`Deuterium-Labeled Terbutaline: Demonstration of a Small
`Isotope Effect
`
`LARS 60RGSTRdM*X, CUES LINDBERG*, SVEN JONSON*, AND KLAS sVENSSON*
`Received November 3, 1 987, from the ‘Pharmacokinetics Laboratory and *Medical Statistics and Data Management, A6 Draco,
`P.O. Box 34, 5-221 00 Lund, Sweden. Accepted for publication July 22, 1988.
`~- . - _ - ~- -. - _ _ _ - - ~-
`Abstract 3 An equimolar mixture of terbutaline and [*H,]terbutaline was
`given as an oral solution to six healthy volunteers (three men and three
`women). Frequent blood samples were collected during a 24-h period
`and the plasma concentrations of unlabeled and deuterium-labeled
`terbutaline were measured by GC-MS. The overall geometric mean
`plasma concentration ratio of terbutaline to [2H6]terbutaline (isotope
`ratio) was 1.04 and differed significantly from unity. The difference can
`be explained by a difference in lipophilicity between the analogues,
`affecting their absorption. No trend in isotope ratio over the experimental
`time was observed. For unknown reasons, the isotope ratio was higher
`for women (1.07) than for men (1 .OO). Deuterium-labeled terbutaline can
`be used, intravenously or orally, as an absolute reference in bioavailabil-
`ity studies on terbutaline. If deuterium-labeled terbutaline is given orally
`in a single-day relative bioavailability study, a correlation should be
`made for the observed isotope effect.
`- - .- . .. - - - __ -. - -_ - - . - . - .- - --
`
`possible to use the labeled analogue in future bioavailability
`studies.
`
`Experimental Section
`Subjects-Six Caucasian subjects (three men and three women)
`participated in the study. Their age was between 23 and 44 years
`(mean 37 years) and they weighed from 53 to 81 kg (mean 64 kg).
`They were judged to be healthy by a physician after physical
`examination and laboratory tests. The study was approved by the
`local Ethics Committee and registered by the Swedish National
`Board of Health and Welfare. The performance was in accordance
`with the Declaration of Helsinki. Informed consent was given in
`writing.
`terbutaline [terbutaline sulphate (batch no.
`Drugs-Unlabeled
`448); powder for oral intake] and deuterium-labeled terbutaline
`([‘H,]terbutaline chloride (batch no. OP2); powder for oral intake)
`were supplied by AB Draco, Lund, Sweden (see structure). Both
`substances are stable and freely soluble, as defined by USP XXI. The
`chemical purity was >99% for both substances. Equimolar amounts
`of the two analogues (9.10 pmol of each), corresponding to a total
`dose of 5 mg of terbutaline sulphate, were dispensed in six 50-mL
`glass bottles. The actual weight of terbutaline and “%61terbutaline
`in each dose was recorded and used in the subsequent calculations.
`Each dose was dissolved in 50 mL of water at the time of administra-
`tion and the bottle was rinsed with another 50 mL of water, which
`was also ingested.
`study was open. After fasting for 10 h,
`Clinical Procedures-The
`the subjects arrived at the clinic in the morning. An indwelling
`catheter was inserted into an antecubital vein for blood sampling
`and a blank blood sample was drawn. The terbutaline: [*H6]terbuta-
`line mixture was administered orally and blood samples were
`obtained at 0.5, 1, 2, 3, 4, 6, 8, 10, 12, and 24 h after the dosing. At
`each sampling time, the first 2 mL of blood from the catheter was
`discarded and the following 10 or 20 mL collected into a heparinized
`
`The bioavailability of a new drug formulation is tradition-
`ally determined in a crossover study with another formula-
`tion as reference. The underlying assumption for this design
`is that absorption, distribution, metabolism, and elimination
`in each subject are stable over the two study periods. As this
`is not always the case, a large number of subjects must be
`included to detect an existing difference between the formu-
`lations. If both formulations are given at the same time, the
`influence of the intraindividual variations can be eliminated,
`which will increase the power of the statistical tests. To
`differentiate between drug substance derived from the two
`formulations, a stable isotope-labeled analogue can be used
`in one of the formulations. Both substances can then be
`determined simultaneously by mass spectrometry. Absolute
`or relative bioavailability of a number of drugs has been
`assessed by this technique, which has been thoroughly re-
`viewed.1 The studies have shown that the number of
`subjects can be drastically reduced without sacrificing the
`power of the statistical tests. The stable isotope-labeled drug
`can also be used as a common reference (pharmacokinetic
`internal standard) in comparative bioavailability ~ t u d i e s . ~
`Incorporation of a heavy isotope, particularly substitution
`of deuterium for hydrogen, can give rise to an isotope effect5
`that could alter the pharmacokinetics of the drug. This effect
`is usually insignificant if the label is placed in a metabolical-
`ly inert position of the molecule. Yet, any isotope effect must
`be investigated to prevent drawing misleading conclusions
`from studies with stable isotope-labeled drugs.
