Exploring the radiological differences of lymph nodes between mismatch repair deficient and mismatch repair proficient colon cancer.

Abstract

e15643 Background: Mismatch repair (MMR) deficient (dMMR) colon cancer (CC) accounts for 15-20% of CC diagnoses and is characterised by clinicopathological features distinct from MMR proficient (pMMR) CC. Regardless of MMR status, CT staging is a critical part of the workup for newly diagnosed CC. Whilst CT can accurately predict T stage, it is less effective for N stage. Incorrect N staging may risk poor prognostication, more invasive surgery and may have implications for emerging neoadjuvant therapies. We recently illustrated the impact of MMR status on CT staging, with N staging accuracy significantly worse in dMMR than pMMR CC. Furthermore, we highlighted the clinical utility of using serum inflammatory markers to predict N stage more accurately in dMMR CC. It is increasingly apparent that MMR status has implications for CT staging and should be considered when developing strategies to improve performance. It is therefore crucial to understand how lymph node appearances differ according to MMR status. Here, we investigated whether radiological features of lymph nodes differ between dMMR and pMMR CC. Methods: CT scans were retrospectively assessed for patients who underwent curative resection from March 2019 to July 2020. Axial images were used to identify nodes > 5mm in longest diameter within the tumour drainage territory. The following were collected for each node: location (pericolic, intermediate, main; according to Japanese Society for Cancer of the Colon and Rectum criteria), long axis diameter and short axis diameter. The following were collected for the largest node: short to long axis ratio (SLAr), attenuation and heterogeneity. Characteristics were compared between the dMMR and pMMR groups. Results: This cohort comprised 92 patients, with 53 and 39 in the pMMR and dMMR groups, respectively. As expected, there were greater proportions of females and right-sided tumours in the dMMR group. Median SLAr and median attenuation of the largest node were significantly greater in the dMMR than the pMMR group (0.7 v 0.57, p = 0.0012 and 53.1HU v 40.4HU, p = 0.013, respectively). Number and distribution of enlarged nodes, and location, long axis and heterogeneity of the largest node did not differ between the groups. Conclusions: dMMR and pMMR CC are increasingly recognised as distinct entities, with implications for prognosis and treatment. There is growing evidence illustrating the impact of MMR status on CT staging, and in particular N staging. Here, we illustrated the similarities and differences of lymph nodes on CT between dMMR and pMMR CC. To the best of our knowledge, this is the first analysis to explore the relevance of the largest node and to illustrate the SLAr and attenuation of the largest node being significantly greater in dMMR CC. Recognising these differences will be crucial as we develop strategies for selecting patients to receive neoadjuvant immunotherapy in dMMR CC.