Hemoadsorption Treatment of Patients with Acute Infective Endocarditis during Surgery with Cardiopulmonary Bypass - A Case Series

Introduction

Infective endocarditis (IE) is a serious heart disease caused by microorganisms that enter the bloodstream and settle on the endocardium, heart valves or intracardiac devices. The cardiac effects of IE may include severe valve dysfunction and myocardial abscesses, which may finally lead to severe congestive heart failure. Therefore, depending on the causative microbes (e.g., staphylococci, enterococci, streptococci) and the clinical symptoms, valve replacement may be indicated for these patients (1, 2). Beside the described intracardiac effects, IE patients are at high risk for developing systemic inflammatory response and septic shock as a result of the bacterial spreadout from valve vegetations. Therefore, a surgical procedure (most often replacement of the affected valve) together with cardiopulmonary bypass (CPB) in a patient with an underlying IE disease represents an intervention with increased risks. The combination of surgical trauma, bacterial spreadout and artificial CPB surfaces results in a release of key inflammatory mediators such as IL-6 and IL-8. This may finally lead to an overshooting systemic hyperinflammatory state, frequently resulting in hemodynamic instability that in turn may induce organ dysfunction such as respiratory failure, acute kidney injury, intestinal ischemia and/or cognitive dysfunction (3). Of note, in case of prolonged CPB surgery, the risk of developing severe systemic inflammatory response syndrome (SIRS) postoperatively may increase even more. Postoperative therapeutic management of these patients includes an appropriate anti-infective therapy in combination with therapeutic approaches maintaining vital organ function.

Since inflammatory mediators are key triggers of inflammation and post-CPB SIRS, intra- or postoperative removal of such mediators from blood using blood purification with a cytokine adsorber has previously been described as an useful approach to control these hyperinflammatory processes, to restore immune homeostasis and potentially to prevent post-CPB SIRS and multiple organ dysfunction syndrome (MODS) (4, 5). Currently, the device most used as an adjunctive treatment to standard therapy in subjects suffering from SIRS, severe sepsis or septic shock to support the removal of cytokines as well as other inflammatory mediators via direct whole blood hemoadsorption is CytoSorb.

Cytosorb (CytoSorbents Corporation) is a polymer bead-based cytokine hemoadsorption cartridge approved in Europe since 2011 that can be used in combination with conventional hemodialysis machines or with CPB systems. In general, with more than 17,000 single treatments performed worldwide to date, CytoSorb application can be considered a safe and biocompatible therapeutic intervention. In this paper we describe the intraoperative application of CytoSorb hemoadsorption in 39 patients during CPB surgery due to IE.

Patients and methods

This case series was conducted in the 12-bed adult cardiothoracic surgery Intensive Care Unit (ICU) at the University Hospital Ulm, Germany. Informed consent for retrospective data evaluation was obtained from all patients or their relatives. From September 2013 until August 2016 we treated and monitored 39 consecutive patients undergoing cardiac surgery with CPB due to acute infective endocarditis. Patient characteristics, diagnoses and individual surgical procedure details are outlined in Table I. Briefly, all patients underwent urgent or emergency cardiac surgery procedures with CPB application (Tab. I). A CytoSorb adsorber cartridge was integrated in a parallel circuit in post-hemofilter position within the extracorporeal CPB circuit. Anticoagulation was achieved using heparin as standard anticoagulant according to routine procedure. Blood flow rates through the adsorber were kept between 200 and 400 mL/min and patients consistently received only CytoSorb treatment during surgery for the entire CPB time and without exchange of the adsorber. Treatment durations are depicted in Table I. Hemodynamic management with catecholamines (i.e., epinephrine, norepinephrine) and volume therapy was performed according to the standard of care protocol. To assess the therapeutic impact of the hemoadsorption treatment we measured laboratory parameters of inflammation (i.e., IL-6 and IL-8) hemodynamics (vasopressor dose, MAP), metabolic variables (lactate, base excess) as well as the extent of postoperative organ support (mechanical ventilation, ECMO, CRRT). Furthermore, we evaluated severity of illness in all patients preoperatively using the European System for Cardiac Operative Risk Evaluation (EuroSCORE II) (6) and additionally assessed the postoperative and 24-hour postoperative Acute Physiology and Chronic Health Evaluation (APACHE II score). ICU length of stay as well as ICU and hospital survival were obtained as outcome parameters.