`Terbutaline is a &-receptor agonist that is widely used for
`the treatment of chronic obstructive lung diseases, often in
`the form of slow-release formulations. The drug undergoes
`extensive first-pass elimination in the gut wa11,6 a fact that
`makes the stable isotope coadministration technique espe-
`
`cially ~ a l u a b l e . ~ In the present study we wanted to compare
`the pharmacokinetics of deuterium-labeled and unlabeled
`terbutaline with the aim of studying whether it will be
`
`H d
`
`Terbutaline
`
`7H3
`R = C . CH3
`CH3
`
`(2He)Terbutaline R = F. CH3
`
`(7H3
`
`CZH3
`
`(7H3
`(2Hg)Terbutaline R = C - C*H3
`CZH3
`
`952 / Journal of Pharmaceutical Sciences
`Vol. 77, No. 11, November 1988
`
`0022-3549/88/17 0@0952$0 1.00/0
`0 1988, American Pharmaceutical Association
`
`Auspex Exhibit 2019
`Apotex v. Auspex
`IPR2021-01507
`Page 1
`
`
`
`tube (Venoject). After sampling, the catheter was flushed with 2 mL
`of heparinized saline (10 IU/mL) to keep it patent. The blood was
`immediately centrifuged a t room temperature and the plasma
`sucked off and stored in polystyrene tubes at -20 “C until assayed.
`Two hours after the start of the experiment, a standardized breakfast
`was served.8
`and deuterium-labeled terbuta-
`Terbutaline Assay-Unlabeled
`line in plasma (free plus protein bound) were determined by GC-MS,
`essentially as described previ~usly.~JO Briefly, after addition of 20
`pmol of [2Hs]terbutaline as internal standard, the plasma sample (2
`mL) was extracted on a disposable C18 column.1° The evaporated
`extract was silylated and the trimethylsilyl derivatives of terbuta-
`line, [2H6]terbutaline, and [2Hslterbutaline were analyzed by ammo-
`nia chemical ionization GC-MSS Separate calibration curves, in the
`range 2-80 pmol, were constructed for terbutaline and [2H6]terbuta-
`line. The lower limit of quantitation was 2 pmol. At this level, the
`within-day variation (CV), as determined during method develop-
`ment, was 3.9% for terbutaline and 3.6% for [2H6]terbutaline.
`[2Hg]Terbutaline interfered to some extent at the mlz value
`recorded for [2H6]terbutaline, and vice versa. Corrections were
`therefore made of the ion intensity ratios rnlz 4421451 and mlz 4481
`451 according to the equations given below:
`
`(1)
`
`(2)
`
`Corrected ratio rnlz 4421451 =
`1442
`1451 - 0.04411-
`1448 - 0.0298Za1
`Corrected ratio rnlz 448f451 =
`1451 - 0.04411448
`where Zu2, 1448, z461 are the measured ion intensities at mlz 442,448,
`and 451, respectively. The ion intensity ratio rnlz 4481451 measured
`for pure [2Hglterbutaline was 0.0298, and the ion intensity ratio d z
`4511448 measured for pure [2H,lterbutaline was 0.0441. There was
`no interference from [2H6]terbutaline or [2H91terbutaline at the rnlz
`value recorded for terbutaline, or vice versa.
