Br. J. clin. Pharmac. (1981), 12, 117-122
`
`SUBLINGUAL BUPRENORPHINE USED POSTOPERATIVELY:
`CLINICAL OBSERVATIONS AND PRELIMINARY
`PHARMACOKINETIC ANALYSIS
`
`R.E.S. BULLINGHAM, H.J. McQUAY, D. DWYER, M.C. ALLEN & R.A. MOORE
`Nuffield Departments of Anaesthetics and Clinical Biochemistry,
`Radcliffe Infirmary,
`Oxford OX2 6HE
`
`1
`Buprenorphine is a long-acting opiate analgesic. This study was designed to investigate the
`pharmacokinetics of this drug when given by the sublingual route to ten postoperative patients. Plasma
`levels of buprenorphine were measured by a specific radioimmunoassay.
`2 Plasma levels of the drug following sublingual administration of 0.4 mg showed an apparent delay in
`absorption and then rose slowly to reach low but significant levels by 3 h. There was considerable
`variation in the time at which peak levels were achieved. The average systemic availability of the drug by
`this route was estimated to be 30% by 3 h.
`3 Analgesic efficacy and duration of sublingual buprenorphine were assessed using demand analgesia.
`The analgesia was of about 9 h duration, similar to that achieved by parenteral administration of 0.3 mg
`of the drug to an equivalent group of patients. The sublingual dose caused a significant fall in the
`postoperatively elevated plasma glucose, and prevented any further rise in plasma cortisol.
`4 Reasons for the efficacy of the sublingual route are discussed and it is suggested that this route may
`be particularly appropriate for highly lipophilic drugs like buprenorphine.
`
`Introduction
`
`Potent opiate analgesics all show a considerable first
`pass effect, with large interindividual variation in its
`extent. The oral route for these drugs is consequently
`unpredictable; the route is therefore unsatisfactory in
`the management of postoperative pain. Sublingual
`doses of potent analgesics avoid initial hepatic pas-
`sage and are as convenient as the oral route for the
`patients and nursing staff. In addition, the high first
`pass effect then acts as a safety factor should the drug
`be swallowed.
`Buprenorphine is a synthetic opiate analgesic with
`mixed agonist-antagonist properties. It has been
`shown to be effective by the sublingual route in
`postoperative pain (Edge, Cooper & Morgan, 1979),
`and the suitability of the drug for this route is sug-
`gested by its high lipophilicity, high first pass effect,
`long duration of action and low addiction potential.
`This study was designed primarily to measure plas-
`ma levels of buprenorphine given sublingually. In
`addition the resulting analgesic effect together with
`the effects of the drug on ventilation and metabolism
`were studied.
`The design and method was the same as that used
`to establish the kinetics and effects of the drug when
`given by the intravenous and intramuscular routes
`0306-5251/81/080117-06 $01.00
`
`(McQuay et al., 1980), to permit comparison of drug
`kinetics and effects when given by the three routes.
`
`Methods
`
`Ethical committee approval was obtained to perform
`this study and signed consent to research forms were
`obtained from each patient.
`Ten patients undergoing elective total hip replace-
`ment at the Nuffield Orthopaedic Centre, Oxford,
`were selected sequentially from operation lists pro-
`vided that their age was between 45 and 75 years and
`their weight was less than 80 kg. Those with serious
`medical disorders or taking cardiovascular drugs
`other than thiazides were excluded. Preoperative full
`blood count and biochemical profile were normal in
`all patients.
`Patients underwent an anaesthetic technique
`which was identical to that reported in detail in a
`similar study utilizing parenteral buprenorphine
`(Bullingham et al., 1980). Patients were premedi-
`cated with 10 mg diazepam orally 2 h preoperatively.