OPEN IN VIEWER

Table I

Patient Characteristics, Surgery Details, Treatment Modalities and Patient Outcome (Hemoadsorption Group)

Case	Age	Gender	BMI	Diagnosis	Microbiological findings	Operation procedure	Emergency	Euro SCORE II	CPB time (min)	X clamp time (min)	CytoSorb treatment time (min)	APACHE II postop	APACHE II 24h postop	Mechanical ventilation (days)	ECMO (days)	CRRT (days)	Hydrocortisone	ICU LOS (d)	ICU survival	Hospital survival
1	43	M	20.7	AV Endocarditis	Staph. aureus	Re-Re ARR, Mechanoconduit	No	18.96	253	151	253			1		2	No	12	Yes	Yes
2	73	F	24.8	AV Endocarditis	Streptococcus mitis/cristatis	Re-AVR	No	31.14	114	69	115	33	28	1			No	7	Yes	Yes
3	75	F	20.8	AV Endocarditis		Bio-Conduit ARR, CABG	Yes	96.73	445	250	445	46		1	1	No	1	No	No
4	67	M	21.3	AV Endocarditis	Abiotropha defectiva	AVR, abscess removal, CABG-RCA	Yes	12.75	138	100	138	26	14	1			No	7	Yes	Yes
5	75	M	28.7	AV Endocarditis	Staph.aureus	AVR	Yes	4.96	71	49	72			1			No	9	Yes	Yes
6	37	M	36.0	AV Endocarditis	Staph. aureus, Stertococcus pyogenes	AVR	Yes	16.68	112	78	112	24	24	12			Yes	19	Yes	Yes
7	52	M	26.5	MV + AV Endocarditis	Streptococcus mitis	MVR, AVR, ARRec, SCAE	No	9.01	200	145	200			1			Yes	11	Yes	Yes
8	69	F	49.5	MV Endocarditis	Staph. aureus	MVR	Yes	64.18	115	70	115	32	26				No	32	No	No
9	62	M	32.3	MV Endocarditis	Streptococcus viridans	MVR	Yes	48.24	171	107	142	31	21	2		3	Yes	6	Yes	Yes
10	75	M	25.1	MV + AV Endocarditis	Streptococcus mitis	AVR, MVR	Yes	5.22	138	104	138	24	12	1			No	3	Yes	Yes
11	75	F	22.8	AV Endocarditis		AVR	Yes	9.62	88	60	88	30	15	1			No	4	Yes	Yes
12	64	M	26.9	MV Endocarditis	Streptococcus agalactiae	MVR	Yes	2.21	117	90	116	28	20	2			No	4	Yes	Yes
13	33	M	25.7	TK Endocarditis	Staph. aureus, fungi	MICTKR	No	4.43	101	61	102	31	17	1			No	5	Yes	Yes
14	38	M	19.8	MV Endocarditis	Streptococcus agalactiae	MIC MVR	Yes	6.2	109	75	91	33	20	1			No	4	Yes	Yes
15	77	F	30.1	MV Endocarditis	Streptococcus bovis	MVR	Yes	53.17	122	89	122	37	33	4		3	Yes	4	No	No
16	60	M	30.0	AV Endocarditis		AVR, MKR, TKR, RFA	Yes	12.83	204	145	205	32	31	2		4	Yes	6	Yes	Yes
17	58	M	28.4	MV Endocarditis	Streptococcus dysgalactiae	MVR, 2x CABG	Yes	12.03	132	90	133	30	12	1			No	6	Yes	Yes
18	79	F	26.8	MV Endocarditis	Streptococcus mitis	MVR	No	3.99	66	45	64	28	18	3			No	4	Yes	Yes
