Exploiting tumor-derived IL-10 activity in lymphoma therapy Free

In this issue of Blood, Garcia-Lacarte et al identify autocrine and paracrine activities of lymphoma cell–derived interleukin-10 (IL-10) in a murine diffuse large B-cell lymphoma (DLBCL) model that may be exploited for tailoring immunotherapy interventions.¹ 

Despite the overall success of anti-CD20 immunochemotherapy (rituximab, cyclophosphamide, doxorubicin, Oncovin [vincristine], and prednisone [R-CHOP]) in the treatment of the non-Hodgkin lymphoma DLBCL, a sizable fraction of patients relapse or are refractory to treatment. DLBCL is a heterogeneous diagnostic category. This heterogeneity is manifest both at the molecular level of the tumor cells and in their microenvironments. In this context, Garcia-Lacarte et al revisited the role of the cytokine IL-10, which is secreted by the lymphoma cells in many DLBCL cases.^2,3 IL-10 promotes lymphoma-cell survival by creating an autostimulatory loop through constitutive STAT3 activation.^3,4 It exerts pleiotropic, potentially immunomodulatory roles in various tumor-surveilling immune cell types in the lymphoma microenvironment (LME).⁵ Naturally, this could impact the efficacy of immunotherapies that are targeted at, or require the activity of, immune cells in the LME.

Garcia-Lacarte et al sought to tackle this unresolved question by employing an animal model of an aggressive subtype of activated B-cell-like DLBCL (ABC-DLBCL) that develops tumors with genetic (constitutive nuclear factor-ĸB activity, Tp53 loss, Prdm1 deletion; called the pBIC model⁶) and phenotypic characteristics of human ABC-DLBCL,⁶ in which the Il10 gene was conditionally deleted in the tumor cells (called pBIC^ΔIl10 model). The suitability of this animal model was assured by demonstrating that lymphoma cells produced IL-10 and required autocrine IL10/STAT3 signaling for growth and for upregulating PD-L1 immune-checkpoint expression. Given the known autocrine function of IL-10 for lymphoma-cell growth,^2-4 it was surprising to observe that IL-10-deficient pBIC^ΔIl10 mice developed more aggressive lymphomas compared with IL-10-proficient pBIC controls. Just as surprising, when Garcia-Lacarte et al administered anti-CD20 antibodies to tumor-bearing animals, mimicking the rituximab component of standard immunochemotherapy regimens, the IL-10-deficient pBIC^ΔIl10 lymphomas responded better than the IL-10-proficient pBIC lymphomas. This finding correlated with enhanced expression of calcium channels that were upregulated in the absence of IL-10/STAT3 signaling and with a dependence on extracellular Ca²⁺ for antibody-mediated cell killing after CD20-cross-linking in vitro (see figure).

Garcia-Lacarte et al then explored the paracrine effects of lymphoma cell–derived IL-10 in vivo by comparing the LME phenotypes in pBIC^ΔIl10 vs pBIC mice. These studies demonstrated that paracrine activities of lymphoma cell–derived IL-10 signaling contributed to the composition of the LME (see figure). Notably, cytotoxic CD8⁺ tumor-infiltrating lymphocytes in IL-10-deficient pBIC^ΔIl10 mice had a hyperexhausted phenotype, and stemlike CD8⁺ T cells were decreased. Thus IL-10 may prevent exhaustion of CD8⁺ tumor-infiltrating and maintain stemlike CD8⁺ T cells in the LME. This is of particular importance as stemlike CD8⁺ T cells are responsive to immune-checkpoint inhibitors.^7,8 Indeed, when tumor-bearing pBIC and pBIC^ΔIl10 mice were treated with combined anti-CD20/anti-PD1 immune-checkpoint inhibitor, the relatively poor response to anti-CD20 alone was enhanced by combination therapy in pBIC mice. In contrast, the superior response to anti-CD20 therapy alone observed in pBIC^ΔIl10 mice was not improved by combination therapy. Thus, in the absence of IL-10, inclusion of anti-PD1-mediated immune checkpoint inhibition provided no additional benefit. Garcia-Lacarte et al also provide evidence for a role of IL-10 in restraining the expansion and activity of growth-promoting tumor-associated macrophages, thus effectively limiting the progression of IL10-proficient pBIC lymphomas. IL-10 also had an antiangiogenic effect. Finally, regulatory T cells (Tregs) were enriched in the LME of IL-10-deficient pBIC^ΔIl10 lymphomas. Given the tumor-promoting role of this cell type, depletion of Tregs could delay tumor development in pBIC^ΔIl10 mice.

Can these unexpected findings be used in human DLBCL biology? Toward this goal, Garcia-Lacarte et al used IL-10-responsive genes distinguishing pBIC and pBIC^ΔIl10 lymphomas to derive a gene expression score for presumptive IL-10 activity. Mirroring the observation that IL-10-deficient pBIC^ΔIl10 lymphomas exhibited increased susceptibility to anti-CD20 immunotherapy compared with their pBIC counterparts, patients with DLBCL from published series⁹ were divided into prognostic groups. Patients with IL-10^hi-score had poorer overall survival with R-CHOP treatment than patients with IL-10^med/low-scores. Thus, the expression score derived from the activated IL-10/STAT3-axis is associated with inferior response to a standard anti-CD20 immunochemotherapy regimen in human DLBCL. Garcia-Lacarte et al propose that the IL-10 score or a corresponding (simpler) biomarker may help in the stratification of patients in the context of developing immunotherapy options. However, such biomarkers face significant hurdles to practical implementation in clinical settings. Expression-based signatures (eg, cell-of-origin or dark zone or molecular high-grade states) are well established but still only have limited adoption outside research settings. Factors that have impacted on this include the relatively unstable commercial provision of platform technologies and limited adoption of gene-expression assessment in clinical practice. Additional challenges lie in progressing a prognostic biomarker identified in retrospective cohort-based evaluation to the setting of a predictive biomarker used to inform therapeutic choice on individual cases in clinical trials and ultimately in routine clinical practice. As the increasingly adopted frontline DLBCL treatment polatuzumab vedotin plus rituximab, cyclophosphamide, doxorubicin, and prednisone (pola-R-CHP), which includes the anti-CD79b-directed drug conjugate polatuzumab vedotin, shows a particular survival benefit in ABC-DLBCL,¹⁰ the utility of the IL-10-score for routine clinical practice is not clear. Although the LME functionality of IL-10 requires further validation at the cellular and functional levels in human lymphoma and in the context of evolving treatment regimens, the findings by Garcia-Lacarte et al illustrate how evaluating the LME could be used to focus combination therapies that include immune-checkpoint inhibitors. Alternatively, one may contemplate such evaluation in the context of bispecific antibodies that act by enforcing the recruitment of lymphoma cell–targeting T cells.

Together, the new findings highlight the importance of understanding how lymphoma cells actively shape the cellular composition and functionality of the surrounding microenvironment and define a link between molecular pathogenesis, tissue-level tumor composition, and treatment response. Such knowledge is essential for establishing a framework that links targeted immunotherapies exploiting the specific features of the LME to the molecular pathogenesis of DLBCL. Clearly, an important pillar of this framework that warrants special consideration is IL-10 activity, as cogently elaborated in this excellent work that puts the spotlight on this peculiar cytokine in DLBCL biology.

Conflict-of-interest disclosure: The authors declare no competing financial interests.