Candida Infective Endocarditis

Candida Infective Endocarditis Report of 15 Cases From a Prospective Multicenter Study Marco Falcone, MD, Nicoletta Barzaghi, MD, Giampiero Carosi, MD, Paolo Grossi, MD, Lorenzo Minoli, MD, Veronica Ravasio, MD, Marco Rizzi, MD, Fredy Suter, MD, Riccardo Utili, MD, Claudio Viscoli, MD, and Mario Venditti, MD, for the Italian Study on Endocarditis (SEI)*

Abstract: Candida species are an uncommon cause of infective endocarditis (IE). Given the rarity of this infection, the epidemiology, prognosis, and optimal therapy of Candida IE are poorly defined. We conducted a prospective, observational study at 18 medical centers in Italy, including all consecutive patients with a definite diagnosis of IE admitted from January 2004 through December 2007. A Candida species was the causative organism in 8 cases of prosthetic valve endocarditis (PVE), 5 cases of native valve endocarditis (NVE), 1 case of pacemaker endocarditis, and 1 case of left ventricular patch infection. Candida species accounted for 1.8% of total cases, and for 3.4% of PVE cases. Most patients (86.6%) had a health careassociated infection. PVE associated with a health care contact occurred after a median of 225 days from valve implantation. Ten patients (66.6%) were treated with caspofungin alone or in combination with other antifungal drugs. The overall mortality rate was 46.6%. Mortality was higher in patients with PVE (5 of 8 cases, 62.5%) than in patients with NVE (2 of 5 patients, 40%). A better outcome was observed in patients treated with a combined medical and surgical therapy. Candida IE should be classified as an emerging infectious disease, usually involving patients with intravascular prosthetic devices, and associated with substantial related morbidity and mortality. Candida PVE usually is a late-onset disease, which becomes clinically evident even several months after an initial episode of transient candidemia. (Medicine 2009;88: 160Y168) Abbreviations: CVC = central venous catheter, ICU = intensive care unit, IE = infective endocarditis, IV = intravenous, NVE = native valve endocarditis, PVE = prosthetic valve endocarditis, SEI = Italian Study on Endocarditis.

INTRODUCTION

F

ungi are an uncommon but emerging cause of infective endocarditis (IE), accounting for 1%Y6% of total cases,28 and

From Dipartimento di Medicina Clinica (MF, MV), Policlinico Umberto I, Universita` degli Studi di Roma BLa Sapienza,[ Rome; UO Terapia Intensiva Cardiochirurgica (NB), ASO S. Croce e Carle, Cuneo; Malattie Infettive e Tropicali (GC), Universita` degli Studi di Brescia, Brescia; Malattie Infettive e Tropicali (PG), Universita` degli Studi dell_Insubria, Varese; Malattie Infettive (LM), IRCCS San Matteo, Pavia; Malattie Infettive (VR, MR, FS), Ospedali Riuniti di Bergamo, Bergamo; Medicina Infettivologica e dei Trapianti (RU), Azienda Ospedaliera Monaldi, Napoli; Clinica Malattie Infettive (CV), Universita` di Genova, Genova, Italy. *See Appendix for list of SEI investigators. Received September 20, 2008, and in revised form January 7, 2009. Accepted for publication March 3, 2009. Reprints: Mario Venditti, MD, Dipartimento di Medicina Clinica, Policlinico Umberto I. University of Rome BLa Sapienza,[ Viale dell_Universita` 37, 00185 Rome, Italy; (e-mail: mario.venditti)uniroma1.it). Copyright * 2009 by Lippincott Williams & Wilkins ISSN: 0025-7974 DOI: 10.1097/MD.0b013e3181a693f8

160

www.md-journal.com

up to 10% of cases of prosthetic valve endocarditis (PVE).12,14 The number of reported cases has increased during the past 2 decades, probably as a result of various factors, including major use of prosthetic intravascular devices, advances in surgical techniques (including reconstructive heart surgery), and an increasing number of immunocompromised patients,10 with a growing number of patients (for example, those admitted in the intensive care unit EICU^) who are at high risk for nosocomial candidemia or other forms of health care-associated invasive fungal infections.26 Candida species are the most frequent causes of fungal endocarditis.10 Candida IE has been reviewed, and the overall mortality rate is more than 50%, despite treatment.10,28 Survival rates are better for patients treated with a combined antifungal and surgical therapy,10,28 but some patients are poor surgical candidates. Given the rarity of this infection, few studies are available in the English literature on Candida IE. Most reports present anecdotal cases, single-center series, autopsy studies, or analysis of existing databases containing cases observed in various geographic areas4,6,10,21,26,28 over a long time. In a 2008 prospective, multicenter, international study, authors3 evaluated the clinical features and intrahospital outcome of 33 patients with Candida IE. Compared to patients with nonfungal endocarditis, patients with Candida IE had a higher mortality at discharge; however, follow-up data were not available. The Italian Study on Endocarditis (SEI) is an ongoing prospective multicenter study, collecting data from patients with a diagnosis of possible or definite IE admitted to 18 Italian hospitals. We conducted the current study to provide additional data on the treatments and outcomes of Candida endocarditis in a national cohort.

