`
`The Role of the o,-Adrenoceptor
`Agonist Dexmedetomidine in Postsurgical
`Sedation in the Intensive Care Unit
`
`Eike Martin*
`Graham Ramsay*
`Jean Mantz*
`S. T. John Sum-Ping’
`
`Dexmedetomidine was evaluated for sedation of 401 post-
`surgical patients in this double-blind, randomized, placebo-
`controlled, multicenter trial. Dexmedetomidine or saline
`was started on arrival in the intensive care unit (ICU) (1.0
`mcg/kg for 10 minutes),
`then titrated at 0.2 to 0.7
`mceg/kg/h to effect. Patients could be given propofol if
`necessary. Morphine was administered for pain. Sixty per-
`cent of the dexmedetomidine patients required no other
`sedative to maintain an RSS > 3; 21% required < 50 mg
`propofol In contrast, 76% of the control group received
`propofol; 59% required > 50 mg. Dexmedetomidine
`patients required significantly less morphine for pain
`relief (P < .001). Continuously given throughout the ICU
`stay, dexmedetomidine had no effect on respiratory rate,
`oxygen saturation, duration of weaning, or times to extu-
`bation. Nurses judged the dexmedetomidine patients were
`easier to manage. Later, fewer dexmedetomidine patients
`remembered pain or discomfort. The majority of
`dexmedetomidine patients maintained blood pressures
`within normal range, without rebound. Hypertension,
`atelectasis, and rigors occurred more frequently in the
`control group, while hypotension and bradycardia
`occurred more frequently in the dexmedetomidine group.
`Preoperative cardiovascular conditions were not risk fac-
`tors for dexmedetomidinepatients.
`
`Key words: a@-adrenoceptors, imidazoles, sympatholysis, anxi-
`ety, cardiac artery bypass graft, respiratory system
`
`From the ‘Klinik fiir Anaesthesiologie Universitatskliniken,
`Heidelberg, Germany; the 'Academisch Ziekenhuis Maastricht,
`Maastricht,
`the Netherlands;
`the ‘Department d’Anesthesie,
`Hé6pital Bichat, Paris, France; and the ‘Department of
`Anesthesiology, Duke University and Durham Veterans
`Administration Medical Centers, Durham, NC.
`
`Received Jun 24, 2002, and in revised form Aug 30, 2002.
`Accepted for publication Sep 5, 2002.
`Address correspondence to Dr Mantz, Department d’Anesthesie,
`HO6pital Bichat, 46 rue Henri Huchard, 75018 Paris, France, or e-
`mail: jean.mantz@bch.ap-hop-paris.fr.
`Martin E, Ramsay G, Mantz J, Sum-Ping STJ. The Role of the a-
`Adrenoceptor Agonist Dexmedetomidine in Postsurgical Sedation
`in the Intensive Care Unit. J Intensive Care Med 2003;18:29-41.
`DOI: 10.1177/0885066602239122
`
`Ideally, postsurgical patient care in the intensive
`care unit (ICU) should minimize stress and sympa-
`thetic nervous system responses, relieve pain and
`anxiety, facilitate diagnostic and therapeutic proce-
`dures, and permit communication with patients
`withoutinterrupting sedation—all without compro-
`mising hemodynamic orrespiratory stability or pro-
`longing time in the ICU [1-3]. Although new seda-
`tion protocols continue to improve the quality of
`ICU sedation and outcomes, fundamental problems
`remain that are inherent to the agents commonly
`used. Propofol and midazolam can depress respi-
`ratory drive [4-6]. For this reason, both are usually
`stopped before extubation. Whereas midazolam
`reduces opioid use, propofol has no analgesic
`properties. The combination of propofol or mida-
`zolam with opioids can result in a disoriented and
`unresponsive patient, at risk for hypoxemia orres-
`piratory depression [7,8].
`Recently, the adrenoceptor agonist dexmed-
`etomidine was approved in the United States for
`short-term sedation of ICU patients. Drugs of this
`class bind to transmembrane G-protein-binding
`receptors, rapidly initiating a cascade of physiolog-
`ical events. When the ,-adrenoceptors, which
`modulate wakefulness in the locus coeruleus are
`activated, a dose-dependent hypnotic sedation
`results. Stimulation of presynaptic &,-adrenoceptors
`reduces sympathetic tone and increases parasym-
`pathetic tone, resulting in a decrease in myocardial
`oxygen requirements [9]. Predictably, this class of
`drugs reduces blood pressure and heart ate.
