- Email: [email protected]
Changes in the pharmacodynamic response to fentanyl in neonates during continuous infusion
PEDIATRIC PHARMACOLOGY AND THERAPEUTICS
Changes in the pharmacodynamic response to fentanyl in neonates during continuous infusion John H, Arnold, MD, Robert D. Truog, MD, Joseph M. Scavone, PharmD, a n d Terence Fenton, EED From the Department of Anesthesia, Children's Hospital and Harvard Medical School, the Department of Biostatistics, Harvard School of Public Health, and the Division of Ctinical Pharmacology, Tufts-New England Medical Center, Boston, Massachusetts Tolerance to opioid-induced sedation has been reported in neonates sedated with fentanyl by continuous infusion while undergoing extracorporeal membrane oxygenation. We undertook a prospective analysis of eight newborn infants sedated with fentanyl during extracorporeal membrane oxygenation therapy for respiratory failure. We recorded daily mean fentanyl infusion rate, measured serial plasma fentanyl concentrations, and documented the occurrence of spontaneous motor activity or respiratory effort. Mean fentanyl infusion rate increased steadily during the period of infusion from a mean of 9.2 • 1.9 (SEM)/~g/kg per hour on day I to a mean of 21.9 +_ 4.5/~g/kg per hour by day 6. Mean plasma fentanyl concentrations increased steadily during the period of fentanyl infusion from 3.1 • 1.1 ($EM) ng/ml on day I to 13.9 _+ 3.2 ng/ml on day 6. All infants exhibited movement in response to gentle tactile stimulation throughout the period of infusion, and seven of eight infants manifested spontaneous movement of the extremities and eye opening. Our results indicate that when fentanyl is used for sedation in neonates, the plasma concentrations required for satisfactory sedation steadily escalate, possibly indicating the rapid development of tolerance to the sedating effects of fentanyl. (J PEDIATR1991; 119:639-43)
Fentanyl is a synthetic opioid that is widely used in neonatal anesthesia and intensive care; it is now the primary anesthetic for both pediatric and adult cardiac surgery. The advantages of fentanyl in the neonatal population include minimal hemodynamic effects from doses of up to 25 ~g/ kg 1 and ablation of pulmonary vascular responsiveness, 2 which may be advantageous for infants with pulmonary hypertension. There is also evidence that fentanyl by continuous infusion may improve survival in newborn infants Presented in part at the annual meeting of the American Society of Anesthesiologists, Las Vegas, Nev., October 1990. Submitted for publication Jan. 2, 1991; accepted April 1, 1991. Reprint requests: John H. Arnold, MD, Multidisciplinary Intensive Care Unit, Farley 5, Children's Hospital, 300 Longwood Ave., Boston, MA 021 l 5. 9/25/29905
with pulmonary hypertension associated with congenital diaphragmatic hernia) We previously reported our experience with the use of fentanyl by continuous infusion for neonates undergoing extracorporeal membrane oxygenation, 4 including the finding that these patients require steadily increasing fentanyl infusion rates to produce a consistent level of sedation. To See related articles, pp. 588, 644, and 649.
[
ECMO
Extracorporeal membrane oxygenation
]
account for alterations in fentanyl pharmacokinetics produced by extracorporeal circulation s, 6 and to elucidate the time course over which pharmacodynamic changes might occur, we prospectively studied a series of infants sedated
639
640
A r n o l d et al.
The Journal o f Pediatrics October 1991
I. Demographic data
Table
Patient No.
Birth weight (gm)
Diagnosis
Age at entry (days)
ECMO duration (days)
1 2 3 4
MAS MAS HMD HMD
3150 3840 3090 2800
1 i 3 1
3,5 4.4 5.8 12.3
5
MAS, GBS sepsis
3000
1
3.5
6 7 8
Blood aspiration Persistentpulmonary hypertension MAS
4200 2000 3160
1 14 5
3.25 5.5 9.3
Comments Uneventful course Uneventful course Uneventful course Day 4: ECMO circuit malfunction produced severe hemolysis, renal failure Day 3: grade 1V intraventricular hemorrhage, ECMO discontinued Uneventful course Uneventful course Uneventful course
GBS, Group B streptococcus;HMD, hyalinemembranedisease; MAS, meconiumaspirationsyndrome. Table
II. Daily lorazepam dose
Patient No.
Dosage (mg/kg per day) Dayl
Day2
Day3
Day4
Day5
Day6
1
0
0
0
0
--
--
2 3 4 5 6 7 8
0.! 0.6 0 0 0.1 0.6 0.5
0 0.9 0 0 0.2 0.4 0.5
0.1 1.0 0 0.2 0.4 0.4 0.4
0.2 1.2 0 0.2 0.4 0.4 0.6
0 1.2 0.3 --0.6 1.0
-1.1 0.3 --0.6 1.0
Dash indicatespatientnot receivingECMO.
