Pharmacokinetics and Sedative Effects of Intranasal Dexmedetomidine in Ambulatory Pediatric Patients

BACKGROUND: Our aim was to characterize the pharmacokinetics and sedative effects of intranasally (IN) administered dexmedetomidine used as an adjuvant in pediatric patients scheduled for magnetic resonance imaging (MRI) requiring sedation. METHODS: This was an open-label, single-period study without randomization. Pediatric patients from 5 months to 11 years of age scheduled for MRI and receiving IN dexmedetomidine for premedication as part of their care were included in this clinical trial. Single doses of 2–3 µg·kg−1 of dexmedetomidine were applied IN approximately 1 hour before MRI. Five or 6 venous blood samples were collected over 4 hours for dexmedetomidine concentration analysis. Sedation was monitored with Comfort-B scores, and vital signs were recorded. Pharmacokinetic variables were calculated with noncompartmental methods and compared between 3 age groups (between 1 and 24 months, from 24 months to 6 years, and over 6–11 years). RESULTS: We evaluated 187 consecutive patients for suitability, of which 132 were excluded. Remaining 55 patients were recruited, of which 5 were excluded before the analysis. Data from 50 patients were analyzed. The average (standard deviation [SD]) dose-corrected peak plasma concentration (Cmax) was 0.011 liter−1 (0.0051), and the median (interquartile range [IQR]) time to reach peak concentration (tmax) was 37 minutes (30–45 minutes). There was negative correlation with Cmax versus age (r = −0.58; 95% confidence interval [CI], −0.74 to −0.37; P < .001), but not with tmax (r = −0.14; 95% CI, 0.14–0.39; P = .35). Dose-corrected areas under the concentration–time curve were negatively correlated with age (r = −0.53; 95% CI, 0.70 to −0.29; P < .001). Median (IQR) maximal reduction in Comfort-B sedation scores was 8 (6–9), which was achieved 45 minutes (40–48 minutes) after dosing. Median (IQR) decrease in heart rate was 15% (9%–23%) from the baseline. CONCLUSIONS: Dexmedetomidine is relatively rapidly absorbed after IN administration and provides clinically meaningful but short-lasting sedation in pediatric patients.

ANESTHESIA & ANALGESIA Pharmacokinetics of Intranasal Dexmedetomidine dexmedetomidine is safe and efficacious in children, but its sympatholytic properties and recovery profile especially after IV use may restrict its use in pediatric ambulatory care. 5 Intranasal (IN) dosing represents a useful alternative dosing route for some drugs, especially in pediatric patients where IV access may not be available. Dexmedetomidine pharmacokinetics after IN dosing have not been studied in pediatric ambulatory patients; however, a recent report describes the pharmacokinetic profile after 1 or 2 μg·kg −1 IN dexmedetomidine dose for the first 2 hours in 12 pediatric patients undergoing heart surgery. 6 The onset for sedation in pediatric patients has been reported to occur 10-33 minutes after 2-3 μg·kg −1 doses. [7][8][9][10] To our knowledge, there are no previous published studies evaluating the sedative dexmedetomidine concentrations on pediatric patients, but in adults, dexmedetomidine plasma concentrations of 0.3-0.7 ng·milliliter −1 have been reported to provide significant sedation. [11][12][13] The pharmacokinetics of dexmedetomidine are still poorly characterized in pediatric patients, and further studies are warranted. 6 Our main objective was to evaluate the absorption and pharmacokinetics of IN dexmedetomidine after 2-3 μg·kg −1 dose in pediatric patients scheduled for magnetic resonance imaging (MRI) requiring sedation. We also wanted to compare the pharmacological effects caused by singe IN dexmedetomidine to pharmacokinetics during pediatric sedation.
Our primary outcome was to determine the peak plasma concentrations (C max ) and time to C max (t max ) after IN dexmedetomidine. We hypothesized that IN 2-3 μg·kg −1 dexmedetomidine leads to previously defined clinically effective plasma concentrations and coincide with the onset of action during MRI sedation in pediatric patients. Our secondary outcomes were area under time-concentration curve from 0 to 4 hours (AUC 0-4h ) and the pharmacological effects caused by single IN dexmedetomidine to pharmacokinetics during pediatric sedation. We also evaluated the effect of age on dexmedetomidine pharmacokinetics and the effect our dosing regimen had on inducing clinically significant sedative effects in this patient population.

