The Effectiveness of Three Regimens of Sedation for Children Undergoing Magnetic Resonance Imaging

A Clinical Study

Abstract

Background: 

Magnetic resonance imaging (MRI) under sedation requires faster recovery for early discharge and feeding resumption in children with neuropsychiatric disorders. The use of dexmedetomidine alone results in delayed recovery. Propofol, when used alone, can cause hypotension and respiratory depression. A new regimen for sedation was evaluated by exploiting the properties of these drugs, to allow faster recovery with minimal adverse events.

Materials and Methods: 

One hundred and fifty children aged 2–12 years requiring MRI were randomly allocated to these three groups. Group P (n = 50) received propofol bolus at 2 mg.kg⁻¹ over 10 min followed by infusion at 100 μg.kg⁻¹.min⁻¹. Group D (n = 50) received dexmedetomidine bolus of 2 μg.kg⁻¹ over 10 min followed by infusion at 1 μg.kg⁻¹.h⁻¹. Group PD (n = 50) received propofol bolus at 2 mg.kg⁻¹ over 10 min followed by dexmedetomidine infusion at 1 μg.kg⁻¹.h⁻¹. Recovery characteristics were noted.

Results: 

Recovery following sedation in Group PD (15 ± 7.0 min) and Group P (17.35 ± 7.4 min) were comparable and significantly (P = 0.03) lesser than Group D (27.58 ± 8.09 min). Emergence delirium scores were significantly less in Group PD (5 ± 1.08) and Group D (5.6 ± 2.4), unlike scores in Group P (9 ± 2.43). About 79.5% (39/49) of children in Group P, 88.2% (45/51) of children in Group D, and 86% (43/50) of children in Group PD completed MRI without any movement. Seven (14.58%) in Group P, 2 (4%) in Group D, and 5 (10.20%) in Group PD required rescue sedation.

Conclusion: 

The regimen with propofol bolus and dexmedetomidine infusion provided adequate sedation and better recovery characteristics in children aged 2–12 years without systemic complications, as compared to the use of either agent alone.

INTRODUCTION

Magnetic resonance imaging (MRI) is an essential investigation modality for children suffering from neuropsychiatric disorders. Spending 30–45 min inside the dark and narrow MRI gantry with loud pitched bizarre noises is an uncomfortable and claustrophobic feeling for adults. With such claustrophobic conditions, it is impossible to conduct an MRI study in children without sedation. Children posted for MRI of the brain/spine under sedation often have preexisting neuropsychiatric illnesses complicated by metabolic disorders posing significant challenges to the anesthesiologist. MRI performed under sedation is a day care procedure that requires faster recovery times to facilitate early discharge and resumption of feeding.

The most commonly used agents for procedural sedation in recent times are intravenous (i.v.) propofol and dexmedetomidine, which are near ideal for sedation procedures but not without drawbacks. Propofol infusion often results in respiratory depression.[¹] Dexmedetomidine a highly selective α₂ agonist commonly used for pediatric procedural sedation[²³] is known to cause a delay in recovery[⁴] and a decrease in heart rate (HR) due to its direct action on the sinus node in children.[⁵]

Drug combinations such as dexmedetomidine and ketamine are used safely in children for MRI,[⁶] cardiac catheterizations,[⁷] and burns dressings.[⁸] This combination was arbitrarily assumed to have a synergistic effect, with minimal adverse events and proven efficacy as compared to stand-alone drugs. Similarly, an assumption was made in the present study wherein we utilized the properties of both propofol and dexmedetomidine to evaluate the efficacy of this new regimen. We used propofol for its rapid onset of sedation and dexmedetomidine for maintenance of sedation, with an expectation of minimal adverse effects compared to either drug when used alone, and rapid offset of sedation facilitating early discharge.

MATERIALS AND METHODS

This randomized prospective study was conducted for 1 year after the institutional ethics committee approval and is registered with the Clinical Trial Registry-India (Numbered CTRI/2018/07/014845). Written informed parental consent was obtained. One hundred and fifty children of either sex aged between 2 and 12 years undergoing MRI for 30–45 min were included. Children with congenital heart diseases, airway abnormalities, recent respiratory tract infections, and a history of allergy to any of the study drugs in the sedation protocol were excluded. Fasting guidelines were followed in accordance with guidelines proposed by the American Society of Anesthesiologists. Prilox (eutectic mixture of a local anesthetic) cream was applied on the dorsum of the child's hand 45 min before establishing i.v. access. Parents accompanied children to the MRI induction room until the initiation of the sedation procedure. A four-point scale, as described by Koroglu et al.[²] was used to categorize if the child had distressed or unstressed behavior. Baseline (Bl) vital parameters were recorded and venous access was established in the MRI induction room. All children received midazolam 0.1 mg.kg⁻¹ i.v as premedication. Before the initiation of sedation, children were randomly allocated to one of the three groups as per the order obtained from a computer-generated table of random numbers.

