Importance of tumor regression assessment in predicting the outcome in patients with locally advanced rectal carcinoma who are treated with preoperative radiotherapy.

BACKGROUND Locally advanced rectal carcinoma has a poor prognosis. However, since the introduction of preoperative radiotherapy, the outcome of patients with rectal carcinoma has been reported to have improved. Nevertheless, to the authors' knowledge few data are available regarding the histopathologic response to radiotherapy as assessed on surgical specimens as a potential predictive factor for outcome. METHODS To estimate the effect of radiotherapy on rectal carcinoma, the authors retrospectively reviewed the surgical specimens of 102 patients with T3-4, N0 or > or = N1 rectal carcinoma and 1 patient with T2 but N1 rectal carcinoma. All patients were treated preoperatively with a hyperfractionated accelerated radiotherapy schedule in a prospective protocol (Trial 93-01). Using a standardized approach, tumor regression was graded using a system that varies from Grade 1 (tumor regression Grade [TRG] 1) when complete tumor regression is observed to Grade 5 (TRG5) when no tumor regression is observed. RESULTS Radiotherapy resulted in tumor downstaging in 43% of the patients. There were 2 pT1 tumors (2%), 21 pT2 tumors (20%), 66 pT3 tumors (64%), and 14 pT4 tumors (14%) after treatment. Regional lymph nodes were involved in 55 patients (53%). None of the patients demonstrated a complete tumor regression after radiotherapy, but in 79% of the specimens a partial tumor regression was observed (TRG1: 0%; TRG2: 20%; TRG3: 39%; TRG4: 20%; and TRG5: 21%). The median actuarial overall survival (OS) and disease-free survival (DFS) were 52 months. Actuarial local recurrence rates at 2 years and 5 years were 6.4% and 7.6%, respectively. Univariate analysis showed the actuarial DFS to be significantly lower in patients with lymph node metastases (P = 0.0004) and advanced pT stages (pT3-4) (P = 0.03). A favorable outcome for OS, DFS, and local control was observed in patients with TRG2-4 (i.e., responders) compared with patients with TRG5 (i.e., nonresponders), but also in patients with low residual tumor cell density (TRG2, 3, and 4). On multivariate analysis, TRG remained an independent prognostic indicator for local tumor control. CONCLUSIONS Tumor regression as well as residual tumor cell density were found to be predictive factors of survival in rectal carcinoma patients after preoperative radiotherapy. Even after preoperative radiotherapy, the pathologic stage of the surgical specimen remained a prognostic factor. The use of a standardized approach for pathologic evaluation must be implemented to allow comparison between the results of various treatment approaches.