exclude: true --- class: top left hide-count background-image: url(img/PXL_20211017_215933273_trim.jpeg) background-size: cover .fade-in[ .move-top10[ .title-text[Deep Sequencing for<br> Respiratory Pathogen Detection<br> & Infection Diagnosis] ] .callout-url-leftcorner[ .title-nametext[ Brendan J. Kelly, MD, MS Infectious Diseases & Epidemiology University of Pennsylvania <br> <br> <!--- 05 November 2021 ---> ] ] ] --- ## Disclosures .pad-left[ - No commercial support or conflicts of interest. - Research supported by: - NIAID: K23 AI121485 - CDC: BAA 200-2016-91964, BAA 200-2018-02919, BAA 200-2021-10986 ] --- ## Respiratory Tract Infection: Public Health Challenge .pad-left[ - Lower respiratory tract infections (LRTIs) were the fourth leading cause of mortality for all ages in 2019 -- even before the global COVID-19 pandemic! - LRTIs account for nearly 4 million deaths annually. - Multidrug-resistant (MDR) bacterial LRTIs have been associated with increased reinfection risk and mortality, relative to non-resistant bacterial infections: - effect of factors that increase risk for MDR (e.g., underlying lung disease)? - effect on time to effective antimicrobial therapy? ... is microbiologic diagnosis complete without resistance profile? ] .footnote[Vos et al. _Lancet_ 2020; Vincent et al. _JAMA_ 2009; Weiner et al. _MMWR_ 2016; Kumar et al. _Crit Care Med_ 2006] --- ## Respiratory Tract Infection: Diagnostic Challenge .pad-left[ - Challenges to sensitivity: - potential etiologies include viruses, bacteria, fungi - low microbial biomass and/or fastidious organisms - difficulty of sampling (e.g., nonproductive cough) - Challenges to specificity: - all non-surgical samples contaminated by the upper airway (bronchoscopic carryover even with protected-specimen brush) ] .footnote[Charlson et al. _Am J Resp Crit Care_ 2011; Charlson et al. _PLoS One_ 2012; Dickson et al. _Ann Rev Physio_ 2016] --- ## Deep Sequencing for Culture-Free Diagnosis .pad-left[ - Approach: - extraction of nucleic acid from upper/lower respiratory specimen → 16S rRNA gene or shotgun metagenomic sequencing - Potential advantages: - rapid (depending on bioinformatics) & less biased (but not unbiased) - Potential disadvantages: - low sensitivity: microbial DNA swamped by host - misclassification risk: metagenomic assignment/assembly - limited accuracy of genotype-based AMR prediction ] --- class: full-screen hide-count background-color: #FCF5E5 <div class="grid-3-1"> <div class="extension-tile gridset-a animated flipInY"><div>Sequence-Based<br>Pathogen<br>Detection</div></div> </div> --- layout: false class: full-screen hide-count background-color: #FCF5E5 <div class="grid-3-1"> <div class="extension-tile gridset-a"><div>Sequence-Based<br>Pathogen<br>Detection</div></div> <div class="extension-tile gridset-b animated flipInX"><div>Microbial<br>Ecology &<br>Infection Diagnosis</div></div> </div> --- layout: false class: full-screen hide-count background-color: #FCF5E5 <div class="grid-3-1"> <div class="extension-tile gridset-a"><div>Sequence-Based<br>Pathogen<br>Detection</div></div> <div class="extension-tile gridset-b"><div>Microbial<br>Ecology &<br>Infection Diagnosis</div></div> <div class="extension-tile gridset-c animated bounceInDown"><div>Dis/Advantages<br>of a Sequence-<br>First Strategy</div></div> </div> --- layout: false class: full-screen hide-count background-color: #FCF5E5 <div class="grid-3-1"> <div class="extension-tile gridset-a"><div>Sequence-Based<br>Pathogen<br>Detection</div></div> <div class="extension-tile gridset-grey"><div>Microbial<br>Ecology &<br>Infection Diagnosis</div></div> <div class="extension-tile gridset-grey"><div>Dis/Advantages<br>of a Sequence-<br>First Strategy</div></div> </div> --- exclude: true class: center middle <iframe src="./figs/ggplot_plotly_raster_demo.html" width = "95%", height = "95%" framBorder="0"></iframe> --- background-image: url(img/kelly_microbiome_2016_heatmap.png) background-size: 70% .footnote[Kelly et al. _Microbiome_ 2016] --- ## 16s rRNA Sequencing for Bacterial Pneumonia .pad-left[ - concordance with clinical cultures: - 4/15 subjects with + LRT