METHODS Patient Sample and Data Collection From January 2004 through December 2007, 18 centers were actively participating in the SEI study; these centers included 7 university-based medical centers, 9 general hospitals, and 2 national research centers. At each institution, all the consecutive patients who met the Duke criteria22 for possible or definite IE were enrolled. A case report form, including more than 200 variables, was used at all sites to collect the data. The following parameters were collected: demographics (age, sex, predisposing factors for IE), medical history, clinical and laboratory findings (fever, cardiac regurgitation, erythrocyte sedimentation rate, C-reactive protein), echocardiographic findings (including number and size of vegetations), microbiologic data (blood cultures, valve or tissue cultures, serologic investigations), surgery, medical therapy, and outcome. All data were entered at the coordinating center in Bergamo, Italy, using a computerized data entry system. Medicine

&

Volume 88, Number 3, May 2009

Copyright @ 2009 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

* 2009 Lippincott Williams & Wilkins

F/71. Mitral valvuloplasty for severe stenosis in August 2006. Steroids therapy for connective tissue disease. Warfarin therapy. Recent hospitalization. M/46. HCV hepatopathy, IVDU, aspirin therapy. In 1999 aortic valve replacement. Diagnosis of Candida IE in 2003 treated with surgery (aortic homograft). F/26. Neurologic disorder. Multiple previous hospital admissions. CVC in place. M/60. One mo prior admission coronary artery bypass procedure, mitral valvuloplasty, PM insertion (complicated by Staphylococcus aureus mediastinitis). CVC in place. Warfarin therapy. F/84. Diabetes mellitus, hospital admission for severe aortic stenosis, 2 mo prior admission aortic valve replacement with prosthetic biologic valve. Warfarin therapy, CVC in place.

2

June 2006/HCA, nosocomial

Prosthetic June 2006/HCA, aortic and nosocomial mitral native valve

CVC distal insertion, tricuspid valve

CVC distal January 2006/HCA, insertion, nosocomial right atrium

Aortic June 2004/HCA, homograft non-nosocomial

Mitral native September 2006/HCA, valve non-nosocomial

Mitral native March 2006/ valve Community

Yes

Yes

Yes

Yes

No

No

Causing Pathogen

Features at Hospital Admission Complication

C. albicans

C. krusei

Multiple vegetations (unspecified size)

Single vegetation (7 mm)

C. albicans

C. tropicalis

Single vegetation C. albicans (29 mm)

No vegetations

Single vegetation (6 mm)

Heart failure (NYHA IV)

Pulmonary embolism

CNS embolism, peripheral embolism (leg), myocardial abscess, valvular dysfunction

Fever, congestive Multiple organ heart failure failure

Fever, hematuria

Fever, dyspnea

Fever, congestive heart failure

Fever, congestive Heart failure heart failure, (NYHA IV), Candida septic shock pluricolonization

Multiple C. parapsilosis Fever, cardiac Heart failure vegetations murmur, dyspnea, (NYHA III) (unspecified size) congestive heart failure

Echocardiograph Finding (Diameter of Vegetation) Therapy

Outcome

20 mo follow-up, cure

Death (June 2004)

www.md-journal.com

Copyright @ 2009 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. Death (July 2006)

(continued on next page)

Caspo

Caspo 4 wk, 8 mo follow-up, followed by death from other suppressive causes Fluco therapy

Caspo 12 wk, then Fluco 4 wk

Ampho B

Fluco 8 wk, 16 mo surgery follow-up, cure (mitral replacement in May2006), Fluco 8 wk Caspo (refused Death surgical (October 2006) intervention)

Volume 88, Number 3, May 2009

6

5

4

3

M/24. HCV hepatopathy, HIV infection, IVDU. Underlying mitral valve disease.

Site of Infection

Prior UTI Diagnosis of IE/ Admission Acquisition* (1 yr)

&

1

Demographics (Sex/Age in yr), Patient Other Characteristics

TABLE 1. Characteristics of 15 Patients With Candida Endocarditis

Medicine Prospective Study of Candida Endocarditis

161

162

www.md-journal.com

13

F/60. In July 2005 aortic Aortic valve replacement prosthetic (mechanical) and PM valve insertion. Recent hospitalization. Warfarin therapy. July 2006/HCA, non-nosocomial

July 2006/HCA, non-nosocomial

August 2006/HCA, non-nosocomial

March 2005/ Community

May 2005/HCA, non-nosocomial

Yes

Yes

Yes

No

No

Yes

Yes

C. albicans

C. albicans

Causing Pathogen

C. glabrata

Complication

Therapy

Outcome

None

Fever, hematuria Spleen embolism

Fever

Caspo plus Vorico

Death (September 2006)

Caspo plus Itra Death (September 2006)

Surgery 24 mo follow-up, (prosthesis cure replacement), Caspo 2 wk, Fluco 6 wk Caspo 4 wk Death plus Ampho B (November 2006) 2 wk

PM removal, 14 mo follow-up, Caspo 6 wk, cure Fluco 6 wk, Posa 6 wk

Spleen embolism, Fluco Death subarachnoid (July 2006) hemorrhage due to rupture of mycotic aneurysm, heart failure (NYHA IV) Hematuria, cardiac Periannular Surgery Improved, lost murmur, dyspnea abscess, fistula, (prosthesis to follow-up prosthesis replacement), dysfunction Caspo 4 wk

Fever, coma, congestive heart failure

Features at Hospital Admission

Multiple C. parapsilosis Fever, new Peripheral vegetations cardiac murmur, embolism (unspecificied hematuria size) Multiple vegetations C. famata Fever, dyspnea Heart failure (max 29 mm) (NYHA IV), pulmonary embolisms, respiratory failure Single vegetation C. parapsilosis Fever, new cardiac Heart failure (max 38 mm) murmur (NYHA IV), pulmonary embolism

No evidence of IE

Multiple C. parapsilosis vegetations (unspecified size)

Periannular abscess

Single vegetation (18 mm)

Echocardiograph Finding (Diameter of Vegetation)

&

CVC, tricuspid valve

Aortic prosthetic valve

Mitral prosthetic valve

PM (distal electrodes)

November 2005/HCA, non-nosocomial

May 2006/HCA, non-nosocomial

Prior UTI Diagnosis of IE/ Admission Acquisition* (1 yr)

Medicine

12

11

10

9

M/24. HCV hepatopathy, HIV infection, IVDU. In May 2005 aortic valve replacement (biologic) for IE. Recent hospitalization. M/38. In 2003 aortic valve replacement (mechanical) for IE. PM implanted 3 mo previously. Warfarin therapy. F/66. HCV hepatopathy. Previous mitral valve replacement (mechanical). Crohn disease. F/59. Diabetes, mitral and aortic valve replacement (biologic prosthesis) 2 mo prior admission. M/73. Pancreatic neoplasm, duodenopancreatectomy 4 mo prior admission. CVC in place.

