Liver segment imaging using monocyte sequestration: a potential tool for fluorescence-guided liver surgery

Automatic Liver Segmentation in CT Images with Enhanced GAN and Mask Region-Based CNN Architectures

Liver image segmentation has been increasingly employed for key medical purposes, including liver functional assessment, disease diagnosis, and treatment. In this work, we introduce a liver image segmentation method based on generative adversarial networks (GANs) and mask region-based convolutional neural networks (Mask R-CNN). Firstly, since most resulting images have noisy features, we further explored the combination of Mask R-CNN and GANs in order to enhance the pixel-wise classification. Secondly, k-means clustering was used to lock the image aspect ratio, in order to get more essential anchors which can help boost the segmentation performance. Finally, we proposed a GAN Mask R-CNN algorithm which achieved superior performance in comparison with the conventional Mask R-CNN, Mask-CNN, and k-means algorithms in terms of the Dice similarity coefficient (DSC) and the MICCAI metrics. The proposed algorithm also achieved superior performance in comparison with ten state-of-the-art algorithms in terms of six Boolean indicators. We hope that our work can be effectively used to optimize the segmentation and classification of liver anomalies.

Low efficiency of leucocyte plugging-based drug delivery to cancer in mice

Cells of the immune system were proposed for use as Trojan horse for tumour-specific drug delivery. The efficacy of such cell-based drug delivery depends on the site-specific cell homing. This present study was aimed to investigate the potential of leucocytes for intratumoural site-specific enrichment using a locoregional application route in experimental liver tumours. Human neutrophils were isolated from peripheral blood and directly labelled with calcein AM or loaded with doxorubicin. The neutrophil loading and release of doxorubicin and the migration and adhesion to ICAM-1 were analysed in vitro. Macrophages were isolated and activated in vitro. Leucocyte plugging and the distribution pattern in the liver microvasculature were studied ex vivo, and the efficacy of leucocyte plugging in tumour blood vessels was analysed in vivo after superselective intra-arterial injection in mouse liver tumour models. Neutrophils were characterised by the high dose-dependent uptake and rapid release of doxorubicin. Doxorubicin loading did not affect neutrophil migration function. Neutrophil plugging in liver microvasculature was very high (> 90%), both after ex vivo perfusion and after injection in vivo. However, neutrophils as well as activated macrophages plugged insufficiently in tumour blood vessels and passed through the tumour microvasculture with a very low sequestration rate in vivo. Neutrophils possess several properties to function as potentially effective drug carriers; however, the tumour site-specific drug delivery after selective locoregional injection was observed to be insufficient owing to low intratumoural microvascular plugging.

Raman optical identification of renal cell carcinoma via machine learning

High pathologic tumor-node-metastasis (pTNM) stage grade or Fuhrman grade indicates poor oncological outcome in renal cell carcinoma (RCC). Early diagnosis and screening of these RCCs and adjust surgical planning accordingly are greatly beneficial to patients. Raman spectroscopy is a highly specific fingerprint spectrum on molecular level, pretty appropriate for label-free and noninvasive cancer diagnosis. In this work we established a Raman spectrum-based supporting vector machine (SVM) model to accurately ex vivo distinguish human renal tumor from normal tissues and fat with an accuracy of 92.89%. The model can also be used to determine tumor boundary, showing consistent results to pathological staining analysis. This method can be additionally used to accomplish the classification purposes of renal tumor subtypes and grades with an accuracy of 86.79% and 89.53%, respectively. Therefore, we prove that Raman spectroscopy has great potential in the rapid and accurate clinical diagnosis of renal cancers.

Demarcation Line Assessment in Anatomical Liver Resection: An Overview

Anatomical liver resection (ALR) is the preferred oncological approach for the treatment of primary liver malignancies, such as hepatocellular carcinoma and intrahepatic cholangiocarcinoma. The demarcation line (DL) is formed by means of selective vascular occlusion and is used by surgeons to guide ALR. Emerging intraoperative technologies are playing a major role to enhance the surgeon’s vision and ensure a precise oncologic surgery. In this article, a brief overview of modalities to assess the DL during ALRs is presented, from the established conventional techniques to future perspectives.

