BACKGROUND: Thermal ablation has recently become a key therapy for the treatment of colorectal liver metastasis (CLM). However, the role of ablation in combination with resection has not yet been firmly established. We hypothesize that in patients with CLM, those who undergo liver resection with ablation (RA) have similar outcomes compared with those who undergo liver resection only.
METHODS: We reviewed a multicenter international database of 906 surgical procedures for CLM from 5 high volume hepatobiliary surgical units. Patients undergoing RA (n = 63) were matched based on the number of lesions and tumor size using a 1:1 balanced propensity score analysis with those having resection only (n = 63). Our primary outcomes were overall survival (OS) and disease-free survival (DFS).
RESULTS: The mean age of our cohort was 58 ± 11 years, with 43% females. With a median follow-up of 70.8 months, patients in the resection and RA group had a median OS of 45.1 and 54.8 months (p = 0.71), respectively. The median DFS was 22.7 and 14.2 months (p = 0.045), respectively. Using a multivariate Cox proportional hazards regression model, the treatment approach was not associated with OS (p = 0.94) or DFS (p = 0.059). A higher number of lesions is independently associated with worse DFS (hazard ratio: 1.12, p < 0.01). When there was disease recurrence, the region of recurrence was similar between the RA versus resection only groups (p = 0.27), but there was a shorter time to recurrence in the RA group (p = 0.002).
CONCLUSIONS: For CLM, the treatment approach was not significantly associated with OS or DFS, while tumor biology likely played an important role. Prospective research on the quality and effectiveness of thermal ablation combined with hepatic resection is warranted.

Colorectal liver metastasis (CRLM) is challenging in the clinical treatment of colorectal cancer. Limited research has been conducted on how CRLM develops. RNA sequencing data were obtained from the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA). Four machine learning algorithms were used to screen the hub CRLM-specific genes, including Least Absolute Shrinkage and Selection Operator (Lasso), Random forest, SVM-RFE, and XGboost. The model for identifying CRLM was developed using stepwise logistic regression and was validated using internal and independent datasets. The prognostic value of hub CRLM-specific genes was assessed using the Lasso-Cox method. The in vitro experiments were performed using SW620 cells. The CRLM identification model was developed based on four CRLM-specific genes (SPP1, ZG16, P2RY14, and PRKAR2B), and the model efficacy was validated using GSE41258 and three external cohorts. Five CRLM-specific prognostic hub genes, SPP1, ZG16, P2RY14, CYP2E1, and C5, were identified using the Lasso-Cox algorithm, and a risk score was constructed. The risk score was validated using the GSE39582 cohort. Three genes have both efficacy in identifying CRLM and prognostic value: ZG16, P2RY14, and SPP1. Immune infiltration and enrichment analyses demonstrated that SPP1 was associated with M2 macrophage polarization and extracellular matrix remodeling. In vitro experiments indicated that SPP1 may act as a cancer-promoting factor. The hub CRLM-specific gene SPP1 can help determine the diagnosis, prognosis, and immune infiltration of patients with CRLM.

OBJECTIVE: Histopathological growth patterns (HGPs) of colorectal liver metastases (CRLMs) have prognostic value. However, the differentiation of HGPs relies on postoperative pathology. This study aimed to develop a magnetic resonance imaging (MRI)-based radiomic model to predict HGP pre-operatively, following the latest guidelines.
METHODS: This retrospective study included 93 chemotherapy-naïve patients with CRLMs who underwent contrast-enhanced liver MRI and a partial hepatectomy between 2014 and 2022. Radiomic features were extracted from the tumor zone (RTumor), a 2-mm outer ring (RT+2), a 2-mm inner ring (RT-2), and a combined ring (R2+2) on late arterial phase MRI images. Analysis of variance method (ANOVA) and least absolute shrinkage and selection operator (LASSO) algorithms were used for feature selection. Logistic regression with five-fold cross-validation was used for model construction. Receiver operating characteristic curves, calibrated curves, and decision curve analyses were used to assess model performance. DeLong tests were used to compare different models.
RESULTS: Twenty-nine desmoplastic and sixty-four non-desmoplastic CRLMs were included. The radiomic models achieved area under the curve (AUC) values of 0.736, 0.906, 0.804, and 0.794 for RTumor, RT-2, RT+2, and R2+2, respectively, in the training cohorts. The AUC values were 0.713, 0.876, 0.785, and 0.777 for RTumor, RT-2, RT+2, and R2+2, respectively, in the validation cohort. RT-2 exhibited the best performance.
CONCLUSIONS: The MRI-based radiomic models could predict HGPs in CRLMs pre-operatively.

