Alfentanil Infusions on the Intensive Therapy Unit

Intensive Care Med (1988) 14:55-59 Intensive Care Medicine © Springer-Verlag 1988

Alfentanil infusions on the intensive therapy unit

M.E. Sinclair ~, J.W. Sear 2, R.J. Summerfield 2 and A. Fisher 2

1Surgical Intensive Care Unit, Department of Anesthesiology, UniversityHospital, Gen~ve, Switzerland, and 2Intensive Therapy Unit, Nuffield Department of Anaesthetics, John Radcliffe Hospital, Headington, Oxford, UK

Received and accepted: 15 February 1987

Abstract. We have investigated the use of alfentanil by benzodiazepines were the first line drugs for use dur- infusion to sedate 14 patients during controlled ven- ing IPPV [2]. A short acting benzodiazepine, such as tilation on the intensive therapy unit (ITU). An initial midazolam, would appear ideal for ITU use, except rate of 24 Ixg"kg- 1. h- 1 was chosen and altered there- for two important drawbacks. Firstly, it does not pro- after according to patient response. Incremental doses vide analgesia and secondly, delayed recovery has been of midazolam (2.5-5.0 rag) were given intravenously reported following its use in ITU patients [3]. Two of (i.v.) if indicated. In 4 patients, the use of a muscle the pure hypnotics that have also been used in the ITU, relaxant was necessary to allow adequate controlled etomidate and Althesin [4, 5], have been recently ventilation of the patient. The mean duration of infu- withdrawn. A more recent survey of ITU's in the sion was 27.9 h (range 10-141 h), and the mean total United Kingdom indicates that the most frequently dose of alfentanil was 69.3 mg (12.5-240 mg). Spon- used sedation regimen has reverted to a combination taneous ventilation was rapidly achieved in 11 patients of opiate (usually morphine) plus a sedative drug dur- after stopping the infusion. The mean arterial carbon ing IPPV [6]. However, serious doubt regarding the dioxide tension (PaCO2)was 5.38kPa, 15-30min safety of morphine and its congeners has arisen after stopping the infusion. The clinical condition of following reports of an association between excretion 2 patients necessitated a change in sedation technique of these drugs and renal function, resulting in pro- and one patient died during the alfentanil infusion. longed recovery [7, 8]. Alfentanil by infusion caused no major cardiovascular Alfentanil, a synthetic opioid, has been shown to effects and did not influence the plasma cortisol have a short duration of action related to its short response to trauma. There was no major alteration in elimination half-life and a small volume of distribu- blood biochemistry or haematology during the infu- tion [9-11]. For these reasons, alfentanil has been sions of alfentanil. The plasma concentrations of assessed and used successfully by infusion to provide alfentanil during infusion showed a wide variability. analgesia during and after peripheral operative pro- These probably relate to both changes in the volume cedures [12, 13]. of distribution of the drug and in hepatic clearance. These studies have shown that recovery from the respiratory depressant effects of alfentanil is rapid and Key words: Intensive therapy unit - Analgesia - no adverse cardiovascular effects were evident. The Sedation - Alfentanil pharmacokinetics of alfentanil remain unchanged in patients with renal failure [14]. It shares the sedative properties of other opioid drugs and therefore, alfentanil appears to fulfill the criteria for the "ideal" agent Yate and colleagues have stated that the ideal agent for for sedation during IPPV, provided there is no sup- sedation during intermittent positive pressure ventila- pression of the adreno-cortical response to stress. tion (IPPV) would provide analgesia combined with Alfentanil has already been used as an infusion on good cardiovascular stability and rapid recovery from the ITU to provide sedation and analgesia in ven- any respiratory depression [1]. A review of 34 intensive tilated patients following open-heart surgery [1]. The therapy units (ITU) in the United Kingdom by Mer- mean infusion rate used was 24 ~tg. kg-1. h-1 and this riman in 1981 showed that opioid analgesics and/or provided acceptable conditions in the patients studied. 56 M.E. Sinclair et al.: Alfentanil infusions on the ITU