METHODS
This study was approved by the institutional review board (IRB) of the Hospital District of Southwest Finland (IRB number: 109/1800/2016). The trial was registered before patient enrollment at clinicaltrials.gov (NCT02955732; Principal investigator: P.U.; Date of registration: December 14, 2016) and in the EudraCT database (2016-002880-33; Principal investigator: P.U.; Date of registration: October 28, 2016). Written informed consent was obtained from the patients' legal guardians, and assent was obtained from all patients >6 years of age. This manuscript adheres to the applicable Transparent Reporting of Evaluations with Nonrandomized (TREND) guidelines.

Study Participants
Our aim was to recruit 50 pediatric patients with normal growth status (standard deviation [SD], −1.5; ageadjusted height between first and 99th percentiles of Finnish children), age between 1 month and 11 years of age, with guardians fluent in Finnish or Swedish (to understand the given information, to be able to give informed consent, and communicate with the study personnel), and scheduled to receive dexmedetomidine for sedation as part of their clinical care during MRI.
Patients with previous history of intolerance to the study drug or to related compounds and additives, previous drug therapy with dexmedetomidine in the 14 days before the study, use of stimulants or any drugs known to cause significant enzyme induction (eg, phenytoin), existing or recent significant disease that could influence the study outcome or cause a health hazard for the subject when participating in the study, clinically significant abnormal findings in physical examination or laboratory screening (routine hematology, renal function tests, and liver function tests [bilirubin]), or patients participating in any other clinical study involving investigational or marketed drug products concomitantly or within 1 month before the entry into this study were excluded.

Study Outline and Dexmedetomidine Administration
An open-label, exploratory study design was used. Patient recruitment and data collection were conducted at Turku University Hospital in Finland. Patients (and their guardians) potentially eligible for the study were approached for information, assessment of eligibility criteria, and consent either during a preceding clinic visit or on arrival in the hospital for the procedure. Venous blood samples and physiological data were collected from 50 pediatric patients who received dexmedetomidine for procedural sedation as part of their care.
Dexmedetomidine dosage used in this study (2-3 μg·kg −1 ) was based on previous reports 3,7,14 attesting to the safety of dexmedetomidine at doses up to 9 μg·kg −1 given IV over 30 minutes. The selection of the actual dose of dexmedetomidine used for each patient was determined by the anesthesiologist taking care of the patient, and it was based on the general condition of the patient and the anticipated need of required level of sedation. A dose of 2-3 μg·kg −1 of dexmedetomidine (dexmedetomidine hydrochloride 118 μg·milliliter −1 , corresponding to dexmedetomidine base 100 μg·milliliter −1 , Dexdor; Orion Pharma, Espoo, Finland) was administered IN using an LMA® MAD Nasal TM device (Teleflex MAD Nasal; Teleflex Inc, Research Triangle Park, NC) approximately 45-60 minutes before the scheduled MRI procedure to achieve sufficient sedative effect as discussed previously. 5,8 The individual dose was rounded to the nearest 10 μg. A volume of 0.5 mL per nares was considered to be the maximum volume, and larger volumes were divided between both nostrils. A suitable forearm vein was cannulated at this stage; its lumen was kept patent with a mandrin. Topical anesthetic cream (lidocaine 2.5% and prilocaine 2.5%, EMLA; AstraZeneca Inc, Södertälje, Sweden) was applied to some patients to facilitate cannulation. IV thiopental (thiopental sodium 25 mg·milliliter −1 , Pentocur; Abcur, Helsingborg, Sweden) 1-10 mg·kg −1 was given as rescue medication to supplement sedation, if required. If rescue medication was needed, effect parameters measured after the dosing were not analyzed. The decision to administer supplemental thiopental and the selection of the actual dose was made by the attending anesthesiologist.
Venous blood samples (1.0 mL each) were drawn into ethylenediaminetetraacetic acid (EDTA) tubes for drug analysis. Initial 2 samples were drawn during the assumed absorption phase, that is, 10 and 20 minutes after drug administration, 2 samples were drawn 30 and 45 minutes after drug administration, and finally 2 samples were drawn during the assumed elimination phase between 120 and 240 minutes. A 5-minute variance in sampling time was allowed for the first 3 samples, and a 10-minute variance was allowed in the sampling time of the fourth blood sample. Two remaining blood samples were collected between 120 and 240 minutes from drug administration to avoid blood sampling during MRI imaging. Plasma was separated within 30 minutes and stored at −70°C until analysis. Concentration analysis of dexmedetomidine in plasma is described in Supplemental Digital Content 1, Data, http://links.lww.com/AA/C843.

