**3. Diagnosis of fungal endocarditis**

The diagnosis of IE is based upon high index of suspicion, careful history and physical examination, echocardiographic or histopathological findings, laboratory results, and chest


endocarditis in echocardiography; new regurgitate murmur, serological



 - Pathological criterion: vegetation or intracardiac abscess, confirmed by histology showing active endocarditis, positive Gram stain results or




heart valve, presented of vegetation or other typical findings of

or culture evidence of infection with Coxiella burnetii.

cultures of specimens obtained from surgery or autopsy



criterion, fever, predisposing valvular conditiona

C-reactive protein hematuria and splenomegaly.

a Prosthetic heart valve or a valve lesion that leads to significant regurgitation or turbulence of blood flow; Vascular phenomenon like emboli to the brain or organs, hemorrhages in the mucous membranes around the eyes; Immunologic phenomenon include lesions such as Roth's spots or "Osler's nodes and

edema, dyspnea, murmur on examination, cardiac failure, and persistent sepsis would also present. Other symptoms include abdominal pain, malaise, weight loss, night sweats, arthritis, finger clubbing, cough, hemoptysis, sudden death, coagulopathy, jaundice, nausea, hypotension, and renal failure. *Candida spp*. is the most etiologic agent of FE; therefore, patients can present endophthalmitis, meningitis, osteomyelitis and other complications of candidemia. The more specific cutaneous or mucocutaneous lesions of IE include Osler's nodes, Roth spots (rare), and Janeway lesions are more specific signs but less common and not diagnostic . Petechiae and splinter hemorrhages (nonblanching, linear reddish-brown lesions found under the nail bed) are not specific but are common skin manifestations. They may be present on the extremities of skin, or on mucous membranes. Other organs may be involved due to embolic events such as splenic or renal infarcts, or immune reactions like arthritis and glomerulonephritis, or spread by the blood passing to other organs like soft tissues, vertebral osteomyelitis, and the brain causing meningitis and/or encephalitis.

The diagnosis of IE is based upon high index of suspicion, careful history and physical examination, echocardiographic or histopathological findings, laboratory results, and chest

Table 1. Definition of three criteria for the diagnosis of infective endocarditis

minor criteria

four days or less

**3. Diagnosis of fungal endocarditis** 

**Duke criteria** (36)

Minor clinical criteria

Major clinical criteria

Definitive

Possible IE

Rejected IE

glomerulonephritis.

radiography. Gold standard tests for the detection of documented infections are the isolation of fungi from the blood, heart biopsy or vegetation by culture and the presence of tissue invasion by histopathology. Isolation of fungi from blood samples is difficult due to non-growth of fungal etiologic agents in blood culture. The rate of culture positive of *Candida* spp. in the blood is about 50% of the documented cases and positive blood culture for *Aspergillus* is rare (38-40). Fungi are cleared rapidly, due to large size, in the blood by the host's reticuloendothelial system; therefore, the blood culture results are negative in many suspicious patients. The use of lysis-centrifugation system (41), or Bactec blood culture (42), may help the isolation of fungal agents but none is recommend as a standard method. Heart tissue is the best sample for the isolation of fungal agents. As resistance to the antifungal agents has been reported in many studies (43-45), in case of positive culture, sensitivity test of the isolated fungi to antifungal agents can contribute to the best management of infections.

Another definitive microbiologic diagnosis depends upon the evidence of fungal tissue invasion with histopathologic investigation. The samples (tissue valve or emboli) are stained with specific stains like Gomori methenamine silver or Periodic acid-Schiff. With histopathology examination, morphological differentiation between *Aspergillus* spp. and other fungi is not completely available.

Given the frequent negative blood cultures, and difficulty in obtaining the material from the surgical sites in the operating rooms, echocardiography, either transthoracic echocardiography (TTE) or transesophageal echocardiography (TEE), are used as the diagnosis tools with the sensitivity of about 77% (15) for the evaluation of FE and the presence of vegetation, based on the major diagnostic Duke criterion. Echocardiography can also detect intra cardiac abscess, new or progressive valvular regurgitation, the size and location of vegetation. The size of vegetation may be small, medium, or large and anatomic site of the vegetation may be aortic valve, on tricuspid, mitral or endocardium, or on the previous aortic valvular surgery.

