Pharmacodynamics and
`Pharmacokinetics of Inhaled Iloprost,
`Aerosolized by Three Different Devices,
`in Severe Pulmonary Hypertension*
`
`Horst Olschewski, MD, PhD; Beate Rohde, MD; Ju¨ rgen Behr, MD, PhD;
`Ralph Ewert, MD, PhD; Tobias Gessler, MD; H. Ardeschir Ghofrani, MD; and
`Thomas Schmehl, PhD
`
`Background: Inhalation of iloprost, a stable prostacyclin analog, is an effective therapy for
`pulmonary hypertension with few side effects. This approach may, however, be handicapped by
`limitations of currently available nebulization devices. We assessed whether the physical
`characterization of a device is sufficient to predict drug deposition and pharmacologic effects.
`Methods: We investigated the effects of a standardized iloprost aerosol dose (5 ␮g; inhaled within
`approximately 10 min) in 12 patients with severe pulmonary hypertension in a crossover design
`employing three well-characterized nebulizers. The nebulizers use different techniques to
`increase efficiency and alveolar targeting (Ilo-Neb/Aerotrap [Nebu-Tec; Elsenfeld, Germany],
`Ventstream [MedicAid; Bognor Regis, UK], and HaloLite [Profile Therapeutics; Bognor Regis,
`UK]). Measurements were performed using a Swan-Ganz catheter and determination of arterial
`iloprost plasma levels.
`Results: During inhalation of iloprost, the pulmonary vascular resistance decreased substantially
`(baseline, approximately 1,250 dyne䡠s䡠cm-5; decrease, ⴚ 35.5 to ⴚ 38.0%) and pulmonary artery
`pressure decreased substantially (baseline, approximately 58 mm Hg; decline, ⴚ 18.4 to ⴚ21.8%),
`whereas the systemic arterial pressure was largely unaffected. Cardiac output and mixed venous
`and arterial oxygen saturation displayed a marked increase. The pharmacodynamic profiles with
`the three devices were superimposable. Moreover, rapid entry of iloprost into the systemic
`circulation was noted, peaking immediately after termination of the inhalation maneuver, with
`very similar maximum serum concentrations (158 pg/mL, 155 pg/mL, and 157 pg/mL), and
`half-lives of serum levels (6.5 min, 9.4 min, and 7.7 min) for the three nebulizers, respectively.
`Interestingly, the “half-life” of the pharmacodynamic effects in the pulmonary vasculature (eg,
`decrease in pulmonary vascular resistance, ranging between 21 and 25 min) clearly outlasted this
`serum level-based pharmacokinetic half-life.
`Conclusions: A standardized dose of aerosolized iloprost delivered by different nebulizer types
`induces comparable pharmacodynamic and pharmacokinetic responses. Pulmonary vasodilation,
`persisting after disappearance of the drug from the systemic circulation, supports the hypothesis
`that local drug deposition largely contributes to the preferential pulmonary vasodilation in
`response to inhaled iloprost.
`(CHEST 2003; 124:1294 –1304)
`
`Key words: iloprost; nebulization device; pulmonary hypertension
`
`Abbreviations: ANOVA ⫽ analysis of variance; AUC ⫽ area under the serum level-time curve; AUCt0-tlast ⫽ area
`under the serum level-time curve calculated from the start of
`inhalation to the last sampling time point;
`Cmax ⫽ maximum serum concentration; NO ⫽ nitric oxide; PAP ⫽ pulmonary artery pressure; PPH ⫽ primary
`⫽ arterial oxygen saturation; SAP ⫽ systemic
`pulmonary hypertension; PVR ⫽ pulmonary vascular resistance; Sao2
`⫽ mixed venous oxygen saturation; SVR ⫽ systemic vascular resistance
`artery pressure; Svo2
`
`P rimary pulmonary hypertension (PPH) is a se-
`
`vere disabling disease with a poor prognosis.1
`Long-term infusion of prostacyclin was the first
`therapy shown to be lifesaving in a controlled study,2
`and such efficacy may also exist for patients with the
`scleroderma spectrum of diseases,3 and other dis-
`eases associated with severe pulmonary arterial hy-
`
`pertension.4,5 However, this approach is hampered
`by the lack of pulmonary selectivity of the vasodila-
`tory effect of prostacyclin causing systemic side
`effects such as pain and systemic hypotension,6 by
`ventilation-perfusion mismatch in predisposed pa-
`tients,7–9 as well as by infectious complications due
`to the continuous use of an IV line.6 Inhalation of
`
`1294
`
`Clinical Investigations
`
`Downloaded From: http://journal.publications.chestnet.org/ by a UIC Library of Health Sciences User on 08/03/2016
`
`WATSON LABORATORIES, INC. , IPR2017-01621, Ex. 1029, p. 1 of 11
`
`

