As infection is a common complication and cause of death in hematological disease, early detection of microorganisms causing infections is particularly important for the improvement of prognosis. The newly emerged metagenomic next-generation sequencing (mNGS) technology has the advantages of simple operation, small sample size required, fast detection, unsusceptible to contamination. Moreover, it has better performance than culture methods in detecting marginal microorganisms.

To determine the causes of infections in patients with hematological diseases via analyzing the bacterial and viral pathogen spectra of infections in such patients detected using mNGS, to inform relevant clinical treatment.

Participants (n=53) were hematological disease patients who had clinical infections or suspected infections symptoms during the hospitalization in Department of Hematology, the First Affiliated Hospital of Guangxi Medical University from August 2018 to December 2020. Samples collected from them included whole blood, cerebrospinal fluid, pleural effusion, tissue, abdominal drainage fluid, joint fluid, sputum, fluid aspirated by puncture of the right lower extremity. For detecting microorganisms causing infections, mNGS technology was used for all patients (58 samples) , culture was used for 52 patients (55 samples) , G/GM test was used for 46 patients (50 samples) , and PCR test was used for 44 patients (48 samples) as well.

The bacterial and viral pathogen spectra of infections detected by mNGS technology showed that the most common Gram-positive bacteria were Propionibacterium acnes (42 times) , Staphylococcus epidermidis (33 times) , and Staphylococcus hominis (32 times) , the most common Gram-negative bacteria were Acinetobacter johnsonii (26 times) , Burkholderia vietnamiensis (20 times) , and Burkholderia ubonensis (19 times) , the most common fungus was Malassezia restricta (28 times) , and the most common viruses were Cytomegalovirus (26 times) , Epstein-Barr virus (14 times) and Torque teno virus (19 times) . Toxoplasma gondii was detected in the blood and cerebrospinal fluid of a patient with severe thalassemia after hematopoietic stem cell transplantation. Compared to culture method, G/GM test, and PCR test, mNGS technology had a higher detection rate (P<0.05) . The medication of 6 patients were adjusted according to the detection results of mNGS, and 4 of them were much improved, but the other 2 cases were still poorly treated.

This study indicates that mNGS technology is contributive to the determination of causes of clinical infections and guidance on treatment in patients with hematological diseases. However, due to possible false positive and false negative rates in the detection by mNGS, mNGS technology is recommended to be used in combination with other detection methods, which could reduce the possibility of misdiagnosis.

感染是血液病患者常见的并发症和死亡原因，病原微生物的早期检测对于改善血液病患者的预后尤为重要。近年来出现的宏基因组高通量测序（mNGS）技术具有操作简单、所需样本量小、检测快、不易受污染、可检测出培养法不能检测的边缘微生物的优点。

探讨mNGS技术检测所揭示的感染谱，以明确血液病患者的感染病因并指导治疗。

标本来源于2018年8月至2020年12月广西医科大学第一附属医院血液科住院期间出现感染或疑似感染症状的53例血液病患者，对53例患者的58个样本（全血、脑脊液、胸腔积液、组织、腹腔引流液、关节液、痰液、右下肢穿刺液）进行mNGS技术检测。在进行mNGS技术检测的同时，52例患者（55个样本）采用培养法检测致病微生物、46例患者（50个样本）采用G/GM试验检测致病微生物、44例患者（48个样本）采用PCR检测致病微生物。

mNGS所揭示的血液病患者的感染谱显示，革兰阳性菌中最常见为痤疮丙酸杆菌（42例次）、表皮葡萄球菌（33例次）、人葡萄球菌（32例次），革兰阴性菌中最常见的为约翰逊不动杆菌（26例次）、越南伯克霍尔德菌（20例次）、乌汶克霍尔德菌（19例次），真菌中最常见的为限制性马拉色菌（28例次）；病毒中最常见的为巨细胞病毒、EB病毒、细环病毒（分别为26、14、19例次）。1例重型地中海贫血造血干细胞移植后患者血液及脑脊液均检出刚地弓形虫。与培养法、G/GM试验、PCR相比，mNGS技术具有较高的阳性率（P<0.05）。本研究中根据mNGS技术检测结果调整用药的患者有6例，其中4例经治疗好转，2例治疗效果欠佳。

本研究利用mNGS技术揭示了血液病患者感染的病原体谱，证实mNGS技术可有效帮助确定血液病患者感染病因和指导治疗，但其可能出现假阳性和假阴性，结合其他检测方法可降低错误诊断的机会。