Rimm: 13 pathologists reading the NCCN/BMS study - only 28-8 andd E1L3N are statistically same. Rest: statistically sig different #AACR17
11:39am April 3rd 2017 via Hootsuite
Rimm: How compared - split at pos/neg cut-points for each test. Only 63% concordance between 4 assays. 37% non-concordance! #AACR17
11:37am April 3rd 2017 via Hootsuite
Rimm: NCCN study - n=90. Looking at different recipes; across 4 Abs. SP142 enhanced for immune cells (?) NCCN similar SP142 #AACR17
11:36am April 3rd 2017 via Hootsuite
Rimm: The Antibody is the Egg; the Assay is the Cake. Blueprint study N=39 ref https://t.co/5mw6evzcjS #AACR17
11:35am April 3rd 2017 via Hootsuite
Rimm: SP142 is also complementary for bladder, lung ca scoring. E1L3N is used as LDT at MSKCC and BWH. "Are they equivalent?" #AACR17
11:32am April 3rd 2017 via Hootsuite
Rimm: Only Dako 22c3 assay is FDA approved as a CDx; >50% tumor cells positive. Dako 28-8 is 'complementary' for >1% tumor cells #AACR
11:31am April 3rd 2017 via Hootsuite
Rimm: Lower PD-L1 have lower response. Each response 'different cutoffs' '16 review https://t.co/o4R34ls7sX #AACR17
11:30am April 3rd 2017 via Hootsuite
David Rimm (Yale CT) PD-L1 based predictive factors as companion diagnostics for immunotherapy #AACR17
11:28am April 3rd 2017 via Hootsuite
Taube: CD8, FoxP3, CD68, Sox10... for additional markers to look at #AACR17
11:27am April 3rd 2017 via Hootsuite
Taube: Using imaging of PD-L1, PD-1, different radii. Conclude: combining PD-L1 w/IFNg and PD-1 promising. Want multiplex IHC #AACR17
11:26am April 3rd 2017 via Hootsuite
Taube: Group at GenOptix presented AACR16, comparing PD-1/PD-L1 interaction score to PD-L1 IHC #AACR17
11:22am April 3rd 2017 via Hootsuite
Taube: Pretreated PD-L1+, IFN-g+ NSCLC may benefit from decrease in tumor size (ASCO poster) #AACR17
11:21am April 3rd 2017 via Hootsuite
Taube: Shows patterns of PD-L1 in SCC cp to non-squamous in lung. https://t.co/jOEP7XRt1n TCGA https://t.co/cpcJKFCK4h #AACR17
11:20am April 3rd 2017 via Hootsuite
Taube: PD-L1+ w/TIL: responsive to immunotherapy. For PD-L1-, no TILs use combination Rx. https://t.co/yp82cEZOqK #AACR17
11:15am April 3rd 2017 via Hootsuite
Taube: Constitutive, +adaptive, 4 patterns: 'not all PD-L1 is created equal' '16 ref https://t.co/dO7o4tMXmT #AACR17
11:12am April 3rd 2017 via Hootsuite
Taube: Two mechanisms for PD-L1 up-reg in tumors: innate (tumor signaling), and adaptive (IFN-g, reflecting immune reaction) #AACR17
11:11am April 3rd 2017 via Hootsuite
Taube: In mice, PD-L1 expression w/o lymphocytes present. '12 STM ref https://t.co/6x6ZxVJalJ 150 melanocytic lesions: TIL necessary #AACR17
11:09am April 3rd 2017 via Hootsuite
Taube: Shows correlation of pre-Rx tumor PD-L1 expression w/ objective response. Need to look at innate vs adaptive exp of PD-L1 #AACR17
11:08am April 3rd 2017 via Hootsuite
Janis Taube (Hopkins MD) Tumor immunoarchitecture: Add'l predictive factors for cancer immunoRx response #AACR17
11:07am April 3rd 2017 via Hootsuite
Hellman: Opportunities to use targeted or WES as a diagnostic? Need to move beyond TMB as a proxy. #AACR17
11:05am April 3rd 2017 via Hootsuite
Hellman: Need to balance minimal tissue use and information derived is 'usable and generalizable', and reproducible 'abs critical' #AACR17
11:04am April 3rd 2017 via Hootsuite
Hellman: Could be other types of mutations (i.e. CNV reports are starting to appear). No clear reason why mutation matters. #AACR17
11:03am April 3rd 2017 via Hootsuite
Hellman: '16 NEJM https://t.co/Xx9q4M9Grk shows mutations in melanoma #AACR17
Hellman: Mutation burden a 'crude characteristic'. '17 Ref loss of PTEN and I-O resistance https://t.co/EaaxfO6zgQ #AACR17
11:01am April 3rd 2017 via Hootsuite
Hellman: Neoantigen burden 'probably not' a meaningful characteristic; immunol. relevant neoAG appear to be relatively few #AACR17
11:00am April 3rd 2017 via Hootsuite
Hellman: Neoantigen burden <> mutation burden. Can est binding affinity in HLA alleles. Accuracy? '14 ref https://t.co/nYQXfQCYZG #AAC
