INH resistance assay

    Direct Detection of Mycobacterium tuberculosis And Drug Resistance In Respiratory Specimen Using Abbott Realtime MTB Detection and RIF/INH Resistance Assay Kingsley King-Gee Tam, Kenneth Siu-Sing Leung, Sabrina Wai-Chi To, Gilman Kit-Hang Siu, Terrence Chi-Kong Lau, Victor Chi-Man Shek, Cindy Wing-Sze Tse, Samson Sai-Yin Wong, Pak-Leung Ho, Wing-Cheong Yam PII: DOI: Reference:

S0732-8893(17)30209-2 doi: 10.1016/j.diagmicrobio.2017.06.018 DMB 14379

To appear in:

Diagnostic Microbiology and Infectious Disease

Received date: Revised date: Accepted date:

17 February 2017 19 June 2017 20 June 2017

Please cite this article as: Tam Kingsley King-Gee, Leung Kenneth Siu-Sing, To Sabrina Wai-Chi, Siu Gilman Kit-Hang, Lau Terrence Chi-Kong, Shek Victor Chi-Man, Tse Cindy Wing-Sze, Wong Samson Sai-Yin, Ho Pak-Leung, Yam Wing-Cheong, Direct Detection of Mycobacterium tuberculosis And Drug Resistance In Respiratory Specimen Using Abbott Realtime MTB Detection and RIF/INH Resistance Assay, Diagnostic Microbiology and Infectious Disease (2017), doi: 10.1016/j.diagmicrobio.2017.06.018

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Direct Detection Of Mycobacterium tuberculosis And Drug Resistance In Respiratory

T

Specimen Using Abbott Realtime MTB Detection and RIF/INH Resistance Assay

RI P

Running Title: Detection of MDR-TB by Abbott Realtime MTB and RIF/INH

SC

Kingsley King-Gee Tam1#, Kenneth Siu-Sing Leung1#, Sabrina Wai-Chi To1, Gilman Kit-Hang Siu2, Terrence Chi-Kong Lau3, Victor Chi-Man Shek4, Cindy Wing-Sze

MA NU

Tse4, Samson Sai-Yin Wong1, Pak-Leung Ho1 and Wing-Cheong Yam1* Authors Affiliations: 1

Department of Microbiology, Queen Mary Hospital, The University of Hong Kong,

Department of Health Technology and Informatics, The Hong Kong Polytechnic

PT

2

ED

Hong Kong Special Administrative Region, China

Department of Biomedical Sciences, City University of Hong Kong, Hong Kong

AC

3

CE

University, Hong Kong Special Administrative Region, China

Special Administrative Region, China 4

Department of Pathology, Kwong Wah Hospital, Hong Kong Special Administrative

Region, China #

These authors contributed equally to this work

*

Corresponding author. Department of Microbiology, Queen Mary Hospital, The

University of Hong Kong, Hong Kong Special Administrative Region, China. Tel: +852-22554821; Fax: +852-28551241; E-mail: [email protected]

1

ACCEPTED MANUSCRIPT ABSTRACT

T

Abbott RealTime MTB (Abbott-RT) in conjunction with Abbott RealTime MTB

RI P

RIF/INH Resistance (Abbott-RIF/INH) is a new, high-throughput automated nucleic

SC

acid amplification platform (Abbott-MDR) for detection of Mycobacterium tuberculosis complex (MTBC) and the genotypic markers for rifampicin (RIF) and

MA NU

isoniazid (INH) resistance directly from respiratory specimens. This prospective study evaluated the diagnostic performance of this new platform for MTBC and multidrug-resistant tuberculosis (MDR-TB) using 610 sputum specimens in a

ED

tuberculosis high-burden setting. Using conventional culture results and clinical

PT

background as reference standards, Abbott-RT exhibited an overall sensitivity and

CE

specificity of 95.2% and 99.8%, respectively. Genotypic RIF/INH resistance of 178

AC

“MTB detected” specimens were subsequently analysed by Abbott-RIF/INH. Compared to phenotypic drug susceptibility test results, Abbott-RIF/INH detected resistance genotypic markers in 84.6% MDR-TB, 80% mono-RIF-resistant and 66.7% mono-INH-resistant specimens. Two of the RIF-resistant specimens carried a novel single, nonsense mutation at rpoB Q513 and in silico simulation demonstrated that the truncated RpoB protein failed to bind with other subunits for transcription. Overall, Abbott-MDR platform provided high throughput and reliable diagnosis of MDR-TB within a TB high-burden region.

2

ACCEPTED MANUSCRIPT 1 INTRODUCTION

T

Mycobacterium tuberculosis complex (MTBC), the aetiological agent of tuberculosis

RI P

(TB), has been a major public health problem for decades. According to World Health

SC

Organization (WHO), 10.4 million new cases and 1.8 million TB-related deaths were estimated in 2015 (WHO, 2016). The global effort of TB control was further

MA NU

hampered by the emergence of multidrug-resistant TB (MDR-TB) and extensively drug-resistant TB (XDR-TB). A nationwide survey on drug-resistant MTB infection in China showed that about 5.7% new cases and 25% retreatment cases were

ED

MDR-TB, which is nearly twice the global occurrence rate (Zhao et al., 2012). The

PT

emergence of MDR-TB has given rise to the need for a fast and robust routine drug

CE

susceptibility testing (DST) service in TB endemic regions for early disease

AC

management. However, phenotypic DSTs are usually time-consuming due to the long turnaround time for mycobacterial cultures (Outhred et al., 2015). Genotypic DSTs, such as GeneXpert® MTB/RIF or the recently launched Xpert® MTB/RIF Ultra, allow early detection of rifampicin resistance (Ioannidis et al., 2011, WHO, 2017), but their low throughput was unable to meet the demands from high TB endemic regions (Enjeti et al., 2015, Weyer et al., 2013).

The latest Abbott-MDR platform is a high throughput and fully automatic PCR-based

3

ACCEPTED MANUSCRIPT MTBC diagnostic system which consists of Abbott Realtime MTB (Abbott-RT) and

T

Abbott Realtime MTB RIF/INH Resistance assay (Abbott-RIF/INH). With the

RI P

flexibility of manipulating 24 to 96 specimens at a time, Abbott-RT can accommodate

SC

medium- to large-scale sample processing in daily clinical practice. Genotypic rifampicin (RIFR) and isoniazid resistance (INHR) markers among selected MTBC

MA NU

specimens can be identified by Abbott-RIF/INH in the reflex mode (Kostera et al., 2016). Recently, Holfmann-Thiel et al. (Hofmann-Thiel et al., 2016) also published the first retrospective study which compared the diagnostic performance of

ED

Abbott-MDR platform with conventional Acid-Fast Bacilli (AFB) culture and

PT

Genotype© MTBDRplus assay (HAIN Lifescience, Nehren, Germany). The study

CE

adopted AFB culture result as the reference standard, and Abbott-RT showed a high

AC

specificity (99.5%) but moderate sensitivity (76.3%) among smear-negative respiratory specimens. Whereas for Abbott-RIF/INH, the assay demonstrated high concordance with Genotype© MTBDRplus assay in detecting INHR and RIFR profile. However, 80% of the RIF-resistant specimens of that study carried identical rpoB S531L mutation. The actual diagnostic performance of Abbott-RIF/INH should be evaluated in greater details if more diverse rpoB mutation patterns were included.