`

`nitric oxide (NO) is selective for the lung vasculature,
`but its pulmonary vasodilatory potency is lower than
`that of prostacyclin,10 –12 and interruption of NO
`inhalation may provoke a rebound hypertensive crisis
`due to the very short half-life of this agent.13,14
`As an alternative approach, repetitive aerosol de-
`livery of iloprost, a stable prostacyclin analog, has
`been used to treat pulmonary hypertension and
`proved to be efficacious in a double-blind, placebo-
`controlled study.15 In PPH and other forms of
`precapillary pulmonary hypertension, it was demon-
`strated that nebulized iloprost decreases pulmonary
`vascular resistance (PVR) and pulmonary artery
`pressure (PAP), concomitant with an increase in
`cardiac output, in the absence of significant systemic
`arterial pressure drop and ventilation-perfusion mis-
`match.9,16 –18 Rescue administration of inhaled ilo-
`prost was undertaken in progressive right-heart fail-
`ure due to severe pulmonary hypertension,19,20 and
`currently available data from a 2-year study with
`iloprost aerosolization suggest beneficial long-term
`effects with minor side effects.21
`The aerosol approach is, however, handicapped by
`limitations of the currently available nebulization
`devices. Substantial loss of the aerosolized drug in
`the device is a common finding, resulting in major
`and possibly variable differences between the dose
`delivered into the device, the nebulized dose, and
`the inhaled dose. Droplet size and aerosol concen-
`tration during the inhalation phase may determine
`the extent and the site of pulmonary drug deposition,
`and thus on delivery of the inhaled agent to the
`pulmonary vasculature.
`In the present study, we investigated the pharma-
`cokinetics and the pharmacodynamic effects of in-
`haled iloprost in patients with severe pulmonary
`hypertension. Three different jet nebulizers manu-
`factured for alveolar drug delivery were employed.
`The first, coupled with a reservoir for alveolar tar-
`geting (Ilo-Neb/Aerotrap; Nebu-Tec; Elsenfeld,
`Germany) has been widely employed in long-term
`
`*From the Department of Internal Medicine (Drs. Olschewski,
`Gessler, Ghofrani, and Schmehl), Justus-Liebig-University, Gies-
`sen; Schering AG (Dr. Rohde), Berlin; Ludwig-Maximilians-
`University (Dr. Behr), Munich; and Deutsches Herzzentrum
`(Dr. Ewert), Berlin, Germany.
`The study was funded by Schering AG, Berlin, Germany.
`Drs. Olschewski, Gessler, and Schmehl are consultants for
`Schering AG and NebuTec Company. Dr. Rohde is an employee
`of Schering AG.
`Manuscript received October 18, 2002; revision accepted Feb-
`ruary 10, 2003.
`Reproduction of this article is prohibited without written permis-
`sion from the American College of Chest Physicians (e-mail:
`permissions@chestnet.org).
`Correspondence to: Horst Olschewski, MD, PhD, Department of
`Internal Medicine II, University Hospital, Klinikstrasse 36,
`D-35392 Giessen; Germany; e-mail horst.olschewski@innere.
`med.uni-giessen.de
`
`iloprost nebulization including the open-label, 2-year
`study in Germany.21 Due to the reservoir, this device
`is relatively large. Additionally, considerable amounts of
`aerosol are deposited inside the device, resulting in an
`efficiency (inhaled dose/filling dose) of only approxi-
`mately 13%. The second nebulizer uses the Venturi
`effect to boost aerosol production during inspiration
`(Ventstream; MedicAid; Bognor Regis, UK) and is
`much smaller as it does not need a reservoir, but loses
`efficiency due to internal drug deposition and high
`filling volumes (efficiency approximately 15%). The
`third nebulizer employs a microchip technique for
`delivery of an aerosol pulse during the first half of the
`inhaled tidal volume, and was used in the pivotal trial
`showing clinical efficacy of inhaled iloprost.15 It com-
`bines increased efficiency (approximately 25%) with
`exact aerosol dosing independent of the breathing
`pattern (HaloLite; Profile Therapeutics; Bognor
`Regis, UK).
`The dosing regimen for each nebulizer was ad-
`justed to deliver a total amount of 5 ␮g of iloprost to
`the respiratory tract of the patients (calculated for
`entry at mouthpiece) during an inhalation maneuver
`lasting approximately 10 min, based on preceding
`biophysical characterization of these devices.22 Fol-
`lowing such a standardized procedure, the pharma-
`codynamic effects of inhaled iloprost were found to
`be virtually superimposable for all devices tested.
`Moreover, this was also true for the kinetics of
`iloprost appearance and disappearance in the sys-
`temic circulation in response to its aerosol delivery.
`Interestingly, the serum levels of inhaled iloprost
`declined much more rapidly than the pulmonary
`vasodilatory effect, which suggests prolonged local
`vasorelaxant potency of the aerosolized agent not
`reflected by the time course of serum levels.
`
`Materials and Methods
`
`Patients
`
`A total of 12 patients (aged ⬎ 18 years) with PPH or secondary
`pulmonary hypertension were enrolled in the study. The patients
`had been treated for at least 3 months with six iloprost inhalations
`per day in three experienced German centers for pulmonary
`hypertension. Only patients who were known to respond to
`iloprost inhalation with a decrease of PVR of at least 20% were
`included. Patients with renal or severe hepatic impairment,
`thromboembolic disease, bleeding disorders, significant restric-
`tive or obstructive lung diseases, pulmonary venous hypertension,
`and diseases directly affecting the pulmonary vessels, according
`to the new diagnostic World Health Organization classification,23
`were excluded from the study. The patients gave written in-
`formed consent before participation in the trial. The study
`protocol was reviewed and approved by the local ethics commit-
`tees of the participating centers.
`
`www.chestjournal.org
`
`CHEST / 124 / 4 /OCTOBER, 2003
`
`1295
`
`Downloaded From: http://journal.publications.chestnet.org/ by a UIC Library of Health Sciences User on 08/03/2016
`
`WATSON LABORATORIES, INC. , IPR2017-01621, Ex. 1029, p. 2 of 11
`
`