with fentanyl by continuous infusion while undergoing ECMO. We hypothesized that there may be significant alterations in the pharmacodynamic response to fentanyl during the first few days of life in infants sedated by continuous infusion. METHODS
Eight neonates requiring ECMO for respiratory failure between July 1989 and March 1990 were enrolled. Patients with congenital diaphragmatic hernia were excluded. Informed consent was obtained in accordance with guidelines established by The Children's Hospital Committee on Clinical Investigation. Our sedation regimen for' infants undergoing ECMO has been described. 4 Briefly, all patients received an initial dose of fentanyl ( 10 to 20 #g/kg) during surgical placement of vascular cannulas, followed by a continuous fentanyl infusion. The infusion rate was adjusted by the patient's physician to render the neonates sedated but arousable. To avoid inadvertent deeannulation, the level of sedation was continuously observed by a nurse at the bedside and excessive movement was immediately treated with an additional dose of fentanyl (2 to 10 #g/kg) and an
increase in the infusion rate (usually 20%). Adequate sedation was indicated by minimal movement and/or apparent discomfort in response to routine patient care, including endotracheal suctioning every 3 hours. Benzodiazepines were frequently added for their synergistic sedative effect] However, benzodiazepines were not the primary mode of sedation and dosages were not adjusted in a consistent fashion in response to changes in level of sedation. To permit periodic neurologic examination, these neonates did not receive muscle relaxants. We examined patient records to determine the mean fentanyl infusion rate (total fentanyl received per 24-hour period expressed as micrograms per kilogram per hour) and dosage of lorazepam (total lorazepam received per 24-hour period expressed as milligrams per kilogram) for each day of ECMO. In addition, episodes of spontaneous motor activity or respiratory effort were recorded. Serial plasma fentanyl and midazolam concentrations were measured in all infants. Heparinized samples were collected in glass tubes and protected from contact with rubber or plasticizer. All samples were obtained shortly after cannulation and then every 24 hours for the duration of ECMO therapy. Samples were aspirated from the venous side of the ECMO circuit at least 60 minutes after a supplemental dose or change in fentanyl infusion rate. Plasma was separated by immediate centrifugation, and the samples were stored at - 2 0 ~ C until analysis. Plasma fentanyl concentrations were determined by gas chromatography and nitrogen-phosphorus detection. Plasma midazolam concentrations were determined by gas chromatography with electron capture detection. Plasma concentrations of fentanyl were linearly related from 0.5 to I00 ng/ml and for midazolam from 1.0 to 200 ng/ml to the peak-height ratio of fentanyl or midazolam versus the internal standard (alfentanfl). The sensitivity limit for fentanyl was approximately 0.5 ng/ml and 1.0 ng/ml for midazolam. The coef-
Volume 119 Number 4
Table
Pharmacodynamic response to fentanyl
64 1
Plasma midazolam concentration
Iii.
Concentration (ng/ml) Patient No.
Day I
Day 2
Day 3
Day 4
Day 5
Day 6
1 2 3 4 5 6 7 8
63,3 * 0 * 41.2 0 0 5.8
99.8 1.7 40.6 153.8 * 85.3 80.1 7.9
184.5 36.8 178.1 252.9 160.9 143.8 * 1.2
193.8 86.7 230.6 280.6 * 384.9 95.3 49.5
-* 232,6 202.0 --176.8 *
---* 333.2 --156.7 26.7
Dash indicates patient r~ot receiving ECMO. *Level not drawn for that day.
IV. Plasma fentanyl concentration and mean fentanyl infusion rate, by E C M O day
Table
Patient No. 1
2 3 4 5 6 7 8
Concentration (ng/ml) Day 1
Day 2
Day 3
Day 4
4.8 * 2.8 * 0 0 3.6 7.1
8.5 2.6 5.1 2.0 * 3.5 3.7 10.l
6,7 3.2 6.1 4.7 6.4 1.1 * 11.2
11.4 5.2 6.4 19.5 * 1.2 8.3 10.9
Mean infusion rate (/~g/kg per hr)
Day 5 --
* 5.8 12.9 --8.7 *
Day 6 --
-* 14.1 --8.2 19.4
Day I
Day2
Day 3
Day 4
Day 5
6.0 7.1 15.3 4.2 8.1 3.I 10.6 18.9
7.9 6.5 10.2 3.6 6.7 4.4 15.0 21.4
10.9 6.9 11,0 5.2 10.1 4.8 15.7 20.2
12.0 7.5 10.8 7.7 12.1 6.3 16.9 21.0
10.0 I 1.5 17.6 --32.9 21.3
--
Day 6 --
-!4.3 18.6 --35.0 19.8
Dash indicated patient not receiving ECMO. *Level not drawn for that day.
