The added value of pelvic surveillance by MRI during postoperative follow-up of rectal cancer, with a focus on abbreviated MRI

Objectives To assess the added value of MRI over CT for the detection of pelvic recurrence during postoperative surveillance after rectal cancer surgery and to compare the diagnostic accuracy for pelvic recurrence achieved with abbreviated MRI (aMRI) with that of conventional enhanced MRI (cMRI). Methods Patients who underwent rectal cancer surgery followed by MRI in addition to the standard CT follow-up protocol were evaluated retrospectively. Two readers independently scored images from CT, cMRI, and aMRI, which consisted of T2-weighted and diffusion-weighted imaging, to rate the likelihood of recurrence. Diagnostic accuracy and ROC curves were calculated. The patients were divided into two groups for risk-adapted surveillance according to risk of recurrence: high-risk ( n = 157) and low-risk ( n = 169) groups. Results In total, 579 MRIs from 326 patients were assessed. A total of 48 pelvic recurrences occurred in 33 patients. The AUC in cMRI, aMRI, and CT were 0.98, 0.99, and 0.84, respectively. The difference in performance between CT and cMRI or aMRI for identifying recurrence was statistically significant ( p < 0.001). Both cMRI and aMRI showed superior performance compared with CT in the high-risk group ( p < 0.001), but this was not the case in the low-risk group ( p = 0.13). Furthermore, the diagnostic accuracy of aMRI was similar to that of cMRI. Conclusions The addition of MRI to the postoperative surveillance protocol may result in an improvement in the detection of pelvic recurrence after rectal cancer surgery. For patients at high risk of recurrence, an aMRI surveillance may be justified to improve the diagnostic yield. Key points • The addition of MRI to the postoperative surveillance protocol improved the diagnostic yield in patients at a high risk of recurrence. • Abbreviated non-enhanced MRI with DWI allows detection of pelvic recurrence with a diagnostic accuracy that is similar to that of contrast-enhanced MRI (AUC, 0.99 and 0.98, respectively; p = 0.12). • Abbreviated MRI that is restricted to high spatial resolution structural imaging and diffusion-weighted imaging takes less time and can be carried out without the need for injection of a contrast agent.