Candida Past, Present,

1
Candida Past, Present, Future A Clinical
Perspective

• John H. Rex, MD
• AstraZeneca Pharmaceuticals
• Vice President, Medical Director for Infection
• University of Texas Medical School-Houston
• Adjunct Professor of Medicine

The slides contain many references. Should you
want the exact cite, a copy of the talk will be
available soon at www.doctorfungus.org
2
Medical MycologyThe Last 50 Years
of drugs
L-AmB ABCD ABLC
Terbinafine
Amphotericin B (1958)
Itraconazole
Griseofulvin
Fluconazole
Ketoconazole
Miconazole
Nystatin
5-FC
3
Medical MycologyPatient Groups as Drivers of
Demand
L-AmB ABCD ABLC

• Having toenails!

Terbinafine

• Chemotherapy

Amphotericin B (1958)
Itraconazole
Griseofulvin
Fluconazole

• HIV Infection

Ketoconazole
Miconazole
Nystatin
5-FC

• ICU Care

4
New Tools Open New Doors

• Novel Diagnostics

• Do I need a pill?

L-AmB ABCD ABLC

• Susceptibility Testing

• Which pill?

Terbinafine

• Fluconazole

• An easy pill!

Amphotericin B (1958)
Itraconazole
Griseofulvin
Fluconazole

• Ketoconazole

• A pill!

Ketoconazole
Miconazole
Nystatin
5-FC

• Amphotericin B

• Lives are saved!

5
Medical Mycology The Next 10 Years

• Thats a lot of progress
• But, there is more to do
• Not a lament for the sake of lament
• Invasive aspergillosis
• Mortality gt 30 Herbrecht NEJM 2002 347408
• Much higher in some settings
• E.g., CNS aspergillosis
• Candidiasis
• Persistent candidemia rate 10
• Even in the most recent studies!
• Onychomycosis
• 20 fail Drake LA, J Am Acad Derm 97 37740

6
Survey Results

• Big Clinical Advances
• Echinocandins
• Understanding resistance mechanisms
• Clinical Wish List
• Adequate diagnostic tests
• Fungicidal azoles. ? New drugs
• Susceptibility methods that better predict
  response
• Better understanding of virulence factors,
  pathogenesis, ways to predict colonization ?
  disease

As I thought about where we need to go next, it
was interesting to hear your thoughts on things
that mattered
7
Molecular Diagnostics

• Current state of the art
• Fungal metabolites
• Can work, but distinctive (unique) metabolites
  have been hard to find
• PCR for fungal DNA
• Plausible, but standardization has been a problem
• Antibody detection
• Epidemiological value, but
• Really frustrating overall
• Antigen detection
• Generally the strongest approach

• Diagnostics
• Best Drug Use
• Runaway Hits
• Finding Drugs

8
Antigen-Based Tools Other Fungi

• All are available commercially, but only two are
  licensed for use by local laboratories
• Cryptococcus (licensed)
• Capsular antigen Long history
• Aspergillus (licensed)
• Galactomannan cell wall antigen
• Histoplasma
• Polysaccharide cell wall antigen Wheat Lab
• Blastomyces
• Polysaccharide cell wall antigen Wheat lab
• Also WI-1 cell wall protein (Klein lab)

The other fungi are models for where Candida
might go
The Histoplasma Blastomyces tests are 510(k)
exempt and marketed without a license. Both are
offered by Immuno-Mycologics, Inc.
9
What about Candida?

