Abstract 3904: Array CGH analysis of cervical cancer

Proceedings: AACR 102nd Annual Meeting 2011‐‐ Apr 2‐6, 2011; Orlando, FL Introduction: We hypothesize that there are characteristic chromosomal gains and losses in invasive cervical cancers that correlate with histologic cell type and HPV genotype. Methods and Materials: Array CGH analysis was performed on DNA extracted from 39 formalin-fixed, paraffin embedded invasive cervical cancer cases using 1.4MB BAC microarray chips (HumArray3.2, UCSF Comprehensive Cancer Center Array Core). Significant gains and losses were determined looking for convergent results from a random permutation approach based on GISTIC (Genomic Identification of Significant Targets in Cancer, Beroukhim R et al: PNAS 104: 20007, 2007), a modification of this approach using weighted sums over neighboring loci, frequency of gains and losses at each locus, and t-tests comparing log2 ratios from the cases to those from four control samples from non-cervical cancer tissue. The log2 ratios derived for each of the 2464 individual loci were aggregated for each locus over the entire set of 39 cases and for defined subsets including histologic subtype (squamous [SC:N=19] cancers, adenocarcinomas [AD:12], adenosquamous carcinomas [AS:6]) and HPV genotype [HPV16:12 and HPV18:14]. Results: Significant gains (108 individual loci) and losses (110 loci) were identified in the overall population. Congregations of significant gains were found on chromosomes 3q (8 loci), 1q (7), 1p (5) and losses on 4q (8 loci), 11q, (6) 4p (5), 3p (5). Overall,17 individual loci were significantly increased and 19 were decreased across the overall population as well as for each of the histologic subsets suggesting associations with cervical cancer per se. Individual loci with the most significant gains (P<0.001) were 3q26.3-3q27 (gained in 95% of cases), 19p13.2 (98%), 1q42-43 (93%) and losses (P<0.001) were 4q22 (100%), 11q22 (100%), 3p14 (90%) and 7p12.1 (100%). Differential changes associated with histologic cell type (P<0.001) included multiple gains in 1p (SC), 20q (AC), and losses in 3p (SC), 4p (AS). Multiple gains in 1p and 3q were related to HPV16 lesions while gains in 5p and 19q were related to HPV18. The HPV associations were somewhat influenced by cell type: 8/9 (89%) of AD (3 untyped) were HPV16, but only 2/18 (11%) of SC (1 untyped) were HPV16. Conclusions: Characteristic chromosome copy number alterations for regions and individual loci were identified for cervical cancer cases including patterns for different histologic subtypes and HPV genotypes. The regions harboring these loci can potentially be developed for screening, assessment of prognosis and therapeutic intervention. 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 3904. doi:10.1158/1538-7445.AM2011-3904