the challenge of antibiotic r&d · the challenge of antibiotic r&d john h. rex, md svp...
TRANSCRIPT
The Challenge of Antibiotic R&D
John H. Rex, MD SVP & Head of Infection Global Medicines Development, AstraZeneca
Non-Executive Director & Consultant, F2G Pharmaceuticals
Slides happily shared – just drop me a note
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 1
Why so few new drugs?
For today, let’s break it down to four things
Three big problems
1. It’s hard to discover new antibiotics
2. It’s hard to develop new antibiotics
3. The economic value of a new antibiotic to a developer can be close to zero
And the idea that
4. Fixing this requires us to see it as an ecosystem
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 2
Agenda
• The 3 big problems
–Discovery is hard
–Development is hard
–Economics are poor
• We must take a systems approach
• Conclusions
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 3
Antibiotics are hard to discover
• Easy to find: Targets
– Multiple bacterial genomes are fully sequenced
• Easy to find: Things that kill bacteria
– Bleach works quite well, as do steam and fire
• Hard to find: Kills bacteria & is safe
– Failures: physical properties, pharmacology, safety
– Need high levels to penetrate bug high doses
• Typical lipid-lowering agent: 5-20 mg/day
• Typical antibiotic: 100-2000 mg/day
• And finally, antibiotics are chemically different… Payne DJ et al. Drugs for bad bugs: confronting the challenges of antibacterial discovery. Nat Rev Drug Discov 2007;6:29-40.
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 4
Chemical space: antibiotics vs. other drugs
Other Drugs
Gram-positive antibiotics
Gram-negative antibiotics
Antibiotics vs. Other Drugs
logD (pH 7.4)
• A useful way to think about
chemical similarity and diversity is
to plot molecules based on
• LogD (a measure of water
solubility at a given pH) vs.
• Molecular Weight (MW)
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics
5 Slide courtesy Matt Cooper & COAD (Community for Open Antimicrobial Discovery)
Chemical space: antibiotics vs. other drugs
Other Drugs
Gram-positive antibiotics
Gram-negative antibiotics
Antibiotics vs. Other Drugs
logD (pH 7.4)
• Antimicrobials are often very
different from other drugs
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics
6 Slide courtesy Matt Cooper & COAD (Community for Open Antimicrobial Discovery)
Chemical space: antibiotics vs. other drugs
Other Drugs
Gram-positive antibiotics
Gram-negative antibiotics
Antibiotics vs. Other Drugs
logD (pH 7.4)
• Antimicrobials are often very
different from other drugs
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics
7 Slide courtesy Matt Cooper & COAD (Community for Open Antimicrobial Discovery)
Chemical space: antibiotics vs. other drugs
Other Drugs
Gram-positive antibiotics
Gram-negative antibiotics
Antibiotics vs. Other Drugs
Australian Biotech
Company Commercially-sourced
diverse set
Typical Corporate Libraries
logD (pH 7.4)
• Antimicrobials are often very
different from other drugs
• Corporate libraries mostly contain
things that don’t look like antibiotics.
Searching here is not likely to be
productive!
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics
8 Slide courtesy Matt Cooper & COAD (Community for Open Antimicrobial Discovery)
Published Academic
Compounds
This insight can be turned into action
Other Drugs
Gram-positive antibiotics
Gram-negative antibiotics
Antibiotics vs. Other Drugs
Australian Biotech
Company Commercially-sourced
diverse set
Typical Corporate Libraries
logD (pH 7.4)
• A program recently funded by the
Wellcome Trust (COAD) will make
testing available to investigators
with compounds
• The goal: new leads in this space
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics
9 Slide courtesy Matt Cooper & COAD (Community for Open Antimicrobial Discovery)
Agenda
• The 3 big problems
–Discovery is hard
–Development is hard
–Economics are poor
• We must take a systems approach
• Conclusions
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 10
The paradox of resistance
• You might think that resistance would make it easy to develop new antibiotics
• But, it’s actually surprisingly difficult to do this work
• To see why, let’s look at antibiotic development as a series of linked challenges…
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 11
You can’t always get what you want
(Rolling Stones)
Development is hard A series of linked challenges
• The superiority-based approaches that work for other areas do not offer a long-term path to a diverse, vibrant antibiotic pipeline
• We have to make non-inferiority (NI) work. How?
