To evaluate the association between the dynamic change in R₂*, obtained from a multi-echo dynamic acquisition, and clinicopathologic data in patients with rectal cancer.MRI is a central tool in rectal cancer management, and current guidelines recommend MRI as part of primary staging¹. Designing the optimal treatment for each patient requires identification of individual tumor aggressiveness prior to commencement of therapy. However, routine clinical, radiological and pathological parameters are currently unable to reliably predict individual tumor aggressiveness. To address unknown causes and mechanisms of rectal cancer aggressiveness, there is now considerable focus on the tumor-microenvironment. Dynamic contrast-based MRI is an increasingly popular method for tumor characterization, enabling quantitative assessment of phenotypic properties of the tumor-microenvironment^2,3. The purpose of this study was to evaluate the association between the dynamic change in transverse relaxation rate, R₂*, obtained from a multi-echo dynamic acquisition, and clinicopathologic data in patients with rectal cancer. To our knowledge, similar data has never before been collected.The study was approved by the regional ethics committee. Preoperative MRI was performed in 20 patients on a Philips Achieva (1,5T). Glucagon and buscopan were administered immediately before the patient was centered in the scanner. The dose of buscopan was divided in two; giving half the dose before the dynamic examination. High spatial-resolution fast spin-echo T₂-weighted images of the pelvic cavity and rectum were obtained in the sagittal- and transversal planes, and perpendicular to the tumor axis. Dynamic imaging was performed using a 3D multi-shot EPI sequence with three echoes using the following parameters: TR=39ms, TE1=4.6ms, echo spacing=9.3ms, flip angle=28°, EPI-factor=9. The acquired matrix size was 92×90 over a 180×180mm field-of-view. 12 slices of 10mm were acquired. Temporal resolution was approximately 1.9 s/imaging volume with a total of 60 dynamic series acquired. ProSet fat suppression was applied along with a parallel imaging (SENSE) factor of 1.7 in the RL-direction. A dose of 0.2mL/kg body weight of gadolinium-based CA (Dotarem® 279.3mg/mL, Guerbet, Roissy, France) was injected as a bolus (3mL/s). All slices were acquired parallel to the T₂-weighted images. Dynamic R₂* image-series were calculated on a voxel-by-voxel basis by assuming a mono-exponential dependence of signal-change on echo-time and parametric images representing the maximum peak change in R₂*, ΔR₂*-peak, were generated. Post-processing was performed using nordicICE (NordicNeuroLab, Bergen, Norway). Tumor volume-of-interests were manually delineated by an experienced radiologist. Staging was performed using the 7th tumor-node-metastasis (TNM) system. T- and N-stages were scored by histopathological evaluation of surgical specimens whereas M staging were determined by CT and/or MRI. Mann-Whitney U tests and receiver operator characteristic (ROC) curve statistics evaluated the associations between ΔR₂*-peak and clinicopathologic data. Statistical analysis was performed in R (R Foundation for Statistical Computing, Vienna, Austria).Figure 1 shows a 73-year-old man histologically diagnosed with rectal adenocarcinoma. The tumor demonstrated a strong increase in R₂* during bolus first-pass (a). This is also illustrated in the parametric map representing ΔR₂*-peak, overlaid a T₂-weighted image (b). Tumor ΔR₂*-peak was significantly different in rectal cancer patients with and without nodal metastases (P=0.01), showing a mean value of 24.6±6.4sec^-1 and 37.9±6.9sec^-1, respectively (Figure 2). Area under the ROC-curve was estimated to 0.94, corresponding to a sensitivity and specificity of 80% and 100%, respectively.The study showed that ΔR₂*-peak estimates from the primary tumor is a sensitive parameter for differentiating rectal cancer patients with and without nodal metastases, representing a main prognostic marker for development of distant metastasis, which is strongly associated with survival. Today, evaluation of lymph nodes is challenging and associated with considerable misinterpretations. Low tumor perfusion, as detected by the low ΔR₂*-peak in patients with nodal metastasis, may be related to tumor hypoxia, a known adverse factor related to metastasis development and poor survival. However, this needs to be further investigated. Actual perfusion parameters, as would be obtained by deconvolution of an AIF with the tissue-response curve, were not estimated in the current study. However, given the short duration of the multi-echo AIF, ΔR₂*-peak was assumed to give a reasonable estimate of the underlying perfusion because in the limiting condition of the AIF being a delta function (zero duration), the ΔR₂*-peak would directly reflect perfusion⁴.We identified low peak change in R₂* in the primary tumor as a predictive marker of nodal metastasis. Recognizing the strong association between nodal metastasis and disease survival, and that reliable staging of lymph nodes is the most challenging issue in rectal cancer staging, peak change in tumor R₂* may provide important information about tumor aggressiveness and serve as a risk-marker for stratification of patients to more individualized treatment.