BiTE-ing into prostate cancer with bispecific T cell engagers

SUMMARY:

Targeting prostate-specific membrane antigen (PSMA) has important therapeutic ramifications, more recently including immune oncology. Data were recently presented on the preclinical efficacy of a half-life extended bispecific T cell engager, AMG 160, which binds PSMA and CD3 to induce T cell-driven cytolytic activity against prostate cancer.

In this issue of Clinical Cancer Research, Deegen and colleagues (1) report pre-clinical evidence demonstrating T cell-mediated cytolytic activity against metastatic castrate resistant prostate cancer (mCRPC) potentiated by AMG 160, a novel half-life extended bispecific T cell engager (BiTE) that binds PSMA and CD3. BiTEs are an immunotherapy modality that induce antigen-specific immune activation by directly binding to the CD3 co-stimulatory domain on T cells while simultaneously engaging a tumor-specific antigen, such as PSMA, via a binding domain on the contralateral side of the molecule (2). Through direct engagement of the co-stimulatory CD3 receptor, this drug class bypasses the need for conventional tumoral T cell engagement at an immune synapse via T cell receptor recognition of tumor epitopes presented by major histocompatibility complexes.

Prostate cancer is immunologically “cold,” with adaptations against immune surveillance including downregulation of genes involved in antigen processing and presentation, upregulation of immune checkpoint proteins, and increased immune cell populations that depress effector T cell function (3). Therapeutic development that can overcome or bypass the challenges of a highly suppressive tumor immune microenvironment (TIME) may restore anti-tumor immune activity, creating possibilities for durable tumor-specific responses. To date, the only approved immunotherapy for mCRPC is sipuleucel-T, a dendritic cell vaccine intended to promote T cell responses targeting prostatic acid phosphatase. Sipuleucel-T has shown a median survival benefit of 4.1 months, despite not demonstrating PSA or imaging responses traditionally indicative of anti-tumor activity. Furthermore, published clinical trials of single-agent checkpoint inhibitors (CPIs) in mCRPC have produced limited results with a 5% overall response rate in unselected, highly pretreated patients (4).

Within this context, AMG 160’s BiTE predecessor, AMG 212 (pasotuximab), demonstrated encouraging results in a phase I trial with dose-dependent PSA reductions and measurable tumor responses in approximately a third of the 68 patients who were enrolled after progression on androgen deprivation therapy with abiraterone or enzalutamide and at least one taxane chemotherapy (5). However, the phase I study had several limitations: >90% of those receiving the subcutaneous formulation developed drug-neutralizing antibodies. Additionally, the short serum half-life (seven hours for the subcutaneous formulation, one to three hours for intravenous administration) necessitated switching to a continuous infusion strategy.

AMG 160 builds upon AMG 212 as a second-generation BiTE (Fig. 1) formulated to have an increased serum half-life through the addition of an Fc fragment attached to the core BiTE molecule, promoting transcytosis and recycling of the molecule through binding the neonatal Fc receptor. Deegan et al. first demonstrated that AMG 160 elicited T cell activation and cell killing in co-cultures of human T cells and prostate cancer cell lines. They subsequently showed dose-dependent anti-tumor effects in subcutaneous human prostate cancer xenograft tumors in immunodeficient mice treated with human T cells followed by AMG 160. They then examined in patient-derived xenografts the impact of a single human T cell infusion followed by AMG 160 in combination with anti-PD-1 CPI compared to AMG 160 or anti-PD-1 blockade alone. The combination with anti-PD-1 therapy demonstrated slightly earlier responses, though the final tumor sizes in the AMG 160 alone and combination AMG 160 and anti-PD1 treatment arms were similar at the end of the experiment, representing a fraction of the starting tumor size. Overall, these studies are consistent with prior human studies demonstrating an overall lack of efficacy of CPI monotherapy for mCRPC. They corroborate prior phase I activity of AMG 212 as a monotherapy, and introduce the potential for synergy with CPI therapies, which is the subject of an ongoing phase I clinical trial.

Figure 1.