Group P (n = 49) received propofol (1%) initially at 2 mg.kg⁻¹ over 10 min, followed by propofol infusion at 100 μg.kg⁻¹.min⁻¹ until the completion of the MRI study.

Group D (n = 51) received dexmedetomidine initially at 2 μg.kg⁻¹ bolus over 10 min followed by dexmedetomidine infusion at 1 μg.kg⁻¹.h⁻¹.

Group PD (n = 50) received propofol 2 mg.kg⁻¹ over 10 min followed by dexmedetomidine infusion at 1 μg.kg⁻¹.h⁻¹ until the completion of MRI.

The study drugs were formulated to a total volume of 50 ml. Propofol 1% or dexmedetomidine 4 μg.mL⁻¹ (200 μg was diluted with 0.9% sodium chloride in a total of 50 mL). After 10 min of bolus administration under monitoring, children were transferred to a magnetic resonance (MR) suite. Procedural monitoring included pulse oximetry, oxygen saturation (SpO₂), electrocardiography (ECG) derived HR, mean arterial pressure (MAP) from noninvasive blood pressure monitoring, ECG, respiratory rate, and expired carbon dioxide (ETCO₂) throughout the study time at 5-min intervals. Dextrose normal saline was used as the maintenance fluid, according to the Holliday-Segar formula. Sedation levels were evaluated using the modified Ramsay sedation score (MRSS)[⁹] at 5-min time intervals. Once the target score of 4 was reached children were transferred to MRI gantry.

The outcome variables – SpO₂, HR, MAP, MRSS levels, and ETCO₂ levels were recorded every 5 min after premedication throughout the study period till complete recovery. At Bl, after (Af) midazolam, at 5 and 10 min after starting the study drug infusions. T₀-at the beginning of MRI and every 5 min until 45 min or completion of MRI study.

Emergence agitation (EA) was assessed at 10 min after complete recovery by an investigator who was different from the primary investigator, using the pediatric emergence delirium scale (PEDS).[¹⁰] Scores of >10 were considered as the presence of emergence delirium. Rescue sedation with midazolam bolus at 0.05 mg.kg⁻¹ i.v. was considered if MRSS >4 was not achieved or if the child woke up midway while in the MRI gantry. A maximum of two boluses were allowed beyond which it was considered a failure in sedation technique.

Recovery time was defined as the time taken from the termination of the maintenance drug until the MRSS reached 1. Children were discharged from the recovery room when the modified Aldrete scores[¹¹] were >9 and MRSS 1. In the recovery room, vital parameters and scores were noted every 5 min. Quality of imaging was assessed by the radiologist and graded as follows, score 0 – no motion, 1 – little motion, 2 – motion that requires repeat scans.

In the event of oxygen desaturation, defined as SpO₂ lesser than 95% during the MRI acquisition, children were repositioned with a pillow under the shoulders to keep the airway open. If there was no improvement in SpO₂ even after repositioning, an oral or nasal airway was inserted to ensure an unobstructed airway. If desaturation persisted, the airway was secured by either laryngeal mask airway (LMA) insertion or with an endotracheal tube and was considered a failed sedation technique.

Bradycardia was defined as HR decrease of more than 20% of the age-adjusted limits and treated with atropine 10 μg.kg⁻¹ i.v. hypotension was defined as a decrease in systolic blood pressure of more than 20% from baseline and was treated with a bolus of 0.9% normal saline at 2 mL.Kg⁻¹. If hypotension persisted i.v. mephenteramine at 0.4 mg.kg⁻¹ was given. Further, the study drug dose was reduced by 50% if hypotension persisted.

Sample size calculation

Time to recovery from sedation was considered the primary outcome. With an error probability of 0.05 and a power of 90%, a sample size of 50 in each group was considered to detect the differences between the three regimens. Therefore, 150 children requiring sedation for MRI were recruited. The sample size was estimated to be 150, as calculated using STATA, Version 9 (StataCorp. 2005. Stata Statistical Software: Release 9. College Station, TX: StataCorp LP.).