8

Aortic prosthetic valve

M/46. IVDU, HCV Prosthetic hepatopathy. In 1998 mitral valve mitral valve replacement (mechanical) for IE. Warfarin therapy. Recent abdominal surgery. CVC in place.

Site of Infection

7

Demographics (Sex/Age in yr), Patient Other Characteristics

TABLE 1. (Continued)

Falcone et al Volume 88, Number 3, May 2009

* 2009 Lippincott Williams & Wilkins

Copyright @ 2009 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Volume 88, Number 3, May 2009

Heart failure (NYHA III) Fever, new cardiac murmur, hematuria C. albicans Single vegetation (max 24 mm) Yes

None Fever C. parapsilosis Single vegetation (max 4 mm) Yes

15

F/2. Three mo prior Ventricular August 2006/HCA, admission heart surgery patch non-nosocomial, (closure of interventricular defect with ventricular patch). CVC in place. M/72. In April 2004 episode Aortic January 2005/HCA, of coagulase-negative prosthetic non-nosocomial staphylococcal IE valve (aortic valve) and valve replacement with biologic prosthetic valve. During the postoperative period candidemia in ICU treated with ampho B for 8 wk. Discharge without evidence of IE on TEE. 14

* 2009 Lippincott Williams & Wilkins

Abbreviations: Ampho B = amphotericin B, Caspo = caspofungin, CNS = central nervous system, Fluco = fluconazole, HCA = health care-associated, HCV = hepatitis C virus, HIV = human immunodeficiency virus, Itra = itraconazole, IVDU = intravenous drug user, NYHA = New York Heart Association, PM = pacemaker, Posa = posaconazole, TEE = transesophageal echocardiography, UTI = urinary tract infection, Vorico = voriconazole. *See Methods section for definitions.

&

Surgery Improved, lost (ventricular to follow-up patch replacement), Vorico 2 wk Surgery 12 mo follow-up, (prosthesis cure.(Death replacement), from other Ampho B 4 wk, causes.) followed by Fluco suppressive therapy

Medicine

Prospective Study of Candida Endocarditis

Follow-up clinical data after hospital discharge were collected from each investigator by means of outpatient visits or phone contacts.

Definitions A Blong-term indwelling central venous catheter (CVC)[ was defined as a tunneled, cuffed catheter, or a subcutaneous port catheter. A CVC was presumed to be a possible source of IE if it was present at the onset of IE symptoms, and if the same Candida species detected at the tip of the catheter was isolated from the blood cultures. An Bintracardiac abscess[ was defined as a thickened area or mass with a heterogeneous echogenic or echolucent appearance by echocardiography, or the presence of pus by direct visualization at the time of surgery.7 PVE was defined as Bearly[ if infection occurred within 60 days of valve replacement, and Blate[ if PVE developed thereafter.39 Health care-associated infection in IE was categorized as either Bnosocomial[ or Bnon-nosocomial.[38 BNosocomial health care-associated infection[ was defined as IE developing in a patient hospitalized for more than 48 hours before the onset of signs or symptoms consistent with IE. BNon-nosocomial health care-associated infection[ was defined as IE diagnosed with onset of specific symptoms within 48 hours of admission in an outpatient with extensive health care contact as reflected by any of the following criteria: 1) patient received intravenous (IV) therapy, wound care, or specialized nursing care at home within the 30 days before the onset of PVE; 2) patient attended a hospital or hemodialysis clinic or received IV chemotherapy within the 30 days before the onset of PVE; 3) patient was hospitalized in an acute care hospital for 2 or more days in the 180 days before the onset of PVE; or (4) patient resided in a nursing home or long-term care facility.11 Diagnosis of IE in pacemaker recipients was based on the Duke criteria.8 Pacemaker lead infection was considered to be definite if transesophageal echocardiography (TEE) or transthoracic echocardiography (TTE) revealed a vegetation attached to at least 1 lead or if results of macroscopic analysis and/or culture of removed leads demonstrated the presence of microorganisms.18

Statistical Analysis The results were analyzed using a commercially available statistical software package (SPSS, version 12.0; SPSS Inc, Chicago, IL). To detect significant differences between groups, we used the chi-square test or the Fisher exact test for categorical variables, and the 2-tailed t test or Mann-Whitney test for continuous variables, as appropriate. Statistical significance was established at e.05. All reported p values are 2-tailed.

RESULTS From January 2004 through December 2007, data for 903 consecutive episodes of IE were registered (822 definite cases and 81 possible cases). The following analysis is restricted to the 822 episodes of definite IE (237 had PVE). Candida IE was diagnosed in 15 patients (8 male, 7 female), with a mean age of 50 years (range, 2Y84 yr). Most patients (86.6%), at the time of diagnosis of IE, had an intravascular device (valvular prosthesis, ventricular patch, pacemaker, or CVC). Eight patients (53.3%) had PVE; 5 patients (33.3%) had NVE (3 CVC-related right-sided endocarditis, and 2 mitral valve endocarditis); 1 patient had pacemaker endocarditis; and 1 patient had an infection of a left ventricular patch. Overall, Candida species accounted for 1.8% of total cases of IE, and for 3.4% of PVE cases. Of the 15 patients included in the study, 11 (73.3%) had a history of ICU admission and 9 (60%) had undergone open heart surgery during the 12 months before the diagnosis of Candida IE. www.md-journal.com

