Computerized Diagnosis of Liver Tumors From CT Scans Using a Deep Neural Network Approach.

The liver is a frequent site of benign and malignant, primary and metastatic tumors. Hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC) are the most common primary liver cancers, and colorectal liver metastasis (CRLM) is the most common secondary liver cancer. Although the imaging characteristic of these tumors is central to optimal clinical management, it relies on imaging features that are often non-specific, overlap, and are subject to inter-observer variability. Thus, in this study, we aimed to categorize liver tumors automatically from CT scans using a deep learning approach that objectively extracts discriminating features not visible to the naked eye. Specifically, we used a modified Inception v3 network-based classification model to classify HCC, ICC, CRLM, and benign tumors from pretreatment portal venous phase computed tomography (CT) scans. Using a multi-institutional dataset of 814 patients, this method achieved an overall accuracy rate of 96%, with sensitivity rates of 96%, 94%, 99%, and 86% for HCC, ICC, CRLM, and benign tumors, respectively, using an independent dataset. These results demonstrate the feasibility of the proposed computer-assisted system as a novel non-invasive diagnostic tool to classify the most common liver tumors objectively.

Ketohexokinase-mediated fructose metabolism is lost in hepatocellular carcinoma and can be leveraged for metabolic imaging.

The ability to break down fructose is dependent on ketohexokinase (KHK) that phosphorylates fructose to fructose-1-phosphate (F1P). We show that KHK expression is tightly controlled and limited to a small number of organs and is down-regulated in liver and intestinal cancer cells. Loss of fructose metabolism is also apparent in hepatocellular adenoma and carcinoma (HCC) patient samples. KHK overexpression in liver cancer cells results in decreased fructose flux through glycolysis. We then developed a strategy to detect this metabolic switch in vivo using hyperpolarized magnetic resonance spectroscopy. Uniformly deuterating [2-13C]-fructose and dissolving in D2O increased its spin-lattice relaxation time (T1) fivefold, enabling detection of F1P and its loss in models of HCC. In summary, we posit that in the liver, fructolysis to F1P is lost in the development of cancer and can be used as a biomarker of tissue function in the clinic using metabolic imaging.

Multi-Center Analysis of Liver Transplantation for Combined Hepatocellular Carcinoma-Cholangiocarcinoma Liver Tumors.

BACKGROUND: Combined hepatocellular-cholangiocarcinoma liver tumors (cHCC-CCA) with pathologic differentiation of both hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma within the same tumor are not traditionally considered for liver transplantation due to perceived poor outcomes. Published results are from small cohorts and single centers. Through a multicenter collaboration, we performed the largest analysis to date of the utility of liver transplantation for cHCC-CCA. STUDY DESIGN: Liver transplant and resection outcomes for HCC (n = 2,998) and cHCC-CCA (n = 208) were compared in a 12-center retrospective review (2009 to 2017). Pathology defined tumor type. Tumor burden was based on radiologic Milan criteria at time of diagnosis and applied to cHCC-CCA for uniform analysis. Kaplan-Meier survival curves and log-rank test were used to determine overall survival and disease-free survival. Cox regression was used for multivariate survival analysis. RESULTS: Liver transplantation for cHCC-CCA (n = 67) and HCC (n = 1,814) within Milan had no significant difference in overall survival (5-year cHCC-CCA 70.1%, HCC 73.4%, p = 0.806), despite higher cHCC-CCA recurrence rates (23.1% vs 11.5% 5 years, p < 0.001). Irrespective of tumor burden, cHCC-CCA tumor patient undergoing liver transplant had significantly superior overall survival (p = 0.047) and disease-free survival (p < 0.001) than those having resection. For cHCC-CCA within Milan, liver transplant was associated with improved disease-free survival over resection (70.3% vs 33.6% 5 years, p < 0.001). CONCLUSIONS: Regardless of tumor burden, outcomes after liver transplantation are superior to resection for patients with cHCC-CCA. Within Milan criteria, liver transplant for cHCC-CCA and HCC result in similar overall survival, justifying consideration of transplantation due to the higher chance of cure with liver transplantation in this traditionally excluded population.

Recurrence After Liver Resection of Colorectal Liver Metastases: Repeat Resection or Ablation Followed by Hepatic Arterial Infusion Pump Chemotherapy.

