Endotracheal administration of drugs such as lidocaine, epinephrine, atropine, naloxone, and vasopressin may be performed for emergency treatment when IV or intraosseous access cannot be established.1,2 The 2010 American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care1 recommend that for endotracheal administration drugs should be dosed 2 to 2.5 times their IV dose diluted with normal saline to a total volume of 5 to 10 mL, and injected directly into the endotracheal tube. Conventional endotracheal drug administration has traditionally been with a nonatomized drug delivery device.
One newer approach to delivering drugs to the tracheal mucosa is through a mucosal atomization device shown in Figure 1.3 Atomized liquid has several advantages over nonatomized liquid, such as electrification4 and an increased surface area because of smaller atomized particles.5 The electrical charge of the atomized particle contributes to increased spread throughout the respiratory tract.6 Atomized drug with its increased surface area5 can distribute over a wider portion of the endotracheal mucosa than nonatomized liquid drug. Thus, the method of lidocaine administration within the trachea (atomized versus nonatomized) may alter its pharmacokinetic profile.
The aim of this study was to examine whether the atomization of lidocaine changed its pharmacokinetic profile after endotracheal administration when compared with nonatomization of lidocaine. A secondary aim was to build a pharmacokinetic model based on measured lidocaine plasma concentrations. We explored whether different estimates of bioavailability or absorption associated with each method of delivery would improve the fit of our pharmacokinetic model to measured plasma concentrations. We hypothesized that endotracheal administration of atomized lidocaine would result in higher plasma concentrations over time. We also hypothesized that the best pharmacokinetic model would include higher bioavailability and faster absorption for atomized lidocaine delivery compared with nonatomized lidocaine delivery.
MATERIALS AND METHODS
This study was approved by our institutional ethics committee and registered to the University Hospital Medical Information Network Clinical Trials Registry (reference number: UMIN000006220). All consecutive patients undergoing elective prostatectomy or mastectomy at the National Defense Medical College Hospital gave their written informed consent before enrollment in the study. The following exclusion criteria were applied: older than 75 years, severe cardiac or pulmonary disease, arrhythmias, severe hepatic or renal disease, obesity (body mass index >30 kg/m2), hypersensitivity to lidocaine or other local anesthetics, and administration of lidocaine for the operation. After recruitment, the patients were randomly allocated into 1 of 2 groups. One group received lidocaine using a mucosal atomization device3 (MADgic®, Wolfe Tory Medical, Salt Lake City, UT). This device consists of a syringe and a 216-mm narrow plastic tube with an atomizer at the tip of the tube. Another group received lidocaine using a spray tube7 (endotracheal spray tube®, Hakko, Tokyo, Japan). This device consists of a syringe and a 160-mm tube that has eighty 1-mm holes over the entire surface at the last 25 mm of the tube. An image of each device is shown in Figure 1. Randomization was generated by physicians who had no involvement in the assessments. The sample size was calculated as 8 for each group to detect a 0.8 μg/mL difference of mean highest plasma lidocaine concentration with a SD of 0.4 μg/mL based on a power of 0.95 and an α of 0.05. We assumed that this difference was reasonable without substantive previous data. To account for patient dropout, we recruited 20 patients, 10 for each group. Six was set as the maximum number of each gender in each group. Ten was set as the maximum number of patients in each group. Investigators were not blinded to interventions, but patients were blinded.
Physical inspection and chest radiographic assessment that are standard parts of our preoperative evaluation were performed to assess for airway abnormality. Premedication was not used. After a patient entered an operating room, standard monitoring was applied. An arterial catheter for blood sampling was inserted into the radial artery under a small dose of 1% mepivacaine. Propofol was administered using a target-controlled infusion pump (Terufusion® TCI pump TE-371, Terumo Corporation, Tokyo, Japan) with an initial target plasma concentration of 3 μg/mL and then titrated as required. After loss of response to verbal command, 0.6 mg/kg rocuronium was administered, and artificial ventilation was performed with 100% oxygen. Anesthesia was maintained with propofol and remifentanil. Ten minutes after the rocuronium administration, 2 mg/kg 4% lidocaine was administered with a laryngoscope into the trachea over 2 seconds under direct vision. The 2 mg/kg dose was used based on the recommendation that endotracheal drug administration should be 2 to 2.5 times the IV lidocaine dose of 1 to 1.5 mg/kg. When administering lidocaine, approximately 10 mm of the tip of the mucosal atomization device or approximately 35 mm of the tip of the spray tube was inserted into the trachea. Gentle manual ventilation at the rate of approximately 10 breaths/min with airway pressure <20 cm H2O was commenced 30 seconds after lidocaine administration. Two minutes after lidocaine was given, the trachea was intubated using an 8.0-mm tracheal tube for male and 7.5 mm for female patients. Mechanical ventilation was started using pressure control ventilation at 15 cm H2O with an inspired oxygen fraction of 50%. Respiratory frequency was adjusted to maintain the end-tidal carbon dioxide between 35 and 40 mm Hg.
