New Approaches for the
Diagnosis of Tuberculosis
Richard O’Brien
Consultant, FIND and Chair, GLI
16th Annual Conference of the
Union North America Region
San Antonio, TX, 24 February 2012
Disclosure
I have no personal financial conflicts of
interest to report . However, my former
employer, FIND, has had contractual
relationships with a number of companies
whose products I will mention in my
presentation. These called on FIND to
support development, evaluation and
demonstration studies of these tests in
low-income settings. In turn, the
companies were to provide the tests at
favorable prices to the public sector in
developing countries.
Outline of Presentation
• Summary of WHO process for
endorsing new TB control tools and
strategies
• Overview of the Stop TB Partnership
Global Laboratory Initiative Working
Group
• Update on Line Probe Assays,
GeneXpert MTB/RIF and LAMP
Global Laboratory Capacity Gap
To reach MDG targets, a global capacity need of 120 million smears, 60 million cultures and
6 million DST investigations must be met by 2015, requiring at least 6.1 billion USD by 2015
# of tests
required (mln)
200
Establish 5 000 new microscopy laboratories; Establish 2 000
new culture and DST laboratories; Train 9 000 new technicians in
smear microscopy; Train 23 000 new technicians in culture and
DST.
USD funding
required (mln)
2500
MDG Target
150
100
2000
1500
1000
Urgent
50
2008
2010
2012
2015
500
TB Diagnostics Pathway*
*Stop TB New Diagnostics Working Group: A blueprint for
the development of TB Diagnostics, 2009
WHO TB diagnostics policy formulation process
Identifying the need
for policy change
• WHO monitoring of country needs
• Partners (researchers, industry, etc)
• Body of evidence available
Reviewing the evidence
• Commissioning of systematic reviews
• QUADAS/other diagnostic accuracy tool
• Meta-analyses (where feasible)
Convening an Expert Group
Assessing policy proposal
and recommendations
Formulating and
disseminating policy
• Experts, methodologists, end-users
• GRADE process for evidence synthesis
• Strategic and Technical Advisory Group
• Endorsement/revision/addition
• Advise to WHO to proceed/not with policy
• Guidelines Review Committee
• Dissemination to Member States
• Promotion with stakeholders & funders
• Phased implementation & scale-up plan
GRADE* evaluation
Clear separation:
1) Recommendation: 2 grades –
strong or conditional(optional/weak) for or against an intervention
– Balance of benefits and downsides, values and preferences, impact,
resource use,
with
2) Quality of evidence: 4 categories –
 (High), (Moderate), (Low), (Very low)
– Methodological quality of evidence
– Likelihood of bias
– By outcome and across outcomes
*Grades of Recommendation Assessment, Development and Evaluation
Recent WHO laboratory policies (1)
• Automated liquid culture and DST (2007): Use of liquid
culture systems in the context of a comprehensive country
plan for strengthening TB laboratory capacity; in a phased
manner starting at national/central reference laboratory level
• Rapid speciation (2007): Strip speciation for rapid
Mycobacterium tuberculosis from non-tuberculosis
mycobacteria; established at regional or central reference
laboratory level in combination with liquid culture
• Line probe assays (2008): Use of line probe assays for
rapid detection of R resistance within the context of country
plans for MDR-TB management, including development of
country-specific screening algorithms and timely access to
quality-assured second-line anti-tuberculosis drugs; do not
eliminate the need for conventional culture and DST
capability; should be phased in, starting at national/central
reference laboratory or those with proven molecular
capability
Recent WHO laboratory policies ()2
• Second-line drug susceptibility testing (2008): Reliable
and reproducible for injectables and fluoroquinolones; to be
conducted in supranational or national/central reference
laboratories using standardised methodology and drug
concentrations
• LED microscopy: (2009): alternative for fluorescence and
conventional light microscope
• Selected non-commercial culture and DST methods
(2010): not alternatives for gold standards, but may provide
interim solution
• Cepheid Gene Xpert (2010): should be used as the initial
diagnostic tests in patients suspected of MDR TB or TB/HIV
• Commercial serologic tests and IGRAs (2011): should not
be used in low and middle income settings
• Line probe assay for XDR TB (expected in 2012)
Stop T B Partnership Working
Groups (WG)
•
•
•
•
•
•
•
DOTS Expansion WG
