Systems and methods for treating heart failure with preserved ejection fraction (HFpEF)

Abstract

Disclosed herein is a method for treating selected patients with **Heart Failure with Preserved Ejection Fraction (HFpEF)** using computer-assisted patient stratification. Also disclosed is a computer-implemented method for classifying patients who are likely to respond to and demonstrate clinical benefit from a Rho kinase inhibitor.

Description

[Technical Field]

[0001] Cross-reference to related applications This application claims the benefit of and benefits from U.S. Provisional Application No. 63 / 482,462, filed January 31, 2023, the entire contents of which are incorporated herein by reference. Government Support Statement

[0002] This invention was made with support from the National Institutes of Health (NIH) under grants R44HL139248 and R43HL164266. The government has certain rights in this invention. Technical Field

[0003] Described herein are systems and methods for treating a subject with heart failure with preserved ejection fraction (HFpEF). In one aspect, the present disclosure describes a method for treating HFpEF by administering to a subject a therapeutically effective amount of a Rho kinase (ROCK) inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof. Introduction

[0004] Heart failure (HF) is the leading cause of cardiovascular morbidity and mortality worldwide. Approximately half of patients with heart failure have heart failure with preserved ejection fraction (HFpEF). While conventional heart failure, or systolic heart failure (HFrEF), is characterized by impaired ventricular contractile function, patients with HFpEF exhibit impaired ventricular function during diastole, rather than systole. While blood is normally ejected from the ventricles in patients with HFpEF, the myocardium does not relax quickly enough to efficiently fill with returning blood. HFpEF is associated with similar morbidity and mortality rates to conventional heart failure (HFrEF), but therapies effective for conventional heart failure are ineffective in treating or preventing HFpEF. Treating and / or preventing HFpEF and its complications represents a significant public health challenge, yet prediction and diagnosis remain challenging due to its complex pathophysiology. summary

[0005] In one aspect, a method for treating heart failure with preserved ejection fraction (HFpEF) is disclosed, comprising administering to a subject a therapeutically effective amount of a Rho kinase inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof.

[0006] Rho kinase inhibitors include the following compounds: Fasudil, Thiazovivin, Y-27632 2HCl, Ripasudil, GSK429286A, RKI-1447, and Azaindole 1 (TC-S 7001). JPEG2026504406000053.jpg183166

[0007] In some embodiments, the Rho kinase inhibitor may be fasudil. Subjects may have diastolic dysfunction, myocardial fibrosis, pulmonary hypertension, a left ventricular ejection fraction (LVEF) of 50% or greater, and may be free of cardiac amyloid deposits.

[0008] The subjects may be humans or other animals, such as mammals.

[0009] In some embodiments, a set of instructions for administering a therapeutically effective amount of a Rho kinase inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof is generated by a computer-implemented method.

[0010] In another aspect, a computer-implemented method is disclosed. The method includes executing a forward propagation artificial intelligence (AI) algorithm with one or more processors and receiving, by the one or more processors, a plurality of subject parameters using a machine learning algorithm. The subject parameters include information regarding patients with and without diastolic dysfunction, patients with and without myocardial fibrosis, patients with and without pulmonary hypertension, patients with a left ventricular ejection fraction (LVEF) greater than or equal to 50% and less than 50%, and patients with and without cardiac amyloid deposits. The subject parameters are analyzed by the one or more processors and configured into neurons corresponding to an input layer of an artificial neural network. Weights are then applied to the input parameters and sent to one or more hidden layers of the feedforward propagation artificial neural network to generate optimized multiple subject data sets. A clustering analysis is then performed by the one or more processors. The clustering analysis includes receiving the optimized multiple subject data sets, analyzing the data sets, identifying multiple populations within the subject population, and generating one or more clusters based on the populations. The one or more processors analyze the clusters of the multiple populations generated by the clustering analysis, and generate a set of instructions for administering a therapeutically effective amount of a Rho kinase inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof to a subject with diastolic dysfunction, myocardial fibrosis, pulmonary hypertension, and a left ventricular ejection fraction (LVEF) of 50% or greater, and no cardiac amyloid deposits, and the subject is then administered the drug in accordance with the instructions.

[0011] In another aspect, a method for reducing myocardial tissue stiffness is disclosed, comprising administering a therapeutically effective amount of a Rho kinase inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof to a subject with diastolic dysfunction, myocardial fibrosis, pulmonary hypertension, and a left ventricular ejection fraction (LVEF) of 50% or greater, and no cardiac amyloid deposits.

[0012] Before proceeding to a detailed description of the embodiments disclosed below, it is to be understood that the present disclosure is not limited in its application to the specific arrangement and design details of components set forth in the following description or drawings. The present disclosure may be applied to other embodiments and may be practiced or carried out in various ways.

[0013] This patent or application contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawing(s) will be provided by the United States Patent and Trademark Office (USPTO) upon request and payment of the necessary fee. Copies of this patent or patent application with color drawing(s) will be provided by the United States Patent and Trademark Office (USPTO) upon request and payment of the necessary fee. [Brief explanation of the drawings]

[0014] [Figure 1] Figure 1 shows the chemical structures of representative Rho kinase (ROCK) inhibitors. [Figure 2] Figure 2 shows a table showing the various output model parameters. [Figure 3A] Figure 3A is a graph showing the distribution of pulmonary artery (PA) diastolic pressure in a patient. [Figure 3B] Figure 3B is a graph showing the distribution of passive left ventricular (LV) stiffness in patients. [Figure 4A] Figure 4A is a schematic diagram illustrating the development of an individualized in vitro heart failure disease model. [Figure 4B] Figure 4B is a schematic diagram illustrating the identification of heart failure patients best represented by in vivo animal models. [Figure 5] FIG. 5 is a block diagram illustrating a computing environment compatible with exemplary computer-implemented methods of the present disclosure. [Figure 6] FIG. 6 is a block diagram illustrating a computer system and a server area network (SAN) that supports exemplary computer-implemented methods of the present disclosure. [Figure 7A] Figure 7A shows a graphical representation of the length-tension relationship during myocardial stress in human microcardiac tissue treated with fasudil. [Figure 7B] Figure 7B shows a graph of the length-tension relationship for diastolic stiffness of human microcardiac tissue treated with fasudil. [Figure 8A] Figure 8A shows a table of gene data showing the changes in the expression profiles of various genes due to the deletion of ROCK1 and ROCK2. The deletion of ROCK2 suppressed the expression of genes involved in fibrosis. [Figure 8B] Figure 8B shows Western blot analysis of procollagen and α-smooth muscle actin (α-SM actin) expression. [Figure 8C] Figure 8C is a bar graph showing that α-smooth muscle actin expression indicates the presence of myofibroblasts in fibrotic tissue. Two types of short hairpin RNA targeting ROCK2 (shRNA: shROCK2_3 and shROCK2_5) suppressed α-SM actin expression. [Figure 8D] Figure 8D is a bar graph showing that inhibiting ROCK2 expression suppresses procollagen I protein expression. [Figure 9A] Figure 9A shows a schematic diagram of the automated generation of cardiomyocytes from induced pluripotent stem cells (iPSCs). [Figure 9B] Figure 9B shows an image of the microheart tissue (NuHeart®) formed in a 96-well plate, observed from the top on Day -1. [Figure 9C] Figure 9C shows the same microheart tissue on Day 0, observed from the top. [Figure 9D] Figure 9D shows a schematic of how cells or tissues grown in microplates can be analyzed using a robotic system. [Figure 10A] Figure 10A shows the dose-response curve of isoprenaline (ISO) on mature microcardial tissue. [Figure 10B]Figure 10B shows the dose-response curve of Bay K8644 on mature microcardial tissue. [Figure 10C] Figure 10C shows the dose-response curve of milrinone on mature microcardiac tissue. [Figure 10D] Figure 10D shows the dose-response curve of dobutamine on mature microcardiac tissue. [Figure 11A] Figure 11A shows a confocal image of 35-day-old NuHeart® microcardiac tissue without the addition of transforming growth factor β (TGFβ). Red: cardiac troponin I, green: α-SM actin, blue: DAPI. Double arrows indicate multinucleation (scale bar = 10 μm). [Figure 11B] Figure 11B is a confocal image of the same 35-day-old microheart tissue with TGFβ added (colors are the same as above). [Figures 11C-11F] Figures 11C–11F show length-tension curves of myocardial contractility (11C, 11D) and diastolic stiffness (11E, 11F), respectively, observed for >30 days in the presence or absence of TGFβ. [Figure 11G] Figure 11G shows a bar graph of the positive inotropic response to ISO (0.1 mM, 1 mM) at the end of tissue development on day 35 (bars = standard deviation, n = 7). [Figure 11H] Figure 11H shows that α-SM actin was elevated in TGFβ-added samples. [Figure 11I-J]Figures 11I and 11J show the contractile force profiles of NuHeart® microcardiac tissues before and after ISO treatment with and without TGFβ. In the absence of TGFβ, ISO increased contractile amplitude and decreased duration, whereas the response was blunted in the presence of TGFβ. Figures 11I and 11J show the myocardial contractile force profiles with and without TGFβ. In the absence of TGFβ, isoprenaline (ISO) treatment increased contractile amplitude and decreased duration (positive inotropic and chronotropic responses). In contrast, the response to ISO was blunted in the presence of TGFβ. Figure 11I shows the myocardial contractile force profile in the absence of TGFβ. Figure 11J shows the myocardial contractile force profile in the presence of TGFβ. [Figures 12A-12B] Figures 12A and 12B show different views of InvivoSciences' representative mechanical assay device, the Palpator®. [Figure 12A] Figure 12A shows a close-up view of the Palpator®, showing microheart tissue being analyzed in a 96-well plate. [Figure 12B] Figure 12B is an overview of the entire "Palpator (registered trademark)." [Figure 12C] Figure 12C shows a force probe stretching the microcardial tissue vertically from above (top) and a response plot of force (mN) versus time (seconds) during the tissue stretching process (bottom). [Figure 13A] Figure 13A shows an example of a microplate reader used to analyze calcium transients and action potentials in microcardiac tissues in a 96-well plate. [Figures 13B-13C]Figures 13B and 13C show biologically stained microcardial tissues, respectively. Figure 13B shows the stained 96-well plate, and Figure 13C shows the calcium transient and action potential analysis output. Figure 13B is a photograph of the actual biologically stained microcardial tissues in a 96-well plate. Figure 13C shows an example of calcium transient and action potential output in the microcardial tissues. [Figure 14A] Figure 14A shows a graph of myocardial stress in microheart tissues derived from human iPSCs derived from familial hypertrophy (FCH) and control groups. Each group was assessed at different stretch levels according to the estimated sarcomere length (physiological range <2.2 μm). [Figure 14B] Figure 14B is a graph showing that there is a statistically significant difference (P < 0.05) between the control group and FCH (hypertrophic cardiomyopathy, HCM) in baseline stress of microcardiac tissue (n = 6, error bars: standard deviation). [Figures 15A-15C]Figures 15A–15C show a closed-loop cardiovascular (CV) model simulating individual patient CV function using 11 clinical parameters collected by transthoracic echocardiography (TTE) and right heart catheterization (RHC). In this simulation, the following nine mechanistic model parameters are used to define the patient's CV phenotype: ELV: left ventricular (LV) contractility, lLV: LV stiffness (elasticity), ERV: right ventricular (RV) contractility, lRV: RV stiffness (elasticity), EPA: pulmonary artery (PA) stiffness (elasticity), EPV: pulmonary vein (PV) stiffness (elasticity), Rpul: pulmonary vascular resistance, ESA: systemic artery (SA) stiffness (elasticity), and Rsys: systemic vascular resistance. Figure 15A is a flowchart illustrating the optimization procedure for adjustable parameters in the dimensionality reduction and clustering algorithm. This algorithm identifies optimal patient clusters with characteristic phenotypes. Figure 15B shows an example of unsupervised clustering of 136 heart failure (HF) patients into 19 groups, including outliers (dark blue, #-1), based on the Uniform Manifold Approximation and Projection (UMAP) dimensionality reduction algorithm. Figure 15C shows plots of the following parameters for each cluster: ejection fraction (EF), mean pulmonary artery pressure (mPAP), pulmonary artery wedge pressure (APWP), left ventricular contractility, left ventricular stiffness, and systemic arterial stiffness. [Figures 16A-16D]Figures 16A–16D show the long-term effects of fasudil. The stretch rate is expressed as "strain" relative to the unstretched tissue length, and error bars indicate standard deviation. Each condition had two or four samples. Figure 16A shows the relationship between contractile stress and strain after 20 days of culture in the absence of TGF-β and administration of fasudil (0 μM, 1 μM, 3 μM, 10 μM: filled circles, open circles, open triangles, open squares). Figure 16B shows the relationship between contractile stress and strain after 20 days of culture in the presence of TGFβ (1 ng / mL) and administration of fasudil (0 μM, 1 μM, 3 μM, 10 μM: filled circles, open circles, open triangles, open squares). Figure 16C shows the relationship between passive stress and strain in engineered cardiac tissue cultured for 20 days in the absence of TGFβ and the addition of fasudil (0 μM, 1 μM, 3 μM, 10 μM: filled circles, open circles, open triangles, open squares). Figure 16D shows the relationship between passive stress and strain in engineered cardiac tissue cultured for 20 days in the presence of TGFβ (1 ng / mL) and the addition of fasudil (0 μM, 1 μM, 3 μM, 10 μM: filled circles, open circles, open triangles, open squares).

[0015] DETAILED DESCRIPTION The exemplary materials, methods, and techniques disclosed and contemplated herein relate generally to systems and methods for treating heart failure with preserved ejection fraction (HFpEF). I. Definitions

[0016] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. In the event of a conflict in terminology, the present specification, particularly definitions, shall control. Materials and methods are described herein below; however, equivalent or similar methods and materials not described herein may also be used in the practice or testing of this disclosure. All publications, patent applications, patents, and other references cited are incorporated herein in their entirety. The materials, methods, and examples disclosed herein are illustrative only and are not intended to be limiting.