BACKGROUND: The aim of this study was to investigate the effectiveness of adjuvant hepatic arterial infusion pump (HAIP) chemotherapy after complete resection or ablation of recurrent colorectal liver metastases (CRLM). METHODS: A retrospective cohort study was conducted of patients from two centers who were treated with resection and/or ablation of recurrent CRLM only between 1992 and 2018. Overall survival (OS) and hepatic disease-free survival (hDFS) were estimated using the Kaplan-Meier method. The Cox regression method was used to calculate hazard ratios (HRs) with corresponding 95% confidence intervals (CI). RESULTS: Of 374 eligible patients, 81 (22%) were treated with adjuvant HAIP chemotherapy. The median follow-up for survivors was 65 months (IQR 32-118 months). Patients receiving adjuvant HAIP were more likely to have multifocal disease and receive perioperative systemic chemotherapy at time of resection for recurrence. A median hDFS of 46 months (95% CI 29-81 months) was found in patients treated with adjuvant HAIP compared with 18 months (95% CI 15-26 months) in patients treated with resection and/or ablation alone (p = 0.001). The median OS and 5-year OS were 89 months (95% CI 52-126 months) and 66%, respectively, in patients treated with adjuvant HAIP compared with 57 months (95% CI 47-67 months) and 47%, respectively, in patients treated with resection and/or ablation only (p = 0.002). Adjuvant HAIP was associated with superior hDFS (adjusted HR 0.599, 95% CI 0.38-0.93, p = 0.02) and OS (adjusted HR 0.59, 95% CI 0.38-0.92, p = 0.02) in multivariable analysis. CONCLUSION: Adjuvant HAIP chemotherapy after resection and/or ablation of recurrent CRLM is associated with superior hDFS and OS.

Spatial mapping of the collagen distribution in human and mouse tissues by force volume atomic force microscopy.

Changes in the elastic properties of living tissues during normal development and in pathological processes are often due to modifications of the collagen component of the extracellular matrix at various length scales. Force volume AFM can precisely capture the mechanical properties of biological samples with force sensitivity and spatial resolution. The integration of AFM data with data of the molecular composition contributes to understanding the interplay between tissue biochemistry, organization and function. The detection of micrometer-size, heterogeneous domains at different elastic moduli in tissue sections by AFM has remained elusive so far, due to the lack of correlations with histological, optical and biochemical assessments. In this work, force volume AFM is used to identify collagen-enriched domains, naturally present in human and mouse tissues, by their elastic modulus. Collagen identification is obtained in a robust way and affordable timescales, through an optimal design of the sample preparation method and AFM parameters for faster scan with micrometer resolution. The choice of a separate reference sample stained for collagen allows correlating elastic modulus with collagen amount and position with high statistical significance. The proposed preparation method ensures safe handling of the tissue sections guarantees the preservation of their micromechanical characteristics over time and makes it much easier to perform correlation experiments with different biomarkers independently.

Extrahepatic recurrence rates in patients receiving adjuvant hepatic artery infusion and systemic chemotherapy after complete resection of colorectal liver metastases.

BACKGROUND: This study investigated the effect of the reduced dose of systemic chemotherapy (SYS) on recurrence patterns in patients receiving adjuvant hepatic artery infusion (HAI) chemotherapy after complete colorectal liver metastases (CRLM) resection. METHODS: Patients undergoing complete CRLM resection between 2000 and 2007 were selected from a prospectively maintained database and categorized as receiving SYS or HAI + SYS. Those with pre and/or intraoperative extrahepatic disease, documented death, or recurrence within 30 days of CRLM resection were excluded. Competing risk, Fine and Gray's tests were used to compare SYS versus HAI + SYS for time-to-organ recurrence. RESULTS: Of 361 study patients, 153 (42.4%) received SYS and 208 (57.6%) received HAI + SYS. The median follow-up for survivors was 100 (range = 12-185) and 156 months (range = 18-217) for SYS and HAI + SYS, respectively. The 5-year cumulative incidence (CI) of any liver recurrence was greater for those receiving SYS (SYS = 41.9% vs. HAI + SYS = 28.6%, p = .005). The 5-year CI of developing any lung or extrahepatic recurrence for SYS patients was 36.2% and 47.9% compared with 44.5% (p = .242) and 51.7% (p = .551), respectively, in patients receiving HAI + SYS. CONCLUSION: Despite the reduced dose of SYS, adjuvant HAI + SYS after CRLM resection is not associated with a significantly increased risk of extrahepatic recurrence.

Addition of adjuvant hepatic artery infusion to systemic chemotherapy following resection of colorectal liver metastases is associated with reduced liver-related mortality.

