Fentanyl is widely used off-label in NICU. Our aim was to investigate its cerebral, cardiovascular and pulmonary effects as well as pharmacokinetics in an experimental model for neonates.
Fentanyl (5 µg/kg bolus immediately followed by a 90 minute infusion of 3 µg/kg/h) was administered to six mechanically ventilated newborn piglets. Cardiovascular, ventilation, pulmonary and oxygenation indexes as well as brain activity were monitored from T = 0 up to the end of experiments (T = 225–300 min). Also plasma samples for quantification of fentanyl were drawn.
A “reliable degree of sedation” was observed up to T = 210–240 min, consistent with the selected dosing regimen and the observed fentanyl plasma levels. Unlike cardiovascular parameters, which were unmodified except for an increasing trend in heart rate, some of the ventilation and oxygenation indexes as well as brain activity were significantly altered. The pulmonary and brain effects of fentanyl were mostly recovered from T = 210 min to the end of experiment.
The newborn piglet was shown to be a suitable experimental model for studying fentanyl disposition as well as respiratory and cardiovascular effects in human neonates. Therefore, it could be extremely useful for further investigating the drug behaviour under pathophysiological conditions.
Citation: Rey-Santano C, Mielgo V, Valls-i-Soler A, Encinas E, Lukas JC, Vozmediano V, et al. (2014) Evaluation of Fentanyl Disposition and Effects in Newborn Piglets as an Experimental Model for Human Neonates. PLoS ONE 9(3): e90728. https://doi.org/10.1371/journal.pone.0090728
Editor: Lynette K. Rogers, The Ohio State Unversity, United States of America
Received: October 23, 2013; Accepted: February 3, 2014; Published: March 4, 2014
Copyright: © 2014 Rey-Santano et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was partially supported by grants from the Spanish Carlos III Health Institute: FIS 10/943 and RD12/0026/0001 (Red SAMID: Maternal, Child Health and Development Research Network, within the framework of the VI Spanish National Plan for R+D+i 2008-2012). It was also supported by a predoctoral grant to one of the authors (EE) from the Research Staff Training Program (Basque Country Government) and by an INNPACTO project of the Spanish Ministry of Science and Innovation — MICINN (ITP-09000-2010-002), which is financed by the European Regional Development Fund (FEDER). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Fentanyl is a potent synthetic opioid which is increasingly used in neonatal intensive care unit (NICU) and pediatric intensive care unit (PICU) as postoperative analgesic, and as a sedative for patients requiring mechanical ventilation. Fentanyl is frequently preferred over morphine because of its wider therapeutic index, lack of histamine release and limited adverse effects on pulmonary and hemodynamic function . Moreover, the higher lipid solubility and less complex receptor binding of fentanyl as compared to morphine allow a rapid penetration of the blood-brain barrier and a rapid onset of action . Its use in pediatric critical care has significantly increased for the past 20 years; however, it remains one of the many medications which have not been properly tested in this population .
In clinical practice, pediatric dosing regimens are often empirically derived from those employed in adults on a body weight basis. However, this procedure is not fully justifiable because ontogenic maturation processes related to drug pharmacokinetics and pharmacodynamics (PK/PD) are not necessarily always body weight correlates. Specifically for fentanyl, the impact of ontogeny is considered to be more outstanding regarding PK, as the drug is subject to hepatic biotransformation  and highly bound to plasma proteins, primarily α1-acid glycoprotein (AAG) , . In this sense, a maturation physiology based predictive PK/PD model for fentanyl in neonatal care was built, starting from the drug disposition in adults and including those developmental changes associated to growth (enzyme activity, organ weights, blood flows, AAG, etc.) .
The relevance of this type of models lies in that it allows minimizing the chance of overdose-associated adverse effects (e.g. skeletal muscle rigidity, hypoxia, desaturations, hypercapnia and limited haemodynamic imbalance) –, which might be more severe in the youngest patients. In sick newborn infants with impaired cerebral autoregulation, those adverse effects might in turn cause fluctuations in cerebral blood flow (CBF) and electroencephalogram abnormalities, which are risk factors for brain injury and long-term disability. Nonetheless, virtually no data concerning the effects of fentanyl administration on cerebral activity, metabolism and circulation is available for neonates.
Given the difficulties of doing pharmacological research in the neonatal population, the performance of experimental studies in suitable animal species is also a common practice, often in addition or as a complement to previously developed theoretical models. In this respect, pigs have been used as experimental animals for a long time, because many of their anatomical and physiological characteristics more closely resemble those of humans than other non-primate species , . Specifically, the newborn piglet is a representative model for the cardiovascular physiologic development of neonates . Furthermore, similarities in porcine drug metabolizing enzymes suggest that pigs may be the most suitable animal model for drug biotransformation studies. In this sense, activity of the most important CYP isoform in humans, CYP3A4, also responsible for hepatic fentanyl metabolism, is present in pigs with comparable levels and activities, as opposed to other routinely used experimental animals , . Similarly, for some drugs, the differences observed between juvenile and adult pig PK were deemed as consistent with ontogenic changes reported for human PK .
