A CD38/CD28xCD3 Trispecific T-Cell Engager (TCE) As a Potentially Active Agent in Multiple Myeloma Patients Relapsed and/or Refractory (RRMM) to Anti-CD38 Monoclonal Antibodies (mAbs)

Background: Despite great advances in the treatment of RRMM, including anti-CD38 mAbs, patients become resistant and therapeutic options are limited. Some factors involved in resistance to anti-CD38 mAbs have been described. On the one hand, their efficacy is partially dependent on baseline CD38 expression in MM plasma cells (MMPC), but the treatment induces a rapid and sustained CD38 downregulation. On the other hand, their tumor-killing activity relies on NK cells, but its frequency rapidly decreases following initiation of treatment, and persistent NK cells exhibit a dysfunctional phenotype. Given the reduction of CD38 levels in MMPC and the impaired NK-cell activity in resistance to anti-CD38 mAbs, we hypothesized that RRMM patients previously treated with these agents could benefit from T-cell-based immunotherapeutic strategies that depend less on CD38 antigen density and NK cell function.

Aim: Evaluate a CD38/CD28xCD3 trispecific TCE (SAR442257) as a potential therapeutic agent in the RRMM setting.

Results: CD38 expression in MMPC and NK-cell frequency in bone marrow were significantly reduced in 29 RRMM patients previously exposed to anti-CD38 mAbs vs 45 newly-diagnosed MM (NDMM) patients (P <.001 and P =.013, respectively). Moreover, the percentage of CD57+ NK cells was significantly lower in RRMM than NDMM patients (P =.006), suggesting that both CD38 downregulation and a contraction of cytotoxic NK cells may be associated with resistance to anti-CD38 mAbs. By contrast, T cells did not change significantly between stages, which ignited further interest in the preclinical efficacy of the CD38/CD28xCD3 TCE.

First, we observed that, contrary to isatuximab and daratumumab, CD38/CD28xCD3 TCE did not reduce surface CD38 levels in the MOLP 8, RPMI 8226 and KMS 12 BM cell lines. Fluorescence microscopy imaging revealed that CD38 was present in cell membrane up to one week of treatment with CD38/CD28xCD3 TCE, while isatuximab induced an internalization of CD38 at 24h. Next, in a co-culture of the OPM-2 cell line with purified T or NK cells from healthy individuals (E:T ratio 2:1), we confirmed that the activity of CD38/CD28xCD3 TCE relied on T but not NK cells. Namely, no direct NK-cell stimulation was found whereas a significant T-cell activation was observed upon CD38/CD28xCD3 TCE treatment. Surprisingly, when OPM-2 were co-cultured with freshly-isolated peripheral blood mononuclear cells (PBMCs, E:T ratio 10:1), CD38/CD28xCD3 TCE induced activation of T- and NK-cells, suggesting a bystander NK-cell activation. This dual stimulation resulted in a significant reduction of OPM-2 cell viability at 48h. Interestingly, we found a comparable tumor-killing activity and T-cell activation in the presence of the CD38 null trispecific isotype (CD3xCD28 ΔCD38) compared with CD38/CD28xCD3 TCE, suggesting that the latter targets CD28 not only in T-cells, but also in OPM-2 cells, which express high CD28 basal levels. Altogether, these data demonstrate the potential dual CD38/CD28 targeting in RRMM patients previously exposed to anti-CD38 mAbs, particularly in those with double positive MMPC. Additionally, CD38 regulation was analyzed in the co-culture system (measured with a non-competing anti-CD38 antibody), and an unaltered and increased CD38 expression was observed in OPM-2 and T-cells, respectively, indicating again T-cell stimulation.

We tested the ex vivo efficacy of CD38/CD28xCD3 TCE (48h, 700 pM) in primary bone marrow samples from eight RRMM patients, which were collected after a median of 10 months since the last anti-CD38 treatment (range, 0.5 – 37). T-cell activation and MMPC lysis were found in six and four samples, respectively, which was accompanied by a reduction in the frequency of CD4 T cells (from 100% to 76%; P =.014). A positive and borderline-significant correlation (r = 0.764, P =.057) was found between the response to CD38/CD28xCD3 TCE and the time since last anti-CD38 treatment, and no differences in CD38 expression were observed between CD38/CD28xCD3 TCE responders and non-responders.

Conclusions: We observed a reduction in CD38 levels in MMPC and an impaired cytotoxic activity of NK cells in RRMM patients previously exposed to anti-CD38 mAbs. We propose a second targeting of CD38 using T-cell-based immunotherapeutic agents whose efficacy depends less on CD38 antigen density, especially after longer washout periods, as a potential therapeutic strategy in the RRMM setting.