This is the first prospective study of Abbott-MDR platform in a TB high-burden region.

4

ACCEPTED MANUSCRIPT The diagnostic performance of this new platform was evaluated using clinical

T

specimens collected from major general hospitals and chest clinics in Hong Kong.

RI P

Both Löwenstein–Jensen (LJ) medium (BioMérieux, Marcy l'Etoile, France) and

SC

BACTEC™ MGIT™ 960 Mycobacterial Detection System (Becton Dickinson Microbiology Systems, Sparks, Md.) were used as reference standards for Abbott-RT

MA NU

analysis. Discrepant results were resolved by clinical background of patients. Genotypic RIFR and INHR detected by Abbott-RIF/INH were verified by Sanger sequencing and compared with phenotypic DSTs using BACTEC™ MGIT™ 960 SIRE

PT

ED

Kit and standard agar proportion method.

CE

2 MATERIAL AND METHODS

AC

The overall workflow of this study is illustrated in Figure 1. Diagnostic performance of Abbott-RT was resolved using standard conventional culture results. For Abbott-RT positive but culture negative specimens, results were resolved as positive for patients exhibiting following clinical background: chest radiograph abnormalities compatible with pulmonary TB, supported by other demographic features including positive contact history or past history of TB. All patients responded to anti-TB therapy. 2.1 Sample collection and processing. Sputum specimens were available from patients with suspected lower respiratory

5

ACCEPTED MANUSCRIPT infection between July 2015 to June 2016 from five general hospitals (inpatients) and

T

13 chest clinics (outpatients) in Hong Kong. All specimens were subjected to AFB

RI P

smear microscopy using Auramine O and Ziehl-Neelsen staining. Approximately 2mL

SC

sputum specimen was processed by conventional N-acetyl-L-cysteine–NaOH method with a final NaOH concentration of 2%. The sediment was re-suspended in phosphate

MA NU

buffered saline to a final volume of 2mL (Pfyffer, 2015). For each processed specimen, 500µL were taken for Abbott-MDR platform and the remaining sediment from each specimen was inoculated on LJ medium and BACTEC™ MGIT™ 960 Mycobacterial

PT

ED

Detection System. .

CE

2.2 Genotypic identification of mycobacterial cultures.

AC

Mycobacterial DNA were extracted from all positive cultures according to our previous publication (Yam et al., 2004, Yam et al., 2006). In brief, a loopful of bacterial colonies from LJ medium was washed in 500μL 0.1M Tris-HCl (pH 7.5). Pellet was then re-suspended in 100μL lysis buffer containing 0.1% w/v NaOH and 0.025% w/v SDS, followed by 45-minute incubation at 60°C. Equal volume of 0.1M Tris-HCl (pH 7.5) was added to the mixture and the resultant DNA extracts were stored at -20°C prior to use. The presence of MTBC was first identified by an in-house polymerase chain reaction (PCR) as described previously (Yam et al., 2006),

6

ACCEPTED MANUSCRIPT followed by 16S rRNA sequencing in case of nontuberculous mycobacteria (NTM)

T

(Yam et al., 2004). Consensus sequences of 16S rRNA were assembled and edited by

RI P

Staden Package (Version 2.0.0) (Bonfield et al., 1995). The final species identification

SC

was based on sequence alignment by BLAST (Basic Local Alignment Search Tools)

MA NU

in NCBI database.

2.3 Phenotypic drug-susceptibility testing (DST).

Phenotypic DSTs were conducted using BACTEC™ MGIT™ 960 SIRE Kit (Becton

ED

Dickinson, Baltimore, MD, USA) and standard agar proportion method on

PT

Middlebrook 7H10 medium. Critical concentrations used in MGIT™ 960 SIRE Kit

CE

were 0.1 mg/L for INH and 1.0 mg/L for RIF. For standard agar proportion method 0.2 mg/L for INH and 1.0 mg/L for RIF were used, in which

AC

(CLSI, 2009),

resistance was defined when growth of inoculums was greater than 1% of the drug-free control. In case of phenotypic DSTs discrepancies or strains harboring novel mutations, minimum inhibitory concentrations (MIC) were determined using MYCOTB Sensititre™ MIC plate (TREK Diagnostic Systems, Cleveland, OH, USA) according to manufacturer’s instructions.

2.4 Abbott RealTime MTB (Abbott-RT) & RealTime MTB INH/RIF Resistance

7

ACCEPTED MANUSCRIPT (Abbott-RIF/INH).

T

A total of 500µL decontaminated sediment from each specimen were mixed with

RI P

1.5mL inactivating reagent (0.4 M NaOH, 60 % Isopropanol and 0.18 % Tween-20)

SC

and incubated for 12 hours at room temperature according to manufacturer’s protocol (Qi et al., 2015). Inactivated samples were loaded onto Abbott m2000sp instrument

MA NU

and DNA was extracted as previously described (Kostera et al., 2016). The plate was manually transferred to Abbott m2000rt real-time PCR thermocycler (Abbott m2000rt) and either “MTB detected” or “MTB not detected” with a manufacturer defined

PT

ED

cut-off cycle number (Cn) of 40.0 would be reported upon completion.

CE

“MTB detected” specimens were subsequently analysed by Abbott-RIF/INH for

AC

genotypic resistance profiling. Equal volumes of DNA eluate (25µL) and Abbott-RIF/INH Amplification Reagent (25µL) were mixed in a new 96-well optical reaction plate prior to loading on Abbott m2000rt. Specimens were reported as “INH susceptible” (INHS) when both wild-type (WT) probes of katG and mabA-inhA were detected. Conversely, specimens were “INH high-level resistant” or “INH low-level resistant” if either katG S315T or mabA-inhA C-15T probe was flagged positive. On the other hand, “RIF resistant” was reported if any of the eight rpoB WT probes covering RIF resistant determining region (RRDR) were not detected. “Below LOD

8

ACCEPTED MANUSCRIPT (limit of detection)” was given by Abbott-RIF/INH if no katG, mabA-inhA or rpoB

RI P

T

target probes were detected.

