Early Diagnosis and Therapy for Fungal Infections

Debra Goff PharmD, FCCP

Clinical Associate ProfessorClinical Associate Professor

Infectious Disease Specialist

The Ohio State University Medical Center

Columbus Ohio, USA

The Ohio State University Medical Center

• James Cancer Center 165 bedsbone marrow transplants

• Ross Heart Hospital 90 bedsHeart and lung transplants

• The Ohio State University Hospital 850 bedssolid organ transplantSICU, MICU, NICU, Burn unit

The Fungal World

FUNGI

YEASTS MOULDS

Candida spp. CryptococcusTrichosporon Zygomycetae Septate Fungi

DimorphicFungi

Rhizopus

Mucor

Absidia

Fusarium

Aspergillus spp.

Paecilomyces

Hyalohyphomycoses

Significance of Candida Infections

• Candida species represent the most common cause of systemic fungal infections and a major cause of mortality among compromised hosts1,2

• Sepsis due to fungi increased >200% in past 20 years3

• Outcomes attributable to candidemia/candidiasis

IDM00126-MK-1 4

• Outcomes attributable to candidemia/candidiasis• Excess medical costs of $216-$281 million/year1

• Increased LOS (up to 34 days)1

• 38%-72% mortality2,4

• However, the true incidence is still underestimatedand frequently is a postmortem diagnosis3

LOS, length of stay.

1. Rentz AM, et al. Clin Infect Dis. 1998;27(4):781-788. 2. Wey SB, et al. Arch Intern Med. 1988;148(12):2642-2645.3. Martin GS, et al. N Engl J Med. 2003;348(16):1546-1554.4. Fraser VJ, et al. Clin Infect Dis. 1992;15(3):414-421.

• In a retrospective study in 100 adult MICU patients:• 81% of antemortem diagnoses confirmed on autopsy

• Most frequent diagnosis: bacterial pneumonia with MODS

• Major missed diagnoses (N = 36)• Class I errors (n = 22)

• Autopsy findings revealed a diagnosis that if known might have

Comparison of Antemortem Clinical Diagnoses in Crit ically Ill Patients and Subsequent Autopsy Findings

• Autopsy findings revealed a diagnosis that if known might have led to a change in therapy

• Invasive fungal infection (5 [22%])• Cardiac tamponade (5), abdominal bleeds (4), MI (3)

• Class II errors (n = 14)• No effect on outcome• Cancer (6), small bowel infarction (3), acute hepatitis (2),

pulmonary embolism (2)

5

MI, myocardial infarction; MICU, medical intensive care unit; MODS, multiple organ dysfunction syndrome.

Roosen J, et al. Mayo Clin Proc. 2000;75(6):562-567.

Incidence and Severity of Invasive Fungal Infections (IFIs): BSIs

Most common pathogens and associated mortality rate sin ICU patients with BSI (N=10,515)

6

Results from a nationwide surveillance study of patients who developed nosocomial BSI either inthe ICU or on a non-ICU ward; the data shown here are for the 10,515 patients with ICU-onset of BSI.

BSI=bloodstream infections; CoNS=coagulase-negative staphylococci; ICU=intensive care unit.

Wisplinghoff H, et al. Clin Infect Dis. 2004;39(3):309-317.

CoNS Staphylococcus

aureus

Candida spp Enterococcus spp

Nosocomial Candidemia:Associated Mortality

Historical Perspective on Candidemia

40

50

60

70

38%

57%

n=88 pairs

40

50

60

70

49%

62%

n=108 pairs

1. Wey SB et al. Arch Intern Med. 1988;148:2642-2645.2. Gudlaugsson O et al. Clin Infect Dis. 2003;37:1172-1177.

Cases Controls

0

10

20

30

40

1983 to 1986

38%

1

19%

(26%-49%)

0

10

20

30

40

1997 to 2001

49%

2

12%

(38%-60%)

Candidemia: LOS and Hospital Charges for Patients W ith and Without Candidemia

With Candidemia Without Candidemia

18.6

10.1 Days*$66,154

$39,331*

8

*Attributable increase (95% CI).LOS=length of stay.

