@ ‘l-HERAPEUTICS
`
`A
`
`VOLUME
`
`64 NUMBER
`
`3
`
`SEPTEMBER
`
`1998
`
`PHARMACOKINETICS
`DRUG DISPOSITION
`
`AND
`
`inhibition of triazolam and
`Ketoconazole
`alprazolam clearance: Differential kinetic
`and dynamic consequences
`
`inhibition
`of metabolic
`with alprazolam
`and
`
`in a study of the
`were evaluated
`triazolam,
`two 3A substrate
`drugs
`
`Backgromzd: Kinetic and dynamic
`consequences
`interaction
`of ketoconazole,
`a P4503A
`inhibitor,
`with different
`kinetic
`profiles.
`placebo
`(A) ketoconazole
`received
`volunteers
`study, healthy
`crossover
`5-way
`Methods:
`In a double-blind,
`(C)keto-
`twice a day plus 1.0 mg alprazolam,
`orally,
`(B) 200 mg ketoconazole
`plus 1.0 mg alprazolam
`conazole
`placebo
`plus 0.25 mg
`triazolam
`orally,
`(D) 200 mg ketoconazole
`twice a day plus 0.25 mg
`tria-
`zolam, and
`(E) 200 mg ketoconazole
`twice a day plus benzodimpine
`placebo. Plasma
`concentrations
`and
`pharmacodynamic
`parameters
`were measured
`after each dose.
`P < .002) and
`(27 versus 86 mL/min,
`Reszlla: For
`trial B versus
`trial A, alprazolam
`clearance was reduced
`(59 versus 15 hours; P c .03), whereas
`peak plasma
`apparent
`elimination
`half-life
`(tX) prolonged
`concen-
`tration
`(C,,)
`was only slightly
`increased
`(16.1
`versus 14.7 ng/mL).
`The &hour
`pharmacodynamic
`effect
`areas
`for electroencephalographic
`(EEG)
`beta activity were
`increased
`by a factor
`of 1.35, and
`those
`for
`
`Jerold S. Harmatz,
`Lisa L. von Moltke,
`C. Eugene Wright,
`J. Greenblatt,
`David
`Lisa M. Harrel,
`Kate Corbett, Molly Counihan,
`Sara Tobias,
`Bruce L. Ehrenberg,
`Richard
`I. Shader Boston, Mass., and Kalamazoo, Mich.
`
`and
`
`Therapeutics
`and Experimental
`of Pharmacology
`the Department
`From
`of Clinical
`Pharmacology,
`lkfts
`University
`School
`and the Division
`of Medicine
`and New England Medical
`Center, Boston,
`and
`the Clin-
`ical Pharmacokinetics
`Unit, Pharmacia
`and Upjohn Co., Kalamazoo.
`Supported
`in part by grant MH-34233
`from
`the Department
`of Health
`and Human
`Services,
`by grant RR-00054
`supporting
`the General
`Clinical
`Research
`Center, Tufts University
`School
`of Medicine
`and
`the New England Medical
`Center Hospital
`(Boston, Mass.),
`and by
`a grant-in-aid
`from Pharmacia
`& Upjohn
`(Kalamazoo,
`Mich.).
`Dr.
`
`Award K21-
`Development
`of a Scientist
`is the recipient
`von Moltke
`from
`the Department
`of Health
`and Human
`Services.
`(MH-01237)
`Received
`for publication
`Oct. 14, 1997; accepted
`Jan. 3, 1998.
`Reprint
`requests:
`David
`J. Greenblatt,
`MD, Department
`of Pharma-
`cology
`and Experimental
`Therapeutics,
`Tufts University
`School
`of
`Medicine,
`136 Harrison
`Ave., Boston, MA 02111. E-mail:
`Dgreen-
`blatt@Infonet.tufts.edu
`Inc.
`Copyright
`0 1998 by Mosby,
`0009-9236/98/$5.00
`+ 0
`13/l/88636
`
`237
`
`1
`
`TEVA1046
`
`