– The tiered framework
• The role of (rapid) diagnostics
• Other issues
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 12
Trial Design Options: Superiority • Superiority designs are very compelling
– And if dramatic, small is possible: 1/40 vs. 39/40 is credible
• The catch: For superiority in prospective, randomized studies to be a reliable path for antibiotics, randomization to potentially ineffective or toxic therapy must be acceptable1,2
– Remember: Untreated infections are lethal
– Unless we have no choice, must not enroll if the pathogen is resistant and the comparator is thus likely ineffective
– For comparator-susceptible pathogens, modern comparators at full dose are very effective
– Superiority is unlikely and active drugs could be discarded
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 13
1Nambiar et al. Clin Pharm Ther 96:147-149, 2014. 2Rex et al. Ann NY Acad Sci doi:10.1111/nyas.12441, 2014.
If clinical superiority is readily achieved, we have a problem
• So, when might you rely on a superiority-based approach?
• Perhaps today (2014) for highly resistant pathogens where a nephrotoxic drug (e.g., colistin) is justified as standard therapy
– Going forward, new drugs should steadily render inferior comparators unethical and superiority trials infeasible
• Related approaches have the same limitation
– Nested superiority in an NI trial?1 Add-on superiority?2
– For both: base therapy must be inadequate (or we must knowingly give inadequate therapy ) for superiority to be reasonably likely
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 14
1Huque MF et al. Hierarchical nested trial design for demonstrating treatment efficacy of new antibacterial drugs in patient
populations with emerging antibacterial resistance. Stat Med (wileyonlinelibrary.com)DOI:10.1002/sim.6233, 2014. 2A study of
NEW + OLD vs. placebo + OLD could seek superiority but would only be likely to succeed if OLD was poor.
Development is hard A series of linked challenges
• The superiority-based approaches that work for other areas do not offer a long-term path to a diverse, vibrant antibiotic pipeline
• We have to make non-inferiority (NI) work. How?
– The tiered framework
• The role of (rapid) diagnostics
• Other issues
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 15
Trial Design Options: Non-Inferiority
• We generally must instead use this design
– Drug X vs. Drug Y, both at meaningful doses
– Exclude if pathogen is resistant to either agent
– No rational expectation of superiority in this study
• You can still study many resistant pathogens
– E.g., if you use a carbapenem as the comparator, you can enroll any Gram-negative other than a carbapenem-resistant organism
• Problem: The trials tend to be large (700+)
– PK-PD to the rescue…
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 16
PK-PD1 & Totality of the Data
17 Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics
PK-PD Index (e.g., AUC:MIC
Ratio)
Eff
ec
t
Emax EC50
MN Dudley, Griffith D. In: Nightingale CH, Murakawa T, Ambrose PG ed. Antimicrobial Pharmacodynamics in Theory and Practice. New York, Marcel Dekker Publishers, 2002.
• Unlike most other drugs…
• Antibiotic blood levels, the
minimum inhibitory concentration
(MIC) of the drug for the bug, and
response have an unusually
predictable relationship
• With rare exceptions,2 blood &
tissue levels that work in a mouse
are very likely to do so in man
1PK-PD = Pharmacokinetic-Pharmacodynamic relationship. See the work of Craig, Drusano, Mouton, Ambrose, Hope, MacGowan,
Nicolau, and many others. 2A classic example is the lack of efficacy of daptomycin in pneumonia that was ultimately found due to the
effects of surfactant on daptomycin (Pertel 2008 CID). 3Peck CC, Rubin DB, Sheiner LB. Hypothesis: a single clinical trial plus causal
evidence of effectiveness is sufficient for drug approval. Clin Pharmacol Ther 2003;73:481-90.
• You still need other data (mostly safety), but PK-PD provides
direct proof of causality that reduces the need for empirical
causality validation3 via multiple trials
Using these ideas, a logical framework can be envisioned
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 18
Quantity of Clinical
Efficacy Data that you can
generate
Acceptance of smaller clinical datasets in response to unmet medical need
Rex et al. Lancet Infect Dis
13: 269-75, 2013.
