Effect of the Implantable Left Ventricular Assist Device on Neuroendocrine Activation in Heart Failure
Background The HeartMate left ventricular assist device has been successfully used as a bridge to cardiac transplantation. Because many patients exhibit marked clinical improvement in their heart failure after HeartMate implantation, we studied the physiological effect of this device on the neurohormonal axis.
Methods and Results In 13 patients awaiting transplant (mean cardiac index, 1.7±0.3 L · min−1 · m−2) who underwent HeartMate implantation, venous atrial natriuretic peptide, epinephrine, norepinephrine, plasma renin activity, angiotensin, and arginine vasopressin were measured immediately before insertion and at explant/transplantation. Mean time to explant was 86±40 days. All patients were taken off inotropic medications within 1 month. Mean cardiac index on support before explant was 3.1±0.9 L · min−1 · m−2. Plasma renin activity decreased from 57±56 ng · mL−1 · h−1 at baseline (before insertion) to 3±3 ng · mL−1 · h−1 at explant (mean percent change, 92%; P<.001). Angiotensin II level decreased from 237±398 U/L at baseline to 14±14 U/L at explant (mean percent change, 73%; P<.001). Plasma epinephrine level fell from 6800±1323 pg/mL at baseline to 46±46 pg/mL at explant (mean percent change, 86%; P<.001). Norepinephrine level decreased from 2953±1457 pg/mL at baseline to 518±290 pg/mL at explant (mean percent change, 79%; P<.001). Atrial natriuretic peptide fell from baseline values of 227±196 to 168±40 pg/mL at explant (mean percent change, −49%; P=.519); and arginine vasopressin level decreased from 6±6 pg/mL at baseline to 0.8±0.5 pg/mL (mean percent change, 69%; P=.002).
Conclusions We provide data supporting that the neurohormonal axis markedly improves after HeartMate implantation, providing biochemical confirmation of the improvement in hemodynamic status.
Congestive heart failure is accompanied by activation of neuroendocrine systems.1 2 3 4 5 6 7 8 9 10 11 12 Plasma NE, ANP, PRA, and AVP levels are elevated in both symptomatic and asymptomatic LV dysfunction, as is well documented in the Studies of Left Ventricular Dysfunction (SOLVD).13
A favorable effect of decreasing neurohormonal activation has been shown with several successful pharmacological therapies for congestive heart failure, including angiotensin-converting enzyme inhibition, digitalis, and some of the third-generation calcium-channel blockers.14 15 16 17 18 19 20 However, the effects of nonpharmacological treatments for heart failure on the neurohormonal axis have not been extensively studied. The aim of this study was to evaluate the effect on neurohormone levels of an implantable long-term LVAD used as a bridge to cardiac transplantation. It has been observed that marked clinical improvement with reversal of heart failure symptoms and use of rehabilitation can be achieved with the HeartMate in place.21 Study of the concomitant physiological effects of the device on neurohormones was the primary objective of this effort.
The HeartMate device (Thermo Cardiosystems, Inc, Woburn, Mass) was implanted in each patient via a median sternotomy incision while on cardiopulmonary bypass.22 The device was placed in a pocket in the left upper abdominal wall. The inflow conduit was inserted into the LV cavity after excision of a core of LV apex, and the outflow conduit was anastomosed to the ascending aorta. Inflow and outflow tissue valves were used. With the HeartMate, blood flows from the LV through the inflow cannula into the pump and is then ejected into the ascending aorta. Because the HeartMate very efficiently unloads the LV, the patient’s aortic valve usually does not open during systole. Air is vented from the device as cardiopulmonary bypass is weaned, and LV assist device operation is begun.
There were 13 patients enrolled in the study who received HeartMate devices, had full neurohumoral data, and eventually were transplanted. The study period extended from December 1, 1992, through March 1994. The study group consisted of 11 men and 2 women, with a mean age of 49±6.6 years (range, 43 to 61 years). All were in cardiogenic shock (mean cardiac index, 1.7±0.3 L · min−1 · m−2) and approved for cardiac transplantation. The origin of the underlying heart disease was idiopathic dilated cardiomyopathy in 4 and ischemic cardiomyopathy in the other 9. Ten of the patients were on intra-aortic balloon counterpulsation support before HeartMate implantation. All HeartMate recipients gave consent in compliance with the Institutional Review Board at Cleveland Clinic.
