Transformer-based multi-task learning for classification and segmentation of gastrointestinal tract endoscopic images.

Despite being widely utilized to help endoscopists identify gastrointestinal (GI) tract diseases using classification and segmentation, models based on convolutional neural network (CNN) have difficulties in distinguishing the similarities among some ambiguous types of lesions presented in endoscopic images, and in the training when lacking labeled datasets. Those will prevent CNN from further improving the accuracy of diagnosis. To address these challenges, we first proposed a Multi-task Network (TransMT-Net) capable of simultaneously learning two tasks (classification and segmentation), which has the transformer designed to learn global features and can combine the advantages of CNN in learning local features so that to achieve a more accurate prediction in identifying the lesion types and regions in GI tract endoscopic images. We further adopted the active learning in TransMT-Net to tackle the labeled image-hungry problem. A dataset was created from the CVC-ClinicDB dataset, Macau Kiang Wu Hospital, and Zhongshan Hospital to evaluate the model performance. Then, the experimental results show that our model not only achieved 96.94% accuracy in the classification task and 77.76% Dice Similarity Coefficient in the segmentation task but also outperformed those of other models on our test set. Meanwhile, active learning also produced positive results for the performance of our model with a small-scale initial training set, and even its performance with 30% of the initial training set was comparable to that of most comparable models with the full training set. Consequently, the proposed TransMT-Net has demonstrated its potential performance in GI tract endoscopic images and it through active learning can alleviate the shortage of labeled images.

Impact of opioid analgesics on the efficacy of immune checkpoint inhibitors in a lung cancer population.

OBJECTIVE: A retrospective clinical study was conducted to compare the prognosis between the opioid analgesic (OA) treated and OA-untreated groups and to evaluate the effect of opioid analgesics on the efficacy of immune checkpoint inhibitors (ICIs) in the treatment of advanced lung cancer patients. In addition, a subgroup analysis of the clinical characteristics of the enrolled patients was performed to explore possible influencing factors. METHODS: This study reviewed the medical records of eligible patients who received ICIs at our institution. The clinicopathological features and clinical outcomes were compared. Also, the use of OA was collected. Patient survival, the incidence of immune-related adverse events (irAEs), and other baseline variables were examined in both cohorts according to whether OA was used. RESULTS: A total of 132 patients were included in the study. Of them, 39 (29.5%) were in the OA-treated group. No significant differences in baseline characteristics were observed between the OA-treated and untreated groups. The combined application of OA treatment significantly shortened progression-free survival (PFS) (P < 0.001) and overall survival (OS) (P = 0.002). However, both groups experienced similar incidences and gradations of irAEs. According to multivariate analysis, OA treatment resulted in significantly worse PFS (HR = 4.994, 95% CI 3.217-7.753, P < 0.001) and OS (HR = 3.618, 95% CI 2.030-6.240, P < 0.001). CONCLUSIONS: Clinical outcomes of ICIs were significantly diminished in a cohort of Chinese patients with advanced lung cancer receiving OA therapy.

Biomass-related PM2.5 induces mitochondrial fragmentation and dysfunction in human airway epithelial cells.

The use of biomass for cooking and heating is considered an important factor associated with chronic obstructive pulmonary disease (COPD), but few studies have previously addressed its underlying mechanisms. Therefore, this research aimed to evaluate the effects of biomass-related PM2.5 (BRPM2.5) exposure on 16HBE human airway epithelial cells and in mice with regard to mitochondrial dysfunction. Our study indicated that BRPM2.5 exposure of 16HBE cells resulted in mitochondrial dysfunction, including decreased mitochondrial membrane potential, increased expression of fission proteins-phospho-DRP1, increased mitochondrial ROS (mtROS), and decreased levels of ATP. BRPM2.5 altered the mitochondrial metabolism of 16HBE cells by decreasing mitochondrial oxygen consumption and glycolysis. However, Mitochondria targeted peptide SS-31 eliminated mitochondrial ROS and alleviated the ATP deficiency and proinflammatory cytokines release. BRPM2.5 exposure resulted in abnormal mitochondrial morphological alterations both in 16HBE and in lung tissue. Taken together, these results suggest that BRPM2.5 has detrimental effects on human airway epithelial cells, leading to mitochondrial dysfunction, abnormal mitochondrial metabolism and altered mitochondrial dynamics. The present study provides the first evidence that disruption of mitochondrial structure and mitochondrial metabolism may be one of the mechanisms of BRPM2.5-induced respiratory dysfunction.

