Diffusion-weighted MRI with ADC mapping for response prediction and assessment of oesophageal cancer: A systematic review
Vollenbrock, Sophie E., Voncken, Francine E.M., Bartels, Lambertus W., Beets-Tan, Regina G.H., Bartels-Rutten, Annemarieke
DOI: https://doi.org/10.1016/j.radonc.2019.07.006
Radiotherapy and Oncology 142 p. 17-26
Abstract
Purpose: The aim was to perform a systematic review on the value of diffusion-weighted MRI (DW-MRI) with apparent diffusion coefficient (ADC) mapping in the prediction and assessment of response to chemo- and/or radiotherapy in oesophageal cancer. Materials and methods: A systematic search was performed on Pubmed, Embase, Medline and Cochrane databases. Studies that evaluated the ADC for response evaluation before, during or after chemo- and/or radiotherapy were included. The Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) was used to assess the quality of the included studies. Results: Fourteen studies, comprising 516 patients, in which the response to treatment in oesophageal cancer was evaluated on ADC maps were included. Acquisition parameter settings for DW-MRI and ROI placement varied substantially. The reference standard was RECIST or endoscopic assessment in eight non-surgery studies and histopathology after surgery in six studies. A high pre-treatment ADC significantly correlated with good response in three out of 12 studies; conversely, one study reported a significantly higher pre-treatment ADC in poor responders. In five out of eight studies good responders showed a significantly larger relative increase in ADC two weeks after the onset of treatment (range 23–59%) than poor responders (range 1.5–17%). After chemo- and/or radiotherapy ADC results varied considerably, amongst others due to large variation in the interval between completion of therapy and DW-MRI. Conclusion: DW-MRI for response evaluation to chemo- and/or radiotherapy in oesophageal cancer shows variable methods and results. A large relative ADC increase after two weeks of treatment seems most predictive for good response.