*4.2. Oncolytic Adenovirus*

Adenoviruses are non-enveloped double-stranded DNA viruses [59]. Oncolytic adenoviruses have been extensively explored and utilized in many clinical trials against several cancers [105]. Conditionally replicating adenoviruses (CRAds) have been modified in diverse ways to target oncogenic pathways frequently mutated cancers such as the retinoblastoma (Rb) or the p53 pathway [58,106]. Dl1520 (ONYX-015) was the first CRAd tested in a phase I clinical trial for recurrent gliomas, in which the safety of local peritumoral injection of the virus was shown [107]. Several possible selectivity mechanisms have been proposed for dI1520, including p53/p14ARF defects, aberrant late mRNA transport and cell cycle disruption, all of which may relate to the functions of the early viral E1B-55k gene which is deleted in this virus [11]. Whether any of these factors can serve as biomarkers for dI1520 response has not been evaluated.

The results of another phase I clinical study against recurrent malignant gliomas using the CRAd delta24-RGD (DNX-2401) were recently published [20]. This OV was engineered to selectively replicate in tumor cells with dysfunctional Rb pathway, which is the case in approximately 80% of GBM tumors [12,58]. Impressive anti-tumor effects were found, with 17% of the treated patients surviving beyond 3 years [20]. Another early clinical trial with this CRAd was conducted in our institute in patients with recurrent GBM (NCT01582516). A subgroup of patients revealed high concentrations of different cytokines in post treatment CSF samples, indicating that delta24-RGD can induce an inflammatory microenvironment, which is potentially key for its therapeutic efficacy [108]. Whether specific cytokines can serve as biomarkers for response requires further investigation.

Very few studies have focused on elucidating the resistance mechanisms in the nonresponder patients in oncolytic adenoviral trials. In one study, it was demonstrated that the IFN signaling pathway was upregulated in Ad5/3-∆24-resistant ovarian tumors compared to untreated tumors [109]. Moreover, the authors showed that the MxA, an ISG which is induced by IFN type I or type III signaling could provide a predictive marker for resistance to oncolytic adenoviral therapies [109,110].

Of great interest are the studies that have focused on the pre-treatment immune status of patients receiving oncolytic adenovirus. Specifically, it was shown that chronic inflammation was a negative predictive marker for response to oncolytic adenovirus therapy in different types of cancer [111]. Furthermore, high-mobility group box 1 (HMGB1), a nuclear protein secreted by immune cells and which is associated with a pro-inflammatory state and immunological cell death, could serve as a predictive and prognostic marker for oncolytic virotherapy with adenoviruses. The study suggested that patients with low serum HMGB1 have more robust anti-tumor responses after oncolytic adenovirus therapy [112]. It was hypothesized that a higher pro-inflammatory state as measured by HMGB1, leads to inhibition of viral replication. These results may suggest that the use of immunosuppressants for a limited amount of time during post virus administration, may improve response rates in this subgroup of patients.

Lastly, in a study using pancreatic cell lines it was shown that high expression of cyclin D1 enhances delta24-RGD-induced cytotoxicity [113]. Cyclin D1 activates the cyclindependent kinase 4 and 6 (CDK4 and CDK6), which then phosphorylates the Rb protein resulting in cell cycle progression. Over-expression of cyclin D1 has been observed in many cancer types like head and neck squamous cell carcinomas (HNSCC), pancreatic and breast cancer [114]. In GBM, the highest expression of cyclin D1 was observed in the proneural subtype, which may suggest that this subtype would benefit more from delta24-RGD treatment.