We have, therefore, used this mean infusion rate as tral venous pressure measurements were routinely our starting point to study the effects of alfentanil in recorded, and the electrocardiogram (ECG) displayed patients where ventilation was anticipated for > 12 h on a Roche 108 trend facility monitor. on the ITU. We have assessed the efficacy and safety of alfentanil for sedation, as well as measuring the Blood samples neuroendocrine response to stress by plasma cortisol Arterial blood was sampled, when appropriate, during concentrations. In addition, plasma levels of alfentanil ventilation and at the following times after the infu- were measured in 10 of the 14 patients during the infusion of alfentanil had been stopped: immediately after sion period. the start of spontaneous ventilation (SV) and 2 and 6 h after. Samples were analysed for PaO2, PaCO 2 and Patients and methods pH using a Radiometer ABL 2 blood-gas analyser. Arterial blood was analysed daily for plasma elec- Permission to perform the study was obtained from trolytes, urea, creatinine liver function tests, the hospital ethical committee. 14 patients admitted to haematology and coagulation. In addition, arterial the ITU and requiring ventilation for > 12h were blood was sampled for analysis of plasma cortisol (by studied. Patients with pre-existing renal or hepatic dis- radio-immunoassay) prior to the study, at 20:00h ease, as determined by conventional laboratory plasma (where appropriate) and at 08:00 h daily and thereafter testing, were excluded. The severity of illness was mea- until the patient left the ITU. sured using the Apache II scoring system [15]. On ad- mission to the study, the patients were already in- Plasma alfentanil assay tubated and ventilated, but were not paralysed. Alfen- tanil was given as a single bolus dose of approximately Arterial blood samples were collected as for cortisol 50 ~tg/kg (2.5-3.5 mg) over 1-5rain depending on and stored at -20°C. The samples were taken at the the patients cardiovascular response. An infusion of following times after the start of the infusion 2, 3, 6, alfentanil (24 ~tg.kg- l.h- 1) was simultaneously com- 9, 12, 18 and 24 h. Thereafter samples were collected menced and adjusted in the following manner: if the at approximately 12 h intervals. Plasma concentra- patient was judged by the attending nurse to be too tions of alfentanil were estimated in duplicate using heavily sedated, the infusion rate was reduced; and if the radioimmunoassay of Michiels et al. [16]. poor conditions of sedation (patient restlessness, "fighting" the ventilator, sweating, increases in blood Statistical analys& of data pressure or pulse rate by > 20°7o that at the time of in- Throughout the text, mean values and range are stitution of ventilation) were present, the infusion rate shown unless otherwise indicated. Drug requirements was increased. In addition, if the attending nurse felt were compared between 4 h epochs, and where appro- that the above criteria were more due to awareness priate between days, by the Wilcoxon matched pairs than pain, incremental doses of midazolam signed ranks test. The same test was used to compare (2.5- 5.0 mg) were given intravenously. the blood gas changes following the cessation of the Mechanical ventilation was provided with either an infusion and controlled ventilation. Plasma cortisol Oxford Penlon or a Siemens Servovent 900 B ven- and alfentanil concentrations showed non-gaussian tilator and the ventilation adjusted to provide an distribution, and the results are therefore expressed as arterial carbon-dioxide tension (PaCO2) of geometric mean and I log SD range. 4.5-5.5 kPa. The inspired oxygen concentration was that needed to maintain clinically acceptable arterial oxygen tensions (PaO2). When clinically indicated, Results the infusion of alfentanil was stopped and the patient 14 patients were studied (11 male, 3 female). The mean allowed to breathe spontaneously. Apache II score was 14.4 (range 4-27). 8 patients had The time from stopping the infusion to the com- suffered multiple trauma, 4 were post-operative fol- mencement of spontaneous ventilation was noted. lowing major surgery, 1 had pneumonia and a sickle Once adequate spontaneous ventilation had com- cell crisis and 1 had carbon monoxide poisoning. The menced, as judged by tidal volume, minute volume mean patient age was 36.6 years (range 16-75) and and blood gas analysis, the patient was extubated. the mean weight 70.1 kg (50- 90 kg). The mean duration of infusion was 27.9 h (range 10-141 h) and the Monitoring total dose 69.3mg (range 12.5-240mg). Effective All patients had the arterial blood pressure con- control of ventilation was achieved in 11 patients. The tinuously monitored via an indwelling cannula. Cen- remaining 3 patients required a change of sedation; 2 M.E. Sinclair et al.: Alfentanil infusions on the ITU 57

Table 1. Dosage requirements over each 4 h period after the start of infusion (tug'h-1)

