Professor Elleke Dresen of UZ Leuven (Belgium) delivered a presentation at ESMO 2023 focusing on the significance of radiological staging for colorectal cancers, particularly in identifying high-risk patients.
Historically, radiologists primarily determined the local extent of disease and the presence of metastases. Treatment paradigms have since shifted from sole curative tumour resection to incorporating chemotherapy in an adjuvant or neoadjuvant setting, contingent on high-risk patient features. Radiologists now play a pivotal role in identifying patients who might benefit from more aggressive therapies. For instance, if a tumour infiltrates mesorectal fat or adjacent structures, creating challenges for achieving clear surgical margins, MRI staging, a standard for rectal cancer, significantly enhances preoperative staging accuracy.
Identifying positive lymph nodes poses a challenge due to the size variability and reactive inflammatory responses, making it difficult to differentiate between positive and negative nodes. In contrast, the T-stage serves as a more reliable indicator of colorectal cancer’s systemic spread risk. Notably, the Extramural Venous Invasion (EMVI), visible on MRI as an irregular, dilated tubular structure emanating directly from the primary tumour, holds critical importance and demands accurate identification.
While MRI has been a longstanding imaging technique for rectal cancer, applying it to colon cancer proves more complex due to mobility issues and the larger field of view, resulting in imaging artefacts. Consequently, CT remains the primary modality for colon cancer imaging, offering approximately 80-85% accuracy in distinguishing advanced T3-T4 tumours from early-stage lesions. Noteworthy features, such as EMVI and specific characteristics of high-risk PT4A tumours with peritoneal spread, are discernible on CT scans.
Ongoing progress suggests that relying solely on visual observation is insufficient. Utilizing radiomics to evaluate voxel and pixel data, coupled with artificial intelligence, promises enhanced information extraction from imaging. However, further extensive studies and external validation are imperative to assess the reliability and efficacy of these advanced techniques.
Reference:
Dresen E., Leuven – How do we optimise imaging?. ESMO2023
With the educational support of:
 |
 |