A novel post-surgical prognostic system for colorectal liver metastases treated by preoperative systemic treatment, using tumoral and non tumoral pathological changes, RAS mutation and Immunoscore

Introduction : Surgical control of colorectal liver metastasis (CRLM) combined with systemic treatment in patients aims to maximize survival. However, around 40% of patients develop recurrence in one-year post operation. Moreover, chemotherapy-related liver injury (CALI), such as sinusoidal obstruction syndrome (SOS), nodular regenerative hyperplasia (NRH), and steatohepatitis, have been reported to worsen operative mortality and morbidity rates. Aim : We aim to develop a prognostic scoring system to stratify patient prognosis post-hepatectomy by identifying the significant prognostic clinico-pathological-molecular factors in patients with resected CRLM. Methods: We investigated 143 patients with 403 CRLM operated from 2005-2013, including patients untreated (19 patients with 29 lesions) and treated with chemotherapy alone (33 patients with 91 lesions), chemotherapy with anti-VEGFR (47 patients with 157 lesions), and chemotherapy with anti-EGFR (44 patients with 126 lesions). All specimens were reviewed to assess tumor regression grading (TRG), histological growth pattern (HGP), and CALI. Genomic DNA from the lesions was extracted and purified from Formalin-fixed, paraffin-embedded slides. Immunoscore was assessed by the immune densities (cells/mm2) of CD3- and CD8 positive lymphocytes in the center and the invasive margin of the tumor by using morphometry. Comparisons were made using the Wilcoxon-Mann-Whitney test. Cumulative disease-free (DFS) and overall survival (OS) were analyzed using the Kaplan-Meier estimator and compared by log-rank tests. Cox proportional hazards models were used for uni- and multi-variate analysis. P-value of less than 0.05 was considered statistically significant. Results : Multivariate analysis showed that a high TRG (TRG 3-5), the worst HGP (replace growth pattern, mixed pattern), ≥ 4 lesions, positive surgical margin, CALI (steatohepatitis, NRH), RAS mutation and the worst Immunoscore were the significant prognostic factors. The prognostic scoring system combining these parameters significantly stratified post-operatively treated patients’ prognosis into three groups (high risk group: 81.3% (63.1%-90.3% 95CI) one-year recurrence rate, intermediate risk group: 41.3% (26.8%-52.9% 95CI) one-year recurrence rate, low risk group: 13.8% (0.3%-25.5% 95CI) one-year recurrence rate). Conclusions : Our novel prognostic scoring system of CRLM assessed by tumor and non-tumor pathological changes, Ras mutation and Immunoscore, allowed us to identify the patient population with high risk of recurrence post-hepatectomy.