The effect of lersivirine, a next-generation NNRTI, on the pharmacokinetics of midazolam and oral contraceptives in healthy subjects
John Davis, Grant Langdon, Gary Layton, Chew Lan Chong, Marie-Noella Ndongo, Manoli Vourvahis*
*Manoli Vourvahis, PharmD, Pfizer Global Research and Development, 219 East 42nd Street, New York, New York 10017, USA. Tel: +1 212 733 8517. Fax: +1 212 416 7663. E-mail:
Supplementary material
The assay precision in Study 1 for midazolam was ≤14.3% coefficient of variance (%CV), and for Study 2 was 3.6–5.4 % CV for ethinylestradiol and 4.1–4.7 % CV for levonorgestrel. The between-day assay accuracy, CV % of the estimated QC samples was ≤7.9% for lersivirine. For midazolam and ethinylestradiol, the calibration range was 0.05–50.0 ng/mL; for levonorgestrel the calibration range was 0.10–20.0 ng/mL; for lersivirine the calibration range was 1.00–2000 ng/mL.
For Study 1, the equation used to calculate the area under the plasma concentration time curve from 0 to infinity (AUCinf was AUCinf=AUClast+(C*last/Kel). AUC from zero to the time of the last measurable concentration (AUClast) used a linear trapezoidal method. Cmax and time to maximal plasma concentration (Tmax) were determined by observation. Terminal elimination half-life (t½) was calculated using t½=ln2/Kel. The terminal elimination rate constant (Kel) was not reported in the study but was determined using linear regression of log-linear concentration-time curve. Total oral clearance (CL/F) was determined using CL/F=Dose/AUC.
A linearmodel was fitted for the relationship between the natural log-transformed AUCinf on Day 15/Day 1 ratios and total daily dose (continuous) of lersivirine including the placebo data. On the natural log scale, the effect of each lersivirine dose was predicted from the model with the predicted placebo value subtracted.Hence the relationship between the AUCinf on Day 15/Day 1 ratios and total daily dose of lersivirine (relative to placebo) becomes non-linear when values are back-transformed. A simplified model was selected as the data did not support elucidation with a full maximum effectiveness (Emax) model. The estimated effect of lersivirine at total daily doses of 250 mg up to 2000mg on the midazolam AUCinf Day 15/Day 1 ratio was obtained, along with 90% confidence intervals (CIs).
For Study 2, natural log-transformed data were analyzed using a mixed-effect model with sequence, period, and treatment as fixed effects and subject within sequence as a random effect. Estimates of the adjusted mean differences (Test/Reference) and corresponding 90% CIs were obtained from the model. The adjusted mean differences and 90% CIs for the differences were exponentiated to provide estimates of the ratio of adjusted geometric means (Test/Reference) and 90%CIs for the ratios. Both geometric (raw) and adjusted geometric (adjusted with respect to the covariates included in the model) means are reported. Descriptive PK parameters for lersivirine were also calculated for Day 10.
OC plus placebo was the Reference treatment and OC plus lersivirine was the Test treatment. Additionally, a post hoc analysis for the trough plasma concentration (Cmin) was performed using a similar mixed-effect model. Cmax was determined by observation and AUC over a 24-hour period was determined using the log-linear trapezoidal method.
For Study 1, a sample size of six subjects per dose (250–600 mg BID) was required to provide 90% CIs for the difference between treatments of no more than ±0.309 on the natural log scale with 80% coverage probability. The common standard deviation of 0.22 for AUCinf used in these calculations was the median value observed from sixprevious sponsor studies (unpublished data). When a suitable dose-response model was fitted to the data, the predicted effect at each dose compared with placebo was based on the model and the CI was, therefore, expected to be narrower.
For Study 2, a sample size of 12 subjects was required to provide 90% CIs for the difference between treatments of no more than 0.118 on the natural log scale with 80% coverage probability. This estimate was based on the estimate of the standard deviation of within-subject differences for the maximum observed plasma concentration (Cmax) for ethinylestradiol (0.195 on the natural logscale), obtained from a previous study in which the ethinylestradiol Cmaxwas the endpoint with the greatest observed variability (unpublished data). To allow for subjects who failed to complete the study, 14 subjects were recruited.