Copyright @ 2009 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

163

Medicine

Falcone et al

&

Volume 88, Number 3, May 2009

TABLE 2. Features of Patients With Candida and Patients With Non-Candida Endocarditis* Feature Vegetation size (mean, mm) Central IV lines CRP increase at diagnosis ESR increase at diagnosis Fever Embolic phenomena Intrahospital mortality

Candida IE (n = 15)

Non-Candida IE (n = 807)

p Value

19.4 (8.8Y29.9) 46.6 (33.7Y59.5) 93.3 (86.8Y99.8) 87.7 (77.9Y95.5) 93.3 (86.8Y99.8) 46.6 (33.7Y59.5) 46.6 (33.7Y59.5)

14 (13.4Y15.1) 9.8 (8.8Y10.8) 87.2 (86Y88.4) 69.8 (68.2Y71.4) 84.9 (83.6Y86.2) 40.9 (39.2Y42.6) 16.1 (14.8Y17.4)

0.04 G0.01 1.0 0.25 0.71 0.79 G0.01

Abbreviations: CRP = C-reactive protein; ESR = erythrocyte sedimentation rate. *Data are percentage of patients, with the relative confidence intervals, unless otherwise indicated.

Demographic and disease characteristics of patients with Candida IE are summarized in Table 1. Infection was categorized as health care-associated in 13 (86.6%) patients. Of these, 10 (77%) were classified as having non-nosocomial infection and 3 (23%) were classified as having nosocomial health care-associated infection. Health careassociated PVE occurred after a median of 225 days (interquartile range, 68Y414 d) from valve implantation. One patient (Patient 7) underwent mitral valve replacement 8 years before admission; he also required a new hospitalization and major abdominal surgery 2 months before the diagnosis of Candida PVE. When this case is excluded from analysis, health careassociated PVE occurred after a median of 180 days (interquartile range, 68Y330 d). Among 807 patients with IE due to other pathogens, health care-associated PVE occurred after a median 176 days (interquartile range, 70Y635 d). C. albicans accounted for 6 cases (40%), followed by C. parapsilosis (5 cases, 33.3%); C. kruzei, C. tropicalis, C. glabrata, and C. famata accounted for 1 case each (6.6% each). Eleven patients (73.3%) had at least 2 positive blood cultures drawn 12 hours apart or all of 3 or a majority of 94 separate blood samples, while in the remaining cases diagnosis was performed by histopathologic examination of a vegetation. Antifungal drugs used for the treatment of Candida endocarditis are described in Table 1. Ten patients (66.6%) were initially treated with caspofungin alone or in combination with amphotericin B (1 case), itraconazole (1 case), or voriconazole (1 case). Two patients (13.3%) were treated with amphotericin B, 2 (13.3%) with fluconazole, and 1 (6.6%) with voriconazole. Implantable intravascular devices (CVC or pacemaker) associated with Candida IE were removed in all cases. Five patients underwent valve replacement, 1 patient refused surgery, while the remaining cases were treated with medical therapy alone. In 5 patients surgery was considered contraindicated due to the severity of underlying clinical conditions. The overall intrahospital mortality was 46.6% (7 cases): mortality rates were higher in patients with PVE (5 of 8 cases, 62.5%) than in patients with NVE (2 of 5 patients, 40%). Among the 5 patients who underwent surgical intervention with valve replacement, 3 were cured (mean follow-up, 17.3 mo; range, 12Y24 mo) and 2 were discharged without signs or symptoms of infection (cases lost to follow-up). The 6 patients who did not undergo surgery because it was contraindicated (n = 5) or who refused surgical intervention (n = 1) eventually died (3 patients had persistent fungemia despite caspofungin, voriconazole, or fluconazole therapy). As reported in Table 1, 4 remaining patients were treated with medical therapy alone. Of these, 3 were cured: the first patient (Patient 9), with pacemaker endocarditis, was

164

www.md-journal.com

treated for 6 weeks with caspofungin IV followed by 12 weeks of oral fluconazole and posaconazole (follow-up, 14 mo); the second patient (Patient 4), with CVC-related right-sided IE, was treated with caspofungin IV for 12 weeks followed by 4 weeks of oral fluconazole therapy (follow-up, 20 mo); the third patient (Patient 5), with CVC-related right-sided IE, was treated with caspofungin IV for 4 weeks followed by chronic suppressive fluconazole therapy (follow-up, 8 mo). The remaining patient (Patient 3) died a few days after hospital admission (a case of relapsing Candida PVE) despite treatment with IVamphotericin B. As described in Table 2, compared to patients with IE due to other pathogens, patients with Candida IE had a higher incidence of central venous lines (p G 0.01), a higher mean vegetation size (p G 0.04), and a significantly higher intrahospital mortality (p G 0.01). Five of 10 cases of Candida IE treated with caspofungin had a favorable response. These included 1 case of pacemaker endocarditis (Patient 9), 2 cases of PVE that received a combined medical-surgical treatment (Patients 8, 10), and 2 cases of CVC-related right-sided endocarditis (Patients 4, 5); the latter patient (Patient 5) died 8 months after the hospital discharge without symptoms and signs of relapsed Candida IE. Three patients with PVE (Patients 6, 11, 13) and 2 with NVE (Patients 2, 12), who received treatment with caspofungin alone, died as a consequence of their infection.

