Extracorporeal Blood Purification with CytoSorb^® Hemoadsorption in Valve Surgery for Infective Endocarditis: A Case Series

Introduction

Infective endocarditis (IE) remains a serious cardiovascular condition, with a prevalence of 3 to 10 cases per 100 000 population and mortality rates of 15–30% [1]. It involves microbial invasion of the endocardium, leading to cardiac damage and a systemic inflammatory response that can cause multi-organ dysfunction [2]. Diagnosis relies on clinical, microbiological, and imaging criteria, but remains challenging, often resulting in delays and increased morbidity [3,4]. Standard treatment includes antibiotics and, when necessary, surgery; however, persistent inflammation limits successful outcomes [5]. In recent years, growing interest has focused on novel adjunctive therapies aimed at managing the systemic inflammatory burden associated with infective endocarditis. Hemoadsorption, an emerging therapy that removes inflammatory mediators from the blood [6], has shown promise in improving hemodynamics and organ function in IE patients with sepsis or hyperinflammation [7–11]. This case series explores the use of the CytoSorb® device (CytoSorbents, Inc., Princeton, NJ, USA) as an adjunctive treatment in the management of IE.

Case Reports

CASE #1:

A 60-year-old woman was transferred from a peripheral hospital with suspected bacterial endocarditis of the bicuspid aortic valve. She was febrile with unknown etiology, accompanied by elevated pro-inflammatory mediators and a previous isolation of Streptococcus oralis from the hemoculture. She exhibited unstable hemodynamic condition (blood pressure 90/65mmHg, heart rate 110/min), fatigue, and shortness of breath that required inotropic support (Sequential Organ Failure Assessment [SOFA] score=5 [3,4]). Four weeks prior to admission, she patient had noticed a skin rash, with hematuria and weight loss. While treated at the peripheral hospital, she experienced a transient cerebral ischemic event with loss of consciousness and additional unclear abdominal pain. A CT scan of the abdomen was performed, and a parenchymal rupture of the spleen due to septic embolization and necrosis was diagnosed. She underwent an emergency splenectomy. After surgery, she required prolonged mechanical ventilatory support in the intensive care unit (ICU) due to respiratory insufficiency. Bilateral pleural effusion necessitated pleural puncture and fluid drainage bilaterally. Transthoracic echocardiography showed the preserved global systolic function of the left ventricle with an ejection fraction of 50%, while the bicuspid aortic valve had significant aortic regurgitation. A floating mass attached to the destroyed aortic valve was found, with a high risk of metastatic embolization, and an urgent surgical procedure was scheduled. The patient denied any allergies and had no additional risk factors for heart disease.

The heart was arrested and a standard on-pump procedure using cardiopulmonary bypass (CPB) was carried out (aortic cross-clamp [ACC] time=55 min, total CPB time=62 min, oxygenator Sorin Inspire 8ss). Intraoperatively, in a single session, a CytoSorb hemoadsorber was additionally installed into the CPB circuit to mitigate the anticipated hyperinflammatory response and to potentially reduce levels of circulating pathogen- and damage-associated molecular patterns (PAMPs and DAMPs). The aortic valve was bicuspid (Sievers type I L-R). Aortic cusps were altered in terms of the presence of active endocarditis lesions, and an older epithelized abscess cavity was detected at the junction of the non-coronary aortic velum. The aortic valve was replaced with a mechanical bi-leaflet prosthesis no. 23 with the closure of the abscess cavity.

Laboratory examinations showed a major improvement in inflammatory parameters, lactic acidosis, and other relevant variables in the 24-hour period after surgery (Table 1). Improvements in fluid balance, hemodynamic stability, and dependence on vasoactive support were also recorded (epinephrine: 0.6 vs 0.4 μg/kg/min; norepinephrine: 0.8 vs 0.5 μg/kg/min; dobutamine: 10 vs 4 μg/kg/min).

The postoperative course was complicated by Klebsiella pneumoniae found in her sputum on postoperative day 1, necessitating noninvasive mechanical respiratory support on a few occasions. She was discharged 17 days after surgery, asymptomatic and without long-term sequelae at 6-month follow-up.