`

`238 Greenblutt
`
`et al.
`
`CLINICAL
`
`PHARMA COLOGY & THERAPEUTICS
`SEPTEMBER 1998
`
`test (DSST) decrement were increased by 2.29 for trial B versus trial A. For trial
`digit-symbol substitution
`D versus trial C, triazolam clearance was reduced (40 versus 4-44 mL/min; P < .002),
`tX was prolonged
`(18.3 versus 3.0 hours; P c .Ol), and C,, was increased (2.6 versus 5.4 ng/mL, P < .OOl). The g-hour
`effect area for EEG was increased by a factor of 2.51, and that for DSST decrement was increased by 4.33.
`Observed in vivo clearance decrements due to ketoconazole were consistent with those anticipated on the
`basis of an in vitro model, together with in vivo plasma concentrations of ketoconazole.
`Conclusion: For tpiazolam, an intermediate-extraction
`compound,
`impaired clearance by ketoconazole has
`more profound clinical consequences than those for alprazolam, a low extraction compound.
`(Clin Phar-
`macol Ther 1998;64:237-47.)
`
`triazolam and alprazo-
`The triazolobenzodiazepines
`lam are extensively prescribed in clinical practice as
`hypnotic and anxiolytic-antipanic agents, respectively.
`Both drugs are biotransformed
`to parallel hydroxylated
`metabolites at the a- and 4-positions on the mole-
`cule.1-4 These reactions are mediated mainly,
`if not
`entirely. by cytochrome P4503A isoforms.l-5 However,
`the human pharmacokinetics of the two compounds dif-
`fer substantially. Triazolam has a hepatic clearance in
`the “intermediate”
`range relative to hepatic blood flow
`and an elimination half-life (t%) in the range of 1% to 5
`hours.637 Absolute bioavailabilitp after oral administra-
`tion is approximately 50%, probably due in large part
`to presystemic extraction, with the likelihood of an
`important contribution by gastrointestinal P4503A iso-
`forms.T>s In contrast, alprazolam has a hepatic clear-
`ance of less than 10% of hepatic blood flow, an elimi-
`nation tYZ in the range of 10 to 20 hours, and absolute
`bioavailability generally exceeding 90%.1,9,10
`The antifungal agent ketoconazole has been estab-
`lished as a highly potent inhibitor of human cytochrome
`P4503A isoforms, leading to large and clinically impor-
`tant drug interactions with a number of 3A substrate
`drugs, including
`triazolam.3*11 Physiologically based
`pharmacokinetic
`theory predicts that the character and
`clinical
`importance of an interaction of a metabolic
`inhibitor such as ketoconazole with an orally adminis-
`tered substrate drug will depend on the hepatic clear-
`ance and presystemic extraction of this substrate.12-l6
`We evaluated the magnitude and clinical consequences
`of the interaction of ketoconazole with triazolam and
`alprazolam, and the extent to which these in vivo inter-
`actions are consistent with
`in vitro
`interactions
`observed in human liver microsomes.
`
`METHODS
`Design. The protocol was reviewed and approved by
`the Human
`Investigation Review Committee serving
`Tufts University School of Medicine and New England
`Medical Center Hospital. Seven healthy male volun-
`teers (age range, 21 to 44 years) participated after each
`
`gave written informed consent. All were active ambu-
`latory nonsmoking adults, with no evidence of medical
`disease and taking no other medications.
`The study had a double-blind,
`single-dose, 5-way
`crossover design. Medications were packaged identi-
`cally in opaque capsules and administered orally, with
`at least 7 days elapsing between trials (Table I). The 5
`treatment conditions were as follows:
`A. Ketoconazole placebo plus 1 .O mg alprazolam
`B. Ketoconazole (200 mg) plus 1.0 mg alprazolam
`C. Ketoconazole placebo plus 0.25 mg triazolam
`D. Ketoconazole (200 mg) plus 0.25 mg triazolam
`E. Ketoconazole (200 mg) plus benzodiazepine placebo
`Procedures. At 8 AM on study day 1, subjects entered
`the outpatient Psychopharmacology Research Unit,
`where they received 200 mg ketoconazole (or placebo)
`and underwent practice trials with
`the rating
`instru-
`ments and psychomotor
`testing procedures described
`below. Subjects took a second dose of ketoconazole (or
`placebo) at home on the evening of day 1, and the third
`dose in the outpatient unit the morning of day 2. In the
`evening of day 2 they were admitted
`to the General
`Clinical Research Center at New England Medical Cen-
`ter Hospital where they received their fourth dose of
`ketoconazole or placebo. They remained in the General
`Clinical Research Center for the next 36 hours.
`Subjects ingested a standardized light breakfast, with
`no caffeine-containing
`beverages, at approximately
`7:30 AM on the morning of day 3. They continued
`to
`fast until 12 noon, after which they resumed a normal
`diet (without caffeine-containing beverages). The fifth
`dose of ketoconazole
`(or placebo) was given at 8 AM,
`and the single challenge dose of triazolam, alprazolam,
`or placebo was given at 9 AM.
`from an
`Venous blood
`samples were drawn
`indwelling
`cannula
`into heparinized
`tubes before
`administration and at the following
`times after admin-
`istration: %, 1, l%, 2,3,4,6,
`8, and 12 hours. Two addi-
`tional samples were drawn at 24 and 48 hours. Samples
`were centrifuged, and the plasma was separated and
`frozen until the time of assay.
`
`2
`
`