Statistical analysis

Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS^® version 20, Chicago, IL, USA). Quantitative variables were represented as mean and standard deviations; qualitative variables were represented as percentages. The Chi-square test was used to analyze qualitative variables as appropriate. Analysis of Variance (ANOVA) test was used to find the significance of study parameters on a categorical scale between the three groups. The repeated measure of ANOVA was employed to find the significance of study parameters on the continuous scale within the group (intragroup analysis) on metric parameters, if significant results were observed, then post hoc analysis with Bonferroni correction was done to correlate the significance. Nonparametric variables were analyzed using Kruskal–Wallis test. Post-hoc analysis was performed using Kruskal–Wallis test to note the degree of significance. P < 0.05 was considered statistically significant.

RESULTS

Demographic characters, presedation midazolam dose, and presedation behavior scores remained comparable between all the three groups [Table 1].

Table 1: Demographic characteristics, presedation behavior, and the total drug dosage in all the three groups

A total of 178 children were assessed for eligibility. 28 children were excluded for not meeting the inclusion criteria. The remaining 150 children were randomly allocated into Group P (49), Group D (51) and Group PD (50) by a computer generated algorithm. Figure 1 depicts the consort diagram. The total time to recovery was significantly (P = 0.000) prolonged in Group D (26.84 ± 8.52 min), but the time to recovery in Group P and PD remained comparable (P = 0.388) and significantly less (17.35 ± 7.48 min in P, 15 ± 7.22 min in PD). Emergence delirium scores assessed from the PEDS scale were 9.04 ± 2.43 in Group P, 5.67 ± 2.40 in Group D, and 5.08 ± 1.80 in Group PD, respectively. Scores in Group D and Group PD were comparable (P = 0.332) and significantly less than the scores in Group P (P = 0.000). Time to discharge from the recovery area was prolonged in Group D (27.58 ± 8.09 min, P = 0.000), unlike that in Group P and Group PD, which remained comparable (P = 0.211) and significantly less (17.55 ± 7.0 min and 17.04 ± 6.6 min).

Figure 1: Consort flow diagram of enrollment, randomization, and analysis

About 88.2% of children (45/51) in Group D, 86% (43/50) in Group PD, and 79.5% (39/49) in Group P completed MRI without any movement. Seven children (14.28%) in Group P, 2 (4%) in Group D, and 5 (10.20%) in Group PD required rescue sedation with an additional dose of midazolam to complete the imaging study, and one child in Group D required a second dose of midazolam.

Overall, three children (2%) failed to complete the MRI study with the designed sedation technique. Two children, one each in Group D and Group PD, could not attain adequate sedation depth despite two bolus doses of midazolam. One child in Group P had significant desaturation requiring placement of LMA. No child in Group D and Group PD had respiratory depression. Shivering was noted in eight children (16.3%) in Group P (P = 0.128), and in one child in Group PD. None of the children in Group D had shivering. sedation scores (mRASS) of 6; nausea and vomiting were not seen in any of the children in either group.

Trends in the variations in HR and MAP are shown in Figure 2. Children in Group D had low normal HR, and higher MAP throughout all time point intervals until complete recovery from sedation. On the contrary, MAP remained in the low normal range in Group P at all time intervals. One child required an additional fluid bolus to treat hypotension. HR and MAP were maintained in the normal range in group PD at all study points. At time T₀ MAP remained higher in all three groups. However, none of the children woke up during the transfer to the MR table. None had bradycardia requiring intervention.

Figure 2: Trends in HR and MAP variations over time in all the groups. HR: Heart rate, MAP: Mean arterial pressure

Mean expired CO₂ levels remained lesser in Group P in comparison to the other groups. Two children (4.08%) in Group P and one child in Group PD (2.04%) had significant desaturation requiring realignment of position, one child in Group P required placement of LMA to maintain a patent airway and was considered failed technique. None of the children in Group D had desaturation.

DISCUSSION

Propofol and dexmedetomidine have been used as sole sedative agents for MRI. Since children will be fasting and may have coexisting metabolic disorders, special consideration needs to be given to variations in blood sugar levels, and any sedation protocol in children should have an accelerated recovery postprocedure so that feeding can be resumed early. The present study was designed to evaluate the effect, of propofol and dexmedetomidine used in combination, exploiting their unique mechanisms of action on the nervous system, to see if we get a better outcome than when either drug is used alone.

