Share this article on:

Prediction of Survival in Patients With Cardiogenic Shock and Multiorgan Failure Treated With Biventricular Assist Device

Potapov, Evgenij V.*; Stepanenko, Alexander*; Kukucka, Marian*; Ba Fadhl, Faiz H.; Qedra, Naser; Weng, Yuguo; Pasic, Miralem; Lehmkuhl, Hans B.; Krabatsch, Thomas; Hetzer, Roland

doi: 10.1097/MAT.0b013e3181dbefd0
Adult Circulatory Support

Mechanical circulatory support (MCS) is valuable in saving the lives of patients with severe cardiogenic shock. However, their survival is limited if multiorgan failure (MOF) proves to be irreversible. Although ventricular assist devices (VADs) have been shown to reverse end-organ failure in some patients, the pathophysiological mechanisms of end-organ failure and its regression are not fully understood, and clinical markers and thresholds for the “point of no return” are lacking. We investigated predictors of 30-day survival in patients supported with a biventricular assist device (BVAD). We studied 157 patients implanted with a Berlin Heart EXCOR BVAD between 1987 and 2006. Children younger than 10 years and cases with postcardiotomy syndrome and transplant failure were excluded from the analysis as well as patients with technical or bleeding problems requiring rethoracotomy. In total, 69 clinical, hemodynamic, echocardiographic, and laboratory parameters were evaluated. Most of the patients suffered from ischemic cardiomyopathy or acute myocardial infarction. In addition, the preoperative multiple organ dysfunction syndrome (MODS) and the sequential organ failure assessment (SOFA) scores were calculated. The patients were divided into two groups regarding procedural success: group I—survival >30 days or heart transplantation or weaning from device (n = 105) and group II—death on system <30 days after surgery (n = 52). The 30-day procedural success rate was 67%. The patients in group I had higher systolic blood pressure (96.7 vs. 90.1 mm Hg, p = 0.027), lower serum creatinine (1.96 vs. 2.4 mg/dl, p = 0.001), and higher arterial pH (7.43 vs. 7.37, p = 0.02). The multivariate analysis recognized age, body temperature, systolic blood pressure, MODS score, and higher arterial pH as significant predictors for 30-day mortality. Standard markers for severity of cardiogenic shock and MOF do not predict survival on BVAD. As expected, older patients are at higher risk for death on BVAD. Acidosis and high MODS score predict unfavorable outcome. However, the prediction of clinical outcome in patients in severe cardiogenic shock supported by BVAD is possible in extreme situations only.

From the Department of Cardiothoracic and Vascular Surgery, Deutsches Herzzentrum Berlin, Berlin, Germany.

Submitted for consideration October 2009; accepted for publication in revised form January 2010.

Presented at the 55th Annual Conference of the ASAIO, May 2009, Dallas.

*Contributed equally to the work.

Reprint Requests: Evgenij V. Potapov, MD, Deutsches Herzzentrum Berlin, Augustenburger Platz 1, 13353 Berlin, Germany. Email:

Mechanical circulatory support (MCS) is a valuable tool in saving patients suffering from severe cardiogenic shock. However, the irreversibility of multiorgan failure (MOF) limits their survival. Although ventricular assist devices (VADs) have been shown to be able to reverse end-organ failure in some patients, the pathophysiological mechanisms of end-organ failure and its regression are not fully understood, and clinical markers and thresholds for the “point of no return” in patients suffering from cardiogenic shock are lacking. Optimal patient selection and timing of MCS implantation are crucial for procedural success. Most of the patients who die after surgery die from irreversible MOF despite re-established optimal body perfusion.

In a small number of previous studies, the prediction of outcome of patients in cardiogenic shock treated with a VAD or total artificial heart was investigated. These studies showed a number of clinical or laboratory parameters specific for adverse outcome,1–5 but some of them do not respect the time of occurrence of death—early death, which is related to preoperative condition to a greater degree, as opposed to late death, which is more related to infection and anticoagulation.6–8 We investigated predictors for 30-day mortality in a homogeneous group of patients suffering from cardiogenic shock and treated with a biventricular assist device (BVAD).

