Jessica Minion, MD PGY-5 Medical Microbiology MSc Epidemiology McGill University, Montreal, QC

Transcription

1 Novel Tools for the Detection of Active TB and Drug Resistance TB Diagnostics Symposium Ontario Agency for Health Protection and Promotion Toronto, Ontario October 8 th, 2009 Jessica Minion, MD PGY-5 Medical Microbiology MSc Epidemiology McGill University, Montreal, QC

2 Global TB Case Detection 2.6 million new smear + cases notified in % of the estimated 4.1 million cases 5.3 million new cases overall notified in % of the estimated 9.3 million cases WHO Report 2009 Global Tuberculosis Control

3

4 Microscopy 1882 Culture 1882 Chest X-ray

5 Thanks to a resurgence of interest (along with massive funding)

6 ... and advances in basic science Omics (genomics, proteomics, etc) Immunology Molecular Biology Biotechnology Nanotechnology

7 There is now a promising diagnostics pipeline Pai et al. Sem Resp Crit Care Med 2008

8 Technology needs for Patient-Centered Care Reference Lab Surveillance Regional Lab 5 % Fraction of patients seen Faster than solid culture 10 % LJ - 40d Auto. LC 15d Manual 15d LPA 1d More sensitive than microscopy Peripheral Lab 25 % Smear - 60% LED +10% Manual NAAT +25% Auto NAAT +40% Clinic / Health Post Simpler than microscopy 60 % Symptoms AG/AB Molecular VOC Source: Foundation for Innovative New Diagnostics (FIND)

9 NEW TECHNOLOGIES

10 Smear Microscopy Simple, cheap High specificity Poor sensitivity Especially with HIV coinfection Especially with extrapulmonary TB

11 Optimizing Smear Microscopy Lancet Infect Dis 2006 Lancet Infect Dis 2006 IJTLD 2007

12 New Policies on Smear Microscopy Front Loaded Specimen Collection? Sputum Processing? LED Fluorescent Microscopy?

13 Fluorescent LED Microscopy Higher Sensitivity than ZN (and possibly conventional FM) 46% time savings vs. ZN (equivalent to conventional FM) Advantages of FM but less expensive, requires less maintenance, no need for a dark room Minion et al. unpublished

14 Commercial LED Microscopes Minion et al. Exp Rev Med Dev 2009

15 Challenges for current EQA program 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% month 0 month 1 month 2 month 3 month 4 month 5 Current External Lab QA program involves sending selected slides into SNRL for rechecking quarterly 3+ pos 2+ pos 1+ pos scant + ABS 100% Auramine-stained slides will be faded and unreliable % Initially Positive Slides Read Positive 90% 80% 70% 60% 50% 40% 30% 20% 10% RT_month INC_month FRG_month Minion, et al. unpublished 0% month 0 month 1 month 2 month 3 month 4 month 5

16 Other Potential Innovations

17 Culture High Sensitivity Slow Turnaround Time, Expensive, Biosafety concerns Push towards liquid culture systems Partly driven by HIV epidemic Improves some problems while exacerbating others

18 New Policies on Culture-based Diagnostics 2007

19

20 New Policies on Culture-based Diagnostics MODS? TLA? NRA? 2009 CRI? Phage?

21 Microscopically Observed Drug Susceptibility (MODS) Direct inoculation of patient specimens detection & DST Liquid culture more sensitive than LJ Microcolony detection faster turnaround time Biosafety? Specificity of ID?

22 Thin Layer Agar (TLA) Direct inoculation of patient specimens detection & DST Solid media easier to manipulate Microcolony detection faster turnaround time Biosafety? Specificity of ID?

23 Nitrate Reductase Assay (NRA) Based on MTB s ability to reduce nitrate to nitrite Simple Sensitive detection of small amount of metabolic biproduct improves turnaround time Prevalence of nitrate reductase negative strains of MTB? KNO 3 - containing media Add reagent to drug-free slant day 7 (repeat day 10, 14) Color development = growth

24 Colorimetric Redox Indicators (CRI) Based on reduction of indicator by growing MTB MIC determination using microdilution Detection of active metabolism improves turnaround time Biosafety concerns? Suitable for reference labs? Incubate microdilution plate 7 days Add indicator to all wells, incubate overnight Color change = growth

25 Mycobacteriophage Assays (FASTPlaque TM ) Based on amplification of phage viruses in live MTB 2 day turnaround for detection & DST Minimal biosafety concerns BUT... High rates of contaminated or uninterpretable tests High rate of false positives

26 GRADE Summary Table Novel Culture Methods Diagnostic (Reference) # Studies (# Participants) Pooled Accuracy Estimates from Meta-Analyses Turnaround Time Contamination Rates Quality of Evidence Costs Resources Sens Spec MODS (1) 9 studies (1474 participants) days 6.3% Moderate Equipment: ++ Consumables: ++ Training: extensive Infrastructure: ++/+++ TLA (1) 3 studies ( days 11.8% Low Equipment: + Training: extensive participants) Consumables: ++ Infrastructure: ++/+++ Phage 12 studies FASTPlaque (2) (2945 participants) CRI (3) 31 studies (2498 participants) NRA (4) 19 studies (2304 participants) WHO-endorsed rapid test for DST (for comparison) days 20.4% Very Low Equipment: ++ Consumables: days 5% Moderate Equipment: + Consumables: days 4.8% Moderate Equipment: + Consumables: ++ Training: moderate Infrastructure: ++/+++ Training: extensive Infrastructure: +++ Training: moderate Infrastructure: ++/+++ LPA (5) 12 studies (4937 participants) days Moderate Equipment: +++ Consumables: +++ Training: moderate Infrastructure: ++/+++ Minion et al. unpublished