`

`Nebulizing Devices
`
`Three different jet nebulizer systems were used: (1) Ilo-Neb
`combined with the Aerotrap reservoir, and the Pulmocar Akku
`compressor (Sanesco Medizintechnik; Vienna, Austria); (2) Vent-
`stream combined with the Freeway Lite compressor (MedicAid);
`and (3) HaloLite. On the basis of preceding in vitro character-
`ization of the aerosols delivered by each device,22 the volume of
`the iloprost solution to be filled into the nebulizer and the
`concentration of the iloprost solution were adjusted for each
`device to provide a total dose of 5 ␮g of iloprost delivered at the
`mouthpiece within a comparable time span. The physical char-
`acteristics of the nebulizers and the dose regimens are depicted
`in Table 1. Accordingly, the inhalation time was defined to be 12
`min and 10 min for the IloNeb and the Ventstream nebulizers,
`respectively. In contrast, the nebulization time of the HaloLite
`system is not prefixed, because this device stops automatically
`when the inhalation of the prescribed dose is completed. The
`basis of this system is a breath-by-breath measurement and
`summing up of the total inhaled quantity. Inhalation times for the
`HaloLite nebulizer were thus variable, depending on the breath-
`ing pattern of the patient, with an average of approximately 11
`min. The iloprost solution administered with the nebulizer was
`prepared immediately before use, according to written detailed
`dosing instructions.
`
`Study Conduct
`
`A randomized, open-label, multicenter, crossover trial design
`was used with six sequences, three treatments and three periods
`(also known as Williams design). Each patient was randomly
`assigned to one of the six possible sequences of treatments. The
`iloprost aerosol (total dose of 5 ␮g at the mouthpiece) was
`administered by three subsequent inhalations using the three
`different devices in randomized order. Every treatment was
`followed by a washout time of 2 h. All treatments were admin-
`istered on a single day during a previously scheduled right-heart
`catheterization. For safety reasons, the following inhalation stop
`criteria were defined: (1) decrease of mean systemic BP to ⬍ 65
`mm Hg or by ⬎ 10% of baseline for at least 2 min, (2) decrease
`of arterial oxygen saturation (Sao2) by⬎ 5% for at least 2 min, (3)
`severe headache, and/or (4) local intolerability of the aerosol.
`Before inhalations were started, a Swan-Ganz catheter (Baxter
`Edwards; Deerfield, IL) and an arterial line were inserted. A
`fiberoptic thermodilution pulmonary artery catheter was used for
`measuring PAP, pulmonary artery wedge pressure, central ve-
`nous pressure, and cardiac output. The arterial line was used for
`continuous measurement of systemic arterial pressure (SAP) and
`drawing arterial blood samples for blood gas analysis and mea-
`surement of iloprost serum levels. For assessment of Po2, Pao2,
`and Paco2, as well as Sao2 and mixed venous oxygen saturation
`
`(Svo2), venous and arterial blood samples were withdrawn at
`baseline, at end of inhalation, and 30 min, 60 min, and 120 min
`after the end of inhalation. Hemodynamics and drug concentra-
`tions were measured simultaneously. Additionally, hemodynam-
`ics were assessed 5 min and 15 min after the end of inhalation,
`and drug concentrations were measured 2 min, 5 min, and 15 min
`after the end of inhalation.
`