Although direct extrapolation of animal data to humans is not appropriate, we hypothesized that preclinical evidence from the newborn piglet model elucidating the potential impact of pharmacotherapy on the developing brain forms could be an essential bridge to human studies and informed clinical practice. The purpose of this study was to investigate the main cardiovascular and pulmonary effects as well as the PK profile of fentanyl when administered as the only narcotic agent in healthy mechanically ventilated newborn piglets. Also, the brain effect of fentanyl was tested by using the two devices available for actual bedside studies in neonates: near-infrared spectroscopy (NIRS), a non-invasive optical technique for assessing circulation and oxygenation in brain, together with amplitude-integrated electroencephalography (aEEG). As a second objective, the study was to provide further confirmatory evidence of the previously developed physiology based predictive PK/PD model for fentanyl in neonatal care .
A prospective study was performed in eight newborn piglets (2–4 days, 1.7±0.2 kg) of either sex. This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals. The experimental protocol met European and Spanish regulations for protection of experimental animals (86/609/EEC and RD 1201/2005) and was approved by the Ethical Committee for Animal Welfare of the Cruces University Hospital.
The animals were sedated with i.m. ketamine (15 mg/kg) and diazepam (2 mg/kg) administration. A face mask was applied and animals were anesthetized with 1.5%–2% of sevoflurane and oxygen.
A tracheotomy was performed; a tracheal tube (4.0 mmID) was inserted and connected to an inhalator anaesthesia system. Animals were then positive pressure ventilated (VIP Bird, Bird Products Corp., Palm Springs, CA) with the following initial settings: fraction of inspired oxygen (FiO2) = 0.25, respiratory frequency (fR) = 20 breaths/min, positive end-expiratory pressure (PEEP) = 3 cmH2O and positive inspiratory pressure (PIP) = 9 cmH2O. Deviations from acceptable blood gases values (PaO2 90–110 mmHg, PaCO2 35–45 mmHg and pH 7.35–7.45) were corrected by changing ventilator parameters and/or by adding sodium bicarbonate as needed.
A thermodilution arterial catheter (5Fr, PiCCO Plus, Pulsion, München, Germany) was inserted into the femoral artery to monitor mean arterial blood pressure (MABP), heart rate (HR) and cardiac output (CO) as well as to obtain blood samples for gas analysis and subsequent determination of fentanyl plasma concentration. Also, a 5 Fr three-lumen catheter was inserted into the internal jugular vein to allow injection of cold saline, measurement of central venous pressure (CVP), maintenance of fluids (5 ml/kg/h) and fentanyl infusion. Jugular vein catheterization was used to infuse vecuronium (3 mg/kg bolus every 1 hour until the end of fentanyl infusion) to avoid breathing and muscle movements that may distort the two-channel bedside aEEG (BRM2; BrainZ Instruments, Auckland, New Zealand), NIRS system (NIRO-200; Hamamatsu Photonics KK, Joko-Cho, Japan) or lung mechanics data. Lung tidal volume and pulmonary dynamic compliance (Cdyn) were continuously monitored by computerized pneumatocography (Mod. M2780A, Philips Medizin Systeme, Boblingen, Germany).
A pilot study in two initial piglets
Testing in the two first piglets was conceived as a pilot study aimed to refine, if necessary, the dosing schedule, the sampling times and/or the duration of the experiment. These animals were administered i.v. fentanyl bolus dose of 4.8 µg/kg and a simultaneously starting 30 minute infusion at a rate of 6 µg/kg/h, while trials were set to be stopped at 90 minutes after infusion discontinuation. The mentioned regimen was selected based on the PK parameters available in the literature for fentanyl in newborn piglets  to achieve a concentration at steady state (Css) of 5 ng/ml. In the absence of specific data for piglets, the concentration reported to produce 50% maximum sedation effect (EC50) in human newborns (i.e. 3 ng/ml)  was assumed to be also true for this species, so the aim was to preclude any possibility of alertness in those animals, in line with the shown ethical concern.
Observations from the pilot phase were compared with the estimates provided by the ontogeny based PK/PD model in the neonate  and served to redefine the study protocol ahead of the experimentation in the remaining planned animals. Only the results corresponding to the final protocol (n = 6) were included in the overall analysis.
Experimental protocol and acquisition of physiological variables
After surgical preparation, baseline conditions (Basal1) were registered, following which the inhalator anaesthesia system was stopped. 5 min later, time considered enough to eliminate the sevoflurane anesthetic effects, basal values were checked again (Basal2). Immediately afterwards, animals received an i.v. fentanyl bolus dose (5 µg/kg) and a simultaneously starting continuous i.v. infusion of fentanyl at a rate of 3 µg/kg/h for 90 min.
