Real-Time Fluorescence Imaging to Identify Cholangiocarcinoma in the Extrahepatic Biliary Tree Using an Enzyme-Activatable Probe.

Introduction: Complete resection is the only possible treatment for cholangiocarcinoma in the extrahepatic biliary tree (eCCA), although current imaging modalities are limited in their ability to accurately diagnose longitudinal spread. We aimed to develop fluorescence imaging techniques for real-time identification of eCCA using an enzyme-activatable probe, which emits fluorescence immediately after activation by a cancer-specific enzyme. Methods: Using lysates and small tissue fragments collected from surgically resected specimens, we selected the most specific probe for eCCA from among 800 enzyme-activatable probes. The selected probe was directly sprayed onto resected specimens and fluorescence images were acquired; these images were evaluated for diagnostic accuracy. We also comprehensively searched for enzymes that could activate the probe, then compared their expression levels in cancer and non-cancer tissues. Results: Analyses of 19 samples (four cancer lysates, seven non-cancer lysates, and eight bile samples) and 54 tissue fragments (13 cancer tissues and 41 non-cancer tissues) revealed that PM-2MeSiR was the most specific fluorophore for eCCA. Fluorescence images of 7 patients were obtained; these images enabled rapid identification of cancerous regions, which closely matched histopathology findings in 4 patients. Puromycin-sensitive aminopeptidase was identified as the enzyme that might activate the probe, and its expression was upregulated in eCCA. Conclusion: Fluorescence imaging with PM-2MeSiR, which may be activated by puromycin-sensitive aminopeptidase, yielded generally high accuracy. This technique may be useful for real-time identification of the spread of eCCA during surgery and endoscopic examinations.

Two-step artificial intelligence algorithm for liver segmentation automates anatomic virtual hepatectomy.

BACKGROUND: Anatomic virtual hepatectomy with precise liver segmentation for hemilivers, sectors, or Couinaud's segments using conventional three-dimensional simulation is not automated and artificial intelligence (AI)-based algorithms have not yet been applied. METHODS: Computed tomography data of 174 living-donor candidates for liver transplantation (training data) were used for developing a new two-step AI algorithm to automate liver segmentation that was validated in another 51 donors (validation data). The Pure-AI (no human intervention) and ground truth (GT, full human intervention) data groups were compared. RESULTS: In the Pure-AI group, the median Dice coefficients of the right and left hemilivers were highly similar, 0.95 and 0.92, respectively; sectors, posterior to lateral: 0.86-0.92, and Couinaud's segments 1-8: 0.71-0.89. Labeling of the first-order branch as hemiliver, right or left portal vein perfectly matched; 92.8% of the second-order (sectors); 91.6% of third-order (segments) matched between the Pure-AI and GT data. CONCLUSIONS: The two-step AI algorithm for liver segmentation automates anatomic virtual hepatectomy. The AI-based algorithm correctly divided all hemilivers, and more than 90% of the sectors and segments.

Effect of the response to preoperative treatment for hepatorenal syndrome on the outcome of recipients of living-donor liver transplantation.

BACKGROUND: The effect of pretransplant hepatorenal syndrome (HRS) on the outcomes of living-donor liver transplantation (LDLT) recipients with special reference to the recovery of HRS before LDLT was investigated. METHODS: The rate of HRS was 43.9% (125/285) among the cohort, and the subjects were divided into three groups: those without HRS (No-HRS group, n = 160), those with HRS but recovered following pretransplant renal function restoration treatment (Responders group, n = 55), and those with persistent HRS (Non-responders group, n = 70). RESULTS: While the 1-, 3-, and 5-year patient survival rates were comparable between those with and without HRS (89.6%, 84.7%, and 84.7% vs 95.6%, 92.2%, and 87.5%), the cumulative incidence of the development of posttransplant chronic kidney disease (CKD) was significantly higher in those with HRS (P < .001). In addition, there was a significant difference between Responders and Non-responders in the development of CKD (P = .01). In the Cox regression model, Non-responders (P = .032, HR 1.79 [95% C.I. 1.05-3.03]) and recipient age (P = .014, HR 1.62 [95% C.I. 1.10-2.37]) were independent predictors for the development of CKD after LDLT. CONCLUSION: Living-donor liver transplantation is safe and effective for patients with HRS, and CKD progression could be reduced among those with HRS who responded to renal restoration treatment.

Artificial intelligence enhances the accuracy of portal and hepatic vein extraction in computed tomography for virtual hepatectomy.

BACKGROUND/PURPOSE: Current conventional algorithms used for 3-dimensional simulation in virtual hepatectomy still have difficulties distinguishing the portal vein (PV) and hepatic vein (HV). The accuracy of these algorithms was compared with a new deep-learning based algorithm (DLA) using artificial intelligence. METHODS: A total of 110 living liver donor candidates until 2017, and 46 donor candidates until 2019 were allocated to the training group and validation groups for the DLA, respectively. All PV or HV branches were labeled based on Couinaud's segment classification and the Brisbane 2000 Terminology by hepato-biliary surgeons. Misclassified and missing branches were compared between a conventional tracking-based algorithm (TA) and DLA in the validation group. RESULTS: The sensitivity, specificity, and Dice coefficient for the PV were 0.58, 0.98, and 0.69 using the TA; and 0.84, 0.97, and 0.90 using the DLA (P < .001, excluding specificity); and for the HV, 0.81, 087, and 0.83 using the TA; and 0.93, 0.94 and 0.94 using the DLA (P < .001 to P = .001). The DLA exhibited greater accuracy than the TA. CONCLUSION: Compared with the TA, artificial intelligence enhanced the accuracy of extraction of the PV and HVs in computed tomography.

Fluorescence Imaging Using Enzyme-Activatable Probes for Real-Time Identification of Pancreatic Cancer.

INTRODUCTION: Radical resection is the only curative treatment for pancreatic cancer, which is a life-threatening disease. However, it is often not easy to accurately identify the extent of the tumor before and during surgery. Here we describe the development of a novel method to detect pancreatic tumors using a tumor-specific enzyme-activatable fluorescence probe. METHODS: Tumor and non-tumor lysate or small specimen collected from the resected specimen were selected to serve as the most appropriate fluorescence probe to distinguish cancer tissues from noncancerous tissues. The selected probe was sprayed onto the cut surface of the resected specimen of cancer tissue to acquire a fluorescence image. Next, we evaluated the ability of the probe to detect the tumor and calculated the tumor-to-background ratio (TBR) by comparing the fluorescence image with the pathological extent of the tumor. Finally, we searched for a tumor-specific enzyme that optimally activates the selected probe. RESULTS: Using a library comprising 309 unique fluorescence probes, we selected GP-HMRG as the most appropriate activatable fluorescence probe. We obtained eight fluorescence images of resected specimens, among which four approximated the pathological findings of the tumor, which achieved the highest TBR. Finally, dipeptidyl-peptidase IV (DPP-IV) or a DPP-IV-like enzyme was identified as the target enzyme. CONCLUSION: This novel method may enable rapid and real-time visualization of pancreatic cancer through the enzymatic activities of cancer tissues.