SC

2.5 Sanger sequencing of rpoA, rpoB, rpoC, katG and mabA-inhA. Sanger sequencing covering RRDR of rpoB, katG codon 315 and inhA upper

MA NU

promoter region were performed on all MTBC culture positive isolates as published previously (Leung et al., 2006, Siu et al., 2011). In case of strains with novel mutations in rpoB, Sanger sequencing on rpoA and rpoC were conducted as

PT

ED

previously described (de Vos et al., 2013).

CE

2.6 RpoB protein structural analysis for Q513* mutant.

AC

The truncated Q513 model was built by homology modelling with the crystal structure of Mycobacterium tuberculosis RpoB (4KBM) (Gulten and Sacchettini, 2013) using the server SWISS-MODEL (Arnold et al., 2006) with default parameters. The model was further analysed using Discovery Studio visualizer (Accelrys), and the Ramachandran plot was examined to ensure that the structure of the model was not in any unfavourable region. The model was then superimposed with the beta and beta’ subunits of the structure of Thermus thermophilus RNA polymerase (2A69) (Artsimovitch et al., 2005) for analysis.

9

ACCEPTED MANUSCRIPT

T

2.7 Statistical analysis.

RI P

The diagnostic performance of Abbott-RT was evaluated primarily against

SC

conventional culture results. The sensitivities, specificities, positive predictive values (PPV) and negative predictive values (NPV) were calculated with reference to both

MA NU

bacteriology and clinical information. Abbott-RIF/INH was compared with phenotypic DSTs and the percentage concordance was calculated based on specimens with positive Abbott-RIF/INH and AFB culture results. For specimens with discordant results

ED

between Abbott-RIF/INH and phenotypic DSTs, the genotypic patterns of the

AC

CE

PT

specimens were confirmed by Sanger sequencing.

10

ACCEPTED MANUSCRIPT 3 RESULTS

T

3.1 Conventional mycobacterial culture.

RI P

In this study, a total of 610 respiratory specimens were prospectively collected from

SC

526 patients with suspected lower respiratory infection. Among these specimens, 187/610 (30.7%) were collected from 144 clinically-defined TB patients, including 161

MA NU

specimens from 131 patients yielded confirmed MTBC positive growth in AFB culture and 26 MTBC culture negative specimens from 13 clinically-defined TB patients under treatment for one to three months. The remaining 423/610 (69.3%)

ED

respiratory specimens were collected from 382 patients with no clinical evidence of

PT

TB. AFB smear positive rate in clinically-defined TB specimens was 46.0% (86/187).

CE

The positive culture rates of LJ medium and MGIT™ 960 were 83.2% (134/161) and

AC

90.1% (145/161) with a mean detection time of 46.2 days (±2.7 days) and 19.5 days (±1.0 days), respectively. By 16S rRNA sequencing, 38/610 (6.2%) specimens were NTM and their identities are listed in Table 1.

3.2 Diagnostic performance of Abbott-RT. Abbott-RT successfully detected MTBC in 178/187 (95.2%) specimens from clinically defined TB patients, including 152 positive MTBC cultures and 26 cases with confirmed TB treatment history. “MTB Not Detected” results were reported in 422/423

11

ACCEPTED MANUSCRIPT (99.8%) MTBC-negative cases and one false positive case was identified with a Cn of

T

39.11. Abbott-RT demonstrated no cross reactivity with NTM and reported as “MTB

RI P

not detected” for all NTM specimens. No inhibition was observed in this evaluation.

SC

Among 101/610 (16.6%) AFB smear positive and 509/610 (83.4%) negative specimens,

and 99.8%, respectively (Table 1).

MA NU

Abbot-RT demonstrated a sensitivity of 98.8% and 92.1% with a specificity of 100%

3.3 Phenotypic DST for M. tuberculosis cultured isolates.

ED

A total of 13 MDR-TB, five RIF-mono resistant (mono-RIFR), 12 INH-mono resistant

PT

(mono-INHR), and 131 pan-susceptible isolates from 161 positive MTBC cultures were

CE

determined based on combined results from MGIT™ 960 SIRE and standard agar

AC

proportion method. Two of the mono-RIFR specimens (HKU 10659 and HKU 12500) with rpoB L533P mutation were RIF susceptible (RIFS) in MGIT™ 960 SIRE but RIFR in standard agar proportion method. MYCOTB Sensititre showed that the RIF MIC for these two specimens were 1 mg/L. Besides, two of the MDR-TB specimens (HKU 11788 and HKU 11789) were later confirmed to be totally drug-resistant tuberculosis (TDR-TB) by MYCOTB Sensititre™ MIC plate and MICs are illustrated in Table 2.

12

ACCEPTED MANUSCRIPT 3.4 Diagnostic performance of Abbott-RIF/ INH.

T

Abbott-RIF/INH was performed on specimens reported as “MTB detected” by

RI P

Abbott-RT, including 152 MTBC-culture positive specimens, 26 cases with TB

SC

treatment history and one false-positive specimen (Table 2). Of 30 phenotypic drug resistant specimens, Abbott-RIF/INH identified 84.6% (11/13) MDR-TB with

MA NU

mutations at both rpoB (Probe 4- [n=7], Probe 5- [n=2], Probe 7- [n=2]) and katG (S315T [n=10], katG- [n=1]), 80% (4/5) mono-RIFR (Probe 4- [n=2], Probe 7- [n=2]) and 66.7% (8/12) mono-INHR (katG S315T [n=5], mabA-inhA C-15T [n=3]). Two

ED

additional mono-RIFR detected by Abbott-RIF/INH (Probe 7- [n=2]) were in fact

PT

MDR-TB specimens without mutations in katG and mabA-inhA. The remaining

CE

MTBC culture positive specimens were either reported as WT (84/152, 55.3%) or

AC

“Below LOD” (38/152, 25%). RIFR and INHR detected by Abbott-RIF/INH were 100% and 79.2% concordant with phenotypic DST, respectively. Genotypic resistance markers of the remaining 26 cases with TB treatment history were also identified by Abbott-RIF/INH. The assay reported 23.1% (6/26) specimens as WT, 19.2% (5/26) specimens carried inhA C-15T mutation and 57.7% (15/26) as “Below LOD”. No concordance with phenotypic DST in these 26 specimens were available due to the absence of AFB positive cultures.