DISCUSSION To our knowledge, the present study is the first prospective, multicenter study analyzing the epidemiology, clinical features, and outcomes of Candida endocarditis in a national cohort. Our results confirm Candida species as an example of health careassociated infection, usually affecting patients with intravascular/ intracardiac devices. Candida endocarditis accounts for 1.8% of the total cases of IE, and 3.4% of the cases of PVE, and, compared to cases of IE due to other etiologies, is associated with a significantly higher intrahospital mortality. The epidemiology, risk factors, and outcome of Candida IE are not well understood, because of the rarity of cases and the lack of large prospective cohorts. The largest studies were reported more than 30 years ago,32 or represent retrospective analyses of existing databases containing cases observed over a long time.4 Candida IE was reported in patients with a median age of 45 years in 2 large reviews,10,28 but patients with PVE are usually older than those with NVE.28 Major risk factors have been found to be previous valve surgery, antibiotic use, rheumatic heart disease, surgery other than cardiac, vascular lines, immunosuppressive treatment, underlying valve disease, previous bacterial * 2009 Lippincott Williams & Wilkins

Copyright @ 2009 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Medicine

&

Volume 88, Number 3, May 2009

Prospective Study of Candida Endocarditis

FIGURE 1. Proposed physiopathology of Candida PVE. *Postoperative candidemia frequently undetected because of the low sensitivity of blood cultures. TPN = total parenteral nutrition.

endocarditis, and injection drug abuse.10,28 During the last decades the number of patients undergoing open heart surgery and carrying prosthetic implants, such as mechanical valves or other intravascular devices, has increased. Moreover, many patients who undergo such operations are critically ill either preoperatively or postoperatively, and require prolonged ICU monitoring, with an increasing risk of developing nosocomial fungal infection.24 In particular, patients undergoing cardiac surgery are at higher risk of candidemia for the following reasons: 1) they usually require prolonged antibiotic therapy, which facilitates intestinal overgrowth of Candida species;5 2) the prolonged duration of the surgical procedure is frequently associated with postoperative splanchnic hypoperfusion, leading to intestinal intramucosal acidosis and increased intestinal mucosal permeability;34 3) ventilator-associated pneumonia and antibiotic therapy-associated endotoxemia may damage the intestinal mucosal barrier;13 and 4) several other risk factors for candidemia are usually simultaneously present, such as CVC, total parenteral nutrition, and acute renal failure.20 Implantation of a contaminated allograft can also occur. All these factors might predispose the spread of Candida species into the systemic circulation and the development of candidemia. Thus, the scenario of a patient with a prosthetic heart valve or other intravascular prosthetic device who develops candidemia is becoming more common. The risk of developing PVE in candidemic patients with prosthetic heart valves was addressed in a retrospective study by Nasser and collaborators.26 Overall, Candida PVE was documented in 11 of 44 (25%) patients: in 7 patients it was * 2009 Lippincott Williams & Wilkins

diagnosed at the same time as candidemia (Group 1), while in the remaining 4 patients (Group 2) it was diagnosed several weeks after an episode of early postoperative candidemia. Reviewing these published data, we noted that in both groups PVE was diagnosed late after valve replacement (after a mean of 270 days in Group 1, and after a mean of 246 days in Group 2),26 a fact suggesting that all the episodes had a similar pathogenic mechanism. We can speculate that in Group 1 patients, an early postoperative candidemia remained microbiologically undetected, but led to colonization of prosthetic valve, with clinical and pathologic evidence of PVE only several weeks after the initial episode of candidemia. Similarly, Group 2 patients, who developed a documented early postoperative candidemia, were diagnosed as having PVE only several weeks later, when the endocardial infection was clinically evident: these patients had negative blood cultures at the time of diagnosis of Candida PVE.26 The clinical course of our Patient 15 (Table 1) is illustrative of this concern: the patient developed early postoperative candidemia treated with liposomal amphotericin B for 8 weeks. Discharged from the hospital without evidence of IE at transesophageal echocardiography, he presented 9 months later with the signs and symptoms of relapsing candidemia and PVE. Among our cohort, 2 additional patients (Patients 3, 13) had confirmed candidemia in the postoperative period with late Candida PVE. On the basis of these data, the physiopathology of Candida PVE may be a 2-step process (Figure 1): the first step is represented by a postoperative transitory candidemia (frequently www.md-journal.com

Copyright @ 2009 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

165

Falcone et al

not detected due to the low sensitivity of blood cultures) occurring during the ICU stay, which leads to colonization of prosthetic valve and subsequent biofilm formation.30 After the initial colonization, the fungus, slowly growing on the prosthesis surface, becomes more resistant to antifungal agents; the clinical syndrome consistent with IE develops late in the course of the infection, often several months after initial candidemia, when the vegetations reach sizes that allow echocardiographic visualization and/or septic embolization. However, because of the small number of published data available, future studies are needed to confirm this hypothesis. Of interest, Candida endocarditis usually has a long duration of illness also in cases of NVE. In 1 review of isolated tricuspid valve endocarditis in non-drug addict patients, the time elapsing between infection and NVE onset was 9 month on average,25 and some species, such as C. parapsilosis, have been reported to recur as late as 43 months after an initial episode of transient candidemia.17,37 The spectrum of Candida species infections has changed over the years as a result of improved diagnostic techniques and modifications in medical practice. Traditionally, C. albicans has been considered the most common species10,28 isolated in patients with Candida IE, but non-albicans Candida species, particularly C. parapsilosis,4,9 have emerged. In the recent analysis by Baddley et al,3 C. albicans accounted for less than 50% of cases, followed by C. parapsilosis, C. glabrata, and C. tropicalis. This finding was confirmed in the current series, where C. albicans represented the causing pathogen in only 40% of cases, followed by C. parapsilosis accounting for 33% of cases. Characteristically, valvular lesions are left heart in distribution (aortic and mitral valves being the most common sites of infection10), and occasionally nonvalvular. Right-sided Candida IE is more frequent in IV drug users and patients with CVCs in place.10,28 As already reported,10 the most common clinical features are fever, changing or new heart murmur, major peripheral embolization, focal or generalized neurologic symptoms, and heart failure. These observations seem to be confirmed in the study by Baddley et al3 and in the current series; however, we frequently observed a combination of more than 1 symptom or sign, especially in patients with PVE. Early diagnosis is difficult, because of the low sensitivity of blood cultures: in the review by Ellis et al,10 a positive mycologic diagnosis was made for 139 of 260 patients by blood cultures (54% sensitivity). However, in a review including cases of fungal IE observed in the period 1995Y2000,28 81.2% of patients with yeast-related IE had positive blood cultures, similar to the 73.3% rate observed in our cohort. These findings may reflect an improvement in the microbiologic techniques used. Candida IE, especially PVE, is usually diagnosed late after initial valvular colonization, when candidemia is detectable and vegetations reach large sizes. As matter of fact, patients with Candida IE had a significantly greater mean diameter of vegetation than patients with IE due to other etiologies, a fact consistent with previous data.10,28 The large size of vegetations is probably also responsible for the high rate of embolic phenomena (46.6% in the present study). However, as also observed in the 2008 report by Baddley et al,3 the incidence of embolic events in patients with Candida IE was not significantly higher than in nonfungal cases. Traditionally, Candida IE has been considered a fatal infection, with mortality approaching 80% in some series,4,10,28 despite aggressive antifungal and surgical therapy. The most recent reviews demonstrated that the crude survival has increased over the years,10,28 with a trend toward improved outcome. Among our patients, the intrahospital mortality rate