Timing of Samplings
Venous plasma neurohormonal samplings of ANP, EPI, NE), PRA, angiotensin II (Ang II), and AVP were drawn in each patient at baseline (within 24 hours before HeartMate insertion) and just before HeartMate explantation and transplantation. All patients were on low sodium (2 g/d) diets before and after implantation.
HeartMate pump output was recorded at the time of each sampling. Treadmill time and distance were also recorded in those undergoing cardiac rehabilitation while on HeartMate support.
Pulmonary artery catheters were in place in all patients at baseline; but within 1 month on HeartMate, support had been discontinued in all patients. Therefore cardiac output and right ventricular pressures were recorded at only two times: baseline before HeartMate insertion and at the time of explantation/cardiac transplantation.
At the time of baseline neurohumoral samplings, all patients were receiving inotropic/catecholamine support in cardiogenic shock. By the time of the 1 month sampling, all 13 patients had been taken off catecholamine support. The baseline inotrope regimens for the group are depicted in Table 1⇓.
Only one of the patients was receiving an angiotensin-converting enzyme inhibitor throughout the HeartMate study period, and only one other patient was receiving digoxin for atrial fibrillation throughout the study period. These drugs were not utilized, therefore, in the majority of the patients because they improved markedly both clinically and hemodynamically with the LVAD alone. All of the patients were receiving diuretic therapy at the time of HeartMate insertion. All but one remained on various maintenance diuretic doses throughout the HeartMate support period until cardiac transplantation.
A short, 18- or 20-gauge, IV cannula was utilized, connected to a 3-way stopcock, to obtain plasma venous samplings. The catheter was filled with dilute heparinized saline solution, with samplings performed after 30 minutes of supine rest. A total of 64 mL of blood was obtained to perform all the neurohormonal assays per patient sampling. For plasma NE and EPI analysis, 10 mL of blood was placed into prechilled tubes containing reduced glutathione and EGTA preservative. The specimen was centrifuged within 1 hour (4°C at 2500 rpm for 15 minutes), then transferred to a polypropylene tube and stored at −70°C. Ten mL of blood for PRA and Ang II measurement was drawn into a tube containing liquid potassium and EGTA. After inversion of the tube several times for mixing, the sample was centrifuged within 1 hour (4°C at 2500 rpm for 15 minutes). Plasma AVP and ANP samples were placed in prechilled EGTA tubes, centrifuged, and stored as in the aforementioned method. All samples were transported on ice for analysis to the Endocrine/Hypertension Research Laboratory, Research Institute, Cleveland Clinic.
PRA was measured by RIA of generated Ang I as previously described.23 Values in 25 normal supine subjects averaged 1.2±0.84 ng · mL−1 · h−1 (mean±SD) and ranged from 0.6 to 2.6 ng · mL−1 · h−1.
Plasma Ang II concentrations were determined by RIA with a detection limit of 1 pg/mL.24 Intra-assay and interassay coefficients of variation for plasma Ang II were 5% and 9%, respectively. Normal range values ranged from <1.3 to 10.5 pg/mL.
Plasma AVP was assayed by RIA according to the methods of Crofton et al.25 Intra-assay and interassay coefficients of variation for plasma AVP were 5% and 8%, respectively. Samples with values below the detectability limit of the assay (<0.5 pg/mL) were assigned a value of 0.4 pg/mL. Normal range values ranged from 0.4 to 3.6 pg/mL.
Plasma ANP was measured using a RIA technique developed in the laboratory of one of us (E.L.B.). The RIA is a 3-day assay involving prior plasma extraction with Bond Elut C-18 columns and a 24-hour preincubation of standards, controls, and samples with antibody at 4°C. Separation of bound from free fractions was achieved by second antibody and normal rabbit serum. The sensitivity of the assay is 12 pg/mL. The intra-assay and interassay coefficients of variation were 6.5% and 14%, respectively. Normal controls (n=18) had plasma concentrations of 30.4±2.5 pg/mL (mean±SEM) on normal salt intake, which decreased to 16.4±1.5 pg/mL on salt deprivation and increased to 55.0±6.7 pg/mL on a high salt diet.
