Cardiopulmonary Support and Physiology
`
`Hache´ et al
`
`Inhaled epoprostenol (prostacyclin) and pulmonary
`hypertension before cardiac surgery
`
`Manon Hache´, MD, Andre´ Denault, MD, FRCPC, Sylvain Be´lisle, MD, FRCPC, Danielle Robitaille, MD, FRCPC,
`Pierre Couture, MD, FRCPC, Peter Sheridan MD, FRCPC, Michel Pellerin, MD, FRCSC, Denis Babin, MSc,
`Nicolas Noe¨l, BPharm, MSc, Marie-Claude Guertin, MSc, PhD, Raymond Martineau, MD, FRCPC, and
`Jocelyn Dupuis, MD, FRCPC
`
`Objective: Pulmonary hypertension is commonly found in patients undergoing
`valvular surgery and can be worsened by cardiopulmonary bypass. Inhaled epopro-
`stenol (prostacyclin) has been used for the treatment of pulmonary hypertension, but
`its effects compared with those of placebo on hemodynamics, oxygenation, echo-
`cardiographic examination, and platelet function have not been studied during
`cardiac surgery.
`
`Methods: Twenty patients with pulmonary hypertension undergoing cardiac surgery
`were randomized in a double-blind study to receive inhaled epoprostenol (60 ␮g) or
`placebo. The inhalation occurred after induction of anesthesia and before surgical
`incision. The effects on left and right systolic and diastolic cardiac functions
`evaluated by means of pulmonary artery catheterization and transesophageal echo-
`cardiography, as well as oxygenation and platelet aggregation, were studied.
`
`Results: Inhalation of epoprostenol significantly reduced indexed right ventricular
`stroke work from 10.7 ⫾ 4.57 g 䡠 m 䡠 m⫺2 to 7.8 ⫾ 3.94 g 䡠 m 䡠 m⫺2 (P ⫽ .003) and
`systolic pulmonary artery pressure from 48.4 ⫾ 18 mm Hg to 38.9 ⫾ 11.9 mm Hg
`(P ⫽ .002). The effect was correlated with the severity of pulmonary hypertension
`(r ⫽ 0.76, P ⫽ .01) and was no longer apparent after 25 minutes. There was no
`significant effect on systemic arterial pressures, left ventricular function, arterial
`oxygenation, platelet aggregation, and surgical blood loss.
`
`Conclusion: Inhaled epoprostenol reduces pulmonary pressure and improves right
`ventricular stroke work in patients with pulmonary hypertension undergoing cardiac
`surgery. A dose of 60 ␮g is hemodynamically safe, and its effect is completely
`reversed after 25 minutes. We did not observe any evidence of platelet dysfunction
`or an increase in surgical bleeding after administration of inhaled epoprostenol.
`
`Pulmonary hypertension (PH) is associated with an increase in mor-
`
`bidity and mortality in patients undergoing cardiac surgery.1 Many
`drugs have been used in recent years to treat PH. Among these are
`vasodilators, including epoprostenol (prostacyclin; PGI2). However,
`its intravenous administration is limited by systemic hypotension
`because of nonselective vasodilation and by hypoxemia through wors-
`ening of intrapulmonary shunt caused by inhibition of hypoxic pulmonary vaso-
`constriction.2,3
`Inhaled PGI2 appears to be a selective pulmonary vasodilator comparable with
`inhaled nitric oxide (iNO) but acting through cyclic adenosine monophosphate
`instead of cyclic guanosine monophosphate.4,5 Its administration can be a simpler
`and less expensive alternative to iNO. Its half-life is 2 to 3 minutes, and at
`physiologic pH, it spontaneously hydrolyses to 6-ketoprostaglandin F1␣ (6-keto-
`PGF1␣). Thus its effect remains localized to ventilated lung units, it can decrease
`pulmonary artery pressure (PAP) without causing systemic hypotension and im-
`
`Dr Hache´
`
`Dr Denault
`
`CSP
`
`From the Montreal Heart Institute, Mon-
`treal, Quebec, Canada.
`
`Received for publication Dec 3, 2001; ac-
`cepted for publication July 16, 2002.
`
`Address for reprints: Andre´ Denault, MD,
`Research Center, Montreal Heart Institute,
`5000 Belanger St East, Montreal, Quebec,
`H1T 1C8 Canada
`(E-mail: denault@
`videotron.ca).
`
`J Thorac Cardiovasc Surg 2003;125:642-9
`
`Copyright © 2003 by The American Asso-
`ciation for Thoracic Surgery
`0022-5223/2003 $30.00⫹0
`doi:10.1067/mtc.2003.107
`
`642 The Journal of Thoracic and Cardiovascular Surgery ● March 2003
`
`Liquidia's Exhibit 1024
`Page 1
`
`

`

`Hache´ et al
`
`Cardiopulmonary Support and Physiology
`
`CSP
`
`prove oxygenation by decreasing ventilation-perfusion mis-
`match.5-8 Its effect on cardiac function when given by
`means of inhalation is controversial but it can increase
`cardiac output when given intravenously.9,10
`Finally, a drawback of PGI2 is that it has been reported
`to alter platelet function,11 which could be hazardous during
`cardiac surgery.
`We have previously reported our retrospective experi-
`ence with the use of inhaled PGI2.12 However, so far no
`study has evaluated its effects in patients undergoing car-
`diac surgery and simultaneously on several important clin-
`ical variables, such as the magnitude of its hemodynamic
`effect and its consequences on echocardiographic indices of
`right ventricular (RV) and left ventricular (LV) systolic and
`diastolic functions, oxygenation, platelet
`function, and
`bleeding.
