Example 2
The following studies were reported by Greenblatt et al., J. Clin. Psychopharmacol., 2018; 38(4):289-295 (doi: 10.1097/JCP.0000000000000892), which is herein incorporated by reference in its entirety for all purposes.
The antipsychotic agent lurasidone is metabolized by Cytochrome P450-3A (CYP3A) enzymes. Coadministration with strong CYP3A inhibitors (such as ketoconazole, posaconazole, and ritonavir) is contraindicated due to the risk of sedation and movement disorders from high levels of lurasidone. This study evaluated the time-course of recovery from the posaconazole drug interaction, and the effect of obesity on the recovery process.
With posaconazole coadministration, lurasidone area under the concentration curve (AUC) increased by an arithmetic mean factor of 6.2 in normals, and by 4.9 in obese subjects. Post-treatment washout of posaconazole was slow in normals (mean half-life 31 hours), and further prolonged in obese subjects (53 hours). Recovery of lurasidone AUC toward baseline was correspondingly slow, and was incomplete. AUC remained significantly elevated above baseline both in normals (factor of 2.1) and obese subjects (factor of 3.4) even at 2 weeks after stopping posaconazole.
Product labeling does not address the necessary delay after discontinuation of a strong CYP3A inhibitor before lurasidone can be safely administered. It is recommended that normal-weight and obese patients be required to limit the dosage of lurasidone, or undergo a washout period after discontinuation of posaconazole, as set forth in the present disclosure.
Methods. Study Site and Institutional Review Board. The study was conducted at Avail Clinical Research, located in DeLand, FL. The study protocol and consent document were reviewed and approved by IntegReview, Austin, TX. All study participants provided written informed consent prior to initiation of any study procedures. In addition, this study was performed in accordance with the Declaration of Helsinki, International Conference on Harmonization Good Clinical Practice guidelines, and applicable regulatory requirements.
Subjects. The study participants consisted of two cohorts, with a total of 34 subjects receiving at least one dose of study drug, and a total of 24 subjects completing the entire study with evaluable pharmacokinetic data. In the first cohort were those of normal body habitus (n=11 completed; BMI 18.5-24.9 kg/m2, inclusive); the second group consisted of subjects of obese body habitus (n=13 completed; BMI ≥35 kg/m2). Subjects were previously known to the research center, or were recruited through notices in the public media. Subjects were matched by gender and age when possible. Sample sizes were based on power calculations.
Potential participants underwent screening and evaluation within 30 days of study initiation. Procedures included medical and psychiatric history, physical examination, electrocardiogram if indicated, hematologic and biochemical screening (including liver function tests such as alanine transaminase, asparagine transaminase, and bilirubin), and urine testing for drugs of abuse. All study participants were healthy, active, non-smoking adults with no history of significant medical or psychiatric disease and taking no prescription medications. Obese subjects were free of metabolic or other complications of obesity. Potentially child-bearing women in both groups had negative pregnancy tests and agreed to avoid the risk of pregnancy during the course of the study. Subjects were instructed to avoid alcohol use throughout the course of the study and underwent a breath alcohol analysis prior to initiation of the study protocol.
Subjects' waist circumference was measured manually. Percent android fat for all subjects was determined by dual energy X-ray absorptiometry (DXA). For three subjects whose weight exceeded the limits of the DXA instrumentation, percent android fat was imputed using population data available from the National Health and Nutrition Evaluation Survey (NHANES). Total android fat (termed total body fat) was calculated as the product of body weight and percent android fat. Ideal body weight (IBW) was determined from actuarial data based on height and gender, and percent ideal body weight calculated as the ratio of actual weight divided by IBW.
Procedures. Subjects received lurasidone (20 mg tablet) on the mornings of study Days 1, 14, 20, 23, 26, and 30. Lurasidone doses were given immediately prior to a continental breakfast provided in the clinical research unit. Venous blood samples were drawn into ethylenediaminetetraacetic acid (EDTA)-containing tubes from an indwelling catheter, or by separate venipuncture, prior to the lurasidone dose and at 1, 2, 3, 4, 8, 12, 18, 24, 48, and 72 hours post-dose. Samples were centrifuged and the plasma was separated and frozen at −70° C. until the time of assay.
