The C. diff Debate: The Role of Diagnostics in Disease Determination

The C. diff Debate: The Role of Diagnostics in Disease Determination Saturday Evening June 3, 2017 7:30 PM ASM Microbe 2017 Bissonet Meeting Room New Orleans Marriott

Program Disclosures Planned and developed by Medavera, Inc. and supported by an educational grant from Alere, Inc. This event is neither sponsored nor endorsed by the American Society for Microbiology. 2

Agenda 7:30 7:55 PM C. difficile: New Developments and Discoveries Moderator Proposition: NAATs are Best for Determining CDI 7:55 8:10 PM NAATs are the Best Diagnostic Approach to CDI 8:10 8:25 PM NAATs are Not the Best Diagnostic Approach to CDI Ciarán P. Kelly, MD Professor Ferric C. Fang, MD Professor Mark H. Wilcox, MD 8:25 8:30 PM Rebuttal Professor Ferric C. Fang, MD 8:30 8:35 PM Rebuttal Professor Mark H. Wilcox, MD 8:35 8:40 PM Closing Arguments 8:40 8:50 PM Panel Discussion: Arguments For and Against 8:50 9:00 PM Q & A 3

Learning Objectives Identify new developments and discoveries with C. difficile Review current guidelines for C. difficile diagnosis and prevention Assess CDI testing methodologies and current controversies Apply findings to determine the appropriate protocol and testing algorithms for CDI for one s institution 4

Continuing Education Credits Educational Review Systems is an approved provider by P.A.C.E. This program is approved for 1.5 hours of CE credit. In order to obtain your CE credit, please visit CdiffDebate.com and follow the instructions. 5

This program is being audio recorded and will be produced along with the slides into an enduring material which will be available within 3 weeks at CdiffDebate.com For those who were unable to attend this live program, they may apply for P.A.C.E. credits after viewing. Please send an email to info@medavera.com if you have questions or need assistance.

C. difficile: New Developments and Discoveries Ciarán P. Kelly, MD Professor of Medicine Harvard Medical School Director Gastroenterology Fellowship Training Beth Israel Deaconess Medical Center Boston, Massachusetts

Disclosures Ciarán P. Kelly, MD Commercial Interest Relationship Role Actelion Honorarium Consultant GSK Honorarium Consultant Institut Merieux Grant Merck Honorarium, Grant Consultant, Speaker National Institutes of Health, U.S. Grant Seres Therapeutics Honorarium Consultant, Speaker Summit Honorarium Consultant Synthetic Biologics Honorarium Consultant Vedanta Honorarium Consultant 8

Setting the Stage Clostridium difficile Infection (CDI) Multiple clinical outcomes An urgent threat in the U.S. Therapies The old The new The future Several diagnostic approaches Find the smoking gun or Catch the bullet? 9

C. difficile Infection - Pathogenesis Not always especially community-acquired CDI Antibiotics or Neonate or IBD Antibiotic therapy Disturbed colonic microflora (Loss of colonization resistance) Nosocomial common But spores are ubiquitous Strains differ in virulence C. difficile exposure and colonization Strains are toxinogenic or not Toxin production Anti-toxin immunity is protective Kelly CP, LaMont TJ. N Engl J Med. 2008;359:1932-40. Kyne L, Warny M, Qamar A, et al. Lancet. 2001;357:189-93. Symptomless carriage Diarrhea and colitis 10

BIDMC C. difficile Infection Results in Variable Clinical Outcomes Aslam S, Hamill RJ, Musher DM. Lancet Infect Dis. 2005;5:549-57. 11

Nosocomial C. difficile Infection and Asymptomatic Carriage Are Common Hospital patients (Acute medical ward) LOS > 2 days Receiving antibiotic 271 enrolled Colonized by C. difficile 84 (31%) Hospital-acquired 47 (17%) Colonized at admission 37 (14%) CDI case 28 (10%) Carrier 19 (7%) CDI case 19 (7%) Carrier 18 (7%) Kyne L, Warny M, Qamar A, et al. N Engl J Med. 2000;342:390-97. * In progress study (also at BIDMC, Boston) In 2017, the carrier rate in this population is 10%.* 12

C. difficile Toxin Production Does regulation of toxin production determine clinical outcomes? Should we quantify bacterial burden or toxin concentrations? Kelly CP, LaMont TJ. N Engl J Med. 2008;359:1932-40. 13

Clostridium difficile An Urgent Threat http://www.cdc.gov/drugresistance/threat-report-2013/ 14