[0017] As used herein, "comprise(s)," "having," "has," "can," "contain(s)," and variations thereof are open-ended transitional terms and do not exclude the possibility of additional acts or structures. Additionally, the singular forms "a," "an," and "the" are intended to include the plural unless the context dictates otherwise. This disclosure also contemplates other embodiments, such as "comprising," "consisting of," and "consisting essentially of," which include elements of this disclosure whether or not they are explicitly stated.

[0018] The term "about" is used herein to indicate that a numerical value may not necessarily be exactly achievable. Therefore, "about" is a term that indicates this range of uncertainty. "About" may mean ±10% of the stated numerical value. For example, "about 10%" could mean a range of 9% to 11%, and "about 1" could mean a range of 0.9 to 1.1. Depending on the context, "about" may also mean rounding to the nearest whole number; for example, "about 1" could mean 0.5 to 1.4. The modifier "about" is also considered to disclose a range defined by the absolute values ​​at both ends. For example, "about 2 to about 4" includes the range "2 to 4."

[0019] When a range of numbers is stated, all intermediate values ​​(of equal precision) are implied. For example, the range "6 to 9" includes 7 and 8 as well as 6 and 9. The range "6.0 to 7.0" also includes 6.0, 6.1, 6.2, …, 6.9, 7.0. As another example, if a pressure range is stated as "from atmospheric pressure to any other pressure," atmospheric pressure is considered to be explicitly included.

[0020] As used herein, the term "subject" refers to a human or animal. Typically, a subject is a mammal. Subjects include primates (e.g., humans, males and females, infants, adolescents, or adults), non-human primates, rats, mice, rabbits, pigs, cows, sheep, goats, horses, dogs, cats, fish, birds, and the like. In one embodiment, the subject is a human. In another embodiment, the subject is an animal.

[0021] As used herein, "treatment" or "treating" refers to preventing, inhibiting, arresting the progression of, reversing, mitigating, ameliorating, or eliminating a disease or condition. Treatment can be administered in either an acute or chronic form. The term "treatment" also includes reducing the severity of a disease or associated symptoms before the onset of the disease.

[0022] "Repressing or ameliorating a disease, disorder, or symptom thereof" refers to administering a cell, composition, or compound described herein to a subject after the disease, disorder, or symptom thereof is clinically manifest, whereas "prophylaxis" or "preventing" refers to administering the same cell, composition, or compound to a subject before the disease, disorder, or symptom thereof is manifest.

[0023] "Suppressing" refers to administering to a subject a cell, composition, or compound described herein after a disease or disorder has been induced but before clinical symptoms appear.

[0024] As used herein, "salt" or "salts" refers to acid addition salts or base addition salts of the compounds of the present disclosure, specifically including "pharmaceutically acceptable salts." By "pharmaceutically acceptable salts," we mean salts that retain the biological effectiveness and properties of the compounds of the present disclosure and that generally do not possess biologically or otherwise undesirable properties. I. Heart failure with preserved ejection fraction (HFpEF)

[0025] Heart failure is broadly classified into two types: heart failure with reduced ejection fraction (HFrEF) (also known as heart failure due to left ventricular systolic dysfunction or systolic heart failure) and heart failure with preserved ejection fraction (HFpEF) (also known as diastolic heart failure or heart failure with normal ejection fraction). HFrEF refers to an ejection fraction (EF) of less than 40%. EF is a measure of the heart's pumping efficiency, or its ability to pump blood from the left ventricle into the systemic circulation. It is usually measured as the percentage of blood volume ejected from the left ventricle during systole of one heartbeat. A heart with normal cardiac function has an EF between 50% and 70%. A preserved EF is generally defined as 50% or greater, while a reduced EF is 40% or less; however, an EF between 40% and 50% is sometimes considered borderline. Another definition defines an EF of less than 50% as reduced and an EF of 50% or greater as preserved.

[0026] Myocardial fibrosis is a central component of the pathology of heart failure, particularly heart failure with preserved ejection fraction (HFpEF), which is characterized by the excessive accumulation of fibrillated collagen as a result of cardiomyocyte death, inflammation, increased cardiac load, hypertrophy, and stimulation by multiple hormones, cytokines, and growth factors.

[0027] Myocardial fibrosis can also refer to abnormal thickening of heart valves, a condition caused by inappropriate proliferation of cardiac fibroblasts, but more commonly refers to the following condition: Myocardial fibrosis is the excessive proliferation of fibroblasts within the myocardium. Fibroblasts normally secrete collagen and provide structural support to the heart. However, excessive activation can cause thickening and fibrosis of the valves, with white tissue accumulating primarily in the tricuspid valve, but occasionally in the pulmonary valve as well. This thickening and loss of flexibility can lead to valve dysfunction and right-sided heart failure.

[0028] Elevated levels of N-terminal pro-B-type natriuretic peptide (NT-proBNP) may also be associated with heart failure with preserved ejection fraction (HFpEF). NT-proBNP is a type of biomarker frequently used in clinical trials to assess the severity of heart failure. Typically, subjects with NT-proBNP levels above 900 ng / L are considered elevated. In some cases, NT-proBNP levels can reach 2,000 ng / L or even 3,000 ng / L.

[0029] The disclosed compounds and compositions can be used in a method for treating heart failure with preserved ejection fraction (HFpEF) by administering to a subject in need of such treatment a therapeutically effective amount of a composition comprising a Rho kinase (ROCK) inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof.

[0030] These compositions can be administered to a subject in need of such treatment to modulate ROCK, an enzyme that controls various biological processes, particularly its role in regulating downstream signaling pathways and biological activity, such as smooth muscle cell contractility. The present disclosure relates to methods for administering compositions to inhibit ROCK. III. Rho Kinase (ROCK) Inhibitors and Other Antifibrotic Compounds

[0031] Treating heart failure with preserved ejection fraction (HFpEF) in a subject can be achieved by administering to the subject a therapeutically effective amount of a Rho-kinase (ROCK) inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof. The subject can exhibit diastolic dysfunction, myocardial fibrosis, pulmonary hypertension, a left ventricular ejection fraction of 50%, and elevated N-terminal pro-B-type natriuretic peptide (NT-proBNP). In various embodiments, the subject can be free of cardiac amyloid deposits, i.e., "amyloid-negative." A. Rho kinase (ROCK) inhibitors

[0032] As used herein, the term "Rho kinase (ROCK) inhibitor" refers to a protein, nucleic acid, small molecule compound, antibody, or other factor that inhibits the expression of ROCK or reduces the activity (e.g., kinase activity) of ROCK. Examples of ROCK inhibitors are disclosed herein.

[0033] ROCK kinase inhibitors (e.g., fasudil) herein may affect myocardial contractility and reduce myocardial tissue stiffness associated with heart failure.

[0034] As used herein, an "effective amount" or "therapeutically effective amount" means a non-toxic and sufficient quantity of a substance, agent, composition, or cell administered to a subject to prevent, treat, or alleviate to some extent one or more symptoms of the disease or condition. This may result in the reduction or elimination of the signs, symptoms, or causes of the disease, or any other desired alteration of a biological system. The effective amount will be determined based on factors specific to the individual subject, such as age, size, type and extent of disease, stage of disease, route of administration, type and extent of adjunctive therapy, ongoing medical condition, and type of treatment desired.

[0035] As used herein, a "therapeutically effective amount" refers to an amount of a Rho kinase (ROCK) inhibitor sufficient to elicit a physiological or medical response in a subject, such as inhibiting or promoting enzyme or protein activity, improving symptoms, alleviating a condition, delaying disease progression, or preventing disease. As a non-limiting example, a "therapeutically effective amount" refers to an amount of a ROCK inhibitor (e.g., fasudil) described herein administered to a subject that is effective to inhibit the expression and / or activity of ROCK. For example, a therapeutically effective amount of a ROCK inhibitor (e.g., fasudil) may be 10-60 mg per kg of body weight per day.

[0036] In various embodiments, the administered ROCK inhibitor may be fasudil and may be administered orally. In some embodiments, fasudil is administered at a dose of 30-60 mg / kg, 35-55 mg / kg, or 40-50 mg / kg per day. In other embodiments, the dose may be 60 mg / kg or less, 55 mg / kg or less, 50 mg / kg or less, 45 mg / kg or less, 40 mg / kg or less, 35 mg / kg or less, or 30 mg / kg or less per day. In yet other embodiments, the dose may be 30 mg / kg or more, 35 mg / kg or more, 40 mg / kg or more, 45 mg / kg or more, 50 mg / kg or more, 55 mg / kg or more, or 60 mg / kg or more per day. B. Other Antifibrotic Compounds

[0037] In experimental screening of anti-fibrotic compounds, three angiotensin II receptor blockers, two ACE inhibitors, two statins, one beta-blocker, and two NSAIDs (nonsteroidal anti-inflammatory drugs) were identified as positive hits. Furthermore, this screening identified a new class of... Further screening of anti-fibrotic compounds revealed that various compound groups with different targets and mechanisms of action showed potential for the treatment of myocardial fibrosis. Many compounds modulate neurotransmitter receptors, including: 1) histamine receptors, 2) serotonin receptors, 3) dopamine receptors, 4) glutamate receptors, and 5) GABA receptors.

[0038] Other classes of antifibrotic compounds include those that modulate targets such as: 1) glucose metabolism, 2) adrenergic receptors, 3) retinoic acid, 4) ion channels, 5) KATP channels, 6) prostaglandins, 7) phosphodiesterases, 8) bisphosphonates, 9) antivirals, 10) DNA function, 11) sex hormones, 12) HIV protease, 13) antibiotics, 14) EGF receptors, 15) thrombin function, 16) Rho kinase, 17) cholecystokinin, and 18) other targets. Representative examples of compounds that may modulate these targets include:

[0039] 005 Losartan: Angiotensin II receptor blocker (AT1)

[0040] 007 Telmisartan: Angiotensin II receptor blocker (AT1)

[0041] 057 Olmesartan: Angiotensin II receptor blocker (AT1)

[0042] 054 Lisinopril: Angiotensin-converting enzyme (ACE) inhibitor

[0043] 082 Ramipril: Angiotensin-converting enzyme (ACE) inhibitor

[0044] 025 Simvastatin: HMG-CoA reductase inhibitor

[0045] 091 Pravastatin: HMG-CoA reductase inhibitor

[0046] 041 Propranolol: β-adrenergic receptor blocker

[0047] 018 Acemetacin: Nonsteroidal anti-inflammatory drug (NSAID)

[0048] 090 Sulindac: Nonsteroidal anti-inflammatory drug (NSAID)

[0049] 001 Metformin: mechanism of action unclear

[0050] 095 Pioglitazone: DPP-4 inhibitor

[0051] 002 Tiotidine: Histamine H2 receptor antagonist

[0052] 006 Cimetidine: Histamine H2 receptor antagonist

[0053] 012 Roxatidine: Histamine H2 receptor antagonist

[0054] 066 Triprolidine: First-generation histamine H1 receptor antagonist

[0055] 003 Tranilast: Inhibits histamine release from mast cells

[0056] Serotonin-related (5-HT, 5-hydroxytryptamine)

[0057] 004 Memantine: NMDA (N-methyl-D-aspartate) receptor antagonist

[0058] 017 Zolmitriptan: 5HT(1B / 1D) receptor agonist

[0059] 037 Ergotamine: 5-HT(1B / D / F) receptor agonist

[0060] 038 Nefazodone: A potent 5-HT(2A / 2C) receptor antagonist

[0061] 042 Methysergide: A 5HT2 antagonist and a 5HT1 agonist

[0062] 013 Deprenyl (selegiline): MAO-B inhibitor, inhibits the breakdown of dopamine

[0063] 021 Bupropion: A noradrenaline and dopamine reuptake inhibitor (NDRI)

[0064] 024 Lobeline: VMAT2 ligand, stimulates dopamine release

[0065] 069 Itopride: Dopamine D2 receptor antagonist

[0066] 030 Apomorphine: a non-ergoline dopamine agonist

[0067] 087 Spiperone: A selective D2 dopamine receptor antagonist

[0068] 078 Quinpirole: A selective D2 and D3 receptor agonist

[0069] 014 Aniracetam: A positive modulator of AMPA-sensitive glutamate receptors

[0070] 023 Bethanechol: A selective agonist of muscarinic acetylcholine receptors

[0071] 034 Carbachol: stimulates both muscarinic and nicotinic receptors

[0072] 040 Tubocurarine: Nicotinic acetylcholine receptor antagonist

[0073] 093 Pancuronium: A competitive antagonist of nicotinic acetylcholine receptors

[0074] 038 Physostigmine: Reversible cholinesterase inhibitor

[0075] 043 Tacrine: Reversible cholinesterase inhibitor

[0076] 050 Rocuronium: Competitively binds to nicotinic cholinergic receptors

[0077] 044 Propofol: A positive modulator of GABA-A receptors

[0078] 053 Ivermectin: GABA receptor agonist

[0079] 081 Tramadol: A selective and weak opioid OP3 receptor agonist

[0080] 015 Epinephrine (adrenaline): Activates adrenaline receptors and relaxes smooth muscles

[0081] 072 Melatonin: A hormone synthesized and secreted by the pineal gland

[0082] 049 Niguldipine: A selective α1A adrenergic receptor antagonist

[0083] 079 Xylazine: A potent alpha-2 adrenergic receptor agonist

[0084] 064 Fluperlapine: High affinity for alpha 1 adrenergic receptors (IC 50 ≒ 10 nM)

[0085] 076 Tulobuterol: Long-acting beta-2 adrenergic receptor agonist

[0086] 080 Salmeterol: Long-acting beta-2 adrenergic receptor agonist

[0087] 008 Retinoic acid: A metabolite of vitamin A1 (all-trans retinol)

[0088] 084 Vitamin A1: Retinol used to treat and prevent vitamin A deficiency

[0089] 009 Dofetilide: IKr (delayed rectifier potassium current) inhibitor

[0090] 068 Indapamide: Antihypertensive and diuretic

[0091] 038 Nifekalant: a nonselective potassium channel blocker

[0092] 010 Lidocaine: Voltage-gated sodium channels (Na + ) is blocked

[0093] 016 Ouabain: A cardiac glycoside similar to digitoxin

[0094] 022 Nimodipine: L-type calcium channel blocker

[0095] 028 Riluzole: Inhibits glutamate release

[0096] 030 Zonisamide: Sodium and calcium channel blocker

[0097] 059 Glimepiride: Blocks KATP channels

[0098] 060 Butyrylcholine: Blocks KATP channels

[0099] 086 Nateglinide: Blocks KATP channels

[0100] 011 Alprostadil: Prostaglandin E1 (PGE1)

[0101] 052 Sulfasalazine: The mechanism of action is unknown, but it inhibits prostaglandin synthesis.

