Andrew Futreal (University of Texas MD Anderson Cancer Center, USA)
“Cancer Genomics and Evolution”
- tumors heterogeneous from one patient to another
- intratumor heterogeneity—a state where subclonal subpopulations within the same tumor accumulate different kinds of mutations over time (65% heterogeneity in RCC)
- 1 mutation/wk, 40 mutations / year
- APOLLO platform https://goo.gl/uRmjHX
- Moon shot platform
- 40% subclonal mutation vs 17% no subclonal mutation
- Immune signature mapping (n=54) → anti-CTLA4 –responder (n=8) and progressor (all sent to PCR-seq check for clonality), for progressors → anti-PD1 blockade → responder (n=17) progressor (n=29)
- TCR beta chain-seq for neoantigen
- Only individual biomarkers are predictive
- Mutation load and neoantigen and common drivers: gene mutations do not predict response to immune checkpoint blockade, but TCR clonality is seen in responders to anti-PD1
- Genome-wide copy loss correlates with responder (KEGG/immune response pathway loss) 這個結論與 Science 2015 這一篇一樣 (LA Garraway group (Van Allen et al., Science, 2015)
- Subclonal evolution under immune pressure (CLL & transplant), multiple rounds of chemotherapy driver genes
- Driver genes : ERG2 appears at later time / p53 also emerges after many years
- Clone T cells → do single cell seq → any special T cells responsible for eliminate some portions of tumor cells
—————————————————————-
From Ming-Han
Goal: Germline genomic vs extrinsic factors (microbe, chemical, environment) vs tumor heterogeneity
General: they separate cancer patients with many groups; whether they are response to anti-CTLA4 or anti-PDL1
@melanoma:
Before treatment (biopsy)-> treat with CTLA4 -> replies (biopsy) -> antiPDL1 -> relapse (biopsy) —> sequence tumor genomic and blood components
Conclusion:
- Patients response to PDL1: high CD8+T cells, tumor express PDL1 at high level.
- If patients’ clonality of T cells is higher: response to PD1 better.
- If genomic copy number tends to “loss”: both CTLA4 and PDL1 work well. If genomic tends to “gain” copy: no response.
- Loss of copy number in cancer genome tends to have the reduced immune response.
@CLL treatment: transfusion allogenic donor’s T cells? (FCR treatment, DLI? WES windows?)
-allosct? DLI?
@TRA framework: the Apollo TRA team:
Try to combine the clinical and genomic and whatever together for the more specific treatment in the future.