Late Recurrence of Prosthetic Valve Endocarditis Due to

Background

Fungal endocarditis due to Candida spp. is classically an uncommon syndrome, although as rates of fungemia increase [1], in conjunction with rising prevalence of cardiac implantable electronic devices, incidence is likely increasing as well. It is a particularly fatal form of infective endocarditis, with mortality rates at discharge measured at 30% in one large multicenter study, compared with 17% in nonfungal endocarditis [2]. Endocarditis is also the most fatal complication of candidemia [3]. Fungal endocarditis also has the highest risk of embolic complications in infective endocarditis, at about 86% in one large study of infective endocarditis, compared with 30% to 40% for bacterial causes. A major known risk factor for Candida endocarditis is intravenous drug abuse. Cardiac implantable electronic devices are also a major risk factor [4], likely due to a combination of poor blood flow to prosthetic devices and the ability of Candida spp. to form biofilms. The Infectious Disease Society of America [5] is clear on the need for lifelong suppressive antifungal therapy for patients with Candida endocarditis in whom cardiac surgery is contraindicated or when the infected prosthetic valve is still in vivo. This case report aims to help define the role of long-term suppressive antifungal therapy following surgery.

Case Report

We present a case of a 32-year-old woman with a history of intravenous drug abuse who presented to our hospital reporting fevers of approximately a 2-month duration, associated with loss of appetite, weight loss, and shortness of breath. She had a history of Staphylococcus aureus endocarditis diagnosed 3 years prior to presentation, for which she underwent mitral valve replacement. She discontinued intravenous drug abuse following the initial valve replacement but was read-mitted 4 months later with recurrent infective endocarditis, at which time cultures grew Candida spp., and she underwent a redo mitral valve replacement. She was treated with intravenous micafungin and was continued on long-term suppressive oral fluconazole, which she stopped for 1 month prior to presentation. She denied any intravenous drug use since her initial diagnosis of infective endocarditis. Detailed records from the previous episode were requested but were not available, and the species of Candida and duration of primary antifungal treatment was unknown. She had no history of congenital cardiac defects or childhood illnesses.

On admission, she was afebrile, with normal vital signs. Physical examination revealed a harsh holosystolic murmur, most prominent over the left sternal border. No track marks were present on comprehensive skin examination.

Initial laboratory studies showed a white blood cell count (WBC) of 14.7 K/mm3, with 74% neutrophils, a C-reactive protein (CRP) level of 4.0 mg/dL, and an erythrocyte sedimentation rate (ESR) of 42 mm/h. High-sensitivity troponin level and brain natriuretic peptide level were within the reference ranges. The urine drug screen was negative.

The electrocardiogram on admission showed a first-degree atrioventricular block. A transthoracic echocardiogram showed a vegetation on the anterior leaflet of a bioprosthetic mitral valve (Figure 1), a normal ejection fraction, mild concentric left ventricular hypertrophy, a borderline dilated right ventricle, and normal left atrial size. No valvular stenosis or regurgitation was seen. Chest X-ray showed no cardiomegaly or interstitial edema.

She was started on ceftriaxone, vancomycin, gentamicin, and micafungin. Transesophageal echocardiogram (TEE) showed a 1.8-cm vegetation on the anterior leaflet of the prosthetic mitral valve. A surgical evaluation was obtained, and she was recommended for a redo mitral valve replacement. Over the following days, she gradually improved clinically, and her WBC count, ESR, and CRP level trended down. Blood cultures grew C. albicans. Sensitivities were pending at this time. Repeat blood cultures 1.5 days following initial cultures and initiation of antifungal therapy were negative.

She was transferred to a higher center for possible redo mitral valve replacement, where repeat blood cultures on admission again grew C. albicans, sensitive to amphotericin B, micafungin, voriconazole, and fluconazole. She then reported left upper quadrant abdominal pain. Contrast-enhanced abdominal computed tomography revealed a splenic infarct of about 20% of splenic volume, which was presumed to be due to septic emboli. Her reports of unilateral foot pain, erythema, and swelling prompted magnetic resonance imaging of the foot, which showed subcutaneous edema suspicious for cellulitis, with a 10-mm hyperintensity suspicious for an abscess. Culture of the aspirate grew C. albicans. Repeat TEE redemonstrated mitral valve vegetations and revealed a 4-mm echodensity on the non-coronary cusp of the aortic valve (Figure 2). The patient was taken to the operating room. Pre-bypass TEE showed good ventricular function and identified the mitral prosthesis, without significant regurgitation but stenosis from heavy vegetation. There was a small vegetation on the aortic valve leaflets, but the leaflets did not look destroyed and had good competency. A third-time redo sternotomy was made, and she was placed on cardiopulmonary bypass. An aortotomy was made, the aortic valve was evaluated, and the vegetation on the aortic valve was removed, aspirated, and irrigated. The base was relatively clean. An aorto-mitral curtain vegetation was found and peeled successfully. Heavy scar removal was performed. A Sondergaard groove was evaluated and opened for the third time. Heavy vegetations were found on the mitral valve, which were removed and sent for culture. Copious irrigation and aspiration of the prosthetic ring was performed. An intraoperative Gram stain was performed prior to placing a mechanical valve, which was sutured and secured in position successfully. The aortotomy was closed, and the patient was rewarmed and weaned from bypass. Post-bypass TEE showed good ventricular function, a well-functioning mechanical valve with no evidence of residual vegetation, and complete function of the aortic valve. The patient tolerated the procedure well and was shifted to the Cardiovascular Intensive Care Unit in stable condition. Operative histopathology of the prosthetic mitral valve showed necrotic tissue and acute inflammation (Figure 3). Grocott methenamine silver staining showed fungal spores, compatible with Candida spp. (Figure 4). Operative cultures grew C. albicans. She was subsequently planned by the infectious disease specialist to continue micafungin intravenously for 8 weeks and was discharged to a rehabilitation facility.

