The Use Of Next-generation Sequencing In Cryptococcal Meningitis

Cryptococcal meningitis(CM)is the most common presentation of cryptococcosis with high morbidity and mortality.Early diagnosis,improved treatment of complications and individualized medication are key to reducing mortality.The clinical manifestations of CM are lack of specificity and the diagnosis mainly depends on cerebrospinal fluid(CSF)etiology.The traditional methods are influenced by the operator’s experience,technical level and reagent,the positive rate is not consistent.Next-Generation Sequencing(NGS)is an unbiased tool that can theoretically identify all pathogens with known sequences and detect a broad range of bacterial,viral,fungal or parasite DNA sequences simultaneously in a CSF sample.As sequencing technology matures and the cost of detection gradually decreases,NGS may become a routine detection method and replace part of the traditional detection.It has been reported in several literatures that NGS can be used to detect pathogen in central nervous system(CNS)infection.However,its application in CM is not clear.ObjectiveCSF samples of CM patients and non-cm patients were tested by NGS to determine the positive criteria,sensitivity and specificity of NGS in the diagnosis of CM.Compare different diagnostic methods of CM and further explore the relationship between NGS results and clinical indicators and treatment effect of patients.Methods1.Samples collection: 534 CSF samples suspected CNS infection were collected retrospectively from January 2012 to May 2019,as well as the clinical information of the corresponding patients’ clinical symptoms,signs and auxiliary examinations.The entry criteria for the case group were with symptoms and signs of CNS infection,and in line with the diagnostic criteria of CM.The entry criteria for control group were with symptoms and signs of CNS infection,in line with the diagnostic criteria for other CNS infections,or brain(membrane)inflammation and myelitis with cytological and biochemical of CSF abnormalities and cryptococcal pathogens negative.2.NGS: Nucleic acid was extracted from the included CSF samples.After library construction,sequencing and bioinformatics analysis were performed by Illumina Mini Seq sequencer.Plasmid DNA at known concentrations and sterile water were used as positive and negative controls,respectively.3.Sensitivity and specificity of NGS and comparison with other methods: The ROC curve was performed according to the unique reads of Cryptococcus in the case group and control group to obtain the positive cut-off value,sensitivity and specificity of NGS in the diagnosis of CM.Compare the positive rate of NGS with other methods.4.Analysis of NGS results: To analyze the correlation between the unique reads of Cryptococcus and the clinical indicators of patients without drugs.To explore whether NGS could judge the therapeutic effect of patients with drugs and whether NGS negative could be used as the standard for drug withdrawal.Results1.Results of sample groupingAmong the 534 samples,51 samples were excluded due to incomplete clinical information.Of the remaining 483 samples,82 were diagnosed with CM,of which 45(54.88%)were not treated with antifungal agents,and 37(45.12%)were treated with antifungal agents and included in the case group.21 samples were diagnosed as other CNS infections and 78 samples of brain(membrane)inflammation or myelitis with cryptococcal pathogens negative were included in the control group(99 samples).2.Sensitivity and specificity of NGSROC curve was performed according to unique reads of the case group and control group.When the number of unique reads was 3,the sensitivity and specificity were the highest,which were 70.7% and 98.9%,respectively.3.Comparison of NGS and other methodsAmong the 82 samples in the case group,the positive rates of NGS,lateral flow immunoassay(LFA),May-Grunwald-Giemsa(MGG)staining,culture,Alcian blue staining and ink staining were: 70.7%,92.65%,77.78%,67.35%,64.81% and 58.23%,respectively.The positive rate of NGS was lower than that of LFA and MGG staining and higher than other methods.Among the 45 samples without antifungal drugs,the positive rates of the six methods were 86.67%,95.35%,79.49%,85.29%,73.53% and 63.64%,respectively.The positive rate of NGS was only lower than that of LFA,there was no statistical difference,and higher than other methods.The positive rate of NGS and culture was statistically different between 45 samples without antifungal drugs and 37 samples with antifungal drugs(P<0.05),but no significant difference was found by other methods.NGS can also detect cryptococcus subtypes and other potential pathogens in a single test.4.Analysis of NGS results(1)Clinical indicators of patients without drugs: For patients without antifungal drugs,the number of unique reads was weakly negatively correlated with fever.The number of unique reads was moderately correlated with intracranial pressure,moderately negatively correlated with white cell count,weakly negatively correlated with protein.The number number of unique reads in HIV-positive patients was significantly higher than that in HIV-negative patients(P<0.05).(2)Therapeutic effect of patients with drugs: The positive rate of clinical symptoms and signs of the NGS negative samples was lower than that of the NGS positive samples.The differences of fever,nausea,vomiting and focal neurological dysfunction were statistically significant(P<0.05).The drug duration,CSF sugar and chloride of the NGS negative samples were higher than that of the NGS positive samples,and the differences were statistically significant(P<0.05).Intracranial pressure and white cell count were lower in the NGS negative samples,but the difference was not statistically significant.Among the 16 NGS negative samples,7 samples were negative for CSF ink staining or MGG staining,while 9 samples were positive,but white cells and biochemistry in CSF were close to normal.Conclusion1.NGS can be used as an effective method in diagnosis of CM.2.NGS can detect cryptococcus subtypes and other potential pathogens in a single test.3.For patients without antifungal drugs,increased intracranial pressure and impaired immune function should be considered when the number of unique reads is high.For patients with antifungal drugs,NGS negative samples were better than that of the NGS positive samples and were closer to withdrawal criteria.A prospective study with a larger sample is needed.