An evaluation of site-specific immune responses directed against first-generation adenoviral vectors administered by convection-enhanced delivery

BACKGROUND: Direct adenoviral vector injections into the brain have been used in clinical trials to treat patients with high-grade gliomas. However, a recent phase 3 trial using first-generation vectors failed to demonstrate significant survival benefits. Malignant gliomas infiltrate extensively through the white matter, making them difficult to treat, and chemotherapy is at best partially effective. Convection enhanced delivery (CED) represents a rationale approach for achieving widespread targeting of infiltrating tumour cells. Previous studies have demonstrated that infusions of particle numbers above a threshold level [10(8) plaque-forming units (pfu)] are associated with a pronounced inflammatory response in rat grey matter, although no such comparisons have been made with CED infusions into the white matter.

METHODS: In the present study, we investigated the distribution and immune response after the administration of 10(7) and 10(9) pfu of a first-generation adenoviral vector (Ad.CMV.EGFP) by CED in both small and large animal models.

RESULTS: We show that Ad.CMV.EGFP can be efficiently distributed by CED over large volumes of brain. A threshold vector dose of between 10(7) and 10(9) pfu was seen in both rat striatum and white matter, above which transgene expression was lost at 30 days. Furthermore, all adenoviral infusions were associated with evidence of significant tissue damage, as demonstrated by loss of neurones and astrocytes or the presence of extensive astrocytosis.

CONCLUSIONS: These results indicate that CED is capable of mediating widespread adenoviral vector distribution, although these vectors are associated with significant tissue toxicity that may render their safe application in clinical trials unfeasible.