Identification of potentially modifiable risk factors during cardiac surgery is important for improving the perioperative course. One such risk factor is intravascular hemolysis, which can be induced by cardiopulmonary bypass (CPB) and is characterized by an acute rise of circulatory cell–free hemoglobin.1 Blood is exposed to nonphysiological high mechanical stress, and the artificial surface in extracorporeal circuits worsens hemolysis.2,3
Haptoglobin is a natural free hemoglobin scavenger that acts by forming haptoglobin–hemoglobin complexes.4 When the binding capacity of haptoglobin is exceeded, free hemoglobin potently induces oxidative stress.5 Free hemoglobin scavenges endothelium-derived nitric oxide, leading to vasoconstriction, decreased microcirculation and blood flow and end-organ injury.4,6–12
A novel hemoadsorption device (CytoSorb, CytoSorbents Europe GmbH, Berlin, Germany) has recently become available for the removal of many key cytokines that cannot be filtered out by current blood purification techniques. CytoSorb is used in cardiac surgery to control the hyperinflammatory systemic reaction caused by extracorporeal circulation.13,14 The company’s claim that the device can remove free hemoglobin has been supported by some reports15,16 but, to date, no randomized controlled trial has been performed.
We performed this post hoc analysis of a previously conducted study13 to investigate whether the hemoadsorption of free hemoglobin can be demonstrated when hemoadsorption is performed during CPB. In addition, we determined whether the natural scavenging system of free hemoglobin is affected by the device. We also investigated whether there is a difference in the occurrence of postoperative severe hemolysis as marker for concentration-dependent adsorption by the device.
Material and Methods
This post hoc analysis of a previously conducted randomized, blinded, controlled, single-center pilot trial13 was approved by the ethics committee of the Medical University of Vienna, Austria (EK Nr: 1095/2013; 14 May 2013). The original study was reported to the Austrian Federal Office for Safety in Health Care (INS-621000-0505) and registered at ClinicalTrials.Gov (NCT01879176) before recruitment started. Written informed consent to participate in the study and consent to publish was obtained from each patient.
Study Design and Patients
The original study13 was conducted between September 10, 2013, and May 6, 2015, at our department. The study enrolled 46 adults undergoing elective open heart surgery with an expected CPB duration of >120 min at the Department of Cardiac Surgery, Medical University of Vienna, Vienna, Austria. Patients with the following interventions or conditions were excluded: declined informed consent; transplant surgery; scheduled insertion of a cardiac assist device; pulmonary endarterectomy; emergency and urgent procedures; serum creatinine >2 mg/dL, C-reactive protein >2 mg/dL, or bilirubin >2 mg/dL; body mass index <18 kg/m2; pregnancy; history of stroke; receiving chemotherapy, antileukocyte drugs, tumor necrosis factor-α blockers, immunosuppressive drugs (e.g., tocilizumab); or diagnosed with any disease state that could produce leukopenia (e.g., acquired immune deficiency syndrome).
The 46 patients were randomized to two groups: 24 received the hemoadsorption device (intervention group) and 22 did not receive the hemoadsorption device (control group). A total of nine patients (five from the intervention group and four from the control group) were excluded from the original study after randomization: eight patients because the procedure was rescheduled to another day or time when the study team would not be available and one patient (intervention group) because of hemodynamic instability after skin incision, requiring cardiopulmonary resuscitation and emergency CPB without use of the adsorber. For this post hoc analysis, we additionally excluded two patients (intervention group) who received postoperative extracorporeal membrane oxygenation support. Supplemental Figure 1 (see Figure 1, Supplemental Digital Content, http://links.lww.com/ASAIO/A353) shows the patient selection process.
Eligible patients were recruited and randomly assigned into the intervention group or control group the day before surgery by one of the physicians involved in the study. The online Randomizer for Clinical Trials 1.7.0 (https://www.meduniwien.ac.at/randomizer) was used to perform block randomization, stratified by sex and procedure.