Rex et al. Ann NY Acad Sci
2014, DOI 10.1111/nyas.12441.
Connecting it all: Totality of the Data
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 19
Quantity of Clinical
Efficacy Data that you can
generate
Acceptance of smaller clinical datasets in response to unmet medical need
Reliance on human PK data combined with preclinical efficacy data
Rex et al. Lancet Infect Dis
13: 269-75, 2013.
Rex et al. Ann NY Acad Sci
2014, DOI 10.1111/nyas.12441.
Tier A: Our traditional approach
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 20
A
P3 x 2
Quantity of Clinical
Efficacy Data that you can
generate
Acceptance of smaller clinical datasets in response to unmet medical need
Tier A:
Two big Phase 3 non-inferiority
studies.
Lots of clinical data. Limited reliance on
PK-PD.
Rex et al. Lancet Infect Dis
13: 269-75, 2013.
Rex et al. Ann NY Acad Sci
2014, DOI 10.1111/nyas.12441.
Reliance on human PK data combined with preclinical efficacy data
Tier D: The animal rule
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 21
A
D
P3 x 2
Animal rule
Quantity of Clinical
Efficacy Data that you can
generate
Acceptance of smaller clinical datasets in response to unmet medical need
Rex et al. Lancet Infect Dis
13: 269-75, 2013.
Rex et al. Ann NY Acad Sci
2014, DOI 10.1111/nyas.12441.
Tier D:
For biothreats such as anthrax
Human efficacy
trials not possible.
Huge reliance on
PK-PD
Reliance on human PK data combined with preclinical efficacy data
Reliance on human PK data combined with preclinical efficacy data
And then we add Tier B & Tier C The pathogen-focused pathways
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 22
A
B
C
D
P3 x 2
Small studies
Animal rule
Quantity of Clinical
Efficacy Data that you can
generate
Acceptance of smaller clinical datasets in response to unmet medical need
P3 x 1 plus small
studies
Pathogen-focused for unmet need
Reliance on human PK data combined with preclinical efficacy data
Rex et al. Lancet Infect Dis
13: 269-75, 2013.
Rex et al. Ann NY Acad Sci
2014, DOI 10.1111/nyas.12441.
Typical Tier B & Tier C Programs1 Good candidate drug for Tier B vs. Tier C
• Tier B: Sufficiently broad spectrum that monotherapy for a syndrome such as intra-abdominal infection is possible
• Tier C: A narrow-spectrum agent that covers but one of many possible pathogens in a syndrome
Phase 3 development program is then
• (Tier B only) Drug X vs. a standard comparator at one body site2 – No (or few) pathogens resistant to the comparator
– PK-PD will provide the link to activity vs. comparator-resistant pathogens
• Resistant pathogen study: Drug X vs. Best Available Therapy (BAT) for highly resistant pathogens for which there is not simple or standard comparator. Prospective, randomized, & open-label. N a few hundred. Multiple body sites.
Rex
JH
- 2
01
4-1
2-0
9 S
tock
ho
lm s
emin
ar -
Ch
alle
nge
of
anti
bio
tics
23 1The example presumes that a clear exposure target is known from preclinical PK-PD and that there is a clear ability to produce a corresponding drug exposure in patients. See literature (Rex et al. Lancet Infect Dis 13: 269-75, 2013) for detailed examples. 2E.g., pneumonia or UTI
These ideas are now mainstream But not yet globally harmonized
• EMA Antibacterial guidance (2013)
– Explicit description of options that match Tiers B & C
• FDA Antibacterial guidance (2013)
– Explicit description of options that match the Tier B ideas
– Unless inferential testing is possible, Tier C is hard for FDA
– We need to work to find flexible ways forward lest important agents not be available in the US
• And also note that antifungal development
– Has effectively long been Tier B-like
– Can now benefit from application of PK-PD ideas
– And are eligible for the US GAIN incentives
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 24
We will need to rethink Health Technology Assessor (HTA) & Payor requirements
• Labeling strongly impacts reimbursement
• This is the problem of evidence vs. access1
– Antibiotic data packages will often be small
– We will not routinely generate superiority data (see above)
• But, HTA & Payor analyses prefer larger data packages with a focus on superiority data
• Reimbursement criteria must be adapted
– E.g., Medicare NTAP2 payments to supplement DRGs must recognize this problem for antibiotics
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 25
1Woodcock J. Evidence vs. access: Can twenty-first century drug regulation refine the tradeoffs? Clin Pharm Ther
91:378-80, 2012. 2NTAP = New Technology Add-On Payments: The quick adjustment to DRGs made when a new
technology emerges.