CASE #2:

A 35-year-old man was transferred from a peripheral hospital as an emergency case with suspected bacterial endocarditis of the tricuspid valve. He had unstable vital signs (blood pressure 90/70mmHg, heart rate 95/min), fatigue, shortness of breath, and chest pain that had already been present for 2 weeks. In the same period, he experienced febricity, chills, shivering, and general weakness. His past medical history was positive for heroin abuse (ceased for 2 months), currently treated with methadone. He was also positive for hepatitis C, and previous hemocultures were positive for Staphylococcus aureus. Apart from this, he had no additional comorbidities or previous serious diseases or allergies. Documentation regarding his previous treatment, including a discharge letter, was missing.

Transthoracic echocardiography showed preserved global systolic function of the left ventricle with an ejection fraction of 62%, but with an indication of a damaged tricuspid valve. A floating mass (1.0×1.5 cm) was attached to the anterior cusp of the valve, resulting in significant tricuspid valve regurgitation.

After 2 weeks of preoperative support and treatment, he was scheduled for cardiac surgery. A CPB-assisted cardiosurgical procedure including integration of an additional hemoadsorption cartridge with CytoSorb was carried out (ACC time=30 min, total CPB time=37 min) using double venous cannulation. A right atriotomy was performed. The right atrium was normal in size, but a floating mass was present at the lateral wall of the right atrium. The tricuspid orifice was also widened. Erosion and perforation of the anterior cusp of the tricuspid valve were observed, as well as a vegetation on the damaged cusp, with a high probability of embolic complications. Direct suturing of the hole on the anterior cusp was performed, and the part of the cusp was reattached to the annulus of the valve. Reductive annuloplasty of the tricuspid valve was also performed. After surgery, his markers of inflammation and lactic acidosis substantially decreased in the early 24-hour postoperative period (Table 1), paralleled by a decrease in dependency on vasoactive substances (epinephrine: 1 vs 0.4 μg/kg/min; norepinephrine: 1 vs 0.5 μg/kg/min; dobutamine: 8 vs 2 μg/kg/min). We also noticed a substantial reduction in ventilator support compared to the average mechanical ventilation duration in all IE patients operated on in our institution (5.3 vs 8.7±2.1 h). The postoperative course was uneventful, and he was discharged 7 days after surgery.

CASE #3:

A 20-year-old man was transferred from a peripheral hospital as an emergency case with suspected bacterial endocarditis of the aortic valve. His vitals were stable (blood pressure 120/70mmHg, heart rate 100/min). The patient stated that he was febrile at up to 38°C for more than 1 month. Lately, within 5 days before admission, he had a feeling of a rapid heartbeat. He reported pain in the muscles of the abdomen and left side of the neck, and chest pain intensified when taking a deep breath. His past medical history was positive for congenital aortic stenosis, operated on 2 months after birth. Previous surgical procedures included commissurotomy of the coronary cusps and removal of the nodules from the non-coronary and left coronary cusps. Importantly, the results of the reconstruction were suboptimal, with a slightly elevated transvalvular pressure gradient of 30 mmHg and consecutive aortic valve regurgitation (grade II). He was regularly checked by echocardiography. However, 1 year ago, the gradient over the aorta had increased (AoPGmax=50 mmHg) along with an aggravation of aortic valve regurgitation (grade IV).

Transesophageal echocardiography (TEE) showed preserved global systolic function of the left ventricle, with an ejection fraction of 60%. The aortic valve was severely damaged, with a 5-mm-long tissue fiber floating in the left ventricle. Hemoculture was positive for Streptococcus viridans.

He was admitted as an emergency case and prepared for re-do cardiac surgery. The heart was arrested, and a standard on-pump procedure plus hemoadsorption was carried out (ACC time=66 min, total CPB time=84 min). The aortic valve was severely eroded, with the presence of active endocarditis lesions. The floating mass was evacuated, and the aortic valve was replaced with a mechanical bi-leaflet prosthesis no. 21. Again, laboratory markers of inflammation and lactate decreased substantially following surgery (Table 1). There was an improvement in fluid balance and hemodynamic stability (epinephrine: 0.6 vs 0.4 μg/kg/min; norepinephrine: 0.8 vs 0.5 μg/kg/min; dobutamine: 10 vs 4 μg/kg/min). Ventilatory support could also be reduced compared to the hospital average (6.1 vs 8.7±2.1 h). He was discharged 7 days after surgery in good medical condition.