`
`Methods
`Population
`After approval by the research and ethics committee and obtaining
`informed consent, 20 patients with PH undergoing cardiac surgery
`with cardiopulmonary bypass were included in the study. Patients
`were considered to have PH if systolic pulmonary artery pressure
`(sPAP) was greater than 30 mm Hg or mean pulmonary artery
`pressure was greater than 25 mm Hg, as measured during the
`preoperative period or estimated by using Doppler echocardiogra-
`phy.13 This was confirmed after insertion of a pulmonary artery
`catheter and before induction of general anesthesia. Patients with
`LV dysfunction (ejection fraction of ⬍30%) or known bleeding
`diathesis were excluded. Further exclusion criteria were contrain-
`dications to transesophageal echocardiography (TEE), including
`esophageal disease or unstable cervical spine. A Parsonnet score
`was calculated for every patient.1
`
`Protocol
`Patients were premedicated with 1 to 2 mg of lorazepam admin-
`istered orally 1 hour before the operation, as well as 0.1 mg/kg
`morphine administered intramuscularly and 0.2 to 0.4 mg of sco-
`polamine administered intramuscularly before being taken to the
`operating room. In the operating room additional midazolam was
`added (0.01-0.05 mg/kg administered intravenously) as needed for
`patient comfort. Usual monitoring was installed, including a 5-lead
`electrocardiogram, pulse oximeter, peripheral venous line, radial
`arterial line, 15-cm 3-lumen catheter (CS-12703, Arrow Interna-
`tional Inc, Reading, Calif), and fast-response thermodilution pul-
`monary artery catheter (Swan-Ganz catheter 7.5F; Baxter Health-
`care Corporation, Irvine, Calif). Anesthesia was induced with 0.04
`mg/kg midazolam and 1 ␮g/kg sufentanil, and muscle relaxation
`was achieved with 0.1 mg/kg pancuronium. After tracheal intuba-
`tion, anesthesia was maintained with 1 ␮g ⫻ kg⫺1 ⫻ h⫺1 sufen-
`tanil and 0.04 mg ⫻ kg⫺1 ⫻ h⫺1 midazolam. No anesthetic gases
`were used. Minute ventilation was adjusted to maintain end-tidal
`carbon dioxide between 30 and 40 mm Hg with an infrared carbon
`dioxide analyzer. A 5.0-MHz TEE omniplane probe (Hewlett-
`Packard Sonos 5500, Andover, Mass) was inserted after induction
`of general anesthesia.
`
`Drug Administration Protocol
`Patients were equally divided into 2 groups to receive either
`inhaled PGI2 or placebo in a double-blind randomized manner by
`using a computer-generated randomization table. Epoprostenol
`(Flolan; Glaxo-Wellcome Inc, Mississauga, Ontario, Canada) was
`given as epoprostenol 1.5 mg of salt dissolved in sterile glycine
`buffer diluent, for a concentration of 15 ␮g/mL. Each patient
`received 4 mL of a solution containing either PGI2 or normal
`saline solution (placebo).
`The study drug was administered through a jet nebulizer (Ref
`8901; Salter Labs, Arvin, Calif) attached to the inspiratory limb of
`the ventilator near
`the endotracheal
`tube. Nebulization was
`achieved with a bypass flow of oxygen at 8 L/min. This high flow
`was used to achieve a high proportion of small particles (⬍5 ␮m).
`Because this added a secondary flow to the patient, minute venti-
`lation was adjusted to maintain peak inspiratory pressures of less
`than 30 cm H2O and a normal end-tidal carbon dioxide.
`
`Measurements
`Hemodynamic parameters included central venous pressure, PAP,
`and pulmonary artery occlusion pressure. Cardiac output was
`assessed by using the thermodilution technique with 3 injections of
`room temperature dextrose 5% (10 mL) at end expiration. Sys-
`temic vascular resistances, pulmonary vascular resistances, RV
`stroke work, and LV stroke work were calculated by using a
`standard formula. Hemodynamic values were indexed for patient
`body surface area.
`Arterial and mixed venous blood gases were obtained to mea-
`⫺, and SO2.
`sure pH, PO2, PCO2, HCO3
`TEE examination was performed to evaluate systolic and dia-
`stolic parameters of LV and RV performance. The TEE examina-
`tion included a midesophageal 4-chamber view, a short-axis trans-
`gastric view at the midpapillary level, and color flow Doppler
`imaging of the mitral valve to detect any unsuspected significant
`mitral valvulopathy. We first obtained a baseline transgastric short-
`axis view of the left ventricle at the midpapillary level, followed by
`a pulsed Doppler examination of pulmonary venous flow, trans-
`mitral flow, transtricuspid flow, and hepatic venous flow. The
`Doppler sample volume (2-mm width) was positioned in the left
`upper pulmonary vein approximately 1 cm proximal to its entrance
`into the left atrium to measure pulmonary venous flow by using
`color Doppler flow to sample maximal flow. When necessary, to
`minimize the angle between the Doppler beam and the pulmonary
`vein’s long axis, we rotated the omniplane probe as far as needed
`from the horizontal plane. This axis was maintained throughout the
`examination. The same approach was used for hepatic venous
`flow. Mitral and tricuspid inflow velocities were measured at the
`tip of the atrioventricular valve leaflets. Three signals were ob-
`tained, and the maximal value was computed for analysis.14 Two
`independent observers were involved: the first one recorded he-
`modynamic parameters, and the other, blinded to the hemody-
`namic data and to the study drug, simultaneously recorded the
`pulsed Doppler and 2-dimensional echocardiographic images. All
`TEE examinations were performed by anesthesiologists who were
`not in charge of the patient. After data recording, a third anesthe-
`siologist blinded to all data reviewed the recorded sequence. All
`2-dimensional images in which the LV and RV endocardial border
`could not be traced adequately by using Schnittger criteria, in
`
`The Journal of Thoracic and Cardiovascular Surgery ● Volume 125, Number 3 643
`
`Liquidia's Exhibit 1024
`Page 2
`
`

`

`Cardiopulmonary Support and Physiology
`
`Hache´ et al
`
`TABLE 1. Demographic data of the study population
`n
`Epoprostenol
`n
`Placebo
`P value
`
`Age, y
`Sex
`Male
`Female
`Weight, kg
`Height, m
`Parsonnet score
`Reoperation
`Aspirin use
`Preoperative heparin
`
`10
`
`5
`5
`10
`10
`10
`3
`0
`3
`
`65 ⫾ 11
`
`10
`
`58 ⫾ 12
`
`50%
`50%
`76 ⫾ 17
`1.65 ⫾ 0.11
`29.2 ⫾ 6.5
`30%
`0%
`30%
`
`40%
`4
`60%
`6
`70 ⫾ 20
`10
`10 1.63 ⫾ 0.08
`10 32.4 ⫾ 10.2
`5
`50%
`3
`30%
`3
`30%
`
`.22
`
`1
`
`.56
`.65
`.4
`.65
`.21
`1
`
`which 80% of the endocardial contour has to be visualized, were
`excluded.15 In addition, the Doppler signals were reviewed and
`rejected if they were not laminar and when a clear contour could
`not be determined for quantification of the velocity-time integral.
