The poor prognosis of pancreatic cancer patients using currently available therapies mandates novel therapeutics that combine anti-neoplastic potency with toxicity-minimizing cancer specificity. Employing an overlapping pathway screen to identify genes exhibiting coordinated expression as a consequence of terminal cell differentiation and replicative senescence, we identified human polynucleotide phosphorylase (hPNPase(old-35)), a 3',5'-exoribonuclease that exhibits robust growth-suppressing effects in a wide spectrum of human cancers. A limitation to the anti-neoplastic efficacy of hPNPase(old-35) relates to its lack of cancer specificity. The promoter of Progression Elevated Gene-3 (PEG-Prom), discovered in our laboratory via subtraction hybridization in a transformation progression rodent tumor model functions selectively in a diverse array of human cancer cells, with limited activity in normal cells. An adenovirus constructed with the PEG-Prom driving expression of hPNPase(old-35) containing a C-terminal Hemaglutinin (HA)-tag (Ad.PEG.hPNPase(old-35)) was shown to induce robust transgene expression, growth suppression, apoptosis, and cell-cycle arrest in a broad panel of pancreatic cancer cells, with minimal effects in normal immortalized pancreatic cells. hPNPase(old-35) expression correlated with arrest in the G(2)/M phase of the cell cycle and up-regulation of the cyclin-dependent kinase inhibitors (CDKI) p21(CIP1/WAF-1/MDA-6) and p27(KIP1). In a nude mouse xenograft model, Ad.PEG.hPNPase(old-35) injections effectively inhibited growth of human pancreatic cancer cells in vivo. These findings support the potential efficacy of combining a cancer-specific promoter, such as the PEG-Prom, with a novel anti-neoplastic agent, such as hPNPase(old-35), to create a potent, targeted cancer therapeutic, especially for a devastating disease like pancreatic cancer.