`To be able to measure the low plasma concentrations between 10
`and 24 h, relatively large volumes of plasma (3.5-8.0 mL) had to be
`extracted. In those cases, the sample was divided into 2-mL portions
`and each portion extracted on a c18 column. The extracts from each
`sample were then combined before the G€-MS analysis. This proce-
`dure did not influence the experimental values. A sign test for a
`difference in the isotope ratio from 1.00 resulted in similar p values
`for measurements performed on 2-mL samples (p <0.001) and larger
`volumes (3.5-8.0 mL) of plasma (p = 0.004).
`The plasma samples were analyzed on two separate days. Samples
`from subject nos. 1,5, and 6 were analyzed during the first day and
`samples from subject nos. 2, 3, and 4 during the second day. The
`geometric mean isotope ratio of the standard samples, used for the
`calibration curves, was 0.996 and did not differ significantly (p =
`0.576) from 1.000. The coefficient of variation of the isotope ratio
`measurement of the standard samples was 3.0%.
`Data Analysis-The measured plasma concentrations of terbuta-
`line and [2H6]terbutaline were normalized, using the actual weight
`of each dose, to a dose of 9.10 pmol, and the ratio of terbutaline to
`[2H61terbutaline (hereinafter referred to as the isotope ratio) was
`calculated for each sample. An isotope effect, if any, will be reflected
`in this ratio. Primary and secondary pharmacokinetic parameters
`(clearance, distribution volume, and area under the curve) are linear
`
`combinations of plasma concentrations and times, and the calcula-
`tion of these pharmacokinetic parameters would add no extra infor-
`mation.
`Differences in isotope ratio between subjects, sexes, and sampling
`times were evaluated by analysis of variance, where the trial was
`viewed as a “split-plot” design. The Greenhouse-Geisser correction
`was used in the calculations. The overall deviation of the experimen-
`tal ratios from unity was also evaluated by analysis of variance. A
`potential trend in ratio over the sampling time was tested by viewing
`the within-subject values as repeated measurements and considering
`the fact that two measurements, close in time, covariate more than
`two distant measurements. The logarithmic values of the isotope
`ratios were used in the statistical evaluations; level of significance
`was set at p = 0.05.
`
`Resu I ts
`The subjects complied well to the study protocol and all
`blood samples were drawn within 3 min of the scheduled
`times. Figure 1 shows the mean plasma concentration-time
`curves of terbutaline and [‘H&erbutaline. The individual
`and mean isotope ratios are presented in Table I and plotted
`in Figure 2 (logarithmic scale on the y-axis). The overall
`geometric mean isotope ratio of the plasma samples was
`1.036 and differed significantly (p <0.001) from 1.000. Thus,
`the body seems to handle unlabeled and deuterium-labeled
`terbutaline slightly differently. A further analysis revealed a
`difference between subjects (p = 0.005); the geometric mean
`ratio was 1.075 for the women and 1.000 for the men. When
`the ratios for male and female subjects were compared with
`each other, a significant difference between the sexes was
`found (p = 0.042). In fact, the overall mean isotope effect
`could be assigned to the observed isotope effect in the women.
`Within each subject, the ratios seemed to be fairly stable
`(Figure 2). There was no trend in the ratios a t different
`sampling times (p = 0.207) when the measurements within
`
`20 24
`16
`12
`8
`4
`Time after dosing (hours)
`Figure 1-Mean plasma concentration-time curves of terbutaline
`(solid line) and [2HG]terbutaline (broken line). The values were corrected
`to a nominal dose of 9.70 pmol of each analogue.
`
`Subject
`No.
`
`Sex
`
`Table I-Plasma Concentration Ratios of Unlabeled versus Deuterated Terbutaline
`Time after Dosing, h
`1 .o
`8.0
`6.0
`24.0
`12.0
`10.0
`4.0
`3.0
`2.0
`0.5
`Meana
`1.0134 1.7118
`1.2753
`1.2331
`1.0729
`1.1128 1.1111
`1.0415 1.1269
`1.0826
`1.0746
`1.0235 1.0313
`1.0637 1.0618
`1.0559 1.0138 0.9746
`1,0330 1.0221
`1.0349 1.0325
`1.0827
`1.1251
`1.1483 1.1016 1.0979
`1.1510 1.0931
`1.0234 0.9698 1.0127 1.1201
`1.0165 1.0412 0.9040 1.0930 0.9389 1.0533 1.0112
`1.0065 1.0228 1.0459
`1.0039
`1.0092 0.9779
`1.0500 0.9937
`0.9820 0.9615 0.9679 0.9574 0.9725 0.9673 0.9191
`1.0222 0.9933 1.0101
`1.0234 1.0276 0.9716 1.0219 0.9707
`1.0645 1.0145 0.9951
`
`1
`2
`3
`4
`5
`6
`
`F
`F
`F
`M
`M
`M
`
`Mean a
`a Mean values are geometric means.