Transesophageal echocardiography should be considered as the standard diagnostic procedure for IE (46). This method is able to evaluate the prosthetic valves, intracardiac complications, inadequate TTE, fungemia or bacteremia, and has superior sensitivity (47), compared to TTE, but significantly more invasive and expensive than it. Transthoracic echocardiography is the first line procedure for the detection of FE especially in native valve and prosthetic valve vegetations, and local extension of infection. The sensitivity of TTE in infants and younger children is about 80 percent (48, 49), therefore, the negative result of it cannot definitively rule out FE and examination should be repeated in respective patients. If there is a high clinical suspicion for FE and the TTE is negative, we should turn to TEE. Once treatment is completed, repeated evaluation may be necessary to establish a new baseline of valvular and myocardial functions for the patient. Unfortunately, both TTE and TEE may yield false negative results if the vegetations are small, or large size of the vegetation suspected as a mural thrombus, vegetation is attached to the mural endocardium and if embolization of the vegetation has occurred.

Chest radiography and echocardiography are not useful in the diagnosis of IE; x-ray may present the septic pulmonary emboli (Minor Duke's Criteria) and echocardiography may show evidence of some complications.

Over the last several decades, non-culture laboratory methods have been directed at the development for the diagnosis of systemic fungal infections such as FE. Serological diagnostic methods can serve as the non-invasive methods for detecting the circulating

Post-Cardiac Surgery Fungal Endocarditis 275

recommendation of combined surgical and medical treatment in patients with FE for better prognosis (7, 8, 85). It is also the suggestion of the 2009 Infectious Diseases Society of America guidelines for the treatment of native and prosthetic valve *Candida* endocarditis (86). However, there are some reports of *Candida* endocarditis in which medical treatment alone proved successful (85, 87-89) with either caspofungin alone or in combination with flucytosine or fluconazole (87-90). The higher dose of antifungal agents than normal dose is recommended for treatment (86). In critically ill patients for whom surgical resection cannot be done, antifungal therapy is recommended for months and even life-long. In combination therapy a minimum of 6 weeks medication after surgery is advocated (91), but the treatment should be continued till signs and symptoms of the infection disappear and radiographic abnormalities are stabilized, and life-long prophylactic therapy is recommended. Relapses are common either with medical or combined therapy (37, 92, 93) and may appear early or late; mean 25 months (92). Due to high relapse rates, patients should receive life-long therapy (92, 93) and careful follow-up

Treatment of FE in immunocompromised patients needs to take into account the underlying disease of patients and the intervention of antifungal agents with the patient's condition. For example, use of amphotericin B deoxycholate, in patients with renal insufficiency or those who are on multiple nephrotoxic drugs is not suggested and for fewer adverse effects, lipid formulations of amphotericin B are recommended (94). Many studies have reported the resistance of some fungi to this antifungal agent (44, 45, 95), therefore, to limit the use of amphotericin B, current treatment options including azole due to its broad antifungal activity, and echinocandins as a new class of antifungal drugs,

In transplant recipients and HIV patients, use of triazoles which have interaction with human P450 cytochromes (97), can block the metabolism of certain anti-HIV drugs and also some drugs such as cyclosporine, statins, and benzodiazepines (98). Therefore, close monitoring of drug levels needs to be calibrated with the dose of immunosuppressive drugs (99, 100). Physicians should prescribe triazole agents in consultation with a pharmacist because inhibitory activities among triazoles are different and fluconazole is of less active inhibition of P450 than other azole agents such as itraconazole, voriconazole, and posaconazole. In patients receiving these antifungals, monitoring of drug levels in respective

Echinocandins; caspofungin, micafungin, and anidulafungin; are new antifungal agents which damage the fungal cell walls by inhibiting the b-(1, 3)-glucan synthesis. Drug–drug interactions between echinocandins such as caspofungin are observed with tacrolimus and cyclosporine, certain anti-HIV drugs and rifampin (96). Use of caspofungin in patients with impaired liver function and those receiving cyclosporine should be carefully considered, because of the common side effects of this agent including increased liver enzymes, pruritus, facial swelling, headache and nausea. They have fungicidal activity against *Candida*  biofilm (101) and most isolates of *Candida* species including *C. glabrata in* vitro and in vivo

If the patients are not responsive to their initial mono-antifungal therapy regimen, the use of the combination antifungal regimen is recommended that include an echinocandins with voriconazole or liposomal amphotericin B. Combination therapy by amphotericin B and a triazole is not suggested in the literature (103). The function of combination

is also essential for successful therapy.

are recommended (96).

sera is suggested (96).

with benign toxicity profile (102).

fungal antigens, fungal metabolites, or antibody in the blood (50, 51). The major limitation of these methods is unavailability to detect some fungi like *Mucor* spp.

