| Recently we showed that the large majority, if not all oral cancers are preceded by preneoplastic mucosal fields of genetically altered cells. A small subgroup of these preneoplastic fields are macroscopically visible as leukoplakia and erythroplakia lesions, but the large majority cannot be distinguished at clinical inspection. These fields cause primary cancers but also local relapses in treated oral cancer patients when not completely excised. This year we showed in a retrospective case-control study by analyzing genetic changes in the surgical margins of treated oral and oropharyngeal cancer patients that preoplastic fields in treated HNSCC patients that are not completely excised are indeed a risk factor for local relapse. In addition we succeeded in setting up a culture of preneoplastic cells. In total 45 mucosal biopsies surrounding oral cancers were cultured and in six cases a culture with genetically altered cells emerged. In one of these an immortalized prenoplastic culture was established. In addition we continued with validating our noninvasive microsatellite PCR assay using allelic loss analysis, which allows detection of preneoplastic fields in small brushed samples. Initial validation studies showed no alterations in 70 non-smoking controls without cancer history of whom 50 were age-matched to our patient cohorts. Of 16 leukoplakia lesions with genetic alterations 10 were found by the noninvasive assay. Most changes encompassed changes at 9p. Based on these data we calculated a sensitivity of 62% and specificity of 100% for this noninvasive assay. At present this is comparable to PAP cytology. Studies in surgically treated oral cancer patients showed preneoplastic fields that stayed behind in 6 of 20 patients of which 2 were detected by the brush assay. In a second high risk group for oral cancers, Fanconi anemia (FA) patients, we detected in 14 of 140 cases a preneoplastic field in the oral cavity and in two cases an invasive cancer. To treat these fields we initiated the development of retargeted conditionally replicating adenoviruses (CRAds). A problem hampering this approach is that the natural adenovirus receptor CAR is hardly expressed in normal and dysplastic mucosal epithelium. Previously we showed in an organotypic culture model closely mimicking the normal mucosa that this barrier can be overcome by retargeting with monoclonal antibodies directed against cell surface antigens on squamous cells, and we generated a bispecific antibody for retargeting of the viral vectors. We have analyzed a large panel of CRAd backbones that would allow replication in premalignant cells while leaving primary keratinocytes and fibroblasts intact, and selected the most promising one. Construction of the retargeted CRAd for preclinical testing is in progress. |
