Kulkarni: Agrees there's an arms race of having more. Reminds the audience that lab people see data, not patients #FDANGSWorkshop
2:54pm February 25th 2016 via Twitter Web Client
Dickson: There is a trend toward more mutations; physicians may think on a reductionist way. #FDANGSWorkshop
2:53pm February 25th 2016 via Twitter Web Client
Sklar: Great pressure on speed and brevity; but need to explain to provide info to outside stakeholders #FDANGSWorkshop
2:49pm February 25th 2016 via Twitter Web Client
Sklar: Gets back to validation efforts by Mfrs. Reporting: need to be a bit more explicit on what it means, explaining #FDANGSWorkshop
Blumenthal: Many resources spent on labeling accuracy, truthful promotion, very tricky. And for a device label? #FDANGSWorkshop
2:47pm February 25th 2016 via Twitter Web Client
Kreuz: Who is the expert with a report? How can the pt be assured of competence? #FDANGSWorkshop
2:46pm February 25th 2016 via Twitter Web Client
Q:How to truthfully, accurately provide disclaimers in labeling? #FDANGSWorkshop
2:44pm February 25th 2016 via Twitter Web Client
Dickson: 80% of the clinics in this country may not have a molecular tumor board; how will clinicians understand? #FDANGSWorkshop
2:43pm February 25th 2016 via Twitter Web Client
Tsimberidou: You need to know sub-clonal drivers 'at the highest sensitivity' as well, looking at global biology #FDANGSWorkshop
2:42pm February 25th 2016 via Twitter Web Client
Sklar: Impt consideration; 'not to mention CNVs', you have to set threshold. #FDANGSWorkshop
2:41pm February 25th 2016 via Twitter Web Client
Sklar: These are quantitative. T790M - subclonal amts in NSCLC, 10% VAF but tumor completely resistant #FDANGSWorkshop
2:40pm February 25th 2016 via Twitter Web Client
Q: Effect of new assay performance (increased sens for example?) Population changes? #FDANGSWorkshop
2:39pm February 25th 2016 via Twitter Web Client
Sklar: MATCH trial is looking at it systematically, but N of 1 means rationale for broad-based seq of tumors #FDANGSWorkshop
2:31pm February 25th 2016 via Twitter Web Client
Sklar: Co's are not clear on what validation they have done, 'hiding behind the fig leaf of RUO'. What pipeline, cell line? #FDANGSWorkshop
2:20pm February 25th 2016 via Twitter Web Client
Kruez: One company released findings, rec'd 'this drug'. What guarantee? Tsimberidou: Careful clinical relevance, trials #FDANGSWorkshop
2:18pm February 25th 2016 via Twitter Web Client
Sklar: 'We usually lose money. What did they find the 2nd time?" BRAF first time, but found more 2nd time #FDANGSWorkshop
2:16pm February 25th 2016 via Twitter Web Client
Kurez: 2nd time: what happened to the first one? Is it a way to make money? What do you tell a patient? #FDANGSWorkshop
Kruez: She had 3 genetic tests, the urgency is 'do whatever it takes'. Don't care about insurance, not covered #FDANGSWorkshop
2:15pm February 25th 2016 via Twitter Web Client
Kulkarni: An expensive drug that may do real harm to a patient. #FDANGSWorkshop
2:14pm February 25th 2016 via Twitter Web Client
Kulkarni: Mixed feelings about this - they have high confidence, 100% concordance, other labs may not reflex like this #FDANGSWorkshop
Kulkarni: For NSCLC, they have intronic baits to pick up ALK fusions; reflex to FISH due to insurance req's #FDANGSWorkshop
2:13pm February 25th 2016 via Twitter Web Client
Sklar: Orthogonal confirmation may be needed. Two levels of validation - the Co that validates the test, and the lab too #FDANGSWorkshop
2:09pm February 25th 2016 via Twitter Web Client
Dickson: Cp of CDx to a new tech: what if a new insertion is found? Sklar: Co's that sell tests need control material #FDANGSWorkshop
Sklar: DNA is 'pretty much DNA'. Interfering substances (melanin) needs to be noted; end-labs need to be involved in design #FDANGSWorkshop
2:08pm February 25th 2016 via Twitter Web Client
Q: Follow-on CDx. What level of agreement is enough for reasonable assurance of safety, effectiveness? Good smpl sources? #FDANGSWorkshop
2:06pm February 25th 2016 via Twitter Web Client
