Abstract 2003: Detecting clonal relationships of multifocal tumors

Abstract

Abstract Understanding the development and progression of tumors is fundamental for the improvement of patient treatment and care. Proliferative verrucous leukoplakia (PVL) is a recurrent, multifocal, precancerous lesion that often transforms into oral squamous cell carcinoma (SCC). Multifocal lesions may be separated in either space (synchronous) or time (metachronous). It is essential to determine the relationship between multifocal lesions in order to properly understand the progression of the disease and treat it accordingly. It has yet to be determined whether PVL lesions arise independently or are derived from a common ancestral clone of cells. We have developed a permutation algorithm to test the null hypothesis that the tumors from the same patient share no more genetic alterations than tumors originating from different patients. The alternative hypothesis is that tumors from the same patient, though separated by time and/or space, share a greater number of chromosomal abnormalities than would be expected by chance and so likely share a common premalignant ancestor cell. This will allow us to determine if the tumors arise as independent events or share some clonal relationship. Biopsy data from nine patients, each with multiple tumors, was used to run individual BAC arrays for each lesion. The breakpoints of genetic alterations were determined using Circular Binary Segmentation (CBS) and served as input for our permutation algorithm which counts the number of shared breakpoint positions between tumors. Though many regions in the genome may be prone to gain or loss in a tumor, the exact start and stop of a breakpoint likely defines a unique event such that two or more lesions sharing this alteration can be considered a signature of clonality. We found that five of the nine patients share a significant number of genetic aberrations between their SCC lesions, leading to the conclusion that in approximately half of patients, multifocal PVL tumors are clonally related. Our algorithm for identifying clonality between multiple samples from a patient could be used on a variety of platforms including high density SNP arrays, measuring loss of heterozygosity and copy number alterations, as well as high throughput sequencing. All that is required is multiple, spatially or temporally distinct samples from each patient and multiple patients. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 2003.