《Enhanced Performance of Next-Generation Sequencing Diagnostics Compared With Standard of Care Microbiological Diagnostics in Patients Suffering From Septic Shock.pdf》由会员分享,可在线阅读,更多相关《Enhanced Performance of Next-Generation Sequencing Diagnostics Compared With Standard of Care Microbiological Diagnostics in Patients Suffering From Septic Shock.pdf(9页珍藏版)》请在得力文库 - 分享文档赚钱的网站上搜索。
1、Downloaded from http:/ by BhDMf5ePHKbH4TTImqenVJKnHXMTrNzpDxfgK8xMyFr3+7Ndz+AKjjLr3rLkpi50pll5g3tDpCQ= on 02/07/2019Downloaded from http:/ by BhDMf5ePHKbH4TTImqenVJKnHXMTrNzpDxfgK8xMyFr3+7Ndz+AKjjLr3rLkpi50pll5g3tDpCQ= on 02/07/2019Critical Care Medicine www.ccmjournal.org 1Objectives: Culture-based
2、 diagnostics represent the standard of care in septic patients, but are highly insensitive and in many cases unspecific. We recently demonstrated the general fea-sibility of next-generation sequencing-based diagnostics using free circulating nucleic acids (cell-free DNA) in plasma samples of septic
3、patients. Within the presented investigation, higher per-formance of next-generation sequencing-based diagnostics was validated by comparison to matched blood cultures.Design: A secondary analysis of a prospective, observational, single-center study.Setting: Surgical ICU of a university hospital and
4、 research labo-ratory.Patients: Fifty patients with septic shock, 20 uninfected patients with elective surgery as control cohort.Interventions: None.Measurements and Main Results: From 256 plasma samples of 48 septic patients at up to seven consecutive time points within the 28-day observation perio
5、d, cell-free DNA was isolated and analyzed by next-generation sequencing and relevance scoring. In parallel, results from culture-based diagnostics (e.g., blood culture) were obtained. Plausibility of blood culture and next-generation sequencing results as well as adequacy of antibiotic therapy was
6、evaluated by an independent expert panel. In con-trast to blood culture with a positivity rate of 33% at sepsis onset, the positivity rate for next-generation sequencing-based pathogen identification was 72%. Over the whole study period, blood cul-ture positivity was 11%, and next-generation sequenc
7、ing positivity was 71%. Ninety-six percent of positive next-generation sequenc-ing results for acute sepsis time points were plausible and would have led to a change to a more adequate therapy in 53% of cases as assessed by the expert evaluation.Conclusions: Our results show that next-generation seq
8、uencing-based analyses of bloodstream infections provide a valuable DOI: 10.1097/CCM.00000000000036581Fraunhofer IGB, Stuttgart, Germany.2Noscendo GmbH, Duisburg, Germany.3Department of Anesthesiology, Heidelberg University Hospital, Heidel-berg, Germany.4Westpfalz-Klinikum GmbH, 1, Hellmut-Hartert-
9、Strae, Kaiserslautern, Germany.Drs. Brenner and Sohn share senior authorship.Drs. S. Grumaz, Decker, Hofer, Brenner, and Sohn conceived of, designed, and supervised the study. Drs. Decker, Hofer, and Brenner col-lected clinical samples and clinical data. Drs. S. Grumaz, C. Grumaz, and Glanz performe
10、d all sample processing and next-generation sequencing experiments. Drs. Vainshtein and Stevens performed bioinformatic data processing and statistical analyses. Drs. S. Grumaz, C. Grumaz, Stevens, and Sohn analyzed the data. Drs. Hofer, Weigand, Brenner, and Sohn provided materials. Drs. S. Grumaz
11、and C. Grumaz prepared the tables and figures. Drs. S. Grumaz and Sohn wrote the article with contributions from all other authors. All authors read, critically revised, and approved the final article.A data visualization tool associated with this article can be viewed here: https:/lippincott.shinya
12、pps.io/Sohn/.Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journals website (http:/ in part, by the Fraunhofer IGB (Stuttgart, Germany), Fraun-hofer Future Foundation, D
13、epartment of Anesthesiology (University of Heidelberg, Germany), Heidelberg Foundation of Surgery (University of Heidelberg, Germany). Our study sponsors were not involved in study design, collection, analysis, or interpretation of data.Drs. Decker and Brenner received project funding from Heidelber
14、g Foun-dation of Surgery. Drs. S. Grumaz, Stevens, and Sohn disclosed that they are co-founders of the company Noscendo active in molecular diagnostics for infectious diseases. The remaining authors have disclosed that they do not have any potential conflicts of interest.For information regarding th
15、is article, E-mail: kai.sohnigb.fraunhofer.deCopyright 2019 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. This is an open access article distributed under the Creative Commons Attribution License 4.0 (CCBY),
16、 which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Enhanced Performance of Next-Generation Sequencing Diagnostics Compared With Standard of Care Microbiological Diagnostics in Patients Suffering From Septic ShockSilke Grumaz, P
