Rare infection of implantable cardioverter-defibrillator lead with Candida albicans : case report and literature review

Abstract

Infection of implanted cardiac devices has a low rate of occurrence. Fungal infections of such devices represent an atypical phenomenon, associated with high mortality. Both medical and surgical therapies are recommended for a successful outcome.

A 60-year-old woman with past medical history of heart failure with reduced ejection fraction, implantable cardioverter-defibrillator (ICD) placement, sarcoidosis and diabetes presented with fevers and atypical pleuritic chest pain. Transthoracic echocardiogram revealed a highly mobile 2.09 cm by 4.49 cm mass associated with the ICD wire. Blood cultures were positive for Candida albicans. The patient underwent sternotomy for removal. The vegetation was 4 cm by 2 cm by 2 cm in size, attached to the right ventricle without interference with the tricuspid valve. The patient was treated with micafungin for 2 weeks and then fluconazole for 6 weeks.

In this case report, we describe the rare infection of an ICD lead with C. albicans, in the form of a fungal ball. This is the 18th reported case of Candida device-related endocarditis and the first reported in a woman. Prior case reports have occurred primarily in pacemaker rather than ICD leads. The vegetation size is also one of the largest that has been reported, measuring 4 cm at its greatest length.

As Candida device-related endocarditis is so rare, and as fatality occurs in half of cases, clinical management can only be derived from sporadic case reports. Therefore, the course of this patient’s disease care will be a useful adjunct to the current literature for determining treatment and prognosis in similar cases.

Introduction

Implanted cardiac devices, specifically pacemakers and implantable cardioverter-defibrillators (ICDs), have a relatively low rate of becoming infected, from 0.8% to 5.7% [Arber et al. 1994; Lai and Fontecchio, 1998; Smith et al. 1998; Eggimann and Waldvogel, 2000; Duval et al. 2004; Uslan et al. 2007]. Multiple risk factors have been associated with infection. These include comorbid conditions such as heart failure, diabetes mellitus and renal dysfunction, immunosuppressive states, oral anticoagulation use, and device-specific issues such as recent device manipulation and temporary pacing prior to permanent device placement [Lai and Fontecchio, 1998; Smith et al. 1998; Eggimann and Waldvogel, 2000; Karchmer and Longworth, 2002; Duval et al. 2004; Bloom et al. 2006; Klug et al. 2007; Uslan et al. 2007; Lekkerkerker et al. 2009; Baddour et al. 2010].

Intracardiac device infections are classified into pocket infections and deeper systemic infections [Kojic et al. 2004]. Cardiac device-related pocket infections occur at a rate of 5.7% and involve subcutaneous infection of the device [Arber et al. 1994; Kojic and Darouiche, 2004]. They are commonly associated with contamination that occurs peri-procedurally and present within a month of placement [Lewis et al. 1985; Smith et al. 1998; Giamarellou, 2002; Karchmer and Longworth, 2002; Massoure et al. 2007; Sohail et al. 2007a, 2007b]. Cardiac device-related deeper infections occur at a rate of 0.8% and are denoted as ‘device-related endocarditis’ [Arber et al. 1994; Kojic and Darouiche, 2004]. They include the intracardiac portion of the device, specifically the electrode tip, the tricuspid valve, and the areas of the endocardium that are in contact with the tip [Arber et al. 1994]. The majority of deeper infections presenting within 2 weeks of implantation are caused by Staphylococcous species; whereas later disease presentation have varied microbial etiologies [Lewis et al. 1985; Smith et al. 1998; Karchmer and Longworth, 2002; Massoure et al. 2007; Sohail et al. 2007a, 2007b]. Approximately 2% of all implanted cardiac device related infections have fungi identified as the etiology [Brown et al. 2001; Hindupur and Muslin, 2005; Ho et al. 2006; Sohail et al. 2007a, 2008]. Specifically, Candida species have a propensity for adherence to and colonization of foreign bodies, forming biofilms [Kojic and Darouiche, 2004; D’enfert, 2006]. Less than 20 cases of Candida device-related endocarditis have been described in case studies [Halawa et al. 2011].