`Measurement of Iloprost Concentrations
`
`Serum levels of iloprost were measured using a specific and
`sensitive radioimmunoassay with a quantitation limit of 25 pg/
`mL. The assay was validated using human samples and gas
`chromatography coupled with mass spectrometer as an alterna-
`tive analysis procedure.24
`
`Pharmacokinetic Evaluation
`
`Pharmacokinetic parameters were obtained from the individ-
`ual serum level-time curves using the TOPFIT program, Version
`2.1 (Goedecke AG, Schering AG, Thomae GmbH; Freiburg,
`Germany). Maximum serum concentration (Cmax) and time to
`reach Cmax were directly taken from the data. The area under
`the serum level-time curve (AUC) was calculated according to
`the linear trapezoidal rule from the start of inhalation to the last
`sampling time point (AUCt0-tlast), which was calculated with
`concentrations above the limit of quantitation (AUCt0-tlast) and
`the partial area from the last sampling time point
`to the
`respective next sampling time point for which the concentration
`was set at zero. This method of extrapolating AUC was chosen
`because AUC measured at the start of inhalation to infinity could
`not be obtained in all cases due to the rapid decrease of iloprost
`serum levels after the end of inhalation. In assessments with
`quantifiable iloprost serum levels in the last blood sample of
`treatment periods 1 and 2, only AUCt0-tlast could be calculated
`and was used for subsequent statistical evaluation. Patient 11 was
`excluded from evaluation of treatment 2 (Ventstream device)
`because no drug exposure was achieved due to a technical
`problem during inhalation.
`For further characterization of the pharmacokinetics of inhaled
`iloprost, the half-life (t1/2) of the iloprost serum level decrease
`was calculated by means of regression analysis of the mean serum
`level-time curve between the end of inhalation and 30 min after
`end of inhalation in a semilogarithmic plot (␭ ⫽slope of regres-
`⫽ 2/␭. Accordingly, the “half-life” of the pharma-
`sion line): t1/2
`codynamic effect was calculated, employing the iloprost-induced
`decrease in PVR. Calculation was performed by means of
`regression analysis of the time curves of absolute changes of PVR
`compared to baseline between the time point of maximum effect
`and 30 min after end of inhalation in a semilogarithmic plot
`⫽ 2/␭.
`(␭ ⫽slope of regression line): t 1/2
`
`Table 1—Baseline Characteristics of the Different Nebulization Devices: Related Dosing Regimens*
`
`Characteristics
`
`Ilo-Neb
`
`Ventstream
`
`HaloLite
`
`In vitro characteristics of nebulizing devices (tidal volume 0.67 L, 15 cycles per min)
`Particle size (MMAD), ␮m
`3.5
`Mass flow (output rate of the nebulizer), ␮L/min
`28
`Dosing regimens for delivery of 5 ␮g of iloprost at the mouthpiece
`Iloprost concentration of the inhalation solution, ␮g/mL
`Volume filled into the inhalation device, mL
`Inhalation time, min
`*MMAD ⫽ mass median aerodynamic diameter.
`†Delivered according to the breathing pattern of the patient (average inhalation time of approximately 11 min in the present study).
`
`4.3
`62
`
`10
`2.5
`approximately 11†
`
`3.8
`65
`
`8
`5
`10
`
`15
`4
`12
`
`1296
`
`Clinical Investigations
`
`Downloaded From: http://journal.publications.chestnet.org/ by a UIC Library of Health Sciences User on 08/03/2016
`
`WATSON LABORATORIES, INC. , IPR2017-01621, Ex. 1029, p. 3 of 11
`
`

