Oncolytic virus-derived type I interferon restricts CAR T cell therapy

The application of adoptive T cell therapies, including those using chimeric antigen receptor (CAR)-modified T cells, to solid tumors requires combinatorial strategies to overcome immune suppression associated with the tumor microenvironment. Here we test whether the inflammatory nature of oncolytic viruses and their ability to remodel the tumor microenvironment may help to recruit and potentiate the functionality of CAR T cells. Contrary to our hypothesis, VSVmIFNβ infection is associated with attrition of murine EGFRvIII CAR T cells in a B16EGFRvIII model, despite inducing a robust proinflammatory shift in the chemokine profile. Mechanistically, type I interferon (IFN) expressed following infection promotes apoptosis, activation, and inhibitory receptor expression, and interferon-insensitive CAR T cells enable combinatorial therapy with VSVmIFNβ. Our study uncovers an unexpected mechanism of therapeutic interference, and prompts further investigation into the interaction between CAR T cells and oncolytic viruses to optimize combination therapy.

Abstract A029: Unexpected antagonism between oncolytic virus derived type I interferon and EGFRvIII CAR T-cells

Although the chimeric antigen receptor (CAR) T-cell platform has experienced clinical success in patients with hematologic malignancies, CAR T-cells specific to solid tumor targets have met with more limited efficacy. The highly inflammatory nature of oncolytic viruses and their ability to remodel the tumor microenvironment suggested to us that they would provide a complementary mechanism of action to both recruit and potentiate the functionality of CAR T-cells. VSVmIFNβ injection into B16EGFRvIII tumors increased the expression of chemokines such as CXCL10 and CCL5, which we would expect to recruit CXCR3+and CCR5+ EGFRvIII specific murine CAR T-cells. However, we did not observe an increase in overall survival or tumor control using the combination strategy compared to monotherapy with CAR T-cells. We recovered fewer viable CD8+ CAR T-cells from tumors injected with VSVmIFNβ early after adoptive transfer, and observed a similar reduction in the number of CD8+ CAR T-cells which persisted long term in the blood. We have shown that type I interferon increases the expression of the CAR from the retroviral LTR and in turn sensitizes these cells to tonic signaling mediated exhaustion and apoptosis. Correspondingly, CAR T-cells prepared from IFNAR KO T-cells were protected from the deleterious effect of type I IFN in vivo. We are currently investigating strategies to overcome the interference between these two modalities and to uncouple the regulation of the expression of the CAR from type I IFN signaling.

Citation Format: Laura Evgin, Amanda L. Huff, Phonphimon Wongthida, Jill Thompson, Timothy Kottke, John Sampson, Luis Sanchez Perez, Richard Vile. Unexpected antagonism between oncolytic virus derived type I interferon and EGFRvIII CAR T-cells [abstract]. In: Proceedings of the Fourth CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; Sept 30-Oct 3, 2018; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2019;7(2 Suppl):Abstract nr A029.

Predictors of luminal narrowing by neointima after angioplasty in atherosclerotic rabbits

OBJECTIVE

The present study was designed to identify the predictors of cross-sectional area narrowing by neointima (%CSAN-N) after balloon angioplasty (BA) in the cholesterol fed rabbit model.

METHODS

Angiographic, histomorphometric, and immunohistochemical data were analyzed from 91 femoral arteries of New Zealand white rabbits. Focal atherosclerosis was induced by air desiccation of the endothelium followed by a 2% cholesterol diet for 28 days. The rabbits received heparin (150 U/kg) at the time of BA (2.5 mm; three, 60-second, 10-atm inflations). Arteries were perfusion-fixed and excised 7 (n = 16), 14 (n = 11), 21 (n = 9), or 28 (n = 20) days after BA. Non-angioplastied arteries were de-endothelialized (cholesterol-fed [n = 12] or normal diet [n = 8]), non-injured but cholesterol-fed (n = 7), or normal (n = 8).

RESULTS

Univariate regression across all groups showed that the absolute area of the lumen by histomorphometry (LA) correlated significantly with the area bounded by the external elastic lamina (EEL) (vessel size), but no correlation was found with the absolute area of neointima or media, the percentage disruption of the internal elastic lamina (IEL), or the percentage of neointima and media occupied by foam cells. However, %CSAN-N correlated significantly with the area bounded by the EEL, significantly with the absolute neointimal area, and negatively with the absolute LA (p < 0.0001). Significant correlations were also found between %CSAN-N and the % IEL disrupted, the area of neointima and media occupied by RAM-11 + foam cells, and the loss of alpha-actin positivity in the media (p < 0.0001).