Incidence of Both Principal and Secondary CDI Diagnoses Continues to Increase All-Listed Diagnoses Lucado J, Gould C, Elixhauser A. HCUP Statistical Brief #124. http://www.hcup-us.ahrq.gov/reports/statbriefs/sb124.pdf. Accessed 5/26/17. 15

Estimated No. of CDI Cases Burden of C. difficile Infection in the U.S. (2011) Primary CDI cases in U.S. in 2011 ~480,000 ~2/3 of CDI cases were healthcare associated infections but only ~1/4 were hospitalized when symptoms began Most community-associated CDI cases (82%) appeared to be associated with outpatient healthcare exposures Community-associated CDI Healthcare-associated CDI Leesa FC, Mu Y, Bamberg WM, et al. N Engl J Med. 2015;372(9):825-34. 16

% of All HAIs C. difficile Infection Is a Common Cause of Healthcare Associated Infection (HAI) Most Common Healthcare-associated Infections 25% 20% 21.8% 21.8% 17.1% Clostridium difficile was the cause of 70.9% of healthcare-associated gastrointestinal infections 15% 12.9% 10% 9.9% 5% 0% Pneumonia Surgical-Site Infection Gastrointestinal Infection Urinary Tract Infection Primary Bloodstream Infection C. difficile is the most commonly reported specific pathogen causing 12.1% of HAI. Magill SS, Edwards JR, Bamberg W, et al. N Engl J Med. 2014;370(26):2542-3. 17

Recurrences and Deaths Associated With CDI: Adjusted U.S. National Estimates for 2011 Recurrence Rate Mortality Rate Unmet medical needs in CDI: Mortality Recurrence Definitive diagnosis #1 High Incidence 25% 20% 15% 10% 5% 0% 13.5% CA-CDI 20.9% HA-CDI 10% 9% 8% 7% 6% 5% 4% 3% 2% 1% 0% 1.3% CA-CDI 9.3% HA-CDI 14 through 56 days following initial CDI Within 30-days of the initial positive C. difficile testing results Leesa FC, Mu Y, Bamberg WM, et al. N Engl J Med. 2015;372(9):825-34. 18

C. difficile Infection Why the Increasing Incidence, Severity and Recurrence Rates? Host factors Age Immune response Underlying disease Bacterial factors Virulence Sporulation Antibiotic resistance Environment Antibiotic use PPI use Burden of contamination by C. difficile spores Kelly CP, LaMont TJ. N Engl J Med. 2008;359:1932-40. Bauer MP, Kuijper EJ, van Dissel JT, et al. Clin Microbiol Infect. 2009;15:1067 79. Cohen SH, Gerding DN, Johnson S, et al. Infect Control Hosp Epidemiol. 2010;31(5):431-55. 19

C. difficile Infection Approaches to Therapy Bacteriotherapies e.g. FMT (and FMT refinements), Non-toxinogenic C. difficile Selective anti-c. difficile antibiotics e.g. fidaxomicin, cadazolid, ridinilazole Toxin binders Immune therapies: Passive anti-toxin infusion or active vaccination Kelly CP, LaMont TJ. N Engl J Med. 2008;359:1932-40. Kyne L, Warny M, Qamar A, et al. Lancet. 2001;357:189-93. Antibiotic therapy Disturbed colonic microflora (Loss of colonization resistance) C. difficile exposure and colonization Toxin production Symptomless carriage Diarrhea and colitis 20

C. difficile Infection Approaches to Therapy Bacteriotherapies e.g. FMT (and FMT refinements), Non-toxinogenic C. difficile Antibiotic therapy Disturbed colonic microflora (Loss of colonization resistance) C. difficile exposure and colonization Toxin production Kelly CP, LaMont TJ. N Engl J Med. 2008;359:1932-40. Kyne L, Warny M, Qamar A, et al. Lancet. 2001;357:189-93. Symptomless carriage Diarrhea and colitis 21

Duodenal Infusion of Donor Feces for Recurrent C. difficile Infection Microbiota Diversity Donors Recipients After Infusion Recipients Before Infusion van Nood E, Vrieze A, Nieuwdorp M, et al. N Engl J Med. 2013;368(5):407-15. 22