`

`Statistics
`
`The target variables were the maximum percentage changes in
`the hemodynamic variables in response to iloprost inhalation and
`the pharmacokinetic variables Cmax and AUC. Analysis of vari-
`ance (ANOVA) for a mixed linear model using fixed effects
`(period, treatment, and first-order carryover) as well as random
`subject effect was performed to compare the target variables.
`Two-sided tests and confidence intervals were adjusted for three
`multiple comparisons according to Tukey-Kramer. The multiple
`significance (confidence) level was set at 0.05 (0.95) for each
`target variable. There was no adjustment for multiple end points.
`
`Results
`
`Nine female and four male patients (aged 26 to 71
`years) were included in the study. Twelve of them
`completed the trial. One male patient was prema-
`turely withdrawn due to an exclusion criterion that
`became evident immediately after start of the first
`treatment (mean SAP ⬍ 65 mm Hg). Eleven of the
`12 patients had PPH, and 1 patient presented with
`isolated pulmonary hypertension with CREST syn-
`drome (calcinosis, Raynaud phenomenon, esopha-
`geal dysfunction, sclerodactyly, telangiectasia with-
`out lung fibrosis or major inner organ involvement).
`Body weight ranged between 41 kg and 86 kg.
`In 33 of 36 inhalation maneuvers, the aerosoliza-
`tion was entirely finished;
`in three cases,
`it was
`stopped prematurely (although close to the pre-
`defined end of inhalation) due to a decrease in mean
`systemic BP by ⬎ 10% of baseline for ⬎ 2 min. The
`patients did not experience any subjective adverse
`events.
`Iloprost inhalation caused an average PVR de-
`crease of 35.5 to 38.0% compared to baseline, with a
`maximum effect either immediately at the end of
`inhalation, or within 5 min (Table 2, Fig 1). The PVR
`decrease was paralleled by an average decrease in
`mean PAP of 18.4 to 21.8%, and an average increase
`in cardiac output of 30.6 to 37.1%. The SAP showed
`only some marginal decrease, while systemic vascu-
`
`lar resistance (SVR) changes were noted in parallel
`with the changes in cardiac output (Fig 2, top, A, and
`center, B; Table 2). The PVR/SVR ratio indicated a
`more pronounced effect of iloprost inhalation on the
`PVR than on the SVR (Fig 2, bottom, C).
`Heart rate was not significantly changed (data not
`given), and no relevant arrhythmia was noted. The
`mean Sao2 and mean Svo2 increased in response to
`iloprost inhalation (Fig 3; Table 2). A critical de-
`crease in Sao2 (⬎ 5% compared to baseline) did not
`occur in any patient.
`Iloprost serum levels became rapidly detectable
`with all three devices employed (Fig 4). Serum
`concentrations reached a maximum either at the end
`of inhalation or within the following 5 min (Fig 4;
`Table 2). Beyond 30 min after the end of inhalation,
`quantifiable serum levels of iloprost were detected in
`only 3 of the 12 patients. Cmax and AUC values were
`comparable using the different nebulizers (Table 2).
`After reaching an early concentration peak, iloprost
`serum levels rapidly decreased with a half-life of 6.5
`to 9.4 min (Table 3). In contrast, the half-life of the
`pharmacodynamic effect, as calculated from the PVR
`decrease, ranged between 21 min and 25 min.
`When statistical analysis (ANOVA) was applied to
`the hemodynamic and pharmacokinetic responses to
`iloprost inhalations performed with the different
`devices, no significant differences between the de-
`vices were revealed, except
`for a slightly more
`pronounced SAP decrease shortly after iloprost in-
`halation using the HaloLite compared to the IloNeb
`system (Table 4). However, this was not associated
`with any typical complaints and not considered
`clinically significant in any case.
`The iloprost aerosol administration was well toler-
`ated by the patients irrespective of the nebulization
`device employed. Minor adverse events such as
`transient headache and flush were noted with similar
`frequencies with all devices: headache occurred in 4
`patients, 3 patients, and 4 patients following inhala-
`
`Table 2—Maximum Hemodynamic Effects, Cmax, and AUC After Inhalation of Iloprost Using Different
`Nebulization Devices*
`
`Effects
`
`Maximum PVR change, %
`Maximum mean PAP change, %
`Maximum cardiac output change, %
`Maximum Svo2 change, %
`Maximum mean SAP change, %
`Maximum SVR change, %
`Cmax, pg/mL
`AUC, pg/h/mL
`*Data are presented as mean ⫾ SEM.
`†n ⫽ 11.
`
`www.chestjournal.org
`
`Ilo-Neb (n ⫽ 12)
`⫺ 36.4 ⫾ 4.8
`⫺ 21.8 ⫾ 3.9
`30.6 ⫾ 7.7
`19.2 ⫾ 5.6
`⫺ 2.3 ⫾ 2.2
`⫺ 17.0 ⫾ 4.7
`158 ⫾ 20
`49.0 ⫾ 9.9
`
`Ventstream (n ⫽ 12)
`⫺ 36.9 ⫾ 3.9
`⫺ 18.5 ⫾ 3.8
`33.3 ⫾ 7.0
`14.1 ⫾ 6.1
`⫺ 3.0 ⫾ 2.8
`⫺ 18.8 ⫾ 7.0
`155 ⫾ 20†
`54.2 ⫾ 13.6†
`
`HaloLite (n ⫽ 12)
`⫺ 38.0 ⫾ 4.9
`⫺ 20.7 ⫾ 3.9
`36.2 ⫾ 10.0
`22.2 ⫾ 9.1
`⫺ 7.8 ⫾ 2.4
`⫺ 24.6 ⫾ 5.7
`157 ⫾ 18
`47.8 ⫾ 10.2
`
`CHEST / 124 / 4 /OCTOBER, 2003
`
`1297
`
`Downloaded From: http://journal.publications.chestnet.org/ by a UIC Library of Health Sciences User on 08/03/2016
`
`WATSON LABORATORIES, INC. , IPR2017-01621, Ex. 1029, p. 4 of 11
`
`