The primary outcome for this post hoc analysis was postoperative differences in serum free hemoglobin and haptoglobin levels between patients treated and not treated with the CytoSorb adsorber during CPB.
Secondary outcomes were differences between the groups in serum lactate dehydrogenase and total bilirubin on postoperative day 1. Additionally, we investigated the difference in the occurrence of postoperative severe hemolysis (defined by free hemoglobin >50 mg/dL17) to analyze a possible concentration-dependent effect of the hemoadsorption device.
Anesthesia and CPB circuit priming were performed according to institutional standards. Cardiopulmonary bypass was performed using nonpulsatile flow at 2.5 L/min/m2, a non–heparin-coated circuit, and a membrane oxygenator (Quadrox; Maquet, Hirrlingen, Germany, or Capiox; Terumo, Eschborn, Germany). Mild hypothermia (32–34°C) was induced at the discretion of the surgeon; Buckberg cardioplegic solution was used for myocardial preservation. Anesthesia was administered by experienced cardiac anesthesia fellows under the supervision of senior cardiac anesthesiologists. Blood transfusion was in accordance with Society of Thoracic Surgeons/Society of Cardiovascular Anesthesiologists (STS-SCA) transfusion guidelines.18,19 Residual pump blood in the CPB circuit reservoir was retransfused after cell salvage. Administration of coagulation factors was based predominantly on rotational thromboelastometry variables and the coagulation profile of the patient.
In the intervention group, the 300 ml CytoSorb adsorber was installed on the CPB machine. CytoSorb is a hemoadsorption device that uses highly porous, biocompatible, nonpolar polymer beads to adsorb mid-molecular weight hydrophobic molecules through size exclusion and nonspecific surface adsorption.13,15,20
Blood was pumped actively through the hemoadsorption cartridge via a side arm coming from the venous outflow tube and returned to the venous reservoir before the oxygenator. The flow through the cartridge was controlled by a roller pump to 200 ml/min to standardize flow conditions in all treated patients. Control group patients received the same intervention, except that the adsorber was not installed. The details of the randomization process and the surgical procedure have been reported previously.13
In this post hoc study, we analyzed prospectively collected and frozen stored plasma samples. These samples had been collected at the following time points: (A) before induction of anesthesia; (B) at the end of CPB; (C) 2 h after CPB; and (D) 24 h after CPB.
Blood samples were drawn in pyrogen-free vials and centrifuged at 3500 rpm (relative centrifugal force (RCF) 1972g) for 10 min; the separated plasma samples were stored at −80°C.
Haptoglobin, free hemoglobin, lactate dehydrogenase, and bilirubin levels in the frozen plasma samples were determined. To validate the results of our defrosted plasma samples, we examined how the test results for free hemoglobin correlated with the values in routinely taken blood test results from all study patients at comparable time points (A, C, D), which were saved electronically in the hospital medical documentation system.
Continuous data were summarized as means and standard deviation and compared between groups using Student’s t-test. Highly skewed distributions were summarized as medians and interquartile range (IQR), and differences between groups were assessed with the nonparametric Wilcoxon rank sum test. Categorical data were summarized as absolute frequencies and compared between groups using the chi-square test. To examine the correlation between the results from frozen plasma samples and the results from routinely taken blood tests, we calculated Pearson’s correlation coefficients. Statistical significance was at P ≤ 0.05. Statistical analysis was performed using R 3.4.1 (http://www.R-project.org/).
A total of 35 patients (71% male; mean age, 66.1 ± 12.6 years) were included in this post hoc analysis: 17 in the intervention group and 18 in the control group. The median transfusion of packed red blood cells was 0 units (IQR, 0–1.0 units) in the intervention group vs. 0.5 units (IQR, 0–1.0 units) in the control group (P = not significant, NS). The demographic and surgical characteristics have been reported previously13; additional information is presented in Supplemental Table 1 (see Table 1, Supplemental Digital Content, http://links.lww.com/ASAIO/A354).
