Abstract

Background: Diffuse Large B Cell Lymphomas are still a significantly heterogeneous disease with few molecularly defined entities. Recently genomic based algorithms have been proposed for DBCL NOS disease classification, including the Lymphgen tool 1. Other large B cell lymphoma entities, including High Grade BCL Dual Hit or plasmablastic lymphoma remain poorly characterized according to this molecular based clustering methods. Here we sought to apply this tool for disease classification in a series of 109 large B cell lymphoma cases derived from GELTAMO clinical trials (NCT01848132 and NCT2015-005390-21) and retrospective multicentric series to confirm its application across different large B cell lymphoma entities. Methods: We retrieved FFPE and liquid biopsy diagnostic samples from 109 cases and performed histopathological confirmation, FISH testing and whole exome sequencing (Agilent Sureselect v6) by NGS. Sequencing data and available fusion data for BCL2 and BCL6 were used to classify the cases according to Lymphgen (https://llmpp.nih.gov/lymphgen). Results: 109 cases were classified as different large B cell lymphoma entities, including 31 DLBCL NOS GCB type, 8 HGBCL/DLBCL DH/NOS, 41 DLBCL NOS non-GCB type, 22 plasmablastic lymphoma, 3 THRLBCL, 1 PMBL, 2 DLBCL NOS and 1 3B Follicular Lymphoma. After WES median Tumor Mutation Burden was 972. Overall genetic classification by Lymphgen was obtained in 53 cases (49%) with the remaining 56 cases considered unclassified (molecular subtype other/NOS). Of the genetically identified cases, 18 were classified as ST2 (33%), 15 were considered EZB (28%), 5 BN2 (10%) and 3 MCD (5%). Interestingly 12 (22%, overall, 11%) were genetically composite, including 8, ST2/EZB; 1, ST2/BN2; 1, ST2/MCD and 2 ST2/EZB/MCD. According to histopathological subtype 17 GCB type DLBCL were classified as EZB (8, 47%), ST2 (5, 30%) and BN2 (1, 6%). 3 cases (17%) were genetically composite. 6 out of 8 cases of HGBCL/DLBCL DH/NOS were classified as either EZB (2), ST2 (2) or composite EZB/ST2 (2). 19 non-GCB type DLBCL were classified as ST2 (5, 26%), BN2 (4, 21 %), EZB (3, 16 %) and MCD (3, 16%). 4 cases were genetically composite (21%). Interestingly, plasmablastic lymphomas showed ST2 genetic features, found in 5 out of 22 (23%) and composite ST2/EZB in 2 cases. Genes overrepresented in the ST2 cluster include TET2, PRRC2C, STAT3,DOCK8 and CLTC. Plasmablastic Lymphomas classified as other showed recurrent mutations in PRRC2C, PABPC1, TRRAP, PRKCB, EP300 and TP53. Conclusions: ST2 and EZB genetic groups stand as the most frequent genetic subtypes of DLBCL NOS. While ST2 is found in either GCB and non-GCB subtypes, EZB is found preferentially in cases with the GCB phenotype, including GCB type DLBCL and HGBCL/DLBCL DH. In comparison with already published series we found a relatively higher prevalence of the ST2 genetic subtype among DLBCL NOS. MCD and BN2 genetic subtypes are restricted to the non-GCB subtype, but its prevalence is relatively low. Importantly, a significant fraction of cases (~50%) is still genetically unclassified (molecular subtype NOS) using WES and FISH-derived fusion data. Proper identification of the A53 cluster and relevant dysregulated pathways in the molecular NOS group require more research. Specific large B cell lymphoma entities such as plasmablastic lymphoma show features of ST2 genetic subtype in a significant fraction of tumors. These features include JAK/STAT3 pathway activation related genes and fit well with already published data 2,3. In addition, recurrent mutations in TET2, involved in epigenetic regulation of gene expression and TRRAP, target of C-MYC transcriptional activation, were enriched in this molecular subtype of plasmablastic lymphoma. Importantly ~70% remained unclassified. Mutations in this group are recurrently found in PRRC2C as well as PABPC1, both involved in VEGFA-VEGFR2 signaling. In summary, DLBCL NOS molecular subtyping supplies exact molecular classification in ~50% of our cohort. ST2 and EZB subtypes constitute ~60% of the cases with a fraction of ~15% of composite ST2/EZB. Application of the molecular classifier to plasmablastic lymphomas shows intraspecific genetic diversity and points toward dysregulation of JAK/STAT, MAPK/Erk and VEGFA-VEGFR2 signaling as well as MYC transcriptional activation. REFERENCES 1. PMID: 32289277 2. PMID: 33225311 3. PMID: 32273478

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call