Oncological outcomes and treatment characteristics of patients with cT3-4 mesorectal fascia positive rectal cancer operated in 2016: Dutch nationwide cohort with standardized magnetic resonance imaging re-assessment.

Abstract

Background: Rectal cancer with mesorectal fascia involvement (MRF+) requires neoadjuvant downstaging and/or induction therapy, followed by (beyond) total mesorectal excision and/or multivisceral resection (MVR) depending on response. Decisions on the need for MVR and plane of dissection vary. This nationwide cross-sectional study aimed to evaluate treatment and oncological outcomes in MRF+ primary rectal cancer and the impact of MVR types.

Methods: Using collaborative research, extensive clinical data on all rectal cancer resections in 2016 across 67 Dutch hospitals were retrospectively collected. Patients with cT3-4 MRF+ primary rectal cancer and standardized re-assessment of primary and restaging magnetic resonance imaging were selected. Patients were categorized by radiologically involved structures or organs and by MVR type. The primary outcome measures were microscopically positive margin (R1) resection rate, 4-year cumulative local recurrence, and 4-year overall survival.

Results: Of 3178 patients registered, 571 cT3-4 MRF+ were selected resections, 125 of which were MVR. Tumours remained MRF+ on restaging magnetic resonance imaging in 76.4%. The overall R1 rate was 11.8% (9.7% in cT3/4a MRF+ and 17.9% in cT4b tumours). The R1 rate after MVR was 20.8%. Overall, the 4-year local recurrence rate was 15.1% and 4-year overall survival 67.5%. When comparing total mesorectal excision with MVR and/or beyond total mesorectal excision, the local recurrence rate was 12.7% versus 24.2% (P = 0.001), and overall survival 71.3% versus 53.7% (P < 0.001). Patients with radiological anterior involvement seemed to have a higher R1 rate compared with other locations. Anterior resections in both men and women appeared to be associated with relatively low local recurrence and high overall survival, whereas other resections resulted in numerically lower R1 rates, but worse long-term oncological outcomes. However, neither anatomical location of MRF+ and/or ingrowth nor MVR type was significantly associated with the primary outcomes. Pelvic sepsis (hazard ratio 1.533, 95% confidence interval 1.0 to 2.3) and positive margins (hazard ratio 1.837, 1.2 to 2.8) were independent predictors of worse overall survival.

Conclusion: Oncological outcomes of MRF+ rectal cancer were not significantly influenced by the anatomical location of MRF involvement or MVR type.