13

ACCEPTED MANUSCRIPT Of all 11 MDR-TB and six mono-RIFR detected by Abbott-RIF/INH, Sanger

T

sequencing confirmed six different RRDR mutation patterns, including H526L [n=3],

RI P

H526D [n=2], H526Y [n=1], S531L [n=8], L533P [n=1] and a novel single mutation

SC

Q513* [n=2]. An additional rpoB L533P mutant was identified in a MTBC culture positive specimen but reported as “Below LOD” by Abbott-RIF/INH (Table 2). On

MA NU

the other hand, Sanger sequencing identified katG S315T [n=15], S315N [n=1] and mabA-inhA C-15T [n=3] mutations in 11 MDR-TB and eight mono-INHR specimens detected by Abbott-RIF/INH. The assay failed to detect mabA-inhA C-15T in another

ED

two mono-INHR specimens and reported as WT or “Below LOD”. The remaining

PT

two MDR-TB and two mono-INHR specimens did not carry any katG codon 315 or

AC

CE

mabA-inhA C-15T mutations and thus inferred as INHS by Abbott-RIF/INH.

3.5 RpoB protein structural analysis and compensatory mutation at rpoA and rpoC and for Q513* mutant. In order to understand the molecular basis of RIF resistance in the Q513* mutant, a molecular model was built by homology modelling with MTB RpoB beta and beta’ subunits structure (4KBM) through SWISS-MODEL. The model was then superimposed on the structure of Thermus thermophilus RNA polymerase in complex with rifapentine (2A69), as described in the molecular model of the SWISS-MODEL

14

ACCEPTED MANUSCRIPT (2A69). As shown in figure 2, the binding pocket of rifapentine was disrupted in the

T

truncated form of RpoB, suggesting that the Q513* mutant is unable to inhibit by

SC

RI P

rifapentine.

To test whether there is compensatory mutations that might restore the bacterial

MA NU

fitness, Sanger Sequencing was performed on rpoA and rpoC for both Q513* mutants. No mutations were detected at rpoA. However, identical rpoC missense mutations at

PT

4 DISCUSSION

ED

D32E and A172V were identified from both Q513* mutants.

CE

Abbott-MDR platform is the latest automatic real-time PCR system designed for rapid

AC

TB diagnosis and determination of RIFR and INHR in MTBC positive specimens. To the best of our knowledge, this is the first prospective study evaluating the overall performance of Abbott-RT and Abbott-RIF/INH in clinical respiratory specimens collected from high TB-burden region.

The performance of Abbott-RT was previously evaluated with clinical specimens collected from different cohorts, including China, Hong Kong and South Africa (Chen et al., 2015, Tang et al., 2015, Wang et al., 2016). However, these evaluations were 15

ACCEPTED MANUSCRIPT based solely on the culture results of less sensitive solid media. Our current study

T

demonstrated that BACTEC™ MGIT™ 960 liquid culture had a higher recovery rate

RI P

and shorter turnaround time than LJ cultures. In order to enhance the reliability of this

SC

study, both solid and liquid media were included in the overall culture results. Under such circumstances, Abbott-RT managed to achieve comparable diagnostic

MA NU

performance as described in previous studies (Chen et al., 2015, Tang et al., 2015). The high sensitivity (92.1%) and specificity (99.8%) of the assay among smear-negative specimens also supports the manufacturer’s claim with a LOD of 17

ED

CFU/mL. In the most recent study conducted by Holfmann-Thiel et al.

PT

(Hofmann-Thiel et al., 2016), Abbott-RT demonstrated a much lower sensitivity

CE

(76.3%) among smear-negative specimens. This could be explained by the

AC

retrospective nature of the study in which decontaminated specimens were first frozen at -30°C up to 6 months prior to analysis. Mycobacterial DNA could have degraded in the frozen decontaminated samples which would reduce sensitivity in smear negative specimens containing low numbers of mycobacteria (Hofmann-Thiel et al., 2016). Nevertheless, no indeterminate results due to PCR inhibition were obtained across both studies, and this can be attributed to the reduction of real-time PCR inhibitors by extracting specimen DNA with magnetic microparticles. When compared with other commercial PCR-based MTBC assays such as GeneXpert® MTB/RIF and Cobas®

16

ACCEPTED MANUSCRIPT Taqman® MTB test with an indeterminate rate of 3.7% and 4.8% respectively

T

(Boehme et al., 2010, Rimek et al., 2002), Abbott-RT demonstrated to be a better

SC

RI P

approach for handling clinical specimens with complex composition.

Despite the excellent diagnostic performance of Abbott-RT, the assay failed to detect

MA NU

the presence of MTBC in nine culture positive specimens. Eight of these specimens were AFB smear negative, indicating that the false negative results might due to the low bacterial load. In addition, we were unable to recover MTBC positive culture in

ED

one Abbott-RT weakly positive specimen [Cn=39.11]. Clinical background showed

PT

that this patient had no symptoms of TB infection in nine months after initial

CE

diagnosis and was not treated for TB. This specimen was considered as the only false

AC

positive in our study. Although not yet well defined, the false positive might be accounted by the persistence of MTBC DNA in normal tissue during latent TB infection (Hernandez-Pando et al., 2000, Soto et al., 2012).

Abbott-RIF/INH, on the other hand, detects RIFR and INHR either as a standalone assay, or as an add-on test on MTBC positive specimens in Abbott-MDR platform. In our study, Sanger Sequencing was performed on all specimens with positive AFB culture result. When compared to the analytical evaluation conducted by Kostera et al.

17

ACCEPTED MANUSCRIPT in which Sanger Sequencing was only performed among specimens with discrepant

T

Abbott-RIF/INH and phenotypic DST results (Kostera et al., 2016), our study was

RI P

able to provide a more comprehensive analysis on the mutation patterns that could be

SC

detected by Abbott-RIF/INH. INHR was regarded as the second-most common resistance pattern in drug-resistant MTB. The global estimated occurrence rate of

MA NU

mono-INHR was around 4-10%, while in Hong Kong the rate was slightly lower at 3.8% (Cattamanchi et al., 2009, HKDH, 2014, Martin et al., 2016). As Xpert® MTB/RIF Ultra assay only provides information on RIFR but not INHR profile (WHO,

ED

2017), MDR-TB identification by Xpert® MTB/RIF Ultra assay might result in the

PT

use of suboptimal anti-TB therapy on mono-INHR cases. In contrast, Abbott-RIF/INH

CE

determines INHR pattern by identifying mutations at katG S315T and mabA-inhA

AC

C-15T. As katG S315T missense mutation causes high-level INHR (MIC>1 mg/L) (Ando et al., 2010) and mabA-inhA C-15T promoter mutation leads to low-level INHR (MIC=0.2-0.8 mg/L) (Abe et al., 2008), Abbott-RIF/INH can distinguish between high- and low-level INHR. Hence, a much more comprehensive drug resistance profile can be provided to clinicians for initiating optimal TB treatment without further delay (Manson et al., 2017).

