Author + information
- Received April 4, 2001
- Revision received August 21, 2001
- Accepted August 31, 2001
- Published online December 1, 2001.
- Ralf Sodian, MD*,* (, )
- Matthias Loebe, MD, PhD*,†,
- Christoph Schmitt, MD*,
- Evgenij V. Potapov, MD*,
- Henryk Siniawski, MD*,
- Johannes Müller, MD*,
- Harald Hausmann, MD*,
- Heinz R. Zurbruegg, MD*,
- Yuguo Weng, MD, PhD* and
- Roland Hetzer, MD, PhD*
- ↵*Reprint requests and correspondence: Dr. Ralf Sodian, Department of Thoracic and Cardiovascular Surgery, German Heart Institute Berlin, Augustenburger Platz 1, 13353 Berlin, Germany
We sought to investigate the relationship between the plasma concentration of brain natriuretic peptide (BNP), echocardiographic findings and the clinical outcome of patients supported with ventricular assist devices (VADs) to determine the role of BNP as a predictor for cardiac recovery.
Ventricular unloading in patients with end-stage heart failure supported by VADs may lead to myocardial recovery. The BNP is produced in the myocardium in response to chronic volume overload, but the effects on it of ventricular unloading by VADs are largely unknown.
Twenty-one patients diagnosed with nonischemic cardiomyopathy and supported by VADs were evaluated for echocardiographic data and blood chemistry including BNP. They were divided into patients who died while on mechanical support (group I; n = 9), patients who were transplanted (group II; n = 8) and patients who were successfully weaned off the system and did not require transplantation (group III; n = 4).
Brain natriuretic peptide plasma concentrations decreased significantly after initiation of mechanical circulatory support (p = 0.017). Furthermore, the changes in BNP plasma concentrations showed a faster decrease to normal levels within the first week after implantation of the VAD in patients who were weaned off the system (group III) compared to patients in group I and group II.
This study shows that ventricular unloading with VADs decreases BNP plasma concentrations in patients who suffer from end-stage heart failure. Furthermore, we hypothesize that an early decrease of BNP plasma concentration may be indicative of recovery of ventricular function during mechanical circulatory support.
Mechanical circulatory support systems have improved the clinical outcome of patients who suffer from end-stage heart failure (1,2). It has been widely confirmed that mechanical circulatory support is a suitable therapy as a bridge to transplantation for patients with cardiac arrest, cardiogenic shock or severe hemodynamic instability (3). The use of mechanical circulatory support systems enables the patient to recover from multiorgan failure. While using this bridging concept, numerous investigators have observed a recovery of cardiac function after prolonged ventricular unloading by mechanical circulatory assistance. Prolonged ventricular unloading appears to help in the restoration of myocardial function and thereby allows some hearts to subsequently function without mechanical circulatory assistance or the need for heart transplantation. However, the mechanisms of cardiac recovery and the patient characteristics are not fully known. Understanding the pathophysiologic mechanisms is of interest for the field and may lead to new therapeutic strategies and selection criteria for patients who suffer from end-stage heart failure (4–6).
Currently, echocardiography is the most widely used method for the detection of cardiac recovery in patients supported with mechanical circulatory support systems (7). In addition, numerous predictors of sustainable myocardial and functional recovery are currently under investigation.
Natriuretic peptides (brain natriuretic peptide [BNP], and atrial natriuretic peptide, [ANP]) are produced in the myocardium in response to chronic volume overload. Both BNP and ANP circulate in the plasma and defend the heart against volume overload by decreasing salt and water retention and inhibiting vasoconstrictor peptides (8). In patients who suffer from congestive heart failure, the concentrations of natriuretic peptides and especially the concentration of BNP, which is a ventricular hormone, correlate with the status of ventricular dysfunction, and high concentrations predict poor long-term survival (9,10).
In this study we investigated the relationship between plasma concentrations of natriuretic peptides (BNP and ANP), echocardiographic findings and clinical outcome so as to determine the potential role of BNP as a predictor for cardiac recovery after prolonged ventricular unloading in patients supported by mechanical circulatory support systems.
