`vascular resistance from 1721 ± 253 dyne/s • c m-5 to 1019
`± 203 dyne/s • c m- 5, and it increased cardiac output from
`2.75 ± 0.21 L/min to 4.11 ± 0.54 L/min, mixed venous
`oxygen saturation from 51.1% ± 3.4% to 66.3% ± 4 . 1 %,
`and arterial oxygen saturation from 90.6% ± 2.7% to
`93.8% ± 23% (P < 0.05 for all changes). Mean systemic
`arterial pressure was only slightly affected. The responses
`lasted for 10 to 30 minutes after inhalation was termi(cid:173)
`nated. Aerosolized iloprost had an identical efficacy pro(cid:173)
`file but was associated with a longer duration of the
`pulmonary vasodilator/ effect (60 min to 120 min). In
`comparison, intravenous prostacyclin reduced pulmonary
`vascular resistance with corresponding efficacy but pro(cid:173)
`duced a more pronounced decline in systemic artery pres(cid:173)
`sure and no clinically significant decrease in pulmonary
`artery pressure. Nitric oxide and 02 were less potent pul(cid:173)
`monary vasodilators in these patients. In one patient, 1
`year of therapy with aerosolized iloprost (100 /u,g/d in six
`aerosol doses) resulted in sustained efficacy of the inhaled
`vasodilator regimen and clinical improvement.
`
`Conclusion: Aerosolization of prostacyclin or its stable
`analog iloprost causes selective pulmonary vasodilatation,
`increases cardiac output, and improves venous and arterial
`oxygenation in patients with severe pulmonary hyperten(cid:173)
`sion. Thus, it may offer a new strategy for treatment of this
`disease.
`
`Ann Intern Med. 1996;124:820-824.
`
`Primary pulmonary hypertension is a progressive
`
`fatal disease of unknown cause, and patients
`with this condition have a median life expectancy of
`less than 3 years after diagnosis (1). Death is most
`closely associated with an increase in pulmonary
`artery pressure and right atrial pressure and a de(cid:173)
`crease in cardiac output due to failure of the right
`side of the heart. The responsiveness of pulmonary
`hypertension to various vasodilator agents led to the
`speculation that, concomitant with vascular remod(cid:173)
`eling processes, persistent vasoconstriction is an im(cid:173)
`portant feature of the disease. Long-term use of
`calcium-channel blockers improves the survival rate
`in approximately 25% of patients; in such patients,
`the response to these drugs is a substantial decrease
`in pulmonary artery pressure and pulmonary vascu(cid:173)
`lar resistance (2). The main hazards of this therapy
`are systemic hypotension and worsening of right
`ventricular function. Intravenous prostacyclin is a
`potent pulmonary vasodilator in patients with pri(cid:173)
`mary pulmonary hypertension, but it requires con(cid:173)
`tinuous intravenous access and, like calcium antag(cid:173)
`onists, lacks selectivity for the lung vasculature (3-6).
`We recently used aerosol techniques for preferential
`distribution of prostacyclin to well-ventilated lung
`areas, thereby achieving selective pulmonary vasodi(cid:173)
`latation with a concomitant improvement of venti-
`
`Aerosolized Prostacyclin and
`Iloprost in Severe Pulmonary
`Hypertension
`Horst Olschewski, MD; Dieter Walmrath, MD;
`Ralph Schermuly, PhD; H. Ardeshir Ghofrani, MD;
`Friedrich Grimminger, MD, PhD; and
`Werner Seeger, MD
`
`Objective: To compare the effects of aerosolization of
`prostacyclin and its stable analog iloprost with those of
`nasal oxygen, inhaled nitric oxide, and intravenous prosta(cid:173)
`cyclin on hemodynamics and gas exchange in patients with
`severe pulmonary hypertension.
`
`Design: Open uncontrolled trial.
`
`Setting: Justus-Liebig-University, Giessen, Germany.
`
`Patients: 4 patients with primary pulmonary hyperten(cid:173)
`sion and 2 patients with severe pulmonary hypertension
`associated with calcinosis, the Raynaud phenomenon,
`esophageal dysfunction, sclerodactyly, and telangiectasia
`(the CREST syndrome). All were classified as New York
`Heart Association class III or class IV.
