[go: up one dir, main page]

WO2025217600A1 - Méthodes et systèmes de gestion de résultats pour des thérapies agonistes du récepteur glp-1 de perte de poids à l'aide de modèles entraînés par des données biochimiques et/ou d'une autre analyse centrée sur le patient - Google Patents

Méthodes et systèmes de gestion de résultats pour des thérapies agonistes du récepteur glp-1 de perte de poids à l'aide de modèles entraînés par des données biochimiques et/ou d'une autre analyse centrée sur le patient

Info

Publication number
WO2025217600A1
WO2025217600A1 PCT/US2025/024402 US2025024402W WO2025217600A1 WO 2025217600 A1 WO2025217600 A1 WO 2025217600A1 US 2025024402 W US2025024402 W US 2025024402W WO 2025217600 A1 WO2025217600 A1 WO 2025217600A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
glp
receptor agonist
patient
patients
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
PCT/US2025/024402
Other languages
English (en)
Inventor
Christopher S.L. Crawford
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Advanced Neuromodulation Systems Inc
Original Assignee
Advanced Neuromodulation Systems Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Advanced Neuromodulation Systems Inc filed Critical Advanced Neuromodulation Systems Inc
Publication of WO2025217600A1 publication Critical patent/WO2025217600A1/fr
Pending legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
    • G16H20/10—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to drugs or medications, e.g. for ensuring correct administration to patients
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients

Definitions

  • GLP-1 Glucagon-like peptide 1
  • GLP-1 is a peptide produced in healthy humans during the digestive process and is secreted after eating.
  • GLP-1 lowers glucose concentrations by augmenting insulin secretion and suppressing glucagon release. GLP-1 produces other effects including slowing of gastric emptying, suppression of appetite and, potentially, inhibition of ⁇ -cell apoptosis.
  • GLP-1 receptor agonists have been developed to induce prolonged physiological responses. Some GLP-1 receptor agonists (such as Exenatide and Lixisenatide) provide short-term receptor activation. Other GLP-1 receptor agonists (for example Albiglutide, Dulaglutide, Exenatide long-acting release, and Liraglutide) have been developed to induce activation of the GLP-1 receptor over longer periods of time.
  • GLP-1 receptor agonists primarily lower postprandial blood glucose levels through inhibition of gastric emptying.
  • longer-acting compounds have a more pronounced effect on fasting glucose levels (primarily insulinotropic and glucagonostatic mechanisms).
  • GLP-1 receptor agonists have been provided to treat type 2 diabetes mellitus and the characteristics of each GLP-1 receptor agonists can be selected to tailor therapy for individual patients. [0004] GLP-1 receptor agonists have been provided to patients to treat obesity.
  • GLP-1 receptor agonists therapy For example, Heppner and Perez-Tilve discuss systemic effects promoting weight-loss caused by GLP-1 receptor agonists therapy (see, GLP-1 based therapeutics: (32736-2072) simultaneously combating T2DM and obesity, Front. Neurosci., 20 March 2015, Sec. Neuroendocrine Science, Volume 9 - 2015).
  • GLP-1 based therapeutics 32736-2072
  • Examples of commercial pharmaceutic products that have been provided to patients for weight loss include Dulaglutide (Trulicity), Exenatide extended release (Bydureon bcise), Exenatide (Byetta), Semaglutide (Ozempic), Liraglutide (Victoza, Saxenda), Lixisenatide (Adlyxin), and Semaglutide (Rybelsus).
  • GLP-1 receptor agonists products have been on the market for a significant amount of time (including a first FDA approval in 2005), there is a body of clinical data available for these products. Generally, the incidence of reported adverse events for GLP-1 receptor agonists is low compared to many pharmaceutic products for other treatment-resistant disorders.
  • FIGS.1-3 depict a system including one or more biochemical sensors according to some embodiments; a system for monitoring patients; and a remote care service system according to some embodiments.
  • FIG.4 depicts patient conditions, physiological states, measured analyte levels, and/or the like which may be correlated to patient activities according to some embodiments.
  • FIG.5 depicts a flowchart representing a series of events for a patient beginning with health coaching through weight-loss through GLP-1 receptor agonist therapy and possibly other operations according to some embodiments.
  • FIG.6 depicts an app interface according to some representative embodiments and FIG.7 depicts a smartwatch with one or more suitable health tracking apps according to some representative embodiments. (32736-2072)
  • FIG.8 depicts patient state scores or classifications for review by clinician(s) and/or patients according to some representative embodiments.
  • FIGS.9-13 depicts biomarker data sets and computational model training using such biomarker data sets according to some representative embodiments.
  • SUMMARY Aspects of the invention are set out in the independent claims and preferred features are set out in the dependent claims. Features associated with one aspect may be applied to other aspects alone or in combination.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies comprising: receiving biochemical sensor data, obtained using biochemical sensing systems, of respective patients including patients before receiving GLP-1 receptor agonist therapies and/or patients undergoing GLP-1 receptor agonist therapies; optionally or alternatively receiving activity data, obtained using a wearable smartwatch or fitness tracking device, of the respective patients including patients before receiving GLP-1 receptor agonist therapies and/or patients undergoing GLP-1 receptor agonist therapies; performing at least one of: providing patient data including instances of the received biochemical data or processed versions of the received biochemical data, and/or instances of the received activity data or processed versions of the received activity data, to first one or more trained computational models to classify the respective pre-GLP-1 receptor agonist therapy patients according to one or more patient classifications to predict potential patient outcomes, wherein the one or more patient classifications are inferred from occurrence of sub-acute systemic damage during GLP-1 receptor agonist therapy; and providing patient data including instances of
  • the method can allow for sub-acute damage to be predicted for pre- therapy patients and/or for patients already undergoing therapy to be monitored or tracked for sub-acute damage during therapy.
  • the first and/or second one or more trained computational models may comprise at least one trained computational model configured to compute at least one output indicative of systemic stress of a patient.
  • the method may further comprise detecting levels of cortisol-related biomarkers in the respective patients, wherein the at least one trained computational model is configured to compute the at least one output indicative of systemic stress based on the detected levels of cortisol-related biomarkers.
  • the first and/or second one or more trained computational models may comprise at least one trained computational model configured to compute at least one output indicative of subacute cardiac damage of a patient.
  • the method may further comprise tracking systemic fatigue of patients by analyzing at least cardiac activity signals, obtained using the wearable smartwatch or fitness tracking devices, of the respective patients.
  • the method may further comprise detecting elevated resting heart rate levels in selected patients undergoing GLP-1 receptor agonist therapies compared to heart rate levels in healthy patients without GLP-1 receptor agonist therapies.
  • the method may further comprise detecting variations in heart rate variability levels in selected patients undergoing GLP-1 receptor agonist therapies compared to heart rate variability levels in healthy patients without GLP-1 receptor agonist therapies.
  • the first one or more trained computational models may include at least one trained computational model configured to classify pre-GLP receptor agonist therapy patients according to one or more classifications indicative of subacute kidney damage using at least creatine-related biomarker data.
  • the method may further comprise establishing a baseline set of biomarker data for the respective patients, wherein the baseline set of biomarker data is indicative of patient state before beginning pre-GLP-1 receptor agonist therapy, for comparison to post-therapy biomarker data.
  • the method may further comprise (32736-2072) processing the post-therapy biomarker data based on the baseline set of biomarker data for provision to the one or more computational models.
  • the method may further comprise training selected ones of the first and second one or more computational models on the received training data of patients before receiving GLP-1 receptor agonist therapies and of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise biochemical sensor data, activity data, and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies comprising: receiving biochemical sensor data, obtained using biochemical sensing systems, of respective patients including patients before receiving GLP-1 receptor agonist therapies, and optionally patients undergoing GLP-1 receptor agonist therapies; receiving activity data, obtained using a wearable smartwatch or fitness tracking device, of the respective patients including patients before receiving GLP-1 receptor agonist therapies and patients undergoing GLP-1 receptor agonist therapies; providing patient data including the received biochemical data, or processed versions of the received biochemical data, and the received activity data, or processed versions of the received activity data, to one or more trained computational models to classify the respective pre-GLP-1 receptor agonist therapy patients according to one or more patient classifications, wherein the one or more patient classifications are inferred from occurrence of muscle loss during GLP-1 receptor agonist therapies in previous patients; and identifying respective patients, based on the one or more patient classifications, for adjunctive therapies before commencement of one or more
  • the method can allow predictions of muscle loss during GLP-1 receptor agonist therapy to be made for patients before starting the therapy, and for patients to be identified for adjunctive therapies to reduce muscle loss accordingly.
  • the method may further comprise estimating levels of blood glucose from data, obtained using a wearable glucose monitor, for provision as biomarker data to the one or more trained computation models.
  • the method may further comprise identifying blood glucose levels in the (32736-2072) respective patients relative to one or more control blood glucose levels from a sample population previously identified as experiencing a non-medically detrimental amount of muscle loss from GLP-1 receptor agonist therapy.
  • the method may further comprise identifying blood glucose levels in the respective patients relative to one or more control blood glucose levels from a sample population previously identified as experiencing detrimental amount of muscle loss from GLP-1 receptor agonist therapy.
  • the method may comprise calculating one or more metrics related to patient activity levels for provision to the one or more computational models.
  • the method may further comprise estimating a body composition metric indicative of a starting point of muscle mass for the respective patients for provision to one or more trained computational models.
  • the method may comprise training the one or more computational models on the received training data of patients before receiving GLP-1 receptor agonist therapies and of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise biochemical sensor data, activity data, and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies comprising: receiving biochemical sensor data, obtained using biochemical sensing systems, of respective patients including patients before receiving GLP-1 receptor agonist therapies and patients undergoing GLP-1 receptor agonist therapies; receiving activity data, obtained using a wearable smartwatch or fitness tracking device, of the respective patients including patients before receiving GLP-1 receptor agonist therapies and patients undergoing GLP-1 receptor agonist therapies; providing patient data including the received biochemical data, or processed versions of the received biochemical data, and the received activity data, or processed versions of the received activity data, to one or more, optionally multiple, trained computational models to classify the respective pre-GLP-1 receptor agonist therapy patients and/or the respective patients undergoing GLP-1 receptor agonist therapies according to separate patient classifications for pre-GLP-1 receptor agonist therapy patients and patients undergoing one or more GLP-1 receptor agonist therapies, wherein one or more of the trained computational models, optionally one or more models of the
  • the method can allow pre-therapy data to be used to classify or track patients after starting therapy.
  • the patient classifications for the pre-therapy patients may be based on predicted success of GLP-1 receptor agonist therapy.
  • the patient classifications for the pre-therapy patients may be based on predicted short-term negative biomarker outcomes.
  • the patient classifications for the pre-therapy patients may be based on at least biomarker profiles and body composition data.
  • the patient classifications for the pre-therapy patients may be based on at least detected levels of physical activity from the received activity data.
  • the method may further comprise comparing pre-therapy prediction classifications to post-therapy classifications.
  • the method may further comprise correlating deviations between pre-therapy and post-therapy classifications to predict sub-acute systemic damage from continued additional dosing of GLP-1 RA pharmaceutic products.
  • the method may further comprise training the one or more computational models on the received training data.
  • the training data may comprise pre-therapy and ongoing-therapy biochemical sensor data and activity data for respective patients being classified during ongoing therapy comprises biochemical sensor data, and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies comprising: receiving biochemical sensor data, obtained using biochemical sensing systems, of respective patients including patients before receiving GLP-1 receptor agonist therapies; receiving activity data, obtained using a wearable smartwatch or fitness tracking device, of the respective patients including patients before receiving GLP-1 receptor agonist therapies; providing patient data including the received biochemical data, or processed versions of the received biochemical data, and the received activity data, or processed versions of the received activity data, of the (32736-2072) respective patients before receiving GLP-1 receptor agonist therapies to one or more first trained computation models to classify the respective patients according to one or more patient classifications, wherein the one or more patient classifications are defined according to patient states correlated to success of GLP-1 receptor agonist therapies; providing outputs to patients and/or clinicians based on the one or more patient classifications, wherein the outputs include one or more predictions of expected benefits of GLP-1 receptor agonist therapies.
  • the method can allow predictions to made in relation to success of GLP-1 receptor agonist therapy for patients before receiving such therapy.
  • the method may further comprise providing outputs to patients and/or clinicians based on the one or more patient classifications, wherein the outputs include one or more lifestyle modification suggestions to practice before commencement of GLP-1 receptor agonist therapy to improve expected outcomes from GLP-1 receptor agonist therapy.
