1.Cellular Partitioning and Plasma Protein Binding of RAD001 in Mouse Blood

Supplemental Material

1.Cellular partitioning and plasma protein binding of RAD001 in mouse blood.

Using methods analogous to those described in Laplanche et al (2007), Figure SM1 demonstrates a typical plot of [3H]RAD001 data, from the erythrocyte partitioning experiments, the in vitro blood distribution of [3H]RAD001 in mouse blood is shown in Table SM1, the distribution of RAD001 between erythrocytes and diluted plasma is shown in Table SM2 and the determined binding parameters of [3H]RAD001 in mouse blood are shown in Table SM3.

Table SM1.In vitro blood distribution of [3H]RAD001 in mouse blood

Nominal concentration (ng/mL) / Actual concentration (ng/mL) / Concentration in plasma (ng/mL) / Fplasma % (mean ± SD) / Cb/Cp
5 / 4.09 / 7.01 / 92.7 1.4 / 0.6
50 / 32.8 / 59.6 / 98.1 1.0 / 0.6
100 / 72 / 133 / 99.5 2.5 / 0.5
500 / 399 / 754 / 102.2 2.2 / 0.5
1000 / 740 / 1417 / 103.4 0.4 / 0.5
5000 / 3870 / 6790 / 94.7 1.6 / 0.6

5-5000 ng/mL [3H]RAD001 was incubated in fresh murine blood at 37°C for 5 min followed by centrifugation at 1600g for 2min for plasma separation. The radioactivity present in the plasma sample was determined by liquid scintillation counting.

Table SM2.Distribution of [3H]RAD001 between erythrocytes and diluted plasma

Concentration in:
Plasma dilution (%) / Actual concentration (ng/mL) / Diluted plasma (ng/mL) / Erythrocytes (ng/mL) / Ces/Cdp / Cdp/Ce
0.1 / 8.31 / 3.25 / 28.53 / 2.55 / 0.11
0.5 / 7.94 / 6.13 / 14.74 / 1.30 / 0.42
1.0 / 8.37 / 7.53 / 11.35 / 1.11 / 0.66
2.0 / 8.48 / 9.30 / 5.17 / 0.91 / 1.80
5.0 / 8.69 / 9.64 / 3.04 / 0.85 / 3.59
7.5 / 8.32 / 9.64 / 3.04 / 0.86 / 3.17
10 / 7.99* / 9.69* / 1.19 / 0.82 / 8.16
20 / 7.84* / 9.76* / 0.62 / 0.80 / 8.16
30 / 8.65 / 10.54 / 1.08 / 0.82 / 9.73
40 / 8.62 / 10.57 / 1.71 / 0.82 / 6.19
60 / 8.72 / 10.71 / 1.68 / 0.81 / 6.38

Data are the means of 3 replicate determinations except where noted with an asterisk where n=2
Ce: concentration in erythrocytes
Cdp: concentration in diluted plasma
Ces: concentration in erythrocyte suspension

Erythrocytes from mice were suspended in diluted plasma at 0.1, 0.5, 1, 2, 5, 7.5, 10, 20, 30, 40 and 60% (hematocrit: 0.2). The erythrocyte suspension spiked at 10ng/mL [3H]RAD001 was incubated at 37°C for 5 min followed by centrifugation at 1600g for 2min for the separation of the supernatant from erythrocytes.

Table SM3.Binding parameters of [3H]RAD001 in mouse blood

Estimation of plasma protein binding of [3H]RAD001 in mouse based on the erythrocyte partitioning method.

Terms / Values / Remark
intercept / 0.082 / 1/Ke
slope / 71.8 / Kpp/Ke
Ke / 12.3 / reciprocal
Kpp / 879.2 / slope x Ke
fu* / 0.001 / 1/(Kpp+1)

* fu is an estimate for 100 % undiluted plasma; setting F=1.

Estimation of plasma protein binding of [3H]RAD001 in mouse based on the erythrocyte partitioning method.