`
`Intervention: Short-term applications of 02, inhaled ni(cid:173)
`tric oxide, intravenous prostacyclin, aerosolized prostacy(cid:173)
`clin, and aerosolized iloprost during repeated catheter
`investigation of the right side of the heart within a
`1-month period. One patient had long-term therapy with
`inhaled iloprost.
`
`Results: Aerosolized prostacyclin decreased pulmonary
`artery pressure in 6 patients from (mean ± SE) 62.3 ± 4.1
`
`From Justus-Liebig-University, Giessen, Germany. For current author ad(cid:173)
`dresses, see end of text.
`
`820
`
`1 May 1996 • Annals of Internal Medicine • Volume 124 • Number 9
`
`Downloaded From: http://annals.org/ by Carolyn Lundin on 12/09/2015
`
`WATSON LABORATORIES, INC. , IPR2017-01621, Ex. 1028, p. 1 of 5
`
`
`
`lation-perfusion matching in mechanically venti(cid:173)
`lated patients with the adult respiratory distress
`syndrome (7). This approach has also been shown
`to be effective in hypoxia-induced pulmonary hyper(cid:173)
`tension in dogs (8). We extended this strategy by
`nebulizing prostacyclin and a stable analog, iloprost,
`in patients with severe pulmonary hypertension.
`
`Methods
`
`Four patients with primary pulmonary hypertension
`and two patients with severe pulmonary hyperten(cid:173)
`sion associated with calcinosis, the Raynaud phenom(cid:173)
`enon, esophageal dysfunction, sclerodactyly, and telan(cid:173)
`giectasia (the CREST syndrome), all of whom were
`classified as New York Heart Association class III or
`class IV. Patients gave informed consent to the test
`trial, which was approved by the institutional ethics
`committee of Justus-Liebig-University, Giessen, Ger(cid:173)
`many. Two of the patients with primary pulmonary
`hypertension (patient A, 30 years of age, and pa(cid:173)
`tient B, 31 years of age) developed the disease
`several months after pregnancy. In the other two
`patients with pulmonary hypertension (patient C, a
`37-year-old man, and patient D, a 56-year-old wom(cid:173)
`an), no trigger was identified. The patients with the
`CREST syndrome (patient E, a 56-year-old woman
`who had received a diagnosis of the syndrome 2
`years earlier, and patient F, a 54-year-old woman
`who had received this diagnosis 15 years earlier)
`had no involvement of the inner organs and had
`never received
`immunosuppressive
`therapy other
`than low-dose corticosteroids. Pulmonary embolism
`was excluded by pulmonary angiograms and venti(cid:173)
`lation and perfusion
`lung scans in each patient.
`Chest radiography and high-resolution computed to(cid:173)
`mography showed no lung fibrosis, and pulmonary
`function testing showed no evidence of obstructive
`or restrictive lung disease. Only one patient (patient
`B) responded to calcium antagonists according to
`the criteria given by Rich and colleagues (2). Long-
`term therapy included anticoagulation and diuretics
`(in all patients) and low-dose steroids (in patients E
`and F).
`thermodilution pulmonary artery
`A
`fiberoptic
`catheter was used to measure central venous pres(cid:173)
`sure, pulmonary artery pressure, pulmonary artery
`wedge pressure, cardiac output, and venous oxygen
`saturation; a femoral artery catheter was used to
`assess mean arterial pressure and arterial oxygen
`saturation. In each test trial, the following were
`administered:
`1. Oxygen, 2 to 8 L/min.
`2. Inhaled nitric oxide, 10 to 28 parts per million.
`3. Intravenous prostacyclin (epoprostenol sodium,
`Wellcome Research Laboratories, Beckenham, Kent,
`
`United Kingdom), increased in increments of 2.5 ng/
`kg-min-1 until patients had discomfort (thoracic op(cid:173)
`pression, heat, or headache) or until mean arterial
`pressure decreased to less than 70 mm Hg. The
`highest tolerable dose (5 to 7.5 ng/kg-min- 1) was
`continued for 20 minutes.
`4. Aerosolized prostacyclin (25 to 50 /xg of gly(cid:173)
`cine buffer per mL), jet-nebulized with room air at
`a pressure of 153 kPa (fluid flux, 0.15 mL/min; mass
`median aerodynamic diameter of particles, 2.9 jutm;
`geometric SD, 3.1, ascertained by impactor tech(cid:173)
`nique) and delivered to a spacer connected to the
`afferent limb of a y-valve mouthpiece for 15 minutes
`(total nebulized dose, 52 to 112 /utg).