`

`CLINICAL
`VOLUME
`
`PHARMA COLOGY & THERAPEUTICS
`64, NUMBER 3
`
`Greenblatt
`
`et al.
`
`239
`
`0.3 3
`8
`6
`5
`4
`3
`12
`-lo
`DOSE
`HOURS AFTER BENZODIAZEPINE
`Figure 1. Mean plasma ketoconazole concentrations when ketoconazole was administered alone
`(trial E), with triazolam (trial D), or with alprazolam (trial B).
`
`I
`36
`
`I
`42
`
`I
`40
`
`7
`
`I
`12
`
`18
`
`24
`
`30
`
`Table I. Summary of study design
`Day I
`
`Day 2
`
`Trial
`
`A
`B
`C
`D
`E
`
`AM
`
`P
`K
`P
`K
`K
`
`PM
`
`P
`K
`P
`K
`K
`
`AM
`
`P
`K
`P
`K
`K
`
`PM
`
`P
`K
`P
`K
`K
`
`HAM
`
`P
`K
`P
`K
`K
`
`P, Placebo; K, 200 mg ketoconazole;
`
`ALP, 1.0 mg alprazolam; TFZ, 0.25 mg triazolam.
`
`Day 3
`
`9AM
`
`ALP
`ALP
`TRZ
`TRZ
`P
`
`PM
`
`P
`K
`P
`K
`K
`
`Day 4
`
`AM
`
`P
`K
`P
`K
`K
`
`PM
`
`P
`K
`P
`K
`K
`
`Table II. Pharmacokinetic interaction of ketoconazole with triazolam and alprazolam
`Control*
`With ketoconazole*
`
`Values of Student
`
`t test
`
`Triazolam study (n = 6)
`C,,
`(ng/mU
`Lax (h after dose)
`Elimination tYZ (h)
`Total AUC (ng/mL h)
`Oral CL (mL/min)
`Alprazolam
`study (n = 7)
`Cm, hW-4
`&
`(h after dose)
`Elimination tl/ (h)
`Total AUC (ng/mL . h)
`Oral CL (mL/min)
`
`C -,
`*Mean
`
`Peak plasma concentration;t,,,,,
`f SE values.
`
`2.6 tt 0.3
`1.2 f 0.2
`3.0 f 0.3
`10.6 i 1.6
`444*
`73
`
`14.7 * 1.6
`1.4 f 0.3
`15.2 f 2.1
`237i43
`86+ 16
`
`5.4 f 0.4
`1.9 f 0.4
`18.3 zt 4.0
`145.4i
`39.1
`39.6 i 10.5
`
`16.1 zt 1.4
`1.5 f 0.2
`59* 17
`944*
`277
`27i7
`
`7.28 (P < .OOl)
`1.96 (NS)
`4.14 (P < .Ol)
`3.58 (P<.O2)
`6.44(P<.OO2)
`
`1.16 (NS)
`0.40 (NS)
`2.85 (PC .03)
`2.85 (PC .03)
`5.49 (P<.OO2)
`
`time to reach C ,,,%; tvz, half-life; AUC, area under
`
`the plasma concentmtion
`
`versus
`
`time curve; CL, clearance.
`
`(EEG) mon-
`An &electrode electroencephalograpbic
`tage was affixed as follows: left and right frontal (F3, F4),
`left and right central (C3, C4), as well as midline frontal
`
`(Fz), central (Cz), parietal (Pz), and occipital (Oz), with
`a nose electrode as reference. Procedures for preparation
`of electrode sites and affixing of electrodes have been
`
`3
`
`

`

`240 Greenblatt et al.
`
`CLINICAL
`
`PHARMACOLOGY
`
`&THERAPEUTICS
`SEl’lTMBER
`1998
`
`6-
`
`Kt3OCONUOLE
`
`0.1 A
`
`0
`
`6
`
`12
`
`16
`
`24
`
`30
`
`36
`
`42
`
`46
`
`0
`
`2
`
`4
`
`6
`
`6
`
`10
`
`12
`
`WITH
`KETOCON&!OlE
`
`16
`
`-
`
`9
`
`-
`
`6-
`
`42
`
`46
`
`0
`
`10
`
`12
`
`0
`
`6
`12
`18
`24
`30
`36
`HOURS AFTER DOSE
`
`2
`4
`6
`6
`HOURS AFTER DOSE
`left panel, Mean * SE plasma triazolam concentrations after 0.25 mg triazolarn
`2. Upper
`Figure
`alone (trial C, control) and with coadministration of ketoconazole (trial D), shown with a logarith-
`mic concentration axis. Upper
`right panel, The first 8 hours after administration, shown with a
`linear concentration axis. Lower
`left panel, Mean f SE plasma alprazolam concentrations after 1.0
`mg alprazolam alone (trial A, control) and with coadministration of ketoconazole (trial B), shown
`with a logarithmic concentration axis. Lower
`right panel, The first 8 hours after administration,
`shown with a linear concentration axis.
`
`described previously.3J7-20 The EEG was recorded in 4-
`second epochs for as long as necessary to ensure at least
`2 minutes of artifact-free information, before administra-
`tion, and at times corresponding to blood samples up to
`24 hours after administration. Data were digitized over
`the power spectrum from 4 to 30 cycles per second (Hz),
`and analyzed by fast Fourier
`transform
`to determine
`amplitude in the total spectrum (4 to 30 Hz) and in the
`beta (12 to 30 Hz) frequency range.
`Subjects’ self-ratings of sedative effects and mood
`state were obtained on a series of 100~mm visual ana-
`
`log scales.3J7,2u,21 Ratings of sedation were also per-
`formed by a trained observer, using the same rating
`instrument, without knowledge of the treatment condi-
`tion. Self- and observer-ratings were obtained
`twice
`before medication administration and at %, 1, 1% 2,3,
`4,6, and 8 after administration.
`test (DSST) was
`The digit symbol substitution
`administered
`twice before dosing and at times corre-
`sponding to rating scales.sJ7,20Jt Subjects were asked
`to make as many correct symbol-for-digit
`substitutions
`as possible within a 2minute
`period. Subjects com-
`
`4
`
`