Development is hard A series of linked challenges
• The superiority-based approaches that work for other areas do not offer a long-term path to a diverse, vibrant antibiotic pipeline
• We have to make non-inferiority (NI) work. How?
– The tiered framework
• The role of (rapid) diagnostics
• Other issues
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 26
1Another example: Take the 50% rate typical of hospital-acquired pneumonia to 75% and the study size shrinks by a third.
Diagnostics increase trial efficiency • In general, we must enroll in a clinical trial before we
have the culture result
– Cultures are often negative (can easily be > 50% rate)
– But, only those subjects with a relevant positive culture contribute fully to analysis of the new drug
• A positive rapid test would serve us well even if it only selected patients more likely to subsequently have a definitive positive culture
– Test might rule in or rule out – doesn’t matter
– If a test moves us from 30 to 50% culture-positive...
– Study size goes down 40%: we save cost & time1
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 27
Key test requirement: SPEED • Speed1 & ease of use (local!) really trumps
– Care processes take time (collect sample, send to lab, get report, order a drug, get drug from pharmacy, etc.)
– Too much prior therapy may invalidate the patient’s data for clinical trial purposes
• This is a challenge for test developers
– Developing a test with adequate sensitivity and specificity is surprisingly hard. We don’t (yet) have the tests we need
• New technologies may help
– Direct-from-specimen methods, next-generation sequencing, and evaluation of host (patient) gene responses all hold promise
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 28
1Contrast this with HIV or cancer where it is often acceptable and appropriate to take a few days to a few weeks to confirm the diagnosis and decide on a course of treatment.
Diagnostics: The dream
• Longitude Prize 20141
• £10m for “best, rapid, accurate, affordable, point-of-care method for diagnosing bacterial infections on a global scale with universal benefit.”
• Would be great to see it be so simple!
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 29
1http://www.longitudeprize.org/. Selected by the British people, the Prize commemorates the 300th anniversary of the 1714 Longitude Act which led to John Harrison’s invention of the reliable shipboard chronometer. The graphics above are from the contest announcement.
Development is hard A series of linked challenges
• The superiority-based approaches that work for other areas do not offer a long-term path to a diverse, vibrant antibiotic pipeline
• We have to make non-inferiority (NI) work. How?
– The tiered framework
• The role of (rapid) diagnostics
• Other issues
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 30
Interpretive Breakpoints • Interpretive breakpoints are concentrations used
when answering the question “Is it susceptible?”
– E.g., “Organisms inhibited by 1 mg/L are susceptible and infections due to them should respond to usual doses.”
• Problem: Historically, we’ve set breakpoints based on the range of observed MICs1 in the clinical trials
– Isolates with the highest MICs are rare and may well not be seen in the (smaller) Tier B /C programs
– Hence, initial clinical trials are usually inadequate to give full insight into best breakpoints
• We must not constrain breakpoints in this way – we really have to use PK-PD to set breakpoints
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 31
1MIC = minimum inhibitory concentration of the drug for the bug in a test tube
Pediatrics
• Focused requirements would speed access
• Example:
– Ceftaroline is a recently registered antibiotic
– Initial1 FDA + EMA pediatric commitments entailed studies of ~750 patients over a 6-year period and at a global cost of > $80m
– Too much, too slow. In addition to PK data by age, safety data requirements need to be reasonable
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 32
1The size of the program was subsequently reduced about 50%.
Agenda
• Where are we?
