Development of non-bias phenotypic drug screening for cardiomyocyte hypertrophy by image segmentation using deep learning.

The number of patients with heart failure and related deaths is rapidly increasing worldwide, making it a major problem. Cardiac hypertrophy is a crucial preliminary step in heart failure, but its treatment has not yet been fully successful. In this study, we established a system to evaluate cardiomyocyte hypertrophy using a deep learning-based high-throughput screening system and identified drugs that inhibit it. First, primary cultured cardiomyocytes from neonatal rats were stimulated by both angiotensin II and endothelin-1, and cellular images were captured using a phase-contrast microscope. Subsequently, we used a deep learning model for instance segmentation and established a system to automatically and unbiasedly evaluate the cardiomyocyte size and perimeter. Using this system, we screened 100 FDA-approved drugs library and identified 12 drugs that inhibited cardiomyocyte hypertrophy. We focused on ezetimibe, a cholesterol absorption inhibitor, that inhibited cardiomyocyte hypertrophy in a dose-dependent manner in vitro. Additionally, ezetimibe improved the cardiac dysfunction induced by pressure overload in mice. These results suggest that the deep learning-based system is useful for the evaluation of cardiomyocyte hypertrophy and drug screening, leading to the development of new treatments for heart failure.

Increasing mixed venous oxygen saturation is a predictor of improved renal function after balloon pulmonary angioplasty in patients with chronic thromboembolic pulmonary hypertension.

Balloon pulmonary angioplasty (BPA) has emerged as an effective treatment for patients with inoperable chronic thromboembolic pulmonary hypertension (CTEPH). Renal function has been identified as a prognostic marker in patients with pulmonary hypertension in previous studies. We, therefore, aimed to investigate the clinical parameters associated with improvements in renal function in patients with CTEPH. A total of 45 consecutive patients with inoperable CTEPH undergoing BPA (mean age 62.2 ± 15.1 years) were included in the study. We evaluated the patients' clinical characteristics at baseline and at 1-year post-BPA, and investigated the association between renal function and hemodynamic parameters, including right heart function. Hemodynamics and renal function showed sustained improvements at 1 year after BPA in 64.4% of patients. Improved estimated glomerular filtration rate (eGFR) was significantly correlated with increased cardiac index (r = 0.433, p = 0.003) and mixed venous oxygen saturation (SvO2; r = 0.459, p = 0.002), and with decreased mean pulmonary arterial pressure (r = - 0.420, p = 0.004) and pulmonary vascular resistance (r = -- 0.465, p = 0.001). Multivariate analysis revealed that an increase in SvO2 immediately after the final BPA was associated with improved eGFR after the 1st year (odds ratio 1.041; 95% confidence interval 1.004-1.078; P = 0.027). The cut-off value for predicting improved eGFR was an increase in SvO2 after the final BPA of >125.4% over the baseline value (specificity 100%, sensitivity 24.1%). In conclusion, BPA improved symptoms, right heart function, hemodynamics, and renal function up to the chronic phase. Increasing SvO2 by >125.4% above baseline in the acute phase is important for improving renal function at 1 year after BPA in CTEPH patients.

A deep learning-based automated diagnosis system for SPECT myocardial perfusion imaging.

Images obtained from single-photon emission computed tomography for myocardial perfusion imaging (MPI SPECT) contain noises and artifacts, making cardiovascular disease diagnosis difficult. We developed a deep learning-based diagnosis support system using MPI SPECT images. Single-center datasets of MPI SPECT images (n = 5443) were obtained and labeled as healthy or coronary artery disease based on diagnosis reports. Three axes of four-dimensional datasets, resting, and stress conditions of three-dimensional reconstruction data, were reconstructed, and an AI model was trained to classify them. The trained convolutional neural network showed high performance [area under the curve (AUC) of the ROC curve: approximately 0.91; area under the recall precision curve: 0.87]. Additionally, using unsupervised learning and the Grad-CAM method, diseased lesions were successfully visualized. The AI-based automated diagnosis system had the highest performance (88%), followed by cardiologists with AI-guided diagnosis (80%) and cardiologists alone (65%). Furthermore, diagnosis time was shorter for AI-guided diagnosis (12 min) than for cardiologists alone (31 min). Our high-quality deep learning-based diagnosis support system may benefit cardiologists by improving diagnostic accuracy and reducing working hours.

