Professor Ajit Lalvani

Transcription

1 Diagnosing TB in the 21 st Century: New Tools to Tackle an Old Enemy Professor Ajit Lalvani, Department of Respiratory Medicine, National Heart and Lung Institute, Faculty of Medicine, St Mary s s Campus, Imperial College London

2 We are fighting a large & growing epidemic with old tools Diagnostics TST Sputum microscopy Chest X-ray s 1920 Vaccine BCG 1921 Treatment Streptomycin Isoniazid Pyrazinamide Ethambutol Rifampin

3 Comstock: TB is an infectious disease with an incubation period from a few weeks to a lifetime Flow Chart: The natural history of M. tuberculosis infection No infection Exposure to M. tuberculosis Infection? Transient infection with clearance TB control strategies are based on the biology and natural history of TB infection Primary TB in the first 1-2 years after infection 5% 95% Latent TB infection Long-term immune control Treat active TB BCG vaccinate Treat LTBI 5% 90% Reactivation TB Lifelong containment

4 Factors promoting progression to active disease Flow Chart: The natural history of M. tuberculosis infection No infection Exposure to M. tuberculosis Infection 5% 95%? Transient infection with clearance Primary TB in the first 1-2 years after infection Latent TB infection Long-term immune control 5% 90% Reactivation TB Lifelong containment

5 Targeted tuberculin testing & treatment of LTBI Focus on high risk-groups who are at increased risk of progression to active TB: Recent infection contacts of TB cases Impaired cellular immunity HIV Children, especially < 5yrs old Concomitant illnesses: renal failure, diabetes, etc Iatrogenic immunosuppression Steroids, organ transplants, methotrexate, ant-tnf

6 The TST is an inadequate tool to achieve the aims of targeted tuberculin testing for LTBI Paradoxically, poor sensitivity in exactly those vulnerable groups targeted by the strategy Poor specificity: Over half the burden of TB is in BCG-vaccinated immigrants as prevalence falls in low-prevalence countries, an increasing proportion of positive TST results will be false-positives These problems apply equally to patients with suspected active TB Problems of in vivo test. Need for return visit Operator variability (inoculation & reading) Standardisation of reagent Painful inflammation & scarring

7 M. tb infection activates many T cell subsets Can we exploit MTB-specific T cells as a marker of infection? Th1-type CD4+ T cells Regulatory T cells Guyot-revol et al AJRCCM 2006 IFN-γ CD8+ T cells Lalvani et al PNAS 1998 CD1 restricted T cells Stenger et al Science 1997 Moody et al Nature 2000 γδ T cells

8 Counting T cells quickly: ELISPOT (Enzyme-linked immunospot for interferon-gamma) Detects IFN-gamma release at the single cell level Each spot is the footprint of a single cell that has reacted to the antigen Very sensitive (detects 1:50,000 antigen-specific T cells) Rapid (ex vivo) Overnight incubation detects effectormemory T cells non-radioactive sterile tissue culture facilities not required

9 Comparative genomics identified regions of difference between BCG and M. bovis/m. tuberculosis The RD1 genomic segment of M. tuberculosis is absent from all strains of BCG and most environmental mycobacteria

10 RD1 encodes the two strongest T cell antigens of M. tuberculosis: ESAT6 and CFP10 = RD1 = ESAT6 & CFP10 M. bovis M. tuberculosis RD1 was lost from BCG early on RD1 is present in all strains of M. tuberculosis BCG BCG BCG M. tuberculosis

11 Hypothesis ESAT-6 and CFP10-specific T cells are an accurate marker of M. tuberculosis infection ESAT6 CFP10

12 Validated in >5,000 people in head-to-head comparisons with tuberculin skin test High, medium & low burden regions Populations with high HIV prevalence Adults & children

13 IFN-gamma-secreting T cells are a good marker of active TB infection Culture-confirmed TB cases vs patients with non-tuberculous illnesses (n>400) High sensitivity: 90-97% (significantly more sensitive than TST) Not confounded by other diseases Detects extrapulmonary as well as pulmonary TB Lalvani et al, Am J Resp Crit Care Med 2001 Lalvani et al, J Inf Dis 2001 Pathan et al, J Immunol Meier et al Eur J Clin Micro Inf Dis 2005 Chapman et al, AIDS Scarpellini et al J Clin Micro 2004 Liu et al, Infect Immun Richeldi et al, Lancet 2006 Liebeschuetz et al, Lancet Mantegani et al Clin Med Res 2006 Richeldi et al, Ann Intern Med Goletti et al, Clin Microbiol Infect 2006 Gooding et al, J Infect 2006