On study Day 4, subjects received two doses of posaconazole (300 mg BID). On the mornings of Days 5-17, they received posaconazole 300 mg once daily. As posaconazole is to be taken with food, subjects were fed a continental breakfast in the clinical research unit after receiving posaconazole and prior to discharge from the unit. Venous blood samples were drawn into EDTA containing tubes prior to the posaconazole dose on Days 4, 7, 11, and prior to the lurasidone dose on Days 14, 20, 23, 26 and 30. An additional blood sample was taken 5 hours after posaconazole dosage on Day 17, for approximate determination of maximum posaconazole plasma concentrations, and on Day 33. Samples were centrifuged and the plasma was separated and frozen at −70° C. until the time of assay.
Analytic Methods. All bioassay analyses were performed by Keystone Bioanalytical, North Wales, PA. For analysis of posaconazole, the internal standard (posaconazole-D4) was added to the biological samples. Plasma samples were precipitated using formic acid in acetonitrile and isolated using a Phree phospholipid removal tube. An aliquot of the sample was injected onto a high-pressure liquid chromatograph with tandem mass spectrometry triple quadrupole mass spectrometer (SCIEX API-5500). The analytical column was a Unison CK-218, 3 μm particle size HPLC column (50×2 mm) from Imtakt USA (Portland, OR).
The mobile phase consisted of an aqueous component (0.25% formic acid and 10 mM ammonium formate in water) and an organic component (0.1% formic acid in acetonitrile) and was delivered by gradient, with the organic component going from 35% to 100%. The m/z transitions monitored were 701.6>614.4 for posaconazole and 705.6>618.4 for the internal standard. The calibration curve ranged from 1-1000 ng/mL (8 concentrations in duplicate).
For analysis of lurasidone, the internal standard (lurasidone-D8) was added to the biological samples. Plasma samples were isolated using a Phree phospholipid removal tube. An aliquot of the sample was injected onto a high-pressure liquid chromatograph with tandem mass spectrometry triple quadrupole mass spectrometer (SCIEX API-5500). The analytical column was a Unison UK-C18, 3 μm particle size HPLC column (50×2 mm) from Imtakt USA (Portland, OR). The mobile phase consisted of an aqueous component (0.025% formic acid and 10 mM ammonium formate in water) and an organic component (0.1% formic acid in acetonitrile) and was delivered by gradient, with the organic component going from 35% to 100%. The m/z transitions monitored were 493.4>166.1 for lurasidone and 501.4>166.1 for the internal standard. The calibration curve ranged from 0.25-200 ng/mL (8 concentrations in duplicate).
Pharmacokinetic and Statistical Methods. For each subject, pre-dose plasma posaconazole concentrations on study Days 14 and 17 were averaged, and used as a steady-state concentration (Css) to calculate apparent steady-state clearance of posaconazole according to the relation: Clearance=(dosing rate)/Css. The apparent washout half-life of posaconazole was calculated by log-linear regression analysis starting with the plasma concentration on Day 20 and ending with the last non-zero value. Differences between normal-weight and obese cohorts were evaluated by Student's t-test for independent groups. The relation between measures of body habitus and posaconazole washout half-life for individual subjects was evaluated by linear regression analysis.
For each lurasidone trial for each subject, the terminal log-linear phase of the plasma concentration curve was identified visually, and the terminal rate constant (beta) was determined by log-linear regression analysis. This was used to calculate the elimination half-life. Area under the plasma concentration curve from time zero until the last non-zero point was determined by the linear trapezoidal method. To this was added the residual area, calculated as the final non-zero concentration divided by beta, yielding the total area under the plasma concentration curve extrapolated to infinity (AUC). Also tabulated was the observed maximum plasma concentration (Cmax). AUC and Cmax both were adjusted, where necessary, for non-zero baseline (pre-dose) concentrations measured in some subjects on the Day 20, 23, 26, and 30 trials.
