Epidemiology Of Central Nervous System Infection And Clinical Pharmacokinetics Of Antimicrobial Agents In Plasma And Cerebrospinal Fluid

ObjectiveCentral Nervous system（CNS）infections is an infectious disease caused by the invasion of pathogens such as bacteria and fungi.Due to the difficulty of treatment,it will prolong the hospitalization time,increase the hospitalization cost,and even lead to the death of patients.Early empirical administration based on bacterial prevalence and drug resistance is critical for the treatment of patients with CNS infection.By analyzing the risk factors affecting patient mortality and multi-drug-resistant（MDR）bacterial infections,it can help doctors adopt corresponding strategies to improve patient prognosis timely.In addition to the selection of appropriate antibiotics,the pharmacokinetic characteristics of drugs in the cerebrospinal fluid（CSF）are also important to the treatment of CNS infections.Therapeutic drug monitoring was performed to understand the distribution of antibiotics in vivo,providing a basis for the standardized and reasonable administration of antibiotics.In this paper,the epidemiological analysis of CNS infection was conducted to understand the bacterial prevalence and antimicrobial sensitivity results,and the risk factors affecting the prognosis of patients and MDR bacterial infection were assessed.Meanwhile,plasma and CSF samples of patients with CNS infection were collected for drug concentration detection.In order to provide help for the treatment of patients with CNS infection.一、Epidemiology of central nervous system infection in a grade 3A hospital from2012 to 2019Methods:1.Collection of clinical information and drug sensitivity data of patients:relevant data of patients with positive CSF culture in a grade 3A hospital from January 2012 to December 2019 were retrospectively collected,including demographic data,CSF microbial culture and drug sensitivity results,and relevant laboratory test results.2.Statistical analysis:Continuous variables were compared between groups using independent sample t test or non-parametric test,and categorical variables were compared between groups using Chi-square test or Fisher’s exact test.Factors with P<0.1 in univariate analysis were further included in a binary logistic regression analysis to determine risk factors for patient mortality and MDR bacterial infection.Results:1.Patients and pathogens information:A total of 72 patients with CNS infection were included,with an average age of 49.7±18.3 years,including 54 males（75%）and52 neurosurgical patients（72.2%）.A total of 86 strains of pathogenic bacteria were found,among which the proportion of Gram-positive bacteria,gram-negative bacteria and fungi were 59.3%,30.2%and 10.5%,respectively.The predominant gram-positive bacteria were Coagulase-negative staphylococcus（Co NS）（43.0%）,Enterococcus faecium（5.8%）and Streptococcus pneumoniae（5.8%）.Acinetobacter baumannii（7.0%）,Escherichia coli（7.0%）and Klebsiella spp.（5.8%）were common gram-negative bacteria.In addition,6 of the 9 fungal strains were Cryptococcus neoformans.2.Distribution of pathogenic bacteria:Gram-positive bacteria were still the main pathogenic bacteria during 2016-2019 compared with 2012-2015 years,but the proportion of gram-negative bacteria increased significantly（from 15.4%to 42.6%）.In particular,Streptococcus pneumoniae and Klebsiella spp.were found only in the latter period.The distribution of pathogenic bacteria in different departments was compared.The results showed that the most common pathogenic bacteria in neurosurgery patients was Co NS,and the predominant pathogenic bacteria in non-neurosurgery patients was Cryptococcus neoforman.There was a significant difference between the two groups.3.Antimicrobial sensitivity data:Among Gram-positive bacteria,Co NS had 80%sensitivity to rifampicin,and 100%sensitivity to vancomycin,teicoranin and linezolid;All Streptococcus pneumoniae were susceptible to levofloxacin,moxifloxacin,vancomycin and teicoranin.The sensitivity of Enterococcus faecium to vancomycin,teicoranin and linezolid was 100%.Among gram-negative bacteria,minocycline（83.3%）and tigecycline（100%）were highly sensitive to Acinetobacter baumannii.Escherichia coli was 100%sensitive to amikacin,meropenem and imipenem.In addition,the sensitivity of Cryptococcus neoformans to 5-fluorouracil,amphotericin B,fluconazole and voraconazole was 100%.4.Risk factors of patient death:in our study,50 patients survived and 22 patients died,with a mortality rate of 30.6%.In univariate analysis,we found signifcant differences between the two groups in the following aspects（P<0.05）:age,pulmonary infection,hypertension,CSF glucose,CSF protein percentage,CSF white blood cell count,blood neutrophils,alanine aminotransferase,aspartate aminotransferase,serum albumin level and urea.Based on Logistic regression analysis,we hypothesized that pulmonary infection and CSF glucose<normal were independent risk factors for death in patients with CNS infection.5.Risk factors for MDR bacterial infection:In univariate analysis,we found that patients infected with MDR bacteria had more surgery,indwelling of ventricular drainage tube/lumbar ciudad drainage tube,intracerebral hemorrhage,and increased adenosine deaminase（ADA）,protein,and white blood cells in CSF.Based on Logistic regression analysis,we hypothesized that CSF ADA>normal value and indignant ventricular drainage tube/lumbar cisternal drainage tube were independent risk factors for MDR bacterial infection.