1. Jean KLASTERSKY, M.D., Ph. D. Institut Jules Bordet, Brussels, Belgium
ANTIFUNGAL THERAPY IN FEBRILE NEUTROPENIC PATIENTS REVIEW OF TREATMENT CHOICES AND STRATEGIES
Jean KLASTERSKY, M.D., Ph. D.
Institut Jules Bordet, Brussels, Belgium

2. G.S. MARTIN et al., NEJM 2003

3. Mc Neil M, CID 2001;13:;

4. Frequency of non-Aspergillus mould infections at Fred-Hutchinson Cancer Research Center (Seattle). The number of patients who developed proven or probable infection with Fusarium species, Zygomycetes and Scedosporium species from 1985 through 1999 are shown

5. fluconazole for prophylaxis (1984-1997).
Pathogenic Candida species in BMT recipients : Candida species that caused candidemia are
compared over 2 decades at the Fred Hutchinson Cancer Research Center. The incidence of
candidemia decreased from 11.4 % in (72) compared to 4.6 % after adoption of
fluconazole for prophylaxis ( ).
Marr K & Bowden R, Transplant Infectious Diseases 1999;1:

6. Antifungal prophylaxis in leukemia patients :
ECIL recommendations

7. Comparative trials of antifungal agents in candidemia and invasive candidasis
Response rate
Overall mortality
Fluconazole vs amphotericin B
70 % vs 79 %
33 % vs 40 %
Itraconazole vs amphotericin B
35 % vs 41 %
40 %
ABLC vs amphotericin B
63 % vs 68 %
41 % vs 39 %
Caspofungin vs amphotericin B
73 % vs 62 %
30 % vs 34 %
Voriconazole vs amphotericin B
41 % vs 41 %
36 % vs 42 %
Micafungin vs Ambisome
89 % vs 89 %
Anidulafungin vs fluconazole
75 % vs 60 %
26 % vs 31 %

8. Why does the frequency of fungal infection increase ?
I.V. devices
Immunosuppression
Neutropenia
Broad spectrum antibiotics
Diabetes

9. Rational for Empirical Antifungal Therapy in
Neutropenic Patients with Persistent Fever
Delayed treatment increases
mortality
Success of antibacterial
empirical therapy
Early diagnosis of many
fungal infections is difficult

10. Follow Daily And Reassess After 72 Hours
An Algorithm for Therapy of Febrile Neutropenia after Initial Empirical Therapy with Broad Spectrum Antibiotics
Follow Daily And Reassess After 72 Hours
Clinical response
Yes No
Continue for 7 days
Pathogen isolated
Yes No
Adjust to sensitivity
Look for localized infection
Use G/GM-CSF?
Repeat cultures and serology
Perform chest CT and BAL
Add amphotericin, and possibly metronidazole, antivirals and/or G/GM-CSF as indicated
Look for non infectious causes of fever 10

11. Possible Causes of Persistent Fever
Approximate frequency in high risk patients (%)
Fungal infections susceptible to empirical therapy 40
Fungal infections resistant to empirical antifungal therapy 5
Bacterial infections (with cryptic foci and resistant organisms) 10
Toxoplasma gondii, mycobacteria, or fastidious pathogens (legionella, mycoplasma, chlamydia, bartonella)
Viral infections (cytomegalovirus, Epstein-Barr virus, human herpes virus 6, varicella-zoster virus, herpes simplex virus) and respiratory pathogens such as parainfluenza virus, respiratory syncytial virus, influenza viruses
Graft-versus-host-disease after hematopoietic stem-cell transplantation
Undefined (e.g. drug fever, toxic effects of chemotherapy, antitumor responses, undefined pathogens) 25
L. COREY and M. BOECKH, NEJM 2002

14. Infectious Complications in Each of the Groups Following Randomization
Randomization group N°
Clinically documented
Bacterial
Fungal
Viral Protozoal
Shock
Discontinue KGC 16 2 3 1 6
Continue KGC alone 5
KGC + Ampho B 18
KGC : cephalotin (Keflin), gentamicin, carbenicillin
P.A. PIZZO et al, Am J Med 1982 14

