2037 - Association Between Rectal Cancer Magnetic Resonance Extramural Venous Invasion (mrEMVI) and Metachronous Liver and Lung Metastases

Purpose/Objective(s):

Baseline pelvic MRI identifies magnetic resonance extramural venous invasion (mrEMVI) in patients with rectal cancer, which is associated with higher rates of synchronous and metachronous metastases. Understanding the pattern of metachronous metastases in patients with rectal cancer and mrEMVI may guide surveillance. The hypothesis is patients with locally advanced rectal cancer and mrEMVI undergoing curative intent radiotherapy predominantly have liver metastases.

Materials/Methods:

We conducted a retrospective study of consecutive patients from our institution with locally advanced rectal adenocarcinoma (January 2015 to January 2024) who had present or equivocal mrEMVI on baseline MRI and received curative intent radiotherapy. Metachronous liver and lung metastases were assessed using annual follow up imaging from years 1 to 6. Patients with baseline distant metastasis, other synchronous cancers or prior malignancies were excluded.

Results:

Of the 656 patients screened, mrEMVI was present in 122 patients and equivocal in 65. After excluding 89 patients with baseline metastasis, metastasis during treatment and/or incomplete curative intent treatment, 98 remained. Of 187 patients with present or equivocal mrEMVI, 21 patients (11.2%) had synchronous metastasis to the liver only, 6 (3.2%) to the lung only, 10 (5.3%) to both the liver and lung, and 2 (1.1%) to other areas. Synchronous metastases were more frequent in the liver than the lung (McNemar test, p=0.007).

Of the 98 patients with present or equivocal mrEMVI without synchronous metastases, follow up CT imaging from years 1 to 3 detected liver metastases in 10 patients (10.2%) and lung in 23 patients (23.5%), with no new events detected beyond year 3. However, follow up adherence decreased over time and some follow up had not been completed at the time of data collection. When comparing paired patient level outcomes over this period, lung metastases were significantly more frequent than liver metastases (McNemar test, p=0.012).

At year 1, liver metastases occurred in 6/42 patients with low rectal tumors versus 2/56 with mid/high (Fisher’s exact, p=0.071), and lung metastasis occurred in 6/42 low versus 5/56 mid/high (Fisher’s exact, p=0.522).

Differences in rates of liver or lung metastasis at year 1 CT were not statistically significant between patients who received multi agent chemotherapy plus concurrent chemoradiation versus concurrent chemoradiation alone (Fisher’s exact; liver p=0.320, lung p=0.655).

Conclusion:

Patients with present or equivocal mrEMVI demonstrated an early pattern of metastasis, with synchronous metastases predominantly affecting the liver and metachronous metastases within 3 years more frequently affecting the lung. No liver or lung metastases were detected beyond 3 years of surveillance. These findings highlight the importance of early and targeted surveillance.