Artikel
High predictive value of hemizygous deletions at the Notch2 locus for survival of primary brain tumour patients
Hemizygote Deletionen beim Notch2 Genlokus haben eine hohe Voraussagekraft für das Überleben von Gliompatienten
Suche in Medline nach
Autoren
-
C. Zweifel
- Molekulare Neuro-Onkologie, Departement Forschung, Universitätsspital Basel; Neurochirurgische Klinik, Universitätsspital Basel
-
J. L. Boulay
- Molekulare Neuro-Onkologie, Departement Forschung, Universitätsspital Basel
-
A. R. Miserez
- Klinik für Innere Medizin, Universitätsspital Bruderholz
-
A. Ghaffari
- Molekulare Neuro-Onkologie, Departement Forschung, Universitätsspital Basel; Neurochirurgische Klinik, Universitätsspital Basel
-
G. Reifenberger
- Institut für Neuropathologie, Heinrich Heine Universität Düsseldorf
-
A. Merlo
- Molekulare Neuro-Onkologie, Departement Forschung, Universitätsspital Basel; Neurochirurgische Klinik, Universitätsspital Basel
| Veröffentlicht: | 4. Mai 2005 |
|---|
Gliederung
Text
Objective
Loss of heterozygosity (LOH) on chromosome 1p predicts the response to chemotherapy in 70% of malignant oligodendrogliomas, pointing to a genetic factor for the response that distinguishes oligodendrogliomas (OGs) from highly resistant glioblastomas (GBMs).
Methods
Eight distinct haplotypes were defined on a somatic deletion map on chromosome 1 of 26 OGs and 50 GBMs. In search for a correlation between survival and particular haplotypes, factor analysis, multivariate analysis and non-parametric Kaplan-Meier curves were computed. Predictive values were determined using receiver operating characteristic (ROC) analysis.
Results
The highly preserved centromeric recombination breakpoint, clustered within 1 centi-Morgan between markers D1S2696 and D1S2344, was prevalent in OGs, but not in GBMs (p<0.0001). Hemizygous deletions at D1S2696, located within intron 12 of the Notch2 gene, strikingly correlated with better outcome (p=0.0001). Interestingly, the coding region of Notch2 harboured overlapping single-copy microdeletions of 45 kb in primary OGs and a subgroup of GBMs. The molecular genetic classification equally well predicted survival time as histological classification including immunohistochemistry (p<0.0001) and was independent of patient age and gender. By ROC analysis, a cut-off of 24 months survival time was defined resulting in a sensitivity for molecular classification of 80.8% and specificity of 92.0% (positive and negative predictive value 84.0% and 90.2%, respectively).
Conclusions
Simple, rapid and highly reproducible molecular classification using marker D1S2696 at Notch2 locus adds independent prognostic information to histological classification and identifies a subgroup of OG-like GBMs with long-term survival to be considered in future clinical trials.