  • the method may further comprise providing outputs to patients and/or clinicians based on the one or more patient classifications, wherein the outputs include an indication when a patient’s health state is expected to lead to positive outcomes from GLP-1 receptor agonist therapy.
  • the respective pre-therapy patients may be classified according to the one or more patient classifications based on predicted short-term negative biomarker outcomes.
  • the respective pre-therapy patients may be classified based on at least biomarker profiles and body composition data.
  • the respective pre-therapy patients may be classified based on at least detected levels of physical activity from the received activity data.
  • the method may further comprise training the one or more computational models on the received training data of patients before receiving GLP-1 receptor agonist therapies.
  • the training data may comprise biochemical sensor data and activity data, and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies comprising: receiving biochemical sensor data, obtained using biochemical sensing systems, of respective patients including patients before receiving GLP-1 receptor agonist therapies, and optionally patients undergoing GLP-1 receptor (32736-2072) agonist therapies and/or patients after undergoing GLP-1 receptor agonist therapies; receiving activity data, obtained using a wearable smartwatch or fitness tracking device, of the respective patients including patients before receiving GLP-1 receptor agonist therapies, and optionally patients undergoing GLP-1 receptor agonist therapies and/or patients after undergoing GLP-1 receptor agonist therapies; providing patient data including the received biochemical data, or processed versions of the received biochemical data, and the received activity data, or processed versions of the received activity data, of the respective patients before receiving GLP-1 receptor agonist therapies to one or more first trained computation models to classify the respective patients according to one or more first patient classifications, wherein the one or more first patient classifications are defined according to patient states
  • the method can allow predictions to be made for pre-therapy patients with respect to both expected success of therapy and adverse effects during the therapy.
  • the method may further comprise providing outputs to patients and/or clinicians based on the one or more first patient classifications, wherein the outputs include one or more lifestyle modification suggestions to practice before commencement of GLP-1 receptor agonist therapy to improve expected outcomes from GLP-1 receptor agonist therapy.
  • the respective pre-therapy patients may be classified according to the one or more first patient classifications based on predicted short-term negative biomarker outcomes.
  • the respective pre-therapy patients may be classified based on at least biomarker profiles and body composition data.
  • the respective pre-therapy patients may be classified based on at least detected levels of physical activity from the received (32736-2072) activity data.
  • the respective pre-therapy patients may be classified based on sleep quality measurements obtained using the respective wearable smartwatch or fitness tracking devices.
  • the respective pre-therapy patients may be classified based on levels of systemic stress.
  • the respective pre-therapy patients may be classified based on cardiac function related to likelihood of generation of systemic stress upon commencement of GLP-1 receptor agonist therapy.
  • the method may further comprise: training the one or more first computational models on the received training data of patients before receiving GLP-1 receptor agonist therapies; and training the one or more second computational models on the received training data of patients undergoing GLP-1 receptor agonist therapies.
  • the training data of patients before receiving GLP-1 receptor agonist therapies and/or the training data of patients undergoing GLP-1 receptor agonist therapies may comprise biochemical sensor data and activity data, and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies comprising: receiving biochemical sensor data, optionally multi-analyte biochemical sensor data, obtained using biochemical sensing systems, of respective patients undergoing GLP-1 receptor agonist therapies; receiving activity data, obtained using a wearable smartwatch or fitness tracking device, of the respective patients undergoing GLP-1 receptor agonist therapies; providing the received patient biochemical data, or processed versions of the received patient biochemical data, and the received activity data, or processed versions of the received activity data, to one or more trained computational models to classify the respective patients undergoing GLP-1 receptor agonist therapies according to one or more patient classifications, wherein the one or more patient classifications are inferred from occurrence of systemic
  • the method can allow occurrence of systemic stress to be detected in patients undergoing GLP-1 receptor agonist therapy and for alerts to be provided accordingly.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data reflecting one or more of elevated resting heart rate, change in daily heart rate variability, change in cardiac rhythm, or change in blood pressure.
  • the multi-analyte biochemical sensor data may comprise data related to measurement of one or more of Serum Amyloid A, C-reactive protein, haptoglobin, fibrinogen, lactate, or norepinephrine.
  • the multi-analyte biochemical sensor data may comprise data indicative of loss of lean muscle mass.
  • the method may further comprise processing the multi-analyte biochemical sensor data and/or the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the multi- analyte biochemical sensor data and/or the activity data to identify one or more of change in creatine levels, variation in pH levels, change in fluid retention levels, or development of irregular heartbeats that are indicative of possible detection of subacute kidney injury or disfunction.
  • the multi-analyte biochemical sensor data may comprise data from measurement of one or more stress biomarkers selected from: catecholamines, cortisol, growth hormone, glucagon, glutamine, glucose, and renin.
  • the method may further comprise processing the multi-analyte biochemical sensor data before provision to the one or more trained computational models, wherein the processing comprises processing the multi-analyte biochemical sensor data to identify biomarker data indicative of elevated lactate levels and comparing that biomarker data to an indication of physical activity, to identify biomarker data indicative of elevated lactate levels that is correlated to little or no physical activity generating elevated lactate levels.
  • the multi-analyte biochemical sensor data may comprise biomarker data from measurement of one or more of ⁇ -hydroxybutyrate, acetyl-CoA, acetylated histones, or non-histone proteins.
  • the multi-analyte biochemical sensor data may comprise biomarker data from measurement of cytokines or pro-inflammatory enzyme levels.
  • the multi-analyte biochemical sensor data may comprise biomarker data from measurement of one or more of fibrinogen, homocysteine, Interleukin 6, magnesium, vitamin D, calcium, phosphate, creatine kinase, myoglobin, FGF19, or FGF21.
  • the method may further comprise prompting one or more of the respective patients, using the respective smartwatch or fitness tracking devices, to input a patient reported outcome related to one or more patient conditions.
  • the patient reported outcome may include one or more of swelling in the legs, ankles, or feet, shortness of breath, fatigue, nausea, weakness, and chest pain or pressure.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to identify one or more of a decrease in steps taken, a decrease in average walking speed, or a change in walking gait to provide an input indicative of fatigue to at least one of the trained computational models.
  • the method may further comprise training the one or more computational models on the received training data of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise multi-analyte biochemical sensor data comprising data indicative of at least one analyte, activity data, and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies comprising: receiving biochemical sensor data, optionally multi-analyte biochemical sensor data, obtained using biochemical sensing systems, of respective patients undergoing GLP-1 receptor agonist therapies; providing the received patient biochemical data or processed versions of the received patient biochemical data to one or more trained computational models to classify the respective patients undergoing GLP-1 receptor agonist therapies according to one or more patient classifications, wherein the one or more patient classifications are inferred from occurrence of systemic stress leading (32736-2072) to long-term negative outcomes of GLP-1 receptor agonist therapies based on previously received training data of patients undergoing GLP-1 receptor agonist therapies; and providing outputs to patients and/or clinicians based on the one or more patient classifications, wherein the outputs include one or more alerts indicative of occurrence of systemic stress associated with long-term negative outcomes of GLP-1 receptor agonist therapies.
  • the method can allow occurrence of systemic stress to be detected in patients undergoing GLP-1 receptor agonist therapy based on biomarker data alone, and for alerts to be provided accordingly.
  • the multi-analyte biochemical sensor data may comprise data related to measurement of one or more of Serum Amyloid A, C-reactive protein, haptoglobin, fibrinogen, lactate, or norepinephrine.
  • the multi-analyte biochemical sensor data may comprise data indicative of loss of lean muscle mass.
  • the multi-analyte biochemical sensor and/or activity data may be processed before provision to one or more trained computational models to identify change in one or more of creatine levels, variation in pH levels, change in fluid retention levels, or development of irregular heartbeats that are indicative of possible detection of subacute kidney injury or disfunction.
  • the multi-analyte biochemical sensor data may comprise data from measurement of one or more stress biomarkers selected from: catecholamines, cortisol, growth hormone, glucagon, glutamine, glucose, and renin.
  • the method may further comprise processing the multi-analyte biochemical sensor data before provision to the one or more trained computational models, wherein the processing comprises processing the multi-analyte biochemical sensor data to identify biomarker data indicative of elevated lactate levels and comparing that biomarker data to an indication of physical activity, to identify biomarker data indicative of elevated lactate levels that is correlated to little or no physical activity generating elevated lactate levels.
  • the multi-analyte biochemical sensor data may comprise biomarker data from measurement of one or more of ⁇ -hydroxybutyrate, acetyl-CoA, acetylated histones, or non-histone proteins.
  • the multi-analyte biochemical sensor data may comprise biomarker data from measurement of cytokines or pro-inflammatory enzyme levels.
  • the multi-analyte biochemical sensor data may comprise biomarker data from measurement (32736-2072) of one or more of fibrinogen, homocysteine, Interleukin 6, magnesium, vitamin D, calcium, phosphate, creatine kinase, myoglobin, FGF19, or FGF21.
  • the method may further comprise training the one or more computational models on the received training data of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise multi-analyte biochemical sensor data comprising data indicative of at least one analyte and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies, comprising: receiving biochemical sensor data, optionally multi-analyte biochemical sensor data, obtained using biochemical sensing systems, of respective patients undergoing GLP-1 receptor agonist therapies; optionally receiving activity data, obtained using a wearable smartwatch or fitness tracking device, of the respective patients undergoing GLP-1 receptor agonist therapies; providing the received patient biochemical data, or processed versions of the received patient biochemical data, and optionally the received patient activity data, or processed versions of the received patient activity data, to one or more trained computational models to classify the respective patients undergoing GLP-1 receptor agonist therapies according to one or more patient classifications, wherein the one or more patient classifications are inferred from occurrence of premature aging leading to long-term negative outcomes of GLP-1 receptor agonist therapies based on previously received training data of patients undergoing GLP-1 receptor agonist therapies; and providing outputs to patients and/or clinicians based on the
  • the method can allow occurrence of premature aging to be detected in patients undergoing GLP-1 receptor agonist therapy and for alerts to be provided accordingly.
  • One or more of the patient classifications may be defined for the training data using invasive medical testing in order to correlate premature aging outcomes to non-invasively measured biomarkers.
  • One or more of the patient classifications may be defined for the training data using invasive measurements of endothelial function.
  • One or (32736-2072) more of the patient classifications may be defined for the training data using biological age calculations using multi-factorial blood testing.
  • the multi-analyte biochemical sensor data may comprise data related to measurement of one or more of Serum Amyloid A, C- reactive protein, haptoglobin, fibrinogen, lactate, or norepinephrine.
  • the multi-analyte biochemical sensor data may comprise data indicative of loss of lean muscle mass.
  • the multi-analyte biochemical sensor and/or activity data may be processed before provision to one or more trained computational models to identify change in one or more of creatine levels, variation in pH levels, change in fluid retention levels, or development of irregular heartbeats that are indicative of possible detection of subacute kidney injury or disfunction.
  • the multi-analyte biochemical sensor data may comprise data from measurement of one or more stress biomarkers selected from: catecholamines, cortisol, growth hormone, glucagon, glutamine, glucose, and renin.
  • the method may further comprise processing the multi-analyte biochemical sensor data before provision to the one or more trained computational models, wherein the processing comprises processing the multi-analyte biochemical sensor data to identify biomarker data indicative of elevated lactate levels and comparing that biomarker data to an indication of physical activity based on the received activity data, to identify biomarker data indicative of elevated lactate levels that is correlated to little or no physical activity generating elevated lactate levels.
  • the multi-analyte biochemical sensor data may comprise biomarker data from measurement of one or more of ⁇ -hydroxybutyrate, acetyl-CoA, acetylated histones, or non-histone proteins.
  • the multi-analyte biochemical sensor data may comprise biomarker data from measurement of cytokines or pro-inflammatory enzyme levels.
  • the multi-analyte biochemical sensor data may comprise biomarker data from measurement of one or more of fibrinogen, homocysteine, Interleukin 6, magnesium, vitamin D, calcium, phosphate, creatine kinase, myoglobin, FGF19, or FGF21.
  • the method may further comprise training the one or more computational models on the received training data of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise multi-analyte biochemical sensor data comprising data indicative of at least one analyte, patient activity data, and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies comprising: receiving biochemical sensor data, optionally multi-analyte biochemical sensor data, obtained using biochemical sensing systems, of respective patients undergoing GLP-1 receptor agonist therapies; optionally receiving activity data, obtained using a wearable smartwatch or fitness tracking device, of the respective patients undergoing GLP-1 receptor agonist therapies; providing the received patient biochemical data, or processed versions of the received patient biochemical data, and optionally the received patient activity data, or processed versions of the received patient activity data, to one or more trained computational models to classify the respective patients undergoing GLP-1 receptor agonist therapies according to one or more patient classifications, wherein the one or more patient classifications are inferred from occurrence of subacute cardiac sequelae leading to long-term negative outcomes of GLP-1 receptor agonist therapies based on previously received training data of patients undergoing GLP-1 receptor agonist therapies; and providing outputs to
  • the method can allow occurrence of subacute cardiac sequelae to be detected in patients undergoing GLP-1 receptor agonist therapy and for alerts to be provided accordingly.
