Abstract

BACKGROUND

Pediatric aggressive large B-cell lymphomas (LBCL) share morphological and phenotypic features with adult types but seem to have better prognosis. Additionally, a specific subtype carrying IRF4 translocations (LBCL-IRF4) has been recently identified in this age group.

In adults, the cell-of-origin (COO) distinction of diffuse large B-cell lymphoma (DLBCL) based on gene expression signatures (germinal center B-cell like, GCB and activated B-cell like, ABC) and more recently clusters of genetic alterations have identified molecularly distinct DLBCL subsets that may benefit from novel therapeutic targets. The integration of pediatric population in these clinically relevant molecular subgroups and its clinical importance is not well known.

The aim of this study was to characterize the molecular heterogeneity of LBCL in pediatric and young adult patients and evaluate their potential clinical impact.

DESIGN

Sixty-one LBCL diagnosed in patients ≤25 years-old (median age 14 years old, male/female 38/23) were included in the study. Molecular analyses included fluorescence in situ hybridization for MYC, BCL2, IRF4 and BCL6, copy number (CN) analysis (Oncoscan, Affymetrix), COO Lymph2Cx assay (NanoString) and targeted next generation sequencing of 96 B-cell lymphoma driver genes (SureSelect XT, Agilent Technologies). CNA and mutational profiles were compared to those previously published in adult DLBCL. Correlation of molecular features and event free survival (EFS) was performed using Kaplan-Meier curves.

RESULTS

Histologically, 33 were DLBCL, 20 LBCL-IRF4 and 8 high grade B-cell lymphomas, not otherwise specified (HGBCL). Nodal disease was present in 57%, mainly in the cervical region. COO distribution was: 68% GCB, 18% ABC and 13% unclassified. Most of LBCL-IRF4 cases were GCB-COO (73%). The IRF4 translocation was demonstrated in 16 out of 19 cases diagnosed as LBCL-IRF4 and an IGH rearrangement was seen in the other 3. Five cases carried MYC-breaks (3 DLBCL and 2 HGBCL) and two cases carried BCL6-breaks (1DLBCL and 1HGBCL). BCL2 rearrangements were absent.

CN analysis detected alterations in 46/51 cases with recurrent gains (>15%) of 1q, 2p16, 11q, trisomies 7 and 12, and recurrent losses (>10%) of 1p36, 6q21-q22, 15q24, 17p13 and 19p13. Recurrent homozygous deletions were observed at 19p13/CD70 (6 cases), 9p13/CDKN2A (3 cases) and 13q14/RB1 (2 cases). Alteration patterns suggestive of chromothripsis were found in 10% (5/51) of the cases. No ABC-DLBCL related alterations such as 3p21-p14, 6q21-q25, 9p21.3 and 17p13 losses were seen.

Targeted sequencing detected a total of 434 variants in 44 of 47 cases (mean 9.2 mutations/case). A pipeline for selection of driver mutations revealed a total of 270 mutations (62%) with potential functional effect. Recurrent mutations found in >15% of the cases affected IRF4, SOCS1, PIM1, CARD11, ACTB and CCND3 genes. In comparison to adult DLBCL, pediatric and young adult cases had significantly lower incidence of MYD88 (5 cases), CREBBP,TP53 (3 cases each) and TNFRSF14 (2 cases) mutations which are strongly associated with the definition of established mutational clusters in adult DLBCL.

In our cohort, the morphological subtypes displayed different molecular profiles. IRF4 variants (some cases with >7 variants) and mutations in NF-kB pathway (CARD11, CD79B and MYD88-non L265P) were found specifically in the LBCL-IRF4 subgroup, whereas mutations in GCB related genes such as SOCS1 and EZH2 were particularly seen in cases with DLBCL diagnosis.

All 49 patients with available follow-up received chemotherapy as first line treatment (41% containing Rituximab). Variables significantly associated with poor EFS in univariate analysis were age >18 years, ABC-COO, Stage IV, high genetic complexity (chromothripsis and/or >10 CN alterations), 2p16/REL gain/amplification, 9p21.3/CDKN2A homozygous deletions, MYC rearrangements and mutations in MYC, TP53 and DDX3X genes.

CONCLUSION

Despite pediatric/young-adult LBCL having overlapping features with adult disease our findings suggest that LBCL in young age have specific molecular mechanisms and highlight potential key drivers of these lymphomas in this age group.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.