Figure SM1.Plot of [3H]RAD001 data from the erythrocyte partitioning experiments

Converted values of concentrations in erythrocyte suspension and diluted plasma were plotted versus plasma dilution (F). The dotted line represents the best fit linear regression line using measurements with F ranging from 0.1 to 5%.

Reference. Laplanche R, Meno-Tetang GML, Kawai R, (2007) Physiologically based pharmacokinetic (PBPK) modeling of everolimus (RAD001) in rats involving non-linear tissue uptake. J Pharmacokinet Pharmacodynam 34:373-400

2.Determination of RAD001 concentrations in tissues.

Tissue sample preparation. Tissue samples were thawed, weighed and an equal weight of PBS buffer, (1:1 to 1:3 dilution depending on tissue type) was added. The tissues were then homogenized using an Ultra-Turrax TP18-10 homogenizer. A 0.15 mL aliquot of tissue homogenate was spiked with the calibration standard, quality control (QC) compound and an internal standard (IS). To this mixture 0.65 mL of the precipitation reagent (0.4 M zinc sulfate:methanol at a ratio of 4:1) was added. This mixture was vortexed 5 mins then centrifuged for 10 mins. The supernatant was loaded onto a Strata x 33U 8E-S100-AGB SPE plate (from Phenomenex), previously conditioned with methanol and ammonium hydroxide buffer pH 8.5). The plate was washed with ammonium hydroxide buffer methanol solution (97.5 / 2.5, v/v). The analyte of interest was eluted from the plate with two 0.35ml aliquots of methylene chloride, evaporated to dryness and reconstituted in 75 µl of mobile phase.

Instrumentation and LC-MS-MS condition. The HPLC system consisted of a LC-10AD pump and a SCL-10AVP controller from Shimadzu (Columbia, MD, USA). An on-line solvent degasser (Metachem, Torrance, CA, USA), a column oven (Analytical Sales, Mahwah, NJ, USA) and a HTS PAL autosampler from Leap Technologies (Carrboro, NC, USA). The chromatography column was Phenomenex, luna C18(2), (30 x 2mm, 3µm particle size) from Keystone Scientific (Bellefonte, PA, USA). The mass analysis was performed on a Quattro Ultima MS (Micromass, Manchester, England) equipped with an Electrospray ionization (ESI) interface and operated in the positive ion mode. The isocratic mobile phase was composed of 80% acetonitrile / 20% 20 mM aqueous ammonium acetate (pH 6.5) (v/v). The LC separation was performed at 45ºC at a flow rate of 0.2 mL/min. The total run time was 2.5 min.

The following mass spectrometric conditions were used: High purity nitrogen (100 psi) was used as the nebulization and desolvation gas, and the capillary and cone voltages were set at 2.75 and 30 V, respectively. The source block and desolvation temperatures were 150 and 300 ˚C, respectively. Argon (50 psi) was used as the collision gas. The collision energy was set at 15 eV with a dwell time of 0.3 s for RAD001. The samples were analyzed via multiple reaction monitoring (MRM) employing the transitions of the ammonium adduct to the parent / daughter ions. The MRM transitions for RAD001 and internal standard (SDZ 223-756) are 975.5 → 908.5 and 989.5 → 922.5, respectively.

The use of ESI-MS/MS in combination with monitoring of the ammonium adduct of the parent ions gave good sensitivity and specificity for RAD001 in all the matrices analysed. Under these conditions a lower limit of quantification (LLOQ) of 0.5 ng/g was achieved. The specificity and selectivity of the method was investigated by preparing and analyzing blank tissue under investigation. No interference was observed in any of the naïve tissue samples analysed.

Calibration curves. The calibration curves (y = ax² + b·x + c) were generated by weighted (1/x2) least-squares regression of the peak area ratios (y) of the analytes to their internal standards versus the concentrations (x) of the calibration standards. The calibration curve for each matrix covered the dynamic range 0.5 to 2500 ng/g, and linear correlation coefficient of > 0.99 was achieved for each matrix analysed.

