Development of the AI Pipeline for Corneal Opacity Detection.

Ophthalmological services face global inadequacies, especially in low- and middle-income countries, which are marked by a shortage of practitioners and equipment. This study employed a portable slit lamp microscope with video capabilities and cloud storage for more equitable global diagnostic resource distribution. To enhance accessibility and quality of care, this study targets corneal opacity, which is a global cause of blindness. This study has two purposes. The first is to detect corneal opacity from videos in which the anterior segment of the eye is captured. The other is to develop an AI pipeline to detect corneal opacities. First, we extracted image frames from videos and processed them using a convolutional neural network (CNN) model. Second, we manually annotated the images to extract only the corneal margins, adjusted the contrast with CLAHE, and processed them using the CNN model. Finally, we performed semantic segmentation of the cornea using annotated data. The results showed an accuracy of 0.8 for image frames and 0.96 for corneal margins. Dice and IoU achieved a score of 0.94 for semantic segmentation of the corneal margins. Although corneal opacity detection from video frames seemed challenging in the early stages of this study, manual annotation, corneal extraction, and CLAHE contrast adjustment significantly improved accuracy. The incorporation of manual annotation into the AI pipeline, through semantic segmentation, facilitated high accuracy in detecting corneal opacity.

Physiologically Based Pharmacokinetic Absorption Model for Pexidartinib to Evaluate the Impact of Meal Contents and Intake Timing on Drug Exposure.

Pexidartinib is a systemic treatment for patients with tenosynovial giant cell tumor not amenable to surgery. Oral absorption of pexidartinib is affected by food; administration with a high-fat meal (HFM) or low-fat meal (LFM) increases absorption by approximately 100% and approximately 60%, respectively, compared with the fasted state. Pexidartinib is currently dosed 250 mg orally twice daily with an LFM (approximately 11-14 g of total fat). We developed a physiologically based pharmacokinetic model to determine the impact on drug exposure of dose timing with respect to meals, meal type, and caloric content. A 15%-16% increase in plasma exposure was predicted when consuming an HFM 1 hour after dosing with an LFM, but almost no effect on pharmacokinetics was predicted when an HFM was consumed 3 hours or more before or after pexidartinib dosing with an LFM. Exposure was not significantly affected when pexidartinib was taken with a 500-kcal LFM over the range of fat (approximately 11-14 g of total fat; 20%-25% calories from fat) for an LFM. These findings on timing of pexidartinib dose with respect to meals should be considered by patients and physicians to reduce the potential for side effects.

Interobserver Reliability of Tear Break-Up Time Examination Using "Smart Eye Camera" in Indonesian Remote Area.

Introduction: Dry eye disease (DED) is a multifactorial disease of tears and ocular surface that results in symptoms of discomfort, visual disturbance, and tear film instability leading to decrease of vision, productivity and quality of life, and blindness. DED diagnosis remains difficult and underdiagnosed regarding inconsistency between subjective symptoms and clinical findings. Tear break-up time (TBUT) is an objective indicator of tear film stability in diagnostic DED. A novel smartphone attachment, namely SEC (smart eye camera), could mimic conventional slit lamp to assess TBUT and beneficial in facilitating DED diagnosis. Reliability between a non and an ophthalmologist in TBUT assessment and DED diagnosis is observed in this study. Purpose: To determine interobserver reliability of TBUT measurement for diagnosing DED using SEC. Design: This a cross-sectional analytic study involving 99 participants (198 eyes) aged 40 years who visited Pratama Gema Santi Hospital, Nusa Penida, from September 2nd to 4th, 2022, with consecutive sampling technique. Methods: Fluoresceined eyes were filmed using the SEC device and apps, continued by masked ophthalmologist and resident assessing TBUT based on the video. The primary outcome is interobserver reliability for TBUT measurement and DED diagnosis. Results: The mean age of participants was 55.22±9.78 years, 48.5% male and 51.5% female. The reliability of interobserver in assessing DED based on TBUT test is 0.78 (95% CI=0.31-1.26, P-value=0.001), and interobserver reliability in diagnosing DED based on OSDI and TBUT showed good agreement (weighted kappa=0.71). Good interobserver reliability underscores that non-ophthalmologists can diagnose DED based on TBUT video using SEC. Conclusion: SEC video has good interobserver reliability to assess TBUT for DED diagnosis. SEC can be used as one of the methods in assessing DED in limited health care facilities. The high reliability of interobserver assessment indicates that DED diagnosis using video taken with SEC may be useful for telemedicine evaluation in remote areas.