Galactomannan (GM) is a more promising circulating fungal antigen used to detect fungal infections especially invasive aspergillosis, but it is an exoantigen released from the tip of mycelium of many fungi spp. during growth (52), therefore, cross-reactivity has been described with other fungi (53-55). False negative reactivity without any known reason (56) and false positive reactivity with use of some drugs and foods have been reported (57-61). Sensitivity range of GM test, for the diagnosis of documented invasive aspergillosis cases was reported to be between 50.0% (62) and 90.6% (63). There are limited studies using this method for the diagnosis of FE. In one study, GM test to establish the diagnosis of invasive aspergillosis was only positive (≥1 ng/mL) in 2/7 patients with endocarditis and mediastinitis(17) and four out of nine cases in another study (33). To diagnose systemic candidiasis, enolase (64), phospholipase and proteinase enzymes (65), *Candida* mannan antigen (66, 67), and β-D-glucan (68-71) have been detected in some studies. However, there are a few reports on the use of such antigens in patients with FE.

Antibody assays can be helpful for some species of fungi which are not the normal flora but there are problems with both specificity and sensitivity when *Candida spp.* is responsible for infections, since it is a part of the body normal flora. Immunosuppressed hosts may be unable to produce strong antibodies; therefore, the sensitivity of the assay in this high-risk population is decreased.

The current focus of non-culture methods is on the development of a polymerase chain reaction (PCR) assay for the detection of fungal infections (72, 73). Panfungal PCR with universal primers (74), nested PCR (75, 76) and real-time PCR (77, 78) can serve as sensitive and quantitative methods to detect fungal DNA in the human blood specimens (74, 79). The sensitivity and specificity of nested PCR for invasive aspergillosis in the blood are 92.8% and 94%, respectively (55). Although these methods have not been standardized and are not widely used, limited studies indicate a good sensitivity for FE diagnosis and close to blood cultures (33, 80). Using the molecular methods with reduced PCR steps like real time- PCR, the result can be released within 6 hours (81). Due to the inhibitory factors in human blood samples, PCR may yield false negative (82) and for the abundant conidia of fungi in the environment, false positive may also be seen, which is rare and limited. The significance of PCR tests is their ability to detect fungal infections in early stages (83).

Other nonspecific laboratory outcomes include: a normochromic normocytic anemia, elevated erythrocyte sedimentation rate and C-reactive protein indicative of inflammation, elevated rheumatoid factor titers (minor Duke criteria), hematuria and proteinuria (minor Duke criteria).

Totally, in patients suspicious to FE, microscopy examination and culture of tissue materials obtained from heart surgery, with antifungal susceptibility test on the isolated fungi are the best methods for the diagnosis and management of FE. In patients with suspected FE in early stage of infection, use of nonaggressive method (i.e., serologic or molecular) is recommended. Combination of serological and microbiological tests is more useful if we are to avoid over-treatment.

### **4. Treatment**

The high mortality rate, difficulty in sterilizing large fungal vegetation or abscesses, and the risk of embolization associated with medical therapy alone (84) are the reasons for the

fungal antigens, fungal metabolites, or antibody in the blood (50, 51). The major limitation of

Galactomannan (GM) is a more promising circulating fungal antigen used to detect fungal infections especially invasive aspergillosis, but it is an exoantigen released from the tip of mycelium of many fungi spp. during growth (52), therefore, cross-reactivity has been described with other fungi (53-55). False negative reactivity without any known reason (56) and false positive reactivity with use of some drugs and foods have been reported (57-61). Sensitivity range of GM test, for the diagnosis of documented invasive aspergillosis cases was reported to be between 50.0% (62) and 90.6% (63). There are limited studies using this method for the diagnosis of FE. In one study, GM test to establish the diagnosis of invasive aspergillosis was only positive (≥1 ng/mL) in 2/7 patients with endocarditis and mediastinitis(17) and four out of nine cases in another study (33). To diagnose systemic candidiasis, enolase (64), phospholipase and proteinase enzymes (65), *Candida* mannan antigen (66, 67), and β-D-glucan (68-71) have been detected in some studies. However, there

Antibody assays can be helpful for some species of fungi which are not the normal flora but there are problems with both specificity and sensitivity when *Candida spp.* is responsible for infections, since it is a part of the body normal flora. Immunosuppressed hosts may be unable to produce strong antibodies; therefore, the sensitivity of the assay in this high-risk

