Tombers, Nicole M.

Object The goals of this retrospective cohort study were as follows: 1) to describe the long-term prevalence and timing of hearing deterioration following low-dose (12- to 13-Gy marginal dose) stereotactic radiosurgery (SRS) for vestibular schwannoma (VS); and 2) to identify clinical variables associated with long-term preservation of useful hearing following treatment. Methods Patients with serviceable hearing who underwent SRS for VS between 1997 and 2002 were studied. Data including radiosurgery treatment plans, tumor characteristics, pre- and posttreatment pure tone average, speech discrimination scores, and American Academy of Otolaryngology–Head and Neck Surgery hearing class were collected. Time to nonserviceable hearing was estimated using the Kaplan-Meier method. Univariate and multivariate associations with time to nonserviceable hearing were evaluated using Cox proportional hazards regression models. Results Forty-four patients met the study criteria and were included. The median duration of audiometric follow-up was 9.3 years. Thirty-six patients developed nonserviceable hearing at a mean of 4.2 years following SRS. The Kaplan-Meier estimated rates of serviceable hearing at 1, 3, 5, 7, and 10 years following SRS were 80%, 55%, 48%, 38%, and 23%, respectively. Multivariate analysis revealed that pretreatment ipsilateral pure tone average (p < 0.001) and tumor size (p = 0.009) were statistically significantly associated with time to nonserviceable hearing. Conclusions Durable hearing preservation a decade after low-dose SRS for VS occurs in less than one-fourth of patients. Variables including preoperative hearing capacity and tumor size may be used to predict hearing outcomes following treatment. These findings may assist in pretreatment risk disclosure. Furthermore, these data demonstrate the importance of long-term follow-up when reporting audiometric outcomes following SRS for VS.

Abstract Chromosome rearrangements that result in gene fusions have important roles in the initial steps of tumorigenesis especially in leukemias and lymphomas, but the biological and clinical impact of gene fusions in the common solid tumor has been less appreciated. We previously generated RNAseq data from 12 oropharyngeal tumors along with matched normal tissue from the same patients. We utilized and compared two bioinformatic pipelines, SnowShoes-FTD and FusionHunter to identify two novel fusion events involving of MGST3-ZMAT5 and MS4A7-DKFZp586P0123. These fusions were validated utilizing. Reverse transcriptase polymerase chain reaction (RT-PCR) and Sanger sequencing validated. In the first translocation the 5′ UTR of MGST3 was fused with exon 1 of ZMAT5, and in the second translocation the exon 2 of MS4A7 was fused with 3′ UTR of DKFZp586P0123. These precise fusions were not present in the matched normal tissue, however, they were only found in the single tumor in which they were identified. Given a growing and powerful set of high-throughput massively parallel sequencing technologies, the future for genome-wide analysis of chromosome rearrangement appears open in many directions. Ultimately, we hope that this integrative methodology will elucidate key aspects of tumor biology as well as generate novel targets for cancer diagnostics and therapy. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 3100. doi:1538-7445.AM2012-3100

Abstract There are a number of human genes which span very large chromosomal regions. Forty genes span greater than 1 megabase and 200 are 500 Kb or larger. We and others have demonstrated that many of the largest genes are derived from within regions of profound genomic instability known as common fragile sites (CFSs). We have been studying oropharyngeal cancers which are derived from the base of tongue, tonsil and larynx. While the traditional etiologic factors underlying the development of this aggressive cancer is a history of smoking and drinking there are an increasing number of younger patients developing this cancer without these risk factors. The most likely cause of these cancers is the presence of human papillomavirus which we detect in over 90% of oropharyngeal cancers derived from the Mayo Clinic. In an attempt to better understand the molecular alterations that underlie oropharyngeal cancers arising from different risk factors we utilized the power of Next Generation sequencing on the Illumina GA IIx to compare the transcriptomes of 12 oropharyngeal cancers to patient matched normal tissue. We first observe a division of patients into two groups of oropharyngeal cancers: those caused by mutations and other alterations induced by a history of smoking and drinking and those associated with HPV infection caused by viral-induced genomic instability. We then examined the expression of the 200 largest human genes in the oropharyngeal tumors and find that each tumor has inactivated the expression of at least 20 different very large genes. There is a distinct pattern of inactivation of these genes in different oropharyngeal cancers depending upon the underlying etiology. Some of the genes most frequently inactivated include Dlg2 (the human homolog of the large discs tumor suppressor gene from drosophila and a mutation target in a number of other cancers); ERBB4 (Her4 and a member of the EGFR family), BAI3 (a brain specific angiogenesis inhibitor that is also mutated in multiple cancers), and ANK2 (mutated in long QT patients). Each of these genes may be functioning as tumor suppressors as has already been demonstrated for the two large CFS genes, FHIT and WWOX. The inactivation of so many of these genes in each oropharyngeal tumor may have a profound effect upon the phenotype of the resulting tumor. When we examine the expression of the large genes most frequently inactivated in each of the two groups of oropharyngeal cancers in panels of breast or lung cancer samples we find that we do not detect significant loss of expression of these genes. We believe that genomic instability promotes alterations in the unstable CFS regions but that there is selective inactivation of different large CFS genes in different cancers. The signature of which specific genes are inactivated may have phenotypic and clinical significance which we are just beginning to investigate. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 3885. doi:10.1158/1538-7445.AM2011-3885