Back to Top | Article Outline


Between January 1987 and March 2006, 410 patients with cardiogenic shock were supported with a pneumatically driven BVAD (Berlin Heart EXCOR, Berlin Heart AG, Berlin, Germany) at our institution. All the patients evaluated for BVAD therapy were at high risk for imminent death from irreversible biventricular failure in the presence of cardiogenic shock. The device and implantation technique as well as postoperative anticoagulation have been described elsewhere.9

Cardiogenic shock was defined as cardiac index (CI) <2.2 L · min−1 · m−2 and pulmonary capillary wedge pressure (PCWP) >18 mm Hg with clinical signs of biventricular failure. Patients with implantation of a BVAD in the case of postcardiotomy heart failure, mechanical support for acute graft failure after heart transplantation, chest reopening because of bleeding and/or technical complications, and implantation of BVAD as redo surgery were excluded from the analysis to prevent additional intraoperative factors influencing the postoperative outcomes. Children younger than 10 years were excluded from the analysis. A total of 157 patients were included in the study. The series comprised 126 men and 31 women with median age of 47 (range 11–69) years. The average body surface area calculated according to the Mosteller formula [BSA (m2) = [height (cm) × weight (kg)/3600]1/2] was 1.92 ± 2.60 m2. Age groups were defined in 20-year ranges; the pH categories were defined as acidosis (pH <7.36), physiological pH (7.36–7.46), and alkalosis (pH >7.46); and the body temperature as hypothermic status (>36°C), normothermic (36–37°C), febrile (37–39°C), highly febrile (39–40°C), and hyperthermic (>40°C).

Most of the patients were suffering from chronic ischemic cardiomyopathy with acutely decompensated end-stage heart failure and a number of coronary interventions in the past (n = 80) or had acute decompensated dilative cardiomyopathy (n = 43). Other diagnoses were hypertrophic cardiomyopathy in 8 patients, acute myocardial infarction in 10, acute myocarditis in 5, and different pathologies (valvular cardiomyopathy, congenital heart disease, and metabolic storage disease) in 11 patients.

Back to Top | Article Outline

Data Collection

Sixty-nine clinical, hemodynamic, echocardiographic, and laboratory parameters were evaluated retrospectively for the purpose of this analysis within a 24-hour window before device implantation. The sequential organ failure assessment (SOFA) score and the multiple organ dysfunction syndrome (MODS) score were additionally calculated preoperatively. Briefly, the SOFA score is calculated as an assessment of the following parameters: PaO2/FiO2 ratio (Horowitz quotient10), serum creatinine, serum bilirubin, pulse-adjusted heart rate, thrombocyte count, and Glasgow coma scale.11 The MODS score is based on the ratio of the partial pressure of arterial oxygen (PaO2) to the fraction of inspired oxygen (FiO2), serum creatinine level, urine output, serum bilirubin level, mean arterial pressure, catecholamine level, thrombocyte count, and Glasgow coma scale.12 The inotropic score was calculated as previously described.13,14 Briefly, the doses of dopamine, dobutamine, and enoximone in micrograms per kilogram body weight per minute were added; the dose of milrinone was multiplied by 15 and doses of epinephrine and norepinephrine by 100 and then added.

Patient data were collected in our hospital electronic chart and MCS institutional database. This study was approved by our local medical ethics committee. To identify risk factors for 30-day survival, univariate analysis of preoperative variables was performed by comparing patients who died on the system within 30 days after surgery (nonsurvival group, n = 52) with those who survived for more than 30 days after the initiation of mechanical support or received heart transplantation or were successfully weaned from the device (survival group, n = 105).

Back to Top | Article Outline

Statistical Analysis

Statistical analysis was performed using SPSS Base 12.0 statistical software (SPSS Inc., Chicago, IL). Categorical variables were expressed as percentages and continuous variables were expressed as the mean ± sd. Categorical variables were compared using the χ2 test or Fisher's exact test. Comparison of continuous variables used the Mann-Whitney test for unpaired groups to avoid the assumption of normality. A two-tailed p value <0.05 was taken to indicate statistical significance. A logistic regression model was used to identify risk factors for 30-day mortality.

