Introduction: Diffuse large B-cell lymphoma, not otherwise specified (DLBCL, NOS), is the most common type of mature B-cell lymphoma. The heterogeneous genetic profile makes it a therapeutic challenge. Approximately 35% of cases of DLBCL, NOS do not respond to standard treatment, leading to relapse or refractory disease and consequent mortality. In recent years, new therapies, in particular immunotherapies, have emerged that offer potential solutions for patients with DLBCL, NOS, especially those with refractory or relapsed disease. The inclusion of immunotherapy, via PD-1/PD-L1 immune switches, in standard treatment protocols for DLBCL, NOS is still uncertain. Elevated expression levels of PD-1 and PD-L1, which influence immune cell function and tumour growth, have been observed in DLBCL, NOS and data on their prognostic significance are conflicting.
Aim: PD-1 and PD-L1 immune switch blockade therapy is becoming increasingly common in the treatment of various neoplasms. There is also great promise in the treatment of lymphomas, where there is a lack of data on the expression of PD-1 and PD-L1 on lymphoma cells (LC) and tumor-immune cells (TIC) in the tumour microenvironment. In our study, we investigated the expression of PD-1 and PD-L1 on LC, and other TIC (reactive lymphocytes, macrophages, monocytes, and NK cells), and examined whether this may influence the disease outcome.
Material and methods: We included patients with primary DLBCL,NOS, treated at the Oncology Institute of Ljubljana (OIL) between February 2004 and August 2022. Patients were aged 18 years or older, and were negative for human immunodeficiency virus and Epsten-Baar virus. Patients were treated with standard immunochemotherapy and any residual disease with radiotherapy. Clinicopathological characteristics were obtained from the hospital information system. The study was divided into two parts, a retrospective and a prospective part. The retrospective part included samples from 283 patients diagnosed with DVCLB, BDO (germinal centre B-cell (GCB) subtype and activated B-cell (ABC) subtype) who underwent lymph node biopsy and histological examination for suspected lymphoma. Haematoxylin and eosin-stained histological slides and immunohistochemistry (IHC) slides for diagnosis were obtained from the archive. Tissue microarrays were prepared from the tissue blocks and stained with PD-1/PAX5 and PD-L1/PAX5 double staining. Brown chromogen was used to label PD-1 or PD-L1, while red chromogen was used to label PAX5. Clinical features and PD-1 and PD-L1 expression on LC and TIC were correlated with progression-free survival (PFS) and overall survival (OS). In the prospective part of the study, we included 34 patients from whom we obtained a lymph node biopsy for suspected lymphoma. The tissue was fixed and embedded in paraffin, and then standard IHC stains for diagnosis were performed, and two additional tissue sections were cut and used for double staining of PD-1/PAX5 and PD-L1/PAX5. On collection of the lymph node biopsy, a small piece of tissue was taken into the in house cell collection medium. This was disintegrated and a cell suspension was prepared. This was used for flow cytometric (FC) measurements, and for cell block preparation. To study the expression of PD-1 and PD-L1 on TIC and LC, we used 3 separate antibody panels, where lymphoma cells, B lymphocytes, T lymphocytes, NK cells, monocytes and macrophages were studied.
Results: Analysis of retrospective results of PD-1 and PD-L1 expression on LC and TIC revealed that PD-1 was expressed on TIC in 38.4% and on LC in 8.8% of cases, while PD-L1 was expressed on TIC in 46.8% and on LC in 6.5% of cases. PD-L1 is more frequently expressed on LC in patients with ABC subtype. We also observed that patients without PD-L1 expression on the LC had a longer PFS. Multivariate analysis showed that the IPI index and PD-L1 expression on the LC had a statistically significant effect on the PFS. Patient age played the most important role in the assessment of OS. The prospective part of the study was mainly methodological, where we compared PD 1 and PD-L1 detection by FC, IHC on tissue blocks and IHC on cell blocks. FC analysis showed a predominance of T lymphocytes, on which PD-1 expression was highest, and PD-L1 was most frequently expressed on macrophages and monocytes. No differences in expression between the ABC and GCB subtypes were demonstrated. In IHC analysis, we found that PD-L1 expression was more frequent in the ABC subtype. When comparing the results between IHC on tissue and cell blocks, we found good to excenet concordance between the methods for PD-1 on LC and TIC and satisfactory concordance for PD-L1 on TIC. When comparing the results of FC and IHC on tissue blocks, we observed satisfactory concordance for PD-1 on TIC and PD-L1 on TIC. However for results of FC and IHK on cell bloks we found satisfactoru concordance only for PD-L1 expression on TIC. The concordance between the results of all three methods was satisfctory for the determination of PD-1 and PD-L1 on TIC
Conclusions: PD-1 and PD-L1 are expressed on the LC and TIC of DLBCL, NOS patients. We have shown an association between PD-L1 expression on the LC and a shorter PFS time, which is mainly characteristic of the ABC subtype of DLBCL, NOS. Double IHC staining with PAX5 has proven to be an effective method for evaluating the expression of PD-1 and PD-L1, as it allows a reliable differentiation between the LC and TIC. The PC method proved to be a useful and accurate method that allows us to further analyse TIC populations and subpopulations. Comparison of IHC on tissue blocks and cell blocks gave comparable results for PD-1 on LC and TIC, and for PD-L1 on TIC. We also demonstrated concordance when comparing FC and IHK on tissue blocks for PD-1 and PD-L1 on TIC, and PD-L1 on TIC when comparing FC and IHK on cell blocks. When comparing all three methods simultaneously, satisfactory concordance was demonstrated for PD-1 and PD-L1 on TIC.
|