In this case report, we discuss the rare infection of an ICD lead with Candida albicans and its successful treatment through both surgical and medical intervention.

Case report

A 60-year-old woman with past medical history of sarcoidosis and diabetes mellitus presented to the emergency department secondary to subjective fevers, chills, sweats, and productive cough for 1–2 weeks, followed by 3–4 days of atypical, pleuritic chest pain. Due to the patient’s depressed left ventricular ejection fraction and episodes of nonsustained ventricular tachycardia, she underwent implantation of a single-chamber ICD 2 years and 2 months prior to the current presentation.

Blood cultures were obtained and a transthoracic echocardiogram (TTE) performed. TTE revealed her ejection fraction had improved to 55% and a highly mobile 2.09 cm by 4.49 cm mass associated with the ICD wire (not shown). Blood cultures were found to be positive for C. albicans. In light of the vegetation size, the patient’s septic state, the multiple comorbidities, and high-risk nature of the infectious agent, open heart surgery for direct extraction of the lead and vegetation was selected.

The patient underwent sternotomy with cardiopulmonary bypass and intraventricular intracardiac exploration for removal of the infected ICD and removal of the ICD generator with revision of the ICD generator pocket along with an intraoperative transesophageal echocardiography (TEE). The vegetation was attached to the right ventricle without interference with the tricuspid valve (Figure 1, Supplemental Video 1: available at: tak.sagepub.com/supplemental). The vegetation measured 4 cm by 2 cm by 2 cm and a fibrous capsule was found to extend up and over the ICD lead measuring 4 cm by 3 cm by 1 cm (Figure 2). Pathology and culture confirmed C. albicans (Figure 3). Following removal of the ICD, the patient was treated initially with micafungin for 2 weeks and then with fluconazole for 6 weeks.

Figure 1.

Transesophageal echocardiogram identifying a 4.49 cm by 2.09 cm echogenic, highly mobile mass attached to the implantable cardioverter-defibrillator lead in the right ventricle. (a) Two-chamber view. (b) Aortic valve long axis. (c) Aortic valve short axis. (d) Off-axis view of the right ventricle. (e) Four-chamber view. (f) Bicaval view.

Figure 2.

Gross specimen of vegetation affixed to the implantable cardioverter-defibrillator (ICD) lead. The vegetation measured 4 cm by 2 cm by 2 cm. A fibrous capsule was found to extend up and over the ICD lead measuring 4 cm by 3 cm by 1 cm.

Figure 3.

Histologic identification of Candida albicans in tissue from extracted vegetation. Hematoxylin and eosin stain of the vegetation with inflammatory cells and fungal elements at (a) 200× and (b) 400× magnification. Periodic acid Schiff stain showing pseudohyphae and yeast-like cells suggestive of Candida at (c) 200× and (d) 400× magnification.

This immunosuppressed patient had multiple risk factors for this rare infection, including sarcoidosis, insulin-dependent diabetes mellitus, and chronic kidney disease. The etiology of her fungemia remains elusive. The presentation of her ICD endocarditis was a deeper infection that did not present until 2 years and 2 months following device placement. This delayed presentation makes seeding at the time of implantation less likely. Nine months prior to her presentation, the patient underwent multiple abdominal surgeries, including ventral hernia repair complicated by incision dehiscence and infection requiring multiple abdominal wall debridements. During this time period she became febrile and blood cultures were found to be positive for C. albicans. TTE was performed to evaluate for a subclinical endocarditis and revealed a reduced ejection fraction of 35–40% but no visible vegetation.

