ( P < .0001), with 96.36% sensitivity and 82.83% specificity. A Glasgow Coma Scale score ≤8, LA, β-2TF/TF ratio, length of hospital stay, intensive care unit admission, poor surgical wound, and craniotomy were associated with poor outcomes ( P < .05). LA and β-2TF were independent risk factors for intracranial infection. Postoperative cerebral infections can be identified using CSF β-2TF as a particular marker protein. CSF TF helps distinguish PNBM from PNAM. Combining CSF LA with them improves diagnostic speed, sensitivity, and accuracy. LA and β-2TF were independent risk factors for cerebral infection.