P1341 European Congress of Clinical Microbiology and Infectious Diseases (ECCMID) 2016 RAI Amsterdam, 1 2 3 4 4 4 4 4 4 Amsterdam, Wilcox MH ; Rahav G ; Dubberke E ; Gabryelski L ; Eves K ; Tipping R ; Guris D ; Kartsonis N ; Dorr MB The Netherlands Assessment of Effi cacy of Bezlotoxumab for Prevention of 1University of Leeds, Leeds, West Yorkshire, UK; 2Sheba Medical Center, Tel Hashomer, Israel; April 09–12, 2016 Clostridium diffi cile Recurrence by Diagnostic Test Method 3Washington University School of Medicine, St Louis, MO, USA; 4Merck & Co., Inc., Kenilworth, NJ, USA

Diagnostic Tests Table 1. Subject baseline characteristics (MODIFY I and II pooled data) • CDI recurrence rates were lower among subjects receiving bezlotoxumab compared Table 3. Diagnosis methods for the recurrent episodes of CDI (MODIFY I and II Introduction and Purpose with placebo (17% vs 27%; adjusted difference Ϫ10.0, 95% confi dence interval pooled data) • The baseline CDI episode was diagnosed based on results of stool tests performed Bezlotoxumab + SOC Placebo + SOC [–14.0,–6.0], p<0.0001) (Figure 3A) at the local laboratory Bezlotoxumab Placebo • It has been estimated that Clostridium diffi cile causes at least 172,000 each Number of subjects (%) Total • Among bezlotoxumab-treated subjects, CDI recurrence rates were higher when + SOC + SOC year across the 27 countries of the European Union among individuals у2 years of • Permitted diagnostic testing methods included N=335 age.1,2 However, these data do not take into account a recently documented 70% Subject characteristics (N=781) (N=773) diagnosis was made by PCR compared with EIA (Figure 3B) n=129 n=206 – Cell cytotoxicity assay Local laboratory Central laboratorya Number of subjects (%) increase in incidence of C. diffi cile infection (CDI) and considerable scope for missed Female 442 (56.6) 449 (58.1) • In contrast, for subjects receiving placebo, CDI recurrence rates were similar for 3 diagnoses across Europe – Culture with toxin detection or strain typing diagnoses using PCR or EIA Positive No result 15 (11.6) 40 (19.4) 55 (16.4) у65 years of age 390 (49.9) 405 (52.4) No result Positive 7 (5.4) 8 (3.9) 15 (4.5) • Although treatment of primary CDI is often successful, up to 35% of patients • Thus, the observed difference in CDI recurrence rates in the bezlotoxumab versus – Commercial toxin EIA kits Positive Negative 3 (2.3) 0 (0) 3 (0.9) experience CDI recurrence after completing initial antibiotic therapy.4,5 After fi rst у1 CDI episodes in past 6 months 216 (27.7) 219 (28.3) placebo groups was larger in subjects diagnosed by detection of free toxin via EIA recurrence, patients have a 45% probability of a second recurrence, with the risk Negative Positive 20 (15.5) 22 (10.7) 42 (12.5) – Commercial PCR assay kits Region of enrollmenta (N=781) (N=773) –12.8% (95% CI –18.5, –7.0) compared with those diagnosed by PCR –6.5% (95% CI increasing with further recurrences6 Positive Positive 84 (65.1) 136 (66.0) 220 (65.6) • All permitted commercial test kits had a labeled specifi city of у94% and had the –12.8, –0.3) North America 354 (45.3) 366 (47.3) Local laboratory test method 102 (100) 176 (100) 278 (100) • MODIFY I and MODIFY II were global, randomized, double-blind, placebo-controlled capacity to detect the presence of C. diffi cile toxin B (EIA kits) or its cognate tcdB • The relative reduction in recurrence for the free toxin (EIA) subgroup was 47% EIA 39 (38.2) 80 (45.5) 119 (42.8) trials of the effi cacy and safety of bezlotoxumab (a human gene (PCR kits) European