Therapeutic Targeting of Cancer: Epigenetic Homeostasis.

A large number of studies have revealed that epigenetics plays an important role in cancer development. However, the currently-developed epigenetic drugs cannot achieve a stable curative effect. Thus, it may be necessary to redefine the role of epigenetics in cancer development. It has been shown that embryonic development and tumor development share significant similarities in terms of biological behavior and molecular expression patterns, and epigenetics may be the link between them. Cell differentiation is likely a manifestation of epigenetic homeostasis at the cellular level. In this article, we introduced the importance of epigenetic homeostasis in cancer development and analyzed the shortcomings of current epigenetic treatment regimens. Understanding the dynamic process of epigenetic homeostasis in organ development can help us characterize cancer according to its differentiation stages, explore new targets for cancer treatment, and improve the clinical prognosis of patients with cancer.

Multi-Task Model for Esophageal Lesion Analysis Using Endoscopic Images: Classification with Image Retrieval and Segmentation with Attention.

The automatic analysis of endoscopic images to assist endoscopists in accurately identifying the types and locations of esophageal lesions remains a challenge. In this paper, we propose a novel multi-task deep learning model for automatic diagnosis, which does not simply replace the role of endoscopists in decision making, because endoscopists are expected to correct the false results predicted by the diagnosis system if more supporting information is provided. In order to help endoscopists improve the diagnosis accuracy in identifying the types of lesions, an image retrieval module is added in the classification task to provide an additional confidence level of the predicted types of esophageal lesions. In addition, a mutual attention module is added in the segmentation task to improve its performance in determining the locations of esophageal lesions. The proposed model is evaluated and compared with other deep learning models using a dataset of 1003 endoscopic images, including 290 esophageal cancer, 473 esophagitis, and 240 normal. The experimental results show the promising performance of our model with a high accuracy of 96.76% for the classification and a Dice coefficient of 82.47% for the segmentation. Consequently, the proposed multi-task deep learning model can be an effective tool to help endoscopists in judging esophageal lesions.

Role of TMPRSS4 Modulation in Breast Cancer Cell Proliferation.

Background: TMPRSS4 is a novel Type II transmembrane serine protease found at the surface of the cells and is involved in the development and cancer progression. However, TMPRSS4 functions in breast cancer remain poor understand. The present study investigated the function of TMPRSS4 in the breast cancer cells and the potential mechanistic action underling. Materials and Methods: The lentiviral vectors causing TMPRSS4 down-regulation and over-expression were established and transfected in MDA-MB-468 and MCF-7 cells, respectively. By using the CCK- 8 assay, cell proliferation was analyzed. Moreover, western blot was used to detect the expression of certain proteins related to cell apoptosis (Bax and Bcl2) signaling pathway and telomere maintenance (POT1, TPP1, and UBE2D3). Cell cycle and cell apoptosis were also analyzed by using the Flow cytometry analysis. TMPRSS4 expression was detected at the mRNA level and protein level by performing qPCR and western blot technique, respectively. Results: TMPRSS4 expression is inhibited in stable transfected MDA-MB-468-shTMPRSS4 cells compared to the control MDA-MB-468-NC and its expression is up-regulated in stable transfected MCF-7-TMPTSS4 compared to its control MCF-7-NC. Moreover, TMPRSS4 silencing in breast cancer reduces cells proliferation by promoting cell cycle arrest in G2/M phase, cell apoptosis, and telomere maintenance impairment while the TMPRSS4 overexpression increases cells proliferation through cell apoptosis reduction and telomere maintenance reinforcement associated with insignificant change in cell cycle progression. Conclusion: TMPRSS4 plays important roles in cancer progression and may be considered as a good therapeutic target for cancer gene therapy especially breast cancer.

