Glutathione S-transferase pi amplification is associated with cisplatin resistance in head and neck squamous cell carcinoma cell lines and primary tumors.

PURPOSE: The purpose is to evaluate the association of glutathione S-transferase pi (GST-pi) amplification and cisplatin resistance in head and neck cancer. EXPERIMENTAL DESIGN: An analysis of chromosomal abnormalities in 10 head and neck cancer cell lines by comparative genomic hybridization was performed. GST-pi amplification and expression were evaluated in head and neck cell lines and paraffin-embedded tissue by fluorescence in situ hybridization (FISH) and immunohistochemistry. RESULTS: Changes in the DNA copy number were seen in all 10 cell lines by comparative genomic hybridization. The most frequent chromosomal alterations were: gain at 3q; loss at 3p; gain at 8q; loss of 18q; gain at 20q; loss at 8p; and gain of 11q11-q13. Using FISH, 9 of 10 cell lines showed increased GST-pi copy number. GST-pi amplification was detected in 7 of 10 cell lines. Five were relatively cisplatin resistant, and 2 were relatively cisplatin sensitive (mean IC(50), 11.2 and 2.75 microM). Two relatively cisplatin-sensitive cell lines showed GST-pi gain and another relatively cisplatin-sensitive cell line had predominantly two copies of the gene. In 10 tumor specimens, 4 had two copies of GST-pi. All 4 had a complete response to neoadjuvant chemotherapy, 3 of whom are alive >50 months from treatment compared with 2 patients showing GST-pi amplification. Neither responded to chemotherapy, and both died of disease <9 months from diagnosis. CONCLUSIONS: Using FISH, GST-pi amplification is a common event in head and neck squamous cell carcinoma and may be associated with cisplatin resistance and poor clinical outcomes in head and neck cancer patients treated with cisplatin-based therapy.

c-Myc promoter activation in medulloblastoma.

%The c-myc oncogene is commonly activated in medulloblastoma. Genomic amplification is a well-documented cause of c-myc activation but does not account for all cases of c-myc activation. In this study, we sought other means by which c-myc is overexpressed in medulloblastoma. Twelve medulloblastoma or PNET cell lines were screened for c-myc genomic amplification, mRNA levels, and protein levels. Two medulloblastoma lines, D283 Med and D721 Med, were identified that expressed c-myc mRNA and protein at high levels without genomic amplification. The c-myc gene's regulatory sequences were normal in those cell lines. However, specific regions of the promoter, independent of the beta-catenin binding sites, were responsible for activation as revealed by promoter assays and site-directed mutagenesis. Transcriptional activation by a beta-catenin-independent pathway is therefore a likely mechanism for c-myc overexpression in a subset of medulloblastomas.

MYCC and MYCN oncogene amplification in medulloblastoma. A fluorescence in situ hybridization study on paraffin sections from the Children's Oncology Group.

CONTEXT: Brain tumors are the most common solid tumor in childhood, and medulloblastoma is the most common malignant brain tumor in this age group. Cytogenetic abnormalities that have been described in childhood medulloblastoma include loss of 17p, amplification of MYCC (c-myc), amplification of MYCN (N-myc), and isochromosome 17q. Data on these tumors indicate that the frequency of MYCC amplification is 5% to 10%. Fluorescence in situ hybridization is a powerful tool for investigating these features on archival material. OBJECTIVES: To determine if intratumoral heterogeneity exists for MYCC and MYCN in medulloblastomas and if tumors with amplified MYCC or MYCN exhibit consistent histologic patterns. DESIGN: In this fluorescence in situ hybridization study, we investigated the frequency and prognostic significance of MYCC and MYCN amplification in 77 medulloblastomas derived from the Children's Oncology Group. RESULTS: MYCC amplification occurred in only 4 (5.2%) of 77 tumors. The 4 patients died of clinically aggressive neoplasms within 7 months of diagnosis. Similarly, 4 of 77 patients' tumors were found to exhibit MYCN amplification, but survival data are incomplete at present, therefore prognostic significance cannot be characterized. CONCLUSIONS: These data establish the frequency of MYCC amplification in a large cohort of children with medulloblastoma and further suggest that MYCC amplification may be a marker of poor prognosis. Intratumoral heterogeneity was identified for these oncogenes in that 1 patient's tumor exhibited evidence of both MYCN and MYCC amplification, and this patient experienced a shortened survival time.

Comprehensive molecular cytogenetic investigation of chromosomal abnormalities in human medulloblastoma cell lines and xenograft.

Cell lines and xenografts derived from medulloblastomas are useful tools to investigate the chromosomal changes in these tumors. Here we used G-banding, fluorescence in situ hybridization (FISH), spectral karyotyping (SKY), and comparative genomic hybridization to study 4 medulloblastoma cell lines and 1 xenograft. Cell line D-425 Med had a relatively simple karyotype, with a terminal deletion of 10q and amplification of MYC in double-minutes (dmins). FISH demonstrated that an apparent isochromosome (17q) by routine karyotyping was actually an unbalanced translocation between 2 copies of chromosome 17. Cell line D-556 Med also had a simple near-diploid stemline with an unbalanced 1;13 translocation resulting in a gain of 1q, an isochromosome (17q), and dmins. These findings were initially described using routine G-banded preparations, and FISH showed that the dmins were an amplification of MYC and the i(17q) was an isodicentric 17q chromosome. The other finding was confirmed by FISH, SKY, and comparative genomic hybridization. Cell lines D-721 Med and D-581 Med had complex karyotypic patterns that could be completely characterized only when FISH and SKY were used. Xenograft D-690 Med also had a complex pattern that FISH and SKY were helpful in completely elucidating. Interestingly, balanced reciprocal translocations were seen as well as complicated unbalanced translocations and marker chromosomes. Comparative genomic hybridization demonstrated only a deletion of 10q22-10q24, supporting the idea that despite the complexity of the chromosomal rearrangements, minimal alterations in the overall chromosomal content had occurred. This study demonstrates that routine cytogenetic preparations are adequate to describe chromosomal abnormalities in occasional medulloblastoma samples, but a broader spectrum of molecular cytogenetic methods is required to completely analyze most of these tumor samples.