Region 313 (40.1) 293 (37.9) compared with a 25% relative reduction for PCR (Figure 4) Cell cytotoxicity assay 1 (1.0) 1 (0.6) 2 (0.7) against C. diffi cile toxin B) alone and in combination with (a human • At the time of a new episode of diarrhea, stool samples were tested by either the local A s i a - P a c i fi c 79 (10.1) 77 (10.0) monoclonal antibody against C. diffi cile toxin A) for the prevention of CDI recurrence F igure 3. The proportion of subjects with CDI recurrence, (A) overall and Culture 5 (4.9) 12 (6.8) 17 (6.1) or central laboratory. The central laboratory diagnostic testing method was anaerobic Other 35 (4.5) 37 (4.8) (B) according to diagnostic test methodology at baseline (MODIFY I and II in adults receiving antibiotic treatment (standard of care; SOC) for primary or stool culture with toxin EIA testing on C. diffi cile isolates PCR 57 (55.9) 83 (47.2) 140 (50.4) 7 pooled data) recurrent CDI PCR ribotypeb (N=490) (N=486) aStool culture with toxin testing. b – In both trials, bezlotoxumab substantially reduced recurrent CDI and had a safety A B Denominators for this category are those subjects who had a positive test at the local laboratory. Outcome Measures 027, 078, or 244 strain 102 (20.8) 115 (23.7) CDI = C. diffi cile infection; EIA = enzyme immunoassay; PCR = polymerase chain reaction; SOC = standard of care. profi le similar to placebo. Furthermore, the addition of actoxumab did not improve e e • Effi cacy outcome measures for this post hoc analysis included clinical cure and EIA PCR e f fi c a c y . 7 Therefore, only the bezlotoxumab alone versus placebo data are 027 strain 89 (18.2) 100 (20.6) 30 30 • Only three stool samples were positive at the local laboratory and then found to be CDI recurrence 26.6 27.3 presented here aOther includes Argentina, Brazil, Chile, Colombia, Mexico, and South Africa. Asia-Pacifi c includes Australia, Japan, Korea, New Zealand and 26.1 negative at the central laboratory (two based on a toxin EIA test and one based on – Clinical cure: achieved when a subject received ୏14-day regimen of SOC therapy Taiwan. Europe includes Austria, Belgium, Czech Republic, Denmark, Finland, France, Germany, Israel, Italy, Poland, Portugal, the Russian 25 25 a PCR test); all three cases were in the bezlotoxumab group • In the MODIFY trials, several different test methods were used to diagnose CDI. Federation, Spain, Sweden, Switzerland, Turkey, and the United Kingdom. North America includes Canada and the United States. (spanning ୏16 calendar days) with no diarrhea (୏2 loose stools per 24 hours) for b Because diagnostic test methods for CDI vary in their predictive value for true CDI,8,9 Denominator is subjects in the FAS population with a positive culture. 19.6 • Of the 278 subjects across all treatment groups with CDI recurrence whose stool 2 consecutive days following completion of SOC therapy for the baseline CDI = C. diffi cile infection; FAS, full analysis set; PCR = polymerase chain reaction; SOC = standard of care. 20 20 the potential impact of test method, toxin detection by enzyme immunoassay (EIA) sample at the time of a new episode of diarrhea was positive at the local laboratory, CDI episode 16.5 versus toxin gene detection by polymerase chain reaction (PCR), on study outcomes 15 15 14.5 50.4% were tested using PCR, and 42.8% were tested with a toxin EIA method was examined – CDI recurrence: the development of a new episode of diarrhea associated with a Table 2. Type of local laboratory test used for baseline CDI diagnosis according positive stool test for