Tables
Supplementary Table 1Subject demographics and baseline characteristics
Study 1 / Study 2Midazolam 7.5 mg QD + / Lersivirine + OC QD
/lersivirine + placebo QD
(n=14)
Placebo / Lersivirine BID
Mean (SD) /
(n=3) / 250 mg
(n=6) / 400 mg
(n=6) / 600 mg
(n=6) / 1000 mg
(n=3)
Age, years / 47 (4.0) / 36 (8.7) / 37 (10.2) / 41 (10.1) / 38 (16.6) / 45.8 (9.0)
Weight, kg / 82.7 (5.9) / 77.8 (6.9) / 78.0 (9.1) / 81.8 (12.2) / 76.0 (12.1) / 58.7 (6.3)
BMI, kg/m2 / 24.8 (0.6) / 24.9 (1.8) / 24.0 (2.7) / 25.1 (3.7) / 26.4 (4.0) / 23.9 (2.5)
Height, cm / 182.7 (5.5) / 177.0 (5.9) / 180.3 (5.1) / 180.8 (4.0) / 170.0 (8.0) / 157.1 (3.5)
BID, twice daily; BMI, body mass index, calculated as weight (kg)/(height [m])2; OC, oral contraceptive; QD, once daily; SD, standard deviation
Supplementary Table 2 Summary of midazolam pharmacokinetics on Day 1 and Day 15, and estimated effect of lersivirine on midazolam AUCinf Day 15/Day 1 ratio using the linear dose-response model, relative to placebo (Study 1)
Day / Placebo BID(n=3) / Lersivirine BID
250 mg
(n=6) / 400 mg
(n=6) / 600 mg
(n=6) / 1000 mg
(n=3)
AUCinf, ng.h/mLa (CV%) / 1 / 97.0 (29.1) / 145.7 (54.3) / 102.0 (47.3) / 95.8 (39.6) / 109.0 (31.3)
15 / 130.2 (9.7) / 128.0 (38.3) / 80.3 (47.3) / 69.0 (26.9) / 53.0 (28.1)
AUClast, ng.h/mLa (CV%) / 1 / 94.3 (29.3) / 143.1 (54.5) / 100.7 (47.2) / 93.8 (39.0) / 104.5 (29.5)
15 / 126.6 (9.1) / 125.2 (37.9) / 79.1 (46.8) / 67.8 (27.2) / 51.8 (30.0)
CL/F, L/ha (CV%) / 1 / 77.2 (31.8) / 51.5 (43.1) / 73.5 (51.3) / 78.2 (30.6) / 68.8 (36.1)
15 / 57.6 (10.0) / 58.6 (48.1) / 93.3 (43.4) / 108.6 (25.3) / 141.4 (28.3)
Cmax, ng/mLa (CV%) / 1 / 35.7 (15.3) / 55.4 (27.7) / 42.1 (43.8) / 38.3 (33.4) / 44.6 (13.9)
15 / 43.4 (19.7) / 43.0 (34.8) / 31.8 (57.5) / 23.1 (27.8) / 27.5 (44.3)
Tmax, hb (range) / 1 / 1.00 (1.0–1.0) / 0.75 (0.5–1.5) / 0.50 (0.5–1.0) / 0.75 (0.5–1.0) / 1.00 (0.5–1.0)
15 / 1.00 (0.5–1.0) / 0.75 (0.5–1.5) / 0.75 (0.5–2.0) / 1.50 (0.5–3.0) / 0.50 (0.5–1.0)
t½, hc (SD) / 1 / 5.9 (2.0) / 4.9 (0.5) / 4.2 (1.3) / 5.0 (1.0) / 6.9 (2.0)
15 / 5.9 (1.5) / 5.3 (1.1) / 4.4 (2.0) / 5.0 (1.1) / 6.1 (1.8)
Total daily dose (mg) of lersivirine / Estimated AUCinf Day 15/Day 1 ratio of midazolam (%)d / 90% CI for ratio (%)
250 / 89 / 85–94
500 / 80 / 72–88
750 / 71 / 62–82
1000 / 64 / 52–77
1500 / 51 / 38–68
2000 / 40 / 27–59