We observed that the combination of propofol and dexmedetomidine infusion for sedation maintenance can be used safely in children with neuropsychiatric illness to facilitate faster onset and offset of sedation with minimal adverse events so that they can resume feeding faster with minimal disturbances in the mentation. Time to recovery from sedation, and discharge from the recovery area were comparable between propofol and propofol-dexmedetomidine groups and much shorter than the dexmedetomidine group. Quality of imaging and EA were comparable between dexmedetomidine and propofol-dexmedetomidine groups and superior to the propofol group. In addition, stable hemodynamics, respiratory parameters, and reduced shivering were seen in the propofol-dexmedetomidine group.

Young children with greater presedation anxiety levels are known to have greater EA;[¹²] hence, midazolam was administered for premedication to reduce EA.[¹³] Propofol bolus dose for induction of sedation was in accordance with the study by Kim et al.[¹⁴]

Although propofol infusion alone at 200 μg.kg⁻¹.min⁻¹ is shown to produce optimal sedation,[¹⁵] we decided to use propofol infusion at 100 μg.kg⁻¹.min⁻¹ for sedation maintenance as midazolam was used for premedication. The dose of 100 μg.kg⁻¹.min⁻¹ of propofol for maintenance of sedation proved adequate in most of the study subjects. Heard et al.[³] had observed inadequacy of dexmedetomidine at a bolus dose of 1–1.5 μg.kg⁻¹ i.v.[³] and the same, when used at higher doses (bolus dose at 3 μg.kg⁻¹ and infusion at 1–2 μg.kg⁻¹.h⁻¹), has shown to be an effective and safe sole sedative agent in children.[¹⁶] Since midazolam was used for premedication, dexmedetomidine bolus dose of 2 μg.kg⁻¹ i.v. and 1 μg.kg⁻¹.h⁻¹ of infusion was considered in our study.

The mean recovery time was prolonged in the dexmedetomidine group and was in accordance with the results observed by Koroglu et al.[²] and Heard et al.[³] This prolonged sedation can be attributed to the effects of dexmedetomidine on locus coeruleus[¹⁷] and the ventrolateral preoptic nucleus. The incidence of EA in groups PD and D remained comparable. Although the children in the propofol group reached desired Aldrete scores earlier but failed to get discharged immediately after recovery as they had higher EA scores that required further monitoring. However, none of the children in either group required further treatment for EA.

The quality of MRI was comparable between Group D and Group PD and better than Group P as opined by radiologists. Children in the dexmedetomidine group had low normal HR and higher MAP at all monitoring time points throughout the study. This is probably due to the α₂ agonistic actions of dexmedetomidine. Propofol-dexmedetomidine group had HR and MAP in the normal ranges at all time points. However, at time T₀ the values were higher in all three groups. This was probably due to the stimulus and manipulations that happened during the transfer of the children to MRI gantry from the holding area. But none of the children in any of the groups woke up during the transfer. Children in the propofol group had lower expired carbon dioxide levels as compared to the children in the dexmedetomidine and propofol-dexmedetomidine groups. This might be because of the low normal MAP values in the propofol group.[¹⁷]

The incidence of desaturation was 4.08% (two children) in Group P and 2.04% (one child) in Group PD, respectively. These results were similar to the study by Heard et al.[³] We can therefore suggest that the propofol-dexmedetomidine combination, can produce desired sedation in children, who are spontaneously breathing without interfering with respiration compared to these drugs when used alone.

Overall, the sedation failure rate was 2% (one child in each of the groups). One child in the propofol group who required placement of LMA, was 4 years old with global developmental delay secondary to maternal toxoplasmosis rubella cytomegalovirus and herpes zoster infection. A 12-year-old child in the dexmedetomidine group with global developmental delay and seizures; had not received the oral antiepileptic drugs (clonazepam, valproate, and risperidone) on the day of MRI. i.v. antiepileptics were administered and sedation initiated, but this child woke up during the scan after 15 min and required two additional boluses of midazolam and continued to move. This child also had an episode of seizure during recovery. A 7-year-old child in the propofol-dexmedetomidine group presented with a severe emotional disorder, diagnosed with autoimmune encephalitis required two bolus doses of midazolam after 16 min of starting the scan, but adequate depths were not attained despite two bolus doses of midazolam. The exact reason for sedation failure in these children could not be explained.