The complement C3-complement factor D-C3a receptor signalling axis regulates cardiac remodelling in right ventricular failure.

Failure of the right ventricle plays a critical role in any type of heart failure. However, the mechanism remains unclear, and there is no specific therapy. Here, we show that the right ventricle predominantly expresses alternative complement pathway-related genes, including Cfd and C3aR1. Complement 3 (C3)-knockout attenuates right ventricular dysfunction and fibrosis in a mouse model of right ventricular failure. C3a is produced from C3 by the C3 convertase complex, which includes the essential component complement factor D (Cfd). Cfd-knockout mice also show attenuation of right ventricular failure. Moreover, the plasma concentration of CFD correlates with the severity of right ventricular failure in patients with chronic right ventricular failure. A C3a receptor (C3aR) antagonist dramatically improves right ventricular dysfunction in mice. In summary, we demonstrate the crucial role of the C3-Cfd-C3aR axis in right ventricular failure and highlight potential therapeutic targets for right ventricular failure.

Anti-senescent drug screening by deep learning-based morphology senescence scoring.

Advances in deep learning technology have enabled complex task solutions. The accuracy of image classification tasks has improved owing to the establishment of convolutional neural networks (CNN). Cellular senescence is a hallmark of ageing and is important for the pathogenesis of ageing-related diseases. Furthermore, it is a potential therapeutic target. Specific molecular markers are used to identify senescent cells. Moreover senescent cells show unique morphology, which can be identified. We develop a successful morphology-based CNN system to identify senescent cells and a quantitative scoring system to evaluate the state of endothelial cells by senescence probability output from pre-trained CNN optimised for the classification of cellular senescence, Deep Learning-Based Senescence Scoring System by Morphology (Deep-SeSMo). Deep-SeSMo correctly evaluates the effects of well-known anti-senescent reagents. We screen for drugs that control cellular senescence using a kinase inhibitor library by Deep-SeSMo-based drug screening and identify four anti-senescent drugs. RNA sequence analysis reveals that these compounds commonly suppress senescent phenotypes through inhibition of the inflammatory response pathway. Thus, morphology-based CNN system can be a powerful tool for anti-senescent drug screening.

Beta blockers versus calcium channel blockers for provocation of vasospastic angina after drug-eluting stent implantation: a multicentre prospective randomised trial.

BACKGROUND: Drug-eluting stent-induced vasospastic angina (DES-VSA) has emerged as a novel complication in the modern era of percutaneous coronary intervention (PCI). Although beta blockers (BBs) are generally recommended for coronary heart disease, they may promote incidence of DES-VSA. This study aimed to compare the effects of calcium channel blockers (CCBs) perceived to be protective against DES-VSA and BBs on subsequent coronary events after second-generation drug-eluting stent implantation. METHODS: In this multicentre prospective, randomised study, 52 patients with coronary artery disease who underwent PCI for a single-vessel lesion with everolimus-eluting stent placement were randomised into post-stenting BB (N=26) and CCB (N=26) groups and followed for 24 months to detect any major cardiovascular events (MACE). A positive result on acetylcholine provocation testing during diagnostic coronary angiography (CAG) at 9 months was the primary endpoint for equivalence. MACE included all-cause death, non-fatal myocardial infarction, unstable angina, cerebrovascular disease or coronary revascularisation for stable coronary artery disease after index PCI. RESULTS: At 9 months, 42 patients (80.8%) underwent diagnostic coronary angiography and acetylcholine provocation testing. Among them, seven patients in each group were diagnosed with definite vasospasm (intention-to-treat analysis 26.9% vs 26.9%, risk difference 0 (-0.241, 0.241)). Meanwhile, the secondary endpoint, 24-month MACE, was higher in the CCB group (19.2%) than in the BB group (3.8%) (p=0.01). In detail, coronary revascularisation for stable coronary artery disease was the predominant endpoint that contributed to the greater proportion of MACE in the CCB group (CCB (19.2%) vs BB (3.8%), p=0.03). CONCLUSIONS: The incidence of acetylcholine-induced coronary artery spasms did not differ between patients receiving BBs or CCBs at 9 months after PCI. However, a higher incidence of 2-year MACE was observed in the CCB group, suggesting the importance of BB administration. TRIAL REGISTRATION NUMBER: This study was registered at the Japanese University Hospital Medical Information Network (UMIN) Clinical Trial Registry (The Prospective Randomized Trial for Optimizing Medical Therapy After Stenting: Calcium-Beta Trial; UMIN000008321, https://upload.umin.ac.jp/cgi-open-bin/ctr_e/ctr_view.cgi?recptno=R000009536).

