A subset of DLBCL cases with 9p24.1 amplification reveals molecular and genetic features similar to PMBCL

Classical Hodgkin lymphoma (cHL) frequently harbors copy number alterations (CNAs) of 9p24.1, including chromosomal amplification, copy number gain, polysomy, or translocation. 9p24.1 CNAs have also been detected in primary mediastinal large B-cell lymphoma (PMBCL), primary central nervous system lymphoma (PCNSL), and primary testicular lymphoma (PTL).

CNAs of 9p24.1 result in the overexpression of the programmed death-ligand 1 (PD-L1), PD-L2, and Janus kinase 2 (JAK2).  The interaction between PD-L1/PD-L2 and the programmed cell death protein 1 (PD-1) receptor on T cells negatively regulate T cell-mediated immunity and a variety of cancers, by upregulating expression of PD-L1, evade immune surveillance. Pembrolizumab or nivolumab, two immune checkpoint inhibitors targeting PD-1, have demonstrated efficacy in relapsed and/or refractory cHL, and also in PMBCL, PCNSL, and PTL.

9p24.1 CNAs have also been observed in diffuse large B-cell lymphoma (DLBCL), the most common type of non-Hodgkin lymphoma (NHL), but little is known about the molecular significance or the impact on clinical outcome. In a paper published in Blood Cancer Journal, Yucai Wang and colleagues analyzed the frequency of 9p24.1 CNA in untreated newly diagnosed DLBCL, its association with the expression of PD-L1, PD-L2 and JAK2, and its correlation with clinical variables and outcome.¹

Study design and results

Patients (n = 199) with newly diagnosed DLBCL from May 2002 to September 2012 were included in the study. All patients were treated with standard immunochemotherapy and followed prospectively.

Frequency of 9p24.1 CNA in untreated newly diagnosed DLBCL

The frequency of 9p24.1 CNA was determined using whole-exome sequencing (WES; n = 57) data, or OncoScan™ (n = 142) and only cases with PDCD1LG2/PD-L2, CD274/PD-L1, and JAK2 gain were evaluated

• Three subgroups of CNAs were identified:
  • No gain of 9p24.1, n = 179
  • Gain, n = 14
  • Amplification, n = 6
• Overall, 10% of DLBCL have a 9p24.1 copy number gain or amplification

Fluorescent in situ hybridisation (FISH) analysis was performed to confirm the WES/OncoScan results. Of the 20 cases with gain or amplification, only nine (amplification, n = 4; gain, n = 5) had samples available for FISH

• The four amplified cases were all confirmed by FISH
• Of the five gain cases, two were confirmed, and of the other three
  • two were polysomy
  • one was an amplification and was considered an amplification for the remaining analysis

Thus, taken together the WES/OncoScan and FISH results identified 6.5% (n = 13) cases of copy number gain in 9p24.1, and 3.5% (n = 7) 9p24.1 amplification cases

Association of 9p24.1 CNA with PD-L1, PD-L2 and JAK2 expression

The expression of CD274/PD-L1, PDCD1LG2/PD-L2, and JAK2 was evaluated using RNASeq data on 31 of the cases used for CNA analysis:

• Amplifications, n = 3
• Gains, n = 2
• No gains, n = 26
• Significantly higher mRNA levels of PD-L1, PD-L2, and JAK2 were observed in patients with 9p24.1 amplification compared with no gain controls, whereas PD-L1, PD-L2, and JAK2 mRNA levels of patients with a 9p24.1 gain were similar to those with no gain. Thus, suggesting that only 9p24.1 amplification leads to increased expression of CD274/PD-L1, PDCD1LG2/PD-L2, and JAK2

Protein levels of PD-L1 were assessed by immunohistochemistry in 19 of the cases used for CNA analysis:

• Amplification, n = 5
• Gain, n = 4
• No gain, n = 10
• The amplification cases had elevated levels of PD-L1 compared with the no gain controls (74% vs 4% positive, respectively; p < 0.001). In the gain cases, PD-L1 increase was not significant compared with the no gain controls (15% vs 4% positive, respectively; p = 0.188)

Thus, suggesting that increased PD-L1 mRNA and protein expression correlate with chromosomal amplification of 9p24.1.

Correlation of 9p24.1 CNA with clinical features in DLBCL

Clinical characteristics of patients with amplifications, gains, or no gain were compared to evaluate whether 9p24.1 CNA correlates with clinical features in DLBCL:

• The average age was 51.4±20.2 for patients with 9p24.1 amplification, 57.2±13.3 for those with 9p24.1 gain and 63.8±11.8 for those with no gain of 9p24.1 (p = 0.008)

• No differences were observed in sex, Eastern Cooperative Oncology Group (ECOG) performance status, lactate dehydrogenase (LDH) level, number of extranodal sites, Ann Arbor stage, and the International Prognostic Index (IPI) score
• Regarding the subtype, the rates of the activated B cell-like (ABC)/non-germinal center B cell-like (non-GCB) subtype were 33.1% vs 66.7% vs 71.4% in no gain vs gain vs amplification, respectively (p = 0.019)

Association of 9p24.1 CNA with the outcome of newly diagnosed DLBCL

Clinical outcome after standard immunochemotherapy (such as R-CHOP), at median follow-up of 95 months, is shown in Table 1. Patients with 9p24.1 amplification showed a trend of better event-free survival (EFS) in comparison with no gain patients. On the contrary, patients with a gain of 9p24.1 had worse EFS than no gain patients, suggesting a different impact of 9p24.1 amplification and gain on clinical outcome.

Table 1. Clinical outcome after standard immunotherapy at a median follow-up of 95 months.

Molecular features of 9p24.1 amplification cases

• Clinical features of 9p24.1 amplification cases were similar to PMBCL: patients were predominantly young females that responded well to frontline standard immunochemotherapy
• An analysis of gene expression profiling (GEP) data from 38 9p24.1 CNA cases (3 amplifications; 3 gains; 32 no gains), revealed that 9p24.1 amplification cases had a GEP consistent with the PMBCL GEP signature
• WES data from 14 9p24.1 CNA cases (4 amplifications; 10 gains) revealed mutations in 60 genes out of the 150 DLBCL driver genes reported in the literature:
  • In the amplification cases, the mutation rate was lower than in the gain cases (p = 0.01)
  • Mutations in the DLBCL driver genes TP53, CD79B, MYD88, CREBBP, or EZH2 were observed in the gain cases but not in the amplification cases
  • A mutation of TNFAIP3 and CD58, occurring frequently in PMBCL, was observed in two amplification cases
• These data suggest that 9p24.1 amplification cases had molecular and genetic features similar to PMBCL

Conclusion

• Consistent with previous studies, these results showed that 9p24.1 copy number gain or amplification occur in 10% of DLBCL cases
• Only amplification cases had high levels of PD-L1, PD-L2, and JAK2 expression and molecular features similar to PMBCL. While a higher expression of PD-L1 is reported to be associated with a poor outcome, it is possible that these patients respond better to immune checkpoint inhibitors targeting the PD-L1/PD-1 pathway
• The association of 9p24.1 with high PD-L1 expression in the ABC/non-GCB subtype might be a mechanism to escape T cell-mediated anti-lymphoma activity and the combination of a PD-1 or PD-L1 antibody with standard immunochemotherapy may improve the outcome of these patients
• Ongoing phase II clinical trials are evaluating the combination of pembrolizumab (NCT02541565) or durvalumab (anti-PDL1 antibody; NCT03003520) with chemotherapy in untreated DLBCL
• Because of the small sample size, further studies are required to validate these results