A study establishing sensitivity and accuracy of smartphone photography in ophthalmologic community outreach programs: Review of a smart eye camera.

Purpose: Diseases affecting the cornea are a major cause of corneal blindness globally. The pressing issue we are facing today is the lack of diagnostic devices in rural areas to diagnose these conditions. The aim of the study is to establish sensitivity and accuracy of smartphone photography using a smart eye camera (SEC) in ophthalmologic community outreach programs. Methods: In this pilot study, a prospective non-randomized comparative analysis of inter-observer variability of anterior segment imaging recorded using an SEC was performed. Consecutive 100 patients with corneal pathologies, who visited the cornea specialty outpatient clinic, were enrolled. They were examined with a conventional non-portable slit lamp by a cornea consultant, and the diagnoses were recorded. This was compared with the diagnoses made by two other consultants based on SEC videos of the anterior segment of the same 100 patients. The accuracy of SEC was accessed using sensitivity, specificity, PPV, and NPV. Kappa statistics was used to find the agreement between two consultants by using STATA 17.0 (Texas, USA). Results: There was agreement between the two consultants to diagnosing by using SEC. Above 90% agreements were found in all the diagnoses, which were statistically significant (P-value < 0.001). More than 90% sensitivity and a negative predictive value were found. Conclusion: SEC can be used successfully in the community outreach programs like field visits, eye camps, teleophthalmology, and community centers, where either a clinical setup is lacking or ophthalmologists are not available.

Artificial intelligence to estimate the tear film breakup time and diagnose dry eye disease.

The use of artificial intelligence (AI) in the diagnosis of dry eye disease (DED) remains limited due to the lack of standardized image formats and analysis models. To overcome these issues, we used the Smart Eye Camera (SEC), a video-recordable slit-lamp device, and collected videos of the anterior segment of the eye. This study aimed to evaluate the accuracy of the AI algorithm in estimating the tear film breakup time and apply this model for the diagnosis of DED according to the Asia Dry Eye Society (ADES) DED diagnostic criteria. Using the retrospectively corrected DED videos of 158 eyes from 79 patients, 22,172 frames were annotated by the DED specialist to label whether or not the frame had breakup. The AI algorithm was developed using the training dataset and machine learning. The DED criteria of the ADES was used to determine the diagnostic performance. The accuracy of tear film breakup time estimation was 0.789 (95% confidence interval (CI) 0.769-0.809), and the area under the receiver operating characteristic curve of this AI model was 0.877 (95% CI 0.861-0.893). The sensitivity and specificity of this AI model for the diagnosis of DED was 0.778 (95% CI 0.572-0.912) and 0.857 (95% CI 0.564-0.866), respectively. We successfully developed a novel AI-based diagnostic model for DED. Our diagnostic model has the potential to enable ophthalmology examination outside hospitals and clinics.

Phase 1 study of DS-1205c combined with gefitinib for EGFR mutation-positive non-small cell lung cancer.