`

`Figure 1. Time course of pulmonary hemodynamic variables in response to iloprost inhalation. In 12
`patients with severe pulmonary hypertension, 5 ␮g of iloprost was inhaled employing three different
`nebulization devices. Time was set at zero at the end of the inhalation maneuver. Data are given as
`mean ⫾ SEM.
`
`1298
`
`Clinical Investigations
`
`Downloaded From: http://journal.publications.chestnet.org/ by a UIC Library of Health Sciences User on 08/03/2016
`
`WATSON LABORATORIES, INC. , IPR2017-01621, Ex. 1029, p. 5 of 11
`
`

`

`Figure 2. Time course of systemic hemodynamic variables in response to iloprost inhalation. In 12
`patients with pulmonary hypertension, 5 ␮g of iloprost was inhaled employing three different
`nebulization devices. Time was set at zero at the end of the inhalation maneuver. Data are given as
`mean ⫾ SEM.
`
`www.chestjournal.org
`
`CHEST / 124 / 4 /OCTOBER, 2003
`
`1299
`
`Downloaded From: http://journal.publications.chestnet.org/ by a UIC Library of Health Sciences User on 08/03/2016
`
`WATSON LABORATORIES, INC. , IPR2017-01621, Ex. 1029, p. 6 of 11
`
`

`

`Figure 3. Time course of Sao2 and Svo2 in response to iloprost inhalation. In 12 patients with
`pulmonary hypertension, 5 ␮g of iloprost was inhaled employing three different nebulization devices.
`Time was set at zero at the end of the inhalation maneuver. Data are given as mean ⫾ SEM.
`
`tions using HaloLite, Ventstream, and IloNeb, re-
`spectively, and flush was observed in 6 patients, 8
`patients, and 10 patients. Both of these adverse
`events were closely related to the peak iloprost
`serum levels, starting during or immediately after the
`end of inhalation and lasting for a few minutes. In
`three patients, mild cough occurred during inhala-
`tion, which did not impair or limit the inhalation.
`Further adverse events possibly related to study drug
`occurred only in single cases: ventricular extrasysto-
`les, palpitation, chest pain, sore throat, rash, tinnitus,
`eye pain, and taste perversion. All adverse events
`were of mild or moderate intensity and disappeared
`completely mostly within minutes.
`
`Discussion
`
`In this study, the dosing regimen of three different
`jet nebulizers was adjusted to deliver a total quantity
`of 5 ␮g of iloprost to the respiratory tract of patients
`with severe pulmonary hypertension during an inha-
`lation maneuver of approximately 10 min. The dose
`calculation was based on detailed biophysical char-
`acterization. Preferential pulmonary vasodilation
`with concomitant increase in cardiac output was
`noted, and assessment of iloprost serum levels doc-
`umented rapid entry of the inhaled agent into the
`systemic circulation. Both the pharmacodynamic and
`the pharmacokinetic effects were superimposable for
`
`1300
`
`Clinical Investigations
`
`Downloaded From: http://journal.publications.chestnet.org/ by a UIC Library of Health Sciences User on 08/03/2016
`
`WATSON LABORATORIES, INC. , IPR2017-01621, Ex. 1029, p. 7 of 11
`
`