`Clonidine has been used for many years to reduce
`blood pressure, opioid use, and the hemodynamic
`stresses associated with surgery [10,11]. Another
`important benefit of this class of drugs is that they
`have nosignificant effect on respiratory drive, even
`when used with opioids [9,12]. A number of stud-
`ies have verified these benefits of dexmedetomi-
`dine whenusedfor sedation in critically ill patients
`
`Copyright © 2003 Sage Publications
`
`29
`
`Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on August 10, 2016
`
`Petition for Inter Partes Review of US 8,455,527
`Amneal Pharmaceuticals LLC — Exhibit 1062 — Page 29
`
`
`
`Martin et al
`
`[9,13]. Because dexmedetomidine has a higher a,
`to a, binding affinity (1300:1) than clonidine (39:1),
`someofthe adverse effects associated with a, stim-
`ulation may be avoided [14]. Studies have shown
`that dexmedetomidine reduces plasma cate-
`cholamine concentrations and hemodynamicstress
`responses to endotrachealinsertion, surgical stress,
`awakening from anesthesia, and extubation [15-19].
`During short-term sedation, it does not inhibit adre-
`nal steroidogenesis [20]. Although propofol is an
`effective sedative,
`it has no analgesic effect, and
`the adequately sedated patient is unresponsive. In
`contrast, the dexmedetomidine-sedated patient is
`easily awakened to participate in diagnostic and
`therapeutic procedures without stopping the seda-
`tive [21,22]. Although there are many demonstrated
`advantages to the use of dexmedetomidine, more
`large randomized studies are necessary to deter-
`mineits value in the clinical setting.
`The objective of this double-blind randomized
`study was to evaluate the efficacy and safety of
`dexmedetomidine for short-term sedation of post-
`surgical, ventilated patients compared to a placebo
`control. Patients received identical
`infusions of
`dexmedetomidine or saline from their entry into the
`ICU through a minimum of 6 hours before extuba-
`tion, and at
`least 6 hours postextubation. They
`were given propofol if their sedation levels were
`considered inadequate by Ramsay sedation score
`(RSS) [23]. The primary endpoint for the study was
`the total dose of propofol required to maintain
`sedation at an RSS 2 3 during assisted ventilation.
`Secondary endpoints were the dose of morphine
`for analgesia, weaning duration, time to extubation,
`and nurses’ assessments of patient management.
`
`Materials and Methods
`
`Subjects for this study were scheduled for major
`surgical procedures that were expected to require
`a minimum of6 hours of postsurgical assisted ven-
`tilation. A total of 401 patients were enrolled in 34
`sites in Europe and Canada.
`Excluded from this study were females if preg-
`nant or lactating and patients whose condition or
`responses could be difficult to evaluate in a blind-
`ed trial (eg, had serious central nervous system
`trauma or intracranial surgery), patients who had
`unstable or uncontrolled diabetes, patients who
`were grossly obese, and patients who were hospi-
`talized for a drug overdose. Discontinued from the
`study were patients who developed excessive
`bleeding that required a return to surgery; patients
`who received neuromuscular blocks, epidural, or
`
`Table 1. Ramsay Sedation Scale
`Score
`Observation
`
`anWwmwNFe
`
`Anxious, agitated, or restless
`Cooperative, oriented, and tranquil
`Responsive to commands
`Asleep, but with brisk response to light
`glabellar tap or loud auditory stimulus
`Asleep, sluggish response to glabellar tap or
`auditory stimulus
`Asleep, no response
`
`spinal analgesia during their ICU stay; and patients
`who had clinically significant arrhythmia or any
`other cardiac condition or other factor that, in the
`investigators opinion, might increase the risk to
`those patients or preclude obtaining satisfactory
`study data.
`informed consent
`The protocol, amendments,
`form, and all other forms of patient information
`related to the study were reviewed and approved
`by an independent ethics committee that complied
`with Food and Drug Administration regulations and
`each country’s regulatory requirements. A volun-
`tary, written informed consent form was signed by
`each patient (or representative) after the nature of
`the study was explained and prior to any study-
`related procedure.
`Dexmedetomidine HCl (100 mcg/mL base) in
`0.9% NaCl and the 0.9% NaCl solutions were sup-
`plied by Abbott Laboratories. Each site provided
`propofol, morphine, and all other supplies and
`equipment. The site pharmacist prepared the cor-
`rect dilutions of dexmedetomidine (4 mcg/mL, or 8
`meg/mL in 0.9% sodium chloride) and labeled the
`syringes according to the randomization code. Both
`solutions were identical in appearance and viscos-
`ity. All laboratory staff, all CRO staff, and all Abbott
`Laboratory personnel involved with the conduct
`and/or analysis of this study were blinded to the
`randomization code. Patient assignments remained
`blinded until after the study was completed, all
`clinical data had been screened, and all patients
`were evaluated. In the event of an emergency, the
`investigator could open the sealed blind-breaker
`envelope.
