Identification of Predominant Histopathological Growth Patterns of Colorectal Liver Metastasis by Multi-Habitat and Multi-Sequence Based Radiomics Analysis

Abstract

Developing an MRI-based radiomics model to effectively and accurately predict the predominant histopathologic growth patterns (HGPs) of colorectal liver metastases (CRLMs). Materials and Methods: In this study, 182 resected and histopathological proven CRLMs of chemotherapy-naive patients from two institutions, including 123 replacement CRLMs and 59 desmoplastic CRLMs, were retrospectively analyzed. Radiomics analysis was performed on two regions of interest (ROI), the tumor zone and the tumor-liver interface (TLI) zone. Decision tree (DT) algorithm was used for radiomics modeling on each MR sequence, and fused radiomics model was constructed by combining the radiomics signature of each sequence. The clinical and combination models were developed through multivariate logistic regression method. The performance of the developed models was assessed by receiver operating characteristic (ROC) curves with indicators of area under curve (AUC), accuracy, sensitivity, and specificity. A nomogram was constructed to evaluate the discrimination, calibration, and usefulness. Results: The fused radiomics tumor and radiomics TLI models showed better performance than any single sequence and clinical model. In addition, the radiomics TLI model exhibited better performance than radiomics tumor model (AUC of 0.912 vs. 0.879) in internal validation cohort. The combination model showed good discrimination, and the AUC of nomogram was 0.971, 0.909, and 0.905 in the training, internal validation, and external validation cohorts, respectively. Han et al. Radiomics in CRLM's HGP Prediction Conclusion: MRI-based radiomics method has high potential in predicting the predominant HGPs of CRLM. Preoperative non-invasive identification of predominant HGPs could further explore the ability of HGPs as a potential biomarker for clinical treatment strategy, reflecting different biological pathways.

Different Forms of Tumor Vascularization and Their Clinical Implications Focusing on Vessel Co-option in Colorectal Cancer Liver Metastases

“…Besides therapeutic challenges, a pre-operative diagnosis of the exact HGP in CRCLM is still an essential topic of research to achieve the maximum clinical benefit for CRCLM patients. As previously described, multi-detector computer tomography, and also radiation-free magnet resonance based radiomics seem to be promising tools to assess the predominant HGP (Cheng et al, 2019;Han et al, 2020). However, in both techniques the predominant HGP was classified as such, when >50% of the tumor border displayed the specific HGP (Cheng et al, 2019;Han et al, 2020).…”

“…As previously described, multi-detector computer tomography, and also radiation-free magnet resonance based radiomics seem to be promising tools to assess the predominant HGP (Cheng et al, 2019;Han et al, 2020). However, in both techniques the predominant HGP was classified as such, when >50% of the tumor border displayed the specific HGP (Cheng et al, 2019;Han et al, 2020). Referring to the study of Galjart et al (2019), the presence of any non-desmoplastic HGP in CRCLM impairs the patients' prognosis.…”

“…The morphological response was assessed by rim enhancement, overall attenuation, and characteristics of the tumor liver interface (Chun et al, 2009). Recently, Han et al (2020) showed that also a magnetic resonance based radiomics model is able to predict the predominant HGP in CRCLM. Especially radiomics of the tumor-liver-interface, combined with clinical characteristics provide a high diagnostic accuracy (Han et al, 2020).…”

“…Recently, Han et al (2020) showed that also a magnetic resonance based radiomics model is able to predict the predominant HGP in CRCLM. Especially radiomics of the tumor-liver-interface, combined with clinical characteristics provide a high diagnostic accuracy (Han et al, 2020). Thus, these findings indicate the possibility to deduce the pathological response of CRCLM from a non-invasive imaging method.…”

Clinico‐metabolic characterization improves the prognostic value of histological growth patterns in patients undergoing surgery for colorectal liver metastases

“…In that sense, having access to noninvasive means, such as CT-or magnetic resonance imaging (MRI)-imaging and radiomics to repeatedly identify the HGP of CRLM in a patient would be of significantly added value, both to improve its reliability as a biomarker and to better understand tumor biology, including response to chemotherapy. 52,53 More particularly, the MRI-technique may allow now to precisely explore the tumor and the peripheral liver parenchyma compartments, offering promising perspectives in this respect. 54 In parallel, recent developments in molecular imaging techniques using specific PET tracers could provide qualitative, quantitative and dynamic evaluation on the nature of intra-and peritumoral cellular infiltrates, such as lymphocytes or fibroblasts.…”