Virtual Library
Start Your Search
J.D. McPherson
Author of
-
+
MINI 14 - Pre-Clinical Therapy (ID 119)
- Event: WCLC 2015
- Type: Mini Oral
- Track: Biology, Pathology, and Molecular Testing
- Presentations: 1
- Moderators:L. Fernandez-Cuesta, A.F. Gazdar
- Coordinates: 9/08/2015, 10:45 - 12:15, Mile High Ballroom 2c-3c
-
+
MINI14.12 - Genomic Profiling of Patient-Derived Xenografts Identify Passenger Aberrations Associated with Better Prognosis in Non-Small Cell Lung Cancer (ID 1735)
11:50 - 11:55 | Author(s): J.D. McPherson
- Abstract
Background:
Patient-derived tumor xenografts (PDXs) increasingly are being used as preclinical models to study human cancers, test novel therapeutics, and identify potential biomarkers, as they more accurately model human cancers than established tumor cell line cultures. However, uncertainty remains as to how well the genomic characteristics of patient non-small cell lung cancer (NSCLC) are recapitulated in these PDX models.
Methods:
PDXs were established by implantation of surgically resected NSCLC patient tumors into the subcutaneous or sub-renal capsule of non-obese diabetic severe combined immune deficient (NOD-SCID mice. Comprehensive genomic profiling including exome, gene copy number, DNA methylation and mRNA expression were conducted on 36 independent PDX models, their matched patient tumors and normal lung tissue. Publicly available cell line and TCGA data were used for comparison. Integrative analysis was performed to identify genomic alterations in PDXs that are associated with significant clinical outcomes in patients.
Results:
From 441 resected NSCLC tumors, 127 serially transplantable and stable PDX models were established. Among 264 NSCLC patients with at least 3-years follow-up, patients whose tumor formed stable PDXs (versus those who did not) showed significantly worse disease free (HR=3.12, 95% CI =2.02-4.83, P<0.0001) and overall survival (HR=4.08, 95% CI =2.16-7.73, P<0.0001), after multivariable adjustment for clinical pathological factors. Genomic and transcriptomic profiling of 36 PDXs showed greater similarity in somatic alterations between PDX and primary tumors than with published cell line data. In addition to known mutations, we found at least 16 non-synonymous somatic mutations in known oncogenes and tumor suppressors that have never been reported. All these mutations had higher observed variant allele frequency in PDXs compared to their matched patient tumors, suggesting that these were tumor sub-clones selected or enriched for growth in the PDXs. Tumor models characterized by a higher number of somatic alterations among 865 frequently altered genes were associated with better overall patient survival (HR=0.15, p=0.00015) compared to patients with corresponding PDXs characterized by higher alteration number; this was validated in the TCGA lung cancer dataset patients (HR=0.28, p=0.000022). These 865 genes were enriched for those encoding for proteins involved in cell adhesion and interactions with the extracellular matrix, and a quarter of the genomic alterations would putatively form neo-antigens implicating a potential role of immune response in the observed improved patient survival.
Conclusion:
PDXs are close preclinical models of patient tumors. Further investigations of passenger mutations may clarify their clinical impact on interactions between tumor cells, stroma, immune microenvironment and patient prognosis.