15. Clinical Response in Persistently (4 days) Febrile Neutropenic Patients
Ampho B
No Ampho B
p Value
Overall
All patients
55/80 (68)
39/77 (50)
0.10
Prior antifungal prophylaxis No
21/27 (78)
9/20 (45)
0.04
Yes
19/31 (61)
24/39 (61) NS
Infection documentation
Clinical
22/29 (75)
14/31 (41)
0.03
Possible
25/39 (64)
20/33 (60)
Granulocyte count at Day 4
< 100/µL
31/45 (69)
20/43 (46)
0.06
/µL
16/23 (70)
14/21 (67)
EORTC-IATCG, Am J Med 1989

16. Causes of Death with or without Empirical Ampho B in Persistently (4 Days) Febrile Neutropenic Patients
Ampho B
No Ampho B
Overall mortality (at day 30) * 11 14
Fungal infection** 4
Bacterial infection 1 2
Pulmonary infection (no bacterial or fungal pathogen identified at autopsy)
Other 9 8
(*) p = NS between the two groups (**) p = 0.05 between the two groups
EORTC-IATCG, Am J Med 1989

17. Randomized Studies Comparing Empirical Treatment with Antifungal Agents for Persisting Fever during Neutropenia
YEAR
STUDY
ANTIFUNGAL AGENTS COMPARED
1982
1989
1996
1998
1999
2000
2001
2002
2004
Pizzo et al
EORTC
Viscoli et al.
Malik et al.
White et al..
Walsh et al.
Winston et al.
Wingard et al.
Boogaerts et al.
Conventional ampho B vs no antifungal therapy
Conventional ampho B vs fluconazole
Conventional ampho B vs ampho B colloidal dispersion
Conventional ampho B vs liposomal ampho B
Liposomal ampho B vs ampho B lipid complex
Conventional ampho B vs itraconazole
Liposomal ampho B vs voriconazole
Liposomal ampho B vs caspofungin

18. Measures of the Success (%) of Empirical Antifungal Therapy with Conventional or Liposomal Amphotericin B, Voriconazole or Caspofungin
Ampho B Liposomal Ampho B Voriconazole Caspofungin (Ampho B) (Vori) (Caspo)
N° of patients
Overall success
Resolution of fever
No breakthrough
fungal infection
Resolution of
baseline infection
Survival for 7 day
No discontinuation
for toxic effects
or lack of efficacy
J. KLASTERSKY, NEJM 2004

19. Comments on the Walsh’s studies
Large prospective controlled trials
Composite score : « common language »
BUT
Survival and fever can be influenced by many other factors than just the nature of the empirical regimen
What is the difference between « baseline » FI (<72h) and « breakthrough » FI (>72h) ?
Discontinuation for toxicity or lack of efficacy :
« mixing apples and pears » ?

20. Toxicity of Empirical Antifungal Therapy %
Conventional amphotericin
Liposomal amphotericin
Voriconazole
Caspofungin
Total n° patients
344
1312
415
564
Chills
Nephrotoxicity* 54 34 24 13 7 14 3
Discontinuation for toxicity 18 9
5** 5
* Creatinine increase 2 x base line
** Visual hallucinations more frequent with voriconazole (4.3 % vs 0.5 %)

21. Measures of the Success (%) of Empirical Antifungal Therapy with Conventional or Liposomal Amphotericin B, Voriconazole or Caspofungin
Ampho B Liposomal Ampho B Voriconazole Caspofungin (Ampho B) (Vorico) (Caspo)
N° of patients
Overall success
Resolution of fever
No breakthrough
fungal infection
Resolution of
baseline infection
Survival for 7 day
No discontinuation
for toxic effects
or lack of efficacy
J. KLASTERSKY, NEJM 2004

22. Cure of Base Line Fungal Infection
Conventional ampho B
Liposomal ampho B
Voriconazole
Caspofungin
Total n° of patients
344
961*
415
556
Response/n° (%)
8/11
(72.7)
20/44 (45.4)**
6/13
(46.2)
14/27
(51.9)
* 2 studies
**Outcome in the study comparing conventional ampho B to
liposomal ampho B was 9/11 (81.8)