  • Classification of the respective patients according to subacute cardiac sequelae by the one or more trained computational models may comprise computing multi-factorial biomarkers related to muscle loss in patients.
  • the multi- factorial biomarkers related to muscle loss may comprise creatine-related biomarkers.
  • Classification of the respective patients according to subacute cardiac sequelae by the one or more trained computational models may comprise computing multi-factorial biomarkers related to development of pro-thrombotic environment in patients.
  • Classification of the respective patients according to subacute cardiac sequelae by the one or more trained computational models may comprise computing multi-factorial biomarkers related to development of oxidative stress in patients. Classification of the respective patients according to subacute cardiac sequelae by the one or more trained (32736-2072) computational models may comprise detecting repeated episodes of hyperglycemia in patients. Classification of the respective patients according to subacute cardiac sequelae by the one or more trained computational models may comprise detecting repeated episodes of excessive ketone-driven metabolic function in patients. [0055] The one or more trained computational models may be configured to process one or more biomarkers indicative of elevated resting heart rate in patients. The one or more trained computational models may be configured to process one or more biomarkers indicative of intermittent cardiac irregularities.
  • the one or more trained computational models may be configured to process one or more biomarkers indicative of changes in QT-intervals in patients.
  • the method may further comprise training the one or more computational models on the received training data of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise multi-analyte biochemical sensor data comprising data indicative of at least one analyte, patient activity data, and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies comprising: receiving biochemical sensor data, optionally multi-analyte biochemical sensor data, obtained using biochemical sensing systems, of respective patients undergoing GLP-1 receptor agonist therapies; optionally receiving activity data, obtained using a wearable smartwatch or fitness tracking device, of the respective patients undergoing GLP-1 receptor agonist therapies; providing the received patient biochemical data, or processed versions of the received patient biochemical data, and optionally the received patient activity data, or processed versions of the received patient activity data, to one or more trained computational models to classify the respective patients undergoing GLP-1 receptor agonist therapies according to one or more patient classifications, wherein the one or more patient classifications are inferred from occurrence of subacute renal sequelae leading to long-term negative outcomes of GLP-1 receptor agonist therapies based on previously received training data of patients undergoing GLP-1 receptor agonist therapies; and providing outputs to patients and/or clinicians
  • the method can allow occurrence of subacute renal sequelae to be detected in patients undergoing GLP-1 receptor agonist therapy and for alerts to be provided accordingly.
  • the one or more trained computational models may be trained using patient training data correlated to measured endothelial functional testing.
  • the one or more trained computational models may be configured to process multi-factorial biomarker data related to subacute renal sequelae comprises creatine-related biomarker data.
  • the one or more trained computational models may be configured to process multi- factorial biomarker data related to subacute renal sequelae comprises patient pH-level biomarker data.
  • the one or more trained computational models may be configured to process data indicative of fluid retention in patients.
  • the one or more trained computational models may be configured to process activity data indicative of fatigue in patients.
  • the method may further comprise training the one or more computational models on the received training data of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise multi-analyte biochemical sensor data comprising data indicative of at least one analyte, patient activity data, and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies, comprising: receiving patient data, obtained using one or more respective wearable devices and/or one or more biochemical sensing systems, of respective patients undergoing GLP-1 receptor agonist therapies for weight loss; providing the received patient data or processed versions of the received patient data to one or more trained computational models to classify the respective patients undergoing GLP-1 receptor agonist therapies according to one or more patient classifications, wherein the one or more patient classifications are inferred from patient results of GLP-1 receptor agonist therapy based on previously received training data of patients undergoing GLP-1 receptor agonist therapies, the training data comprising one or more representations of systemic fatigue in patients undergoing one or more GLP-1 receptor agonist therapies; (32736-2072) and providing outputs to patients and/or clinicians based on the one or more patient classifications.
  • the method can allow occurrence of systemic fatigue to be detected in patients undergoing GLP-1 receptor agonist therapy, for example based on patient data obtained using a wearable device such as a smartwatch.
  • the received patient data may comprise activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of elevated resting heart rate.
  • the received patient data may comprise physiological data obtained using the respective wearable devices.
  • the method may further comprise processing the physiological data before provision to the one or more trained computational models, wherein the processing comprises processing the physiological data to generate data indicative of changes in daily heart rate variability.
  • the received patient data may comprise physiological data obtained using the respective wearable devices.
  • the method may further comprise processing the physiological data before provision to the one or more trained computational models, wherein the processing comprises processing the physiological data to generate data indicative of changes in cardiac rhythm.
  • the received patient data may comprise activity data obtained using the respective wearable devices.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of decreases in steps taken.
  • the received patient data may comprise activity data obtained using the respective wearable devices.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of decreases in average walking speed.
  • the received patient data may comprise activity data obtained using the respective wearable devices.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of changes in walking gait.
  • the received patient data may comprise activity data obtained using the respective wearable devices.
  • the method may further comprise (32736-2072) processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of sleep quality or sleep duration.
  • the method may further comprise training the one or more computational models on the received training data of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise patient data and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies comprising: receiving biochemical sensor data, obtained using biochemical sensing systems, of respective patients including patients before receiving GLP-1 receptor agonist therapies and patients undergoing GLP-1 receptor agonist therapies; providing the received patient biochemical-related data or processed versions of the received patient biochemical-related data to one or more trained computational models to classify the respective patients undergoing GLP-1 receptor agonist therapies according to one or more patient classifications, wherein the one or more patient classifications are inferred from occurrence of one or more subacute adverse effects of patients leading to long-term negative outcomes of GLP-1 receptor agonist therapies based on previously received training data of patients before receiving GLP-1 receptor agonist therapies and of patients undergoing GLP-1 receptor agonist therapies, wherein the one or more trained computational models are configured to classify the respective patients based on changes in biochemical levels of the respective patients undergoing therapy relative to baseline pre-therapy levels; and providing outputs to patients and
  • the method can allow predictions of expected benefits of continued GLP- 1 receptor agonist therapy for patients already undergoing such therapy based on their pre-therapy baseline biochemical levels.
  • the method may further comprise: receiving physiological data, obtained using a respective wearable smartwatch or fitness tracking device, of the respective patients undergoing GLP-1 receptor agonist therapies; and (32736-2072) processing the physiological data before provision to the one or more trained computational models, wherein the processing comprises processing the physiological data to generate data reflecting one or more of elevated resting heart rate, change in daily heart rate variability, change in cardiac rhythm, or change in blood pressure.
  • the patient data may comprise activity data obtained using the respective wearable smartwatch or fitness tracking device.
  • the method may further comprise: processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to identify one or more of a decrease in steps taken, a decrease in average walking speed, or a change in walking gait; and providing, based on the processing of the activity data, an input indicative of fatigue to at least one of the trained computational models.
  • the biochemical sensor data may comprise data related to measurement of one or more of Serum Amyloid A, C-reactive protein, haptoglobin, fibrinogen, lactate, or norepinephrine.
  • the biochemical sensor data may comprise data indicative of loss of lean muscle mass.
  • the method may further comprise processing the biochemical sensor data and/or activity data of the respective patients before provision to the one or more trained computational models, wherein the processing comprises processing the biochemical sensor data and/or the activity data to identify one or more of change in creatine levels, variation in pH levels, change in fluid retention levels, or development of irregular heartbeats that are indicative of possible detection of subacute kidney injury or disfunction.
  • the biochemical sensor data may comprise data from measurement of one or more stress biomarkers selected from: catecholamines, cortisol, growth hormone, glucagon, glutamine, glucose, and renin.
  • the method may further comprise processing the biochemical sensor data before provision to the one or more trained computational models, wherein the processing comprises processing the biochemical sensor data to identify biomarker data indicative of elevated lactate levels and comparing that biomarker data to an indication of physical activity, to identify biomarker data indicative of elevated lactate levels that is correlated to little or no physical activity generating elevated lactate levels.
  • the multi-analyte biochemical sensor data may comprise biomarker data from measurement of one or more of ⁇ -hydroxybutyrate, acetyl-CoA, acetylated histones, or non-histone proteins.
  • the multi-analyte biochemical sensor data may comprise biomarker data from measurement of cytokines or pro-inflammatory enzyme levels.
  • the multi-analyte biochemical sensor data may comprise biomarker data from measurement of one or more of fibrinogen, homocysteine, Interleukin 6, magnesium, vitamin D, calcium, phosphate, creatine kinase, myoglobin, FGF19, or FGF21.
  • the method may further comprise prompting one or more of the respective patients, using the smartwatch or fitness tracking device, to input a patient reported outcome related to one or more patient conditions.
  • the patient reported outcome may include one or more of swelling in the legs, ankles, or feet, shortness of breath, fatigue, nausea, weakness, and chest pain or pressure.
  • the method may further comprise training the one or more computational models on the received training data of patients before receiving GLP-1 receptor agonist therapies and of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise biochemical sensor data and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies comprising: optionally receiving biochemical sensor data, obtained using biochemical sensing systems, of patients including patients undergoing weight-loss regimens without receiving weight-loss promoting pharmaceutical products; receiving biochemical sensor data, obtained using biochemical sensing systems, of respective patients including patients undergoing GLP-1 receptor agonist therapies; providing the received patient biochemical-related data or processed versions of the received patient biochemical-related data to one or more trained computational models to classify the respective patients undergoing GLP-1 receptor agonist therapies according to one or more patient classifications, wherein the one or more patient classifications are inferred from occurrence of one or more subacute adverse effects of patients leading to long-term negative outcomes of GLP-1 receptor agonist therapies based on previously received training data of patients undergoing GLP-1 receptor agonist therapies and of patients (32736-2072) undergoing weight-loss regimens without receiving weight-loss promoting pharmaceutical products, wherein the one
  • the method can allow predictions of expected benefits of continued GLP- 1 receptor agonist therapy for patients already undergoing such therapy based on baseline biochemical levels for patients undergoing non-pharmaceutical weight-loss regimens.
  • the method may further comprise: receiving physiological data, obtained using a wearable smartwatch or fitness tracking device, of patients undergoing GLP-1 receptor agonist therapies; and processing the physiological data before provision to the one or more trained computational models, wherein the processing comprises processing the physiological data to generate data reflecting one or more of elevated resting heart rate, change in daily heart rate variability, change in cardiac rhythm, or change in blood pressure.
  • the patient data may comprise activity data obtained using the respective wearable smartwatch or fitness tracking device.
  • the method may further comprise: processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to identify one or more of a decrease in steps taken, a decrease in average walking speed, or a change in walking gait; and providing, based on the processing of the activity data, an input indicative of fatigue to at least one of the trained computational models.
  • the biochemical sensor data may comprise data related to measurement of one or more of Serum Amyloid A, C-reactive protein, haptoglobin, fibrinogen, lactate, or norepinephrine.
  • the biochemical sensor data may comprise data indicative of loss of lean muscle mass.
  • the method may further comprise processing the biochemical sensor data and/or activity data of the respective patients before provision to the one or more trained computational models, wherein the processing comprises processing the (32736-2072) biochemical sensor data and/or the activity data to identify one or more of change in creatine levels, variation in pH levels, change in fluid retention levels, or development of irregular heartbeats that are indicative of possible detection of subacute kidney injury or disfunction.
  • the biochemical sensor data may comprise data from measurement of one or more stress biomarkers selected from: catecholamines, cortisol, growth hormone, glucagon, glutamine, glucose, and renin.
  • the method may further comprise processing the biochemical sensor data before provision to the one or more trained computational models, wherein the processing comprises processing the biochemical sensor data to identify biomarker data indicative of elevated lactate levels and comparing that biomarker data to an indication of physical activity, to identify biomarker data indicative of elevated lactate levels that is correlated to little or no physical activity generating elevated lactate levels.
  • the biochemical sensor data may comprise biomarker data from measurement of one or more of ⁇ -hydroxybutyrate, acetyl-CoA, acetylated histones, or non-histone proteins.
  • the biochemical sensor data may comprise biomarker data from measurement of cytokines or pro-inflammatory enzyme levels.
  • the biochemical sensor data may comprise biomarker data from measurement of one or more of fibrinogen, homocysteine, Interleukin 6, magnesium, vitamin D, calcium, phosphate, creatine kinase, myoglobin, FGF19, or FGF21.