CASE #4:

A 63-year-old man was transferred from a peripheral hospital as an emergency case with suspected bacterial endocarditis of the aortic valve. His vitals were unstable (blood pressure 120/50 mmHg, heart rate 120/min) with signs of hemodynamic decompensation. He reported having shortness of breath, fatigue, and night dyspnea. Soon thereafter, the patient demonstrated incipient cardiogenic shock (SOFA=3). A month previously, he had had the flu, which was accompanied by cough and expectoration, followed by increased body temperature for a few days. Further medical history included diabetes type II and heavy smoking. A wide-ranging cardiological panel investigation, including selective coronarography was negative for coronary artery disease, and he was discharged without the need for invasive treatment.

Transthoracic echocardiography performed in the peripheral hospital had already shown aortic valve damage, and infective endocarditis of the aortic valve was confirmed. A TEE showed impaired global systolic function of the left ventricle with an ejection fraction of less than 40%, complete destruction of the aortic valve, with a rupture of the non-coronary cusp. Large vegetations were found on the right and non-coronary cusp, resulting in severe aortic regurgitation (AR=IV degree). He was scheduled for an emergency on-pump cardiac surgery procedure (ACC time=61 min, total CPB time=71 min). A CytoSorb hemoadsorber was installed into the CPB circuit and ran for the entire phase of the cardiopulmonary bypass. His aortic valve was found to be severely destroyed, with the presence of active endocarditis lesions. Floating masses were evacuated, and the aortic valve was replaced with a mechanical bi-leaflet prosthesis no. 21.

Laboratory examinations showed an excellent improvement 24 hours after surgery, with decreasing levels of WBC, CRP, and lactate (Table 1). The fluid balance and hemodynamics improved in parallel with a decrease in vasopressors (epinephrine: 1.1 vs 0.5 μg/kg/min; norepinephrine: 1.0 vs 0.5 μg/kg/min) and dobutamine support (10 vs 5 μg/kg/min). Compared to the hospital average, a reduction in ventilator support was also recorded (8.1 vs 8.7±2.1h). He was discharged 7 days after surgery and showed no complications at the 6-month follow-up.

The change in median levels of C-reactive protein, procalcitonin, lactate, and bilirubin in all 4 cases revealed a statistically significant difference from baseline to 24 hours after surgery (Figure 1). In all cases, there was no need for re-operation within the following 48 hours, nor was there a need for postoperative support with extracorporeal membrane oxygenation or intra-aortic balloon pump.

Discussion

An inflammatory response during IE is usually associated with 2 different mechanisms of action: direct heart valve destruction and systemic sepsis-mediated organ damage. While the first is caused by direct bacterial destruction of the involved tissue, the second is caused by the presence of bacterial wall components (known as PAMPs) and elements of the immune response products (DAMPs) [12]. In some people, these particles activate an unbalanced and exaggerated systemic inflammatory response, leading to uncontrolled release of pro- and anti-inflammatory cytokines responsible for hemodynamic instability and multi-organ dysfunction. This phenomenon is known as a “cytokine storm,” which is associated with high morbidity and mortality [12–14].

Cardiac valvular surgery is routinely performed worldwide with cardiopulmonary bypass and extracorporeal circulation. This procedure is also associated with a wide-ranging inflammatory response, including activation of cytokines, coagulation, and the alternative pathway of complement activation [14,15]. There is no consensus about the best extracorporeal blood purification treatment for patients with IE. The association between improved hemodynamics, reduced vasoactive medication support, and better organ function and cytokine hemoadsorption with CytoSorb found in the literature, especially for cardiac surgery patients under high mortality risk [16], helped us to hypothesize that the application of this hemoadsorptive therapy could improve overall outcomes after surgical treatment of severe cases of IE.