Sample Acquisition and Drug Assay
Arterial blood samples (2 mL each) were taken before; at 1, 3, 5, 7, 10, 15, 20, 30, 45, and 60 minutes; and then every 60 minutes after the administration of lidocaine until the end of the operation. Blood samples were centrifuged, and the plasma was transferred to a polyethylene tube and kept at −20°C until assayed.
Plasma lidocaine concentrations were measured using high-performance liquid chromatography (HPLC; SPD-6AV, CTO-10AS, LC-10AD, SIL-10AD, SCL-10A, and DGU-14A; Shimadzu, Kyoto, Japan) with ultraviolet detection at a wavelength of 210 nm. The mixture of plasma 150 μL, an internal standard (ropivacaine, 10 mg/mL) 100 μL, 1 M sodium carbonate 1 mL, and ethyl acetate 4 mL was stirred for 5 minutes and centrifuged at 3000 rpm for 10 minutes. Three milliliters of the supernatant was transferred to a tube containing 0.0025 M sulfuric acid 200 μL, stirred for 5 minutes, and centrifuged at 3000 rpm for 10 minutes. The sulfuric layer was stirred with 10 mM sodium hydroxide 100 μL, and 280 μL of the solution was then evaporated to dryness under reduced pressure. The residue was dissolved in 100 μL of the mobile phase of HPLC (acetonitrile:methanol:0.05 M NaH2PO4 = 1:3:6), and 20 μL of the sample was injected to an HPLC apparatus. A linear relationship was obtained between lidocaine concentration and peak area (R2 > 0.999) over the range of 0.05 to 10 μg/mL. The limit of detection for lidocaine at a signal-to-noise ratio of 3 was 0.018 μg/mL, and the limit of quantification for lidocaine in plasma was 0.060 μg/mL.
Measured plasma lidocaine concentrations were plotted over time. A pharmacokinetic 2-compartment model with a bioavailability and a first-order absorption model were fit to the raw data8 (Fig. 2) using the first-order conditional estimation method with interaction using a mixed-effects population approach on the NONMEM version VII (ICON Development Solution, Ellicott City, MD). A proportional variation model was used for interindividual variability, and a constant coefficient of variation model was used for intraindividual variability. For the ith subject, an individual pharmacokinetic parameter (V1, V2, Cl2, F, and ka) was determined to be θi. The parameter, θi, was calculated using the following equation: ·, where θTV is the typical value of the parameter in the population and ηi is a random variable in the ith individual with a mean of 0 and a variance of ω2. To determine the bioavailability (F) for endotracheal lidocaine administration, total body clearance (Cl1) was assumed to be 0.646 L/min. This value is the previously published total body clearance for an IV lidocaine infusion.10 A first-order absorption model was used to describe the time delay required for the bioavailable drug to reach the arterial sampling site. A first-order absorption rate constant (ka) was used to estimate drug absorption to the central compartment (Fig. 2). An absorption time was estimated as the reciprocal of ka.
We compared our population pharmacokinetic model using 1 F and 1 ka for all data from both groups (atomized and nonatomized lidocaine) with our population pharmacokinetic model using different Fs for each group: F atomizer for the mucosal atomization device and F spray for the spray tube, and 1 ka. Once the best model of bioavailability was identified, we then modified it with different kas: ka atomizer for the mucosal atomization device and ka spray for the spray tube. For model change, the objective function value was calculated as −2 log likelihood. A decrease in objective function value by 7.88 (P < 0.005, χ2 test with 1 degree of freedom) was considered as a significant improvement of the model.