`

`CLINICAL
`VOLUME
`
`PHARMA COLOGY &THERAPEUTICS
`64, NUMBER 3
`
`Greenblatt
`
`et al.
`
`241
`
`16
`
`14
`
`12
`
`10
`
`8
`
`6
`
`4
`
`2
`
`0
`
`0
`
`20
`
`24
`
`-2
`
`L
`
`0
`
`20
`
`24
`
`16
`12
`a
`4
`16
`12
`8
`4
`HOURS AFTER DOSE
`HOURS AFTER DOSE
`Figure 3. Changes over predose baseline in percentage of electroencephalographic (EEG) ampli-
`tude falling in the beta frequency range; values are the mean of left and right frontal leads. Each
`point is the mean for all subjects at the corresponding time. Standard errors for individual data
`points, omitted for clarity, are available from the authors on request. Left panel, Results with tria-
`zolam alone (trial C), triazolam plus ketoconazole (trial D), and ketoconazole alone (trial E). Right
`panel, Results with alprazolam alone (trial A), alprazolam plus ketoconazole (trial B), and keto-
`conazole alone (trial E).
`
`Table III. AUC of EEG change versus time, based on mean values of percent beta amplitude from left and right
`frontal leads
`
`AUC(O-8)
`
`AUC(O-12)
`
`AUC(O-24)
`
`Triazolam study (n = 6)
`Trial C (triazolam)*
`Trial D (triazolam and ketoconazole)*
`Trial E (ketoconazole)*
`Value of F
`Alprazolam
`study (n = 7)
`Trial A (alprazolam)*
`Trial B (alprazolam and ketoconazole)*
`Trial E (ketoconazole)*
`Value of F
`
`29.7 zt 8.7
`74.5(*10.1)
`1.4 f 4.4
`18.1 (P < .OOl)t
`
`53.0 i 12.3
`71.3 f 11.1
`1.2 f 3.7
`21.4 (P < .OOl)§
`
`27.5 zt 10.9
`106.9 f 9.8
`1.0 LII 8.0
`48.1 (P < .OOl)t
`
`67.8 ic 12.1
`98.1 LII 11.7
`1.9 zt 6.9
`32.1 (P < .OOl)t
`
`15.8 f 12.1
`174.9 i 14.6
`-8.0 ic 13.2
`52.51 (P < .OOl)$
`
`97.9 f 19.4
`151.0 f 11.0
`-4.3 f 11.7
`30.5 (P < .OOl)t
`
`AUC(O-8), AUC
`
`from 0 to 8 hours; AUC(O-12),
`
`EEG, Electroencephalographic;
`*Mean
`f SE AUC values.
`.05) from
`(P i
`(Student-Newman-Keuls
`the other
`tBach
`trial significantly
`different
`$Trial D signiticantly
`different
`from C and E; C and E not significantly
`different
`§Trials A and B significantly
`different
`from E; A and B not significantly
`different
`
`comparison).
`(Student-Newman-Keuls
`(Student-Newman-Keuls
`
`comparison).
`comparison).
`
`AUC
`
`from 0 to 12 hours; AUC(O-24).
`
`AUC
`
`from 0 to 24 hours.
`
`pleted equivalent DSST variants, with no individual
`taking the same test more than once.
`Acquisition and recall of information were evaluated
`using a word-list free recall procedure that was admin-
`istered 1% hours after benzodiazepine
`or placebo
`administration.s~17JtJ* Sixteen words,
`taken from 4
`different
`categories, were read in random order
`in
`“shopping-list”
`fashion. Recall was tested immediately
`after presentation of the list, as subjects wrote list items
`in any order. The list was then presented in a different
`
`random order, with subjects again writing down the
`items immediately. This was repeated a total of 6 times.
`At 24 hours after dosing, subjects were asked to remem-
`ber as many words as possible from the list (free recall);
`thereafter, the same list was read in 6 different random
`sequences.
`Analysis of data. Plasma concentrations of triazolam
`and alprazolam were determined by gas chromatogra-
`phy with electron-capture detection.svQsJ‘t Plasma con-
`centrations of ketoconazole were determined by HPLC.3
`
`5
`
`