Plasma NE and EPI were measured by a radioenzymatic assay technique described by de Champlain et al.26 It is sensitive to 50 pg/300 μL of plasma of either NE or EPI. The intra-assay and interassay coefficients of variation are 3% and 6%, respectively. Normal plasma NE concentration is 218±92 pg/mL (mean±SD) and for EPI, 42±18 pg/mL (mean±SD). For concentrations <50 pg, the assays were repeated with a larger plasma volume to enhance detectability.
The neurohormonal and hemodynamic data are expressed as mean values±SD. Mean percent change between baseline and follow-up neurohormone samplings was calculated as last sampling value minus first sampling value divided by first value. Comparisons between any two groups were drawn using Wilcoxon’s signed rank test.
All 13 patients were receiving inotropic support at the time of HeartMate implantation, but all were successfully weaned off the inotropes within 1 month. All of the patients exhibited improvement clinically while on HeartMate support as they underwent cardiac rehabilitation, ambulating on the treadmill from 17 to 100 minutes, varying by patient, with walking distances ranging from 0.4 to 3 miles.
The HeartMate device generated flow rates in the group ranging from 3.4 to 8.2 L/min, which allowed adequate hemodynamic pressures and flows. The patients exhibited significant hemodynamic improvement while receiving HeartMate support, with an increase in mean cardiac index from 1.7±0.3 L · min−1 · m−2 at baseline to 3.1±0.9 L · min−1 · m−2 at explant (P<.001) and a decrease in mean pulmonary arterial pressure from 38.4±8.5 to 19.9±6 mm Hg (P<.001) (Table 2⇓).
The results of the baseline and explant/transplant neurohormone levels, including normal ranges, after HeartMate implantation are shown in Table 3⇓. Mean time to explant was 86±40 days. PRA decreased from 57±56 ng · mg−1 · h−1 at baseline to 3±3 ng · mL−1 · h−1 at explant (mean percent change, 92%; P<.001). Ang II decreased from 237±398 U/L at baseline to 14±14 U/L at explant (mean percent change, 73%; P<.001). Plasma EPI fell from 6800±1323 pg/mL at baseline to 46±46 pg/mL at explant (mean percent change, 86%; P<.001). NE decreased from 2953±1457 pg/mL at baseline to 518±290 pg/mL at explant (mean percent change, 79%; P<.001). ANP decreased from baseline values of 227±196 to 168±40 pg/mL at explant (mean percent change, −49%; P=.519); and AVP decreased from 6±6 pg/mL at baseline to 0.8±0.5 pg/mL (mean percent change, 69%; P=.002). There was no significant change in blood urea nitrogen or serum creatinine from baseline to explant and, accordingly, there also was no relation of neurohormone levels to renal function.
Clinically, the hemodynamic profiles in these critically ill patients improved markedly during HeartMate support, with a rise in mean cardiac index from 1.7±0.3 L · min−1 · m−2 before implantation to 3.1±0.9 L · min−1 · m−2 at time of explant for cardiac transplantation, with corresponding improvements in pulmonary arterial pressure as discussed earlier. Implantable LVAD support has also been previously shown to effect improvement in renal, hepatic, and pulmonary function as patients await cardiac transplantation.22 The neurohormonal samplings gathered as part of this study were not used to determine whether the patient was “ready” to undergo transplantation. The values were stored and analyzed periodically. No clinical decisions were made based on the patient’s neurohormonal levels.
