Randomized, Double-Blind Trial of Inhaled Nitric Oxide in LVAD Recipients With Pulmonary Hypertension

Transcription

1 ORIGINAL ARTICLES: CARDIOVASCULAR Randomized, Double-Blind Trial of Inhaled Nitric Oxide in LVAD Recipients With Pulmonary Hypertension Michael Argenziano, MD, Asim F. Choudhri, BS, Nader Moazami, MD, Eric A. Rose, MD, Craig R. Smith, MD, Howard R. Levin, MD, Arthur J. Smerling, MD, and Mehmet C. Oz, MD Departments of Surgery, Medicine, and Anesthesiology, Columbia University College of Physicians and Surgeons, New York, New York Background. Pulmonary vascular resistance is often elevated in patients with congestive heart failure, and in those undergoing left ventricular assist device (LVAD) insertion, it may precipitate right ventricular failure and hemodynamic collapse. Because the effectiveness of inotropic and vasodilatory agents is limited by systemic effects, right ventricular assist devices are often required. Inhaled nitric oxide (NO) is an effective, specific pulmonary vasodilator that has been used successfully in the management of pulmonary hypertension. Methods. Eleven of 23 patients undergoing LVAD insertion met criteria for elevated pulmonary vascular resistance on weaning from cardiopulmonary bypass (mean pulmonary artery pressure >25 mm Hg and LVAD flow rate <2.5 L min 1 m 2 ) and were randomized to receive either inhaled NO at 20 ppm (n 6) or nitrogen (n 5). Patients not manifesting a clinical response after 15 minutes were given the alternative agent. Results. Hemodynamics for the group at randomization were as follows: mean arterial pressure, 72 6 mm Hg; mean pulmonary artery pressure, 32 4 mm Hg; and LVAD flow, L min 1 m 2. Patients receiving inhaled NO exhibited significant reductions in mean pulmonary artery pressure and increases in LVAD flow, whereas none of the patients receiving nitrogen showed hemodynamic improvement. Further, when the nitrogen group was subsequently given inhaled NO, significant hemodynamic improvements ensued. There were no significant changes in mean arterial pressure in either group. Conclusions. Inhaled NO induces significant reductions in mean pulmonary artery pressure and increases in LVAD flow in LVAD recipients with elevated pulmonary vascular resistance. We conclude that inhaled NO is a useful intraoperative adjunct in patients undergoing LVAD insertion in whom pulmonary hypertension limits device filling and output. (Ann Thorac Surg 1998;65:340 5) 1998 by The Society of Thoracic Surgeons Presented at the Thirty-third Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Feb 3 5, Address reprint requests to Dr Argenziano, Division of Cardiothoracic Surgery, Milstein Hospital, Rm 7-435, 177 Fort Washington Ave, New York, NY ( ma66@columbia.edu). Pulmonary vascular resistance (PVR) is often elevated in congestive heart failure. Although the precise mechanism of this form of secondary pulmonary hypertension is not known, chronic elevations in left atrial pressure and pulmonary venous congestion as well as alterations in sympathetic tone are thought to play a role [1]. Patients with heart failure and pulmonary hypertension, whether it is primary or secondary, are at increased risk of right ventricular failure after cardiac transplantation [2]. For this reason, demonstration that PVR can be decreased pharmacologically in such patients (ie, that the pulmonary hypertension is not due to fixed vascular resistance) is generally a requirement before transplantation [3]. In patients receiving a left ventricular assist device (LVAD) for advanced heart failure, pulmonary hypertension often limits the adequacy of device filling, which leads to right ventricular failure in up to 40% of patients [4]. Because the effectiveness of inotropic and nonspecific vasodilator agents in the management of these patients is frequently limited by systemic hypotension and inhibition of alveolar oxygenation [5], the institution of right ventricular mechanical support is sometimes required [6]. Although right ventricular assist devices effectively improve cardiac output in these situations, their use is complicated by prolongation of operative time, the need of reestablishment of cardiopulmonary bypass, and an increased risk of hemorrhage [7]. Furthermore, biventricular assistance greatly inhibits patient mobility and necessitates additional operation for device removal. Nitric oxide (NO), initially described as an endotheliumderived relaxation factor [8], has been implicated in a wide variety of physiologic and pathophysiologic processes. Inhaled NO has been shown to cause pulmonary vasodilation in primary pulmonary hypertension [9], in pulmonary hypertension secondary to congenital heart disease [10], and in the adult respiratory distress syn by The Society of Thoracic Surgeons /98/$19.00 Published by Elsevier Science Inc PII S (97)

