The vast majority of human cancers are incurable once metastatic. Chemotherapy and radiotherapy can induce tumour growth inhibition or regression in some cases, but cancer progression and resistance to these standard therapeutic modalities inevitably develops. Further dose escalation to overcome resistance is generally limited by the resultant toxicity to normal tissues. Therefore, new agents with larger therapeutic indices between cancer cells and normal cells are needed. Such agents would have improved efficacy and/or reduced toxicity when compared to currently available modalities. Oncolytic viruses are being developed which may meet the above criteria through replication and cell lysis which is limited to cancer cells. The therapeutic index should be greatly enhanced through replication of these viruses. After one or two rounds of viral replication the amount of virus present in tumour tissue will increase by several logs whle the viral load and toxicity in normal tissue will be minimal.

The idea that viruses might be used as selective anticancer agents dates back almost a century. Patients with various malignancies were noted to have transient remissions following viral illnesses or rabies vaccination [l]. Although direct viral lysis of cancer cells was not proven in these patients, these case reports set off a flurry of experiments aimed at identlfying oncolytic viruses. By 1970, an estimated 38 different viruses with in vivo antiturnoural activity in animals or man had been identified [2-51. Why then, given the favourable attributes of a selectively replicating oncolytic virus,