Vaccine therapy + dasatinib for the treatment of patients with stage IIIB–IV melanoma

Melanoma remains a significant health concern worldwide, with disease incidence continuing to rise over the past four decades [1]. In the USA alone in 2016, the American Cancer Society estimates that 76,380 new cases of melanoma will be diagnosed, with 10,130 patients succumbing to this disease [1]. Although early stages of the disease are surgically curable and adjuvant therapy of high-risk disease is moderately effective in preventing recurrence, advanced-stage, metastatic melanoma has historically exhibited poor durable response rates to existing therapy options and high disease-associated fatality rates [2]. The landscape of first-line treatment options for advanced-stage IIIB–IV melanoma patients has been radically altered over the past decade with the advent of targeted small molecule inhibitors (SMIs; i.e., BRAFi, MEKi) and immune checkpoint inhibitors (ICB; i.e., antagonist antibodies against CTLA4, PD1 or PD-L1) [3]. SMIs have demonstrated profound, yet often transient, antitumor efficacy in the clinical setting, with the majority of treated patients ultimately developing treatment-refractory progressive disease [3]. Although superior rates of durable clinical response have been noted with ICBs and when combining ICBs such as ipilimumab (anti-CTLA4) and nivolumab (anti-PD1), such trials have also evidenced high rates (>50% incidence grade 3–4 for ipilimumab/nivolumab) and severity of immune-related toxicities, including autoimmune colitis [4]. Developing research for patient biomarkers predictive of durable objective clinical response or reduced risk for pathologic autoimmune sequelae should lead to more restrictive inclusion criteria and patient safety monitoring, allowing for an optimal clinical benefit-to-toxicity ratio to be realized for existing first-line and evolving combination approaches utilizing these agents. Despite optimism for such developing therapies, there remains an urgent need to develop second-line treatment options for the many advanced-stage melanoma patients that have or will fail current SMI- and/or ICB-based clinical trials [5].

Melanoma vaccine: promise & limitations

The frequent, but often short-lived, clinical benefits of ICB in the melanoma setting provide optimism for the curative potential of the patient’s immune system, despite a consensus perception of its dysfunctional state in patients with progressive disease [6]. In particular, there remains great enthusiasm for therapeutic vaccines designed to promote the activation and expansion of tumor-reactive T-cells (and B cells producing protective antibodies), and for adoptive transfer approaches implementing enriched populations of endogenous or genetically-engineered anti-tumor T-cells (i.e., adoptive cell therapy; ACT) [3,5]. Melanoma vaccines have evolved a rich history in the clinical setting over the past 20 years, demonstrating a high-degree of safety [7,8]. However, a meta-analysis of 56 Phase II/III melanoma vaccine trials (involving a total of 4375 patients) revealed that while vaccines were commonly effective in promoting tumor antigen-specific T-cell responses as measured in the peripheral blood of patients, they did not provide better overall disease control or extend overall survival versus standard of care [8].

So why have not melanoma vaccines performed better in the clinic? Clear limitations in such approaches include omnipresent concerns for immune checkpoint and regulatory cell (i.e., myeloid-derived suppressor cells [MDSC], tumor-associated macrophages [TAM] and regulatory T-cells [Treg])-mediated immunosuppression of antitumor T-cell responses in vivo. An additional limitation relates to the inability of vaccine-induced (melanoma-reactive) T-cells to be effectively recruited into tumor sites throughout the body in sufficient numbers to manifest clinically meaningful benefit(s) to the patient [3,6,7]. Indeed, the presence of persistent vaccine depots in the patient’s dermis serves as a sink to preferentially outcompete the tumor microenvironment (TME) in recruiting vaccine-induced T-effector cells [9]. The physical milieu of the TME (i.e., hypoxia, acidosis, high interstitial fluid and mechanical pressure) is also hostile to antitumor T-cell survival and function [10]. Furthermore, T-effector cells exhibit a primary dependency on glycolysis as a source of cellular energy, an addiction that is not sustainable in the poorly perfused, nutrient-deficient TME [10,11].

Vaccines targeting melanoma-associated blood vessels

An additional possible reason for the limited clinical success of vaccines targeting melanoma-associated antigens reflects the heterogeneity in antigenic phenotypes among cancer cell subpopulations within a given patient lesion, as well as across multiple lesions that may exist in a single patient. More specifically, tumor cell subpopulations are known to exhibit deficiencies in their expression of vaccine-targeted antigens, of MHC class I/II antigen-presenting molecules or of components in the antigen processing machinery, with such cells having clear survival advantages in the face of vaccine-induced immune surveillance [5,7]. Indeed, such clonal subpopulations of tumor cells are believed to serve as the foundation of immunotherapy-resistant progressive disease in previously treated patients [5,7].