  • the method may further comprise prompting one or more of the respective patients, using the smartwatch or fitness tracking device, to input a patient reported outcome related to one or more patient conditions.
  • the patient reported outcome may include one or more of swelling in the legs, ankles, or feet, shortness of breath, fatigue, nausea, weakness, and chest pain or pressure.
  • the method may further comprise training the one or more computational models on the received training data of patients undergoing GLP-1 receptor agonist therapies and of patients undergoing weight-loss regimens without receiving weight-loss promoting pharmaceutical products.
  • the training data may comprise biochemical sensor data, and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies comprising: receiving biochemical sensor data, obtained using biochemical sensing systems, of respective patients including patients before receiving GLP-1 receptor agonist therapies and patients undergoing GLP-1 receptor agonist therapies; optionally receiving physiological data, obtained using a respective wearable smartwatch or fitness tracking device, of the respective patients including patients before receiving GLP-1 receptor agonist therapies and patients undergoing GLP-1 receptor agonist therapies; providing patient data including the received biochemical data, or processed versions of the received biochemical data, and optionally the received physiological data, or processed versions of the received activity data, to one or more, optionally multiple, trained computational models to classify the respective pre-GLP-1 receptor agonist therapy patients according to one or more patient classifications, wherein the one or more patient classifications are inferred from occurrence of sub-acute systemic damage during GLP-1 receptor agonist therapy, wherein the one or more trained computational models include one or more
  • the method can allow predictions relating to occurrence of subacute systemic damage (e.g. metrics relating to systemic stress, systemic inflammation, cardiac health and/or metabolic function) to be made for patients before undergoing GLP-1 receptor agonist therapy.
  • the method may further comprise processing the physiological data before provision to the trained computational models, wherein the processing comprises processing the physiological data to generate data reflecting one or more of elevated resting heart rate, change in daily heart rate variability, change in cardiac rhythm, or change in blood pressure.
  • the biochemical sensor data may comprise data related to measurement of one or more of Serum Amyloid A, C-reactive protein, haptoglobin, fibrinogen, lactate, or norepinephrine.
  • the biochemical sensor data may comprise data indicative of loss of lean muscle mass.
  • the method may further comprise processing the biochemical sensor data and/or the physiological data before provision to the trained computational models, wherein the processing comprises processing the biochemical sensor data and/or the physiological data to identify one or more of change in creatine levels, variation in pH levels, change in fluid retention levels, or development of irregular heartbeats that are indicative of possible detection of subacute kidney injury or disfunction.
  • the biochemical sensor data may comprise data from measurement of one or more stress biomarkers selected from: catecholamines, cortisol, growth hormone, glucagon, glutamine, glucose, and renin.
  • the method may further comprise processing the biochemical sensor data before provision to the trained computational models, wherein the processing comprises processing the biochemical sensor data to identify biomarker data indicative of elevated lactate levels and comparing that biomarker data to an indication of physical activity, to identify biomarker data indicative of elevated lactate levels that is correlated to little or no physical activity generating elevated lactate levels.
  • the biochemical sensor data may comprise biomarker data from measurement of one or more of ⁇ -hydroxybutyrate, acetyl-CoA, acetylated histones, or non-histone proteins.
  • the biochemical sensor data may comprise biomarker data from measurement of cytokines or pro-inflammatory enzyme levels.
  • the biochemical sensor data may comprise biomarker data from measurement of one or more of fibrinogen, homocysteine, Interleukin 6, magnesium, vitamin D, calcium, phosphate, creatine kinase, myoglobin, FGF19, or FGF21.
  • the method may further comprise prompting one or more of the respective patients, using the respective smartwatch or fitness tracking device, to input a patient reported outcome related to one or more patient conditions.
  • the patient reported outcome may include one or more of swelling in the legs, ankles, or feet, shortness of breath, fatigue, nausea, weakness, and chest pain or pressure.
  • the method may further comprise receiving activity data obtained by using the respective wearable smartwatch or fitness tracking device.
  • the method may further comprise: processing the activity data before provision to the trained computational models, wherein the processing comprises processing the activity data to (32736-2072) identify one or more of a decrease in steps taken, a decrease in average walking speed, or a change in walking gait; and providing, based on the processing of the activity data, an input indicative of fatigue to at least one of the trained computational models.
  • the method may further comprise training the computational models on the received training data of patients before receiving GLP-1 receptor agonist therapies and of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise biochemical sensor data, and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies comprising: receiving biochemical sensor data, optionally multi-analyte biochemical sensor data, obtained using biochemical sensing systems, of respective patients undergoing GLP-1 receptor agonist therapies, wherein the multi-analyte biochemical sensor data includes data relevant to stress-hormone levels in interstitial fluid of the respective patients and wherein the stress-related hormone levels are selected from one or more of: cortisol levels and norepinephrine levels; providing the received patient biochemical data or processed versions of the received patient biochemical data to one or more trained computational models to classify the respective patients undergoing GLP-1 receptor agonist therapies according to one or more patient classifications, wherein the one or more patient classifications are inferred from occurrence of systemic stress leading to long-term negative outcomes of GLP-1 receptor agonist therapies based on previously received training data of patients undergoing GLP-1 receptor agonist therapies; and providing outputs to
  • the method can allow occurrence of systemic stress and potential long- term negative outcomes to be detected in patients undergoing GLP-1 receptor agonist therapy.
  • the method may further comprise receiving activity data, obtained using respective wearable smartwatch or fitness tracking devices, of the respective patients undergoing GLP-1 receptor agonist therapies.
  • the method may further comprise: (32736-2072) correlating physiological activity data of the received activity data, obtained using the respective wearable smartwatch or fitness tracking devices, of the respective patients that is relevant to systemic stress to measured levels of the data relevant to stress- hormone levels in interstitial fluid of the respective patients; and providing data indicative of the correlation between the physiological activity data of the respective patients relevant to systemic stress and the measured levels of the data relevant to stress- hormone levels in interstitial fluid of the respective patients to the one or more trained computational models.
  • the physiological activity data may comprise cardiac activity.
  • the cardiac activity may include heart rate.
  • the method may further comprise: calculating temporal variations of the data relevant to stress-hormone levels in interstitial fluid of the respective patients; and providing the calculated temporal variations to the one or more trained computational models.
  • the method may comprise: estimating blood glucose levels for the respective patients based on the multi-analyte biochemical sensor data; correlating the estimated blood glucose levels to estimated stress-hormone levels in interstitial fluid of the respective patients; and providing data indicative of the correlation between variations between the estimated blood glucose levels and the estimated stress-hormone levels in interstitial fluid of the respective patients to the one or more trained computational models.
  • the method may comprise processing the multi-analyte biochemical sensor data before provision to the one or more trained computational models, wherein the processing comprises processing the multi-analyte biochemical sensor data to identify biomarker data indicative of elevated lactate levels and comparing that biomarker data to an indication of physical activity, to identify biomarker data indicative of elevated lactate levels that is correlated to little or no physical activity generating elevated lactate levels.
  • the method may further comprise training the one or more computational models on the received training data of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise multi-analyte biochemical sensor data comprising data relevant to stress-hormone levels in interstitial fluid of patients, and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies, comprising: receiving biochemical sensor data, obtained using biochemical sensing systems, of respective patients undergoing GLP-1 receptor agonist therapies; providing the received patient biochemical data or processed versions of the received patient biochemical data to one or more trained computational models to classify the respective GLP-1 receptor agonist therapy patients according to one or more patient classifications, wherein the one or more patient classifications are inferred from previously received sets of training data, the training data comprising biochemical data for patients categorized by occurrence or non-occurrence of sagging or aging of facial skin identified after GLP-1 receptor agonist therapy for weight loss and one or more patient characteristics including one or more of the following: an aggregate weight loss metric, starting-therapy weight metric, total amount of weight-loss metric, a muscle-loss metric, a body mass or body composition metric; and providing outputs to patients and/or clinicians based on
  • the method can allow indications of the risks/benefits of continuing GLP- 1 receptor agonist therapy to be provided for patients undergoing such therapy based on detected occurrence of sagging or aging of facial skin. Sagging or aging of facial skin may be detected based on detecting a depletion of facial volume and/or fat (e.g. resulting in wrinkles).
  • the biochemical sensor data may comprise data related to measurement of one or more of Serum Amyloid A, C-reactive protein, haptoglobin, fibrinogen, lactate, or norepinephrine.
  • the biochemical sensor data may comprise data indicative of loss of lean muscle mass.
  • the method may further comprise processing the biochemical sensor data of the respective patients before provision to the one or more trained computational models, wherein the processing comprises processing the biochemical sensor data to identify one or more of change in creatine levels, variation in pH levels, change in fluid retention levels. (32736-2072)
  • the biochemical sensor data may comprise data from measurement of one or more stress biomarkers selected from: catecholamines, cortisol, growth hormone, glucagon, glutamine, glucose, and renin.
  • the method may further comprise processing the biochemical sensor data before provision to the one or more trained computational models, wherein the processing comprises processing the biochemical sensor data to identify biomarker data indicative of elevated lactate levels and comparing that biomarker data to an indication of physical activity, to identify biomarker data indicative of elevated lactate levels that is correlated to little or no physical activity generating elevated lactate levels.
  • the biochemical sensor data may comprise biomarker data from measurement of one or more of ⁇ -hydroxybutyrate, acetyl-CoA, acetylated histones, or non-histone proteins.
  • the biochemical sensor data may comprise biomarker data from measurement of cytokines or pro-inflammatory enzyme levels.
  • the biochemical sensor data may comprise biomarker data from measurement of one or more of fibrinogen, homocysteine, Interleukin 6, magnesium, vitamin D, calcium, phosphate, creatine kinase, myoglobin, FGF19, or FGF21.
  • the method may further comprise: receiving patient data, obtained using one or more respective wearable devices, of the respective patients undergoing GLP-1 receptor agonist therapies for weight loss; and providing patient physiological data based on the received patient data of the respective patients undergoing GLP-1 receptor agonist therapies for weight loss to the one or more computation models.
  • the patient data may comprise activity data obtained using wearable device.
  • the method may further comprise the step of processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of elevated resting heart rate.
  • the method may further comprise processing the physiological data before provision to the one or more trained computational models, wherein the processing comprises processing the physiological data to generate data indicative of changes in daily heart rate variability.
  • the method may further comprise processing the physiological data before provision to the one or more trained computational models, wherein the processing comprises processing the physiological data to generate data indicative of changes in cardiac rhythm.
  • the method (32736-2072) may further comprise processing the physiological data to identify development of irregular heartbeats that are indicative of possible detection of subacute kidney injury or disfunction.
  • the patient data may comprise activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of decreases in steps taken.
  • the patient data may comprise activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of decreases in average walking speed.
  • the patient data may comprise activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of changes in walking gait.
  • the patient data may comprise activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of sleep quality or sleep duration.
  • the method may further comprise training the one or more computational models on the received training data of patients undergoing GLP-1 receptor agonist therapies, wherein the training data comprises biochemical sensor data, one or more patient characteristics, and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies, comprising: receiving biochemical sensor data, obtained using biochemical sensing systems, of respective patients undergoing GLP-1 receptor agonist therapies; providing the received patient biochemical data or processed versions of the received patient biochemical data to one or more trained computational models to classify the respective GLP-1 receptor agonist therapy patients according to one or more patient classifications, wherein the one or more patient classifications are inferred from (32736-2072) previously received sets of training data, the training data comprising biochemical data for patients categorized by changes in one or more endothelial function outcomes measured using quantitative angiography before and after GLP-1 receptor agonist therapy for weight loss; and providing outputs to patients and/or clinicians based on the one or more patient classifications, wherein the provided outputs are indicative of one or more risks or benefits of continuing therapy.
  • the method can allow indications of the risks/benefits of continuing GLP- 1 receptor agonist therapy to be provided for patients undergoing such therapy based on changes in endothelial function.
  • the biochemical sensor data may comprise data related to measurement of one or more of Serum Amyloid A, C-reactive protein, haptoglobin, fibrinogen, lactate, or norepinephrine.
  • the biochemical sensor data may comprise data indicative of loss of lean muscle mass.
  • the method may further comprise processing the biochemical sensor data and/or activity data of the respective patients before provision to the one or more trained computational models, wherein the processing comprises processing the biochemical sensor data and/or the activity data to identify one or more of change in creatine levels, variation in pH levels, change in fluid retention levels, or development of irregular heartbeats that are indicative of possible detection of subacute kidney injury or disfunction.
  • the biochemical sensor data may comprise data from measurement of one or more stress biomarkers selected from: catecholamines, cortisol, growth hormone, glucagon, glutamine, glucose, and renin.