• The 3 big problems
–Discovery is hard
–Development is hard
–Economics are poor
• We must take a systems approach
• Conclusions
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 33
If we want a diverse, vibrant pipeline…
• We must find ways to fund & incentivize this work – “We can’t make companies do this work … we have to
make them want to do this work”1
• Our answer must address several basic tensions – We want to minimize use of all antibiotics – We want to have new(er) antibiotics available on demand – We want those antibiotics developed before the epidemic
• How can we do this? – Noting that “All models are wrong, but some are useful”2…
– … let’s now look at a model that may be instructive
1Spellberg B. The antibacterial pipeline: Why is it drying up, and what must be done about it? Appendix A in Antibiotic Resistance: Implications
for Global Health and Novel Intervention Strategies: Workshop Summary, Institutes of Medicine, 2010. Accessed online at
http://www.nap.edu/catalog/12925.html on 11 July 2013. 2GEP Box and NR Draper in Empirical Model-Building and Response Surfaces, 1987,
John Wiley & Sons, New York, NY.
34 Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics
The cost of creating an antibiotic An EU-based analysis
• The typical antibiotic lifecycle can be modeled from start to finish1
• The model allows for failed drugs
• Spend and revenue by year are based on industry average data
• Note the Phase 3 bump in spend
• And then a sales curve: ~10 years of protected sales and then ~10 years of declining sales
-$450
-$300
-$150
$0
$150
$300
Disc. Ph 1-3 On market
5 yr 8 yr 20 years
(Spend) Revenue by year
• Approximate total spend (years 1-13): $600m
• Approximate total sales (next 20 years): $2,500m
• But, we’ve forgotten about NPV!
1Sharma, P. & Towse, A. New drugs to tackle antimicrobial resistance: analysis of EU policy options. OHE website, 2011;
Spellberg et al. Nat Rev Drug Discov 11: 168., 2012 Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 35
0
25
50
75
100
Year 0 +1 +2 +3 +4 +5 +6 +7 +8 +9 +10
Sidebar: NPV (Net Present Value) How much is an investment worth in today’s terms?
• Cash today is worth more than a promise of cash tomorrow (or in ten years)
• Based on cost of capital, risk, etc., it is typical to discount 10% per year
• The math is the inverse of interest on a loan:
• $100 today = $100; $100 in a year = $90; $100 in two years = $81, etc.a
At 10% per year discount, $100 in 10yrs time is only worth $39 today
• A project’s NPV is calculated by
• Computing sales less costs for each year (Annual Net Cash Flow)
• Each future year’s Cash Flow is discounted to today
• The total across all years is the Net Present Value
• Any NPV > 0 means you’ve created (at least some) value
Before we go
further, we
interrupt this
presentation...
Now, back to the story… Rex JH - 2014-09-06 ICAAC Keynote - New tools & pathways for antibacterials a. Actually, I’ve simplified a bit here – the actual values are $91, $83 … but this simpler way of thinking
about it is close enough for illustrative purposes 36
The very real effects of NPV math
-$450
-$300
-$150
$0
$150
$300
-$450
-$300
-$150
$0
$150
$300
Disc. Ph 1-3 On market
5 yr 8 yr 20 years
(Spend) Revenue by year
Disc. Ph 1-3 On market
5 yr 8 yr 20 years
• Now, consider this in NPV terms
• From the standpoint of year 0 (the
day you decide to start discovery),
the graph below shows spend &
revenue discounted 10%/year
• The line below is the cumulative
NPV
• But in NPV terms, it totals to a loss
of around $50m
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 37
Recent US-based analysis: same result • Comprehensive model for drugs for 6 key indications
(ABOM, ABSSSI, CAP, cIAI, cUTI, HAP-VAP)1
• NPV of the new drug always < $40 million
– All 90% confidence intervals on estimate went below zero
• Value to the patient was MUCH higher
– Just based on the value of days of work and life restored, the value to society ranged from $500m to $12 billion
• Thus, these EU- and US-based models show that
– Starting antibacterial R&D is not financially rational, at least not with traditional R&D costs and approaches
– We (society) undervalue these drugs
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 38
1Sertkaya A, Eyraud J, Birkenbach A, Franz C, Ackerley N, Overton V, Eastern Research Group. Analytical framework for examining the
value of antibacterial products. Report to US DHHS. http://aspe.hhs.gov/sp/reports/2014/antibacterials/rpt_antibacterials.cfm
That’s a problem we must solve
• To restore vitality to the pipeline and ensure we have the life-saving drugs we will need in the future,
• We have to move these models back into positive territory
And, we’re now doing just that…
39 Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics
Agenda
• Where are we?
