Diffusion-weighted MR imaging for the selection of good and complete responders after chemoradiation treatment for locally advanced rectal cancer

- Background: - In the Netherlands, ±3300 patients/year are diagnosed with rectal cancer. Around 50% of these patients have locally advanced tumors that require surgical treatment preceded by chemoradiation therapy (CRT) in order the maximise the chance of cure. In ±35% of patients, only a small residual tumor is found after CRT (good response) and in 20% the tumor completely disappears (complete response). For these patients, standard surgery may not be necessary. The good responders may also benefit from a local excision, while the complete responders may even undergo a wait-and-see policy without surgery. These treatments are, however, only feasible when the response to CRT can be accurately assessed and the right candidates – i.e. the good and complete responders – can be reliably selected. Response evaluation is generally done with digital examination and endoscopy + biopsy, but these are not infallible. The problem is the assessment of fibrotic scar tissue at the site of the irradiated tumor. Standard imaging techniques (CT, FGD-PET/CT or MRI) also lack sufficient accuracy, because they can not discriminate residual tumor in these areas of fibrosis. An imaging tool that can improve patient selection could thus have a substantial impact on treatment decision making. Recent studies have shown potential for diffusion-weighted MR imaging (DWI) for the assessment of treatment response. DWI is a functional MRI technique that uses differences in the extracellular movement of water protons to differentiate between tissues of normal cellularity (healthy tissue) and increased cellularity (tumor). The diffusion can be analysed qualitatively (visually) as the signal intensity on the DWI images or can be quantitatively measured as the ‘apparent diffusion coefficient’ (ADC). If DWI proves efficient to identify the good and complete responders, tailoring of treatment into minimally-invasive treatment strategies (local excision and wait-and-see) after CRT may become feasible and unnecessary invasive treatment may be avoided. - Aim: - The aim of this study is to investigate the diagnostic performance of DWI for response assessment after chemoradiation treatment, in specific for the selection of good and complete responders, who may be eligible for minimally invasive treatments (local excision or wait-and-see) as an alternative to standard surgery. A good response is defined as regression of the tumor to ypT1-2 (limited to the rectal wall) and sterilisation of all metastatic lymph nodes (ypN0). A complete response is defined as complete regression of both tumor and lymph nodes (ypT0N0). The concrete study questions are as follows: 1. Is DWI accurate for the selection of patients in whom all metastatic nodes are sterilised, the ypN0? 2. Is DWI accurate for the evaluation of tumor response? a.Is DWI accurate for the selection of patients with a complete response of the primary tumor (ypT0)? b.Is DWI accurate for selection of patients with a good response of the primary tumor (ypT1-2)? - Plan of investigation: - The study will include patients with locally advanced rectal cancer undergoing CRT followed by surgery at Maastricht University Medical Center. All patients will undergo MRI + DWI at two time-points: [1] before onset of treatment, and [2] 8 weeks after completion of chemoradiation. Histology of the surgical specimen will serve as the reference standard. The response of the primary tumor will be categorised as ‘poor response’ (T3-4 residual tumor), ‘good response’ (T1-2) and ‘complete response’ (T0). The response of the lymph nodes will be categorised as N0 (no remaining metastatic nodes) and N+ (1 or more metastatic nodes). 1. Is DWI accurate for selection of ypN0? On the post-CRT DWI, each visible lymph node will visually be analysed and the ADC of each individual node will be measured. The visual scores and ADC measurements will be compared with histology on a node-by-node basis. The diagnostic accuracy of visual DWI and ADC analyses, respectively, for identification of metastatic nodes will be analysed. 2a. Is DWI accurate for selection of ypT0? On the post-CRT DWI, the likelihood of a complete tumor response will visually be scored. The diagnostic accuracy for the visual identification of a complete response will be analysed. Furthermore, mean ADC of the primary tumor will be measured both before and after CRT. The diagnostic accuracy of measuring tumor ADC for prediction of a complete response will be analysed for both time points. 2b. Is DWI accurate for selection of ypT1-2? On the post-CRT DWI, the likelihood of a good tumor response will visually be scored. The diagnostic accuracy for the visual identification of a good response will be analysed. Furthermore, mean ADC of the primary tumor will be measured both before and after CRT. The diagnostic accuracy of measuring tumor ADC for prediction of a good response will be analysed for both time points.