`Thiopentone 4mg/kg and pancuronium 0.1 mg/kg
`
`© Macmillan Publishers Ltd 1981
`
`Page 1
`
`RB Ex. 2028
`BDSI v. RB PHARMACEUTICALS LTD
`IPR2014-00325
`
`

`

`118
`
`R.E.S. BULLINGHAM, HJ. McQUAY, D. DWYER, M.C. ALLEN & R.A. MOORE
`
`were followed by intubation and mechanical ventila-
`tion on the Bain circuit; the tidal volume was
`10 ml/kg and fresh gas flow 70 ml/kg using nitrous
`oxide and oxygen (2:1) and halothane 0.5%. Fluid
`administration was with 5 ml kg-1 h-I of Hartmann's
`solution for the first 2 h, and then with blood transfu-
`sion to a specific schedule depending on the
`preoperative haemoglobin and the blood loss. At the
`end of surgery the anaesthetic gases were discon-
`tinued and residual neuromuscular blockade was re-
`versed by atropine 1.2mg and neostigmine 2.5 mg
`given intravenously.
`Buprenorphine (0.3 mg) diluted to 10 ml with nor-
`mal saline was injected intravenously over 1 min
`following induction of anaesthesia. Blood samples
`for plasma buprenorphine levels following the in-
`travenous dose were taken at 0, 30, 150 and 180 min,
`and for blood gases and hormones at 30 and 150 min.
`Postoperatively, and 3 h after the first dose of
`buprenorphine, a further 0.4 mg (two 0.2 mg tablets)
`was given sublingually, provided that patients were
`able to open their eyes to command. Patients were
`instructed to keep the tablets under the tongue, and
`neither to chew nor swallow them. Blood samples for
`plasma buprenorphine levels were taken at 0, 5, 10,
`15, 20, 40, 60, 80,120,150 and 180min. Samplesfor
`blood gas and hormone assays were taken at 10, 60,
`120 and 180 min.
`Blood samples were taken through a radial artery
`cannula (0.53 mm i.d., Longdwell), which was used
`for intraoperative direct arterial pressure monitor-
`ing. The samples were collected into lithium heparin
`and potassium fluoride tubes; after centrifugation the
`plasma was separated and stored at -20OC until
`analysed. Buprenorphine was measured by radio-
`immunoassay, using the method of Bartlett et al.
`(1980), cortisol by the method of Beardweli, Burke
`& Cope (1968) and glucose by a standard glucose
`oxidase procedure. Blood gases were measured using
`a Radiometer ABL2 blood gas analysis system.
`
`Pharmacokinetic analysis
`
`1 The intraoperative plasma buprenorphine values
`of the ten patients were compared individually
`against the mean intravenous kinetics established
`previously in an equivalent study (Bullingham et al.,
`1980). All values were within two standard devia-
`tions from this mean curve. The mean decay curve for
`the present patient group was therefore assumed to
`be as previously established.
`2 The contribution of the initial (i.v.) buprenorphine
`dose to the measured plasma levels after the subling-
`ual dose was computed from the known mean termi-
`nal exponential decay rate (Bullingham et al., 1980)
`of the first dose, applied to the values obtained in that
`individual. By subtracting these computed values
`
`from the measured plasma levels after the sublingual
`dose, plasma levels were obtained which represent
`the contribution of the sublingual dose alone. This
`procedure assumes linear behaviour of the phar-
`macokinetic system. The excellent fit to a triexpo-
`nential model for parenteral doses provides justifica-
`tion for this (Bullingham et al., 1980).
`3 The individual stripped plasma buprenorphine
`values obtained in (2) were averaged. Area under the
`curve for the first 3 h after the sublingual dose
`(AUC_W) was estimated from these by numerical
`integration (Greville, 1969). The sublingual AUC,
`was then compared with an intravenous postopera-
`tive total AUCI, (zero to infinity). This was computed
`as the mean of the total AUCi,s obtained from a
`different but strictly comparable ten patients who
`had received 0.3mg intravenously as their post-
`operative dose (Bullingham et al., 1980); the AUCiV
`was multiplied by 4/3 to obtain dose equivalence.
`Systemic availability of the sublingual dose to 3 h was
`then calculated as AUC3d1.33AUCi,.
`Demand analgesia
`
`Six hours after the first dose of buprenorphine the
`patients were connected to a demand analgesia sys-
`tem via a separate 0.6 mm i.d. intravenous cannula
`(21 g Butterfly). The system had been explained to
`them the previous day. The apparatus used a mod-
`ified Mill Hill infusion pump (Muirhead Ltd, 34
`Croydon Rd, Beckenham, Kent), which delivered
`0.25 mg diamorphine each time that the patient pres-
`sed a button. This was the only analgesia received by
`the patients until they left the recovery room the
`following morning. The analgesic demands were re-
`corded automatically over this period.