2 Ann Thorac Surg ARGENZIANO ET AL 1998;65:340 5 INHALED NO IN LVAD RECIPIENTS 341 drome [11]. Recently, inhaled NO has been used to selectively lower PVR in patients undergoing cardiac surgical procedures [12]. Prompted by the successful use of inhaled NO in 2 LVAD recipients with right ventricular failure, we initiated the present randomized, placebocontrolled study to investigate the short-term effects of inhaled NO in patients with hemodynamically significant pulmonary hypertension after insertion of an LVAD for end-stage heart failure. Material and Methods Study Patients and Protocols Over an 8-month period, 23 consecutive patients undergoing LVAD insertion were evaluated, and informed consent for the randomized, placebo-controlled trial was obtained. Standard care included perioperative administration of dobutamine hydrochloride and dopamine hydrochloride and the loading of amrinone lactate at initiation of cardiopulmonary bypass. On weaning from cardiopulmonary bypass, 11 of these patients met inclusion criteria for hemodynamically significant pulmonary hypertension: mean pulmonary artery pressure (PAP) greater than 25 mm Hg and LVAD assisted cardiac index less than 2.5 L min 1 m 2 despite maximal medical therapy (volume loading to a central venous pressure 15 mm Hg and administration of intravenous amrinone, g/kg). The study protocol was approved by the Institutional Review Board of Columbia University. Hemodynamics were recorded for the 23 LVAD recipients on arrival in the operating room, throughout the device insertion procedure, and during the postoperative intensive care unit stay. On meeting the two selection criteria 5 minutes after weaning from cardiopulmonary bypass, LVAD patients were blindly randomized to receive inhaled NO at a delivered concentration of 20 ppm or nitrogen (N 2 ) at an equivalent flow rate. A clinical response was defined as a decrease in mean PAP of 5 mm Hg or greater, an increase in LVAD output of more than 20%, or both in the absence of other pharmacologic or surgical interventions. In the absence of a clinical response to the assigned agent within 15 minutes, patients were given the alternative agent, again in a blinded manner. If a clinical response was observed, the assigned agent was continued postoperatively. On arrival in the intensive care unit, the patient was weaned from the administered gas mixture (NO or N 2 ) to maintain mean PAP at less than 25 mm Hg and LVAD flow at higher than 2.5 L min 1 m 2. Nitric Oxide Administration Nitric oxide, an NO stock gas mixture (800 ppm) (Airco Special Gases, Riverton, NJ) diluted appropriately with oxygen, was administered intraoperatively through the operating room ventilator system. The breathing circuit was equipped with an in-line oxygen analyzer downstream as well as a chemiluminescence monitor attached as close to the patient as possible to monitor the delivered concentration of oxygen and NO. Delivered inhaled NO concentration was controlled by adjusting the ratio of NO flow to ventilator flow. All exhaled gases were scavenged appropriately. In the intensive care unit, NO, again an 800-ppm stock gas mixture, was administered through a modified ventilator circuit. Analysis of Data Hemodynamic and clinical data are reported as the mean the standard error of the mean. Paired variables were analyzed by the paired Student s t test, and unpaired variables were compared using the Wilcoxon nonparametric test. A p value of less than 0.05 was considered significant. Results Eleven of the 23 LVAD recipients met inclusion criteria for hemodynamically significant pulmonary hypertension on weaning from cardiopulmonary bypass; they had increased PAP (32 4 mm Hg) and decreased LVAD flow index ( L min 1 m 2 ). The study group consisted of 7 men and 4 women with a mean age of 55 years. The cause of heart failure was ischemic cardiomyopathy in 7 patients, idiopathic dilated cardiomyopathy in 2, myocarditis in 1 patient, and acute cardiac allograft failure in 1. Preoperative ejection fraction was , and patients received the Thermo Cardiosystems Inc (Woburn, MA) pneumatic (7 patients) or vented electric (3 patients) device or the ABIOMED (Danvers, MA) BVS 5000 pneumatic device (1 patient). Demographics and prebypass hemodynamics for