Latent Diffusion Models with Image-Derived Annotations for Enhanced AI-Assisted Cancer Diagnosis in Histopathology.

Artificial Intelligence (AI)-based image analysis has immense potential to support diagnostic histopathology, including cancer diagnostics. However, developing supervised AI methods requires large-scale annotated datasets. A potentially powerful solution is to augment training data with synthetic data. Latent diffusion models, which can generate high-quality, diverse synthetic images, are promising. However, the most common implementations rely on detailed textual descriptions, which are not generally available in this domain. This work proposes a method that constructs structured textual prompts from automatically extracted image features. We experiment with the PCam dataset, composed of tissue patches only loosely annotated as healthy or cancerous. We show that including image-derived features in the prompt, as opposed to only healthy and cancerous labels, improves the Fréchet Inception Distance (FID) by 88.6. We also show that pathologists find it challenging to detect synthetic images, with a median sensitivity/specificity of 0.55/0.55. Finally, we show that synthetic data effectively train AI models.

Biodegradable core-multishell nanocarrier: Topical tacrolimus delivery for treatment of dermatitis.

Two challenges in topical drug delivery to the skin include solubilizing hydrophobic drugs in water-based formulations and increasing drug penetration into the skin. Polymeric core-multishell nanocarrier (CMS), particularly the novel biodegradable CMS (bCMS = hPG-PCL1.1K-mPEG2k-CMS) have shown both advantages on excised skin ex vivo. Here, we investigated topical delivery of tacrolimus (TAC; > 500 g/mol) by bCMS in a hydrogel on an oxazolone-induced model of dermatitis in vivo. As expected, bCMS successfully delivered TAC into the skin. However, in vivo they did not increase, but decrease TAC penetration through the stratum corneum compared to ointment. Differences in the resulting mean concentrations were mostly non-significant in the skin (epidermis: 35.7 ± 20.9 ng/cm2 for bCMS vs. 92.6 ± 62.7 ng/cm2 for ointment; dermis: 76.8 ± 26.8 ng/cm2vs 118.2 ± 50.4 ng/cm2), but highly significant in blood (plasma: 1.1 ± 0.4 ng/ml vs 11.3 ± 9.3 ng/ml; erythrocytes: 0.5 ± 0.2 ng/ml vs 3.4 ± 2.4 ng/ml) and liver (0.01 ± 0.01 ng/mg vs 0.03 ± 0.01 ng/mg). bCMS were detected in the stratum corneum but not in viable skin or beyond. The therapeutic efficacy of TAC delivered by bCMS was equivalent to that of standard TAC ointment. Our results suggest that bCMS may be a promising carrier for the topical delivery of TAC. The quantitative difference to previous results should be interpreted in light of structural differences between murine and human skin, but highlights the need as well as potential methods to develop more a complex ex vivo analysis on human skin to ensure quantitative predictive value.

Visualizing Oxidative Cellular Stress Induced by Nanoparticles in the Subcytotoxic Range Using Fluorescence Lifetime Imaging.

Nanoparticles hold a great promise in biomedical science. However, due to their unique physical and chemical properties they can lead to overproduction of intracellular reactive oxygen species (ROS). As an important mechanism of nanotoxicity, there is a great need for sensitive and high-throughput adaptable single-cell ROS detection methods. Here, fluorescence lifetime imaging microscopy (FLIM) is employed for single-cell ROS detection (FLIM-ROX) providing increased sensitivity and enabling high-throughput analysis in fixed and live cells. FLIM-ROX owes its sensitivity to the discrimination of autofluorescence from the unique fluorescence lifetime of the ROS reporter dye. The effect of subcytotoxic amounts of cationic gold nanoparticles in J774A.1 cells and primary human macrophages on ROS generation is investigated. FLIM-ROX measures very low ROS levels upon gold nanoparticle exposure, which is undetectable by the conventional method. It is demonstrated that cellular morphology changes, elevated senescence, and DNA damage link the resulting low-level oxidative stress to cellular adverse effects and thus nanotoxicity. Multiphoton FLIM-ROX enables the quantification of spatial ROS distribution in vivo, which is shown for skin tissue as a target for nanoparticle exposure. Thus, this innovative method allows identifying of low-level ROS in vitro and in vivo and, subsequently, promotes understanding of ROS-associated nanotoxicity.

Dendritic Core-Multishell Nanocarriers in Murine Models of Healthy and Atopic Skin.