• Metabolites
• D-arabinitol Limited, but not bad
• Not produced by C. glabrata or krusei
• Must adjust for renal clearance
• Candidemia 74 ve (Walsh, Am J Med, 95)
• PCR Plausible
• Works for candidemia, blood culture
• Not clear if helpful beyond that
• Antigens The great hope
• Undefined antigens? Variable data
• Cand-Tec, mannan(s) not so good
• Defined antigens? More promise
• Enolase, glucan lets look at these

Good reviews (all in Clin Micro Reviews) Yeo SF,
15465-84, 2002 Reiss E, 6311-323, 1993 (old,
but instructive) Jones JM, 332-45, 1990 (old,
but instructive)
10
Candida Enolase

• Enolase 48-kd cytoplasmic antigen
• Walsh 1991 High-risk cancer pts
• With repeat testing, found in 18 of 24 pts (75)
  with proven fungal infection
• Deep tissue or blood, not mucosal
• Not present in 140/146 who lacked proven disease
  (specificity 96)
• Mitsutake 96 39 pts w/ candidemia
• Sensitivity Specificity
• Enolase 72 100
• Glucan 88 84
• Pastorex mannan 100 26
• Cand-Tec 88 77
• This is not any current test but a homebrew

For details, see Walsh, NEJM 3241026-31,
1991 Mitsutake, J Clin Micro 341918, 1996
11
Beta-D-Glucan

• Two kits for this cell wall component
• Fungitec-G (marketed in Japan)
• Obayashi, Lancet 34517-20, 95
• Definite fungal infections 90 sensitive
• Glucatell (approaching US licensure)
• Various abstracts (see at left)
• Same antigen, different assay
• Principle for Fungitec-G Glucatell
• Limulus amebocyte lysate reaction
• Endotoxin triggers clot through Factor C
• Beta-D-glucan triggers through Factor G
• Kits Factor C removed ? Glucan specific
• Kits Different species of horseshoe crabs!
• South China Sea vs. North America
• Cut-offs differ, results NOT identical!

Glucatell abstracts ICAAC 2001, 841 ICAAC 2002,
M-902 ICAAC 2003, M-1028 ICAAC 2003,
M-1021 ICAAC 2002, M-1034a (latebreaker)
12
Glucatell for b-D-glucan Performance

• Positive cut-off 60 pg/ml
• Established from candidemia pts
• Ostrosky, ICAAC 03, M1034a
• Multicenter study of 189 with and 170 w/o IFI
  (invasive fungal infection)
• Based on a single sample/patient
• Sensitivity 70 Specificity 87
• Odabasi (Clin Infect Dis, in press)
• 283 during chemo for leukemia/MDS
• Samples 2x/week (7.3 samples/pt)
• One ve sample 100 sensitive for IFI
• Two ve samples 96 specific for IFI

Sampling frequency is critical!
IFI by EORTC/MSG (Ascioglu et al. CID 347, 2002
13
All beta-D-glucan assays share certain properties

• Other fungi
• Beta-D-glucan is present in released from
  Aspergillus, Fusarium, many other moulds
• (ICAAC 03, M-1021 Miyazaki, J Clin Micro
  333115, 95)
• But, not from Cryptococcus
• It mostly uses an alpha-glucan
• So, a positive result requires thought What
  fungus is likely?
• False positives Still learning.
• Gauze, some hemodialysis filters, some
  cellulose-filtered drugs

14
Candida Molecular Diagnostics Summary

• It is possible
• Beta-D-glucan, enolase, and even Cand-Tec do
  detect disease
• Glucan is not Candida-specific
• You have to think a little with it!
• What is needed next?
• Commercial support
• Validation, validation, validation
• Data in a variety of patient groups, various
  clinical settings, etc.
• And finally Can one of these become a surrogate
  marker for use in clinical trials?

The real goal for advancing the field Be able to
use to diagnose and monitor therapy. A high
hurdle!
15
Surrogate markers Some background

• Our current tools for diagnosis are poor.
  Miserable, to be honest.
• This cripples clinical studies
• Solid proof of infection and response is very
  hard to demonstrate. Enrollment is slow. Costs
  are high.
• A surrogate marker would help
• But, just having a diagnostic test is not enough!
  You must prove that the test links to disease.
  How?
• ICH E9 (Stats), Section 2.2.6
• Biological plausibility (YES)
• Link of marker to outcome (Not yet)
• Treatment effects on marker match clinical
  outcome (Not yet)

Personally, I think this is the central problem
of clinical mycology at present
ICH International Conference On Harmonisation
16
Best use of existing drugs

• Now playing at local theatres
• Amphotericin B, multiple flavors
• Fluconazole, itraconazole, voriconazole,
  ketoconazole
• 5-FC
• Terbinafine
• Caspofungin
• Coming attractions
• Micafungin, anidulafungin
• Posaconazole, ravuconazole
• Rumors intrigues
• Sordarins, other azoles, PLD-118, nikkomycin,
  Mycograb and other antibodies, and more.