Arterial blood samples for gas analysis (pH, PaO2, PaCO2, and arterial oxygen saturation: SaO2) and 1 ml for subsequent quantification of fentanyl plasma levels were withdrawn at the following timepoints: Basal 1, Basal 2, immediately after bolus administration (Bolus), 1, 10, 30, 90, 95, 120 and 180 min after the start of the infusion and then every 30 min until the end of the experiment.
Also, at each timepoint cardiovascular parameters (MABP, HR, and CO), ventilator parameters (PIP, PEEP, fR and FiO2) and lung mechanics were registered. Moreover, OI: mean airway pressure (cmH2O)*FiO2/PaO2 (mmHg) and ventilator efficiency index (VEI): 3800/PIP-PEEP*fR (breath/min)*PaCO2 (mmHg) were also calculated.
The end of the experiment was arbitrarily established as the moment when application of a painful stimuli resulted in a significant HR increase (≥15 beats/min) and/or an increment in the aEEG amplitude activity greater than 30% over the previous measurement. At the end of each experiment, the piglets were sacrificed with an overdose of sevoflurane, vecuronium and potassium chloride.
Change in cerebral perfusion-oxygenation was assessed using NIRS system. The sensor was placed on the skull in the midline fronto-parietal area. Tissue oxygen index (TOI) and variations in tissue hemoglobin index (THI) were continuously monitored. TOI represents the tissue oxygen saturation and is measured in percent; it was used to calculate the fractional tissue oxygen extraction (FTOE) : FTOE = (SaO2-TOI)/SaO2. THI is an absolute figure of the total Hb content in brain; its changes were used to estimate changes in cerebral blood volume . The cerebral intravascular oxygenation (CIO), equivalent to the difference between oxyhemoglobin and deoxyhemoglobin, was used as surrogate of CBF changes .
Brain activity was monitored using a two-channel bedside aEEG monitor with five needle electrodes. The aEEG background activity and the aEEG amplitude were measured. NIRS and aEEG parameters were continuously recorded throughout the experimental period.
Handling and measurement of fentanyl in plasma samples
Arterial blood samples for analysis of plasma fentanyl concentrations were collected in EDTA tubes, and kept on ice until their immediate centrifugation at 3000 rpm for 10 minutes. Once separated, plasma was transferred to cryovials and initially frozen at –20°C, followed by a subsequent storage at –80°C until analyzed. Plasma fentanyl levels were measured by a selective and sensitive high-performance liquid chromatography-tandem mass spectrometry (HPLC/MS-MS) method. The assay, specific for fentanyl and not accounting for its metabolites, had a sensitivity of 0.2 ng/ml, as represented by the lower limit of quantification (LLOQ). Intra and interassay coefficients of variation did not exceed 5% and 15%, respectively.
Evaluation of the pharmacokinetic variables
Noncompartmental pharmacokinetic analysis (NCA) was performed on the plasma concentration data using PhoenixTM WinNonLin (version 6.1, Pharsight Corporation, St. Louis, MO, USA). Solely those plasma samples with fentanyl levels above the LLOQ were considered for the analysis. Peak plasma concentration (Cmax), time to reach Cmax (tmax), last quantifiable concentration (Clast) and the corresponding timepoint for Clast (Tlast) were obtained directly from the measured concentration-time curves. The area under the plasma concentration-time curve from time 0 to the last measurable concentration (AUC0-t) was calculated by the trapezoidal formula. Systemic clearance (CL) was calculated from the ratio between the infusion dose and the exposure (AUC0-t).
Furthermore, an exploratory analysis aimed to elucidate the relationship between these PK variables and the observed effect was performed. Evaluation was focused on the degree of sedation, represented by aEEG measured brain activity, as the main effect targeted by drug administration.
Determination of plasma AAG as a potential source of PK variability
In an attempt to partially account for interindividual PK variability (IIV) on fentanyl disposition and to assess if plasma levels of this acute phase protein were modified by the experimental procedure, concentrations of AAG in plasma samples (basal, 90 min and last timepoint) from six piglets were determined using a Porcine AAG ELISA kit (AMS Biotechnology Ltd., Oxfordshire, UK). Analyses were performed by duplicate and the absorbance was measured in a microplate reader (POLARstar, BMG, Ortenberg, Germany) at 450 nm. For further comparison, AAG was also quantified in blank plasma samples from additional newborn piglets acting as controls.
Data was analyzed using JMP statistical discovery software (version 8, SAS Institute Inc., North Carolina, USA). Comparison of measured values before and after fentanyl administration was assessed by t test of paired means. Simple linear regression analysis was done to assess the relationship between plasma AAG levels and fentanyl exposure. A p< 0.05 was considered statistically significant. Values are expressed as mean ± SD.