Free hemoglobin levels peaked immediately after CPB in both groups to reach 39.45 mg/dL (IQR, 30.7–56.0 mg/dL) in the intervention group vs. 34.4 mg/dL (IQR, 20.4–54.7 mg/dL) in the control group and then decreased steadily. On postoperative day 1 (POD1) free hemoglobin levels were 3.63 mg/dL (IQR, 3.1–4.5 mg/dL) in the intervention group vs. 4.3 mg/dL (IQR, 3.0–5.7 mg/dL) in the control group. At both time points, the differences between the groups were not statistically significant.
Free hemoglobin levels in frozen plasma samples and routinely collected blood tests showed a very strong positive correlation (r = 0.97).
Haptoglobin levels decreased from the baseline level of 119.5 mg/dL (IQR, 64.1–140.2 mg/dL) in the intervention group vs. 87.2 mg/dL (IQR, 72.4–114.4 mg/dL) in the control group (P = NS) to a nadir 2 h post-CPB when they were 32.2 mg/dL (5.0–38.6 mg/dL) in the intervention group and 17.8 mg/dL (5.0–26.3 mg/dL) in the control group (P = NS). They then increased and on POD1 were 58.4 mg/dL (41.5–64.5 mg/dL) in the intervention group vs. 17.9 mg/dL (5.0–31.7 mg/dL) in the control group (P < 0.01). Post hoc analysis of haptoglobin level was possible in only 106/140 (76%) of stored plasma samples; analysis was not possible in 34 samples because of insufficient material. Figure 1 and Table 1 show the details.
Lactate dehydrogenase level was significantly different between the groups on POD1: 353.0 U/L (IQR, 324–384 U/L) in the intervention group vs. 432.0 U/L (IQR, 371–558 U/L) in the control group (P < 0.05). Total bilirubin level was comparable in the two groups on POD1: 0.89 mg/dL (IQR, 0.8–1.4 mg/dL) in the intervention group vs. 0.86 mg/dL (0.6–1.2 mg/dL) in the control group (P = NS). Figure 2 and Table 2 show the details.
There was no significant difference between the groups in the incidence of postoperative severe hemolysis: 5/17 (29%) patients in the intervention group vs. 5/18 (28%) patients in the control group (P = NS). The extent of severe hemolysis was also comparable in the two groups: free hemoglobin levels were 58.8 mg/dL (IQR, 56.5–59.8 mg/dL) in the intervention group vs. 60.9 mg/dL (58.5–106.1 mg/dL) in the control group (P = NS). The details are presented in Supplemental Table 1 (see Table 1, Supplemental Digital Content, http://links.lww.com/ASAIO/A354).
In this post hoc analysis, we tested primarily the possible effects of the CytoSorb hemoadsorption cartridge on hemolysis during CPB.
We did not find any differences in postoperative free hemoglobin elevation between patients who were treated with the hemoadsorption device and those who were not, although we did observe significant differences between the two groups in the levels of haptoglobin and lactate dehydrogenase on POD1.
Postoperative free hemoglobin levels showed an 11-fold increase from baseline values in the intervention group and a 15-fold increase in the control group but returned to preoperative levels on POD1 in both groups. We also found concordant declines in haptoglobin levels in the two groups 2 h post cardiopulmonary bypass: a fourfold decline in the intervention group versus a fivefold decline in the control group, but these recovered in the intervention group on POD1. The difference in haptoglobin on the first postoperative day may be a sign of overwhelming exhaustion by free hemoglobin in the control group; the difference in lactate dehydrogenase levels between the groups supports this hypothesis. However, we did not observe significant differences between the groups in the incidence of postoperative severe hemolysis.
Hemoglobin is a hydrophilic protein and is found in a dimer–tetramer equilibrium in humans.21 The CytoSorb hemadsorption cartridge supposedly adsorbs hydrophobic molecules with molecular weight up to 60 kDa. Thus, both in terms of solubility and mass, hemoglobin seems to be out of the adsorption range of CytoSorb.