166

www.md-journal.com

Medicine

&

Volume 88, Number 3, May 2009

was less than 50% (46.6%), but was significantly higher than that of patients with nonfungal IE (16.1%; p G 0.01). Surprisingly, Baddley et al3 reported an even lower mortality rate (30.3%). Possible reasons for this improved survival are better echocardiographic techniques, earlier diagnosis of endocarditis, and better supportive care of ill patients. However, as confirmed in the current study, Candida IE still represents an infection complicated by an excess of intrahospital mortality. Ten patients (66.6%) in the present study were treated with caspofungin, alone or in combination with other antifungals. To our knowledge, this study represents the largest experience available on echinocandin therapy of Candida endocarditis. The Infectious Diseases Society of America 2004 guidelines on the treatment of candidiasis recommend IV amphotericin B and 5fluorocytosine in addition to valve surgery as first-line treatment for Candida IE.27 However, recent in vitro studies have shown reduced activity of amphotericin B against Candida biofilm,19Y31 and poor penetration into vegetations and blood clots in experimental models of IE.33 On the contrary, caspofungin displays potent in vitro activity against sessile Candida cells within biofilms,2Y16 and has been successfully used in anecdotal cases of Candida IE.1,15,23,29 Nevertheless, we here outline that caspofungin therapy alone, followed by an oral azole, achieved a favorable clinical response in only 3 cases (2 had CVC-related right-sided NVE, and 1 had pacemaker endocarditis). Among the remaining 7 cases, 2 patients with PVE improved or were cured with combined caspofungin and surgical therapy, while 5 patients treated with medical therapy alone eventually died (3 with PVE and 2 with NVE). These results suggest that, regardless of the antifungal agent used, surgical therapy remains the cornerstone of therapy, as confirmed in the meta-analysis of Steinbach et al,35 who found the antifungal monotherapy without adjunctive surgery to be the variable associated with the poorest patient outcome. The potential role of caspofungin therapy, as also recently described in a small cohort of patients with Candida IE,36 remains to be defined. Echinocandins followed by long-term therapy with fluconazole may be effective in cases of right-sided Candida NVE; however, there are not sufficient clinical data to support the superiority of newer agents compared to the Infectious Diseases Society of America recommendations. These results, despite some limitations of our cohort (small number of cases, and most of the centers are referring centers with Cardiosurgery units), lead to the following considerations about the management of Candida endocarditis: 1) An aggressive approach (with preemptive use of fungicidal agents such as echinocandins) should be considered in patients with prosthetic intravascular devices and significant risk factors for nosocomial candidemia; 2) Optimal duration of antifungal therapy for nosocomial candidemia in patients with prosthetic heart valves remains to be defined, but more prolonged courses (for example, 6Y8 wk) with fungicidal agents are probably required because of the possibility of late-onset endocarditis; 3) Clinicians searching for evidence of PVE in patients with prosthetic heart valves who become candidemic must be diligent, because clinical evidence of fungal PVE is frequently delayed; and 4) A combined medical and surgical approach is needed for successful treatment of infection. In conclusion, Candida IE should be classified as an emerging infectious disease, usually involving people with intravascular prosthetic devices, and associated with substantial related morbidity and mortality. The diagnosis of Candida PVE * 2009 Lippincott Williams & Wilkins

Copyright @ 2009 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Medicine

&

Volume 88, Number 3, May 2009

is difficult and usually delayed, because endocardial vegetations become evident at echocardiography investigation even several months from an initial episode of early postoperative candidemia. Once Candida PVE is diagnosed, a combined medical and surgical approach is required for optimal therapy. With the continued expansion of medical and surgical techniques, we expect that an increasing number of these infections will occur. ACKNOWLEDGMENT This article is dedicated to the memory of Pietro Martino, Professor of Medicine. As a physician and as a researcher he was a pioneer in the fields of invasive fungal infections and infective endocarditis. As a man he will remain forever a beloved friend.