27

28 NAAT High specificity and PPV Sensitivity is lower and highly variable Especially in extrapulmonary Especially in smear negative reduced utility in HIV + Expensive limited applicability in developing countries

29

30 2009 Updated CDC Guidelines

31 Line Probe Assays Detection of MTB & RIFresistance (rpob) Requires extraction, amplification Colorimetric development using immobilized probes Inno-LiPA Rif.TB assay Innogenetics, Belgium Expensive Feasible in peripheral settings? GenoType MTBDRplus assay Hain Lifescience GmbH, Germany

32 New Policy on Line Probe Assays 2008

33 New Initiatives to Expand Access

34 Loop Mediated Isothermal Amplification (LAMP) Simplified NAAT, does not require a thermocycler, detection by fluorescence Rapid (1 hour) High throughput Sensitivity 97%, Specificity 99% (culture reference) Feasible in high burden settings? Eiken Chemical Co., Tokyo, Japan

35

36 Serology Attractive... especially if point of care (POC) option >80 antigenic targets evaluated and several commercial assays developed All existing serologic tests have failed to demonstrate adequate accuracy Although still marketed and sold by many companies and used in developing countries! Thorax 2007 PLoS Medicine 2007 Clin Vaccine Immunol 2009

37 WHO/TDR evaluation of 19 commercial serologic tests for TB: poor accuracy WHO/TDR Diagnostics Evaluation Series 2009

38 Antigen Detection Urinary Lipoarabinomannan (LAM) ELISA-based test Clearview (R) TB (Inverness, UK) Optimal specimen, rapid turnaround (2.5 hrs) Potential for POC dipstick Initially evaluations were promising Boehme et al. 2005: 80% sensitivity; 99% specificity BUT... Subsequent studies have failed to demonstrate similar performance Indicated for HIV+? Improved sensitivity with low CD4?

39 Disappointing LAM performance Minion et al. Unpublished

40 Challenges for diagnostic research Avoiding Optimism Bias Inappropriate study designs Inappropriate study populations Transferability and Reliability Selective reporting of results Unjustified positive conclusions Publication Bias Rapid licencing, commercialization and marketing of tests before adequate evaluation

41 If exposure and disease are not associated False positive study Hot topic Bias Publication Bias 100 studies will be designed If = 0.05 THE FALSE POSITIVE RESEARCH CYCLE (Choi, 1998) 5 studies show false positive results Positive results bias 5 studies will be published Likely to be meta-analyzed Editor s bias Courtesy: Bernard Choi, PHAC

42 Challenges for diagnostic research Ensuring Quality Bias due to inappropriate reference standards Spectrum bias Verification bias Partial verification Differential verification Review bias (lack of blinding) Incorporation bias

43 Reporting Standards & Quality Assessment Tools

44 Challenges for diagnostic research Moving Beyond Accuracy accuracy is a surrogate for patientimportant outcomes Results in low quality evidence for policy recommendations

45 Blueprint for new diagnostics development Demonstration of test accuracy will not be enough Stop TB Partnership s New Diagnostic Working Group 2009

46 Stop TB Research Movement Landscape of TB Diagnostic Research: PRELIMINARY RESULTS Majority of TB diagnostic studies are focused on accuracy Abstracts included in the preliminary results 3922 Abstracts available 3230 (82%) Original Study 3129 (97%) Abstracts excluded: No abstract available 692 (18%) Abstracts excluded: Not original study 101 (3%) 1. Underpinning Research 302 (10%) 2. Aetiology 1028 (33%) 3. Prevention of Disease and Conditions, and Promotion of Well-Being 120 (4%) 4. Detection, Screening and Diagnosis 420 (13%) 5. Development 6. Evaluation of 8. Health and 7. Management of Treatments Treatments and Social Care Of Diseases and Therapeutic Therapeutic Services and Conditions Interventions Interventions Research 232 (7%) 202 (7%) 162 (5%) 104 (3%) 9. Case Reports or Case Series 559 (18%) 1.1Biological 272 (91%) 1.2 Psycological, Socioeconomic 0 (0%) 1.3 Chemical, Physical 0 (0%) 1.4 Methodology 23 (8%) 1.5 Ressources 3 (1%) 2.1 Biological 319 (31%) 2.2 Environmental 195 (19%) 2.3 Psychological, social, economic 65 (6%) 2.4 Distribution 402 (39%) 2.5 methodology 49 (5%) 2.6 Ressources 2 (0.2%) 3.1 Behaviour 0 (0%) 3.2 Environmental 11 (9%) 3.3 Nutrition, Chemoprevention 0 (0%) 3.4 Vaccines 109 (91%) 3.5 Ressources 0 (0%) 4.1 Preclinical 32 (8%) 4.2 Evaluation 372 (89%) 4.3 Impact 4 (1%) 4.4 Population Screening 2 (0.5%) 4.5 Resources 2 (0.5%) 4.6 Cost 8 (2%) 7.1 Individual care needs 96 (41%) 7.2 End of life Care 1 (1%) 7.3 Management 133 (57%) 7.4 Ressources 2 (1%) 8.1 Organisation Delivery of Services 54 (52%) 8.2 Health, welfare Economics 15 (14%) 8.3 Policy, ethics Research Governance 34 (33%) 8.4 Methodology 1 (1%) 8.5 Ressources 0 (0%)

47 In conclusion, much progress has been made in improving diagnosis but...

48 Are we close to the big goal of a simple point of care test?

49 What specifications should a POC test for TB have?

50 Revolution through competition!

51 Revolution through competition! X Prize May be Announced for a TB Diagnostic Test!

52 Acknowledgements Madhukar Pai Fran Jaimeson Shana Kelley