[0102] 019 Zaprinast: A selective inhibitor of cGMP-dependent phosphodiesterase (PDE)

[0103] 020 Vardenafil: cGMP-dependent phosphodiesterase type 5 (PDE5) inhibitor

[0104] 032 Trequincin: A powerful PDE3 inhibitor

[0105] 074 Zardaverine: A dual selective inhibitor of phosphodiesterase III / IV

[0106] 083 Siguazodan: A selective inhibitor of phosphodiesterase 3

[0107] 029 Lorglumide: Cholecystokinin A and B receptor antagonist

[0108] Bisphosphonates (known relative efficacy: zoledronic acid > pamidronate = alendronate)

[0109] 026 Zoledronic Acid: A Bisphosphonate for the Treatment of Osteoporosis

[0110] 065 Pamidronate: A bisphosphonate for the treatment of osteoporosis

[0111] 092 Alendronate: A bisphosphonate for the treatment of osteoporosis

[0112] 027 Oseltamivir: Inhibitor of influenza virus

[0113] 046 Rimantadine: M2 ion channel inhibitor

[0114] 047 Allopurinol: Xanthine oxidase inhibitor

[0115] 048 Itraconazole: Azole antifungal drug

[0116] 063 Idoxuridine: An antiviral drug used in keratitis caused by herpes simplex virus.

[0117] 089 Ribavirin: A synthetic guanosine nucleoside and antiviral drug

[0118] 031 Acycloguanosine: A nucleoside analogue DNA polymerase inhibitor

[0119] 035 Shikonin: DNA inhibitor

[0120] 067 Fluorouracil: Antineoplastic antimetabolite

[0121] 073 Lomustine: An alkylating agent used to treat neoplastic diseases

[0122] 033 Goserelin: Gonadotropin-releasing hormone agonist

[0123] 062 Hexestrol: Estrogen receptor agonist

[0124] 085 Stanozolol: Anabolic steroid

[0125] 036 Imiquimod: Toll-like receptor 7 agonist

[0126] 039 Nelfinavir: HIV-1 protease inhibitor

[0127] 075 Amprenavir: HIV-1 protease inhibitor

[0128] 045 Streptomycin: Antibiotic

[0129] 056 Florfenicol: Antibiotic

[0130] 071 Pefloxacin: Antibiotic

[0131] 094 Rifamycin: Antibiotic

[0132] 051 Tanshinone: Phenanthrenequinone derivative

[0133] 055 Suramin: mechanism of action unknown

[0134] 058 Mebendazole: an inhibitor of microtubule synthesis

[0135] 077 Argatroban: A synthetic direct thrombin inhibitor derived from L-arginine

[0136] 088 Lapatinib: A dual inhibitor of EGFR and HER2 receptor tyrosine kinases

[0137] 096 Fasudil: A potent Rho kinase (ROCK) inhibitor

[0138] In some cases, the ROCK inhibitor may be fasudil. C. Pharmaceutically Acceptable Salts, Esters and Solvates

[0139] The compounds of the present disclosure may exist as pharmaceutically acceptable salts, esters, or solvates.

[0140] As used herein, the term "pharmaceutically acceptable salt" refers to a salt or zwitterion that is soluble or dispersible in water or oil. The compounds of the present disclosure may exist in the form of pharmaceutically acceptable salts, esters, or solvates that are effective for their intended use at a reasonable benefit / risk ratio without excessive toxicity, irritation, or allergic reactions. These salts may be prepared during the final isolation and purification process of the compound, or separately, by reacting the amino group of the compound with a suitable acid. For example, the compound may be dissolved in a suitable solvent, such as methanol or water, and treated with one or more equivalents of an acid (e.g., hydrochloric acid), resulting in the precipitation of a salt, which may be isolated by filtration and drying under reduced pressure. Alternatively, the solvent and excess acid may be removed under reduced pressure to obtain the salt. Representative salts include acetate, adipate, alginate, citrate, aspartate, benzoate, benzenesulfonate, bisulfite, butyrate, camphorate, camphorsulfonate, digluconate, glycerophosphate, hemisulfate, heptanoate, hexanoate, formate, isethionate, fumarate, lactate, maleate, methanesulfonate, naphthalenesulfonate, niacinate, oxalate, pamoate, pectinate, persulfate, 3-phenylpropionate, picrate, pivalate, propionate, succinate, tartrate, trichloroacetate, trifluoroacetate, glutamate, p-toluenesulfonate, undecanoate, hydrochloride, hydrobromide, sulfate, phosphate, and the like. The amino group may also be quaternized with alkyl chlorides, bromides, or iodides (e.g., methyl, ethyl, propyl, isopropyl, butyl, lauryl, myristyl, stearyl) to provide acceptable and effective salt forms for appropriate therapeutic applications.

[0141] As used herein, the term "pharmaceutically acceptable ester" refers to an ester that is hydrolyzed in vivo to yield the parent compound or a salt thereof. Examples of suitable ester groups include those derived from pharmaceutically acceptable aliphatic carboxylic acids, particularly alkanoic acids, alkenoic acids, cycloalkanoic acids, and alkanedicarboxylic acids, each of which preferably has six or fewer carbon atoms in its alkyl or alkenyl group. Specific examples include, but are not limited to, formates, acetates, propionates, butyrates, acrylates, and ethylsuccinates.

[0142] As used herein, the term "pharmaceutically acceptable solvate" refers to a substance formed by combining a target compound with a solvent molecule. For example, the compounds described herein may form hydrates (combined with water). D. Pharmaceutical Compositions

[0143] The ROCK inhibitors (e.g., fasudil) and other compounds disclosed herein may be incorporated into pharmaceutical compositions suitable for administration to a subject (e.g., a human or non-human patient). The pharmaceutical composition may contain a "therapeutically effective amount" or a "prophylactically effective amount." A "therapeutically effective amount" refers to an amount effective, at a dosage and for a period of time necessary, to achieve the desired therapeutic effect. A therapeutically effective amount of a composition can be determined by one of skill in the art and may vary depending on factors such as the disease state, age, sex, weight, and the ability of the composition to elicit a desired response in a subject. A therapeutically effective amount is also an amount in which the toxic or adverse effects of a disclosed compound (e.g., fasudil) are less than its therapeutically beneficial effects. A "prophylactically effective amount" refers to an amount effective, at a dosage and for a period of time necessary, to achieve the desired prophylactic effect. Typically, prophylactic administration is used in subjects before the onset of disease or at an early stage of disease, so the prophylactically effective amount is often less than the therapeutically effective amount.

[0144] The amount of a pharmaceutical composition to be used can be determined by one skilled in the art and may vary depending on factors such as the state of the disease, age, sex, weight, and the ability of the composition to elicit the desired response in an individual. A "therapeutically effective amount" refers to an amount in which the therapeutic benefit outweighs the toxic or harmful effects of the compound (e.g., fasudil). A "prophylactically effective amount" refers to an amount effective at a dosage and for a period of time necessary to achieve the desired preventive effect. Generally, prophylactic administration is performed before the progression of a disease or at an early stage, so the prophylactically effective amount tends to be lower than the therapeutically effective amount. A pharmaceutical composition may contain a pharmaceutically acceptable carrier. A "pharmaceutically acceptable carrier" refers to any non-toxic and inert solid, semi-solid, or liquid filler, diluent, encapsulating material, formulation auxiliary ingredient, or the like. Examples of substances that can be used as pharmaceutically acceptable carriers include, but are not limited to, sugars such as lactose, glucose, and sucrose, starches such as corn starch and potato starch, cellulose and its derivatives such as sodium carboxymethylcellulose, ethyl cellulose, and cellulose acetate, powdered tragacanth, malt, gelatin, talc, excipients such as cocoa butter and suppository wax, oils such as peanut oil, cottonseed oil, safflower oil, sesame oil, olive oil, corn oil, and soybean oil, glycols such as propylene glycol, esters such as ethyl oleate and ethyl laurate, agar, buffers such as magnesium hydroxide and aluminum hydroxide, alginic acid, and pyrogenic substances (see below).

[0145] Thus, the compounds disclosed herein, and their physiologically acceptable salts and solvates, may be formulated into solids, eye drops, oily topical preparations, injections, inhalants (oral or nasal), implants, or forms suitable for oral, buccal, injectable, or rectal administration. Techniques and formulation examples are generally found in literature such as "Remington's Pharmaceutical Sciences" (Meade Publishing Co., Easton, Pa.). Therapeutic formulations typically must be sterile and stable under the conditions of manufacture and storage.

[0146] The type of carrier to be used will depend on the route of administration of the disclosed compounds and the formulation of the composition. The compositions may be prepared in a variety of forms suitable for, for example, systemic administration (oral, rectal, nasal, sublingual, buccal, implant, injection, etc.) or topical administration (cutaneous, pulmonary, nasal, aural, ocular, liposome delivery system, iontophoresis, etc.).

[0147] Carriers for systemic administration include at least a diluent, lubricant, binder, disintegrant, colorant, flavoring agent, sweetener, antioxidant, preservative, glidant, solvent, suspending agent, wetting agent, surfactant, or combinations thereof. All of these carriers can be used as optional ingredients. Suitable diluents include sugars such as glucose, lactose, dextrose, sucrose, and diols such as propylene glycol.

[0148] Suitable lubricants include silica, talc, stearic acid and its magnesium and calcium salts, and calcium sulfate. Liquid lubricants include polyethylene glycol and vegetable oils (e.g., peanut oil, cottonseed oil, sesame oil, olive oil, corn oil, and cocoa butter). The amount of lubricant in a systemic or topical composition is generally about 5-10%.

[0149] Suitable binders include polyvinylpyrrolidone, magnesium aluminum silicate, starches such as corn starch and potato starch, gelatin, tragacanth, and cellulose and its derivatives (e.g., sodium carboxymethylcellulose, ethyl cellulose, methyl cellulose, microcrystalline cellulose, etc.) The amount of binder in a systemic composition is generally about 5-50%.

[0150] Suitable disintegrants include agar, alginic acid and its sodium salt, effervescent mixture, croscarmellose, crospovidone, sodium carboxymethyl starch, sodium starch glycolate, clay, and ion exchange resins. The amount of disintegrant in a systemic or topical administration composition is generally about 0.1 to 10%. Suitable coloring agents include FD&C dyes. When used, the amount of coloring agent is generally about 0.005 to 0.1%. Suitable flavoring agents include menthol, peppermint, and fruit flavors, and the amount of flavoring is generally about 0.1 to 1.0%.

[0151] Suitable sweeteners include aspartame and saccharin. The amount of sweetener in a systemic or topical composition is typically about 0.001-1%. Suitable antioxidants include butylated hydroxyanisole (BHA), butylated hydroxytoluene (BHT), vitamin E, and the like. The amount of suitable antioxidant in a systemic or topical composition is typically about 0.1-5%. Suitable preservatives include benzalkonium chloride, methylparaben, sodium benzoate, and the like. The amount of preservative used is typically about 0.01-5%. Suitable lubricants include silicon dioxide, and the amount of lubricant used is typically about 1-5%. The amount of antioxidant included in a systemic or topical formulation is typically about 0.1-5%. Suitable preservatives include benzalkonium chloride, methylparaben, sodium benzoate, and the like. The amount of preservatives included in systemic or topical formulations is typically about 0.01 to about 5%. Suitable lubricants (glidants) include silicon dioxide (silica). The amount of lubricant included in systemic or topical formulations is typically about 1 to about 5%.

[0152] Suitable solvents include water, isotonic saline, ethyl oleate, glycerin, hydroxylated castor oil, alcohols such as ethanol, and phosphate buffers. The amount of solvent used in a systemic or topical composition is typically 0-100%. Suitable suspending agents include AVICEL RC-591 (FMC, Philadelphia, Pennsylvania) and sodium alginate. The amount of suspending agent used is typically about 1-8%. Suitable surfactants include lecithin, polysorbate 80 (Tween 80), sodium lauryl sulfate, and TWEEN products manufactured by Atlas Powder Co., Wilmington, Delaware. Information about surfactants is also found in the following literature: CTFA Cosmetic Ingredient Handbook, 1992, pp. 587-592; Remington's Pharmaceutical Sciences, 15th Edition, 1975, pp. 335-337; McCutcheon's Volume 1, Emulsifiers & Detergents, 1994, North American Edition, pp. 236-239. The amount of surfactant used is generally about 0.1 to 5%.

[0153] The proportion of each component in a composition for systemic administration may vary depending on the type of formulation, but generally contains 0.01 to 50% of the active ingredient (eg, fasudil) and 50 to 99.99% of the carrier component.

[0154] A composition for administration for injection (composition for parenteral administration) generally contains 0.1 to 10% of an active ingredient and 90 to 99.9% of a carrier including a diluent and a solvent.

[0155] Oral dosage forms can take a variety of forms. For example, solid forms include tablets, capsules, granules, and bulk powders. These oral dosage forms contain a safe and effective amount of the active ingredient (usually about 5% or more, more specifically about 25-50%), with the carrier component accounting for about 50-95%, more specifically about 50-75%.