The use of CytoSorb in the context of infective endocarditis aligns with the broader understanding of its application in sepsis and hyperinflammatory conditions. CytoSorb is a polystyrene-based hemoadsorption device capable of removing serum hydrophobic particles of up to 60 kDa in molecular weight, which is well-matched with the size of the particles released in a cytokine storm [6]. Studies have demonstrated the device’s capacity to adsorb a broad spectrum of cytokines, toxins, and other inflammatory mediators, including DAMP particles [17,18], leading to improvements in hemodynamic stability and organ function [19–22]. Relevant randomized controlled trials (RCTs) are scarce. Recent systematic reviews and meta-analyses suggest a clinical benefit, highlighting the need for well-designed prospective trials with appropriate selection criteria, and more information is needed on the timing and dosing of hemoadsorption therapy for IE [16,23,24].

This case series investigated the application of extracorporeal blood purification with the CytoSorb device in 4 high-risk patients with severe IE, proposing possible selection criteria for future RCTs: acute and/or complicated IE, persistent fever, and urgent surgery. The results demonstrated an impressive improvement in markers of inflammation and lactic acidosis within the first 24-hour postoperative period (Figures 1–3). This was accompanied by a notable reduction in dependency on vasoactive substances, a decrease in ventilator support, and an overall uneventful postoperative course, leading to planned discharges for all patients within 7–17 days after surgery. The reduction in markers of inflammation, lactic acidosis, and vasoactive substances observed in this case series is consistent with the reported effects of CytoSorb in mitigating the cytokine storm associated with severe infections.

The observations in this case series support the hypothesis that, beyond infection source control with surgery and antimicrobial therapy, an additional therapeutic option may be re-establishing an optimal immune response by removing excessive inflammatory mediators from the bloodstream. Such a strategy could be useful for preventing or controlling cytokine storms to decrease overall morbidity and mortality in cardiac surgery patients with IE.

Despite its seemingly valuable effect, CytoSorb clinical use in this setting is still being assessed [25–27]. Observational studies reported improved outcomes associated with intraoperative use of CytoSorb in patients with high mortality risk (EuroSCORE >8%) [8,11] and Staphylococcus aureus IE [7]. However, a post hoc analysis of the largest RCT on the topic failed to show benefit in an S. aureus subgroup of patients [25]. These conflicting results emphasize the need for proper patient selection, relying on a specific “hot endocarditis phenotype” with predominant vasodilation, shock, vasoplegia, multiple organ dysfunction, and prolonged ICU stay, as suggested by Marczin et al [28], which are in line with our proposal of “severe IE” criteria described above. Moreover, the benefit of hemoadsorption use in sepsis and other critical conditions was assessed in recent meta-analyses, again with conflicting conclusions – from advice against its use outside of clinical trials [29] to results of improved short-term mortality [30]. Schultz et al proposed a dose-dependent model of the “amount of blood purified” and suggested that hemoadsorption can improve survival, provided that the applied dose is high enough [20]. This can be translated to severe IE patients who may need the postoperative continuation of hemoadsorptive therapy, as the intraoperative “dose” may have been suboptimal. In summary, more clinical evidence is required to standardize hemoadsorption in specific IE patients. However, the potential benefit of its use in this population was recognized in the most recent 2024 guidelines on cardiopulmonary bypass in adult cardiac surgery from the European Association for Cardio-Thoracic Surgery (EACTS), the European Association of Cardiothoracic Anaesthesiology and Intensive Care (EACTAIC), and the European Board of Cardiovascular Perfusion (EBCP) [31].

While this case series provides valuable insights, limitations include its small sample size and the absence of a control group. Larger, randomized controlled trials with carefully defined inclusion criteria are essential to conclusively establish the efficacy of CytoSorb in IE patients.

Conclusions

The results of this case series suggest that extracorporeal blood purification with CytoSorb holds promise as adjunctive therapy in the surgical management of infective endocarditis. The observed improvements in inflammatory markers, hemodynamic stability, and postoperative course merit further investigation, encouraging future research to elucidate the specific mechanisms and optimal application of CytoSorb in this complex clinical scenario. Randomized controlled trials with appropriately defined inclusion criteria are needed to confirm the benefit of this treatment.