The baseline neurohormone levels at the time of cardiogenic shock leading to HeartMate implantation in our study reflect marked neurohormonal activation as previously described in congestive failure states.1 2 3 4 5 6 7 8 9 10 11 12 Elevations of NE, AVP, PRA, and ANP have been well documented in states of LV dysfunction both with and without symptomatic congestive heart failure, as described in the SOLVD study.13 Moreover, sequential rises in neurohormonal activation have been demonstrated in patients with chronic congestive heart failure.1 Recently, neurohormonal activation has also been reported at the time of hospital discharge in myocardial infarction patients with postinfarction LV dysfunction.27 According to the neurohormonal hypothesis, congestive heart failure progresses as the activated neurohumoral systems cause ongoing deleterious cardiovascular effects, either by exacerbation of hemodynamic derangements or by a direct myocardial toxic effect.2 High concentrations of Ang II and NE have in fact been shown to exert deleterious effects on myocytes.28 29
Numerous studies have shown that the severity of elevation of the various hormones significantly predicts prognosis in severe chronic congestive heart failure.3 7 10 Accordingly, the success of several pharmacological agents in improving clinical status and survival has been attributed, at least in part, to downregulation of these neurohormones. Although multiple mechanisms of action for digitalis have been described, its effect on decreasing neurohumoral activation may prove to be one of its most potent mechanisms.14 15 16 17 18 In addition to clinical data, direct measurement of efferent sympathetic nerve activity in humans with heart failure demonstrates marked sympathoinhibitory effects of digitalis.14 Similarly, although the angiotensin-converting enzyme inhibitors have multiple mechanisms of action, it is believed that their favorable effect on decreasing neurohormones may be a main reason for their positive impact on clinical improvement and survival in the setting of LV dysfunction.10 19 β-Blockers similarly have been shown to exert favorable effects in heart failure patients, further supporting the neurohormonal hypothesis. Subgroup analysis of large trials has demonstrated reduced mortality in patients with LV dysfunction due to ischemic heart disease who received β-blocker therapy.30 The third-generation calcium channel blockers amlodipine and felodipine have been shown to decrease neurohormonal output in heart failure and are currently being investigated for clinical efficacy.20 However, the effects of nonpharmacological therapies for heart failure, such as ventricular support devices, on neurohormones are not well described.
Some medications for heart failure and shock can affect levels of neurohormones.31 Inotropes and catecholamines, including dobutamine, dopamine, NE, and milrinone, were required in various combinations because of cardiogenic shock at the time of HeartMate implantation (Table 1⇑). Accordingly, baseline plasma EPI levels were likely elevated further by those receiving intravenous EPI; and baseline plasma NE was probably also elevated further in those receiving intravenous NE. Also, baseline plasma EPI was probably further elevated in those receiving intravenous dopamine because dopamine undergoes conversion to EPI. Within 1 month, however, all inotropes had been discontinued in these patients.
It is not possible to definitively show that the drop in neurohormones is due purely to the support device rather than to discontinuation of inotropes as well, because it was not possible to interrupt inotropic supportive therapy for measurement of baseline hormone levels in these patients at their most critical point of decompensation requiring LV assist. However, historical baseline data from the VA Cooperative Studies Group show markedly elevated plasma NE levels >900 pg/mL in the sickest heart failure patients on medical therapy without intravenous inotropes.5 Moreover, our focus was not the elevation of the catecholamine levels at baseline but rather their marked decline toward the normal range on mechanical support in the presence of a still-diseased LV.
We refer to the LV as “still-diseased,” based on data from other studies.32 Although hemodynamic data showed marked improvement on LVAD support (Table 2⇑), from a histological standpoint changes of continued myocyte abnormalities previously were described.32 Comparison of cores of LV apexes at implantation and explantation has revealed a reduction in myocyte wavy fibers and contraction band necrosis on long-term LVAD support, accompanied by an increase in myocardial fibrosis. Echocardiographically, fractional shortening has also not been found to be significantly changed on long-term LVAD support.32
It has previously been shown that diuretic use can account solely for PRA elevation.13 33 All but one of our patients continued to receive some diuretic throughout their HeartMate support. The diuretic use can explain why the PRA and Ang II levels remained above normal range. However, despite continued use of diuretics, there was still significant reduction in Ang II and PRA levels, suggesting that other factors corrected by the LVAD contribute to the rise in Ang II levels in patients with cardiac failure.
Right and left atrial pressures significantly decreased on HeartMate support. In a parallel fashion, ANP levels also decreased on ventricular support, although not to a significant degree. There are several potential explanations for this. The right atrial pressures, though markedly decreased, were still somewhat elevated above normal (Table 2⇑). The mean right atrial pressure at explant actually exceeded the mean left atrial pressure, probably reflecting the unloading of the LV by the assist device. Second, the diseased LV can be a source of ANP.34 Third, it is conceivable that the HeartMate may cause ANP release, although it would be difficult to dissect this possibility from the concomitant heart failure conditions.