`5. Aerosolized iloprost (Ilomedin, Schering AG,
`Berlin, Germany) (5 to 10 /ig of saline per mL),
`administered according to prostacyclin (total nebu(cid:173)
`lized dose, 9 to 21 /xg).
`Patients were tested on a separate day for the
`presence of calcium antagonists (data not shown).
`All other measurements were taken before, during,
`and after application of each test trial challenge.
`The time between the different maneuvers was suf(cid:173)
`ficient for a new stable baseline period. The number
`of challenges required 10 hours for one entire trial
`sequence. These trials were done twice within a
`1-month period in all patients, and the data were
`averaged (except for the data from patient B, who
`took part in only one test trial). One patient (pa(cid:173)
`tient E) subsequently began receiving
`long-term
`treatment with aerosolized iloprost (100 jmg/d, di(cid:173)
`vided into six aerosol doses), which has now been
`continued for 1.5 years.
`Values before and after challenge in six patients
`are given for each challenge and are expressed as
`mean ± SD. We tested these values for significance
`using the two-tailed Student Mest for intrapair dif(cid:173)
`ferences; these levels of significance are given.
`
`Results
`
`In all six patients, the high pulmonary artery
`pressure and pulmonary vascular resistance values,
`the low cardiac output and venous oxygen satura(cid:173)
`tion values, and the increased central venous pres(cid:173)
`sure value indicated advanced disease (Table 1). As
`we had anticipated, arterial and venous oxygen sat(cid:173)
`uration values increased with oxygen, but hemody(cid:173)
`namics improved only moderately. Inhaled nitric
`oxide substantially decreased pulmonary artery pres(cid:173)
`sure, pulmonary vascular resistance, and central ve(cid:173)
`nous pressure and increased cardiac output and ve(cid:173)
`nous oxygen saturation. Inhalation only moderately
`decreased systemic vascular resistance. After cessa(cid:173)
`tion of nitric oxide therapy, values returned to base(cid:173)
`line within 2 to 5 minutes. Intravenous prostacyclin
`
`1 May 1996 • Annals of Internal Medicine • Volume 124 • Number 9 821
`
`Downloaded From: http://annals.org/ by Carolyn Lundin on 12/09/2015
`
`WATSON LABORATORIES, INC. , IPR2017-01621, Ex. 1028, p. 2 of 5
`
`
`
`Table 1. Hemodynamics ar
`ange in Resp
`id Gas Exch<
`ite Exposure
`onse to Act
`
`to Vasoactr
`\fe Agents*
`
`Variable
`
`Before
`o2
`
`During
`o2
`
`Before
`NO
`
`During
`NO
`
`Before Intra(cid:173)
`venous PG
`
`During Intra- Before Aero- During Aero(cid:173)
`venous PG
`solized PG
`solized PG
`
`54.2 ± 4.3t
`60.4 ± 3.1
`2.80 ±0.17 3.48 ± 0.21t
`
`59.8 ±5.6
`62.7 ± 4.5
`2.94 ± 0.24 4.52 ± 0.42*
`
`50.8 ± 5.5*
`62.3 ±4.1
`2.75 ±0.21 4.11 ±0.54*
`
`Pulmonary artery pressure, mm Hg 58.2 ± 2.6
`55.2 ± 2.4
`Cardiac output, L/min
`2.76 ± 0.20 2.78 ±0.12
`Pulmonary vascular resistance,
`dyne/s- cm~5
`Central venous pressure, mm Hg
`Mean artery pressure, mm Hg
`Systemic vascular resistance,
`dyne/s -cm~5
`Heart rate, beats/min
`Arterial oxygen saturation, %
`Venous oxygen saturation, %
`
`1537 ± 173 1465 ± 138 1578 ± 146 1141 ± 141*
`13.3 ± 1.2
`13.7 ± 1.1
`10.2 ± 0 . 8*
`13.7 ±0.9
`96 ±4
`97 ± 5
`97 ± 5
`97 ± 5
`
`2459 ± 191 2496 ± 183 2413 ± 160 2064 ± 178*
`91.0 ± 3.4
`87.9 ± 3.1
`96.4 ± 3.2
`93.7 ± 3.2
`90.4 ± 1.9
`97.6 ± 1.0* 89.4 ± 2.9
`90.0 ± 3.1
`57.7 ±2.6
`65.1 ± 1.9*
`57.0 ± 3.8
`62.7 ± 5.0*
`
`1551 ± 186 1000 ± 158*
`13.8 ± 1.7 12.9 ±1.5
`87 ± 6*
`100 ± 6
`
`1721 ±253 1019 ± 203*
`10.2 ± 1.6*
`13.6 ±1.6
`96 ± 6
`90 ± 6
`
`2453 ± 206 1398 ± 169*
`101.8 ±4.2 106.5 ± 5.0*
`89.0 ± 3.3 88.8 ± 3.4
`66.7 ± 5.2*
`53.4 ± 4.5
`
`2400 ± 173 1680 ± 2 2 9*
`99.1 ±4.8
`102.1 ±4.2
`90.6 ± 2.7 93.8 ± 2.3*
`51.1 ±3.4
`66.3 ± 4 . 1*
`
`
`Values are prese
`
`nted as mean ± SE for six patients . The intravenous
`
`* 02 = nasal oxygen, 2 to 8 L/min; NO
`= inhaled nitric (
`PG = prostacyclin.