`
`1,0297 1.0087
`
`1.0122 1.0494 1.0544 1.0316 1.0032 1.0826 1.0599
`
`1.0354 1.0364
`
`Journal of Pharmaceutical Sciences / 953
`Vol. 77, No. 11, November 1988
`
`Auspex Exhibit 2019
`Apotex v. Auspex
`IPR2021-01507
`Page 2
`
`
`
`The observed difference of the isotope ratio from unity was
`not due to contamination of the deuterium-labeled terbuta-
`line with unlabeled drug, since the same batch of deuterated
`terbutaline was used to prepare standard samples for the
`calibration curves. The imprecision of the analytical method
`(3.0%, CV) was of the same magnitude as the difference of the
`mean isotope ratio from the theoretical value (4%). Detection
`of this small difference by statistical methods was possible
`because as many as six subjects were included in the study.
`Similar investigations on other drugs reported in the litera-
`ture have only comprised one to three subjects, and, with
`such small panels, only relatively large isotope effects can be
`detected. A study of the pharmacokinetic equivalence of
`metaproterenol, another &receptor agonist similar in struc-
`ture and lipophilicity to terbutaline, and a deuterated ana-
`logue in two volunteers was recently p~b1ished.l~ Although it
`was concluded that no isotope effect was at hand, the data
`indicated a reduced absorption of the deuterium-labeled
`compound.
`
`Conclusion
`A small in vivo isotope effect of deuterium-labeled terbuta-
`line was observed. The effect can be explained by reduced
`absorption of deuterium-labeled terbutaline from the gut,
`caused by a somewhat lower lipophilicity of this analogue.
`The results show that the deuterated analogue can be used
`in bioavailability studies on terbutaline when given intrave-
`nously. The labeled analogue can also be used, intravenously
`or orally, as a pharmacokinetic internal standard in studies
`on the relative bioavailability of different terbutaline formu-
`lations, comprising two or more study periods. In these cases,
`no corrections of the experimentally obtained data with the
`presently observed isotope ratios for men and women need to
`be performed. If unlabeled and deuterium-labeled terbuta-
`line are given orally in a single-day bioavailability study, the
`observed isotope ratios for men and women should be taken
`into consideration.
`
`References and Notes
`Wolen, Robert L. J. Clin. Pharmacol. 1986,26, 419-424.
`Baillie, Thomas A.; Rettenmeier, Albert W.; Peterson, Lisa A.;
`Castagnoli, Neal, Jr., In Annual Reports in Medicinal Chemis-
`try; Bailey, Denis M., Ed.; Academic Press: New York, 1984; pp
`273-282.
`Eichelbaum, Michel; von Unruh, Gerd E.; Somogyi, Andrew
`Clin. Pharmucokinet. 1982, 7, 490-507.
`Heck, Henry d'A.; Buttrill, Sidney E., Jr.; Flynn, Norman W.;
`Dyer, Robert L.; Anbar, Michael; Cairns, Thomas; Dighe, Shri-
`kant; Cabana, Bernard E. J. Pharmucokinet. Biopharm. 1979,7,
`233-248.
`van Langenhove, Agnes J . Clin. Pharmacol. 1986,26,383389.
`f e E ti. E" ur. J . ResD. Dis. 1984.65 (surd. 134). 93-100.
`Te 6 , K . Nilsson, H. T.; Persson, C. G. A.; Persson, K.; Ryr-
`Eichelbaum, M.; Dengler, H. J.;'Somog$, A,; von Unruh, G. E.
`Eur. J. Clin. Pharmacol. 1981.19, 127-131.
`Borgstrom, L.; Johansson, C. G.; Larsson, H.; Lenander, R. J.