3.Determination of [3H]RAD001 and metabolites by Liquid Chromatography – Reverse Isotope Dilution (LC-RID)

Stock solutions and calibration curves. Stock solution A (12.5 mg unlabelled RAD001 dissolved in 125 mL acetonitrile). Stock B ([3H]RAD001 in toluene) was diluted with stock solution A to give 107 dpm/mL. Calibration curves were prepared by diluting aliquots of Stock B with Stock A to give a range of dpm/mL values (which were confirmed by scintillation counting). For determining RAD001 metabolites, a similar protocol was used, but peaks other than the characteristic RAD001 peak which contained radioactivity were analysed.

Sample extraction. Aliquots (0.2 to 0.5 mL) of blood, plasma or tissue were diluted to 250 mL with HPLC-grade water. Homogenates (10 to 20 % w/v in water) were added to a glass tube, diluted with 200 µL of stock solution A, then 500 µL of Tritisol buffer (Merck, pH 9), followed by 4 mL of diethyl ether. The samples were shaken for 10 min, centrifuged (3000 x g, 10 mins) and the supernantant transferred to a fresh glass tube and evaporated to dryness in a vacuum centrifuge (Univapo, model 150H, Zivy, Basel). The residue was dissolved in 250 µL mobile phase and 100 µL n-hexane added and the immiscible mixture vortexed for 30 sec. Following centrifugation (3000 x g, 1 min), the top hexane layer was removed and 200 µL of the lower aqueous phase used for HPLC analysis.

HPLC system. A Kontron 450-MT2 HPLC apparatus was used. Samples were injected onto a Browlee Spheri-5 RP2 column (220 x 4.6 mm) fitted with a precolumn which was eluted with linear gradient of A) water/methanol/tetramethylammonium-hydrogen sulfate (TMaHs, 10g/L) at 50/50/1 v/v/v, and B) methanol/TMaHs 100/1 v/v. A linear gradient (0% B to 50% B) within 15 mins at 1.2 mL/min and the column washed with solution A for 5 mins. The solutes were detected at 278 nm and the peaks corresponding to RAD001 (or metabolites) were collected and the dpm determined by liquid scintillation counting. The external standard curve was linear, and the ratio of measured/expected had a mean of 0.724 (CV = 1.83%), the precision of the assay was good (CV 0.6 to 3.9%), and recovery of the quality control samples demonstrated a mean absolute deviation from the expected concentrations was 5.9%.

Calculations. Based upon the principle of reverse isotope dilution, As/Acs = Aad/Acd or As = Aad x Acs/Acd where As is the amount of analyte in the sample (unknown), Acs is the amount of carrier added, Aad is the amount of analyte detected and Acd is the amount of carrier detected. The amount of analyte detected was calculated as Aad = Rp/SR, where Rp is the amount of radioactivity in the collected peak fraction, and SR is the specific radioactivity and the ratio is the mean of the ratios of radioactivity in the peak fraction/expected radioactivity. The amount of carrier detected was calculated as Acd = Area/RF, where RF is the response factor.

4.Additional pharmacokinetic and pharmacodynamic data

Figure SM2.Pharmacokinetics of RAD001 in non-tumor and tumor bearing BALB/c nu/nu mice

A. Non-tumor bearing mice / B. Tumor-bearing mice

Female BALB/c athymic nude mice were used, either with (Panel A) or without tumors. Mice were given a single p.o. administration of RAD001 at 5 mg/Kg. Plasma and tissues were obtained at various time points following administration, and RAD001 levels determined using HPLC/MS. Data are expressed as mean ± sem, n = 4. Horizontal lines represent the in vitro IC50 values which were taken from (Tanaka et al [2008]). Identifying optimal biologic doses of everolimus (RAD001) in patients with cancer based on the modeling of preclinical and clinical pharmacokinetic and pharmacodynamic data.