Observations from the pilot study (n = 2)
NCA of the plasma profiles in two initial piglets revealed an average CL of 0.022 L/min, which is of the same order of magnitude as the one calculated by the above mentioned ontogeny predictive model for human newborns (i.e., 0.028 L/min) . Based on this fact, fentanyl disposition in piglets was assumed to be comparable to human newborns and, consequently, the protocol was amended to adapt the dosing schedule from that usually employed in the NICU for sedation (i.e., a 2–5 µg/kg bolus immediately followed by an infusion of 1–3 µg/kg/h). Namely, the highest dose level in the range was found suitable, via simulation, to provide a Css of roughly 3 ng/ml, which was deemed as an appropriate target concentration in view of the somewhat excessive degree of sedation attained for a Css of 5 ng/ml, as per the observed aEEG (data not shown). In addition, the observation that animals were sufficiently controlled when receiving fentanyl as the sole narcotic agent encouraged to enlarge both the infusion duration and the observation period after infusion discontinuation.
Overall data corresponding to the final protocol (n = 6)
Baseline conditions were consistent with those previously reported for newborn piglets . The animals remained stable throughout the study and presented a reliable degree of sedation for 120–150 min after fentanyl infusion; only some agitation was noted at the end of the experiment. In this sense, the study period upon completion of fentanyl administration was 150 min in two of the six animals while it was ≥180 min in the remaining.
Arterial pH and PaCO2 transiently changed (Table 1) due to a significant decrease in Cdyn (Figure 1A) observed after fentanyl bolus administration. Due to changes performed in the ventilatory setting of mechanical ventilator, gas exchange was maintained within normal ranges for all the study. However, VEI and OI were significantly altered during the first minute and during initial 10 min, respectively, following fentanyl bolus administration (Figure 1B). Those parameters remained unchanged for 120 min after completion of the fentanyl infusion. At the end of the experiment (last 30 min in each animal procedure) the gas exchange (Table 1), Cdyn (Figure 1A) and ventilation indexes (Figure 1B) showed great variations due to the agitation perceived in animals.
A- Mean dynamic compliance, Cdyn, B- mean oxygenation index, OI, (white squares) and mean ventilator efficiency index,VEI, (black squares). Values are expressed as mean±SD. Normal values: OI<10; VEI>0.3; Cdyn: 0.8–1.5 ml/cmH2O/kg. The letter “B1” indicates the baseline measurement during inhaler anesthesia and letter “B2” indicates the baseline measurement after elimination of inhaler anaesthesia. *p <0.05 vs. baseline 2 level.
Cardiovascular response is shown in Table 1. Compared to pre-treatment values (Basal 2), HR increased from T = 90 min to the end of the experiment with no significant change in MABP, CO or CVP.
Fentanyl administration produced a continuous decrease of CIO over time, reaching statistical significance at the end of the infusion (T = 90 min) (Figure 2A). This alteration in NIRS variables was maintained for the following 120–150 min, i.e. until initial awakening of the animals. In contrast, FTOE was only partially increased during fentanyl administration (Figure 2A).
A-Cerebral perfusion-oxygenation using NIRS system. Mean cerebral intravascular oxygenation, CIO, (black squares) and mean fractional tissue oxygen extraction, FTOE, (white squares). B- Brain activity using aEEG system. Normal values: FTOE: 0.3–0.6; CIO: (–10) –10. The letter “B1” indicates the baseline measurement during inhaler anesthesia and letter “B2” indicates the baseline measurement after elimination of inhaler anaesthesia. Mean aEEG background activity (black squares) and mean aEEG amplitude (white squares). *p <0.05 vs. baseline 2 level.
The median aEEG background activity pattern decreased from 37±12 µV to 20±5 µV after fentanyl bolus administration, which represents a reduction of 41±19% (Figure 2B). This depressed activity remained unchanged during fentanyl infusion and for approximately 120–150 min afterwards. The aEEG activity was partially recovered from T = 210 min to the end of the experiment, reaching levels only 25% lower than the basal values at the last timepoint measured in each individual study. This trend was also applicable to the aEEG amplitudes registered at each timepoint, which followed the same behaviour as the median aEEG amplitude. Indeed, this parameter was significantly reduced after fentanyl bolus (35±12%) (Figure 2B) and remained unchanged until T = 210–240 min, when the differential amplitude began to increase, reaching levels below the basal values by only 12%.
The time course of fentanyl plasma concentrations (Cp) measured in the six piglets’ population is shown in Figure 3. The mean fentanyl Cp was 14.8±9.0 ng/ml immediately after bolus administration, 5.9±1.4 ng/ml at 1 min, 3.5±0.9 ng/ml at 10 min, 2.8±1.5 ng/ml at 30 min and 2.4±1.6 ng/ml at 90 min during fentanyl infusion. 5, 30 and 60 min after infusion was stopped, mean Cp was 1.9±1.3, 1.5±1.3 and 1.1±0.8 ng/ml, respectively, which continued decreasing until the last timepoint measured (up to a maximum of 210 min after stopping infusion). Cp at the latest sampling times was mostly very close to the LLOQ of the analytical technique (i.e. 0.2 ng/ml), with solely two piglets showing Cp below this limit at times beyond 150 and 210 min, respectively. Moreover, the fact that plasma fentanyl levels had been cleared to a high extent by the end of experiments (T = 225–300min) is consistent with animals showing initial signs of awakening at this timepoint.