Study design and recruitment of patients
We studied 21 patients who presented at the Department of Cardiothoracic and Vascular Surgery, German Heart Institute Berlin, between October 1, 1997, and July 30, 1999, according to the guidelines of the Committee on Human Investigation at our institution. Patients suffered from end-stage nonischemic cardiomyopathy and were treated with mechanical circulatory support systems. Seven patients with end-stage nonischemic cardiomyopathy were supported with a Novacor left ventricular assist device (LVAD) (Baxter Healthcare, Oakland, California), 2 patients with an LVAD (Berlin Heart, Berlin, Germany) and 12 patients with a Berlin Heart biventricular assist device (BVAD). Initial screening criteria for the implantation of an LVAD included the following:
1. nonischemic cardiomyopathy (angiography);
2. refractory cardiogenic shock (cardiac index <2.0 l/min per m2with systolic blood pressure <95 mm Hg and pulmonary wedge pressure >20 mm Hg and dependent on high-dose inotropes or cardiac arrest; and
3. acceptable right ventricular function shown by transesophageal echocardiography (TEE).
Patients additionally suffering from severe right heart failure (central and peripheral edema and echocardiographic evidence of right ventricular dilation with a right ventricular ejection fraction <20%) underwent placement of a BVAD. The device selection criteria have been previously described by Loebe et al. (11). Patients were retrospectively divided into three groups on the basis of postmechanical circulatory support outcome. Group I consisted of patients who died while on mechanical circulatory support. Group II consisted of patients who could not be weaned and then were bridged to heart transplantation, and group III consisted of patients who could be weaned off mechanical circulatory support. Patients in group III were weaned off the circulatory support because cardiac function had recovered based on echocardiographic data obtained during the weaning protocol.
Plasma samples for determination of standard blood chemistry and natriuretic peptides were taken directly before the implantation of a mechanical circulatory support system and following a standard protocol after implantation of a ventricular assist system (first postoperative day; first postoperative week; second postoperative week; third postoperative week; fourth postoperative week; every four weeks) until the patient underwent heart transplantation or could be weaned off the system or died during circulatory support. All patients underwent regular echocardiographic examination to detect ventricular recovery while supported by an LVAD or a BVAD. In addition, the preoperative and postoperative administration of inotropes was analyzed (inotrope administration for longer than seven days after implantation of the assist device).
Quantitative analysis of natriuretic peptides
Plasma BNP and ANP levels were determined in blood samples before and after implantation of the mechanical circulatory support system. All samples were taken by one physician from a peripheral vein at the same time (in the morning). Brain natriuretic peptide and ANP contents were determined by a semiquantitative analysis of blood samples. Biochemical analysis for BNP content was performed according to the method of Patrono et al. (12)using a kit (Peninsula Laboratories, San Carlos, California). Biochemical analysis for ANP content was carried out according to the method of de Bold et al. (13)with a kit (Immundiagnostik, Bensheim, Germany). The intraassay variability was 4% to 6%. Investigators who were blinded to the sample source performed the biochemical analysis.
The quantitive data are presented as mean ± SD. Comparisons among groups were calculated using the Kruskal-Wallis H-test and the Pearson chi-square test for nonparametric data. Comparisons of BNP values and ANP values before assist implantation and eight weeks after implantation of circulatory support systems were performed using the Wilcoxon signed rank test. Statistical significance was defined as p < 0.05. Because of various circumstances, including death, transplantation and weaning of the patients over the follow-up period, the number of patients was too small to test for differences over the entire follow-up period and to test for group effects and interaction terms.
Demographic data are shown in Table 1. No significant differences existed in age or gender among the groups, and the cause of end-stage heart failure was nonischemic cardiomyopathy in all cases. The age range was from 3 months to 60 years, and the duration of mechanical circulatory support varied in the groups (group I: 1 to 559 days; group II: 2 to 601 days, and group III: 13 to 224 days). Nine patients underwent placement of an LVAD system and 12 patients were supported by a BVAD system. Of the nine patients in group I, four were placed on the LVAD system (Novacor, n = 3; Berlin Heart, n = 1) and five patients on a BVAD. In group II, two patients received an LVAD (Novacor) and six patients were supported by a BVAD. Of the four patients who were succesfully weaned off the device, three patients were supported with an LVAD system (Novacor, n = 2; Berlin Heart, n = 1) and one patient with a BVAD. In group I, eight patients died of multiorgan failure and one other died of thromboembolic complications. No difference was seen among the groups concerning the pre- and postoperative administration of inotropes. In groups I and III all patients received inotropes preoperatively, and in group II, seven of eight patients received inotropes. Moreover, in group I, eight of nine patients, in group II, five of eight patients, and in group III, three of four patients received inotropes postoperatively. All patients who were weaned off the device were discharged.
Decreased plasma concentrations of natriuretic peptides
As shown in Figure 1A, we found higher plasma concentrations of BNP before the implantation of a mechanical circulatory support system in all patients when compared with BNP levels during circulatory support.
