A comparative genomic analysis between human breast cancer and paired tumor lines derived from transplanation of tumor biopsies into NOD/SCID mice.

Abstract

Abstract Abstract #41 Background: Successful preclinical studies of experimental anticancer agents require model systems that recapitulate breast cancer biology and genomic abnormalities as accurately as possible. The use of primary human tumor explants engrafted into the humanized mammary fat pad of NOD SCID mice (HIM technique for "human in mouse") is promising in this regard, but the evidence that the genomic profile of the tumor is stable during serial translatation of human tumors into mice and between mouse passages has not beeen clearly demonstrated.
 Methods: 4th mammary glands of NOD/SCID mice were humanized by removing mouse mammary epithelia and implanting immortalized human breast fibroblasts. Human breast tumor tissues were dissociated into single cell suspensions and injected into the “humanized” mammary fat pads of NOD/SCID mice. Patient tumors were micro-dissected from the surrounding normal tissue by laser capture microdissection and tumor RNA and DNA was isolated. In addition, tumor RNA and DNA was isolated from xenografts at passage 1, 2, and 3. Agilent 4 x 44K whole genome expression array and 244K array-based comparative genomic hybridization (aCGH) was performed on established xenografts (passage 1-3) as well as the original tumor samples. Exon 9 and 20 of PIK3CA and exon 4 to 9 of TP53 were targeted for mutation detection since these genes are commonly mutated in human breast cancer.
 Results: To date, eight HIM tumor lines have been successfully established and serially passaged. The HIM number and source of tumor are as follows: HIM2 (primary tumor); HIM 3 (primary tumor); HIM 4 (abdominal wall metastasis); HIM 5 (brain metastasis from the patient used to establish HIM 2); HIM6 (primary breast cancer); HIM7 (chest wall metastasis); HIM 8 (chest wall metastasis); HIM9 (chest wall metastasis). Passage 3 HIM tumors serially transplanted in NOD/SCID mice have similar histopathological features to their original human counterparts (HCP). Sequence analyses of the eight HIM lines reveals mutations in the TP53 gene in three tumor lines and a mutation in PIK3CA gene in one tumor line that are also present in the HCP. aCGH analysis, where available, also demonstated that the positions of the major gene amplifications and deletions in HIM lines are also consistent with the HCP. Furthermore, aCGH shows that genomic structure of the grafts are stable with passage. Immunohistochemistry shows that HIM8 and its HCP are positive for HER2. Molecular subtyping by PAM50 gene list indicates that HIM 2 – 8 tumor lines and their HCP are basal type breast cancer. HIM9 tumor line and its HCP are luminal subtypes, although the proliferation signature was activated in the mouse graft.
 Conclusions: The HIM mouse system faithfully reproduces the genotypic features of the respective HCP's and is therefore a valuable research engine for the preclinical development of biomarkers, imaging techniques and for the assessment of novel therapeutic approaches, particularly for Basal-type breast cancer. Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 41.