22.Tanaka C, O'Reilly T, Kovarik JM, et al (2008). Identifying optimal biologic doses of everolimus (RAD001) in patients with cancer based on the modeling of preclinical and clinical pharmacokinetic and pharmacodynamic data. J Clin Oncol. 26:1596-602.

Figure SM3.Kp values for RAD001 in KB-31 tumor-bearing nude mice (Tissue penetration as a ratio of blood levels)

Tissue penetration as a function of time post RAD001 administration
Kp Tumor / Kp Skin / Kp Muscle
Tissue penetration as related to plasma RAD001 levels during the elimination phase

KB-31 tumor-bearing nude mice were treated once with 5 mg/Kg RAD001, p.o.. At the indicated times, blood and tissues were removed and the concentration or RAD001 determined. Kp values were based upon Ctissue/Cblood.

Table SM4RAD001 PK parameters estimated from modeling

BALB/c nude mouse / Lewis rat
Regimen / PK parameter / Plasma (%CV) / Tumor (%CV) / Blood (%CV) / Tumor (%CV)
Non-compartmental model
5 mg/kg, p.o. / Cmax (ng/mL or ng/g) / 2513 / 102 / 212 / 770
Tmax (h) / 0.5 / 2
AUC (ngh/mL or ngh/g) / 19415 / 1639 / 1332 / 15821
Cl/f (mL/h or g/h) / 0.24 / 2.14 / 3.5 / 0.29
Regression r
Correlation coefficient / 0.98 / 0.96 / 0.97 / 0.996
1 Compartmental model
5 mg/kg, p.o. / Cmax
(ng/mL or ng/g) / 2430
(4.2) / 86
(6) / 245
(14) / 595
(13)
Tmax (h) / 0.52
(23) / 3
(15) / 0.34
(16) / 3.3
(29)
AUC
(ngh/mL or ngh/g) / 20534
(9.8) / 2871
(17) / 1316
(14) / 17049
(31)
Cl/f (mL/h or g/h) / 0.24
(9.8) / 1.7
(17) / 3.6
(7) / 0.29
(30)
AIC / 156 / 87 / 157 / 120
SC / 158 / 88 / 159 / 121
Correlation coefficient / 0.99 / 0.96 / 0.99 / 0.90
2 Compartment modelC
5 mg/kg, p.o / Cmax
(ng/mL or ng/g) / 2643
(7) / 93
(9)
Tmax (h) / 0.3
(16) / 3
(28)
. / AUC
(ngh/mL or ngh/g) / 22110
(22) / 3062
(20)
Cl/f (mL/h or g/h) / 0.23
(14) / 1.6
(21)
AIC / 157 / 89
SC / 159 / 91
Correlation coefficient / 0.99 / 0.97

Female KB-31 tumor-bearing nude mice and CA20948 tumor-bearing Lewis rats were administered a single p.o. dose of 5 mg/Kg RAD001. Plasma or whole blood and tumor tissue were obtained at various time points following administration, and RAD001 levels determined using HPLC/MS.

Initial PK values were calculated by non-compartmental analysis of extravascular dosing (WinNonlin) using mean values. Correlation coefficient indicates the relationship between the observed parameters and those predicted by the model. AIC, Akaike Information Criterion, SC, Schwarz Criterion

Figure SM4.PK modeling of RAD001 in mice: Comparison of observed and predicted profiles

A. Plasma / B. Tumor

Plasma used the 1-compartment model and tumor a 2-compartment model(lines) described in Table SM4. Circles represent the measured RAD001 concentrations.

Figure SM5.Predicted PK profiles of RAD001 in plasma and tumor tissue of KB-31 tumor bearing BALB/c nude mice.

Female KB-31 tumor-bearing nude mice were administered a single p.o. dose of 5 mg/Kg RAD001. Whole blood and tissues were obtained at various time points following administration, and RAD001 levels determined using HPLC/MS. These data were then used in PK modeling. Initial PK values were calculated by non-compartmental analysis of extravascular dosing (WinNonlin) using mean values. Subsequently, the data were modeled using a the 1-compartment and 2-compartment models described in Table SM4.