BACKGROUND: Tyrosine kinase inhibitors (TKIs) are effective for the treatment of non-small cell lung cancer (NSCLC) patients with activating mutations of the epidermal growth factor receptor (EGFR), but responses are not durable as tumors develop resistance. DS-1205c is a novel, specific, orally bioavailable, small-molecule AXL receptor TKI. In preclinical studies, DS-1205c restored TKI antitumor activity in a TKI acquired-resistance EGFR-mutant NSCLC tumor xenograft model. METHODS: This first-in-human, multicenter, open-label Phase 1 study (registered at ClinicalTrials.gov: NCT03599518) primarily evaluated the safety and tolerability of combination therapy with DS-1205c and gefitinib in Japanese patients with metastatic or unresectable EGFR-mutant NSCLC and tumor progression during treatment with EGFR-TKIs. Patients (n = 20) received DS-1205c monotherapy (200-1200 mg twice daily [BID]) in a 7-day safety monitoring period before combination DS-1205c/gefitinib (250 mg once daily) in 21-day cycles. RESULTS: The observed common treatment-emergent adverse events (TEAEs) were increased aspartate aminotransferase (35%), increased alanine aminotransferase (30%), rash maculo-papular (30%), and diarrhea (25%). No serious TEAEs were reported. Plasma concentrations and pharmacokinetic parameters of DS-1205a (free form of DS-1205c) were unaffected by concomitant administration of gefitinib. No patient achieved a complete or partial response and 5 patients (25%) had stable disease. CONCLUSION: DS-1205c was generally safe and well tolerated at all dose levels, but the safety profile of ≤800 mg BID was more favorable than 1200 mg BID. The recommended dose for dose-expansion cohorts of DS-1205c in combination therapy with gefitinib was 800 mg BID.

Programmable molecular transport achieved by engineering protein motors to move on DNA nanotubes.

Intracellular transport is the basis of microscale logistics within cells and is powered by biomolecular motors. Mimicking transport for in vitro applications has been widely studied; however, the inflexibility in track design and control has hindered practical applications. Here, we developed protein-based motors that move on DNA nanotubes by combining a biomolecular motor dynein and DNA binding proteins. The new motors and DNA-based nanoarchitectures enabled us to arrange the binding sites on the track, locally control the direction of movement, and achieve multiplexed cargo transport by different motors. The integration of these technologies realized microscale cargo sorters and integrators that automatically transport molecules as programmed in DNA sequences on a branched DNA nanotube. Our system should provide a versatile, controllable platform for future applications.

[A Case of a Young Patient with Nested Variant of Urothelial Carcinoma of the Bladder].

A 26-year-old man visited our hospital with a complaint of macrohematuria. Cystoscopy revealed a nodular tumor around the right ureteral orifice. Transurethral resection of bladder tumor was performed, and the tumor was pathologically diagnosed as the nested variant of urothelial carcinoma (NVUC). Radical cystectomy and modified Studer orthotopic neobladder reconstruction were performed. The pathological stage was pT2a, pN2. The patient received 2 courses of adjuvant chemotherapy consisting of gemcitabine and cisplatin. The patient is currently free from disease at 31 months after the treatment. To our knowledge, this case report represents the youngest case of NVUC.

Pathological complete response after nivolumab therapy following angiogenesis inhibitors in a patient with metastatic renal cell carcinoma.

INTRODUCTION: Nivolumab is effective for advanced renal cell carcinoma; however, reports are limited wherein nivolumab is combined with sequential therapy of angiogenesis inhibitors and metastasectomy. CASE PRESENTATION: A 65-year-old man was diagnosed with left renal cell carcinoma of cT2aN0M1 with lung metastasis. The patient underwent nephrectomy and sequential therapy with interferon-α and angiogenesis inhibitors. Lung metastasis decreased by angiogenesis inhibitors, but new right adrenal gland metastasis appeared. Nivolumab as the fifth systemic therapy remarkably shrank the metastasis. After discontinuing nivolumab therapy, the metastasis continued to shrink. The patient underwent adrenalectomy, and pathological analysis revealed no remnant cancer cells in the specimen, confirming a pathological complete response. Twenty months postoperatively, he remains in good health without recurrence. CONCLUSION: We report a rare case with renal cell carcinoma of a pathological complete response by nivolumab after angiogenesis inhibitors.

