RESUMO
Background: With the continuous innovation of magnetic resonance imaging (MRI) hardware and software technology, amide proton transfer-weighted (APTw) imaging has been applied in liver cancer. However, to our knowledge, no study has evaluated the feasibility of a three-dimensional amide proton transfer-weighted (3D-APTw) imaging sequence for hepatocellular carcinoma (HCC). This study thus aimed to conduct an image quality assessment of 3D-APTw for HCC and to explore its feasibility. Methods: 3D-APTw MRI examinations were completed in 134 patients with clinically suspected HCC. According to the uniformity of APTw signal in the liver and within the lesion and the proportion of artifact and missing signal regions, APTw images were subjectively scored using a 5-point scale. The scanning success rate of liver APTw imaging was calculated as the ratio of the number of cases with a quality assurance measurement of more than 3 to the total number of HCC cases. The intra- and interobserver quality assurance measurements for APTw images were compared via the Kappa consistency test. Within the HCC cases with a minimum image quality threshold of 3 points, the APT values of HCC and the liver parenchyma, signal-to-noise ratio of APT-weighted images (SNRAPTw), and contrast-to-noise ratio of HCC (CNRHCC) were measured by two observers. The intra- and interobserver agreement was assessed using the intraclass correlation coefficient (ICC). The differences in APT values between HCC and liver parenchyma was determined using the Mann-Whitney test. Results: Sixty-six HCC cases with a quality assurance measurement of APTw imaging were included in the final analysis, and the calculated success rate was 70.21% (66/94). The subjective APT image quality scores of the two observers were consistent (3.66±1.18, 3.50±1.19, and 3.68±1.18), and no intergroup or intragroup statistical differences were found (P=0.594, and P=0.091), but the consistency of inter- and intraobserver was not as satisfactory (κ=0.594 and κ=0.580). The APT values in HCC lesion were significantly higher than those in liver parenchyma (2.73%±0.91% vs. 1.62%±0.55%; P<0.001). The APT values in HCC showed favorable intra- and interobserver consistency between the two observers (ICC =0.808 and ICC =0.853); the APT values in liver parenchyma, SNRAPTw, and CNRHCC values had moderate intraobserver consistency (ICC =0.578, ICC =0.568, and ICC =0.508) and interobserver consistency (ICC =0.599, ICC =0.199, and ICC =0.650). The coefficients of variation of the APTw values in the HCC lesion and in liver parenchyma were 33.4% and 34.4%, respectively. The SNRAPTw and CNRHCC were 30.75±18.74 and 3.56±3.19, with a coefficient of variation of 60.9% and 74.9%, respectively. Conclusions: Liver 3D-APTw imaging was preliminarily demonstrated to be clinically feasible for evaluating HCC.
RESUMO
BACKGROUND: Liver cancer is a highly lethal and aggressive form of cancer that poses a significant threat to patient survival. Within this category, liver hepatocellular carcinoma (LIHC) represents the most common subtype of liver cancer. Despite decades of research and treatment, the overall survival rate for LIHC has not significantly improved. Improved models are necessary to differentiate high-risk cases and predict possible treatment options for LIHC patients. Recent studies have identified a set of genes associated with neutrophil extracellular traps (NETs) that may contribute to tumor growth and metastasis; however, their prognostic value in LIHC has yet to be established. This study aims to construct a prognostic signature based on a set of NET-related genes (NRGs) for patients diagnosed with LIHC. METHODS: The transcriptomic data and clinical information concerning LIHC patients were procured from the Cancer Genome Atlas (TCGA) and International Cancer Genome Consortium LIHC (ICLIHC) databases, respectively. To determine the NRG subtypes, the k-means algorithm was employed, along with consensus clustering. The aforementioned analysis aided the construction of a prognostic signature utilizing the last absolute shrinkage and selection operator Cox analysis. To validate the prognostic model, an external dataset, receiver operating characteristic curve, and principal component analysis were utilized. Moreover, the immune microenvironment and the proportion of immune cells between high- and low-risk cases were scrutinized by ESTIMATE and CIBERSORT algorithms. Finally, gene set enrichment analysis was executed to investigate the potential mechanism of NRGs in the pathogenesis and prognosis of LIHC. RESULTS: Two molecular subtypes of LIHC were identified based on the expression patterns of differentially expressed NRGs (DE-NRGs). The two subtypes demonstrated significant differences in survival rates and immune cell expression levels. The study results demonstrated the role of NRGs in antigen presentation, which led to the promotion of tumor immune escape. A risk model was developed and validated with strong overall survival prediction ability. The model, comprising 34 NRGs, showed a strong ability to predict prognosis. CONCLUSION: We built a dependable prognostic signature based on NRGs for LIHC. We identified that NRGs could have a significant interaction in LIHC's immune microenvironment and therapeutic response. This finding offers insight into the molecular mechanisms and targeted therapy for LIHC.