• The 3 big problems
• We must take a systems approach
–Global leadership
–Public-private partnerships
–Economics
• Conclusions
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 40
Global Leadership: A partial list 2003 et seq: IDSA: “Bad Bugs, No Drugs”
17 Sep 2009: (EU) Swedish presidency • “Innovative Incentives for Effective Antibacterials”
7 April 2011: WHO World Health day on AMR • “No action today, no cure tomorrow”
17 Nov 2011: (EU) ND4BB program • PPP for Discovery & Development
2011 forward: (US & EU) FDA & EMA • A steady stream of new guidances
2012: (US) GAIN Act (see subsequent slide)
3-4 Oct 2013: (EU) Chatham House Conference • “Antimicrobial resistance: Incentivizing
Change Towards a Global Solution”
2014: (US) PCAST Report
41 Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics
Public-Private Partnerships In the US: NIAID & BARDA
• NIAID: Antibacterial Resistance Program
– Extensive array of preclinical services
– Phase 1 clinical units
– ARLG (Antibacterial Resistance Leadership Group)
– Modeled on ideas such as I-SPY, master protocols are being considered as a way to provide infrastructure that would support development efforts
• BARDA (Biomedical Advanced Research & Development Authority)
– Several public-private partnerships established to date
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 42
In the EU: IMI’s ND4BB program (New Drugs For Bad Bugs)
ND4BB cross topic collaboration and dissemination
Topic 1: COMBACTE a) Enabling
Clinical Collaboration and refining clinical
trial design b) Clinical
Development of GSK1322322
c) Clinical Development of
MEDI4893
Topic 2 : TRANSLOCATION
Research penetration and
efflux Gram-negatives Data Hub and
Learning from R&D experience
Topic 4: Driving re-
investment in R&D and
Responsible use of
Antibiotics
Topic 5: Clinical
development of antibacterial agents for
Gram-negative antibiotic resistant
pathogens
Topic 6: Systemic molecules
against HAIs due to
clinically challenging
Gram-negative pathogens
Call 6
Call 8
Call 9
Call 11
Topic 7: Inhaled
Antibacterials in CF and non-CF BE
Discovery Economics & stewardship
Development
Topic 3 : ENABLE
Discovery & development of
new drugs combatting
Gram–negative infections
Development
ND4BB Information Centre – All data generated is submitted and is accessible to all consortium partners
IMI = Innovative Medicines Initiative
Rex JH - 2014-09-06 ICAAC Keynote - New tools & pathways for antibacterials 43
Example: TRANSLOCATION (Topic 2)
Discovery: Improve understanding of drug penetration into Gram-negatives Efficiency: Use a data center to compile and analyze ND4BB related information Management: holding all together & ensuring best practices / communication
Discovery Efficiency Management
44 44 Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics
And finally, economics In the US…
• 2012: GAIN (part of FDASIA)
– Generate Antibiotic Incentives Now
– Extended exclusivity, priority review, fast track, and a requirement to generate new guidances
• 2014: ADAPT (LPAD), DISARM
– Two more pieces of legislation now in discussion
– ADAPT (LPAD): Further support for streamlining antibacterial development
– DISARM: A fix for the NTAP1 problem
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 45
1NTAP = New Technology Add-On Payments: The quick adjustment to DRGs made when a new technology
emerges.
And in the EU… ND4BB Topic 4: The DRIVE-AB Project
• Launch meeting: 6 Oct 2014 • “Driving re-investment in R&D and Responsible use of antibiotics”
• Aim: Address the tension between economics & stewardship – Create a multi-disciplinary, multi-stakeholder community (16 public
partners and 7 private partners from 12 countries) with an in-depth comprehension of challenges of antibiotic development
– Develop evidence-based measures for responsible antibiotic use
– Create implementable options for new commercial models that address the needs of multiple stakeholders
– Validate options through modelling
• We expect DRIVE-AB to explore a broad range of approaches – In particular, we hope to see ways to separate (delink) usage from
reward to the innovator. That is, reward should not be sales-based
– Let’s look at two possible examples...