Cardiac surgery is associated with the development of intravascular hemolysis, which is an additional risk factor for adverse outcomes.9 Hemolysis, with resulting release of free hemoglobin into the plasma, is caused by mechanical destruction of erythrocytes due to contact with the bypass circuit surface, high blood flow and pressure conditions, and cardiotomy suctioning.1,22,23 Previous publications have reported that CytoSorb can adsorb molecules such as myoglobin,24 and the similarity between the chemical structures of myoglobin and hemoglobin molecules suggests that hemoglobin adsorption cannot be ruled out.1
False-positive test results for free hemoglobin can be caused, for example, by hemolysis during blood sampling or interference by high total bilirubin levels, and so a positive free hemoglobin test alone does not establish the occurrence of hemolysis.25 We therefore examined how our test results correlated with the results of routinely analyzed samples and also considered haptoglobin levels.
Storage of red blood cells in standard conditions can affect the integrity of the erythrocyte membrane and therefore significantly affect hemolysis parameters.26 This effect was ruled out in the current study because there was no significant difference between the groups in the proportion of patients receiving transfusion of packed red blood cells (PRBC), although we found two patients (12%) within the intervention group receiving 6 and 7 PRBC units, respectively.
We also investigated lactate dehydrogenase and total bilirubin levels. Both reflect the degree of hemolysis in the systemic circulation and were used as additional markers for late cardiac surgery–associated hemolysis.25,27–29 Lactate dehydrogenase levels were significantly lower in the intervention group on POD1. Decrease in lactate dehydrogenase has been related to hemolysis in other studies.30 Cowger et al31 found that in patients with hemolysis, lactate dehydrogenase elevation can occur even in the absence of free hemoglobin elevation and that the cumulative incidence of hemolysis defined by lactate dehydrogenase was higher. However, severe hemolysis, indicated by lactate dehydrogenase >600 U/L, was seen in only one patient in our control group and none in the intervention group.
Hyperbilirubinemia after cardiac surgery has been reported to occur in 25% of patients.32 Various reasons have been proposed, including decreased hepatic capacity for bilirubin disposal and bile transport due to liver congestion, blood transfusion, and increased CPB time.32 Patients with preoperative elevated total bilirubin levels were excluded from this study, and there were no differences between the groups in the perioperative administration of transfusions or infusion. However, preoperative right ventricular function was not monitored, and so we cannot rule out liver congestion as a reason for hemolysis.
This study has certain limitations. First, this was a post hoc analysis and therefore was not designed to detect differences in free hemoglobin or haptoglobin levels. We were not able to analyze all samples, mainly due to the inadequacy of stored plasma. Also, for power of 0.8 for detection of difference in free hemoglobin levels, a sample size of at least 48 would have been necessary.
Second, the plasma preparation protocol of the original study was designed to identify differences in cytokines. Although we found a strong correlation (r = 0.97) between our test results and the results of routinely analyzed samples, we cannot rule out that some red blood cells remained in the supernatant after centrifugation and were lysed during defrosting and may thus have affected free hemoglobin values.
Third, there may be confounding by unobserved factors; for example, we did not monitor the differences in use of valve prostheses, which has been known to cause hematologic changes,28 or preoperative right ventricular function, which could cause liver congestion. Fourth, the original study on which this post hoc analysis was based was not a double-blind study (only patients were blinded). Blinding of the surgeon is rarely feasible in studies involving operative management and use of medical devices.
Finally, as reported before,13 we cannot be sure what proportion of blood was actually treated by the device; in our study,13 the CytoSorb treated about 3–4% of the total blood volume each minute, and we can only assume that >99% of the blood volume passed through the hemoadsorption device over the mean treatment time of 180 min.