Structural Basis for the Function of the ß-Barrel Assembly-Enhancing Protease BepA.

The ß-barrel assembly machinery (BAM) complex mediates the assembly of ß-barrel membrane proteins in the outer membrane. BepA, formerly known as YfgC, interacts with the BAM complex and functions as a protease/chaperone for the enhancement of the assembly and/or degradation of ß-barrel membrane proteins. To elucidate the molecular mechanism underlying the dual functions of BepA, its full-length three-dimensional structure is needed. Here, we report the crystal structure of full-length BepA at 2.6-Å resolution. BepA possesses an N-terminal protease domain and a C-terminal tetratricopeptide repeat domain, which interact with each other. Domain cross-linking by structure-guided introduction of disulfide bonds did not affect the activities of BepA in vivo, suggesting that the function of this protein does not involve domain rearrangement. The full-length BepA structure is compatible with the previously proposed docking model of BAM complex and tetratricopeptide repeat domain of BepA.

Remote Coupled Drastic ß-Barrel to ß-Sheet Transition of the Protein Translocation Motor.

The membrane protein SecDF, belonging to the RND superfamily, enhances protein translocation at the extracytoplasmic side using a proton gradient. Here, we report the crystal structure of SecDF in a form we named Super-membrane-facing (Super F) form, demonstrating a ß-barrel architecture instead of the previously reported ß-sheet structure. Through this structural insight and supporting results of an in vivo crosslinking experiment, we propose a remote coupling model in which a structural change of the transmembrane region drives a functional, extracytoplasmic conformational transition.

Construction of an in Vitro Gene Screening System of the E. coli EmrE Transporter Using Liposome Display.

Liposome display is a method that enables the directed evolution of membrane proteins in vitro. The method is based on the syntheses of membrane proteins using an in vitro transcription-translation system (IVTT) inside cell-sized phospholipid vesicles from a single copy of template DNA. So far, a large number of membrane proteins have been synthesized by IVTT; however, none of these proteins, except for α-hemolysin, has been tested for use in gene screening with liposome display. Here, using EmrE, a multidrug transporter from Escherichia coli,, as a model protein, we developed an in vitro screening system of the transporter gene based on its function, which was made possible by using liposome display. The screening was performed based on two functions of EmrE: substrate transport activity and membrane integration activity. Starting from a mock gene library prepared by mixing an active and an inactive gene, 10- to 35-fold enrichment of the active genes was obtained, which was in the same range as theoretically calculated values. In addition, starting from a random mutagenized gene library of wild-type EmrE, a gene pool exhibiting ethidium bromide (EtBr) transport activity higher than that of the wild-type was obtained, indicating the validity of the established screening system.

Systems Pharmacology: An opinion on how to turn the impossible into grand challenges.

A pharmacology that hits single disease-causing molecules with a single drug passively distributing to the target tissue, was almost ready. Such a pharmacology is not (going to be) effective however: a great many diseases are systems biology diseases; complex networks of some hundred thousand types of molecule, determine the functions that constitute human health, through nonlinear interactions. Malfunctions are caused by a variety of molecular failures at the same time; rarely the same variety in different individuals; in complex constellations of OR and AND logics. Few molecules cause disease single-handedly and few drugs will cure the disease all by themselves when dosed for a limited amount of time. We here discuss the implications that this discovery of the network nature of disease should have for pharmacology. We suggest ways in which pharmacokinetics, pharmacodynamics, but also systems biology and genomics may have to change so as better to deal with systems-biology diseases.

Synthesis and SAR of 1,3-thiazolyl thiophene and pyridine derivatives as potent, orally active and S1P3-sparing S1P1 agonists.