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 46
Two intriguing economic ideas • (Push) Refundable tax credits
– For some percentage (e.g., 50%) of qualified expenses, the company either gets a tax credit (if the company has income) or receives a payment of that amount
– Has immediate impact on NPV while also ensuring the company has “skin in the game” that ensures delivery
• (Pull) Insurance-based approaches
– National acquisition at a fixed, predictable rate (e.g., US buys $100m/year of a new antibiotic for 5 years)
– Annual fee guarantees availability of a certain number of courses of therapy, whether used or not
– We should be pleased to buy but not use the drug, just as we are pleased when our life insurance does not pay off
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 47
We’re now tackling the entire model! • The typical antibiotic lifecycle can be
modeled from start to finish1
• The model allows for failed drugs
• Spend and revenue by year are best
on industry average data
• Note the Phase 3 bump in spend
• And then a sales curve: ~10 years of
protected sales and then ~10 years
of declining sales -$450
-$300
-$150
$0
$150
$300
Disc. Ph 1-3 On market
5 yr 8 yr 20 years
• With support from NIAID, BARDA, ND4BB, & others plus
the tiered approach, we are truly taking a systems
approach to this problem
• The Discovery and Development support + the tiered
approach is already having an impact
• Last step: Rethinking value and business models
NIAID, BARDA, ND4BB NIAID, BARDA, ND4BB
The tiered approach
ND4BB & New business models
48 Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics
Agenda
• Where are we?
• The 3 big problems
• We must take a systems approach
• Conclusions
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 49
Our head is round so that our thinking can change direction
(Francis Picabia)
There’s no time to lose It takes 10-20 years to make a new antibiotic
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 50
*Hit to Phase 2 based on novel mechanism AB discovery (GSK) #Based on Paul, et al (2010), Nature Reviews Drug Discovery 9: 203-214.
So few companies are still in the game
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 51
Graphic adapted from Kinch MS et al. Drug Discovery Today, July 2014.
Active corporate
anti-bacterial programs
Corporate Entries
Corporate Exits
Over the past 15 years
(1998-2013)
• 14 entries
• 36 exits
I’m pleased to see new
companies emerging in
2014, but we have a lot of
ground to reclaim
30
20
10
0
+4
-4
-8
0
# of
companies
2000 1980 1960 1940 2013
2000 1980 1960 1940 2013
IDSA has set the challenge
52
• Global commitment to develop 10 new systemic antibacterial drugs by 2020 (CID; April 2010)
http://www.idsociety.org/10x20/ Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics
4 oritavancin The Medicines Company, Approved: Aug 6, 2014
4
IDSA has set the challenge
53
• Global commitment to develop 10 new systemic antibacterial drugs by 2020 (CID; April 2010)
http://www.idsociety.org/10x20/ Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics
4 oritavancin The Medicines Company, Approved: Aug 6, 2014
4
These four focus principally
on Gram-positives. Of great
interest, several agents for
Gram-negatives are now
approaching registration.
We must take a systems approach • Collaborative R&D
• Streamlined pathways1,2,3
– By acting quickly to create approaches to describe and manage the uncertainty of small data packages,
– We will provide patients with timely access to urgently needed, life-saving antibiotics, and
– We will avoid the paradoxical situation of being forced in the future to accept even greater degrees of therapeutic uncertainty as antimicrobial resistance progresses
• New economic thinking
– We will fail if we don’t unleash the power of private investment
Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics 54
1Hersh et al., Clinical Infectious Diseases 54:1677, 2012. Among 562 respondents in a 2011 survey of the Emerging Infections Network (EIN), 64% reported using colistin during the previous year and 63% reported caring for a patient with an infection resistant to all available antibacterials. 2Boucher HW et al. Clin Infect Dis 56:1685-94, 2013. 3Rex et al. Lancet Infect Dis 2013; 13: 269-75.
In closing
• The incredibly thin pipeline has many causes
– A path to a diverse, sustainable pipeline must be found
• We’ve made a lot of progress! Thank you!
– Discovery is hard: NIAID, BARDA, ND4BB are opening doors
– Development is hard: Now improved! Tiered framework!
– Economics are poor: Tiered framework; New business models
• There’s still more to do
– But, I think (and to borrow a line from Mr. Churchill), we are now at the end of the beginning
55 Rex JH - 2014-12-09 Stockholm seminar - Challenge of antibiotics
With many thanks to all who are part of these amazing endeavors!