0-4 h a 5-8h 9-12h 13-16h 17-20h

Mean 1.28 1.67 1.62 1.60 1.80 Range 0.4 - 2.4 0.8-2.3 0.7-2.4 0.8-2.6 0.8-3.2 Mean midazolam requirements (range) 2.10 2.05 2.73 1.88 4.17 per 4 h (mg) (0-7.5) (o- 5) (o- 1o) (o- 5) (o- 1o) Number of patients 11 11 11 8 6

a This dose requirement excludes the initial bolus dose of alfentanil

received bolus doses of morphine and muscle relax- 2.04 mg-h J (SD+ 1.2 mg.h -1) and for the 2nd 24 h ants to facilitate controlled ventilation, and the third in the same patients, 2.43 mg-hr-1 (SD+ 1.6 mg was changed to diamorphine as terminal care was felt • h-I). The mean overall rate of infusion in all 7 pa- appropriate. tients was 1.71mg.h 1 (SD___I.2). Eleven patients left hospital; the 3 deaths (Apache After the alfentanil infusion, arterial blood gases II scores of 6, 16 and 22 respectively) being due to the were analysed in 9 patients (Table 3). There was no initial trauma (ruptured splenic pedicle and thoracic evidence of any residual respiratory depression, nor of aorta, and rupture of the interventricular septum; any renarcotisation leading to deterioration of mediastinitis following perforation of oesophageal respiratory function. In one patient (male aged 17 carcinoma during oesophageal intubation; and respi- years, with chest trauma following a road traffic acci- ratory failure following traumatic lung contusion). dent) re-intubation and further ventilation was neces- Table 1 shows the dose requirements of alfentanil sary due to fatigue and deterioration of cardiovascular in 11 patients in 4 h time periods for the first 20 h, the and respiratory parameters. Alfentanil was again used other 3 patients requiring the coincidental administra- for the control of ventilation. The patient was suc- tion of midazolam and muscle relaxants to allow ade- cessfully extubated after 20 h. quate control of ventilation. The requirements of The biochemical and haematological changes seen midazolam during the same period are also shown. during the infusion of alfentanil were unremarkable. There was no significant relationship between the in- Raised creatinine levels were observed in one patient, fusion rate of alfentanil, and midazolam require- who subsequently developed acute renal failure ments. Table 2 lists the dose requirements per 24 h for necessitating dialysis. Elevated urea concentrations those 7 patients receiving alfentanil by infusion for (greater than 6.7 mmol.1 -j) were seen in 11 patients. >24 h. Increased liver enzymes (alkaline phosphatase and There was no evidence of tolerance to alfentanil, as gamma glutamyl transpeptidase) were observed in 4 assessed by increasing opioid dose requirements, or an patients, one of whom developed hepatic dysfunction. increasing need for midazolam supplementation. The During the infusion, the concentration of cortisol mean infusion rate of alfentanil for the 1st 24 h was in the plasma ranged from 103-5100nmo1-1 -j (ex-

Table 2. Dose requirements for 24-h periods in 7 patients receiving alfentanil (mg) by infusion for more than 24 h

Patient Alfentanil Other drugs b

Bolus dose Day Day (rag) 1 2 3 4 5 6 1 2 3 4 5 6

1 3.5 1.06 1.87 Inf. Off - - - M25 M 17.5 - - 2 3.5 0.78 0.68 0.26 a 0.62 0.71 0.51 0 0 0 M20 M22.5 M5 3 2.5 1.35 1.30 1.47 1.61 0.78 Inf. not available 4 nil 4.17 5.23 -~ Off. M13.75P6 M23.5 - - 5 3.0 2.22 2.54 ) Sedation changed M85 A60 M90 A90 - - 6 nil 2.14 2.9l M22 - - - 7 (a) 3.0 1.51 Inf, Off. - - - - M10 - - - (b) 3.0 - - 1.64 Inf. Off. - - - - M5 -

a Spontaneous ventilation with alfentanil infusion continuing for analgesia, b Daily doses (mg) of: M = midazolam; P = pancuronium; A = alcuronium 58 M.E. Sinclair et al.: Alfentanil infusions on the ITU

Table 3. During ventilation and post-infusion/ventilation blood gas analysis, mean values (_+ SD); except pH, median (range) (n = 9)

During IPPV Immediately on + 2 h + 6 h resuming SV (0 h)