Back to Top | Article Outline


Overall 30-day survival after device implantation was 67%. The most frequent causes of death were MOF (n = 15), sepsis (n = 16), and profound vasoplegia (n = 12). In addition, 9 patients died for the following reasons: lethal intracranial bleeding (n = 3), lung (n = 2) and liver failure (n = 1), brain death (n = 1), and unreported causes (n = 2). There was no difference between the groups over time in the survival rate. The relationship between 30-day mortality and age is shown in Figure 1. The 90-day mortality in group I was 23%.

The following mortality rates were observed: of patients with myocardial infarction 50% died during the first 30 days of support; 40% of patients with dilative cardiomyopathy and 45% of the cohort with “other pathologies” died, as did 60% of patients with acute myocarditis and hypertrophic obstructive cardiomyopathy and 25% of ischemic cardiomyopathy patients.

Fifty-five percent of patients were mechanically ventilated preoperatively (group I 63% vs. group II 37%, p = 0.18). The median Horowitz quotient in group I was 238 (range 62–710) and thereby significantly higher than that in group II in which it was 145 (range 40–495) (p = 0.003). Patients with lower Horowitz quotient (PaO2/FiO2) had higher mortality. Patients' age at the time of device implantation was similar in both groups (42.4 ± 14.9 vs. 46.1 ± 13.8 years, p = 0.13).

Preoperative body temperature and body mass index (BMI) were generally higher in nonsurvivors, but the p value did not reach significance. Figure 2 shows 30-day mortality and body temperature. The mortality rate significantly increases with increasing body temperature. Arterial pH level recorded by the last available arterial blood gas analysis preoperatively was significantly lower in nonsurvivors (7.37 ± 0.011) than in survivors (7.43 ± 0.011), p = 0.002. The mortality rate significantly decreases with increasing pH (p < 0.05). The relationship between 30-day mortality and arterial pH level is shown in Figure 3.

Preoperatively, high SOFA and MODS scores were directly associated with increased mortality, with similar area under the receiver operating characteristic curve (0.63 ± 0.05 vs. 0.64 ± 0.05). None of the three patients with MODS score of more than 15 points survived 30 days after surgery (Figures 4 and 5). Preoperative systolic blood pressure and mean blood pressure were significantly lower in nonsurvivors than in survivors (96.7 ± 17.6 and 90.1 ± 17.7 mm Hg, p = 0.027; 67.7 ± 11 and 63 ± 11 mm Hg, p = 0.014, respectively), without differences in iv inotropic level between the two groups.

Selected clinically relevant preoperative demographic, hemodynamic, and laboratory parameters are presented in Table 1.

The multivariate analysis recognized younger age (p = 0.007), alkalosis (pH > 7.47), body temperature (p = 0.037), MODS score (p = 0.011), and systolic blood pressure (p = 0.006) as significant predictors for 30-day survival.

Back to Top | Article Outline


Mechanical circulatory support effectively reverses MOF and is one possibility to save patients' lives during MOF caused by cardiogenic shock. The study showed that employing standard parameters available in the clinical setting does not allow optimal prediction of clinical outcome in patients suffering from cardiogenic shock and treated by MCS. Only few parameters among the 69 investigated in this study independently predicted negative outcome: most clinically significant are advanced age and low pH. Furthermore, MODS score >15 units predicted 100% of deaths.

The question of the effectiveness of MCS and predictors of outcome has arisen because broad use has been made of MCS as a salvage tool in patients suffering from cardiogenic shock. There were mostly conflicting results, showing only few or no parameters to be predictive for outcome. Farrar2 identified elevated blood urea nitrogen and previous surgery as risk factors for survival until heart transplantation. This study was performed in the early era of MCS (>15 years ago) and included postcardiotomy implantations and patients treated with a left ventricular assist device (LVAD). Therefore, the study had two biases: in patients supported with an LVAD, right ventricular failure would impact the outcome, whereas postcardiotomy patients were mostly not in cardiogenic shock before surgery.