14 IFN-gamma-secreting T cells are a good marker of latent TB infection (LTBI) Most TST-positive household contacts of smear-positive TB are ELISpot-positive (Am J Resp Crit Care Med, 2001) Without a gold standard reference test, it is not possible to directly quantify the sensitivity and specificity of new blood tests or the TST MTB transmission is promoted by prolonged and close contact with an infectious case: key determinant of infection is amount of time spent sharing room air with source case If the ELISpot is a more sensitive and specific test, it should correlate more closely with level of exposure to MTB than TST, but should be independent of BCG status.

15 Lancet 2003

16 TST, but not ELISpot, was confounded by prior BCG vaccination (14 years later previously) ELISpot is more specific than TST for detecting LTBI

17 IFN-gamma-secreting T cells are a better marker of LTBI than TST ELISpot correlates better with TB exposure and is independent of BCG vaccination status (n>1,000) Lalvani et al, Lancet 2001 Ewer et al, Lancet 2003 Richeldi et al, Am J Resp Crit Care Med 2004 Shams et al, Am J Resp Crit Care Med 2005 Zellweger et al, Int J Tub Lung Dis Diel et al, Eur Resp J 2006

18 Regulatory-approved in vitro diagnostic test: T-SPOT.TB T-SPOT TB is: Based on Lalvani ELISpot (same assay, same peptides) Complete system - kit + instrumentation Validated and produced to international quality standards Standardized Quality-controlled CE marked for in vitro diagnostic use

19 Use blood tests, if available, in: - Individuals who have tested positive by TST - Individuals in whom TST is unreliable 2006 NICE reviewed 11 studies published before July 2005: 6 papers used ELISpot, 5 used ELISA. They will revise the recommendations if new significant evidence emerges ALSO ENDORSED BY NATIONAL GUIDELINES IN: Italy, Switzerland, Canada

20

21 Diagnostic sensitivity in active TB: ELISpot vs ELISA Immunocompetent Adults TIGRA Setting n Sensitivity (%) Reference(s) ELISpot Prospective Kang et al Chest 2007 High burden country Wang et al EID 2007 Low burden country:routine clinical practice Ferrara et al Lancet 2006 High burden country:case control Lee et al ERJ 2006 Low burden country:case control Meier et al EJMID 2005 Low burden country:case control (ESAT-6) Lalvani et al AJRCCM 2001 Low burden country:case control (ESAT-6) Pathan et al JI (83 to 97%) QFT-G Low burden country Dewan et al CID 2007 Prospective Kang et al Chest 2007 High burden country Pai et al Infection 2007 Low burden country: prospective multicentre cross-sectional Mazurek et al CID 2007 Case control Kobashi et al CID 2006 Prospective Ferrara et al Lancet 2006 Prospective Lee et al ERJ Ferrara et al AJRCCM 2005 Case control Kang et al JAMA 2005 Prospective Ravn et al Clin Diagn Lab Immunol 2005 Case control Mori et al AJRCCM (64 to 89%) QFT-IT Prospective Pai et al Infection 2007 High burden country Tsiouris et al J Clin Microbiol (73 to 76%) ELISpot has higher diagnostic sensitivity than QFT-G and QFT-IT and both test formats have higher diagnostic sensitivity than the TST

22 Diagnostic sensitivity in active TB: ELISpot vs ELISA Immunocompromised Adults TIGRA Setting n Sensitivity (%) Reference(s) ELISpot High burden country: case control HIV-positive Chapman et al AIDS 2002 Prospective Clark et al Clin Exp Immunol QFT-IT High burden country: HIV-positive Tsiouris et al J Clin Microbiol 2006 TIGRAs more sensitive than TST and ELISpot more sensitive than QFT in active TB with HIV More data required on all tests in immunocompromised populations Children TIGRA Setting n Sensitivity (%) Reference(s) ELISpot Low burden country Detjen et al CID 2007 High burden country: case control Nicol et al CID 2005 Routine clinical practice: kwazulu-natal Liebeschuetz et al Lancet (70 to 93%) QFT-G Prospective Connell et al Thorax 2006 QFT-IT Low burden country Detjen et al CID 2007 Prospective 8 63 Dogra et al J Infect (63 to 93%) ELISpot is more sensitive than TST and QFT for diagnosis of active TB More data required on all tests in children