cultures - 16S rRNA identifies all as dominant 24-72hrs before culture - dominance of unknown clinical significance: - _Enterococcus faecalis_ domination post RSV pneumonia - _Ureaplasma parvum_ in an immunocompromised host - the "normal flora" problem: standards for ignoring expected taxa? ] .footnote[Kelly et al. _Microbiome_ 2016] --- background-image: url(img/pendleton_ajrccm_2017_top.png) background-size: contain .footnote[Pendleton et al. _AJRCCM_ 2017] --- background-image: url(img/pendleton_ajrccm_2017_bottom.png) background-size: contain .footnote[Pendleton et al. _AJRCCM_ 2017] --- ## 16s rRNA & Metagenomic Pathogen Detection .pad-left[ - high sensitivity & early detection of culture-positive cases - require tailored methods for _Mycobacteria_, fungi, viruses - lots of dominance of unclear clinical significance: - correlates with markers of inflammation and lung injury - should this receive directed antibiotic treatment? - byproduct of antibiotic treatment? ] .footnote[Graf et al _JCM_ 2016; Abbas _Am J Transpl_ 2017; Kitsios et al. _Front Micro_ 2018; Langelier et al. _PNAS_ 2018; Zinter et al. _CID_ 2019; Dickson et al. _AJRCCM_ 2020; McGinniss et al. _J Heart Lung Transpl_ 2021] --- ## Conclusions: Sequence-Based Pathogen Detection .pad-left[ - Advantages: - rapid & less biased - Disadvantages: - need new standard for "normal flora" - no single extraction/sequencing modality: _Mycobacteria_, fungi, viruses (efforts made to develop pipelines -- e.g., SURPI*) - AMR prediction in its infancy - Best of both worlds with multiplex NAAT? ] .footnote[Kelly et al. _Microbiome_ 2016; Kitsios et al. _Front Micro_ 2018; *Langelier et al. _PNAS_ 2018; Dickson et al. _AJRCCM_ 2020] --- layout: false class: full-screen hide-count background-color: #FCF5E5 <div class="grid-3-1"> <div class="extension-tile gridset-grey"><div>Sequence-Based<br>Pathogen<br>Detection</div></div> <div class="extension-tile gridset-b"><div>Microbial<br>Ecology &<br>Infection Diagnosis</div></div> <div class="extension-tile gridset-grey"><div>Dis/Advantages<br>of a Sequence-<br>First Strategy</div></div> </div> --- background-image: url(img/kelly_microbiome_2016_alpha_simple.png) background-size: 70% .footnote[Kelly et al. _Microbiome_ 2016] --- ## Dominance, Diversity & Infection Risk .pad-left[ - more comprehensive and quantitative approach permits: - <u>richness</u>: number of different species - <u>evenness</u>: balance across observed species - <u>diversity</u>: summary of richness & evenness - <u>dominance</u>: > 30% or > 50% of sequence reads - 16S rRNA qPCR or other standard-based quantitation of DNA/RNA abundance (use of cycle threshold values for SARS-CoV-2) ] .footnote[Dickson et al. _Lancet Resp Med_ 2014; Kelly et al. _Microbiome_ 2016; Kitsios et al. _Front Micro_ 2018; Dickson et al. _AJRCCM_ 2020; Kitsios et al. _AJRCCM_ 2020; Harrigan et al. _CID_ 2021] --- background-image: url(figs/p_combined_mdi_crosssection_counterfact.png) background-size: contain .footnote[Harrigan et al. _CID_ 2021] --- background-image: url(figs/p_minimpute_persistent_mdi.png) background-size: contain .footnote[Harrigan et al. _CID_ 2021] --- background-image: url(figs/p_minimpute_compare_outcome_definitions_persistence.png) background-size: contain .footnote[Harrigan et al. _CID_ 2021] --- ## Dominance, Diversity & Infection Risk .pad-left[ - 83 subjects dependent on mechanical ventilation - Sampled 1066 days to compare microbiome disruption indices (MDIs): - cross-sectional MDIs (low Shannon diversity, high 16S rRNA gene qPCR) associated with risk for VA-LRTI, but poor predictors - persistent low Shannon diversity had per-day LRTI odds ratio of 1.36 (95% credible interval 1.10 to 1.72) - observed association was consistent across multiple definitions of VA-LRTI ] .footnote[Harrigan et al. _CID_ 2021] --- ## Conclusions: Microbial Ecology of Infection .pad-left[ - Predictive utility of microbial community structure: - challenge of gold standard for pneumonia diagnosis - optimal utility with longitudinal measures/sampling - independent of pathogen identification (strength & weakness) - Advantage of deep