Our working group showed, in a study of 40 BVAD patients in 1992, that serum creatinine, liver enzymes, and pulmonary gas exchange could not be identified as predictive indicators of irreversible organ damage.15 In this study, the number of affected organs or systems was predictive for postimplant recovery. In a study performed in 199416 with a larger cohort of patients (including 9% of LVAD-supported patients), preoperative coagulation parameters (fibrinogen, antithrombin III, platelet count, and kinetics) correlated to postoperative blood loss and outcome. Patients who had mild or no coagulation disorders because of a shorter phase of low cardiac output before implantation of the assist device proved to gain faster restitution of organ function and most underwent transplantation.

In a small group of patients supported with a BVAD, Reinhartz et al.5 showed preoperatively elevated bilirubin to be a predictor for survival to discharge, whereas El-Banayosy et al.1 did not find any significant predictors of survival. In a recent study, Kirsch et al.4 investigated a fairly homogeneous group consisting of 71 patients treated with BVAD or LVAD for cardiogenic shock. The study showed the iv use of adrenalin only to be a sole and independent risk factor for death in the intensive care unit (ICU). All these studies have some limitations—the impact of device-related complications was not eliminated, and the studies included patients who underwent reoperation for bleeding, which may have significantly influenced the postoperative recovery from MOF and survival, and therefore the situation would not adequately mirror the depth of cardiogenic shock that may be reversed by MCS. Furthermore, patients implanted with an LVAD and suffering from right heart failure may remain in MOF because of low perfusion state.

To keep the study population more homogeneous, we included only patients supported by the Berlin Heart EXCOR BVAD. To exclude the impact of device-related complications on outcome, patients with postoperative re-exploration for device-related bleeding and patients who died from device-related complications (technical defects and air embolism) were also excluded from the analysis. In this homogeneous group, pH, as an integrative parameter of the compensatory process of MOF, was found to be an independent predictive factor. In our study, the arterial blood pressure was significantly lower in patients who died in the first 30 days after surgery in univariate and multivariate analysis. This finding is in agreement with the results recently published by Kirsch et al. and mirrors the more advanced stage of shock, with vasoplegia in patients who died in the first 30 days. In these patients, more pronounced and decompensated MOF caused severe acidosis with subsequent vasoplegia and necessity for adrenalin use.

The interference between the inflammatory response and hemodynamic deterioration in patients suffering from severe end-stage heart failure is still unclear.17 Systemic inflammatory response syndrome may contribute to worse outcome.4 It has also been shown that E-selectin, a marker for inflammatory status, predicts hemodynamic decompensation 2–3 days before it occurs.17 In our study, C-reactive protein was significantly higher in patients who died.

One of the aims of the study was to develop a score to optimize patient selection and timing of MCS implantation and to identify patients in the study population who would not profit from the reestablishment of optimal organ perfusion. This goal was not achieved based on the current level of knowledge; however, MODS score >15 points predicted unfavorable outcome in 100% of patients.

Optimal patient selection and timing of MCS implantation are closely related to ethical and financial issues. There are some open questions: based on which criteria can we predict the clinical course and when are we allowed to decide about restriction of access to MCS resources in severely ill patients with limited prognosis? In our opinion, we should offer all patients the chance at reasonable expense. The use of short-term devices such as extracorporeal membrane oxygenation (ECMO) or the new generation of centrifugal pumps as “bridge to bridge” or “bridge to decision” seems to be an optimal concept.8,18–23 The use of ECMO allows rapid initiation of support (e.g., femoro-femoral), even in a hospital without VAD experience, followed by transfer to a specialized institution for further treatment. Hoefer et al.24 found the bridge to bridge concept successful for selected patients with cardiogenic shock. During ECMO support, patients can be evaluated for comorbidities. The authors found that a combination of risk factors (status post cardiopulmonary resuscitation, elevated lactate levels, and impaired liver function) indicates unfavorable outcome and that, in these cases, VAD implantation should be considered very carefully. The magnetically levitated new centrifugal pump Levitronix (Waltham, MA) is effective in rescuing critically ill “moribund” patients and can provide an opportunity for low-cost support and optimization of their condition before deciding whether a more expensive device should be placed or whether transplantation should be undertaken. During support, better candidate selection for further procedures is then possible.21