17、hD1; Christian Grumaz, PhD1; Yevhen Vainshtein, PhD1; Philip Stevens, PhD2; Karolina Glanz1; Sebastian O. Decker, MD3; Stefan Hofer, MD4; Markus A. Weigand, MD3; Thorsten Brenner, MD3; Kai Sohn, PhD1S. Grumaz et al2 www.ccmjournal.org XXX 2019 Volume XX Number XXX diagnostic platform for the identif
18、ication of clinically relevant pathogens with higher sensitivity and specificity than blood cul-ture, indicating that patients might benefit from a more appropriate therapy based on next-generation sequencing-based diagnosis. (Crit Care Med 2019; XX:0000)Key Words: blood culture; cell-free nucleic a
19、cids; deep sequencing; molecular diagnostic techniques; sepsis; septic shockSepsis is a major health concern with increasing fre-quency and an estimated global mortality rate of 5.3 million deaths per year (1). Time is crucial in the man-agement of septic patients and early treatment, including an-t
20、ibiotic administration and source control are the decisive first steps that influence patients outcome dramatically (24). Current guidelines recommend the initiation of antimicrobial therapy as early as possible and preferably within 1 hour (5). However, most of early treatments are empirical, and 4
21、6% of empirical antibiotic treatments were shown to be inappro-priate and associated with 35% mortality (6). About 50% was either unnecessary or too broad spectrum, increasing the risk for resistance and toxicity. Early recognition of the infecting microorganism is therefore crucial for a targeted a
22、ntimicrobial therapy, reducing side effects for the patient and improving patients outcome. The current standard of care blood culture (BC) however is limited by long time to positivity, low sensi-tivity, and low specificity.We recently published a proof of concept study, in which the general applic
23、ability of microbial circulating cell-free DNA (cfDNA) to diagnose causative pathogens in sepsis using a sig-nificance scoring system and to identify single-gene resistances via next-generation sequencing (NGS) was demonstrated (7).The aim of the presented study was to evaluate the perfor-mance of t
24、he NGS-based diagnostic approach with a larger cohort of patients and benchmark it by direct comparison to the current standard of care BC. Due to the limitations of BC, our findings were reviewed by an independent expert jury for plausibility, and the antimicrobial treatment regimen was reas-sessed
25、 for putative changes.METHODSStudy DesignData result from a secondary analysis of septic patients (n = 50) participating in a previously published, prospective ob-servational clinical study of our workgroup, which was con-ducted in the surgical ICU of Heidelberg University Hospital, Germany between
26、November 2013 and January 2015 (German Clinical Trials Register: DRKS00005463) (8). The focus of this primary study was the immune response to fungal infections in patients suffering from septic shock, and three patients (S16, S25, and S35) were already described with special focus on fungal infecti
27、ons in detail, including NGS results as heat maps within this study. All study patients or their legal designees gave written informed consent. In total 50 patients suffering from septic shock according to the criteria of the Surviving Sepsis Campaign: International Guidelines for Management of Seve
28、re Sepsis and Septic Shock 2012 were enrolled in this study (9). Treatment of patients with septic shock included early-goal directed therapy (10), elimination of the septic focus, and broad-spectrum antibiotic therapy (1012). Blood samples were collected at sepsis onset (T0) and 1 day (T1), 2 days
29、(T2), 7 days (T3), 14 days (T4), 21 days (T5), and 28 days (T6) thereafter. Relevant baseline data (demographic data and primary site of infection), clinical data (disease severity scores such as Simplified Acute Physiology Score II, Sequen-tial Organ Failure Assessment SOFA score, Acute Physiology
30、and Chronic Health Evaluation II score, surgical procedures, antifungal therapy, and outcome variables) as well as routine infection variables (e.g., leukocytes, C-reactive protein CRP, procalcitonin, and body temperature) were collected. In addi-tion, 20 postoperative patients following major abdom
31、inal surgery without any evidence of infection were included as controls. Routine infection variables (e.g., leukocytes, CRP, procalcitonin, and body temperature), BCs, and other clinical microbiological specimens were without pathologic findings in this group. Plasma samples from the post-surgery g
32、roup were collected prior to surgery (T0), immediately after the surgical procedure (T1), and 24 hours later (T2). Two septic patients as well as three patients of the post-surgery control group were retrospectively excluded from further analyses due to technical reasons, resulting in 48 septic pati
33、ents and 17 post-surgery control patients. A workflow diagram to il-lustrate the study design and the NGS diagnostics workflow in context with clinical data is provided in Figure S1 (Supple-mental Digital Content 1, http:/ Additional details about patients, study time points, and samples are provide