19	62	F	27.7	MV Endocarditis	Staph. aureus	MVR	Yes	31.46	141	83	141	33	28	7		6	Yes	13	Yes	Yes
20	73	M	23.5	AV Endocarditis	Propionibacterium	AVR	No	4.15	101	67	102	32	11	1			No	4	Yes	No
21	30	M	26.2	AV Endocarditis	Staph. aureus	Re-AVR, TKE, VA ECMO	Yes	31.69	280	129	282	30	30	2	2	1	Yes	2	No	No
22	76	F	29.3	MV Endocarditis	Enterococcus faecalis	MVR	No	74.69	159	69	160	33	30	8	4	1	Yes	8	No	No
23	57	M	27.1	MV Endocarditis	Streptococcus agalactiae	MVR	Yes	33.1	128	95	128		27	2		3	Yes	11	Yes	Yes
24	51	F	23.4	AV Endocarditis	Streptococcus mitis	AVR, SCAE	Yes	4.49	224	161	224	25	14	1			No	3	Yes	Yes
25	56	M	27.0	AV Endocarditis	Staphylococcus aureus	bio Bentall, CABG, SCAE	No	9.62	340	257	340	22	10	8			No	8	Yes	Yes
26	37	M	25.0	AV Endocarditis	Streptokokken	Bentall OP, SCAE, abscess removal	No	9.01	180	132	180			2			No	5	Yes	Yes
27	59	M	24.6	AV Endocarditis	Staph aureus	Re-AVR, RFA, LAA closure	No	60.49	128	74	128	30	23	3		2	Yes	3	Yes	No
28	48	M	23.4	AV Endocarditis	Enterococcus faecalis	mAVR, mMVR, SCAE	Yes	11	148	126	148	30	16	1			No	4	Yes	Yes
29	67	F	34.5	MV + AV Endocarditis		RE AVR, MVR, ARR	No	50.07	302	210	302	36	33	2	2	2	Yes	2	No	No
30	60	F	44.5	AV Endocarditis	Staph. aureus	Re-AVR	No	22.23	125	84	125	32	13	3			No	6	Yes	Yes
31	77	M	23.7	AV Endocarditis	Enterococcus faecalis	AVR, TKR, RFA, VA ECMO	Yes	78.23	208	114	208	39	36	7	2	7	Yes	7	No	No
32	51	M	30.5	AV Endocarditis	Streptococcus sanguinis	AVR, SCAE	Yes	16.17	117	94	116	32	29	1			No	5	Yes	Yes
33	46	M	29.3	AV Endocarditis	Staph. aureus	AVR	No	62.96	157	87	157			5	5	5	No	5	No	No
34	56	M	22.6	AV Endocarditis	Streptococcus agalactiae	AVR	No	2.15	90	61	90	26	10	1			No	3	Yes	Yes
35	61	M	26.6	AV prothesis Endocarditis		Re-AVR	No	5.2	134	96	134	19	5	1			No	4	Yes	Yes
36	68	M	39.2	AV Endocarditis		AVR	Yes	5.67	82	54	82			3		3	No	4	Yes	Yes
37	70	M	33.8	AV Endocarditis	Streptococcus mitis	AVR	Yes	28.3	110	73	110			1		3	No	10	Yes	Yes
38	61	M	16.9	MV Endocarditis	No microbiologiacal finding	MVR, ACVB	Yes	4.99	104	66	105			2			No	2	Yes	Yes
39	75	F	39.7	MV Endocarditis	Staphylococcus aureus	MVR	No	55.76	131	87	132	31	19	3		3	No	11	Yes	Yes

ARR = aortic root replacement; AVR = aortic valve replacement; MVR = mitral valve replacement; ARRec = aortic root reconstruction; TKR = tricuspid valve replacement.