the 11 patients are summarized in Table 1. Six patients were randomized to receive inhaled NO and 5 to inhaled N 2 ; pretreatment hemodynamics of the two groups are summarized in Table 2. In the 6 patients randomized to inhaled NO, this agent rapidly ( 10 minutes) and significantly decreased mean PAP from 35 6mmHgto24 4mmHg(p 0.02) (Fig 1) and increased LVAD flow index from L min 1 m 2 to L min 1 m 2 (p 0.02) (Fig 2). Mean arterial pressure increased slightly in this group, but this change was not significant (Fig 3). The 5 patients randomized to the inhaled N 2 placebo showed no significant changes in mean PAP (see Fig 1), LVAD flow index (see Fig 2), or mean arterial pressure (see Fig 3) after 15 minutes. When they were crossed over to receive inhaled NO at 20 ppm, mean PAP rapidly decreased from 31 4mmHgto22 3mmHg(p 0.02) (see Fig 1), and LVAD flow index increased from to L min 1 m 2 (p 0.002) (see Fig 2). Overall, 11 LVAD recipients with pulmonary hypertension after cardiopulmonary bypass received inhaled NO (either at primary randomization or after crossover from the inhaled N 2 group), and this treatment resulted in dramatic and highly significant decreases in mean PAP and increases in LVAD flow, as summarized in Table 3. Importantly, inhaled NO effected significant pulmonary vasodilation without decreasing systemic arterial pressure, thus underscoring the exquisite pulmonary selectivity of this agent.

3 342 ARGENZIANO ET AL Ann Thorac Surg INHALED NO IN LVAD RECIPIENTS 1998;65:340 5 Table 1. Demographic Characteristics and Prebypass Hemodynamics of 11 Patients With Left Ventricular Assist Device and Pulmonary Hypertension a Variable Value Age (y) 55 3 Male to female ratio 7:4 Indication for LVAD support Ischemic cardiomyopathy 7 (64) Idiopathic dilated cardiomyopathy 2 (18) Myocarditis 1 (9) Acute cardiac allograft failure 1 (9) Device TCI pneumatic 7 (64) TCI vented electric 3 (27) ABIOMED BVS (9) Preoperative ejection fraction Prebypass hemodynamics MAP (mm Hg) 71 5 MPAP (mm Hg) 42 2 CO (L/min) PVR (Wood units) a Data are shown as the mean the standard error of the mean or as the number of patients with percentages in parentheses. CO cardiac output; LVAD left ventricular assist device; MAP mean arterial pressure; MPAP mean pulmonary artery pressure; PVR pulmonary vascular resistance; TCI Thermo Cardiosystems Inc. Fig 1. Effects of inhaled nitric oxide (NO I ) and nitrogen (N 2 )on mean pulmonary artery pressure (MPAP). For patients maintaining a mean PAP lower than 25 mm Hg and an LVAD flow index greater than 2.5 L min 1 m 2, inhaled NO was slowly reduced from 20 ppm to 2 ppm and then discontinued. The median duration of inhaled NO support was 24 hours (range, 12 hours to 6 days). No complications were associated with inhaled NO administration. In 1 patient dependent on inhaled NO, a ventilator malfunction on postoperative day 2 caused abrupt discontinuation of the gas, resulting in hemodynamic collapse and ventricular fibrillation. This patient was resuscitated and received a right ventricular assist device, which was removed successfully 3 days later, again with inhaled NO support. The patient was weaned from inhaled NO over the next 2 days and subsequently underwent transplantation. There were two perioperative deaths. One occurred on postoperative day 1 and was due to intractable hemorrhage in a patient with multiple-system failure. The other occurred on postoperative day 3 and was due to brain death in a patient with an intraoperative cerebrovascular embolic event. Comment Pulmonary hypertension is a common consequence of long-standing congestive heart failure. Although mechanisms responsible for this process are poorly understood, chronic elevations in left atrial pressure, pulmonary venous congestion, intravascular volume expansion, and alterations in sympathetic tone are thought to play a role [1]. Interestingly, recent evidence suggests that pulmonary hypertension may, in fact, result from impairments in the release of NO by the pulmonary vascular endothe- Table 2. Hemodynamics Before Treatment in Patients Randomized to Receive Inhaled Nitric Oxide or Nitrogen a Variable Total (n 11) NO I (n 6) N 2 (n 5) MAP (mm Hg) MPAP (mm Hg) LVAD index (L min 1 m 2 ) a Data are shown as the mean the standard error of the mean. LVAD left ventricular assist device; MAP mean arterial pressure; MPAP mean pulmonary artery pressure; N 2 nitrogen; NO I inhaled nitric oxide. Fig 2. Effects of inhaled nitric oxide (NO I ) and nitrogen (N 2 ) on left ventricular assist device (LVAD) flow index.

