Supplemental Information – Supplemental Figures S1-S4, Supplemental Tables 1-3,
and Supplemental Methods
Supplemental Figure S1
CONSORT Flow Diagram
Figure S1. CONSORT Flow Diagram depicting the study algorithm and allocation of study patients.
Supplemental Figure S2
Figure S2. Treatment effect of RO4929097 on DCE perfusion MRI parameters. A, Plasma volume (Vp) map at baseline and after 7 days of RO4929097 (RO) administration in a study patient, showing marked decrease in blood perfusion. B and C, Vp ratio and permeability coefficient (Ktrans) at baseline and 7 days after exposure to RO4929097 (n=8 patients). Values are expressed as mean ± S.E.M. (*p<0.05).
Supplemental Figure S3
Figure S3. Immunohistochemistry of Ki-67 in matched study patients. Images show the Ki-67 immunohistochemical labeling index in each patient before and after exposure to RO4929097 with graphic depiction of the quantification thereof. A minimum of 1,000 cells was counted per condition (Scale bar = 100 µm).
Supplemental Figure S4
A
B
Figure S4. Gene expression profile in treated study patients. A, Gene expression panel showing fold change values comparing matching post-treatment (7 days RO4929097) and pre-treatment tissue (n=5). (NSC = neural stem cell markers; MS = mesenchymal markers; Notch-R = Notch-Receptors; VEGF-R = VEGF-Receptors). B, Gene expression panel showing fold change values comparing recurrent glioblastoma tissue after prolonged drug therapy versus matched pre-treatment tumor (n=3). (NSC = neural stem cell markers; MS = mesenchymal markers; Notch-R = Notch-Receptors; VEGF-R = VEGF-Receptors).
Supplemental Table 1
Definitely, likely or possibly related to treatment / Grade 3 / Grade 4Anemia / 2
Lymphopenia / 14 / 2
Neutropenia / 3
Leukopenia / 3
Thrombocytopenia / 4 / 1
Alanine Transaminase elevation / 2
Hypophosphatemia / 3
Hypokalemia / 2
Hyperkalemia / 1
Hypocalcemia / 1
Hyponatremia / 1
Hypernatremia / 1
Hypermagnesemia / 1
Unrelated or unlikely related to treatment / Grade 3 / Grade 4
Hyperglycemia / 6 / 2
Prolonged INR / 1
Hypoalbuminemia / 2
Intracranial Hemorrhage / 1
Intra-operative Hemorrhage / 1
Brain abscess / 1
Pulmonary embolism / 2
Intra-abdominal hemorrhage / 1
Muscle Weakness / 2
Gait disturbance / 1
Seizure / 1
Confusion / 2 / 1
Hydrocephalus / 1
Supplemental Table 1. Reported grade 3 and 4 adverse events in study patients. Number of patients who experienced grade 3 and 4 adverse events among the 21 patients evaluable for toxicity, regardless of attribution, are shown in this table. No grade 5 treatment-specific events were observed.
Supplemental Table 2. RO4929097 Pharmacokinetic Data a
Dailydose / Time / Cmax / AUC24
(mg) / point / (ng/mL) / (ngh/mL)
10 / RT-D1 / 192 / ± / 190 (4) b / ND / c
RT-D35 / 172 / ± / 91 (3) / ND
ADJ-D1 / 158 / ± / 102 (3) / ND
15 / RT-D1 / 331 / ± / 87 (5) / 2.804 / ± / 249 (3)
RT-D35 / 310 / ± / 83 (3) / ND
ADJ-D1 / 249 / ± / 18 (3) / 1.869 / ± / 258 (3)
20 / RT-D1 / 409 / ± / 44 (8) / 5.035 / ± / 1,222 (8)
RT-D35 / 381 / ± / 106 (6) / ND
ADJ-D1 / 353 / ± / 91 (6) / 3.291 / ± / 1,851 (5)
Supplemental Table 2. RO4929097 Pharmakokinetic Data. Values are reported as the geometric mean ± SD. Number in parentheses are the number of patients. Abbreviations: RT-D1 = radiotherapy day 1, RT-D35 = radiotherapy day 35, ADJ-D1 = adjuvant treatment day 1, ND = not determined
Supplemental Table 3.
