Richard E. Chaisson, MDCenter for Tuberculosis Research

Johns Hopkins University

What Can Be Done to Control Tuberculosis?

A History of TB Control: Mission Accomplished?

• 1882 – Koch discovers the tubercle bacillus*• 1907 – von Pirquet adapts Koch’s tuberculin• 1919 – Calmette and Guerin produce BCG vaccine• 1943 – Schatz and Waksman discover streptomycin*• 1948 – BMRC trial of streptomycin vs bed rest• 1952 – Development of INH• 1966 – Development of rifampin• 1978 – Short-course TB therapy 6 months

*Awarded Nobel Prize

Federal Funding for Tuberculosis Research and Control, 1962 - 1990

BR Bloom and CJL Murray, Science 1992;257:1055-64

Centers for Disease Control and PreventionReported Tuberculosis in the United States, 1999

Reported Tuberculosis Cases in the United States, 1953 - 1999

Year of notification1950 1960 1970 1980 1990 2000

Num

ber o

f cas

es (l

og s

cale

)

20000

40000

80000

Estimated Global Incidence of Tuberculosis, 1990 - 2005

120

125

130

135

140

1990 1995 2000 2005

Cases per 100,000

WHO 2007

WHO Estimates of Global Burden of Tuberculosis as of December 31, 2008

Estimated number of

cases

Estimated number of

deaths

1.3 million*9.4 million

~150,000511,000

All forms of TB

MDR-TB

XDR-TB 50,000 30,000*excludes 500,000 HIV-TB deaths

TB Incidence in Africa, 1990 and 2005

Chaisson and Martinson, N Engl J Med 2008;358:1089

XDR TB outbreak in Tugela Ferry, South Africa

Days since Sputum Collected

2402101801501209060300

Prop

ortio

n Su

rviv

ing

1.1

1.0

.9

.8

.7

.6

.5

.4

.3

.2

.1

0.0

-.1

Gandhi et al., Lancet 2006; 368:1575-80

Survival of XDR TB Patients at the Church of Scotland Hospital (N=53)

TB in HIV+ Patients at Chris Hani Baragwanath Hospital, Soweto, South Africa

– TB admissions 2005-2008:• 6500-6800 per year• 18-20 new cases per day

– 85% HIV+– >90% fully drug susceptible– Inpatient mortality = 18%

• ~1200 deaths per year from drug-susceptible TB

Edgington et al., Int J Tuberc Lung Dis 2006;10:1018; ML Wong, pers. comm.; Martinson et al., AIDS. 2007;21:2043 Shah et al., JAIDS 2009, epub

Tools to Control of TuberculosisWhat went wrong?

• Failure to apply tools broadly– Weaknesses in health systems

• Inadequacies of existing tools– Smear detection of cases ~50%– Adherence to regimens is very poor – BCG vaccine does not prevent adult TB

• Changing epidemiological situation– HIV epidemic and other co-morbidities– MDR

• Lack of understanding of best epidemiologic approaches

Step 3

Step 4

Step 1

Step 2

UNINFECTED

INFECTED

CASES

An Epidemiological Model of Tuberculosis in the United States

Infected

Cases

Ferebee, SH. Natl Tuberc Assoc Bull 1967;53:4

Step 3

Step 4

Step 1

Step 2

UNINFECTED

INFECTED

CASES

An Epidemiological Model of Tuberculosis in the United States

Ferebee, SH. Natl Tuberc Assoc Bull 1967;53:4

Interventions to control TB

Find and treat cases: Steps 2 and 4

Treat latent TB: Steps 1 and 3

Vaccinate susceptible: Step 2

The Kolin Study of Tuberculosis Control with Mass Case Finding and Treatment

“It is concluded from the study that in developed countries priority should be given to adequate treatment of all persons with active tuberculosis, and to early diagnosis in persons consulting physicians and in the high-risk populations.”