Non-Toxinogenic C. difficile Spores Nature s Tailor-Made Probiotic? NTCD (Non-toxinogenic C. difficile) Spores of strain VP20621 Protects hamsters against colonization by toxinogenic C. difficile and against CDI Phase II Trail All with CDI or oral vancomycin Placebo (N = 43) NTCD (Total N = 125) 10 4 x 7 days (N = 41) 10 7 x 7 days (N = 43) 10 7 x 14 days (N = 41) P < 0.0001 P < 0.01 2% Recurrence rate in those colonized by NTCD Gerding DN, Meyer T, Lee C, et al. J Am Med Assoc. 2015;313(17):1719-27. 23

C. difficile Infection Approaches to Therapy Antibiotic therapy Selective anti-c. difficile antibiotics e.g. fidaxomicin, cadazolid, ridinilazole Disturbed colonic microflora (Loss of colonization resistance) C. difficile exposure and colonization Toxin production Kelly CP, LaMont TJ. N Engl J Med. 2008;359:1932-40. Kyne L, Warny M, Qamar A, et al. Lancet. 2001;357:189-93. Symptomless carriage Diarrhea and colitis 24

Fidaxomicin vs. Vancomycin for Clostridium difficile Infection Initial Response Recurrence Sustained Response Louie TJ, Miller MA, Mullane KM, et al. N Engl J Med. 2011: 364(5):422-31. FDA approved for CDI therapy 2011 25

C. difficile Infection Approaches to Therapy Antibiotic therapy Disturbed colonic microflora (Loss of colonization resistance) Toxin binders C. difficile exposure and colonization Toxin production Kelly CP, LaMont TJ. N Engl J Med. 2008;359:1932-40. Kyne L, Warny M, Qamar A, et al. Lancet. 2001;357:189-93. Symptomless carriage Diarrhea and colitis 26

Phase 3 Multicenter Trials of Tolevamer (Oral Toxin-Binding Agent) vs. Metronidazole vs. Vancomycin for CDI * P < 0.001, tolevamer (T) vs. metronidazole (M) and T vs vancomycin (V) ** P = 0.020, M vs. V > 1,000 patients randomized 2:1:1 to T, M, & V between 2005 & 2007 Johnson S, Louie TJ, Gerding DN, et al. Clin Infect Dis. 2014;59:345-54. 27

C. difficile Infection Approaches to Therapy Antibiotic therapy Disturbed colonic microflora (Loss of colonization resistance) C. difficile exposure and colonization Immune therapies: Passive anti-toxin infusion or active vaccination Kelly CP, LaMont TJ. N Engl J Med. 2008;359:1932-40. Kyne L, Warny M, Qamar A, et al. Lancet. 2001;357:189-93. Toxin production Symptomless carriage Diarrhea and colitis 28

Bezlotoxumab (Neutralizing Anti-Toxin B IgG HuMab) Infusion During Standard of Care CDI Therapy Reduces Later CDI Recurrence Reduction in CDI recurrence: Absolute: 10% Relative: 37% Wilcox MH, Gerding DN, Poxton IR, et al. N Engl J Med. 2017;376(4):305-17. 29

Antitoxin Immunization to Break the Cycle of Dysbiosis in Recurrent C. difficile Infection Antibiotic therapy Disturbed Restored colonic colonic microflora microflora (Loss of colonization resistance) (return of colonization resistance) X Recurrence Passive anti-toxin immunotherapy C. difficile exposure and colonization Kelly CP, LaMont TJ. N Engl J Med. 2008;359:1932-40. Kyne L, Warny M, Qamar A, et al. Lancet. 2001;357:189-93. Toxin production X Symptomless carriage Diarrhea and colitis X X Dysbiosis Antibiotic treatment X 30

Diagnostic Tests May Detect Spores, Vegetative Forms, and/or Toxins Phase Spore Vegetative Toxin producing and releasing Detection assay Culture NAAT (e.g. PCR) Culture NAAT (e.g. PCR) GDH (Glutamate DeHydrogenase) Tissue culture cytotoxicity Enzyme immunoassay (EIA) New ultra sensitive & quantitative EIAs 31

CDI Diagnostics Debate CDI is a very common HAI - a growing and urgent threat Colonization by toxinogenic C. difficile may be fatal or symptomless Hence any stool test result must be interpreted in the context of the clinical picture Question: Which approach to stool testing for CDI is best? Find: The smoking gun (toxinogenic organism) or The bullet (toxins) Without accurate and clinically meaningful diagnostics our current or newer, more effective but also more expensive, therapies cannot be employed effectively Aslam S, Hamill RJ, Musher DM. Lancet Infect Dis. 2005;5:549-57. 32