[0156] Tablets include compressed tablets, lozenges, enteric-coated tablets, sugar-coated tablets, film-coated tablets, and multilayer compressed tablets. Tablets typically contain an active ingredient and carrier components selected from diluents, lubricants, binders, disintegrants, colorants, flavoring agents, sweeteners, glidants, and the like. Specific diluents include calcium carbonate, sodium carbonate, mannitol, lactose, cellulose, and the like. Specific binders include starch, gelatin, sucrose, and the like. Specific disintegrants include alginic acid and croscarmellose. Specific lubricants include magnesium stearate, stearic acid, and talc. Specific coloring agents include FD&C dyes, which are used to improve appearance. Chewable tablets preferably contain a sweetener such as aspartame or saccharin, or a flavoring such as menthol, peppermint, or fruit flavors (or a combination thereof).

[0157] Capsules (including implantable, sustained-release, and extended-release types) typically contain an active ingredient such as fasudil and a carrier containing one or more of the diluents listed above, and are filled into gelatin capsules. Granules preferably contain the disclosed compound and a lubricant such as silicic acid to improve flowability. Implants can be biodegradable or non-biodegradable.

[0158] The selection of ingredients for the carrier (base) for oral formulations depends on secondary considerations such as taste, cost, and shelf stability, which are not essential for the purposes of this disclosure. Solid formulations may be coated by conventional methods, usually with pH- or time-dependent coatings, to release the disclosed compounds in the gastrointestinal tract proximate the desired site of application, or at various sites and times to prolong the desired effect. The coatings typically contain one or more components selected from the group consisting of cellulose acetate phthalate, polyvinyl acetate phthalate, hydroxypropyl methylcellulose phthalate, ethyl cellulose, EUDRAGIT® coating (available from Rohm & Haas GmbH, Darmstadt, Germany), waxes, and shellac.

[0159] Oral formulations may be in liquid form. For example, suitable liquid forms include aqueous solutions, emulsions, suspensions, solutions reconstituted from non-effervescent granules, suspensions reconstituted from non-effervescent granules, effervescent preparations reconstituted from effervescent granules, elixirs, tinctures, syrups, etc. Liquid formulations for oral administration typically contain a disclosed compound and a **carrier (base)**, where the carrier is selected from diluents, colorants, flavorings, sweeteners, preservatives, solvents, suspending agents, and surfactants. Oral liquid formulations preferably contain one or more components selected from colorants, flavorings, and sweeteners.

[0160] Other formulations useful for achieving systemic delivery of the disclosed compounds include sublingual, buccal, and nasal dosage forms. Such formulations typically contain one or more soluble excipients (diluents including sucrose, sorbitol, mannitol, etc.) and binders (acacia, microcrystalline cellulose, carboxymethylcellulose, hydroxypropylmethylcellulose). Additionally, lubricants, colorants, flavors, sweeteners, antioxidants, and glidants may be included. IV. COMPUTER-IMPLEMENTED METHODS

[0161] In various embodiments, computer-implemented methods are used to identify personalized treatment strategies to improve disease phenotypes in heart failure patients, including AI-Assisted, Systems-biology Integrated Patient Stratification Technology (AASIST), as described in more detail in the Examples section below.

[0162] An exemplary computer-implemented method includes executing an artificial intelligence with one or more processors. Executing the artificial intelligence includes receiving, by the one or more processors, a plurality of patient parameters (data). The plurality of patient parameters received by the artificial intelligence may include: patients with and without diastolic dysfunction (diastolic asystole), patients with and without myocardial fibrosis, patients with and without pulmonary hypertension, patients with a left ventricular ejection fraction greater than or equal to 50% and patients less than 50%, and patients with and without myocardial amyloid deposits.

[0163] An exemplary computer-implemented method includes executing a feedforward artificial neural network by one or more processors. Executing the feedforward artificial neural network includes receiving, by the one or more processors, a plurality of patient parameters, including data and metadata, including: patients with and without diastolic dysfunction, patients with and without myocardial fibrosis, and pulmonary. The patient parameters include: patients with and without diastolic dysfunction, patients with and without myocardial fibrosis, patients with and without pulmonary hypertension, patients with a left ventricular ejection fraction greater than or equal to 50% and patients less than 50%, and patients with and without myocardial amyloid deposits.

[0164] An exemplary computer-implemented method may further include analyzing a plurality of patient parameters by placing them in neurons of an input layer of an artificial neural network. Then, weights are assigned to the plurality of patient parameters placed in the input layer, and the weighted parameters are placed in one or more hidden layers of a feedforward artificial neural network, thereby generating a plurality of optimized patient parameter data sets. Next, a cluster analysis is performed to receive and analyze the plurality of optimized patient parameter data sets. The analysis identifies a plurality of populations from the optimized data sets, and one or more clusters are generated and analyzed by the plurality of processors.

[0165] In each embodiment of the present invention, the computer-implemented method includes the AI-Assisted Systems Biology Integrated Patient Stratification Technology (AASIST) described in the Examples section below.

[0166] Figure 5 is a functional block diagram that schematically illustrates a representative computing environment 500. As shown in Figure 5, the representative computing environment 500 may include a computer system 520 and a storage area network 530 connected via a network 510.

[0167] A representative computer system 520 may include an AASIST program 522 and a computer interface 524. The storage area network 530 may include a server system 532 and a database 534. In various examples, the computer system 520 may be a standalone device, a server, a laptop computer, a tablet, a netbook, a personal computer (PC), a personal digital assistant, a desktop computer, or any other programmable electronic device capable of receiving, transmitting, and processing data. Generally, the computer system 520 represents any programmable electronic device capable of executing machine-readable program instructions and communicating with other computer systems (not shown), or a combination thereof. In one embodiment, the computer system 520 may represent a computing system that uses multiple clustered computers and components to function as a seamless resource pool. The computer system 520 may be accessible to multiple other computing systems (not shown) and may execute the AASIST program 522 and the computer interface 524. For example, the computer system 520 may include various internal and external hardware components (not shown).

[0168] 5, in one embodiment, AASIST program 522 and computer interface 524 may be stored on computer system 520. However, in another embodiment (not shown), AASIST program 522 and computer interface 524 may be stored externally and accessed via a communications network (e.g., network 510).

[0169] In general, network 510 may be any combination of connections and protocols for supporting communications between computer system 520, storage area network 530, and multiple other computer systems (not shown) in accordance with a preferred embodiment of the present invention. Network 510 may be, for example, a local area network (LAN), a wide area network (WAN, e.g., the Internet), or a combination thereof.

[0170] In various examples, the other computer systems (not shown) may be standalone devices, servers, laptop computers, tablet computers, netbook computers, personal computers (PCs), desktop computers, or any programmable electronic device capable of receiving, transmitting, and processing data. In other embodiments, these computer systems represent computing systems utilizing clustered computers and components to function as a single pool of seamless resources. Generally, these other computer systems are accessible to computer system 520 and network 510 and may run AASIST program 522 and computer interface 524. These other computer systems may include internal and external hardware components such as those shown and described with respect to FIG. 5.

[0171] As shown in FIG. 5 , in one embodiment, AASIST program 522 utilizes, at least in part, data stored in database 534 to manage access to computer system 520 in response to digital media recognition requests from users (i.e., users of computer system 520, hereinafter also referred to as “requesters”). More specifically, AASIST program 522 defines one or more artifacts and weights, which represent types of input / output (I / O). These constitute a feedforward artificial neural network that generates a dataset of multiple optimized patient parameters. This dataset is then used in cluster analysis to generate an instruction set for administering an effective amount of a drug, such as a Rho kinase (ROCK) inhibitor, an antifibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof. For example, a “weight” may be a value metric, and an “artifact” may be a data point.

[0172] Computer system 520 may include numerous logics and / or programs managed according to AASIST program 522. Generally, computer system 520 manages access to AASIST program 522, which may be physical or virtual resources. In various cases, AASIST program 522 includes information and code that, when executed, enables computer system 520 to perform specific operations on other physical or virtual resources. For example, in one embodiment, AASIST program 522 manages at least some functionality that enables computer system 520 to perform actions on one or more physical or virtual resources. In another embodiment, AASIST program 522 controls physical and / or virtual resources. In still other embodiments, AASIST program 522 may embody any or a combination of the elements described above. To illustrate various aspects of the present disclosure, the following examples of AASIST programs 522 are provided, including but not limited to: patient profile transactions, predictor value estimation profiles, artificial neural network (ANN) requests, and clustering analyses. AASIST programs 522 may also include other types of transactions known in the art.

[0173] Storage area network (SAN) 530 may be a storage system including server system 532 and database 534. SAN 530 may include one or more (but not limited to) computing devices, servers, server clusters, web servers, databases, and storage devices. SAN 530 may operate to communicate with computer system 520 and various other computing devices or systems (not shown) over a network, such as network 510. For example, SAN 530 may communicate with AASIST program 522 to transfer data between computer system 520 and various other computing devices or systems (not shown) connected to network 510. SAN 530 may be any computing device or combination communicatively connected to a local IoT network, i.e., a network consisting of various computing devices, including, but not limited to, computer system 520 and various other computing devices. SAN 530 may include the internal and external hardware components described with respect to FIG. 6. This disclosure recognizes that Figure 5 may include any number of computing devices, servers, databases, and / or storage devices, and this disclosure is not limited to only those shown in Figure 5. Accordingly, in various embodiments, some or all of the features and functionality of SAN 530 may be included as part of computer system 520 and / or various other computing devices or computer systems. Similarly, in various embodiments, some of the features and functionality of computer system 520 may be included as part of SAN 530 and / or various other computing devices or computer systems.

[0174] Additionally, in some cases, the storage area network (SAN) 530 represents a cloud computing platform. Cloud computing is a service delivery model that enables on-demand, convenient access to a shared pool of configurable computing resources (e.g., networks, network bandwidth, servers, processors, memory, storage, applications, virtual machines, and services) over a network. These resources can be rapidly provisioned and released with minimal management effort or interaction with a service provider. The cloud model may include characteristics such as on-demand self-service, broad network access, resource pooling, rapid elasticity, and metered services based on usage, and is expressed through service models such as Platform as a Service (PaaS), Infrastructure as a Service (IaaS), and Software as a Service (SaaS). It may also be implemented as a variety of deployment models, including private clouds, community clouds, public clouds, and hybrid clouds.

[0175] In various examples, SAN 530 may include any number of databases managed according to the functions of applications executing on SAN 530. In one embodiment, database 534 represents data, and server system 532 represents code for performing specific actions on other physical or virtual resources and managing the use and modification of the data. In another embodiment, AASIST program 522 may represent any or a combination of the above functions, and database 534 may be accessible by applications executing on SAN 530. To illustrate various aspects of the disclosure, an embodiment is presented in which AASIST program 522 represents one or more of the following, including but not limited to, an example application executing on SAN 530: a local IoT network, a digital media awareness monitoring system.

[0176] 5, in some cases database 534 is stored on SAN 530. However, in other embodiments (not shown), database 534 may be stored externally and accessed via a communications network, such as network 510 described above.

[0177] The AASIST program 522 may include a stored data and / or statistical data analysis program for performing a clustering analysis. In some examples, the AASIST program 522 may analyze, at least in part, (i) the output data of the multilayer neural network and (ii) the output weight changes. The AASIST program 522 may generate a plurality of optimized data sets based on a plurality of patient parameters, including (1) data and (2) metadata regarding the patient parameters. Additionally, the AASIST program 522 may perform a clustering analysis and generate a set of instructions for administering an antifibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof, to a target subject in an effective amount based on at least the optimized data set of patient parameters.

[0178] 6 is a block diagram 600 illustrating an exemplary arrangement of components associated with computer system 520 and SAN 530. Note that FIG. 6 is intended as an example of one embodiment and is not intended to suggest any limitation with respect to the environments in which other embodiments may be implemented. Many modifications to the depicted environment may be made.

[0179] 6, computer system 520 and SAN 530 may include a communications fabric 602. This communications fabric 602 provides communication between: a computer processor 604, memory 606, persistent storage 608, a communications unit 610, and an input / output (I / O) interface 612. Communications fabric 602 may be implemented with any architecture designed to transmit data and / or control information between a processor (e.g., a microprocessor, a communications / network processor, etc.), system memory, peripherals, and other hardware components. For example, communications fabric 602 may be implemented using one or more buses.

[0180] The memory 606 and the persistent storage 608 are computer-readable storage media. Generally, the memory 606 can include any suitable volatile or non-volatile computer-readable storage media. In some cases, the memory 606 can include a random access memory (RAM) 614 and a cache memory 616.

[0181] In some embodiments, the AASIST program 522, computer interface 524, server system 532, and database 534 may be stored in persistent storage 608 and accessed and executed by one or more processors 604 through one or more of the memories 606. The persistent storage 608 may include a magnetic hard disk drive or, in addition to a magnetic hard disk drive, may include other storage media such as a solid-state drive (SSD), semiconductor memory, read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, or any other computer-readable storage medium capable of storing program instructions or digital information.

[0182] The media used for persistent storage 608 may be removable. For example, a removable hard disk may be used as persistent storage 608. Other examples include optical disks, magnetic disks, USB flash drives, and smart cards, which may be inserted to transfer data to another computer-readable storage medium and thereby become part of persistent storage 608.

[0183] Communications unit 610 provides for communication with other data processing systems or devices, including communication with resources of network 110. Communications unit 610 may include one or more network interface cards. Communications may be provided via wired and / or wireless links. The AASIST program 522, computer interface 524, server system 532, and client application 534 may be downloaded to persistent storage 608 through communications unit 610.