34、d in Data S2 (Supplemental Digital Con-tent 1, http:/ Whole blood for BC and for plasma preparation for NGS testing was drawn on the same day (different EDTA tubes). Study and control patients or their legal designees signed written informed consent. All study procedures were approved by the local e
35、thics committee (Ethics Committee of the Medical Faculty of Heidelberg, Trial Code No. S-097/2013).Clinical MicrobiologyAt Heidelberg University Hospital BC testing is routinely per-formed as described previously (13). Quantification of herpes simplex virus 1 DNA and cytomegalovirus DNA from plasma
36、or tracheal secretion was carried out via quantitative real-time polymerase chain reaction as described previously (14). Culti-vation of wound swabs, catheter, and stool samples was carried out as previously described (15, 16).Plasma Preparation and Nucleic Acid IsolationPlasma was prepared, and nuc
37、leic acid isolation was per-formed as described previously (8). If plasma volumes were below 1,000 L after centrifugation, the respective samples were excluded (Data S2, Supplemental Digital Content 1, http:/ Contamination controls Online Clinical InvestigationCritical Care Medicine www.ccmjournal.o
38、rg 3were prepared following the same procedure, and quality control steps were carried out as already described (8).Preparation of NGS Libraries and SequencingLibrary preparation and sequencing were carried out as previ-ously described (7) from 1 ng cfDNA using the Nextera XT li-brary preparation ki
39、t (Illumina, San Diego, CA) with a Biomek FXP liquid handling robot (Beckman Coulter, Brea, CA). The utilized raw data for NGS-based diagnostics are deposited in the European Nucleotide Archive under the following acces-sion numbers: PRJEB21872 and PRJEB30958.BioinformaticsBioinformatic processing a
40、nd sepsis indicating quantifier (SIQ) score calculation were carried out as already described (7). Briefly, after bioinformatic removal of human sequences and taxonomic classification of nonhuman sequences, the SIQ score provides a quantitative and probabilistic assessment of every detected mi-crobe
41、 in the respective sample based on a noninfected control group and permits a comparison between different samples, when identically processed. After normalization of read counts to the library size, a likelihood estimate of the probability to ob-serve a certain species in the control group is genera
42、ted. Under the assumption that all read counts for a certain species are Pois-son distributed based on data generated from the control group, a p value that assesses the likelihood for read abundances outside this Poisson distribution is calculated. This p value along with a species-specific factor
43、and the normalized read abundance gives rise to the individual SIQ score of a certain species in a patient sample. Control samples from elective surgery patients, which passed the quality control restrictions were added to our data-base to serve as the noninfected control group. The same criteria as
44、 previously described (8) were applied to exclude species hits (Fig. S3, Supplemental Digital Content 1, http:/ with the following exception: thresholds for bacte-rial, viral, and fungal hits were 10 normalized reads.Evaluation Clinical Expert PanelTo assess the plausibility of the calculated SIQ sc
45、ores, a panel of eight clinicians specialized in intensive care, clinical micro-biology, and infectiology were asked to answer a questionnaire (Fig. S4, Supplemental Digital Content 1, http:/ The participants were from Heidelberg University Hospital but had no role in the study. As most positive BCs
46、 were obtained at sepsis onset, only the first three time points (T0, T1, T2) were subjected to evaluation. Every case was introduced with the patients anamnesis, antibiotic therapy and all results from BC and NGS at T0, T1, and T2 as well as microbiological results from other specimens, if availabl
47、e. The concept of the SIQ score was explained to the clinical experts as well as the visualization of the results in form of heat maps, with highest scores in dark red and low scores without coloring. Color scal-ing was however only comparable within the results from one patient. The experts were as
48、ked to make an overall assessment of the plausibility of reported species for the individual time point, not for each species individually and note if they deemed the overall result plausible with respect to the patient history of underlying primary diseases, surgeries, complications, and mi-crobiol
49、ogical results from other specimens. Majority rules were obtained as further described in the figure legend of Figure S4 (Supplemental Digital Content 1, http:/ How NGS results and clinical data were integrated is fur-thermore illustrated in Figure S1 (Supplemental Digital Con-tent 1, http:/ and Stu
50、dy DesignIn total, 48 patients with septic shock were included in the investigation, whose characteristics are described in detail in Table 1. The primary septic focus was the abdomen (n = 43; 90%), followed by the lung (n = 4; 8%) as well as the genitou-rinary tract (n = 1; 2%). In 31 patients (65%