INTRODUCTION: Median survival for high grade gliomas is limited, at least partly explained by intratumoral heterogeneity contributing to treatment resistance. Radiological evaluation of treatment response is in most cases limited to assessment of tumor size months after the initiation of therapy. Diffusion weighted MRI and its estimate apparent diffusion coefficient, ADC, has been widely investigated, as it reflects tumor cellularity and proliferation. The aim of the present study was to investigate the potential of applying texture analysis in conjunction with multivariate ADC image for identifying early imaging biomarkers. MATERIALS AND METHODS: 23 consecutive high-grade glioma patients aged ≥18 years after diagnostic biopsy or resection were treated with radiotherapy (2 Gy/60 Gy) with concomitant and adjuvant temozolomide. MRI assessment three months after completion of radio-chemotherapy was used for classifying tumor progression or regression. A baseline exam was performed within one week before treatment start, day one, two weeks and six weeks into the treatment. ADC maps and T1 weighted anatomical images with and without contrast enhancement were collected, and (residual) tumor contrast enhancement was delineated. A gray-level co-occurrence matrix analysis was performed on the ADC maps in a cuboid encapsulating the tumor in coronal, sagittal and transversal planes, giving a total of 66 textural descriptors for each tumor. Principal component analysis, PCA, was applied to reduce dimensionality of the data, and the five largest components, or scores, were used in subsequent analyses. RESULTS: The score scatter plots revealed that the first, fourth and fifth components combined exhibited a pattern that strongly correlated to survival. Two groups could easily be identified: One with a median survival after diagnosis of 908 days and one with 299 days, p = 0.0005. MRI three months after completion of radiotherapy, age, surgical procedure and tumor grade all harbored prognostic information, but not to the same extent. CONCLUSION: By combining PCA and texture analysis ADC texture characteristics are identified, that seems to host pretreatment prognostic information, independent of known prognostic factors such as age, stage and surgical procedure. Validation of these data in a larger patient cohort is ongoing at our department.
展开▼