The C. diff Debate: The Role of Diagnostics in Disease Determination Saturday Evening June 3, 2017 7:30 PM ASM Microbe 2017 Bissonet Meeting Room New Orleans Marriott

NAATs Are the Best Diagnostic Approach for Clostridium difficile Infections Ferric C. Fang, MD Professor of Laboratory Medicine and Microbiology Adjunct Professor of Medicine (Infectious Diseases) Director, Harborview Medical Prof. Center Ferric Clinical C. Fang Microbiology Laboratory University of Washington School School of Medicine of Medicine Seattle, Seattle, Washington, Washington USA

Disclosures Ferric C. Fang, MD Commercial Interest Relationship Role Biofire Honorarium, Reagents Consultant, Speaker, Researcher Cepheid Honorarium Consultant, Speaker 35

A True Statement About C. difficile A. Patients with positive nucleic acid amplification test (NAAT) and negative toxin EIA assays are merely colonized. B. Glutamate dehydrogenase (GDH) is as sensitive as NAAT and can be used as a cost-effective screening test. C. Patients with C. difficile diarrhea have higher toxin levels than those who are asymptomatically colonized. D. Patients with C. difficile diarrhea have higher organism burdens than those who are asymptomatically colonized. E. All of the above. F. None of the above. 36

C. difficile Infection (CDI) Is Caused by Toxins REF: Kelly and Lamont. New Engl J Med, 2008. 38

Percent The Presence of Toxigenic C. difficile or Toxin Is Necessary but Not Sufficient for Disease 100 80 Toxigenic Cx Toxin 60 40 20 0 Asymptomatic (65) Diarrhea (31) Colitis (3) McFarland LV, Elmer GW, Stamm WE, et al. Infect Immun. 1991;2456-62. 39

Cytotoxin Assays Are Insensitive for CDI N Toxin Toxigenic Cx Reference 149 70% 98% Arch Intern Med. 1986;146:95. 144 67% 97% Diagn Micro Infect Dis. 1988;10:85. 38 74% 92% Infect Immun. 1991;59:2456. 19 70% 93% J Clin Microbiol. 1992;30:1837. 85 71% 100% Eur J Clin Microbiol Infect Dis. 1992;11:246. 43 81% ND Ann Intern Med. 1995;123:835. 123 82% 95% J Clin Microbiol. 1996;34:2718. 41 78% 100% Pathol Biol. 1999;47;415. 1691 85% 100% J Hosp Infect. 2001;48:233. 85 74% 94% J Clin Microbiol. 2001;39:1996. 40

Enzyme Immunoassays Are Less Sensitive Than Tissue Culture Cytotoxin Assays Peterson LR, Kelly PJ. Infect Dis Clin North Am. 1993;7(2):277-93. 41

NAAT Sensitivity Is Comparable to Toxigenic Culture Direct Culture (No. of Samples) Enrichment Culture (No. of Samples) NAAT Positive Negative Total Positive Negative Total Positive 245 188 433 316 117 422 Negative 3 1860 1863 22 1841 1863 Total 248 2048 2296 338 1958 2296 In a prospective, multicenter trial, NAAT was 99% sensitive and 91% specific compared with direct culture. Tenover FC, Novak-Weekley S, Woods CW, et al. J Clin Microbiol. 2010;48(10):3719 24. 42

NAAT Is More Sensitive Than EIA/Cytotoxin Assays Site No. Site Assay N Sensitivity (%) NAAT Site 1 Toxin A/B EIA 1023 94.1 67.5 2 GDH-EIA 268 91.4 74.3 3 Toxin A/B EIA 293 92.3 53.8 4 Toxin A/B EIA 312 91.4 54.3 5 GDH-EIA-PCR 114 92.3 61.5 6 Toxin A/B EIA 173 97.0 33.3 7 Cytotoxin 110 90.9 54.5 Tenover FC, Novak-Weekley S, Woods CW, et al. J Clin Microbiol. 2010;48(10):3719 24. 43

NAAT Is the Most Cost-Effective Diagnostic Strategy for C. difficile 2-Step GDH Algorithm NAAT Schroeder LF, Robilotti E, Peterson LR, et al. J Clin Microbiol. 2014;52(2):489-96. 44

GDH Sensitivity Is Strain-Dependent Sensitivity (%) P-Value Ribotype N C. difficile PCR NAAT GDHbased Algorithms 027 11 90.9 90.9 1.0 Non-027 36 91.7 69.4% 0.001 Tenover FC, Novak-Weekley S, Woods CW, et al. J Clin Microbiol. 2010;48(10):3719 24. 45

