Purpose: Pancreatic cancer is a devastating disease that is almost universally fatal because of the lack of effective treatments.We recently constructed a novel oncolytic virus (FusOn-H2) from the type 2 herpes simplex virus. Because the replication potential of FusOn-H2 depends on the activation of the Ras signaling pathway, we evaluated its antitumor effect against pancreatic cancer, which often harbors K-ras gene mutations. Experimental Design: Human pancreatic cancer xenografts were established in nude mice either s.c. or orthotopically (n = 8/group). FusOn-H2 was injected either directly (s.c. tumors) or by the i.v. or i.p. route (orthotopic tumors).Tumor volume, weight, and survival time were recorded for each animal. Statistical analyses were done by Student's t test. Results: A single intratumor injection of FusOn-H2 completely eradicated s.c. pancreatic cancers in all animals. Systemic injection of the oncolytic virus produced clear antitumor effects but did not abolish tumors in any animal. The most striking antitumor effect was seen when the virus was given i.p. Delivery of FusOn-H2 by this route completely eradicated established orthotopic tumors in 75% of the animals and completely prevented local metastases. Conclusions: FusOn-H2 has potent activity against human pancreatic cancer xenografts and may be a promising candidate for investigative virotherapy of this malignancy.Pancreatic cancer, which is usually diagnosed at an advanced stage, carries the highest fatality rate among all human cancers and accounts for the fourth highest number of cancer deaths in the U.S. (1). The dismal prognosis associated with this tumor has not improved over the past three decades, largely because effective treatments for metastatic disease have not been developed. This outlook may be changing due to the recently acquired ability to isolate or to genetically engineer viruses that could act as oncolytic agents in the treatment of solid tumors such as pancreatic cancer.Oncolytic viruses infect, replicate in, and destroy cancer cells without harming normal cells. Nonengineered oncolytic viruses are naturally occurring viruses that preferentially target tumor cells (2), but they are limited in number and are not available for all types of cancers. This disadvantage can be overcome with ''engineered'' oncolytic viruses that have been genetically modified to target cancer cells. In general, there are three ways to engineer viruses to target malignant cells. One approach is to alter surface viral proteins involved in the entry of viruses into cells, so that the virus is specifically targeted to tumor cells (3, 4). Another approach is to modify the virus to replicate only in the absence of a functional tumor suppressor gene, such as p53, which is mutated in many cancer cells (5). In the third approach, the virus is altered so that it will replicate only in dividing cells (6). Several viruses modified for oncolytic purposes are currently in clinical trials for a variety of solid tumors of different tissue origins. On...