JCM Accepted Manuscript Posted Online 27 September 2017 J. Clin. Microbiol. doi:10.1128/JCM.01135-17 Copyright © 2017 American Society for Microbiology. All Rights Reserved.

Evaluation of the cobas® Cdiff test for the Detection of

2

Toxigenic Clostridium difficile in Stool Samples

3

Revision 1

4

Lance R. Peterson, MD,a,# Stephen A. Young, PhD,b,c Thomas E. Davis, Jr, MD, PhD,d

5

Zi-Xuam Wang, PhD,e John Duncan, MS,f Christopher Noutsios, MS,f

6

Oliver Liesenfeld, MD,f John C. Osiecki, PhD,f and Michael A. Lewinski, PhDf

7

a

8

Evanston, IL 60201

9

b

Department of Pathology, University of New Mexico HSC, Albuquerque, NM 87131

10

c

TriCore Reference Laboratories, Albuquerque, NM 87102

11

d

Department of Pathology and Laboratory Medicine, Indiana University School of Medicine,

12

Indianapolis, IN 46202

13

e

Thomas Jefferson University, Philadelphia, PA 19107

14

f

Medical and Scientific Affairs, Roche Molecular Diagnostics, Pleasanton, CA 94588

15

This research project was funded by Roche Molecular Diagnostics, Pleasanton, CA

16

Running title: cobas® Cdiff Test for toxigenic Clostridium difficile

17

#Correspondent footnote:

18

Lance R. Peterson, M.D.

19

Department of Pathology and Laboratory Medicine

20

NorthShore University HealthSystem

21

2650 Ridge Avenue

22

Evanston, IL 60201, U.S.A.

23

Phone: 847-570-1637

24

Fax: 847-733-5093

Department of Laboratory Medicine and Pathology, NorthShore University Health System,

E-mail: [email protected]

25 1

Downloaded from http://jcm.asm.org/ on September 29, 2017 by UNIV OF ESSEX

1

ABSTRACT

27

Nucleic acid amplification tests are reliable tools for the detection of toxigenic Clostridium

28

difficile from unformed (liquid or soft) stool samples. The objective of this study was to evaluate

29

performance of the cobas® Cdiff Test on the cobas® 4800 System using prospectively collected

30

stool specimens from patients suspected of C. difficile infection (CDI). Performance of the

31

cobas® Cdiff Test was compared to the combined results of direct and broth enriched toxigenic

32

culture in a large, multicenter clinical trial. Additional discrepant analysis was performed using

33

the Xpert® C. difficile Epi Test. Sample storage was evaluated on contrived and fresh samples

34

before and after storage at -20°C. Testing was from 683 subjects (306 males and 377 females);

35

113 (16.5%) of 683 were positive for toxigenic C. difficile by direct toxigenic culture and 141 of

36

682 were positive using the combined direct and enriched toxigenic culture (reference method),

37

for a prevalence rate of 20.7%. The sensitivity and specificity of the cobas® Cdiff Test

38

compared to combined direct and enriched culture was 92.9% (131/141; 95% CI: 87.4% to

39

96.1%) and 98.7% (534/541; 95% CI: 97.4% to 99.4%), respectively. Discrepancy analysis

40

using a retested sample results from a second NAAT (Xpert® C. difficile/Epi test (Cepheid,

41

Sunnyvale, CA) found no false negative and 4 false positive cobas® Cdiff Test results. There

42

was no difference in positive and negative results when comparing fresh and stored samples.

43

These results support the use of the cobas® Cdiff Test as a robust aid in the diagnosis of CDI.

44 45

Key Words. Clostridium difficile infection (CDI), Nucleic acid amplification test, Healthcare-

46

Associated Infection, toxin B (tcdB) gene, active surveillance testing (AST)

47

2

Downloaded from http://jcm.asm.org/ on September 29, 2017 by UNIV OF ESSEX

26

INTRODUCTION

49

Clostridium difficile is a Gram-positive, anaerobic, spore-forming bacillus identified as an

50

etiologic agent of antibiotic-associated diarrhea and pseudomembranous colitis in the late

51

1970s (1-2). Historically, the incidence of C. difficile infection (CDI) ranged from 30-40 cases

52

per 100,000 population in acute care hospitals in the United States (3). The incidence rose to

53

more than 140 cases per 100,000 by 2013, with 60.7% being clearly healthcare associated (4).

