Abstract 2671

Introduction:

Primary mediastinal diffuse large B-cell lymphoma (PMBCL) is an uncommon disease with an aggressive clinical course and potential curability. Growing evidence suggests that host antitumor immunity suppression may play a role in resistant cases. The most studied candidate molecules are ligands PD-L1 (CD 274) and PD-L2 (CD 273) expressed by lymphoma cells, which effectively suppress host T cells. The PD ligand genes are located on chromosome 9p24.1 close to Janus kinase 2 (JAK2) gene. The clinical impact of PD-L1/PD-L2 protein expression has not been described in PMBCL.

Methods:

Tumor samples of 27 previously untreated patients were analyzed. Clinical characteristics were as follows: median age at diagnosis 35 years (20–74), female-to-male ratio 1.7:1, most patients (70%) had limited mediastinal disease and a mean tumor diameter of 10.7 cm. The IPI and aaIPI scores were low in 67% and 37%, intermediate-low in 26% and 41% and intermediate-high in 7% and 22%, respectively. No patients were assigned to a high risk group. All patients were treated with anthracycline-based chemotherapy 15% with CHOP and 85% with third-generation intensive regimen. Therapy was intensified in 70% of the cases with high-dose therapy and autologous stem cell transplantation. Most of the patients (70%) received rituximab and 15% were also treated with IF radiotherapy. Formalin-fixed and paraffin-embedded tissues were processed in routine tissue sections (approx. 5 micrometers) and placed on plus slides. After antigen retrieval with the use of the enzymatic or microwave oven processes, indirect immunohistochemistry with commercially available primary antibodies in optimized dilution was performed: CD20 clone L27, CD23 clone 1B12, CD30 clone Ber-H2, CD10 clone G27-P, Bcl-2 clone 100, Bcl-6 clone PGB6p, MUM1/IRF4 clone MUM1p, CD274 polyclonal, CD273 polyclonal, and HLADR clone TAL.1B5. For visualization, a secondary antibody with the standard avidin-biotin (ABC) method was applied. Results were expressed as a percentage of positive tumor cells and H-score (product of percentage of positive cells and staining intensity). Cytogenetic analysis with a locus-specific FISH probe (9p24) and arrayCGH was carried out in 15 (56%) of the patients.

Results:

Final treatment response was assessed in 26 (96%) patients (1 patient did not passed restaging procedures yet), CR was achieved in 22 (85%), PR in three and one patient progressed. After a median follow-up of 73 months (6.1 yrs), 22/26 (84%) patients are alive in the 1st CR, and only three patients died. Five-year PFS was 82.6% (95% CI 0.67–0.98) and five-year overall survival was 90.9% (95% CI 0.79–1.00). All samples expressed PD-L2 in (a median of) 80% of tumor cells with a median H-score of 90. PD-L2 protein expression was very low - six cases were negative and in positive cases, median expression was only 5% (H-score 5). HLA-DR expression was detected in all cases with a median positivity of 70% (H-score 140). Cytogenetic analysis detected amplification of 9p24.1 in 8/15 (53%) of the cases. When analyzing clinical characteristics, only correlation of high HLA-DR expression with limited clinical stage (p=0.04) and low IPI (p<0.01) was found. There was no correlation between treatment response quality and HLA-DR or PD-L2 expression, but high PD-L1 expression (above the median) correlated with non-CR status after treatment (p=0.07). Due to a low number of relapses, there was no relationship between protein expression and survival. No difference was found between cases with or without JAK-2 copy gain in terms of PD-1L expression (71% vs. 73%, p=0.92) or PD-L1 H-score (80 vs. 73, p=0.55); expression of PD-2L was higher (4% vs. 9%, p=0.19) in cases with JAK-2 amplification.

Conclusion:

Frontline intensive therapy is very effective in PMBCL patients. This is why no clear prognostic impact of protein expression of PD ligands or HLA expression was observed. There was constant high PD-L1 protein expression in PMBCL, low PD-2L expression and a high proportion of cases with JAK-2 gene amplification. Preliminary data show relationship between tumor immunogenicity (HLA-DR expression) and lymphoma aggressiveness. High PD-1L protein expression may probably influence treatment response quality. Further analyses are needed to clarify correlation between 9p24.1 amplification and PD-L protein expression. Supported by grants: MZ ÈR IGA NT 11103, LF-2012-007 and MSM 6198959205.

Disclosures:

Prochazka:Roche: Travel grants Other.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution