Advances in the Diagnosis and Treatment of Tuberculosis San Antonio, Texas

Transcription

1 Advances in the Diagnosis and Treatment of Tuberculosis San Antonio, Texas Molecular Detection of Drug Resistance Beverly Metchock, DrPH, D(ABMM) February 22, 2012 Beverly Metchock, DrPH, D(ABMM) has the following disclosures to make: No conflict of interests No relevant financial relationships with any commercial companies pertaining to this educational activity 1

2 Molecular Detection of Drug Resistance Beverly Metchock, DrPH, D(ABMM) Team Lead, Reference Laboratory/Division of Tuberculosis Elimination February 22, 2012 National Center for HIV/AIDS, Viral Hepatitis, STD & TB Prevention Division of Tuberculosis Elimination Objectives Describe tests for the detection of drug resistance in Mycobacterium tuberculosis complex (MTBC) Conventional culture based methods Molecular methods Use case based scenarios to explain the use of molecular test results and to illustrate the benefits and limitations of these tests 2

3 Amplificationbased Tests 21 st Century Algorithm Process Specimen 1 day AFB Microscopy Inoculate Media 2 to 6 weeks Culture Positive Molecular DST Species Identification 2 to 3 weeks Drug Susceptibility Drug Susceptibility Testing (DST) of MTBC Current recommendations (Clinical and Laboratory Standards d Institute t [CLSI] M24 A2) Initial isolate should be tested against primary or first line drugs (FLD) INH, RMP, PZA, EMB For RMP resistant it tisolates, or resistance it to any 2 FLD, test second line drugs (SLD) To include FQ, AMK, KAN, CAP 3

4 Agar Proportion Method for MTBC DST The method of proportion using Middlebrook 7H10 agar has been considered the gold standard method in the U.S. for several decades used in RLT Plate bacteria onto media containing no drugs critical concentrations of a drug Incubate for 3 weeks Count colonies Isolate is resistant if the number of colonies on drugcontaining media is >1% of the colonies on drug free media DST in Broth Systems Selection of critical (testing) concentrations based on comparison of results with agar proportion = equivalent critical concentrations Much more rapid results (5 7 days) than agar proportion (21 28 days) FDA cleared for first line drugs MGIT IRES, Z TREK IRE, Z Published evaluations of second line drugs 4

5 Current Practice for DST For FLD, broth based methods are routine and widely available Results generally available within 28 days of specimen receipt in laboratory Molecular assays (RMP, INH) are available in a few jurisdictions Laboratory developed tests or research use only tests Performed directly on clinical specimens or on culture isolates and results available within 1 2 days For SLD, testing is often is performed in piecemeal fashion through referral algorithms; few laboratories with technical expertise and capacity Slow turn around time Direct agar proportion (e.g., sputum sediment) takes ~21 days Indirect agar proportion takes ~28 days after isolation from culture Some laboratories have verified and validated methods for broth based testing Problems/Concerns with Current DST Practices Most testing algorithms based on referrals of specimens/isolates Lack of confidence/reluctance of labs to report resistance prior to confirmation Discordant results inter and intra lab, dff different methods, etc. Shift in laboratory workforce and loss of expertise for conventional TB testing methodologies 5

6 Genotype MTBDRplus ( HAIN test ) NAA and hybridization based test use immobilized DNA probes on nitrocellulose membranes (line probe assay [LPA]) Colorimetric change indicates hybridization Read the bands to determine MTBC or not and to detect resistance associated mutations for RMP and INH Cepheid Xpert MTB/RIF Assay Automated commercial system for identification of M. tuberculosis complex and mutations in rpob Uses real time PCR with molecular beacons 5 probes for wild type RRDR in rpob and 1 probe for amplification control (B. globigii) Decontamination, digestion, DNA extraction, amplification, and detection in same cartridge; Limited biosafety requirements Results in ~2 hours Minimal hands on manipulation technically simple Platform is random access 6

7 Cepheid Xpert MTB/RIF Assay CDC s MDDR Service (Molecular Detection of Drug Resistance) Implemented in September 2009 (CLIA compliant) Comprehensive clinical service to domestic TB control programs and clinicians Rapid confirmation of RMP resistant and MDR TB Laboratory testing data available about SLD resistance in cases of RMP resistant or MDR TB New technologies may fill the role in the future but demand exists now 7