  • the method may further comprise processing the biochemical sensor data before provision to the one or more trained computational models, wherein the processing comprises processing the biochemical sensor data to identify biomarker data indicative of elevated lactate levels and comparing that biomarker data to an indication of physical activity, to identify biomarker data indicative of elevated lactate levels that is correlated to little or no physical activity generating elevated lactate levels.
  • the biochemical sensor data may comprise biomarker data from measurement of one or more of ⁇ -hydroxybutyrate, acetyl-CoA, acetylated histones, or non-histone proteins.
  • the biochemical sensor data may comprise biomarker data from (32736-2072) measurement of cytokines or pro-inflammatory enzyme levels.
  • the biochemical sensor data may comprise biomarker data from measurement of one or more of fibrinogen, homocysteine, Interleukin 6, magnesium, vitamin D, calcium, phosphate, creatine kinase, myoglobin, FGF19, or FGF21.
  • the method may further comprise: receiving patient data, obtained using one or more respective wearable devices, of the respective patients undergoing GLP-1 receptor agonist therapies for weight loss; and providing patient physiological data based on the received patient data of the respective patients undergoing GLP-1 receptor agonist therapies for weight loss to the one or more computation models.
  • the patient data may comprise activity data obtained using the respective wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of elevated resting heart rate.
  • the method may further comprise processing the physiological data before provision to the one or more trained computational models, wherein the processing comprises processing the physiological data to generate data indicative of changes in daily heart rate variability.
  • the method may further comprise processing the physiological data before provision to the one or more trained computational models, wherein the processing comprises processing the physiological data to generate data indicative of changes in cardiac rhythm.
  • the patient data may comprise activity data obtained using the respective wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of decreases in steps taken.
  • the patient data may comprise activity data obtained using the respective wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of decreases in average walking speed.
  • the patient data may comprise activity data obtained using respective the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of changes in walking gait.
  • the patient data may comprise (32736-2072) activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of sleep quality or sleep duration.
  • the method may further comprise training the one or more computational models on the received training data of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise biochemical sensor data and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies, comprising: receiving biochemical sensor data, obtained using biochemical sensing systems, of respective patients undergoing GLP-1 receptor agonist therapies; providing the received patient biochemical data or processed versions of the received patient biochemical data to one or more trained computational models to classify the respective GLP-1 receptor agonist therapy patients according to one or more patient classifications, wherein the one or more patient classifications are inferred from previously received sets of training data, the training data comprising biochemical data for patients categorized by changes in one or more metabolic function outcomes measured before and after GLP-1 receptor agonist therapy for weight loss; and providing outputs to patients and/or clinicians based on the one or more patient classifications, wherein the provided outputs are indicative of one or more risks or benefits of continuing therapy.
  • the method can allow indications of the risks/benefits of continuing GLP- 1 receptor agonist therapy to be provided for patients undergoing such therapy based on changes in metabolic function. For example, patients who exhibit no metabolic response (to therapy) may be associated with a therapy failure classification.
  • the biochemical sensor data may comprise data related to measurement of one or more of Serum Amyloid A, C-reactive protein, haptoglobin, fibrinogen, lactate, or norepinephrine.
  • the biochemical sensor data may comprise data indicative of loss of lean muscle mass.
  • the method may further comprise processing the biochemical sensor data of the respective patients before provision to the one or more trained computational models, wherein the processing comprises processing the biochemical sensor data and/or (32736-2072) the activity data to identify one or more of change in creatine levels, variation in pH levels, change in fluid retention levels.
  • the biochemical sensor data may comprise data from measurement of one or more stress biomarkers selected from: catecholamines, cortisol, growth hormone, glucagon, glutamine, glucose, and renin.
  • the method may further comprise processing the biochemical sensor data before provision to the one or more trained computational models, wherein the processing comprises processing the biochemical sensor data to identify biomarker data indicative of elevated lactate levels and comparing that biomarker data to an indication of physical activity, to identify biomarker data indicative of elevated lactate levels that is correlated to little or no physical activity generating elevated lactate levels.
  • the biochemical sensor data may comprise biomarker data from measurement of one or more of ⁇ -hydroxybutyrate, acetyl-CoA, acetylated histones, or non-histone proteins.
  • the biochemical sensor data may comprise biomarker data from measurement of cytokines or pro-inflammatory enzyme levels.
  • the biochemical sensor data may comprise biomarker data from measurement of one or more of fibrinogen, homocysteine, Interleukin 6, magnesium, vitamin D, calcium, phosphate, creatine kinase, myoglobin, FGF19, or FGF21.
  • the method may further comprise: receiving patient data, obtained using one or more respective wearable devices, of the respective patients undergoing GLP-1 receptor agonist therapies for weight loss; and providing patient physiological data based on the received patient data of the respective patients undergoing GLP-1 receptor agonist therapies for weight loss to the one or more computation models.
  • the patient data may comprise activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of elevated resting heart rate.
  • the method may further comprise processing the physiological data before provision to the one or more trained computational models, wherein the processing comprises processing the physiological data to generate data indicative of changes in daily heart rate variability.
  • the method may further comprise processing the physiological data before provision to the one or more (32736-2072) trained computational models, wherein the processing comprises processing the physiological data to generate data indicative of changes in cardiac rhythm.
  • the method may further comprise processing the physiological data to identify development of irregular heartbeats that are indicative of possible detection of subacute kidney injury or disfunction.
  • the patient data may comprise activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of decreases in steps taken.
  • the patient data may comprise activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of decreases in average walking speed.
  • the patient data may comprise activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of changes in walking gait.
  • the patient data may comprise activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of sleep quality or sleep duration.
  • the method may further comprise training the one or more computational models on the received training data of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise biochemical sensor data and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method for managing weight-loss patients including patients undergoing one or more GLP-1 receptor agonist therapies, comprising: receiving biochemical sensor data, obtained using biochemical sensing systems, of respective patients undergoing GLP-1 receptor agonist therapies; providing the received patient biochemical data or processed versions of the received patient biochemical data to one or more trained computational models to classify the respective GLP-1 receptor agonist therapy patients according to one or more (32736-2072) patient classifications, wherein the one or more patient classifications are inferred from previously received sets of training data, the training data comprising biochemical data for patients categorized by changes in one or more calculated biological age metrics from blood testing measured before and after GLP-1 receptor agonist therapy for weight loss; and providing outputs to patients and/or clinicians based on the one or more patient classifications, wherein the provided outputs are indicative of one or more risks or benefits of continuing therapy.
  • the method can allow indications of the risks/benefits of continuing GLP- 1 receptor agonist therapy to be provided for patients undergoing such therapy based on changes in biological age.
  • the biochemical sensor data may comprise data related to measurement of one or more of Serum Amyloid A, C-reactive protein, haptoglobin, fibrinogen, lactate, or norepinephrine.
  • the biochemical sensor data may comprise data indicative of loss of lean muscle mass.
  • the method may further comprise processing the biochemical sensor data of the respective patients before provision to the one or more trained computational models, wherein the processing comprises processing the biochemical sensor data and/or the activity data to identify one or more of change in creatine levels, variation in pH levels, change in fluid retention levels.
  • the biochemical sensor data may comprise data from measurement of one or more stress biomarkers selected from: catecholamines, cortisol, growth hormone, glucagon, glutamine, glucose, and renin.
  • the method may further comprise processing the biochemical sensor data before provision to the one or more trained computational models, wherein the processing comprises processing the biochemical sensor data to identify biomarker data indicative of elevated lactate levels and comparing that biomarker data to an indication of physical activity, to identify biomarker data indicative of elevated lactate levels that is correlated to little or no physical activity generating elevated lactate levels.
  • the biochemical sensor data may comprise biomarker data from measurement of one or more of ⁇ -hydroxybutyrate, acetyl-CoA, acetylated histones, or non-histone proteins.
  • the biochemical sensor data may comprise biomarker data from measurement of cytokines or pro-inflammatory enzyme levels.
  • the biochemical sensor (32736-2072) data may comprise biomarker data from measurement of one or more of fibrinogen, homocysteine, Interleukin 6, magnesium, vitamin D, calcium, phosphate, creatine kinase, myoglobin, FGF19, or FGF21.
  • the method may further comprise: receiving patient data, obtained using one or more respective wearable devices, of the respective patients undergoing GLP-1 receptor agonist therapies for weight loss; and providing patient physiological data based on the received patient data of the respective patients undergoing GLP-1 receptor agonist therapies for weight loss to the one or more computation models.
  • the patient data may comprise activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of elevated resting heart rate.
  • the method may further comprise processing the physiological data before provision to the one or more trained computational models, wherein the processing comprises processing the physiological data to generate data indicative of changes in daily heart rate variability.
  • the method may further comprise processing the physiological data before provision to the one or more trained computational models, wherein the processing comprises processing the physiological data to generate data indicative of changes in cardiac rhythm.
  • the method may further comprise processing the physiological data to identify development of irregular heartbeats that are indicative of possible detection of subacute kidney injury or disfunction.
  • the patient data may comprise activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of decreases in steps taken.
  • the patient data may comprise activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of decreases in average walking speed.
  • the patient data may comprise activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to (32736-2072) generate data indicative of changes in walking gait.
  • the patient data may comprise activity data obtained using the wearable device.
  • the method may further comprise processing the activity data before provision to the one or more trained computational models, wherein the processing comprises processing the activity data to generate data indicative of sleep quality or sleep duration.
  • the method may further comprise training the one or more computational models on the received training data of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise biochemical sensor data and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method of conducting glucose monitoring in patients using personal glucose monitoring systems comprising: receiving biochemical sensor data, obtained using biochemical sensing systems, of respective patients undergoing GLP-1 receptor agonist therapies; providing the received patient biochemical sensor data or processed versions of the received patient biochemical sensor data to one or more trained computational models to classify the respective patients undergoing one or more GLP-1 receptor agonist therapies according to one or more patient classifications, wherein the one or more patient classifications are inferred from modifications in metabolic activity of patients in response to one or more GLP-1 receptor agonist therapies based on previously received training data of patients undergoing GLP-1 receptor agonist therapies; subsequent to completion of the one or more GLP-1 receptor agonist therapies for the respective patients, selecting for each respective patient a post-therapy metabolic model from a plurality of post- therapy metabolic models based on the one or more patient classifications; and providing the respective selected post-therapy metabolic models to corresponding personal glucose monitoring systems of each respective patient.
  • the method can allow changes in metabolic activity to be detected and classified for patients undergoing GLP-1 receptor agonist therapy, which can be used to update a metabolic model for a glucose monitoring system.
  • the method may comprise providing one or more outputs to one or more patients and/or clinicians based on the classification.
  • the method may further comprise: calculating doses (32736-2072) of insulin or timing of doses to be provided by one or more medical devices of each respective patient based on monitored glucose levels of the respective patients; and controlling operation of the one or more medical devices according to the calculated doses or timing of doses.
  • the one or more medical devices may include insulin infusion devices.
  • the one or more medical devices may include insulin pen devices.
  • the method may further comprise: receiving patient data, obtained using one or more respective wearable devices, of the respective patients undergoing GLP-1 receptor agonist therapies for weight loss; and providing patient physiological data based on the received patient data of the respective patients undergoing GLP-1 receptor agonist therapies to the one or more computation models.
  • the method may further comprise: estimating changes in insulin sensitivity experienced by the respective patients; and providing the estimated changes in insulin sensitivity to the one or more computational models. Estimating the changes in insulin sensitivity may comprise: identifying patient physical activity, using the one or more respective wearable devices, of the respective patients; and correlating the patient physical activity to glucose levels of the respective patients.
  • the method may comprise: estimating changes in insulin secretion capability experienced by the respective patients; and providing the estimated changes in insulin secretion capability to the one or more computational models.
  • the method may further comprise training the one or more computational models on the received training data of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise biochemical sensor data and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer-implemented method of conducting operations of implantable cardiac devices in patients comprising: receiving biochemical sensor data, obtained using biochemical sensing systems, of respective patients undergoing GLP-1 receptor agonist therapies; providing the received patient biochemical sensor data or processed versions of the received patient biochemical sensor data to one or more trained computational models to classify the respective patients undergoing one or more GLP-1 receptor agonist therapies according to one or more patient classifications, wherein the one or more patient classifications are inferred from modifications in cardiac activity of patients in response to one or more one or more GLP-1 (32736-2072) receptor agonist therapies based on previously received training data of patients undergoing GLP-1 receptor agonist therapies; subsequent to completion of the one or more one or more GLP-1 receptor agonist therapies for the respective patients, selecting for each respective patient a post-therapy cardiac model from a plurality of post-therapy cardiac models based on the one or more patient classifications; providing the respective selected post-therapy cardiac models to corresponding implantable cardiac devices of each respective patient; monitoring cardiac activity in
  • the method can allow changes in cardiac activity to be detected and classified for patients undergoing GLP-1 receptor agonist therapy, which can be used to update a cardiac model for cardiac rhythm management by an implantable cardiac device.
