DCE-CT in the follow-up of targeted therapy in GISTs

Abstract

Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal malignancy that initiate from the gastrointestinal tract. GISTs most often metastasize to the liver or the peritoneum. The effect of conventional chemotherapy and radiotherapy do only have very limited therapeutic effect in GISTs. However, the introduction of targeted therapy has dramatically changed the management of advanced GISTs when improved tumor control, progression-free survival, and overall survival are considered. The size reduction of GIST lesions during targeted therapy regimens most often is modest and lesions may even enlarge during successful treatment. Progressive disease does not necessitate the presence of new lesions, and the phenomenon of a nodule within a mass may represent progressive disease. So, the most widely used assessment of response of cancer treatment based on the sum of the largest diameter of target lesions (RECIST) (1) may underor overestimate the effect of targeted therapies in GISTs; due to that fact Choi et al. (2) introduced computed tomography (CT) criteria that also focused on changes in lesion attenuation. These criteria which rely on additional CT enhancement are a better predictor of prognosis than criteria which rely only on changes in size. 18-FDG PET has shown to be very sensitive in detecting early response and as a predictor of long-term response to targeted therapy in GIST. PET/CT may combine metabolic and morphologic issues. However, the term PET/CT shall be reserved to protocols that indeed perform diagnostic CT with diagnostic acquisition parameters and a state of the art contrast media policy. A CT scan used for attenuation correction in a “PET/CT” cannot be considered diagnostic. However, due to limited availability and high costs, and due to the fact that up to 20% of GIST lesions on pretreatment scans do not show increased uptake of glucose, in general CE-CT is used as the standard modality in the follow-up of treatment in GISTs supplemented with MRI or PET/CT when needed. The morphological response assessment in GISTs using CT as modality is not only based on size and density measurements. Pseudo progression and pseudo regression have to be evaluated visually and integrated in the final assessment. Since the introduction of DCE-CT by Miles (3) the discipline has evolved dramatically due to enhanced computer power, the development of MDCT, and the development of software tools. In a recent report recommendations and guidelines for DCE-CT are presented (4). It was concluded that DCE-CT scans can robustly assess tumor vascular support and that DCE-CT has reached technical maturity for use in therapeutic trials in oncology. Betz et al. (5), in this issue of Acta Radiologica, present a pictorial review demonstrating that CE-CT easily can be combined with a DCE-CT protocol, so pseudo responses in the follow-up of treatment with targeted agents in GISTs probably can be diminished. DCE-CT can produce a range of physiological parameters that can be displayed as parametric maps, and as shown in (5) be used for quantitative as well as qualitative evaluation. The subjective lesion delineation is an issue that hampers imaging disciplines which rely on signal changes over time in user defined ROIs/VOIs. No guidelines for the delineation of such lesions have yet been agreed upon. Another issue for techniques which rely on time-signal curves to create surrogate molecular imaging markers is the problem with movement artifacts. In the presented manuscript by Betz et al. (5) image registration software that correct for movement artifacts has been included. However, it is obvious that such tools can be improved, especially when tissue near the diaphragm is considered. Although semi-automatic segmentation of a whole lesion volume now is technical possible, and a sample volume in the z-axis that in practice will cover the whole lesion considered, the technique cannot cover the whole body and may serve as complementary to CE-CT as all tumors and metastases do not necessarily behave the same way during treatment. The parameters derived from the DCE-CT discipline depends on the CT system used, acquisition parameters, contrast media policy, and data processing. Although very illustrative parametric color display shows details of tumor characteristics that supplement the findings seen in the corresponding morphological CE-CT, there is a need to have statistics for reproducibility and standards for how to define a ROI in GISTs when considering quantitative measurements. A study that illustrates the impact of the supplementary DCE-CT figures on endpoints as progression-free survival and overall survival compared to morphological response evaluation criteria in longitudinal studies is also warranted. Another important question is if pre-therapeutic DCE-CT figures can act as prognostics in GISTs.