`Dxide, 10 to 28 p.
`arts per million;
`min - 1; the aeros>
`prostacyclin dose was 5 to 7.5 ng/kg •
`olized prostacyclin
`dose was 52 to
`112 ^ g.
`
`t P < 0.05 for comparisons between vali jes before and al
`fter drug therapy,
`
`$ P < 0.01 for comparisons between vali jes before and al
`fter drug therapy.
`
`substantially decreased pulmonary vascular resis(cid:173)
`tance and increased cardiac output in all patients,
`resulting in a modest decline of pulmonary artery
`pressure but a substantial decrease in systemic ar(cid:173)
`terial pressure and an increase in heart rate due to
`peripheral vasodilatation. Aerosolized prostacyclin
`achieved the same reduction in pulmonary vascular
`resistance with a smaller increase in cardiac output
`but a significant decline in pulmonary artery pres(cid:173)
`sure. Moreover, arterial oxygen saturation values
`were increased with aerosolized prostacyclin but re(cid:173)
`mained unchanged, on average, after
`intravenous
`prostanoid application. In all patients, the stable
`prostacyclin analog iloprost caused nearly identical
`changes in hemodynamics and gas exchange (exam(cid:173)
`ple given in Figure 1, further data not shown). The
`effects of iloprost, however, lasted longer: After ter(cid:173)
`mination of aerosolization, prostacyclin-induced
`changes returned to baseline within 10 to 30 min(cid:173)
`utes and iloprost-induced changes, within 60 to 120
`minutes.
`Because response to inhaled prostanoids was fa(cid:173)
`vorable, we initiated regular long-term therapy with
`aerosolized iloprost in one patient (patient E; see
`Figure 1). Daily inhalation of the prostanoid was
`well tolerated, and sustained responsiveness of the
`pulmonary vasculature was seen over a 1-year pe(cid:173)
`riod. This patient's New York Heart Association
`class improved from IV to III, and her CREST-
`related skin lesions healed.
`
`Discussion
`
`Prostacyclin is one of the most potent pulmonary
`vasodilatory agents available for clinical purposes.
`Moreover, in severe pulmonary hypertension, local
`deficiencies of prostacyclin may play a role in the
`genesis and progression of vascular remodeling (4).
`Accordingly, long-term therapy with high doses of
`
`prostacyclin resulted in increased life expectancy
`and exercise tolerance (6). We describe a technique
`of intra-alveolar deposition of prostacyclin in pa(cid:173)
`tients with severe pulmonary hypertension. The
`acute effects on pulmonary vascular resistance seen
`with this technique corresponded to those of intra(cid:173)
`venous prostacyclin, but selectivity for the pulmo(cid:173)
`nary circulation was achieved, as indicated by a
`substantial decrease of pulmonary artery pressure
`and a smaller effect on systemic arterial pressure.
`Moreover, preferential distribution of the aerosol(cid:173)
`ized vasodilator to the best-ventilated lung areas,
`which improved ventilation-perfusion matching, was
`suggested by an increase in arterial oxygen satura(cid:173)
`tion. This effect was even superior to that of inhaled
`nitric oxide, which has previously been shown to
`cause selective pulmonary vasodilatation in patients
`with severe pulmonary hypertension (9, 10). We
`found that the nitric oxide-evoked decrease in pul(cid:173)
`monary vascular resistance was less than that seen
`in response to inhaled or intravenous prostacyclin in
`all patients.