A limitation inherent in this study is the potential confounding effect of the medications. However, the effect of HeartMate mechanical support on reversing neurohormonal activation in heart failure is evident. The extreme initial neurohumoral elevations in these critically ill heart failure patients on inotropic support are not at all unexpected. Rather, the marked, significant decrease in these neurohormone levels toward normal ranges in patients with continued severe LV dysfunction before transplantation receiving HeartMate support is the remarkable finding. The mechanism for this hormonal decrease remains speculative at this time; our data suggest that the improvement of the hemodynamic derangements afforded by mechanical support is contributory.
Another limitation of this study is the small number of patients studied. Not all of the HeartMate LVAD patients from the Cleveland Clinic was entered into this study. The study began after the initial group of LVAD patients at the Cleveland Clinic was implanted, and other patients who underwent emergency HeartMate LVAD insertion could not have baseline measurements drawn. Finally, we did not report on a small group of patients who did not survive until transplantation because their data were incomplete and patients died early; therefore sequential data were not available. Most importantly, we thought that this information would serve as a benchmark for patients who receive the LVAD for permanent implantation. These neuroendocrine study results, therefore, can be extrapolated to patients with permanent LVAD implantation, versus those with cardiac transplantation or those with medical therapy for congestive heart failure.
This study did not address the subsequent course of neurohormonal levels after cardiac transplantation. This has, however, been investigated by several other groups. Findings on NE and PRA levels after transplant vary somewhat from study to study, whereas the data on ANP after transplantation are more consistent. Spes et al35 report that the elevations of NE, EPI, and PRA seen in heart failure reverse after heart transplantation, probably as a consequence of normalization of cardiac function. On the other hand, Quigg et al36 describe persistent elevation of NE and PRA levels after transplantation. In their study, the NE levels appear to increase over time after transplant and significantly correlate with serum creatinine. They conclude that the elevated NE may be related to cyclosporine-induced renal insufficiency. Among various studies, ANP levels appear to remain persistently elevated after transplantation. These authors offer various reasons as to why this occurs: rejection; a compensatory response to cyclosporine-induced hypertension; effects of the surgical transplant procedure or antirejection therapy; or persistence of factors that were present preoperatively.35 36 37
We have shown that mechanical ventricular support of patients with end-stage cardiomyopathy results in a dramatic reduction in neurohormonal stimulation concomitant with improvement in hemodynamic status. To the extent that catecholamine excess may itself have a detrimental effect in cardiomyopathy, this neurohumoral stabilization may be of benefit in “resting” the ventricle mechanically in potentially reversible cardiomyopathies.
Selected Abbreviations and Acronyms
|Ang II||=||angiotensin II|
|ANP||=||atrial natriuretic peptide|
|LV||=||left ventricle/left ventricular|
|LVAD||=||left ventricular assist device|
|PRA||=||plasma renin activity|
|SOLVD||=||Studies of Left Ventricular Dysfunction|
The neuroendocrine assays were funded by the Cleveland Clinic Department of Cardiothoracic Surgery.
Reprint requests to Karen B. James, MD, FACC, Cleveland Clinic Foundation, Department of Cardiology F25, 9500 Euclid Ave, Cleveland, OH 44195.
- Copyright © 1995 by American Heart Association
Packer M, Lee WH, Kessler PD, Gottlieb SS, Bernstein JL, Kukin M. Role of neurohormonal mechanisms in determining survival in patients with severe chronic heart failure. Circulation. 1987;75(suppl IV):IV-80-IV-92.
Francis GS, Cohn JN, Johnson G, Rector TS, Goldman S, Simon A, for the V-HeFT VA Cooperative Studies Group. Plasma norepinephrine, plasma renin activity, and congestive heart failure. Circulation. 1993;87(suppl VI):VI-40-VI-48.
Francis GS, Goldsmith SR, Levine TB, Olivera MT, Cohn JT. The neurohumoral axis in congestive heart failure. Ann Intern Med. 1984;101:370-377.
Rector TS, Olivari MT, Levine TB, Francis GL, Cohn JN. Predicting survival for an individual with congestive heart failure using plasma norepinephrine concentration. Am Heart J. 1987; 114: 148-152.
Swedberg K, Eneroth P, Kjekshus J, Wilhelmsen L, for the CONSENSUS Trial Study Group. Hormones regulating cardiovascular function in patients with severe congestive heart failure and their relation to mortality. Circulation. 1990;82:1730-1736.