`
`i) Duration ofanalgesia The automatic recordings of
`each patient's postoperative analgesic demands were
`analysed to provide quantification of the duration of
`action of buprenorphine. Time to the fifth demand
`made by each patient was used. The choice of the fifth
`demand was made because of the potentially exp-
`loratory nature of the initial presses, though a
`number of other criteria (time to 2, 3 or 4 demands
`and regression to zero demand axis at the time when
`demands became frequent) gave similar results.
`
`ii) Sex difference The difference in drug analgesic
`effect between men and women was shown by a
`comparison of regression slopes obtained from plots
`of the averaged cumulative demands for each 15 min
`period against time. The demands made between 60
`and 600 min were used for the analysis and values
`between 0 and 60 min were excluded because, in this
`initial period, the slopes were markedly different
`from the succeeding 9 h.
`
`Page 2
`
`

`

`Results
`
`The age, weight, sex distribution, duration of surgery
`and intraoperative Paco2 values are shown in Table
`1. The patient group showed no significant differ-
`ences for these variables from the previously de-
`scribed parenteral groups (McQuay et al., 1980). The
`intraoperative course was uneventful with no unex-
`pected changes in pulse and blood pressure.
`
`Table 1
`
`Patient data (n = 10)
`
`Age (years)
`Weight (kg)
`Duration of surgery (min)
`Intraoperative PacO2 (kPa)
`Blood loss (ml)
`Sex distribution
`
`Results are mean ± s.e.mean
`
`59.8± 3.7
`68.8±3.0
`99.9 ± 3.8
`4.90±0.18
`510±91.0
`6M:4F
`
`Following the sublingual dose, the onset of
`analgesia, estimated by the observer, occurred be-
`tween 15 and 45 min. There was less drowsiness after
`the sublingual dose than had been noted for paren-
`teral administration. One patient vomited 10mi
`after the sublingual dose was given, and one of the
`tablets was returned in the vomitus. This patient
`showed typical absorption of the sublingual bup-
`renorphine, and his results are included for all anal-
`yses.
`
`Plasma buprenorphine concentrations
`
`The mean measured values after the sublingual dose
`are shown in Figure 1, together with the mean values
`obtained from the parenteral study. It can be seen
`that sublingual administration of buprenorphine re-
`sults in a gradual rise in plasma level, producing
`values similar to those following intravenous and
`intramuscular administration at the 80 min sampling
`time. At the 150 and 180 min sampling times the
`plasma concentration of buprenorphine was signific-
`antly higher than following parenteral administration
`(0.3 mg).
`Table 2 contains the values obtained by averaging
`the stripped values computed as described in
`Methods, pharmacokinetic analysis, (2). There was
`large individual variation in the plasma buprenor-
`phine profile following sublingual administration in
`contrast to the consistent decay profiles observed
`when the drug was administered parenterally (Bul-
`lingham et aL, 1980). An average peak level of
`0.74 ng/ml was reached at 150 min, but four patients
`had not achieved a peak level by the end of the 3 h
`sampling period. One patient achieved peak level at
`20min.
`
`SUBLINGUAL BUPRENORPHINE
`
`119
`
`Table 2 Plasma buprenorphine levels after sublingual ad-
`ministration
`
`Time (min)
`
`0
`
`Plasma buprenorphine (ng/ml)
`
`0
`
`0
`
`5
`0
`10
`15
`0
`0.07±0.12
`20
`0.10±0.12
`40
`60
`0.28±0.15
`0.41 ±0.17
`80
`120
`0.51±0.13
`0.74±0.16
`150
`0.71 ±0.14
`180
`Measured plasma levels minus computed contribution from
`the first (i.v.) dose. Results are mean ± s.e.mean, from the
`ten patients.
`
`The systemic availability of the sublingual dose,
`relative to the intravenous dose, from 0 to 180 min,
`estimated from the area under the curve, was 31%.