In this study, only 84% (21/25) phenotypic INHR specimens carried mutations at katG

18

ACCEPTED MANUSCRIPT codon S315 or mabA-inhA. This proportion was similar to the study conducted by

T

Kostera et al., where 89% of the phenotypic INHR specimens carried mutations at

RI P

katG codon S315 or mabA-inhA (Kostera et al., 2016). Although mutations at katG

SC

and mabA-inhA are the most common causes for INHR, multiple studies have reported that mutations at these two genes only account for 66-70% of INHR occurrence

MA NU

(Zhang and Yew, 2009). Mutations at other genes such as aphC, kasA, furA, or at other regions of katG have been reported to confer INH resistance (Kiepiela et al., 2000, Shekar et al., 2014, Siu et al., 2014, Zhang et al., 2005). Further investigations

PT

ED

are required to identify other potential INH-resistance related gene candidates.

CE

Abbott-RIF/INH identified six different RIFR patterns in our study cohort, including a

AC

rare rpoB L533P mutation in one RIFR specimen (HKU10659) which was reported as Pb4- in Abbott-RIF/INH. Sanger sequencing later confirmed that another RIFR specimen (HKU12500) carried the same mutation but was reported as "Below LOD" by Abbott-RIF/INH. Interestingly, these two specimens were regarded as RIFS in MGIT™ SIRE test but RIFR in standard agar proportion method. MYCOTB Sensititre result showed that the two specimens only exhibited slight elevation in RIF resistance in microbroth culture (MIC=1.0 mg/L). Previous studies reported that rpoB L533P mutation often resulted in borderline RIFR, which explained the discordant

19

ACCEPTED MANUSCRIPT susceptibility results between MGIT™ SIRE test and standard agar proportion method

T

(Hauck et al., 2009, Rigouts et al., 2013, Van Deun et al., 2009). In comparing to the

RI P

evaluation by Holfmann-Thiel et al (Hofmann-Thiel et al., 2016), the greater variety

SC

of RIFR-associated rpoB mutation patterns in our study provided a more

performance of Abbott-RIF/INH.

MA NU

representative RRDR mutation pattern for evaluating the actual diagnostic

A nonsense mutation at rpoB Q513 were observed in the cultures of two specimens

ED

from the same patient (HKU11788 and HKU11789) with extremely slow growth rate,

PT

in which positive culture was only detected 50 days after incubation in MGIT™ 960.

CE

These specimens were later confirmed to be TDR-TB by MYCOTB™ Sensititre assay

AC

with high-level resistance towards 12 different anti-TB drugs. It was suspected that the slow growth rate was caused by RpoB protein truncation, as the truncated protein might become functionless in the context of binding to other subunits for transcription. In silico simulation in Figure 2 shows that the truncated protein could not bind to rifapentine, a synthetic chemical with structure resembling RIF, which explained the potential mechanism of RIFR in two TDR-TB strains identified in the present study. The survival and virulent mechanisms for HKU11788 and HKU11789 remained unclear, but various putative mutations at different genes, such as rpoA and rpoC,

20

ACCEPTED MANUSCRIPT were proven with potentials to compensate for the poor growth fitness of rpoB

T

mutated strains (Comas et al., 2012). In this study, the identification of missense

RI P

mutations at rpoC were located at codon 32 and 172, which is distant from the

SC

putative compensatory mutation sites as suggested by similar studies (Brandis et al., 2012, Comas et al., 2012, de Vos et al., 2013). The loss of rpoB function and slow

MA NU

growth rate in the TDR-TB strains may be restored by a similar or even novel pathway. In-depth investigation will be conducted in order to delineate the molecular characteristics associated with the abnormal growth pattern of these two TDR- TB

PT

ED

strains.

CE

Overall, the operation of Abbott-MDR was deemed to be user friendly with minimal

AC

manual handling. The total manual time required approximately 55 minutes with a turnaround time of 2 days. As Abbott-RIF/INH can be selectively performed on a maximum of 22 specimens per batch, the assay can potentially offer a more resource-saving choice for resistance detection. Since the DNA extract has a maximum shelf life of 90 days and amplification reagents can tolerate multiple freeze-thaw cycles, the frequency for resistance test can be adjusted according to the daily workload of clinical laboratory.

21

ACCEPTED MANUSCRIPT Abbott-RIF/INH has its own limitations. The assay renders a specimen as RIFR when

T

the signal for any one of the eight fluorescent RRDR-wild type probe failed to be

RI P

detected. Without prior knowledge of the exact binding sites for each wild-type rpoB

SC

RRDR probes, possible mutation sites could not be discriminated simply from Abbott-RIF/INH results. Moreover, without the use of mutant rpoB RRDR probes to

MA NU

detect common rpoB mutations, specimens carrying silent mutations might be regarded as RIFR. It is also crucial to note that mutations within rpoB RRDR can only explain 95% of the RIFR cases globally. Although not detected in this study, rare RIFR

ED

mutations outside rpoB RRDR such as rpoB V146F and I572F missense mutation

CE

PT

would be reported as RIFS by Abbott-RIF/INH (Siu et al., 2011).

AC

In this study, a total of 30.9% (55/178) “MTB-detected” specimens were reported as “Below LOD” by Abbott-RIF/INH, including one specimen with rpoB L533P mutation and one specimen with mabA-inhA C-15T mutation. According to manufacturer’s instruction, Abbott-RIF/INH has a higher LOD requirement (60CFU/mL) than Abbott-RT (17CFU/mL). We noticed that 48/55 (87.3%) “Below LOD” specimens were AFB smear-negative, suggesting the detection failure might be due to low bacterial load. The LOD differences can be explained by the nature of multiplex real-time PCR reaction as Abbott-RIF/INH uses five fluorescence probes

22

ACCEPTED MANUSCRIPT instead of three in Abbott-RT. A higher DNA concentration, therefore, would be

T

required by Abbott-RIF/INH in order to compensate for the reduced amplification

RI P

efficiency with a more demanding PCR condition. Furthermore, the duo-target system

SC

of Abbott-RT detects MTBC using multiple copies insertion sequence IS6110 while Abbott-RIF/INH detects mutations only in single copy rpoB, katG and inhA genes.

MA NU

The difference in gene copy number might also explain the higher sensitivity of Abbott-RT.

ED

In conclusion, Abbott-MDR platform demonstrated high sensitivity and specificity in

PT

MTBC diagnosis and provided reliable and accurate INHR and RIFR profile among MDR-TB specimens. Together with the simple and flexible nature of its workflow,

CE

Abbott-MDR platform can potentially be deployed as a rapid and high-throughput

China.