BACKGROUND: The prognostic predictive tool for patients with colorectal liver metastasis (CRLM) is limited and the criteria for administering preoperative neoadjuvant chemotherapy in CRLM patients remain controversial.
METHODS: This study enrolled 532 CRLM patients at West China Hospital (WCH) from January 2009 to December 2019. Prognostic factors were identified from the training cohort to construct a WCH-nomogram and evaluating accuracy in the validation cohort. Receiver operating characteristic (ROC) curve analysis was used to compare the prediction accuracy with other existing prediction tools.
RESULTS: From the analysis of the training cohort, four independent prognostic risk factors, namely tumor marker score, KRAS mutation, primary lymph node metastasis, and tumor burden score were identified on which a WCH-nomogram was constructed. The C-index of the two cohorts were 0.674 (95% CI: 0.634-0.713) and 0.655 (95% CI: 0.586-0.723), respectively, which was better than the previously reported predication scores (CRS, m-CS and GAME score). ROC curves showed AUCs for predicting 1-, 3-, and 5-year overall survival (OS) of 0.758, 0.709, and 0.717 in the training cohort, and 0.860, 0.669, and 0.692 in the validation cohort, respectively. A cutoff value of 114.5 points was obtained for the WCH-nomogram total score based on the maximum Youden index of the ROC curve of 5-year OS. Risk stratification showed significantly better prognosis in the low-risk group, however, the high-risk group was more likely to benefit from neoadjuvant chemotherapy.
CONCLUSIONS: The WCH-nomogram demonstrates superior prognostic stratification compared to prior scoring systems, effectively identifying CRLM patients who may benefit the most from neoadjuvant chemotherapy.

Colorectal cancer (CRC) is the third most prevalent cancer globally, and liver metastasis (CRLM) is the primary cause of death. Hence, it is essential to discover novel prognostic biomarkers and therapeutic drugs for CRLM.
This study developed two liver metastasis-associated prognostic signatures based on differentially expressed genes (DEGs) in CRLM. Additionally, we employed an interpretable deep learning model utilizing drug sensitivity databases to identify potential therapeutic drugs for high-risk CRLM patients. Subsequently, in vitro and in vivo experiments were performed to verify the efficacy of these compounds.
These two prognostic models exhibited superior performance compared to previously reported ones. Obatoclax, a BCL-2 inhibitor, showed significant differential responses between high and low risk groups classified by prognostic models, and demonstrated remarkable effectiveness in both Transwell assay and CT26 colorectal liver metastasis mouse model.
This study highlights the significance of developing specialized prognostication approaches and investigating effective therapeutic drugs for patients with CRLM. The application of a deep learning drug response model provides a new drug discovery strategy for translational medicine in precision oncology.

This study aims to investigate applicable robust biomarkers that can improve prognostic predictions for colorectal liver metastasis (CRLM) patients receiving simultaneous resection. A total of 1323 CRLM patients from multiple centres were included. The preoperative aspartate aminotransferase to platelet ratio index (APRI) level from blood of patients were obtained. Patients were stratified into a high APRI group and a low APRI group, and comparisons were conducted by analyzing progression-free survival (PFS), overall survival (OS) and postoperative early recurrence. Tumour samples of CRLM were collected to perform single-cell RNA sequencing and multiplex immunohistochemistry/immunofluorescence (mIHC/IF) to investigate the association of APRI levels and the tumour microenvironment of CRLM. Compared with APRI <0.33, PFS disadvantage (IPTW-adjusted HR = 1.240, P = 0.015) and OS disadvantage (IPTW- adjusted HR = 1.507, P = 0.002) of APRI ≥0.33 were preserved in the IPTW-adjusted Cox hazards regression analyses. An APRI ≥0.25 was associated with a significantly increased risk of postoperative early recurrence after adjustment (IPTW-adjusted OR = 1.486, P = 0.001). The external validation showed consistent results with the training cohort. In the high APRI group, the single-cell RNA sequencing results revealed a heightened malignancy of epithelial cells, the enrichment of inflammatory-like cancer-associated fibroblasts and SPP1+ macrophages associated with activation of malignant cells and fibrotic microenvironment, and a more suppressed-function T cells; mIHC/IF showed that PD1+ CD4+ T cells, FOXP3+ CD4+ T cells, PD1+ CD8+ T cells, FOXP3+ CD8+ T cells, SPP1+ macrophages and iCAFs were significantly increased in the intratumoral region and peritumoral region. This study contributed valuable evidence regarding preoperative APRI for predicting prognoses among CRLM patients receiving simultaneous resection and provided underlying clues supporting the association between APRI and clinical outcomes by single-cell sequencing bioinformatics analysis and mIHC/IF.