[0184] The I / O interface 612 allows data to be input and output to and from the computer system 520, the SAN 530, and other devices connected to the client device 540. For example, the I / O interface 612 provides connection to external devices 618, such as a keyboard, a keypad, a touch screen, or other suitable input devices. The external devices 618 may also include portable computer-readable storage media, such as a USB memory stick (thumb drive), a portable optical disk, or a magnetic disk. Software and data used in embodiments of the present disclosure (e.g., AASIST program 522, computer interface 524, server system 532, client application 534) may be stored on these portable storage media and loaded into persistent storage 608 via I / O interface 612. Additionally, I / O interface 612 may also be connected to display 620.

[0185] Display 620 provides a means for displaying data to a user. For example, a computer monitor or television screen.

[0186] A computer program product may include a computer-readable storage medium (or media) having computer program instructions recorded thereon that configure a processor to carry out aspects of the present disclosure.

[0187] A computer-readable storage medium may be a tangible device capable of retaining and storing instructions for use by an instruction execution device. The computer-readable storage medium may be, for example, but is not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination thereof. More specific, non-limiting examples include portable computer diskettes, hard disks, RAM (random access memory), ROM (read-only memory), EPROM (erasable programmable ROM) or flash memory, SRAM (static RAM), portable CD-ROMs, DVDs, memory sticks, floppy disks, mechanically encoded devices (such as punch cards or ridge structures in grooves) and any suitable combination thereof. Note that, as used herein, computer-readable storage medium does not refer to transient signals per se, such as radio waves or other electromagnetic waves propagating through free space, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses propagating through optical fibers), or electrical signals transmitted over electrical wires.

[0188] The computer-readable program instructions described herein may be downloaded from a computer-readable storage medium to each computing / processing device, or may be downloaded to an external computer or external storage device via a network (e.g., the Internet, a local area network, a wide area network, and / or a wireless network). The network may include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the instructions for storage in a computer-readable storage medium within the respective computing / processing device.

[0189] The computer-readable program instructions for carrying out the operations of the present disclosure may be written in any combination of assembly language instructions, instruction set architecture (ISA) instructions, machine language instructions, machine-dependent instructions, microcode, firmware instructions, state-setting data, or source or object code in one or more programming languages ​​(such as object-oriented programming languages ​​like Smalltalk or C++, or traditional procedural programming languages ​​like C). These computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on both the user's computer and a remote computer, or entirely on a remote computer or server. In the latter scenario, the remote computer may be connected to the user's computer via any type of network, such as a local area network (LAN) or wide area network (WAN), or may be connected to an external computer (e.g., a computer on the Internet via an Internet Service Provider). In some cases, the electronic circuitry includes programmable logic circuitry, a field programmable gate array (FPGA), or a programmable logic array (PLA), which are individualized using state information from computer-readable program instructions and configured to perform aspects of the disclosure.

[0190] Aspects of the present disclosure are described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products. It will be understood that each block of each flowchart illustration or block diagram, or combinations thereof, can be implemented by computer-readable program instructions.

[0191] These computer-readable program instructions, when executed via a processor on a computer or other programmable data processing apparatus, may form means for implementing the functions / acts set forth in the flowchart diagrams and / or block diagrams. These instructions may also be stored on a computer-readable storage medium and configured as an article of manufacture for instructing a computer, programmable data processing apparatus, or other device to function in a particular manner. In other words, a storage medium having instructions stored thereon constitutes an article of manufacture for implementing the functions / acts set forth in the flowchart diagrams and / or block diagrams with the instructions.

[0192] These computer-readable program instructions may be loaded into a computer or other programmable data processing device or other device and cause the computer or other device to perform a series of operational steps to create a computer-implemented process, where execution of the instructions causes the computer or other device to perform the functions / acts shown in the flowchart illustrations and / or block diagrams.

[0193] The flowcharts and block diagrams shown in FIGS. 5 and 6 illustrate the architecture, functionality, and operation of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowcharts or block diagrams represents a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative embodiments, functions may occur out of the order depicted in the figures. For example, two blocks shown consecutively in the figures may actually occur substantially concurrently or in the reverse order, depending on the functionality involved. It should also be noted that each block in the block diagrams and / or flowchart diagrams, and combinations of blocks therein, may be implemented by a special-purpose hardware-based system that performs a particular function or process, or a combination of special-purpose hardware and computer instructions.

[0194] The programs described herein are identified in particular embodiments based on the application in which they are implemented. However, it should be noted that the names of the programs in this disclosure are used for convenience only and should not be construed as limiting the programs to a particular application.

[0195] For example, terms such as "Smalltalk" may be the subject of trademark rights in various jurisdictions around the world, and such terms are used herein only to refer to applicable products or services to the extent such trademark rights exist.

[0196] In some cases, the computer-implemented method may further include analyzing data collected using existing diagnostic testing techniques. For example, the method may include analyzing data collected by existing diagnostic testing techniques used to investigate cardiac disease, such as electrocardiograms (ECGs), Holter monitors, echocardiograms, exercise or stress tests, cardiac catheterizations, cardiac CT scans, and cardiac MRIs. For example, multiple parameters obtained from right heart catheterization (RHC) and transthoracic echocardiography (TTE) are used to classify heart failure patients into different groups. Specifically, TTE measurements provide cardiac output (CO) based on heart rate (HR), left ventricular outflow tract velocity time integral (LVOT VTI), and aortic valve cross-sectional area for each patient. RHC measures systolic and diastolic pressures in the right ventricle (RV) and pulmonary artery (PA), as well as CO, HR, and pulmonary capillary wedge pressure (PCW). A quantitative study was conducted to compare what these clinical data sets reveal about right and left ventricular hemodynamics, as well as systemic and pulmonary circulation, with closed-loop models of the circulatory system. One such closed-loop model is the TriSeg ventricular mechanical model, which incorporates the mechanical interactions between the left and right ventricular free walls and the interventricular septum (Lumens, J. et al., Ann. Biomed. Eng. 2009, Vol. 37, No. 11, pp. 2234-2255). Various representative model parameters are shown in Figure 2.

[0197] AASIST may be used to analyze right heart catheterization (RHC) and transthoracic echocardiography (TTE) data to estimate biomechanical parameters such as left ventricular (LV) diastolic stiffness and contractile force by optimizing output model parameters that best represent individual cardiovascular phenotypes. To this end, a genetic algorithm with a population size of 500 and a stall generation limit of 10 generations was used to identify output model parameters that match clinical parameters obtained from individual patients. After analyzing a large number of heart failure (HF) patients, supervised or unsupervised machine learning algorithms are used to classify patients into different groups, thereby training an AI system that can effectively classify patients. A method for estimating personalized model parameters from patient input parameters using the modified TriSeg model has been described in a previously published study. For example, clinical parameters, such as left ventricular end-systolic and end-diastolic volumes, right ventricular systolic and diastolic pressures, pulmonary artery systolic and diastolic pressures, mean pulmonary artery wedge pressure, systemic arterial systolic and diastolic pressures, systemic venous pulse pressure, and cardiac output measured by right heart catheterization, were used as input data for 79 patients. Although no clear differences in input parameters were observed in patients with HFpEF (dilated heart failure with EF), differences in the output model parameter, LV passive stiffness, led to the formation of a stiffer HFpEF2 group. AI classification identified at least three distinct phenotypes, including an unclassifiable group (see Figure 3). Furthermore, extending RHC and TTE measurements with and without exercise load allows for the estimation and analysis of pressure-volume loops under various inotropic conditions, enabling more accurate and effective assessment of the end-systolic pressure-volume relationship (ESPVR) and end-diastolic pressure-volume relationship (EDPVR).

[0198] Based on clinical diagnostic parameters, AASIST is used to create personalized in vitro heart failure disease models that match specific heart failure phenotypes. This disease modeling process consists of the following steps: a. Using a closed-loop cardiovascular computational model to individually estimate the mechanical and functional parameters of cardiovascular components (e.g., left ventricle (LV) and right ventricle (RV)). b. Training a machine learning algorithm using data from a large number of patients. c. Using machine learning to classify heart failure patients into subgroups based on the mechanical and functional parameters. d. Creating microhearts that recapitulate the functional phenotypes of the subgroups by optimizing tissue development conditions. Potential strategies include: i. Using cells or tissues isolated from patients belonging to the subgroup; ii. Applying disease-inducing factors (e.g., hyperglycemia) identified by blood tests; and iii. Analyzing cardiac tissue slices when human samples are available.

[0199] In this study, we first identify heart failure patient groups with specific phenotypes through AI-assisted patient stratification techniques using each patient's clinical data. Next, we apply a closed-loop cardiovascular model or other mechanistic systems biology model to define the tissue mechanical properties (e.g., stress, strain, and elastic modulus) of cardiovascular components, such as the left and right ventricular myocardium. By defining these tissue mechanical property parameters, clinical input parameters (e.g., pressure and volume) can be converted into in vitro measurable parameters using patient-derived or patient-isolated cells and tissues. By analyzing the clinical data of a large number of patients, machine learning algorithms can be trained to classify patients. In a pilot study, three patient groups were identified (Jones, E. et al., J. Physiol. 2021, 599(22): 4991-5013). Furthermore, left ventricular active contractility and passive stiffness are key factors in classification. Therefore, normalization of these parameters may lead to standardization of cardiovascular physiology.

[0200] Microheart tissues developed using cardiomyocytes differentiated from patient-derived induced pluripotent stem cells (iPSCs) and non-myocyte cardiac cells such as fibroblasts and endothelial cells can be used to evaluate changes in physiological and mechanical properties following interventions such as small molecule compounds, surgical procedures, and electrical stimulation (Figure 4A). The mechanical properties of tissue samples (artificial myocardial tissue or myocardial slices isolated from human or animal hearts) are measured by stretching the tissue and recording contractile force and baseline tension. Contractile force indicates cardiac contractility, while the baseline length-force relationship indicates tissue stiffness. Furthermore, measurements of action potentials, calcium transients, and metabolic activity (e.g., mitochondrial membrane potential, NAD / NADH ratio) reveal excitation-contraction-energy coupling in cardiac muscle or its equivalents. In addition to artificial myocardium, myocardial slices obtained from human donor hearts or isolated diseased hearts can also be used for the above analyses.

[0201] The method for determining drug candidates that reduce the increase in diastolic myocardial stiffness and the computer-assisted patient stratification method for identifying subgroups of heart failure patients generally involve the following steps: a. Identify a group of drug candidates or specific drug candidates that reduce the stiffness of 3D microtissues composed of myofibroblasts (e.g., human cardiac fibroblasts cultured in 10% fetal bovine serum) through, for example, complex compound library screening. For example, as described in the example below, one of the drug candidates, fasudil (hydrochloride hydrate), was shown to be effective in reducing the diastolic stiffness of left ventricular tissue. b. Computer-assisted patient stratification methods are used to estimate the clinical efficacy of drug candidates. For example, as described in the following example, fasudil was expected to have a clinical effect of improving cardiac reserve and exercise tolerance in patients with very high left ventricular stiffness parameters. As shown in Figure 15C, some patients with HFpEF (heart failure with preserved ejection fraction) have significantly higher left ventricular stiffness than other patients. Therefore, because fasudil administration is expected to reduce left ventricular stiffness, the efficacy of fasudil in the stratified patient group was predicted to be higher than that of the overall HFpEF patient population.

[0202] A similar approach can be applied to in vivo heart failure animal models to identify personalized models that match specific heart failure phenotypes based on patient clinical diagnostic parameters. This process involves the following steps: a. Using a closed-loop cardiovascular computational model, the mechanical and functional parameters of cardiovascular components such as the left ventricle (LV) and right ventricle (RV) are estimated individually. b. Training machine learning algorithms using large amounts of patient data. c. Classifying heart failure patients into subgroups based on the estimated parameters. d. A similar closed-loop cardiovascular computational model is used to estimate parameters in an animal model of heart failure. i. Compare the estimated disease parameters of animal models with those of humans. ii. Identify the phenotypic group that the animal model best represents. iii. Prepare cardiac microtissues or cardiac slices using cells or hearts isolated from the relevant animal model. iv. Screening a drug candidate library using the in vivo model and analyzing and confirming the efficacy of identified drug candidates.

[0203] Fasudil inhibits Rho kinase (ROCK). Abnormal ROCK activation has been shown to be involved in elevated vascular tone (J Cardiovasc Dis Res. 2010. 1(4): pp.165-170). A recent multicenter study, PROMIS-HFpEF, revealed a high prevalence of coronary microvascular dysfunction (CMD) in approximately 75% of HFpEF patients (Eur. Heart. J. 2018. 39(37):3439-3450), suggesting that ROCK activation is a predictor of CMD. Therefore, fasudil is expected to normalize abnormal ROCK activity and improve CMD. Furthermore, Figure 15C illustrates various levels of elevated systemic arterial stiffness in HFpEF patients. Because fasudil is expected to reduce vascular tone, HFpEF patients with abnormally high vascular stiffness combined with elevated left ventricular stiffness may respond better to fasudil than other HFpEF patients. Therefore, in clinical practice, the efficacy of fasudil in improving hemodynamic parameters and exercise tolerance in stratified patient groups may be higher than in general HFpEF patients.

[0204] Various animal models for heart failure exist. However, it is difficult to reproduce the diverse phenotypes of heart failure, especially using inbred animal models. The disease phenotype measured in each strain must be reproducible, and statistically significant differences in disease indicators compared to controls must be obtained. Therefore, each animal model should reproduce the disease phenotype common to a specific patient group. Through the above process, it is possible to identify the HF patient group that a particular animal model best reproduces (see Figure 4B). This approach is expected to improve the translatability of in vivo studies to human clinical practice. B. Closed-Loop Cardiovascular Model

[0205] In some examples, a closed-loop cardiovascular model was used, and its parameters were adjusted to match individual clinical parameters to accurately represent each patient's unique cardiovascular condition. The model parameters define the mechanical characteristics of the patient's cardiovascular physiology. Because the data was highly complex (11 dimensions), a nonlinear dimensionality reduction algorithm such as UMAP (Uniform Manifold Approximation and Projection) was used to simplify the data structure. This multidimensional data was then processed using a nonlinear dimensionality reduction algorithm such as UMAP. Subsequently, unsupervised machine learning algorithms such as DBSCAN (Density-Based Spatial Clustering of Applications with Noise) were used to perform clustering, allowing for the identification of distinct clusters.