`
`Study Design
`
`received
`randomized to group A_
`Patients
`dexmedetomidine intravenously; group B patients
`received 0.9% saline (see Fig 1
`for the study
`design). Both solutions were called “study drug.”
`
`30
`
`Journal of Intensive Care Medicine 18(1); 2003
`
`Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on August 10, 2016
`
`Petition for Inter Partes Review of US 8,455,527
`Amneal Pharmaceuticals LLC — Exhibit 1062 — Page 30
`
`
`
`Dexmedetomidine for Postsurgical Patient Care
`
`Study Design
`
`Morphine
`
`Additional medications as needed:
`
`Propofol* (sedation)
`
`Dexmedetomidine Infusion
`
`(24 hours maximum)
`
`JV
`
`entilator
`(minimum 6 hours)
`
`Postextubation period
`(minimum 6 hours)
`
`24-hourfollow-up
`(from infusion end)
`
`
`
`* Continuousinfusion optional after 3 bolus doses within any 2-hour period
`
`Fig 1. Study design. Study drug infusions began within 1 hour after the patients entered the intensive care unit (ICU)
`and continued uninterrupted for up to 24 hours. Study protocols required a minimum of 6 hoursofassisted ventilation
`before extubation and a minimum of 6 hours with study drug administration after extubation.
`
`There were no restrictions on intraoperative drug
`use. If a patient required sedation to assist in the
`transport from the operating room to the ICU prior
`to the start of study drug, a 0.2-mg/kg bolus of
`propofol could be given. Study drug infusions were
`started as soon as possible after the patient entered
`the ICU, but within 1 hour. Baseline values were
`the last measurements recorded before the start of
`study drug infusion. The infusion pump wasset to
`give 1.0 mcg/kg (dexmedetomidine or saline) for
`10 minutes (loading dose) and then was reduced to
`0.4 mcg/kg/h. The latter rate could be adjusted
`within the range of 0.2 to 0.7 mcg/kg/h as neces-
`sary to achieve and maintain an RSS 2 3 while the
`patient was ontheventilator. If sedation could not
`be maintained within the protocol-defined range
`and the infusion rate was at the maximum of0.7
`mcg/kg/h, patients could be given a bolus of 0.2
`mg/kg propofol intravenously. If sedation wasstill
`considered inadequate after 3 bolus doses, a con-
`tinuous infusion of propofol wasstarted (at a rate
`of 0.5 to 4.0 mg/kg/h). Following extubation, the
`infusion rate was to be adjusted to achieve a
`Ramsay score 2 2.
`Incremental
`(2 mg) doses of morphine were
`given intravenously to patients in both groups as
`necessary for pain relief at any time during the
`study. No pain scale was used. The need for anal-
`
`gesia was assessed individually, either by direct
`communication or based on indirect symptoms of
`pain (eg, sweating, tachycardia, hypertension, or
`excessive movement).
`All patients were kept on the ventilator for a min-
`imum of 6 hours after entry into the ICU. Infusions
`of dexmedetomidine or saline were continued
`without
`interruption while in the ICU,
`through
`weaning and extubation, and for a minimum of 6
`hours postextubation; total time was < 24 hours.
`Blood samples for hematology, blood gases, and
`blood chemistry assessments were collected before
`dosing and approximately 24 hours after the end of
`study drug infusion. Systolic and diastolic arterial
`blood pressure, heart rate, and respiratory rate
`were recorded at protocol-specified time points
`(Table 2). SpO, was monitored continuously.
`Efficacy was evaluated by measuring the amount
`of propofol,
`in addition to dexmedetomidine or
`saline (placebo), that was required to achieve and
`maintain an RSS = 3 during assisted ventilation.
`Secondary efficacy variables were as follows.
`
`1. The total dose of morphine administeredfor pain.
