Distant metastasis after treatment of pT1 colorectal carcinomas

Aims For decades, lymph node metastases (LNM) were considered to be stepping stones for distant metastasis (DM) in submucosal invasive colorectal cancers (T1 CRCs). Despite completion surgery, DM still developed in 2-4% of T1 CRCs, suggesting a direct route of metastases. Extensive research has been performed to identify risk factors to predict LNM. However, data on the relationship between these risk factors, DM and survival are very limited. The aim of this study was to investigate risk factors for DM in T1 CRCs, and how this affects CRC specific survival (CSS).

Methods In a large cohort of consecutive T1 CRCs treated between 2000 and 2017 in 21 Dutch hospitals the proportion of DM, together with 3-year overall survival (OS), CSS, and adverse event-free survival (AFS) were determined. AFS was defined as having no DM and no CRC treatment related death. Risk factors for these outcomes were analyzed with cox proportional hazard models. Multiple imputation was used for missing data. High-risk pT1 CRC was defined as the presence of lymphovascular invasion (LVI), high-grade differentiation, and/or positive (R1)/unassessable (Rx) resection margin.

Results A total of 3812 T1 CRCs patients were included (mean age 69 years, 41% female, 18% ASA III/IV, 61% non-pedunculated), with a total median follow-up time of 51 months (IQR 23 – 64). After oncological resection, LNM were detected in 223/2099 (10.6%) patients. DM were observed in 134/3812 (3.5%) patients during follow-up. After surgery, 53/76 (69.7%) patients with DM showed no LNM, of whom 29/53 (54.7%) had 1-9 LNs resected and 24/53 (45.3%) had≥10 LNs resected. In multivariate analysis, non-pedunculated morphology (HR 1.52, 95% CI 1.01 – 2.27), size>40 mm (HR 2.52, 95% CI 1.31 – 4.84), present LVI (HR 1.96, 95% CI 1.19 – 3.23), high-grade differentiation (HR 2.27, 95% CI 1.22 – 4.21), and≥10 resected lymph nodes (HR 0.58, 95% CI 0.36 – 0.94) were independently associated with distant metastases. The 3-yrs OS, CSS and AFS were 92.3% (95-CI 89.3 – 95.4%), 97.9% (95-CI 94.8 – 100%), and 96.5% (95-CI 93.4 – 99.6%) respectively. CRC-specific mortality (126/3812; 3.3%) was determined by the development of DM (73/3812; 1.9%) and treatment-related death (53/3812; 1.4%). The 3-yrs AFS did not differ between the high risk pT1 CRCs who had follow-up versus surgery (556/579 (96.0%) vs 765/800 (95.6%), HR 0.93, 95% CI 0.51 – 1.69 after adjusting for ASA score, age, LVI, differentiation grade, gender, location, and non-pedunculated morphology).

Conclusions As 70% of DM after surgery developed in pT1N0 patients, this study provides support that an accountable proportion of DM may develop via a direct route instead of via the lymph nodes. In this study, CRC specific mortality of pT1 CRCs was related to the development of distant metastasis (1.9%), but also to treatment-related deaths (1.4%). With a similar 3-yrs AFS between high risk pT1 CRC patients with follow-up or surgery, the benefit of resecting the draining lymph nodes is questionable.

Conflicts of interest

Authors do not have any conflict of interest to disclose.

Publication History

Article published online:
15 April 2024