Model performance was assessed by visual inspection of goodness-of-fit plots, which depict the predicted Cp versus measured Cp of lidocaine or time course of measured Cp divided by predicted Cp, and by calculating median absolute prediction error (MDAPE; defined as the average of individual MDAPEs) for inaccuracy and median prediction error (MDPE; defined as the average of individual MDPEs) for bias of the model predictions as previously described.11 MDAPE < 30% and MDPE between −20% and 20% was regarded as acceptable.12 To assess the systemic bias of the developed pharmacokinetic models, we also calculated MDAPE and MDPE for each device.
Bootstrap analysis was used for advanced internal evaluation of the final pharmacokinetic models. One thousand bootstrap resampling data sets were created by sampling data from the original data set with replacement. The sample size of resampling data sets was the same as that of the original data set. The 2-compartment model with a bioavailability and a first-order absorption model were fitted to the resampling data sets. Confidence intervals of 95% were obtained for each parameter estimate as the 2.5th and 97.5th percentiles.
Predicted peak plasma concentrations and time to predicted peak plasma concentrations were estimated using the post hoc Bayesian individual pharmacokinetic parameter estimates determined by NONMEM.9
Computer simulations were performed to estimate the dose of endotracheal-administered lidocaine required to achieve a near equivalent time course of lidocaine Cp for each delivery technique using the final population pharmacokinetic model.
Variables between groups were compared using an unpaired t test with Welch correction for parametric data or Mann-Whitney U test or Fisher exact test for nonparametric data. P < 0.05 was defined as statistically significant. Data are presented as mean ± SD or median (range). Statistical analyses were done using Prism 6.04 (GraphPad Software, La Jolla, CA).
All data from all patients (10 patients in each group) were included in the analysis. There was no difference in age, gender, weight, and height between groups (Table 1). There were no obvious airway abnormalities found in physical inspection or chest radiographic assessment. Manual ventilation before tracheal intubation was performed without any difficulty in all patients. No patients experienced hemodynamic instability or severe adverse events throughout the study. The volume of lidocaine administered into the trachea was 3.0 mL (2.3−3.7 mL) for the mucosal atomization device or 2.8 mL (2.2−3.3 mL) for the spray tube. The duration of arterial blood sampling was 180 minutes (45−240 minutes). The plasma lidocaine concentration of 1 sample was below the limit of quantification at 1 minute after administration using the mucosal atomization device.
The individual time courses of measured Cp for both groups are presented in Figure 3. The highest measured Cp by the mucosal atomization device (1.9 [1.4−3.2] μg/mL) was significantly greater (P = 0.0021) than that by the spray tube (1.1 [0.6−2.0] μg/mL). In both groups, the highest measured Cps were observed most frequently at 7 minutes after lidocaine administration.
Parameter estimates of the developed pharmacokinetic models and the model performance are given in Table 2. Different Fs, F atomizer and F spray (middle model in Table 2), significantly improved the model compared with 1 F (left model in Table 2) with a decrease of the objective function value by 12.2 (P = 0.0005). The estimated bioavailability for the mucosal atomization device was 24% higher than that for the spray tube. Different ka, ka atomizer, and ka spray (right model in Table 2) did not improve the model compared with 1 ka (middle model in Table 2) with a decrease of the objective function value by 2.9 (P = 0.0880). In the model with different ka, the estimated absorption time was 11.1 minutes for the mucosal atomization device and 14.4 minutes for the spray tube. The final pharmacokinetic model included different Fs and 1 ka.
In the model with 1 F and 1 ka (left model in Table 2), MDPE values were 13.7% for the mucosal atomization device and −16.1% for the spray tube, indicating systematic bias of the model prediction between the devices (P = 0.0180). In the model with different Fs, F atomizer and F spray, and 1 ka (the final model, middle model in Table 2), MDPE values were −5.2% for the mucosal atomization device and 0.5% for the spray tube, indicating no significant systematic bias (P = 0.6431).