UNASSIGNED: No well-performing nomogram has been developed specifically to predict individual-patient cancer-specific survival (CSS) and overall survival (OS) among patients with resectable colorectal liver metastasis (CRLM) who undergo simultaneous resection of primary and hepatic lesions without neoadjuvant chemotherapy (NAC). We aim to investigate the prognosis of patients with resectable CRLM undergoing simultaneous resection of primary and hepatic lesions without NAC.
UNASSIGNED: Data of patients with CRLM in the Surveillance, Epidemiology and End Results Program (cohort, n = 225) were collected as the training set, and data of patients with CRLM treated at the National Cancer Center (cohort, n = 180) were collected as the validation set. The prognostic value of the clinicopathological parameters in the training cohort was assessed using Kaplan‒Meier curves and univariate and multivariate Cox proportional hazards models, and OS and CSS nomograms integrated with the prognostic variables were constructed. Calibration analyses, receiver operating characteristic (ROC) curves, and decision curve analyses (DCAs) were then performed to evaluate the performance of the nomograms.
UNASSIGNED: There was no collinearity among the collected variables. Three factors were associated with OS and CSS: the pretreatment carcinoembryonic antigen (CEA) concentration, pathologic N (pN) stage, and adjuvant chemotherapy (each p < 0.05). OS and CSS nomograms were constructed using these three parameters. The calibration plots revealed favorable agreement between the predicted and observed outcomes. The areas under the ROC curves were approximately 0.7. The DCA plots revealed that both nomograms had satisfactory clinical benefits. The ROC curves and DCAs also confirmed that the nomogram surpassed the tumor, node, and metastasis staging system.
UNASSIGNED: The herein-described nomograms containing the pretreatment CEA concentration, pN stage, and adjuvant chemotherapy may be effective models for predicting postoperative survival in patients with CRLM.

This study aimed to investigate the causal associations between several liver traits (liver iron content, percent liver fat, alanine transaminase levels, and liver volume) and colorectal cancer (CRC) risk using a Mendelian randomization (MR) approach to improve our understanding of the disease and its management.
Genetic variants were used as instrumental variables, extracted from genome-wide association studies (GWAS) datasets of liver traits and CRC. The Two-Sample MR package in R was used to conduct inverse variance weighted (IVW), MR Egger, Maximum likelihood, Weighted median, and Inverse variance weighted (multiplicative random effects) MR approaches to generate overall estimates of the effect. MR analysis was conducted with Benjamini-Hochberg method-corrected P values to account for multiple testing (P < 0.013). MR-PRESSO was used to identify and remove outlier genetic variants in Mendelian randomization (MR) analysis. The MR Steiger test was used to assess the validity of the assumption that exposure causes outcomes. Leave-one-out validation, pleiotropy, and heterogeneity testing were also conducted to ensure the reliability of the results. Multivariable MR was utilized for validation of our findings using the IVW method while also adjusting for potential confounding or pleiotropy bias.
The MR analysis suggested a causal effect between liver volume and a reduced risk of CRC (OR 0.60; 95% CI, 0.44-0.82; P = 0.0010) but did not provide evidence for causal effects of liver iron content, percent liver fat, or liver alanine transaminase levels. The MR-PRESSO method did not identify any outliers, and the MR Steiger test confirmed that the causal direction of the analysis results was correct in the Mendelian randomization analysis. MR results were consistent with heterogeneity and pleiotropy analyses, and leave-one-out analysis demonstrated the overall values obtained were consistent with estimates obtained when all available SNPs were included in the analysis. Multivariable MR was utilized for validation of our findings using the IVW method while also adjusting for potential confounding or pleiotropy bias.
The study provides tentative evidence for a causal role of liver volume in CRC, while genetically predicted levels of liver iron content, percent liver fat, and liver alanine transaminase levels were not associated with CRC risk. The findings may inform the development of targeted therapeutic interventions for colorectal liver metastasis (CRLM) patients, and the study highlights the importance of MR as a powerful epidemiological tool for investigating causal associations between exposures and outcomes.