A significant decrease of plasma BNP levels occurred within eight weeks after implantation of the ventricular assist device (VAD) (p < 0.02). Patients in group I (1,650 ± 770 pg/ml) and group II (2,080 ± 1,400 pg/ml) had higher BNP levels preoperatively compared with patients in group III (1,080 ± 740 pg/ml). Furthermore, we found a faster reduction of BNP levels within the first week after implantation of VAD in patients who were weaned off the system compared with patients in groups I and II (Fig. 1B).
We found higher plasma levels of ANP before the implantation of a mechanical circulatory support system compared with ANP levels during circulatory support. However, no significant differences surfaced over the follow-up period or between the groups (Fig. 2A and 2B).
Table 2shows the echocardiographic findings before implantation of a mechanical circulatory support system and the most up-to-date echocardiographic data directly before a patient underwent heart transplantation or was weaned off the system or died during circulatory support. Patients in group III showed adequate ventricular function at the time of device explantation and had a significant increase of the left ventricular ejection fraction (LVEF) (p < 0.048) (mean LVEF [group III] before implantation of the system 17 ± 2%; mean LVEF before the explantation of the system 61 ± 5%) compared with patients in group I (mean LVEF before implantation of the system 19 ± 6%; mean LVEF before the patient died 35 ± 14%) and group II (mean LVEF before implantation of the system 15 ± 5%; mean LVEF before the patient underwent heart transplantation 33 ± 6%).
It has been widely confirmed that implantable mechanical circulatory support systems have improved the survival of patients suffering from end-stage heart failure who are awaiting heart transplantation and of those who can be successfully weaned off the system to continue a normal life without the necessity of heart transplantation (14,15).
The strategy of weaning patients off the system is relatively new and little is known about the optimum selection criteria and monitoring of patients who could be potential candidates for weaning from mechanical circulatory assistance. Currently, the decision as to whether a patient should be weaned off the system is mainly based on echocardiographic data (16,17). Meanwhile, an increasing number of investigators have reported morphological, biochemical and immunological mechanisms that may be responsible for recovery of cardiac function after prolonged assistance with a mechanical circulatory support device (18–20). Several groups hypothesize that long-term ventricular unloading by the implantation of an LVAD or a BVAD reduces the stress on the failing myocardium and may lead to the reversal of cellular abnormalities in nonischemic cardiac muscle. This has been suggested even for patients who suffer from end-stage cardiomyopathy (21,22).
Usefulness of repeated measurements of neurohumoral peptides in patients supported with mechanical circulatory assist systems
To find a potential predictor or a parameter for adequate ventricular unloading, we investigated the potential role of natriuretic peptides as physiological regulators of hypertension and plasma volume expansion. Both BNP and ANP have similar cardiovascular effects, which include reduction of cardiac preload through action in the brain, adrenal glands, kidneys, and vasculature. Both circulate in the plasma, and the concentrations correlate with the degree of ventricular dysfunction, especially the level of BNP, which is mostly synthesized in ventricular myocardium. Increased plasma levels of BNP were found in patients who had only mild heart disease (New York Heart Association [NYHA] functional class I to II), which subsequently worsened during the follow-up period (23,24). Moreover the BNP plasma concentrations were reported to be correlated with a decrease of LVEF and an increase of left ventricular end-diastolic pressure (25). Therefore, BNP might be a better marker than ANP for failing myocardium and might potentially predict the progression of the disease, which could help in its early diagnosis.
Accordingly, we hypothesized that prolonged and optimal ventricular unloading by mechanical circulatory support may lead to decreased plasma levels of natriuretic peptides and especially a decrease of BNP concentrations, which are pathophysiologically relevant to disease progression and clinical outcome.
In this study we demonstrated that prolonged ventricular unloading by mechancal circulatory assistance leads to decreased concentrations of peripheral natriuretic peptide levels and even to a significant reduction of BNP plasma levels within the first two months after implantation of the system. This may be caused by a reduction of ventricular volume overload by mechanical circulatory support. The reduction of volume-induced stress to the myocardial cells may regulate the expression of certain genes, cytokines and peptides as previously reported by other investigators (26–28). Moreover, our results showed that in patients who could be weaned off the system, the preoperative BNP plasma levels were only 50% of the BNP plasma levels of patients who underwent heart transplantation or who died while on support. Patients in group III also showed a faster reduction of BNP levels than that demonstrated in the patients in group I and group II. These findings suggest that patients with a relatively low plasma level of BNP are theoretically in an earlier stage of heart failure and therefore have a better chance for recovery from cellular abnormalities after ventricular unloading by mechanical circulatory support. Additionally, one may hypothesize that a possibility for recovery of cardiac function can be detected in an early stage of circulatory support. To clarify this hypothesis, more conclusive data will be necessary from the results of a larger number of patients.