`

`242
`
`Greenblatt
`
`et al.
`
`CLINICAL
`
`I’ HAIWIACOLOGY
`
`& THERAfEUTICS
`SEPTEMBER 1998
`
`-25
`
`0
`
`7
`
`8
`
`0
`
`7
`
`8
`
`6
`5
`4
`3
`12
`6
`5
`4
`3
`12
`HOURS AFTER DOSE
`HOURS AFTER DOSE
`Figure 4. Changes over predose baseline in scores on the Digit Symbol Substitution Test (DSST).
`Each point is the mean for all subjects at the corresponding time. Standard errors for individual
`data points, omitted for clarity, are available from the authors on request. Left panel, Results with
`triazolam alone (trial C), triazolam plus ketoconazole (trial D), and ketoconazole alone (trial E).
`Right panel, Results with alprazolam alone (trial A), alprazolam plus ketoconazole (trial B), and
`ketoconazole alone (trial E).
`
`Table IV Area under the B-hour effect change curves (AUC) for selected pharmacodynamic variables
`Observer-rated
`sedation
`
`Feeling
`
`“spacey ”
`
`DSST score
`
`Self-rated sedation
`
`Triazolam study (n = 6)
`Trial C (triazolam)*
`Trial D (triazolam and ketoconazole)*
`Trial E (ketoconazole)*
`Value of F
`Alprazolam
`study (n = 7)
`Trial A (alprazolam)*
`Trial B (alprazolam and ketoconazole)*
`Trial E (ketoconazole)*
`Value of F
`
`102.2 i 32.6.t
`329.6 * 61.57
`5.7 f 4.4
`23.0 (P < 001)
`
`89.5 f 21.7t
`103.9 * 16.4t
`3.9 f 4.1
`15.21 (P < ,001)
`
`58.9 iz 30.3
`153.4 f 48.9?
`18.7 zt 5.5
`5.79 (P < .025)
`
`28.2 LIZ 42.1
`118.3 zt 37.3t
`16.1 ct 5.4
`2.39 (NS)
`
`68.2 zt 62.3
`116.1 LIZ 35.6t
`5.9 f 9.3
`1.99 (NS)
`
`11.8 i 45.2
`103.9 f 57.9
`5.1 f 7.9
`1.22 (NS)
`
`-31.7 f lO.O$
`-137.4 i 30.6t
`-1.8 ZIZ 12.9
`11.03 (P < .005)
`
`-21.0 ZIZ 15.2
`-47.9 f 12.1 t
`-1.1 f 11.0
`3.34 (P < .OS)
`
`f SE effect AUC values.
`*Mean
`tP -z .05 (significance
`of difference
`$P < .l (significance
`of difference
`
`from Trial E).
`from Trial E).
`
`log-linear phase of
`The slope (beta) of the terminal
`each plasma concentration versus time curve was deter-
`mined by linear regression analysis. This slope was
`used to calculate
`the apparent elimination
`tl/. Area
`under the plasma concentration versus time curve from
`time zero until
`the last detectable concentration was
`determined by the linear trapezoidal method. To this
`area was added the residual area extrapolated to infin-
`ity, calculated as the final concentration divided by beta,
`yielding
`the total area under the plasma concentration
`versus time curve (AUC). The peak plasma concentra-
`tion and the time of peak concentration were used as
`
`measures of the rate of appearance of drug in systemic
`circulation. Oral clearance was calculated as the admin-
`istered dose divided by the total AUC.
`For self- and observer-ratings on visual analog
`scales, the 2 predose baseline ratings were averaged,
`and all scores after admininistration were expressed as
`the increment or decrement relative to the mean pre-
`dose value. Scores on the DSST were similarly ana-
`lyzed. The word-list memory test was analyzed as the
`absolute number of words correctly recalled.
`For each EEG recording session, the relative beta
`amplitudes (beta divided by total, expressed as percent)
`
`6
`
`