`Iloprost showed an identical efficacy profile but a
`longer duration of pulmonary vasodilatation in com(cid:173)
`parison with prostacyclin. It was well tolerated in
`the patient who received long-term therapy with
`daily inhalations of this drug, and sustained respon(cid:173)
`siveness of the pulmonary vasculature was seen. In(cid:173)
`terestingly, the beneficial effects were achieved by a
`total aerosolized
`iloprost quantity of 100 /xg/d,
`which corresponds to a dosage of only approxi(cid:173)
`mately 1.1 ng/kg • min- 1. This is remarkable because
`the aerosol fraction being deposited in the alveolar
`spaces is commonly estimated to range from less
`than 10% to 15%. In addition, contrasting with the
`current experience with continuous prostacyclin in(cid:173)
`fusion (6), we noted no need to increase the dosage.
`Our study is limited in that only a small number of
`patients were investigated, and long-term applica(cid:173)
`tion was only used in one patient. Moreover, the
`
`822
`
`1 May 1996 • Annals of Internal Medicine • Volume 124 • Number 9
`
`Downloaded From: http://annals.org/ by Carolyn Lundin on 12/09/2015
`
`WATSON LABORATORIES, INC. , IPR2017-01621, Ex. 1028, p. 3 of 5
`
`
`
`Figure 1. Short- and long-term effects of iloprost inhalation in one patient. Left. Example of the hemodynamic and gas exchange response to one
`15-minute iloprost inhalation (white bar, ILO) in patient E, who received long-term treatment with inhalations of 100 /xg of iloprost daily. During inhalation,
`cardiac output (CO) and venous oxygen saturation (Svo2) increased sharply. Simultaneously, mean pulmonary artery pressure (PAP) and central venous pressure
`(CVP) decreased. Note the almost constant mean arterial pressure (MAP) and increasing arterial oxygen saturation (Sao2) in response to inhalation. Right. Time
`course of hemodynamics and arterial oxygen saturation in response to long-term iloprost aerosolization, 100 ju,g/d in six aerosol doses, in patient E. Baseline
`values (open squares, assessed in the morning before daily inhalation) of Sao2, pulmonary vascular resistance (PVR), PAP, and CO, and those values obtained
`immediately after aerosolization of the first 17 /u,g of the daily iloprost dose are shown. Note that the acute responses to inhalation (arrows) did not decrease
`during long-term therapy.
`
`short duration of pulmonary vasodilatation even in
`response to the stable prostacyclin analog requires
`frequent dosage of the aerosolized drug. The long-
`term effects of repetitive alveolar prostacyclin dep(cid:173)
`osition in addition to pulmonary vascular changes
`need to be carefully addressed in further studies.
`In conclusion, aerosolization of prostacyclin and
`its stable analog iloprost, substances with well-
`known pharmacologic profiles, caused marked pul(cid:173)
`monary vasodilatation in patients with severe pul(cid:173)
`monary hypertension. Inhalation appeared to be
`reliable and safe. In contrast to continuous infusion
`of these prostanoids, which is cumbersome and
`prone to hazards caused by intravenous line com(cid:173)
`plications, selectivity for the pulmonary circulation
`was achieved concomitantly with an improvement of
`arterial oxygenation. This strategy may thus be suit(cid:173)
`able for bridging to lung or heart-lung transplanta(cid:173)
`tion. Moreover, considering that local deficiencies
`of prostacyclin may play a role in the genesis of
`severe pulmonary hypertension, high doses of lo(cid:173)
`
`cally deposited prostanoids might favorably affect
`vascular remodeling processes in addition to having
`immediate vasodilatory efficacy.
`
`Acknowledgments: The authors thank Dr. R.H. Bodeker, Institute
`of Medical Statistics, for statistical discussion and Dr. R. Snipes,
`Institute of Anatomy, Justus-Liebig-University, Giessen, Ger(cid:173)
`many, for linguistic revision of the manuscript.
`
`Requests for Reprints: Werner Seeger, MD, Department of Med(cid:173)
`icine, Justus-Liebig-University, Klinikstrasse 36, D-35392 Giessen,
`Germany.