Study limitations and future perspectives in the treatment of patients supported with mechanical circulatory assist systems
The implication of our studies is notthat BNP is the sole parameter for indicating cardiac recovery, but it may be an important and additional factor for the selection of patients or for monitoring patients for adequate ventricular unloading during mechanical circulatory assistance. This parameter can be used as a potential marker of cardiac recovery for the selection of patients who are possible candidates for weaning off the system (e.g., device selection, surgical technique). In our institution, the decision as to whether a patient will be weaned off the system is still based on clinical (e.g., NYHA classification) and functional (e.g., echocardiography) data. These data were most important among the patients studied with end-stage nonischemic cardiomyopathy but could not predict whether a patient would recover cardiac function during mechanical circulatory support. In this regard, the plasma levels of natriuretic peptides and especially BNP plasma levels can also be an important marker for the detection of cardiac recovery in an early stage and consequently be a marker for the preselection of patients for weaning. Moreover, BNP plasma levels could potentially be helpful in finding an optimal time-point for the implantation of mechanical circulatory support systems in patients suffering from end-stage heart failure, ideally before multiorgan failure occurs (29,30). For these patients a rapid BNP test is currently under investigation in our institution.
Furthermore, a rapid BNP test could be beneficial in the postoperative monitoring of patients supported with VADs outside the hospital and might improve the long-term follow-up by monitoring the current status of ventricular unloading by the system. In the last few years only a small number of reports about cardiac recovery under mechanical circulatory support have been published (31,32). Most of these were case reports, and worldwide there are only a few patients who were weaned from mechanical circulatory support without requiring heart transplantation. The number of patients in our study is small, which is certainly a limitation, but the overall number of such patients worldwide is also very small. Whether the observations reported in this study can be extended to patients with ischemic cardiomyopathy is not known. In our current study design we decided not to include ischemic cardiomyopathies because of a different pathophysiology for end-stage heart failure compared with patients suffering from nonischemic cardiomyopathy. We hypothesize that the ischemic myocardium has wide areas of fibrotic tissue and large scars, which are potentially not able to recover and may not synthesize neurohumoral peptides. This question has to be clarified in future studies and is currently under investigation in our institution.
The data presented in our study support the hypothesis that normalization of loading conditions in heart failure patients with nonischemic cardiomyopathy leads to a physiologic regulation of cytokines and peptides and may restore myocardial function. Additionally, we showed that relatively low preoperative levels of BNP may be an important marker for cardiac recovery, and a reduction of plasma BNP levels over the follow-up period may predict recovery under ventricular unloading. To support these hypotheses conclusively, more data from a larger number of patients are necessary, and a more comprehensive study is currently underway.
We acknowledge the assistance of Ms. Anne Gale (editorial), Ms. Julia Stein (statistical) and Ms. Annette Gaussmann (graphics) in the preparation of the manuscript.
☆ This work was supported in part by Freunde und Foerderer des Deutschen Herzzentrum e.V.
- atrial natriuretic peptide
- brain natriuretic peptide
- biventricular assist device
- left ventricular assist device
- left ventricular ejection fraction
- New York Heart Association
- transesophageal echocardiography
- ventricular assist device
- Received April 4, 2001.
- Revision received August 21, 2001.
- Accepted August 31, 2001.
- American College of Cardiology
- Muller J.,
- Wallukat G.,
- Weng Y.G.,
- et al.
- Burnett J.C. Jr.,
- Kao P.C.,
- Hu D.C.,
- et al.
- Yasue H.,
- Yoshimura M.,
- Sumida H.,
- et al.
- Loebe M.,
- Drews T.,
- Potapov E.,
- Ngo D.V.,
- zu Dohna R.,
- Hetzer R.
- Patrono C.,
- Peskar B.A.
- Torre-Amione G.,
- Kapadia S.,
- Lee J.,
- et al.
- Zafeiridis A.,
- Jeevanandam V.,
- Houser S.R.,
- Margulies K.B.
- Linck B.,
- Boknik P.,
- Eschenhagen T.,
- et al.
- Bozkurt B.,
- Kribbs S.B.,
- Clubb F.J. Jr.,
- et al.
- Dipla K.,
- Mattiello J.A.,
- Jeevanandam V.,
- Houser S.R.,
- Margulies K.B.
- Tsutamoto T.,
- Wada A.,
- Maeda K.,
- et al.
- Brunner-La Rocca H.P.,
- Kaye D.M.,
- Woods R.L.,
- et al.