  • the cardiac rhythm management operations may include one or more of: pacing operations, cardioversion operations, and defibrillation operations.
  • the method may comprise providing one or more outputs to one or more patients and/or clinicians based on the classification.
  • the method may further comprise training the one or more computational models on the received training data of patients undergoing GLP-1 receptor agonist therapies.
  • the training data may comprise biochemical sensor data and an indication of an outcome of GLP-1 receptor agonist therapies.
  • a computer program, computer program product, or computer readable medium comprising instructions which, when executed by a computer, cause the computer to carry out any of the methods described herein.
  • a computer system configured to perform any of the methods described herein.
  • GLP-1 receptor agonists have been shown to result in weight loss without the risks associated with an invasive surgical procedure. As previously noted, GLP-1 receptor agonists induce adverse effects in patients which can be uncomfortable or unpleasant but, in most cases as reported in clinical literature, do not have unduly high probabilities of severe adverse effects (in the short term at least). Cases of morbidity from gastroparesis and/or bowel blockages have been reported, but such cases are rare given the population size that has been treated with these pharmaceutic products for diabetes and/or weight loss. Although published literature generally suggests that GLP-1 receptor agonists do not pose significant risk to patients in the short term, Applicant posits that all risk should not be completely dismissed.
  • troponin is a protein that is released when cardiac tissue is damaged and troponin can be readily detected via a blood test.
  • cardiac performance can be evaluated using an in-clinic electrocardiogram evaluation by a cardiologist.
  • low level or sub-acute cardiac damage that may occur in patients receiving GLP-1 receptor agonist therapies may not be detected through conventional medical tests in patients receiving GLP-1 receptor agonist therapies if, for example, significant increases in troponin levels are not induced.
  • GLP-1 receptor agonist therapies cause profound system effects which are not completely understood, patients receiving these therapies may exhibit remodeled physiological functions that differ from the physiological functions in other individuals with similar levels of obesity that did not receive such therapies. For example, patients who are able to reduce their body fat levels to 35% or below from higher levels of body fat with these drugs may exhibit different metabolic function than other individuals that are naturally at a body fat level of 35%. For example, Applicant posits that certain (but not necessarily all) patients receiving GLP-1 agonist therapies may exhibit negative impact on their insulin sensitivity and other metabolic processes years after receiving therapy. [0141] Further, in certain patients, GLP-1 receptor agonist therapy induced weight-loss has anecdotally generated changes in patient appearance.
  • Applicant posits that the occurrence of significant changes (32736-2072) in facial appearance in patients with lower amounts of weight lost and/or who have begun therapy from a lower BMI level could not be solely attributed to loss of facial fat and reduction in skin elasticity. Instead, Applicant posits that reports of changes in facial appearance may indicate that systemic stress may contribute to Ozempic face. Ozempic face may also be a symptom of premature aging. [0143] Accordingly, although published clinical data supports the proposition that GLP-1 agonist therapies benefit lifespan and quality of life overall in a population of patients, Applicant proposes that personalized patient management and other methods as described herein is likely to improve these therapies.
  • FIG.1 is system for monitoring patients using wearable biochemical sensor 102 that may be worn by patients considering, undergoing, or have undergone GLP-1 agonist therapy.
  • the mechanical housing, structure, and application of wearable sensor system 102 may follow the design of Abbott’s FREESTYLE LIBRE TM system that performs continuous glucose monitoring for patients.
  • Wearable sensor system 102 may differ from Abbott’s FREESTYLE LIBRE TM system in that the FREESTYLE LIBRE TM system is (32736-2072) intended for continuous glucose monitoring.
  • Wearable sensor system 102 may include additional or alternative biochemical sensing components to support sensing other than or in addition to sensing of glucose.
  • System 102 may include one or more additional micro-needles to support one or more sensing sites (e.g., for oxidase or other reactions for the biochemical analyte detection).
  • sensing system 100 includes sensor control device 102 and reader device 120 that are configured to communicate with one another over a local communication path or link, which may be wired or wireless, uni- or bi-directional, and encrypted or non-encrypted.
  • Reader device 120 may constitute an output medium for viewing analyte concentrations and alerts or notifications determined by device 102 or a processor associated therewith, as well as allowing for one or more user inputs, according to some embodiments.
  • Reader device 120 may be a multi-purpose smartphone or a dedicated electronic reader instrument. While only one reader device 120 is shown, multiple reader devices 120 may be present in certain instances.
  • Reader device 120 may also be in communication with remote terminal 170 and/or trusted computer system 180 via communication path(s)/link(s) 141 and/or 142, respectively, which also may be wired or wireless, uni- or bi-directional, and encrypted or non-encrypted.
  • Reader device 120 may also or alternately be in communication with network 150 (e.g., a mobile telephone network, the internet, or a cloud server) via communication path/link 151.
  • Network 150 may be further communicatively coupled to remote terminal 170 via communication path/link 152 and/or trusted computer system 180 via communication path/link 153.
  • device 102 may communicate directly with remote terminal 170 and/or trusted computer system 180 without an intervening reader device 120 being present.
  • device 102 may communicate with remote terminal 170 and/or trusted computer system 180 through a direct communication link to network 150, according to some embodiments, as described in U.S. Patent Application Publication 2011/0213225 and incorporated herein by reference in its entirety.
  • Any suitable electronic communication protocol may be used for each of the communication paths or links, such as near field communication (NFC), radio frequency identification (RFID), BLUETOOTH® or BLUETOOTH® Low Energy protocols, WiFi, or the like.
  • Remote terminal 170 and/or trusted computer system 180 may be (32736-2072) accessible, according to some embodiments, by individuals other than a primary user who have an interest in the user's analyte levels.
  • Reader device 120 may comprise display 122 and optional input component 121.
  • Display 122 may comprise a touch-screen interface, according to some embodiments.
  • Sensor control device 102 includes sensor housing 103, which may house circuitry and a power source for operating sensor(s) 104.
  • a processor may be communicatively coupled to sensor(s) 104, with the processor being physically located within sensor housing 103 or reader device 120.
  • Sensor(s) 104 protrudes from the underside of sensor housing 103 and extends through adhesive layer 105, which is adapted for adhering sensor housing 103 to a tissue surface, such as skin, according to some embodiments.
  • sensor(s) 104 are adapted to be at least partially inserted into a tissue of interest, such as within the dermal or subcutaneous layer of the skin.
  • Sensor 104 may comprise a sensor tail of sufficient length for insertion to a desired depth in a given tissue.
  • the sensor tail may comprise, for example, at least one working electrode and a creatinine-responsive active area disposed thereon.
  • the main housing 103 Upon application to the patient using a suitable tool, the main housing 103 is held in place against the patient’s skin with patch or component 105 with adhesive properties. Also, upon application, one or more micro-needles supporting sensor(s) 104 may be automatically deployed that permit access to the patient’s interstitial fluid.
  • An enzyme substrate or other sensing surface is used to detect glucose or any other suitable biochemical molecule, compound, or other relevant analyte using, for example, an oxidase reaction at the enzyme site within the patient’s interstitial fluid.
  • a system controller component (not shown), e.g., including one or more processors and relevant firmware instructions, within device 102 controls operations of system 102 to take measurements of the relevant analyte levels in interstitial fluid at respective intervals. Further, the system controller component may process the measurements according to one or more suitable algorithms (although this may be performed on an external device). The relevant raw and/or processed measurement data is stored in memory.
  • An external device 120 e.g., a smartphone or smartwatch with one (32736-2072) or more suitable “apps” 122) may obtain the stored measurement data from system 102 using Bluetooth or other suitable communication methods.
  • patients considering, undergoing, or have undergone GLP-1 agonist therapy interact with one or more clinicians using a “remote care management system.”
  • a remote care management system is the NEUROSPHERETM DIGITAL CARE / VIRTUAL CLINIC TM system which is a connected care management platform compatible only with ABBOTT (Plano, TX) products.
  • ABBOTT Plano, TX
  • the virtual clinic/remote care system can be modified to support remote support of patients by clinicians for any type of disorder and/or interacting with a variety of patient devices.
  • such a system may be modified to support patients considering, undergoing, or have (32736-2072) undergone GLP-1 agonist therapy and interacting with wearable and/or implantable IMD or other devices of such patients.
  • a remote care service system 200 configured to support remote patient therapy as part of an integrated remote care service session according to one or more embodiments of the present patent disclosure.
  • a “remote care system” may describe a healthcare delivery system configured to support a remote care service over a network in a communication session between a patient and a clinician wherein telehealth or telemedicine applications involving remote medical consultations.
  • the remote medical consultations may be guided using data from a variety of patient devices 270 (including, for example, sensor system 100 as shown in FIG.1 and a variety of other patient devices).
  • patient devices 270 including, for example, sensor system 100 as shown in FIG.1 and a variety of other patient devices.
  • the architecture of remote care system 200 includes patient device 250 and clinician device 280, each having a corresponding remote care service application module, e.g., a patient application 252 and a clinician application 282, executed on a suitable hardware/software platform for supporting a remote care service that may be managed by a network entity 255.
  • the network entity 255 may comprise a datacenter or cloud-based service node (e.g., disposed in a public cloud, a private cloud, or a hybrid cloud, involving at least a portion of the Internet) operative to host a remote care session management service 257.
  • patient application 252 and clinician application 282 may each include a respective remote session manager 254, 284 configured to effectuate or otherwise support a corresponding communication interface 260, 290 with network entity 255 using any known or heretofore unknown communication protocols and/or technologies.
  • interfaces 260, 290 are each operative to support an AV channel or session 263A, 263B and a remote device channel or session 265A, 265B, respectively, with an AV communication service 261A and a remote therapy session service 261B of the remote care session management service 257 as part of a common bi- directional remote care session 259, 299 established therewith.
  • patient application 252 and clinician application 282 may each further include or otherwise support suitable graphical user interfaces (GUIs) and associated controls 256, 286, as well as corresponding AV managers 258, 288, each of (32736-2072) which may be interfaced with respective remote session managers 254, 284.
  • GUIs graphical user interfaces
  • Remote care session manager 254 of the patient application 252 and remote care session manager 284 of the clinician application may each also be interfaced with a corresponding logging manager 262, 286 according to some embodiments.
  • Remote care session manager 254 of patient application 252 is further interfaced with a security manager 264, which may be configured to facilitate secure or trusted communication relationships with the network entity 255.
  • remote care session manager 284 of clinician application 282 may also be interfaced with a security manager 288 that may be configured to facilitate secure or trusted communication relationships with the network entity 255.
  • Each security manager 264, 288 may be interfaced with a corresponding communication manager 266, 290 with respect to facilitating secure communications between the clinician device 280 and the patient device 250.
  • Therapy communication manager 266 of the patient application 252 may also interface with a local communication module 268 operative to effectuate secure communications with the patient's IMD(s) 270 using suitable wireless communication (e.g., Bluetooth).
  • security managers 264, 288 of patient and clinician applications 252, 282 may be configured to interface with the remote care session management service 257 to establish trusted relationships between patient device 250, clinician device 280 and/ or IMD(s) 270 based on the exchange of a variety of parameters, e.g., trusted indicia, cryptographic keys and credentials, etc.
  • the integrated remote care session management service 257 may include a session data management module 271, an AV session recording service module 275 and a registration service module 283, as well as suitable database modules 273, 285 for storing session data and user registration data, respectively.
  • Skilled artisans will realize that the example remote care system architecture 200 set forth above may be advantageously configured to provide both telehealth medical consultations while the patient and the clinician/provider are not in close proximity of each other (e.g., not engaged in an in-person office visit or consultation).
  • a remote care service of some embodiments may form an integrated healthcare delivery service effectuated via a common application user interface that not only allows healthcare professionals to use electronic communications to evaluate and diagnose patients remotely but also facilitates (32736-2072) interaction with and/or programming of the patient's IMD(s) or other devices to support patient care.
  • Virtual clinic and remote programming platforms are discussed in detail in the following applications which are incorporated herein by reference: U.S. Patent App. Pub. No.20200398062; U.S. Patent App. Pub. No.20230317303; U.S. Patent App. Pub. No.20230100246; U.S. Patent App. Pub. No.20220105350A1; U.S. Patent App. Pub.