The current focus of non-culture methods is on the development of a polymerase chain reaction (PCR) assay for the detection of fungal infections (72, 73). Panfungal PCR with universal primers (74), nested PCR (75, 76) and real-time PCR (77, 78) can serve as sensitive and quantitative methods to detect fungal DNA in the human blood specimens (74, 79). The sensitivity and specificity of nested PCR for invasive aspergillosis in the blood are 92.8% and 94%, respectively (55). Although these methods have not been standardized and are not widely used, limited studies indicate a good sensitivity for FE diagnosis and close to blood cultures (33, 80). Using the molecular methods with reduced PCR steps like real time- PCR, the result can be released within 6 hours (81). Due to the inhibitory factors in human blood samples, PCR may yield false negative (82) and for the abundant conidia of fungi in the environment, false positive may also be seen, which is rare and limited. The significance of

Other nonspecific laboratory outcomes include: a normochromic normocytic anemia, elevated erythrocyte sedimentation rate and C-reactive protein indicative of inflammation, elevated rheumatoid factor titers (minor Duke criteria), hematuria and proteinuria (minor

Totally, in patients suspicious to FE, microscopy examination and culture of tissue materials obtained from heart surgery, with antifungal susceptibility test on the isolated fungi are the best methods for the diagnosis and management of FE. In patients with suspected FE in early stage of infection, use of nonaggressive method (i.e., serologic or molecular) is recommended. Combination of serological and microbiological tests is more useful if we are

The high mortality rate, difficulty in sterilizing large fungal vegetation or abscesses, and the risk of embolization associated with medical therapy alone (84) are the reasons for the

these methods is unavailability to detect some fungi like *Mucor* spp.

are a few reports on the use of such antigens in patients with FE.

PCR tests is their ability to detect fungal infections in early stages (83).

population is decreased.

Duke criteria).

**4. Treatment** 

to avoid over-treatment.

recommendation of combined surgical and medical treatment in patients with FE for better prognosis (7, 8, 85). It is also the suggestion of the 2009 Infectious Diseases Society of America guidelines for the treatment of native and prosthetic valve *Candida* endocarditis (86). However, there are some reports of *Candida* endocarditis in which medical treatment alone proved successful (85, 87-89) with either caspofungin alone or in combination with flucytosine or fluconazole (87-90). The higher dose of antifungal agents than normal dose is recommended for treatment (86). In critically ill patients for whom surgical resection cannot be done, antifungal therapy is recommended for months and even life-long. In combination therapy a minimum of 6 weeks medication after surgery is advocated (91), but the treatment should be continued till signs and symptoms of the infection disappear and radiographic abnormalities are stabilized, and life-long prophylactic therapy is recommended. Relapses are common either with medical or combined therapy (37, 92, 93) and may appear early or late; mean 25 months (92). Due to high relapse rates, patients should receive life-long therapy (92, 93) and careful follow-up is also essential for successful therapy.

Treatment of FE in immunocompromised patients needs to take into account the underlying disease of patients and the intervention of antifungal agents with the patient's condition. For example, use of amphotericin B deoxycholate, in patients with renal insufficiency or those who are on multiple nephrotoxic drugs is not suggested and for fewer adverse effects, lipid formulations of amphotericin B are recommended (94). Many studies have reported the resistance of some fungi to this antifungal agent (44, 45, 95), therefore, to limit the use of amphotericin B, current treatment options including azole due to its broad antifungal activity, and echinocandins as a new class of antifungal drugs, are recommended (96).

In transplant recipients and HIV patients, use of triazoles which have interaction with human P450 cytochromes (97), can block the metabolism of certain anti-HIV drugs and also some drugs such as cyclosporine, statins, and benzodiazepines (98). Therefore, close monitoring of drug levels needs to be calibrated with the dose of immunosuppressive drugs (99, 100). Physicians should prescribe triazole agents in consultation with a pharmacist because inhibitory activities among triazoles are different and fluconazole is of less active inhibition of P450 than other azole agents such as itraconazole, voriconazole, and posaconazole. In patients receiving these antifungals, monitoring of drug levels in respective sera is suggested (96).

Echinocandins; caspofungin, micafungin, and anidulafungin; are new antifungal agents which damage the fungal cell walls by inhibiting the b-(1, 3)-glucan synthesis. Drug–drug interactions between echinocandins such as caspofungin are observed with tacrolimus and cyclosporine, certain anti-HIV drugs and rifampin (96). Use of caspofungin in patients with impaired liver function and those receiving cyclosporine should be carefully considered, because of the common side effects of this agent including increased liver enzymes, pruritus, facial swelling, headache and nausea. They have fungicidal activity against *Candida*  biofilm (101) and most isolates of *Candida* species including *C. glabrata in* vitro and in vivo with benign toxicity profile (102).

