5/28/2018 New regimen for treat TB
1/18
Director, Market Access, TB Alliance
William Wells
Alignment of new TBdrug regimens and drugsusceptibility testing: a
framework for action
March 25, 2013
5/28/2018 New regimen for treat TB
2/18
The TB drug development pipeline
Resistance and DST algorithms
Turning theory into diagnostic development
Outline
2
5/28/2018 New regimen for treat TB
3/18
3
No new drugs for TB in more than 40 years
DrugSensitive TB
4 drugs takenfor 6 or more
months
Shorter,
simplertherapy
M(XDR)-TB
Injections anddrugs taken for
more than 2years, poorly
tolerated
More effective,shorter, safer
simplerregimens
TB/HIV con-infection
Drug-druginteractionswith ARVs
Co-
administrationwith ARVs
Latent TBInfection
9 months ofisoniazid
Shorter, more
easily toleratedtherapy
Children
Formulationsnot adequately
dosed
Regimens andformulationswith correct
dose
Current Therapy and Unmet Needs
Current
Therapy
Unm
et
Nee
ds
5/28/2018 New regimen for treat TB
4/18
5/28/2018 New regimen for treat TB
5/18
2013+: Bedaquiline (TMC-207; Janssen/Tibotec) and delamanid
(OPC-67683; Otsuka) for MDR-TB Based on Phase II results: adding new drug for 6 months on top of existing
MDR-TB regimen
Phase III trial will take several years
2015: Dispersible first-line (HRZ) FDCs for pediatric use
2015: REMoxTB regimens (moxifloxacin) for drug-sensitive TB
2018+: PaMZ for drug-sensitive TB AND some MDR-TB
??: Universal first line regimen with at least 3 new chemical entities(e.g., bedaquiline, delamanid or PA-824, sutezolid, and others)
Timeline for country availability of new TBdrugs and regimens
5
5/28/2018 New regimen for treat TB
6/18
6
Randomized, Double-blind; Non-inferiority
Phase 3 REMox TB Trial Design
1 2 3 4 5 6Treatment Duration (months)
HRHRZE630 participantsStandard Regimen
ContinuationIntensive
Placebos
HRZM HRM
Placebos
630 participants
Moxifloxacin
for
Ethambutol
MRZE MR
Placebos
630 participants
Moxifloxacin
for
Isoniazid
Al l par t ic ipants fo l lowed for 12 months post- t reatment
H = isoniazid; M = moxifloxacin; R = rifampin; Z = pyrazinamide; E = ethambutol
1931 patients enrol led in Ch ina, Ind ia, Kenya, Malaysia, Mexic o, South
A fric a, Tanzania, Thailand , Zamb ia
5/28/2018 New regimen for treat TB
7/18
Treatment outcome benefits Shorter regimens may increase adherence and yield higher effective cure rates
and less emergence of MDR-TB
Less resistance to drugs in regimen (e.g. isoniazid vs moxifloxacin)
Health Systems benefits Reduced cost in healthcare utilization; patient cohort size reduced by one third
Patient benefits Reduced out of pocket costs (fewer visits)
Less time exposed to side effects
PaMZ: similar benefits + one dose for all, no ARV-rif interaction, and
treats ~25% of MDR-TB
Shorten Treatment from 6 to 4 months
Benefits of REMox Regimen
7
5/28/2018 New regimen for treat TB
8/188
Clinical, Regulatoryand Market Accessactivities in brief
Timeline for REMox TB
5/28/2018 New regimen for treat TB
9/18
What is the background,population-basedresistance to drugs inthe combination?
Based on modeling, what individualdrug susceptibility testing (DST)
with what kind of sensitivity and specificitywill be needed?
What is the likely availability of those DST diagnostics? Developed
In field use
At national & district levels
Current DST is focused on the needs of the current regimenbut
what will be needed for the future?
