摘要目的 建立高通量测序技术检测重症肺炎病原体的流程和方法,并评估其可行性.方法 临床对照研究. 收集2015年3月至2016年12月深圳儿童医院就诊的76例重症肺炎的患者(实验组)和18例气管软化并排除细菌感染的患者(对照组),采集肺泡灌洗液标本,同时进行临床检测和高通量测序,并对高通量测序的结果采用PCR方法进行确认,然后对基于高通量测序的重症肺炎病原体检测方法进行评估.采用χ2检验对高通量测序组和非高通量测序组的检出率进行相关性分析.结果 建立了基于高通量测序的重症肺炎病原体的检测流程及方法,该流程包括:标本收集、DNA提取、文库构建、上机测序及生物信息学分析.在76例重症肺炎患者中,66例患者的肺泡灌洗液标本的高通量测序结果为阳性,灵敏度86.84%,显著高于传统临床检验所用的细菌培养、免疫荧光法和定量PCR 3种方法综合检测的灵敏度(68.42%,52/76,χ2=7.426,P<0.001).66份高通量测序阳性标本中,共检测到13种病原体,其中主要的病原体包括肺炎支原体、肺炎链球菌、流感嗜血杆菌和腺病毒等,非高通量测序法共检测到9种病原体.比较临床检验和高通量测序结果,在76份实验组标本中,发现61份(80.26%)标本两者结果完全一致,15份(19.74%)标本结果不完全一致,这些不一致检出的标本主要集中在高通量测序检测出腺病毒、肺炎链球菌和流感嗜血杆菌而临床培养和免疫荧光方法不能检出这些病原体.PCR验证结果显示高通量检测结果与PCR的检测结果差异没有统计学意义(χ2=0.517,P=0.472),高通量检测结果与培养及免疫荧光方法检测结果差异有统计学意义(χ2=11.671,P<0.001).结论 基于高通量测序的重症肺炎病原体的检测方法灵敏度高,能检测出未知病原体,优于传统临床检验所用方法.

abstractsObjective To establish the pipeline and evaluate the feasibility of high-throughput sequencing method used in the detection of severe pneumonia pathogens.Methods Clinical control study was used.Bronchi alveolar lavage fluids (BALF) samples from 76 patients with severe pneumonia (treatment group) and 18 patients with tracheal malacia (control group) and without clinical detected pathogens were collected during March 2015 to December 2016 in Shenzhen Children′s Hospital.The pathogens in the samples were detected using clinical tests and high-throughput sequencing respectively.The results of high-throughput sequencing were confirmed by real-time quantitative PCR and the high-throughput sequencing method used in the detection of severe pneumonia pathogens was evaluated.The χ2 test was used to analyze the correlation of detection rate between the high-throughput sequencing group and the non high-throughput sequencing group.Results The pipeline and method of high-throughput sequencing used in the severe pneumonia pathogens detection was established.The pipeline included sample collection, DNA extraction, library construction, sequencing, and bioinformatic analysis.In 76 cases of patients with severe pneumonia, the results of high-throughput sequencing in 66 cases of bronchoalveolar lavage fluid specimens were positive.The sensitivity was 86.84%, which was significantly higher than the total sensitivity of traditional clinical detection methods including bacterial culture, immunofluorescence and quantitative PCR(68.42%,52/76),χ2=7.426,P<0.001.A total of 13 pathogens were detected in 66 positive samples of high-throughput sequencing, including Mycoplasma pneumoniae, Streptococcus pneumoniae, Haemophilus influenzae and adenovirus, etc.Nine kinds of pathogens were detected in these samples through non-high-throughput sequencing.In the experimental group, the results obtained by clinical test and high-throughput (80.26%) were entirely consistent in 61 samples and not completely consistent in 15 samples (19.74%) specimens.These inconsistent results were mainly concentrated in the detection of adenovirus, Streptococcus pneumoniae and Haemophilus influenzae through high-throughput sequencing, whereas clinical cultures and immunofluorescence methods were not able to detect these pathogens.PCR validation showed that there was no significant difference between the results of high-throughput sequencing and the PCR tests (χ2=0.517,P=0.472), and the difference between the results of high-throughput sequencing and traditional clinical detection methods was statistically significant (χ2=11.671,P<0.001).Conclusion The method for the detection of severe pneumonia pathogens based on high-throughput sequencing is highly sensitive and can detect unknown pathogens, which is superior to those used in traditional clinical detection.