Intracavitary ‘T4 immunotherapy’ of malignant mesothelioma using pan-ErbB re-targeted CAR T-cells

Highlights

• Mesotheliomas express EGF receptor in 80% and ErbB4 in 50% of cases.

• T-cells from mesothelioma patients are amenable to efficient genetic engineering.

• ErbB-targeted patient CAR T-cells mediate potent activity against mesothelioma.

• Tumor regression is not accompanied by toxicity.

• Clinical evaluation of this approach is warranted.

Abstract

Malignant mesothelioma remains an incurable cancer. We demonstrated that mesotheliomas expressed EGFR (79.2%), ErbB4 (49.0%) and HER2 (6.3%), but lacked ErbB3. At least one ErbB family member was expressed in 88% of tumors. To exploit ErbB dysregulation in this disease, patient T-cells were engineered by retroviral transduction to express a panErbB-targeted chimeric antigen receptor (CAR), co-expressed with a chimeric cytokine receptor that allows interleukin (IL)-4 mediated CAR T-cell proliferation. This combination is referred to as T4 immunotherapy. T-cells from mesothelioma patients were uniformly amenable to T4 genetic modification and expansion/enrichment thereafter using IL-4. Patient-derived T4⁺ T-cells were activated upon contact with a panel of four mesothelioma cell lines, leading to cytotoxicity and cytokine release in all cases. Adoptive transfer of T4 immunotherapy to SCID Beige mice with an established bioluminescent LO68 mesothelioma xenograft was followed by regression or eradication of disease in all animals. Despite the established ability of T4 immunotherapy to elicit cytokine release syndrome in SCID Beige mice, therapy was very well tolerated. These findings provide a strong rationale for the clinical evaluation of intracavitary T4 immunotherapy to treat mesothelioma.

Introduction

Malignant mesothelioma is a locally aggressive tumor that is primarily attributable to inhalation of asbestos fibers [1], prompting stringent regulation of this mineral in many territories. Nonetheless, incidence continues to increase in many countries, reflecting the long latency period of disease progression [2]. Mesothelioma remains incurable with a median survival of 9–17 months [3], [4].

Immunotherapy has long been posited as a logical approach to treat mesothelioma [5]. Promising data have emerged from clinical trials of immune checkpoint blockade in this disease [6]. In parallel, there has been increasing interest in the development of adoptive immunotherapeutic approaches using genetically engineered T-cells. Chimeric antigen receptors (CARs) are bespoke fusion molecules that couple an antigen recognition element to a tailored signaling domain, thereby enabling the HLA independent retargeting of T-cell specificity [7]. Interest has been fueled by the repeated demonstration of unprecedented therapeutic efficacy when autologous CD19 targeted CAR T-cells are administered to patients with refractory acute lymphoblastic leukemia [8]. The most successful CAR configurations employ a second generation design in which CD28 or 4-1BB costimulatory modules are fused to CD3ζ [9], [10], [11]. While initial experience in patients with solid tumors has been disappointing, mesothelioma may represent a special case since tumors are amenable to regional CAR T-cell delivery to the affected body cavity. Preclinical data have provided strong support for this approach using CAR T-cells retargeted against mesothelin [12], [13] and fibroblast activation protein [14], [15]. Clinical trials employing both intra-pleural [16], [17] and systemic delivery [18] are ongoing.

An increasing concern with CAR T-cell immunotherapy is the emergence of therapeutic resistance due to target loss [19], [20]. This highlights the desirability of selecting multiple targets that contribute to disease pathogenesis and thus are subject to selective pressure for retention by tumors. In mesothelioma, the extended ErbB receptor family represents an attractive candidate to meet these requirements. Epidermal growth factor receptor (EGFR; ErbB1) is upregulated and phosphorylated in the majority of cases, compared to normal pleura [21], [22], [23], [24], [25], [26], [27]. Disappointingly however, clinical evaluation of EGFR tyrosine kinase inhibitors has proven ineffective in this disease [24], [25], while trials involving EGFR inhibitory antibodies are ongoing (www.clinicaltrials.gov, accessed February 5th, 2017). Co-expression of HER2 (ErbB2) [24], [28] has also been reported, although ErbB3 and ErbB4 have not previously been evaluated.

To redirect T-cell specificity against the ErbB family, we have engineered a second generation CAR named T1E28z. T1E28z engages 8 of 9 possible ErbB homo and heterodimers [29] and enables T-cells to elicit antitumor activity in preclinical models of ovarian, breast and head and neck cancers [29], [30], [31]. The CAR is co-expressed with a chimeric cytokine receptor named 4αβ that delivers an interleukin (IL)-2/IL-15 signal upon binding of IL-4, thereby enabling the selective enrichment of CAR T-cells during ex vivo expansion [32]. Importantly, culture of the resulting ‘T4⁺’ T-cells in IL-4 does not alter their differentiation and maintains type 1 polarity [32]. T4 immunotherapy is currently undergoing Phase 1 evaluation by intra-tumoral delivery in patients with head and neck cancer [33]. This approach has a further theoretical advantage in mesothelioma treatment since IL-4 is commonly found in the tumor microenvironment, providing an additional source of stimulation for T4⁺ T-cells [34]. Here, we tested the hypothesis that T4⁺ T-cells derived from patients with mesothelioma would exert meaningful antitumor activity against mesothelioma, justifying the clinical evaluation of this approach.