TB/HIV WG
MDRTB WG
New Diagnostics WG
WG on New TB Drugs
WG on New TB Vaccines
Global Laboratory Initiative WG (Nov 08)
Global Laboratory Initiative
Platform of coordination and communication, providing the
required infrastructure, focused on TB laboratory
strengthening, in the areas of:
• Global policy guidance (norms, standards, best practices)
• Laboratory capacity development
• Interface with other laboratory networks, enabling integration
• Standardised laboratory quality assurance
• Coordination of technical assistance
• Effective knowledge sharing
• Advocacy and resource mobilisation
GLI structure & governance
WHO Stop TB Department
GLI Secretariat
GLI Secretariat
Stop TB Partnership
GLI Working Group
Evaluates, approves, governs
projects; Advises GLI Secretariat
GLI Core Group
Technical Working Groups
GLI Partners Committee
Advises and approves
strategic agenda of GLI; Monitors project progress
Supranational Ref Lab Network Sub-WG
Human resource development strategy
Laboratory strengthening
Laboratory accreditation
Laboratory biosafety
Priority projects and activities
Time limited
Partner approach
GLI strategic priorities*
• Accelerating evidence-based policy development on
diagnostics and laboratory practices
• Promoting a structured framework/roadmap for TB
laboratory strengthening within the context of national
laboratory plans at country level
• Developing a comprehensive set of tools, norms and
standards based on international standards and bestpractice
• Advancing laboratory strengthening through global,
regional and local partnerships
• Developing multi-level laboratory human resource
strategies to address the capacity crisis
• Accelerating new diagnostics into countries
*GLI Website: http://www.stoptb.org/wg/gli/
GenoType® MTBDRplus test procedure
3) Hybridization
1) DNA
Extraction
From NALC/NaOH
Processed sputum
Reverse hybridization of
amplified nucleic acids
to specific DNA probes
bound on strips
2) Amplification
by PCR
4) Evaluation
identification of
Mtb complex and
RIF-R and INH-R
MTBDRplus Strips
Conclusions from Cape Town
Evaluation Study*
• Overall performance of MTBDRplus assay is
superior to conventional culture/DST: speed,
accuracy, interpretable results, high throughput
• Cost of testing may be less than culture/DST
• Can substantially reduce the need for culture/
DST when screening for MDR TB
• May be much easier to establish than new
culture-capable laboratory
• New sputum processing methods may both
increase sensitivity and facilitate transport
*Barnard M, et al. AJRCCM 2008:177:787-792
FIND Hain Demonstration Projects
Samara, RF
China
Turkey
Vietnam
India
Brazil
Uganda
South Africa
Completed
Underway
Thailand
WHO Policy Statement on LPAs
• Integrated national plan for
LPAs together with MDR-TB
management and lab
capacity strengthening
• Use of LPAs on smearpositive sputum and cultures
(insufficient evidence on
smear negatives)
• LPAs do not replace culture
+ DST
• Commercial assays
recommended
• Lab infrastructure,
procedures and biosafety
• Human resources
• Training, technical support
and supplies
• EQA
MTBDRsl Line Probe Assay - Background
• June 2008 WHO approves LPAs for
MDR TB screening
• Mach 2009 Hain introduces LPA for
XDR TB – based on single published
study
• April 2010 FIND enters into
agreement with Hain
– FIND to conduct evaluation (accuracy)
studies of the test
– Hain to provide the test at a favorable
price to the public sector in developing
countries
• FIND supported studies at CDC, UCT
and Korea ITRC
• Data from 2 additional studies (SA
and Nepal) provided to FIND
Conclusions from FIND and Other Unpublished
Evaluation Studies
• The studies (excepting UCT) found a high level of valid test
results when performed on culture isolates
• Test sensitivity for FQ-R and injectables met FIND’s predefined performance targets at three of the study sites
• Test specificity was over 95% for both drug classes at all study
sites, except for CDC where it was 80-94% depending on the
reference standard
• Test performance (sensitivity) for the injectable drugs was
unacceptably low in the SAMRC study
• Banding problems with rrs probes at 2 SA sites presently
unexplained but not thought to be related to strain differences
• None of the studies evaluated the test on sputum specimens
• Operational issues, cost, and patient impact not assessed
Published Evaluation Studies of MTBDRsl Test
• PNTH, HCMC: 62 isolates (41 FQ-R, 21 MDR/FQ-S)
• Pitié-Salpêtrière, Paris: 52 isolates (41 MDR, 8 XDR)
• NRL, Borstel: 106 isolates (63 resistant) and 64 frozen sputa*