`

`Figure 4. Time courses of iloprost serum levels in response to iloprost inhalation. In 12 patients with
`pulmonary hypertension, 5 ␮g of iloprost was inhaled employing three different nebulization devices.
`Time was set at zero at the end of the inhalation maneuver. Data are given as mean ⫾ SEM (error bars
`are missing when falling into a symbol).
`
`all three devices employed. The serum half-life of
`iloprost was, however, much shorter than the half-
`life of the pulmonary vasodilatory effects exerted by
`this agent, suggesting that
`the inhaled iloprost
`caused prolonged local vasodilation not reflected by
`the serum levels.
`The patients were selected based on a docu-
`mented acute PVR fall ⬎ 20% in response to a
`previous iloprost inhalation. The rationale behind
`
`Table 3—Half-life of Iloprost Serum Levels and
`Pulmonary Vasodilatory Effect After Inhalation
`of Iloprost
`
`Half-Life
`
`Ilo-Neb
`(n ⫽ 12)
`
`Ventstream
`(n ⫽ 11)
`
`HaloLite
`(n ⫽ 12)
`
`this was to improve the chance to show differences
`between the devices. In a previous study12 compar-
`ing NO and inhaled iloprost in nonselected patients
`with PPH, only approximately 20% had a poor acute
`vasodilatory response to inhaled iloprost consisting of
`⬍ 20% PVR decrease, and the mean PVR response
`to inhaled iloprost was 32%. This suggests that the
`patient population of the present study represents
`the majority of patients with PPH.
`The devices currently investigated employed dif-
`ferent mechanisms for alveolar targeting. In the
`IloNeb/Aerotrap system, the aerosol is continuously
`produced, but during expiration it is stored in the
`Aerotrap reservoir. At the beginning of the next
`inspiration,
`the aerosol content of
`the reservoir
`together with the continuously produced aerosol
`enters the lung, resulting in enhanced aerosol con-
`centration. In contrast, at the end of the inhalation
`phase, only the freshly produced (less concentrated)
`aerosol is carried to the respiratory tract, with parts
`of it filling the dead space of the respiratory tract.
`The main disadvantages of this device are the dimen-
`sions of the reservoir and the drug loss within the
`apparatus, mainly due to aerosol deposition at the
`
`6.5
`
`24
`
`9.4
`
`21
`
`Half-life of iloprost serum
`levels, min*
`“Half-life” of pulmonary
`hemodynamic effect, min†
`*Calculated from the mean iloprost serum concentration time curve
`from the end of inhalation up to 30 min after the end of inhalation.
`†Calculated from the time course of the absolute changes of mean
`PVRs from the end of inhalation up to 30 min after the end of
`inhalation.
`
`7.7
`
`25
`
`www.chestjournal.org
`
`CHEST / 124 / 4 /OCTOBER, 2003
`
`1301
`
`Downloaded From: http://journal.publications.chestnet.org/ by a UIC Library of Health Sciences User on 08/03/2016
`
`WATSON LABORATORIES, INC. , IPR2017-01621, Ex. 1029, p. 8 of 11
`
`