`Severe mitral stenosis or regurgitation were exclusion criteria for
`the measurement of mitral inflow. All the anesthesiologists per-
`forming the TEE measurements were board certified in perioper-
`ative TEE. If disagreement occurred between 2 reviewers, a third
`echocardiographer was asked to review the echocardiographic
`sequence. Our experience and interobserver variability in the mea-
`surement of systolic and diastolic function has been published
`previously.16,17
`Platelet aggregation studies were performed on whole blood by
`using a Chronolog 560 whole blood lumi-aggregometer (Chro-
`nolog Corp, Havertown, Pa). Sodium citrate, 0.5 mL of a 3.2%
`solution, was added to 4.5 mL of venous blood. The citrated blood
`was diluted 1:1 with normal saline solution. After the solution had
`been cooled to 20°C to 25°C, a 900-␮L sample was placed in a
`cuvette containing a silicone stir bar. After 3 minutes, 100 ␮L of
`chrono-lume (luciferase luciferon reagent, Chronolog Corp) was
`added. After another 2 minutes, 2 mmol/L of adenosine triphos-
`phate (ATP) was added. The ATP standard was then measured for
`each patient. Two minutes later, an aggregant was added (20
`␮mol/L adenosine diphosphate, 5 ␮g/mL collagen, or 1 nmol/L
`arachidonic acid), and platelet aggregation and ATP release (lu-
`minescence) were measured. Blood loss was measured for the
`intraoperative period, as well as for the first 24 hours postopera-
`tively.
`Hemodynamic parameters were measured before (T1) and 10
`minutes after (T2) induction of anesthesia, after nebulization of
`PGI2 or placebo (T3), and 15 (T4) and 25 (T5) minutes after
`nebulization. Arterial and mixed venous blood gases were obtained
`at the same times, except for T5. TEE examination and platelet
`aggregation studies were performed before and after administra-
`tion of PGI2 or placebo. Patients were observed until discharge
`from the intensive care unit.
`
`Statistical Analysis
`Population size was calculated for a power of 80% and an ␣ error
`of .05, assuming an sPAP of 40 ⫾ 4 mm Hg to decrease by 20%
`in the PGI2 group and remain stable in the placebo group.
`Continuous variables were analyzed with the Student t test and
`categoric variables with the ␹2 or Fisher exact test. Two-factor
`(time and group) repeated-measures analysis of variance was used
`
`to determine time variations between the 2 groups. In case of
`significant interaction, time ⫻ group comparison was performed
`with Bonferroni corrections.
`The Pearson correlation test was performed to determine the
`relationship between the level of sPAP and the degree of reduction
`of sPAP after PGI2 and placebo administration.
`
`Results
`Twenty-seven patients were enrolled in the study. Four were
`later not randomized because they failed to meet the inclu-
`sion criteria for PH on arrival to the operating room. Two
`were excluded because the operation was subsequently re-
`scheduled and another because of agitation. Demographic
`variables were similar in both groups (Table 1).
`
`Hemodynamics
`Baseline hemodynamic variables were similar between the
`2 groups (Table 2). Baseline sPAP was 61.2 ⫾ 19 mm Hg
`in the PGI2 group and 54.3 ⫾ 14.6 mm Hg in the placebo
`group (P ⫽ .4). These decreased significantly after induc-
`tion of anesthesia in both groups to 48.4 ⫾ 18 mm Hg and
`42.7 ⫾ 12.8 mm Hg, respectively. After administration of
`the study drug, a decrease was noted in the PGI2 group to
`38.9 ⫾ 11.9 mm Hg (P ⫽ .002), as opposed to that seen in
`the placebo group. Fifteen minutes after the end of nebuli-
`zation, these values were stable (42 ⫾ 12.4 vs 43.6 ⫾ 12.5
`mm Hg). After 25 minutes, they returned to baseline in the
`PGI2 group to 53.3 ⫾ 17.6 mm Hg and remained stable in
`the placebo group at 43.5 ⫾ 13 mm Hg (Figure 1). There
`was a significant positive correlation between the severity of
`PH before the administration of inhaled PGI2 and the mag-
`nitude of the decrease in sPAP (r ⫽ 0.76, P ⫽ .01). Heart
`rate decreased in the PGI2 group from 64.4 ⫾ 8.8 to 58.5 ⫾
`11 beats/min (P ⫽ .002). It stayed stable thereafter until 25
`minutes after inhaled PGI2, when it increased to 62.8 ⫾ 11.6
`beats/min (P ⫽ .001). Mean PAP did not change signifi-
`cantly after PGI2 administration nor did systemic arterial
`pressures. Cardiac indexes remained unchanged throughout
`the study. Compared with placebo, indexed RV stroke work
`decreased after PGI2 inhalation from 10.7 ⫾ 4.57 to 7.8 ⫾
`3.94 g 䡠 m 䡠 m⫺2 (P ⫽ .003) and remained stable thereafter
`for 25 minutes.