8 MDDR Service Description Molecular Assays* 9 Gene targets amplified by 10 PCR reactions Sanger Sequencing g q g Forward and reverse strands; ABI 3100xl or ABI 3500xl Comparison of sequences generated to fullly sequenced and annotated pan susceptible strains Conventional DS T performed in parallel * Campbell, PJ, et al Antimicrob.Agent. Chemother. 55: MDDR Service: Drugs and Genes for Panel Rifampin rpob (81bp region) Isoniazid MDR TB inha ( 15) Isoniazid katg (Ser315) Ethambutol embb (Met306, Gly406) Pyrazinamide pnca (promoter and coding regions) Fluoroquinolones gyra (coding region) Amikacin, Kanamycin, Capreomycin Kanamycin Capreomycin XDR TB rrs (nt1401/1402,1484) eis (promoter region) tlya (coding region) 8

9 Criteria for MDDR Testing Isolate available (solid or liquid media) NAAT (+) specimens upon consultation High risk patients (RMP R, MDR TB) From population with high rates of drug resistance Exposed to DR case Failing therapy Cases of public health importance Impacton public healthmeasures & public healthresponse Known RMP Resistance Conventional or molecular test by submitter Mixed or non viable cultures Other Reasons MDDR Algorithm Isolate Received for MDDR Molecular Analysis Conventional DST 2 day turnaround time Molecular Results (Interim Report) ~35 day turn around time Molecular + Conventional DST Results (Final Report) 9

10 C T Most commonly observed rpob mutation: TCG>TTG Ser531Leu Advantages Conventional DNA Sequencing Long L sequence reads >500bp Easy to customize Ability to detect mixed sequences LOD ~30% Ability to find new mutations Actual DNA sequence is determined Disadvantages Labor intensive hands on Equipment cost and maintenance Reagent cost Throughput limitations sample # vs loci # ( Specimens) Isolates 10

11 Pyrosequencing Direct DNA sequencing of PCR products Unique chemistry detection of released pyrophosphate Visible light is generated that is proportional to the number of incorporated nucleotides Biotin labeled PCR product (1 strand) Biotinylated DNA strand captured on beads Beads hybridized with sequencing primer Instrument carries out DNA sequencing reaction and analysis <2 hrs 11

12 Advantages Rapid sequencing reactions Easy to customize Ability to detect mixed sequences High throughput Actual DNA sequence is determined Pyrosequencing Disadvantages Short Sequence Reads <100bp Labor intensive hands on Separate PCR rx s to sequence both strands Poor performance in homopolymer stretches (i.e.,>3 same nucleotide) Specimens Isolates How to report results? Weighing Genotypic versus Phenotypic Results The term Gold Standard can be misleading. New, previously uncharacterized or poorly characterized mutations Reported as clinical significance unknown Anecdotal information may be reported Functional genetic analysis is necessary to definitely determine effect of mutation on resistance 12

13 Other Example Interpretive Comments No mutations in inha or katg Cannot rule out INH resistance. (90% of INH R isolates in our in house evaluation of 254 clinical isolates have a mutation at one or both of these loci.) Ser531Leu mutation in rpob Rifampin resistant. (100% of isolates in our in house evaluation of 254 clinical isolates with this mutation are RMP R R.) Gly285Cys mutation in katg Cannot rule out INH resistance. The mutation found has not been detected previously and the significance of this mutation regarding predicting resistance to INH is unknown. 13

14 Benefits of Molecular Detection of Drug Resistance Rapid results within days as compared to weeks for conventional testing Expedite p further conventional testing (e.g., second line drug susceptibility testing) High throughput Some assays are closed systems reduces potential for cross contamination Development of technologies requiring limited biosafety infrastructure; does not require BSL 3 once DNA is extracted Information provided by some platforms may be used to enhance accuracy of conventional DST Limitations and Considerations (review) Not all mechanisms of resistance are known and the lack of a mutation susceptibility Limited genes and sites are targeted Emerging resistance (mixed populations) may not be detected; limit of detection Not all mutations are associated with phenotypic resistance Silent (synonymous) mutations no no change in protein (e.g., Phe514Phe and Arg528Arg in rpob) Neutral polymorphisms (e.g., gyra codon 95 may be Ser or Thr ) 14

15 Limitations and Considerations (review 2) Still filling in gaps in knowledge about drug resistance (phenotypic and genotypic testing) Gold standard DST may not be perfect Mutations resulting in elevated MICs but S at critical concentration (e.g., Leu511Pro in rpob) Clinical utility Do results impact patient care? Will clinicians trust these results or wait for the conventional DST result? Expertise of staff and clinicians Output from the assay depends on the platform; Need to understand platform to understand limitations Educational partnerships (laboratory, program, and clinicians) need to be developed Understanding Discordance Between different phenotypic DST results e.g., MGIT 960 and agar proportion Between phenotypic and genotypic (molecular) results 15