`
`Analgesia
`
`A continuous record of demands against time was
`obtained for each patient. The minimum record was
`for 11.9 h, the maximum 17 h. The total demands
`made by each patient at 7 and 13 h after the sublingu-
`al dose is shown in Figure 2. There were no significant
`differences between the three groups for the number
`of demands made, at either 7 or 13 h after their
`buprenorphine
`respective
`postoperative
`doses
`(Mann-Whitney U test).
`
`i) Duration of analgesia The median estimate of
`duration of action was 534 min from the time the
`sublingual dose was given.
`
`ii) Sex differences Men made significantly more de-
`mands than women; the male demand rate (mg
`diamorphine x 10-3/patient/min)
`8.3±0.1
`was
`females was
`(mean ± s.e.mean),
`for
`and that
`1.5 ± 0.1. This difference in rate for 60 to 600 min of
`highly
`significant
`demand
`system
`was
`use
`(P< 0.000 1).
`
`Blood gas analysis
`
`The blood gas results are shown in Table 3. Post-
`operatively the Paco2 values were raised above the
`normal range; there were no differences between
`men and women.
`
`Hormones and metabolites
`
`The measured plasma glucose and cortisol are shown
`in Table 3. Plasma glucose concentrations rose dur-
`
`Page 3
`
`

`

`120
`
`R.E.S. BULLINGHAM, H.J. McQUAY, D. DWYER, M.C. ALLEN & R.A. MOORE
`
`2.0
`
`E
`
`1.0
`
`co
`
`C C
`
`0Z
`
`U 0
`
`9. 0.8
`
`06
`
`0.4
`
`Ec
`
`oG
`
`0.2-
`
`0
`
`20
`
`40
`
`60
`
`100
`80
`Time (min)
`FIgum 1 Mean measured plasma buprenorphine concentrations foliowing aministration of the second dose by
`different routes. (0) buprenorphine 0.3mg i.v., (0) buprenorphine 0.3mg i.m., (A) buprenorphine 0.4mg
`sublingually. Ten patients in each group: sublingual data as reported here and data after intravenous and
`intramuscular 0.3 mg doses from Bullingham et al. (1980). Plasma levels after the sublingual 0.4 mg dose were
`significantly higher (P< 0.01) at the 150 and 180 min sampling points.
`
`120
`
`140
`
`160
`
`180
`
`Tabe 3 Blood gas, cortisol and glucose values
`
`Time
`(min)
`0
`30
`150
`190
`240
`300
`360
`
`Paco(kPaJ
`4.90±0.18
`5.90±0.38
`6.27 ± 0.22
`6.39±0.19
`6.11 ±0.20
`6.26±0.19
`
`Pao
`jOVa5
`23.4±1.2
`15.4±1.1
`15.4±1.1
`16.0±1.1
`15.3±1.3
`14.9± 1.6
`
`Results are mean ± s.e.mean
`
`Plasma
`cortisol
`(nmolll)
`337 ± 53
`439±86
`832±33
`923±53
`927±51
`1062±55
`1022±32
`
`Plasma
`glucose
`(mmol/l)
`4.46±0.09
`5.56±0.40
`7.03±0.46
`7.23 ±0.45
`7.15 ±0.43
`6.96±0.29
`6.65 ± 0.25
`
`ing operation and fell in the 3 h after the second dose
`by a mean of 0.58 mmoUl (0.05> P< 0.1, paired
`t-test). This change was most apparent in the last
`hour of measurement, where the fall was 0.36
`mmol/l, which was highly significant (P< 0.01),
`using the paired data. Plasma cortisol levels rose
`during operation, but after the second dose of bup-
`renorphine levels did not change significantly.
`
`Disculion
`
`The averaged kinetic data presented here for subling-
`
`Page 4
`
`

`

`SUBLINGUAL BUPRENORPHINE
`
`121
`
`The limiting factor in this argument becomes the
`pharmacologically effective level of receptor occu-
`pancy. A high level of occupancy will require high
`drug levels at the receptor and thus implies high
`plasma levels, at least at some time during absorp-
`tion. However, buprenorphine is very lipophilic and
`its association to the opiate receptor is relatively
`slow, but avid (Hambrook & Rance, 1976). Under
`these circumstances, a major determinant of loading
`of the opiate receptor is likely to be the time during
`which plasma buprenorphine levels are sustained. An
`effective level may be relatively low, and apparently
`variable, but it must be prolonged.