Fecal Toxin Assays Cannot Distinguish Asymptomatic Carriage and CDI Significant Diarrhea No Significant Diarrhea Significant Diarrhea No Significant Diarrhea Anikst VE, Gaur RL, Schroeder LF, et al. Diagn Microbiol Infect Dis. 2016;84(4):343-6. 46

Organism Load Cannot Distinguish Asymptomatic Carriage and CDI Anikst VE, Gaur RL, Schroeder LF, et al. Diagn Microbiol Infect Dis. 2016;84(4):343-6. 47

A Negative Cytotoxin Assay Cannot Rule Out CDI 48

Percent Response Toxin+ and Toxin- Patients Respond Similarly to CDI Treatment 100 Vancomycin Placebo 80 60 40 20 0 NA Cx+ Tox+ Cx+ Tox- Cx- Tox- Keighley MR, Burdon DW, Arabi Y, et al. Br Med J. 1978;2(6153):1667-9. 49

Cytotoxin Assays Can Be Negative Even in Severe CDI Fulminant CDI had a 70% mortality. 12.5% had negative toxin assays Patients diagnosed at autopsy were twice as likely to have a false-negative toxin assay. Dallal RM, Harbrecht BG, Boujoukas AJ, et al. Ann Surg. 2002 ;235(3):363-72. 50

NAAT Reduces Empiric Antibiotic Treatment of C. difficile Clinical and economic outcomes EIA (n = 79) NAAT (n = 87) P Duration of antibiotic therapy in days, mean (CI) 2.31 (1.48-3.15) 0.88 (0.45-1.33) 0.007 Diagnostic test performed per patient, mean (CI) 2.73 (2.64-2.83) 1.16 (1.04-1.28) <0.001 Duration of contact insolation in days, mean (CI) 1.46 (0.61-2.32) 0.62 (0.08-1.32) 0.131 Total treatment cost per patient 1 (CI) 69.54 (43.36-95.73) 65.97 (46.61-85.34) 0.828 Diagnostic test cost 1 13.67 (13.08-14.26) 37.15 (32.51-41.79) <0.001 Antibiotic therapy cost 1 (CI) 36.95 (12.70-61.20) 20.64 (5.08-36.20) 0.262 Contact isolation cost 1 (CI) 19.39 (8.07-30.71) 8.19 (1.14-17.52) 0.131 Antimicrobial prescribing EIA (n = 79) (n = 87) P Doses administered, mean (CI) 7.24 (4.71-9.77) 3.91 (1.71-6.13) 0.051 No treatment, n (%) 54 (68.4) 71 (81.6) 0.071 Metronidazole monotherapy, n (%) 16 (20.2) 11 (12.6) 0.211 Vancomycin monotherapy, n (%) 7 (8.9) 3 (3.5) 0.195 Combination therapy, n (%) 2 (2.5) 2 (2.3) 0.999 Peppard WJ, Ledeboer NA. Hosp Pharm. 2014;49(7):639-43. 51

Switching From EIA to NAAT May Result in Reduced Length-of-Stay The impact of PCR on Clostridium difficile detection and clinical outcomes Mona Akbari, Alina Vodonos, George Silva, Manida Wungjiranirun, Daniel A. Leffler, Ciarán P. Kelly and Victor Novak Characteristic EIA NAAT P value Gender [n (%)] Female 1125 (52.0%) 153 (56.46%) 0.167 Male 1038 (48.0%) 118 (43.54%) Abdominal pain [n (%)] No 2138 (98.84%) 258 (95.20%) <0.001 Yes 25 (1.16%) 13 (4.80%) Age [mean ± SD] 67 ±17 65.3 ±18 0.114 BUN [(mg dl 1 ) mean ± SD] 28.91 ±23.38 29.01 ±23.15 0.945 Cr [(mg dl 1 ) mean ± SD] 1.64 ±1.64 1.72 ±1.80 0.424 Albumin [(g dl 1 ) mean ± SD] 3.09 ±0.65 3.24 ±0.70 0.002 WBC [(x 10 3 cells µl 1 ) mean ± SD] 12.65 ±12.01 12.11 ±9.42 0.474 Hgb [(g dl 1 ) mean ± SD] 11.03 ±1.90 10.87 ±2.02 0.191 Hct [(%) mean ± SD] 33.07 ±5.44 33.40 ±5.94 0.354 Na [(meq dl 1 ) mean ± SD] 137.92 ±5.21 137.38 ±5.29 0.104 HCO 3 [(meq dl 1 ) mean ± SD] 24.62 ±4.81 24.95 ±4.92 0.285 Median length of stay [days (range)] 10 (3 146) 8 (3 151) 0.004 Akbari M, Vodonos A, Silva G, et al. J Med Microbiol. 2015;64(9):1082-6. 52