In the beginning of cardiogenic shock, the left ventricle is most severely affected. In this situation, the new microaxial device Impella Recover (Abiomed, Danvers, MA), which is implanted through the femoral artery and placed across the aortic valve, may adequately support the hemodynamics for transport and until further decisions.25,26 The TandemHeart LVAD is another device for rapid percutaneous LVAD support.27 The other possibility for treatment in a specialized hospital is the implantation of long-term cannulas in connection with short-term pumps. In the case of organ recovery in patients suitable for long-term support, the pumps may be switched to long-term pumps (e.g., Berlin Heart EXCOR or Thoratec) in the ICU. This approach avoids the risks associated with repeat sternotomy and the use of cardiopulmonary bypass and decreases the total costs of patient care.18

Back to Top | Article Outline


This was a retrospective observational, single-center study and was affected by the limitations and biases inherent to such reports. Clinical data were obtained by chart review and therefore limited by questions of accuracy and completeness of the recorded data. Furthermore, a small number of patients supported by ECMO were not included in the study.

Back to Top | Article Outline


Mechanical circulatory support reverses MOF; however, the standard markers for severity of cardiogenic shock and grade of MOF do not predict survival on BVAD. Therefore, based on our experience, rapid re-establishment of adequate body perfusion as a means of gaining time, with subsequent employment of a decision-making algorithm, is necessary in patients suffering from severe cardiogenic shock and MOF before further progression of MOF. However, patients with advanced age and/or disastrous condition, as indicated by severe acidosis, have poor prognosis and should be carefully evaluated in terms of ethical and financial issues.

Back to Top | Article Outline


The authors thank Anne Gale, Editor in the Life Sciences, Deutsches Herzzentrum, Berlin, Germany, for editorial assistance, Julia Stein for statistical calculations, and the MCS team of DHZB, Dr. Ewald Hennig and Friedrich Kaufmann for technical support.