PK variables calculated by NCA are shown in Table 2. As expected, Cmax corresponded to the sample withdrawn immediately after bolus administration in all cases. Nonetheless, an important IIV was evidenced by these data, given that coefficient of variation (CV) was as high as 60% for Cmax, 56% for plasma concentration-time curve (AUC), and 64% for CL.
Notwithstanding, calculated CL was 0.022 L/min on average, which is again in line with the value predicted for human neonates (i.e., 0.028 L/min) . Moreover, this is endorsed by the fact that average plasma concentration of fentanyl at the end of infusion tended to be 3 ng/ml as expected, even if a 90 min continuous infusion was not enough for reaching steady state (ss) conditions (Figure 3).
Exploratory analysis revealed that none of the PK variables calculated by NCA was significantly correlated with the duration or depth of the sedative effect, nor with any of the other measured effects.
Plasma AAG as a potential source of PK variability.
In most cases, there was no appreciable change on the measured AAG levels between the three plasma samples analyzed within each animal (basal, 90 minute and final) (Figure 4A). Moreover, the absence of significant differences when compared to control animals (data not shown) supports the observation that experimental procedure did not trigger an acute-phase response as represented by elevated plasma AAG levels in these animals, at least within the short period set by the study protocol (i.e., up to 5 hours).
A- Individual concentrations of AAG determined by the ELISA method in plasma samples drawn from the six piglets at baseline, T = 90 min and Tlast. B-Correlation between the mean plasma AAG level in each animal (calculated from the measurements at baseline, T = 90min and Tlast) and the corresponding calculated drug exposure (AUC0-t).
On the other hand, IIV could partially be explained by the differences in measured AAG levels. In this sense, a positive correlation was observed between the mean plasma AAG level in each animal and the corresponding calculated drug exposure (AUC0-t), correlation coefficient (R2) being 0.749 (Figure 4B).
In clinical practice, pediatric dosing regimens of many medications used in NICU and PICU  are often empirically derived from those employed in adults on a body weight basis. However, developmental changes may affect the relationship between dose and exposure (PK) and/or the relationship between exposure and response (PD), which can lead either to an efficacy or safety concern. Development of scientific models predictive of the drug behaviour at the different pediatric stages by considering the changes associated to growth and maturation represents a valid approach towards dosing individualization . In addition or as a complement to theoretical models, the performance of experimental studies in suitable animal species is also a necessary practice. Specifically, the newborn piglet is a representative model for newborn human cardiovascular physiologic development,  and also has similar biotransformation pathways to those in humans, including comparable levels and activities for CYP3A4, the isoform responsible for hepatic fentanyl metabolism –.
In the present study, the dose of fentanyl administered to newborn piglets (5 µg/kg bolus and 3 µg/kg/h infusion according to the final protocol) is similar to that used in the NICU when fentanyl is given as the sole or principal analgesic agent to neonates with respiratory distress syndrome maintained in mechanical ventilation , . The dosing regimen as well as the sampling times and the duration of the study were designed based on the results of a pilot study performed in the first two animals. The systemic drug clearance calculated from this study was 0.022 L/min on average, which is of the same order of magnitude as the one provided by a previously developed ontogeny based model predictive of fentanyl disposition in the newborn (i.e., 0.028 L/min) . Those results not only support the validity of the developed scientific model but also openly endorse, to the best of our knowledge for the very first time, the suitability of the newborn piglet as an experimental model for human neonates concerning fentanyl PK.
Since its development in 1960, it is known that fentanyl and other synthetic opioids may cause skeletal muscle rigidity in neonates and infants after low-dose fentanyl administration –. Similar findings were noted in our newborn piglets in that fentanyl-induced chest wall rigidity occurred almost immediately after the administration of low doses of fentanyl. Accordingly, newborn animals showed a transient alteration of gas exchange, a significantly increased OI and a decreased VEI, corresponding with a simultaneous decrease in Cdyn accompanied by the need for increased PIP to maintain an adequate oxygenation and tidal volume.
Robinson and Gregory revolutionized practice after reporting the first use of fentanyl, as the principal anesthetic agent in neonates undergoing ductus ligation surgery . Using HR and MABP responses as an index of adequate anesthesia, these investigators demonstrated that the administration of fentanyl could provide anesthesia with minimal hemodynamic consequences. Subsequent studies also reported hemodynamic stability associated with no change in HR, MABP or CO –. However, changes in MABP without alteration in HR or vice versa have also been observed , . In all reported studies, hemodynamic instability is rare, as long as a vagolytic agent (pancuronium and/or atropine) is administered concomitantly.
