Pfeifer: Labs need to demonstrate extraction of nucleic acids of high quality. But where is the line drawn? #FDANGSWorkshop
10:25am February 25th 2016 via Twitter Web Client
Pfeifer: Looking for changes - cell lines can be mixed, often see the constituent components, unrelated to biology #FDANGSWorkshop
10:24am February 25th 2016 via Twitter Web Client
Schetter: Q4: How contrived samples may be used to demonstrate analytical validity. QC metrics, how to more closely mimic #FDANGSWorkshop
10:23am February 25th 2016 via Twitter Web Client
Rossi: For myeloma: enrichment for cells is critical. Specific input from hematopathologists #FDANGSWorkshop
10:22am February 25th 2016 via Twitter Web Client
Aisner: Bony specimens are handled specifically; communication streams impt for decalcification. Necrotic samples #FDANGSWorkshop
10:20am February 25th 2016 via Twitter Web Client
Pfeifer: And the ordering physician can't tell the difference of the limitations of a given test. #FDANGSWorkshop
10:19am February 25th 2016 via Twitter Web Client
Pfeifer: And another provider will provide a cheaper assay w/o any details on size of indels, other complexities #FDANGSWorkshop
Pfeifer: People will provide a test w/o specifying the kind of variants (i.e. rearrangements, translocations) #FDANGSWorkshop
10:18am February 25th 2016 via Twitter Web Client
Rossi: 'We use genomics to get at biology' Aisner: Rather than tumor types, should look at matrix interference as well #FDANGSWorkshop
10:17am February 25th 2016 via Hootsuite
Berger: 60 general types, 300 specific types, practical considerations. (Echos Raja's sentiments.) #FDANGSWorkshop
10:15am February 25th 2016 via Twitter Web Client
Luthra: Their experience (MDA) has been not tumor type, but rather how it was prepared (FNA or resection) and FF or FFPE #FDANGSWorkshop
10:14am February 25th 2016 via Hootsuite
Schetter: Q3 Pan-Cancer claims - what kinds of studies needed for mfs to claim panels for different cancers #FDANGSWorkshop
10:13am February 25th 2016 via Twitter Web Client
Aisner: Yet constraints into what is submitted to the FDA, and what may accurately reflect the biology (vs the test) #FDANGSWorkshop
10:11am February 25th 2016 via Twitter Web Client
Schetter: The need is to accurately report what kind of CDx work has been done, and truthfully report that #FDANGSWorkshop
10:10am February 25th 2016 via Twitter Web Client
Pfeifer: And the primary vs. metastasis at a different site: different findings due to different sampling. #FDANGSWorkshop
10:06am February 25th 2016 via Twitter Web Client
Pfeifer: Where is the tumor sampled? Wonders if the FDA is thinking about this; studies are doing multiple sampling #FDANGSWorkshop
10:04am February 25th 2016 via Twitter Web Client
Rossi: Need to get pathologists to get involved 'from the very beginning'. #FDANGSWorkshop
Aisner: Labs that are not enriching for (tumor cellularity) need to look at this. Remnant material retained to go back to #FDANGSWorkshop
10:01am February 25th 2016 via Twitter Web Client
Pfeifer: We worry a lot about cellularity. And library complexity. #FDANGSWorkshop
10:00am February 25th 2016 via Twitter Web Client
Schetter: Topic 2: What specific concerns around sample processing of FFPE? What level of validation? Isolation methods? #FDANGSWorkshop
9:57am February 25th 2016 via Twitter Web Client
Schetter: For vendors, what metrics are needed? Berger: coverage, yet terminology may cause confusion #FDANGSWorkshop
9:55am February 25th 2016 via Twitter Web Client
Pfeifer: May not necessarily be an extra step, could be add'l analysis #FDANGSWorkshop
9:54am February 25th 2016 via Twitter Web Client
Pfeifer: Contamination an impt point - haplotype analysis. At NIH trials - genotyping needed (a 'provenance step') #FDANGSWorkshop
Berger: They look at tumor-normal pairs, look for sources of contamination; it's 'not all artifact' #FDANGSWorkshop
9:53am February 25th 2016 via Twitter Web Client
Berger: Enrichment technologies (hyb vs amplicon) have different nomenclature ('duplicates') need to consider #FDANGSWorkshop
9:51am February 25th 2016 via Twitter Web Client