AC

diagnostic tool for routine services in high TB burden regions such as Hong Kong and

Ethics approval This study has been approved by the Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster (Ref. number: UW 12-309).

23

ACCEPTED MANUSCRIPT Funding

T

This research did not receive any specific grant from funding agencies in the public,

SC

RI P

commercial, or not-for-profit sectors.

Conflict of Interests

MA NU

None.

Acknowledgement

ED

We are grateful to Chi-Chiu Leung and Kwok-Chiu Chang from the Department of

PT

Health, TB and Chest Service, Hong Kong SAR, China for generously providing

AC

CE

clinical respiratory specimens and clinical information for the study.

24

ACCEPTED MANUSCRIPT Reference

SC

RI P

T

Abe C, Kobayashi I, Mitarai S, Wada M, Kawabe Y, Takashima T, et al. Biological and molecular characteristics of Mycobacterium tuberculosis clinical isolates with low-level resistance to isoniazid in Japan. Journal of Clinical Microbiology 2008;46(7):2263-8. Ando H, Kondo Y, Suetake T, Toyota E, Kato S, Mori T, et al. Identification of katG mutations associated with high-level isoniazid resistance in Mycobacterium tuberculosis. Antimicrobial Agents and Chemotherapy 2010;54(5):1793-9. Arnold K, Bordoli L, Kopp J, Schwede T. The SWISS-MODEL workspace: a

ED

MA NU

web-based environment for protein structure homology modelling. Bioinformatics 2006;22(2):195-201. Artsimovitch I, Vassylyeva MN, Svetlov D, Svetlov V, Perederina A, Igarashi N, et al. Allosteric modulation of the RNA polymerase catalytic reaction is an essential component of transcription control by rifamycins. Cell 2005;122(3):351-63. Boehme CC, Nabeta P, Hillemann D, Nicol MP, Shenai S, Krapp F, et al. Rapid Molecular Detection of Tuberculosis and Rifampin Resistance. New England Journal of Medicine 2010;363(11):1005-15. Bonfield JK, Smith KF, Staden R. A new DNA sequence assembly program. Nucleic

AC

CE

PT

Acids Research 1995;23(24):4992-9. Brandis G, Wrande M, Liljas L, Hughes D. Fitness-compensatory mutations in rifampicin-resistant RNA polymerase. Molecular Microbiology 2012;85(1):142-51. Cattamanchi A, Dantes RB, Metcalfe JZ, Jarlsberg LG, Grinsdale J, Kawamura LM, et al. Clinical characteristics and treatment outcomes of patients with isoniazid-monoresistant tuberculosis. Clinical Infectious Diseases 2009;48(2):179-85. Chen JH, She KK, Kwong TC, Wong OY, Siu GK, Leung CC, et al. Performance of the new automated Abbott RealTime MTB assay for rapid detection of Mycobacterium tuberculosis complex in respiratory specimens. European Journal of Clinical Microbiology & Infectious Diseases 2015;34(9):1827-32. CLSI. Susceptibility testing of mycobacteria, nocardia, and other aerobic actinomycetes; approved standard, 2nd ed. CLSI document M24-A2. Clinical and Laboratory Standards Institute, Wayne, PA.; 2011. Comas I, Borrell S, Roetzer A, Rose G, Malla B, Kato-Maeda M, et al. Whole-genome sequencing of rifampicin-resistant Mycobacterium tuberculosis strains identifies compensatory mutations in RNA polymerase genes. Nature Genetics 2011;44(1):106-10. de Vos M, Muller B, Borrell S, Black PA, van Helden PD, Warren RM, et al. Putative compensatory mutations in the rpoC gene of rifampin-resistant Mycobacterium 25

ACCEPTED MANUSCRIPT

SC

RI P

T

tuberculosis are associated with ongoing transmission. Antimicrobial Agents and Chemotherapy 2013;57(2):827-32. Enjeti A, Granter N, Ashraf A, Fletcher L, Branford S, Rowlings P, et al. A longitudinal evaluation of performance of automated BCR-ABL1 quantitation using cartridge-based detection system. Pathology 2015;47(6):570-4. Gulten G, Sacchettini JC. Structure of the mtb CarD/RNAP β-Lobes complex reveals the molecular basis of interaction and presents a distinct DNA-binding domain for mtb CarD. Structure 2013;21(10):1859-69. Hauck Y, Fabre M, Vergnaud G, Soler C, Pourcel C. Comparison of two commercial assays for the characterization of rpoB mutations in Mycobacterium tuberculosis and

ED

MA NU

description of new mutations conferring weak resistance to rifampicin. Journal of Antimicrobial Chemotherapy 2009;64(2):259-62. Hernandez-Pando R, Jeyanathan M, Mengistu G, Aguilar D, Orozco H, Harboe M, et al. Persistence of DNA from Mycobacterium tuberculosis in superficially normal lung tissue during latent infection. The Lancet 2000;356(9248):2133-8. HKDH. Annual Report of Tuberculosis & Chest Service of the Department of Health 2014. Department of Health, Hong Kong Special Adminstirative Region.; 2014. Hofmann-Thiel S, Molodtsov N, Antonenka U, Hoffmann H. Evaluation of the Abbott RealTime MTB and RealTime MTB INH/RIF Assays for direct detection of

AC

CE

PT

Mycobacterium tuberculosis complex and resistance markers in respiratory and extrapulmonary specimens. Journal of Clinical Microbiology 2016;54(12):3022-7. Ioannidis P, Papaventsis D, Karabela S, Nikolaou S, Panagi M, Raftopoulou E, et al. Cepheid GeneXpert MTB/RIF assay for Mycobacterium tuberculosis detection and rifampin resistance identification in patients with substantial clinical indications of tuberculosis and smear-negative microscopy results. Journal of Clinical Microbiology 2011;49(8):3068-70. Kiepiela P, Bishop K, Smith A, Roux L, York D. Genomic mutations in the katG, inhA and aphC genes are useful for the prediction of isoniazid resistance in Mycobacterium tuberculosis isolates from Kwazulu Natal, South Africa. Tubercle and Lung Disease 2000;80(1):47-56. Kostera J, Leckie G, Tang N, Lampinen J, Szostak M, Abravaya K, et al. Analytical and clinical performance characteristics of the Abbott RealTime MTB RIF/INH Resistance, an assay for the detection of rifampicin and isoniazid resistant Mycobacterium tuberculosis in pulmonary specimens. Tuberculosis 2016;101:137-43. Leung ETY, Ho PL, Yuen KY, Woo WL, Lam TH, Kao RY, et al. Molecular characterization of isoniazid resistance in Mycobacterium tuberculosis: identification of a novel mutation in inhA. Antimicrobial Agents and Chemotherapy 2006;50(3):1075-8. 26