Pharmacokinetic Analysis
Actual sampling times were used for pharmacokinetic calculations. C max and t max were observed directly from the data. The area under the dexmedetomidine plasma concentration-time curve during 0-4 hours (AUC 0-4h ) was estimated by means of the trapezoidal rule. We used the linear trapezoidal rule for increasing values and the logarithmic trapezoidal rule for decreasing concentrations. The apparent volume of distribution of dexmedetomidine was calculated during the early elimination phase (apparent volume of distribution [V d ·F −1 ]) by the use of noncompartmental methods based on statistical moment theory. The pharmacokinetic analysis was performed with WinNonlin program (version 4.1; Pharsight Corp, Mountain View, CA).

Drug Effects
The psychomotor effects of IN dexmedetomidine were assessed by 2 pediatric anesthesiologists using the Comfort-B sedation scale (CBSS), which has been used to monitor sedation in procedural and intensive care sedation of pediatric patients. Sedation was assessed clinically acceptable if CBSS decreased ≥6 points. 15,16 Peripheral oxygen saturation (Spo 2 ) and heart rate (HR) were monitored continuously after drug administration. We analyzed the pharmacological effect parameters only until supplemental thiopental was administered to rule out the effects of thiopental.

Assessment of Local Tolerability and Safety
The local tolerability of IN dexmedetomidine was assessed and recorded real time by visual inspection (crying, nasal irritation, and running nose) by the investigator immediately during administration of IN dexmedetomidine and thereafter at 1, 2, 3, and 4 hours after dosing and at the end of the clinical observation period. We identified adverse events of special interest, including bradycardia, hypotension, and respiratory complications within 4 hours of dexmedetomidine administration. We defined bradycardia as a decrease in HR ˃30% of patient's median baseline HR or requiring intervention, including administration of atropine. We defined hypotension as decreased systolic or mean arterial blood pressure ˃30% from patient's median baseline blood pressure or requiring intervention, including administration of a fluid bolus, or initiation of vasopressor therapy. We defined respiratory complications as a drop of Spo 2 below 88% or apnoea. Adverse events were assessed real time and adjudicated by the primary investigator (P.U.) based on review of the patient's medical chart. Potential adverse events were reported immediately during the study.

Statistics and Sample Size
The C max and t max of IN dexmedetomidine were the primary outcome measures of the study, and all other pharmacokinetic parameters and pharmacological effects were secondary variables. The sample size was based on previous experience in similar pharmacokinetic studies. [17][18][19] To evaluate the difference between age groups and to account for a larger variability in absorption after IN dosing, we calculated that 14 patients were needed to demonstrate a 50% difference in the C max between the age groups with a type I error of 5% and a statistical power of 80%. The data were evaluated for normality of distribution using probit plots and the Shapiro-Wilk W ANESTHESIA & ANALGESIA Pharmacokinetics of Intranasal Dexmedetomidine test. Concentrations were normalized to a 1 µg·kg −1 dose and log transformed before computing pharmacokinetic parameters for each individual, but nontransformed results are reported. The data were analyzed first as one group, after which the patients were allocated to 3 age groups to analyze the effect of age. The age groups were as follows: between 1 and 24 months of age (AG1), from 24 months to 6 years of age (AG2), and over 6-11 years of age (AG3). Maximal reductions in CBSS, HR, and Spo 2 readings were calculated before analysis. Pharmacokinetic parameters and pharmacological effects were compared with 1-way analysis of variance (ANOVA), while t max and the other time parameters were compared with the Kruskal-Wallis test. Levene test for homogeneity of variances was used before ANOVA. The associations of dexmedetomidine concentrations in plasma and the pharmacokinetic parameters with pharmacological effects were evaluated using Pearson product moment correlation coefficient. The results are expressed as mean values and SD, except for categorical and time-related parameters that are reported as medians and interquartile range (IQR). The statistical significance level was set to P < .05. R software (version 3.4.2; www.r-project.org) and ggplot2 (version 2.2.1; ggplot2.tidyverse.org) were applied for statistical analysis and graphical presentations.