Number of Cases C. difficile NAAT Can Facilitate Infection Control Cases 1.1 140 Rate 120 1.0 100 80 0.9 60 40 0.8 20 NAAT 0 0.7 2008 2010 2012 2014 2016 Year Unpublished internal institutional data. Rate per 1000 Patient-Days 53

C. difficile Diagnostic Algorithms > 3 soft/liquid stools in 24 hours GDH Immunoassay Toxin Immunoassay Insensitive GDH Immunoassay NAAT + + + Toxin Immunoassay Toxin Immunoassay Potential Underdiagnosis NAAT Potential Overdiagnosis NAAT Unpublished from various laboratories. Moderately Sensitive Potential Overdiagnosis Complex Reporting 54

Conclusions CDI requires a toxigenic organism and a susceptible host diagnostic assays must be interpreted in clinical context Two-step algorithms (GDH/NAAT) are less expensive but also less sensitive and increase the complexity of reporting NAATs are rapid, cost-effective, and sensitive, and facilitate effective infection control and stewardship Overdiagnosis can be minimized by limiting testing to patients with clinical presentations consistent with CDI Reliance on toxin assays can result in underdiagnosis, including some patients with severe and even life-threatening infections 55

Fast is fine, but accuracy is everything. Wyatt Earp 56

NAATs Are Not the Best Diagnostic Approach to CDI Professor Mark Wilcox, MD Leeds Teaching Hospitals & University of Leeds, UK Public Health England Size matters!

Disclosures Mark H. Wilcox, MD Commercial Interest Relationship Role Abbott Honorarium Consultant, Speaker Actelion Honorarium Consultant, Speaker Astellas Honorarium Consultant, Speaker Biomerieux Honorarium Consultant, Speaker Cubist Honorarium Consultant, Speaker Da Volterra Honorarium Consultant, Speaker European Tissue Symposium Honorarium Consultant, Speaker Optimer Honorarium Consultant, Speaker Pfizer Honorarium Consultant, Speaker Qiagen Honorarium Consultant, Speaker Sanofi Pasteur Honorarium Consultant, Speaker Summit Honorarium Consultant, Speaker Alere Honorarium Consultant Basilea Pharmaceutica Honorarium Consultant Durata Therapeutics Honorarium Consultant Merck Honorarium Consultant Nabriva Therapeutics Honorarium Consultant

Size Matters The size of the population at risk of Clostridium difficile infection (CDI) The size of the overlap between CDI and non-cdi cases The size of the CDI diagnosis error rate The size of the consequences of erroneous CDI testing The true size of the CDI challenge 59

Common Causes of Diarrhoea Infectious Clostridium difficile Norovirus Other viruses: rotavirus, adenovirus Other bacteria: Salmonella, Campylobacter, Shigella, Escherichia coli Non-infectious Inflammatory bowel disease Gastrointestinal neoplasia Irritable bowel syndrome Celiac disease Anxiety Food allergy Lactose intolerance Iatrogenic Use of laxatives Nasogastric feeding Post-gastrointestinal surgery Radiotherapy Antibiotic-assoc. diarrhoea (non-infective) Other drugs: selective serotonin reuptake inhibitors, NSAIDs, statins 61

http://nitcollaborative.org.uk/wp/hoodini-hospital-onset-diarrhoea-investigation/ 63

Day Prevalence of Hospital Onset Diarrhoea (HOD) (N = 4,995 Patients on 141 Wards in 32 Hospitals) Hospital type HOD point prevalence 95% C.I. All 4.6% 4.1 5.3% Non-teaching 2.9% 2.3 3.6% Specialty 10.0% 4.2 21.9% Teaching 6.1% 5.3 7.1% HOD: 2 episodes of Bristol Stool Type 5-7 in 24 hours before evaluation day; onset > 48 hours after admission. Mawer D et al. ECCMID 2017. Abstract OS0230. 64

Size Matters The size of the population at risk of CDI The size of the overlap between CDI and non-cdi cases The size of the CDI diagnosis error rate The size of the consequences of erroneous CDI testing The true size of the CDI challenge 65