Drug
Sensitivity
Specificity
Agreement
FQ
83%
99%
93%
AMK
87%
100%
96%
KAN
83%
100%
97%
CAP
86%
98%
94%
EMB
65%
97%
77%
*42 AFB+, 94% valid results (4 invalid on AFB+ specimens)
September 2010 WHO EGM on SL LPA
• The EG was considering a “conditional”
recommendation for use of the test as an
alternative to conventional SL DST in screening for
XDR TB – but on culture isolates only
– Based on acceptable performance in evaluation studies testing
culture isolates
• Additional data needed on:
– Direct testing on sputum specimens
– Performance in other settings/populations (Taiwan specifically
mentioned)
– Availability of genetic sequencing data on strains with
discrepant LPA and DST results
• EG accepted a suggestion to defer a
recommendation until these additional data are
available
MTB / Rif-resistance test
Essential features of test
Workflow
• sputum
• simple 1-step external sample prep. procedure
• time-to-result < 2 h
• throughput: > 16 tests / day / module
• no need for biosafety cabinet
• integrated controls
• true random access
GeneXpert System
module
MTB
Performance
• specific for MTB
• sensitivity better than smear, similar to culture
• detection of rif-resistance via rpoB gene
Product and system design
• test cartridges for GeneXpert System
• several GeneXpert modules can be combined
in 1 workstation
• swap replacement of detection unit
• ~1 day technician training for non-mycobacteriologists
cartridge
Automated
Sample Prep,
Amplification
and Detection
90 minutes
GeneXpert® - a Molecular Lab in a Cartridge
Fully-Integrated Sample Preparation,
Amplification and Detection
• Universal sample prep
• Closed test system
provides:
– Nested PCR
– Reflexive test
capability
• Same Basic Cartridge
works With all Tests and
GeneXpert® Systems
Concentrates bacilli &
removes inhibitors
End of hands on work
3
Sample is
automatically
filtered & washed
4
Ultrasonic lysis of filtercaptured organisms to
release DNA
5
DNA molecules are
mixed with dry PCR
reagents
Transfer of 2 ml
after 15 min
GeneXpert
WHO Expert Group Review Endorsement
6 Dec 10
WHO
Feasibility
Development
Evaluation
Demonstration
Expert
Group
Meeting
1. Multi-centre clinical evaluation
studies
STAG
Meeting
Collecting
evidence
for scale-up
Scaleup
Global
Consultation
 1,730 subjects in five
evaluation sites (four
countries)
2. Multi-centre demonstration studies  6,648 subjects in nine
evaluation sites (six
countries)
3.
Single-centre evaluation studies
 4,575 subjects in 12 studies
(nine countries)
Evaluation Study Results*
• 1730 TB/MDR suspected patients enrolled in Azerbaijan,
India, Peru, South Africa
• A single, direct Xpert detected 92.2% of all C+ patients.
• Sensitivity in S-C+ patients was 72.5% and increased to
90.2% when three samples were tested. Specificity was
99%.
• A single, direct Xpert identified a greater proportion of
culture-positive patients than did a single LJ culture.
• Xpert MTB/RIF detected rifampicin resistance with 99.1%
sensitivity and excluded resistance with 100% specificity.
*Boehme
C, et al. NEJM 2010:363:1005-14
Multicenter Implementation Study: Single, direct Xpert in
routine settings Performance similar to solid culture
Sensitivity
All C+
96.6%
Sensitivity
S+C+
99.3%
Sensitivity
S-C+
88.1%
Specificity
Non-TB
99.6%
Baku, Azerbaijan
88.6%
97.8%
74.7%
98.7%
Cape Town, SA
86.3%
100.0%
79.1%
99.7%
Kampala, Uganda
83.4%
97.8%
57.7%
100.0%
Vellore, India
100.0%
100.0%
100.0%
97.7%
Manila, Philippines
91.9%
96.2%
56.3%
97.9%
90.3%
(933/1033)
98.3%
(637/648)
76.9%
(296/385)
99.0%
(2846/2876)
Lima, Peru
TOTAL




Routine smear microscopy (2-3 smears per patient) had a sensitivity of 61%.
Xpert identified at least as many patients as a single LJ culture (89.8%; CI 87- 92%).
High positive and negative predictive values in all settings
2.5% indeterminate rate; 0.3% after repetition. Culture indet. rate 4.7%.
Sensitivity and specificity of rifampin resistance detection
Total
%
(Correct / total)
[CI]
Sensitivity in
RIF-resistant
Specificity in
RIF-sensitive
94.4%
(236/250)
[90.8% - 96.6%]
98.3%
(796/810)
[97.1% - 99.0%]
 Rifampin resistance was a good marker for MDR at all sites.
 Positive predictive value suboptimal in low MDR-prevalence settings
 Confirmation of resistance by culture at present recommended for low MDR
prevalence settings
 Optimization ongoing
WHO recommendation on use of Xpert MTB/RIF
A phased roll out
MOH, SA
World TB Day 2011
Through Dec 2011, 460
GeneXpert instruments
(2401 modules) have been
ordered by 47 countries
45% of the capacity is in
South Africa!
