Rectal cancer: a methodological approach to matching PET/MRI to histopathology

Purpose:To enable the evaluation of locoregional disease in the on-going RECTOPET (REctal Cancer Trial on PET/MRI/CT) study; a methodology to match mesorectal imaging findings to histopathology is presented, along withinitial observations. Methods:FDG-PET/MRI examinations were performed in twenty-four consecutively included patients with rectaladenocarcinoma. In nine patients, of whom five received neoadjuvant treatment, a postoperative MRI of thesurgical specimen was performed. The pathological cut-out was performed according to clinical routine with theaddition of photo documentation of each slice of the surgical specimen, meticulously marking the location, size,and type of pathology of each mesorectal finding. This allowed matching individual nodal structures frompreoperative MRI, via the specimen MRI, to histopathology. Results:Preoperative MRI identified 197 mesorectal nodal structures, of which 92 (47%) could be anatomicallymatched to histopathology. Of the matched nodal structures identified in both MRI and histopathology, 25% werefound to be malignant. These malignant structures consisted of lymph nodes (43%), tumour deposits (48%), andextramural venous invasion (9%). One hundred eleven nodal structures (55%) could not be matched anatomically.Of these, 97 (87%) were benign lymph nodes, and 14 (13%) were malignant nodal structures. Five were malignantlymph nodes, and nine were tumour deposits, all of which had a short axis diameter < 5 mm. Conclusions:We designed a method able to anatomically match and study the characteristics of individualmesorectal nodal structures, enabling further research on the impact of each imaging modality. Initial observationssuggest that small malignant nodal structures assessed as lymph nodes in MRI often comprise other forms ofmesorectal tumour spread. Trial registration: Clinical Trials Identifier: NCT03846882