`Morphine use was compared between groups dur-
`ing thefirst 6.5 hours of study drug infusion (when
`all patients received assisted ventilation and the
`period of most
`intense analgesic requirements)
`and from 6.5 hours after the start of study drug
`
`Journal of Intensive Care Medicine 18(1); 2003
`
`31
`
`Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on August 10, 2016
`
`Petition for Inter Partes Review of US 8,455,527
`Amneal Pharmaceuticals LLC — Exhibit 1062 — Page 31
`
`
`
`Martin et al
`
`Table 2. Study Procedures and Schedule of Assessments
`
`Assessment
`
`After First
`Baseline
`Hour to
`First Hour
`(Prior to Start
`Screening
`Stop of Study
`of Study
`of Study
`(< 7 Days
`Before Dosing) Drug Infusion) Drug Infusion Drug Infusion
`
`Prior to
`Discharge
`(24 Hours
` Postinfusion
`Stop)
`
`Informed consent/medical history
`Laboratory testing
`Physical examination
`12-lead electrocardiogram
`xX
`Cardiac telemetry monitoring
`q60 min>
`xX
`SBP, DBP, HR, RR
`q60 min?
`x
`CVP
`q60 min>
`x
`Spo,
`q6 hours
`xX
`Temperature
`At end of infusion
`xX
`Blood gases
`
`
`Ramsay ql0 min*®=.q60 min?x
`Pain assessments
`xX
`PRN
`PRN
`xX
`
`X
`
`X
`X
`
`xX
`
`xX
`
`xX
`
`Continuous
`ql0 min*
`ql0 min#
`ql0 min*
`
`xX
`xX
`xX
`
`x*
`
`xe
`xe
`
`Cardiac output and central venous pressure (CVP) were assessed asclinically indicated and only at a subgroup of sites. SBP = systolic
`blood pressure, DBP = diastolic blood pressure, HR = heart rate, RR = respiratory rate, SpO, = oxygen saturation by pulse oximeter,
`PRN = as needed.
`a. Assessed every 10 + 2 minutes for 30 minutes after the start of the study drug infusion and at 60 minutes after the start of study drug.
`b. Assessed every 60 + 5 minutes from 1 hourafter the start of the study drug infusion to the end of study drug infusion.
`c. Assessed every 3 hours until study completion.
`d. Also assessed prior to and 10 minutes after any rate change in study drug or administration of additional sedative.
`
`(postextubation for most patients) to the end of
`study drug infusion.
`2. The duration of weaning (the difference between
`initiation of weaning from the ventilator and
`readiness for extubation). Guidelines for weaning
`were as follows. After the patient was stabilized in
`the ICU,the fraction of inspired oxygen (FiO) was
`to be decreased no more frequently than every 30
`minutes by 0.10 decrements (if SpO, exceeded
`92%) to a goalFiO, of no more than 0.40. Patients
`were weaned from the ventilator when they were
`responsive, were hemodynamically stable, had no
`shivering or bleeding requiring treatment, and reg-
`istered a temperature above 36°C.
`3. Time to extubation (the difference between ICU
`arrival and when thepatient was considered ready
`for extubation). The endotracheal tube could be
`removed if the following criteria were met and the
`investigator deemed it appropriate: (1) the patient
`was awake or arousable, neurologically intact,
`cooperative, and comfortable and (2) the patient
`had an FiO, value < 0.4, positive end-expiratory
`pressure (PEEP) < 5 cm H,O,and pressure support
`< 10 cm HO.The patient’s lung mechanics were
`as follows: minute ventilation expired > 4 L/min
`but < 15 L/min,tidal volume > 5 mL/kg, and spon-
`taneousrespiratory rate < 25 breaths/min.
`4. Nurses’ assessments (see the appendix). For any
`nursing shift that started or ended during assisted
`ventilation, nurses recorded their ratings of the
`quality of sedation, the ability of the patientto tol-
`erate the endotracheal tube/ventilator and the ICU,
`the ease of communication with the patient, and
`ease of management. Scores from each of these
`assessments were summed to arrive at a compos-
`ite score defined as the Patient ManagementIndex.
`We used the Cochran-Mantel-Haenszelstatistic to
`test the significance of the differences between
`
`each assessment, simultaneously adjusting forsite-
`specific nursing practices.
`5. Patient questionnaires (see the appendix). At the
`end of their stay in the ICU, each patient was
`asked about their recall of their ICU experience,
`discomfort, and overall satisfaction.
`
`Throughoutthe study, the investigator closely mon-
`itored each patientfor clinical or laboratory evidence
`of an adverse event (AE). The investigator rated the
`severity of each AE as mild (transient andeasily tol-
`erated), moderate (causing discomfort), or severe
`(causing considerable interference with the patient’s
`usual activities, incapacitating, or life threatening).