The patient was closely followed by infectious disease and cardiology specialists. She had multiple episodes of recurrent fevers and chest discomfort. Blood cultures were monitored at regular intervals and remained negative. Seven months after her original presentation, the patient again presented to the hospital with fever and chest discomfort. TTE and TEE revealed a moderate size vegetation on the tricuspid valve (Figure 4, Supplemental Video 2: available at: tak.sagepub.com/supplemental). Pending the results of blood cultures, the patient was treated with fluconazole. Blood cultures at this time continued to be negative and she was treated for ‘culture-negative endocarditis’ with a planned 6-week course of vancomycin and ceftriaxone. The patient remained febrile and returned to the hospital 4 weeks later with respiratory complaints and a left upper lobe consolidation on chest computed tomography (CT). Follow-up blood cultures, along with sputum culture and bronchoalveolar lavage were positive for C. albicans. The patient was treated with micafungin for 1 week and then continued with a 6-week course of amphotericin. Follow-up TTE demonstrated resolution of the vegetation (not shown). Close patient follow up for the subsequent 2 years demonstrated the patient to be C. albicans culture negative and follow-up echocardiograms did not reveal vegetations. Due to continued improvement in the patient’s congestive heart failure symptoms and ejection fraction (45–50%), along with the absence of cardiac arrhythmias, she has not undergone reimplantation of an ICD. She is currently on optimal medical therapy and without cardiac-related complaints.

Figure 4.

Echocardiograms identifying a 1 cm echogenic mass attached to the tricuspid valve. (a) Transesophageal echocardiogram. (b) Transthoracic echocardiogram.

Discussion

Device-related endocarditis is relatively uncommon yet increasingly recognized given the expanded indications for device therapy. The most common etiology is Staphylococcus [Lewis et al. 1985; Smith et al. 1998; Karchmer and Longworth, 2002; Massoure et al. 2007; Sohail et al. 2007a, 2007b]. An extensive search of the literature using PubMed and Medline confirmed the scarcity of case reports of fungal infection of implantable cardiac devices [Lewis et al. 1985; Smith et al. 1998; Karchmer and Longworth, 2002; Massoure et al. 2007; Sohail et al. 2007a, 2007b; Halawa et al. 2011]. In the early years of its identification, fungal device-related endocarditis was fatal and found only postmortem [Bryan et al. 1978; Bluhm et al. 1982; Phibbs and Marriott, 1985; Heimberger and Duma, 1989]. The first antemortem diagnosis and survivor was reported in 1993 [Wilson et al. 1993]. Fungal device-related endocarditis is often elusive due to commonly negative blood cultures and low index of suspicion [Isotalo et al. 2001]. Associated mortality ranges from 50% to 94% [Mehta, 1990; Bohm et al. 1996; Ellis, 1997; Pierrotti and Baddour, 2002]. Specific risk factors for the development of Candida device-related endocarditis have not yet been identified [Kojic and Darouiche, 2004]. General risk factors for Candida infections include immunosuppressed states, diabetes mellitus, intravenous drug use, existing cardiac abnormalities, antibiotic therapy, and indwelling devices [Joly et al. 1997; Isotalo et al. 2001; Pierrotti and Baddour, 2002; Kojic and Darouiche, 2004].

Whereas infective endocarditis has defined diagnostic criteria [Durack et al. 1994], there is no classic presentation or constellation of symptoms that suggest device-related endocarditis. The time from implantation to onset of infection can vary between days and years [Joly et al. 1997; Sohail et al. 2007a], however one-third of reported cases occur within 6 months [Cacoub et al. 1998; Duval et al. 2004; Klug et al. 2007]. Multiple studies have demonstrated nonspecific findings such as leukocytosis, anemia, high sedimentation rate, and fever/chills. Positive blood cultures along with other signs of sepsis have been reported [Kojic and Darouiche, 2004; Sohail et al. 2007a; Grunberg et al. 2013], however cases with negative blood cultures have also been reported [Sohail et al. 2007a; Halawa et al. 2011]. Variation in symptoms has also been dependent on transvenous versus epicardial electrode involvement and time from onset of device implantation [Cacoub et al. 1998; Schulze et al. 2005; Klug et al. 2007; Sohail et al. 2007a]. Altogether, there are no definitive findings that identify device-related endocarditis.

