Abstract
BACKGROUND: Pathological response (PR) to preoperative chemotherapy for colorectal liver metastases (CLM) is recognised as a prognostic factor of outcome. However, the optimal system to assess this parameter is still debated. This study focuses on current methods and proposes a possibly better method for assessing PR. METHODS: Among 223 patients resected for CLM between 2004 and 2011, after more than three cycles of chemotherapy, the percentage of tumour cells, necrosis and fibrosis, and the tumour regression grade were assessed for each of 802 nodules. Pathological response was evaluated according to validated methods and their combinations. A new method combined the percentage of tumour cells and the size of all nodules as follows: , where n is each separate nodule, % is the percentage of remaining tumour cells within nodule n (%) and s is the size of nodule n (cm).The prognostic value of each method was calculated. RESULTS: After a median follow-up of 47 months (3-106), the cumulative 5-year overall survival rate after liver resection was 59%. The proposed method categorised as follows: 0 residual tumour; 0.1-6-cm residual tumour; >6-cm residual tumour, and necrosis rate >50% stratified prognosis (P=0.0027; P=0.02), while the other methods did not. At multivariate analysis, our method remained an independent predictor of outcome (P=0.001). CONCLUSIONS: Combining the percentage of tumour cells multiplied by the size of each separate tumour seems to be a better method for assessing PR. External validation is required.