ficient of variation for replicate samples at concentrations ranging from 1 to 50 n g / m l of fentanyl and 5 to 100 n g / m l of midazolam d i d not exceed 7%. Statistical methods included an analysis of variance model assessing day-to-day variations in mean infusion rates and plasma fentanyl concentrations, with patient identity entered as a random effect to adjust for t h e correlation resulting from rePeated measures on the same patient, The SAS statistical package was used for this analysis. Post hoc paired comparisons with Bonferroni corrections for multiple t tests were made by comparing daily fentanyl infusion rates and plasma fentanyl concentrations. RESULTS The study group included two boys and six girls with a mean birth weight of 3160 _+_ 620 (SD) gin. The average duration of E C M O therapy was 5.9 + 3.0 (SD) days. Demographic data and notable events occurring during the period of study are recorded in Table I. Lorazepam was administered routinely (0.1 to 0.2 m g / k g every 6 to 8 hours) to seven infants, and the mean daily dosage was steadily increased (Table II). Midazolam was occasionally administered in individual doses together with fentanyl in response
to inadequate sedation. More commonly, however, midazolam was given prophylactically before endotracheal suctioning, to minimize agitation. Mean daily plasma midazolam concentrations are presented in Table III; a steady increase in plasma concentration occurred from days 1 through 5 and a small decrease from day 5 to day 6. The mean fentanyl infusion rate increased steadily during the period of infusion from 9.2 _+ 1.9 ( S E M ) ~zg/kg per hour on day 1 to 21.9 + 4.5 u g / k g per hour by day 6. Analysis of variance revealed a highly significant effect of day of infusion on mean infusion rate (p = 0.0024). Post hoc testing revealed that the infusion rates on days 5 and 6 were significantly higher than the infusion rates measured on days 1, 2, and 3 (p < 0.05). Mean plasma fentanyl concentrations increased steadily during the per!od of fentanyl infusion, from 3.1 _+ 1.1 ( S E M ) n g / m l on day t to 13.9 ___ 3.2 n g / m l on day 6. Analysis of variance revealed a highly significant effect of day of infusion on plasma fentanyl concentration (p = 0.007). Post hoc testing revealed that by day 4, measured fentanyl concentrations were significantly higher than those on day 1 (p < 0.05). The plasma concentrations on day 6 were significantly higher than the plasma concentra-
642
Arnold et al.
tions measured on days 1, 2, and 3. Individual data for plasma fentanyl concentration and daily mean infusion rate are presented in Table IV. All infants were sedated to a common endpoint: asleep but arousable and exhibiting minimal movement and/or discomfort in response to routine patient care. Review of patient records revealed further information regarding level of sedation. All infants exhibited movement in response to gentle tactile stimulation throughout the period of infusion, and seven of eight infants had spontaneous movement of the extremities and eye opening. Six of eight infants were described as "awake," "irritable," or "agitated" during fentanyl infusion despite attempts to attain more profound levels of sedation. One patient was noted to breathe spontaneously during the period of infusion despite plasma fentanyl levels between 2.8 and 6.4 ng/ml (patient 3). DISCUSSION Fentanyl, a synthetic opioid, is widely used in pediatric and neonatal anesthesia and intensive care because of its wide margin of safety1, s, 9 and potent inhibition of the stress responsefl' 10 Our data indicate that significant changes in the concentration of fentanyl are required to produce sedation in neonates receiving it by continuous infusion. The pharmacodynamics of fentanyl in the neonatal population have not been well delineated. Collins et al. 11 reported that when used as the sole anesthetic in neonates undergoing ligation of the ductus arteriosus, fentanyl provided satisfactory anesthesia at plasma concentrations between 7.7 and 13.6 ng/ml (mean 10.6 ng/ml)) 1 By comparison, we found that after 6 days of continuous infusion, newborn infants required a mean fentanyl plasma concentration of 13.9 ng/ml just to achieve moderate levels of sedation. Furthermore, all our patients required supplementation with benzodiazepines to sustain this level of sedation. Gregory et al. 12 demonstrated that fetal lambs have significantly lower anesthetic requirements than do adult ewes; rising anesthetic requirements in the first 12 hours of life were inversely correlated with declining progesterone concentrations. These studies were carried out with halothane as an anesthetic; unfortunately, no comparable data are available for the opioids. Perinatal changes in progesterone concentrations may partially explain the pharmacodynamic changes we observed in our patients during the first week of life. There is also evidence that B-endorphin-like immunoreactivity is detectable at much higher concentrations in human neonates than in adults and that these levels rapidly decrease to the adult range by 4 days of life. 13 Endogenous opioid concentrations could have a potent effect on the requirement for exogenous opioid administration in the perinatal period. Given the increasing daily lorazepam dosages and steady
The Journal of Pediatrics October 1991
increases in plasma midazolam concentrations during the period of study, it is unlikely that the increasing fentanyl requirement we observed was due to inadvertent benzodiazepine withdrawal. It would be imprudent to speculate on the pharmacodynamics of the benzodiazepines in these patients as benzodiazepines were not the primary mode of sedation and their dosage was not consistently increased in response to changes in the level of sedation. Several aspects of our study design must be acknowledged to interpret these pharmacodynamic findings properly. First, the assessment of adequacy of sedation in critically ill newborn infants is a difficult process complicated by subjective impression, interobserver variability, and quantitative methods not easily adapted to clinical practice. 14 A recent review of sedation and analgesia in intensive care details a number of scoring systems that have been utilized in the treatment of adults, is However, neither internal validity nor interobserver reliability have been established for these systems and there is no published experience with the use of such scores in neonates. Therefore our prospectively defined clinical endpoint (i.e., sedated but arousable) and assessment of the response to a reproducible stimulus (endotracheal suctioning) every 3 hours are probably as reliable as current methods allow. Second, alterations in fentanyl pharmacodynamics may have been related to changes in the clinical status of the infants during ECMO. Two important clinical factors known to affect anesthetic requirements are hypotension I6 and hypoxemia. 17 Criteria for the initiation of ECMO in the management of neonatal respiratory failure in our institution include significant hypoxemia, often accompanied by hypotension.18 However, after stabilization of our patients' clinical status with ECMO, blood pressure and arterial oxygenation were under rigorous control. Therefore hypotension and hypoxemia could not have contributed to the steady increase in anesthetic requirements in our patients throughout the period of ECMO therapy. Nevertheless, it is possible that other undetected alterations in their clinical status may have played a role in the pharmacodynamie changes observed. Third, there is evidence in animals that continuous opioid administration produces tolerance much more rapidly than intermittent administration,19, zo presumably by uninterrupted agonist activity at the opioid receptor. 2z, 22 Therefore continuous opioid administration by infusion may well promote the rapid development of tolerance. The only other reports of tolerance to fentanyl in human beings have been demonstrated in patients receiving fentanyl by infusion. 23,z4 Therefore in our patients continuous opioid administration may well have altered the pharmacodynamie response by inducing the rapid development of tolerance~
Volume 119 Number 4
In summary, we have demonstrated that when fentanyl is used for sedation in neonates undergoing E C M O , the plasma concentrations required for satisfactory sedation rapidly escalate. The observed increase in fentanyl requirement may reflect a pharmacodynamic transition from fetal to postnatal life. Alternatively, continuous administration by infusion may well have promoted the rapid development of tolerance. Nonetheless, rapid increases in the plasma concentration of fentanyl required to produce adequate sedation may limit the usefulness of fentanyl as a sedative for intensive care of the neonate.