[0206] The goal of patient data clustering is to identify distinct phenotypic groups based on multiple parameters characterizing cardiovascular function. These parameters include both mechanical and clinical factors, with a particular focus on: factors that change with disease progression (e.g., heart failure) and response to therapeutic intervention. Such an approach can identify patient groups that may be effectively modified by targeted therapies. V. Experimental Examples

[0207] Without intending to limit the scope of the present disclosure, various experimental examples related to the above-described embodiments have been prepared, the results of which are presented below. A. Drug candidates for heart failure

[0208] Fasudil (see Figure 1) is a small molecule inhibitor of Rho-kinase (ROCK) that regulates downstream signaling pathways and biological activities such as smooth muscle cell contractility. InvivoSciences' biomechanical phenotyping assay involves stretching bioengineered tissue constructs and recording isometric force. (See U.S. Patent No. 10,732,174.) These tissue constructs, reconstituted with human fibroblasts, mimic connective tissue. When reconstituted with myofibroblasts, they mimic connective tissue with fibrosis. Using this assay and tissue constructs, fasudil was observed to reduce contractile force and tissue stiffness.

[0209] When the same biomechanical assay was applied to micro-heart tissue constructs, fasudil treatment resulted in a dose-dependent decrease in stiffness (Figure 7B). Fasudil did not alter the myocardial contractility of the micro-heart tissue (see Figure 7A). This micro-heart tissue was generated using induced pluripotent stem cell (iPSC)-derived human cardiomyocytes and primarily non-myocyte cells, such as fibroblasts. Therefore, fasudil is considered a promising drug candidate to reduce elevated diastolic stiffness in patients with heart failure with preserved ejection fraction (HFpEF). Myocardial fibrosis is observed in various types of cardiomyopathies and heart failure (HF). For example, myocardial interstitial fibrosis contributes to left ventricular dysfunction and leads to the development of heart failure. Fibrotic myocardial tissue is stiffer than healthy tissue, causing diastolic dysfunction. Fasudil acts by targeting ROCK (Rho kinase), which regulates the activity of nonmuscle myosin independently of calcium homeostasis. Human cardiac fibroblasts express ROCK1 and ROCK2, both of which regulate the contractility of nonmuscle cells. In particular, inhibition of ROCK2 significantly suppressed the expression of anti-fibrotic genes and proteins (see Figure 7A-B). Furthermore, fasudil treatment reduced the expression of smooth muscle markers in human cardiac fibroblasts. B. Creation of Individualized Fibrotic and Mature Cardiac Microtissues

[0210] The cells that make up the heart include cardiomyocytes (CMs), cardiac fibroblasts (cFBs), endothelial cells (ECs), smooth muscle cells (SMCs), and immune cells (ICs). These cell types can be generated using various differentiation protocols using iPS cells derived from human patients or donors. Many differentiation protocols have been automated and optimized for mass production. Patient-derived iPS cells are useful for recapitulating specific heart failure susceptibility and disease phenotypes. Myocardial tissue models generated using these iPS cells can recapitulate disease phenotypes such as hypertrophic cardiomyopathy and diastolic cardiomyopathy. Well-studied genetic mutations are registered in databases such as ClinVar. However, the relationship between disease phenotypes and genetic background remains unclear, and further research is expected to clarify this. Nevertheless, iPS cells and their derivatives obtained from patient populations predicted to develop specific types of heart failure are extremely useful for in vitro disease modeling and drug discovery research.

[0211] Mature microcardial tissue was produced in a 96-well format using a semi-automated, multi-point quality control (QC) process. Patient-derived or gene-edited iPS cells passed quality tests for pluripotency and purity before proceeding to the next step. This robust differentiation protocol efficiently produced cardiomyocytes and non-myocytes. Various parameters were optimized to obtain highly pure (>95%) cardiomyocytes and cardiac non-myocytes. The ACTRO® robot, differentiation media, and other differentiation techniques were used for cardiomyocyte differentiation and maintenance. Figures 9A-9D show the mass-produced microcardial tissue and its mature phenotype, and Figures 10A-10D evaluate its response to various positive inotropic and chronotropic drugs. To grow mature microcardial tissue, the following was applied: physiological stretch cycles (15% strain ≈ sarcomere length 2.13 μm), low-energy biphasic electrical stimulation (5–10 V, 1 ms duration), and maturation medium containing fatty acids, triiodothyronine, dexamethasone, taurine, creatine, L-carnitine, etc. at appropriate concentrations.

[0212] Myocardial fibrosis is more pronounced in mature microcardial tissue exhibiting spontaneous beating heartbeats in patients with HFpEF. **TGF-β (Transforming Growth Factor β)** is a major profibrotic factor. Microcardial tissue treated with TGF-β exhibited activated fibroblasts (myofibroblasts) stained with α-smooth muscle actin antibodies, and increased diastolic stress (stiffness) and myocardial contractility were observed (see Figure 11A-J). Mature microcardial tissue exhibited positive inotropic and chronotropic responses to isoproterenol (ISO), but chronic TGF-β treatment inhibited these ISO-induced positive responses (see Figure 10). Western blot analysis confirmed increased expression of α-smooth muscle actin. These data indicate that TGF-β treatment results in the following phenotypes: increased diastolic stiffness, hypercontractility, and loss of myocardial contractile reserve (decreased responsiveness to inotropic and chronotropic stimuli). These characteristics, i.e., increased diastolic stiffness and loss of contractile reserve, are common features in patients with HFpEF. Some patients with HFpEF (including those with hypertrophic cardiomyopathy) may exhibit a hypercontractile phenotype. The TGF-β signaling pathway is also well known in the pathogenesis of diabetes mellitus (DM). Furthermore, in rat models, the presence of diabetes is typically observed in the setting of overexpression of TGF-β and cardiac hypertrophy. It has been reported that TGF-β induces the induction of fetal-type genetic programs. Therefore, human iPS cell-derived microcardial tissue (NuHeart®) will be used in this study based on TGF-β treatment. C. Mechanical Assay

[0213] To assess the myocardial contractility and stiffness of microcardial tissues, we used a force probe stretching method to record the length-tension relationship. A microforce transducer (see Figure 11) mounted on the Palpator® mechanical assay device measured the resistance required to stretch microcardial tissues cultured in 96-well plates. Using a probe pulled by a motor-driven arm, the myocardial contractile force and basal tension of myocardial tissue samples were measured at various tissue lengths. Sarcomere length was calibrated based on the stretch ratio (percentage of stretch relative to the original length) according to a protocol described previously (Asnes, CF et al. Biophys. J. 2006. 91(5): pp. 1800-1810). Furthermore, biological indices such as calcium transients (CaTs) and action potentials (APs) were measured (by Fourier analysis of sarcomere patterns in fixed microcardial tissues). These were recorded using indicator dyes and a high-throughput microplate reader (see Figure 13A-C). As an example, we introduced the cardiac troponin T R92Q mutation (familial hypertrophic cardiomyopathy) into micromyocytes generated from gene-edited iPSC-derived cardiomyocytes. This hypertrophic cardiomyopathy (HCM) is known to exhibit a hypercontractile phenotype, and this FCH (family hypertrophic cardiomyopathy) micromyocyte exhibited hypercontractility as evidenced by increased myocardial stress in length-tension plots (Figure 14A-B). Additionally, this micromyocyte exhibited increased diastolic stiffness, a characteristic observed in hypertrophic cardiomyopathy. D. Stratification of HFpEF Patients

[0214] The physiological and pathological heterogeneity of HFpEF poses a significant challenge for validating the efficacy of drug candidates, especially in large-scale Phase III clinical trials. This technology classifies HFpEF patients and stratifies them into distinct groups through AI-assisted clustering of diagnostic parameters. E. Diagnostic Parameters

[0215] This stratification technique begins by analyzing well-defined and widely accepted diagnostic parameters, such as echocardiography, magnetic resonance imaging (MRI), and cardiac catheterization. These parameters include right ventricular systolic and diastolic pressures, as well as other pulsatile characteristics and pressure measurements. Other parameters measured include pulmonary artery pressure, mean pulmonary capillary wedge pressure (PCW pressure), systolic and diastolic systemic blood pressure, heart rate during catheterization, cardiac output (CO), weight, height, and sex.

[0216] To measure these parameters, a Swan-Ganz catheter, for example, is inserted through the jugular vein and pressure is measured until the catheter tip reaches the pulmonary artery. In addition to pressure information, cardiac output (CO) can be estimated by thermodilution or Fick's method (JAMA Cardiol. 2017. 2(10): pp. 1090-1099). Thermodilution estimates CO by injecting a bolus of chilled saline into the proximal end of the catheter and measuring its diffusion at the distal end. Fick's method estimates CO by measuring venous and arterial oxygen saturation and whole-body oxygen consumption (VO2) based on weight, height, and sex.

[0217] Transthoracic echocardiography (TTE) is one of the modalities used to analyze the following: ventricular volumes (systolic and diastolic left ventricular volumes) and heart rate. Left ventricular volumes are measured using either the single-diameter method, which uses the left ventricular internal diameter just below the mitral valve leaflets, or the volumetric method, which traces the left ventricular contour from apical two- and four-chamber views. The single-diameter method considers the left ventricle as a truncated prolate spheroid and estimates volume based on the nonlinear relationship between internal diameter and ventricular length. Volumetric methods from two- and four-chamber views use the Simpson method (disk method). If the image quality is deemed suitable for the Simpson method, a cardiologist evaluates the following: heart rate and time-integrated flow velocity (VTI) in the left ventricular outflow tract (LVOT), which then estimates cardiac output (CO). Cardiac magnetic resonance imaging (CMR) is considered the gold standard for measuring ventricular volumes and fibrosis. If necessary, CMR may be performed in this analysis. F. Cardiovascular Model

[0218] Various closed-loop cardiovascular models have been introduced. The complexity of the model depends on the computational resources used and the amount of data available to define the model input parameters. The following simplifications were made: the pericardial compartment was removed, and the zero-pressure volume (dead space volume) in each vascular and ventricular compartment was set to zero. A coefficient was also introduced to simulate blood volume remobilization in heart failure, based on the assumption that 30% of the nominal total blood volume is remobilized as "stressed blood volume" (Fudim et al., J. Am. Heart Assoc. 2017, 6(8), e006817). However, no additional blood volume was remobilized in either HFpEF or HFrEF patients in this study. The mathematical formulation of the simplified cardiovascular model used in this study has been previously described (Smith, BW et al., Med. Eng. Phys. 2004, 26(2): pp. 131-139). Mathematical representations of other types of closed-loop cardiovascular systems are also available.

[0219] In this example model, 16 model parameters (Figure 2) were adjusted during the optimization process. The descriptions of these parameters are shown in Table 1 of Figure 2. The following steps were taken to define some of the parameters: 1) the left ventricular elastic modulus was calculated from the measured volume and estimated systolic pressure; 2) the left ventricular diastolic stiffness was calculated from the measured volume and estimated diastolic pressure; 3) the right ventricular elastic modulus was calculated from the measured pressure and estimated systolic volume; 4) the right ventricular diastolic stiffness was calculated from the measured pressure and estimated diastolic volume; 5) the systemic vascular elastic modulus was calculated based on the measured and estimated pulse pressure in the arterial compartment; and 6) the pressure and estimated stress volume in the venous compartment were determined. 7) Pulmonary vascular elasticity was calculated from measured and estimated pulse pressure and estimated stress volume. 8) Systemic and pulmonary vascular resistance was calculated from measured systemic and pulmonary mean arterial pressure, estimated systolic venous pressure, and cardiac output measured by right heart catheterization (RHC). Total blood volume was calculated based on each patient's height, weight, and sex according to the method reported by Nadler et al. (Surgery. 1962. 51(2): pp. 224-32). The initial distribution of stressed and unstressed blood volume was determined based on the study by Beneken (Reeve, BM et al., American Heart Journal. 1968. 75(3): pp. 432-433), assuming an overall stress volume of 18.75%. G. Parameter Optimization

[0220] For each patient, adjustable parameters were estimated by minimizing the error between the simulated and measured values ​​using the least squares error method for the following parameters: right ventricular pressure (systolic and diastolic), pulmonary artery pressure (systolic and diastolic), mean pulmonary capillary wedge pressure, systemic arterial pressure (systolic and diastolic), cardiac output (CO) during right heart catheterization (RHC), left ventricular volume (systolic and diastolic), and CO during transthoracic echocardiography (TTE).

[0221] Because heart rates may vary during right heart catheterization (RHC) and transthoracic echocardiography (TTE), separate simulations were performed for each condition. However, the same set of parameters was used for both simulations, as it was assumed that parameters representing cardiac function within a single patient would not change significantly between procedures. Clinical measurements were evaluated by adjusting the data and model to minimize error while maintaining the validity of the cardiac function model specific to each patient.