• Diagnostics
• Best Drug Use
• Runaway Hits
• Finding Drugs

17
But, none of this helps todays patients

• What studies would really make a difference?
  (Multicenter, please!)
• Validation of serodiagnostic kits
• Make every effort to tie to outcome!
• Candidal prophylaxis in ICU
• Placebo-controlled with reproducible entry
  rules outcome measures
• Would also validate a predictive rule!
• Combination therapy
• Aspergillosis Combine drugs
• Caspofungin vs. vori vs. C V
• Its not Candida, but we need it!
• Candida Combine with cytokine?
• Hypothesis need less clear at present

Those drugs may or may not come to pass. What
more we do with our current set of tools?
18
Brouwer et al. ISHAM 0292003

• 64 adults, HIV, cryptococcal meningitis
• Serial quantitative CSF cultures
• Four therapy arms, 16 patients each
• AmB, AmB flucytosine, AmB fluconazole, all
  three together
• All drugs at full therapeutic doses
• Measured mean rate of fall of colony forming
  units (CFU) for crypto in CSF
• Greatest for AmB flucytosine
• It beat the other arms
• Better than AmB alone, P 0.004
• Better than AmB fluconazole, P 0.04
• Better than triple combo, P 0.02
• This result is especially meaningful due to data
  linking CFU clearance to outcome

This shows the power of a surrogate marker!
19
Immune System Fungal Infections

• Pappas et al. (JID, in press)
• IFN-gamma as adjuvant therapy for cryptococcal
  meningitis
• CSF cultures more often neg. at 2 wk
• Serum crypto antigen lower

Big hurdle must show superiority of the adjuvant
therapy
This is just one example of work in this area
20
A comment on susceptibility testing

• Better susceptibility testing was a frequent
  wish in the survey
• Surprise! Susceptibility testing for fungi is
  as good as for bacteria!
• At least for fluconazole and Candida
• Other bug-drug pairs are coming
• Key an MIC can only do so much
• The 90-60 rule
• Susceptible isolate? 90 response
• Resistant isolate 60 response
• Or, in short
• Some improve despite our therapy
• Some wont improve with any therapy

I really dont have time to do this topic
justice. Please see these papers for more Clin
Infect Dis 200235982-989 Antimicrob Agents
Chemother 200448693-715
21
Future Drugs What next?

• Another view Where are the holes?
• PO drug echinocandin-like Candida coverage
• Stunning. Vaginitis, OPC, IV follow-up
• If IV, could overtake IV candins
• IV PO for Aspergillus, Zygomycetes
• Without all the high-end azole quirks
• E.g., no interactions with vincas, cyclosporin,
  rifampin
• With Candida coverage, would rule cancer
  prophylaxis
• Better onychomycosis therapy
• Even just different would be good!

• Diagnostics
• Best Drug Use
• Runaway Hits
• Finding Drugs

22
Why is finding new drugs so hard?

• The desire for new antimicrobial drugs is intense
• CDC in 2001
• 2001 Public Health Action Plan to Combat
  Antimicrobial Resistance
• Action Item facilitate development by
  industry of novel therapeutic agents
• IDSA in 2003
• 12 March The Impending Crisis
• Since 1998, only 7 new antibacterial agents have
  been approved
• Of 290 agents listed in current major
  pharmaceutical pipelines, only 4 (1.4) are
  antibacterial agents
• But, why is it still so slow and hard?