A. Second Generation Bispecific T Cell Engagers (BiTEs) build upon prior technology through the addition of an F_C Fragment to the core antigen recognition domains. B. Second Generation BiTEs engage with their target antigens in a similar fashion to first generation BiTEs. C. Second Generation BiTEs have improved pharmacokinetic properties through recycling mediated by the neonatal F_c receptor (F_cR_N) as compared to first generation BiTEs.

To address potential hurdles in future clinical studies, Deegen et al. demonstrated in murine xenograft models that AMG-160 does not interfere with the signal intensity of ⁶⁸Ga-PSMA-11 PET/CT when compared to a non-PSMA-specific BiTE. This experiment provides evidence for the validity of PSMA-targeted nuclear imaging in the setting of a competing ligand, which in turn has important implications for response assessment. They subsequently showed in non-human primate dose-toxicology studies that AMG 160 has superior pharmacokinetic properties compared to AMG 212 with a terminal half-life of 6.1 days. They also explored a step-dose design to reduce potential acute toxicities from drug-induced cytokine release during infusion.

Limitations of these studies include the difficulty in extrapolating an effective dosing regimen in consideration for early phase human trials – the dosing regimen for their combination experiments with anti-PD-1 inhibitors in patient-derived xenografts (10 mg/kg administered every two days for a total of 12 doses) was escalated compared to their prostate cancer cell line xenografts (0.02-2 mg/kg administered weekly), representing a 20-fold cumulative dose increase compared to the initial cumulative maximum dose. Some insights into these issues have been addressed by initial results from the dose-exploration portion of a phase I study of AMG-160 in heavily pre-treated metastatic prostate cancer patients recently reported at the European Society of Medical Oncology (ESMO) Virtual Congress in September 2020 (6). Patients were divided into six dosing cohorts ranging from 0.003-0.9 mg every 2 weeks. While this study is not powered for efficacy, 68.4% of the 35 patients enrolled had a PSA decline, with 34.3% of patients achieving a PSA decline >50%. The greatest depth of PSA responses was observed in those receiving 0.03-0.9 mg every 2 weeks.

We applaud the authors’ anticipatory work modelling toxicities and pharmacokinetics in NHP models as well as potential impacts on PSMA PET imaging. The dose utilized to assess for interference was at the lowest pre-clinical dosing range of AMG 160 (0.2 mg/kg). While a maximum tolerated dose has not yet been reached in the above phase I study, the highest dosing cohort thus far is well below 0.2 mg/kg.

Additional biomarkers are needed to determine the heterogeneity of responses and the factors that drive exceptional responses in a subset of patients and ongoing barriers in non-responders. The dynamics of tumor PSMA expression in response to AMG 160 as measured by PSMA PET imaging, combined with conventional imaging, will be highly informative. Biopsies of metastatic lesions will also yield important insights into BiTE penetration and T cell infiltration and activity in the tumor microenvironment which are focuses of ongoing optimization. Furthermore, high resolution tumor immune profiling as well as genomic and transcriptomic analyses may enhance our perspective on worthy targets for combination therapies beyond CPIs.

The timing of these drugs with respect to other conventional prostate cancer therapies should be a focus of ongoing investigation. The goal of immunotherapy in the context of metastatic disease is to awaken an intrinsic self-sustaining and tumor-specific response. The optimal positioning of these drugs may be before the implementation of conventional agents like cytotoxic chemotherapy which have less potential for enduring responses, are not self-sustaining, and have cumulative systemic toxicities. Although novel immunotherapies are currently being investigated within the heavily pre-treated metastatic setting, there is an unexplored potential for synergy with other treatment modalities at earlier disease stages. For example, the use of immunotherapy in the adjuvant setting after radiation therapy in high-risk patients may take advantage of unique radiation-induced changes to the TIME that may prime responses against micrometastatic disease.

In summary, this data provides compelling evidence that bispecific T cell engagers like AMG 160 have significant activity against mCRPC, warranting further investigation in clinical trials with a focus on combinatorial efficacy with other immune oncology drugs.