CONCLUSIONS

These studies show that neointimal formation contributes significantly to luminal narrowing 1 month after angioplasty in this model, that the degree of vascular injury and the extent of foam cell accumulation in the neointima and media are significant independent predictors of neointimal formation, and that the area of the neointima, and the percent narrowing by neointima, are important predictors of remodeling itself (EEL area). These predictors were not identifiable when the analysis was focused on the determinants of absolute luminal area alone.

Geometric remodeling is not the principal pathogenetic process in restenosis after balloon angioplasty. Evidence from correlative angiographic-histomorphometric studies of atherosclerotic arteries in rabbits

BACKGROUND

Restenosis after balloon angioplasty of coronary arteries is thought to be a proliferative response of the arterial wall to injury. Recently, it has been suggested that geometric remodeling of the arterial wall, rather than intimal fibromuscular hyperplasia, may be the major pathophysiological mechanism underlying restenosis. In this study, we evaluated the relative contribution of a geometric decrease in arterial size versus neointimal growth to luminal narrowing associated with restenosis after balloon angioplasty of atherosclerotic femoral arteries in rabbits.

METHODS AND RESULTS

Focal femoral atherosclerosis was induced by endothelial desiccation injury followed by a 2% cholesterol diet. After 1 month on the high cholesterol diet, the animals were subjected to one of four strategies: (1) balloon angioplasty, (2) balloon angioplasty followed by treatment with the factor Xa inhibitor antistasin, (3) combined laser and balloon angioplasty, or (4) no angioplasty. Animals were killed 2 hours or 28 days after angioplasty, and excised femoral artery segments were prepared for histomorphometric analysis. Angiography was performed serially before and immediately after angioplasty and before the animals were killed. An initial postprocedural gain in luminal diameter at sites of angioplasty was followed by a significant reduction in diameter by angiography and a significant increase in luminal cross-sectional area narrowing by plaque by histomorphometry 28 days after angioplasty compared to adjacent nonangioplastied segments of the same arteries, to nonangioplastied control arteries, or to angioplastied segments of animals treated with the factor Xa inhibitor antistasin. By contrast, the overall arterial size (cross-sectional area bounded by the external elastic lamina) at sites of restenosis was not significantly different from adjacent nonangioplastied segments in the majority of arteries excised at 28 days, and the mean overall arterial size at sites of restenosis was not significantly different from corresponding segments of nonangioplastied control arteries or from angioplastied segments of animals treated with antistasin. In the minority of angioplastied arteries in which the arterial size did change, most got larger.

CONCLUSIONS

Geometric remodeling resulting in a decrease in overall cross-sectional arterial size does not appear to be the principal pathogenetic mechanism for restenosis after balloon angioplasty with or without laser in this experimental model.

Response of femoral arteries of cholesterol-fed rabbits to balloon angioplasty with or without laser: emphasis on the distribution of foam cells

Very little is known about the structural composition of the restenotic plaque in evolution. The responses of atherosclerotic femoral arteries of rabbits to balloon angioplasty (BA), thallium/holmium/chromium: YAG infrared laser angioplasty (LA), combined LA and BA, or no angioplasty were compared by blinded quantitative histomorphometry and angiography. The endothelium was injured by nitrogen/air desiccation, and the animals were fed a 2% cholesterol diet for 1 month prior to the angioplasty procedure. Animals were sacrificed 2 hr or 28 days after angioplasty by pressure perfusion with 10% formaldehyde (100 mm Hg), and arterial segments (4-5 cm) were excised bilaterally. The frequency of thrombus was greatest in arteries with LA. Arteries with combined LA and BA had the greatest initial gain in luminal diameter by angiography, but they also had the greatest reduction in luminal diameter from 2 hr to 28 days and the greatest cross-sectional area narrowing by plaque at 28 days. The principal component of the intimal plaques in all groups was fibrous tissue (approximately 90%), with the remainder consisting primarily of "foam cells." By multiple regression analysis, the strongest predictors of cross-sectional area narrowing were contiguity of foam cells between the intima and media, depth of the tear, percentage of foam cells in the plaque, and the intervention of LA followed by BA. The principal predictors of foam cells in the plaque, irrespective of treatment, were also cross-sectional area narrowing, contiguity of foam cells between plaque and media, and the depth of tear. It is suggested that a large proportion of the foam cells of the intima may be derived from foam cells of the media and adventitia rather than from the lumen. These observations may be of particular importance regarding angioplasty in young people where foam cells occupy a significantly greater proportion of the atherosclerotic plaque.