Long-Term Survival and Local Relapse Following Surgery Without Radiotherapy for Locally Advanced Upper Rectal Cancer: An International Multi-Institutional Study

Abstract

Controversy remains regarding whether preoperative chemoradiation protocol should be applied uniformly to all rectal cancer patients regardless of tumor height. This pooled analysis was designed to evaluate whether preoperative chemoradiation can be safely omitted in higher rectal cancer.An international consortium of 7 institutions was established. A review of the database that was collected from January 2004 to May 2008 identified a series of 2102 patients with stage II/III rectal or sigmoid cancer (control arm) without concurrent chemoradiation. Data regarding patient demographics, recurrence pattern, and oncological outcomes were analyzed. The primary end point was the 5-year local recurrence rate.The local relapse rate of the sigmoid colon cancer (SC) and upper rectal cancer (UR) cohorts was significantly lower than that of the mid/low rectal cancer group (M-LR), with 5-year estimates of 2.5% for the SC group, 3.5% for the UR group, and 11.1% for the M-LR group, respectively. A multivariate analysis showed that tumor depth, nodal metastasis, venous invasion, and lower tumor level were strongly associated with local recurrence. The cumulative incidence rate of local failure was 90.6%, 92.5%, and 94.4% for tumors located within 5, 7, and 9 cm from the anal verge, respectively.Routine use of preoperative chemoradiation for stage II/III rectal tumors located more than 8 to 9 cm above the anal verge would be excessive. The integration of a more individualized approach focused on systemic control is warranted to improve survival in patients with upper rectal cancer.

FIGURE 1

Estimated local recurrence (A), disease-free survival (B), and overall survival (C) for all patients according to tumor location.

FIGURE 2

Predicted model for cumulative local relapse rate according to the tumor height: A, curve was drawn using raw data; B, curve was drawn after removing effects of 3 variables, venous invasion, mesenteric LN metastasis, and depth of tumor that were statistically significant in multivariate analysis; at last, C, curve was drawn after removing effects of all variables.

FIGURE 3

Estimated local recurrence (A), disease-free survival (B), and overall survival (C) for subgroup analysis: intraperitoneal versus extraperitoneal tumor.