Benefit and harm of intravenous vasodilators across the clinical profile spectrum in acute cardiogenic pulmonary oedema patients.

BACKGROUND: The absence of high quality, large-scale data that indicates definitive mortality benefits does not allow for firm conclusions on the role of intravenous vasodilators in acute heart failure. We aimed to investigate the associations between intravenous vasodilators and clinical outcomes in acute heart failure patients, with a specific focus on patient profiles and type of vasodilators. METHODS: Data of 26,212 consecutive patients urgently hospitalised for a primary diagnosis of acute heart failure between 2009 and 2015 were extracted from a government-funded multicentre data registration system. Propensity scores were calculated with multiple imputations and 1:1 matching performed between patients with and without vasodilator use. The primary endpoint was inhospital mortality. RESULTS: On direct comparison of the vasodilator and non-vasodilator groups after propensity score matching, there were no significant differences in the inhospital mortality rates (7.5% vs. 8.8%, respectively; P=0.098) or length of intensive/cardiovascular care unit stay and hospital stay between the two groups. However, there was a substantial difference in baseline systolic blood pressure by vasodilator type; favourable impacts of vasodilator use on inhospital mortality were observed among patients who had higher systolic blood pressures and those who had no atrial fibrillation on admission. Furthermore, when compared to nitrates, the use of carperitide (natriuretic peptide agent) was significantly associated with worse outcomes, especially in patients with intermediate systolic blood pressures. CONCLUSIONS: In acute heart failure patients, vasodilator use was not universally associated with improved inhospital outcomes; rather, its effect depended on individual clinical presentation: patients with higher systolic blood pressure and no atrial fibrillation seemed to benefit maximally from vasodilators. TRIAL REGISTRATION: UMIN-CTR identifier, UMIN000013128.

Impact of tacrolimus versus cyclosporin A on renal function during the first year after heart transplant.

AIMS: Nephrotoxicity of calcineurin inhibitors (CNIs) is associated with adverse events in patients undergoing heart transplant (HTx), although studies directly comparing tacrolimus (TAC) versus cyclosporin A (CsA), especially in combination with everolimus and low-dose CNIs approach, are limited. Thus, we sought to investigate the associations of TAC and CsA with clinical outcomes in HTx recipients, with specific focus on renal function. METHODS AND RESULTS: From August 2007 to February 2017, 72 consecutive patients (39 treated with TAC vs. 33 with CsA) receiving de novo HTx in a single transplant centre were retrospectively evaluated. We used the instrumental variable method to account for unmeasured confounding. The study outcomes were percentage change in estimated glomerular filtration rates (eGFR) (safety endpoint) and biopsy-proven acute rejection (efficacy endpoint) within the first year after HTx. The enrolled patients (median age 40 years) were predominantly men (68%). There were no significant differences in baseline characteristics, including eGFR (64.8 [45.7-96.4] mL/min/1.73 m2 in TAC vs. 65.6 [57.9-83.0] mL/min/1.73 m2 for CsA; P = 0.48), other than sex (male, 49% for TAC vs. 91% for CsA; P < 0.001) between the two groups. Within the first year after HTx, 23 (59%) in the TAC group switched mycophenolate mofetil to everolimus, whereas 16 (48%) in the CsA group (P = 0.52). At 12 months, the rates of mortality and end-stage renal disease requiring renal replacement therapies were both 0%. In the instrumental variable analysis, no differences in renal function as well as graft rejection for 1 year after HTx existed between the TAC and CsA groups. These results were similar when taking into account of everolimus use. CONCLUSIONS: Irrespective of everolimus use with low-dose CNIs, our analysis using the instrumental variable method showed no differences in renal function as well as graft rejection during the first year after HTx between HTx recipients who received TAC or CsA.

Association of renin-angiotensin system inhibitors with long-term outcomes in patients with systolic heart failure and moderate-to-severe kidney function impairment.