`
`Oxygenation
`Oxygenation data are given in Table 3. The Pao2 value at
`baseline was significantly different on arrival in the operat-
`ing room with a nasal cannula at 4 L/min of oxygen
`(173.1 ⫾ 65.2 mm Hg in the PGI2 group vs 251.5 ⫾ 95.4
`mm Hg in the placebo group, P ⫽ .05), but this difference
`disappeared after induction of anesthesia with a fraction of
`inspired oxygen of 100% (428.2 ⫾ 65.2 vs 443.7 ⫾ 64.9,
`respectively; P ⫽ .6). Otherwise, oxygenation variables did
`not change throughout the study.
`
`644 The Journal of Thoracic and Cardiovascular Surgery ● March 2003
`
`CSP
`
`Liquidia's Exhibit 1024
`Page 3
`
`

`

`Hache´ et al
`
`Cardiopulmonary Support and Physiology
`
`TABLE 2. Hemodynamic variations of the population throughout the study
`T1
`T2
`
`T3
`
`T4
`
`T5
`
`Epoprostenol
`HR (beats/min)
`SAP (mm Hg)
`MAP (mm Hg)
`CVP (mm Hg)
`PAOP (mm Hg)
`sPAP (mm Hg)
`MPAP (mm Hg)
`CI (L 䡠 min⫺1 䡠 m⫺2)
`EV (mL 䡠 beat⫺1 䡠 m⫺2)
`SVRI (dyne 䡠 s 䡠 cm⫺5)
`PVRI (dyne 䡠 s 䡠 cm⫺5)
`LVSWI (g 䡠 m 䡠 m⫺2)
`RVSWI (g 䡠 m 䡠 m⫺2)
`Placebo
`HR (beats/min)
`SAP (mm Hg)
`MAP (mm Hg)
`CVP (mm Hg)
`PAOP (mm Hg)
`sPAP (mm Hg)
`MPAP (mm Hg)
`CI ((L 䡠 min⫺1 䡠 m⫺2)
`EV (mL 䡠 beat⫺1 䡠 m⫺2)
`SVRI (dyne 䡠 s 䡠 cm⫺5)
`PVRI (dyne 䡠 s 䡠 cm⫺5)
`LVSWI (g 䡠 m 䡠 m⫺2)
`RVSWI (g 䡠 m 䡠 m⫺2)
`
`66.1 ⫾ 9.7
`134.4 ⫾ 28.6
`85.9 ⫾ 14.6
`11.1 ⫾ 6.8
`22.8 ⫾ 8.1
`61.2 ⫾ 19
`41.5 ⫾ 9
`2.5 ⫾ 0.7
`69.4 ⫾ 20
`2474 ⫾ 424.2
`683.5 ⫾ 449.1
`33.1 ⫾ 16.1
`15.5 ⫾ 6.3
`
`67.8 ⫾ 12.5
`142.2 ⫾ 26.4
`92 ⫾ 14.2
`14.4 ⫾ 7.1
`21.7 ⫾ 11.3
`54.3 ⫾ 14.6
`38.4 ⫾ 9.9
`2.8 ⫾ 0.8
`73.2 ⫾ 15.7
`2282 ⫾ 635.2
`521.7 ⫾ 310
`40.4 ⫾ 18.4
`13.4 ⫾ 6.2
`
`64.4 ⫾ 8.8
`105.3 ⫾ 15
`67.9 ⫾ 11.5
`11 ⫾ 4.9
`21.8 ⫾ 6.6
`48.4 ⫾ 18§
`32.9 ⫾ 9.2§
`2.2 ⫾ 0.6
`65.4 ⫾ 17.3
`2078 ⫾ 566.6
`451.7 ⫾ 315.6
`21.7 ⫾ 8.6
`10.7 ⫾ 4.6§
`
`68.5 ⫾ 14.4
`118.4 ⫾ 24.1
`77.3 ⫾ 13.6
`11.3 ⫾ 5.2‡
`19.7 ⫾ 8
`42.7 ⫾ 12.8‡
`30 ⫾ 9.6‡
`2.7 ⫾ 0.8
`67.7 ⫾ 18.9
`2128 ⫾ 708.3
`311.8 ⫾ 147.1
`30.3 ⫾ 11.7
`9.5 ⫾ 4.9‡
`
`58.5 ⫾ 11*
`101.2 ⫾ 19.3
`63.8 ⫾ 10.5
`10.8 ⫾ 4.8
`19 ⫾ 6.4
`38.9 ⫾ 11.9¶
`28.2 ⫾ 8.2
`2 ⫾ 0.5
`62.9 ⫾ 17.6
`2260 ⫾ 381.3
`404.9 ⫾ 250
`22.5 ⫾ 10.8
`7.8 ⫾ 3.9#
`
`65.5 ⫾ 16
`119.6 ⫾ 23.1
`78.4 ⫾ 14.3
`11.7 ⫾ 4.7
`19.9 ⫾ 8.4
`41.3 ⫾ 15.2
`29.6 ⫾ 13.4
`2.5 ⫾ 0.7
`68.5 ⫾ 20.2
`2290 ⫾ 776.4
`299.5 ⫾ 161.8
`31 ⫾ 9.6
`8.8 ⫾ 5
`
`56.7 ⫾ 10.6
`108.6 ⫾ 22
`69.7 ⫾ 13.7
`11.5 ⫾ 4.9
`20.6 ⫾ 6.6
`42 ⫾ 12.4
`30.4 ⫾ 8.4
`2 ⫾ 0.6
`66 ⫾ 19.2
`2460 ⫾ 719.7
`398.2 ⫾ 266.6
`24.3 ⫾ 11.6
`8.3 ⫾ 4.2
`
`61.4 ⫾ 14.8
`117.4 ⫾ 24.6
`76.7 ⫾ 16.1
`12.6 ⫾ 4.1
`21.1 ⫾ 6.5
`43.6 ⫾ 12.5
`30.4 ⫾ 9.1
`2.4 ⫾ 0.7
`68.1 ⫾ 18.5
`2497 ⫾ 786.8
`352.9 ⫾ 208.6
`32 ⫾ 11.9
`9.5 ⫾ 5
`
`62.8 ⫾ 11.6†
`124.9 ⫾ 28
`82.6 ⫾ 19
`13.3 ⫾ 4.3
`24.6 ⫾ 9
`53.3 ⫾ 17.6㛳
`36 ⫾ 11.1㛳
`1.9 ⫾ 0.5
`56.7 ⫾ 17.7
`3157 ⫾ 1249
`555.4 ⫾ 347.3
`24.6 ⫾ 12.3
`8.6 ⫾ 3.1
`
`61.5 ⫾ 11.7
`119.1 ⫾ 22.1
`78.6 ⫾ 11
`12.9 ⫾ 4.6
`21.1 ⫾ 5.8
`43.5 ⫾ 13
`29.6 ⫾ 7.6
`2.2 ⫾ 0.6
`63.6 ⫾ 15.4
`2492 ⫾ 924.3
`324.5 ⫾ 276.7
`27.2 ⫾ 6.6
`7.8 ⫾ 3.9
`
`CSP
`
`HR, Heart rate; SAP, systemic arterial pressure; MAP, mean arterial pressure; CVP, central venous pressure; PAOP, pulmonary artery occlusion pressure;
`MPAP, mean pulmonary arterial pressure; CI, cardiac index; EV, ejection volume; SVRI, systemic vascular resistance index; PVRI, pulmonary vascular
