Candidemia in the ICU
Candidemia, Invasive candidiasis, candida bloodstream infection
1. Description of the problem
Scope of the Problem
Candida species represent the fourth most common bloodstream infections in hospitals.
Candida species cause 8-10% of all nosocomial bloodstream infections.
Rates of candidemia are highest in the ICU : 1-4% of all patients.
The attributable mortality rate remains as high as 35%, even with treatment.
Risk Factors for Candidemia
Central venous catheter.
Colonization with candida (urine, sputum, wounds).
Total parenteral nutrition.
Immunosuppression (e.g. corticosteroids, solid organ and stem cell transplantation).
Wide spectrum/varied presentation from fever alone to hemodynamic instability/septic shock.
Think about candidemia in the ICU patient with risk factors and persistent fevers despite receiving broad-spectrum antibiotics.
Blood cultures may be negative in up to 1/3 of patients.
New hemodynamic instability.
Skin lesions representing hematogenous dissemination: varied lesions: pale pink macules, pustules, nodules, necrotic lesions (Figure 1).
Biopsy of a suspicious skin lesion may be the only way to make the diagnosis of disseminated candidiasis (Figure 2).
Visual changes due to candida retinitis.
Hepatic and/or splenic hypodensities or small pulmonary nodules seen in hematogenous disseminated candidiasis, particularly in the neutropenic host (Figure 3).
See Table I. Organisms Responsible
|C. albicans||50-60%||usually fluconazole sensitive|
|C. glabrata||15-20%||rising fluonconazole resistance (~20%), which may be sometimes overcome by increasing the dose (dose-dependent)|
|C. parapsilosis||10-20%||higher MICs for echinocandins (clinical significance unclear)|
|C. kruseii||1-3%||intrinsically resistant to fluconazole|
may be resistant to amphotericin B
2. Emergency Management
Diagnosis of Invasive Candida
Isolation of candida from any normally sterile body site is most consistent with invasive disease.
Isolation of candida from sites not normally sterile (e.g. sputum, wound or drain cultures) must be interpreted within the clinical context.
While any organism can be a contaminant in a blood culture, isolation of candida from blood cultures should be interpreted as evidence of candidemia.
-Candiduria in the presence of pyuria and symptoms
-Asymptomatic candiduria is not usually significant, except in high-risk hosts (e.g., neutropenic, stem cell transplant).
Uncommon. Candida from sputum, bronchoscopic specimens most often represents colonization of the oropharynx.
When Candida are a cause of pneumonia, the mechanism is hematogenous spread to the lung with multiple small pulmonary nodules, most commonly in neutropenic hosts.
4. Bone/joint involvement
Isolation of candida species from a specimen obtained sterilly. Candida species isolated from open wounds must be interpreted within the clinical context.
5. Eye involvement
Dilated funduscopic exam may reveal creamy white lesions in the choroid or retina. There may be inflammatory infiltrates in the vitreous. Diagnosis may be made visually, or sometimes by aspiration of the vitreous. Patients may have no visual symptoms or may have visual loss, depending on the location of the lesions.
Seen more commonly in patients with prosthetic heart valves. Suspect when candidemia fails to clear on appropriate therapy. Echocardiography may reveal suggestive lesions.
7. Hepatosplenic candidiasis
Seen in neutropenic cancer patients, or at the time of neutrophil recovery. This is an indolent process wtih seeding of multiple organs from a previous candidemia (which may have not been recognized). It should be suspected when there is fever, RUQ pain/tenderness, anorexia, rising liver-associated tests (alkaline phosphatase is usually elevated out of proportion to transaminases or bilirubin).
Evaluation of patients with risk factors in the ICU and fevers despite broad-spectrum antibiotics
Assessment of risk factors.
Careful physical exam–look for skin lesions (biopsy suspicious lesions).
Dilated funduscopic exam to look for evidence of candida retinitis.
Blood cultures (routine Bactec-type blood cultures offer as good a yield as isolator blood cultures)–cultures may take up to 4-5 days to turn positive.
Consider monitoring for colonization (sputum, urine, wound cultures) to evaluate level of risk.
Consider sending (1,3) beta-D-glucan for serodiagnosis (a non-culture method that measures beta-glucan, a component of fungal cell walls). Published sensitivities have been 75-100% and specificities 88-100%.
Imaging when appropriate (e.g., for a neutropenic patient); consider CT abdomen to look for disseminated (hepatosplenic) candidiasis.
Unfortunately, WBC and other laboratory parameters are not particularly helpful in the diagnosis of candidemia.
Obtain baseline creatinine and liver-associated tests to guide choice of antifungal therapy and follow for signs of toxicity.