REFERENCES 1. Hickey PR, Hansen DD, Wessel DL, et al. Pulmonary and systemic hemodynamic responses to fentanyl in infants. Anesth Analg 1985;64:483-6. 2. Hickey PR, Hansen DD, Wessel DL, et al. Blunting of stress responses in the pulmonary circulation of infants by fentanyl. Anesth Analg 1985;64:1137-42. 3. Vacanti JP, Crone RK, Murphy JD, et al. The pulmonary hemodynamic response to perioperative anesthesia in the treatment of high-risk infants with congenital diaphragmatic hernia. J Pediatr Surg 1984;19:672-9. 4. Arnold JH, Truog RD, Orav E J, et al. Tolerance and dependence in neonates sedated with fentanyl during extracorporeal membrane oxygenation. Anesthesiology 1990;73:1 t 36-40. 5. Holley FO, Ponganis KV, Stanski DR. Effect of cardiopulmonary bypass on the pharmacokinetics of drugs. Clin Pharmacokinet 1982;7:234-5 I. 6. Rosen D, Rosen K, Davidson B, et al. Fentanyl uptake by the Scimed membrane oxygenator. J Cardiothorae Anesth 1988; 2:619-26. 7. Vinik HR, Bradley EL, Kissin 1. Midazolam-alfentanil synergism for anesthetic induction in patients. Anesth Analg 1989; 69:213-7. 8. Bovill JG, Sebel PS, Stanley TH. Opioid analgesics in anesthesia: with special reference to their use in cardiovascular anesthesia. Anesthesiology 1984;61:731-55. 9. Janssen P. The past, present, and future of opioid anesthetics. J Cardiothorac Anesth 1990;4:259-65.
Pharmacodynamic response to fentanyl
643
10. Anand KJ, Sippell WG, Aynsley Green A. Randomised trial of fentanyl anaesthesia in preterm babies undergoing surgery: effects on the stress response. Lancet 1987;1:62-6. 11. Collins C, Koren G, Crean P, et al. Fentanyl pharmacokinetics and hemodynamic effects in preterm infants during ligation of patent ductus arteriosus. Anesth Analg 1985;64:1078-80. 12. Gregory GA, Wade JG, Beihl DR, et al. Fetal anesthetic requirement (MAC) for halothane. Anesth Analg 1983;62:9-14. 13. Moss IR, Conner H, Yee WF, et al. Human beta-endorphinlike immunoreactivity in the perinatal/neonatal period. J PEDIATR 1982;101:443-6. 14. Marshall RE. Neonatal pain associated with caregiving procedures. Pediatr Clin North Am 1989;36:885-903. 15. Hamilton-Farrell MR, Hanson GC. General care of the ventilated patient in the intensive care unit. Thorax 1990;45: 962-9. 16. Tanifuji Y, Eger E. Effect of arterial hypotension on anaesthetic requirement in dogs. Br J Anaesth 1976;48:947-52. 17. Cullen D J, Eger E. The effects of hypoxia and isovolemic anemia on the halothane requirement (MAC) of dogs. I. The effect of hypoxia. Anesthesiology 1970;32:28-34. 18. ORourke PP, Crone RK, Vacanti JP, et al. Extracorporeal membrane oxygenation and conventional medical therapy in neonates with persistent pulmonary hypertension of the newborn: a prospective randomized study. Pediatrics 1989;84:95763. 19. Cheney DL, Goldstein A. Tolerance to opioid narcotics: time course and reversibility of physical dependence in mice. Nature I971 ;232:477-8. 20. Dewey WL. Various factors which affect the rate of development of tolerance and physical dependence to abused drugs. NIDA Res Monogr 1984;54:39-49. 21. Hovav E, Weinstock M. Temporal factors influencing the development of acute tolerance to opiates. J Pharmacol Exp Ther 1987;242:251-6. 22. Cochin J, Mushlin BE. Effect of agonist-antagonist interaction on the development of tolerance and dependence. Ann N Y Acad Sci 1976;281:244-51. 23. Sharer A, White PF, Schuttler J, et al. Use of a fentanyl infusion in the intensive care unit: tolerance to its anesthetic effects. Anesthesiology 1983;59:245-8. 24. McQuay H J, Bullingham RE, Moore RA. Acute opiate tolerance in man. Life Sci 1981;28:2513-7.