`If life-threatening or prolonging hospitalization, the
`severe AE wasalso rated serious. Before breaking
`the blind, the investigator also assessed the possi-
`ble relationship of an AE to study drug (both
`dexmedetomidine and saline were called study
`drug to maintain the study blind). Hypertension
`and hypotension were defined by each investigator
`according to his or her evaluation of each patient’s
`age, history, and condition, as well as the absolute
`change in blood pressure. These were further rated
`in terms of severity—mild, moderate, or severe.
`
`Statistical Analyses
`
`To detect a clinically significant difference between
`the dexmedetomidine and placebo groups in the
`total dose of propofol administered at the .05 (2-
`tailed) level with 80% power, the calculated sample
`
`32
`
`Journal of Intensive Care Medicine 18(1); 2003
`
`Downloaded fromjic.sagepub.com at UNIV OF PITTSBURGH on August 10, 2016
`
`Petition for Inter Partes Review of US 8,455,527
`Amneal Pharmaceuticals LLC — Exhibit 1062 — Page 32
`
`
`
`Dexmedetomidine for Postsurgical Patient Care
`
`Table 3. Patient Demographics and Disposition
`
`Intent-to-treat patients
`Meanage, years (range)
`Gender n (%)
`Female
`Male
`Weight (kg + SD)
`Female
`Male
`Evaluable patients
`Reasons for nonevaluability
`Insufficient time on ventilator
`Received disallowed medication during study drug therapy
`Enrolled twice
`
`Dexmedetomidine
`
`Control
`
`203
`60.2 (17-88)
`
`62 81)
`141 (69)
`
`64.3 + 11.12
`76.8 + 12.31
`200
`
`0
`3
`0
`
`198
`62.5 (17-87)
`
`64 (32)
`134 (68)
`
`64.1 + 11.61
`76.5 + 12.35
`191
`
`1
`5
`1
`
`size was 150 patients per treatment group. This was
`based on the following assumptions: propofol use
`over 24 hours would be 70 mg/kg for the placebo
`group and 20 mg/kg for the dexmedetomidine
`group, the effect size would be 0.35, and 90% of
`the patients enrolled would be evaluable.
`Treatment groups were compared using analysis
`of variance. Differences in the distributions of
`patients in each category between groups were
`tested with a chi-square statistic. Treatment differ-
`ences for weaning duration and time to extubation
`were also analyzed by Kaplan-Meier survival
`curves with log-rank analysis. If a patient was dis-
`continued from the study for any reason, duration
`of weaning and time to extubation were based on
`the length of infusion at discontinuation. If extuba-
`tion had not occurred by 24 hours, extubation time
`was considered 24 hours. Where indicated in the
`text, differences were also compared by Fisher’s
`exact test. Statistical software used was SAS version
`
`6.12.
`All results in this report are based on the intent-
`to-treat population (patients who received any
`amount of study drug).
`
`Results
`
`The intent-to-treat data set consisted of 401 post-
`surgical patients. There were nosignificant differ-
`ences between groups in baseline demographics or
`clinical characteristics (Table 3). Surgeries were of
`4 types: cardiac (45%), laparotomy (30%), head and
`neck (7%), and other (18%). The majority of the
`cardiac surgeries were coronary artery bypass grafts
`(CABGs). Complications resulted in discontinuation
`from the study for 13 dexmedetomidine and 7 con-
`trol group patients. Study drug was considered a
`
`possibly contributing factor for 4 of 13 dexmedeto-
`midine and 2 of 7 control group patients. Four
`patients died during the study (3 dexmedetomi-
`dine,
`1 control). None of the events leading to
`death were related to study drug. Most complica-
`tions observed during the study were mild or mod-
`erate in severity, and events considered severe
`occurred at the same rate in both groups (12%).
`During assisted ventilation, both groups were
`sedated to similar levels (3.4 + 0.04, dexmedetomi-
`dine; 3.1 + 0.04, control). A statistically significant
`center effect was observed (in magnitude, not
`direction), but results at all centers were within the
`range of 3 to 6. Three percent of patients in the
`dexmedetomidine group had an RSS of 1 at least
`once compared to 7% in the control group. To
`maintain the protocol-defined target range RSS of 2
`3, dexmedetomidine patients required significantly
`less propofol than did patients in the control group,
`by mean total dose and mg/h (Table 4). After extu-
`bation, the dexmedetomidine patients received 8.4
`+ 6.3 mg of propofol compared to 46.6 + 16.3 mg
`in the control group (meantotal dose + SEM, P =
`.028). Sixty percent (122/203) of the dexmedetomi-
`dine patients required no propofol, 21% received
`less than 50 mg, and 19% received 2 50 mg. In con-
`trast, 76% of the control group received propofol;
`17% received less than 50 mg propofol and 59%
`received 2 50 mg. Differences by country in the use
`of propofol ranged from all patients in both treat-
`ment groups receiving some propofol (Austria) to
`only control patients receiving propofol (Sweden).