`resistance index; LVSWI, LV stroke work index; RVSWI, RV stroke work index.
`*P ⫽ .002 T2 to T3 in prostacyclin group.
`†P ⫽ .001 T4 to T5 in prostacyclin group.
`‡P ⬍ .001 T1 to T2 in placebo group.
`§P ⬍ .001 T1 to T2 in prostacyclin group.
`㛳P ⬍ .001 T3 to T5 in prostacyclin group.
`¶P ⫽ .002 T2 to T3 in prostacyclin group.
`#P ⫽ .003 T2 to T3 in prostacyclin group.
`
`Echocardiography
`Echocardiography data are given in Table 4. Of 240 images
`taken, 43 were rejected according to the predefined criteria.
`There was a 3% interobserver discordance. TEE examina-
`tions were similar in both groups before and after inhalation
`of PGI2 or placebo, but in the PGI2 group a tendency toward
`improved fractional area change or systolic function was
`noted for both the left and right ventricles. Also, in the PGI2
`group the systolic portion of hepatic flow tended to increase,
`and the diastolic portion remained stable compared with that
`seen with placebo, which suggests improvement in RV
`diastolic function, but this did not reach statistical signifi-
`cance.
`
`Platelet Aggregation
`Platelet aggregation data are shown in Table 5. Platelet
`aggregation studies showed a significant difference in plate-
`
`let luminescence produced by collagen (0.8 ⫾ 0.2 nm in the
`PGI2 group vs 1.2 ⫾ 0.5 nm in the placebo group, P ⫽
`.046), but this difference remained stable throughout the
`study (0.7 ⫾ 0.3 vs 1.1 ⫾ 0.5 nm, respectively; P ⫽ .055).
`Otherwise, platelet aggregation studies showed no dif-
`ference between the 2 groups with either adenosine diphos-
`phate, collagen, or arachidonic acid used as aggregants.
`Surgical blood loss was similar in both groups for both the
`intraoperative and postoperative periods.
`
`Discussion
`This is the first randomized controlled trial to examine
`hemodynamic, echocardiographic, oxygenation, and plate-
`let function effects of inhaled PGI2 versus placebo in pa-
`tients after anesthesia induction and before cardiac surgery.
`It confirms that inhaled PGI2 can be a selective pulmonary
`vasodilator and decrease indexed RV stroke work. Our TEE
`
`The Journal of Thoracic and Cardiovascular Surgery ● Volume 125, Number 3 645
`
`Liquidia's Exhibit 1024
`Page 4
`
`

`

`Cardiopulmonary Support and Physiology
`
`Hache´ et al
`
`Figure 1. sPAP variations of each patient before (T1) and 10 minutes after (T2) induction of anesthesia, after
`nebulization of PGI2 or placebo (T3), and 15 (T4) and 25 (T5) minutes after nebulization.
`
`findings also suggest a tendency toward improvement of
`both RV and LV systolic functions, as well as RV diastolic
`function. The dose administered was safe, with no systemic
`hypotension or any effect on platelet aggregation.
`Inhaled PGI2 decreases PAP. This has been confirmed in
`many animal18-20 and human12,21-24 studies. Because of this,
`it reduces RV afterload and can improve RV systolic and
`diastolic functions. In dogs having PH after hypoxic pul-
`monary vasoconstriction, Zwissler and colleagues21 demon-
`
`strated improvement of RV contraction indices through
`reduction in RV afterload with a small dose of inhaled PGI2.
