The implementation of clinical-decision support algorithms for medical imaging faces challenges with reliability and interpretability. Here, we establish a diagnostic tool based on a deep-learning framework for the screening of patients with common treatable blinding retinal diseases. Our framework utilizes transfer learning, which trains a neural network with a fraction of the data of conventional approaches. Applying this approach to a dataset of optical coherence tomography images, we demonstrate performance comparable to that of human experts in classifying age-related macular degeneration and diabetic macular edema. We also provide a more transparent and interpretable diagnosis by highlighting the regions recognized by the neural network. We further demonstrate the general applicability of our AI system for diagnosis of pediatric pneumonia using chest X-ray images. This tool may ultimately aid in expediting the diagnosis and referral of these treatable conditions, thereby facilitating earlier treatment, resulting in improved clinical outcomes. VIDEO ABSTRACT.
An effective blood-based method for the diagnosis and prognosis of hepatocellular carcinoma (HCC) has not yet been developed. Circulating tumour DNA (ctDNA) carrying cancer-specific genetic and epigenetic aberrations may enable a noninvasive 'liquid biopsy' for diagnosis and monitoring of cancer. Here, we identified an HCC-specific methylation marker panel by comparing HCC tissue and normal blood leukocytes and showed that methylation profiles of HCC tumour DNA and matched plasma ctDNA are highly correlated. Using cfDNA samples from a large cohort of 1,098 HCC patients and 835 normal controls, we constructed a diagnostic prediction model that showed high diagnostic specificity and sensitivity (P < 0.001) and was highly correlated with tumour burden, treatment response, and stage. Additionally, we constructed a prognostic prediction model that effectively predicted prognosis and survival (P < 0.001). Together, these findings demonstrate in a large clinical cohort the utility of ctDNA methylation markers in the diagnosis, surveillance, and prognosis of HCC.
The recent success of immunotherapies has highlighted the power of leveraging the immune system in the fight against cancer. In order for most immune‐based therapies to succeed, T cell subsets with the correct tumor‐targeting specificities must be mobilized. When such specificities are lacking, providing the immune system with tumor antigen material for processing and presentation is a common strategy for stimulating antigen‐specific T cell populations. While straightforward in principle, experience has shown that manipulation of the antigen presentation process can be incredibly complex, necessitating sophisticated strategies that are difficult to translate. Herein, the design of a biomimetic nanoparticle platform is reported that can be used to directly stimulate T cells without the need for professional antigen‐presenting cells. The nanoparticles are fabricated using a cell membrane coating derived from cancer cells engineered to express a co‐stimulatory marker. Combined with the peptide epitopes naturally presented on the membrane surface, the final formulation contains the necessary signals to promote tumor antigen‐specific immune responses, priming T cells that can be used to control tumor growth. The reported approach represents an emerging strategy that can be used to develop multiantigenic, personalized cancer immunotherapies.
The rs1061170T/C variant encoding the Y402H change in complement factor H (CFH) has been identified by genome-wide association studies as being significantly associated with age-related macular degeneration (AMD). However, the precise mechanism by which this CFH variant impacts the risk of AMD remains largely unknown. Oxidative stress plays an important role in many aging diseases, including cardiovascular disease and AMD. A large amount of oxidized phospholipids (oxPLs) are generated in the eye because of sunlight exposure and high oxygen content. OxPLs bind to the retinal pigment epithelium and macrophages and strongly activate downstream inflammatory cascades. We hypothesize that CFH may impact the risk of AMD by modulating oxidative stress. Here we demonstrate that CFH binds to oxPLs. The CFH 402Y variant of the protective rs1061170 genotype binds oxPLs with a higher affinity and exhibits a stronger inhibitory effect on the binding of oxPLs to retinal pigment epithelium and macrophages. In addition, plasma from non-AMD subjects with the protective genotype has a lower level of systemic oxidative stress measured by oxPLs per apolipoprotein B (oxPLs/apoB). We also show that oxPL stimulation increases expression of genes involved in macrophage infiltration, inflammation, and neovascularization in the eye. OxPLs colocalize with CFH in drusen in the human AMD eye. Subretinal injection of oxPLs induces choroidal neovascularization in mice. In addition, we show that the CFH risk allele confers higher complement activation and cell lysis activity. Together, these findings suggest that CFH influences AMD risk by modulating oxidative stress, inflammation, and abnormal angiogenesis.
Through whole-genome sequencing of 2,230 Icelanders, we detected a rare nonsynonymous SNP (minor allele frequency = 0.55%) in the C3 gene encoding a p.Lys155Gln substitution in complement factor 3, which, following imputation into a set of Icelandic cases with age-related macular degeneration (AMD) and controls, associated with disease (odds ratio (OR) = 3.45; P = 1.1 × 10(-7)). This signal is independent of the previously reported common SNPs in C3 encoding p.Pro314Leu and p.Arg102Gly that associate with AMD. The association of p.Lys155Gln was replicated in AMD case-control samples of European ancestry with OR = 4.22 and P = 1.6 × 10(-10), resulting in OR = 3.65 and P = 8.8 × 10(-16) for all studies combined. In vitro studies have suggested that the p.Lys155Gln substitution reduces C3b binding to complement factor H, potentially creating resistance to inhibition by this factor. This resistance to inhibition in turn is predicted to result in enhanced complement activation.
Platelets exhibit distinct surface moieties responsible for modulating their adhesion to various disease‐relevant substrates involving vascular damage, immune evasion, and pathogen interactions. Such broad biointerfacing capabilities of platelets have inspired the development of platelet‐mimicking drug carriers that preferentially target drug payloads to disease sites for enhanced therapeutic efficacy. Among these carriers, platelet membrane–coated nanoparticles (denoted “PNPs”) made by cloaking synthetic substrates with the plasma membrane of platelets have emerged recently. Their “top‐down” design combines the functionalities of natural platelet membrane and the engineering flexibility of synthetic nanomaterials, which together create synergy for effective drug delivery and novel therapeutics. Herein, the recent progress of engineering PNPs with different structures for targeted drug delivery is reviewed, focusing on three areas, including targeting injured blood vessels to treat vascular diseases, targeting cancer cells for cancer treatment and detection, and targeting drug‐resistant bacteria to treat infectious diseases. Overall, current studies establish PNPs as versatile nanotherapeutics for drug targeting with strong potentials to improve the treatment of various diseases.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.