二、Methods for monitoring the concentration of antibioticsMethods:1.To establish HPLC method for determination of meropenem:the mobile phase consisted of methanol-PBS(0.005 mol·L^-1KH₂PO₄and 0.005 mol·L^-1Na₂HPO₄)with gradient elution.Chromatographic colum:Agela Technologies（4.6 mm×150 mm,5μm）.Column temperature at 30℃.Detection wavelength:298 nm.Injection volume 20μL.Flow rate 1 m L·min^-1.Plasma and CSF samples were treated with methanol using ceftazidime as internal standard.The specificity,linearity,limit of quantitation,recovery,precision and stability of the method were investigated.2.Vancomycin was determined by enzyme amplification immunoassay,detection reagent and matching calibration solution were purchased from Siemens;Tigecycline was determined by two-dimensional liquid chromatography,which was established by Hunan Demeter Company.The required mobile phase,sample treatment solution,freeze-dried quality control products and freeze-dried calibration products were purchased from the company.Results:The linear range of plasma melopenem was 1-80μg·m L^-1,the lower limit of quantitation was 1μg·m L^-1,and the standard curve equation was Y=0.0182x+0.0013,r=0.9998.The linear range of CSF was 0.25-20μg·m L^-1,the lower limit of quantification was 0.25μg·m L^-1,and the standard curve equation was Y=0.0917x-0.0173,r=0.9998.The results of methodological investigation showed that the method had good specificity,with intra-day and intra-day RSD<15%.The plasma and CSF samples of meropenem were stable after being placed at 4℃for 24h and cryopreserved at-80℃for 1 month and repeated freeze-thaw for 3 times.三、Pharmacokinetic study of meropenem,vancomycin and tigecycline in the cerebrospinal fluid in patients with CNS infection after neurosurgeryMethods:1.Patients diagnosed with or highly suspected of CNS infection after neurosurgery and treated with meropenem,vancomycin or tigecycline were selected,and plasma and CSF samples and relevant clinical data were collected after medication.2.The concentrations of each drug in plasma and CSF were determined.Meropenem was determined by HPLC.Vancomycin was determined by enzyme amplification immunoassay.Tigecycline was determined by two-dimensional liquid chromatography.DAS2.0 software was used to fit the drug concentration-time curve in CSF and calculate relevant pharmacokinetic parameters.Results:1.A total of 10 patients with CNS infection after neurosurgery were included in this study,including 2 patients receiving meropenem alone,2 patients receiving vancomycin alone,1 patient receiving tigecycline alone,4 patients receiving meropenem combined with vancomycin,and 1 patient receiving meropenem combined with tigecycline.A total of 10 plasma samples and 55 CSF samples were determined from 10 patients.2.The half-lives of meropenem in the CSF of were 1.717h,1.978h,4.896h,and the peak time was 2-3h after the end of administration.The peak concentrations were2.01,3.52,5.5μg·ml^-1,and the AUC was 13.96,18.16,21.19 h·μg/m L.The BBB penetration were 56.7%（1.22/2.15）and 34.3%（1.08/3.15）,respectively.3.The mean trough concentration of CSF vancomycin was 2.22±0.87μg·m L^-1（0.7-3.3）in patients by intravenous administration.Trough concentrations of CSF vancomycin were 4.7 and 0.8/0.5μg·m L^-1,in the two patients receiving intravenous combined intrathecal vancomycin,respectively.The BBB penetration was 39.8%（3.3/8.3）in a patient by compared CSF and plasma vancomycin trough concentrations within the same dose interval.4.The half-life of tigecycline in CSF was 5.251h and 8.253h,respectively,and the peak time was 3-4h after the end of administration,with peak concentrations of 34.307and 77.862ng·ml^-1and AUC of 295.39 and 806.78 h·ng/m L.5.Most of the patients included in this study showed significant improvement in CNS infection symptoms after medication.Conclusion:1.Gram-positive bacteria are still the predominant pathogenic bacteria of CNS infection in our hospital,but the proportion of Gram-negative bacteria has increased significantly in recent years.Gram-positive bacteria were 100%sensitive to vancomycin,while among gram-negative bacteria,Acinetobacter baumannii and Klebsiella pneumoniae were 100%sensitive to tigecycline.2.There are significant differences in the pathogenic bacteria of patients with CNS infection in different departments,and appropriate antibacterial agents should be selected based on different departments.3.Age>50 years,complicated with pulmonary infection,CSF glucose<normal value was an independent risk factor for death.CSF ADA>normal values and presence of external ventricular drainage/lumbar cistern drainage were associated with MDR bacterial infection.4.This study established a HPLC method for the determination of meropenem concentration in plasma and CSF.The method has strong specificity and high accuracy,which can be used for the concentration monitoring of meropenem in clinical patients.5.Among the 3 patients receiving meropenem,the peak time of drug in CSF was3-4 h after the end of administration,and there were significant differences in the half-life,peak concentration,AUC and BBB transmittance of melopenem in CSF.6.One patient in this study who received vancomycin did not achieve higher concentrations of the drug in CSF even with combined intrathecal administration.7.In the 2 patients receiving tigecycline,the drug peak time in CSF was 2～3h after the end of administration,and the half-life,peak concentration and AUC of tigecycline in CSF were significantly different.