4. Specific Treatment
See Table II. Therapies for fungal infections
|Candidemia||Fluconazole 400 mg/d x 14 daysCaspofungin 70 mg load x1, then 50 mg/d x 14 daysAnidulafungin 200 mg load x1, then 100 mg/d x 14 daysMicafungin 100 mg/day x 14 days||Liposomal amphotericin 3-5 mg/kg/day x 14 daysVoriconazole 400 mg (6 mg/kg) bid for 2 doses, then 200 mg (3 mg/kg) bid x 14 days||-Fluconazole if known sensitive species or clinically stable, non-neutropenic patient, and not already receiving an azole-Voriconazole can be considered if need to also cover mold (e.g., neutropenic fever in stem cell transplant recipient).-Consider changing to fluconazole if starting with another Rx and isolate is sensitive to fluconazole.-Oral fluconazole and voriconazole are well absorbed and are reasonable alternatives to IV administration.|
|Urinary tract infection||Fluconazole 200 mg/d x 2 weeks||-Amphotericin B 0.3-0.6 mg/kg/d x 7 days-Amphotericin bladder irrigation-Echinocandin||-Asymptomatic candiduria is usually not significant (neutropenic hosts are an exception).-Remove indwelling Foley catheter if possible, and repeat urine culture.-Treatment if symptoms, fever and candiduria and pyuria|
|Chronic, disseminated candidiasis||Fluconazole 400 mg/d until resolution of signs and symptoms||An echinocandinAmphotericinVoriconazole||Response to therapy may be protracted and require a long duration of Rx.|
|Candida endophthalmitis||Amphotericin B 0.7-1 mg/kg/d + 5-FC 25 mg/kg QIDFlucoanzole 6-12 mg/kg/d||Liposomal amphotericin 3-5 mg/kg/dVoriconazole 6 mg/kg q 12h for 2 doses, then 3-4 mg/kg q 12hrEchinocandiin||-Some patients require surgical intervention.-Treat for at least 4-6 weeks and until resolution.|
|Candida endocarditis||Liposomal amphotericin 3-5 mg/kg/d +/- 5-FC 25 mg/kg QIDEchinocandin||Fluconazole 400-800 mg (6-12 mg/kg) daily||-Fluconazole once blood cultures are negative if organism is susceptible-Valve replacement recommended for cure. If not possible, consider lifelong suppression with fluconazole.-Consider higher doses of echinocandin (caspofungin 50-150 mg/d).|
|Bone and joint infection||Fluconazole 400 mg daily for at least 6 weeks for arthritis and 6-12 months for osteomyelitis||EchinocandinAmphotericin||-Surgical debridement is recommended.|
|Suppurative thrombophlebitis||Liposomal amphotericin 3-5 mg/kg/d x 2 weeksFluconazole 400-800 mg/dEchinocandin||
Surgical incision and drainage or vein resection is recommended.
Antifungal agents: dosing, side effects and drug interactions
Amphotericin B preparations & dosing for candidemia
Amphotericin B deoxycholate 0.5-1.0 mg/kg/d IV
Liposomal amphotericin B 3-5 mg/kg//d IV
Amphotericin B lipid complex 3-5 mg/kg/d IV
Amphotericin B colloidal dispersion 3-5 mg/kg/d IV
Most candida species (except for
Candida lusitaniae), broad-spectrum activity against many other yeasts and molds.
Nephrotoxicity (usually reversible), less frequent with lipid preparations.
Infusion-related toxicity: fever, chills, pulmonary infitrates.
Renal tubular acidosis–usually mild.
Pharmacokinetics and Bioavailability:
Reduced delivery to kidney in lipid preparations compared to amphotericin
May have additive nephrotoxicity with other agents–avoid co-administration with aminoglycosides.
Adjustments in Patients with Reduced Creatinine Clearance:
Avoid if possible.
Maintain adequate hydration
If infusion-related toxicity, consider premedication with Tylenol +/- Benadryl.
Infusion-related toxicity tends to become milder with subsequent administrations.
Lipid preparations are more expensive and associated with less toxicity, but have not been demonstrated to have improved efficacy.
Triazoles & dosing for candidemia
Fluconazole: 800 mg x 1 load, then 400 mg QD.
Itraconazole: 200 mg tid x 3 days, then 200 mg QD or bid. Not well studied in candidemia. Usually reserved for mucosal disease only.
Voriconazole: 400 mg bid x 1 day load, then 200 mg po bid; IV preparation: 6 mg/kg q 12 h x 1 day, then 3-4 mg/kg Q 12 h.
Posaconazole: Approved for prophylaxis in stem cell transplant (200 mg QID or 400 mg bid oral suspension), but not for candidemia.
Fluconazole: resistance mostly in C. glabrata and C. kruseii. Resistance less common but can potentially occur with other species. No activity against molds.
Itraconazole: similar to fluconazole against candida, with activity against some molds
Voriconazole: similar to fluconazole, but may also have activity against C.kruseii or fluconazole-resistant C. glabrata. Activity against a variety of molds (not Rhizopus or Mucor).
Posaconazole: similar to voriconazole with added activity against Rhizopus and Mucor.
As a class: hepatotoxicity, QT prolongation, rash.
Voriconazole: phototoxicity, visual symptoms: most commonly altered color perception or halos around lights; rarely, visual hallucinations.
Pharmacokinetics and Bioavailability:
Fluconazole and voriconazole: excellent oral bioavailability, considered equivalent to IV administration. Oral absorption not affected by gastric pH, but may be diminished with food.