16 What causes discordant DST results? Human error/lab error Transcription, labeling errors Cross contamination/specimen mix up Different inoculum /bacterial l population e.g., isolates from different specimens; sampling from same specimen; original isolate vs. subculture Size of inoculum/clumps Different growth kinetics Different method or media equivalent critical concentrations calling result too soon The bug MIC is close to the critical concentration Evaluations performed with highly resistant bugs What causes discordance between molecular and phenotypic DST results? Not all mechanisms of resistance are known and the lack of a mutation susceptibility Limited genes and sites are targeted Emerging resistance (mixed populations) may not be detected; limit of detection Not all mutations are associated with phenotypic resistance Silent (synonymous) mutations and neutral polymorphisms Gold standard DST may not be perfect Mutations resulting in elevated MICs but S at critical concentration 16

17 Cases Case # 1 Is it RMP R? (RMP Discordance between molecular and conventional results) Smear (+) pulmonary TB; prisoner At hospital Xpert (X2) RMP Resistance Detected DST (MGIT) INH R and RMP S AP DST pending at State lab At CDC, rpobdna sequence Phe514Phe 17

18 SILENT MUTATIONS Nucleotide changes that do not result in a change in amino acid sequence CGC>CGT Arg528Arg TTC>TTT Phe514Phe rpob Case # 2 Is it RMP R? (RMP discordance between broth and AP) State PHL DST results: Bactec 460 R to INH; S to RMP (2 µg/ml) AP (7H10) 100% R to INH; 80% R to RMP (1 µg/ml) MDDR: rpob Asp516Tyr; RMP resistant inha C( 15)T; INH resistant CDC AP: 40% R to RMP 18

19 Case # 3 Is it RMP R? (RMP Discordance between molecular and conventional results) Pulmonary TB; Burma (Nepal camp) State Lab DST (MGIT) INH R and RMP S CDC rpob Asp516Tyr; RMP resistant CDC AP RMP S rpob mutations associated with highly discordant DST results (van Deun, A. et al. 2009, J Clin Microbiol. 47: ) Low level or borderline resistance Probably clinically relevant resistance Probably clinically relevant resistance Resistance often missed by standard, growth based systems, especially automated broth systems Critical concentration may be too high to cover all clinically relevant resistance, or Maybe the methods need modification (e.g., prolonged incubation, larger inoculum size) to detect resistance Frequency of these strains unknown Mutations : Asp516Tyr, Leu511Pro, Leu533Pro, His526Leu, His526Ser, Ile572Phe 19

20 CDC MPEP Strain H (6/2008) and Strain T (5/2010)* H (6/2008) T (5/2010) Method No. RMP % No. RMP % R/No. results R/No. results LJ Proportion 7/7 100 n/a n/a Agar 19/ /23 65 Proportion BACTEC / /19 37 MGIT 13/ /61 15 VersaTREK 0/3 0 0/5 0 Total 54/ / *His526Leu mutation in rpob Asp516Tyr (CDC MDDR) (cases #2 and 3) Original (retrospective) validation RMP R R by DST RMP S by DST # with mutation/total # with mutation/total 2/152 0/102 Prospective validation 0/17 1/63 Service (9/2009 2/2011) 2/84 3/143 Total 4/253 4/308 20

21 Clinical failures associated with rpob mutations in phenotypically occult MDR TB (Williamson, DA, et al. 2012, Int J Tuberc Lung Dis, 16: ) Significant g association between the presence of rpob mutations that are not detected in DST and treatment failure 94 patients INH R, RMP S by MGIT DST ( ) 4 had rpob mutations (GeneXpert) RRDR sequenced Leu511Pro/Met515Ile, His526Asn/Ala532Val, Asp516Tyr, His526Leu 3 of 4 were treatment failures; other was unknown Case # 4 Is it RMP R? Isolate submitted for MDDR HIV+, prison, Mexico, intermittent therapy, funky RMP on Bactec 460 CDC rpob wildtype; probably RMP S CDC AP contaminated Resubmit isolate (A) and a newer isolate (B) AP (A) RMP R (5%) AP (B) RMP R (12%) rpob on colonies His536Tyr (100% of isolates with this mutation are RMP R) Under limit of detection of MDDR 21

