A new magnetic resonance imaging tumour response grading scheme for locally advanced rectal cancer
Menée auprès de 1 033 patients atteints d'un cancer rectal de stade localement avancé traité par chimioradiothérapie néoadjuvante, cette étude évalue la concordance entre un système de mesure par IRM du degré de régression tumorale et le système de classification de l'"American Joint Committee on Cancer" et du "College of American Pathologists"
Background : The potential of using magnetic resonance image tumour-regression grading (MRI-TRG) system to predict pathological TRG is debatable for locally advanced rectal cancer treated by neoadjuvant radiochemotherapy.
Methods : Referring to the American Joint Committee on Cancer/College of American Pathologists (AJCC/CAP) TRG classification scheme, a new four-category MRI-TRG system based on the volumetric analysis of the residual tumour and radiochemotherapy induced anorectal fibrosis was established. The agreement between them was evaluated by Kendall’s tau-b test, while Kaplan–Meier analysis was used to calculate survival outcomes.
Results : In total, 1033 patients were included. Good agreement between MRI-TRG and AJCC/CAP TRG classifications was observed (k = 0.671). Particularly, as compared with other pairs, MRI-TRG 0 displayed the highest sensitivity [90.1% (95% CI: 84.3–93.9)] and specificity [92.8% (95% CI: 90.4–94.7)] in identifying AJCC/CAP TRG 0 category patients. Except for the survival ratios that were comparable between the MRI-TRG 0 and MRI-TRG 1 categories, any two of the four categories had distinguished 3-year prognosis (all P < 0.05). Cox regression analysis further proved that the MRI-TRG system was an independent prognostic factor (all P < 0.05).
Conclusion : The new MRI-TRG system might be a surrogate for AJCC/CAP TRG classification scheme. Importantly, the system is a reliable and non-invasive way to identify patients with complete pathological responses.
British Journal of Cancer , résumé, 2022