  • patient physiological data, medical history, device data, and/or other data for patients considering, undergoing, or have undergone GLP-1 agonist therapy are aggregated and subjected to processing to train and/or for use by one or more AI/ML or other computational models.
  • data collected from patient devices and other sources may be employed for patient management using, for example, a cloud-centric digital healthcare network architecture as illustrated in FIG.3 according to some embodiments.
  • Example architecture 360 may include one or more virtual clinic platforms 314, remote data logging platforms 316, patient/clinician report processing platforms 318, as well as data analytics platforms 320 and security platforms 322, at least some of which may be configured and/or deployed as an integrated digital health infrastructure 312.
  • One or more populations of patients are collectively shown at reference numeral 304, wherein individual patients may be provided with one or more suitable IMDs, partially implantable medical devices, other personal biomedical devices, etc., depending on respective patients' health conditions and/or treatments.
  • a plurality of clinician devices 308, patient devices 310, and authorized third-party devices 31 associated with respective users may be deployed as external devices 306 that may be configured to interact with patients' IMDs and/or sensor/tracking devices for effectuating therapy, monitoring, data logging, secure file transfer, etc., via local communication paths or over network-based remote communication paths established in conjunction with the digital health infrastructure network 312. (32736-2072)
  • One or more remote data logging platforms 316 of system 300 may be configured to obtain, receive or otherwise retrieve data from patient devices, clinician devices and other authorized third-party devices.
  • Patient aggregate data 350 is available for processing, analysis, and review to optimize patient outcomes for individual patients, for a patient population as a whole, and for relevant patient sub-populations of patients.
  • Patient aggregate data (PAD) 350 may include basic patient data including patient name, age, and demographic information, etc.
  • PAD 350 may also include information typically contained in a patient's medical file such as medical history, diagnosis, results from medical testing, medical images, etc.
  • the data may be inputted directly into system 300 by a clinician or medical professional. Alternatively, this data may be imported from digital health records of patients from one or more health care providers or institutions.
  • a patient may employ one or more patient “apps” on the patient's smartphone or other electronic device to control or interact with patient's biochemical sensor system, IMD(s), and/or minimally invasive medical device.
  • analyte measurements may be retrieved from one or more patient devices for aggregation by system 300.
  • patient interaction with one or more apps may be monitored and logged.
  • the patient app may be adapted to log such events (“Device Use/Events Data”) and communicate the events to system 300 to maintain a history for the patient for review by the patient's clinician(s) to evaluate and/or optimize the patient's care as appropriate.
  • PAD 350 may include “Patient Self-Report Data” obtained using a digital health care or wellness app operating on patient devices 310.
  • the patient self-report data may include patient reported levels of various conditions, patient well-being scores, emotional states, activity levels, and/or any other relevant patient reported information.
  • PAD 350 may include sensor data such as sensor analyte data. Data captured using such sensors can be communicated from the medical devices to patient controller devices and then stored within patient/clinician data logging and monitoring platform 316. Patients may also possess wearable devices such as health monitoring products (heart rate monitors, fitness tracking devices, smartwatches, etc.). Any data (32736-2072) available from wearable devices may be likewise communicated to monitoring platform 316.
  • PAD data 350 may include video analytic data for individual patients, patient sub- populations, and the overall patient population for each supported therapy.
  • the data may comprise various data logs that capture patient-clinician interactions (“Remote Programming Event Data” in PAD 350), e.g., individual patients' therapy/program settings data in virtual clinic and/or in-clinic settings, patients' interactions with remote learning resources, physiological/behavioral data, daily activity data, and the like.
  • Clinicians may include clinician reported information such as patient evaluations, diagnoses, etc. in PAD 350 via platform 316 in some embodiments.
  • data obtained via remote monitoring, background process(es), baseline queries and/or user-initiated data transfer mechanisms may be (pre)processed or otherwise conditioned in order to generate appropriate datasets that may be used for training, validating and testing one or more AI/ML-based models or engines for purposes of some embodiments.
  • patient input data may be securely transmitted to the cloud-centric digital healthcare infrastructure wherein appropriate AI/ML-based modeling techniques may be executed for evaluating the progress of the therapy trial, predicting efficacy outcomes, providing/recommending updated settings, etc.
  • AI/ML Data may be used as a qualitative term for a collection of datasets comprising the relevant data (including, but not limited to, PAD 350) for management of patient therapy as described herein. Because “AI/ML Data” available with respect to patients' health data, (32736-2072) physiological/behavioral data, sensor data gathered from patients and respective ambient surroundings, daily activity data, therapy settings data, health data collected from clinicians, etc.
  • platform 320 may employ suitable infrastructure implementations to support AI/ML and computational model processing.
  • processes may be implemented in a “massively parallel processing” (MPP) architecture with software running on tens, hundreds, or even thousands of servers.
  • MPP massively parallel processing
  • data analytics platform 320 may be configured to train various AI/ML-based models, computational models, or decision engines for purposes of some example embodiments of the present patent disclosure.
  • Various supervised and unsupervised learning and/or reinforcement techniques such as support vector machines (SVMs), support vector networks (SVNs), Naive Bayes (NB), neural networks (e.g., ANNs/CNNs), k-nearest neighbor, decision tree (DT), back-propagation neural network (BPNN), support vector regression (SVR), multiple linear regression (MLR), partial least square (PLS), k-Means, hierarchical algorithm (HA), mean-shift, density-based spatial clustering of application with noise (DBSCAN), feature selection, feature extraction, Q-learning, temporal difference, value iteration, and/or Markov decision processing.
  • SVMs support vector machines
  • SVNs support vector networks
  • NB Naive Bayes
  • neural networks e.g., ANNs/CNNs
  • DT back-propagation neural network
  • BPNN back-propagation neural network
  • SVR support vector regression
  • MLR multiple linear regression
  • PLS partial least square
  • supervised learning may comprise a type of machine learning that involves training a predictive model based on decision trees built from a training sample to go from observations about a plurality of features or attributes and separating the members of the training sample in an optimal manner according to one or more predefined features.
  • Tree models where a target variable can take a discrete set of values are referred to as classification trees, with terminal nodes or leaves representing class labels and nodal branches representing conjunctions of features and thresholds that indicate the most likely class labels.
  • an embodiment of the present patent disclosure may advantageously employ supervised learning that involves ensemble techniques where more than one decision tree (typically, (32736-2072) a large set of decision trees) are computed.
  • a boosted tree technique may be employed by incrementally building an ensemble by training each tree instance to emphasize the training instances previously mis-modeled or mis-classified.
  • bootstrap aggregated (i.e., “bagged”) tree technique may be employed that builds multiple decision trees by repeatedly resampling training data with or without replacement of a randomly selected feature.
  • some example embodiments of the present patent disclosure may involve a Gradient Boosted Tree (GBT) ensemble of a plurality of regression trees and/or a Random Forest (RF) ensemble of a plurality of classification trees, e.g., in patient health score classification and modeling.
  • GBT Gradient Boosted Tree
  • RF Random Forest
  • the ‘081 describes analysis of patient activities for the purpose of controlling a neurostimulation system, similar processing of patient activities may be employed to manage patients receiving a GLP-1 receptor agonist therapy as described herein for some embodiments.
  • patient conditions, physiological states, measured analyte levels, and/or the like e.g., patient data levels 401 and 402 plotted in time
  • activities of the patient are monitored and detected using one or more sensors of an implantable device or an external device.
  • the sensors may include sensors for sensing physiological conditions, sensors for detecting movement or location, and/or any other suitable sensors.
  • an activity profile for the patient is determined that represents expected times when the patient will engage in a plurality of different activities of the patient.
  • patient activity may be detected using location determining circuitry and location-based algorithms to correlate location to activity.
  • Microlocation processing algorithms may be employed to determine patient activity within the patient's domicile as one example.
  • the monitoring of activities of a patient may include repetitively detecting a location of the patient using location determining circuitry of the external device of the patient.
  • the circuitry for location-based (32736-2072) activity tracking may include cellular communication circuitry, WiFi circuitry, and Bluetooth circuitry.
  • Location-based activity detection may include detecting an amount of time spent at an identified location.
  • monitoring activities of the patient may comprise obtaining data pertaining to physiological signals of the patient using a wearable device or an implanted device.
  • the physiological signals may include heart rate data, electrocardiogram data, a sleep quality data, body temperature data, blood oxygen saturation data, blood glucose data, other measured analyte levels, and/or any other measured physiological-related data.
  • an external controller or other user device may receive user input from the patient by the external controller that is indicative of patient activities being performed by the patient.
  • the patient may provide user input by selecting respective ones of activity icons displayed on or more user interface screens where each respective icon represents a distinct patient activity.
  • the user interface screen(s) may receive input from the user indicative of ease or difficulty for the patient in performing a respective activity. Also, the user interface(s) may receive input from the user indicative of one or more patient states (e.g., a wellness level, an overall “energy” level, a level of pain experienced by the patient at a respective point in time, etc.).
  • a patient activity profile is generated from the activity data collected from the implanted and/or external devices of the patient.
  • the patient activity data is communicated to a remote care management system, wherein the remote care management system determines the activity profile for the patient.
  • the remote care management system may perform an averaging calculation of observed times for patient activities of the activity profile; calculate average start times of respective activities for the activity profile; may calculate average end times of respective activities for the activity profile; apply a calculation of frequency of performance of activities to determine the activity profile; and/or apply an averaging calculation to determine average duration of activities for the activity profile.
  • Such (32736-2072) suitable processing of patient activity data into activity metrics may be employed to create a patient activity profile.
  • FIG.5 depicts a flowchart representing a series of events for a patient beginning with health coaching through weight-loss through GLP-1 receptor agonist therapy, and, possibly, to implant of one or more medical devices.
  • the patient creates an account and downloads a “Health App” for the patient’s smartphone or other device, and uses the app.
  • the Health App may provide general health guidance, nutritional guidance, and similar content to the patient.
  • the Health App may employ tracking of patient data (e.g., diet, exercise, physiological states such as cardiac activity, blood oxygenation, sleep, etc.) based on user input and/or sensor data. Such patient data may be aggregated as discussed herein. Also, the patient data may be processed to generate automated suggestions for the patient based on the gathered data.
  • patient uses the Health App to obtain health coaching from one or more clinicians using virtual clinical features of the app.
  • the patient can discuss specific health issues with a doctor or other clinician without requiring an in-clinic appointment using app interface 601.
  • the patient consults with one or more clinicians and obtains biochemical sensor(s) (see, e.g., discussion of device in FIG.1 herein) and/or fitness tracking wearable device (e.g., smartwatch 701 with one or more suitable health tracking apps as shown in FIG.7) to assist tracking the patient’s physiological states and/or patient activities.
  • the patient links the device(s) to patient (32736-2072) account.
  • the patient data can be processed and/or analyzed on an individual basis and subjected to aggregated processing and/or analysis with data from other patients.
  • the aggregated (sample-level) data processing may include training and/or reinforcement of one or more AI/ML or other computational models.
  • the patient data may be subject to processing based on AI/ML or other computational models trained from sample-level data aggregated by the systems described herein. Based on the output from AI/ML or other computational models, the patient may receive automated guidance and/or clinician guidance. [0192]
  • the patient’s clinician identifies the patient as a candidate for GLP-1 receptor agonist therapy.
  • the patient begins a pre-therapy preparation phase.
  • the biochemical sensor may include glucose monitoring and such data may be used to provide guidance for diet and exercise choices for the patient.
  • the clinician or automated system output may make recommendations to the patient to optimize the outcome of the GLP-1 receptor agonist therapy.
  • recommendations to the patient may include alterations of the patient dietary habits to increase or supplement the patient’s ordinary diet with nutritional components to minimize the probability of losing lean muscle mass during the GLP-1 receptor agonist therapy.
  • the biochemical sensor may include measurements of one or more other analytes to provide a profile of the patient which can be subsequently used for baseline analysis after the patient commences GLP-1 receptor agonist therapy. It may be beneficial to conduct baseline measurements after initial commencement of pre- therapy preparation phase. For example, changes in the patient diet or other lifestyle changes (e.g., protein supplementation) may alter any such baseline measurements. By proceeding in this manner, it may be beneficial to separate changes from such changes from changes subsequently induced by the GLP-1 receptor agonist therapy. (32736-2072) [0195] In 506, the patient begins GLP-1 receptor agonist therapy.
  • patient data from one or more biochemical sensors and/or fitness tracking device is obtained and communicated to the patient account.
  • patient data from the respective devices e.g., biochemical sensor and/or fitness tracking device
  • AI/ML or other computational models are applied to one or more AI/ML or other computational models.