We have previously disclosed 1,2,4-oxadiazole derivative 3 as a potent S1P(3)-sparing S1P(1) agonist. Although compound 3 exhibits potent and manageable immunosuppressive efficacy in various in vivo models, recent studies have revealed that its 1,2,4-oxadiazole ring is subjected to enterobacterial decomposition. As provisions for unpredictable issues, a series of alternative compounds were synthesized on the basis of compound 3. Extensive SAR studies led to the finding of 1,3-thiazole 24c with the EC(50) value of 3.4 nM for human S1P(1), and over 5800-fold selectivity against S1P(3). In rat on host versus graft reaction (HvGR), the ID(50) value of 24c was determined at 0.07 mg/kg. The pharmacokinetics in rat and monkey is also reported. Compared to compound 3, 24c showed excellent stability against enterobacteria.

Prediction of human plasma concentration-time profiles of monoclonal antibodies from monkey data by a species-invariant time method.

We have previously reported that human total body clearance (CL) and steady-state volume of distribution (Vss) of monoclonal antibodies (mAbs) could be predicted reasonably well from monkey data alone using simple allometry with scaling exponents of 0.79 and 1.12 (for soluble targets), and 0.96 and 1.00 (for membrane-bound targets). In the present study, to predict the plasma concentration-time profiles of mAbs in humans, we employed simple dose-normalization and species-invariant time methods (elementary Dedrick plot and complex Dedrick plot), based on the monkey data and the scaling exponents we previously determined. The results demonstrated that the species-invariant time methods were able to provide higher accuracy of prediction than simple dose-normalization, regardless of the type of target antigens (soluble or membrane-bound). The accuracy between elementary Dedrick plot and complex Dedrick plot was nearly equivalent. The predicted human CL and Vss using species-invariant time methods were within mostly 2-fold differences from the observed values. The prediction not only of pharmacokinetic (PK) parameters but also of the plasma concentration-time profile in humans can serve as guidelines for better planning of clinical studies on mAbs.

Combination of GSH trapping and time-dependent inhibition assays as a predictive method of drugs generating highly reactive metabolites.

Covalent binding (CB) of reactive metabolites (RMs) is potentially involved in severe adverse drug reactions. Because the CB assay is of low throughput and costly, a qualitative trapping assay using agents such as [(35)S]GSH is often performed in the early stages of drug discovery. However, trapping methods alone cannot replace the CB assay. We hypothesized that the time-dependent inhibition (TDI) assay might be complementary to the [(35)S]GSH trapping assay in detecting RMs. We performed CB assays, [(35)S]GSH trapping assays, and TDI assays for 42 structurally diverse compounds. First, we showed that the [(35)S]GSH trapping assay alone does not correlate with the extent of CB. Four compounds that the [(35)S]GSH trapping assay failed to detect but that showed high extent of CB were inactivators of the enzyme in the TDI assay. There was a tendency for compounds judged as positive in the TDI assay to show a high degree of CB irrespective of the result of the [(35)S]GSH trapping assay. Finally, to combine parameters from the two assays, we introduced intrinsic clearance to describe the formation of RMs (CL(int, RMs)). The Spearman rank correlation coefficient between the extent of CB and CL(int, RMs) was 0.77 (p < 0.0001), which was better than that for the formation rates of [(35)S]GSH adducts. Therefore, we demonstrated that a combination of the [(35)S]GSH trapping and TDI assays is an effective method for detecting compounds potentially capable of generating highly reactive metabolites in the early stages of drug discovery.

A zone classification system for risk assessment of idiosyncratic drug toxicity using daily dose and covalent binding.