APPENDIX SEI investigators: A. Biglino, P. Crivelli, Unita` Operativa Malattie Infettive, Ospedale Civile di Asti; R. Stellini, S. Magri, Clinica Malattie Infettive e Tropicali, Universita` degli Studi di Brescia; V. Del Bono, F. Vitale, Clinica Malattie Infettive, Universita` di Genova; A. Scalzini, G. Gattuso, U.O. Malattie Infettive, Azienda Ospedaliera BC. Poma,[ Mantova; E. Durante Mangoni, E. Ragone, S. Cuccurullo, UOC Medicina Infettivologica e dei Trapianti, AO Monaldi, Napoli; M. F. Tripodi, II Div. Medicina Interna ed Epatologia, II Universita` di NapoliVPoliclinico Universitario; G. Spoladore, Malattie Infettive, IRCCS San Matteo, Pavia; M. B. Pasticci, L. Martinelli, U.O. Mal. Infettive, Azienda Ospedaliera di Perugina; N. Petrosillo, P. Chinello, II Div. Malattie Infettive, INMI BL. Spallanzani,[ Roma; P. Delle Foglie, Sezione Malattie InfettiveVPresidio Ospedaliero S. Chiara, Ospedale di Trento; A. Vaglia, Unita` Operativa Malattie Infettive, Ospedale S. Maria di Ca_ Foncello, Treviso; P. Viale, M. Crapis, Clinica Malattie Infettive, Policlinico Universitario, Udine; R. Tambini, L. Tomasello, Clinica Malattie Infettive e Tropicali, Universita` degli Studi dell_Insubria, Varese; G. P. Pellizzer, R. Nicolin, UO Malattie Infettive e Tropicali, Ospedale San Bortolo, Vicenza; E. Concia, P. Del Bravo, A. Tedesco, Clinica Malattie Infettive, Universita` degli Studi di Verona. REFERENCES 1. Bacak V, Biocina B, Starcevic B, Gertler S, Begovac J. Candida albicans endocarditis treatment with caspofungin in an HIV-infected patientVcase report and review of literature. J Infect. 2006;53:e11Ye14. 2. Bachmann SP, VandeWalle K, Ramage G, Patterson TF, Wickes BL, Graybill JR, Lopez-Ribot JL. In vitro activity of caspofungin against Candida albicans biofilm. Antimicrob Agents Chemother. 2002; 46:3591Y3596. 3. Baddley JW, Benjamin DK Jr, Patel M, Miro J, Athan E, Barsic B, Bouza E, Clara L, Elliott T, Kanafani Z, Klein J, Lerakis S, Levine D, Spelman D, Rubinstein E, Tornos P, Morris AJ, Pappas P, Fowler VG Jr, Chu VH, Cabell C. International Collaboration on EndocarditisProspective Cohort Study Group (ICE-PCS). Candida infective endocarditis. Eur J Clin Microbiol Infect Dis. 2008;27:519Y529. 4. Benjamin DK Jr, Miro JM, Hoen B, Steinbach WJ, Fowler VG Jr, Olaison L, Habib G, Abrutyn E, Perfect J, Zass A, Corey GR, Eykyn S, Thuny F, Jimenez-Exposito MJ, Cabell CH; ICE-MD Study Group. Candida endocarditis: contemporary cases from the International Collaboration of Infectious Endocarditis Merged Database (ICE-MD). Scand J Infect Dis. 2004;36:453Y455.

* 2009 Lippincott Williams & Wilkins

Prospective Study of Candida Endocarditis

5. Bernhardt H, Zimmermann K, Knoke M. The continuous flow culture as an in vitro model in experimental mycology. Mycoses. 1999; 42(S2):29SY32S. 6. Challa S, Prayaga AK, Vemu L, Sadasivan J, Jagarlapudi MK, Digumarti R, Prabhala R. Fungal endocarditis: an autopsy study. Asian Cardiovasc Thorac Ann. 2004;12:95Y98. 7. Daniel WG, Mugge A, Martin RP, Lindert O, Hausmann D, Nonnast-Daniel B, Laas J, Lichtlen PR. Improvement in the diagnosis of abscesses associated with endocarditis by transesophageal echocardiography. N Engl J Med. 1991;324:795Y800. 8. Durack DT, Lukes AS, Bright DK. New criteria for diagnosis of infective endocarditis: utilization of specific echocardiographic findings. Duke Endocarditis Service. Am J Med. 1994;96:200Y209. 9. Ellis M. Fungal endocarditis. J Infect. 1997;35:99Y103. 10. Ellis ME, Al-Abdely H, Sandridge A, Greer W, Ventura W. Fungal endocarditis: evidence in the world literature, 1965Y1995. Clin Infect Dis. 2001;32:50Y62. 11. Fernandez-Hidalgo N, Almirante B, Tornos P, Pigrau C, Sambola A, Igual A, Pahissa A. Contemporary epidemiology and prognosis of health care-associated infective endocarditis. Clin Infect Dis. 2008;47:1287Y1297. 12. Gordon S, Keys T. Bloodstream infections in patients with prosthetic cardiac valves. Sem Thoracic Cardiovasc Surg. 1995;7:2Y6. 13. Holzheimer RG. Antibiotic induced endotoxin release and clinical sepsis: a review. J Chemother. 2001;13:159Y172. 14. Ivert T, Dismukes W, Cobb C, Blackstone EH, Kirklin JW, Bergdahl LA. Prosthetic valve endocarditis. Circulation. 1984;69:223Y232. 15. Jimenez-Exposito MJ, Torres G, Baraldes A, Jimenez-Exposito MJ, Torres G, Baraldes A, Benito N, Marco F, Pare JC, Moreno A, Claramonte X, Mestres CA, Almela M, Garcia de la Maria C, Perez N, Schell WA, Corey GR, Perfect J, Jimenez de Anta MT, Gatell JM, Miro JM. Native valve endocarditis due to Candida glabrata treated without valvular replacement: a potential role for caspofungin in the induction and maintenance treatment. Clin Infect Dis. 2004;39:e70Ye73. 16. Katragkou A, Chatzimoschou A, Simitsopoulou M, Dalakiouridou M, Diza-Mataftsi E, Tsantali C, Roilides E. Differential activities of newer antifungal agents against Candida albicans and Candida parapsilosis biofilms. Antimicrob Agents Chemother. 2008;52:357Y360. 17. Khan UM, Ali S, Baig MA, Rafig MA, Vasavada BC, Khan IA. Candida parapsilosis endocarditis 8 months after transient candidemia. Int J Cardiol. 2007;118:e58Ye59. 18. Klug D, Balde M, Pavin D, Hidden-Lucet F, Clementy J, Sadoul N, Rey JL, Lande G, Lazarus A, Victor J, Barnay C, Grandbastien B, Kacet S; PEOPLE Study Group. Risk factors related to infections of implanted pacemakers and cardioverter-defibrillators. Results of a large prospective study. Circulation. 2007;116:1349Y1355. 19. Kuhn DM, George T, Chandra J, Mukherjee PK, Ghannoum MA. Antifungal susceptibility of Candida biofilms: unique efficacy of amphotericin B lipid formulations and echinocandins. Antimicrob Agents Chemother. 2002;46:1773Y1780. 20. Leon C, Ruiz-Santana S, Saavedra P, Almirante B, Nolla-Salas J, Alvarez-Lerma F, Garnacho-Montero J, Leon MA; EPCAN Study Group. A bedside scoring system (BCandida Score[) for early antifungal treatment in nonneutropenic critically ill patients with Candida colonization. Crit Care Med. 2006;34:730Y737. 21. Levy I, Shalit I, Birk E, Klinger G, Sirota L, Linder N. Candida endocarditis in neonates: report of five cases and review of the literature. Mycoses. 2006;49:43Y48. 22. Li JS, Sexton DJ, Mick N, Nettles R, Fowler VG Jr, Ryan T, Bashore T, Corey GR. Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis. Clin Infect Dis. 2000;30:633Y638. 23. Lye DC, Hughes A, O_Brien D, Athan E. Candida glabrata prosthetic