In addition we retrieved clinical parameters and outcome data from a historical control group (from the years 2013 and 2014) of patients with infective endocarditis who had surgery with CPB but without CytoSorb hemoadsorption intraoperatively. However, in this group perioperative cytokine levels were not routinely measured and are therefore not available. The data of this comparative historical group are given in Table II.

OPEN IN VIEWER

Table II

Patient characteristics, surgery details, treatment modalities and patient outcome (comparative historical group)

Case	Age	Gender	BMI	Diagnosis	Microbiological findings	Operation procedure	Emergency	EuroSCORE II	CPB time (min)	X clamp time (min)	Mechanical ventilation (days)	ECMO (days)	CRRT (days)	Hydrocortisone	ICU LOS (d)	ICU survival	Hospital survival
Con01	57	M	31	MV endocarditis	Staph. Aureus	MVR	Yes	3.3	100	59	3			No	21	Yes	Yes
Con02	72	M	26	AV endocarditis	Granulicatella adiacens	aortic root conduit repair	Yes	26.0	202	134	26	7	30	Yes	96	No	No
Con03	79	M	27	AV endocarditis	Enterococcus faecalis	AVR	No	23.0	190	86	3		10	No	9	Yes	Yes
Con04	65	F	24	MV endocarditis	Streptococcus pneumoniae	MVR	Yes	3.4	232	75	7			No	6	Yes	Yes
Con05	75	M	23	MV endocarditis	Streptococcus mitis	MVR, TV repair	No	14.9	127	84	7			Yes	9	Yes	Yes
Con06	64	M	30	AV prothesis endocarditis	none	AVR	Yes	25.9	138	96	1		4	No	10	Yes	Yes
Con07	62	M	28	AV endocarditis	Streptococcus gallolyticus	AVR	No	5.3	92	63	1			No	6	Yes	Yes
Con08	59	M	23	MV endocarditis	Staphylococcus aureus	MV repair	Yes	2.4	110	67	1			No	5	Yes	Yes
Con09	56	M	26	AV and MV endocarditis	Streptococcus mitis	AV and MVR	No	2.2	142	105	1			No	4	Yes	Yes
Con10	76	M	31	AV prothesis endocarditis	Enterococcus faecalis	AVR	No	8.6	115	86	6		2	No	14	Yes	Yes
Con11	72	F	33	MV prothesis endocarditis, AV endocarditis	Corynebacterium species	MVR, AVR	Yes	33.9	231	167	8	8	8	Yes	8	No	No
Con12	85	F	35	AV prothesis endocarditis	Aerococcus urinae	AVR	No	16.1	143	103	1		4	No	7	Yes	Yes
Con13	80	M	23	MV endocarditis	Citrobacter koseri	MVR, AVR	No	7.5	145	106	4			Yes	14	Yes	Yes
Con14	79	F	28	MV endocarditis	Staphylococcus aureus	MVR	No	24.6	164	106	2		7	Yes	8	Yes	Yes
Con15	72	F	28	AV endocarditis	Escherichia coli	AVR	No	4.0	39	27	1			No	9	Yes	Yes
Con16	77	M	23	AV endocarditis	Staphylococcus haemolyticus	AVR	No	10.9	116	81	1			No	3	Yes	Yes
Con17	88	M	27	AV endocarditis	Citrobacter koseri	AVR	No	12.0	105	66	1			No	5	Yes	Yes
Con18	82	F	21	MV endocarditis	Escherichia coli	MVR	No	15.6	125	90	2		5	Yes	12	Yes	Yes
Con19	75	F	21	MV endocarditis	Steptococcus mutans	MVR	No	11.8	120	80	1		3	No	4	Yes	Yes
Con20	51	M	29	MV endocarditis	Staphylococcus lugdunensis, Enterococcus faecalis	MVR	No	3.3	108	72	2			No	6	Yes	Yes
Con21	37	M	26	TV endocarditis	Staphylococcus aureus	TVR	No	1.3	78	54	5			No	10	Yes	Yes
Con22	63	M	30	MV and Av endocarditis	Enterococcus faecalis	MV and AVR	No	9.5	138	108	2			Yes	2	Yes	Yes
Con23	41	M	20	MV endocarditis	Streptococcus mitis	MVR	No	1.7	179	146	1			No	4	Yes	Yes
Con24	76	F	32	MV endocarditis	Streptococcus agalctiae	MVR	No	14.2	93	66	1			No	7	Yes	Yes
Con25	82	M	21	AV prothesis endocarditis	Enterococcus faecalis	AVR	No	13.2	182	82	2			No	10	Yes	Yes
Con26	38	M	26	AV endocarditis	none	AVR	No	2.2	73	51	1			No	3	Yes	Yes
Con27	61	M	25	AV prothesis endocarditis	Corynebacterium species	AVR	No	6.9	96	53	3			No	4	Yes	Yes
Con28	36	M	28	TV endocarditis	Staphylococcus aureus	TVR	Yes	19.2	122	58	41		46	Yes	57	Yes	Yes