Figure SM6.Pharmacokinetics of RAD001 in rats.

A. CA20948 pancreatic tumor bearing Lewis rats / B. Sprague-Dawley rats

Panel A. CA20948 tumor-bearing Lewis rats received a single administration of 5 mg/Kg RAD001 (microemulsion-formulated) when the umors reached 1000 mm3. At the indicated times, blood and tumor tissue were removed and RAD001 extracted and determined using HPLC-MS/MS. Unbound RAD001 in blood was determined by correcting for the degree of partitioning into erythrocytes (60 %) and protein binding (92 %). Panel B. Tumor-free SD rats were administered a single administration of RAD001 (p.o., emulsion formulated at 1.5 or 15 mg/Kg; or i.v. formulated in Cremophor EL-ethanol [2/1 w/w] at 1 and 10 mg/Kg), whole blood was taken at the indicated times and the RAD001 concentration determined by HPLC-MS/MS.

Figure SM7.RAD001 tissue levels after repeated administration to Wistar rats.

A. RAD001 tissue concentrations / B. Kp values

Wistar rats were administered RAD001 at 0.5 or 5 mg/Kg p.o., once per day for 14 days. Data represent tissue RAD001 levels 24h after the last administration (mean ± SEM, n=4). The limit of detectability was 1.0 ng/mL or 5 ng/g (horizontal lines). The Kp values are based upon the ratio of total tissue RAD001 levels/estimated unbound RAD001 levels in blood. The unbound fractions were estimated to be 0.2 ± 0.03 at 0.5 mg/Kg and 2.0 ± 0.1 at 5 mg/Kg.

Figure SM8.PK modeling of RAD001 in tumor-bearing Lewis rats: Comparison of observed and predicted profiles

A. Plasma / B. Tumor

Modelling of plasma and tumor RAD001 PKused 1-compartment modelsas described in Table SM4. Circles represent the measured RAD001 levels.

Figure SM9.Predicted PK profiles of RAD001 in plasma and tumor tissue of CA pancreatic tumor-bearing Lewis rats.

a

Female CA920495 tumor-bearing Lewis rats were administered a single p.o. dose of 5 mg/Kg RAD001. Whole blood and tissues were obtained at various time points following administration, and RAD001 levels determined using HPLC/MS. These data were then used in PK modeling. Initial PK values were calculated by non-compartmental analysis of extravascular dosing (WinNonlin) using mean values. Subsequently, the data were modeled using the 1-compartment and 2-compartment models described in Table SM4.

Table SM5.Comparison of antitumor activity with RAD001 initial plasma and tumor PK in mice.