Type 1960 1961 1962 1963 1964New Cases 46 132 61 88 47Relapses 29 25 18 13 16

Chronic 29 40 17 11 8

Other 26 26 42 14 20Total 150 233 138 126 91

Styblo et al., Bull WHO 1967;37:819-74

Rates per 100,000 Population

Mass case finding

World Health Organization Strategies for Tuberculosis Control, 1974 – present

• Passive case detection and treatment– DOTS targets added in 1991

• 70% case detection, 85% cure rate

• BCG vaccination of all children at birth• Isoniazid preventive therapy for young

children exposed to smear-positive cases (PPD+)

Epidemiologic Basis for TB ControlKey Considerations for Strategies

• Where are the seedbeds of tuberculosis?– Who has latent TB infection?– Who amongst these is most likely to develop disease?

• Who has active TB and how can they be reached?– What proportion of cases are detected, and when?– Is treatment effective in controlling spread?

• Where is TB transmission occurring, and how can it be curtailed?– Who are the contacts of cases who become infected?– What measures can be taken to reduce transmission?

• What can be done to reduce susceptibility?– Vaccination– Antiretrovirals– Control of co-morbidities, e.g., diabetes

The Origin of TB Cases:Prevalence of Risk Factors in Patients with Culture-Confirmed

Pulmonary TB in Baltimore

Characteristic No. (Total = 139) %

Foreign born 12 9%

HIV Infection 31 24%

IDU 28 20%

Diabetes 18 14%

Renal Failure 12 9%

Recent Cancer 8 6%

Steroid Use 7 6%

Oursler et al., CID 2002;34:729-9

TB Incidence and Prevalence of Diabetes, 2010 and 2030

Dooley and Chaisson, Lancet Infect Dis, 2009; 9: 737–46

Smoking and incident TB in HIV-infected adults in Soweto, South Africa

Pack years Incidence IRR

<16.5

(5.9-7.4)REF

1-59.0

(6.8-11.7)1.36

(1.01-1.82)

>512.8

(9.7-16.7)1.95

(1.44-2.60)

Pack years Incidence IRR

Never6.7

(6.0-7.6)REF

Past7.8

(5.8-1.2)1.15

(0.9-1.55)

Current10.7

(8.4-13.4)1.59

(1.21-2.05)

Martinson et al CROI 2008

Biomedical Tools and Public Health

Control of infectious diseases requires:- effective biomedical tools (diagnostics, drugs and vaccines) and - effective public health strategies for applying and utilizing the tools at the population level to reduce disease burden

Current Tools for Controlling TB

Target Available tools

Diagnostics Sputum smear, culture, x-rays, (molecular assays)

Drugs Isoniazid, rifampin, PZA, ethambutol, 2nd line drugs

Vaccine BCG (>10 strains)

Current Strategies for Controlling TB

• Passive case finding (DOTS)– Relying on sputum smear (~50% sensitivity)– No drug susceptibility testing– Most cases not diagnosed or effectively treated

• INH preventive therapy– Rarely used outside US and Europe

• BCG Vaccination– Most widely used vaccine, but not effective

• Infection control– Little attention has been paid to controlling

transmission

A Platform for Controlling Global Tuberculosis

• FIND the TB that is there– Passive case detection is not sufficient

• TREAT the TB that is found– Treatment success is unacceptably low– Treatment for M/XDR is abysmal– New drugs and treatment strategies urgently needed

• Prevent the TB that hasn’t occurred yet– Preventive therapy essential for high risk populations– Infection (transmission) control critical– Control susceptibility (antiretrovirals, diabetes control)– New vaccine essential

FIND TB

• Identify TB suspects– Symptomatic screening in health facilities– Campaigns to identify prevalent cases

• Community-based active case finding

• Evaluate TB suspects– Better use of existing technologies – New technologies

Estimated TB case detection rates in 2008, by WHO region

78 7870

6557

47

0102030405060708090

America

s

Europe

Wester

n Pacific

South East A

sia EMRAfri

ca

EMR = Eastern Mediterranean Region

Cas

e de

tect

ion

rate

(%)