Usually, in studies carried out in animal models, most cardiovascular variables including HR, MABP and CO, remained unchanged after administration of a wide range of doses , . The present findings confirmed stability of CO and MABP after low doses of fentanyl while a progressive increase of HR was observed. This effect had been previously noticed in different experimental models receiving fentanyl, thus suggesting that the observed continuous increase of HR was induced by the drug , . Although the mechanism for this is not known, in the absence of more supportive data available, we hypothesized that it may be initially mediated by a withdrawal of vagal tone rather than increased sympathetic output . This hypothesis was based on the unresponsiveness (i.e., no changes in HR, MABP or aEEG) observed in the present study after application of specific painful stimuli at different time intervals from T = 90 min to T = 210 min. However, there was a response to a painful stimulus administered beyond 210 min, coinciding with the initial disappearance of fentanyl effects, which indicated that a sympathetic component may be contributing to maintain elevated HR values in this case.
Besides hemodynamic stability, little is known about the effect of fentanyl in CBF, brain metabolism and aEEG activity in the newborn , . In our study, a decrease of tissue hemoglobin index (THI, reflecting changes in cerebral blood volume) (data not shown) and CIO (reflecting changes in CBF), with only a partial increase of FTOE (oxygen consumption) was observed over time as long as the effect of fentanyl was maintained (not response to pain stimuli); thus suggesting that there may be not enough oxygen supply to meet the metabolic requirements of the brain , . Moreover, aEEG background activity pattern as well as the aEEG amplitude was diminished during the entire fentanyl effective period. The depressing effect of opioids in aEEG had been previously described in preterm and newborn infants –.
Those brain effects returned to basal (pre-infusion) values when the effect of fentanyl ended (response to pain stimuli) suggesting only a transient effect related with fentanyl administration. However, our study was carried out in healthy newborn piglets without any cardiac, hepatic or pulmonary injury, while in clinical practice fentanyl is administered to critically ill newborn infants, who may respond to analgesics in a different manner, given that they usually present hemodynamic parameters already altered at baseline. Thus, more attention should be paid to the effects of sedation and analgesia in sick newborn with impaired cerebral autoregulation, in which fluctuations in CBF and EEG abnormalities could be aggravated by fentanyl administration, even at very low doses. We suggest that the described experimental model could be extremely useful within the purpose of further investigating the impact of fentanyl administration under pathophysiological conditions, as its validity for evaluating respiratory and cardiovascular effects in human neonates has been sufficiently supported by a number of reports.
Otherwise, the observed interindividual PK variability (IIV) seemed not related to sex or body weight, and it may be due to distribution or metabolism related factors. In order to assess the impact of metabolism, further evaluation of individual hepatic blood flow and enzyme activity would be needed, even if maintenance of temperature constant at 37.5–38.5°C throughout the study allowed to preclude variations derived from the strong temperature-dependence of hepatic CYP3A4 activity, as previously shown in juvenile pigs .
On the other hand, the mentioned variability could partially be explained by the differences in measured AAG levels, in line with the following pieces of evidence: 1) Fentanyl is highly (80–85%) bound to plasma proteins, primarily AAG, the principal binding protein for basic drugs , ; 2) Protein binding is one of the determinant factors in drug disposition, known to influence both drug distribution and body clearance, because only unbound drug can transfer among plasma, interstitial fluid and tissue fluid , ; 3) AAG is one of the most important acute- phase proteins with significant clinical implications, as it becomes elevated in plasma in response to a variety of insults and pathophysiological conditions including surgical interventions, inflammation and stress . Even if the third statement could not be confirmed, at least within the short time frame of this protocol, two first assumptions were supported by the clear correlation observed between AAG plasma levels and drug exposure (AUC0-t), which is mainly related to systemic clearance. On the contrary, differences in AAG levels were not sufficient to explain the observed variability in Cmax, most probably because it is influenced by a number of additional factors, not easily measurable in the experimental model, affecting the volume of distribution, as well as by the difficulties linked to accurate extraction of first blood sample immediately after bolus administration. Notwithstanding, elevated plasma peak concentration obtained after the initial bolus dose was not associated with the appearance of adverse effects and rapidly declined (Figure 3) in an exponential manner (i.e., rate of elimination directly proportional to the concentration value), thus having little influence at the latest times, where plasma concentration becomes dependent on the constant rate infusion.
