Gene Expression Profiling Using Nanostring™ Reveals Differences According to Location, Histology and MIPI Score in Mantle Cell Lymphoma

Abstract

Background Mantle cell lymphoma (MCL) is an aggressive B cell neoplasm characterized the by t(11;14)(q13;q32)/CCND1-IGH resulting in overexpression of cyclin D1. Although indolent variants are recognized, MCL is generally aggressive and incurable. The MCL International Prognostic Index (MIPI), gene expression-based proliferation signatures, Ki-67 proliferation index, p53 expression, and aggressive histologic features are used for risk stratification, but the prognostic groups defined by each of these markers are clinically heterogeneous, demonstrating that each parameter alone does not fully account for the clinical behavior of these tumors. In this study, we investigated the gene expression profile of 60 cases of MCL and evaluated which genes had expression patterns that correlate with different clinical and histopathological parameters. Design A total of 60 excisional biopsy specimens of MCL were selected for gene expression profiling. Forty (66%) and 16 (26%) cases were untreated and relapse/persistent samples, respectively; in 4 this unknown was unknown. Biopsy sites included lymph nodes (n=38), spleen (n=10), tonsil (n=6) and various other extranodal tissue sites (n=6). The morphologic features were classic in 27 and aggressive (blastoid/pleomorphic) in 33 patients. MIPI was calculated in selected patients with nodal involvement and whose biopsy was an obtained prior to treatment. Each block contained tumor cells representing ≥ 80% of all cells in the biopsy specimen. RNA was extracted from formalin-fixed paraffin-embedded tissue using the Qiagen Allprep FFPE Kit after deparaffinization, according to the manufacturer's instructions. Gene expression was quantified in 200ng of RNA on the Nanostring™ platform. Normalization for RNA loading was performed using the geometric mean of 40 housekeeping genes with a cutoff value 20. Standard QC and data processing were performed using the nSolver™ Analysis Software. Adjusted p-values were used according to the Benjamini-Hochberg procedure. Results Gene expression profiling was different according to the site of involvement of MCL. Compared with nodal involvement, MCL involving spleen showed overexpression of VEGFA (p <0.01) and NOS (p <0.05). These genes are associated with angiogenesis via the PI3K signaling pathway. Similarly, tonsillar MCL showed overexpression of SFN and HSPB1 (p <0.01) compared with nodal MCL; these genes are related to the MAPK pathway. Different gene expression profiles were observed with respect to histology. Compared to classic variant, blastoid variant demonstrated overexpression of CDK4 (p <0.01) and underexpression of KAT2B and PIK3CG (p <0.01) in all cases. However, no significant differences were found in the gene expression profiles between the classic and pleomorphic variants. We also compared gene expression profiles based on the MIPI score. Compared to cases with a low MIPI score, those with a high MIPI score showed significant overexpression of several genes including SFN, COL11A2, HDAC6 and TP53 (p <0.01). Discussion MCL is a heterogeneous neoplasm, both clinically and genetically. In the present study, significant differences in gene expression profiles were observed based on site of involvement (spleen vs lymph node vs tonsil). However, we cannot exclude the possibility that the result reflects gene expression of non-lymphoma cells in the microenvironment since we did not extract RNA from lymphoma cells. We observed different gene expression profiles in the blastoid variant, but not in the pleomorphic variant compared to the classic variant, suggesting that the biology of pleomorphic variant is probably between the blastoid and the classic variants, possibly closer to the latter. Different gene expression profiles were found with respect to MIPI score, providing scientific support that the MIPI score, in addition to being a useful stratification factor in MCL patients, is reflecting differences in underlying biology of MCL. Disclosures No relevant conflicts of interest to declare.