The goodness-of-fit plots for the final population and post hoc Bayesian individual pharmacokinetic models are shown in Figure 4. The MDAPE and MDPE were 22.4% and −2.3% for the final model, respectively (Table 2). The goodness-of-fit plots and these calculated indices indicate that the model provided an accurate description of the observed data. One thousand bootstrap runs were obtained with the same number of patients for both devices in the original data set for the final model. The number of successfully converged resamples was 998. The confidence intervals of 95% for parameter estimates are shown in Table 3. The intervals for θ exclude 0 values. These bootstrap results revealed that all the final estimated model parameters have acceptable typical values. The post hoc Bayesian individual estimates between the groups were similar excluding F (Table 3).
The individual time courses of predicted Cp using post hoc final pharmacokinetic parameter estimates are depicted in Figure 5. The predicted peak Cp for the mucosal atomization device (1.7 ± 0.4 μg/mL) was significantly higher than that for the spray tube (1.0 ± 0.3 μg/mL, P = 0.0006).
Figure 6 presents the time courses of predicted lidocaine Cp after 1 bolus into the trachea using the final population pharmacokinetic model. The time courses of the predicted Cp are similar between the 2-mg/kg bolus for the mucosal atomization device and the 3-mg/kg bolus for the spray tube.
Our results confirmed our study hypothesis: plasma lidocaine concentrations were higher over time for atomized lidocaine delivery when compared with nonatomized lidocaine delivery. Of note, all measured plasma lidocaine concentrations were well below the toxic level (6.0 μg/mL).13 This was expected given that our doses of 2 mg/kg were well below the maximum lidocaine doses of 5 mg/kg.14 The clinical implications of this finding may be of some consequence in terms of treating malignant arrhythmia known to respond to IV lidocaine. We found the median highest concentrations of 1.9 vs 1.1 ng/mL depending on which device was used. Plasma lidocaine concentrations reduce or terminate ventricular ectopy over a range of 1.4 to 6.0 μg/mL.15 Thus, the more efficient delivery of lidocaine through a mucosal atomization device may be more likely to achieve therapeutic concentrations when compared with conventional spray devices.
On the basis of our population pharmacokinetic model analysis, we found that atomized lidocaine resulted in higher estimates of bioavailability (80.1%) than nonatomized lidocaine (55.9%) with no significant difference in the absorption rate constant (ka) between the 2 delivery systems (Table 2). These results indicate that the atomization of lidocaine changes its pharmacokinetic profile as lidocaine is transferred from the trachea into the systemic circulation.
By using model parameter estimates of bioavailability presented in Table 3, we estimated near equivalent doses for each device. The mucosal atomization device reduced the lidocaine dose by 70% as defined by F spray/F atomizer (0.559/0.801). Our simulations of lidocaine Cp after a 3 mg/kg dose using the spray tube is near equivalent to a 2.1 mg/kg dose using the mucosal atomization device (Fig. 6). These results suggest that atomized drug administration may be an effective alternative for endotracheal drug administration at a reduced dose.
Of note, mucosal atomization devices with a longer tube are available for use in intubated patients.16 If atomized lidocaine is administered to deeper parts of the respiratory tract during cardiopulmonary resuscitation, the pharmacokinetic profile may be different from what is presented in this study. Vasoconstriction and low cardiac output during resuscitation2 may result in decreased absorption and lower plasma lidocaine concentrations. Reduced clearance during resuscitation may lead to increased plasma lidocaine concentrations.17 Thus, our findings are not generalizable to emergent conditions such as acute cardiac arrest where the pathophysiology is likely to substantially alter the disposition of lidocaine.
To estimate bioavailability with blood samples obtained after endotracheal lidocaine administration, 2 options were available: (1) measure plasma concentrations after IV lidocaine administration18 or (2) assume a total body clearance. We chose to estimate the total body clearance using a previously published value of 0.646 L/min.10 Thus, our estimates of bioavailability based on this assumption may be different from the true bioavailability in any given patient enrolled in our study. Patients in both groups had similar demographics (Table 1), which suggests that total body clearance in each group was similar. Thus, although our bioavailability estimates may themselves be inaccurate, the percent decrease (30%) in bioavailability using the spray device versus the mucosal atomization device may in fact represent a reasonable estimate of how bioavailability is changed using atomized drug delivery.