4 Ann Thorac Surg ARGENZIANO ET AL 1998;65:340 5 INHALED NO IN LVAD RECIPIENTS 343 Fig 3. Effects of inhaled nitric oxide (NO I ) and nitrogen (N 2 )on mean arterial pressure (MAP). lium [13] and that cardiopulmonary bypass related endothelial dysfunction may be responsible for further perioperative elevations in PVR [14]. Whatever the mechanism, secondary elevations in PVR are initially reactive in nature and are usually reversible if underlying pathologic processes (eg, congestive heart failure and hypoxia) are corrected early in their course [15]. Persistent elevations in PAP, however, are associated with an increased risk of right heart failure and death after cardiac transplantation [2, 3] and LVAD placement [4]. Accordingly, reversibility of pulmonary hypertension with vasodilators such as sodium nitroprusside and most recently by inhaled NO [16] is generally considered to be a requirement in the selection of candidates for cardiac transplantation (and therefore LVAD recipients) [17]. Right ventricular failure refractory to pharmacologic therapy occurs in 20% to 40% of patients supported with an LVAD [4, 18, 19]. Left ventricular assist device support induces a variety of hemodynamic changes with important effects on right ventricular physiology and function. Improvements in systemic perfusion result in augmented venous return to the right ventricle, which increases preload but may precipitate right ventricular failure because of volume overload [18]. In addition, although left ventricular unloading may passively reduce right ventricular afterload by decreasing post pulmonary capillary pressures [20], resultant acute increases in pulmonary blood flow may transiently increase PVR and afterload [18]. Finally, although LVAD support may improve right ventricular systolic function by increasing coronary perfusion, the disturbances to ventricular interaction caused by leftward displacement of the interventricular septum and reduction of left ventricular pressure generation result in a net decrease in right ventricular contractility [21]. Medical management of right ventricular failure in LVAD recipients includes the provision of adequate right ventricular preload and the maintenance of sinus rhythm [22]. Volume infusion is usually limited, however, by the risk of volume overload and right ventricular decompensation. Although pharmacologic management with inotropic and vasodilatory drugs is often successful [23], the effectiveness of these agents may be limited by an increased incidence of cardiac arrhythmias, profound systemic hypotension, and derangement of alveolar gas exchange by inhibition of hypoxic vasoconstriction [5]. Inhaled NO is an extremely effective, specific pulmonary vasodilator in animals, infants, and adults with pulmonary hypertension [9 12]. After easily crossing the alveolus, NO is absorbed into the bloodstream, rapidly bound to hemoglobin, and converted into relatively inactive nitrate and nitrite by the enzyme methemoglobin reductase. At clinically useful concentrations, fewer than 80 ppm, there does not seem to be systemic vasodilatation because the inhaled gas affects only the vascular smooth muscle subjacent to ventilated bronchioles and alveoli. In addition, because the vasodilatory effect of inhaled NO gas is limited to ventilated areas of lung, it selectively dilates vessels in well-ventilated lung regions, thus improving ventilation-perfusion matching. The present study identified 11 patients undergoing LVAD placement for end-stage heart failure who manifested hemodynamically significant elevations in PVR and signs of right ventricular failure despite maximal medical therapy on weaning from cardiopulmonary bypass. Patients randomized to receive inhaled NO at a concentration of 20 ppm demonstrated rapid, significant reductions in PVR that were manifested as decreases in PAP and increases in LVAD assisted cardiac output. Patients receiving N 2 placebo showed no response but demonstrated impressive hemodynamic improvement when crossed over to the inhaled NO group. The beneficial effects of inhaled NO were not associated with systemic hypotension, hypoxia, or other adverse consequences. Importantly, all patients were successfully weaned from inhaled NO in less than 1 week. The vasodilator acted as a hemodynamic bridge while reactive elevations in PVR and temporary right ventricular Table 3. Hemodynamics of 11 Recipients of Left Ventricular Assist Device Before and After Treatment With Inhaled Nitric Oxide a,b Variable Before NO I NO I After MPAP (mm Hg) c LVAD flow index (L min 1 m 2 ) d MAP (mm Hg) a Six patients were initially randomized to NO I, and 5 patients crossed over from nitrogen to NO I. b Data are shown as the mean the standard error of the mean. c Significance: p d Significance: p LVAD left ventricular assist device; MAP mean arterial pressure; MPAP mean pulmonary artery pressure; NO I inhaled nitric oxide.