Primary retroperitoneal liposarcoma with extensive ossification: A case report.

RATIONALE: Primary retroperitoneal liposarcoma, which originates from mesenchymal tissues, can rarely present with extensive ossification. PATIENT CONCERNS: A 41-year-old male patient presented with a chief complaint of discomfort around the waist for 2 months. DIAGNOSES: Computerized tomography (CT) and magnetic resonance imaging suggested a lesion of approximately 5.6 × 5.1 × 8.7 cm in front of the psoas major muscle, which was considered to be a mesenchymal or neurogenic tumor. INTERVENTIONS: The hard mass was removed by laparotomy, and the pathological investigation revealed that this was an atypical lipomatous tumor/well-differentiated liposarcoma, with extensive ossification. OUTCOMES: The patient was discharged from the hospital after surgery. There was no sign of reoccurrence after 1 year of follow-up. LESSONS: Retroperitoneal liposarcomas with extensive ossification are rare tumors that can present with nonspecific symptoms, and are difficult to diagnose. CT is the most common imaging technique, and surgical resection has been considered to be the most effective treatment. This rare case can be challenging for diagnosis and treatment.

ARNT/HIF-1ß links high-risk 1q21 gain and microenvironmental hypoxia to drug resistance and poor prognosis in multiple myeloma.

1q21 gain is a common cytogenetic abnormality featuring high-risk multiple myeloma (HRMM). However, the molecular mechanism underlying the adverse prognostic effect of 1q21 gain remains largely unclear. Here, we report that ARNT/HIF-1ß, a 1q21 gene, is highly expressed in HRMM and induced by microenvironmental hypoxia, which confers drug resistance and correlates with inferior outcome. Analysis of the gene expression profile database revealed that ARNT expression was upregulated in MM and increased with disease progression or in HRMM subtypes (particularly 1q21 gain), while correlated to shorter overall survival. In a cohort of 40 MM patients, qPCR further validated that ARNT expression was higher in MM patients than normal donors. MM cells carrying 1q21 gain or acquired drug resistance displayed a robust increase in HIF-1ß protein level. Hypoxia induced HIF-1ß expression via a NF-κB-dependent process. Notably, HIF-1ß overexpression impaired bortezomib sensitivity, whereas shRNA knockdown of ARNT reversed hypoxia-mediated drug resistance. Together, these findings suggest that ARNT/HIF-1ß might represent a novel biomarker for risk stratification and prognosis of HRMM patients, as well as a potential therapeutic target for overcoming 1q21 gain- or microenvironment-mediated and acquired drug resistance in MM.

Promoting tumor penetration of nanoparticles for cancer stem cell therapy by TGF-ß signaling pathway inhibition.

Cancer stem cells (CSCs), which hold a high capacity for self-renewal, play a central role in the development, metastasis, and recurrence of various malignancies. CSCs must be eradicated to cure instances of cancer; however, because they can reside far from tumor vessels, they are not easily targeted by drug agents carried by nanoparticle-based drug delivery systems. We herein demonstrate that promoting tumor penetration of nanoparticles by transforming growth factor ß (TGF-ß) signaling pathway inhibition facilitates CSC therapy. In our study, we observed that although nanoparticles carrying siRNA targeting the oncogene polo-like kinase 1 (Plk1) efficiently killed breast CSCs derived from MDA-MB-231 cells in vitro, this intervention enriched CSCs in the residual tumor tissue following systemic treatment. However, inhibition of the TGF-ß signaling pathway with LY364947, an inhibitor of TGF-ß type I receptor, promoted the penetration of nanoparticles in tumor tissue, significantly ameliorating the intratumoral distribution of nanoparticles in MDA-MB-231 xenografts and further leading to enhanced internalization of nanoparticles by CSCs. As a result, synergistic treatment with a nanoparticle drug delivery system and LY364947 inhibited tumor growth and reduced the proportion of CSCs in vivo. This study suggests that enhanced tumor penetration of drug-carrying nanoparticles can enhance CSCs clearance in vivo and consequently provide superior anti-tumor effects.