`In 8 of 11 countries represented in this study, the
`majority (60% to 85%) of the dexmedetomidine
`patients received no propofol.
`During assisted ventilation (the first 6.5 hours of
`the study), patients in the dexmedetomidine group
`received significantly less morphine for painrelief,
`
`Journal of Intensive Care Medicine 18(1); 2003
`
`33
`
`Downloaded from jic-sagepub.com atUNIV OF PITTSBURGH on August 10, 2016
`
`Petition for Inter Partes Review of US 8,455,527
`Amneal Pharmaceuticals LLC — Exhibit 1062 — Page 33
`
`
`
`Martin et al
`
`Table 4. Propofol Administered
`
`During assisted ventilation
`n
`Total dose (mg)
`n*
`Meanrate (mg/h)
`During study drug administration
`n
`Total dose (mg)
`Mean rate (mg/h)
`
`Dexmedetomidine
`
`Control
`
`P Value
`
`203
`71.6 + 17.51
`198
`8.6 + 1.9
`
`203
`80.0 + 21.3
`5.3+41.2
`
`198
`513.2 + 55.6
`195
`65.6 + 6.8
`
`198
`559.8 + 60.5
`39.1 + 4.1
`
`< .001
`
`< .001
`
`< .001
`< .001
`
`Values are expressed as mean total dose + SEM. The P values are from an analysis of variance.
`a. Exact time of extubation missing for 5 dexmedetomidine patients and 3 control patients.
`
`Table 5. Nursing Assessments and Patient Management Index
`
`Overall sedation and tolerance of the intensive care unit*
`Tolerance of endotracheal tube/ventilator>
`Ease of communication with patient‘
`Ease of managementofthe patient
`Patient Management Index!
`
`Dexmedetomidine
`
`Control
`
`n
`
`180
`180
`179
`178
`177
`
`Score
`
`1.5 + 0.04
`1.3 + 0.03
`2.1 + 0.07
`1.2 + 0.03
`6.1 + 0.12
`
`n
`
`176
`175
`176
`175
`174
`
`Score
`
`1.9 + 0.06
`1.5 + 0.04
`2.4 + 0.08
`1.6 + 0.05
`7.3 + 0.18
`
`a. 1 = very easy, 2 = easy, 3 = moderate, 4 = difficult.
`b. 1 = good, 2 = moderate, 3 = poor.
`cl
`= very easy, 2 = easy, 3 = moderate, 4 = difficult, 5 = not possible.
`d.
`The P value from the Cochran-Mantel-Haenszel row mean score statistic adjusted for center differences was < .001.
`
`4.09 + 0.47 mg versus 8.46 + 0.64 mgin the control
`group (meantotal dose + SEM, P < .001). From
`extubation to the end of study drug administration,
`the total morphine administered was 1.31 + 0.19
`mg (0.14 + 0.02 mg/h) in the dexmedetomidine
`group compared to 4.14 + 0.45 mg (0.50 + 0.06
`mg/h) in the control group (mean total dose and
`dose rate/hour, P < .001).
`Mean duration of weaning was 30.4 + 12.3 min
`in the dexmedetomidine group and 63.1 + 14.5 min
`for control group patients. Times to extubation
`were 471.5 + 15.9 min and 498.1 + 43.85 min for
`dexmedetomidine and control,
`respectively. By
`Kaplan-Meier estimates, no significant differences
`were predicted in duration of weaning (15 minutes
`for both groups) or median times to extubation
`(395 minutes and 385 minutes for dexmedetomi-
`dine and control, respectively).
`Dexmedetomidine patients received significantly
`(P < .001) lower Patient Management Index scores
`(6.1 + 0.12, n = 177) compared to control group
`patients (7.3 + 0.18, n = 174), with lower scores
`corresponding to greater apparent calm; greater
`tolerance of the endotracheal tube, the ventilator,
`and the ICU; greater ease of communication; and
`overall manageability of care (see Table 5 and the
`appendix).
`
`When patients were surveyed with respect to their
`experience as participants in the study, responses
`were generally similar between dexmedetomidine
`and control grouppatients (eg, 36% vs 31%, respec-
`tively, said that they were completely comfortable).