Hospital Onset Diarrhoea Patients: Key Messages HOD patients were elderly (median age 77) Hospitalised for a median 8 days before diarrhoea onset A quarter had 1 previous HOD episodes that admission 85% had multiple possible causes (range 2-13) 54% were receiving antimicrobials 85% were taking other medication that can cause diarrhoea No difference in the number of potential causes of HOD per patient between those tested for CDI vs. not tested vs. CDI (median 3 per group; P = 0.55) 66

The Wrong Interpretation of The More You Test, the More You Find We introduced PCR testing and our CDI rate went up. No, it didn t. The CDI rate stayed the same; you just issued more positive results. 68

Setting CDI Test Change Inflation of CDI Reported Rate 3 states in CDC s Emerging Infections Program CDI surveillance EIA PCR 43% - 67% Multi-hospital EIA PCR 56% Single center EIA PCR 57% Single center EIA PCR 110% Single center GDH + EIA + CCNA PCR 50% Single center EIA GDH + PCR 97% Single center GDH + EIA GDH + EIA + PCR 70% EIA: enzyme immunoassay; PCR: polymerase chain reaction; GDH: glutamate dehydrogenase; CCNA: cell cytotoxicity neutralization assay Gould CV, Edwards JR, Cohen J, et al. Clin Infect Dis. 2013;57(9):1304-7. Moehring RW, Lofgren ET, Anderson DJ. Infect Control Hosp Epidemiol. 2013;34:1051-66. Goldenberg SD. Infect Control Hosp Epidemiol. 2011;32:1231-32. Williamson DA, Basu I, Freeman J, et al. Am J Infect Control. 2012;41:270-72. Longtin Y, Trottier S, Brochu G, et al. Clin Infect Dis. 2013;56(1):67-73. de Jong E, de Jong AS, Bartels CJ, et al. Eur J Clin Microbiol Infect Dis. 2012;31:2219-25. Fong KS, Fatica C, Hall G, et al. Infect Control Hosp Epidemiol. 2011;32:932-33. 69

12,240 samples analyzed from 10,691 episodes 12,186 patients 8,026 in-patient results from 6,665 episodes 6,355 in-patients Outcome results available for 6,522 inpatient episodes 435 cell cytotoxin assay (CTA) positive 207 cytotoxigenic culture (CC) positive and negative 5,880 both CC and CTA negative 5,927 patients survived, 595 patients died Planche TD, Davies KA, Coen PG, et al. Lancet Infect Dis. 2013;13:936-45. 70

41% Increase Cytotoxin +ve (731, 5.9%) Cytotoxigenic culture +ve (1030, 8.3%) Toxin +ve NAAT +ve, Toxin -ve NAAT -ve, Toxin -ve WCC x 10 9 /L 12.4 a 9.9 a 10.0 Death by day 30 72/435 (16.6%) b 30/311 (9.7%) b 349/3943 (8.9%) 81% Increase (95% CI: 77-85) NAAT +ve (10.7%) a P < 0.0001 b P = 0.004 NAAT: nucleic acid amplification test Planche TD, Davies KA, Coen PG, et al. Lancet Infect Dis. 2013;13:936-45. 71

Impact of C. difficile Testing Method on Accuracy of Diagnosis 1321 stool samples were tested during 95,750 patient-days 85 hospital associated-cdi cases detected by PCR 56 cases by EIA/CCA (P = 0.01) Overall incidence rate was 8.9 per 10,000 patient-days (95% CI, 7.1-10.9) by PCR; 5.8 per 10,000 patient-days (95% CI, 4.4-7.4) by EIA/CCA P = 0.001; 53% higher by PCR Overall complication rate was 27% (23/85) when CDI was diagnosed by PCR 39% (22/56) by EIA/CCA (P = 0.16) Cases detected by PCR only were MORE THAN 10x less likely to develop a complication of CDI compared with cases detected by both PCR and EIA/CCA (3% vs. 39%; P < 0.001) Longtin Y, Trottier S, Brochu G, et al. Clin Infect Dis. 2013;56:67-73. 72

Outcomes by C. difficile Test Group Outcome C. difficile Positive C. difficile Negative Tox+/PCR+ (N = 131) Tox-/PCR+ (N = 162) C. difficile-related Complication or Death Within 30 d, No. (%) a P-value for significance across 3 groups. b Intensive care unit care, colectomy, or megacolon related to CDI. c All-cause mortality within 30 days was 10.7%, 14.2%, & 8.7%, respectively, for the 3 groups. Tox-/PCR- (N = 1123) P-Value a Complication b 10 (7.6) 0 3 (0.3) < 0.001 Death c 11 (8.4) 1 (0.6) 0 < 0.001 Complication or death 18 (13.7) 1 (0.6) 3 (0.3) < 0.001 Polage CR, Gyorke CE, Kennedy MA, et al. JAMA Intern Med. 2015;175:1792-801. 73