`

`CLINICAL
`VOLUME
`
`P HARMACOLOGY
`64, NUMBER 3
`
`& THERAPEUTICS
`
`Greenblatt
`
`et al. 243
`
`16
`
`E m 4
`I
`z2
`
`n KETOCONUOLE
`
`16
`
`6
`
`--
`FREE RECALL
`
`0 TRIALS
`
`0 TRIALS
`1.5 HOURS
`
`FREE RECALL
`
`0 TRIALS
`
`I) TRIALS
`1.5 HOURS
`24 HOURS
`24 HOURS
`Figure 5. Mean * SE number of words recalled immediately after 6 learning trials at 1% hours
`after benzodiazepine (or placebo) administration; free recall was tested at 24 hours after adminis-
`tration, followed by 6 relearning trials. Left panel, Comparison of ketoconazole (trial E), triazolam
`(trial C), and triazolam plus ketoconazole (trial D). At 1% hours, the 3 trials differed significantly
`from each other (F = 7.44; P < .OOl); the Student-Newman-Keuls procedure indicated that trials C
`and D did not differ from each other and that trials C and E did not differ from each other, but tri-
`als D and E were significantly different. At free recall, trials were significant from each other (F =
`4.77; P < .OOl); based on the Student-Newman-Keuls procedure, trial E differed significantly from
`trials C and D, but trials C and D did not differ from each other. The 3 trials were not significantly
`different with relearning at 24 hours (F = 2.69). Right panel, Comparison of ketoconazole (trial
`E), alprazolam (trial A) and alprazolam plus ketoconazole (trial B). At 1% hours, the 3 trials were
`significantly different (F = 7.19; P < .OOl); based on the Student-Newman-Keuls procedure, trial E
`differed from trials A and B, but trials A and B did not differ from each other. At free recall, the tri-
`als differed significantly (F = 5.55; P < .02); trials A and E differed from each other, but trials B
`and E did not differ significantly from each other, and trials B and A did not differ significantly
`from each other. Upon relearning at 24 hours, the 3 trials did not differ significantly (F = 2.18).
`
`in the predose recordings were used as baselines. All
`values after administration were expressed as the incre-
`ment or decrement over that trial’s mean predose base-
`line value.
`For each pharmacodynamic variable, the area under
`the &hour plot of effect change score versus time was
`calculated. For EEG measures, area under the effect
`curve was also calculated up to 12 and 24 hours after
`benzodiazepine or placebo administration.
`Statistical procedures included linear and nonlinear
`regression, ANOVA, and the Student-Newman-Keuls
`procedure. ANOVA indicated no significant difference
`among the 5 treatment conditions
`in predose baseline
`values of any of the pharmacodynamic variables.
`In vitro studies. Biotransformation of triazolam and
`alprazolam to their respective hydroxylated metabolites
`
`liver microsomal
`in vitro using human
`was studied
`preparations.334 Changes in reaction velocity caused by
`coincubation with ketoconazole were used to calculate
`in vitro
`inhibition
`constants (Ki)
`for ketoconazole.
`Using the plasma concentrations of ketoconazole mea-
`sured in this study, an in vitro-in vivo scaling model
`was used to estimate the extent to which in vivo inter-
`actions were consistent with
`those observed
`in
`vitro.3,4326
`
`RESULTS
`Subjective effects were limited to the benzodiazepine
`agonist effects described below. There were no adverse
`reactions. Data for all 7 subjects were available
`for
`analysis of the alprazolam-ketoconazole
`interaction
`(trials A, B, and E). Due to a medication error, one sub-
`
`7
`
`