AUCinf, area under the plasma concentration time curve from 0 to infinity; CI, confidence interval; CV %, coefficient of variance; AUClast, area under the plasma concentration time curve from 0 to the time of last measurable concentration; BID, twice daily; CL/F, total oral clearance; Cmax, maximum observed plasma concentration; Tmax, time to first occurrence of Cmax; t½,terminal elimination half-life; SD, standard deviation; ageometric mean; bmedian; carithmetic mean; dthe AUCinf Day 15/Day 1 ratio of midazolam was natural log-transformed prior to analysis
Supplementary Table 3 Effect of lersivirine on the pharmacokinetics of oral contraceptives on Day 10 (Study 2)
levonorgestrel150 µg QD)
+ lersivirine (1000 mg QD)
/ OC (ethinylestradiol30 µg/
levonorgestrel150 µg QD)
+ placebo
/ Ratio of adjusted
geometric means / 90% CI
for ratio
Geometric mean (CV%)
(n=131) / Adjusted geometric mean
(n=14) / Geometric mean (CV%)
(n=131) / Adjusted geometric mean
(n=131)
Ethinylestradiol
AUC24, pg.h/mL / 974 (27) / 978.6 / 905 (33) / 894.1 / 1.10 / 0.92–1.31
Cmax, pg/mL / 94.2 (30) / 94.9 / 88.2 (34) / 88.2 / 1.08 / 0.85–1.36
Cmin, pg/mL / 19.0 (34) / 19.0 / 20.1 (32) / 19.5 / 0.97 / 0.84–1.13
Levonorgestrel
AUC24, ng.h/mL / 63.8 (29) / 63.9 / 76.1 (34) / 73.5 / 0.87 / 0.78–0.97
Cmax, ng/mL / 5.4 (31) / 5.4 / 6.6 (30) / 6.4 / 0.84 / 0.69–1.02
Cmin, ng/mL / 1.5 (40) / 1.5 / 2.0 (42) / 2.0 / 0.79 / 0.69–0.90
AUC24, area under the plasma concentration time curve from 0 to 24 hours; Cmax, maximum observed plasma concentration; Cmin, minimum observed plasma concentration; CI, confidence interval; CV, coefficient of variation; OC, oral contraceptives; QD, once daily
1n=12 for Cmin
Supplementary Table 4 Treatment-emergent (all-causality) adverse events occurring in ≥3 subjects in any treatment group
MedDRA preferred term, n / STUDY 1 / STUDY 2Placebo BID + midazolam (7.5 mg)
(n=3) / Lersivirine BID + midazolam (7.5 mg) / Total / OC (EE 30 µg/LN 150 µg QD) + lersivirine (1000 mg QD)
(n=13) / OC (EE 30 µg/LN 150 µg QD) + placebo
(n=14)
250 mg (n=6) / 400 mg (n=6) / 600 mg (n=6) / 1000 mg (n=3)
Somnolence / 2 / 5 / 6 / 5 / 3 / 21 / 1 / 2
Fatigue / 0 / 1 / 3 / 4 / 2 / 10 / 0 / 0
Flatulence / 0 / 4 / 2 / 0 / 1 / 7 / 0 / 0
Diarrhea / 0 / 1 / 0 / 2 / 3 / 6 / 0 / 1
Nasopharyngitis / 0 / 2 / 0 / 3 / 0 / 5 / 0 / 0
Headache / 0 / 2 / 2 / 2 / 2 / 8 / 7 / 4
Nausea / 0 / 0 / 0 / 0 / 2 / 2 / 8 / 2
Dizziness / 0 / 0 / 0 / 1 / 2 / 3 / 6 / 3
Withdrawal bleed / 0 / 0 / 0 / 0 / 0 / 0 / 3 / 3
Increased appetite / 1 / 0 / 0 / 0 / 0 / 1 / 3 / 1
BID, twicedaily; EE, ethinylestradiol; LN, levonorgestrel; MedDRA, Medical Dictionary for Regulatory Activities; OC, oral contraceptives; QD, once daily