Relation between Ki-67, ER, PR, Her2/neu, p21, EGFR, and TOP II-α expression in invasive ductal breast cancer patients and correlations with prognosis.

The aim of the present study was to investigate the expression of the transcription factor Ki-67, ER, PR, Her2/neu, p21, EGFR, and TOP II-α in the tumor tissue of patients with invasive ductal carcinoma(IDC); in addition, we examined correlations between these markers. Two hundred and sixteen IDC patients, who were not previously been treated with chemo- or radiotherapy, were included in the study. All tumors were grade I-III. Expression of molecular markers was determined by immunohistochemical analysis on paraffin-embedded tissue sections. Follow-up data were collected for 3 months to 10 years and analyzed for tumor recurrence, survival time, and prognostic risk factors. We determined Ki-67 expression correlates with the expression of ER, PR, HER-2, EGFR, and TOP-α, as well as lymph node involvement, high tumor grade, lymphovascular invasion, high tumor stage, and high TNM stage in IDC. Positive Ki-67 expression was a risk factor for rapid tumor recurrence and may help tumor progression, leading to poor prognosis in IDC. Ki-67 was directly correlated with EGFR, TOP II-α, lymph node involvement, high tumor grade, lymphovascular invasion, high tumor stage, and high TNM stage in the hormone receptor subtypes of breast cancer. In triple negative breast cancer, Ki-67 correlated with TOP II-α. Expression of Ki-67 correlated with that of ER, PR, HER-2, EGFR, TOP II-α, and p21. In addition, the biomarker Ki-67 has a role as a prognostic factor and indicates a poor prognosis in IDC.

Photodynamic effect of functionalized single-walled carbon nanotubes: a potential sensitizer for photodynamic therapy.

Single-walled carbon nanotubes (SWNTs) possess unique physical and chemical properties, which make them very attractive for a wide range of applications. In particular, SWNTs and their composites have shown a great potential for photodynamic therapy (PDT). SWNTs have usually been used for photothermal therapy; herein, the photodynamic effect of two functionalized SWNTs are detected under visible light illumination in vitro and in vivo. The results indicated that the photodynamic effect is not entirely dependent on illumination time, but also on the modification method of the SWNTs. The ability of SWNTs complexes to combine with photodynamic therapy significantly improved the therapeutic efficacy of cancer treatment, and the combined treatment demonstrated a synergistic effect. These findings suggest that the SWNTs composite has great potential as sensitizer for PDT.

A tumoral acidic pH-responsive drug delivery system based on a novel photosensitizer (fullerene) for in vitro and in vivo chemo-photodynamic therapy.

Fullerene has shown great potential both in drug delivery and photodynamic therapy. Herein, we developed a doxorubicin (DOX)-loaded poly(ethyleneimine) (PEI) derivatized fullerene (C60-PEI-DOX) to facilitate combined chemotherapy and photodynamic therapy in one system, and DOX was covalently conjugated onto C60-PEI by the pH-sensitive hydrazone linkage. The release profiles of DOX from C60-PEI-DOX showed a strong dependence on the environmental pH value. The biodistributions of C60-PEI-DOX were investigated by injecting CdSe/ZnS (Qds) labeled conjugates (C60-PEI-DOX/Qds) into tumor-bearing mice. C60-PEI-DOX/Qds showed a higher tumor targeting efficiency compared with Qds alone. Compared with free DOX in an in vivo murine tumor model, C60-PEI-DOX afforded higher antitumor efficacy without obvious toxic effects to normal organs owing to its good tumor targeting efficacy and the 2.4-fold greater amount of DOX released in the tumor than in the normal tissues. C60-PEI-DOX also showed high antitumor efficacy during photodynamic therapy. The ability of C60-PEI-DOX nanoparticles to combine local specific chemotherapy with external photodynamic therapy significantly improved the therapeutic efficacy of the cancer treatment, the combined treatment demonstrating a synergistic effect. These results suggest that C60-PEI-DOX may be promising for high treatment efficacy with minimal side effects in future therapy.