[0222] Calculations were performed throughout the cardiac cycle after pulsatile pressure and blood flow reached steady state. This steady state was ensured by running a 50-beat simulation. After steady state was reached, the maximum and minimum values ​​of the pressure and volume data over the last five beats were used to calculate the residual error. Capillary wedge pressure and cardiac output (CO) were expressed as average values ​​over the cardiac cycle; therefore, these values ​​were calculated as cardiac cycle averages and then compared with right heart catheterization (RHC) and transthoracic echocardiography (TTE) measurements. Adjustable parameter estimation was optimized using a genetic algorithm implemented in MATLAB. H. Artificial Intelligence

[0223] After optimizing and determining the model parameters representing the individual models, three different clustering methods were used to classify individuals into groups based on similar characteristics. Clinical data and optimized parameter values ​​were organized into matrices of 0 and 1, respectively, with each row representing a specific patient and each column representing a clinical measurement or optimized parameter value. Before applying the clustering methods, mean-centering was performed by subtracting the mean of each column from its elements. Clinical data and optimized parameters with different units were also normalized by dividing by their respective standard deviations. After conducting principal component analysis (PCA) to identify the principal components of the multidimensional data, the HFpEF stratification algorithm extracted identifiable groups using various clustering methods, including k-means clustering and hierarchical clustering. This approach is called "AI-Assisted, Systems-biology Integrated Patient Stratification Technology (AASIST)" for identifying specific HFpEF patient groups. After the parameters representing the individual models were optimized and determined,

[0224] As the number of HFpEF patients increased, the training data for the stratification algorithm AASIST was expanded, enabling more reliable differentiation of HFpEF patient groups. This training was performed by analyzing existing data from an electronic health record (EHR) database. A fully trained AASIST algorithm groups HFpEF patients based on their cardiovascular function. Application of AASIST to HFpEF patient data showed that left ventricular active contractility and passive stiffness significantly differed between the clustered groups. Furthermore, fasudil has been shown to reduce passive stiffness in iPS cell-derived human myocardial microtissues. Therefore, treatment with fasudil or other ROCK inhibitors (Rho kinase inhibitors) may improve the high passive stiffness of HFpEF patients stratified by AASIST. Using AASIST for such patient selection is expected to enhance the efficacy of fasudil and other ROCK inhibitors. I. Gene Expression Analysis

[0225] Cardiac amyloidosis involves the deposition of monoclonal immunoglobulin (AL amyloid) and transthyretin (TTR amyloid), which may lead to HFpEF. Screening of HFpEF patients has revealed that approximately 13% of hospitalized patients with HFpEF are biopsy-confirmed to be TTR-positive, and approximately 17% of autopsy samples from patients diagnosed with HFpEF antemortem have also shown TTR deposits. Histological detection of amyloid deposits in tissue biopsy specimens is generally the only means of diagnosing amyloidosis. An exception is cardiac transthyretin amyloidosis (ATTR-CM), which may be diagnosed without a biopsy if strict diagnostic criteria are met. With the recent approval of Tafamidis for the treatment of cardiac TTR amyloidosis, cardiologists are now implementing diagnostic tests to detect amyloid deposits.

[0226] In patients with HFpEF, including those with amyloid-positive disease, even those with amyloid-negative disease, plasma levels of N-terminal pro-B-type natriuretic peptide (NT-proBNP) are elevated above approximately 2,000 ng / L, suggesting an advanced stage of heart failure (HF). It is unclear whether NT-proBNP levels correlate with the degree of myocardial fibrosis. Gene expression analysis of patients with HFpEF with elevated NT-proBNP levels but no amyloid deposits has confirmed activation of characteristic signaling pathways, including the MAPK signaling pathway, the Hippo signaling pathway, and regulation of the actin cytoskeleton. (Source: Hahn, VS et al., Circulation. 2021, 143(2): pp. 120-134) These signaling pathways are involved in regulating fibrosis, and therefore, treatment with ROCK inhibitors (e.g., fasudil) or other antifibrotic compounds may be effective in reducing fibrosis. Furthermore, gene expression analysis of biopsy samples may identify a panel of 1–10 genes that identifies a group of patients likely to respond to fasudil treatment. J. Combining AASIST with gene expression analysis

[0227] By stratifying HFpEF patients by AASIST and combining it with gene expression analysis, we expect the results to largely overlap. However, by combining the results of both analyses, we can further narrow down the HFpEF patient population that will maximize the therapeutic benefit in clinical trials. K. Rho kinase (ROCK)-targeted inhibitors

[0228] As shown in Figure 1, inhibition of ROCK activity exhibits pharmacological effects similar to those of fasudil. Other small molecule inhibitors targeting ROCK include Thiazovivin, Y-27632 2HCl, GSK419286A, Ripasudil, RKI-1447, and Azaindole 1 (also known as TC-S 7001), all of which are known to exhibit similar pharmacological effects for the treatment of myocardial fibrosis. L. Closed-Loop Cardiovascular Model

[0229] Various types of Closed-Loop Cardiovascular Models are available. Non-limiting examples are provided below. 1. Windkessel Models

[0230] The Windkessel model is one of the simplest cardiovascular models. These models represent the compliance (elasticity) and resistance of blood vessels by analogy with electrical circuits and are used to model blood pressure and blood flow, especially in large blood vessels such as the aorta. The Windkessel model primarily focuses on the behavior of the arterial system. 2. Lumped Parameter Models

[0231] The lumped parameter model divides the cardiovascular system into several sections and defines the resistance, compliance, and inertia (inductance) for each section. It allows for a more detailed representation than the Windkessel model and is sometimes used for dynamic analysis of the entire cardiovascular system. 3. Distributed Parameter Models

[0232] Distributed parameter models take into account spatial variations along the blood vessel. They use partial differential equations to describe blood flow and pressure and are more complex and detailed than lumped parameter models. They are particularly useful for studying wave propagation phenomena in arteries. 4.0-dimensional (0D), 1-dimensional (1D), 2-dimensional (2D), and 3-dimensional (3D) models

[0233] These are numerical models with different spatial dimensions. For example, 0D models are similar to lumped parameter models, while 1D models represent axial changes within a blood vessel. Furthermore, 2D and 3D models are sometimes used in computational fluid dynamics (CFD) studies to accurately reproduce blood flow and vessel wall dynamics. 5.Patient-Specific Models

[0234] This model is specialized for the cardiovascular physiology of each individual patient and is constructed using medical imaging data, etc. It is used in personalized medicine for the purposes of diagnosis and treatment. 6. Multi-Scale Models

[0235] Multiscale models integrate different levels of physiological detail across the cellular, tissue, whole-organ, and system levels. These models are used, for example, to study how changes at the molecular or cellular level affect overall cardiovascular function. 7. Hybrid Models

[0236] Hybrid models combine elements of lumped and distributed parameter models, or computational fluid dynamics (CFD) and structural mechanics models (fluid-structure interaction), and are used to integrate multiple aspects of cardiovascular physiology for complex simulations.

[0237] To match individual clinical parameters, a closed-loop cardiovascular model, as described above, was used, and its parameters were adjusted to accurately represent each patient's unique cardiovascular condition. The model parameters define the mechanical characteristics of each patient's cardiovascular physiology. Because this data is complex, with 11 dimensions, a nonlinear dimensionality reduction algorithm, specifically UMAP (Uniform Manifold Approximation and Projection), was used to simplify the data structure. Unsupervised machine learning algorithms, such as Density-Based Spatial Clustering of Applications with Noise (DBSCAN), were then applied to this dimensionality-reduced data to identify distinct clusters.

[0238] As previously mentioned, the goal of clustering patient data is to identify distinct phenotypic groups based on a combination of parameters characterizing cardiovascular function. These parameters may include both mechanistic and clinical factors, with a particular emphasis on parameters that change with the progression of heart failure and response to therapeutic interventions. This approach allows for the identification of specific patient populations that can be effectively targeted for treatment based on these parameters.

[0239] The algorithms used, UMAP (Uniform Manifold Approximation and Projection) and DBSCAN (Density-Based Spatial Clustering Application with Noise), consisted of several adjustable parameters to optimize the structuring and clustering of the data. A cost function was defined, which measures the variability of selected parameter values ​​in each cluster. These parameters included: left ventricular contractility, left ventricular stiffness, systemic arterial stiffness, mean pulmonary artery pressure (mPAP), and pulmonary artery wedge pressure (PAWP). The goal of the optimization process was to minimize this cost function, which allows identifying clusters with minimal variability in parameter values.

[0240] The analysis shown in Figures 15A–C demonstrates an example of stratification of heart failure patients. 137 heart failure patients were classified into 19 phenogroups based on a combination of mechanical and clinical parameters. As mentioned above, heart failure patients who underwent invasive right heart catheterization (RHC) were considered to have a high probability of pulmonary hypertension (PH). Right heart catheterization (RHC) parameters analyzed included mean pulmonary artery pressure (mPAP) and pulmonary artery occlusion pressure (PAWP). Patient data included cases with ejection fractions of less than 50% and greater than 50%, but many patients had mean pulmonary artery pressures >25 mmHg and pulmonary artery occlusion pressures >15 mmHg. According to classical diagnostic criteria, these patients could be classified as having pulmonary hypertension (PH) with heart failure with preserved ejection fraction (HFpEF). Left heart disease (LHD), both heart failure with reduced ejection fraction (HFrEF) and heart failure with preserved ejection fraction (HFpEF), is one of the most common causes of pulmonary hypertension, with a reported prevalence of 23%–80% in patients with LHD (Al-Omary, MS et al., Hypertension, 2020, 75(6): pp. 1397–1408). It has also been shown that elevated left ventricular filling pressures can lead to combined precapillary and postcapillary pulmonary hypertension (CpcPH) in patients with heart failure with preserved ejection fraction (HFpEF).

[0241] The pulmonary arterial hypertension (PH) phenotype causes pulmonary arterial vasoconstriction and remodeling. Patients classified by right heart catheterization (RHC) data of mPAP ≥ 25 mmHg and PAWP > 15 mmHg are consistent with trends reported in the literature (Dixon, DD et al., Heart Failure Reviews, 2016, 21(3): pp. 285-297). Similarly, in this analysis, patients with RHC data of mPAP ≥ 25 mmHg and PAWP > 15 mmHg were present in both reduced ejection fraction (HFrEF) and preserved ejection fraction (HFpEF) subgroups, but were predominantly overrepresented in HFpEF. The analysis shown in Figure 15C demonstrates an example of a phenotype-based stratification approach, focusing on two HFpEF subgroups (Groups 3 and 4)**, indicated by the solid and striped arrows in Figure 15C. While these two groups had similar left ventricular contractility and systemic arterial stiffness, there were differences in left ventricular stiffness. Although this preliminary analysis was based on a limited data set, it suggests the possibility of identifying distinct patient subgroups that are often overlooked by traditional classification methods. In particular, the patient group indicated by the solid arrow may benefit more from fasudil treatment in improving left ventricular stiffness by affecting the contractility of left ventricular myofibroblasts. M. The Significance of Stratification in Clinical Trials (Enrichment by Clustering)

[0242] Patient stratification is particularly useful for targeted therapies known to modify specific cardiovascular parameters. For example, for antihypertensive drugs that reduce systemic arterial stiffness, clinical trials may be conducted in patients with significantly elevated levels of this parameter. Compared to the non-stratified group, a statistically significant difference may be obtained with a smaller number of patients. For drugs like fasudil that simultaneously modify multiple parameters, such as left ventricular contractility, left ventricular stiffness, and systemic arterial stiffness, stratification is expected to further increase the power to detect treatment effects.

[0243] Therefore, clinical trials should specifically select patients with elevated left ventricular contractility, left ventricular stiffness, and arterial stiffness. This differs from the conventional approach of indiscriminately targeting different clusters of patients. As an example of this patient stratification process, we performed the following experiment. Engineered heart tissues (EHTs) were generated from human induced pluripotent stem cell (iPSC)-derived cardiomyocytes (10% cardiac fibroblasts). Five days after generation, half of the EHTs were cultured in normal culture (Figures 16A and 16C), while the other half were treated with 1 ng / mL TGF-β1 (Figures 16B and 16D). TGF-β1 treatment increased passive stiffness in EHTs containing iPSC-derived cardiac fibroblasts. This effect was clearly demonstrated by an increase in the slope of the passive stress-strain curve. In particular, this trend can be confirmed by the data points for 0 μM fasudil, indicated by the black circles in Figures 16C and 16D.

[0244] Fifteen days after preparation, EHTs were administered 1, 3, and 10 μM fasudil, with or without TGF-β1 treatment (see Figures 16A–16D). According to a literature review (Satoh et al., Life Sci, 2001, 69(12): pp. 1441–1453), the plasma concentration of intravenously administered fasudil is approximately 0.36 μM. Therefore, a high dose of 10 μM may be too high and may significantly reduce myocardial contractility by day 29. Nevertheless, fasudil effectively reduced the passive stiffness of EHTs with or without TGF-β1. However, the effect was more pronounced in TGF-β1-treated tissues. Furthermore, the effect of fasudil on myocardial contractility was not significant 1 day after administration (data not shown). However, by day 29, myocardial contractility was confirmed to be reduced regardless of the presence or absence of TGF-β1.

[0245] In summary, short-term administration of fasudil (<24 hours) effectively reduced passive stiffness of EHT without significantly affecting myocardial contractility. On the other hand, long-term administration (3-14 days) reduced both myocardial contractility and passive stiffness of EHT. The underlying mechanism for these effects is Rho-kinase inhibition by fasudil. However, while Rho-kinase inhibition may have a direct effect on myocardial contractility, previous studies have shown that it had minimal effects on contractility in human ventricular muscle strips (Grimm, M. et al., Cardiovasc Res, 2005, 65(1): pp. 211-220).

[0246] These results suggest the existence of a mechanical coupling between the myocardial contractile units and the nonlinear passive elastic component. These components are thought to be arranged in parallel and in series. The decrease in passive stiffness may be due to a decrease in myofibroblast contractility, and this decrease in the passive mechanical component may also lead to a decrease in myocardial contractility. This may be related to the enhancement of myocardial contractility through increased myofibroblast contractility and fibrotic remodeling of the extracellular matrix.

[0247] From a clinical perspective, fasudil may be most effective in heart failure patients with increased diastolic stiffness and myocardial contractility, such as the patient group indicated by the solid arrow in Figure 15C. Furthermore, fasudil's ability to reduce smooth muscle contractility makes it a promising treatment for vascular and pulmonary hypertension. Thus, targeting heart failure patients whose symptoms are most responsive to fasudil may improve therapeutic outcomes. This patient enrichment strategy is useful for identifying individuals who are most likely to benefit from fasudil. Similar strategies can be utilized to maximize the therapeutic effects of other treatments.