Dendritic hPG-amid-C18-mPEG core-multishell nanocarriers (CMS) represent a novel class of unimolecular micelles that hold great potential as drug transporters, e.g., to facilitate topical therapy in skin diseases. Atopic dermatitis is among the most common inflammatory skin disorders with complex barrier alterations which may affect the efficacy of topical treatment.Here, we tested the penetration behavior and identified target structures of unloaded CMS after topical administration in healthy mice and in mice with oxazolone-induced atopic dermatitis. We further examined whole body distribution and possible systemic side effects after simulating high dosage dermal penetration by subcutaneous injection.Following topical administration, CMS accumulated in the stratum corneum without penetration into deeper viable epidermal layers. The same was observed in atopic dermatitis mice, indicating that barrier alterations in atopic dermatitis had no influence on the penetration of CMS. Following subcutaneous injection, CMS were deposited in the regional lymph nodes as well as in liver, spleen, lung, and kidney. However, in vitro toxicity tests, clinical data, and morphometry-assisted histopathological analyses yielded no evidence of any toxic or otherwise adverse local or systemic effects of CMS, nor did they affect the severity or course of atopic dermatitis.Taken together, CMS accumulate in the stratum corneum in both healthy and inflammatory skin and appear to be highly biocompatible in the mouse even under conditions of atopic dermatitis and thus could potentially serve to create a depot for anti-inflammatory drugs in the skin.

Stratum corneum targeting by dendritic core-multishell-nanocarriers in a mouse model of psoriasis.

Inflammatory disorders of the skin pose particular therapeutic challenges due to complex structural and functional alterations of the skin barrier. Penetration of several anti-inflammatory drugs is particularly problematic in psoriasis, a common dermatitis condition with epidermal hyperplasia and hyperkeratosis. Here, we tested in vivo dermal penetration and biological effects of dendritic core-multishell-nanocarriers (CMS) in a murine skin model of psoriasis and compared it to healthy skin. In both groups, CMS exclusively localized to the stratum corneum of the epidermis with only very sporadic uptake by Langerhans cells. Furthermore, penetration into the viable epidermis of nile red as a model for lipophilic compounds was enhanced by CMS. CMS proved fully biocompatible in several in vitro assays and on normal and psoriatic mouse skin. The observations support the concept of CMS as promising candidates for drug delivery in inflammatory hyperkeratotic skin disorders in vivo.

Influence of Th2 Cytokines on the Cornified Envelope, Tight Junction Proteins, and ß-Defensins in Filaggrin-Deficient Skin Equivalents.

Atopic dermatitis is a chronic skin condition with complex etiology. It is characterized by skin barrier defects and T helper type 2 (Th2)-polarized inflammation. Although mutations in the filaggrin gene are known to be prominent genetic risk factors for the development of atopic dermatitis, the interdependency between these and an altered cytokine milieu is not fully understood. In this study, we evaluated the direct effects of filaggrin deficiency on the cornified envelope, tight junction proteins, and innate immune response, and report the effects of Th2 cytokines in normal and filaggrin-deficient skin equivalents. Supplementation with IL-4 and IL-13 led to distinct histologic changes and significantly increased skin surface pH, both of which were enhanced in filaggrin knockdown skin equivalents. We detected a compensatory up-regulation of involucrin and occludin in filaggrin-deficient skin that was dramatically disturbed when simultaneous inflammation occurred. Furthermore, we found that a lack of filaggrin triggered an up-regulation of human ?-defensin 2 via an unknown mechanism, which was abolished by Th2 cytokine supplementation. Taken together, these results indicate that defects in the epidermal barrier, skin permeability, and cutaneous innate immune response are not primarily linked to filaggrin deficiency but are rather secondarily induced by Th2 inflammation.

Tracking CNS and systemic sources of oxidative stress during the course of chronic neuroinflammation.

The functional dynamics and cellular sources of oxidative stress are central to understanding MS pathogenesis but remain elusive, due to the lack of appropriate detection methods. Here we employ NAD(P)H fluorescence lifetime imaging to detect functional NADPH oxidases (NOX enzymes) in vivo to identify inflammatory monocytes, activated microglia, and astrocytes expressing NOX1 as major cellular sources of oxidative stress in the central nervous system of mice affected by experimental autoimmune encephalomyelitis (EAE). This directly affects neuronal function in vivo, indicated by sustained elevated neuronal calcium. The systemic involvement of oxidative stress is mirrored by overactivation of NOX enzymes in peripheral CD11b(+) cells in later phases of both MS and EAE. This effect is antagonized by systemic intake of the NOX inhibitor and anti-oxidant epigallocatechin-3-gallate. Together, this persistent hyper-activation of oxidative enzymes suggests an "oxidative stress memory" both in the periphery and CNS compartments, in chronic neuroinflammation.