`

`244
`
`Greenblatt
`
`et al.
`
`CLINICAL
`
`PHARMA COLOGY & THERAPEUTICS
`SElTEMBER
`1998
`
`14
`
`-
`
`E
`z
`g
`
`12
`
`-
`
`10
`
`-
`
`2
`iti
`Y6
`Iii
`m
`
`5
`w
`
`s
`f
`
`6-
`
`4-
`
`2-
`/
`
`o.-'
`
`.
`
`-2 4
`
`0
`
`'Ef-
`
`,eG-*
`
`D
`
`./,
`
`;9
`
`/’
`
`,I’
`I
`
`14 -
`
`12
`
`-
`
`10
`
`-
`
`a -
`0 -
`
`a3
`/
`
`,
`
`'/
`
`,47
`A
`
`k,Z
`
`4
`
`Yy
`,
`
`<?
`
`q
`
`-1
`
`;:
`
`-j~p,
`
`1
`
`2
`
`3
`
`4
`
`5
`
`PLASMA TRIAZOLAM
`
`(nglml)
`
`I
`6
`
`-2
`
`2
`
`4
`
`10
`a
`6
`PLASMA ALPRAZOLAM
`
`12
`
`14
`
`16
`
`(rig/ml)
`
`in EEG amplitude
`to mean change over baseline
`6. Relation of mean plasma concentration
`Figure
`to a sigmoid E,,,,
`at the corresponding
`time. The concentration-response
`relationships were fitted
`to both
`trials. Left
`values applied
`model, with
`the assumption
`for each drug
`that the same E,,,
`panel, Triazolam
`(trial C) and triazolam plus ketoconazole
`(trial D). The plasma concentration
`cor-
`responding
`to 50% of maximum effect (EC5c) value for trial D was larger
`than that for trial C (2.5
`versus 1.6 ng/mL), but the difference was not significant. Right panel, Alprazolam
`(trial A) and
`values
`for
`the 2 trials were nearly
`identical
`alprazolam
`plus ketoconazole
`(trial B). EC,,
`(11.1 ng/mL).
`
`in trial C; therefore data for 6
`ject did not participate
`subjects were available for analysis of the triazolam-
`ketoconazole interaction.
`Plasma ketoconazole concentrations. There were no
`significant differences among
`trials B, D, and E in
`plasma ketoconazole concentrations at individual
`time
`points or in area under the ketoconazole plasma con-
`centration curve for the first 8 hours after benzodi-
`azepine (or placebo) administration
`(Figure 1).
`Pharmacokinetics
`of triazolam and alprazolam.
`Coadministration
`of
`ketoconazole
`significantly
`increased the triazolam peak plasma concentration, pro-
`longed the elimination t%, increased AUC, and decreased
`apparent oral clearance (Table II and Figure 2). The frac-
`tional decrement in triazolam clearance attributable
`to
`ketoconazole for each subject was calculated as 1 minus
`the ratio of clearance with ketoconazole coadministra-
`tion, divided by clearance without ketoconazole;
`this
`quantity was expressed in percent.3,4,25,26 Using this
`method, the mean f SE decrement in triazolam clear-
`ance compared to control was 92% * 2%. Ketoconazole
`also significantly prolonged alprazolam elimination
`t%,
`increased AUC, and reduced clearance (Table II and Fig-
`ure 2); the mean decrement in alprazolam clearance was
`70% f 4%. However, ketoconazole did not significantly
`increase alprazolam peak plasma concentration.
`
`Relation of in vitro results to in vivo kinetic inter-
`action. The in vitro-iu vivo interaction scaling model pre-
`dicted greater than 95% impairment of triazolam clear-
`ance at plasma ketoconazole concentrations exceeding 0.1
`pg/mL;
`the mean observed degree of impairment in this
`study was 92%. For alprazolam, the predicted degree of
`impaired clearance by ketoconazole was 78% at plasma
`ketoconazole equal to 0.1 l.tg/mL, and greater than 95%
`at 1 .O l.tg!ti,
`the mean observed impairment was 70%.
`Pharmucodynamic
`results. Triazolam alone (trial C)
`and alprazolam alone (trial A) both produced signifi-
`cant increases in EEG beta amplitude, whereas keto-
`conazole alone (trial E) produced no significant EEG
`changes, (Table III and Figure 3). Effects were most
`clearly evident in the frontal leads, but similar changes
`were also observed for the other lead placements. Coad-
`ministration of ketoconazole significantly enhanced
`EEG effects of triazolam (trial D versus C) and alpra-
`zolam (trial B versus A). The quantitative extent of EEG
`effect enhancement by ketoconazole was much greater
`for triazolam than for alprazolam (Table III and Figure
`3). When triazolam was given with ketoconazole (Trial
`D), EEG beta amplitude remained significantly above
`baseline even at 24 hours after administration.
`the
`For other pharmacodynamic variables, including
`DSST, the consequences of the kinetic interaction with
`
`8
`
`