(µM) / Disrupted BBB
(µM) / Plasma
(µM) / Intact BBB/Plasma Ratio / Disrupted BBB/Plasma Ratio
10 / M01 / 0.13 / 0.30 / 0.12 / 1.11 / 2.59
10 / M02 / 0.36 / 0.55 / 0.31 / 1.18 / 1.81
10 / M03 / 0.43 / 0.89 / 0.34 / 1.27 / 2.66
15 / M05 / 0.49 / 1.26 / 0.36 / 1.35 / 3.47
15 / M18 / 0.36 / 0.62 / 0.42 / 0.87 / 1.48
20 / M21 / 0.17 / 0.36 / 0.34 / 0.51 / 1.08
20 / M23 / 0.36 / 0.62 / 0.37 / 0.97 / 1.66
20 / M24 / 0.36 / 1.26 / 0.44 / 0.82 / 2.84
15 / M07 / 0.47 / 0.62 / 0.28 / 1.68 / 2.21
15 / M09 / 1.36 / 1.32 / 0.21 / 6.64 / 6.43
20 / M14 / 0.75 / 0.72 / 0.86 / 0.86 / 0.84
20 / M24 / 0.28 / 0.43 / 0.33 / 0.84 / 1.30
Supplemental Table 3. Individual molar drug levels and tumor-to-plasma ratios. Samples shown in bold font were obtained later in the course of the disease, at the time of progression.
Supplemental Methods
Inclusion and Exclusion Criteria
Inclusion criteria consisted of newly diagnosed glioblastoma or AA, age >18, indication for additional debulking surgery as part of the initial treatment, ECOG performance status <2, and adequate marrow and organ function. Exclusion criteria consisted of patients with anaplastic oligodendroglioma or 1p/19q co-deletion; prior chemotherapy, RT or experimental therapies; patients taking medications with narrow therapeutic indices metabolized by CYP P450 or receiving strong CYP3A4 inducers/ inhibitors; malabsorption, hepatitis B or C; uncontrolled electrolyte abnormalities or intercurrent illness; increased risk for QT interval prolongation (QTc >450 msec in males or >470 msec in females); and contraindication for a brain surgical procedure.
MRI acquisition parameters and processing
MRI sequences were acquired with a 1.5 T or 3.0 T GE scanner (Milwaukee, WI) and a standard GE 8 channel head coil. The standard imaging protocol consisted of sagittal and axial T1-weighted images; axial T2-weighted images; axial gradient echo or susceptibility-weighted images; axial diffusion-weighted images (DWI) with apparent diffusion coefficient (ADC) maps; and contrast-enhanced coronal, sagittal and axial T1-weighted images.
The kinetic enhancement of tissue during and after injection of Gd-DTPA (0.2 ml/kg to maximum 20 ml gadopentetate dimeglumine; Bayer HealthCare Pharmaceuticals, Wayne, NJ) was obtained using a 3D T1-weighted fast spoiled-gradient (SPGR) echo sequence (TR 4 - 5 ms; TE 1 – 2 ms; slice thickness 3 mm; FA 25 o; FOV 24 cm; matrix 256×128; temporal resolution (Δt) of 5 – 6 s). Ten to 14 slices covering the lesion were scanned. The first 10 phases pre-injection and 30 phases post-injection were obtained. Pre-contrast T1 value was estimated as 1,000 ms. Matching T1-weighted (TR/TE = 600/8 ms; thickness = 4.5 mm) and T2-weighted (TR/TE = 4,000/102 ms; thickness = 4.5 mm) spin-echo images were obtained.
Data processing and the analysis of pharmacokinetic variables were performed using NordicICE software (NordicNeuroLab, Bergen, Norway). Pre-processing included removal of background noise and minimizing high frequency noise.