In evaluating the small series of published case reports, there was a significant advantage of combining medical treatment with antimicrobials and surgical removal of the entire implanted cardiac device system with associated vegetation over antimicrobials alone [Mehta, 1990; Arber et al. 1994; Bohm et al. 1996; Joly et al. 1997; Klug et al. 1997; Cacoub et al. 1998; Victor et al. 1999; Bracke et al. 2000; Chua et al. 2000; Kurup et al. 2000; Giamarellou, 2002; Baddour et al. 2003; Del Rio et al. 2003; Rallidis et al. 2003; Darouiche, 2004; Sohail et al. 2007a; Uslan et al. 2007; Talarmin et al. 2009]. This may be due in part to the maintenance of microbial biofilms following cessation of antimicrobials [Kojic and Darouiche, 2004]. One study found a sevenfold increase in 30-day mortality and threefold increase in mortality at 1 year in patients who did not have device removal [Le et al. 2011; Athan et al. 2012]. Still, there have been few case reports that achieved successful eradication of infection with medical therapy alone when surgery was contraindicated [Mehta, 1990; Roger et al. 2000]. Furthermore, some investigators have recommended thoracotomy in cases when vegetations are greater than 1 cm due to risk of symptomatic embolization [Kennelly and Piller, 1974; Mansour et al. 1985; Klug et al. 1997; Cacoub et al. 1998; Voet et al. 1999; Dumont et al. 2003; Meier-Ewert et al. 2003; Rallidis et al. 2003; Duval et al. 2004; Klug et al. 2004, 2007; Massoure et al. 2007]. However, others have not seen significant pulmonary embolism in percutaneous removal of vegetations up to 7 cm [Victor et al. 1999; Nguyen et al. 2000; Del Rio et al. 2003; Ruttmann et al. 2006; Sohail et al. 2007a].

In 2007, the Mayo Clinic published an algorithm and guidelines outlining the approach to and management of a patient with clinical suspicion of device-related endocarditis [Sohail et al. 2007a]. At initial evaluation, blood cultures should be drawn [Sohail et al. 2007a]. In the scenario of positive cultures or negative cultures with recent antibiotic use, TEE rather than TTE should be used to evaluate possible device-related endocarditis because of the superiority of the specificity of TEE, 90% versus 30% respectively [Rallidis et al. 2003; Sohail et al. 2007a]. Following identification of device infection, complete removal should be performed [Sohail et al. 2007a]. The duration of antibiotics after device explantation is based on the presence of positive blood cultures and complications [Sohail et al. 2007a]. If blood cultures are persistently positive, antimicrobials should be used for 4 weeks, regardless of TEE results [Sohail et al. 2007a]. If complications arise, such as septic venous thrombosis or osteomyelitis, treatment should be extended [Sohail et al. 2007a]. In cases with both positive blood cultures and positive lead vegetation on TEE, reimplantation is recommended only after repeat blood cultures are negative for 72 h [Sohail et al. 2007a]. However, some reports have advocated delay of reimplantation up to 2 weeks after a pocket infection and 6 weeks after a deeper infection [Chua et al. 2000; Darouiche, 2004]. Still, the use of temporary transvenous cardiac devices has been shown to increase the risk of ensuing device-related endocarditis [Klug et al. 2003; Sohail et al. 2005], and therefore delay of reimplantation cannot be extended for too long.

As previously mentioned, less than 20 cases of Candida device-related endocarditis have been reported. In a documented literature review in 2011 of 15 of those cases, 8 of the 15 died while receiving treatment for the infection [Halawa et al. 2011]. Recommendations in agreement with the Infectious Diseases Society of America suggest the use of amphotericin B alone or in combination with 5-flucytosine or an echinocandin. Should the infection be susceptible, the patient become stable, or blood cultures become negative, the regimen can be switched to fluconazole. Treatment is recommended for 4 to 6 weeks after device removal or indefinitely should there be a contraindication to explantation [Pappas et al. 2009].

In review of the case reports, a high index of suspicion along with early diagnosis and intervention are necessary for patient survival. This case is unique in being the first woman with Candida device-related endocarditis, one of the largest vegetations at 4 cm, and one of the few that is ICD associated. This rare presentation of Candida device-related endocarditis was successfully treated with a combination of medical and surgical intervention. As Candida device-related endocarditis is so rare, with less than 20 cases being reported, and fatality occurring in half, clinical management can only be derived from sporadic case reports. Therefore, the course of this patient’s disease and care will be a useful adjunct to the current literature for determining treatment and prognosis in similar cases.