Variables were aggregated as arithmetic mean and SD or SE. Lurasidone Cmax and AUC were also aggregated as geometric mean and 90% confidence interval (90% CI). Differences in kinetic variables between study Day 1 and Days 14, 20, 23, 26, and 30 (control vs after posaconazole administration) were evaluated either from the untransformed data using Dunnett's t-test, or by comparison of geometric means and the 90% CI of the difference.
The relation between lurasidone AUC and plasma posaconazole concentration for individual subjects across the 5 DDI trials (Days 14, 20, 23, 26, and 30) was analyzed by nonlinear regression (SAS PROC NLIN). The following function was fitted to data points:
Y=Y0+B XA where Y is the lurasidone AUC value corresponding to X, the plasma posaconazole concentration at the start of relevant AUC measurement period. Iterated variables were: Y0, A, and B.
Results
Subject Characteristics. Screening procedures yielded 34 subjects who were potential study participants. Of these, 8 initiated participation but did not complete the study for personal or administrative reasons not related to the study or study medications. Data from 2 other subjects could not be analyzed due to apparent protocol deviations. A total of 24 subjects (11 normal-weight and 13 obese) completed the study and were included in the pharmacokinetic analysis (Table 5). The groups were comparable in age, gender composition, height, and IBW. The obese group had significantly higher values of weight, percent IBW, BMI, waist circumference, percent android fat, and total body (android) fat (Table 5). The mean weight in the obese group s 140 kg (309 pounds), and the mean BMI was 49.3 kg.
TABLE 5
DEMOGRAPHIC CHARACTERISTICS OF STUDY PARTICIPANTS
Independent
t-test: Normal
Normal-weight*Obese*vs obese
Number1113
Age (years)34 ± 8 33 ± 7 N. S.
Male/female6/56/7
Weight
(Kg)67.9 ± 9.1 140.4 ± 32 P < 0.001
(Pounds)149 ± 29 309 ± 70 P < 0.001
Height
(Cm)171 ± 10 168 ± 11 N. S.
(Inches)67.3 ± 4.0 66.3 ± 4.3 N. S.
BMI (kg/m2)23.1 ± 1.8 49.3 ± 9.6 P < 0.001
Waist circumference
(Cm)80.4 ± 6.8 129.3 ± 22.4 P < 0.001
(Inches)31.7 ± 2.7 50.9 ± 8.8 P < 0.001
Ideal body weight (kg)64.5 ± 12.361.9 ± 11.4N. S.
Percent ideal body weight106 ± 11 230 ± 46 P < 0.001
Percent android fat33 ± 1266 ± 4 P < 0.001
Total body fat (kg)22.5 ± 8.0 81.3 ± 25.8P < 0.001
*Mean ± SD
Adverse Events. Five subjects experienced adverse events considered possibly or probably related to one or both study medications. These were gastrointestinal disturbances in two cases, and one each of dry mouth, somnolence, and headache. All resolved without specific treatment.
Posaconazole Pharmacokinetics. Plasma posaconazole concentrations had reached steady-state by study Day 14 (FIG. 1). Mean Css was significantly lower, and posaconazole clearance was significantly higher, in the obese cohort compared to controls (Table 6). However, weight-normalized posaconazole clearance was not significantly different between the groups.
Washout of posaconazole after discontinuation of treatment was significantly slower in the obese group compared to controls (P<0.005) (FIG. 1). Mean washout half-life values in the two groups were 2.19 days (52.5 hours) and 1.28 days (31 hours), respectively (Table 6).
Among all subjects, the correlation between posaconazole washout half-life and each of the measures of body habitus was statistically significant, but the degree of obesity explained only a small fraction of variance in washout half-life (r2<0.32). The attenuated associations were in part attributable to two obese subjects with very long half-life values (121 hours).
