Diehn: How to localize a tumor ID'd via ctDNA? May be able to det tissue of origin: pattern of muts, methylation, exosomes, protein #Tricon
12:21pm February 23rd 2017 via Hootsuite
Diehn: Problem of hematopoeisis - majority of ccfDNA is from blood cells. #Tricon
12:20pm February 23rd 2017 via Hootsuite
Diehn: Benign conditions w/somatic mutations are nevi, leiomyomas (esp uterine fibroids), clonal hematopoiesis #Tricon
12:18pm February 23rd 2017 via Hootsuite
Diehn: Unclear what fraction of pre-invasive lesions release much ctdDNA. 1/11 colonic adenomas w/det ctDNA #Tricon
12:16pm February 23rd 2017 via Hootsuite
Diehn: From NIPT, incidental detection of ctDNA from 125K tests, 0.008% was occult cancer. But rel late stg. #Tricon https://t.co/V940xolHSS
12:15pm February 23rd 2017 via Hootsuite
Diehn: Showed results from SafeSeqS pub '14; Stg I sens at 42%; median was 1 fragment per 5mL #Tricon
12:13pm February 23rd 2017 via Hootsuite
Diehn: Showed cases from '14 ref - progression, amount of mutant cfDNA cp to tumor volume. #Tricon
12:12pm February 23rd 2017 via Hootsuite
Diehn: Showed ROC curve from Newman '14 comparing Stg I 50% sens to Stg II-IV 100% sens even when mutations were known. #Tricon
12:11pm February 23rd 2017 via Hootsuite
Diehn: For colorectal and lung, early-detection has improved survival. ctDNA may be an ideal biomarker. #Tricon
12:10pm February 23rd 2017 via Hootsuite
Diehn: Tumors at early stg assoc'd with better 5-y OS; 95% for CR Stg I vs 10% at Stg IV; for NSCLC, 60% at Stg I, 5% at Stg IV. #Tricon
12:09pm February 23rd 2017 via Hootsuite
Diehn: Also went into xenografts with MET ampl and dose-response; showed pathway inh restores sens to rociletinib #Tricon
12:07pm February 23rd 2017 via Hootsuite
Diehn: Also found MET ampl mediates innate and acquired resistance; showed progress over time. Also 5 other muts in same pt #Tricon
12:05pm February 23rd 2017 via Hootsuite
Diehn: Mutation via structure analysis, showed binding pocket affected, showed in-vitro dose-response in Ba/F3 cells #Tricon
12:04pm February 23rd 2017 via Hootsuite
Diehn: Difference was a different third-gen TKI. Found a novel EGFR L798I, showed progression over time. #Tricon
12:03pm February 23rd 2017 via Hootsuite
Diehn: Showed emergence of EGFR C797S in a single pt, and how it progressed over time. Found in 2%; other report 32% #Tricon
Diehn: For Rociletinib, found significant intra- and inter-pt heterogeniety. Weakness of PM - selection of multiple resistance #Tricon
12:02pm February 23rd 2017 via Hootsuite
Diehn: All pts have EGFR activating, T780M; 46% had more than 1 mechanism '16 ref https://t.co/huMro79E4v #Tricon
12:00pm February 23rd 2017 via Hootsuite
Diehn: Looking at resistance heterogeneity; is first-line Rx lead to intra-pt heterogeneity in resistance? #Tricon
11:59am February 23rd 2017 via Hootsuite
Diehn: 10% had more than one mechanism of resist; 49% T790M; 30% unk '11 ref https://t.co/GFseY5snQ3 #Tricon
11:58am February 23rd 2017 via Hootsuite
Diehn: T790M resistance is most frequent - 50-60%; Third-gen TKI is osimertinib, rociletinib, others #Tricon
11:56am February 23rd 2017 via Hootsuite
Diehn: Applied to NSCLC, EGFR mutants. 15-50% of lung ca; sens to EGFR TKI's. First line Rx 9-12 mos resistance #Tricon
Diehn: Looking at clinical samples, 50 advanced NSCLC pts, showed Sens 93%, Spec 100%, PPV 100%, NPV 95% #Tricon
11:55am February 23rd 2017 via Hootsuite
Diehn: Observed and expected AF% based on ddPCR, linear R=0.93 down to well below 0.5% #Tricon
11:54am February 23rd 2017 via Hootsuite
Diehn: iDES=integrated Digital Error Correction. Other methods require a lot more input DNA. Chart stochastic and stereotype errors #Tricon
11:53am February 23rd 2017 via Hootsuite
Diehn: Looked at background errors from healthy individuals, then developed both barcoding and polishing they call iDES #Tricon
11:52am February 23rd 2017 via Hootsuite