54

Currently CDI is the most commonly diagnosed infectious cause for healthcare-associated

55

diarrhea and is a significant cause of morbidity and mortality (5). Toxigenic strains of C. difficile

56

typically produce two toxins, toxin A that is an enterotoxin and toxin B that is a cytotoxin (6). A

57

small percentage of strains only produce toxin B (7). There are no known naturally occurring

58

toxin A positive, toxin B negative strains associated with clinical disease (8-9). Diagnosis of CDI

59

may be established by the presence of toxin B, or toxins A and B in stool samples from patients

60

with appropriate clinical symptoms, driving the need for accurate testing.

61

Early detection of CDI with rapid implementation of contact precautions is considered

62

critical for infection control to prevent transmission in the healthcare setting (10). The diagnostic

63

approach for CDI is varied (11), and real-time PCR (qPCR) is considered a reliable, rapid, single

64

test for detection of C. difficile toxin (12-15). The purpose of this study was to compare

65

performance of the cobas® Cdiff Test to direct and enriched toxigenic culture for detection of

66

toxigenic C. difficile in stool samples.

67 68

MATERIALS AND METHODS

69

Study Population

70

Stool samples were collected from 683 eligible patients over a 5 month period at 4 sites in the

71

United States (NorthShore University HealthSystem, Evanston, IL; the hospitals using TriCore

72

Reference Laboratories located in Albuquerque, NM; Indiana University School of Medicine,

73

Indianapolis, IN; and Thomas Jefferson University, Philadelphia, PA). Patients older than 12

3

Downloaded from http://jcm.asm.org/ on September 29, 2017 by UNIV OF ESSEX

48

months of age being tested for the diagnosis of CDI were identified and enrolled in the study.

75

Each patient entering the study had an unformed stool sample greater than 5 mLs submitted for

76

testing. Patients were excluded from the study if they (i) submitted a formed stool sample, (ii)

77

had been administered antibiotic therapy against C. difficile (e.g. oral and/or parenteral

78

metronidazole, oral vancomycin, or fidaxomicin) during their present hospitalization, (iii) were

79

previously enrolled in this study, (iv) were enrolled within the previous 12 months in any other

80

study focused on the prevention or treatment of CDI, or (v) had a contraindication to collection

81

of stool samples according to the institution policies and procedures.

82 83

Clinical Trial Design

84

The clinical performance of the cobas® Cdiff Test was evaluated in a prospective, multi-site,

85

investigation comparing the results with toxigenic culture using remnant, de-identified, unformed

86

stool samples from eligible subjects. Toxigenic culture was performed at a single site

87

designated as the reference laboratory (NorthShore University HealthSystem). The cobas®

88

Cdiff Test was performed at the other three sites, following the manufacturer’s instructions.

89

This study was conducted in compliance with International Conference on

90

Harmonization (ICH), Good Clinical Practice (GCP) Guidelines, and regulations of the US Food

91

and Drug Administration (FDA). Approval by the institutional review board at all four sites was

92

obtained. This study was performed to obtain clearance of the cobas® Cdiff Test by the FDA.

93 94

Stool Sample Processing

95

On the day of sample collection, fresh stool specimens were transported to the collection site

96

laboratory at temperatures between 2°C and 30°C. Upon receipt, fresh stool samples were

97

processed within 48 hours of collection. The accepted liquid and unformed samples of fresh

98

stool were stirred to homogenize, and the aliquots were then prepared by transferring

99

approximately 1.0 mL portions to 4 separate, sterile 5 mL cryogenic vials as depicted in Figure

4

Downloaded from http://jcm.asm.org/ on September 29, 2017 by UNIV OF ESSEX

74

100

1. Aliquots 1 and 2 were shipped on cold packs to one of the designated test sites for C. difficile

101

PCR testing within 5 days of sample collection. Aliquots 3 and 4 were shipped on ice to the

102

reference laboratory to arrive within the same time frame.

104

Toxigenic Culture

105

The toxigenic culture included an initial and repeat direct and enrichment culture of stool

106

followed by cytotoxicity testing. The direct culture was done by inoculating a portion of the

107

sample to pre-reduced anaerobically sterilized (PRAS) selective media, cycloserine-cefoxitin-

108

fructose agar with horse blood and taurocholate (CCFA-HT; Anaerobe Systems, Morgan Hill,

109

CA), followed by cytotoxicity testing on recovered C. difficile. Briefly, after incubation at 35oC,

110

suspected colonies were identified as C. difficile by Gram stain, aerotolerance testing, and by

111

the Pro Disk test (Hardy Diagnostics, Santa Maria, CA), and then inoculated into anaerobic