Kruez: Confusion, debates, impatience. (Wow. What a reminder of why we do what we do.) #FDANGSWorkshop
2:03pm February 25th 2016 via Twitter Web Client
Kruez: On chemo, ended her 30y career (here https://t.co/Et7GK0Qu7t ) sister died of lu ca. Mixed emotions #FDANGSWorkshop
Panel: Greta Kruez (Patient Advocate) Stage IV NSCLC survivor (never-smoker) '12 Dx br ca lumpectomy, Late '13 Dx #FDANGSWorkshop
2:01pm February 25th 2016 via Twitter Web Client
Panel: Drs. Kulkarni & Dickson, Jeff Sklar (Yale), Apostolia-Maria Tsimberidou (MDA), Gideon Blumenthal (FDA/CDER) #FDANGSWorkshop
1:59pm February 25th 2016 via Twitter Web Client
Dickson: Makes the case for standardized NGS testing #FDANGSWorkshop
1:57pm February 25th 2016 via Twitter Web Client
Dickson: "General consensus of 'standard of care without supporting data'" Like digoxin w/o support for 100y #FDANGSWorkshop
1:56pm February 25th 2016 via Twitter Web Client
Dickson: Shows a bit of a Venn to illustrate EGFR driver in NSCLC, the NGS test a subset/overlap, CDx a subset/overlap #FDANGSWorkshop
1:54pm February 25th 2016 via Twitter Web Client
Dickson: 'We need clinical outcomes to determine overall benefit'. Increased sens. may not lead to better outcomes #FDANGSWorkshop
1:53pm February 25th 2016 via Twitter Web Client
Dickson: We 'can't aggregate clinical utility data unless we standardize the testing'; knowing the 'exact test' #FDANGSWorkshop
1:52pm February 25th 2016 via Twitter Web Client
Dickson: PA (pre-analytic), SQ (sequencing), IP (informatics) with different enrichment, seq platform, pipeline #FDANGSWorkshop
1:51pm February 25th 2016 via Twitter Web Client
Dickson: We're still in the testing phase for all these questions. #FDANGSWorkshop
1:48pm February 25th 2016 via Twitter Web Client
Dickson: NGS a 'Swiss Army Knife' but with a ? or , or ! - depends. We need to test analytical equiv, clinical equiv #FDANGSWorkshop
1:47pm February 25th 2016 via Twitter Web Client
Dane Dickson (pathologist, Molecular Evid Dev Consortium): Tenets: obs to questions to hypothesis to testing to theory #FDANGSWorkshop
1:46pm February 25th 2016 via Twitter Web Client
Kulkarni: Inpt work with ClinGen, important to build database of variants expertly curated. #FDANGSWorkshop
1:41pm February 25th 2016 via Hootsuite
Kulkarni: Huge challenges w/databases; cell lines w/issues (gain aneuploidy or other artifacts). #FDANGSWorkshop
1:40pm February 25th 2016 via Hootsuite
Kulkarni: L5 is a known SNP (ESP, ExAC, TGP w/MAF >1%) #FDANGSWorkshop
Kulkarni: Leve 1: predictive, prognostic (BRAF V600E); L2 in other tumor types (IDH1 in CRC) L3: Reported (unk sig) L4: VUS #FDANGSWorkshop
1:39pm February 25th 2016 via Hootsuite
Kulkarni: AMP and ACMG and ASCO have joint committees for giving guidelines 'in a few months' #FDANGSWorkshop
1:38pm February 25th 2016 via Hootsuite
Great to meet @VanAllenLab IRL at the #FDANGSWorkshop today. Live-tweeting has its rewards!!
1:37pm February 25th 2016 via Hootsuite
RT @kennamshaw: Shashi Kulkarni from wash u speaks first on clinical interpretation/signifiance of variants #FDA #NGS #FDANGSWorkshop
1:36pm February 25th 2016 via Hootsuite
RT @kennamshaw: Shashi Kulkarni from wash u speaks first on clinical interpretation/signifiance of variants #FDA #NGS
1:35pm February 25th 2016 via Hootsuite
Q: New discovery, gene gets added to existing panel A: Klees: NY expects to see accuracy, specificity, and rest of panel #FDANGSWorkshop
12:26pm February 25th 2016 via Twitter Web Client
Van Allen: May be a data representation issue Hegde: VAF is important, and then to comparison to ExAC and other germline dbs #FDANGSWorkshop
12:23pm February 25th 2016 via Twitter Web Client
Q: Reservations about VAF judgements to make conclusions (50% may not be germline). Then reporting... #FDANGSWorkshop
12:22pm February 25th 2016 via Twitter Web Client
Hegde: If there's a problem with the cellularity? Hard to get to. Eberhard: May disclaim with caution, or refuse #FDANGSWorkshop
12:17pm February 25th 2016 via Twitter Web Client
