Virtual Library
Start Your Search
Victor Pontén
Author of
-
+
P2.02 - Biology/Pathology (ID 616)
- Event: WCLC 2017
- Type: Poster Session with Presenters Present
- Track: Biology/Pathology
- Presentations: 2
- Moderators:
- Coordinates: 10/17/2017, 09:30 - 16:00, Exhibit Hall (Hall B + C)
-
+
P2.02-023 - Targeted Gene Expression Profiling to Evaluate Minimal Diagnostic FFPE-Biopsies from NSCLC-Patients (ID 9786)
09:30 - 09:30 | Author(s): Victor Pontén
- Abstract
Background:
The molecular analysis of non-small cell lung cancer (NSCLC) is limited by the availability of only small biopsies or cytological specimens that are procured for diagnostic purpose. The nuclease protection method provides the possibility to analyze minimal amount of formalin fixed paraffin embedded (FFPE) tissue without previous extraction steps. We tested this technique and compared it to the traditional methods RNA sequencing (RNAseq) and immunohistochemistry (IHC).
Method:
The nuclease protection method (HTGmolecular) in combination with next-generation sequencing was used to measure gene expression of 549 immune-oncology genes in FFPE samples from NSCLC-patients. Standardized minimal tissue amounts were used for 12 samples (4 tissue circles, 4µm thick, 1 mm in diameter, from a tissue microarray). Of these tissue sections also two corresponding original tumor biopsy were analyzed. RNA sequencing data was available from a corresponding fresh frozen tissue as well as IHC annotation of the immune markers FOXP3, CDH1, CD20, CD44, CD3, CD4, CD8 and PD-L1 on the analyzed tissue cores.
Result:
Of the 12 core preparations, 9 samples were successfully analyzed and fulfilled the quality criteria in the first run, the three others in a second re-analysis. The mRNA expression profiles of 12 samples measured with HTG on minute FFPE samples and RNAseq from fresh frozen tissue showed most often good correlations (r=0.41-0.87). HTG based mRNA data correlated with IHC expression for 5 of 8 genes (PD-L1 r=0.76, CD44 r=0.75, CDH1 r=0.61, CD8 r=0.60, CD4 r=0.54). RNAseq data showed good correlations with IHC for only 3 of 8 genes (CD44 r=0.91, PD-L1 r=0.86, CD8 r=0.67). Also, the HTG data of the two biopsies demonstrated very good correlations to the corresponding tissue cores and the RNAseq data (r>0.91). Finally, technical replicates of 10 of the minimal tissue core samples measured in different laboratories revealed relatively good concordance (r=0.71-0.94).
Conclusion:
The applied nuclease protection technique opens the possibility to multiplex and analyze the immune profile of 549 genes in minimal diagnostic biopsies with a high success rate. This is of great value for clinical use or in NSCLC clinical studies where the amount of tissue often is a limiting factor in companion diagnostics.
-
+
P2.02-066 - Identification of Crucial Gene Targets in the in Situ Environment of Cancer by Google Network Ranking (ID 10151)
09:30 - 09:30 | Presenting Author(s): Victor Pontén
- Abstract
Background:
The vast majority of cancer driver genes in non-small cell lung cancer (NSCLC) are characterized by activating mutations or high gene copy number amplifications. To identify tumorigenic genes without any genomic aberrations remains difficult. Network analysis of gene expression provides the possibility to describe the relations of genes to each other and by that to estimate their importance.
Method:
To analyze gene networks of NSCLC we applied the PageRank algorithm that was established by Google primarily to order the importance of websites. Data from NSCLC cancer tissue (n=1002) and normal lung (n=110) were retrieved from the cancer genome atlas (TCGA) and the highest expressed genes (n=16000) were ranked according to their importance in normal lung tissue as well as in NSCLC tissue. Subsequently, the difference in rank between normal and cancer was analyzed. Four comparative categories were defined and were analyzed with respect to their cellular function (GO annotation) and survival. Additionally, organ specific (n=163), housekeeping (n=68) and lung cancer related genes (n=62) were compared in the networks.
Result:
Genes with the highest importance (top 100) in normal lung tissue were connected with cellular metabolic processes or membrane transport. In cancer, most genes (top 100) were related to cell cycle and mitosis, chromosomal localization and DNA processing. There was no overlap between the two lists. Organ specific genes increased in average in their rank (p<0.001) while housekeeping genes decreased (p<0.001). Notably, cancer related genes did not significantly change their relevance in the network. Among the genes (top 100) that increased rank from normal to cancer, many were related to antigen processing and presentation.
Conclusion:
The PageRank algorithm provides the possibility to unbiasedly evaluate the importance of genes in the gene expression network of cancer. Surprisingly, not traditional cancer related genes but several hitherto not recognized genes were identified to be of regulatory importance and may be target for therapy. Once again, our results indicate the significance of the immune response in changes related to cancer.