`In human subjects one study showed improvement of RV
`ejection fraction in patients having PH caused by pulmonary
`fibrosis.23 Haraldsson and colleagues25 showed improve-
`ment of RV performance with inhaled PGI2 in patients
`having PH after cardiac surgery. Our study demonstrates a
`reduction in indexed RV stroke work in patients before
`cardiac surgery. The reduction in RV stroke work index was
`
`646 The Journal of Thoracic and Cardiovascular Surgery ● March 2003
`
`CSP
`
`Liquidia's Exhibit 1024
`Page 5
`
`

`

`Hache´ et al
`
`Cardiopulmonary Support and Physiology
`
`TABLE 3. Arterial and venous blood gas evolution
`T1
`
`T2
`
`T3
`
`T4
`
`Epoprostenol
`PaO2
`Arterial pH
`PaCO2
`Bicarbonate
`BE
`PvO2
`Venous pH
`PvCO2
`PaCO2 ⫺ PvCO2
`Placebo
`PaO2
`Arterial pH
`PaCO2
`Bicarbonate
`BE
`PvO2
`Venous pH
`PvCO2
`PaCO2 ⫺ PvCO2
`BE, Base excess; PvO2, venous oxygen partial pressure; PvCO2, venous carbon dioxide partial pressure; PaCO2 ⫺ PvCO2, arteriovenous difference in carbon
`dioxide partial pressure.
`*P ⫽ .05 between both groups.
`†P ⬍ .001 T1 to T2 in both groups.
`
`432.2 ⫾ 82
`7.46 ⫾ 0.02
`39.4 ⫾ 3.7
`28.1 ⫾ 2.1
`4.7 ⫾ 1.7
`46 ⫾ 4.6
`7.42 ⫾ 0.02
`46.9 ⫾ 4
`7.5 ⫾ 2.2
`
`425.9 ⫾ 88.75
`7.45 ⫾ 0.03
`40.4 ⫾ 4.6
`27.8 ⫾ 2
`4.1 ⫾ 1.7
`51.9 ⫾ 6.6
`7.40 ⫾ 0.03
`47 ⫾ 3.9
`6.6 ⫾ 2.4
`
`431.6 ⫾ 68.4
`7.45 ⫾ 0.03
`41.7 ⫾ 6
`28.9 ⫾ 2.5
`5.1 ⫾ 1.8
`45.6 ⫾ 5.7
`7.41 ⫾ 0.02
`49 ⫾ 5.4
`7.3 ⫾ 1.9
`
`438.3 ⫾ 80.62
`7.42 ⫾ 0.05
`43 ⫾ 6.4
`28.1 ⫾ 1.6
`3.9 ⫾ 1.3
`54.1 ⫾ 9.5
`7.40 ⫾ 0.04
`49.1 ⫾ 6.4
`6.1 ⫾ 2.2
`
`173.1 ⫾ 65.2*
`7.41 ⫾ 0.03
`46 ⫾ 4.2
`29.8 ⫾ 2.8
`5.2 ⫾ 2.4
`41.9 ⫾ 6.1
`7.39 ⫾ 0.03
`53.1 ⫾ 3.8
`6.4 ⫾ 4.2
`
`251.5 ⫾ 95.4*
`7.41 ⫾ 0.02
`44.9 ⫾ 3.3
`28.7 ⫾ 1.8
`4.2 ⫾ 1.6
`44.8 ⫾ 3.8
`7.39 ⫾ 0.03
`49.6 ⫾ 4.8
`4.7 ⫾ 2.8
`
`428.2 ⫾ 65.2†
`7.44 ⫾ 0.02
`42 ⫾ 2.3
`28.3 ⫾ 1.6
`4.3 ⫾ 1.6
`50 ⫾ 7
`7.39 ⫾ 0.02
`50.5 ⫾ 3
`8.5 ⫾ 3.5
`
`443.7 ⫾ 64.94†
`7.42 ⫾ 0.02
`43.1 ⫾ 2.8
`27.8 ⫾ 1.9
`3.6 ⫾ 1.7
`55.9 ⫾ 8
`7.39 ⫾ 0.02
`48.6 ⫾ 3.2
`5.5 ⫾ 2.3
`
`CSP
`
`associated with a tendency to improve RV fractional area
`change and also RV diastolic function.
`Although the decrease in PAP we observed was modest,
`we observed that its magnitude correlated with the severity
`of PH. This suggests that the benefit of inhaled PGI2 might
`be greater with patients with more advanced disease.
`There was a statistically significant decrease in heart rate
`after inhaled PGI2 administration as opposed to that seen
`after placebo administration, but this was of no clinical
`significance.
`inhaled PGI2 should improve systemic
`Theoretically,
`oxygenation by dilating pulmonary vessels in ventilated
`areas. This has been demonstrated in many models of hyp-
`oxia, including adult respiratory distress syndrome.6 It has
`been noted, however, that in patients who are not hypox-
`emic, as in our population, this effect is not always ob-
`served.4,25
`PGI2 is a powerful inhibitor of platelet aggregation. In
`vitro, small concentrations of PGI2 can inhibit platelet ag-
`gregation.11 However, when administered by means of in-
`halation, this effect has not been shown to occur consistent-
`ly.6,26,27 All previous platelet aggregation studies were done
`on platelet-rich plasma. We chose to evaluate platelet ag-
`gregation on whole blood because the platelet is then sur-
`rounded by other blood constituents that might play a role in
`modulating platelet aggregation. This might be more reflec-
`tive of the in vivo condition.28 We did not show a significant
`effect on platelet aggregation in our study. This suggests
`
`that not enough drug actually reached systemic circulation
`to have an effect on platelets. This, together with the similar
`surgical blood losses observed in both groups, suggests that
`the dose of inhaled PGI2 in the present study is safe.
`Two patients died after the operation. One patient in the
`PGI2 group had mitral valve replacement and a history of
`prior aortic valve replacement, paroxysmal atrial fibrilla-
`tion, and sleep apnea. The second patient, from the placebo
`group, underwent aorta-coronary bypass surgery and had a
`history of mitral commissurotomy, mitral valve replace-
`ment, and chronic obstructive lung disease. They both died
`of multiple-system organ failure. A 10% mortality rate was
`expected in our population.1
`
`Study Limitations
`The main limitation of this study is the small sample size
`that was calculated to observe a 20% decrease in sPAP. This
`might not have been sufficient to observe a significant effect
`on echocardiographic measurements or platelet function.