`Phurrnacokznet. Biopharm. 1981,9, 431-441.
`Jacobsson Sven-Erik; Jonsson, Sven; Lindberg, Claes; Svens-
`son, Leif-Ake Biomed. Mass Spectrom. 1980, 7, 265-268.
`Borgstrom, L.; Nyberg, L.; Jonsson, S.; Lindberg, C.; Paulson, J.
`Br. J . Clin. Phurmacol., in press.
`Westlake, W. J. J. Pharm. Sci. 1973, 62, 1579-1589.
`Davies, D. S. Eur. J . Resp. Dis. 1984, 65 (suppl. 1341, 111-117.
`El Tayar, Nabil; van de Waterbeemd, Han; Gryllaki, Markou-
`lina; Testa, Bernard; Trager, William F. Znt. J. Pharm. 1984,19,
`271-281.
`Baillie, Thomas A. Pharmacol. Rev. 1981,33, 81-132.
`Hatch, Frank; McKellop, Keith; Hansen, Gordon; MacGregor,
`Thomas J. Pharm. Sci. 1986, 75,886-890.
`
`1.
`2.
`
`3.
`
`4.
`
`5.
`6.
`
`7.
`
`8.
`
`9.
`
`10.
`
`11.
`12.
`13.
`
`14.
`15.
`
`0.5L.
`0 4
`
`-.
`
`I
`
`8 12 16 20 24
`Time after dosing (hours)
`Figure 2-Individual
`(dotted lines) and mean (solid line) isotope ratios
`after simultaneous administration of unlabeled and deuterium-labeled
`terbutaline. The line corresponding to an isotope ratio of 7.00 is also
`shown for comparison.
`
`each subject were regarded as repeated measurements and
`the covariation of the experimental values at adjacent sam-
`pling times was considered.
`Discussion
`Stable isotope labeling of a drug can alter its physico-
`chemical properties such as pK, and lipid ~olubility.~ These
`changes may influence the fate of the drug at different steps
`along its passage through the body. Absorption, distribution,
`metabolism, or excretion can be changed. Absorption and
`distribution are processes that depend primarily on the
`molecular size and the lipophilicity of the substance. Terbu-
`taline is a hydrophilic substance with incomplete absorption
`(-50%) after oral intake.12 It has been shown that substitu-
`tion of deuterium for hydrogen makes a molecule less lipo-
`philic.13 A lower lipophilicity of [2H6]terbutaline could re-
`duce its absorption compared with terbutaline. Such an effect
`should give an isotope ratio above unity, constant over time
`after the initial absorption phase. As no time-dependent
`trend in the isotope ratio was observed, a change in absorp-
`tion is a plausible explanation of the observed isotope effect.
`The reason for the observed sex difference in isotope ratio is
`unknown. A difference in distribution of the two terbutaline
`analogues should result in a smaller distribution volume,
`and a faster rate of elimination, of the less lipophilic deuter-
`ated compound. This would show up as an isotope ratio below
`unity and as a trend in the isotope ratio over time. These
`characteristics were not found.
`Drug metabolism can give rise to large isotope effects if the
`breaking of a chemical bond to a deuterium atom is the rate-
`limiting step in the proce~s.'~ [2H6]Terbutaline was labeled
`in the t-butyl group, a position in the terbutaline molecule at
`which no metabolic reactions are known to occur.6 Renal
`excretion depends on the physicochemical properties of the
`molecule, but also on the molecular structure in case an
`active transport process exists. Metabolism and excretion are
`both processes that, in contrast to absorption, are effective
`during the whole study period. If excretion, or metabolism,
`differed for unlabeled and labeled drug, this would appear as
`a trend in the isotope ratio over time. No such trend was
`revealed when the early (0-4 h) ratios were contrasted with
`the late (6-24 h) ratios. A power analysis showed that the
`chance to detect a difference in ratios of - 0.06 between the
`first and last values was 80%. It is therefore unlikely that the
`observed isotope effect is due to differences in excretion or
`metabolism of the terbutaline analogues.
`
`954 i Journal of Pharmaceutical Sciences
`Vol. 77, No. 1 1 , November 1988
`
`Auspex Exhibit 2019
`Apotex v. Auspex
`IPR2021-01507
`Page 3
`
`