BACKGROUND: After resection of colorectal liver metastases (CRLM), recurrent disease in the liver is a major cause of death but may be reduced with the addition of adjuvant hepatic arterial infusion (HAI) chemotherapy to systemic chemotherapy (SYS). OBJECTIVE: This study investigates organ-specific causes of death in patients receiving adjuvant HAI and SYS compared to adjuvant SYS alone. METHODS: Between 2000 and 2007, patients undergoing complete CRLM resection were identified from a prospectively maintained liver resection database and categorized as receiving HAI + SYS or SYS only. Using newly constructed definitions, mortality was attributed to specific organs (liver, lung, peritoneum, and brain) or infection. Univariate models and cumulative incidence functions were generated using competing risk methods. RESULTS: Of 361 eligible patients, 208 (57.6%) received HAI + SYS and 153 (42.4%) received SYS. The median follow up among survivors was 142 months (range = 12-217 months). Ten-year overall survival was 50.6% in the HAI + SYS group compared to 30.9% in those receiving SYS (P = .004). The 5-year cumulative incidence of liver-related mortality was 6.8% in the HAI + SYS group compared to 14.3% in the SYS group (P = .007). CONCLUSION: The addition of HAI to SYS after CRLM resection is associated with a 50% reduction in liver-related mortality at 5 years.

Correction to: Hypophosphatemia as a Predictor of Organ-Specific Complications Following Gastrointestinal Surgery: Analysis of 8034 Patients.

In the original version of this article, Martin R. Weiser's last name was spelled incorrectly. It is correct as reflected here. The original article has been updated.

Hypophosphatemia as a Predictor of Organ-Specific Complications Following Gastrointestinal Surgery: Analysis of 8034 Patients.

BACKGROUND: Organ-specific complications (OSC) remain serious potential sequela of gastrointestinal surgery. Hypophosphatemia correlates with poor outcomes and may be a harbinger of OSC after gastrointestinal surgery. Our goal was to describe and evaluate the relationship between postoperative phosphate levels and OSC. METHODS: Consecutive patients who underwent pancreatic, colorectal, or gastric resections were analyzed. OSC were defined as those resulting from the failure of at least one anastomosis performed during the primary resection, manifesting as an anastomotic leak, fistula, and/or intra-abdominal abscess. Postoperative serum phosphate levels and other recognized OSC risk factors were compared among patients who did and did not develop OSC. RESULTS: A total of 8034 patients who underwent pancreatic (n = 397), colorectal (n = 5808), or gastric (n = 1829) resections were included in the study. In each resection group, the majority of patients experienced hypophosphatemia postresection with the nadir on postoperative day (POD) 2, and the subgroups that developed OSC exhibited lower phosphate levels on POD3-7. On multivariate analysis, lower phosphate level on POD3 remained significantly associated with OSC following pancreatic resection [median (interquartile range) mmol/L, 0.65 (0.53-0.76) vs. 0.71 (0.61-0.84), p = 0.045] and colorectal resection [0.71 (0.61-0.87) vs. 0.77 (0.65-0.94), p = 0.006], and lower phosphate level on POD4 remained associated with OSC following gastric resection [0.87 (0.74-1.03) vs. 0.96 (0.81-1.13), p = 0.049]. CONCLUSION: This study identified a consistent trajectory of serum phosphate levels following 3 different gastrointestinal operations and association between early postoperative phosphate levels and OSC. Persistent lower phosphate levels should raise the level of concern for evolving postoperative leak and may lead to earlier radiographic evaluation and treatment.

A new technique to improve the mechanical and biological performance of ultra high molecular weight polyethylene using a nylon coating.

A new patent pending technique is proposed in this study to improve the mechanical and biological performance of ultra high molecular weight polyethylene (UHMWPE), i.e., to uniformly coat nylon onto the UHMWPE fiber (Firouzi et al., 2012). Mechanical tests were performed on neat and new nylon coated UHMWPE fibers to examine the tensile strength and creep resistance of the samples at different temperatures. Cytotoxicity and osteolysis induced by wear debris of the materials were investigated using (MTT) assay, and RT-PCR for tumor necrosis factor alpha (TNFα) and interleukin 6 (IL-6) osteolysis markers. Mechanical test results showed substantial improvement in maximum creep time, maximum breaking force, and toughness values of Nylon 6,6 and Nylon 6,12 coated UHMWPE fibers between average 15% and 60% at 25, 50, and 70°C. Furthermore, cytotoxicity studies have demonstrated significant improvement in cell viability using the nylon coated UHMWPE over the neat one (72.4% vs 54.8%) for 48h and (80.7 vs 5%) for 72h (P<0.01). Osteolysis test results have shown that the expression levels of TNFα and IL-6 markers induced by the neat UHMWPE fiber were significantly higher than those induced by the Nylon 6,6 coated UHMWPE (2.5 fold increase for TNFα at 48h, and three fold increase for IL-6 at 72h (P<0.01)). This study suggests that UHMWPE coated with nylon could be used as a novel material in clinical applications with lower cytotoxicity, less wear debris-induced osteolysis, and superior mechanical properties compared to neat UHMWPE.