WHO 2009

Routine detection of TB in HIV-infected patients in Vietnam and Thailand

Diagnosis N detected/N PercentAny TB 147//1060 14%

Pulmonary onlyExtrapulmonary onlyBoth

61/14721/14765/147

42%14%44%

Pulmonary CasesSmear +Liquid culture +

47/126124/126

37%98%

Lymph node aspiratesSmear +Culture +

16/5234/82

31%42%

Monkongdee et al., AJRCCM 2009, epub

Outside the HIV Clinic:Prevalence of Active TB with Case Finding in HIV-Infected

Populations

Population StudiedSetting Time

PeriodPrevalence of Active TB in Population

HIV+ Women in HIV MTCT Program(N=438)

SowetoSouth Africa

2001 13% OF TST+ 3% overall

Patients in HIV Home Care Program (N=441)

Phnom Penh

2001 9% of patients

Adults in HIV VCT Program (N=5000)

Cape Town

2000-1 8% of HIV+

Adult residents of an urban shantytown (N=10,900)

Cite SoleilHaiti

1991-2 6% of HIV+ 2% of HIV–

Nachega 2003; Kimerling 2002; Coetzee 2005; Desourmeaux 1996

Should we consider campaigns to detect prevalent, untreated TB cases?

Impact of mass radiography on TB case detection, incidence and survival

• 1945-1948 – USPHS screened 6 million people in 21 communities with CXR– 85-90% of cases unknown to local health depts.– TB mortality decreased significantly

• 1950’s – USPHS used mass x-ray on ships, trains, airplanes and dog sleds to screen Alaskan natives– Tuberculosis mortality declined

• 665/100,000 in 1950• 116/100,000 in 1957

Golub et al., Active case finding of tuberculosis: historical perspective and future prospects. IJTLD 2005;9:1183

Rocinha favela, Rio de Janeiro

A cluster-randomized trial of door-to-door active case finding for TB in Rio de Janeiro

(14 clusters, 58,587 residents)

Analysis Time Household Case Finding TB incidence

Pamphlet Only

TB incidence

Rate ratio (95% CI)

Intervention only 9.34/1000 py 6.04/1000 py 1.55 (1.10, 1.99)

Intervention plus 60 days

5.16/1000 py 4.93/1000 py 1.05 (0.56, 1.54)

Post-interventionTB incidence

Pre-interventionTB incidence

Rate ratio(95% CI)

Entire Study 4.5/1000 py 3.4/1000 py 1.3 (0.97, 1.77)

Miller et al., IJTLD in press

Availability of culture and drug susceptibility testing in TB/HIV high-burden countries

Country Culture Drug Susceptibility TestingN of Labs N per 5 million N of Labs N per 10 million

South Africa 15 1.5 10 2.1Nigeria 2 0.1 1 0.1Ethiopia 1 0.1 1 0.1DR Congo 1 0.1 1 0.2Kenya 5 0.7 1 0.3Tanzania 3 0.4 1 0.2Uganda 3 0.5 2 0.6Zimbabwe 1 0.4 1 0.7Mozambique 1 0.2 1 0.5Cambodia 3 1.0 1 0.7

WHO Global TB Report, 2009

New TB Diagnostic Tools

• LED fluorescent microscopy

• Liquid culture (e.g. MGIT)

• Capilia TB– Rapid strip test that detects a TB-specific antigen from

culture

• Molecular assays (e.g. Cepheid GeneXpert, Hain GenoType MTBDRplus)– Rapid detection of TB and drug-resistance

Impact of Improving Case Finding and Treatment on Tuberculosis Control: A Mathematical Model

0100200300400500600700800900

1000

0 20 40 60 80 100

Year/Case Detection Rate (%)