Patient Data
We evaluated 187 consecutive patients for suitability, of which 132 were excluded. A consent was obtained for 55 patients, who all were administered IN dexmedetomidine, but 5 patients were later excluded (Supplemental Digital Content 2, Figure 1 Table 1.

Pharmacological Effects
The impact of IN dexmedetomidine administration on the pharmacological effect parameters is summarized in Figure 2 and Table 3. Decrease in CBSS was significantly correlated with C max (r = 0.34; 95% CI, 0.067-0.56; P < .016), but not with t max . The CBSS reductions were not significantly associated with age or dose of IN dexmedetomidine. Based on clinical judgment, almost all patients required additional sedation, and 47 patients were dosed with thiopental (Table 1).

Cardiorespiratory Effects
Dexmedetomidine decreased HR and the lowest HRs before administration of supplemental thiopental in  Abbreviations: CI, confidence interval; C max , peak concentration in plasma; D, dexmedetomidine dose (µg); t max , time to peak concentration; AUC 0-4 , area under the plasma concentration-time curve from 0 to 4 h; V d F −1 , apparent volume of distribution.

Safety and Tolerability
IN dexmedetomidine was well tolerated. Nasal irritation was not observed in or reported by any of the patients, but crying was observed in 16 patients (32%) and runny nose in 8 patients (16%). One patient (5 years, 17 kg) received a single dose of atropine for bradycardia of 36 minute −1 , occurring soon after dosing of thiopental, 57 minutes after dexmedetomidine.
No serious adverse events related to study drug were reported. The reported adverse effects were mild and mostly related to an unpleasant feeling in nasal mucosa (3 of 50 children) immediately after administration of the nasal spray. One event of vomiting was recorded.

DISCUSSION
IN dexmedetomidine was rapidly absorbed, and C max was achieved in a median time of 37 minutes.
Interindividual variability was, however, large, as previously reported in healthy adult volunteers. 11,20 A comparison with previous results 11 suggests that C max may continue to decrease with increasing age because the dose-normalized C max calculated for adult healthy volunteers was smaller (0.0060 vs 0.0086 1·liter −1 ) than that now found in our 6-to 11-year-old pediatric patients. This is most likely linked to the larger dosing volumes needed in older subjects. The actual delivered dose volumes in our study varied from 240 to 1200 µL, which by far exceeds the volumes usually considered ideal. 21 This calls for the development of improved drug delivery systems compared to the one used in this study, if IN application is to be clinically applicable.
To evaluate the effect of age on dexmedetomidine pharmacokinetics, patients were allocated to 3 age groups. Although newborns differ from infants (0-2 years), these patients could not be included because ambulatory MRI scans are not scheduled for newborns. Similarly, there were quite few ambulatory MRI scans performed on the youngest age group. Two other age groups were based on observations on developmental pharmacology. 22 The youngest patient group had significantly higher dose-normalized peak