What countries may need to know about drug resistance
Questions for Adoption
Confidential 9
5/28/2018 New regimen for treat TB
10/18
Typically diagnosis in LMICs is with smear microscopy and no DST
Current treatment algorithm: many countries
HRZE
Xpert (R)
Rs Rr
Hain sl or cultureHRZE
MDR24
cure fail
Confidential 10
5/28/2018 New regimen for treat TB
11/18
Undiagnosed R resistance, treated with HRZE: Effectively means treatment with HZE or worse. Weak regimen; high likelihood
of treatment failure and mortality for the individual, and of continuing
transmission of drug-resistant strain to the community
During this ineffective treatment, it is very likely that the individual acquires
further resistance (to H, Z and E)
Failure to implement DST is courting disaster not just for the
individual, but for the TB epidemic as a whole
Consequences of having no DST
Confidential 11
5/28/2018 New regimen for treat TB
12/18
Especially in low income countries, complicated techniques such asculture are feasible only at a very few, centralized locations
Sample referral systems either do not exist or are slow; many
patients are lost to follow up
Mtb grows slowly. So fast DST is almost certainly molecular DST.
Challenges: mutations may be spread over a whole gene (pncAfor
Z resistance) or more than one gene (fluoroquinolone resistance);
this limits both sensitivity and specificity
Good DST = simple and fast DST
Confidential 12
5/28/2018 New regimen for treat TB
13/18
Xpert uptake not just for DST ability, but often for TB detection
Many high burden countries have insufficient resources to do DST
on all suspects or patients
If resistance prevalence is low, false positives (and need for
confirmatory testing) would be high
Trade-off: DST increases knowledge of patients clinical status
But DST increases burden on health system and patient
If DST results in greater travel costs, multiple visits and diagnosticdelays, it will reduce patient retention and may result in worse
outcomes.
Its not that simple
Test everybody for everything?
13
5/28/2018 New regimen for treat TB
14/18
For M: Upfront DST may not be needed or advisable in sub-Saharan Africa, if they have M
resistance in new cases of ~1%
This may be a closer call in Asia, if they have higher prevalence of M resistance among
new cases.
For Z: Baseline resistance among new cases may be ~3% everywhere (i.e., somewhat higher
than for M). This, plus concern about exposing M to an inadequate regimen, may
increase pressure for upfront Z DST.
How low would Z resistance have to be in new cases (algorithm #2) or non-MDR-TB
cases (algorithm #3) for presumptive PaMZ treatment to be OK?
For both: Fast test = molecular test. Achieving a high PPV with a molecular test may be more
challenging for M and Z than it is currently for R (not all mutations known, or known to
be specific to resistant strains). If PPV is low, either reserve DST for high risk
subpopulation and/or need confirmatory testing of positives .
Conclusions on DST algorithms for PaMZ
Confidential 14
5/28/2018 New regimen for treat TB
15/18
Specifications: which drugs;
surveillance, screening or stand-alone test;
what sensitivity and specificity required;
What decentralization required.
Market size/demand: What resistance prevalence is the cut-off for adoption; therefore which
countries adopt
Where the diagnostic is placed in algorithmstherefore what percentage of
TB suspects or patients get tested
What diagnostics developers need
15
5/28/2018 New regimen for treat TB
16/18
Prediction Predicting new regimen adoption is hard;
Predicting new DST adoption is hard;
Predicting new regimen + new DST adoption is much harder.
Chicken and egg Diagnostic companies only willing to develop new DST if new regimens are
adopted.
But adoption of new regimens relies on availability of new DST assays
Two-fold challenge in predicting demand
16
5/28/2018 New regimen for treat TB
17/18
Enabling science Define which mutations correlate with in vitro resistance andhave clinical impact
Establish a strain bank to use for testing of new assays
Surveillance Establish more baseline values for fluoroquinolone resistance among new patients, and
Z resistance among new and MDR-TB patients Modeling impact
Model trade-offs between speed, accuracy, price, and technical specs of DST assays
Model different DST algorithms (e.g., DST for all, DST for retreatment/failure cases only,
or use of novel regimens without DST). Above what threshold of resistance prevalence
would more widespread DST be advisable?
Assay development Finalize target product profiles, so developers have a clear pathway forwards
Use both existing platforms (e.g., FQr via Xpert) and new platforms
Co-development of drugs and diagnostics is the way forwards!
Pathway for action
17
5/28/2018 New regimen for treat TB
18/18
18
Thanks.to Peter Kim, Marco Schito, and all other
members of the Tuberculosis Diagnostics Research
Forum; and to you for listening.