[0248] The detailed description and accompanying examples set forth herein are merely illustrative and do not limit the scope of the invention, which is defined solely by the appended claims and equivalents thereof.

[0249] Various changes and modifications to the embodiments described herein will be apparent to those skilled in the art, including, without limitation, changes and modifications to chemical structures, substituents, derivatives, intermediates, synthetic methods, compositions, formulations, methods of use, and the like related to the present invention, provided that these changes do not depart from the spirit and scope of the present invention.

[0250] For reference purposes, various aspects of the present invention are set forth in the following numbered claims: Clause 1: A method of treating heart failure with preserved ejection fraction (HFpEF), comprising: Methods include: Administering a therapeutically effective amount of a Rho kinase inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof to a subject having: -diastolic dysfunction -cardiac fibrosis -pulmonary hypertension -Left ventricular ejection fraction 50% However, cardiac amyloid deposits are absent. Clause 2: The method of clause 1, wherein the Rho kinase inhibitor is any of the following: Fasudil Thiazovivin Y-27632 2HCl Ripasudil GSK429286A RKI-1447 Azaindole 1 (TC-S 7001) JPEG2026504406000054.jpg183166 Clause 3: The method according to clause 1 or 2, wherein the Rho kinase inhibitor is fasudil. Clause 4: The method according to any one of clauses 1 to 3, wherein the subject is a human. Clause 5: The method according to any one of clauses 1 to 4, wherein the subject is an animal. Clause 6: A method according to any one of claims 1 to 5, wherein before administering a therapeutically effective amount of a Rho kinase inhibitor, an antifibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof to a subject, selecting the subject and generating an instruction set for administering the Rho kinase inhibitor, the antifibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof is performed by a computer-implemented method. Clause: The method according to any one of clauses 1 to 5, wherein the computer-implemented method comprises: 1. An implementation of a feedforward artificial neural network by one or more processors, comprising: receiving, by one or more processors, a plurality of patient parameters, the plurality of patient parameters including: - 1 or more patients with diastolic dysfunction and 1 or more patients without diastolic dysfunction, - One or more patients with myocardial fibrosis and one or more patients without myocardial fibrosis, - 1 or more patients with pulmonary hypertension and 1 or more patients without pulmonary hypertension, - At least one patient with a left ventricular ejection fraction (LVEF) of 50% or greater and at least one patient with a LVEF less than 50%. - One or more patients with myocardial amyloid deposits and one or more patients without myocardial amyloid deposits. analyzing the plurality of patient parameters by one or more processors; by one or more processors, mapping the plurality of patient parameters to neurons in an input layer of an artificial neural network; applying, by one or more processors, weights to the plurality of patient parameters disposed in the input layer; placing, by one or more processors, the weighted plurality of patient parameters into one or more hidden layers of the feedforward artificial neural network; generating, by one or more processors, a plurality of optimized patient parameter data sets; · Receiving, by one or more processors, the optimized plurality of patient parameter data sets, wherein the plurality of optimized patient parameter data sets includes data for a plurality of patients. analyzing, by one or more processors, the plurality of optimized patient parameter data sets; identifying, by one or more processors, a plurality of populations from the plurality of optimized patient parameter data sets; classifying, by one or more processors, the plurality of populations into one or more clusters. analyzing, by one or more processors, one or more clusters produced by said cluster analysis; generating, by one or more processors, an administration instruction set for administering a therapeutically effective amount of a Rho kinase inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof to a subject based on the results of the analysis. Clause 8: The method of clause 7, wherein the plurality of patient parameters comprises data and metadata. Clause 9: The method according to clause 7 or 8, wherein the plurality of patient parameters further comprises age, weight, sex, height, and blood glucose level. Clause 10: A computer-implemented method, the computer-implemented method including: Execution of the feedforward artificial intelligence or machine learning algorithms by one or more processors includes: receiving a plurality of patient parameters by one or more processors; wherein the plurality of patient parameters includes: - one or more patients with diastolic dysfunction and one or more patients without diastolic dysfunction, - one or more patients with cardiac fibrosis and one or more patients without cardiac fibrosis, - one or more patients with pulmonary hypertension and one or more patients without pulmonary hypertension, - At least one patient with a left ventricular ejection fraction (LVEF) of 50% or greater and at least one patient with a LVEF less than 50%. - One or more patients with cardiac amyloid deposits and one or more patients without cardiac amyloid deposits. analyzing, by one or more processors, the plurality of patient parameters; by one or more processors, placing the plurality of patient parameters into neurons in an input layer of an artificial neural network; applying, by one or more processors, weights to the plurality of patient parameters disposed in the input layer; placing, by one or more processors, the weighted plurality of patient parameters into one or more hidden layers of the feedforward artificial neural network; · generating, by one or more processors, a plurality of optimized patient parameter data sets; Running cluster analysis on one or more processors, Performing the cluster analysis includes: receiving, by one or more processors, the optimized plurality of patient parameter data sets; Here, the plurality of optimized patient parameter data sets includes data for a plurality of patients. analyzing, by one or more processors, the plurality of optimized patient parameter data sets; identifying, by one or more processors, a plurality of populations from the plurality of optimized patient parameter data sets; Classifying, by one or more processors, the plurality of populations into one or more clusters. analyzing, by one or more processors, one or more clusters produced by said cluster analysis; generating, by one or more processors, an administration instruction set for administering to a subject: Rho kinase inhibitors, anti-fibrotic compounds, or · Pharmaceutically acceptable salts, esters, or solvates thereof. The subject has: Diastolic dysfunction, Myocardial fibrosis, Pulmonary hypertension, ·Left ventricular ejection fraction 50% or more, However, the subject does not have myocardial amyloid deposits. and administering to the subject a therapeutically effective amount of a Rho kinase inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof. Clause 11: The computer-implemented method of clause 10, wherein the Rho kinase inhibitor is selected from: JPEG2026504406000055.jpg151163 Fasudil, Thiazovivin, Y-27632 2HCl, Ripasudil, GSK429286, ARKI-1447, Azaindole 1 (TC-S 7001). Clause 12. The computer-implemented method of clause 10 or 11, wherein the Rho kinase inhibitor is fasudil. Clause 13. The computer-implemented method of any one of clauses 10 to 12, wherein the subject is a human. Clause 14. The computer-implemented method of any one of clauses 10 to 12, wherein the subject is an animal. Clause 15. The computer-implemented method of any of clauses 10-14, wherein the plurality of patient parameters includes data and metadata. Clause 16. The computer-implemented method of any of clauses 10-15, wherein the plurality of patient parameters further comprises age, weight, sex, height, and blood glucose level. Clause 17: A method of reducing stiffness of cardiac tissue, comprising the steps of: A therapeutically effective amount of a Rho kinase inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof is administered to a subject with diastolic dysfunction, myocardial fibrosis, pulmonary hypertension, and a left ventricular ejection fraction of 50%, provided that the subject does not have cardiac amyloid deposits. Clause 18. The method according to clause 17, wherein the Rho kinase inhibitor is one of the following: JPEG2026504406000056.jpg156168 Fasudil Thiazovivin Y-27632 2HCl Ripasudil GSK429286A RKI-1447 Azaindole 1 (TC-S 7001) Clause 19: The method according to clauses 17 and 18, wherein said Rho kinase inhibitor is Fasudil. Clause 20. The method of clauses 17-19, wherein prior to administering to a subject a therapeutically effective amount of a Rho kinase inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof, instructions directing said administration are generated by a computer-implemented method.

Claims

1. 1. A method of treating heart failure with preserved ejection fraction (HFpEF), comprising: Administering a therapeutically effective amount of a Rho kinase inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof to a subject: Diastolic dysfunction Cardiac fibrosis Pulmonary hypertension Left ventricular ejection fraction greater than 50%; and left ventricular ejection fraction greater than 50%. The subject does not have cardiac amyloid deposits.

2. The method of claim 1, wherein the Rho kinase inhibitor is:

3. The method of claim 1, wherein the Rho kinase inhibitor is Fanidil.

4. The method of claim 1, wherein the subject is a human.

5. The method of claim 1 , wherein the subject is an animal.

6. 2. The method of claim 1, wherein the set of instructions for administering the therapeutically effective amount of a Rho kinase inhibitor, the anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof to the subject is generated by a computer-implemented method prior to administering the therapeutically effective amount of a Rho kinase inhibitor, the anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof to the subject.

7. The method of claim 6, wherein the computer-implemented method includes: executing, by one or more processors, a feedforward artificial neural network, said executing comprising: receiving, by one or more processors, a plurality of patient parameters, the plurality of patient parameters comprising: one or more patients with diastolic dysfunction and one or more patients without diastolic dysfunction; one or more patients with cardiac fibrosis and one or more patients without cardiac fibrosis; one or more patients, including one or more patients with pulmonary hypertension and one or more patients without pulmonary hypertension one or more patients with a left ventricular ejection fraction greater than 50% and one or more patients without a left ventricular ejection fraction greater than >50%; and one or more patients with a left ventricular ejection fraction greater than 50%. one or more patients with cardiac amyloid deposits and one or more patients without cardiac amyloid deposits; analyzing, by one or more processors, the plurality of patient parameters; and mapping, by one or more processors, the plurality of patient parameters to neurons in an input layer of an artificial neural network; applying, by one or more processors, weights to the plurality of patient parameters disposed in the input layer; placing, by one or more processors, the weighted plurality of patient parameters into one or more hidden layers of the feedforward artificial neural network; and generating, by one or more processors, a plurality of optimized data sets of patient parameters; performing, by one or more processors, a cluster analysis, said performing comprising: receiving, by one or more processors, a plurality of optimized data sets of patient parameters; wherein the plurality of optimized data sets of patient parameters comprises data for a plurality of patients. analyzing, by one or more processors, the plurality of optimized data sets of patient parameters; identifying, by one or more processors, a plurality of populations of the plurality of optimized data sets of patient parameters; generating, by one or more processors, one or more clusters of the plurality of populations; analyzing, by one or more processors, the one or more clusters of the plurality of populations produced by the cluster analysis; generating, by one or more processors, a set of instructions for administering a therapeutically effective amount of a Rho kinase inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof to a subject;

8. The method of claim 7 , wherein the plurality of patient parameters comprises data and metadata.

9. The method of claim 7 , wherein the plurality of patient parameters further comprises age, weight, sex, height, and blood glucose level.

10. 1. A computer-implemented method comprising: Executing, by one or more processors, a feedforward artificial intelligence, the executing the machine learning algorithm comprising: receiving, by one or more processors, a plurality of patient parameters, the plurality of patient parameters comprising: one or more patients with diastolic dysfunction and one or more patients without diastolic dysfunction; one or more patients with cardiac fibrosis and one or more patients without cardiac fibrosis; one or more patients, including one or more patients with pulmonary hypertension and one or more patients without pulmonary hypertension one or more patients with a left ventricular ejection fraction greater than 50% and one or more patients without a left ventricular ejection fraction greater than >50%; and one or more patients with a left ventricular ejection fraction greater than 50%. one or more patients with cardiac amyloid deposits and one or more patients without cardiac amyloid deposits; analyzing, by one or more processors, the plurality of patient parameters; and mapping, by one or more processors, the plurality of patient parameters to neurons in an input layer of an artificial neural network; applying, by one or more processors, weights to the plurality of patient parameters disposed in the input layer; placing the weighted plurality of patient parameters in one or more hidden layers of the feedforward artificial neural network; and placing the weighted plurality of patient parameters in one or more hidden layers of the feedforward artificial neural network. generating, by one or more processors, a plurality of optimized data sets of patient parameters; performing, by one or more processors, a cluster analysis, said performing comprising: receiving, by one or more processors, a plurality of optimized data sets of patient parameters; wherein the plurality of optimized data sets of patient parameters comprises data for a plurality of patients. analyzing, by one or more processors, the plurality of optimized data sets of patient parameters; identifying, by one or more processors, a plurality of populations of the plurality of optimized data sets of patient parameters; generating, by one or more processors, one or more clusters of the plurality of populations; analyzing, by one or more processors, the one or more clusters of the plurality of populations produced by the cluster analysis; administering, by one or more processors, an effective amount of a Rho kinase inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof: Diastolic dysfunction Cardiac fibrosis Pulmonary hypertension left ventricular ejection fraction greater than 50%; generating a set of instructions for administering to a subject experiencing the subject does not have cardiac amyloid deposits; and Administering to a subject a therapeutically effective amount of a Rho kinase inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof.

11. 11. The computer-implemented method of claim 10, wherein the Rho kinase inhibitor is:

12. 11. The computer-implemented method of claim 10, wherein the Rho kinase inhibitor is Fanidi.

13. The computer-implemented method of claim 10 , wherein the subject is a human.

14. The computer-implemented method of claim 10 , wherein the subject is an animal.

15. The computer-implemented method of claim 10 , wherein the plurality of patient parameters comprises data and metadata.

16. The computer-implemented method of claim 10 , wherein the plurality of patient parameters further comprises age, weight, sex, height, and blood glucose level.

17. 1. A method of reducing cardiac tissue stiffness, comprising: Administering a therapeutically effective amount of a Rho kinase inhibitor, an anti-fibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof to a subject: Diastolic dysfunction Cardiac fibrosis Pulmonary hypertension Left ventricular ejection fraction greater than 50%; and left ventricular ejection fraction greater than 50%.

10. The method of claim 1, wherein the subject does not have cardiac amyloid deposits.

18. 18. The method of claim 17, wherein the Rho kinase inhibitor is:

19. 18. The method of claim 17, wherein the Rho kinase inhibitor is Fanidil.

20. 18. The method of claim 17, wherein the set of instructions for administering the therapeutically effective amount of a Rho kinase inhibitor, the antifibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof to the subject is generated by a computer-implemented method prior to administering the therapeutically effective amount of a Rho kinase inhibitor, the antifibrotic compound, or a pharmaceutically acceptable salt, ester, or solvate thereof to the subject.

Images