`

`Table 4—ANOVA of Maximum Percentage Changes of Hemodynamics and Pharmacokinetics*
`
`Variables
`
`Treatment vs Treatment
`
`Difference in
`Maximum
`Percentage
`Change
`
`Adjusted
`p Value
`
`Ratio of
`Treatment
`Effects, %
`
`Adjusted
`Lower CI
`Limit
`
`Adjusted
`Upper CI
`Limit
`
`Adjusted
`Lower CI
`Limit, %
`
`Adjusted
`Upper CI
`Limit, %
`
`Hemodynamics
`PVR
`
`Mean PAP
`
`Cardiac output
`
`SVR
`
`Mean SAP
`
`HaloLite vs Ventstream
`Ilo-Neb vs Ventstream
`HaloLite vs Ilo-Neb
`HaloLite vs Ventstream
`Ilo-Neb vs Ventstream
`HaloLite vs Ilo-Neb
`HaloLite vs Ventstream
`Ilo-Neb vs Ventstream
`HaloLite vs Ilo-Neb
`HaloLite vs Ventstream
`Ilo-Neb vs Ventstream
`HaloLite vs Ilo-Neb
`HaloLite vs Ventstream
`Ilo-Neb vs Ventstream
`HaloLite vs Ilo-Neb
`
`0.9788
`0.9881
`0.9986
`0.6879
`0.1687
`0.5401
`0.9419
`0.7493
`0.9181
`0.7574
`0.4011
`0.1342
`0.0884
`0.8529
`0.0318
`
`⫺ 0.80
`⫺ 0.60
`⫺ 0.20
`⫺ 2.54
`⫺ 5.83
`3.28
`⫺ 1.67
`⫺ 3.67
`2.00
`⫺ 4.82
`8.83
`⫺ 13.65
`⫺ 6.52
`1.50
`⫺ 8.02
`
`⫺ 11.19
`⫺ 10.96
`⫺ 10.59
`⫺ 10.33
`⫺ 13.60
`⫺ 4.50
`⫺ 14.61
`⫺ 16.52
`⫺ 10.93
`⫺ 22.00
`⫺ 8.20
`⫺ 30.83
`⫺ 13.90
`⫺ 5.59
`⫺ 15.40
`
`9.58
`9.77
`10.18
`5.24
`1.95
`11.07
`11.26
`9.17
`14.93
`12.36
`25.86
`3.53
`0.85
`8.59
`⫺ 0.65
`
`Pharmacokinetics
`AUC
`
`Cmax
`
`129.1
`67.9
`93.7
`0.8615
`HaloLite vs Ilo-Neb
`128.2
`67.4
`93.0
`0.8320
`HaloLite vs Ventstream
`134.5
`73.3
`99.3
`0.9979
`Ilo-Neb vs Ventstream
`125.9
`76.2
`97.9
`0.9753
`HaloLite vs Ilo-Neb
`128.6
`77.8
`100.0
`1.0000
`HaloLite vs Ventstream
`129.6
`80.5
`102.1
`0.9718
`Ilo-Neb vs Ventstream
`*The maximum percentage changes observed in selected hemodynamic variables and the log-transformed (base e) pharmacokinetic variables were
`compared between inhalations with two different devices. Adjustment according to Tukey-Kramer with a multiple significance level of 0.05 and
`a simultaneous confidence level of 0.95. CI ⫽ 95% confidence interval.
`
`inspiratory valve. The Ventstream nebulizer uses a
`valve system to provide an additional inspiratory side
`flow for enhancing the aerosol output during inha-
`lation by the Venturi principle. Consequently, aero-
`sol formation is reduced while the patient is exhaling,
`thereby matching aerosol delivery with tidal volume
`and suppressing drug wastage. Nevertheless, due to
`continuous operation of the nebulizer, there is con-
`siderable aerosol loss during expiration. Moreover,
`there is substantial intradevice drug wastage due to a
`high minimum filling volume required in the aerosol
`chamber. The HaloLite is a new electronically con-
`trolled device, applying aerosol pulses only in a
`preset period during early inspiration, with delivery
`adjusted to the breathing pattern. These aerosol
`pulses are added up, and the device stops automat-
`ically when the target dose has been delivered. The
`main advantages of the system are the virtual ab-
`sence of aerosol delivered to the airway dead space,
`the fact that the predefined drug dose will be applied
`irrespective of the breathing pattern, and the low
`volume of inhalation solution necessary for sufficient
`nebulization. However, a high driving pressure is
`needed for this device, which requires continuous
`connection to mains, and the filling volume consid-
`erably exceeds the maximum nebulized volume. For
`
`clinical practice, it is important to note that the filling
`volume and the inhaled volume differ considerably
`and that their ratio critically depends on the device
`itself and on the mode of use. Therefore, recommen-
`dations for devices for use of inhaled iloprost have to
`consider all these factors.
`It is a main finding of the present study that—
`irrespective of the substantial technical differences
`of
`the devices employed—the pharmacodynamic
`effects of inhaled iloprost were virtually superimpos-
`able. As previously described,9,16,16 –20 preferential
`pulmonary vasodilation was noted, with a decrease in
`both PVR and PAP, whereas the SAP was largely
`unaffected. In parallel, the cardiac output and both
`Svo2 and Sao2 consistently increased, with all effects
`leveling off within approximately 60 min. Moreover,
`rapid entry of the inhaled drug into the systemic
`circulation was observed: for all devices employed, a
`Cmax of approximately 160 pg/mL of iloprost was
`detected immediately after termination of inhalation,
`with subsequent rapid disappearance of this agent
`from the systemic vasculature. Two conclusions may
`be drawn from these findings: (1) detailed biophys-
`ical characterization of the aerosol devices provides a
`sound basis for reliable calculation of the dosage of a
`drug delivered by nebulization to the alveolar com-
`
`1302
`
`Clinical Investigations
`
`Downloaded From: http://journal.publications.chestnet.org/ by a UIC Library of Health Sciences User on 08/03/2016
`
`WATSON LABORATORIES, INC. , IPR2017-01621, Ex. 1029, p. 9 of 11
`
`

`

`partment or even via inhalation to the systemic
`circulation, and (2) the different technical solutions
`developed for alveolar drug targeting in the three
`devices tested do result in comparable hemodynamic
`and pharmacokinetic effects of the standardized dose
`of 5 ␮g of iloprost delivered to the mouthpiece,
`irrespective of the differences in aerosol concentra-
`tion during the inhalation cycle discussed above.
`Moreover, tolerability of the inhalation maneuver
`was comparable for all devices, with only minor
`adverse effects such as flush, which is a typical
`systemic side effect of iloprost.
`In the pr