Cancer cells escape immune recognition by exploiting the programmed cell-death protein 1 (PD-1)/programmed cell-death 1 ligand 1 (PD-L1) immune checkpoint axis. Immune checkpoint inhibitors that target PD-1/PD-L1 unleash the properties of effector T cells that are licensed to kill cancer cells. Immune checkpoint blockade has dramatically changed the treatment landscape of many cancers. Following the cancer paradigm, preliminary results of clinical trials in lymphoma have demonstrated that immune checkpoint inhibitors induce remarkable responses in specific subtypes, most notably classical Hodgkin lymphoma and primary mediastinal B-cell lymphoma, while in other subtypes, the results vary considerably, from promising to disappointing. Lymphomas that respond to immune checkpoint inhibitors tend to exhibit tumor cells that reside in a T-cell-rich immune microenvironment and display constitutive transcriptional upregulation of genes that facilitate innate immune resistance, such as structural variations of the PD-L1 locus, collectively referred to as T-cell-inflamed lymphomas, while those lacking such characteristics are referred to as noninflamed lymphomas. This distinction is not necessarily a sine qua non of response to immune checkpoint inhibitors, but rather a framework to move the field forward with a more rational approach. In this article, we provide insights on our current understanding of the biological mechanisms of immune checkpoint evasion in specific subtypes of B-cell and T-cell non-Hodgkin lymphomas and summarize the clinical experience of using inhibitors that target immune checkpoints in these subtypes. We also discuss the phenomenon of hyperprogression in T-cell lymphomas, related to the use of such inhibitors when T cells themselves are the target cells, and consider future approaches to refine clinical trials with immune checkpoint inhibitors in non-Hodgkin lymphomas.

Central nervous system (CNS) involvement is rare in primary mediastinal large B‐cell lymphoma (PMLBCL). We aimed to evaluate the incidence of CNS relapse as first treatment failure event and the effect of the induction chemotherapy regimen, central nervous system ‐ international prognostic index (CNS‐IPI) and other clinical and laboratory variables on the risk of CNS relapse in 564 PMLBCL patients treated with immunochemotherapy. Only 17 patients (3.0%) received CNS prophylaxis. During a 55‐month median follow‐up only 8 patients experienced CNS relapse as first event, always isolated. The 2‐year cumulative incidence of CNS relapse (CI‐CNSR) was 1.47% and remained unchanged thereafter. The CI‐CNSR was not affected by the chemotherapy regimen (R‐CHOP or R‐da‐EPOCH). None of the established International Prognostic Index factors for aggressive lymphomas predicted CNS relapse in PMLBCL. The 2‐year CI‐CNSR in patients with versus without kidney involvement was 13.3% versus 0.96% (p < 0.001); 14.3% versus 1.13% with versus without adrenal involvement (p < 0.001); and 10.2% versus 0.97% with versus without either kidney or adrenal involvement. CNS‐IPI was also predictive (2‐year CI‐CNSR in high‐risk vs. intermediate/low‐risk: 10.37% vs. 0.84%, p < 0.001). However, this association may be driven mainly by kidney and/or adrenal involvement. In conclusion, in PMLBCL, CNS relapse is rare and appears to be strongly associated with kidney and/or adrenal involvement.

No abstract

No abstract