| ObjectiveTo explore the diagnostic value of metagenomic next-generation sequencing(mNGS)in severe lower respiratory tract infection(LRTI)with different immune statuses,and its actual impact on guiding clinical decision-making and prognosis.MethodsThe clinical data of 201 patients with severe LRTI admitted to the respiratory intensive care unit(RICU)of the First Affiliated Hospital of Zhengzhou University from November 2018 to December 2020 were retrospectively analyzed.The patients were divided into immunocompromised hosts(ICHs)and non-immunocompromised hosts(NICHs)according to whether there was an underlying disease that caused immunodeficiency,and they were also divided into the improvement group and the deterioration group according to the ICU prognosis.Bronchoalveolar fluid(BALF)samples were obtained from bronchoscopy for mNGS analysis as well as smear microscopy/culture.The diagnostic performance of pathogens was compared between mNGS and smear microscopy/culture.The distribution of pathogens in patients with different immune statuses was compared.The initial anti-infective therapies of all patients were analyzed according to the etiological results,and the adjustments of treatment regimens based on the results of mNGS or traditional microbiological tests were compared between ICHs and NICHs.Multivariate logistic regression was used to analyze the effect of adjusting treatment regimens according to mNGS results on the prognosis of severe LRTI patients with different immune statuses.Results1.A total of 201 patients with severe LRTI were included,including 124(61.7%)males and 77(38.3%)females,aged 15-97(56.35±17.19)years.120(59.7%)cases were ICHs,and 81(40.3%)cases were NICHs.The deterioration rate of ICHs was significantly higher than that of NICHs(50.8%vs 34.6%,P<0.05).2.Of 201 patients,the overall microbial yield was 90.5%for mNGS versus 58.7%for smear microscopy/culture and 67.2%for comprehensive traditional pathogen methods(P<0.05).Compared with smear microscopy/culture,mNGS showed a significantly higher positive rate in detecting Gram-positive bacteria(15.4%vs 4.5%,P<0.05)and Pneumocystis(32.3%vs 3.0%,P<0.05).There were no statistically significant differences in the detection rate of Gram-negative bacteria,Mycobacterium tuberculosis,non-tuberculous Mycobacteria and other fungi(all P>0.05).Viruses were detected by mNGS in BALF specimens of 134(66.7%)patients.In this study,PCR was only performed for viral nucleic acid detection in 100 patients,of which 34(34.0%)cases yielded positive results.3.Compared with NICHs,the detection rate of bacteria was lower in ICHs(50.0%vs 65.4%,P<0.05),while that of fungi(68.3%vs 49.4%,P<0.05)and viruses(75.0%vs 54.3%,P<0.05)was higher.There was no significant difference in the prevalence of various bacteria between the two groups.In terms of fungi,the detection rate of Pneumocystis in ICHs was significantly higher than that in NICHs(51.7%vs 3.7%,P<0.05).According to mNGS results,the detection rate of Human herpes viruses and Torque teno viruses in ICHs was higher than that in NICHs(Human herpes viruses:64.2%vs 30.9%,P<0.05;Torque teno viruses:15.8%vs 6.2%,P<0.05),while the detection rate of respiratory viruses was lower(16.7%vs 30.9%,P<0.05).4.Among 201 patients,the pathogenic microorganisms were completely covered by the initial anti-infection regimens in 79(39.3%)cases,partially covered in 78(38.8%)cases,and uncovered in 30(14.9%)cases;while 14(7.0%)cases failed to identify the etiology.Among the 187 patients with clear etiology,67(35.8%)cases maintained the initial treatment regimen,66(35.3%)cases were adjusted based on mNGS results alone,34(18.2%)cases were adjusted according to the results of mNGS as well as traditional etiological detections,and 20(10.7%)cases were adjusted based on the traditional etiological detection results alone.The complete coverage rate of initial therapies in ICHs was higher than that in NICHs(46.7%vs 28.4%,P<0.05).Correspondingly,after obtaining the pathogen results,more patients in ICHs continued to maintain the initial regimen than those in NICHs(41.7%vs 26.4%,P<0.05),and fewer patients in ICHs adjusted it based on the traditional etiology test results alone(7.0%vs 16.7%,P<0.05).Meanwhile,the proportion of adding uncovered medications in NICHs was higher than that in ICHs(58.3%vs 40.0%,P<0.05).5.The multivariate logistic regression analysis revealed that lower PaO2/FiO2,higher APACHE Ⅱ score,lower CD4+T cell count,and mixed infection were independent risk factors for ICU prognosis of ICHs with severe LRTI(all P<0.05),while adjusting the treatment regimen based on mNGS results was not finally included in the logistic regression equation of ICHs.For NICHs,a history of the cardiovascular disease,lower PaO2/FiO2,higher APACHE Ⅱ score,and lower CD4+T cell count were independent risk factors for ICU deterioration(all P<0.05),and adjusting the medication regimen based on mNGS results was a protective factor(OR=0.172,95%CI 0.030-0.989,P=0.049).Conclusions1.For patients with severe LRTI,the mNGS detection of BALF could improve the positive rate of pathogens,especially for Gram-positive bacteria,Pneumocystis,and viruses.2.The spectrum of detected pathogens differs between ICHs and NICHs.3.The identification of pathogens through mNGS has a certain guiding significance for the adjustment of clinical decision-making in patients with severe LRTI.4.Mixed infection is an independent risk factor for the ICU deterioration of ICHs with severe LRTI;however,this study fails to prove the beneficial effect of adjusting the medication regimen according to mNGS results on their ICU prognosis.For NICHs with severe LRTI,timely adjusting the treatment regimen based on mNGS results could help to improve their ICU prognosis. |