ANESTHESIA & ANALGESIA
Pharmacokinetics of Intranasal Dexmedetomidine concentrations and drug exposure (AUC 0-4h ) compared to the older patients.
A recent study evaluated pharmacokinetics of IN dexmedetomidine in pediatric patients (age between 6 and 44 months) undergoing cardiopulmonary bypass. 6 The patients had congenital heart disease and were anesthetized before administration of IN dexmedetomidine, both of which could have affected the disposition of dexmedetomidine. In our study, the peak concentrations (0.54 ng·kg −1 after a mean dose of 2.8 µg·kg −1 ) were similar to previous findings, but t max was slightly shorter. Use of IN dexmedetomidine as premedication has been studied in nonsurgical pediatric patients with dosages of 2-3 µg·kg −1 and the onset of IN dexmedetomidine has been described to be 10-33 minutes. [8][9][10] The onset of action was similar in our study (Figure 2), which coincided with the time to maximal effect. Because most of the patients received additional sedation during MRI, we were not able to evaluate the duration of sedation caused by dexmedetomidine alone.
Previously, the pharmacokinetics of IN dexmedetomidine has mainly been investigated in healthy adult volunteers. 11,20 Assuming that pediatric patients have similar nasal bioavailability as adults, we can estimate that the pediatric patients have higher AUC 0-4h than adults because the dose-corrected AUC 0-4h after IN dosing in the current study was 1.4 minutes·milliliter −1 , whereas in adults, the dose-corrected mean AUC extrapolated to infinity was 0.88 minutes·milliliter −1 . Considering the age dependency of the exposure, the difference seems to be even bigger in younger patients. Although we were not able to measure the absolute bioavailability in this single-period observational study, our results show that nasally applied dexmedetomidine is efficiently absorbed in children.
The average C max in our study was 0.54 ng·milliliter −1 , which suggests that sedative concentrations were achieved after IN 2-3 µg·kg −1 doses of   16 We found no evidence of differences in the sedative effects between the different age groups.
However, the youngest age group consisted of only 8 patients, and we cannot draw any conclusions based on this finding. We chose to use the Comfort-B assessment scale because this scale is routinely used in our pediatric intensive care unit. More importantly, CBSS is an objective monitoring tool without any need for verbal or physical stimulation, which can be considered as an advantage in MRI sedation where lightly sedated spontaneously breathing pediatric patients are expected to lie immobile.
Dexmedetomidine causes individually variable hemodynamic responses and decreases HR due to its sympatholytic effects. 12,13 Our result shows that IN dexmedetomidine decreased HR on the average by 16 beats/min (by 8.9%), and this is in accordance with previously published results. 11,20 IN administered dexmedetomidine reduced HR less compared to IV dosing, which produces up to 30% decreases in HR. 2,5 Similar to the previous reports on IN dexmedetomidine, HR was reduced most in the youngest patients of 0-2 years of age. The risk of bradycardia should Figure 3. Vital parameters after intranasal 2-3 µg·kg −1 dexmedetomidine dose as nasal spray in 50 ambulatory pediatric patients in the 3 age groups. Heart rate and Spo 2 were recorded at 6 predefined times after dexmedetomidine dosing. Gray curves show heart rate or Spo 2 versus time for all 50 patients, and green, orange, and blue for 0-2, 2-6, and 6-11 y old patients, respectively. Spo 2 indicates peripheral oxygen saturation.

ANESTHESIA & ANALGESIA
Pharmacokinetics of Intranasal Dexmedetomidine be kept in mind when using high dexmedetomidine doses in young patients.
Our study has several limitations. In particular, these include small number of patients in the youngest age group, the short sampling period for dexmedetomidine plasma samples, and the common need for administration of supplemental thiopental to the patients at the later timepoints of the study to achieve acceptable sedation. Blood samples for dexmedetomidine concentration measurements were collected only up to 4 hours, based on the adult data. 11 It would be more informative to collect further blood samples more precisely and during a longer followup period, but because our patients were discharged after 4 hours, further follow-up was not possible. Furthermore, we had to allow a 5-to 10-minute variance in the sampling time. Nevertheless, we were able to capture informative concentration-time profiles from each patient. Our sample size was based on previous experience in the dexmedetomidine pharmacokinetics studies, but the total number of patients was low especially in the youngest age group. The selected sample size thus affects the power to compare the 3 age groups, which should be taken into consideration when evaluating the results. However, the CI evaluation shows that our primary end points have acceptable precision.
In conclusion, our results demonstrate that IN dexmedetomidine is relatively rapidly absorbed and causes significant sedation in pediatric patients. Pharmacokinetics of IN dexmedetomidine in pediatric patients show quite similar characteristics compared to adults, and our findings regarding the sedative effects are consistent with previous results published for pediatric patients. The results suggest that IN dexmedetomidine as sole agent might not be sufficient for procedural MRI sedation of pediatric patients, and combination with other sedative agents may be needed. Because the volumes of the preparation used in our study exceed the recommended volumes for IN administration, there appears to be a need for a more concentrated dexmedetomidine solution. The optimal dose and drug delivery system for IN dexmedetomidine dosing remain to be evaluated. E