112

chopped meat broth (Anaerobe Systems) that was incubated for 5 to 7 days at 35°C for

113

cytotoxicity testing. Supernatant from the anaerobic chopped meat broth was processed for

114

detection of C. difficile toxin B using cell culture cytotoxicity testing according to the

115

manufacturer instructions (C. DIFFICILE TOX-B TEST; TECHLAB®, Blacksburg, VA). Enriched

116

toxigenic culture was done simultaneously and used anaerobic cycloserine-cefoxitin-mannitol

117

broth with taurocholate, lysozyme and cysteine (CCMB-TAL; Anaerobe Systems), followed by

118

subculture on anaerobic Brucella agar plates (Anaerobe Systems), with identification and

119

cytotoxicity testing of all recovered C. difficile.

120

A specimen was considered positive for toxigenic C. difficile if a toxin producing

121

organism was recovered from stool by direct or enriched toxigenic culture (any positive rule). If

122

C. difficile was isolated from the direct culture and the isolate tested positive by cell cytotoxicity

123

assay, the enrichment culture was not further analyzed. Using aliquot 4 specimens that were

124

frozen at -70oC for up to 5 months, 580 study samples were retested with direct and enriched

125

culture. If a stool specimen was positive by the cobas® Cdiff test but negative by the initial

5

Downloaded from http://jcm.asm.org/ on September 29, 2017 by UNIV OF ESSEX

103

culture then repeat culture was done (n = 29). Also, all other samples with a negative initial

127

direct culture (n = 541) had a repeat direct and enriched culture performed. Additionally, 10

128

specimens with a positive initial direct culture were tested as positive controls. Specimens were

129

classified as negative for toxigenic C. difficile only if they tested negative by both initial plus

130

repeat direct and enrichment culture.

131 132

Testing of patient samples on the cobas® 4800 System

133

After receipt of patient specimens in the laboratory, test orders were logged into the cobas®

134

4800 System either manually, with a work list, or via a Laboratory Information System (LIS)

135

interface and then tested. The cobas® 4800 System is a walk-away platform performing nucleic

136

acid purification, real-time PCR (qPCR) set-up into a 96 well plate, followed by manual sealing

137

and transfer of the plate to the amplification and detection system (16). The system currently is

138

available for MRSA/SA (17), C. difficile (current data from this report), herpes simplex virus 1

139

and 2 (18-19), CT/NG (Chlamydia trachomatis/Neisseria gonorrhoeae) testing (20-25), high risk

140

human papilloma virus (26-28) and BRAF (29), EGFR (30), and KRAS (31) mutation detection.

141

The cobas® MRSA/SA Test, cobas® Cdiff Test and cobas® HSV 1 and 2 Test can be performed

142

simultaneously in a single testing run. This system uses the cobas® x 480 Instrument that

143

includes an automated pipetter to extract, purify, and prepare target nucleic acid for qPCR (e.g.,

144

automated specimen preparation). Bacterial lysis of the stool specimen is achieved with

145

proteinase K, detergent, and a chaotropic agent. The purified DNA is bound to magnetic glass

146

particles (MGPs), washed, and eluted. The cobas® x 480 Instrument next transfers the working

147

master mix reagent and moves the specimen into the qPCR microwell plate. The plate is then

148

transferred to the cobas® z 480 instrument where amplification and detection of target DNA and

149

controls occurs using specific primers and probes. Data analysis and report generation are

150

performed by the cobas® 4800 Software. The cobas Cdiff Test uses positive (+) and negative

151

(-) controls (cobas 4800 Cdiff Controls and Cofactor Kit), and an internal control (cobas 4800

6

Downloaded from http://jcm.asm.org/ on September 29, 2017 by UNIV OF ESSEX

126

152

System Internal Control Kit). One set of positive and negative controls are included in each run.

153

Valid results must be obtained for both the positive and negative control for the cobas 4800

154

Software to display a reportable result.

156

Discrepant Analysis

157

Discrepant analysis using the Xpert® C. difficile/Epi test (Cepheid, Sunnyvale, CA) was

158

performed on all samples with discordant results between the cobas® Cdiff Test and combined

159

direct and enrichment culture. Eight concordant positive and eight concordant negatives were

160

also tested as controls. The results of the Xpert® C. difficile/Epi test determined the final

161

reference result for the discrepancy analysis.