`
`Current Author Addresses: Drs. Olschewski, Walmrath, Schermuly,
`Ghofrani, Grimminger, and Seeger: Department of Medicine,
`Justus-Liebig-University, Klinikstrasse 36, D-35392 Giessen, Ger(cid:173)
`many.
`
`References
`
`1. D'Alonzo GE, Barst RJ, Ayres SM, Bergofsky E, Brundage BH, Detre
`KM, et al. Survival in patients with primary pulmonary hypertension. Results
`from a national prospective registry. Ann Intern Med. 1991;115:343-9.
`
`1 May 1996 • Annals of Internal Medicine • Volume 124 • Number 9
`
`823
`
`Downloaded From: http://annals.org/ by Carolyn Lundin on 12/09/2015
`
`WATSON LABORATORIES, INC. , IPR2017-01621, Ex. 1028, p. 4 of 5
`
`
`
`2. Rich S, Kaufmann E, Levy PS. The effect of high doses of calcium-channel
`blockers on survival in primary pulmonary hypertension. N Engl J Med. 1992;
`327:76-81.
`3. Higenbottam T. Wheeldon D, Wells Ff Wallwork J. Long-term treatment
`of primary pulmonary hypertension with continuous intravenous epoprostenol
`(prostacyclin). Lancet. 1984;1:1046-7.
`4. Higenbottam T. The place of prostacyclin in the clinical management of
`primary pulmonary hypertension. Am Rev Respir Dis. 1987;136:782-5.
`5. Rubin U, Mendoza J, Hood M, McGoon M, Barst R, Williams WB, et
`al. Treatment of primary pulmonary hypertension with continuous intrave(cid:173)
`nous prostacyclin (epoprostenol). Results of a randomized trial. Ann Intern
`Med. 1990;112:485-91.
`6. Barst RJ, Rubin U, McGoon MD, Caldwell EJ, Long WA, Levy PS.
`Survival in primary pulmonary hypertension with long-term continuous intra(cid:173)
`
`venous prostacyclin. Ann Intern Med. 1994;121:409-15.
`7. Walmrath Df Schneider T, Pilch J, Grimminger F, Seeger W. Aerosolised
`prostacyclin in adult respiratory distress syndrome. Lancet. 1993;342:961-2.
`8. Welte M, Zwissler B, Habazettl H, Messmer K. PGI2 aerosol versus nitric
`oxide for selective pulmonary vasodilation in hypoxic pulmonary vasoconstric(cid:173)
`tion. Eur Surg Res. 1993;25:329-40.
`9. Pepke-Zaba J, Higenbottam TW, Dinh-Xuan AT, Stone D, Wallwork J.
`Inhaled nitric oxide as a cause of selective pulmonary vasodilatation in pul(cid:173)
`monary hypertension. Lancet. 1991;338:1173-4.
`10. Zapol W M, Rimar S, Gillis N, Marietta M, Bosken CH. Nitric oxide and
`the lung. Am J Respir Crit Care Med. 1994;149:1375-80.
`
`1996 American College of Physicians
`
`In my mechanical way I read all I could find about the affliction, and each word more
`brought its own gloom. Emphysema, it emerged, could become a torture of the body
`beyond even my mother's suffering or the holocaust of cancer itself.... Somehow
`through all the null medical words—generalized overdistension, difficulty of exhalation,
`excess mucus—I came to picture the disaster happening in my father's lungs as a
`pattern like the splotched burning of a sagebrush fire. Perhaps it was the years of
`blue haze drifting up from his cigarettes that made me think of smolder and
`flamelick. For whatever reason, the image came to me of the black turf that such a
`fire spreads in its steady searing fan across the land, and the thought too that there
`would be no grass-bright greening in this fire's wake as time passed. Only char and
`more char.
`
`Ivan Doig
`This House of Sky: Landscape of a Western Mind
`
`Submitted by:
`Erik R. Swendon, MD
`Department of Veterans Affairs Medical Center
`Seattle, WA 98108
`
`Submissions from readers are welcomed. If the quotation is published, the sender's name will be acknowl(cid:173)
`edged. Please include a complete citation, as done for any reference.—The Editor
`
`824
`
`1 May 1996 • Annals of Internal Medicine • Volume 124 • Number 9
`
`Downloaded From: http://annals.org/ by Carolyn Lundin on 12/09/2015
`
`WATSON LABORATORIES, INC. , IPR2017-01621, Ex. 1028, p. 5 of 5
`
`