www.md-journal.com

Copyright @ 2009 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

167

Falcone et al

valve endocarditis treated successfully with fluconazole plus caspofungin without surgery: a case report and literature review. Eur J Clin Microbiol Infect Dis. 2005;24:753Y755. 24. Michalopoulos AS, Geroulanos S, Mentzelopoulos SD. Determinants of candidemia and candidemia-related death in cardiothoracic ICU patients. Chest. 2003;124:2244Y2255.

Medicine

&

Volume 88, Number 3, May 2009

32. Rubenstein E, Noriega ER, Simberkoff MS, Holzman R, Rahal JJ Jr. Fungal endocarditis: analysis of 24 cases and review of the literature. Medicine (Baltimore). 1975;54:331Y344. 33. Rubinstein E, Noriega ER, Simberkoff MS, Rahal JJ. Tissue penetration of amphotericin B in Candida endocarditis. Chest. 1974;66:376Y377.

25. Nandakumar R, Raju G. Isolated tricuspid valve endocarditis in nonaddicted patients: a diagnostic challenge. Am J Med Sci. 1997; 314:207Y212.

34. Ryan T, McCarthy JF, Rady MY, Serkey J, Gordon S, Starr NJ, Cosgrove DM. Early bloodstream infection after cardiopulmonary bypass: frequency rate, risk factors, and implications. Crit Care Med. 1997; 25:2009Y2014.

26. Nasser RM, Melgar GR, Longworth GL, Gordon SM. Incidence and risk of developing fungal prosthetic valve endocarditis after nosocomial candidemia. Am J Med. 1997;103:25Y32.

35. Steinbach WJ, Perfect JR, Cabell CH, Fowler VG, Corey GR, Lie JS, Zaasb AK, Benjamin DK Jr. A meta-analysis of medical versus surgical therapy for Candida endocarditis. J Infect. 2005;51:230Y247.

27. Pappas PG, Rex JH, Sobel JD, Filler SG, Dismukes WE, Walsh TJ, Edwards JE; Infectious Diseases Society of America. Guidelines for treatment of candidiasis. Clin Infect Dis. 2004;38:161Y189.

36. Talarmin JP, Boutoille D, Tattevin P, Abgueguen P, Ansart S, Roblot F, Raffi F. Candida endocarditis: role of new antifungal agents. Mycoses. 2009;52:60Y66.

28. Pierrotti LC, Baddour LM. Fungal endocarditis, 1995-2000. Chest. 2002;122:302Y310.

37. Tonomo K, Tsujino T, Fujioko Y. Candida parapsilosis endocarditis that emerged 2 years after abdominal surgery. Heart Vessels. 2004;19: 149Y152.

29. Rajendram R, Alp NJ, Mitchell AR, Bowler ICJW, Forfar JC. Candida prosthetic valve endocarditis cured by caspofungin therapy without valve replacement. Clin Infect Dis. 2005;40:e72Ye74. 30. Ramage G, Martinez JP, Lopez-Ribot JL. Candida biofilms on implanted biomaterials: a clinically significant problem. FEMS Yeast Res. 2006;6:979Y986. 31. Ramage G, VandeWalle K, Bachmann SP, Wickes BL, Lopez-Ribot JL. In vitro pharmacodynamic properties of three antifungal agents against preformed Candida albicans biofilms determined by time-kill studies. Antimicrob Agents Chemother. 2002;46:3634Y3636.

168

www.md-journal.com

38. Wang A, Athan E, Pappas PA, Fowler VG Jr, Olaison L, Pare C, Almirante B, Munoz P, Rizzi M, Naber C, Logar M, Tattevin P, Iarussi DL, Selton-Suty C, Jones SB, Casabe J, Morris A, Corey GR, Cabell CH; International Collaboration on Endocarditis-Prospective Cohort Study Investigators. Contemporary clinical profile and outcome of prosthetic valve endocarditis. JAMA. 2007;297:1354Y1361. 39. Wilson WR, Jaumin PM, Danielson GK, Giuliani ER, Washington JA II, Geraci JE. Prosthetic valve endocarditis. Ann Intern Med. 1975; 82:751Y756.

* 2009 Lippincott Williams & Wilkins

Copyright @ 2009 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.