5 344 ARGENZIANO ET AL Ann Thorac Surg INHALED NO IN LVAD RECIPIENTS 1998;65:340 5 dysfunction gradually improved. The use of inhaled NO averted the need of right ventricular assist device insertion in all but 1 patient, and in this patient, hemodynamic collapse resulted from abrupt discontinuation of this agent. Prior to the availability of inhaled NO at our center, 4 of 19 LVAD recipients with right ventricular failure required right ventricular assist device support, with an associated mortality rate of 50% [6]. In conclusion, inhaled NO is a potent and specific pulmonary vasodilator. When administered to patients receiving LVADs, it results in significant improvements in PVR and right ventricular function without adverse sequelae. Because pulmonary hypertension is a risk factor for the development of right ventricular failure and death in patients undergoing orthotopic cardiac transplantation, mitral valve operations, and other cardiac operations, further study of this agent as an intraoperative adjunct in the management of cardiac surgical patients is warranted. Doctor Oz is an Irving Scholar of Columbia University. References 1. Francis GS, Goldsmith SR, Levine TB, Olivari MT, Cohn JN. The neurohumoral axis in congestive heart failure. Ann Intern Med 1984;101: Kirklin J, Naftel D, Kirklin J, Blackstone E, White-Williams C, Bourge R. Pulmonary vascular resistance and the risk of heart transplantation. J Heart Transplant 1988;7: Addonizio L, Gersony W, Robbins R. Elevated pulmonary vascular resistance and cardiac transplantation. Circulation 1987;76 (Suppl 5): Frazier OH, Rose EA, Macmanus Q, et al. Multicenter clinical evaluation of the HeartMate 1000 IP left ventricular assist device. Ann Thorac Surg 1992;53: Rademacher PA, Santak B, Becker H. Prostaglandin E 1 and nitroglycerin reduce pulmonary capillary pressure but worsen ventilation-perfusion distributions in patients with adult respiratory distress syndrome. Anesthesiology 1989;70: Oz MC, Argenziano M, Catanese KA, et al. Bridge experience with long-term implantable left ventricular devices: are they an alternative to transplantation? Circulation 1997;95: Oz MC, Slater JP, Edwards N, et al. Desaturated venous-toarterial shunting reduces right-sided heart failure after cardiopulmonary bypass. J Heart Lung Transplant 1995;14: Ignarro L, Lippton H, Edwards J. Mechanism of vascular smooth muscle relaxation by organic nitrates, nitrites, nitroprusside and nitric oxide: evidence for involvement of S- nitrosothiols as active intermediates. J Pharmacol Exp Ther 1981;218: Pepke-Zaba J, Higgenbottam TW, Dinh Xuan AT, Stone D, Wallwork J. Inhaled nitric oxide as a cause of selective pulmonary vasodilatation in pulmonary hypertension. Lancet 1991;338: Roberts JD Jr, Lang P, Bigatello LM, Vlahakes GJ, Zapol WM. Inhaled nitric oxide in congenital heart disease. Circulation 1993;87: Rossaint R, Falke KJ, Lopez F, Slama K, Pison U, Zapol WM. Inhaled nitric oxide for the adult respiratory distress syndrome. N Engl J Med 1993;328: Snow DJ, Gray SJ, Ghosh S. Inhaled nitric oxide in patients with normal and increased pulmonary vascular resistance after cardiac surgery. Br J Anaesth 1994;72: Dinh Xuan AT, Higgenbottam TW, Clelland C, Pepke-Zaba J, McGoldrick J, Wallwork J. Impairment of endotheliumdependent vasorelaxation in secondary pulmonary hypertension. Eur Respir J 1989;2:667S. 14. Komai H, Adatia IT, Elliott MJ, de Leval MR, Haworth SG. Increased plasma levels of endothelin-1 after cardiopulmonary bypass in patients with pulmonary hypertension and congenital heart disease. J Thorac Cardiovasc Surg 1993;106: Bhatia S, Kirshembaum J, Shemin R. Time course of resolution of pulmonary hypertension and right ventricular remodeling after orthotopic cardiac transplantation. Circulation 1987;76: Thompson ME. Selection of candidates for cardiac transplantation. J Heart Transplant 1983;3: Rimar S, Gillis CN. Selective pulmonary vasodilation by inhaled nitric