Changes in the pharmacodynamic response to fentanyl in neonates during continuous infusion John H, Arnold, MD, Robert D. Truog, MD, Joseph M. Scavone, PharmD, a n d Terence Fenton, EED From the Department of Anesthesia, Children's Hospital and Harvard Medical School, the Department of Biostatistics, Harvard School of Public Health, and the Division of Ctinical Pharmacology, Tufts-New England Medical Center, Boston, Massachusetts Tolerance to opioid-induced sedation has been reported in neonates sedated with fentanyl by continuous infusion while undergoing extracorporeal membrane oxygenation. We undertook a prospective analysis of eight newborn infants sedated with fentanyl during extracorporeal membrane oxygenation therapy for respiratory failure. We recorded daily mean fentanyl infusion rate, measured serial plasma fentanyl concentrations, and documented the occurrence of spontaneous motor activity or respiratory effort. Mean fentanyl infusion rate increased steadily during the period of infusion from a mean of 9.2 • 1.9 (SEM)/~g/kg per hour on day I to a mean of 21.9 +_ 4.5/~g/kg per hour by day 6. Mean plasma fentanyl concentrations increased steadily during the period of fentanyl infusion from 3.1 • 1.1 ($EM) ng/ml on day I to 13.9 _+ 3.2 ng/ml on day 6. All infants exhibited movement in response to gentle tactile stimulation throughout the period of infusion, and seven of eight infants manifested spontaneous movement of the extremities and eye opening. Our results indicate that when fentanyl is used for sedation in neonates, the plasma concentrations required for satisfactory sedation steadily escalate, possibly indicating the rapid development of tolerance to the sedating effects of fentanyl. (J PEDIATR1991; 119:639-43)
Fentanyl is a synthetic opioid that is widely used in neonatal anesthesia and intensive care; it is now the primary anesthetic for both pediatric and adult cardiac surgery. The advantages of fentanyl in the neonatal population include minimal hemodynamic effects from doses of up to 25 ~g/ kg 1 and ablation of pulmonary vascular responsiveness, 2 which may be advantageous for infants with pulmonary hypertension. There is also evidence that fentanyl by continuous infusion may improve survival in newborn infants Presented in part at the annual meeting of the American Society of Anesthesiologists, Las Vegas, Nev., October 1990. Submitted for publication Jan. 2, 1991; accepted April 1, 1991. Reprint requests: John H. Arnold, MD, Multidisciplinary Intensive Care Unit, Farley 5, Children's Hospital, 300 Longwood Ave., Boston, MA 021 l 5. 9/25/29905
with pulmonary hypertension associated with congenital diaphragmatic hernia) We previously reported our experience with the use of fentanyl by continuous infusion for neonates undergoing extracorporeal membrane oxygenation, 4 including the finding that these patients require steadily increasing fentanyl infusion rates to produce a consistent level of sedation. To See related articles, pp. 588, 644, and 649.
[
ECMO
Extracorporeal membrane oxygenation
]
account for alterations in fentanyl pharmacokinetics produced by extracorporeal circulation s, 6 and to elucidate the time course over which pharmacodynamic changes might occur, we prospectively studied a series of infants sedated
639
640
A r n o l d et al.
The Journal o f Pediatrics October 1991
I. Demographic data
Table
Patient No.
Birth weight (gm)
Diagnosis
Age at entry (days)
ECMO duration (days)
1 2 3 4
MAS MAS HMD HMD
3150 3840 3090 2800
1 i 3 1
3,5 4.4 5.8 12.3
5
MAS, GBS sepsis
3000
1
3.5
6 7 8
Blood aspiration Persistentpulmonary hypertension MAS
4200 2000 3160
1 14 5
3.25 5.5 9.3
Comments Uneventful course Uneventful course Uneventful course Day 4: ECMO circuit malfunction produced severe hemolysis, renal failure Day 3: grade 1V intraventricular hemorrhage, ECMO discontinued Uneventful course Uneventful course Uneventful course
GBS, Group B streptococcus;HMD, hyalinemembranedisease; MAS, meconiumaspirationsyndrome. Table
II. Daily lorazepam dose
Patient No.
Dosage (mg/kg per day) Dayl
Day2
Day3
Day4
Day5
Day6
1
0
0
0
0
--
--
2 3 4 5 6 7 8
0.! 0.6 0 0 0.1 0.6 0.5
0 0.9 0 0 0.2 0.4 0.5
0.1 1.0 0 0.2 0.4 0.4 0.4
0.2 1.2 0 0.2 0.4 0.4 0.6
0 1.2 0.3 --0.6 1.0
-1.1 0.3 --0.6 1.0
Dash indicatespatientnot receivingECMO.