`Thus the sublingual route may be particularly ap-
`propriate for buprenorphine, where the slow absorp-
`tion rate combines with its associative properties with
`the opiate receptor to provide analgesia of long dura-
`tion. An additional benefit may be that side-effects
`related to plasma level may differ by the sublingual
`route from those found with parenteral use. An
`indication of this was found here, where there was
`less sedation with sublingual buprenorphine. A simi-
`lar effect has been noted for sublingual etorphine
`(Blane & Robbie, 1972).
`The finding of a lower postoperative analgesic
`requirement in females was also true for parenteral
`administration (McQuay et al., 1980). There was no
`difference in the plasma levels between the sexes in
`either study and there was no difference in the
`weights between the sexes. Other workers have seen
`a similar phenomenon with methadone (Kaiko et aL,
`1977), and this suggests that the explanation lies in an
`intrinsic difference, rather than in a specific effect of
`the drug.
`The Paco2 levels were raised postoperatively and
`this was not surprising in view of the relatively elderly
`population studied. There was no significant differ-
`ence between the values obtained in this study and
`those following parenteral buprenorphine (McQuay
`etal., 1980).
`The intraoperative rise in plasma glucose and cor-
`tisol was similar to that seen in other groups of
`patients (Brandt et aL, 1978). Plasma glucose con-
`centrations started to fall after the second dose of
`buprenorphine, and although the fall in the first 3 h
`was not as dramatic as that seen with intramuscular
`buprenorphine (0.58 mmolVl v 1.51 mmoll), the rate
`of fail was increasing and largest in the last hour of
`prolonged
`observations
`More
`measurement.
`(Moore, McQuay & Bullingham, 1980) show that
`these metabolic trends continue well beyond the 3 h
`study period. Plasma cortisol did not fall. A con-
`tinued rise is expected (McQuay et al., 1979; Mc-
`Quay etaaL, 1980), but did not occur after the subling-
`ual dose.
`This study has demonstrated substantial efficacy of
`sublingual buprenorphine given at a near parenteral
`
`Intravenous
`Intrfmuscular
`
`Subingual,
`
`/
`
`.7._
`
`£
`180
`
`a.
`
`a0
`
`40
`
`20
`
`Z X
`
`E X
`
`Ew
`a
`
`_"
`
`.7
`
`..
`
`'J
`
`!3
`
`.
`.-
`secon do"-
`Time (hi afte
`F gue 2 Postoperative analgesic demands made by
`each patient after a second dose of buprenorphine. Ten
`patients in each group. Data after postoperative in-
`travenous and intramuscular doses from McQuay et al.
`(1980). No significant differences between the demands
`made by each group at the 7 and 13 h points (Mann-
`Whitney U test).
`
`ual buprenorphine shows both delay in absorption
`and large intersubject variation in the buprenorphine
`plasma levels. These two observations would lead to
`the prediction of slow onset and variable efficacy.
`In contrast, the analgesic onset time, duration and
`individual variation was not measurably different
`from that found after parenteral postoperative bup-
`renorphine doses (McQuay et aL, 1980). Other
`analgesic measures have also shown sublingual bup-
`renorphine to be effective in postoperative pain
`(Edge et al., 1979). The surprising analgesic efficacy,
`despite low plasma levels, of the sublingual route may
`be explained by a general consideration of the dose-
`plasma level-response relation with respect to the
`sublingual route, and by specific features of bup-
`renorphine.
`Wagner (1969), in a theoretical paper, considered
`a single compartment model with a first order elimi-
`nation rate constant (kj) and a first order absorption
`rate constant (kab). He showed that as the ratio kj/k5b
`fell, the plasma profiles changed as anticipated, i.e.
`achieved lower, and later, peaks, but with more
`prolonged levels after peak. The change was most
`pronounced when k4/k,b became less than one. As-
`suming a drug receptor concentration/response rela-
`tionship of a form analogous to the Michaelis-
`Menten relationship for enzyme kinetics, he also
`showed that the early responses could be alike de-
`spite the very different plasma profiles. In essence,
`this is because of the non-linear nature of the drug
`receptor concentration/response relationship.