Fluorescence-guided fiber-optic micronavigation using microscopic identification of vascular boundary of liver segment and tumors

Background: The exact identification of tumor boundaries and related liver segments is especially important for liver tumor surgery. This study aimed to evaluate a new approach for vascular boundary assessment and surgical navigation based on fiber-optic microscopy and microvascular fluorescence labeling.

Methods: Antibody clones with fast binding ability were identified and selected using immunofluorescence. We evaluated the endothelial capture efficacy for an anti-mouse CD31 antibody labeled with different fluorophores and different degrees of labeling ex vivo. Segment boundary identification and navigation potential using endothelial capture were explored by two different fiber-optic microscopy systems. Finally, microvasculature labeling and fiber-optic microscopy were used to identify and treat microscopic liver tumors in vivo.

Results: The following monoclonal antibodies were selected: anti-mouse CD31 (clone 390), anti-mouse CD54 (YN1/1.7.4), anti-human CD31 (WM59), and anti-human CD54 (HA58). These clones showed fast binding to endothelial cells and had long half-lives. The fluorophore choice and the degree of antibody labeling did not significantly affect capture efficacy in an isolated liver perfusion model. The microvascular system was clearly identified with wide-field fiber-optic microscopy after labeling the endothelium with low doses of specific antibodies, and the specifically labeled liver segment could be microscopically dissected. High antibody doses were required for confocal laser endomicroscopy. After microscopically identifying the vascular margin in vivo, tumor thermoablation strongly reduced tumor size or totally eliminated tumors.

Conclusions: We demonstrated that vascular boundaries of liver tumors and locally perfused liver segments were accurately identified and surgical micronavigation was facilitated with fiber-optic microscopy and selected endothelium-specific antibodies.

Endothelial capture using antibodies and nanoparticles in human tissues: Antigen identification and liver segment imaging

BACKGROUND

The unique phenomenon of endothelial antibody capture (endocapt) leads to site-specific accumulation of antibodies on the endothelium after its locoregional injection. The potential of this phenomenon has already been demonstrated in animal models. In the present study, the translational potential of several human endothelium-specific antibodies for their use in the endocapt-based approach was analysed.

METHODS

The binding of different endothelium-surface specific monoclonal antibody clones was analysed in human tissue and in endothelial cells using image-based immunofluorescence and the determination of half-maximal effective concentration (EC₅₀). The potential of endocapt-based locoregional application of antibodies or antibody-coated liposomes was analysed ex vivo using isolated mouse liver perfusion and in vivo using superselective injection in tumour models.

RESULTS

Eight out of ten antibody clones were assigned to the group of "fast binding antibodies". Different antibody clones showed various binding kinetics to the same endothelial marker whereas the binding kinetics of single antibody clones was independent from the tissue type. Anti-CD49e, anti-CD31, anti-CD34 and anti-CD102 antibodies showed the lowest EC₅₀ of antibody binding concentration and constant results in EC₅₀ determination of antibody binding to cells and human tissue. Experimental studies using anti-mouse CD49e antibody and coated immunoliposomes confirmed their effective capture by endothelial cells in vitro and in vivo by which fluorescent liver segment labelling was achieved.

CONCLUSIONS

Our findings identify the high potential of several human antibody clones, especially anti-CD49e, -CD31, -CD34 and -CD102, for endocapt technology. We also propose important translational implications of these antibodies for image-guided liver surgery and for use of nanoparticles/immunoliposomes. Toxicological studies are indispensable for further translational development of new antibodies for endocapt.

STATEMENT OF SIGNIFICANCE

The phenomenon of endothelial antibody capture (endocapt) leads to site-specific accumulation of antibodies on the endothelium after its locoregional injection. This phenomenon broadly prevents systemic circulation of the antibody or antibody-drug conjugates. In the present study, our findings identify several human antibody clones promising for endothelial capture technology. This study provided the experimental demonstration of liver segment labelling ex vivo using isolated mouse liver perfusion and in vivo using superselective injection in tumor models. In addition, this study proposed the important translational implications of selected antibodies for image-guided liver surgery and for use of nanoparticles/immunoliposomes.