Arming ColoAd1 with tumor necrosis factor alpha and lymphotoxin alpha

Colon and rectum cancers (CRCs) are the third most prevalent cause of cancer-related mortality. Advanced metastatic CRC has a very poor prognosis; indicating the need for improved therapy. Tumour necrosis factor (TNF) can induce an immune response against tumours and cause cell death of the tumour-associated vasculature. However, dose-limiting toxicity occurs with systemic TNF treatment. In order to assess the effects of expressing TNF and lymphotoxin α (LTA) locally in the tumour microenvironment, a syngeneic CT26 CRC model expressing soluble (sm), full length (fm) or membranebound (mbm) murine TNF or LTA under a doxycyclin-dependent promoter. Moreover, an oncolytic adenovirus (ColoAd1) was modified to express fm or mbm TNF or LTA under its major late promoter (MLP). ColoAd1 is a novel chimeric species B adenovirus based on serotypes Ad11p and Ad3 and is currently in clinical phase 2 trials. In the CT26 model, expression of mbm TNF decreases tumour growth and increases survival compared to control mice not receiving doxycyclin. Expression of sm or fm TNF had no effect on tumour growth or survival. Increased immune infiltration, especially myeloid cells (CD45+ CD11b+) was seen in mice bearing tumours expressing sm, fm and mbm TNF. ColoAd1 can express functional TNF constructs without a substantial impact on virus replication or yield in vitro. Arming ColoAd1 with TNF did not increase efficacy or innate immune infiltration compared to parental ColoAd1, in vivo. Administration of ColoAd1, ColoAd1 fm TNF and ColoAd1 mbm TNF increases immune cell infiltration compared to PBS, especially dendritic (CD45+ CD11b+ F4/80- CD11c+), monocytic (CD45+ CD11b+ Ly6Chigh Ly6G- ) and granulocytic (CD45+ CD11b+ Ly6Clow Ly6G+) cell populations were enriched in virus treated tumours. Expression of TNF from ColoAd1 was well tolerated and systemic toxicity was not observed, in vivo. A statitistically significant derecrease in ColoAd1 mbm TNF genome copies and nonstatistically significant decrease in ColoAd1 fm TNF genome copies were found in tumours compared to ColoAd1. This may indicate increased viral clearance or decreased replication of these armed viruses. The decrease in virus progeny in tumours may have been a major hurdle for increasing the efficacy of ColoAd1 TNF. Synergy between adenoviruses and radiation has been observed for species C virus but little is known about the effects of species B virus on the DNA damage response. In vitro, induction of γH2AX and phospho-ATM was seen upon ColoAd1 infection of DLD cells and compared to Ad5, less Rad50 and Mre11 degradation occurs in ColoAd1 or Ad11p infected cells. Like all other species of Ad studied, ligase IV is degraded during ColoAd1 infection. Combination of ColoAd1 with radiation therapy, in vivo, led to an increase in viral genome copies in the tumour lysate, 10 days post radiation. However, no synergistic effects of combining virus with radiation was observed. Additionally, contrary to other Ad5 viruses armed with TNF, no evidence for synergy between mTNF expressing ColoAd1 and radiation was found. In conclusion, ColoAd1 can deliver therapeutic proteins to the tumour microenvironment and could potentially be a very useful approach to treat refractive cancers. However, the limitations of relevant pre-clinical models will need to be overcome or new phase 0 mechanistic trial designs will be required to move the technology to reach full potential.