Background: Surgery is considered curative in Dukes' B rectal cancer; however, many patients present with early relapse. To identify additional staging information, venous and lymphatic invasion were evaluated as potential prognostic factors.
Methods: Patients with Dukes' B or C rectal disease treated between 1976 and 2001 at a single institution were compared. Patient and treatment characteristics and vascular invasion were documented. The impact of vessel invasion was determined using Cox proportional hazards model.
Results: There were 256 Dukes' B patients and 74 Dukes' C cases without vascular invasion. Five-year survival was 76.5% for Dukes' B and 57.1% for Dukes' C patients. Vessel involvement increased the risk of recurrence (hazard ratio [HR] = 3.27, P = .0003) and death (HR = 3.11, P = .002) in B2 patients. The magnitude of these associations were comparable to that of C1 patients for recurrence (HR = 2.81, P = .004) and death (HR = 3.05, P = .005), as well as C2 patients for recurrence (HR = 3.45, P = .0008) and death (HR = 3.87, P = .0005).
Conclusion: Vascular invasion may be useful in characterizing patients with aggressive Dukes' B disease, who might benefit the most from adjuvant systemic therapy.