Using a two-compartment model with kinetic modeling and arterial input function (AIF)-based vascular deconvolution as proposed by Murase et al. (47), the fractional plasma volume (Vp) and volume transfer coefficient (permeability coefficient or Ktrans) between plasma and extravascular extracellular space were measured. Curves showing an optimal relationship between AIF and concentration-time curve were carefully selected. All regions of interest (ROI) were manually delineated by a neuroradiologist (SK) on an axial slice using T2 FLAIR images and matching T1 post-contrast images as an anatomical guideline. ROI included the tumor tissue, but excluded areas of hemorrhage, calcification, cystic/necrotic change and vessels. The ROI was then overlaid onto DCE parametric maps including Ktrans and Vp. A same-sized ROI was placed on the contralateral white matter and the ratio was calculated. For statistical analysis, Wilcoxon signed-rank test, at significance level of p<0.05 with two tailed hypothesis, was performed to evaluate the differences of values of Ktrans and Vp values between baseline and chemotherapy treatment.
Pharmacokinetic Studies
High-performance liquid chromatography with tandem mass spectrometry was utilized to determine levels of RO4929097 in blood and tumor tissue.
For patients enrolled prior to the surgical resection, RO4929097 was given for 7 days pre-operatively according to their dosage cohort. Day 7 coincided with surgical tumor resection, on which RO4929097 was administered 2-3 h before surgery. During surgical resection, patients underwent tumor sampling from the surgical suite at Memorial Sloan Kettering Cancer Center (MSKCC). Areas with and without contrast-enhancement, as defined by intraoperative MRI and navigation systems (Brainlab AG, Muenchen, Germany), were differentially sampled. Histopathological diagnosis was confirmed by a neuropathologist at MSKCC. Tissue samples (minimum sample size: 200 mg) were rapidly snap-frozen, shipped on dry ice to the analysis site, and stored at -80 ºC until analysis.
Standard PK evaluations were utilized to evaluate RO4929097 PK parameters during concomitant radiotherapy and temozolomide, and adjuvant temozolomide, with drug levels determined at the following time points:
Treatment concomitant with Radiotherapy / Sample Collection time (hour)Week 1, Day 0**
**RO4929097 not administered this day / 0 (before Temozolomide), 30 minutes, 1, 2, 4, 6, 8
Week 1, Day 1 / 0 (before Temozolomide and RO4929097), 30 minutes, 1, 2, 4, 6, 8
Week 1, Day 2 / 0 (before Temozolomide and RO4929097)
Week 5, Day 1 / 0 (before Temozolomide and RO4929097), 30 minutes, 1, 2, 4, 6, 8
Adjuvant Treatment after Radiotherapy / Sample Collection time (hour)
Week 1, Day 0 / 0 (before Temozolomide), 30 minutes, 1, 2, 4, 6, 8
Week 1, Day 1 / 0 (before Temozolomide and RO4929097), 30 minutes, 1, 2, 4, 6, 8
Week 1, Day 2 / 0 (before Temozolomide and RO4929097)
Week 2, Day 1 / 0 (before RO4929097 )
Viability and Flow Cytometry Analysis of Organotypic Explants
For the organotypic explant experiments, tumor tissue was rinsed twice in ice-cold sterile 1x phosphate-buffered saline solution containing 3x penicillin/streptomycin, cut by scalpel into ~1-2 mm3 cubic pieces, and placed on sterile plate inserts (Millipore, Billerica, MA) overnight. A total of 1 ml of sterile N2 medium was added to each dish, with 50 µl dispensed on top of each explant. A total of 1 µM of RO4929097 was added to each treatment group and the medium was changed every other day. For the combination group, explants were pretreated with 1 µM RO4929097, followed by irradiation with 10 Gy.
