Back to Annual Meeting Program

In patients with colorectal liver metastases (CRC-LM) preoperative chemotherapy (preCTx) is frequently used in a perioperative setting and/or to downsize irresectable CRC-LM. Especially in the perioperative setting, however, the exact role of preCTx is not well defined. To improve patient selection for preCTx factors predicting response and/or prognosis should be identified. In our study we analyzed the pathohistological response to preCTx in almost 150 patients who had had any form of CTx before liver resection, and correlated this response with survival.
Methods: We could evaluate the outcome of 147 patients who had hepatic resection for CRC-LM during the last decade in our institution, with pathohistological assessment of tumor regression grade (TRG) of metastatic disease and sufficient follow-up. Preoperatively the patients had undergone various regimens of CTx (42% FU-based, 40% Oxaliplatin and/or Irinotecan, 18% CTx plus antibodies/targeted therapy). 55% of the resections were at least a hemihepatectomy, 45% segmental or wedge resections. Free hepatic margins were achieved in 90%, free overall margins (including extrahepatic disease) in 81%. For this study all pathological specimens were reanalysed to classify TRG (grade 1= total regression to grade 5=no regression/vital tumor). Survival was estimated by Kaplan-Meier- and Cox-methods.
Results: Only 3% of the patients showed TRG grade 1 (total regression), 28% had good or moderate regression (TRG 2/3), and 69% had minor or no regression (TRG 4/5). TRG was better after CTx + targeted therapy (11% TRG1, 33% TRG 2/3; p<0.01 vs CTx alone). TRG was also better in patients receiving more than 6 months of preCTx (p<0.03 vs preCTx < 6 months). Overall 5 year survival (5y-Surv) was 42%. 5y-Surv was formally 100% in the few patients with total tumor regression (TRG 1), but comparable in the groups with TRG2/3 or TRG 4/5 (44% / 38%; p=0.1). In univariate and multivariate analysis in this subgroup of 147 patients with assessment of TRG only the hepatic margin (p<0.01) or the overall margin (p<0.001) significantly influenced survival. Type and duration of preCTx, extent of hepatic resection, nodal disease of primary CRC, number and size of metastases, and gender did not influence survival.
Conclusion: Only few patients with preoperative chemotherapy show total regression of the resected colorectal liver metastases. Addition of targeted therapy to CTx may enhance pathohistological tumor regression. Total response to preoperative chemotherapy may be associated with a clearly improved prognosis.