PaO2 (kPa)/FrO 2 39.4 36.5 37.1 35.9 (10.3) (9.6) (7.9) (3.8) PaCO 2 (kPa) 3.95 5.38 4.90 4.69 (0.6) (1.4) (0.9) (0.5) pH 7.48 7.40 7.40 7.44 (7.43 - 7.64) (7.19 - 7.48) (7.33 - 7.49) (7.38 - 7.46) PaO 2 (kPa) 16.8 18.7 21.2 13.9 (7.6) (14.4) (22.3) (5.4)

cluding one patient who received dexamethasone). rate. The geometric concentration per rag. hr-1 at 2, Mean plasma cortisol concentrations in any single pa- 3, 6, 9, 12, 18 and 24 h were 89.7, 82.3, 123.3, 178.6, tient did not fall below 219nmol.l-L However, low 194.3, 153.7 and 137.4 ng.m1-1 respectively. The esti- levels (< 200 nmol-1-~) were seen in 2 patients during mated clearance, based on plasma alfentanil levels their stay in ITU; in one of the patients this was at- sampled between 12 and 66 h after the start of the in- tributable to the use of etomidate for intubation. fusion (n = 7 patients), and determined as average in- Figure i shows cortisol levels at 20:00 h on the first day fusion rate divided by plasma drug concentration, of infusion, and at 08:00 h on days 2 and 3. Data was varied between 2.23 and 18.741. h- 1 (geometric mean: only obtained at 20:00 on the first day if the infusion 5.941.h-1; 1 log SD range 2.50-12.61). had been in progress for more than 4 h (n = 8). By the third day, there was no difference in cortisol values be- Discussion tween those in whom the infusion was continuing (n = 5) and those in whom it had stopped (n = 5). Of The results confirm that a mean alfentanil infusion the remaining two patients, one died and no sample rate of approximately 26 ~tg. kg- 1. h- 1 in conjunction was obtained for the other. with increments of midazolam will provide satisfacto- Plasma alfentanil levels were measured in 9 pa- ry sedation for ventilated patients. This is close to the tients during the infusion of the drug. In general, 20 ~tg-kg -1-h -1 used in 2 studies to provide post- changes in drug concentration paralleled changes in operative pain relief in spontaneously breathing pa- the infusion rate. There was however considerable tients [12, 13] and similar to the 24 ~tg.kg -1.h -1 in- variation in plasma concentrations - a 10 to 12 fold fusion rate used by Yate et al. [1]. Although the re- spread, existing for the same, dose-corrected, infusion quirement between patients showed considerable variability, there was no significant change in the rate 24OO- per 4 h period during the first 24 h of sedation. There 2200- was no relationship between the incremental doses of 2000- midazolam and the alfentanil infusion rate. This is '-- 1800- possibly due to the differing properties of the two 5 F 1600- drugs; alfentanil, being an opioid with some sedative 1400- properties, does not appear to relieve anxiety. Reports ~2oo- from former patients indicate that the use of sedative -..,= 1000- drugs alone is not always the most appropriate or 800- desirable way to relieve distress [16, 17]. After the infu- '~ 600- sion was stopped, recovery was rapid - despite total .400- doses in excess of 50 rag. Spontaneous ventilation was a._ 200- acceptable in all patients, as assessed by arterial blood O-J Pre- 12h 08:00 h 08:00h 08:00h gas analysis, 1/4-1/2 h after stopping IPPV. infusion after start (n=12) Stilt Stopped (n=6} infusion receiving receiving Alfentanil did not appear to influence the (n=8} alfentaniI alfentond neuroendocrine response to stress in our patients; (n:5) {n=5) although Moiler et al. have shown an intraoperative Doy I Day 2 Day 3 infusion of 3 ~tg.kg -1.rain -~ to delay the onset of Fig. 1. Plasma cortisol levels in patients sedated with alfentanil the increase in cortisol secretion in response to stress (Geometric mean and 1 Log. SD range) [19]. Then low cortisol levels recorded in two of our M.E. Sinclair et al.: Alfentanil infusions on the ITU 59