We presumed a single absorption phase of lidocaine on the pharmacokinetic evaluations after the visual inspection of measured Cp versus time (Fig. 3) although 2 independent absorption phases have been observed in previous studies.19–21 In the previous study, a unique difference in the experimental design was that the authors used 5 forceful manual hyperventilation breaths immediately after administering endotracheal lidocaine. Also different from our study, the lidocaine was diluted into 10 mL normal saline. Distribution of endotracheal-administered drug using forceful hyperventilation22 may result in 2 different absorption rates in the endotracheal/endobronchial mucosa and the alveolar-capillary membrane.20 On the basis of our pharmacokinetic model analysis, we estimated the absorption of lidocaine to be monophasic. Thus, it is likely that most of the lidocaine was absorbed in a consistent manner through the endotracheal/endobronchial mucosa and/or the alveolar-capillary membrane with either device.
When building our pharmacokinetic model, we used a 1-stage approach to estimate population pharmacokinetic parameters with post hoc Bayesian individual pharmacokinetic parameter estimates. We internally validated our model by visual inspection (Fig. 4) and a bootstrap analysis. In addition, post hoc individual Bayesian estimates of distribution volume (V1 and V2) and clearance (Cl2) with a fixed total body clearance (Cl1) were similar between the groups (Table 3). Because these results were expected due to the similar patient characteristics between groups, they suggest that the final model was appropriately developed. With the dense cluster of measured samples during the first 20 minutes for all individuals, we could have used a standard 2-stage approach.23 We elected not to because of the varied duration of blood sampling (45−240 minutes). If we used a 2-stage approach, it is likely we would have had differences in the number of compartments for models that were fit to individual plasma lidocaine concentrations making comparisons between groups difficult. Thus, to avoid the influence of model structure on the parameter estimates of bioavailability and absorption rate, we used a 1-stage approach.
A limitation of this study was that we administered lidocaine before tracheal intubation. It is important to point out that endotracheal drug administration would usually occur after securing the airway with an endotracheal tube during resuscitation. Thus, further study is warranted to investigate endotracheal administration of atomized lidocaine through an endotracheal tube using a longer mucosal atomization device16 compared with the injection of lidocaine directly into the endotracheal tube.
In conclusion, we found that endotracheal administration of lidocaine using a mucosal atomization device resulted in higher plasma concentrations than with conventional spray tubes. We also found that our population pharmacokinetic model estimates of bioavailability were different between the mucosal atomization device and spray tube, whereas the absorption rate was not significantly different between the 2 devices. These results suggest that when using atomized delivery of lidocaine, less drug is required to achieve a near equivalent plasma lidocaine concentration in stable hemodynamic conditions. Atomized drug administration may be an attractive alternative method for endotracheal drug administration. Further study is warranted to explore the potential utility of atomized drug administration to intubated patients or cardiac arrest animals.
Name: Yumiko Takaenoki, MD.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Yumiko Takaenoki has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Kenichi Masui, MD, PhD.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Kenichi Masui has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
Name: Yutaka Oda, MD, PhD.
Contribution: This author helped analyze the data and write the manuscript.
Attestation: Yutaka Oda reviewed the analysis of the data and approved the final manuscript.
Name: Tomiei Kazama, MD, PhD.
Contribution: This author helped analyze the data and write the manuscript.
Attestation: Tomiei Kazama reviewed the analysis of the data and approved the final manuscript.
This manuscript was handled by: Ken B. Johnson, MD.