Zaoutis TE, et al. Clin Infect Dis. 2005;41(9):1232-1239.

Mea

n D

ays

8.5

Total Hospital LOS

Mea

n C

harg

es

Total Charges/Patients

$26,823

Additional Factors That Increase Risk of Invasive Candidiasis

Prolonged antibiotic use 3Total parenteral

Broad-spectrum antibiotics 2

More than 2 days of mechanical

ventilation 2

Neutropenia 1

Immunosuppression 1,2

Risk for InvasiveRisk for Invasive

9

APACHE, Acute Physiology and Chronic Health Evaluation.

1. Pappas PG, et al. Clin Infect Dis. 2009;48(5):503-535.2. Ostrosky-Zeichner L, et al. Crit Care Med. 2006;34(3):857-863.3. Pappas PG, et al. Clin Infect Dis. 2004;38(2):161-189.

antibiotic use 3

ProlongedICU stay 2,3

High APACHE II score 2Central venous

catheter 1-3

Total parenteral nutrition 2,3

Candidacolonization

≥≥≥≥2 sites 2,3

Risk for InvasiveRisk for InvasiveCandidiasisCandidiasis

Diagnostic Challenges:Aspergillosis and Candidiasis

• Variable and nonspecific clinical presentation1

• Lack of availability of one universally applicable diagnostic test1

Delay in initiation of therapy1

• Lack of distinct clinical presentation3

• Tissue culture4

• A positive culture, considered definitive, may require invasive biopsy, which may not be

AspergillosisAspergillosis Candidiasis Candidiasis

• Delay in initiation of therapy1

• Risk for potentially fatal progression of disease

• Diagnostic criteria exist for proven and probable IA2

• Although useful in clinical trials, less valuable in practice

biopsy, which may not be feasible in critically ill patients

• Blood culture4,5

• Relatively insensitive

• Nonsterile site cultures may reflect colonization or contamination vs infection4

• Catheter culture sensitivity varies with methodologies3,5

1. Denning DW. Clin Infect Dis. 1998;26(4):781-805. 2. Subirà M, et al. Ann Hematol. 2003;82(2):80-82.3. Rodriguez LJ, et al. Adv Pharmacol. 1997;37:349-400. 4. Dean DA, et al. Am J Surg. 1996;171(3):374-382.5. Rex JH, et al. Adv Intern Med. 1998:43:321-371.

Diagnostic Challenges of Candidiasis

• No distinct manifestations1

• Colonization vs infection2

• Blood culture positivity rates ~40%-60%3

• ↓ in patients on prior antifungals4

• ↓ if a non-albicans Candida species (~20%-40%)4

• High index of suspicion3

• High-risk group5

1. Rodriguez LJ, et al. Adv Pharmacol. 1997;37:349-400. 2. Dean DA, et al. Am J Surg. 1996;171(3):374-382.3. Vazquez JA, Sobel JD. Candidiasis. In: Dismukes WE, et al, eds.

Clinical Mycology. New York, NY: Oxford University Press; 2003:143-187.4. Kami M, et al. Br J Haematol. 2002;117(1):40-46.5. Pfaller MA, et al. J Clin Microbiol. 2001;39(9):3254-3259.

New Diagnostic Techniques

• Antibodies1

• Disappointing so far• Metabolites1

• D-arabinitol: useful, but not practical• Fungal cell wall components2,3

• New assays for β-D-glucan and galactomannan show promisepromise

• Fungal PCR4

• Real-time PCR can contribute to rapid diagnosis• PNA FISH5

• Potential for rapid identification of C. albicans

1. Yeo SF, Wong B. Clin Microbiol Rev. 2002;15(3):465-484.2. Odabasi Z, et al. Clin Infect Dis. 2004;39(2):199-205.3. Sulahian A, et al. Cancer. 2001;91(2):311-318.4. Schabereiter-Gurtner C, et al. J Clin Microbiol. 2007;45(3):906-914. 5. Wilson DA, et al. J Clin Microbiol. 2005;43(6):2909-2912.