TB LAMP
• Isothermal
• Rapid
• Closed system
• Visible readout
Status TB LAMP
90°C for
5min
60ul
Shake
Sputum
Heating
tube
Heating
block
Absorbent
tube
Dried
reagent
Injection
cap
Reaction
tubes
Mix
Detect fluorescence
signal
Add 25 – 35 µl
67℃ 40min
Oct 2011
WHO
submission
August
2005
Development
Evaluation
Demonstration
Evidence for
scale up
Evaluation Study (1063 TB suspects):
Sensitivity and specificity remain below target
Performance of single direct LAMP
Reference standard: 2 direct smears; 2 LJ and 2 MGIT cultures per
patient
Peru
South Africa
Vietnam
Brazil
Overall
Sensitivity in C+
Sensitivity in S+C+
Sensitivity in
S-C+
Specificity in S-C-
88.7% (86/97)
98.3% (58/59)
73.7% (28/38)
95.7% (180/188)
[80.8%-93.5%]
[91.0%-99.7%]
[58.0%-85.0%]
[91.8%-97.8%]
77.5% (62/80)
90.6% (48/53)
51.9% (14/27)
92.1% (175/190)
[67.2%-85.3%]
[79.7%-95.9%]
[34.0%-69.3%]
[87.4%-95.2%]
67.3% (113/168)
100% (59/59)
49.5% (54/109)
96.1% (123/128)
[59.8%-73.9%]
[93.9%-100%]
[40.3%-58.8%]
[91.2%-98.3%]
85.3% (81/95)
98.7% (74/75)
35.0% (7/20)
95.6% (109/114)
[76.8%-91.0%]
[92.8%-99.8%]
[18.1%-56.7%]
[90.1%-98.1%]
77.7%
97.2%
53.1%
94.7%
(342/440)
(239/246)
(103/194)
(587/620)
[73.6%-81.4%] [94.2%-98.6%] [46.1%-60.0%] [92.6%-96.2%]
TB LAMP performance: Improved in demo studies
Sensitivity in C+
Sensitivity in S+C+
Sensitivity in S-C+
Specificity in S-C-
Lima,
Kampala,
Sevegram,
All
Peru
Uganda
India
94.3%
76.1%
87.3%
83.9%
(100/106)
(140/184)
(89/102)
(329/392)
[88.2% - 97.4%]
[69.4% - 81.7%]
[79.4% - 92.4%]
[80.0% - 87.2%]
100.0%
95.5%
95.9%
96.8%
(67/67)
(105/110)
(70/73)
(242/250)
[94.6% - 100.0%]
[89.8% - 98.0%]
[88.6% - 98.6%]
[93.8% - 98.4%]
84.6%
47.3%
65.5%
61.3%
(33/39)
(35/74)
(19/29)
(87/142)
[70.3% - 92.8%]
[36.3% - 58.5%]
[47.3% - 80.1%]
[53.1% - 68.9%]
97.8%
97.2%
93.7%
96.2%
(364/372)
(520/535)
(448/478)
(1332/1385)
[95.8% - 98.9%]
[95.4% - 98.3%]
[91.2% - 95.6%]
[95.0% - 97.1%]
• Sensitivity target in S+C+ (>95%) met
• Point estimates close to meeting S-C+ target (65%) and S-C- target (97%)
Low indeterminate rate for LAMP compared to
culture
Lima, Peru
0.0%
(0/480)
Kampala, Uganda
0.0%
(0/770)
Sevegram, India
4.4%
(29/665)
All
1.5%
(29/1915)
LJ indeterminate rate
0.2%
(1/480)
6.5%
(50/770)
4.2%
(28/662)
4.1%
(78/1912)
MGIT indeterminate rate
2.1%
(10/480)
13.0%
(100/769)
5.5%
(37/669)
7.6%
(145/1918)
Overall LAMP
Indeterminate rate
• No serious DNA contamination events reported
• Negative control 1x positive in Uganda; 1 x in India
User appraisal: "Easier than smear microscopy"
After Training
14
12
10
More
8
Same
6
Less
4
2
0
LAMP difficulty compared to
smear microscopy
LAMP hands-on time compared
to smear microscopy
End of Study
14
12
10
More
8
Same
6
Less
4
2
0
LAMP difficulty compared to
smear microscopy
LAMP hands-on time compared
to smear microscopy
Consolidating LAMP – the way forward
Deliverables
Key notes
Next steps
1. Validate minor changes and move to demonstration phase II
2. Involve multiple partners for phase II
3. Submit evidence base to WHO in 2012/2013
Platform expansion
1. Malaria
2. HAT
3. Leishmaniasis
Acknowledgement
Catharina Boehme, FIND