LOW MOLECULAR DIVERSITY AND GENOTYPIC-PROGNOSTIC ASSOCIATIONS IN FOLLICULAR LYMPHOMA (FL): IDENTIFICATION OF NEW BIOLOGICAL MARKERS BY LARGE-SCALE GENE MICROARRAY

Abstract

Introduction: FL has heterogeneous biological behavior, variable clinical outcomes, and a natural history marked by histological transformation (HT). The management of FL and its outcomes are highly dependent on staging (CS) and tumor burden (TB). The identification of new molecular biomarkers is crucial, not only to understand FL's biology, but also, can guide better prognostic stratification, and determine new therapeutic targets. This study aims to outline the GEP in tissue samples from FL categorized by CS and TB using a large-scale gene microarray platform with subsequent expansion by RT-qPCR. Methods: This study involved 219 FL cases. FL patients were categorized into early-stage disease (I/II) (ES), advanced-stage (AS) (III/IV) with low tumor burden (LTB), and AS with high tumor burden (HTB) according to GELF criteria. The experimental phase involved 126 FFPE samples. Firstly, 42 samples (12 ES, 11 AS-LTV, 12 AS-HTV, 6 control lymph nodes and 1 control HELA) were submitted to the array using the Clariom D Array Human platform. The identification of differentially expressed genes (DEGs) between the 4 groups used the Transcriptome Analysis Console by the eBay ANOVA method. Bidimensional PCA sought to identify discriminatory DEGs clusters. A comparison between pairs was subsequently performed. The 7 most discriminatory DEGs were selected for a cohort expansion by RT-qPCR. Results: According to CS and TB, 14.6% of FL cases were categorized as ES, 17.8% as AS-LTB, and 67.6% as AS-HTB. With a median follow-up of 80.7 months, the median OS was 208.2, 130.5, and 126.3 months for ES, AS-LTB, and AS-HTB,p = 0.249. The median PFS was 113.3, 110.7, and 59.5 months, for the same groups, p = 0.018. The array identified 135,750 genes comparisons between the control, ES, AS-LTB, and AS-HTB. PCA analysis was not able to identify discriminatory DEGs clusters among the 4 groups. Pairwise comparative analysis identified 1913 DEGs among the 4 groups. Of these, 1759 (91.9%) were DEGs between FL and healthy controls, but only 218 (11.3%) were DEGs among the 3 different FL subgroups, evidencing a low molecular diversity in FL when categorized by CS and TB. In the cohort expansion by qRT-PCR, 7 pre-selected DEGs based on the microarray were tested, including the genes GNG7, ITGAM, AZN1, TIMD4, IL7R, PTGS2, and MTTL10. The comparison between the median expression of these genes among the 3 FL's groups showed that the expression of the ILR7 was downregulated with the progression of the disease, with marked hypoexpression in AS-HTB patients. These data may suggest downregulation of IL7R gene in the FLs natural history. We found that hypoexpression of the GNG7 (p = 0.024) and PTGS2 (p = 0.090) were associated with higher rates of disease progression/relapse, and hypoexpression of the GNG7 (p = 0.006) was associated with a higher probability of HT, highlighting the potential prognostic impact of these molecular biomarkers. Finally, we established phenotypic-genotypic associations, as hypoexpression of ITGAM was associated with higher rates of B-symptoms (p = 0.017) and bulky disease (p = 0.026), as well as hypoexpression of IL7R was correlated with a higher probability of bulky (p = 0.025) and extranodal involvement (p = 0.016). Conclusion: Although CS and TB are strong determinants of the prognostic heterogeneity of FL, this study did not find high molecular diversity by this categorization. We observed a relatively homogeneous GEP with low biological diversity in FL presenting different disease evolution phases. Moreover, new molecular biomarkers potentially implicated in FLs natural history, prognostication and specific clinical phenotypes were revealed.