• Diagnostics
• Best Drug Use
• Runaway Hits
• Finding Drugs

23
From a distance, drug development looks like this
PK Studies
Slide courtesy Francis Tally
24
But, theres more
Microbiology Studies
Discover drug
Slide courtesy Francis Tally
25
Consider just one of those other components
Example Hurdle 1

• Formulation development
• Tablet? Capsule?
• Salt? Prodrug?
• Even worse if you must account for two
  circulating forms of the drug
• Stability
• If it is going to sit on the shelf for 2 years,
  you have to prove it is stable
• Accelerated stability testing (heat, cool,
  repeat) gives some clues, but
• The real studies take 2 years
• Final formulation must go through full stability
  and safety testing

It is really hard to make changes once you are
even part way down this path
26
Why are (safe) antibiotics so difficult to find?
Example Hurdle 2

• Are there special issues here?
• Non-human target should be easy!
• But, curiously, it is not
• Can avoid mechanism-based toxicity by picking
  targets well
• Need to hit intra-microbial targets adds another
  barrier (efflux, etc.)
• Thus, we often use drug levels that are high
  relative to other drugs
• Digoxin typical levels 1 ng/ml
• Beta-lactams up to 100-200 mg/ml
• This puts you at risk for non-mechanism
  side-effects (hERG, etc.)

27
Paying for it all (1)
Net effect of all those hurdles

• How do CEOs make choices?
• How do you choose what to fund?
• Net Present Value (NPV)
• NPV is current value of future profits
• It can be risk-adjusted to account for relative
  likelihood of failure
• NPV of an antibiotic is often less than in other
  areas. Why?
• Antibiotics are used only briefly
• Unlike chronic use of (say) insulin
• Areas with greatest need may have less ability to
  afford
• Big problem for TB and malaria
• Government pricing pressures

A sensitive area, but one that must be addressed.
28
Paying for it all (2)

• Risk-adjusted NPV (million US) for a new
• Musculoskeletal agent 1,150
• Oncology agent 300
• IV-only Gram-positive agent 100
• Effect of possible requirement for tighter
  statistics (10 delta rule)
• Increase in time cost of project
• This causes significant drop in NPV
• Musculoskeletal agent 1,150
• Oncology 300
• IV-only Gram-positive agent 35

A positive NPV means that you do make a
profit. Bigger NPV is better.
Projan S. Curr Opin Microbiol 20036427-30
29
Paying for it all (3)

• Now, there is more to it than NPV
• Companies do want to spread risk
• All your drugs cant be in one class
• Antibiotics have the nice feature of low risk of
  development failure once youve shown them to be
  non-toxic

Antibiotics
Other drugs
Source Pharmaprojects, McKinsey Co. Graphic
courtesy Francis Tally
30
Paying for it all (4)

• There is more to it than NPV (cont.)
• Companies do want to fill unmet needs
• Social responsibility
• Ultimately, we are all patients!
• But, profit is required
• Without profit on drug 1, there will be no drug
  2
• Recall that cost of 900 million to bring ONE
  drug to market
• Lots of failures along the way
• The process takes years and people
• And lots of optimism!

It is frustrating to hear The drugs are too
expensive. Actually, the disease is
expensivegood drugs are both life-saving and
often a relative bargain.
New drug 800 95 post-approval commitments
895. DiMasi JAJ Health Econ 200322151-85.
31
Paying for it all (5) Balance!

• First, be curious!
• Drugs can create a niche
• Caspofungin 200m/yr to 1bn/yr?

Science Unmet need
Avoid paralysis by analysis! First is not always
best. Me-Too can have value (NEJM 350211, 04)
Commercial Analysis
32
Finding new drugs Summary

• Drug discovery and development is
• Hugely expensive
• Always a gamble
• It is also a partnership
• Regulatory requirements process
• The agencies have an impossible job there is no
  way to guarantee perfectly safe drugs at
  perfectly affordable prices
• Government pricing agreements
• Antimicrobial discovery development is indeed
  underway
• But it must be carefully nurtured!

There are a lot of efforts ongoing to regularize
and smooth the overall process
33
Thank you!

• Any solution to a problem changes the problem.
• R. W. Johnson
• Life would otherwise be boring, no?