23 Clinical utility in active TB TB Infection Asymptomatic: LTBI Symptomatic: Active TB Uninfected Immune-based tests are tests of infection, not active disease and cannot yet distinguish between the two TB infection is a prerequisite for TB disease But by ruling out infection, you can ruleout active TB Requires test of high diagnostic sensitivity, e.g. ELISpot (Lancet 2004) Age-related prevalence of LTBI means that negative results rare in adults but more common in children

24 Diagnosis of LTBI: ELISpot vs ELISA Immunocompetent Adults TIGRA Setting n Reference(s) ELISpot Contact tracing 785 Arend et al AJRCCM 2007 Screening 390 Porsa et al Clin Vaccine Immunol 2007 Institutional outbreak: contact tracing 88 Zellweger et al IJTLD 2005 Contact tracing 413 Shams et al AJRCCM 2005 Institutional outbreak: contact tracing 24 Richeldi et al AJRCCM 2004 Contact tracing 50 Lalvani et al Lancet 2001 TB suspects 393 Ferrara et al Lancet QFT-G Screening 856 Mazurek et al CID 2007 Case control 273 Kang et al JAMA QFT-IT Contact tracing 785 Arend et al AJRCCM 2007 Contact tracing 190 Kobashi et al Intern Med 2007 Cross-sectional study 726 Pai et al JAMA ELISpot and QFT correlate with TB exposure More data on ELISpot ELIspot correlates better with TB exposure than TST, suggesting higher sensitivity

25 Clinical utility in LTBI Depends crucially on how tests perform in high risk groups predisposed to false-negative TST results due to impaired cellular immunity...and who are at highest risk of progression to TB Flow Chart: The natural history of M. tuberculosis infection No infection Exposure to M. tuberculosis Infection 5% 95%? Transient infection with clearance - HIV - Children, especially < 5yrs old Primary TB in the first 1-2 years after infection Latent TB infection - Concomitant illness (renal failure, diabetes etc) - Iatrogenic immunosuppression Long-term immune control 5% 90% Reactivation TB Lifelong containment

26 Diagnosis of LTBI: ELISpot vs ELISA Children TIGRA Setting n Reference(s) ELISpot Institutional outbreak: contact tracing 92 Richeldi et al AJRCCM 2004 Institutional outbreak: contact tracing 535 Ewer et al Lancet 2003 Cross-sectional 718 Hill et al Pediatrics 2006 Contact tracing 41 Richeldi et al AJRCCM 2004 Prospective community-based 979 Soysal et al Lancet QFT-G Prospective 101 Connell et al Thorax 2006 Cross-sectional screening 75 Andersen et al AJRCCM 2006 Contact tracing 125 Brock et al AJRCCM QFT-IT Convenience sample 184 Tsiouris et al IJTLD 2006 Case control 270 Nakaoka et al EID ELISpot and QFT correlate with TB exposure More data on ELISpot, including infants ELISpot correlates better with TB exposure than TST suggesting higher sensitivity

27 Diagnosis of LTBI: ELISpot vs ELISA Immunocompromised Adults TIGRA Setting n Reference(s) ELISpot Cross-sectional study: HIV+ 160 Rangaka et al AJRCCM 2007 Immunosuppressed haematology patients 138 Piana et al ERJ 2006 Hemodialysis patients 203 Passalent et al Clin J Am Soc Nephrol QFT-G Cross-sectional study 160 Rangaka et al AJRCCM QFT-IT 294 Luetkemeyer et al AJRCCM 2007 HIV 590 Brock et al Resp Res 2006 Patients treated for rheumatic conditions 142 Matulis et al Ann Rheum Dis ELISpot and QFT appear more sensitive than TST for diagnosis of LTBI ELISpot but not QFT is independent of CD4 count More data required on all 3 test in immunocompromised populations