ACCEPTED MANUSCRIPT

SC

RI P

T

Manson AL, Cohen KA, Abeel T, Desjardins CA, Armstrong DT, Barry Iii CE, et al. Genomic analysis of globally diverse Mycobacterium tuberculosis strains provides insights into the emergence and spread of multidrug resistance. Nature Genetics 2017;advance online publication. Martin LJ, Roper MH, Grandjean L, Gilman RH, Coronel J, Caviedes L, et al. Rationing tests for drug-resistant tuberculosis – who are we prepared to miss? BMC Medicine 2016;14(1):1-9. Outhred AC, Jelfs P, Suliman B, Hill-Cawthorne GA, Crawford AB, Marais BJ, et al. Added value of whole-genome sequencing for management of highly drug-resistant TB. Journal of Antimicrobial Chemotherapy 2015;70(4):1198-202.

ED

MA NU

Pfyffer GE. Mycobacterium: General Characteristics, Laboratory Detection, and Staining Procedures*. Manual of Clinical Microbiology, Eleventh Edition: American Society of Microbiology; 2015. Qi C, Wallis C, Pahalawatta V, Frank A, Ramdin N, Viana R, et al. Evaluation of the efficiency of the sample inactivation reagent in the Abbott RealTime MTB assay for inactivation of Mycobacterium tuberculosis. Journal of Clinical Microbiology 2015;53(9):3001-2. Rigouts L, Gumusboga M, de Rijk WB, Nduwamahoro E, Uwizeye C, de Jong B, et al. Rifampin resistance missed in automated liquid culture system for Mycobacterium

AC

CE

PT

tuberculosis isolates with specific rpoB mutations. Journal of Clinical Microbiology 2013;51(8):2641-5. Rimek D, Tyagi S, Kappe R. Performance of an IS6110-based PCR assay and the COBAS AMPLICOR MTB PCR system for detection of Mycobacterium tuberculosis complex DNA in human lymph node samples. Journal of Clinical Microbiology 2002;40(8):3089-92. Shekar S, Yeo ZX, Wong JC, Chan MK, Ong DC, Tongyoo P, et al. Detecting novel genetic variants associated with isoniazid-resistant Mycobacterium tuberculosis. PloS One 2014;9(7):e102383. Siu GKH, Yam WC, Zhang Y, Kao RY. An upstream truncation of the furA-katG operon confers high-level isoniazid resistance in a Mycobacterium tuberculosis clinical isolate with no known resistance-associated mutations. Antimicrobial Agents and Chemotherapy 2014;58(10):6093-100. Siu GKH, Zhang Y, Lau TC, Lau RW, Ho P-L, Yew W-W, et al. Mutations outside the rifampicin resistance-determining region associated with rifampicin resistance in Mycobacterium tuberculosis. Journal of Antimicrobial Chemotherapy 2011;66(4):730-3. Soto ME, Ávila-Casado MDC, Huesca-Gómez C, Alarcon GV, Castrejon V, Soto V, et al. Detection of IS6110 and HupB gene sequences of Mycobacterium tuberculosis and 27

ACCEPTED MANUSCRIPT

SC

RI P

T

bovis in the aortic tissue of patients with Takayasu’s arteritis. BMC Infectious Diseases 2012;12(1):1. Tang N, Frank A, Pahalawatta V, Lampinen J, Coblenz-Korte A, Dunn C, et al. Analytical and clinical performance of Abbott RealTime MTB, an assay for detection of Mycobacterium tuberculosis in pulmonary specimens. Tuberculosis 2015;95(5):613-9. Van Deun A, Barrera L, Bastian I, Fattorini L, Hoffmann H, Kam KM, et al. Mycobacterium tuberculosis strains with highly discordant rifampin susceptibility test results. Journal of Clinical Microbiology 2009;47(11):3501-6. Wang SF, Ou XC, Li Q, Zheng HW, Wang YF, Zhao YL. The Abbott RealTime MTB

ED

MA NU

assay and the Cepheid GeneXpert assay show comparable performance for the detection of Mycobacterium tuberculosis in sputum specimens. International Journal of Infectious Diseases 2016;45:78-80. Weyer K, Mirzayev F, Migliori GB, Van Gemert W, D'Ambrosio L, Zignol M, et al. Rapid molecular TB diagnosis: evidence, policy making and global implementation of Xpert MTB/RIF. European Respiratory Journal 2013;42(1):252-71. WHO. Global tuberculosis report 2016. World Health Organization, Geneva, Switzerland; 2016. WHO. Meeting Report of a Technical Expert Consultation: Non-inferiority analysis of

AC

CE

PT

Xpert MTB/RIF Ultra compared to Xpert MTB/RIF. Meeting Report of a Technical Expert Consultation: Non-inferiority analysis of Xpert MTB/RIF Ultra compared to Xpert MTB/RIF. World Health Organization, Geneva, Switzerland; 2017. Yam WC, Cheng VCC, Hui WT, Wang LN, Seto WH, Yuen KY. Direct detection of Mycobacterium tuberculosis in clinical specimens using single-tube biotinylated nested polymerase chain reaction-enzyme linked immunoassay (PCR-ELISA). Diagnostic Microbiology and Infectious Disease 2004;48(4):271-5. Yam WC, Yuen KY, Kam SY, Yiu LS, Chan KS, Leung CC, et al. Diagnostic application of genotypic identification of mycobacteria. Journal of Medical Microbiology 2006;55(5):529-36. Zhang M, Yue J, Yang Y-p, Zhang H-m, Lei J-q, Jin R-l, et al. Detection of mutations associated with isoniazid resistance in Mycobacterium tuberculosis isolates from China. Journal of Clinical Microbiology 2005;43(11):5477-82. Zhang Y, Yew W. Mechanisms of drug resistance in Mycobacterium tuberculosis The International Journal of Tuberculosis and Lung Disease 2009;13(11):1320-30. Zhao Y, Xu S, Wang L, Chin DP, Wang S, Jiang G, et al. National survey of drug-resistant tuberculosis in China. New England Journal of Medicine 2012;366(23):2161-70.