23. Choice of a suitable denominator for the evaluation of empirical therapy based on microbiological results
(T. Walsh)
(J. Klastersky)
Success of therapy
Failure of therapy
No breakthrough FI
Total n° patients -
Cure of baseline FI
N° of baseline FI
Breakthrough FI
Total of n° of patients
No cure of baseline FI

24. Outcome of Empirical Antifungal Therapy in Microbiologically Demonstrated Fungal Infections (FI)
Ampho B
Liposomal ampho B
Voriconazole
Caspofungin
(344)
(961)
(415)
(556)
Breakthrough FI
37 (10.8)
45 (4.6)
8 (1.9)
29 (5.2)
No cure of base line FI
3 (0.8)
22 (2.2)
7 (1.6)
13 (2.3)
Total failures*
40 (11.6)
67 (6.9 %)
15 (3.6 %)
42 (7.7 %)
*p = 0.03

25. % of invasive fungal infections
Summary of trials of empirical antifungal therapy that evaluated alternatives to amphotericin B
Study drugs
% of invasive fungal infections n
Arm 1
Arm 2
687
AmB
L-Amb
8.7
5.0
384
Itraconazole
2.7
837
L-AmB
Voriconazole
1.9*
1095
Caspofungin
4.3
5.2
* p< 0.5
J.R. Wingard, CID, 2004

26. 26

27. Aspergillus fumigatus

29. Galatomannan detection for the diagnosis of invasive aspergillosis
Approved by FDA; standard : optical density index > 0.5 in 2 consecutive samples ?
Positivity can preceed radiological findings
In probable or proven cases levels are often higher and increase within days
81 % sensitivity; 89 % specificity; NPV:98 %;
PPV:7-94 %
False negatives : prophylactic use of mold-active azoles, early antifungal therapy and others ? False positives : use of piperacillin-tazobactam and others ?

30. Aspergillus Infections (AI) in Empirical Therapy
Liposomal ampho B
Voriconazole
Caspofungin
(961)
(415)
(556)
Breakthrough AI 21 4 10
No cure of base line AI°
13*
4** 7
Total failures
24 (2.4 %)
8 (1.9 %)
17 (3.0 %)
° Assuming that 1/2 infections in the liposomal ampho B arm (*) and
4/7 infections in the voriconazole arm (**) were caused by Aspergillus

31. R. Herbrecht et al, N Engl J Med, 2002

32. T.J. Walsh et al, N Engl J Med 2002
1.9 %
4.9 % (p = 0.02)
T.J. Walsh et al, N Engl J Med 2002

33. T.J. Walsh et al, N Engl J Med 2004

34. T.J.Walsh et al, N Engl J Med 2004

35. 12 CR
6 PR voriconazole
2 SD rescue
12 No response (7 died < IFI) 8/12 : CR

36. K.A. MARR et al, CID 2004

37. N. Singh, Transplantation, 2006

38. The prevalence of fungal infections in patients receiving empirical therapy is 4-8 %

39. Prevalence of fungal infections in persistently neutropenic patients not receiving empirical therapy
Pizzo et al. (1982) 18
EORTC (1989)
28 *
Guiot et al. (1993)
26*
Corey and Boeckh (2002) 45
Maertens et al. (2005) 21
* Autopsy-based data

41. CID, 2005

42. The preemptive approach (adapted from Maertens et al
The preemptive approach (adapted from Maertens et al.) « Possible IFI »
Diagnostic evaluation for IFI
Reasons
Procedures
. FN refractory for 5 days
. FUO relapsing after 48 h with N
. Signs or/and symptoms suggestive of IFI
. Isolation of molds/hypae in URT
. Galactomannan assays positive (> 0.5)
. HRCT of the chest (+ sinuses)
. Bronchoscopy +BAL
- Smears + cultures for bacteria, fungi, mycobacteria, Legionella
- PCR for CMV, HSV, VZV, Toxoplasma, Pneumocystis, Mycoplasma, Chlamydia

43. The preemptive approach (adapted from Maertens et al
The preemptive approach (adapted from Maertens et al.) « Possible IFI »
Liposomal amphotericin B (5 mg/kg) IF
> 2 consecutives EIA for galactomannan assays
> 0.5) OR
CT suggestive of IFI supported by microbiology