In addition, this study demonstrated the feasibility of maintaining animals adequately controlled solely by fentanyl. This is in line with others reports in the scientific literature, where fentanyl was infused to newborn piglets as the only narcotic agent for as long as 6 h, even if at higher doses . In the present study, Cp <3 ng/ml (the target EC50 described for humans newborn) produced adequate degree of sedation in piglets, which may be due to several reasons, including the high observed IIV, the degree of plasma protein binding or a lack of equilibrium between plasma and the site of the effect in the central nervous system (CNS). In this sense, a more comprehensive PK/PD analysis of the data, which is outside the scope of this paper, is needed in order to define a formal concentration-response relationship. Moreover, factors suggested by the present study to have a probable impact on this issue, such as the drug concentration at the site of effect (via sequential sampling from the CNS) and the unbound drug fraction, including the degree of functionality and drug binding affinity for AAG at the neonatal period, should be considered in future experimental protocols.
In conclusion, the attained level of sedation in fentanyl treated newborn piglets was associated with some degree of chest wall rigidity and depressed brain activity, which were only transient as they mostly returned to basal values by the end of experiments. This experimental model is generally regarded as suitable for evaluating respiratory and cardiovascular effects in human neonates, and could therefore be useful to further investigate the impact of fentanyl administration under pathophysiological conditions. Also, fentanyl PK was shown to be comparable between newborn piglets and human neonates, which served to confirm the adequacy of a previously developed predictive model for this age.
The authors of this manuscript would like to gratefully acknowledge the excellent technical assistance of Ariadna Ochoa and Estibaliz Ruiz-del-Yerro as well as the outstanding contribution of Dr. Rosario Calvo. We also want to thank the staff of the Analytical Chemistry Department of the Science and Technology Faculty at the University of the Basque Country (UPV/EHU) for their assistance in the development and validation of the HPLC/MS-MS method.
Conceived and designed the experiments: CRS VEM AVS ES JCL. Performed the experiments: CRS VEM EE. Analyzed the data: EE VEM ES CRS VEM. Contributed reagents/materials/analysis tools: AVS VV JCL. Wrote the paper: CRS VEM EE ES.
- 1. Rosow CE, Moss J, Philbin DM, Savarese JJ (1982) Histamine release during morphine and fentanyl anesthesia. Anesthesiology 56: 93–96.
- 2. Rajan V, Beharry KD, Williams P, Modanlou HD (1998) Pharmacodynamic effects and pharmacokinetic profile of continuous infusion fentanyl in newborn piglets. Biol Neonate 74: 39–47.
- 3. Hsu B, Brazelton T (2009) Off-label medication use in an academic hospital pediatric critical care unit. Wisconsin Medical Journal 108: 343–348.
- 4. Labroo RB, Paine MF, Thummel KE, Kharasch ED (1997) Fentanyl metabolism by human hepatic and intestinal cytochrome P450 3A4: implications for interindividual variability in disposition, efficacy, and drug interactions. Drug Metab Dispos 25: 1072–1080.
- 5. McNamara PJ, Alcorn J (2002) Protein binding predictions in infants. AAPS PharmSci 4: E4.
- 6. Sen S, Ytrebø LM, Rose C, Fuskevaag OM, Davies NA, et al. (2004) Albumin dialysis: a new therapeutic strategy for intoxication from protein-bound drugs. Intensive Care Med 30: 496–501.
- 7. Encinas E, Calvo R, Lukas JC, Vozmediano V, Rodriguez M, et al. (2013) A predictive pharmacokinetic/pharmacodynamic model of fentanyl for analgesia/sedation in neonates based on a semi-physiologic approach. Pediatric Drugs 15: 10.
- 8. Dewhirst E, Naguib A, Tobias JD (2012) Chest wall rigidity in two infants after low-dose fentanyl administration. Pediatr Emer Care 28: 465–468.
- 9. Fahnenstich H, Steffan J, Nikolaus K, Bartmann P (2000) Fentanyl-induced chest wall rigidity and laryngospasm in preterm and term infants. Crit Care Med 28: 836–839.
- 10. Prakash S, Mehra V, Gogia AR (2010) Fentanyl induced rigidity in an infant. J Anaesthesiol Clin Pharmacol 26: 567–568.
- 11. Irazuzta J, Pascucci R, Perlman N, Wessel D (1993) Effects of fentanyl administration on respiratory system compliance in infants. Crit Care Med 21: 1001–1004.
- 12. Wells S, Williamson M, Hooker D (1994) Fentanyl-induced chest wall rigidity in a neonate: a case report. Heart Lung 23: 196–198.
- 13. Schieber RA, Stiller RL, Cook DR (1985) Cardiovascular and pharmacodynamic effects of high-dose fentanyl in newborn piglets. Anesthesiology 63: 166–171.
- 14. Anzenbacher P, Soucek P, Anzenbacherová E, Gut I, Hrubý K, et al. (1998) Presence and activity of cytochrome P450 isoforms in minipig liver microsomes. Comparison with human liver samples. Drug Metab Dispos 26: 56–59.
- 15. Roth WJ, Kissinger CB, McCain RR, Cooper BR, Marchant-Forde JN, et al. (2013) Assessment of juvenile pigs to serve as human pediatric surrogates for preclinical formulation pharmacokinetic testing. AAPS J 15: 763–774.
