Deep learning-based phenotyping reclassifies combined hepatocellular-cholangiocarcinoma.

Primary liver cancer arises either from hepatocytic or biliary lineage cells, giving rise to hepatocellular carcinoma (HCC) or intrahepatic cholangiocarcinoma (ICCA). Combined hepatocellular- cholangiocarcinomas (cHCC-CCA) exhibit equivocal or mixed features of both, causing diagnostic uncertainty and difficulty in determining proper management. Here, we perform a comprehensive deep learning-based phenotyping of multiple cohorts of patients. We show that deep learning can reproduce the diagnosis of HCC vs. CCA with a high performance. We analyze a series of 405 cHCC-CCA patients and demonstrate that the model can reclassify the tumors as HCC or ICCA, and that the predictions are consistent with clinical outcomes, genetic alterations and in situ spatial gene expression profiling. This type of approach could improve treatment decisions and ultimately clinical outcome for patients with rare and biphenotypic cancers such as cHCC-CCA.

Hyperthermia-mediated changes in the tumor immune microenvironment using iron oxide nanoparticles.

Iron oxide nanoparticles (IONPs) have often been investigated for tumor hyperthermia. IONPs act as heating foci in the presence of an alternating magnetic field (AMF). It has been shown that hyperthermia can significantly alter the tumor immune microenvironment. Typically, mild hyperthermia invokes morphological changes within the tumor, which elicits a secretion of inflammatory cytokines and tumor neoantigens. Here, we focused on the direct effect of IONP-induced hyperthermia on the various tumor-resident immune cell subpopulations. We compared direct intratumoral injection to systemic administration of IONPs followed by application of an external AMF. We used the orthotopic 4T1 mouse model, which represents aggressive and metastatic breast cancer with a highly immunosuppressive microenvironment. A non-inflamed and 'cold' microenvironment inhibits peripheral effector lymphocytes from effectively trafficking into the tumor. Using intratumoral or systemic injection, IONP-induced hyperthermia achieved a significant reduction of all the immune cell subpopulations in the tumor. However, the systemic delivery approach achieved superior outcomes, resulting in substantial reductions in the populations of both innate and adaptive immune cells. Upon depletion of the existing dysfunctional tumor-resident immune cells, subsequent treatment with clinically approved immune checkpoint inhibitors encouraged the repopulation of the tumor with 'fresh' infiltrating innate and adaptive immune cells, resulting in a significant decrease of the tumor cell population.

Differential contribution of bone marrow-derived infiltrating monocytes and resident macrophages to persistent lung inflammation in chronic air pollution exposure.

Chronic exposure to particulate matter < 2.5µ (PM2.5) has been linked to cardiopulmonary disease. Tissue-resident (TR) alveolar macrophages (AΦ) are long-lived, self-renew and critical to the health impact of inhalational insults. There is an inadequate understanding of the impact of PM2.5 exposure on the nature/time course of transcriptional responses, self-renewal of AΦ, and the contribution from bone marrow (BM) to this population. Accordingly, we exposed chimeric (CD45.2/CD45.1) mice to concentrated PM2.5 or filtered air (FA) to evaluate the impact on these end-points. PM2.5 exposure for 4-weeks induced an influx of BM-derived monocytes into the lungs with no contribution to the overall TR-AΦ pool. Chronic (32-weeks) PM2.5 exposure on the other hand while associated with increased recruitment of BM-derived monocytes and their incorporation into the AΦ population, resulted in enhanced apoptosis and decreased proliferation of TR-AΦ. RNA-seq analysis of isolated TR-AΦ and BM-AΦ from 4- and 32-weeks exposed mice revealed a unique time-dependent pattern of differentially expressed genes. PM2.5 exposure resulted in altered histological changes in the lungs, a reduced alveolar fraction which corresponded to protracted lung inflammation. Our findings suggest a time-dependent entrainment of BM-derived monocytes into the AΦ population of PM2.5 exposed mice, that together with enhanced apoptosis of TR-AΦ and reorganization of transcriptional responses, could collectively contribute to the perpetuation of chronic inflammation.

PTPmu-targeted nanoparticles label invasive pediatric and adult glioblastoma.

Poor prognosis for glioblastoma (GBM) is a consequence of the aggressive and infiltrative nature of gliomas where individual cells migrate away from the main tumor to distant sites, making complete surgical resection and treatment difficult. In this manuscript, we characterize an invasive pediatric glioma model and determine if nanoparticles linked to a peptide recognizing the GBM tumor biomarker PTPmu can specifically target both the main tumor and invasive cancer cells in adult and pediatric glioma models. Using both iron and lipid-based nanoparticles, we demonstrate by magnetic resonance imaging, optical imaging, histology, and iron quantification that PTPmu-targeted nanoparticles effectively label adult gliomas. Using PTPmu-targeted nanoparticles in a newly characterized orthotopic pediatric SJ-GBM2 model, we demonstrate individual tumor cell labeling both within the solid tumor margins and at invasive and dispersive sites.

Imaging breast cancer using a dual-ligand nanochain particle.

Nanoparticles often only exploit the upregulation of a receptor on cancer cells to enhance intratumoral deposition of therapeutic and imaging agents. However, a single targeting moiety assumes that a tumor is homogenous and static. Tumoral microenvironments are both heterogenous and dynamic, often displaying variable spatial and temporal expression of targetable receptors throughout disease progression. Here, we evaluated the in vivo performance of an iron oxide nanoparticle in terms of targeting and imaging of orthotropic mouse models of aggressive breast tumors. The nanoparticle, a multi-component nanochain, was comprised of 3-5 iron oxide nanoparticles chemically linked in a linear chain. The nanoparticle's surface was decorated with two types of ligands each targeting two different upregulated biomarkers on the tumor endothelium, P-selectin and fibronectin. The nanochain exhibited improved tumor deposition not only through vascular targeting but also through its elongated structure. A single-ligand nanochain exhibited a ~2.5-fold higher intratumoral deposition than a spherical nanoparticle variant. Furthermore, the dual-ligand nanochain exhibited higher consistency in generating detectable MR signals compared to a single-ligand nanochain. Using a 7T MRI, the dual-ligand nanochains exhibited highly detectable MR signal within 3h after injection in two different animal models of breast cancer.