PURPOSE: Although guidelines recommend that patients with heart failure with reduced ejection fraction (HFrEF) should be treated with renin-angiotensin system (RAS) inhibitors, the long-term efficacy of RAS inhibitors in HFrEF patients with moderate-to-severe chronic kidney disease (CKD) remains unclear. METHODS: The present study included consecutive patients hospitalized for acute heart failure across five Japanese teaching hospitals. The impact of RAS inhibitors on 2-year all-cause mortality was evaluated in patients with an ejection fraction ≤40% and CKD, defined as an estimated glomerular filtration rate (eGFR) <45 mL/min/1.73 m2, at discharge. Its severity was subclassified from 3B to 5 according to eGFR. RESULTS: Overall, 553 patients (age, 76 ±â€¯11 years; 68% male) were included. RAS inhibitors were prescribed more frequently in 227 patients with stage 3B (71.2%) than in 107 patients with stage 4 or 5 CKD (45.7%). All-cause mortality was recorded in 119 patients (23.4%) (55 [18.5%] patients with stage 3B; 64 [30.3%] patients with stage 4 or 5 CKD), within the median follow-up period of 609 (220-983) days. After many-to-one propensity score matching (87 pairs in stage 3; 60 pairs in stage 4 or 5 CKD), those with RAS inhibitors had reduced mortality rate in stage 3B (hazard ratio [HR], 0.39; 95% confidence interval [CI], 0.19-0.83) but not in stage 4 or 5 CKD (HR, 1.08; 95% CI, 0.57-2.03). CONCLUSIONS: In HFrEF patients with CKD, RAS inhibitors are associated with reduction in mortality in stage 3B CKD, but the association is less clear in stage 4 or 5 CKD.

Effects of body habitus on contrast-induced acute kidney injury after percutaneous coronary intervention.

BACKGROUND: Limiting the contrast volume to creatinine clearance (V/CrCl) ratio is crucial for preventing contrast-induced acute kidney injury (CI-AKI) after percutaneous coronary intervention (PCI). However, the incidence of CI-AKI and the distribution of V/CrCl ratios may vary according to patient body habitus. OBJECTIVE: We aimed to identify the clinical factors predicting CI-AKI in patients with different body mass indexes (BMIs). METHODS: We evaluated 8782 consecutive patients undergoing PCI and who were registered in a large Japanese database. CI-AKI was defined as an absolute serum creatinine increase of 0.3 mg/dL or a relative increase of 50%. The effect of the V/CrCl ratio relative to CI-AKI incidence was evaluated within the low- (≤25 kg/m2) and high- (>25 kg/m2) BMI groups, with a V/CrCl ratio > 3 considered to be a risk factor for CI-AKI. RESULTS: A V/CrCl ratio > 3 was predictive of CI-AKI, regardless of BMI (low-BMI group: odds ratio [OR], 1.77 [1.42-2.21]; P < 0.001; high-BMI group: OR, 1.67 [1.22-2.29]; P = 0.001). The relationship between BMI and CI-AKI followed a reverse J-curve relationship, although baseline renal dysfunction (creatinine clearance <60 mL/min, 46.9% vs. 21.5%) and V/CrCl ratio > 3 (37.3% vs. 20.4%) were predominant in the low-BMI group. Indeed, low BMI was a significant predictor of a V/CrCl ratio > 3 (OR per unit decrease in BMI, 1.08 [1.05-1.10]; P < 0.001). CONCLUSIONS: A V/CrCl ratio > 3 was strongly associated with the occurrence of CI-AKI. Importantly, we also identified a tendency for physicians to use higher V/CrCl ratios in lean patients. Thus, recognizing this trend may provide a therapeutic target for reducing the incidence of CI-AKI.

Automated Deep Learning-Based System to Identify Endothelial Cells Derived from Induced Pluripotent Stem Cells.

Deep learning technology is rapidly advancing and is now used to solve complex problems. Here, we used deep learning in convolutional neural networks to establish an automated method to identify endothelial cells derived from induced pluripotent stem cells (iPSCs), without the need for immunostaining or lineage tracing. Networks were trained to predict whether phase-contrast images contain endothelial cells based on morphology only. Predictions were validated by comparison to immunofluorescence staining for CD31, a marker of endothelial cells. Method parameters were then automatically and iteratively optimized to increase prediction accuracy. We found that prediction accuracy was correlated with network depth and pixel size of images to be analyzed. Finally, K-fold cross-validation confirmed that optimized convolutional neural networks can identify endothelial cells with high performance, based only on morphology.