Kao W, Gheorghiade M, Hall V, Goldstein S. Relation between plasma norepinephrine and response to medical therapy in men with congestive heart failure secondary to coronary artery disease or idiopathic dilated cardiomyopathy. Am J Cardiol. 1989; 64: 609-613.
Francis GS, Benedict C, Johnstone DE, Kirlin PC, Mickas J, Liang C, Kubo SH, Rduin-Toretsky E, Yusuf S, for the SOLVD Investigators. Comparison of neuroendocrine activation in patients with left ventricular dysfunction with and without congestive heart failure: a substudy of the Studies of Left Ventricular Dysfunction (SOLVD). Circulation. 1990;82:1724-1729.
Ferguson DW, Berg WJ, Sanders JS, Roach PJ, Kempf JS, Kienzle MG. Sympatho-inhibitory responses to digitalis glycosides in heart failure patients: direct evidence from sympathetic neural recordings. Circulation. 1989;80:65-77.
Alicandri C, Fariello R, Boni E, Zaninelli A, Castellano M, Beschi M, Rosei A, Muiesan G. Captopril versus digoxin in mild-moderate chronic heart failure: a crossover study. J Cardiovasc Pharmacol. 1987;9(suppl 2):S61-S67.
Ribner HS, Plucinski DA, Hsieh A-M, Bresnahan D, Molteni A, Askenzai J, Lesch M. Acute effects of digoxin on total systemic vascular resistance in congestive heart failure due to dilated cardiomyopathy: a hemodynamic-hormonal study. Am J Cardiol. 1985; 56: 896-904.
The CONSENSUS Trial Study Group. Effect of enalapril on mortality in severe congestive heart failure. N Engl J Med. 1987; 316: 1429-1435.
Suzuki H, Ferrario CM, Septh RC, Brosnihan KB, Smeby RR, DeSilva P. Alterations in plasma and cerebrospinal fluid norepinephrine and angiotensin II during the development of renal hypertension in conscious dogs. Hypertension. 1983;5(suppl I): I-139-I-149.
Crofton JT, Share L, Wang BC, Shade RE. Pressor responsiveness to vasopressin in the rat with DOC-salt hypertension. Hypertension. 1980;2:424-431.
de Champlain J, Farley L, Cousineau D, Van Ameringen M. Circulating catecholamine levels in human and experimental hypertension. Circ Res. 1976;38:109-114.
Rouleau JL, de Champlain J, Klein M, Bichet D, Moye L, Packer M, Dagenais GR, Sussex B, Arnold JM, Sestier F, Parker JO, McEwan P, Bernstein V, Cuddy TE, Lamas G, Gottlieb SS, McCans J, Nadeau C, Delage F, Hamm P, Pfeffer MA. Activation of neurohumoral systems in postinfarction left ventricular dysfunction. J Am Coll Cardiol. 1993;22:390-398.
Mann DL, Kent RL, Parson B, Cooper G. Adrenergic effects on the biology of the adult mammalian cardiocyte. Circulation. 1992; 85: 790-804.
Tan LB, Jalil JE, Pick R, Janicki JS, Weber KT. Cardiac myocyte necrosis induced by angiotensin II. Circ Res. 1991;69:1185-1195.
Chadda K, Goldstein S, Byington R, Curb JD. Effect of propranolol after acute myocardial infarction in patients with congestive heart failure. Circulation. 1986;73:503-510.
Bayliss J, Norell M, Canepa-Anson R, Sutton G, Poole-Wilson P. Untreated heart failure: clinical and neuroendocrine effects of introducing diuretics. Br Heart J. 1987;57:17-22.
Rodeheffer RJ, Naruse M, Atkinson JB, Naruse K, Burnett JC, Merrill WH, Frist WH, Demura H, Inagami T. Molecular forms of atrial natriuretic factor in normal and failing human myocardium. Circulation. 1993;88:364-371.
Quigg R, Sica D, Griffith L, Roberge P, Salyer J, Gehr T, Quint R. Persistent elevation of neurohormones despite correction of systolic heart failure by cardiac transplantation. J Am Coll Cardiol. 1994;140A. Abstract.