28 Utility in LTBI HIV Young kids LTBI Other immunosuppression Uninfected BCG vaccinated High specificity of both tests will lessen the number of people inappropriately treated on the basis of false-positive TST results caused by prior BCG vaccination Rule-in as well as rule-out test High sensitivity in young children and immunosuppressed populations, as evidenced for ELISpot, allows more accurate targeting of latently infected people at highest risk of progression

29 What have we learnt about natural history of MTB infection by using the new tools? Transient T cell responses in TST-negative contacts with low TB exposure suggest Self-resolving infection Ewer et al, AJRCCM 2006

30 Clinical practice in context of European guidelines Dichotomy between USA (replace TST with TIGRA) and NICE: TIGRA to confirm LTBI in those who have tested positive by TST Direct use of TIGRA for those in whom TST is unreliable Rationale: cost effectiveness analysis Consequence for practice: excellent Clinical significance of TST-negative, TIGRA-positive result: unknown. Small proportion of recent TB contacts who are TST-negative, TIGRApositive at screening subsequently turn TIGRA-negative: they may have self-resolving infection, i.e. no forward risk of progression to TB disease (Ewer, Millington et al AJRCCM 2006; Editorial: Here today, gone tomorrow: the case for transient TB infection) NICE: uninfected immunocompetent contacts will not be treated However, high sensitivity of ELISpot, and its robustness to immune suppression, mean that people at risk of LTBI with suppressed or immature (i.e. infants) immune systems who test TST-negative but ELISpot-positive are likely to have false-negative TST results and should be considered infected. High risk of progression to TB in any case requires a low threshold for treatment of LTBI. NICE: Infected, immunosuppressed (or child) contacts will be treated

31 What evidence is required to completely replace TST with TIGRAs? Only a small proportion of TST-positive TB contacts progress to active TB (5% over 2-5 y); most of those we treat would not in any case have progressed to TB, but there is still a net treatment benefit at the population level Therefore, just as with TST, there will be TIGRA-positive contacts who will benefit from preventive therapy and others who will not. Key determinant: what proportion of TIGRA-positive contacts will benefit from preventive therapy? i.e. what is the overall population risk of progression to TB disease in TIGRA-positive contacts? If longitudinal studies with clinical outcomes (i.e. development of TB) show a definitive risk of progression, this implies a net treatment benefit at the population level and provides a clear mandate for completely replacing TST with TIGRA In the meantime, NICE guidelines are an excellent compromise until prognostic value of TIGRAs in recent TB contacts is elucidated. Europe is following NICE

32 Does a positive ELISpot result predict progression to active TB? Asymptomatic newborn son of index case Heavily exposed to the mother while she was infectious for 1 week at home before her diagnosis Chemoprophylaxis with ETB and PZN for 9 months TST and ELISPOT 3, 6, 12, 18 and 24 months after exposure Pediatrics 2007

33 ESAT6 CFP10 D ice (O p t :1.0 0 % ) (T o l 2. 0 % -2. 0% ) (H > 0.0% S > 0.0 % ) [ 0. 0 % % ] IS R F L P 100 I S R F L P M y c ob ac te riu m tub er c ul os is 31 4 M o d en a M y c ob ac te riu m tub er c ul os is 20 1M od en a RFLP analysis of MTB isolate from bronchoalveolar lavage: identical strain to that of the mother Pediatrics 2007

34 M. tb-specific IFN-γ ELISpot response declines with treatment Pathan et al. JI 2001 Lalvani et al. JID 2001 Millington et al. JI 2007 Size IFN-γ response Follow-up 8m No measure of bacterial load in TB Follow-up 5m Since effector cells measured by ELISpot are in dynamic equilibrium with MTB antigen load in vivo, they might be a useful marker of bacterial load Follow-up 28m Big variation in rate of decline Minority are negative by end of therapy Many remain positive long after treatment completion

35 IFN-g ELISpot remains positive after treatment of active TB in a substantial proportion of patients 2000 ESAT CFP-10 ESAT-6-specific IFN-g SFC / 10 6 PBMC CFP-10-specific IFN-g SFC / 10 6 PBMC Follow up, months Follow up, months Millington et al, submitted (invited revisions in progress)