`Fewer dexmedetomidine than control patients
`remembered pain (23% vs 34%), discomfort from
`the endotracheal tube (33% vs 37%), people (36%
`vs 46%), and noise (23% vs 34%). More dexmedeto-
`midine than control patients had no memory of
`their ICU experience (31% vs 25%, respectively).
`During study drug infusion (after extubation),
`there were no statistically significant differences
`between groups in mean respiratory rate. Overall
`changes in respiratory rate were < 3 breaths per
`minute. When analyzed by surgery type (data not
`shown),
`there were no consistent differences
`between the treatments. After extubation, decreas-
`es in SpO, during study drug infusion were similar
`in both groups and remained within normal ranges
`(Fig 2). There were nosignificant differences in the
`variability of oxygen saturation during study drug
`infusion (P = .846). The incidenceofall treatment-
`emergent respiratory system disorders was similar
`between groups (11% vs 14% for dexmedetomidine
`and control, respectively). The only statistically sig-
`nificant difference in this category (P < .010)
`
`34
`
`Journal of Intensive Care Medicine 18(1); 2003
`
`Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on August 10, 2016
`
`Petition for Inter Partes Review of US 8,455,527
`Amneal Pharmaceuticals LLC — Exhibit 1062 — Page 34
`
`
`
`Dexmedetomidine for Postsurgical Patient Care
`
`Table 6. Most Common Treatment-Emergent Adverse Events
`Dexmedetomidine
`(n = 203)
`
`All Treated Patients
`
`Patients with at least
`1 treatment-emergent
`adverse event
`Hypotension
`Hypertension
`Nausea
`Bradycardia
`Vomiting
`Hypoxia
`Mouth dry
`Fever
`Tachycardia
`Hemorrhage
`Atrial fibrillation
`Acidosis
`Confusion
`Agitation
`Atelectasis
`Rigors
`
`121 (60%)
`61 30%)
`24 (12%)
`22 (11%)
`18 (9%)
`10 6%)
`8 (4%)
`7 B%)
`6 B%)
`4 (2%)
`3 (1%)
`3 A%)
`3 (1%)
`3 1%)
`2 (< 1%)
`1 (< 1%)
`1 (< 1%)
`
`Control
`(n = 198)
`
`112 (57%)
`20 0%)
`45 (23%)
`19 (10%)
`4 (2%)
`11 (6%)
`5 3%)
`1 (< 1%)
`7 (4%)
`6 B%)
`7 (4%)
`5 B%)
`5 3%)
`6 3%)
`6 B%)
`9 (5%)
`8 (4%)
`
`P Value
`
`545
`< .001
`-005
`743
`-003
`826
`575
`.068
`-785
`539
`.216
`499
`.499
`.333
`171
`.010
`.019
`
`Adverse events experienced by 2 3% of patients in either group. P values were calculated by Fisher’s Exact Test. Terms are from the
`World Health Organization-Adverse Reaction Terms.
`
`Dexmedetomidine; ------
`Control;
`~=———
`
`
`
`MeanOSChange(x)+/—StdEvrur
`
`Time Since 8352) ine (hours)
`
`To determine whether patients experienced a
`reboundeffect, we also examined the mean change
`from baseline SBP after study drug was stopped. By
`the 15-hour time point (Fig 3), at which time study
`drug was stopped for 97% to 98% ofall patients,
`SBP in the dexmedetomidine group had returned
`to baseline. After the first 24 hours, there were
`essentially no differences between the two groups.
`Hypertension occurred almost twice as often in
`the control group compared to the dexmedetomi-
`dine group (23% vs 12%, respectively, P = .005)
`(Table 6). More than half of the incidents of hyper-
`tension in the control group occurred after the first
`hour, and almost all episodes were considered by
`the investigator to be due to pain, the surgical pro-
`cedure, or exacerbation of preexisting hyperten-
`sion.
`In contrast,
`among dexmedetomidine
`patients, two thirds of the incidents occurred dur-
`ing the 10-minute loading infusion of dexmedeto-
`midine. These were generally described by investi-
`gators as mild to moderate, lasting less than 1 hour,
`and resolving with no treatment or medication.
`Mild to moderate hypotension occurred in 30%
`of the dexmedetomidine patients versus 10% in the
`control group (P < .001). These generally resolved
`with no treatment or with changes in positioning
`and/or fluids or medication. Thirty-five of the 61
`incidents (57%)
`in the dexmedetomidine group
`occurred during or minutes after the loading infu-
`sion; another 4 occurred within the first hour. In
`contrast,
`in the control group, only 1
`incident
`occurred during the first hour. Severe hypotension
`
`Fig 2. Mean changefrom baseline for oxygen saturation
`while receiving study drug. Control group valuesare off-
`set +6 minutes for comparison of standard error bars.