TB In

cide

nce,

per

100

,000

0.0%-1.0%

-0.3%

-1.5%-0.3%-1.8%

-0.7%-2.5% -1.0%

-2.9% -1.4%

-3.3%

Hold CDR constant

Increase CDR by 1%/year

Increase CDR by 2%/year

-11.2%-0.8% -0.7%

-0.6%

-2.1%

-2.8%

-3.4%

-4.2%-4.6%

Increase CDR to 70% immediately, then hold constant

22 High burden countries

Dowdy and Chaisson, Bull WHO 2009: 87:296–304

TREAT TB

• Assure treatment completion for all patients

• Prevent the emergence of drug resistance

• Manage co-morbidities – HIV, diabetes, ESRD, IDU

• New drugs to improve therapy– M/XDR TB treatment– Treatment-shortening regimens

Rocinha Family Health Program Community Health Worker Team

Results of DOTS: Health Center-Based versus Community-Based using CHWs

39/355 (11%)

36/66 (55%)

106/134 (79%)

236/289 (82%)

272/355 (77%)

Non-slum Patients

DOTS at Health Center N= 565

16/331 (5%)27/210 (13%)Abandoned

58/70 (83%)27/46 (59%)Cure, retreatment cases

135/153 (88%)82/98 (84%)Cure, new smear + cases

230/261 (88%)138/164 (84%)Cure, new cases

288/331 (87%)165/210 (78%)Cure, all cases

Slum PatientsSlum Patients

DOTS CHWs N=331

Cavalcante et al., Int J TB Lung Dis 2007;11:544-9

New Drugs for TB

• Fluoroquinolones– Moxifloxacin

• Rifapentine• TMC 207 (ATP synthase inhibitor)

– In Phase 2 trials for MDR TB in Africa• Nitroimidazopyrans

– PA-824– OPC-67683

• Phase 2 trial for MDR TB beginning in January 2008• Diamines (SQ-109)

– Phase 1 studies complete, awaiting Phase 2• Oxazolidinones• >12 other new compounds in development

Moxifloxacin vs. Ethambutol as 4th Drug in Initial Phase of TB Therapy:

Culture Conversion by Week

0

20

40

60

80

100

0 1 2 3 4 5 6 7 8

Week of Treatment

% Culture Negative

MOX EMB

0.04

0.0010.0001

0.01 0.010.30

0.140.02

Conde et al., Lancet 2009; 373:1183-9

0 2 4 6 8 10 12 14 16012345678

R10H25Z150R10M100Z150P10M100Z150

Weeks

Log 1

0C

FU p

er lu

ng

2 logs

4 logs

Rosenthal et al., PLoS Medicine 2007;4(12):e344

RIF10 INH PZARIF10 MOX PZARPT10 MOX PZA

Bactericidal activity of daily regimens in mice

TMC 207 for MDR TBCulture conversion at 2 months

p = 0.003

8.7% culture

negative

47.5% culture

negative

Diacon et al., N Engl J Med 2009;360:2397

PREVENT TB

• TB preventive therapy for high-risk individuals without active TB

• Contact evaluation and treatment• Prevention of nosocomial transmission

Isoniazid Preventive Therapy: The (George) Comstock Lode

TB Rates in the 6 Years After 1-year of Treatment with INH or Placebo in the Bethel Trial

02468

10121416

1 2 3 4 5 6

Year After Randomization

TB

per

1,0

00 P

Y

PlaceboINH

GW Comstock, Ferebee SH, Hammes LM.. Am Rev Respir Dis 1967;95:935-43.

Cumulative reduction 5.1% 2.1% = 60%

Efficacy of IPT in HIV+ Adults: Risk of TB

• 11 randomised trials with 8,130 HIV+ participants overall reduction in TB = 36%, reduction PPD+ = 62%

Woldehanna and Volmink, Cochrane Review 2006

0.95

0.64

TB incidence

Death

Relative Risk (Fixed)95% CI

Reference1.0

TB screening, treatment and IPT 2002-2008

59 63 287 3 384

72 26

14

17 10

44

5825

71

77

42 28

63

59

0.0

200.0

400.0

600.0

800.0

1000.0

1200.0

Screened for TB Diagnosed with TB IPT

Thou

sand

s of

pat

ient

s

2002200320042005200620072008

Data for 2008 is preliminary and does not include data from European region

By 2008, 1 out of 4 estimated HIV positive TB patients were identified and put on TB treatment

Infection Control?