ARR = aortic root replacement; AVR = aortic valve replacement; MVR = mitral valve replacement; ARRec = aortic root reconstruction; TKR = tricuspid valve replacement.

Of note, all sets of data were statistically analyzed and graphically presented by means of the GraphPad Prism 7.01 software showing the median and interquartile range.

Results

Preoperative EuroSCORE II values in the CytoSorb group were rather heterogeneous, ranging between 2.2 and 96.7 (median 11). In the CytoSorb group, the median EuroSCORE II values for ICU survivors (n = 31) and ICU non-survivors (n = 8) were 11 and 64, respectively. In the comparative historical control group, the median EuroSCORE II values for ICU survivors (n = 26) and ICU non-survivors (n = 2) were 9 and 30, respectively. CPB times and X clamp times are depicted in Table I. All patients in the CytoSorb group obtained CytoSorb treatments that ranged from 64 up to 445 minutes duration (median 132 minutes). All patients showed a marked intraoperative increase of inflammatory mediators IL-6 and IL-8 followed by peak levels measured directly after completion of the surgical procedure. This was followed by a clear decrease in levels of IL-6 and IL-8 on postoperative day 1 and a return to preoperative baseline levels on postoperative day 3 (Fig. 1). Metabolic variables (i.e., lactate and base excess) showed a comparable pattern with a most pronounced change postoperatively and a return to baseline levels on postoperative day 3 (Fig. 1). Corresponding courses of the same metabolic parameters of the comparative historical control group are depicted in Figure 1A. Moreover, we observed a stabilization of hemodynamic parameters, as demonstrated by a consistent and maintained increase in MAP postoperatively with a concomitant reduction of catecholamine need at the same time (epinephrine and norepinephrine) (Fig. 2). On postoperative day 3, 72% and 82% of the patients were free from norepinephrine and epinephrine support, respectively. On postoperative day 5, these percentages increased to 82% and 95% for norepinephrine and epinephrine, respectively (data not shown). Interestingly, 15 out of the 39 IE patients initially requiring vasopressor support in their postoperative phase did not require any further vasopressor support 18 hours post surgery (3 patients with high EuroSCORE II between 31–97; 2 patients with mid EuroSCORE II between 16–31; 10 patients with low EuroSCORE II between 0–16).

OPEN IN VIEWER

Fig. 1

(A) CytoSorb group: Levels of IL-6 and IL-8 as well as metabolic parameters (lactate and base excess [median with IQR]), throughout the observation period. Values were assessed prior to treatment (baseline), during surgery, immediately after as well as on days 1 and 3 post treatment during CPB. (B) Historical control group: Metabolic parameters (lactate and base excess [median with IQR]), throughout the observation period. Values were assessed prior to treatment (baseline), during surgery, immediately after as well as on day 1 and 3 post CPB.