TABLE 6
POSACONAZOLE PHARMACOKINETICS
Mean ± SD Value of
value for Group:Student’s t:
NormalObeseNormal vs Obese
Steady-state2377 ± 11881462 ± 649 3.33 (P < 0.005)
concentration (ng/mL)
Steady-state clearance
mL/min101 ± 71 175 ± 91 2.19 (P < 0.04)
mL/min/kg1.48 ± 1.021.25 ± 0.61N. S.
Washout half-life (hours) 31 ± 6.752.5 ± 31.12.25 (P <0.04)
Lurasidone Pharmacokinetics. Coadministration of lurasidone with posaconazole resulted in a highly significant increase in lurasidone Cmax and AUC (FIG. 2, Table 7). Comparing Day 14 values to the Day 1 pre-posaconazole values based on ratio of geometric means, Cmax increased by a factor of 4.0 in normal-weight subjects and by 2.9 in the obese subjects. Corresponding increases in AUC were greater than increases in Cmax. Geometric mean AUC increased by a factor of 5.75 in the normal-weight cohort, and by 4.34 in the obese cohort (Table 7). When calculated as arithmetic mean ratios, values were 6.2 in controls and 4.9 in obese subjects.
TABLE 7
SUMMARY OF LURASIDONE PHARMACOKINETICS
Arithmetic mean ± Geometric meanRatio of geometric means
standard error(90% CI)(RGM) vs Day 1 (90% CI)
Corrected Cmax (ng/mL)Corrected Cmax (ng/mL)Corrected Cmax
NormalObeseNormalObeseNormalObese
Day 117.1 ± 1.619.8 ± 416.3(13.5-19.6)15.1(10.2-22.6)Day 144.00(3.09-5.19)2.91(1.89-4.47)
Day 1469.4 ± 8.3*47.0 ± 5*65.2(53.5-79.5)44.1(35.9-54.2)Day 202.98(2.06-4.33)2.42(1.55-3.76)
Day 2055.9 ± 7.840.0 ± 5*48.6(34.6-68.4)36.6(29-46.3)Day 232.32(1.68-3.21)1.85(1.2-2.84)
Day 2342.5 ± 6.3*30.0 ± 337.8(28.5-50.2)28.0(22.7-34.6)Day 261.63(1.1-2.4)1.78(1.13-2.79)
Day 2632.2 ± 6.630.0 ± 426.5(18.3-30.9)26.9(21-34.5)Day 301.47(1.09-1.99)1.42(0.89-2.26)
Day 3026.2 ± 3.225.0 ± 4.424.0(18.6-30.9)21.6(16.4-28.4)
Total AUC (ng/mL × hr)Total AUC (ng/mL × hr)Total AUC (ng/mL × hr)
NormalObeseNormalObeseNormalObese
Day 157.9 ± 5.850.8 ± 954.5(43.3-68.6)42.0(30.4-57.9)Day 145.94(4.64-7.46)4.66(3.28-6.59)
Day 14 333 ± 24* 205 ± 19*324(282-372)195(166-230)Day 204.34(3-6.28)4.90(3.45-6.96)
Day 20 265 ± 27* 217 ± 20*237(175-321)205(173-244)Day 233.38(2.39-4.78)3.82(2.68-5.47)
Day 23 204 ± 27* 170 ± 17*184(139-242)160(133-193)Day 262.24(1.45-3.46)3.33(2.3-4.83)
Day 26 148 ± 27* 152 ± 19*122(83-179)140(113-173)Day 302.10(1.46-3.01)3.34(2.33-4.78)
Day 30 129 ± 20* 150 ± 17*114(85-154)140(116-170)
*P < 0.05 compared to Day 1 value, Dunnett’s t test
Kinetic variables for lurasidone recovered toward the pre-posaconazole baseline values during the posaconazole washout period. Based on ratios of geometric mean values versus the Day 1 baseline, Cmax remained elevated above Day 1 even on Day 30 (ratio=1.47, 90% CI=1.09-1.99) in the normal-weight control subjects. In the obese cohort, Cmax remained above baseline up to Day 26. Recovery of AUC in both groups was even less complete, with Day 30 ratios of 1.9 in the normal-weight group and 2.8 in the obese subjects (arithmetic mean ratios: 2.1 and 3.4, respectively). Consistent with the slower washout of posaconazole in the obese group, the rate of recovery of lurasidone AUC toward baseline values was correspondingly slower in the obese cohort compared to controls (FIG. 3).