sequencing over multiplex NAAT. - Is it worth the work? ] --- layout: false class: full-screen hide-count background-color: #FCF5E5 <div class="grid-3-1"> <div class="extension-tile gridset-grey"><div>Sequence-Based<br>Pathogen<br>Detection</div></div> <div class="extension-tile gridset-grey"><div>Microbial<br>Ecology &<br>Infection Diagnosis</div></div> <div class="extension-tile gridset-c"><div>Dis/Advantages<br>of a Sequence-<br>First Strategy</div></div> </div> --- ## Deep Sequencing First? .pad-left[ - Disadvantages: - need to redefine "normal flora" (specificity questions) - broad sensitivity demands complex wet-bench & bioinformatic pipeline - metagenomic AMR prediction is in early stages - Advantages: - speed and sensitivity - added predictive/diagnostic value of microbial community structure - __pathogen tracking and infection control__ -- application to public health ] --- background-image: url(figs/covid_cases_chloropleth.png) background-size: 30% background-position: 85% 50% ## SARS-CoV-2 Variant Tracking .pull-left[ - The value of extracted nucleic acids! - Draw on broad catchment of PennMedicine, led by Kyle Rodino, PhD: - ~ 1500 SARS-CoV-2 NAATs per day - positives with Ct < 28, stratified by county and zip code - ARTIC primer set to amplify → Illumina - assembly, alignment, PANGO lineage ] .pull-right[ ] --- background-image: url(img/marques_arma_model.png) background-size: 80% background-position: 50% 90% .move-top5[ <img src="./img/marques_arma_model_key.png" width="70%" style="display: block; margin: auto;" /> ] .footnote[Marques et al. _medRxiv_ 2021] --- background-image: url(img/marques_vaccine_model.png) background-size: contain .footnote[Marques et al. _medRxiv_ 2021] --- layout: false class: full-screen hide-count background-color: #FCF5E5 <div class="grid-3-1"> <div class="extension-tile gridset-a animated flipInX"><div>Sequence-Based<br>Pathogen<br>Detection</div></div> <div class="extension-tile gridset-b animated flipInY"><div>Microbial<br>Ecology &<br>Infection Diagnosis</div></div> <div class="extension-tile gridset-c animated flipInX"><div>Dis/Advantages<br>of a Sequence-<br>First Strategy</div></div> </div> --- ## Conclusions .pad-left[ - Specificity challenges deep sequencing for respiratory pathogen identification: - nature of upper airway contamination - need to redefine standards for "normal flora" - Novel insights into microbial ecology provide alternate framework for infectious disease diagnosis: skip the CT scan and monitor lower respiratory bacterial community diversity/dominance? - As multiplex NAAT and diagnostic sequencing applications expand, opportunity to collaborate with infection control and public health colleagues to build bioinformatics infrastructure. ] --- ## Acknowledgements .pull-left[ - __ARES Group @ Penn__ Laurel Glaser, Kyle Rodino, Sean Loughrey, Laura Cowden, Magda Wernovsky, Jennifer Han, Erik Clarke, Michael David, Matt Ziegler, Lauren Dutcher, Ebbing Lautenbach, Jim Harrigan, Hatem Abdallah - __Bushman Laboratory @ Penn__ Arwa Abbas, Aoife Roche, Andrew Marques, Aubrey Bailey, Jacob Leiby, Rick Bushman - __PennCHOP Microbiome Program__ Lisa Mattei, Casey Hofstaedter, Huanjia Zhang, Kyle Bittinger ] .pull-right[ - __Collman Laboratory @ Penn__ Ize Imai, Aurea Simon Soro, John McGinniss, Ron Collman - __Division of ID @ Penn__ Ian Frank, Pablo Tebas, Robert Gross, Emily Blumberg - __Rutgers University & Penn DBEI__ Jason Roy, Arman Oganisian - __CDC Prevention Epicenters__ Clifford McDonald, Alison Laufer Halpin - __Funding__ CDC BAAs 200-2016-91964, 200-2018-02919, 200-2021-10986; NIAID: K23 AI121485 ] .center[ ### brendank@pennmedicine.upenn.edu ] --- class: middle center hide-count background-image: url(img/PXL_20211017_215933273_trim.jpeg) background-size: cover .title-subtext[Questions?] .callout-url-bottom[ <!-- .fade-in[<span style="font-size:0.5em; color:#113864">slides ↓ </span>] --> <span style="color:black; link-color:black;">[bjklab.org](http://www.bjklab.org)</span> <br> ] --- exclude: true ## bjklab.org .center[ <iframe src="http://bjklab.org" width="95%" height="400px" data-external="1"></iframe> ]