PCR, polymerase chain reaction; PNA FISH, peptic nucleic acid fluorescence in situ hybridization.

Nonculture Diagnostics: Benefits and Limitations

• β-D-glucan from cell wall1 (Fungitell BG)

• Detects Candida spp. and Apergillus• Sensitivity for Candida pathogens 78%-90%, depending on cut-off values (>80

pg/ml is positive)• High false + in patients with bacterial infections (54-68%)• False + with hemodialysis, surgical gauze, albumin, immunoglobulin

• Galactomannan from Aspergillus cell wall2,3 Bio-Rad Platelia EIA

• Sensitivity and specificity of BAL GM >90%

13

• Positive and Negative predictive values are 76% and 96%• False + in pts receiving pip/tazo, amoxicillin or other beta-lactamase inhibitors

• PCR for fungal DNA4

• Able to distinguish among clinically significant Candida and Aspergillus spp.• Real-time PCR provides rapid results < 6 hours• Positive predictive value 100% Negative predictive value 99%

• PNA FISH5

• Probe for identifying C. albicans• Positive predictive value, 100%; negative predictive value, 99%

1. Ostrosky-Zeichner L, et al. Clin Infect Dis. 2005;41(5):654-659. 2. Sulahian A, et al. Cancer. 2001;91(2):311-318.3. Maertens J, et al. Blood. 2001;97(6):1604-1610. 4. Schabereiter-Gurtner C, et al. J Clin Microbiol. 2007;45(3):906-914. 5. Wilson DA, et al. J Clin Microbiol. 2005;43(6):2909-2912.

Multiplex PCR detection

enhancement of bacteremia and

• Objective : test a multiplex RT-PCR method for simultaneous detection of multiple organisms in bloodstream infections

bacteremia and fungemia

• Methods : Prospective observational study of

200 patients at risk of BSI with signs of SIRS.

Louie R. et al 2008 CCM :36(5);1487-1492.Tsalik E et al 2010 JCM 48(1); 26-33.

Organisms detected by multiplex PCR

Louie R. et al CCM 2008:36(5);1487-1492.

Results

PCR detected bacteria/fungi in 45 cases vs 37 by blood culture.

PCR detected mecA in all 3 PCR detected mecA in all 3 culture confirmed MRSA

PCR did not detect E. faecalisin 5 BC confirmed cases

7 samples could be tested simultaneously in 6.54 hours

Louie R. et al CCM 2008:36(5);1487-1492.

Conclusion• Despite limitations of both blood culture

and RT multiplex PCR methods1.PCR could be an adjunct to BC2. PCR can facilitate early detection2. PCR can facilitate early detection3. Early detection can facilitate evidence-based treatment decisions

Louie R. et al 2008 CCM 36(5);1487-1492.Tsalik E et al 2010 JCM 48(1); 26-33.

Overview of Antifungal Management Strategies

• Highest-risk patient (eg, HSCT for Candida)1

• No infection

• High-risk patient with persistent fever despite antibiotics2

• Possible infection

• High index of suspicion (based on signs and symptoms) but without definitive diagnostic proof3

• Probable infection

• Full-blown disease4

• Proven infection

• For patients refractory to or intolerant of primary therapy1

18

Prophylaxis Empirical Therapy

Presumptive Therapy

Refractory Therapy

Increasing certainty of fungal infection

Specific Treatment

1. Pappas PG, et al. Clin Infect Dis. 2009;48(5):503-535. 2. Hayes-Lattin B, et al. Leuk Lymphoma. 2004;45(4):669-680.3. Dean DA, et al. Am J Surg. 1996;171(3):374-382. 4. Denning DW. Clin Infect Dis. 1998;26(4):781-805.

A Multi-Institutional Study of Antifungal Use in Surgical Intensive Care Units

Indication for Antifungal Therapy*• Empiric 44%

• Preemptive 43%• Preemptive 43%

• Definitive 12%

*Evaluated on the start date of antifungal therapy.Garey KW et al. Mycoses. 2006;49:226–231.