OPEN IN VIEWER

Fig. 2

(A) CytoSorb group: Mean arterial pressure (MAP), catecholamine doses (norepinephrine and epinephrine) throughout the observation period (median with IQR). Values were assessed prior to treatment (baseline), at end of surgery, at 6, 12, 18 and 36 hours as well as on day 1, 2 and 3 postoperatively. Please note that data sets were not completed for every patient. (B) Historical control group: Mean arterial pressure (MAP), catecholamine doses (norepinephrine and epinephrine) throughout the observation period (median with IQR). Values were assessed prior to treatment (baseline), at end of surgery, at 6, 12, 18 and 36 hours as well as on day 1, 2 and 3 postoperatively.

The severity of illness in the short-term postoperative period using the APACHE II also showed a trend to improvement from a median of 31 directly post operation to a median of 20 on day 1 post-surgery (Tab. I). A total of 18 patients were able to be weaned from mechanical ventilation within 24 hours after surgery, whereas 21 patients had a prolonged ventilation ranging from 1 to 12 days. Extracorporeal membrane oxygenation (ECMO) was mandatory in 5 patients for up to 5 days. High grade AKI necessitating CRRT was applied in 16 patients for up to 4 days (Tab. I).

In the comparative historical control group, 12 patients were able to be weaned from mechanical ventilation within 24 hours after surgery, whereas 16 patients had a prolonged ventilation ranging from 2 to 41 days. Extracorporeal membrane oxygenation (ECMO) was mandatory in 2 patients for up to 8 days. High grade AKI necessitating CRRT was applied in 10 patients for up to 46 days (Tab. II).

Length of ICU stay in the CytoSorb group ranged between 1 and 32 days (median 5). Of the 39 patient treatments summarized in this case series, 8 patients died during their ICU stay (7 between ICU days 1 and 8, 1 patient on ICU day 32) and 2 patients later died during their hospitalization period (Tab. I). Of note, the death of these patients could not be attributed to any specific treatment. One patient died from mesenteric ischemia with no option for surgical treatment, 1 patient had therapy withdrawn in accordance with the patient's advance directive, 5 patients died of refractory multiple organ failure, and 1 patient with refractory cardiac failure.

The length of ICU stay in the comparative historical control group ranged between 2 and 96 days (median 7.5 days). Of the 28 patient evaluated as a historical control group, 2 patients died during ICU stay (days 8 and 96) (Tab. II)

Intraoperative hemoadsorption treatment appeared to be well-tolerated, without device-related adverse events during or after treatment. No technical problems with the implementation of CytoSorb as part of the CPB circuit were observed.

Discussion

This retrospective case series reports on the application of the hemoadsorption cartridge CytoSorb during the intraoperative treatment of 39 patients with IE undergoing cardiac surgery with CPB. The clinical and laboratory parameters measured along this case series revealed (i) a consistent balanced control of the inflammatory response postoperatively, shown by a marked reduction of IL-6 and IL-8 plasma levels, (ii) the rapid adjustment of metabolic processes indicated by a normalization of lactate and base excess back to preoperative baseline levels within 3 days and (iii) hemodynamic stability before, during, and after the operation accompanied by a rapid decrease in need for vasopressors.