oxide is due to hemoglobin inactivation. Circulation 1993;88: Farrar DJ, Compton PG, Hershon JJ, Fonger JD, Hill JD. Right heart interaction with the mechanically assisted left heart. World J Surg 1985;9: Elbeery JR, Owen CH, Savitt MA. Effects of the left ventricular assist device on right ventricular function. J Thorac Cardiovasc Surg 1990;99: Farrar DJ, Compton PG, Dajee H, Fonger JD, Hill JD. Right heart function during left heart assist and its effects on volume loading in a canine preparation. Circulation 1984;70: Woodard JC, Chow E, Farrar DJ. Isolated ventricular systolic interaction during transient reductions in left ventricular pressure. Circ Res 1992;70: Higgins RSD, Elefteriades JA. Right ventricular assist devices and the surgical treatment of right ventricular failure. Cardiol Clin 1992;10: Kieler-Jensen N, Lundin S, Ricksten SE. Vasodilator therapy after heart transplantation: effects of inhaled nitric oxide and intravenous prostacyclin, prostaglandin E 1, and sodium nitroprusside. J Heart Lung Transplant 1995;14: DISCUSSION DR ROLAND HETZER (Berlin, Germany): I think this is a very important study, and I congratulate you. We have been using nitric oxide (NO) in our assist device patients routinely during the last 3 years with an annual rate of between 40 and 60 implantations. However, there are still patients for whom NO alone is not sufficient to overcome the transient pulmonary hypertension, in particular patients with frank pulmonary edema and acidosis. This is why in such patients we have returned to using biventricular assist devices primarily. Have you found any criteria for selecting patients who are in extremely advanced heart failure whom you would not include in the left ventricular assist device and NO group? DR ARGENZIANO: I appreciate your comments and am happy to hear that others are using NO as successfully as we are. With respect to the efficacy of the agent, one certain contraindication to the use of inhaled NO is, of course, fixed pulmonary hypertension. Because our patients with a left ventricular assist device, at least currently, are required to be heart transplant candidates, they must have at least a partially reversible

6 Ann Thorac Surg ARGENZIANO ET AL 1998;65:340 5 INHALED NO IN LVAD RECIPIENTS 345 component of pulmonary hypertension. For that reason, we have had success with inhaled NO. Groups that have tried to use inhaled NO in patients with more fixed and more prolonged causes of pulmonary hypertension have had less success because, understandably, the fixed variety of this disease has a structural component to it. With respect to pulmonary edema with decreased alveolar ventilation, there is decreased delivery of inhaled NO to the subalveolar capillary membrane. I suspect that the use of techniques developed to ventilate patients with oxygen who have pulmonary edema, such as increasing the inspired concentration, may allow sufficient quantities of the agent to arrive at the subalveolar membrane. We have not encountered such a patient, however. Perhaps patients in pulmonary edema will require higher concentrations of NO, maybe in the range of 100 ppm. DR LUDWIG K. VON SEGESSER (Lausanne, Switzerland): I congratulate you on this important work and have two questions. First, in regard to hypoxia, have all patients been under 100% oxygen before application of NO? Second, why did you select 20 ppm as the NO level to treat these patients? We know that normally 40 ppm is the optimal value for adult patients. DR ARGENZIANO: Thank you very much for your comments. To answer the second question first, with respect to the choice of concentration, the literature is replete with reports of the use of NO in concentrations ranging from 1 to 2 ppm up to 80 ppm. We found that in the cardiac surgical population, the lowest efficacious dose was 20 ppm in most cases, and we elected to use the lowest concentration that we could. We found that effective pulmonary