The percentage of viable cells was determined using the CellTiter Glo Luminescent Assay following standard protocol. For flow cytometry, explants were cut manually and then dissociated into single cells with Accumax (Innovative Cell Technologies, San Diego, CA) as previously described (45, 46). Single cells were blocked with human FcR blocking reagent (1:10, Miltenyi Biotec. Bergisch Gladbach, Germany) at 4 °C for 20 min before incubation with APC-conjugated anti-CD133 antibodies (1:30 Miltenyi Biotec; 1:1 AC141:AC133 epitopes) for 30 min on ice. Tumor cells were then washed and resuspended in FACS buffer (containing 1× Ca2+/Mg2+-free HBSS [Invitrogen, Waltham, MA], 10 mM of HEPES, 0.156 % of glucose and 0.5 % of low-endotoxin BSA fraction V [all from Sigma-Aldrich, St. Louis, MO]; pH =7.2) before analysis. Flow cytometry was performed using FACSCalibur flow cytometer (BD Biosciences, East Rutherford, New Jersey), excluding dead cells by labeling with 7-amino-actinomycin D (BD Biosciences). A minimum of 10,000 events was counted for the analysis with FlowJo software (FlowJo, Ashland, OR).
Semi-Quantitative Real-Time PCR
Total RNA from brain tumor tissue was extracted using TRIzol (Invitrogen). A total of 1.5 µg of RNA was reverse-transcribed using the High Capacity cDNA RT Kit (Applied Biosystems, Waltham, MA) for reverse transcription-polymerase chain reaction. PCR amplification of resulting cDNA was performed with TaqMan Gene Expression Master Mix (Applied Biosystems) on custom designed TaqMan Array 96-Well Fast Plates containing predesigned primers and probe sets (Applied Biosystems). Transcript expression was normalized by 18S RNA.
Immunohistochemistry
Brain tumor tissue was embedded in paraffin. Twenty µm sections of the tissue were cut with a cryostat-microtome. Immunofluorescent detection of NICD (Cleaved Notch-1; Cell Signaling Technology, Danvers, MA; cat# 4147, ) and CD31 (Dako, Carpinteria, CA; cat#M0823) was performed at the Molecular Cytology Core Facility of MSKCC using Discovery XT processor (Ventana Medical Systems, Tucson, AZ). The tissue sections were deparaffinized with EZPrep buffer (Ventana Medical Systems), antigen retrieval was performed with CC1 buffer (Ventana Medical Systems) and sections were blocked for 30 minutes with Background Buster solution (Innovex , Needham, MA). Anti-NICD antibodies were applied at 5 µg/ml, anti-CD31 at 5 µg/ml, and sections were incubated for 5 hours, followed by 60 minutes incubation with biotinylated goat anti-rabbit IgG (Vector Laboratories, Burlingame, CA; cat#PK6101) for NICD or horse anti-mouse IgG (Vector Laboratories, cat# BMK-2202) at 1:200 dilution. The detection was performed with Streptavidin-HRP D (Ventana Medical Systems), followed by incubation with Tyramide Alexa Fluor 488 (Invitrogen, cat#T20922) for NICD or Tyramide Alexa Fluor 568 (Invitrogen, cat# T20914) for CD31. Slides were stained with DAPI and cover-slipped using Mowiol. All images were acquired with an Eclipse i80 microscope (Nikon, Tokyo, Japan).
Morphometry
The distribution percentage of NICD+ cells was calculated by using ImageJ. A minimum of 1,000 cells per sample was counted. To determine the percentage ratio of NICD+ blood vessels, the entire tissue section of each group was scanned and the total number of tumor vessels, as determined by CD31+ staining, counted. CD31-positive tumor vessels co-expressing NICD were counted to calculate the percentage ratios.
RNA Sequencing and Profiling
Formalin-fixed paraffin-embedded tissue from the initial diagnostic biopsy, as well as from surgical resection and re-resection of tumor was used for mRNA extraction. A customized gene panel was used for the profiling assay and the data were analyzed with a custom script in R. The raw sequence counts were pre-processed and normalized by first adding 1 (pseudo count) taking the log base 2 and then normalizing using the quartile normalization method from the LIMMA R/Bioconductor package. P-values were computed using a simple t-test and then corrected for multiple testing using the FDR method.