with fentanyl by continuous infusion while undergoing ECMO. We hypothesized that there may be significant alterations in the pharmacodynamic response to fentanyl during the first few days of life in infants sedated by continuous infusion. METHODS
Eight neonates requiring ECMO for respiratory failure between July 1989 and March 1990 were enrolled. Patients with congenital diaphragmatic hernia were excluded. Informed consent was obtained in accordance with guidelines established by The Children's Hospital Committee on Clinical Investigation. Our sedation regimen for' infants undergoing ECMO has been described. 4 Briefly, all patients received an initial dose of fentanyl ( 10 to 20 #g/kg) during surgical placement of vascular cannulas, followed by a continuous fentanyl infusion. The infusion rate was adjusted by the patient's physician to render the neonates sedated but arousable. To avoid inadvertent deeannulation, the level of sedation was continuously observed by a nurse at the bedside and excessive movement was immediately treated with an additional dose of fentanyl (2 to 10 #g/kg) and an
increase in the infusion rate (usually 20%). Adequate sedation was indicated by minimal movement and/or apparent discomfort in response to routine patient care, including endotracheal suctioning every 3 hours. Benzodiazepines were frequently added for their synergistic sedative effect] However, benzodiazepines were not the primary mode of sedation and dosages were not adjusted in a consistent fashion in response to changes in level of sedation. To permit periodic neurologic examination, these neonates did not receive muscle relaxants. We examined patient records to determine the mean fentanyl infusion rate (total fentanyl received per 24-hour period expressed as micrograms per kilogram per hour) and dosage of lorazepam (total lorazepam received per 24-hour period expressed as milligrams per kilogram) for each day of ECMO. In addition, episodes of spontaneous motor activity or respiratory effort were recorded. Serial plasma fentanyl and midazolam concentrations were measured in all infants. Heparinized samples were collected in glass tubes and protected from contact with rubber or plasticizer. All samples were obtained shortly after cannulation and then every 24 hours for the duration of ECMO therapy. Samples were aspirated from the venous side of the ECMO circuit at least 60 minutes after a supplemental dose or change in fentanyl infusion rate. Plasma was separated by immediate centrifugation, and the samples were stored at - 2 0 ~ C until analysis. Plasma fentanyl concentrations were determined by gas chromatography and nitrogen-phosphorus detection. Plasma midazolam concentrations were determined by gas chromatography with electron capture detection. Plasma concentrations of fentanyl were linearly related from 0.5 to I00 ng/ml and for midazolam from 1.0 to 200 ng/ml to the peak-height ratio of fentanyl or midazolam versus the internal standard (alfentanfl). The sensitivity limit for fentanyl was approximately 0.5 ng/ml and 1.0 ng/ml for midazolam. The coef-
Volume 119 Number 4
Table
Pharmacodynamic response to fentanyl
64 1
Plasma midazolam concentration
Iii.
Concentration (ng/ml) Patient No.
Day I
Day 2
Day 3
Day 4
Day 5
Day 6
1 2 3 4 5 6 7 8
63,3 * 0 * 41.2 0 0 5.8
99.8 1.7 40.6 153.8 * 85.3 80.1 7.9
184.5 36.8 178.1 252.9 160.9 143.8 * 1.2
193.8 86.7 230.6 280.6 * 384.9 95.3 49.5
-* 232,6 202.0 --176.8 *
---* 333.2 --156.7 26.7
Dash indicates patient r~ot receiving ECMO. *Level not drawn for that day.
IV. Plasma fentanyl concentration and mean fentanyl infusion rate, by E C M O day
Table
Patient No. 1
2 3 4 5 6 7 8
Concentration (ng/ml) Day 1
Day 2
Day 3
Day 4
4.8 * 2.8 * 0 0 3.6 7.1
8.5 2.6 5.1 2.0 * 3.5 3.7 10.l
6,7 3.2 6.1 4.7 6.4 1.1 * 11.2
11.4 5.2 6.4 19.5 * 1.2 8.3 10.9
Mean infusion rate (/~g/kg per hr)
Day 5 --
* 5.8 12.9 --8.7 *
Day 6 --
-* 14.1 --8.2 19.4
Day I
Day2
Day 3
Day 4
Day 5
6.0 7.1 15.3 4.2 8.1 3.I 10.6 18.9
7.9 6.5 10.2 3.6 6.7 4.4 15.0 21.4
10.9 6.9 11,0 5.2 10.1 4.8 15.7 20.2
12.0 7.5 10.8 7.7 12.1 6.3 16.9 21.0
10.0 I 1.5 17.6 --32.9 21.3
--
Day 6 --
-!4.3 18.6 --35.0 19.8
Dash indicated patient not receiving ECMO. *Level not drawn for that day.
ficient of variation for replicate samples at concentrations ranging from 1 to 50 n g / m l of fentanyl and 5 to 100 n g / m l of midazolam d i d not exceed 7%. Statistical methods included an analysis of variance model assessing day-to-day variations in mean infusion rates and plasma fentanyl concentrations, with patient identity entered as a random effect to adjust for t h e correlation resulting from rePeated measures on the same patient, The SAS statistical package was used for this analysis. Post hoc paired comparisons with Bonferroni corrections for multiple t tests were made by comparing daily fentanyl infusion rates and plasma fentanyl concentrations. RESULTS The study group included two boys and six girls with a mean birth weight of 3160 _+_ 620 (SD) gin. The average duration of E C M O therapy was 5.9 + 3.0 (SD) days. Demographic data and notable events occurring during the period of study are recorded in Table I. Lorazepam was administered routinely (0.1 to 0.2 m g / k g every 6 to 8 hours) to seven infants, and the mean daily dosage was steadily increased (Table II). Midazolam was occasionally administered in individual doses together with fentanyl in response
to inadequate sedation. More commonly, however, midazolam was given prophylactically before endotracheal suctioning, to minimize agitation. Mean daily plasma midazolam concentrations are presented in Table III; a steady increase in plasma concentration occurred from days 1 through 5 and a small decrease from day 5 to day 6. The mean fentanyl infusion rate increased steadily during the period of infusion from 9.2 _+ 1.9 ( S E M ) ~zg/kg per hour on day 1 to 21.9 + 4.5 u g / k g per hour by day 6. Analysis of variance revealed a highly significant effect of day of infusion on mean infusion rate (p = 0.0024). Post hoc testing revealed that the infusion rates on days 5 and 6 were significantly higher than the infusion rates measured on days 1, 2, and 3 (p < 0.05). Mean plasma fentanyl concentrations increased steadily during the per!od of fentanyl infusion, from 3.1 _+ 1.1 ( S E M ) n g / m l on day t to 13.9 ___ 3.2 n g / m l on day 6. Analysis of variance revealed a highly significant effect of day of infusion on plasma fentanyl concentration (p = 0.007). Post hoc testing revealed that by day 4, measured fentanyl concentrations were significantly higher than those on day 1 (p < 0.05). The plasma concentrations on day 6 were significantly higher than the plasma concentra-
642
Arnold et al.