10:56am April 3rd 2017 via Hootsuite
Hellman: Does mut burden assoc w/PD-L1? Perhaps, but not convincing. Is candidate neoantigen burden meaningful? #AACR17
10:55am April 3rd 2017 via Hootsuite
Hellman: Shows data between high- and low-burden; shares Seiwert ASCO SITC data, Nanostring GEP and mut burden not high corr. #AACR17
10:54am April 3rd 2017 via Hootsuite
Hellman: Yes, WES a reasonable proxy for tumor mutation burden. Points out that WES targeted panels were not designed for purpose #AACR17
10:51am April 3rd 2017 via Hootsuite
Matthew Hellman (Mem Sloan Kettering NY) DNA-based approaches to examining determinant of response and resistance to PD-1 blockade #AACR17
10:50am April 3rd 2017 via Hootsuite
Beechem: Can invert the mask, to select the remainder (tumor microenvironment). Poster 3810/6 #AACR17 with D Rimm (Yale)
10:38am April 3rd 2017 via Hootsuite
Beechem: The mask will then selectively illuminate only the tumor cells, PanCK labeled; masking then all tags are released #AACR17
10:36am April 3rd 2017 via Hootsuite
Beechem: Able to do 'phenotypic reading and writing' - took colorectal ca slice; PanCK imaging. Color image = masked #AACR17
10:35am April 3rd 2017 via Hootsuite
Beechem: Micromirror-controlled cleavage is specific to the spot; shows quantitation of each area, single-molecule counting #AACR17
10:33am April 3rd 2017 via Hootsuite
Beechem: The cleaved tags are then going through the regular analysis - a localized analysis. 4-color microscope #AACR17
Beechem: Based on that region of interest ID'd by microscopy, can ID and select, UV-cleave, aspirate oligos and collect. #AACR17
10:31am April 3rd 2017 via Hootsuite
Beechem: A design for a digital, spatial profiling microscope. Able to label antibodies, illuminate region of interest and cleave #AACR17
10:30am April 3rd 2017 via Hootsuite
Mills: WES 'mutation was there' but not called. Can inform immuno-oncology studies by simultaneous analysis of DNA, RNA, protein #AACR17
10:27am April 3rd 2017 via Hootsuite
Mills: Cross-platform between nCounter SNV vs deep WES; NRAS Q61R missed via WES. T200 platform (targeted) found at 25% AF #AACR17
10:25am April 3rd 2017 via Hootsuite
Mills: Work is in-press (Genome Med, Reuben et al); #AACR17 poster 5563. RNA/Protein analysis; PI3K/AKT activation, ImmunoRx resistant
10:24am April 3rd 2017 via Hootsuite
Mills: FFPE looking at 104 SNVs, 770 RNAs, 29 protein targets. 5ng DNA, 100ng RNA, can work with 1 5um slide #AACR17
10:23am April 3rd 2017 via Hootsuite
Mills: 3D biology utilization - protein vs CNV; protein vs mRNA - plots show where only protein indicated (no CNV, mRNA present) #AACR17
10:22am April 3rd 2017 via Hootsuite
Mills: RP protein array (RPPA) vs nCounter: overall 'high correlation' (found PhophoEGFR Ab x-reacts w/phophoHER2) #AACR17
10:20am April 3rd 2017 via Hootsuite
Mills: Shows technical, biological replicate data: r=0.99, r=0.96. 'Almost unbelievable correlation'. Low counts (5 or 10) reprod #AACR17
10:19am April 3rd 2017 via Hootsuite
Mills: Can easily 30-40 antibodies at a time; using sparse unbiased pathway representation - can impute pathway activity #AACR17
10:18am April 3rd 2017 via Hootsuite
Mills: Their protein work: cleavable linker DNA on Ab - validated via IHC, RP protein array; have immune-oncology platform in-play #AACR17
10:17am April 3rd 2017 via Hootsuite
Mills: Chart of DIN to AF, 97-98% concordance. Input is only 7ng (maximum); can design 'very spec probes' hard NGS var's to detect #AACR17
10:16am April 3rd 2017 via Hootsuite
Mills: cfDNA down to 0.5%; detection rate 97.8% (1 FN, 45/46 samples). NRAS mutation missed - may be technical problem #AACR17
10:14am April 3rd 2017 via Hootsuite
Mills: 25 genes, have known samples w/ultra-deep seq, some work on cfDNA. FFPE has AF's from 2.4% - 50%; DIN from 1.7-6.2 #AACR17
10:13am April 3rd 2017 via Hootsuite
Mills: Used probesets, with perfect / imperfect matches. Can displace with mutant probe. Vantage 3D: picked up DNA SNVs #AACR17
10:12am April 3rd 2017 via Hootsuite