Michael Rossi (Emory U): Discussion has been DNA-centric. Assessment of RNA too; hard cutoffs are not practical #FDANGSWorkshop
Aisner: Situations w/poor prob. of library success - hard decision to go to lower-complexity test. Many times: in parallel #FDANGSWorkshop
9:49am February 25th 2016 via Twitter Web Client
Luthra: With limited sample, they will try. Have to look at the 'beginning to end' product #FDANGSWorkshop
9:48am February 25th 2016 via Twitter Web Client
Raja Luthra (MD Anderson): 15% of cases used fail due to DNA conc input requirement; relaxing it they go down to 5%. #FDANGSWorkshop
9:46am February 25th 2016 via Twitter Web Client
Berger: Because the data could be beneficial to the patient. #FDANGSWorkshop
9:45am February 25th 2016 via Twitter Web Client
Berger: They move forward even if after pre-analtyical QC, they will still sequence. If LOD is only 15%, still get data #FDANGSWorkshop
Michael Berger (MSKCC): they run a 400-gene panel (10K samples); all around post-seq QC. Assay shown to be robust #FDANGSWorkshop
9:44am February 25th 2016 via Twitter Web Client
Pfeifer: The library complexity metric - addresses complexity of the sample. But still doesn't capture upstream var #FDANGSWorkshop
9:42am February 25th 2016 via Twitter Web Client
Pfeifer: They (WashU) look at % reads mapped, % reads on-target, % unique reads (library complexity measurement) #FDANGSWorkshop
9:41am February 25th 2016 via Twitter Web Client
Panel discussion: Pfeifer: 'We take a "proof in the pudding approach" - look at the sequence metrics after the input' #FDANGSWorkshop
9:39am February 25th 2016 via Twitter Web Client
Aaron Schetter: What pre-analytical QC to eval DNA/RNA prior for NGS oncology panels? QC metrics? When not needed? #FDANGSWorkshop
9:38am February 25th 2016 via Twitter Web Client
Aisner: Points out the weakness of macrodissection, risk of false negatives, unless careful microdissection occured #FDANGSWorkshop
9:29am February 25th 2016 via Twitter Web Client
Aisner: 'This should not be over-prescribed' - shows slides of melanoma, difference between resolutions #FDANGSWorkshop
Aisner: Enrichment commentary: will vary widely. Microdissection, 'this is the practice of medicine'. #FDANGSWorkshop
9:28am February 25th 2016 via Twitter Web Client
Aisner: Organ tissue: Is skin and lung really that different? May look at core tissue, and ECM / other var's separately #FDANGSWorkshop
9:25am February 25th 2016 via Twitter Web Client
Aisner: For ERBB2 to stipulate cold ischemic time - has caused a 'lot of tumult' in the medical practice context #FDANGSWorkshop
9:23am February 25th 2016 via Twitter Web Client
Aisner: Transfer media, temp, cold ischemic time, target organ, procurement method: FDA may not be able to stipulate #FDANGSWorkshop
9:22am February 25th 2016 via Twitter Web Client
Dara Aisner (U CO) #FDANGSWorkshop From tissue acquisition - how assessed, enriched, extracted, assessed, and quantified
9:21am February 25th 2016 via Twitter Web Client
Pfeifer:Bioinformatic approaches? in-silico datasets to evaluate pipelines. #FDANGSWorkshop
9:20am February 25th 2016 via Twitter Web Client
Pfeifer: How to look at different VAFs, sequence artifacts possible, but still not comprehensive. #FDANGSWorkshop
9:19am February 25th 2016 via Twitter Web Client
Pfeifer: Brings up constructs / cell lines - what's the appropriate mix of mutations? (@SeraCare's approach) #FDANGSWorkshop
9:18am February 25th 2016 via Twitter Web Client
Pfeifer: Points to Boland's '13 WES cross-platform comparisons Venn diagam #FDANGSWorkshop
Pfeifer: DNA quality https://t.co/LYdMwlkCnc #FDANGSWorkshop measuring complexity as a fn of input amounts. 10ng - may have lost info
9:17am February 25th 2016 via Twitter Web Client
Pfeifer: Assessment of tumor cellularity. 2 labs, same specimens, but different results b/c of enrichment (microdissection) #FDANGSWorkshop
9:15am February 25th 2016 via Twitter Web Client
Pfeifer: The DNA you do get - lower (usable) quanity. What people don't look at: Add'l cytology specimens (FNAs) #FDANGSWorkshop
9:14am February 25th 2016 via Twitter Web Client
