Article
Location-dependent prognostic factors and recurrence patterns in IDH wildtype glioblastoma
Search Medline for
Authors
-
Christine Jungk
- Division of Experimental Neurosurgery, Department of Neurosurgery, University Hospital Heidelberg, Heidelberg, Deutschland
-
Rolf Warta
- Division of Experimental Neurosurgery, Department of Neurosurgery, University Hospital Heidelberg, Heidelberg, Deutschland
-
Bettina Hug
- Department of Neuroradiology, Universitätsmedizin Mannheim, University of Heidelberg, Mannheim, Deutschland
-
Martin Bendszus
- Division of Neuroradiology, Department of Neurology, University Hospital Heidelberg, Heidelberg, Deutschland
-
Andreas von Deimling
- Department of Neuropathology, Institute of Pathology, University of Heidelberg, Heidelberg, Deutschland
-
Christel Herold-Mende
- Division of Experimental Neurosurgery, Department of Neurosurgery, University Hospital Heidelberg, Heidelberg, Deutschland
-
Andreas Unterberg
- Division of Experimental Neurosurgery, Department of Neurosurgery, University Hospital Heidelberg, Heidelberg, Deutschland
| Published: | June 9, 2017 |
|---|
Outline
Text
Objective: Recent studies suggest that glioblastomas (GBM) contacting the subventricular zone (SVZ) confer a dismal prognosis and may be linked to a distinct cell of origin compared to their cortical counterparts. However, existing analyses have disregarded the unique molecular and prognostic phenotype associated with isocitrate dehydrogenase (IDH) mutant GBM. We therefore sought to examine location-dependent prognostic factors and recurrence patterns exclusively in IDH wildtype GBMs.
Methods: Out of a consecutive cohort of 302 GBM patients treated at our department from 2004 until 2011 for whom clinical, molecular and imaging data were available, 288 patients (95%) did not harbor IDH mutations as ruled out by immunohistochemistry or direct sequencing. Based on preoperative contrast-enhancing MR images, patients were allotted to GBM with (SVZ+) and without (SVZ-) SVZ involvement or with (cortex+) and without (cortex-) cortical involvement, respectively, and compared for demographic, treatment-related, imaging and survival data at first diagnosis and recurrence. Multivariate survival analysis was performed employing a Cox proportional hazards model with stepwise forward selection of covariates.
Results: IDH mutations were more common in SVZ- GBMs (7% vs. 2% in SVZ+ GBMs; p=0.04). Consequently, only IDH wildtype GBM were analyzed. Demographic (age, sex) and treatment-related factors (Karnofsky performance status, completion of Stupp regimen, treatment within clinical trials) were balanced between groups with two exceptions: SVZ+ GBMs were found to have a lower frequency of complete resections at first diagnosis (p<0.0001) and lower non-surgical treatment intensity at recurrence (p=0.0009). Regarding growth and recurrence patterns, multifocal disease both at first diagnosis (p=0.031) and at recurrence (p=0.0006) was more common in cortex+ GBMs, together with a higher frequency of distant recurrences (28% vs. 2% in cortex- GBMs; p<0.0001). Besides established clinical prognostic factors (higher age, subtotal resection, lower treatment intensity at first diagnosis and recurrence), multivariate survival analysis identified SVZ involvement as the only independent prognosticator of inferior overall survival (p=0.018; HR=1.58) and survival after relapse (p=0.015; HR=1.58).
Conclusion: In this large-scale analysis of IDH wildtype GBMs, SVZ involvement significantly contributed to their dismal prognosis. In contrast, cortical involvement determined growth and recurrence patterns. This holds out the prospect for location-tailored prognostication and treatment planning. Interestingly, IDH mutations were enriched in non-SVZ GBM that, together with distinct clinical features, supports the hypothesis of location-dependent cells of origin.