44. The preemptive approach of persistent febrile neutropenia in 88 patients (adapted from Maertens et al.)
Persistent fever 35/117 (29 %) episodes
Prevalence of IFI : 22 % (mortality 36 %)
No aspergillar infection was missed
Early therapy could be initiated in clinically not suspected cases
Significant (78 %) reduction in use of antifungals

45. CID, 2005

46. Empirical versus preemptive therapy (liposomal amphotericin B) in patients with persisting fever and neutropenia
Adapted from Maertens’ study
35 patients with persistent fever
EMPIRICAL APPROACH *
PREEMPTIVE APPROACH
1 episode of IA (2.4 %)
10 episodes of IA (28 %)
? Death
2 deaths (20 %)
need of a controlled study
* Estimation from J. Klastersky, NEJM 2004

47. Blood, 2006

48. 293 patients c hematological malignancies
Empirical (E ) vs pre-emptive approach (PE)
Design
293 patients c hematological malignancies R
Antifungal therapy E PE
. Persistent or recurrent fever
. Persistent or recurrent fever +
. pneumonia
. mucositis
. septic shock
. sinusitis
. skin lesions
. aspergillus colonisation
. + gamacto-mannan

49. Empirical (E ) vs pre-emptive approach (PE)
Results E
150 patients PE
143 patients
Diagnosed IFI
4 (2.6 %)
13 (9.0) p < 0.02
Overall survival
147 (98 %)
136 (95 %) NS
IFI related mortality
0 (0 %)
3 (2.1 %) p = 0.12
Mean cost (euros)
3595 NS

50. Estimated prevalence of invasive fungal infection (IFI) in neutropenic patients according to the management strategy in exemplative studies
N° patients
N° IFI
Prevalence %
Controls ¹
341 14
26.0
Prophylaxis ² (posaconazole)
291 7
2.4
Empirical therapy ³
(voriconazole)
415 15
3.6
Pre-emptive therapy (polyenes) 4
143 13
9.0
1 Guiot, CID, 1994
2 Ullmann, NEJM, 2007
3 Walsh, NEJM, 2002
4 Cordonnier, Blood, 2006

51. (amphotericin B < lipid preparations < voriconazole = caspo)
Which Antifungal Drug is Best for the Empirical Treatment of Patients with Febrile Neutropenia?
Few adverse effects
(amphotericin B < lipid preparations < voriconazole = caspo)
Superiority of action :
conventional amphotericin looks to be the least effective
voriconazole might be superior in reducing microbiologically proven failures
Cost
Resistance

52. Average Daily Cost of Antifungal Therapy (Jules Bordet Bruxelles)
Dose
Euros
Conventional Amphotericin B
1 mg/kg 8
Liposomal amphotericin B
3 mg/kg
629
Voriconazole
400 mg IV
407
400 mg PO 84
Caspofungin
70 mg IV
644

53. Sensitivity of common pathogenic fungi
Amphotericin B
Voriconazole
Caspofungin
Candida sp. S
C. lusitaniae R
C. krusei
S**
C. glabrata
Trichosporon sp.
Aspergillus sp.*
Fusarium sp.
Mucorales
C. neoformans
*Except terreus
**MIC are higher than for C. albicans but still < 3 ng/ml

54. Antifungal therapy in persistently febrile neutropenic patients
Conclusions
Voriconazole is presently the optimal choice for therapy and empirical therapy of suspected aspergillar infections
Early therapy (empirical) is needed since mortality remains high (30-50 %)
Empirical therapy reduces the incidence of IFI from + 25 % to %; the corresponding figures for liposomal ampho B and caspofungin being + 7 %
Experience with the preemptive approach is still limited; it is demanding in terms of diagnostic procedures but it reduces the rate of overtreatment.
There is evidence so far that pre-emptive therapy is inferior in efficacy and not more cost effective than the empirical approach

55. Needs for Improvement of Neutropenic Febrile Patients with Persistent Fever to Avoid Overtreatment
Better diagnostic tools for EARLIER microbiological and clinical diagnosis : galactomannan, PCR, those caused by Aspergillus sp.
More accurate IDENTIFICATION OF PATIENTS at high risk of fungal infections (cf MASCC score)

56. Thank you for
your kind attention
and
« Au revoir »