tions measured on days 1, 2, and 3. Individual data for plasma fentanyl concentration and daily mean infusion rate are presented in Table IV. All infants were sedated to a common endpoint: asleep but arousable and exhibiting minimal movement and/or discomfort in response to routine patient care. Review of patient records revealed further information regarding level of sedation. All infants exhibited movement in response to gentle tactile stimulation throughout the period of infusion, and seven of eight infants had spontaneous movement of the extremities and eye opening. Six of eight infants were described as "awake," "irritable," or "agitated" during fentanyl infusion despite attempts to attain more profound levels of sedation. One patient was noted to breathe spontaneously during the period of infusion despite plasma fentanyl levels between 2.8 and 6.4 ng/ml (patient 3). DISCUSSION Fentanyl, a synthetic opioid, is widely used in pediatric and neonatal anesthesia and intensive care because of its wide margin of safety1, s, 9 and potent inhibition of the stress responsefl' 10 Our data indicate that significant changes in the concentration of fentanyl are required to produce sedation in neonates receiving it by continuous infusion. The pharmacodynamics of fentanyl in the neonatal population have not been well delineated. Collins et al. 11 reported that when used as the sole anesthetic in neonates undergoing ligation of the ductus arteriosus, fentanyl provided satisfactory anesthesia at plasma concentrations between 7.7 and 13.6 ng/ml (mean 10.6 ng/ml)) 1 By comparison, we found that after 6 days of continuous infusion, newborn infants required a mean fentanyl plasma concentration of 13.9 ng/ml just to achieve moderate levels of sedation. Furthermore, all our patients required supplementation with benzodiazepines to sustain this level of sedation. Gregory et al. 12 demonstrated that fetal lambs have significantly lower anesthetic requirements than do adult ewes; rising anesthetic requirements in the first 12 hours of life were inversely correlated with declining progesterone concentrations. These studies were carried out with halothane as an anesthetic; unfortunately, no comparable data are available for the opioids. Perinatal changes in progesterone concentrations may partially explain the pharmacodynamic changes we observed in our patients during the first week of life. There is also evidence that B-endorphin-like immunoreactivity is detectable at much higher concentrations in human neonates than in adults and that these levels rapidly decrease to the adult range by 4 days of life. 13 Endogenous opioid concentrations could have a potent effect on the requirement for exogenous opioid administration in the perinatal period. Given the increasing daily lorazepam dosages and steady
The Journal of Pediatrics October 1991
increases in plasma midazolam concentrations during the period of study, it is unlikely that the increasing fentanyl requirement we observed was due to inadvertent benzodiazepine withdrawal. It would be imprudent to speculate on the pharmacodynamics of the benzodiazepines in these patients as benzodiazepines were not the primary mode of sedation and their dosage was not consistently increased in response to changes in the level of sedation. Several aspects of our study design must be acknowledged to interpret these pharmacodynamic findings properly. First, the assessment of adequacy of sedation in critically ill newborn infants is a difficult process complicated by subjective impression, interobserver variability, and quantitative methods not easily adapted to clinical practice. 14 A recent review of sedation and analgesia in intensive care details a number of scoring systems that have been utilized in the treatment of adults, is However, neither internal validity nor interobserver reliability have been established for these systems and there is no published experience with the use of such scores in neonates. Therefore our prospectively defined clinical endpoint (i.e., sedated but arousable) and assessment of the response to a reproducible stimulus (endotracheal suctioning) every 3 hours are probably as reliable as current methods allow. Second, alterations in fentanyl pharmacodynamics may have been related to changes in the clinical status of the infants during ECMO. Two important clinical factors known to affect anesthetic requirements are hypotension I6 and hypoxemia. 17 Criteria for the initiation of ECMO in the management of neonatal respiratory failure in our institution include significant hypoxemia, often accompanied by hypotension.18 However, after stabilization of our patients' clinical status with ECMO, blood pressure and arterial oxygenation were under rigorous control. Therefore hypotension and hypoxemia could not have contributed to the steady increase in anesthetic requirements in our patients throughout the period of ECMO therapy. Nevertheless, it is possible that other undetected alterations in their clinical status may have played a role in the pharmacodynamie changes observed. Third, there is evidence in animals that continuous opioid administration produces tolerance much more rapidly than intermittent administration,19, zo presumably by uninterrupted agonist activity at the opioid receptor. 2z, 22 Therefore continuous opioid administration by infusion may well promote the rapid development of tolerance. The only other reports of tolerance to fentanyl in human beings have been demonstrated in patients receiving fentanyl by infusion. 23,z4 Therefore in our patients continuous opioid administration may well have altered the pharmacodynamie response by inducing the rapid development of tolerance~
Volume 119 Number 4
In summary, we have demonstrated that when fentanyl is used for sedation in neonates undergoing E C M O , the plasma concentrations required for satisfactory sedation rapidly escalate. The observed increase in fentanyl requirement may reflect a pharmacodynamic transition from fetal to postnatal life. Alternatively, continuous administration by infusion may well have promoted the rapid development of tolerance. Nonetheless, rapid increases in the plasma concentration of fentanyl required to produce adequate sedation may limit the usefulness of fentanyl as a sedative for intensive care of the neonate.