New paradigms

• What can be done now to control TB in communities with high burdens of TB?

• How can multiple interventions be combined to yield maximal effects?

C R E A E

MissionTo organize, implement and evaluate novel public health strategies to reduce tuberculosis incidence in populations with high rates of HIV and TB co-infection.

Funded by the Bill and Melinda Gates Foundation

The CREATE Portfolio of Population-Level Studies

Study Intervention Design (N)

Mass TB preventive therapy for S.A. gold

miners

Cluster randomized trial

(~60,000)Intensified TB case

finding, contact evaluations in Zambia

and S.A.

Community randomized trial(~1.2 million)

Preventive therapy and ARVs for HIV patients

in Rio de Janeiro

Phased implementation trial

(15,000)

Prevalence of INH resistance after receiving IPT vs. controls without IPT

7/58 (12.1%) First episodes

Retreatment episodes

12/200 (6.0%)

32/270 (11.8%)

1/13 (7.7%)

14/75 (18.7%)

van Halsema, IAS 2009

ZAMSTAR

• 24 communities in South Africa and Zambia

• Designed to reduce TB prevalence through– improved TB case finding by increased access

to TB diagnostics for symptomatic persons– Household interventions for families of TB

patients, which include HIV testing and treatment, TB screening, and IPT for contacts of TB cases

PIs: P. Godfrey-Faussett, H. Ayles, N. Beyers

ZAMSTARPreliminary Results

• Baseline population surveys– TB prevalence 900/100,000 in Zambia and

2,200/100,000 in South Africa!• Enhanced Case Finding

– 15-26% of all TB cases in intervention communities detected by ECF

• Household evaluations– >2-5% of households have secondary TB case– ~50% contacts HIV tested

• HIV prevalence high - ~50%• >50% HIV+ started on ART

The THRio Study:A Clinic - Randomized Trial of INH

Preventive Therapy in HIV+ Patients• 29 HIV clinics randomized to time that IPT

program initiated• TB rates compared pre- and post- intervention

Intervention

Control

Month1 2 3 4 5 30 36 42

Follow-upClinic

1

2

3

4

29

TB Rates by ART and INH Treatment Status, 2003-2005

Exposure category

Person-Years

TB Cases

Incidence Rate(per 100 PYs)

PercentReduction

No Rx 3,865 155 4.01 (3.40-4.69) -

ART only 11,627 221 1.90 (1.66-2.17) 52%

IPT only 395 5 1.27 (0.41-2.95) 68%

Both 1,253 10 0.80 (0.38-1.47) 80%

Total 17,140 391 2.28 (2.06-2.52)

Golub et al., AIDS 2007;21:1441-8

Tools and Strategies for Reducing the Global Burden of TB

• Improved diagnostics ( case finding)– Better tests– Campaigns to find prevalent cases

• Improved therapy ( treatment completion)– Shorter duration regimens to assure adherence– New drugs for MDR/XDR TB

• Prevention– INH preventive therapy– Reduction of susceptibility (ART, diabetes, smoking)

– Effective vaccine• Combination of approaches essential

Lessons from Tuberculosis Control for Communicable Disease Control

• Good biomedical tools are essential but not sufficient for disease control

• Understanding the epidemiology and dynamics of the disease is essential for targeting interventions

• Trials of control that measure population level impacts are important

• It ain’t over till it’s over!

With thanks to…

Jonathan Golub Neil MartinsonAnn Miller Peter Godfrey-FaussettSolange Cavalcante Gavin ChurchyardBetina Durovni Liz CorbettJacques Grosset Eric NuermbergerLarry Moulton The CREATE TeamMarcus Conde Generous funders - BMGFDavid Dowdy NIH, CDC, FDA

Thank you