toxigenic C. diffi cile at the local or central laboratory following to treatment group (MODIFY I and II pooled data) • A numerically higher proportion of recurrences in the bezlotoxumab group were 10 10 clinical cure of the baseline CDI episode through 12 weeks following infusion of diagnosed at the local laboratory with a PCR test (55.9%) compared with the placebo Bezlotoxumab Placebo Overall (47.2%) group Objective study medication + SOC + SOC 5 5 I and II – % of subjects with CDI recurrenc I and II – % of subjects with CDI recurrenc n=781 n=773 N=1554 n= 372 357 385 337 • To examine the impact of CDI diagnostic testing methodology on effi cacy outcomes in Data Analysis 0 0 Conclusions subjects receiving bezlotoxumab or placebo in the MODIFY I and II trials Number of subjects (%) Bezlotoxumab + Placebo + Bezlotoxumab + Placebo + • Using the pooled effi cacy analysis population (full analysis set) for the bezlotoxumab MODIFY SOC (N=781) SOC (N=773) MODIFY SOC (N=729) SOC (N=722) • The toxin B-selective monoclonal antibody, bezlotoxumab, was associated and placebo treatment groups: EIA 372 (47.6) 385 (49.8) 757 (48.7) Treatment Treatment with clinically meaningful reductions in rates of CDI recurrence, regardless Cell cytotoxicity assay 10 (1.3) 6 (0.8) 16 (1.1) Culture and cell cytotoxicity assay data not shown, Methods – Effi cacy outcomes (clinical cure and CDI recurrence) were estimated based bezlotoxumab + SOC n=52, placebo + SOC n=51. of the diagnosis test method used on the diagnostic test method used to confi rm the diagnosis of CDI for the Culture 42 (5.4) 45 (5.8) 87 (5.6) CDI = C. difficile infection; EIA = enzyme immunoassay PCR = polymerase chain reaction; SOC = standard of care. • The magnitude of reduction of CDI recurrence associated with Study Design baseline episode PCR 357 (45.7) 337 (43.6) 694 (44.7) bezlotoxumab was higher if toxin detection rather than PCR was used to • MODIFY I (MK-3415A-001, NCT01241552) and II (MK-3415A-002, NCT01513239) diagnose CDI – For the subgroup of subjects with an outcome of CDI recurrence during the Figure 4. Relative reduction in recurrent CDI rate for bezlotoxumab versus were randomized, double-blind, placebo-controlled, multicenter, Phase 3 studies CDI = C. diffi cile infection; EIA = enzyme immunoassay; PCR = polymerase chain reaction; SOC = standard of care. follow-up period, the test method and location of the laboratory (local or central) placebo according to diagnosis method for baseline CDI episode (MODIFY I and • These fi ndings are consistent with recent data confi rming the importance that performed the diagnostic test are summarized of the detection of fecal toxin for the diagnosis of CDI, including recurrent • The design of both studies is summarized in Figure 1 Figure 2. The proportion of subjects achieving clinical cure, (A) overall and II pooled data) CDI8,9 (B) according to diagnostic test methodology at baseline (MODIFY I and II 50 47 Figure 1. Trial design fl ow diagram pooled data)

Results s References A B 40 38 1. Barbut F, et al. 2013. Clostridium Diffi cile infection in Europe: a CDI Europe report. http://www.multivu.com/assets/60637/ EIA PCR documents/60637-CDI-HCP-Report-original.pdf. Baseline CDI diagnosis 100 100 Study Population 2. Bauer MP, et al. Lancet. 2011;377(9759):63–73. CDI episode Diarrhea + Toxigenic C. difficile 3. Davies KA, et al. Lancet Infect Dis 2014;14(12):1208–1219. • A total of 1554 subjects were included in the full analysis set for the pooled analysis. 