The risk of idiosyncratic drug toxicity (IDT) is of great concern to the pharmaceutical industry. Current hypotheses based on retrospective studies suggest that the occurrence of IDT is related to covalent binding and daily dose. We determined the covalent binding of 42 radiolabeled drugs in three test systems (human liver microsomes and hepatocytes in vitro and rat liver in vivo) to assess the risk of IDT. On the basis of safety profiles given in official documentation, tested drugs were classified into the safety categories of safe, warning, black box warning, and withdrawn. The covalent binding in each of the three test systems did not distinguish the safety categories clearly. However, when the log-normalized covalent binding was plotted against the log-normalized daily dose, the distribution of the plot in the safety categories became clear. An ordinal logistic regression analysis indicated that both covalent binding and daily dose were significantly correlated with safety category and that covalent binding in hepatocytes was the best predictor among the three systems. When two separation lines were drawn on the correlation graph between covalent binding in human hepatocytes and daily dose by a regression analysis to create three zones, 30 of 37 tested drugs were located in zones corresponding to their respective classified safety categories. In conclusion, we established a zone classification system using covalent binding in human hepatocytes and daily dose for the risk assessment of IDTs.

Covalent binding and tissue distribution/retention assessment of drugs associated with idiosyncratic drug toxicity.

Bioactivation of a drug to a reactive metabolite and its covalent binding to cellular macromolecules is believed to be involved in clinical adverse events, including idiosyncratic drug toxicities (IDTs). For the interpretation of the covalent binding data in terms of risk assessment, the in vitro and in vivo covalent binding data of a variety of drugs associated with IDTs or not were determined. Most of the "problematic" drugs, including "withdrawn" and "warning" drugs, exhibit higher human liver microsome (HLM) in vitro covalent binding yields than the "safe" drugs. Although some of the problematic drugs that are known to undergo bioactivation other than cytochrome P450-mediated oxidation exhibited only trace levels of HLM covalent binding like safe drugs, a rat in vivo covalent binding study could assess the bioactivation of such drugs. Furthermore, the tissue distribution/retention of the drugs was also examined by rat autoradiography (ARG). The residual radioactivity in the liver observed at 72 or 168 h postdose was found to be well correlated with the rat in vivo covalent binding to liver proteins; thus, the in vivo covalent binding yields of the drugs could be extrapolated from the retention profiles observed by means of ARG. Long-term retention of radioactivity in the bone marrow was observed with some drugs associated with severe agranulocytosis, suggesting a spatial relationship between the toxicity profile and drug distribution/retention. Taken together, the covalent binding and tissue distribution/retention data of the various marketed drugs obtained in the present study should be quite informative for the interpretation of data in terms of risk assessment.

Role of mitochondrial permeability transition in diclofenac-induced hepatocyte injury in rats.

Hepatotoxicity of diclofenac has been known in experimental animals and humans but its mechanism has not been fully understood. The present study examined the role of mitochondrial permeability transition (MPT) in the pathogenesis of diclofenac-induced hepatocyte injury by using isolated mitochondria and primary culture hepatocytes from rats. Incubation of energized mitochondria with succinate in the presence of Ca(2+) and diclofenac resulted in mitochondrial swelling, leakage of accumulated Ca(2+), membrane depolarization, and oxidation of nicotinamide adenine dinucleotide phosphate and protein thiol. All of these phenomena were suppressed by coincubation of the mitochondria with cyclosporin A, a typical inhibitor of MPT, showing that diclofenac opened the MPT pore. It was also suggested that reactive oxygen species probably generated during mitochondrial respiration and/or voltage-dependent mechanism was involved in MPT, which are proposed as mechanisms of MPT by uncouplers of mitochondrial oxidative phosphorylation. Culture of hepatocytes for 24 hours with diclofenac caused a decrease in cellular ATP, leakage of lactate dehydrogenase and membrane depolarization. The hepatocyte toxicity thus observed was attenuated by coincubation of the hepatocytes with cyclosporin A and verapamil, a Ca(2+) channel blocker. In conclusion, these results showed the important role of MPT in pathogenesis of hepatocyte injury induced by diclofenac and its possible contribution to human idiosyncratic hepatotoxicity.