Estimated RAD001 PK parameter at ED50 Dose
Unbound Plasma / Total Tumor / Maximal effects
Cell line / In vitro
IC50 (ng/mL) / In vivo ED50
(mg/Kg, p.o., q24h) / Cmax
ng/mL / C24h
ng/mL / AUC
ng h/mL / Cmax
ng/mL / Cmin
ng/mL / AUC
ng h/mL / Dose
mg/Kg / T/C
Based upon mouse PK
A549 / 2.3 / 0.11 / 0.3 / 0.1 / 0.4 / 2.2 / 1.0 / 36 / 2.5 / 0.0004
KB-31 / 1704 / 0.32 ± 0.04 / 0.8 / 0.2 / 1.2 / 7 / 2.9 / 105 / 2.5 / 0.25
B16 / 0.7 / 0.42 / 1.0 / 0.2 / 1.6 / 9 / 3.8 / 138 / 5 / 0.007
H596 / 4.8 / 0.50 / 1.2 / 0.2 / 1.9 / 10 / 4.5 / 164 / 5 / 0.16
H-520 / 40.8 / 0.52 / 1.3 / 0.3 / 2.0 / 11 / 4.7 / 171 / 5 / 0.14
KB-8511 / 1427 / 0.7 ± 0.01 / 1.7 / 0.3 / 2.7 / 14 / 6.3 / 230 / 5 / 0.29
HCT116 / 3953 / 2.38 ± 0.03 / 5.6 / 1.1 / 8.7 / 47 / 20.5 / 748 / 5 / 0.35
Based upon rat PK
A549 / 2.3 / 0.11 / 0.7 / 0.01 / 0.8 / 17 / 0.3 / 355 / N/A / N/A
KB-31 / 1704 / 0.32 / 2.2 / 0.02 / 2.3 / 49 / 1.0 / 1033 / N/A / N/A
B16 / 0.7 / 0.42 / 2.8 / 0.03 / 3.1 / 65 / 1.3 / 1355 / N/A / N/A
H596 / 4.8 / 0.5 / 3.4 / 0.04 / 3.6 / 77 / 1.6 / 1614 / N/A / N/A
H-520 / 40.8 / 0.52 / 3.5 / 0.04 / 3.8 / 80 / 1.6 / 1678 / N/A / N/A
KB-8511 / 1427 / 0.7 / 4.7 / 0.05 / 5.1 / 108 / 2.2 / 2259 / N/A / N/A
HCT116 / 3953 / 2.38 / 15.5 / 0.16 / 16.6 / 351 / 7.1 / 7358 / N/A / N/A
HUVEC
VEGF-driven / 0.12
bFGF-driven / 0.8

In vitro IC50 determinations were performed using monolayer cultures containing various dilutions of RAD001, incubated for 4 d and the cell mass determined by methylene blue staining, with the IC50 being the dilution where RAD001 has inhibited the cell proliferation by 50 % as compared with untreated controls. ED50 values represent the RAD001 dose (p.o., once per day) where the tumor growth is reduced by 50 % as compared with controls and the values were estimated from the dose response curves using the Hill equation. Unbound plasma and total RAD001 tumor levels at the ED50 were extrapolated from the non-compartmental PK data obtained at 5 mg/Kg RAD001 administered p.o., once, assuming linear kinetics over the ED50 dose range.

As shown previously, following a single 5 mg/Kg administration, the blood and tumor RAD001 levels in both mice bearing KB-31 tumors and rats bearing CA 40928 tumors only transientlyexceeded the in vitro anti-proliferative median IC50 during the entire administration cycle (24 hr) but remain well above the IC50 for the HUVEC cells (see Lane et al, 2009). Despite this, tumors considered relatively insensitive to RAD001 in vitro, display at least moderate responses to RAD001 at optimal doses (e.g. T/C of 0.20 to0.40) and can display ED50 values similar to relatively sensitive tumor cell lines. This suggests a constant anti-vascular effect occurs and that direct anti-tumor cell activity may be more or less contributing to the overall anti-tumor effect of RAD001.

As shown in the Table, at ED50 doses of RAD001, the predicted tumor Cmax levels in mice can reach the in vitro IC50 with some cell lines (e.g. B16/B16 murine melanoma), but generally do not. Even when considering drug predicted accumulation in tumor (~ 2-fold). This would suggest that the anti-tumor effect of RAD001 when administered to mice is largely due to the antiangogensis effect. However, with the estimates of unbound RAD001 in blood and total levels in tumor tissue based upon rat PK, Cmax values can be higher than the IC50 level, especially when considering predicted drug accumulation. Cautiously extending these data would suggest that in rat tumors, as the RAD001 levels can greatly exceed the IC50 levels for many tumors over the entire administration cycle, and therefore antitumor effects observed in rats may be more often due to combined anti-tumor cell and anti-vascular effects.

Lane HA, Wood JM, McSheehy PM, Allegrini PR, Boulay A, Brueggen J, Littlewood-Evans A, Maira SM, Martiny-Baron G, Schnell CR, Sini P, O’Reilly T. (2009) mTOR inhibitor RAD001 (everolimus) has antiangiogenic/vascular properties distinct from a VEGFR tyrosine kinase inhibitor. Clinical Cancer Research 15:1612-1622.