As circulating proinflammatory mediators like IL-6 and IL-8 play a central role in the development of SIRS and sepsis, the application of hemoadsorption devices represents a potential interesting interventional tool to avoid detrimental cross-talk between mediators, DAMPS and PAMPS with the immune system. CytoSorb has been shown to effectively remove hydrophobic molecules from 5 kDa up to approximately 55 kDa such as cytokines, chemokines, myoglobin and various other substances (7, 8). Moreover, cytokine reduction by means of CytoSorb application was reported for critically ill and cardiac surgery patients, as supported by a number of published preclinical and clinical data (4, 9, 10). After an initial intraoperative increase of cytokine levels, the application of CytoSorb hemoadsorption in this set of patients was associated with a decrease in the postoperative course. Of note, we are aware of the fact that the standard treatment regimen of these critically ill patients including hydrocortisone and CRRT could also have resulted in an additional decrease of these inflammatory parameters. Despite the exclusively intraoperative use of the cytokine adsorber, a reduction of cytokine levels was observable on postoperative day 1, returning back to preoperative levels on day 3, an effect that could possibly have been supported by the CytoSorb treatment. This long-term effect of even 1 CytoSorb treatment might be explained by the fact that hemoadsorption using CytoSorb might function at the level of the circulating immune effector cells, resulting in decreased activation of NfκB (in neutrophils and Kupffer cells) by a diminished cytokine load in the circulation and a subsequent decrease in cytokine production (11).

In addition, removal of substances is concentration-dependent. While low cytokine plasma concentrations are not affected, high cytokine plasma levels are reduced effectively. Supporting this line of argumentation, a recent blinded, randomized controlled trial in cardiosurgical patients with CPB compared (i) the CytoSorb application during CPB with (ii) a control group without hemoadsorption during CPB (12). Therefore, for instance, the IL-6 level monitored in the plasma of CPB patients did not exceed 254 pg/mL throughout the measurement duration. It is important to note that these were patients (and procedures) with only moderately increased risk who did not suffer from IE at the time of CPB surgery. In contrast, IE patients in our study undergoing CPB had a much more pronounced IL-6 cytokine release level of up to 5,000 pg/mL post treatment. This relevant difference shows that cytokine adsorption might preferably be effective in patients who are in a state of hyperinflammation (e.g., in infective endocarditis). This notion should be even more underlined as Bernardi et al stated that the authors did not find any differences for IL-6 in patients with or without CytoSorb treatment during CPB (12).

The decrease in cytokine levels in our case series was paralleled by a stabilization of hemodynamic parameters, during, and after the operation, as demonstrated by reduced catecholamine support (epinephrine and norepinephrine) and an increase in MAP. This effect has been observed in pre-clinical studies as well as in case report and series (4, 5, 9, 13).

Of note, it should be considered that the historical control group in our study showed a markedly lower risk profile as compared to the CytoSorb group as evidenced by the EuroSCORE II. This important limitation of historical case control analyses needs to be taken into account when comparing outcome data of both groups.

An important point to justify such a preventive treatment approach is the proof of potential outcome benefits despite the additional costs associated with Cytosorb treatment. From our perspective, such treatment might result in a mitigated inflammatory response postoperatively and hence preserve organ function and result in faster recovery during the postoperative course. While systematic data on these cost/benefit questions are still lacking, there is preliminary evidence available on improved organ function after CytoSorb use. Next to descriptions of unexpectedly fast hemodynamic stabilization (5, 14, 15), there is also a recent report indicating (16) a protected vascular barrier function after CytoSorb treatment, which might play an important role in earlier recovery of organ function in systemic hyperinflammation.

Since the question of whether there is a reproducible positive benefit/cost ratio to generally justify preventive CytoSorb treatment in patients with infective endocarditis undergoing cardiac surgery cannot definitely be answered from our case series, it will have to be established in future prospective studies.

Conclusions

With these clinical data and outcomes from 39 patients suffering from IE and undergoing cardiac surgery with CPB in combination with a CytoSorb adsorption device we were able to confirm and extend the results published earlier (5). Treatment with the CytoSorb device was safe and well-tolerated with no device- related adverse events during or after the treatment sessions. Even though clinical experience from this case series looks interesting, it is hard to draw any definite conclusions from this uncontrolled, retrospective, observational trial as to whether the effects seen in these patients were a primary therapy effect of CytoSorb or the consequence of a combination of all conducted treatments. With the insight of this recent case series, randomized controlled trials are warranted to further stress the potential benefits of this new treatment option for IE patients receiving cardiac surgery with CPB.