Neural network-based hypothesis for sedation with combination drugs

Understanding the strength of intraneural and interneural connectivity between different brain networks helps in deriving the probable hypothesis for sedation with the combination of propofol and dexmedetomidine.

With propofol sedation

At the onset of sedation

During the awake state, the posterior cingulate cortex (PCC) is under the inhibition of the arousal center in the brainstem. With propofol, the connection between the brainstem and PCC weakens, as observed by Långsjö et al.[¹⁸] in their study, resulting in activation of PCC and subsequent increase in the functional connectivity (FC) to other areas of the brain, namely the anterior cingulate cortex (ACC) and cuneate nucleus, suppressing them and causing unconsciousness [Figure 3].

Figure 3: Neural network-based hypothesis for the loss of consciousness by propofol. ^†Anterior cingulate cortex, ^‡Posterior cingulate cortex

Return of consciousness

With the fall in the plasma concentrations of propofol, the connection between the brainstem arousal center and PCC is re-established, resulting in an increase in the inhibitory effect of the arousal center on the PCC. This, in turn, reactivates upstream areas, mainly ACC and cuneate nucleus,[¹⁸] causing awakening. Reactivation of the cuneate nucleus is specifically observed with the waning of propofol sedation[¹⁸] and results in arousal.

With dexmedetomidine sedation

At the onset of sedation

Like propofol, the connection from the arousal center in the brain stem to PCC weakens with Dexmedetomidine as well.[¹⁸] However, the FC with other areas such as the thalamus, ACC, and brain stem[¹⁹²⁰] is better preserved. The thalamic connectivity to the mesopontine region is shown to be more preserved with dexmedetomidine sedation [Figure 4].[¹⁸] This area is the key region in ascending reticular arousal system. A better-preserved FC can reduce EA at recovery.

Figure 4: Neural network-based hypothesis for the loss of consciousness by dexmedetomidine. *Dexmedetomidine, ^†Anterior cingulate cortex, ^‡Posterior cingulate cortex

With the return of consciousness

Långsjö et al.[¹⁸] observed that changing plasma concentrations of dexmedetomidine does not affect the state of consciousness.[¹⁸] It is the alteration in the strengths of FC in the aforementioned networks that determines consciousness with respect to dexmedetomidine. Långsjö et al.[¹⁸] had observed no change in the plasma concentration of dexmedetomidine at the onset of sedation and at return to consciousness.

With propofol and dexmedetomidine combination [Figure 5]

Figure 5: Neural network-based hypothesis for the loss of consciousness by a combination of propofol and dexmedetomidine. *Dexmedetomidine, ^†Anterior cingulate cortex, ^‡Posterior cingulate cortex

We hypothesize that with propofol bolus, the connection between the brainstem and PCC had weakened. After the propofol levels decrease in the plasma, the return of connectivity between the brainstem, thalamus, and medial prefrontal cortex would be established, causing arousal. In this study, children were on dexmedetomidine which was initiated along with the propofol bolus. Dexmedetomidine would have altered the FC before propofol is redistributed out of the brain in the PD group. Hence, there might be a predominant effect of dexmedetomidine on the neural network along with propofol due to these considerations [Figure 5].

• The PCC secondary to inhibition from the brain stem would have increased FC to other areas of the brain
• FC between the thalamus, ACC, and brainstem is better preserved.

Study limitations

1. The fore mentioned sedation regimen could not be considered for illnesses such as mitochondrial disorders, movement disorders, and dysmorphic syndromes
2. The exact pharmacokinetics and pharmacodynamic principles favoring the combination of drugs, as a better regimen could not be explained in our study.

Future implications

In the available literature to date, there are no studies using the combination of propofol bolus and dexmedetomidine infusion for pediatric sedation in MRI. The exact interactions between the drugs could be explained by functional imaging to assess the interactions between various neural networks and also by chemical shift imaging studies. Further studies can correlate the changes in the neural networks with this combination of propofol and dexmedetomidine.

CONCLUSION

We would like to conclude that in children, a sedation regimen with a combination of propofol bolus and dexmedetomidine infusion can produce optimal sedation, good recovery characteristics, complete imaging, stable hemodynamics, and respiratory parameters. This combination regimen has demonstrated a better sedation profile as compared to using propofol or dexmedetomidine alone.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

Acknowledgment

We are thankful to the parents of children who consented to enroll their children in this study. We also thank all anesthesia technicians who helped us at various levels ensuring patient safety.