Initial assessment of a model relating intratumoral genetic heterogeneity to radiological morphology.

Tumour heterogeneity has major implications for tumour development and response to therapy. Tumour heterogeneity results from mutations in the genes responsible for mismatch repair or maintenance of chromosomal stability. Cells with different genetic properties may grow at different rates and exhibit different resistance to therapeutic interventions. To date, there exists no approach to non-invasively assess tumour heterogeneity. Here we present a biologically inspired model of tumour growth, which relates intratumoral genetic heterogeneity to gross morphology visible on radiological images. The model represents the development of a tumour as a set of expanding spheres, each sphere representing a distinct clonal centre, with the sprouting of new spheres corresponding to new clonal centres. Each clonal centre may possess different characteristics relating to genetic composition, growth rate and response to treatment. We present a clinical example for which the model accurately tracks tumour growth and shows the correspondence to genetic variation (as determined by array comparative genomic hybridisation). One clinical implication of our work is that the assessment of heterogeneous tumours using Response Evaluation Criteria In Solid Tumours (RECIST) or volume measurements may not accurately reflect tumour growth, stability or the response to treatment. We believe that this is the first model linking the macro-scale appearance of tumours to their genetic composition. We anticipate that our model will provide a more informative way to assess the response of heterogeneous tumours to treatment, which is of increasing importance with the development of novel targeted anti-cancer treatments.

The use of time to maximum enhancement to indicate areas of ablation following the treatment of liver tumours with high-intensity focused ultrasound.

The aim of this study was to investigate the use of time to maximum enhancement (t(max)) for each voxel in contrast-enhanced MRI (CE-MRI) as a non-invasive tool to determine areas of necrosis following treatment of liver tumours with high-intensity focused ultrasound (HIFU) and, having established the utility of t(max) maps, to develop a three-dimensional (3-D) representation to display this information concisely. 3-D T(1) weighted fast spoiled gradient echo images of the liver were acquired before and after administration of contrast agent. The CE-MR images were aligned to the pre-contrast volume and an estimate of t(max) was obtained for each voxel. Such pre- and post-contrast image sets were acquired before and after ablation. The t(max) maps before and after HIFU treatment were correlated with the procedure notes, radiological reports and gross histological specimen. Finally, 3-D t(max) maps of the whole liver were reconstructed to show all areas of abnormal tissue perfusion. Normal, healthy liver tissue uniformly enhances maximally after approximately 1 min. The computed t(max) maps accurately delineated areas of abnormal contrast agent uptake, corresponding to tumour deposits. Changes in t(max) and non-enhancing voxels after treatment correlate well with volumes targeted during ablation and the necrotic regions seen on gross histological specimens. Alignment of the contrast-enhanced images with the pre-contrast volume greatly improved the conspicuity of the t(max) maps. We conclude that t(max) maps and their 3-D views can be used as a non-invasive tool to assess and potentially to quantify the success of HIFU ablation, and concisely represent the large number of CE-MRI data.