82.3 83.1 30 80.0 80.3 79.0 77.4 25 4. Aslam S, et al. Lancet Infect Dis. 2005;5(9):549–557. Baseline demographics and characteristics were similar in the bezlotoxumab (n=781) 80 80 5. Kelly CP, LaMont JT. N Engl J Med 2008;359:1932–1940. and placebo (n=773) pooled analysis groups (Table 1) 6. Surawicz CM. Nat Clin Pract Gastroenterol Hepatol. 2004;1(1):32–38. 20 7. Wilcox M, et al. Open Forum Infect Dis. 2015;2 (Suppl 1):S3. Antibiotic treatment • Oral metronidazole 60 60 • Oral vancomycin ± IV metronidazole 8. Polage CR, et al. JAMA Intern Med. 2015;175(11):1792–1801. 10–14-day regimen Diagnostic Tests Used for Baseline Episode 9. Planche T, Wilcox MH. Infect Dis Clin North Am. 2015;29(1):63–82. • Oral fidaxomicin ± IV metronidazole (bezlotoxumab vs placebo) 10 • An EIA kit was the most commonly used diagnostic test at baseline, followed by PCR, 40 40 % Relative reduction in rCDI rate Acknowledgments culture with toxin detection or strain typing, and cell cytotoxicity assay (Table 2) Medical writing and/or editorial assistance was provided by Edward Rochford, PhD, Complete Medical Communications, Macclesfi eld, UK. This assistance was funded by Merck & Co., Inc., Kenilworth, NJ, USA. Study period • Overall, the bezlotoxumab and placebo groups were balanced in the proportion of 0 20 20 All EIA PCR subjects diagnosed with the different test methods I and II – % of subjects with clinical cure I and II – % of subjects with clinical cure Detection method • Actoxumab (MODIFY I only) Disclosures Day 1: Day 85 n= 372 357 385 337 CDI = C. difficile infection; EIA = enzyme immunoassay; PCR = polymerase chain reaction; rCDI = recurrent CDI. Funding for this research was provided by Merck & Co., Inc., Kenilworth, NJ, USA. • Bezlotoxumab 0 0 Single IV Endpoint Estimates by Test Method MODIFY MODIFY MHW has received consulting fees from Actelion, Astellas, bioMerieux, MedImmune, Merck, Pfi zer, Sanofi -Pasteur, Seres, • Actoxumab + Bezlotoxumab Bezlotoxumab + Placebo + Bezlotoxumab + Placebo + infusion Summit, Synthetic Biologics, and Valneva; lecture fees from Alere, Astellas, Merck, and Pfi zer; and grant support from Overall, clinical cure rates were broadly similar between the bezlotoxumab (80.0%) SOC (N=781) SOC (N=773) SOC (N=729) SOC (N=722) Actelion, Astellas, bioMerieux, Da Volterra, Merck, and Summit. • Placebo (0.9% normal saline) • Diagnosis Method for Recurrent Episodes and placebo groups (80.3%) (Figure 2A) Treatment Treatment LG, KE, RT, DG, NK, and MBD are current employees of Merck Sharp & Dohme Corp., a subsidiary of Merck & Co., Culture and cell cytotoxicity assay data not shown, • Across both treatment groups, the majority of subjects with a CDI recurrence were Inc., Kenilworth, NJ, USA, and may own stock and/or stock options. bezlotoxumab + SOC n=52, placebo + SOC n=51. Record loose stools daily; test for toxigenic C. difficile if diarrhea returned • Within each treatment group, clinical cure rates were slightly lower among subjects confi rmed positive for toxigenic C. diffi cile at both the central and local laboratories GR is a consultant and scientifi c advisor to Merck, Pfi zer, and AstraZeneca. whose diagnosis was based on PCR compared with toxin EIA (Figure 2B) EIA = enzyme immunoassay; PCR = polymerase chain reaction; SOC = standard of care. ED has conducted research for Merck & Co., Inc., Rebiotix, and Sanofi -Pasteur and is or has been a consultant to Merck & (220/335, 65.7%, Table 3) Co., Inc., Rebiotix, Sanofi -Pasteur, Summitt, and Daiichi.

Copyright © 2016 Merck Sharp & Dohme Corp., a subsidiary of Merck & Co., Inc. All rights reserved.