28

ACCEPTED MANUSCRIPT

MTBC positive (n=81) a

73

MTBC negative (n=428) Under treatment (n=20) b Culture negative (n=380) NTM (n=28)e

21 20 1d 0

IP

Negative (n=509)

8

407 0 379 28

Sensitivity 98.8 [92.8-99.9]

CR

MTBC culture result MTBC positive (n=80) a MTBC negative (n=21) Under treatment (n=6)b Culture negative (n=5) NTM (n=10)c

ED

AFB Smear Positive (n=101)

MA NU S

Abbott-RT (%) Positive Negative 79 1 6 15 6 0 0 5 0 10

T

Table 1. Rapid diagnosis of M. tuberculosis using Abbott-RT for 610 sputum specimens

92.1 [84.5-96.3]

Resolved performance (%) [95% CI] Specificity PPV 100 100 [74.7-100] [94.6-100]

99.8 [98.4-100.0]

98.9 [93.4-99.9]

NPV 93.8 [67.7-99.7]

98.1 [96.1-99.1]

PT

a. Specimens yielded MTB culture positive from the original specimen or subsequent specimen from the same patient within the same week. b. Specimens were collected from patients with chest radiograph abnormalities compatible with pulmonary TB, supported by other

CE

demographic features including positive contact history or past history of TB. All patients responded to anti-TB therapy. c. 10 specimens were confirmed NTM by 16S rRNA sequencing: M. abscessus (n=5), M. avium (n=4) and M. kansasii (n=1).

AC

d. 1 patient with AFB smear negative/ m2000qPCR positive/ MTB culture negative result had no indication of TB infection nor receiving anti-TB treatment in the following 9-month period. e. 28 mycobacterial isolates were confirmed NTM by 16S rRNA sequencing: M. arupense (n=1), M. avium (n=3), M. intracellulare (n=5), M. abscessus (n=3), M. celatum (n=1), M. colombiense (n=1), M. chelonae (n=3), M. fortuitum (n=1), M. gordonae (n=2), M. kansasii (n=3), M. neoaurum (n=1), M. parascrofulaceum (n=1), M. yongonense (n=3). Abbreviations: AFB, acid-fast bacilli; CI, confidence interval; MTBC, Mycobacterium tuberculosis complex; NTM, Nontuberculous mycobacteria; PPV, positive predictive value; NPV, negative predictive value.

29

Table 2. Rapid diagnosis of drug resistance using Abbott-RIF/INH Resistance detection on 187 specimens from 144 clinically-defined TB patients

ACCEPTED MANUSCRIPT

Isolates

MA NU S

S315T S315T wt wt wt wt wt Below LOD Below LOD Below LOD wt wt Below LOD N/A

wt wt wt wt C-15T wt wt Below LOD Below LOD Below LOD C-15T wt Below LOD N/A

T

Sanger sequencing katG S315T S315N S315T S315T wt wt wt S315T wt wt wt wt wt N/A N/A wt

IP

S531L H526Y S531L H526D wt wt wt L533PB wt wt N/A N/A N/A wt

CR

Probe 4Probe 7Probe 4Probe 7wt wt wt Below LOD Below LOD Below LOD wt wt Below LOD N/A

ED

1 1 1 1 1 1 34 1 1 31 1D 4D 15D 8E

PT

Negative [n=101]

rpoB S531L S531L Q513*C H526D H526L L533PB H526L wt wt wt wt wt wt N/A N/A wt

CE

Positive [n=86]

Specimens Abbott RIF/INH Resistance detection rpoB katG inhA Probe 4S315T wt A Probe 4katGwt Probe 5S315T wt Probe 7S315T wt Probe 7wt wt Probe 4wt wt Probe 7wt wt wt S315T wt wt wt C-15T wt wt wt wt wt wt wt wt wt Below LOD Below LOD Below LOD wt wt C-15T wt wt wt N/A N/A N/A

AC

AFB smear result

No. of specimens (n=187) 5 1 2 1 2 1 1 5 2 1 1 50 7 4D 2D 1E

S315T S315T wt wt wt wt wt wt wt wt N/A N/A N/A wt

inhA wt wt wt wt wt wt wt wt C-15T C-15T wt wt wt N/A N/A wt

MGIT™ 960 SIRE RIF INH R R R R R R R R R R R S R S S R S R S R S R S S S S N/A N/A N/A N/A S S

wt wt wt wt C-15T wt wt wt C-15T wt N/A N/A N/A wt

R R R R S S S R S S N/A N/A N/A S

R R S S R R S S R S N/A N/A N/A S

A. Neither katG wild-type probe nor the S315T mutant were detected in the specimen (HKU11619). B. 2 specimens with rpoB L533P (HKU10659 and HKU12500) reported as RIF susceptible in MGIT™960 SIRE but RIF resistant in standard agar proportion method. C. 2 specimens with rpoB Q513* mutation (HKU11788 and HKU11789) confirmed to be TDR-TB by MYCOTB® Sensititre MIC plate with the following MICs: Rifampicin [≥16 mg/L], Isoniazid [≥4 mg/L]; Ethambutol [≥32 mg/L]; Streptomycin [32 mg/L]; Ofloxacin [≥32 mg/L]; Moxifloxacin [8 mg/L]; Kanamycin [≥40 mg/L]; Amikacin [4 mg/L]; Ethionamide [≥40 mg/L]; Rifabutin [≥16 mg/L]; Cycloserine [≥256 mg/L]; and Para-aminosalicylic acid [≥64 mg/L]. D. Clinically defined MTB cases with negative AFB culture results. Hence, Sanger sequencing and phenotypic DST on MTB isolates were unavailable. E. Clinically defined MTB cases flagged as “MTB not detected” in Abbott-RT. Hence Abbott-RIF/INH were not performed on theses specimens.

30

ACCEPTED MANUSCRIPT Fig. 1 Study plan for Abbott-MDR platform evaluation.

AC

CE

PT

ED

MA

NU

SC

RI P

T

Fig. 2 In silico simulation of RIF binding pocket of RpoB in (A) wild-type, and (B) truncated structure. The wild–type beta subunit and truncated beta subunit beta is colored in purple and yellow, respectively; rifapentine was shown in ball-and-stick format.

31

ED

MA

NU

SC

RI P

T

ACCEPTED MANUSCRIPT

AC

CE

PT

Figure 1

32

ED

MA

NU

SC

RI P

T

ACCEPTED MANUSCRIPT

AC

CE

PT

Figure 2

33

ACCEPTED MANUSCRIPT

AC

CE

PT

ED

MA

NU

SC

RI P

T

Highlights  Evaluation of Abbott RealTime MTB and RealTime MTB RIF/INH Resistance in clinical settings  Abbott RealTime MTB provided reliable MTB diagnosis in respiratory specimens  Abbott RealTime MTB RIF/INH Resistance accurately identified RIF/INH resistance markers in MDR-TB from respiratory specimens

34