Incorrect CDI Diagnosis Is a Global Issue Incorrect diagnosis of Clostridium difficile infection in a university hospital in Japan. Mori N, Yoshizawa S, Saga T, et al. J Infect Chemother. 2015;21:718-22. Wilcox, et al. ECCMID 2016. P1341. U.S. CDI burden 2011: 453,000 286,300 (no NAAT); 701,000 (all NAAT) Lessa FC, Mu Y, Bamberg W, et al. N Engl J Med. 2015;372:825-34. 75

Investigation Summary: C. difficile Diagnosis Analysis was conducted to evaluate both the role of C. difficile diagnostic testing in defining the correct SER-109 Phase 2 study entry population, and in the proper diagnosis of C. difficile recurrences during the study. In the Phase 2 study, 81% of study subjects (72 of 89 subjects) were enrolled based on polymerase chain reaction (PCR) based testing for C. difficile, as well as clinical evaluation. An important and increasingly well-appreciated limitation of PCR testing is that while a positive result indicates that C. difficile cytotoxin genes are present, a positive PCR test does not necessarily indicate thatthe organism is viable and producing disease causing cytotoxins, nor that C. difficile is the source of clinical symptoms. 1 Two separate observations were made pertaining to the effects of discordant results from PCR and cytotoxin assay on the SER- 109 trial. The qualifying stool samples evaluated for Phase 2 study entry were not available for retesting for cytotoxin, however, the company was able to retest the samples associated with patients entering the open label extension trial for the presence of the C. difficile cytotoxin and determined that only 44% of samples (15 of 31 subjects) that tested positive by PCR testing also tested positive based on C. difficile cytotoxin assay. These results suggest that a substantial proportion of patients who entered the SER-109 Phase 2 study may have been C. difficile carriers and, therefore, C. difficile infection may not have been the source of the clinical symptoms. In addition, data from this analysis suggest that the use of PCR to measure C. difficile may have overestimated study recurrences in both treatment arms of the Phase 2 trial, further complicating interpretation of study results. This was shown by reanalysis of samples with cytotoxin assay, from patients diagnosed as recurrent in the Phase 2 study. In this retesting, between one quarter andone half of presumed study recurrences may not have been true C. difficile infections leading to pathology. http://ir.serestherapeutics.com/phoenix.zhtml?c=254006&p=irol-newsarticle&id=2240833 76

U.S. Clostridium difficile Infection Burden (2011) Lessa FC, Mu Y, Bamberg W, et al. N Engl J Med. 2015;372:825-34. 78

NHSN CDI Risk Adjusted SIR Calculation: Determining the Expected Number of CDI LabID Events Dudeck MA, Weiner LM, Malpiedi PJ, et al. Risk Adjustment for Healthcare Facility-Onset C. difficile and MRSA Bacteremia Laboratory-identified Event Reporting in NHSN. http://www.cdc.gov/nhsn/pdfs/mrsa-cdi/riskadjustment-mrsa-cdi.pdf. Accessed 5/16/17. 79

Two Stage Testing Best of Both Worlds Best for sensitivity GDH or NAAT Best for specificity toxin Most patients get ONE test Identify potential C. difficile excretors cross infection risk Reduce risk of patients being labelled as CDI when they are not Reduce unnecessary treatment of C. difficile colonised patients Reduce inflated CDI rates UK: 2007-12 toxin testing; 2012-17 2-stage testing; 80% fall in CDI rates 81

Size Matters The size of the population at risk of CDI substantial The size of the overlap between CDI and non-cdi cases substantial The size of the CDI diagnosis error rate substantial The size of the consequences of erroneous CDI testing clear & far-reaching The true size of the CDI challenge large challenge without inflation 82

The C. diff Debate: The Role of Diagnostics in Disease Determination Saturday Evening June 3, 2017 7:30 PM ASM Microbe 2017 Bissonet Meeting Room New Orleans Marriott

Q & A

This program was audio recorded and will be produced along with the slides into an enduring material which will be available within 3 weeks at CdiffDebate.com For those who were unable to attend this live program, they may apply for P.A.C.E. credits after viewing. Please send an email to info@medavera.com if you have questions or need assistance.