`Only one dose was studied. We did not explore the
`dose-response curve of inhaled PGI2. The dose used was
`based on previous experience with the drug in both the
`operating room and the intensive care unit. When adminis-
`tering drugs by means of nebulization, it is very difficult to
`determine the effective dose for many reasons. The quantity
`of drug that actually reaches the lungs is highly variable. It
`depends on the characteristics of the drug itself, the nebu-
`lizer used, the flow of carrier gas to the nebulizer, the
`
`The Journal of Thoracic and Cardiovascular Surgery ● Volume 125, Number 3 647
`
`Liquidia's Exhibit 1024
`Page 6
`
`

`

`Cardiopulmonary Support and Physiology
`
`Hache´ et al
`
`TABLE 4. TEE examination results
`Left ventricle
`n
`
`T2
`
`n
`
`Epoprostenol
`FAC (%)
`LV ESA (cm2)
`LV EDA (cm2)
`MEW (cm/s)
`MEW VTI (cm)
`MAW (cm/s)
`MAW VTI (cm)
`PSW (cm/s)
`PSW VTI (cm)
`PDW (cm/s)
`PDW VTI (cm)
`PAW (cm/s)
`PAW VTI (cm)
`Placebo
`FAC (%)
`LV ESA (cm2)
`LV EDA (cm2)
`MEW (cm/s)
`MEW VTI (cm)
`MAW (cm/s)
`MAW VTI (cm)
`PSW (cm/s)
`PSW VTI (cm)
`PDW (cm/s)
`PDW VTI (cm)
`PAW (cm/s)
`PAW VTI (cm)
`
`8
`8
`8
`6
`5
`6
`5
`9
`8
`9
`9
`6
`6
`
`10
`10
`10
`8
`8
`6
`6
`10
`10
`10
`10
`8
`8
`
`0.56 ⫾ 0.16
`10.7 ⫾ 6.5
`23.5 ⫾ 9.7
`138.8 ⫾ 66.8
`34.9 ⫾ 30.5
`78.9 ⫾ 39.7
`10 ⫾ 5.6
`18.6 ⫾ 55.9
`7.4 ⫾ 12.7
`51.1 ⫾ 26.2
`12.8 ⫾ 4.9
`33.8 ⫾ 18.9
`3.4 ⫾ 1.5
`
`0.54 ⫾ 0.13
`8.48 ⫾ 5.81
`17.51 ⫾ 9.05
`135.5 ⫾ 34.56
`29.31 ⫾ 14.15
`92.6 ⫾ 42.66
`10.67 ⫾ 6.68
`39.28 ⫾ 16.08
`9.84 ⫾ 6.13
`50.06 ⫾ 12.71
`12.99 ⫾ 3.51
`22.76 ⫾ 7.52
`2.31 ⫾ 1.07
`
`8
`8
`8
`6
`5
`6
`5
`8
`8
`9
`9
`6
`6
`
`10
`10
`10
`8
`8
`6
`6
`10
`10
`10
`10
`7
`7
`
`T3
`
`0.63 ⫾ 0.16
`8.5 ⫾ 5.9
`20.8 ⫾ 9.2
`133.7 ⫾ 46.4
`39.3 ⫾ 40.7
`72.9 ⫾ 38.9
`8.8 ⫾ 3.3
`25.7 ⫾ 62.3
`8.4 ⫾ 13.2
`50.9 ⫾ 28.1
`14.6 ⫾ 6.8
`37.1 ⫾ 19.2*
`4.5 ⫾ 2.4†
`
`0.55 ⫾ 0.05
`8.30 ⫾ 4.18
`18.35 ⫾ 9.07
`130.9 ⫾ 37.04
`27.51 ⫾ 21.07
`84.43 ⫾ 35.49
`10.08 ⫾ 2.32
`43.22 ⫾ 24.59
`10.41 ⫾ 7.4
`49.84 ⫾ 12.07
`12.88 ⫾ 3.32
`19.71 ⫾ 8.8*
`1.87 ⫾ 0.9†
`
`FAC (%)
`RV ESA (cm2)
`RV EDA (cm2)
`HSW (cm/s)
`HSW VTI (cm)
`HDW (cm/s)
`HDW VTI (cm)
`HAW (cm/s)
`HAW VTI (cm)
`TEW (cm/s)
`TEW VTI (cm)
`TAW (cm/s)
`TAW VTI (cm)
`
`FAC (%)
`RV ESA (cm2)
`RV EDA (cm2)
`HSW (cm/s)
`HSW VTI (cm)
`HDW (cm/s)
`HDW VTI (cm)
`HAW (cm/s)
`HAW VTI (cm)
`TEW (cm/s)
`TEW VTI (cm)
`TAW (cm/s)
`TAW VTI (cm)
`
`Right ventricle
`n
`
`T2
`
`T3
`
`0.47 ⫾ 0.07
`7.2 ⫾ 1.5
`13.4 ⫾ 1.8
`11.9 ⫾ 26.7
`3.4 ⫾ 6.9
`20.7 ⫾ 11.8
`5.9 ⫾ 2.8
`17.5 ⫾ 3.2
`2.1 ⫾ 0.9
`55.7 ⫾ 24.5
`11.3 ⫾ 3.7
`40.8 ⫾ 2.9‡
`5.3 ⫾ 1.6
`
`0.45 ⫾ 0.08
`7.56 ⫾ 3.52
`13.9 ⫾ 6.83
`16.28 ⫾ 12.08
`3.96 ⫾ 3.8
`17.33 ⫾ 5.99
`4.89 ⫾ 2.2
`17.08 ⫾ 9.82
`2.21 ⫾ 0.7
`47.56 ⫾ 17.26
`9.46 ⫾ 5.2
`35.44 ⫾ 11.02
`5.3 ⫾ 1.69
`
`7
`7
`7
`10
`9
`10
`9
`5
`4
`3
`3
`2
`2
`
`7
`7
`7
`10
`10
`9
`9
`6
`6
`6
`6
`4
`4
`
`0.54 ⫾ 0.08
`6.2 ⫾ 2.4
`13.2 ⫾ 3.4
`14 ⫾ 17.8
`2.7 ⫾ 3
`19.7 ⫾ 8.7
`5.4 ⫾ 2.2
`17.7 ⫾ 6.7
`2.5 ⫾ 1.6
`52.3 ⫾ 17.5
`11.3 ⫾ 3.8
`43.8 ⫾ 3.8‡
`5.9 ⫾ 0.3
`
`0.48 ⫾ 0.12
`6.83 ⫾ 3.77
`12.44 ⫾ 4.52
`13.7 ⫾ 8.72
`3.46 ⫾ 2.84
`31.51 ⫾ 47.42
`4.39 ⫾ 2.14
`12.57 ⫾ 4.15
`1.99 ⫾ 0.65
`48.68 ⫾ 10.77
`10.4 ⫾ 2.4
`32.05 ⫾ 6.8
`4.49 ⫾ 1.07
`
`n
`
`7
`7
`7
`9
`9
`9
`9
`5
`5
`5
`4
`3
`2
`
`8
`8
`8
`10
`9
`10
`9
`8
`7
`8
`7
`7
`6
`
`FAC, Fractional area change; ESA, end-systolic area; EDA, end-diastolic area; MEW, mitral E wave; MAW, mitral A wave; PSW, pulmonary S wave; PDW,
`pulmonary D wave; PAW, pulmonary A wave flow; HSW, hepatic S wave; HDW, hepatic D wave; HAW, hepatic A wave; TEW, tricuspid E wave; TAW,
