Smarter combos, stronger T cells in lymphoma Free

In this issue of Blood, Sam et al¹ leverage humanized mouse models to rationally test candidate combination strategies to enhance the depth and durability of glofitamab-based therapies in non-Hodgkin lymphomas.

Bispecific antibodies (BsAbs) have emerged as a potent therapeutic strategy in B-cell lymphomas. These Abs simultaneously bind target cells expressing a defined antigen (such as, CD20) while binding and activating T cells through their CD3 epsilon subunit. They have shown substantial clinical activity in patients with both untreated and relapsed disease, including those whose disease progressed after chimeric antigen receptor (CAR) T-cell therapy, leading to their approval for both follicular lymphoma and diffuse large B-cell lymphoma (DLBCL). However, in contrast to CAR T-cell therapy, in which several mechanisms of response and resistance have been described, knowledge of the mechanisms of action and resistance of BsAbs is limited. The primary mechanism of adaptive resistance to treatment appears to be either genetic or post-transcriptional loss of CD20 expression.^2,3 The understanding of immune cell properties associated with response or resistance to therapy is also substantially limited. BsAbs induce a transient decrease in peripheral blood CD3⁺ cells (considered to reflect T-cell redistribution),^4-6 which is more pronounced in responders. However, in ex vivo studies, BsAb therapy induced greater tumor kill in combination with peripheral blood T cells rather than intratumoral T cells,⁷ suggesting that tumor-resident T cells may be at least in part resistant to BsAb-dependent activation. Currently, there are extremely limited data to support rational combination strategies to either limit antigen escape or maintain T-cell functionality in the setting of BsAb therapy. Despite this limitation, combination strategies are already being empirically used with several partners including naked Abs, Ab-drug conjugates, and small-molecule inhibitors (see table).

In DLBCL, the appropriate choice of partnering cytotoxic therapies is of particular importance because these are the mainstay of therapies for DLBCL, both in the frontline and relapsed settings. To determine the impact of glofitamab on the efficacy of conventional chemotherapeutic strategies, the authors used fully humanized mice with xenografts of DLBCL cell lines known to be partially resistant to approved multiagent chemotherapy regimens. They found that the administration of rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) substantially attenuates the efficacy of glofitamab, in part due to an impairment in the intratumoral accumulation of endogenous T cells that serve as the effectors of BsAb-mediated responses. In contrast, the addition of glofitamab to either polatuzumab vedotin monotherapy or the approved combination of R-CHP-Pola led to enhanced tumor control in vivo. The synergy between rituximab, cyclophosphamide, doxorubicin, prednisone, and polatuzumab vedotin, but not R-CHOP, with glofitamab was surprising considering that vincristine does not typically cause lymphopenia or attenuate T-cell responses. Instead, the authors found that polatuzumab vedotin therapy led to an increase in CD20 expression, which they suggested might sensitize tumors to glofitamab therapy. The mechanism by which polatuzumab vedotin enhances CD20 expression and whether this effect is maintained in the context of CD20-targeting therapy are intriguing subjects for future investigation.

The authors also observed substantial synergy when combining glofitamab with the combination of gemcitabine and oxaliplatin in both mouse models and blood samples from patients enrolled in a clinical trial, suggesting that this is a more suitable multiagent chemotherapy strategy to combine with glofitamab. This result is consistent with reports showing that gemcitabine does not attenuate T-cell activation. Moreover, gemcitabine-based strategies have shown substantial synergy with anti–programmed death-1 (PD-1) therapy in Hodgkin lymphoma, including rescuing efficacy in anti–PD-1 monotherapy-refractory patients.⁸ 

Finally, the authors provide evidence to support 2 potentially complementary methods to overcome glofitamab resistance. First, they suggest that the cotargeting of CD19 and CD20 provides additional benefits by limiting antigen loss-driven escape, which has been reported in patients treated with CD3-CD20 BsAbs. This finding may in part explain the observed synergy of glofitamab with polatuzumab vedotin, which targets CD79b. Moreover, given their observed increase in both the markers of T-cell exhaustion and accumulation of immunosuppressive regulatory T cells, the authors tested the addition of either regulatory T-cell–depleting Abs or PD-1/LAG-3–targeting inhibitory Abs and observed further synergy with glofitamab.

The results of Sam et al provide a rational framework upon which clinical trial combinations with BsAbs can be developed and underscore the importance of targeting mechanisms of resistance that are likely to emerge with prolonged therapy, including antigen loss and both autocrine and paracrine drivers of T-cell suppression.⁹ Given that BsAb therapy provokes an inflammatory reaction known as cytokine release syndrome, which can cause significant morbidity, combination approaches that further disinhibit T-cell immunity must be approached with caution and may be better suited for patients with relapsed disease rather than in the up-front setting. One of the caveats of this study includes the use of immortalized human cell lines, whose growth kinetics and genetics do not fully reflect human biology. This potential shortcoming could possibly be explored using recently published human lymphoma organoid methods.¹⁰ Moreover, several open questions remain whose answers will be key to continue advancing and understanding BsAb therapy in lymphomas. A deeper mechanistic understanding of how chemotherapy drug classes impact BsAb-driven T-cell activation, as well as the impact on the sequencing of chemotherapeutic agents with BsAb administration, will be critical to maximize the efficacy of these combinations.

Conflict-of-interest disclosure: S.A.V. is a consultant for Generate Biomedicines and receives research support from Bristol Meyers Squibb.