Purpose: After preoperative (radio)chemotherapy, histologic determinants for prognostication have changed. It is unclear which variables, including assessment of tumor regression, are the best indicators for local recurrence and survival.
Experimental design: A series of 201 patients with locally advanced rectal cancer (cT3/T4, M0) presenting with an involved or at least threatened circumferential margin (CRM) on preoperative imaging (<2 mm) were evaluated using standard histopathologic variables and four different histologic regression systems. All patients received neoadjuvant radiochemotherapy or radiotherapy. The prognostic value of all factors was tested with univariate survival analysis of time to local recurrence and overall survival.
Results: Local recurrence occurred in only 8% of the patients with a free CRM compared with 43% in case of CRM involvement (P < 0.0001). None of the four regression systems were associated with prognosis, not even when corrected for CRM status. However, we did observe a higher degree of tumor regression after radiochemotherapy compared with radiotherapy (P < 0.001). Absence of tumor regression was associated with increasing invasion depth and a positive CRM (P = 0.02 and 0.03, respectively).
Conclusions: Assessment of CRM involvement is the most important pathologic variable after radiochemotherapy. Although tumor regression increases the chance on a free CRM, in cases with positive resection margins prognosis is poor irrespective of the degree of therapy-induced regression.