`The plasma curves presented here resemble the
`case when k1/kab becomes less than one. Study of
`plasma levels beyond the 3 h reported here is re-
`quired to estimate an absorption rate constant.
`
`Page 5
`
`

`

`122
`
`R.E.S. BULLINGHAM, H.J. McQUAY, D. DWYER, M.C. ALLEN & R.A. MOORE
`
`dose. Low plasma levels resulted and yet the
`analgesic, ventilatory and metabolic effects were
`similar to those following parenteral administration.
`Full description of drug pharmacokinetics by this
`route, and of the potential metabolic benefit, will
`require extended study.
`
`We thank the surgeons of the Nuffield Orthopaedic Centre
`for allowing us to study their patients, the nursing staff for
`their help and the staff of the Nuffield Department of
`Clinical Biochemistry for their technical assistance.
`H.J. McQuay was funded by the M.R.C.
`Reckitt and Colman Pharmaceutical donated the drug
`and assay materials and funded M.C.A. We gratefully ack-
`nowledge their support.
`
`BARTLETr, A.J., LLOYD-JONES, J.G., RANCE, M.J.,
`FLOCKHART, I.R., DOCKRAY, G., BENNETT, M.R.D. &
`MOORE, R.A. (1980). The radioimmunoassay of bup-
`renorphine. Eur. J. clin. Pharmac, 18, 339-345.
`BEARDWELL, C.G., BURKE, C.W. & COPE, C.L. (1968).
`Urinary free cortisol measured by competitive protein
`binding. J. Endocrinol., 42,79-83.
`BLANE, G.F. & ROBBIE, D.S. (1972). In Agonist and An-
`tagonist actions of narcotic analgesic drugs. (Proceedings
`of the Symposiumof theBritishPharmaceutical Society,
`Aberdeen, 1971.) ed. Kosterlitz, H.W., Collier, A.OJ.
`& Villareal, J.E., pp.120-127. MacMillan: London.
`BRANDT, M.R., FERNANDES, F., MORDHORT, R. &
`KEHLET, H. (1978). Epidural analgesia improves post-
`operative nitrogen balance. Br. med. J., 1, 1106-1108.
`BULLINGHAM, R.E.S., McQUAY, HJ., MOORE, R.A. &
`BENNEIT, M.R.D. (1980). Buprenorphine kinetics.
`Clin. Pharmac. Ther., 28, 667-672.
`EDGE, W.G., COOPER, G.M. & MORGAN, M. (1979).
`Analgesic effects of sublingual buprenorphine. Anaes-
`thesia, 34,463-467.
`GREVILLE, T.N.E. (1969). Theory and application ofspline
`functions. pp.156-167. London: Academic Press.
`HAMBROOK, J.M. & RANCE, M.J. (1976). The interaction
`
`of buprenorphine with the opiate receptor. In Opiates
`and endogenous opioid peptides. ed. Kosterlitz,H.W.,
`pp.295-301. Amsterdam: Elsevier/North Holland
`Biomedical Press.
`KAIKO, R.F., INTURRISI, C.E., ROGERS, A., HOUDE, R.W.
`& WALLENSTEIN, S.L. (1977). Plasma methadone and
`analgesia in cancer patients. Abstracts, First Annual
`Scienific Meeting, Eastern U.S.A. Regional Chapter,
`I.A.S.P.
`McQUAY, HJ., BULLINGHAM, R.E.S., PATERSON, G.M.C.
`& MOORE, R.A. (1980). Theclinicaleffectsof buprenor-
`phine during and after operation. Br. J. Anaesth, 52,
`1013-1019.
`McQUAY, H.J., MOORE, R.A., PATERSON, G.M.C. &
`ADAMS, A.P. (1979). Plasma fentanyl concentrations
`and clinical observations during and after operation. Br.
`J. Anaesth., 51, 543-550.
`MOORE, R.A., McQUAY, H.J. & BULLINGHAM, R.E.S.
`(1980). Reversal of postoperative hyperglycaemia by
`buprenorphine. Lancet, i, 597-598.
`WAGNER, J.G. (1968). Kineticsof pharmacologic response.
`I. Proposed relationships between response and dry
`concentration in the intact animal and man. J. theoret.
`Biol., 20,173-201.
`
`(Received August 4, 1980)
`
`Page 6
`
`