`tricuspid A wave; VTI, velocity time interval; LV, left ventricular; RV, right ventricular.
`*P ⫽ .05 at T3 between both groups.
`†P ⫽ .02 at T3 between both groups.
`‡P ⬍ .01 T2 to T3 in epoprostenol group.
`
`CSP
`
`density of the carrier gas, the amount of drug remaining in
`the nebulizer, and the humidity and temperature in the
`nebulizer circuit.29,30 Particles of 1 to 5 ␮m are considered
`to be in the respirable range.29 We used high nebulizer flows
`because this increases the proportion of particles produced
`in the respirable range, and during mechanical ventilation, it
`is known that less than 10% of the medication actually
`reaches the alveolar epithelium. The 60-␮g dose we nebu-
`lized over 10 minutes is approximately equal to 85 ng ⫻
`kg⫺1 ⫻ min⫺1.
`PGI2 might be an alternative to iNO. Both can induce
`selective pulmonary vasodilatation. The effect of iNO re-
`mains localized through inactivation by hemoglobin as soon
`as it reaches the circulation. However, disadvantages related
`to the use of iNO have been identified, such as the produc-
`tion of toxic metabolites and methemoglobinemia when
`given at high concentrations for a long period of time. It
`
`requires costly and specialized equipment. Epoprostenol, on
`the other hand, has no known toxic metabolites and very
`few side effects. It can be administered by means of simple
`nebulization with minimal equipment, which is an advan-
`tage for use in the operating room. A full 24 hours of
`treatment is estimated to cost approximately $115.00 CAN
`(approximately $70.00 US).
`
`Conclusion
`
`Inhaled PGI2 reduces PAP in the preoperative period. By
`doing so, improvements in RV function through a decrease
`in indexed RV stroke work can be achieved. It is a safe
`medication that does not increase platelet dysfunction or
`perioperative bleeding. This exploratory trial confirms that
`inhaled PGI2 can be safely studied in the postbypass period,
`where treatment for unexpected PH can be encountered.
`
`648 The Journal of Thoracic and Cardiovascular Surgery ● March 2003
`
`Liquidia's Exhibit 1024
`Page 7
`
`

`

`Hache´ et al
`
`Cardiopulmonary Support and Physiology
`
`CSP
`
`10. Rich S, McLaughlin VV. The effects of chronic prostacyclin therapy
`on cardiac output and symptoms in primary pulmonary hypertension.
`J Am Coll Cardiol. 1999;34:1184-7.
`11. van Heerden PV, Gibbs NM, Michalopoulos N. Effect of low con-
`centrations of prostacyclin on platelet function in vitro. Anaesth In-
`tensive Care. 1997;25:343-6.
`12. Hache M, Denault AY, Belisle S, et al. Inhaled prostacyclin (PGI2) is
`an effective addition to the treatment of pulmonary hypertension and
`hypoxia in the operating room and intensive care unit [L’inhalation de
`prostacycline (PGI2) est un traitement comple´mentaire efficace de
`l’hypertension pulmonaire et de l’hypoxie observe´es en salle
`d’ope´ration et a` l’unite´ des soins intensifs]. Can J Anaesth. 2001;48:
`924-9.
`13. Moraes D, Loscalzo J. Pulmonary hypertension: newer concepts in
`diagnosis and management. Clin Cardiol. 1997;20:676-82.
`14. Fyrenius A, Wigstrom L, Bolger AF, et al. Pitfalls in Doppler evalu-
`ation of diastolic function: insights from 3-dimensional magnetic
`resonance imaging. J Am Soc Echocardiogr. 1999;12:817-26.
`15. Schnittger I, Fitzgerald PJ, Daughters GT, et al. Limitations of com-
`paring left ventricular volumes by two dimensional echocardiography,
`myocardial markers and cineangiography. Am J Cardiol. 1982;50:
`512-9.
`16. Bernard F, Denault A