In conclusion, we found no significant decrease or increase of free hemoglobin in patients treated with hemoadsorption. Therefore, no effect on hemolysis was found. Greater decrease in haptoglobin level on postoperative day 1 in patients not treated with hemadsorption may be an indication that the device has some free hemoglobin adsorbing effect. Similarly, relatively higher lactate dehydrogenase in control patients may indicate more hemolysis. We believe that this article can serve as a hypothesis-generating study. Studies with larger samples are needed to clarify the significance of the moderate differences in hemolytic effects observed in this study.
We thank all the medical staff of the Division of Cardiac Thoracic Vascular Anesthesia and Intensive Care Medicine and the Department for Cardiac Surgery for their contributions toward the conduct of this study. Special thanks go to our medical students Klaus Dragosits, David Hirschl, Christian Lamm, Falk Preißing, and Christoph Steinkellner for their invaluable help in data collection. The article was reviewed by a professional English-speaking editor.
1. Wetz AJ, Richardt EM, Schotola H, Bauer M, Bräuer A. Haptoglobin and free haemoglobin during cardiac surgery-is there a link to acute kidney injury? Anaesth Intensive Care 2017.45: 58–66.
2. Vieira Junior FU, Antunes N, Vieira RW, Alvares LM, Costa ET. Hemolysis
in extracorporeal circulation: relationship between time and procedures. Rev Bras Cir Cardiovasc 2012.27: 535–541.
3. Ding J, Niu S, Chen Z, Zhang T, Griffith BP, Wu ZJ. Shear-induced hemolysis
: species differences. Artif Organs 2015.39: 795–802.
4. Kato GJ. Haptoglobin halts hemoglobin’s havoc. J Clin Invest 2009.119: 2140–2142.
5. Rother RP, Bell L, Hillmen P, Gladwin MT. The clinical sequelae of intravascular hemolysis
and extracellular plasma hemoglobin: a novel mechanism of human disease. JAMA 2005.293: 1653–1662.
6. Gladwin MT, Crawford JH, Patel RP. The biochemistry of nitric oxide, nitrite, and hemoglobin: role in blood flow regulation. Free Radic Biol Med 2004.36: 707–717.
7. Caulfield JL, Singh SP, Wishnok JS, Deen WM, Tannenbaum SR. Bicarbonate inhibits N-nitrosation in oxygenated nitric oxide solutions. J Biol Chem 1996.271: 25859–25863.
8. Haase M, Haase-Fielitz A, Bellomo R. Cardiopulmonary bypass
, free iron, acute kidney injury and the impact of bicarbonate. Contrib Nephrol 2010.165: 28–32.
9. Kubota K, Egi M, Mizobuchi S. Haptoglobin administration in cardiovascular surgery patients: its association with the risk of postoperative acute kidney injury. Anesth Analg 2017.124: 1771–1776.
10. Shih HM, Chen YC, Pan CF, Lin HC, Wu CJ, Chen HH. Hemolysis
-induced acute kidney injury following cardiac surgery: a case report and review of the literature. Hemodial Int 2013.17: 101–106.
11. Vercaemst L. Hemolysis
in cardiac surgery patients undergoing cardiopulmonary bypass
: a review in search of a treatment algorithm. J Extra Corpor Technol 2008.40: 257–267.
12. Vermeulen Windsant IC, de Wit NC, Sertorio JT, et al. Hemolysis
during cardiac surgery is associated with increased intravascular nitric oxide consumption and perioperative kidney and intestinal tissue damage. Front Physiol 2014.5: 340.
13. Bernardi MH, Rinoesl H, Dragosits K, et al. Effect of hemoadsorption
during cardiopulmonary bypass
surgery - a blinded, randomized, controlled pilot study using a novel adsorbent. Crit Care 2016.20: 96.
14. Träger K, Skrabal C, Fischer G, et al. Hemoadsorption
treatment of patients with acute infective endocarditis during surgery with cardiopulmonary bypass
—a case series. Int J Artif Organs 2017.40: 240–249.