For patients with initially unresectable colorectal liver metastasis (IU-CRLM) receiving conversion therapy, disease relapse after conversion hepatectomy is common. However, few studies have focused on the assessment and management of relapse following conversion hepatectomy for IU-CRLM.
In the retrospective cohort study, 255 patients with IU-CRLM received conversion therapy and underwent subsequent R0 resection. The treatment effects of repeated liver-directed treatment (RLDT) versus non-RLDT for liver relapse were examined. Survival analysis was evaluated with the use of Cox proportional hazards methods. The importance of RLDT was further confirmed in the propensity score matching (PSM) and subgroup analyses.
The 5-year overall survival (OS) rate after conversion hepatectomy was 34.9%. Liver relapse was observed in 208 patients. Of these patients, 106 underwent RLDT (65 underwent repeated hepatectomy and the remainder underwent ablation treatment), while 102 received only palliative chemotherapy. The relapse patients who underwent RLDT had a significantly longer OS than those who did not (hazard ratio (HR): 0.382, 95% CI: 0.259-0.563; P<0.001). In a multivariable analysis, RLDT was independently associated to prolonged survival (HR: 0.309, 95%CI: 0.181-0.529; P<0.001). In the PSM and subgroup analyses, RLDT consistently showed evidence of prolonging OS significantly.
For IU-CRLM patients with liver relapse following conversion hepatectomy, the RLDT is essential for cure and prolonged survival. To avoid missing the opportunity for RLDT, intensive disease surveillance should be proposed.

OBJECTIVE: This study aims to investigate the efficiency of a radiomics model in identifying high-frequency microsatellite instability (MSI-H) and microsatellite stability (MSS) of colorectal liver metastasis (CRLM) according to machine learning radiomics features of enhanced CT liver images.
METHODS: A total of 12 patients with MSI-H CRLM and 96 patients with MSS CRLM were randomly divided into the training group and internal validation group according to the ratio of 7: 3 (training: 75 cases, validation: 33 cases). From the enhanced CT (portal phase) image data of patients, 788 radiomics features were extracted, and a random forest model was established with the optimal features selected. The receiver operating characteristics (ROC) curve analysis was performed to assess the model\'s diagnostic efficacy.
RESULTS: The training group comprised 8 patients with MSI-H CRLM and 67 patients with MSS CRLM, and the internal validation group included 4 patients with MSI-H CRLM and 29 patients with MSS CRLM. After feature selection, 7 radiomics features good for distinguishing MSI-H CRLM and MSS CRLM were screened out. The ROC curve analysis demonstrated that the random forest model had the AUC (area under the ROC curve) value 0.88, accuracy 0.85, sensitivity 0.85, specificity 0.92, and F1 score 0.88 in the training group. The model had an AUC value of 0.75, accuracy of 0.74, sensitivity of 0.81, specificity of 0.85, and F1_score of 0.78 in the internal validation group in identifying the MSI-H from the MSS CRLM. In order to evaluate the robustness of the overall model, the 788 features obtained were all applied to the 5-fold cross-validation, with the model being built on the random forest and analyzed with the ROC curve analysis. The AUC value of the model was 0.86 (P＜0.05), accuracy value 0.91, sensitivity 0.60, and specificity 0.95.
CONCLUSIONS: The random forest prediction model built on the radiometric features extracted from enhanced CT images can be used to identify the MSI-H from the MSS CRLM and may provide effective guidance for clinical immunotherapy of CRLM patients with unknown MSI status.