36 The future of T cell-based monitoring of TB.. Active TB During & after treatment

37 IFN-g/IL-2 functional signatures in antiviral CD4 T cell immunity Acute Ag exposure Ag re-exposure High antigen load 1 o CMV, 1 o HIV-1, progressive HIV-1 Ag load Protracted Ag exposure Low antigen load Chronic EBV, HSV, HCV, CMV, HIV-1 in LTNP Time IFN-γ IFN-γ/IL-2 IL-2 Ag clearance Influenza Pantaleo & Harari 2006 Nature Reviews Immunology

38 Relationship between IFN-γ/IL-2 profile & Ag load Viral infections Tuberculosis Tetanus toxoid IL-2: central memory Ag load cleared Total cytokine-secreting CFP-10-specific CD4 T cells in the enriched fraction (%) Follow up, months IFN-γ IFN-γ/IL-2 IL-2 antibiotic treated Participant number B141 B41 B13 B49 B36 Chronic CMV IL-2/IFN-γ Ag persistence acute HIV Ag levels high Total cytokine-secreting CFP-10-specific CD4 T cells in the enriched fraction (%) IFN-γ: effector memory NPH7 NPH29 NPH40 B133 Participants IFN-γ IFN-γ/IL-2 IL-2 active disease

39 Evaluation of risk factors for M. tb infection using a more accurate tool than TST Hypothesis Because ELISpot distinguishes between M.tb infection and BCG vaccination, ELISpot can be employed to assess whether BCG vaccination affects the risk of infection Aim Investigate risk factors for TB infection in 979 children from Istanbul, Turkey Study Prospective community-based study of child contacts recently exposed to tuberculosis in the household Detect infection with ELISpot and TST 443 index patients with sputum smear-positive pulmonary TB recruited at 7 TB clinics in Istanbul 1024 child HH contacts enrolled at central study clinic 45 contacts excluded Study population: 979 child contacts 414 index patients 395 households (2.5 child contacts per HH)

40 Independent risk factors (multivariate analysis) Increasing number of index patients in HH ELISpot P=0.006 TST P=0.017 Being the child of the index patient ELISpot P=< TST P=< Increasing age from 0-1y 2-5 ELISpot P=0.001 TST P= ELISpot P=< TST P=0.001 Absence of BCG scar ELISpot P=0.003 TST (10mm) P=0.33

41 Conclusions & Implications Presence of a BCG scar associated with a 24% reduction in risk of being infected This protective effect not seen when infection measured with TST The BCG vaccine can prevent infection as well as prevent progression to disease Now a 3-pronged approach to TB control Exposure Infection Disease BCG vaccination BCG vaccination or preventative treatment Treatment

42 WE_logotype_2col_CMYK.jpg TB Immunology Group Imperial College London Kerry Millington Suzie Hingley-Wilson Muhunthan Thillai Deena Blumenkrantz Louisa Gnatiuc Davinder Dosanjh Tim Hinks Valerie Guyot-Revol Heartlands Hospital, Birmingham John Innes Sarah Hackforth University of Birmingham Cancer Research UK Damien Montamat-Sicotte Ben Wilcox Public Health & Epidemiology Jon Deeks Veterinary Laboratories Agency, Weybridge Martin Vordermeier Northwick Park Hospital, Harrow Geoffrey Pasvol Hansa Varia University of Oxford Paul Klenerman University of Texas, Tyler Peter Barnes Marmara University, Istanbul Mustafa Bakir Ahmet Soysal MRC, Harare & LSHTM Liz Corbett Anthony Butterworth Policlinico di Modena Luca Richeldi Leo Fabbri King George V Hospital & University of KZN, Durban Instituto Nacional de Medicina Genomica Irma Aguilar CMC, Vellore C Gnanamuth Hinduja Hospital, Mumbai Zarir Udwadia Camilla Rodrigues

43 WE_logotype_2col_CMYK.jpg TB Immunology Group Imperial College London Kerry Millington Suzie Hingley-Wilson Muhunthan Thillai Deena Blumenkrantz Louisa Gnatiuc Davinder Dosanjh Tim Hinks Valerie Guyot-Revol