[HTML][HTML] Explorative data analysis of MCL reveals gene expression networks implicated in survival and prognosis supported by explorative CGH analysis

S Blenk, JC Engelmann, S Pinkert, M Weniger… - BMC cancer, 2008 - Springer
S Blenk, JC Engelmann, S Pinkert, M Weniger, J Schultz, A Rosenwald…
BMC cancer, 2008Springer
Background Mantle cell lymphoma (MCL) is an incurable B cell lymphoma and accounts for
6% of all non-Hodgkin's lymphomas. On the genetic level, MCL is characterized by the
hallmark translocation t (11; 14) that is present in most cases with few exceptions. Both gene
expression and comparative genomic hybridization (CGH) data vary considerably between
patients with implications for their prognosis. Methods We compare patients over and below
the median of survival. Exploratory principal component analysis of gene expression data …
Background
Mantle cell lymphoma (MCL) is an incurable B cell lymphoma and accounts for 6% of all non-Hodgkin's lymphomas. On the genetic level, MCL is characterized by the hallmark translocation t(11;14) that is present in most cases with few exceptions. Both gene expression and comparative genomic hybridization (CGH) data vary considerably between patients with implications for their prognosis.
Methods
We compare patients over and below the median of survival. Exploratory principal component analysis of gene expression data showed that the second principal component correlates well with patient survival. Explorative analysis of CGH data shows the same correlation.
Results
On chromosome 7 and 9 specific genes and bands are delineated which improve prognosis prediction independent of the previously described proliferation signature. We identify a compact survival predictor of seven genes for MCL patients. After extensive re-annotation using GEPAT, we established protein networks correlating with prognosis. Well known genes (CDC2, CCND1) and further proliferation markers (WEE1, CDC25, aurora kinases, BUB1, PCNA, E2F1) form a tight interaction network, but also non-proliferative genes (SOCS1, TUBA1B CEBPB) are shown to be associated with prognosis. Furthermore we show that aggressive MCL implicates a gene network shift to higher expressed genes in late cell cycle states and refine the set of non-proliferative genes implicated with bad prognosis in MCL.
Conclusion
The results from explorative data analysis of gene expression and CGH data are complementary to each other. Including further tests such as Wilcoxon rank test we point both to proliferative and non-proliferative gene networks implicated in inferior prognosis of MCL and identify suitable markers both in gene expression and CGH data.
Springer