Risks of Organ Preservation in Rectal Cancer: Data From Two International Registries on Rectal Cancer

Abstract

Purpose: Organ preservation has become an attractive alternative to surgery (total mesorectal excision [TME]) among patients with rectal cancer after neoadjuvant therapy who achieve a clinical complete response (cCR). Nearly 30% of these patients will develop local regrowth (LR). Although salvage resection is frequently feasible, there may be an increased risk for development of subsequent distant metastases (DM). The aim of this study is to compare the risk of DM between patients with LR after Watch and Wait (WW) and patients with near-complete pathologic response (nPCR) managed by TME at the time of reassessment of response.

Methods: Data from patients enrolled in the International Watch & Wait Database (IWWD) with cCR managed by WW and subsequent LR were compared with patients managed by TME (with ≤10% cancer cells-nPCR) from the Spanish Rectal Cancer Project (VIKINGO project). The primary end point was DM-free survival at 3 years from decision to WW or TME. The secondary end point was possible risk factors associated with DM.

Results: Five hundred and eight patients with LR were compared with 893 patients with near-complete response after TME. Overall, DM rate was significantly higher among LRs (22.8% v 10.2%; P ≤ .001). Independent risk factors for DM included LR (v TME at reassessment; P = .001), ypT3-4 status (P = .016), and ypN+ status (P = .001) at the time of surgery. 3-year DM-free survival was significantly worse for patients with LR (75% v 87%; P = .001). When stratified for pathologic stage, patients with LR did significantly worse through all stages (P ≤ .009).

Conclusion: Patients with LR appear to have a higher risk for subsequent DM development than patients with nPCR managed by TME at restaging irrespective of final pathology. Leaving the primary undetectable tumor in situ until development of LR may result in worse oncologic outcomes.

FIG 1.

(A) DMFS for patients with LR and TME at the time of reassessment of response (TME) using time-zero₁—decision to WW (in the LR group) and date of surgery in the TME group—showing statistically significant differences between groups. (B) Estimates for DMFS using time-zero₂—the date of removal of primary tumor in both groups (salvage of LR and TME at reassessment)—also resulted in significant differences between groups. Of note, using time-zero₂, LR estimates does not start at 100% to account for synchronous (distant metastases and LRs) observed in a proportion of patients already at time zero. DMFS, distant metastases-free survival; LR, local regrowth; TME, total mesorectal excision; WW, Watch and Wait.

FIG 2.

Distant metastases-free survival estimates for patients with LR and TME at reassessment stratified by final ypT status. Patients with LR show significantly worse survival estimates for subgroups ypT1-2 and ypT3-4. DMFS, distant metastases-free survival; LR, local regrowth; TME, total mesorectal excision.

FIG 3.

Distant metastases-free survival estimates for patients with LR and TME at reassessment stratified by final ypN status. Patients with LR show significantly worse survival estimates for subgroups ypN0 and ypN+. DMFS, distant metastases-free survival; LR, local regrowth; TME, total mesorectal excision.

FIG 4.

(A-C) Distant metastases-free survival estimates for patients with LR and TME at reassessment stratified by final pathologic UICC stage. Patients with LR show significantly worse survival estimates for subgroups (A) ypT1-2N0, (B) ypT3-4N0, and (C) ypT0-4N+. DMFS, distant metastases-free survival; LR, local regrowth; TME, total mesorectal excision; UICC, Union for International Cancer Control.

FIG A1.

Flow diagram demonstrating the study population in each database (IWWD and VIKINGO) and patients excluded from the analyses. IWWD, International Watch & Wait Database; nCRT, neoadjuvant chemoradiotherapy; PME, partial mesorectal excision; TME, total mesorectal excision; TRG, tumor regression grade; VIKINGO, the Spanish Rectal Cancer Project; WW, Watch and Wait.

FIG A2.

(A, B) DMFS analysis considering only patients who clearly have the clinical complete response description available in the IWWD. (A) DMFS for patients with LR and TME at the time of reassessment of response (TME) using time-zero₁—decision to WW (in LR group) and date of surgery in TME group—showing statistically significant differences between groups. (B) Estimates for DMFS using time-zero₂—the date of removal of primary tumor in both groups (salvage of LR and TME at reassessment)—also resulted is significant differences between groups. Of note, using time-zero₂, LR estimates does not start at 100% to account for synchronous (distant metastases and LRs) observed in a proportion of patients already at time zero. DMFS, distant metastases-free survival; LR, local regrowth; TME, total mesorectal excision; WW, Watch and Wait.

FIG A3.

Reverse Kaplan-Meier curves for both cohorts (LR and TME at reassessment). LR, local regrowth; TME, total mesorectal excision.