vasodilatation was achieved in all patients at that dose. Could we have obtained more vasodilation with a higher dose? Perhaps. But because our goal was to get the patients out of the operating room and to manage their pulmonary hypertension acutely, we thought that a dose that maintained their stability was sufficient. With respect to your first question, we routinely treat our patients with 100% oxygen in the operating room. Consequently, when we infuse inhaled NO through the ventilator circuit, this requires that the inspired oxygen fraction be reduced because the total gas delivery is a combination of the oxygen and the inhaled NO. Although we start with an inspired oxygen fraction of 100%, it settles out at about 85% to 90% after we begin administration of inhaled NO. DR BARTLEY P. GRIFFITH (Pittsburgh, PA): Have you tried this preoperatively either to avoid use of the left ventricular assist device or to predict which patients would benefit the most postoperatively? Could you be a little more specific about your weaning cycle? Your problem with the patient with a right ventricular assist device is not infrequent, and it is difficult sometimes to decide whether to insert a device or continue with inhaled NO. DR ARGENZIANO: We have not used inhaled NO preoperatively to manage our patients because in general, left-sided failure has been the presenting manifestation. We have found that inhaled NO is useful after the restitution of cardiac output by the device in fact brings about a number of well-described physiologic challenges to the right ventricle. With respect to assessing patients preoperatively, a number of institutions are now using inhaled NO to assess the reversibility of pulmonary hypertension in the evaluation of cardiac transplant candidates. Our institution has not yet begun this practice. In regard to the NO delivery and the weaning cycle, we begin at 20 ppm, and then at 6-hour intervals postoperatively, we reduce the inhaled concentration to 15, 10, and 5 ppm using pulmonary artery pressures and left ventricular assist device flows as our guides. When we get to about 5 ppm and have hemodynamic stability, we turn the NO off. INVITED COMMENTARY I was pleased to be asked to comment on this interesting study of the use of inhaled nitric oxide in left ventricular assist device (LVAD) recipients with pulmonary hypertension. Preliminary reports [1, 2] suggested a beneficial effect allowing maintenance of the pulmonary circulation without undue difficulty in the case of acute right ventricular failure after LVAD implantation. In their well-controlled randomized, double-blind study Dr Argenziano and his associates demonstrated clearly the value of inhaled nitric oxide in resolving the problem of insufficient LVAD filling caused by acute pulmonary hypertension after rapid right ventricular failure. Moreover, they showed that use of inhaled nitric oxide can result in a decreased need of a right ventricular assist device. Only 1 of 11 LVAD recipients with right ventricular failure required a right ventricular assist device after administration of inhaled nitric oxide versus 4 of 19 such patients before the availability of this vasodilator. I believe this may prove to be a landmark study demonstrating the feasibility of applying this still controversial therapy to right ventricular failure after LVAD implantation. Naoki Yahagi, MD Surgical Intensive Care Unit National Cardiovascular Center Suita, Osaka 565, Japan References 1. Mertes PM, Pinelli G, Hubert T, et al. Impact of nitric oxide inhalation on right ventricular failure after implantation of Novacor left ventricular assist system. J Thorac Cardiovasc Surg 1995;109: Yahagi N, Kumon K, Nakatani T, et al. Inhaled nitric oxide for the management of acute right ventricular failure in patients with a left ventricular assist system. Artif Organs 1995;19: by The Society of Thoracic Surgeons /98/$19.00 Published by Elsevier Science Inc PII S (97)