Infections Caused by Non-albicans Candida Are Increasing

60

80

100

% o

f Tot

al N

umbe

r of

Cas

es

C. parapsilosisC. tropicalis

C. albicansOther

20

0

20

40

1997-1998 1999 2000 2001 2002 2003� Neither C. glabrata nor C. krusei showed a consistent increase or decrease in isolation rates overall� Increased rates of isolation of C. tropicalis (4.2% to 7.5% increase) and C. parapsilosis (4.6% to 7.3%

increase) were observed between 1997 and 2003

% o

f Tot

al N

umbe

r of

Cas

es

Pfaller MA, Diekema DJ. Clin Microbiol Rev. 2007;20(1):133-163.

C. tropicalisC. glabrataC. krusei

Why Should Candida Spp. Be Identified?

• C. albicans1

• C. glabrata—less susceptible to all antifungals1,2

• C. parapsilosis—catheter related1

• Reduced echinocandin susceptibility

• C. tropicalis1

• C. krusei—“neutropenics”1

• Intrinsic azole resistance, less susceptible• Decreased susceptibility to AMB

• C. guillermondii• C. lusitaniae AMB resistance 1

AMB, amphotericin B.

1. Pfaller MA, Diekema DJ. Clin Microbiol Rev. 2007;20(1):133-163. 2. Pfaller MA, et al. Clin Microbiol Infect. 2004;10(suppl 1):11-23.

Importance of C. glabrata?

• US incidence: 20%-24% of all BSIs1

• Susceptibility2

• 10%-15% resistant to fluconazole• 46%-63% resistant to itraconazole• Also less susceptible to all antifungals, including AmB

IDM00126-MK-1 22

• IDSA recommended antifungal therapy3:• Echinocandins—drug of choice• Voriconazole or AmB

OR • Echinocandin followed by azole

1. Pfaller MA, et al. Clin Microbiol Rev. 2007;20(1):133-163.2. Pappas PG, et al. Clin Infect Dis. 2004;38(2):161-189. 3. Pappas PG, et al. Clin Infect Dis. 2009;48(5):503-535.

Prediction of Non- Albicans Candidemia

• Retrospective case series of 245 patients (60% in t he ICU) in 2 academic, tertiary care centers

• C albicans in 52% of infections and C glabrata in 2 0% • No variable, including both previous fluconazole

exposure and severity of illness, correlated with t he

Shorr AF et al. Crit Care Med. 2007;35:1077-1083.

exposure and severity of illness, correlated with t he fungemia due to a non-albicans species

• Conclusion: Simple clinical factors do not allow the clinician to effectively identify patients likely i nfected with non-albicans pathogens or with possible fluconazole-resistant fungi

Challenges in the Management of Candidiasis

• The longer the wait, the higher the mortality rate1

• nonspecific clinical presentation2

• No single, sensitive, universally applicable test available for establishing diagnosis1,3

24

available for establishing diagnosis1,3

• Timing of therapy1

• Waiting for definitive proof increases the risk for potentially fatal progression of disease

1. Wheat LJ, et al. Clin Chest Med. 2009;30:367-377.2. Rodriguez LJ, et al. Adv Pharmacol. 1997;37:349-400.3. Vazquez JA, Sobel JD. Candidiasis. In: Dismukes WE, et al, eds.

Clinical Mycology. New York, NY: Oxford University Press; 2003:143-187.

Mortality Related to Untreated Candidemia

• Times between admission and onset of candidemia� Median 13 days� Mean 20.8 days

•

Gudlaugsson O et al. Clin Infect Dis. 2003;37:1172-1177.