162 163

Testing Fresh and Archived (Frozen) Stool Specimens

164

To evaluate the effect of specimen storage on toxigenic culture and cobas® Cdiff Test results,

165

aliquots from contrived specimens and unformed stool prospectively collected from patients

166

suspected of CDI were tested fresh and following storage at -20°C or colder. Briefly, four studies

167

were performed; specifically, (1) toxigenic culture was performed on fresh samples at the

168

reference laboratory before shipping the samples frozen on dry ice to Roche Molecular

169

Systems, Inc., (Pleasanton, CA) for PCR, and following PCR, a subset was returned for repeat

170

culture; (2) a panel of 100 contrived specimens was prepared using fresh negative stool

171

specimens spiked with toxigenic C. difficile at varying densities (2.5, 5.0, 10, and 20-fold the

172

assay Limit of Detection) and tested fresh plus following storage for 32 days at -20±5°C; (3) the

173

assay also was run on the 580 prospectively collected samples both fresh and following storage

174

at -70°C for up to 5 months; and (4) repeat toxigenic culture was performed on a subset of

175

clinical samples (15 culture-positive and 72 culture-negative) stored at -20°C or colder for ≥60

176

days.

177 7

Downloaded from http://jcm.asm.org/ on September 29, 2017 by UNIV OF ESSEX

155

Analytical performance

179

The technical performance verification (TPV) study was conducted using a panel composed of

180

C. difficile cultures diluted into negative stool matrix in cobas® PCR Media to a density below the

181

Limit of Detection (LOD), near LOD, above LOD as well as a negative level composed of only

182

the stool suspension in cobas® PCR media. We used three unique lots of cobas® Cdiff Test

183

reagents and three instruments for a total of 36 runs over 12 days.

184

The reproducibility of the cobas® 4800 System was established in a multi-site

185

investigation using simulated clinical samples evaluated across lot, site/instrument, operator,

186

day and within-run. Reproducibility test panels consisted of four specimens, with three

187

replicates each, using varying densities of C. difficile strain ATCC 43255 (Negative, Below LOD,

188

1 x LOD, and 3 x LOD) into pooled, C. difficile-negative, unformed stool matrix in cobas® PCR

189

media. These were tested at three sites by two operators/day (six total operators) for five days

190

per lot over two lots for an overall total of 712 valid tests. We modeled the target cycle

191

threshold (Ct) as the dependent variable using random effects lot, site, operator, and day.

192

The LOD was determined by analyzing quantified cultures of seven toxigenic C. difficile

193

strains diluted into pooled negative stool specimen matrix in cobas® PCR media. All densities

194

were analyzed using three unique reagent lots. At least 21 replicates per reagent lot were

195

tested at each density. To assess inclusivity, the cobas® Cdiff Test was evaluated with 28

196

additional strains of toxigenic C. difficile tested at a minimum of three densities and 40 replicates

197

per density level. The lowest density that had at least a 95% hit (positive detection) rate was

198

considered the LOD for each strain.

199

Analytical specificity was evaluated with a panel that consisted of 103 bacteria, fungi,

200

and viruses as well as one human tissue cell line (32). In addition, 28 Clostridium genus

201

organisms, including non-toxigenic C. difficile, were tested (32). Bacteria were quantified as

202

colony forming units (CFU)/mL, human cells were quantified as cells/mL, and viruses were

203

quantified as plaque forming units (PFU)/mL. The only exceptions were for the following

8

Downloaded from http://jcm.asm.org/ on September 29, 2017 by UNIV OF ESSEX

178

microorganisms: Chlamydia trachomatis was quantified as elementary bodies (EB)/mL, and

205

Cytomegalovirus, Human Echovirus, and Human Enterovirus were quantified as IU/mL.

206

Bacteria and human cells were spiked to 1 x 106 CFU/mL or cells/mL, respectively, and all

207

viruses were spiked to 1 x 105 PFU/mL, except for Adenovirus Type 40, Cytomegalovirus

208

(HHV5), and Human Rotavirus that were spiked to lower densities due to stock density

209

limitations. Testing was performed with the organisms alone or with two C. difficile isolates

210

present individually at 3 x LOD.

211

Interference was evaluated for 26 medications, fecal fat, whole blood, and mucin. The

212

effects of the potentially interfering substances were evaluated in the presence and absence of

213

two toxigenic C. difficile isolates spiked to approximately 3 x LOD of the cobas® Cdiff Test (54

214

CFU/mL and 113 CFU/mL for the two strains).