If the patients are not responsive to their initial mono-antifungal therapy regimen, the use of the combination antifungal regimen is recommended that include an echinocandins with voriconazole or liposomal amphotericin B. Combination therapy by amphotericin B and a triazole is not suggested in the literature (103). The function of combination

Post-Cardiac Surgery Fungal Endocarditis 277

[1] Newman WH, Cordell AR. *Aspergillus* endocarditis after open-heart surgery.report of a case and review of the literature. J Thorac Cardiovasc Surg 1964;48:652-60. [2] El-Hamamsy I, Du¨ rrleman N, et al. Cluster of Cases of *Aspergillus* Endocarditis After

[3] Vaideeswara P, Mishra P, et al. Infective endocarditis of the Dacron patch-a report of 13

[4] Ryu KM, Seo PW, et al. Surgical Treatment of Native Valve *Aspergillus* Endocarditis and fungemic vascular complications. J Korean Med Sci 2009; (1): 170-2. [5] Bayer AS, Scheld M. Endocarditis and intravascular infections.In: Mandell GL, Bennett

infectious diseases.Philadelphia, PA: Churchill Livingstone, 2000; 857–902. [6] Karchmer AW. Infections on prosthetic valves and intravasculardevices.In: Mandell GL,

infectious diseases.Philadelphia, PA: Churchill Livingstone 2000; 903–17. [7] Pierrotti LC, Baddour LM. Fungal Endocarditis, 1995–2000 Chest. 2002; 122(1):302-10. [8] Tunkel AR, Kaye D. Endocarditis with negative blood cultures. N Engl J Med

[11] Fernandez-Guerrero M, Verdejo C, et al. Hospital acquired infective endocarditis not

[12] Woods GL, Wood RP, et al. *Aspergillus* endocarditis in patients without prior

[13] Barst RJ, Prince AS, et al. *Aspergillus* endocarditis in children: case report and review of

[14] Paterson DL, Domingues EA, et al. Infective endocarditis in solid organ transplant

[15] Ellis ME, Al-Abdely H, et al. Fungal endocarditis: evidence in the world literature, 1965-

[16] Baddley JW, Benjamin DK, et al. *Candida* infective endocarditis. Eur J Clin Microbiol

[17] Jensen J, Guinea J, et al. Post-surgical invasive aspergillosis: an uncommon and under-

[19] Gordon SM, Keys TF. Bloodstream infections in patients with prosthetic cardiac valves.

[20] Watanakunakorn C. Prosthetic valve infective endocarditis. Prog Cardiovasc Dis

[21] Rubinstein E, Noriega ER, et al. Fungal endocarditis: analysis of 24 cases and review of

[22] Challa S, Prayaga AK, et al. Fungal endocarditis: An autopsy study. Asian Cardiovasc

[23] Hauser M, Hess J, et al. Treatment of *Candida albicans* endocarditis: case report and a

[18] Giamarellou H. Nosocomial cardiac infections. J Hosp Infect 2002; 50: 91–105.

associated with cardiac surgery: an emerging problem. Clin Infect Dis 1995; 20:16–23.

cardiovascular surgery: report of a case in a liver transplant recipient and review.

[9] Moreillon P, Que YA. Infective endocarditis. Lancet 2004;10;363(9403):139-49.

[10] Rubinstein E, Lang R. Fungal endocarditis. Eur Heart J 1995;16:84-9.

JE, Dolin R, eds.Mandell, Douglas and Bennett's principles and practice of

Bennett JE, Dolin R, eds. Mandell, Douglas and Bennett's principles and practice of

Cardiac Surgery. Ann Thorac Surg 2004;77:2184–6.

cases at autopsy. Cardiovascular Pathology 2010;32 (3):1-8.

**8. References** 

1992;326(18):1215-17.

Rev Infect Dis 1989;11:263–72.

the literature. Pediatrics 1981;68:73–8.

1995.Clin Infect Dis 2001; 32:50-62.

Infect Dis 2008; 27:519-29.

Thorac Ann 2004;12(2):95-8.

review. Infection 2003;31(2):125-7.

1979;22:181-92.

recipients. Clin Infect Dis 1998: 26: 689-94.

appreciated entity. J Infect 2010;60(2):162-7.

Semin Thorac Cardiovasc Surg 1995;7:2–6.

the literature. Medicine 1975;54(4):331-4.

antifungal therapy is controversial due to probable increase in side effects and toxicity level (104, 105).