Discussion

This case serves to explore the risk factors associated with invasive candidiasis, the clinical manifestations of such infection, and the possible role of chronic suppressive therapy. Candida spp. cause approximately half of fungal endocarditis cases, with half of those caused by C. albicans [5,6]. Reported risk factors include structural heart defects, antibiotic use, immunocompromised states, chronic liver disease, malignancies with or without neutropenia, and indwelling central venous catheters [7–9]. In our case, the patient had both a history intravenous drug abuse, although she denied active use, and a prosthetic mitral valve. Candida infective endocarditis is optimally treated with surgical and antifungal therapy, based on case series, observational studies, and clinical experience [4]. It is worth noting that successful treatment with medical therapy alone has been reported in the literature [9]; however, current standard of care includes surgical treatment, unless contraindicated [5]. Controlled studies are difficult to perform given the rarity of the condition. Acceptable antifungal regimens used for primary therapy include conventional or liposomal amphotericin B, with or without flucytosine or echinocandins [4,5]. Liposomal amphotericin B and echinocandins are preferred over conventional amphotericin B, owing to increased in vitro activity against biofilms and better penetration into vegetations [10], as well as a superior adverse effect profile [11,12]. Primary azole monotherapy is not preferred owing to high rates of recurrence and mortality, although it is the most common choice for chronic suppression [13,14].

Other known complications of candidemia include endophthalmitis [15], osteoarticular infections [16], central nervous system infection [17], chronic hepatosplenic candidiasis [18], intracardiac [19] and distant abscesses, distant embolization, including cerebrovascular infarcts [17], solid organ infarcts, and lung infarcts in right-sided disease [20]. In our case, the patient was found to have a splenic infarct and a bone abscess. Candida vegetations are known to be especially prone to embolization [21,22].

As illustrated in the present case, Candida prosthetic valve endocarditis is difficult to eradicate over the long term. Several reasons were considered as an explanation for her recurrence. She had an interruption in her suppressive fluconazole just before symptoms started. The interruption was nearly 2 years following primary surgical treatment that included valve replacement. Although a thorough debridement was done prior to valve implantation, it is difficult to exclude with total confidence that a small focus of infection was not missed. She also had a bio-prosthetic valve, which can be associated with a higher risk for endocarditis, and especially Candida. One prospective study noted a higher rate of recurrent endocarditis at 5 years with bioprosthetic valves, although it was not statistically significant [23]. A 2022 study of the Swedish Registry on Infective Endocarditis enrolling 780 prosthetic valve endocarditis cases showed 11 cases of Candida prosthetic valve endocarditis with tissue valves, compared with 1 with mechanical valves, although no P value was reported [24]. It must be noted that it is possible that our patient had resumed intravenous drug abuse in the interim, although she vehemently denied it, the family corroborated, and no physical, laboratory, or imaging signs were suggestive. In addition to surgical valve replacement, the Infectious Disease Society of America recommends long-term suppressive antifungal therapy following primary treatment of Candida prosthetic valve endocarditis [5]. It is often continued lifelong when surgical treatment fails, is suboptimal, or is contraindicated. Even when surgical treatment is successful, some clinicians advocate lifelong suppression to prevent late recurrence [25,26].

Conclusions

The rarity of fungal endocarditis due to Candida spp. makes it difficult to identify the exact risk factors associated with late recurrence. This report has highlighted the association between endocarditis and intravenous drug abuse, including prosthetic valve endocarditis and opportunistic fungal infections such as C. albicans. Lifelong oral fluconazole can be considered for all patients with C. albicans prosthetic valve endocarditis, especially in the setting of the presence of other risk factors such as intravenous drug abuse, as demonstrated in our case. Further studies are needed to determine differences in outcomes.