
Assistant Member @MSKPathology | Computational Oncology | Translational Biomarker Discovery | Cancer Genomics | Germline Predisposition to Cancer
New York, NY
Joined December 2012
- Tweets57
- Following300
- Followers242
- Likes999
Excited to share our latest in @NatureMedicine✨
We present a detailed clinico-genomic landscape of pancreatic adenocarcinomas
With important implications for RAS 🎯 therapies
Expertly co-led by @AnnaVarghese4 and @MariaAPerry. w/@EileenMOReilly & colleagues
@MSKCancerCenter
KEAP1 stood out as it is also a well-known prognostic biomarker in LUAD.
So, we asked if the outcomes to standard of care Tx in LUAD also differed by KEAP1 zygosity.
Indeed they did!
Only KEAP1 biallelic but not monoallelic patients had shorter PFS.
🙏@AdamJSchoenfeld /7
1
62
But, does TSG zygosity matter for patient outcomes?
It appears it does.
For several TSGs, we find that zygosity is strongly associated with overall survival (OS). These include Class 1 (TP53, PTEN, KEAP1, etc) as well as Class 2 (eg: SMARCA4) genes. /6
1
2
63
We are excited to share a new portal for integrated somatic and germline interpretation. signaldb.org. Thanks to the phenomenal scientific engineering team led by @gaojj with @inodb, Onur Sümer, Xiang Li. 15/n
Overall, nearly 27% of all cancers with high penetrance variants lack the lineage association or the somatic phenotype indicative of the tumor’s dependence on the germline allele. 13/n
In a clear case of tumorigenesis independent of the pathogenic allele, we highlight a patient with Lynch Syndrome who presented with two clonally independent primaries. We find biallelic inactivation and concomitant MSI in one, but neither in the other. 11/n
Biallelic inactivation of an MMR gene is an obligate event for MSI phenotype. However, not all biallelic losses lead to MMR deficiency. Here again, we observe the effect of lineage, together with zygosity, in modulating the penetrance of this somatic phenotype 10/n
How many drivers are required for tumorigenesis? Germline cancers, with already one-hit inherited, require on average one fewer somatic oncogenic driver mutations. (@OncoKB) 9/n