Diehn: Looked at distribution of errors for all types G>T etc. https://t.co/X1KajWfFTp went to published data at 0.03%, saw same #Tricon
11:51am February 23rd 2017 via Hootsuite
Diehn: Assay affected by number of molecules recovered; mutations covered; and error rate of sequencing #Tricon
11:49am February 23rd 2017 via Hootsuite
Diehn: Stg I-IV: shows ROC curve from '14 Nature Med ref https://t.co/Z1E17xGury as little as 1.9pg/mL #Tricon
Diehn: Before did tissue first then ctDNA; now they can skip the tissue. Later draws, look for new mutations (resistance) that arise #Tricon
11:47am February 23rd 2017 via Hootsuite
Diehn: Look at population-level analysis; look for recurrent mutations; drill down to panel. Tumor/normal, monitor over time #Tricon
11:46am February 23rd 2017 via Hootsuite
Diehn: Assumptions for the above: 30ng cfDNA; 50% efficiency; 90% probability of detection. #Tricon
11:45am February 23rd 2017 via Hootsuite
Diehn: WES down to 5%; WGS down to 1%; amplicons down to 0.3%; barcoded amplicon (SafeSeqS) to about 0.05%, CAPP-Seq 0.002% #Tricon
11:44am February 23rd 2017 via Hootsuite
Diehn: Cp to how low you can go: Sanger, Pyro is sens to 10%; Allele-spec PCR is 0.1-0.05%; dPCR down to 0.05% or less #Tricon
11:43am February 23rd 2017 via Hootsuite
Diehn: Limited input (a few thousand copies per mL of blood). Then from the cancer, only a few molecules. #Tricon
11:42am February 23rd 2017 via Hootsuite
Diehn: Then if recurring, re-test for additional therapy, and then monitor again. #Tricon
11:41am February 23rd 2017 via Hootsuite
Diehn: Illustrates screening, non-invasive genotyping, local treatment response, detect minimal residual disease, surveillance... #Tricon
Diehn: Mutations are involved in pathogenesis, part of the biology of the disease. And some muts are actionable #Tricon
11:40am February 23rd 2017 via Hootsuite
Diehn: Shows diagram from Crowley et al, shed into bloodstream; ctDNA attractive due to specificity (cp to protein markers). #Tricon
11:39am February 23rd 2017 via Hootsuite
Diehn: Will describe CAPP-Seq, then clin appl to adv lung ca, and clin appl to early stage lung ca #Tricon
11:38am February 23rd 2017 via Hootsuite
Max Diehn (Stanford CA) Deep sequencing of circulating tumor DNA for personalized cancer detection and monitoring. #Tricon
11:37am February 23rd 2017 via Hootsuite
Patel: Lots of effort both academically and commercially (i.e. Grail) where we'll learn much more in the years to come. #Tricon
11:36am February 23rd 2017 via Hootsuite
Patel: Finding occult metastatic disease may not have same impact as early-detection. Known mutations w/aging: clonal hematopoeisis #Tricon
11:35am February 23rd 2017 via Hootsuite
Patel: Second biological challenge: are the same mutations found in healthy individuals? #Tricon
Patel: Technical challenges - broad coverage is needed, bckgrnd errors need to be suppressed. Biological ones: enough DNA? and... #Tricon
11:34am February 23rd 2017 via Hootsuite
Patel: Promise of ctDNA: very little physiologic background; in principle, all the signal is tumor-derived #Tricon
11:33am February 23rd 2017 via Hootsuite
Patel: For protein biomarkers - there's non-zero background. Ex: PSA: tumor-derived PSA hard to detect early (little cp to bckgnd) #Tricon
11:32am February 23rd 2017 via Hootsuite
Abhijit Patel (Yale CT) Opening remarks for Circulating Cell-Free DNA symposia at #Tricon
11:31am February 23rd 2017 via Hootsuite
FYI @SeraCare @DaleYuzuki presents at #Tricon 10:15am "Improved material for developing, validating, and monitoring liquid biopsy assays"
10:20am February 23rd 2017 via Hootsuite
MT @TwoPoreGuys: Very excited to share our tech! Two Pore Guys to Present at #TRICON @BusinessWire https://t.co/g9IlTsTktA
10:17am February 23rd 2017 via Hootsuite
RT @JamesDHayden: Great article on the use of #machinelearning in #clinicaltrials and #drugdev @IBMWatsonHealth https://t.co/yNW4TWdFWp
10:00pm February 22nd 2017 via Hootsuite
