Editorial
Pretreatment MRI: One of the Key Components of Treatment Planning
;
The strategies for treatment of rectal cancer have changed fundamentally in recent years, becoming far more complex in the process (1, 2, 3, 4). A central part is played by imaging: high-resolution magnetic resonance imaging (MRI), with its optimal visualization of the anatomic structures of the true pelvis, is now the gold standard procedure for the diagnosis of local spread (1, 2, 5, 6). A new German study explores the accuracy of MRI for locoregional staging prior to treatment (7).
A multicenter study
Stelzner and colleagues re-evaluated the pretreatment MRI scans of 50 men and women with rectal cancer <12 cm from the rectal verge who had been treated in the Optimized Surgery and MRI (Optimierte Chirurgie und MRT, OCUM) study. In this multicenter observational study, MRI was used as a selection criterion to determine whether or not neoadjuvant radiotherapy should be performed (8). Therefore, the new study included only patients who had been treated with primary total mesorectal excision (TME), i.e., without neoadjuvant therapy. Their MRI scans, obtained between 2008 and 2016 in the course of the OCUM study, were re-evaluated by a total of 74 radiologists. The (newly assessed) pretreatment MRI findings were then compared with the postoperative histopathology results, focusing on the following aspects: T category, N category, distance between tumor and mesorectal fascia (MRF), and presence/absence of extramural vascular invasion (EMVI). If the T or N category could not be determined, classification as Tx or Nx, respectively, was possible. Moreover, the participating radiologists were divided into three groups on the basis of their professional experience —high, intermediate, low—to evaluate whether years of experience affected the outcome quality.
Advantages regarding MRF and EMVI, weaknesses regarding T and N categories
The raters completed a median of 45.5 re-evaluations. It emerged that the MRI-based UICC stage agreed with the histopathological assessment in only 33.6% of cases (overstaging: 25.3%, understaging: 15.7%, unclassifiable: 25.4%). A higher rate of agreement (48.3%) was observed for the T category when considered separately, with overstaging in 42.2% and understaging in 9.4% of cases. On closer examination, distinguishing between T2 and T3a/b proved particularly problematic. Correct N staging was achieved in 46.3% of cases (overstaging: 15.3%, understaging: 13.1%), with 25.3% of cases unclassifiable. Much better rates of agreement were found for MRF and EMVI: The distance to the MRF was assessed correctly in 86.8% of cases, with an overstaging rate of 13.2% (the OCUM study had included only patients in whom MRF involvement was ruled out, so no understaging was possible). As for EMVI, the rate of agreement of the MRI with the microscopic findings was 79.9% (overstaging: 10.4%, understaging: 9.7%). Somewhat surprisingly, the raters’ degree of experience had only a slight effect on outcome quality.
Critical evaluation
The study by Stelzner and colleagues highlights an important clinical problem and confirms the particular difficulty of radiological N staging, in contrast to the much more reliable assessment of the distance between tumor and MRF and evaluation of EMVI (1, 2, 5, 6). A positive aspect is that the evaluations were conducted by a very high number of radiologists. Most of them were not specialized experts, but this reflects the widespread reality of patient care. One somewhat problematic factor is that the re-evaluated MRI scans did not include any diffusion-weighted images (DWI) and came from only three sites (Munich, Karlsruhe, Dresden), so that there is a possibility of selection bias.
Conclusion and prospects
MRI plays a key role in the planning of treatment for rectal cancer, not just as a selection criterion for neoadjuvant approaches, but also after completion of neoadjuvant therapy. In this context, MRI is crucial, in combination with endoscopic and clinical evaluation: a) in assessing whether a clinical complete response has been achieved and whether it may be possible to switch to an organ-preserving watch and wait approach, and b) for patients within a watch and wait approach for monitoring local tumor regrowth (9, 10). In this context, radiological appraisal is even more difficult than before treatment (1, 5, 6, 9).
The study by Stelzner and colleagues confirms that MRI also has limitations, and that because of its special importance for diagnosis and treatment planning, particularly high standards should be enforced with regard to scanning procedures and image interpretation. It therefore remains a clinical necessity to further improve image quality and to implement molecular biomarkers. Finally, from the clinical viewpoint, the treatment indications should ideally be determined not only on the basis of the radiological images: the decision must also take account of the endoscopic findings and, especially, the results of clinical evaluation and critical discussion in an interdisciplinary tumor board.
Conflict of interest statement
The authors declare that no conflict of interest exists.
Manuscript received on 9 February 2026, revised version accepted on 9 February 2026.
Translated from the original German by David Roseveare.
Corresponding author
Prof. Dr. med. Michael Ghadimi
mghadimi@med.uni-goettingen.de
Cite this as:
Grade M, Ghadimi M: Pretreatment MRI: One of the key components of treatment planning. Dtsch Arztebl Int 2026; 123: 127–8. DOI: 10.3238/arztebl.m2026.0023
|