• Deaths in 12/15 who never received antifungal therapy � 11/12 deaths within 72 hours after the first Candid a-

positive blood cultures were obtained (and before positive culture results were available)

Delaying Empirical Treatment of Positive Candida Bloodstream Infections Until Positive Cultures Are Available

• Retrospective cohort analysis of 157 patients with candidemia� Deaths in 50 (31.8%) of patients

• Definition of “inappropriate treatment” in this stu dy:� “The absence of antifungal agents at the time that f ungus-positive

blood samples for culture were drawn”� “Fluconazole treatment with the subsequent isolation of either � “Fluconazole treatment with the subsequent isolation of either

Candida krusei or Candida glabrata”

• Timing of the administration of antifungal therapy: � Within 12 hours in 9 (5.7%) patients� Between 12 and 24 hours in 10 (6.4%) patients� Between 24 and 48 hours in 86 (54.8%) patients� Greater than 48 hours in 52 (33.1%) patients

Morrell M et al. Antimicrob Agents Chemother. 2005;49:3640-3645.

Delaying Antifungal Treatment Has Been Associated With Increased Mortality

Delaying

Hospital mortality based on timing of antifungal th erapy

15

20

25

30

35

40

Per

cent

Hos

pita

l Mor

talit

y

0

5

10

15

<12 12 to 24 24 to 48 >48

Delay in Start of Antifungal Treatment (hours)

Per

cent

Hos

pita

l Mor

talit

y

• Independent determinants (by multivariate analysis) of hospital mortality� APACHE II score (1-point increments) (P <.001)� Prior antibiotic treatment (P = .028)� Administration of antifungal treatment 12 hours aft er having the first positive

blood sample for cultures (P = .018)APACHE = acute physiology and chronic health evalua tion.Morrell M et al. Antimicrob Agents Chemother. 2005;49:3640-3645.

Impact on Mortality of Candidemia Based on Time to Initiation of Antifungal Therapy

• Retrospective cohort study of 230 patients from 4 medical centers

• 162 patients (70%) with nonsurgical hospital admiss ion• C albicans most commonly isolated (56% of patients)• 192 patients with no previous fluconazole treatment• 192 patients with no previous fluconazole treatment• Mortality rates based on time of initiation of

fluconazole (P = .0009 for trend)

Garey KW et al. Clin Infect Dis. 2006;43:25-31.

Day 0 Day 1 Day 2 Day ≥314/92 patients

(15%)9/38 patients

(24%)12/33 patients

(37%)12/29 patients

(41%)

Effect of Antifungal Therapy Timing on Mortality in Patients with Candidemia

106 episodes analyzedIncubation: time from BC collection to positivityProvider notification: time from positivity to provider notificationAntifungal initiation: time from provider notification to to 1st dose

Taur Y, et al. Antimicrob Agents Chemother 2010;54(1):184-90.

Delay in Therapy Increases MortalityHR 1.025, P = 0.001

HR 0.516, P = 0.391

Associated with mortality

HR 0.989, P = 0.562

Taur Y, et al. Antimicrob Agents Chemother 2010;54(1):184-90.

Impact of Inadequate Antifungal Therapy on Crude Mortality in Candidemia

54%

(38/70)

34%

(37/108)

42%

(90/214)

40

50

60

Cru

de

Mo

rta

lity

(%

)

Region 1(Connecticut)

Region 2(Baltimore/Baltimore County)

56%

(44/78)

31%

(56/179)

39%

(206/529)40

50

60

IDM00126-MK-1 31

*Adequate treatment was defined as any systemic antifungal medication administeredfor a minimum of 7 days after the first Candida-positive blood culture.

Adapted from Morgan J, et al. Infect Control Hosp Epidemiol. 2005;26(6):540-547.

(37/108)

23%

(182/789)

0

10

20

30

Cru

de

Mo

rta

lity

(%

)

Controls All Candidemia

Candidemia

+ Adequate*Treatment

+ InadequateTreatment

(56/179)

15%

(309/2065)

0

10

20

30

Controls All Candidemia

Candidemia

+ Adequate*Treatment

+ InadequateTreatment

Surviving Sepsis Campaign: International Guidelines for Management of Severe Sepsis and Septic Shock, 2008

• “Clinicians should also consider whether candidemia is a likely pathogen when choosing initial therapy. ”

• “When deemed warranted, the selection of empirical antifungal therapy will be tailored to the local pattern of antifungal therapy will be tailored to the local pattern of the most prevalent Candida species and any prior administration of azoles drugs.”

• “Risk factors for candidemia should also be considered when choosing initial therapy.”