Back to Top | Article Outline


1. El-Banayosy A, Arusoglu L, Kizner L, et al: Predictors of survival in patients bridged to transplantation with the Thoratec VAD device: A single-center retrospective study on more than 100 patients. J Heart Lung Transplant 19: 964–968, 2000.
2. Farrar DJ: Preoperative predictors of survival in patients with Thoratec ventricular assist devices as a bridge to heart transplantation. Thoratec Ventricular Assist Device Principal Investigators. J Heart Lung Transplant 13: 93–100, 1994.
3. Farrar DJ: The thoratec ventricular assist device: A paracorporeal pump for treating acute and chronic heart failure. Semin Thorac Cardiovasc Surg 12: 243–250, 2000.
4. Kirsch M, Vermes E, Radu C, et al: Impact of preoperative hemodynamic support on early outcome in patients assisted with paracorporeal Thoratec® ventricular assist device. Eur J Cardiothorac Surg 34: 262–267, 2008.
5. Reinhartz O, Farrar DJ, Hershon JH, et al: Importance of preoperative liver function as a predictor of survival in patients supported with Thoratec® ventricular assist devices as a bridge to transplantation. J Thorac Cardiovasc Surg 116: 633–640, 1998.
6. Lahpor J, Khaghani A, Hetzer R, et al: European results with a continuous-flow ventricular assist device for advanced heart-failure patients. Eur J Cardiothorac Surg 37: 357–361, 2009.
7. Potapov EV, Jurmann MJ, Drews T, et al: Patients supported for over 4 years with left ventricular assist devices. Eur J Heart Fail 8: 756–759, 2006.
8. Potapov EV, Loforte A, Weng Y, et al: Experience with over 1000 implanted ventricular assist devices. J Card Surg 23: 185–194, 2008.
9. Drews T, Loebe M, Hennig E, et al: The ‘Berlin Heart’ assist device. Perfusion 15: 387–396, 2000.
10. Horovitz JH, Carrico CJ, Shires GT: Pulmonary response to major injury. Arch Surg 108: 349–355, 1974.
11. Vincent JL, Moreno R, Takala J, et al: The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med 22: 707–710, 1996.
12. Marshall JC, Cook DJ, Christou NV, et al: Multiple organ dysfunction score: A reliable descriptor of a complex clinical outcome. Crit Care Med 23: 1638–1652, 1995.
13. Kormos RL, Gasior TA, Kawai A, et al: Transplant candidate's clinical status rather than right ventricular function defines need for univentricular versus biventricular support. J Thorac Cardiovasc Surg 111: 773–782, 1996.
14. Potapov EV, Hennig F, Wagner FD, et al: Natriuretic peptides and E-selectin as predictors of acute deterioration in patients with inotrope-dependent heart failure. Eur J Cardiothorac Surg 27: 899–905, 2005.
15. Friedel N, Viazis P, Schiessler A, et al: Recovery of end-organ failure during mechanical circulatory support. Eur J Cardiothorac Surg 6: 519–522, 1992.
16. Schiessler A, Friedel N, Weng Y, et al: Mechanical circulatory support and heart transplantation. Pre-operative status and outcome. ASAIO J 40: 476–481, 1994.
17. Potapov EV, Wagner FD, Hennig FK, et al: Natriuretic peptides as predictors of clinical course in patients with end-stage heart failure. J Heart Lung Transplant 22: S206, 2003.
18. Potapov EV, Weng Y, Hausmann H, et al: New approach in treatment of acute cardiogenic shock requiring mechanical circulatory support. Ann Thorac Surg 76: 2112–2114, 2003.
19. Combes A, Leprince P, Luyt CE, et al: Outcomes and long-term quality-of-life of patients supported by extracorporeal membrane oxygenation for refractory cardiogenic shock. Crit Care Med 36: 1404–1411, 2008.
20. Idelchik GM, Simpson L, Civitello AB, et al: Use of the percutaneous left ventricular assist device in patients with severe refractory cardiogenic shock as a bridge to long-term left ventricular assist device implantation. J Heart Lung Transplant 27: 106–111, 2008.
21. De Robertis F, Rogers P, Amrani M, et al: Bridge to decision using the Levitronix CentriMag short-term ventricular assist device. J Heart Lung Transplant 27: 474–478, 2008.
22. John R, Liao K, Lietz K, et al: Experience with the Levitronix CentriMag circulatory support system as a bridge to decision in patients with refractory acute cardiogenic shock and multisystem organ failure. J Thorac Cardiovasc Surg 134: 351–358, 2007.
23. Shuhaiber JH, Jenkins D, Berman M, et al: The Papworth experience with the Levitronix CentriMag ventricular assist device. J Heart Lung Transplant 27: 158–164, 2008.
24. Hoefer D, Ruttmann E, Poelzl G, et al: Outcome evaluation of the bridge-to-bridge concept in patients with cardiogenic shock. Ann Thorac Surg 82: 28–33, 2006.
25. Jurmann MJ, Siniawski H, Erb M, et al: Initial experience with miniature axial flow ventricular assist devices for postcardiotomy heart failure. Ann Thorac Surg 77: 1642–1647, 2004.
26. Rossiter-Thornton M, Arun V, Forrest AP, et al: Left ventricular support with the Impella LP 5.0 for cardiogenic shock following cardiac surgery. Heart Lung Circ 17: 243–245, 2008.
27. Burkhoff D, Cohen H, Brunckhorst C, O'Neill WW: A randomized multicenter clinical study to evaluate the safety and efficacy of the TandemHeart percutaneous ventricular assist device versus conventional therapy with intraaortic balloon pumping for treatment of cardiogenic shock. Am Heart J 152: 469.e1–469.e8, 2006.
Copyright © 2010 by the American Society for Artificial Internal Organs