22 Case #5 Discordant results for INH and RMP between 2 laboratories Lab A; INH R and RMP R R (MGIT) Lab B; INH S and RMP S (MGIT and AP) CDC MDDR rpob, inha, katg all wildtype (no mutations) CDC AP DST S to INH, S to RMP Question What are the possible reasons for the discordant results? 1.? Specimen mix up 2. Technical error in lab A 3. Lab B and CDC did not test same bug as lab A 4. All of the above 22

23 Case #6 Discordant results for RMP and EMB between 2 laboratories Lab A; RMP R R and EMB R (MGIT) CDC MDDR rpob deletion mutation; probably resistant embb Asp354Ala; probably resistant (80% of isolates with this mutation R at CDC) CDC AP DST S to RMP and EMB (X2)? AP DST missed R Issues with recent MPEP EMB MPEP is a free performance evaluation program to assess the participating laboratory s drug susceptibility testing process for M. tuberculosis. Participation in the program is voluntary and individual laboratories are not identified in the aggregates reports generated for each shipment. In 2009 and 2010, a total of 19 isolates of M. tuberculosis complex were included in 4 panel shipments. Six of these isolates were resistant to EMB 23

24 EMB results reported by MPEP participants for six isolates of EMB resistant M. tuberculosis complex, Isolate embb mutation CDC AP (% R) MPEP Participant Results (No. reported R) / (No. reported results) [% reported R] AP BACTEC 460 MGIT VersaTrek 2009L Met306Ile 12.5% 19/24 [79] 23/24 [96] 20/56 [36] 1/4 [25] 2009P Glu378la c 100% 21/21 [100] 21/22 [96] 37/63 [59] 4/5 [80] 2010A a Gly406Asp 40% 9/22 [41] 12/20 [60] 1/65 [2] 0/5 [0] 2010E a Gly406Asp 40% 9/22 [41] 13/20 [65] 1/63 [2] 0/5 [0] 2010C b Met306Val 100% 23/26 [88] 19/20 [95] 13/62 [21] 5/6 [83] 2010D b Met306Val 100% 24/26 [92] 19/20 [95] 15/63 [24] 5/6 [83] a same strain; b same strain; C neutral mutation not associated with resistance. Second line drugs 24

25 Fluoroquinolones High degree of cross resistance between ofloxacin (2 µg/ml), levofloxacin (1 µg/ml), and moxifloxacin (0.5 µg/ml) [agar proportion] Testing moxifloxacin at 0.5 µg/ml and levofloxacin at 1.0 µg/ml [agar proportion] will reliably identify most isolates resistant to these drugs Some isolates resistant to moxifloxacin at 0.5 µg/ml are susceptible at 2.0 µg/ml; this is associated with particular gyra mutations Mutations in gyrbcan also lead to fluoroquinolone resistance M. Wilby Keystone Meeting Fluoroquinolones CDC MDDR service 11 isolates with gyra mutations; all were R by agar proportion 4 additional isolates were R by agar proportion (<40% resistance when compared to growth control) 4 discordant isolates were grown in the presence of drug and then gyra sequenced 2 had mutations in gyra Limit of detection 25

26 Cross Resistance Among Second Line Injectable Drugs locus KAN CAP AMK eis R tlya R rrs R (R) R Situations in which the definition of drug resistance is challenging* New drug Availability of different members of the same class that have different in vitro potency (e.g., fluoroquinolones) Lack of clinical validation of drug resistance (e.g., many 2 nd and 3 rd line drugs) New testing methods (phenotypic vs. genotypic tests) Multidrug l d therapy * Dr. William Burman, APHL TB Laboratory Meeting, June

27 Possible reasons that resistance may be under estimated by phenotypic testing* Mixed infection Lack of detection of specific forms of resistance (e.g., RMP) Minority variants resistance mutations having a fitness cost * Dr. William Burman, APHL TB Laboratory Meeting, June 2011 Conclusions Paradigm shift in DST Molecular tests do not replace conventional DST Need to develop cost effective algorithms for incorporating new technology; timely referral Results from genotypic and phenotypic tests need to be used in conjunction with one another (may depend on drug and genetic locus) molecular (genotypic) tests may Elucidate truth in certain cases Elucidate truth in certain cases Add to confusion in certain cases Help us fine tune conventional DST Communication is essential 27

28 Acknowledgements DTBE Laboratory Branch Tracy Dalton, PhD Jeff Driscoll, PhD James Posey, PhD Angela Starks, PhD (404) The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention. National Center for HIV/AIDS, Viral Hepatitis, STD & TB Prevention Division of Tuberculosis Elimination 28