Immunotherapy of cancer with CD3-bispecific antibodies is an approved therapeutic option for some hematological malignancies and is under clinical investigation for solid cancers. However, the treatment of solid tumors faces more pronounced hurdles, such as increased on-target off-tumor toxicities, sparse T-cell infiltration and impaired T-cell quality due to the presence of an immunosuppressive tumor microenvironment, which affect the safety and limit efficacy of CD3-bispecific antibody therapy. In this review, we provide a brief status update of the CD3-bispecific antibody therapy field and identify intrinsic hurdles in solid cancers. Furthermore, we describe potential combinatorial approaches to overcome these challenges in order to generate selective and more effective responses.

BackgroundT-cell-engaging CD3-bispecific antibodies (CD3-bsAbs) are promising modalities for cancer immunotherapy. Although this therapy has reached clinical practice for hematological malignancies, the absence of sufficient infiltrating T cells is a major barrier for efficacy in solid tumors. In this study, we exploited oncolytic reovirus as a strategy to enhance the efficacy of CD3-bsAbs in immune-silent solid tumors.MethodsThe mutant p53 and K-ras induced murine pancreatic cancer model KPC3 resembles human pancreatic ductal adenocarcinomas with a desmoplastic tumor microenvironment, low T-cell density and resistance to immunotherapy. Immune-competent KPC3 tumor-bearing mice were intratumorally injected with reovirus type 3 Dearing strain and the reovirus-induced changes in the tumor microenvironment and spleen were analyzed over time by NanoString analysis, quantitative RT-PCR and multicolor flow cytometry. The efficacy of reovirus in combination with systemically injected CD3-bsAbs was evaluated in immune-competent mice with established KPC3 or B16.F10 tumors, and in the close-to-patient human epidermal growth factor receptor 2 (HER2)+ breast cancer model BT474 engrafted in immunocompromised mice with human T cells as effector cells.ResultsReplication-competent reovirus induced an early interferon signature, followed by a strong influx of natural killer cells and CD8+ T cells, at the cost of FoxP3+ Tregs. Viral replication declined after 7 days and was associated with a systemic activation of lymphocytes and the emergence of intratumoral reovirus-specific CD8+ T cells. Although tumor-infiltrating T cells were mostly reovirus-specific and not tumor-specific, they served as non-exhausted effector cells for the subsequently systemically administered CD3-bsAbs. Combination treatment of reovirus and CD3-bsAbs led to the regression of large, established KPC3, B16.F10 and BT474 tumors. Reovirus as a preconditioning regimen performed significantly better than simultaneous or early administration of CD3-bsAbs. This combination treatment induced regressions of distant lesions that were not injected with reovirus, and systemic administration of both reovirus and CD3-bsAbs also led to tumor control. This suggests that this therapy might also be effective for metastatic disease.ConclusionsOncolytic reovirus administration represents an effective strategy to induce a local interferon response and strong T-cell influx, thereby sensitizing the tumor microenvironment for subsequent CD3-bsAb therapy. This combination therapy warrants further investigation in patients with non-inflamed solid tumors.

Beta-lactoglobulin (BLG)-derived peptides may facilitate oral tolerance to whey and prevent cow's milk allergy (CMA). Loading of BLG-peptides in poly(lactic-co-glycolic acid) (PLGA) nanoparticles (Pep-NP) may improve this. Here we studied the uptake of NP and the capacity of NP and Pep-NP to activate bone marrow dendritic cells (BMDC). Furthermore, CMA prevention was evaluated by orally exposing three-week-old female C3H/HeOuJ mice to Pep-NP, NP or free peptides (PepMix) for 6 days before oral sensitization with whole whey protein and effects on the spleen and small intestine lamina propria (SI-LP) were studied. In BMDC, NP and Pep-NP enhanced CD40 expression and IL-6 and TNF-α secretion, while tended to decrease CD80 expression and prevented PepMix-induced IL-12 secretion. In vivo, oral exposure to Pep-NP, but not NP or PepMix, prior to whey sensitization tended to partially prevent the acute allergic skin response to whole whey protein. Splenocytes of NP-pre-exposed mice secreted increased levels of whey-specific IL-6, but this was silenced in Pep-NP-pre-exposed mice which also showed reduced TNF-α and IFN-γ secretion. In the SI-LP, Pep-NP pre-exposure reduced the CD4 T cell frequency in CMA mice compared to PBS pre-exposure. In addition, while NP increased whey-specific IL-6 secretion in the SI-LP, Pep-NP did not and maintained regulatory TGF-β secretion. This study presents a proof-of-concept that PLGA nanoparticles facilitate the capacity of BLG peptides to suppress the allergic response to whole whey protein. Hence, PLGA nanoparticles may be further developed as an adjunct strategy for BLG-peptide-based oral tolerance induction and CMA prevention.