`

`CLINICAL
`VOLUME
`
`PHARMACOLOGY
`64, NUMBER 3
`
`& THERAPEUTICS
`
`Greenblatt
`
`et al.
`
`245
`
`to that observed with the
`ketoconazole were similar
`EEG (Table IV and Figure 4). Triazolam and alprazo-
`lam given with ketoconazole placebo (trials A and C)
`produced benzodiazepine
`agonist effects
`that were
`greater than with ketoconazole alone (trial E). Effects
`were enhanced when benzodiazepines were coadmin-
`istered with ketoconazole
`(trials B and D). Statistical
`significance was not always attained due to the degree
`of variability and the small sample size.
`The test of information acquisition and recall showed
`significantly
`impaired
`immediate
`recall of words
`learned at 1% hours after administration,
`as well as
`greatly impaired
`free recall at 24 hours after adminis-
`tration, after triazolam (trial C) or alprazolam (trial A)
`compared with ketoconazole
`(trial E; Figure 5). Tria-
`zolam effects were enhanced, although not significantly
`so, by coadministration
`of ketoconazole
`(trial D),
`whereas effects of alprazolam were not enhanced by
`ketoconazole
`(trial B). Relearning of the same list of
`words at 24 hours after administration was not signifi-
`cantly different among the 3 trials.
`Plasma triazolam or alprazolam concentrations were
`significantly
`related to pharmacodynamic effects (Fig-
`ure 6). For alprazolam, the concentration-response rela-
`tionships were not altered by ketoconazole administra-
`tion. However, in the case of triazolarn the data suggested
`a “rightward” shift in the concentration-response
`rela-
`tionship during coadministration of ketoconazole, but the
`extent of the shift was not statistically significant.
`
`DISCUSSION
`Triazolam and alprazolam are substrates for human
`cytochrome P4503A isoformsl-5
`In vitro studies have
`shown that the azole antifungal agent ketoconazole, a
`highly potent
`inhibitor of human P4503A activity,
`inhibits biotransformation
`of both of these drugs, with
`Ki values consistently below 0.05 pmol/L
`for triazo-
`lam3 and below 0.1 pmol/L
`for alprazolam.4 Plasma
`ketoconazole concentrations during usual therapeutic
`doses in humans typically range from 0.1 to 8.0 pg/mL
`(0.19 to 15 ymolL),
`as was the case in the present
`study.27328 Assuming that ketoconazole concentrations
`at the intrahepatic and gastrointestinal
`sites of drug
`metabolism are similar to concentrations in plasma (an
`assumption
`that is supported by considerable experi-
`mental data 3,*6), therapeutic doses of ketoconazole are
`anticipated to produce levels that exceed Ki values for
`triazolam and alprazolam biotransformation
`by a con-
`siderable margin. This relationship
`thereby predicts a
`large impairment of triazolam or alprazolam clearance
`in vivo due to of coadministration of ketoconazole, as
`was observed
`in the present study and in previous
`
`interaction.3,lr
`reports of the triazolam-ketoconazole
`Since triazolam undergoes substantial hepatic or gas-
`trointestinal presystemic extraction after oral adminis-
`tration, coadministration of ketoconazole substantially
`increased triazolam peak plasma concentration, as well
`as prolonging elimination
`t% and increasing total AUC.
`Triazolam clearance was reduced on average by more
`than 90%. Ketoconazole also prolonged alprazolam
`elimination
`tl/ and increased AUC, with a mean clear-
`ance reduction of 70%. However, ketoconazole did not
`significantly
`increase alprazolam peak plasma concen-
`tration, as would be predicted based on the observation
`that orally administered alprazolam undergoes minimal
`presystemic extraction. The magnitude of the pharma-
`cokinetic
`interactions of ketoconazole with triazolam
`and alprazolam was consistent with
`that anticipated
`based on in vitro-in
`vivo scaling approaches,
`the
`assumptions and
`limitations
`of which have been
`described previously.3,4,2s,26,29-s2
`inter-
`The clinical importance of a pharmacokinetic
`action depends on the kinetic characteristics and con-
`centration-response
`relationship
`for the affected drug,
`as well as on the magnitude of the interaction
`itself.29
`We used a number of pharmacodynamic measures to
`evaluate the time course and intensity of benzodiazepine
`agonist effects after administration of triazolam and
`alprazolam alone and when coadministered with keto-
`conazole. These measures included the fraction of EEG
`amplitude falling in the beta frequency range,33,34 per-
`formance on the DSST, immediate and delayed recall
`of information, and a number of subjective and observer
`ratings of drug effects. The influence of ketoconazole
`on pharmacodynamic effects of triazolam was dramatic,
`since plasma triazolam concentrations were greatly
`increased at all time points after administration. How-
`ever, ketoconazole enhanced
`the memory-impairing
`effects of triazolam to a somewhat lesser extent, possi-
`bly because information acquisition and recall was sub-
`stantially impaired by triazolam alone. Also of interest
`was the apparent rightward shift in the concentration-
`response relationship for triazolam alone (trial C) com-
`pared with
`triazolam plus ketoconazole
`(trial D).
`Although
`the extent of the shift did not attain statistical
`significance,
`it is consistent with a previous observa-
`tions The mechanism of this phenomenon
`is not estab-
`lished but it might be explained by the finding that keto-
`conazole acts as a neutral ligand at the y-aminobutyric
`acid-benzodiazepine
`receptor complex.35
`Alprazolam alone (trial A) also produced significant
`benzodiazepine agonist effects on all pharmacodynamic
`measures. Consistent with its elimination
`tYZ, effects of
`alprazolam alone were of longer duration than those of
`
`9
`
`

`

`246 Greenblatt
`
`et al.
`
`CLINICAL
`
`PHARMACOLOGY
`
`&THERAPEUTICS
`SEPTEMBER 1998
`
`triazolam alone (trial C). Agonist effects of alprazolam
`also were enhanced by coadministration of ketocona-
`zole (trial B), although to a lesser degree than for tria-
`zolam. Since alprazolam undergoes
`relatively
`little
`presystemic extraction after oral administration,g~lu
`plasma alprazolam concentrations during the first few
`hours after administration
`were only modestly
`increased by ketoconazole. Kinetic-dynamic
`relation-
`ships for alprazolam did not clearly support a shift in
`configuration attributable to ketoconazole.
`
`We are grateful
`Liza B. Durol.
`
`for
`
`the assistance
`
`of Su Xiang Duan
`
`and Anna
`
`References
`1. Greenblatt DJ, Wright CE. Clinical pharmacokinetics of
`alprazolam: therapeutic implications. Clin Pharmacokinet
`1993;24:453-71.
`2. Kronbach T, Mathys D, Umeno M, Gonzalez FJ, Meyer
`UA. Oxidation of midazolam and triazolam by human liver
`cytochrome P45OIIIA4. Mol Pharmacol 1989;36:89-96.
`3. von Moltke LL, Greenblatt D