`Baseline in the dexmedetomidine group was 98.7% and
`in the control group was 98.5%.
`
`occurred in the incidence of atelectasis, which was
`observed in 9 of 198 (5%) control patients and 1 of
`203 (< 1%) dexmedetomidinepatients.
`Mean changes from baseline in systolic blood
`pressure (SBP) during study drug infusion in the
`dexmedetomidine group averaged about 7 mm Hg
`lowerthan in the control group (Fig 3). Differences
`between groups were significant at 20 minutesafter
`the start of study drug through the 1-hour time
`point and from the 4-hour through the 20-hour
`time points. No significant difference was observed
`between the groupsin the variability of the actual
`SBP values (mean standard deviation for
`the
`dexmedetomidine group was 16.3 mm Hg vs 16.0
`mm Hgfor the control group).
`
`Journal of Intensive Care Medicine 18(1); 2003
`
`35
`
`Downloaded from jic.sagepub.com at UNIV OF PITTSBURGH on August 10, 2016
`
`Petition for Inter Partes Review of US 8,455,527
`Amneal Pharmaceuticals LLC — Exhibit 1062 — Page 35
`
`
`
`Martin et al
`
`15.0
`
`10.0
`
`
`
`meanchange+SEM(mmHg)
`
`ts are
`CPLR
`
`ee|ee
`Mm/;
`soepheae|
`POSEYglTY
`ft
`
` =
`
`10min(N=195/200)
`
`20min(N=197/199)
`
`195/197) ee
`30min(N
`
`—C— control
`
`—® dexmedetomidine
`
`
`—_~ Se ROR ODDO DDD
`SF SSESFaSSegsgeeTs
`sees Ss 822255 &
`SzFzaegsseseeesIs
`=
`
`AA SS 0% 6 nov oi Zz&
`
`Hd gd dao ao oo Zé4
`CCC LCCZZ44Z C44 228
`See e2ee2 22282 6222 99 4
`=SFS2eFaagekegekeag *
`aAawot ti da
`See2enani&8B8
`
`time (N=control/dexmedetomidine)
`
`Fig 3. Mean change from baseline in systolic blood pressure during the entire study period. This figure includes all random-
`ized intent-to-treat patients, whetherstill receiving study drug or not. N equals the numberof control/dexmedetomidine
`patients with data available at each time point. Study drug was stopped after 12 hours for most patients (large dashed
`arrow). Vital signs were collected at 3-hour intervals for another 24 hours after study drug infusions were stopped. By
`hour 16, study drug had been stopped for approximately two thirds of the patients in both groups.
`
`occurred in 10 of 203 (5%) dexmedetomidine
`patients and 4 of 198 (2%) control patients. Most
`required medication to resolve (Table 7).
`Differences in heart rates between groups were
`significant from 10 minutes through 15 hours after
`the start of study drug (Fig 4). Mean heart rate in
`the dexmedetomidine group decreased from base-
`line between —1.3 and —7.8 bpm. In the control
`group, mean heart rate increases above baseline
`ranged between 2.1 and 12.8 bpm.After study drug
`infusions were stopped (for the majority of patients
`between 12 and 15 hours), heart rates returned to
`baseline levels.
`Bradycardia occurred significantly more fre-
`quently (P = .003) in the dexmedetomidine group
`(Table 6). In the dexmedetomidine group, brady-
`cardia occurred 7/18 times during the first hour; 5
`of these occurred during the loading infusion. Six
`of the 18 were considered severe; 12 were consid-
`ered possibly or probably related. None of the 4
`incidents of bradycardia in the control group were
`considered severe; 3 of 4 were classified as related
`
`to the study drug (placebo). In both groups, brady-
`cardia generally resolved spontaneously or with
`medication (eg, atropine).
`Preoperative hypotension, hypertension, brady-
`cardia, or tachycardia were not risk factors for
`patients who received dexmedetomidine.
`
`Discussion
`
`Ourresults indicate that dexmedetomidineis effec-
`tive for short-term sedation of postsurgical patients.
`Sixty percent of the dexmedetomidine patients
`required no other sedative to maintain a Ramsay
`sedation score 2 3 while on the ventilator. Propofol
`doses in the dexmedet