patients may be attributable to the use of etomidate 2. Merriman HM (1981) The techniques used to sedate ventilated and phenoperidine in one patient [4, 20]; but cannot patients. A survey of methods used in 34 intensive care units in be explained in the other. The wide range of cortisol Great Britain. Intensive Care Med 7:217 3. Byatt CM, Lewis LD, Dawling S, Cochrane GM (1984) Ac- levels at 08:00 h on the second day of alfentanil infu- cumulation of midazolam after repeated dosage in patients sion is comparable with those reported by Drucker receiving mechanical ventilation in an intensive care unit. Br and Shandling in acute medical admissions to the ITU Med J 289:799 [21]. 4. Wagner RL, White PF, Kan PB, Rosenthal MH, Feldman D It is interesting to note the wide range of plasma (1984) Inhibition of adrenal steroidogenesis by the anesthetic etomidate. N Engl J Med 310:1415 alfentanil levels, with increments of midazolam, need- 5. Editorial (1984) Sedation in the intensive care unit. Lancet 1:748 ed to ensure adequate conditions for IPPV 6. Gast PH, Fisher A, Sear JW (1984) Intensive care sedation now. (40-1063 ng.ml-l). Drug kinetics in ITU patients Lancet 2:863 are liable to considerable variability due to the multi- 7. Moore A, Sear J, Baldwin D, Allen M, Hunniset A, Bullingham MR, McQuay H (1984) Morphine kinetics during and after ple intercurrent therapy (e.g. vasoactive drugs, renal transplantation. Clin Pharmacol Ther 35:641 steroids, H 2 receptor antagonists and enzyme induc- 8. McQuay H, Moore A (1984) Be aware of renal function when ing drugs) and concurrent diseases (renal, hepatic and prescribing morphine. Lancet 2:284 pulmonary). Besides changes in intrinsic hepatic en- 9. Kay B, Pleuvry B (1980) Human volunteer studies of alfentanil zyme activity, ITU patients also suffer significant (R39209), a new short-acting analgesic. Anaesthesia 35:352 10. Kay B, Stephenson DK (1980) Alfentanil (R39209): initial ex- alterations in liver blood flow. Drug metabolism may perience with a new narcotic analgesic. Anaesthesia 35:1197 be affected by other factors - including the arterial 11. Bower S, Hull C (1982) Comparative pharmacokinetics of fen- oxygen tension. Cummings [22] has shown a signifi- tanyl and alfentanil. Br J Anaesth 54:87 cant relationship between the PaO: and antipyrine 12. Andrews CJH, Sinclair ME, Prys-Roberts C, Dye A (1983) Ven- clearance. The small volume of distribution of alfen- tilatory effects during and after continuous infusions of fen- tanyl of alfentanil. Br J Anaesth 55:211s tanil will be subject to alterations secondary to change 13. O'Connor M, Escarpa A, Prys-Roberts C (1983) Ventilatory in the blood volume (i.e. due to haemorrhage, fluid re- depression during and after infusion of alfentanil in man. Br J placement of vasoactive drug administration). All of Anaesth 55:217s these factors together may result in a decline in drug 14. Sear JW, Bower S, Potter D (1986) Disposition of alfentanil in patients with chronic renal failure (A.R.S.). Br J Anaesth 58:812 clearance, and elevation of plasma alfentanil concen- 15. Knaus WA, Draper EA, Draper DP (1984) Apache II: final trations. form and national validation results of a severity of disease In summary, alfentanil by infusion at a rate of classification system. Crit Care Med 12:975 about 26 ~tg-kg 1.h-1, supplemented where neces- 16. Michiels M, Hendriks R, Heykants J (1984) Radioimmunoassay sary by incremental doses of midazolam, provides of the new opiate analgesics alfentanil and sufentanil. Preliminary pharmacokinetic profile in man. J Pharm Phar- another technique available for sedation for ventilated macol 35:86 patients. Cardiovascular stability, satisfactory adrenal 17. Henschell EO (1977) The Guillain Barr6 syndrome. A personal cortical function and acceptance of IPPV were good experience. Anesthesiology 47:228 features but further studies for longer periods of time 18. Shovelton DS (1979) Reflections on an intensive therapy unit. Br Med J 1:737 are necessary to assess any undesirable side effects. 19. Moiler IW, Krantz T, Wandall E, Kehlet H (1985) Effect of alfentanil anaesthesia on the adrenocortical and hyperglycaemic Acknowledgements. We wish to acknowledge the help and coopera- response to abdominal surgery. Br J Anaesth 57:591 tion of the nursing and medical staff of the Intensive Care Unit, 20. Stoddart JC, Watson MJ (1983) Etomidate. Lancet 2:168 John Radcliffe Hospital, Oxford. Plasma alfentanil concentrations 21. Drucker D, Shandling M (1985) Variable adrenocortical func- were assayed by Dr. R. Woestenborghs, Janssen Pharmaceutica BV, tion in acute medical illness. Crit Care Med 13:477 Beerse, Belgium. We also would like to thank Mrs E. Agrebi who 22. Cummings JF (1976) The effect of arterial oxygen tension on typed the manuscript. antipyrine half life in plasma. Clin Pharmacol Ther 19:168

Dr. Mike E. Sinclair Nuffield Department of Anaesthetics References John Radcliffe Hospital Headington 1. Yate PM, Thomas D, Sebel PS (1984) Alfentanil infusion for Oxford 0)(3 9DU sedation and analgesia in intensive care. Lancet 2:396 UK