Baseline values of lurasidone elimination half-life averaged 9.4 hours in normal-weight subjects and 10.9 hours in the obese group. These values are in the range of what has been reported previously. The half-life values were significantly prolonged during and after administration of posaconazole, and were still substantially longer than baseline values even on the Day 30 trial (FIG. 2, Table 8). Mean half-life values were longer in obese subjects compared to controls. However, half-life determinations were complicated by estimates that exceeded the sampling duration in some subjects.
TABLE 8
LURASIDONE ELIMINATION HALF-LIFE (HOURS)
Arithmetic mean ± S.E.
NormalObese
Day 19.4 ± 1.510.9 ± 4
Day 1437 ± 4*38 ± 2*
Day 2039 ± 3*48 ± 4*
Day 2348 ± 5*52 ± 3*
Day 2650 ± 7*61 ± 4*
Day 3045 ± 9*71 ± 5*
*P < 0.05 compared to Day 1 based on Dunnett’s t test
Relation of Plasma Posaconazole to Lurasidone AUC. Based on analysis of data from all subjects, individual variations in plasma posaconazole concentrations accounted for 66% of the variance in lurasidone AUC at the corresponding times (r2=0.66), indicating that posaconazole exposure is a principal determinant of the magnitude of the posaconazole-lurasidone DDI (FIG. 4).
Discussion. The present study evaluated the pharmacokinetic DDI between lurasidone as victim (substrate) and the strong CYP3A inhibitor posaconazole as perpetrator (precipitant), both in volunteers of normal body weight and in an otherwise healthy group of subjects with BMI ≥35 kg/m2. A particular focus of the study was the time-course of recovery from the DDI during the two weeks after discontinuation of posaconazole.
Coadministration of lurasidone with typical doses of posaconazole resulted in increased lurasidone exposure (total AUC) by a factor averaging in the range of 4 to 6 in both groups of subjects. After posaconazole was discontinued, the effect on lurasidone exposure did not return quickly to baseline. Rather, the DDI persisted for at least 2 weeks after the last dose of posaconazole, and probably well beyond the study duration. The slow recovery from the DDI was consistent with the long elimination half-life of posaconazole. With all data aggregated, plasma posaconazole concentration accounted for 66% of the variability in lurasidone AUC associated with the DDI.
The pharmacokinetic properties of posaconazole were significantly modified in the cohort of obese subjects compared to those of normal body size. The clearance of posaconazole—not corrected for body weight—was higher in obese subjects compared to controls, resulting in lower values of Css when the same daily dosage was administered to both groups. Despite the higher clearance, the washout half-life was significantly prolonged in the obese subjects compared to controls. This is likely explained by the disproportionate distribution of the lipophilic drug posaconazole into excess adipose tissue, thereby causing a prolongation of elimination half-life. As a result of the longer half-life and persistence of posaconazole in blood, the duration of the lurasidone DDI was correspondingly longer. At two weeks after the last dose of posaconazole, lurasidone AUC was still elevated above baseline by a mean factor of 3.3 in the obese subject group.
This study involved a relatively small number of subjects, but the findings were statistically robust. Although lurasidone was administered as single test doses, the kinetics of lurasidone are linear, and single-dose kinetic properties will be predictive of behavior during multiple dosing as is customary in the treatment of schizophrenia.
Conclusions. The posaconazole-lurasidone DDI persists long after posaconazole is discontinued, resulting in a sustained risk of a potentially hazardous DDI. The duration of persistent risk is further prolonged in obese individuals due to the effect of obesity on the elimination kinetics of posaconazole. Revision of product labeling is needed to assure patient safety. Based on the findings of this study, it is recommended to require normal-weight and obese patients to limit the dosage of lurasidone, or undergo a washout period, as set forth in the present disclosure.