IV voriconazole is compexed to a cyclodextrim molecule, which may accumulate in patients with renal dysfunction.
Itraconazole: oral suspension has much better bioavailability than capsules (capsule absorption inhibited by drugs that raise gastric pH; absorption improved with acidic and carbonated drinks). Oral solution better absorbed without food; capsule has better absorption with food.
All inhibit cytochrome P450 enzymes. Carefully review all other medications for potential drug interactions based on P450.
Adjustments in patients with reduced creatinine clearance:
Voriconazole: no adjustment for oral preparation. Intravenous voriconazole is not recommended if CrCl < 50 mL/min.
Large variety in blood levels with voriconazole (see Therapeutic Drug Monitoring)
Echinocandins &dosing for candidemia
Caspofungin: 70 mg load x 1 , then 50 mg/d.
Anidulafungin: 200 mg load x 1, then 100 mg daily.
Micafungin: 100 mg daily.
Activity against most candida species.
C. parapsilosis has diminished in vitro activity; clinical significance is unclear.
Partial list of side effects:
Minimal/well tolerated: hepatotoxicity (usually mild), infusion related.
Pharmacokinetics and Bioavailability:
Available only for IV administration.
Adjustments in patients with reduced creatinine clearance:
No dose adjustment for renal dysfunction or dialysis, but dose reductions necessary with caspofungin only in patients with hepatic dysfunction.
fWhen to request susceptibility testing
Standardized susceptibility testing is available for Candida species to fluconazole, itraconazole, voriconazole, flucytosine, and the echinocandins.
Consider susceptibility testing in the following patients:
Failure to respond to treatment.
C. glabrata isolates.
Therapeutic drug monitoring
Available for itraconazole, voriconazole and posaconazole.
For itraconazole, steady-state levels may take up to 2 weeks to acheive. Drug levels should not be obtained prior to this time period.
For voriconazole, serum concentrations vary widely. Drug levels can be obtained after 4-5 days.
Suspected candidemia: empiric antifungal therapy in the ICU
Neutropenic patients with peristent fever after 4-7 days should receive empiric antifungal therapy per published guidelines.
Empiric treatment for candidemia in non-neutropenic ICU patients is controversial. There are conflicting studies in the literature.
Consider empiric therapy in patients with:
Persistent fever on broad-spectrum antibiotics.
Multiple risk factors for candidemia.
Colonization with candida from multiple non-sterile sites, or positive beta-D-glucan.
5. Disease monitoring, follow-up and disposition
Obtain daily blood cultures until they are clear.
Dilated funduscopic exam should be performed in all patients to look for candida retinitis/endophthalmitis.
Follow renal and hepatic function as well as CBC for signs of toxicity.
Clinical monitoring for signs/sx’s of recurrance or metastati cfoci of infection–further work-up is indicated if recurrent fever, anorexia or focal symptoms
Follow-up imaging of all previously abnormal findings related to candida infection. For example, in disseminated, hepatosplenic candidiasis, obtain periodic abdominal CT or US to follow the response to therapy. For endophthalmitis, perform periodic funduscopic exam.
Endocarditiis – more common in patients with prosthetic heart valves
Disseminated infection: deep organ infection
Arthritis, osteomyelitis—knees and vertebrae most common sites
What's the evidence
Wisplinghoff, H, Bischoff, T, Tallent, SM. “Nosocomial bloodstream infections in US hospitals: analysis of 24,179 cases from a prospective nationwide surveillance study”. Clin Infect Dis. vol. 39. 2004. pp. 309(Epidemiology of Candidemia.)
Chow, JK, Golan, Y, Ruthazer, R. “Risk factors for albicans and non-albicans candidemia in the intensive care unit”. Crit Care Med. vol. 36. 2008. pp. 1993(Risk factors for candidemia.)
Blumberg, HM, Jarvis, WR, Soucie, JM. “Risk factors for candidal bloodstream infections in surgical intensive care unit patients: the NEMIS prospective multicenter study. The National Epidemiology of Mycosis Survey”. Clin Infect Dis. vol. 33. 2001. pp. 177(Risk factors for candidemia.)
Pappas, PG, Kauffman, CA, Andes, D. “Clinical practice guidelines for the management of candidiasis: 2009 update by the Infectious Diseases Society of America”. Clin Infect Dis. vol. 48. 2009. pp. 503(Guidelines for the treatment of candidemia.)
Schuster, MG, Edwards, JE, Sobel, JD, Darouiche, RO, Karchmer, AW. “Empirical fluconazole versus placebo for intensive care unit patients: a randomized trial”. Annals of Internal Medicine. vol. 149. 2008. (Empiric antifungal therapy in high risk ICU patients.)
Copyright © 2017, 2013 Decision Support in Medicine, LLC. All rights reserved.
No sponsor or advertiser has participated in, approved or paid for the content provided by Decision Support in Medicine LLC. The Licensed Content is the property of and copyrighted by DSM.
- 1. Description of the problem
- 3. Diagnosis
- 4. Specific Treatment
- 5. Disease monitoring, follow-up and disposition
- What's the evidence