Dellinger RP et al. Crit Care Med. 2008;36:296-327.

IDSA Guidelines for Management of Candidiasis

TherapyTherapy

Condition Primary Alternative†

CandidemiaNonneutropenic adults

Fluconazole800-mg (12-mg/kg) loadingdose, then 400 mg (6 mg/kg) daily or an

echinocandin* (A -I)

LFAmB 3-5 mg/kg daily; or AmB-d 0.5-1 mg/kg daily; or voriconazole 400 mg (6 mg/kg) bid for 2 doses, then 200 mg (3 mg/kg) bid (A-I)

33

* Echinocandin dosing in adults is as follows: anidulafungin, 200-mg loading dose, then100 mg/day; caspofungin, 70-mg loading dose, then 50 mg/day; and micafungin, 100 mg/day.

† AmB-d (0.5-1.0 mg/kg daily) or LFAmB (3-5 mg/kg daily) are alternatives if there isintolerance to or limited availability of other antifungal agents (A-I).

LFAmB, lipid formulation of amphotericin B.

Pappas PG, et al. Clin Infect Dis. 2009;48(5):503-535.

echinocandin* (A -I)

Neutropenic adults An echinocandin* (A-II) Fluconazole 800-mg (12-mg/kg) loading dose, then 400 mg(6 mg/kg) daily; or voriconazole 400 mg (6 mg/kg) bid for 2 doses then 200 mg (3 mg/kg) bid (B-III)

IDSA 2009 Candidiasis GuidelinesNon-neutropenic Initial Therapy for Adult Patients

fluconazole800mg LD

400mg daily

NO

echinocandinA1

Yes

Moderate to Severe illnessor

recent azole exposure

Key piece of data

Must considerLocal pathogen prevalence data

For Internal Use Only, Not for Distribution, Display or Promotion

fluconazole

C. albicasC. tropicalis

culture result

400mg dailyA1

for pt who initially receivedechinocandin and follow up

cultures are negcontinue echinocandin

fluconazoleB-III

C. parapsilosis

echinocandinB-III

C glabrata

culture result

Key piece of data

Must know pathogenPrevalence in ICU

Ref: Pappas et al CID 2009;48:503-35

Epidemiology and Susceptibility toCandida species

Species Frequency Ampho B

Flu or Itra

Vori or Posa

Echinocandins

C. albicans 40-60% S S S S

C. glabrata 20-30% S-I S-DD to R

S to S-DD

S

C. 10-20% S S S S to IC. parapsolosis

10-20% S S S S to I

C. tropicalis 20-30% S S S S

C. lusitania 0-5% R S S S

C. krusei 5-10% S-I R S to S-DD

S

Ref: Ostrosky-Zeichner et al CCM 2006 34:3

The Ohio State University Medical Center

• 1000 bed tertiary care academic medical center

• Antimicrobial Stewardship Program (ASP)

• Multi-disciplinary team

Candidemia Management Algorithm

2-5 days

10 minutes10 minutes

24-48 hours

2-4 hours

Candidemia Management Algorithm

Germ Tube Test for C. albicans

Early forms of hyphae are referred to as germ tubes

C. albicans when placed in a nutrient environment, is able to form germ tubesin less than 3 hours

This is a quick and inexpensive method todifferentiate albicans from non-albicansspecies of Candida.

The downside of this method isit is very time sensitiveRequires experienced lab personnelSamples incubated too long can cause false positivesHeavy inoculum may produce false negatives

So How Do I Choose?Treatment of Invasive Candidiasis

• Consider alternatives to fluconazole when:

• recent exposure to fluconazole or other azole• broader spectrum is desirable

(e.g., persistently neutropenic patient)• non-albicans species isolated during or

immediately following azole therapy

• unstable or severely immunocompromisedpatient

• Hospital has high rate of non-albicans

Conclusion

• Timing is everything!• Select the most effective antifungal agent

based on local pathogen prevalence data • Review the ordering process at your hospital• You can select effective therapy but if it isn’t

given in a timely fashion the outcome may not be optimal