215

Potential cross-contamination was assessed by testing a series of high positive toxigenic

216

C. difficile samples and negative samples in a checkerboard configuration on the cobas® 4800

217

system. High positive samples were prepared by spiking pooled negative stool matrix sufficient

218

to generate a Ct value (i.e., the number of PCR cycles until the PCR growth curve was

219

classified as positive) of the 95th percentile of the clinical specimen population. A total of 94

220

samples were included per run, with nine runs performed using three cobas® 4800 systems

221

having alternating checkerboard configurations. A total of three runs of negative samples

222

following these checkerboard runs were performed to assess carry-over contamination.

223 224

Statistical Analysis

225

SAS/STAT® software (Cary, NC) was used to perform all analyses. The statistical analyses of

226

clinical samples were chosen based on recommendations in the statistical guidance from the

227

FDA (33-34), and in accordance with the guidelines published in Clinical Laboratory Standards

228

Institute EP12-A2 for evaluating qualitative test performance (35). For the clinical reproducibility

229

study, 95% two-sided exact binomial confidence intervals were used. In the studies designed to

9

Downloaded from http://jcm.asm.org/ on September 29, 2017 by UNIV OF ESSEX

204

230

examine the difference between results from fresh and frozen samples, the statistical analysis

231

used the two-sided Fisher’s exact test and the likelihood ratio of the chi-square test.

232 RESULTS

234

Study Population

235

The clinical utility study collected 683 specimens across 4 collection sites and all specimens had

236

valid test results from the cobas® Cdiff Test as well as the direct culture from initial testing. One

237

sample lacked sufficient volume for repeat direct and enrichment culture and was not included

238

in the final statistical analysis. Demographics and baseline characteristics of evaluable subjects

239

are shown in Table 1. The age range of the patients was between 3 and 99 with a median age

240

of 59 years. The gender distribution was 306 males (44.8%) and 377 females (55.2%).

241 242

Comparison of the cobas® Cdiff Test with Direct Culture

243

The performance of the cobas® Cdiff Test compared to direct culture is shown in Table 2. The

244

overall percent agreement (OPA) was 95.3%. The three specimens with false-negative cobas®

245

Cdiff Test results relative to direct culture were all negative in the Xpert® Cdiff test. Of the 29

246

specimens with false-positive cobas® Cdiff Test results relative to direct culture, 15 were positive

247

in the Xpert® Cdiff test; one sample was not tested because of insufficient specimen volume.

248

The clinical performance of the cobas® Cdiff Test compared with the combined results of

249

initial and repeat direct and enriched toxigenic culture are shown in Table 3. This data

250

demonstrates the benefit of using enrichment culture to enhance the recovery of C. difficile

251

when using culture as a reference method. The sensitivity and specificity of the cobas® Cdiff

252

Test was 92.9 and 98.7%, respectively; the positive predictive value (PPV) and negative

253

predictive value (NPV) was 94.9 and 98.2% respectively. In the discrepancy analysis, of the 7

254

specimens with false-positive cobas® Cdiff Test results relative to combined direct and

255

enrichment culture, 3 were positive in the Xpert Cdiff test. The 10 specimens with a false-

10

Downloaded from http://jcm.asm.org/ on September 29, 2017 by UNIV OF ESSEX

233

256

negative cobas® Cdiff Test were all negative by Xpert® C. difficile Epi Test.

257 Impact of sample storage

259

The results of 124 random samples returned to the reference laboratory for retesting had a

260

positive percent agreement, negative percent agreement and the overall percent agreement

261

between fresh or frozen stool specimens of 97.1% (33/34), 97.8% (88/90), and 97.6%

262

(121/124), respectively (chi-square statistic = 0.02; P = 0.89). For the 100 contrived specimens,

263

the positive percent agreement between the cobas® Cdiff Test results before and after one

264

freeze-thaw cycle was 100.0% (100/100; 95% CI = 96.4–100.0%). There was no difference in

265

the distribution of positive and negative cobas® Cdiff Test results when comparing fresh and

266

frozen specimens from the multicenter trial of 580 samples (Fisher’s exact test P = 1; chi-square

267

test likelihood ratio P = 0.93). Finally, there was 100% agreement between direct and repeat

268

toxigenic culture results from 87 samples cultured fresh and following storage for at least 60

269

days at -20°C or colder (Kappa statistic: 1.0).

270 271

Analytical performance

272

Analysis of the precision study data found that at

Evaluation of the cobas® Cdiff test for the Detection of Toxigenic Clostridium difficile in Stool Samples.

Nucleic acid amplification tests are reliable tools for the detection of toxigenic Clostridium difficile from unformed (liquid or soft) stool samples...
682KB Sizes 0 Downloads 16 Views

Recommend Documents