1. Neumar RW, Otto CW, Link MS, Kronick SL, Shuster M, Callaway CW, Kudenchuk PJ, Ornato JP, McNally B, Silvers SM, Passman RS, White RD, Hess EP, Tang W, Davis D, Sinz E, Morrison LJPart 8: adult advanced cardiovascular life support: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care.Circulation2010122S729–67
2. Prengel AW, Rembecki M, Wenzel V, Steinbach GA comparison of the endotracheal tube and the laryngeal mask airway as a route for endobronchial lidocaine administration.Anesth Analg2001921505–9
3. Hamilton ND, Hegarty M, Calder A, Erb TO, von Ungern-Sternberg BSDoes topical lidocaine before tracheal intubation attenuate airway responses in children? An observational audit.Paediatr Anaesth201222345–50
4. Chapman SThe charges on droplets produced by the spraying of liquids as revealed by the millikan oil drop method.Physics19345150
5. Rebuck AS, Braude ACAssessment of drug disposition in the lung.Drugs198428544–53
6. Rytting E, Nguyen J, Wang X, Kissel TBiodegradable polymeric nanocarriers for pulmonary drug delivery.Expert Opin Drug Deliv20085629–39
7. Yufune S, Takamatsu I, Masui K, Kazama TEffect of remifentanil on plasma propofol concentration and bispectral index during propofol anaesthesia.Br J Anaesth2011106208–14
8. Pétricoul O, Cosson V, Fuseau E, Marchand MEtte EI, Williams PJPopulation models for drug absorption and enterohepatic recycling.Pharmacometrics: The Science of Quantitative Pharmacology2007New JerseyJohn Wiley & Sons, Inc.345–82
9. Owen JS, Fiedler-Kelly JApplications Using Parameter Estimates from the Individual Introduction to Population Pharmacokinetic/Phamacodynamic Analysis with Nonlinear Mixed Effects Models20141st edNew JerseyJohn Wiley & Sons, Inc.198–211
10. Kuipers JA, Boer F, de Roode A, Olofsen E, Bovill JG, Burm AGModeling population pharmacokinetics of lidocaine: should cardiac output be included as a patient factor?Anesthesiology200194566–73
11. Varvel JR, Donoho DL, Shafer SLMeasuring the predictive performance of computer-controlled infusion pumps.J Pharmacokinet Biopharm19922063–94
12. Masui K, Upton RN, Doufas AG, Coetzee JF, Kazama T, Mortier EP, Struys MMThe performance of compartmental and physiologically based recirculatory pharmacokinetic models for propofol: a comparison using bolus, continuous, and target-controlled infusion data.Anesth Analg2010111368–79
13. Gianelly R, von der Groeben JO, Spivack AP, Harrison DCEffect of lidocaine on ventricular arrhythmias in patients with coronary heart disease.N Engl J Med19672771215–9
14. Finnerup NB, Biering-Sørensen F, Johannesen IL, Terkelsen AJ, Juhl GI, Kristensen AD, Sindrup SH, Bach FW, Jensen TSIntravenous lidocaine relieves spinal cord injury pain: a randomized controlled trial.Anesthesiology20051021023–30
15. Collinsworth KA, Kalman SM, Harrison DCThe clinical pharmacology of lidocaine as an antiarrhythymic drug.Circulation1974501217–30
16. Chen KT, Lin HJ, Guo HR, Lin MT, Lin CCFeasibility study of epinephrine administration via laryngeal mask airway using a porcine model.Resuscitation200669503–7
17. Chow MS, Ronfeld RA, Hamilton RA, Helmink R, Fieldman AEffect of external cardiopulmonary resuscitation on lidocaine pharmacokinetics in dogs.J Pharmacol Exp Ther1983224531–7
18. Schnider TW, Minto CFEvers AS, Maze MPharmacokinetic and pharmacodynamic principles of drug action.Anesthetic Pharmacology Physiologic Principles and Clinical Practice2004New YorkChurchill Livingstone1–30
19. West GB, Brown JH, Enquist BJA general model for the origin of allometric scaling laws in biology.Science1997276122–6
20. Prengel AW, Lindner KH, Hähnel J, Ahnefeld FWEndotracheal and endobronchial lidocaine administration: effects on plasma lidocaine concentration and blood gases.Crit Care Med199119911–5
21. Hähnel JH, Lindner KH, Schürmann C, Prengel A, Ahnefeld FWWhat is the optimal volume of administration for endobronchial drugs?Am J Emerg Med19908504–8
22. Greenberg MI, Spivey WHComparison of deep and shallow endotracheal administration of dionosil in dogs and effect of manual hyperventilation.Ann Emerg Med198514209–12
© 2016 International Anesthesia Research Society
23. Sheiner LB, Beal SLEvaluation of methods for estimating population pharmacokinetic parameters. II. Biexponential model and experimental pharmacokinetic data.J Pharmacokinet Biopharm19819635–51