REFERENCES 1. Hickey PR, Hansen DD, Wessel DL, et al. Pulmonary and systemic hemodynamic responses to fentanyl in infants. Anesth Analg 1985;64:483-6. 2. Hickey PR, Hansen DD, Wessel DL, et al. Blunting of stress responses in the pulmonary circulation of infants by fentanyl. Anesth Analg 1985;64:1137-42. 3. Vacanti JP, Crone RK, Murphy JD, et al. The pulmonary hemodynamic response to perioperative anesthesia in the treatment of high-risk infants with congenital diaphragmatic hernia. J Pediatr Surg 1984;19:672-9. 4. Arnold JH, Truog RD, Orav E J, et al. Tolerance and dependence in neonates sedated with fentanyl during extracorporeal membrane oxygenation. Anesthesiology 1990;73:1 t 36-40. 5. Holley FO, Ponganis KV, Stanski DR. Effect of cardiopulmonary bypass on the pharmacokinetics of drugs. Clin Pharmacokinet 1982;7:234-5 I. 6. Rosen D, Rosen K, Davidson B, et al. Fentanyl uptake by the Scimed membrane oxygenator. J Cardiothorae Anesth 1988; 2:619-26. 7. Vinik HR, Bradley EL, Kissin 1. Midazolam-alfentanil synergism for anesthetic induction in patients. Anesth Analg 1989; 69:213-7. 8. Bovill JG, Sebel PS, Stanley TH. Opioid analgesics in anesthesia: with special reference to their use in cardiovascular anesthesia. Anesthesiology 1984;61:731-55. 9. Janssen P. The past, present, and future of opioid anesthetics. J Cardiothorac Anesth 1990;4:259-65.
Pharmacodynamic response to fentanyl
643
10. Anand KJ, Sippell WG, Aynsley Green A. Randomised trial of fentanyl anaesthesia in preterm babies undergoing surgery: effects on the stress response. Lancet 1987;1:62-6. 11. Collins C, Koren G, Crean P, et al. Fentanyl pharmacokinetics and hemodynamic effects in preterm infants during ligation of patent ductus arteriosus. Anesth Analg 1985;64:1078-80. 12. Gregory GA, Wade JG, Beihl DR, et al. Fetal anesthetic requirement (MAC) for halothane. Anesth Analg 1983;62:9-14. 13. Moss IR, Conner H, Yee WF, et al. Human beta-endorphinlike immunoreactivity in the perinatal/neonatal period. J PEDIATR 1982;101:443-6. 14. Marshall RE. Neonatal pain associated with caregiving procedures. Pediatr Clin North Am 1989;36:885-903. 15. Hamilton-Farrell MR, Hanson GC. General care of the ventilated patient in the intensive care unit. Thorax 1990;45: 962-9. 16. Tanifuji Y, Eger E. Effect of arterial hypotension on anaesthetic requirement in dogs. Br J Anaesth 1976;48:947-52. 17. Cullen D J, Eger E. The effects of hypoxia and isovolemic anemia on the halothane requirement (MAC) of dogs. I. The effect of hypoxia. Anesthesiology 1970;32:28-34. 18. ORourke PP, Crone RK, Vacanti JP, et al. Extracorporeal membrane oxygenation and conventional medical therapy in neonates with persistent pulmonary hypertension of the newborn: a prospective randomized study. Pediatrics 1989;84:95763. 19. Cheney DL, Goldstein A. Tolerance to opioid narcotics: time course and reversibility of physical dependence in mice. Nature I971 ;232:477-8. 20. Dewey WL. Various factors which affect the rate of development of tolerance and physical dependence to abused drugs. NIDA Res Monogr 1984;54:39-49. 21. Hovav E, Weinstock M. Temporal factors influencing the development of acute tolerance to opiates. J Pharmacol Exp Ther 1987;242:251-6. 22. Cochin J, Mushlin BE. Effect of agonist-antagonist interaction on the development of tolerance and dependence. Ann N Y Acad Sci 1976;281:244-51. 23. Sharer A, White PF, Schuttler J, et al. Use of a fentanyl infusion in the intensive care unit: tolerance to its anesthetic effects. Anesthesiology 1983;59:245-8. 24. McQuay H J, Bullingham RE, Moore RA. Acute opiate tolerance in man. Life Sci 1981;28:2513-7.