- 16. Fritz HG, Holzmayr M, Walter B, Moeritz KU, Lupp A, et al. (2005) The effect of mild hypothermia on plasma fentanyl concentration and biotransformation in juvenile pigs. Anesth Analg 100: 996–1002.
- 17. Aranda JV, Carlo W, Hummel P, Thomas R, Lehr VT, et al. (2005) Analgesia and sedation during mechanical ventilation in neonates. Clinical Therapeutics 27: 877–899.
- 18. Alvarez FJ, Lafuente H, Rey-Santano C, Mielgo VE, Gastiasoro E, et al. (2008) Neuroprotective effects of the nonpsychoactive cannabinoid cannabidiol in hypoxic-ischemic newborn piglets. Pediatric Research 64: 653–658.
- 19. Tsuji M, Duplessis AJ, Taylor GA, Crocker R, Volpe JJ (1998) Near infrared spectroscopy detects cerebral ischemia during hypotension in piglets. Pediatr Res 44: 591–595.
- 20. Guinsburg R, Kopelman BI, Branco de Almeida MF, de Araujo Peres C, Miyoshi MH (1996) Physiological, hormonal, and behavioral responses to a single fentanyl dose in intubated and ventilated preterm neonates. J Pediatr 132: 954–959.
- 21. Robinson S, Gregory GA (1981) Fentanyl-air-oxygen anesthesia for ligation of patent ductus arteriosus in preterm infants. Anesth Analg 60: 331–334.
- 22. Yaster M (1987) The dose response of fentanyl in neonatal anesthesia. Anesthesiology 66: 433–435.
- 23. Anand KJ, Sippell WG, Aynsley-Green A (1987) Randomised trial of fentanyl anaesthesia in preterm babies undergoing surgery: effects on the stress response. Lancet 1: 62–66.
- 24. Hickey PR, Hansen DD, Wessel DL, Lang P, Jonas RA (1985) Pulmonary and systemic hemodynamic responses to fentanyl in infants. Anesth Analg 64: 483–486.
- 25. Friesen RH, Henry DB (1986) Cardiovascular changes in preterm neonates receiving isoflurane, halothane, fentanyl, and ketamine. Anesthesiology 64: 238–242.
- 26. Collins C, Koren G, Crean P, Klein J, Roy WL, et al. (1985) Fentanyl pharmacokinetics and hemodynamic effects in preterm infants during ligation of patent ductus arteriosus. Anesth Analg 64: 1078–1080.
- 27. Modanlou HD, Pan T, Riaz H, Samson T, Sheikh R, et al. (1996) Effects of continuous infusion fentanyl citrate on cerebrovascular and systemic prostanoids in postsurgical newborn piglets. Journal of Investigative Medicine 44: 362–369.
- 28. Yaster M, Koehler RC, Traystman RJ (1987) Effects of fentanyl on peripheral and cerebral hemodynamics in neonatal lambs. Anesthesiology 66: 524–530.
- 29. Smith RP, Miller SL, Igosheve N, Peebles DM, Glover V, et al. (2004) Cardiovascular and endocrine responses to cutaneous electrical stimulation after fentanyl in the ovine fetus. Am J Obstet Gynecol 190: 836–842.
- 30. Norman E, Wikstr”m S, Ros‚n I, Fellman V, Hellstr”m-Westas L (2013) Premedication for intubation with morphine causes prolonged depression of electrocortical background activity in preterm infants. Pediatric Research 73: 87–94.
- 31. Scott JC, Ponganis KV, Stanski DR (1985) EEG quantitation of narcotic effects: the comparative pharmacodynamics of fentanyl and alfentanil. Anesthesiology 62: 234–241.
- 32. da Silva O, Alexandrou D, Knoppert D, Young GB (1999) Seizure and electroencephalographic change in the newborn period induced by opiates and corrected by naloxone infusion. Journal of Perinatology 19: 120–123.
- 33. Bernet V, Latal B, Natalucci G, Doel C, Ziegler A, et al. (2010) Effect of sedation and analgesia on postoperative amplitude-integrated EEG in newborn cardiac patients. Pediatric Research 67: 650–655.
- 34. Kuroha M, Son DS, Shimoda M (2001) Effects of altered plasma alpha-1-acid glycoprotein levels on pharmacokinetics of some basic antibiotics in pigs: simulation analysis. J Vet Pharmacol Ther 24: 423–431.
- 35. Calvo R, Lukas JC, Rodriguez M, Leal N, Suarez E (2006) The role of unbound drug in pharmacokinetics/pharmacodynamics and in therapy. Curr Pharm Des 12: 977–987.
- 36. Itoh H, Tamura K, Izumi M, Motoi Y, Kidoguchi K, et al. (1993) The influence of age and health status on the serum alpha 1-acid glycoprotein level of conventional and specific pathogen-free pigs. Can J Vet Res 57: 74–78.