15. Datzmann T, Träger K. Extracorporeal membrane oxygenation and cytokine adsorption. J Thorac Dis 201810: S653–S660.
16. Cirstoveanu CG, Barascu I, Mc Kenzie Stancu S. Hemadsorption with Adult CytoSorb
® in a low weight pediatric case. Case Rep Crit Care 2017.2017: 6987167.
17. Omar HR, Mirsaeidi M, Socias S, et al. Plasma free hemoglobin is an independent predictor of mortality among patients on extracorporeal membrane oxygenation support. PLoS One 2015.10: e0124034.
18. Ferraris VA, Brown JR, et al.; Society of Thoracic Surgeons Blood Conservation Guideline Task Force; 2011 update to the society of thoracic surgeons and the society of cardiovascular anesthesiologists blood conservation clinical practice guidelines. Ann Thorac Surg 2011.91: 944–982.
19. Ferraris VA, Ferraris SP, et al.; Society of Thoracic Surgeons Blood Conservation Guideline Task Force; Perioperative blood transfusion and blood conservation in cardiac surgery: The Society of Thoracic Surgeons and the Society of Cardiovascular Anesthesiologists clinical practice guideline. Ann Thorac Surg 2007.83: S27–86.
20. Malard B, Lambert C, Kellum JA. In vitro
comparison of the adsorption of inflammatory mediators by blood purification devices. Intensive Care Med Exp 2018.6: 12.
21. Abraham B, Hicks W, Jia Y, Baek JH, Miller JL, Alayash AI. Isolated Hb Providence β82Asn and β82Asp fractions are more stable than native HbA(0) under oxidative stress conditions. Biochemistry 2011.50: 9752–9766.
22. Haase M, Haase-Fielitz A, Bagshaw SM, Ronco C, Bellomo R. Cardiopulmonary bypass
-associated acute kidney injury: a pigment nephropathy? Contrib Nephrol 2007.156: 340–353.
23. Loef BG, Epema AH, Navis G, Ebels T, van Oeveren W, Henning RH. Off-pump coronary revascularization attenuates transient renal damage compared with on-pump coronary revascularization. Chest 2002.121: 1190–1194.
24. Wiegele M, Krenn CG. Cytosorb
™ in a patient with Legionella pneumonia-associated rhabdomyolysis: a case report. ASAIO J 2015.61: e14–e16.
25. Heilmann C, Geisen U, Benk C, et al. Haemolysis in patients with ventricular assist devices: major differences between systems. Eur J Cardiothorac Surg 2009.36: 580–584.
26. Donadee C, Raat NJ, Kanias T, et al. Nitric oxide scavenging by red blood cell microparticles and cell-free hemoglobin as a mechanism for the red cell storage lesion. Circulation 2011.124: 465–476.
27. Ng RRG, Chew STH, Liu W, et al. The inflammatory response between miniaturised and conventional cardiopulmonary bypass
after cardiac surgery in an Asian population. Perfusion 2015.30: 487–494.
28. Jegger D, Horisberger J, Jachertz M, et al. A novel device for reducing hemolysis
provoked by cardiotomy suction during open heart cardiopulmonary bypass
surgery: A randomized prospective study. Artif Organs 2007.31: 23–30.
29. Monsefi N, Zierer A, Beiras-Fernandez A, et al. Hematologic effects of heart valve prostheses and vascular grafts after aortic root replacement. Thorac Cardiovasc Surg 2015.63: 452–458.
30. Gerrah R, Shargal Y, Elami A. Impaired oxygenation and increased hemolysis
after cardiopulmonary bypass
in patients with glucose-6-phosphate dehydrogenase deficiency. Ann Thorac Surg 2003.76: 523–527.
31. Cowger JA, Romano MA, Shah P, et al. Hemolysis
: A harbinger of adverse outcome after left ventricular assist device implant. J Heart Lung Transplant 2014.33: 35–43.
32. An Y, Xiao YB, Zhong QJ. Hyperbilirubinemia after extracorporeal circulation surgery: a recent and prospective study. World J Gastroenterol 2006.12: 6722–6726.