*4.6. Mammalian Orthoreovirus (Reovirus)*

Wild type reovirus has been tested clinically for various types of cancer including GBM, both by local and systemic administration [135,136]. Reovirus is a non-enveloped virus with 10 segments of dsRNA enclosed in its capsid. The main receptor that wild type reoviruses use to enter the cells is the junctional adhesion molecule A (JAM-A) [15]. Van den Hengel et al. tested a panel of primary GBM cell cultures, showing that they exhibit a large intertumoral variability in JAM-A expression, suggesting that reovirus efficacy may be hampered in low JAM-A expressing tumors [56]. Recently a second receptor was shown to mediate reovirus infection in the central nervous system (CNS), the Nogo receptor NgR1 [137]. The NgR1 expression in GBM cell lines has been established in various studies, however in a recent study it was shown that the NgR maturation, and thus expression to the cell membrane, is inhibited by transforming growth factor-β (TGF-β) 1, which is highly expressed by GBM cells [138–140]. The cell surface expression of NgR1 in primary GBM cells is yet to be elucidated.

Reovirus' natural tropism to replicate and kill tumor cells makes it an ideal candidate for oncolytic virotherapy. Early on, researchers attempted to elucidate the mechanism of tumor selectivity. It was found that the constitutively active Ras signaling pathway potentiates the reovirus replication via inhibition of PKR [141]. Additional studies have supported that Ras pathway activating mutations enhance reovirus replication [142,143]. Contrary to these findings, Twigger et al. showed that reovirus oncolysis does not depend on the status of EGFR/Ras/MAPK pathway in squamous cell carcinoma primary cell lines [144]. Another factor involved in host cell sensitivity to reovirus is determined by the levels of inhibitors of proteases, such as cathepsin B, that are required to disassemble the virus in the cytoplasm. Inhibition of these proteases restricts disassembly and inhibits viral replication [145].

In a clinical phase I trial for high grade gliomas and brain metastasis, the safety of reovirus after intravenous injection prior to brain surgery was demonstrated. The reovirus capsid σ3 protein and the virus RNA detection in tumors varied between the nine trial patients and correlated with the high proliferation index of the tumor cells (Ki67 expression) [136]. The efficacy of reovirus was also evaluated in a clinical trial for non-small cell lung cancer (NCT01708993), in which reovirus (Reolysin) was injected intravenously. A post-hoc analysis of the data obtained from this clinical trial showed a favorable trend for patients with p53 and EGFR mutations [146]. Another clinical trial (NCT01199263) testing oncolytic reovirus in ovarian cancer patients did not show any clinical benefit and authors stated that one explanation could be that only 20% of the ovarian cancer patients harbor Ras mutations [147].