A theoretical means by which to promote anti-tumor immunity, while coordinately circumventing the immune-phenotypic heterogeneity of melanomas, involves the development of vaccines targeting tumor-associated stromal cell populations, such as (myo)fibroblasts, vascular pericytes and vascular endothelial cells (VECs) [12,13]. We have recently shown that vaccination of HLA-transgenic mice against tumor blood vessel-associated antigens (peptides) results in therapeutic immunity against melanoma under conditions that are nonpermissive for direct T cell recognition of tumor cells [12]. Instead, these vaccines elicit CD8⁺ T-effector cell targeting of tumor (but not normal tissue) associated blood vascular pericytes and/or VEC [12]. Since the vaccine-inclusive peptides (derived from tumor vascular antigens) can be recognized by CD8⁺ T-cells isolated from the blood of HLA-A2⁺ melanoma patients after in vitro stimulation [12], this vaccine strategy was amenable to translation into clinical trials for the treatment of HLA-A2⁺ patients with advanced-stage melanoma.

Therapeutic ‘vascular normalization’

Based on core findings by Jain and colleagues [14], we believe that vaccine efficacy may be improved by the inclusion of agents that promote therapeutic vascular normalization (TVN), leading to the improved delivery of antitumor T-cells into the TME, with coordinate support for their sustained functionality in vivo. TVN results in the trimming of abundant nonfunctional venules within the tumor mass, maturation of vascular pericytes and VEC, improved tissue perfusion/oxygenation, and elimination of acidosis and high IFP in the TME [14]. A broad range of in-clinic drugs have now been demonstrated to promote at least transient TVN, including VEGF/VEGFR antagonists, tyrosine kinase inhibitors and even the anti-malarial drug chloroquine [14–17]. Interestingly, recent reports that integrate immune monitoring as a therapeutic end point have shown that TVN is associated with improved T-effector cell recruitment into tumors (based on upregulated expression of the CXCL9, CXCL10, CXCL11 and/or CCL5 chemokines in the TME), the removal of suppressor MDSC, TAMs and Treg in the TME, and a reduction in the expression of hypoxia-responsive immune checkpoint molecules (i.e., PD-L1) that inhibit tumor-infiltrating T-cell activity/survival [13–15]. Unfortunately, anti-angiogenic agents when used as monotherapies exhibit high rates of acquired drug-resistance (based on the activation of compensatory pro-angiogenic signaling pathways), leading to only transient clinical benefits for several months [14].

Dasatinib as a vaccine ‘adjuvant’ to promote TVN

Dasatinib (BMS-354825; NSC-732517) is a potent, broad spectrum ATP-competitive inhibitor of five critical oncogenic tyrosine kinases/kinase families: BCR-ABL, SRC, KIT, PDGFR and ephrin receptor kinases, known to impact tumor growth and metastasis [18]. Although dasatinib monotherapy has previously manifested disappointing clinical activity in patients with advanced-stage melanoma patients [19], we observed that it functions well as an adjuvant to specific vaccination in the MO5 murine melanoma model [16]. Hence, dasatinib + vaccine combined therapy promoted: superior expansion and melanoma recruitment of therapeutic T-cells (via locoregional production of CXCL9, CXCL10 and CXCL11, reduced tumor hypoxia and MDSC and Treg presence in the TME, evidence of ‘broadening/spreading’ in the anti-tumor T-cell repertoire (to include specificities that were not included in the vaccine formulation) and extended overall survival [16].

Development of combined vaccine therapy + dasatinib for the treatment of patients with stage IIIB–IV melanoma

As our preclinical data support:

• The ability of dasatinib to favorably condition the host for improved responsiveness to vaccination, and to better direct and sustain vaccine-induced T-effector cells into/within the TME via TVN [16]; and

• The therapeutic efficacy of vaccine targeting tumor-associated blood vessels under conditions where the immune system need not directly recognize tumor cells [12].

We developed a pilot clinical trial at the University of Pittsburgh Cancer Institute combining these modalities for the treatment of patients with advanced stage melanoma (NCT01876212).

This is an open single-center, prospective randomized Phase II study to determine the activity and safety of intradermal administration of Type-1, autologous dendritic cells (DCs) loaded with a mixture of six tumor vascular antigen (DLK1, EPHA2, HBB, NRP1, RGS5, TEM1)-derived peptides combined +/- daily oral administration of dasatinib (70 mg BID) applied as an immune adjuvant/conditioning agent [19]. This trial will evaluate the activity, safety and immune effects of the combination vaccine in up to 28 HLA-A2⁺ patients (≥18 years of age) with measurable, advanced stage IIIB–IV melanoma. Based on possible efficacy for dasatinib among patients with c-kit mutant melanomas [20], these patients will be excluded from study. Randomization will also be stratified to balance the two study treatment cohorts based on melanoma BRAFV600E status.

The primary objective of this study is to evaluate the effects of combination therapy of dasatinib and vaccine on specific immune response rates against the six tumor-associated vascular antigens. Secondary objectives include the evaluation of the objective clinical response rate, overall survival, progression-free survival and exploratory immunological end points, including the number of CD8⁺ T-cells, MDSC/Treg regulatory cells and blood vessels in melanoma biopsies, and the level of CXCL10 chemokine in patient serum pre- versus post-treatment.