PEGylated fullerene/iron oxide nanocomposites for photodynamic therapy, targeted drug delivery and MR imaging.

Recently, fullerene and fullerene derivatives owning to their highly enriched physical and chemical properties have been widely explored for applications in many different fields including biomedicine. In this study, iron oxide nanoparticles (IONPs) were decorated onto the surface of fullerene (C60), and then PEGylation was performed to improve the solubility and biocompatibility of C60-IONP, obtaining a multi-functional C60-IONP-PEG nanocomposite with strong superparamagnetism and powerful photodynamic therapy capacity. Hematoporphyrin monomethyl ether (HMME), a new photodynamic anti-cancer drug, was conjugated to C60-IONP-PEG, forming a C60-IONP-PEG/HMME drug delivery system, which demonstrated an excellent magnetic targeting ability in cancer therapy. Compared with free HMME, remarkably enhanced photodynamic cancer cell killing effect using C60-IONP-PEG/HMME was realized not only in a cultured B16-F10 cells in vitro but also in an in vivo murine tumor model due to 23-fold higher HMME uptake of tumor and strong photodynamic activity of C60-IONP-PEG. Moreover, C60-IONP-PEG could be further used as a T2-contrast agent for in vivo magnetic resonance imaging. Our work showed C60-IONP-PEG/HMME had a great potential for cancer theranostic applications.

The application of hyaluronic acid-derivatized carbon nanotubes in hematoporphyrin monomethyl ether-based photodynamic therapy for in vivo and in vitro cancer treatment.

Carbon nanotubes (CNTs) have shown great potential in both photothermal therapy and drug delivery. In this study, a CNT derivative, hyaluronic acid-derivatized CNTs (HA-CNTs) with high aqueous solubility, neutral pH, and tumor-targeting activity, were synthesized and characterized, and then a new photodynamic therapy agent, hematoporphyrin monomethyl ether (HMME), was adsorbed onto the functionalized CNTs to develop HMME-HA-CNTs. Tumor growth inhibition was investigated both in vivo and in vitro by a combination of photothermal therapy and photodynamic therapy using HMME-HA-CNTs. The ability of HMME-HA-CNT nanoparticles to combine local specific photodynamic therapy with external near-infrared photothermal therapy significantly improved the therapeutic efficacy of cancer treatment. Compared with photodynamic therapy or photothermal therapy alone, the combined treatment demonstrated a synergistic effect, resulting in higher therapeutic efficacy without obvious toxic effects to normal organs. Overall, it was demonstrated that HMME-HA-CNTs could be successfully applied to photodynamic therapy and photothermal therapy simultaneously in future tumor therapy.

Photodynamic therapy of a 2-methoxyestradiol tumor-targeting drug delivery system mediated by Asn-Gly-Arg in breast cancer.

Fullerene (C60) has shown great potential in drug delivery. In this study we exploited modified fullerene (diadduct malonic acid-fullerene-Asn-Gly-Arg peptide [DMA-C60-NGR]) as an antitumor drug carrier in order to build a new tumor-targeting drug delivery system. We also investigated the synergistic enhancement of cancer therapy using photodynamic therapy (PDT) induced by DMA-C60-NGR and 2-methoxyestradiol (2ME). Cytotoxicity tests indicated that DMA-C60-NGR had no obvious toxicity, while our drug delivery system (DMA-C60-2ME-NGR) had a high inhibition effect on MCF-7 cells compared to free 2ME. The tumor-targeting drug delivery system could efficiently cross cell membranes, and illumination induced the generation of intracellular reactive oxygen species and DNA damage. Furthermore, DMA-C60-2ME-NGR with irradiation had the highest inhibition effect on MCF-7 cells compared to the other groups. DMA-C60-NGR combined with 2ME showed a good synergistic photosensitization effect for inhibiting the growth of MCF-7 cells, demonstrating that DMA-C60-2ME-NGR may be promising for high treatment efficacy with minimal side effects in future therapy.