  • the device data can be augment by any other data such as in-person medical testing. Multiple such AI/ML or other computational models are discussed herein.
  • One or more of AI/ML or other computational models may output one or more patient state scores or classifications for review by clinician(s) and/or patients (see, e.g., screen 800 in FIG.8).
  • one or more of the AI/ML or other computational models may be trained to classify whether the patient data is properly classified as representing systemic stress of the patient indicative of premature aging.
  • One or more of AI/ML or other computational models may output one or more recommendations or evaluations of therapy as discussed herein.
  • AI/ML or other computational models may be trained to classify other patient responses to be avoided, such as immune-system mediated reactions to the therapy or endothelial damage.
  • the output of the one or more AI/ML or other computational models is provided to patient’s clinician(s) to assist the medical determination whether stop therapy, modify dosage amount and/or dosage timing, and/or provide adjunctive therapy/therapies.
  • the output of the one or more AI/ML or other computational models provides one or more metrics representing a patient state, condition, or response to therapy.
  • the output of the one or more AI/ML or other computational models provides information relative to a given patient’s state, condition, or response to therapy relative to another matched sample of patients (e.g., patients who experienced weight loss without GLP-1 receptor agonist therapy, patients who experienced weight loss with GLP-1 receptor agonist therapy successfully and/or with suitably evaluated responses, etc.).
  • the output of the one or more AI/ML or other computational models classifies the patient into one or more respective categories (responder, non-responder, subject to sub-acute damage, etc.).
  • the output of the one or more AI/ML or other computational models provides one or more recommendations (e.g., stop (32736-2072) therapy, continue therapy, change to other GLP-1 receptor agonist product, change timing of therapy, change dosage of therapy, etc., provide adjunctive therapy to reduce probability of patient injury or sequelae, etc.).
  • Example patient output screen 801 generated using one or more AI/ML or other computation model is shown in FIG.8 according to some embodiments.
  • the patient may benefit from an implantable medical device, semi-implantable medical device, or other personal medical device. For example, the patient may develop diabetes requiring insulin injections.
  • Automated personal insulin delivery devices in the form of “insulin pens”) capable of delivering variable amounts of insulin (possible with other pharmaceutical compounds) may benefit such patients.
  • Smart insulin delivery systems may be employed such as the BIGFOOT UNITY TM Diabetes Management System (available from Abbott, Abbott Park, Illinois).
  • automated insulin pumps may assist management of the patient’s diabetes.
  • the patient may eventually benefit of implantable cardiac rhythm management (CRM) devices such as pacemakers and cardioverter/defibrillator devices.
  • CCM implantable cardiac rhythm management
  • pacemakers and cardioverter/defibrillator devices implantable cardiac rhythm management
  • the operation of such devices may be modified or programmed using data specific to the patient collected using the system described herein, including data related to the patient’s prior GLP-1 receptor agonist therapy.
  • Applicant posits that certain patients that have previously received GLP-1 receptor agonist therapy may exhibit differences in insulin sensitivity relative to other patients and/or differences in metabolic function.
  • AI/ML or other computational models (trained using the data aggregated using the systems described herein) may be employed to identify such patients and to account for such differences in the operation of the patient’s medical device(s).
  • one or more biomarkers or other patient data are employed for training AI/ML or other computational models and/or as inputs to AI/ML or other computational models to evaluate patient response to GLP-1 receptor agonist therapy.
  • the computational models may be configured to perform classifications with respect to success of GLP-1 receptor agonist therapy and/or adverse effects during GLP-1 receptor agonist therapy.
  • Adverse effects may relate to one or more of sub-acute systemic damage, sub-acute cardiac sequelae, sub-acute renal sequelae, muscle loss, systemic stress, oxidative stress, sub-acute cardiac stress, systemic (32736-2072) inflammation, systemic fatigue, premature aging, sagging or aging of facial skin, biological age, blood test metrics, adverse effects compared to a pre-therapy baseline, adverse effects compared to a non-pharmaceutical weight-loss regime, endothelial function (e.g. endothelial damage or changes in endothelial function), metabolic function (e.g. mitochondrial dysfunction, metabolic impairment/dysfunction or other changes in metabolic function), immune system dysregulation, cardiac function (e.g.
  • a number of AI/ML or other computational models are discussed herein, and various AI/ML or other computational models may utilize different combinations of patient data (e.g. biomarkers, activity data and/or physiological data).
  • the patient data may include biochemical biomarkers (such as biochemical sensor data or multi-analyte biochemical sensor data), physiological data (such as patient physiological activity biomarkers), patient activity data, patient reported outcomes, physician reported outcomes, patient physical activity, and/or the like.
  • Biomarkers such as (multi-analyte) biochemical sensor data may be obtained using one or more biochemical sensing systems.
  • Biomarkers such as biochemical sensor data can relate to one or more analytes, so may include biochemical sensor data relating to a single analyte or may include multi-analyte biochemical sensor data.
  • Other types of patient data such as activity data and/or physiological data, may be obtained using one or more monitoring devices associated with the patient, including wearable devices such as a smartwatch and/or fitness tracking device.
  • Patient data received for the purposes of classification using one or more of the trained computational models can be data relating to a single patient (e.g. at multiple points in time) or data relating to one or more groups of patients. [0199]
  • One or more of the computational models may be trained using data of the same type or a similar type as the received patient data.
  • models may be trained using additional types or sources of data (e.g. using comparative data sources for baseline levels). (32736-2072)
  • one or more AI/ML or other computational models are trained using one or more biomarkers to generate a model output from patient data that is reflective of systemic stress of a patient who is receiving or yet to receive GLP-1 receptor agonist therapy.
  • one or more AI/ML or other computational models are trained using one or more biomarkers to generate a model output from patient data that is reflective of premature aging of a patient who is receiving or yet to receive GLP-1 receptor agonist therapy.
  • one or more AI/ML or other computational models are trained using one or more biomarkers to generate a model output from patient data that is reflective of oxidative stress of a patient who is receiving or yet to receive GLP-1 receptor agonist therapy.
  • one or more AI/ML or other computational models are trained using one or more biomarkers to generate a model output from patient data that is reflective of systemic inflammation of a patient who is receiving or yet to receive GLP-1 receptor agonist therapy.
  • one or more AI/ML or other computational models are trained using one or more biomarkers to generate a model output from patient data that is reflective of endothelial damage or dysregulation of a patient who is receiving or yet to receive GLP-1 receptor agonist therapy.
  • one or more AI/ML or other computational models are trained using one or more biomarkers to generate a model output from patient data that is reflective of immune system dysregulation of a patient who is receiving or yet to receive GLP-1 receptor agonist therapy.
  • one or more AI/ML or other computational models are trained using one or more biomarkers to generate a model output from patient data that is reflective of subacute cardiac stress, cardiac function impairment, and/or other cardiac dysregulation of a patient who is receiving or yet to receive GLP-1 receptor agonist therapy. (32736-2072) [0207] In some embodiments, one or more AI/ML or other computational models are trained using one or more biomarkers to generate a model output from patient data that is reflective of subacute kidney function impairment or other dysregulation of a patient who is receiving or yet to receive GLP-1 receptor agonist therapy.
  • one or more AI/ML or other computational models are trained using one or more biomarkers to generate a model output from patient data that is reflective of mitochondrial dysfunction of a patient who is receiving or yet to receive GLP-1 receptor agonist therapy.
  • one or more AI/ML or other computational models are trained using one or more biomarkers to generate a model output from patient data that is reflective of metabolic impairment or dysfunction of a patient who is receiving or yet to receive GLP-1 receptor agonist therapy.
  • any of the described outputs of the one or more AI/ML or other computational models may be combined, processed, or provided as inputs to one or more AI/ML or other computational models to generate an output for a clinician to continue, discontinue, pause, or otherwise modify GLP-1 receptor agonist therapy for a patient and/or to provide adjunctive therapy or therapies for the patient while receiving GLP-1 receptor agonist therapy.
  • the computational model(s) may be configured to perform classification based on patient data (e.g. biomarkers, activity data and/or physiological data) of patients before receiving GLP-1 receptor agonist therapy, patients undergoing GLP-1 receptor agonist therapy, and/or patients after undergoing GLP-1 receptor agonist therapy.
  • the computational model(s) may be configured to perform classifications in relation to patients who are undergoing GLP-1 receptor agonist therapy (e.g. for classifying or monitoring patients after starting therapy) and/or who are pre-GLP- 1 receptor agonist therapy (e.g. for predicting expected outcomes, benefits and/or risks of therapy for pre-therapy patients).
  • one or more first computational models may be configured to perform classifications (e.g. predictions) for pre-therapy patients
  • one or more second computational models may be configured to perform classifications (e.g. detections) for (32736-2072) patients undergoing therapy.
  • multiple models may be used to perform multiple classifications for the same group of patients.
  • one or more first computational models may be configured to perform positive classifications (e.g. relating to likelihood of therapy success) for patients, and one or more second computational models may be configured to perform negative classifications (e.g. relating to adverse effects of therapy) for those same patients.
  • the output(s) of the one or more AI/ML or other computational models may be measured relative to or otherwise compared to a population of other patients who have also received GLP-1 receptor agonist therapy/therapies.
  • the population or population of patients may be categories of patients experiencing negative responses to therapy and/or positive responses to therapy.
  • the population of patients may alternatively or additionally include patients who underwent weight loss without GLP-1 receptor agonist therapy/therapies.
  • the input data and/or the output data of the one or more AI/ML or other computational models may be interpreted, enriched, attenuated, amplified, combined or otherwise modified based on baseline biomarker condition(s) of the patient.
  • Classifications produced by the computational models described herein may be used to provide outputs to patients and/or clinicians. For example, based on the classification(s) produced by one or more computational models, one or more predictions of expected risks and/or benefits of continuing GLP-1 receptor agonist therapy (or continuing such therapy) may be provided. Additionally or alternatively, the outputs may include one or more alerts indicative of occurrence of one or more adverse effects associated with long-term negative outcomes of GLP-1 receptor agonist therapies.
  • the alert(s) may be indicative of one or more of systemic stress, premature aging, sub-acute cardiac sequelae and sub-acute renal sequelae.
  • treatment decisions or recommendations may be provided for a patient, such as recommendations to stop or pause therapy, modify a dosage amount and/or timing, and/or provide one or more adjunctive therapies.
  • one or more patients may be identified for adjunctive therapy to reduce adverse effects (e.g. muscle loss) associated with GLP-1 receptor agonist therapy prior to (32736-2072) commencement of GLP-1 receptor agonist therapy.
  • a post-therapy model (such as a metabolic model or cardiac model) may be selected for a patient.
  • a selected post-therapy metabolic model may be provided to a personal glucose monitoring system of the patient, and/or a selected post-therapy cardiac model may be provided to an implantable cardiac device (or implantable cardiac device controller) for monitoring cardiac activity and/or modifying cardiac management operations.
  • an implantable cardiac device or implantable cardiac device controller
  • GLP-1 receptor agonists generally reduce systemic inflammation and, perhaps, for this reason, it is believed (at least in the short term), patients undergoing such therapies exhibit a lower likelihood of severe cardiac events.
  • GLP-1 receptor agonists concluded that such therapies showed little if any cardiovascular benefit that could not have been a consequence of glucose lowering and further suggested that the net effects of some agents might be harmful on vascular disease (see, e.g., Nissen SE, Wolski K. Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes.
  • GLP-1 receptor agonist therapy for weight loss may present both systemic benefits and systemic stress.
  • the benefits may result of short-term improvements while systemic stress associated with long-term changes in and/or reprogramming of e.g., metabolism and inflammation, that may contribute to the development of other health conditions.
  • Applicant further posits that the risk/benefit profile of receptor agonist therapy for weight loss for a given patient may depend upon the rate at which its long-term effects are generated.
  • BIOMARKER EXAMPLES Applicant posits that the use of one or more AI/ML or other computational models to classify patient response to GLP-1 receptor agonist therapy will improve patient outcomes.
  • suitable biomarkers e.g., of systemic stress
  • suitable patient classification and model training may provide insight into the long-term outcomes for patients undergoing GLP-1 receptor agonist therapies.
  • biomarkers discussed herein may change in a manner indicating improvement of a patient’s overall condition upon commencement of therapy relative to pre-therapy levels. Such biomarkers may also be useful for identifying patients who might not benefit from GLP-1 receptor agonist therapy and/or may experience long-term negative effects. For example, a relative reduction in biomarker improvement in one set of patients versus another set of patients may provide a discriminating parameter to make assessments of the risk/benefit ratio of GLP-1 receptor agonist therapy in the AI/ML or other computational models according to some embodiments. (32736-2072) [0219] In some embodiments, biomarkers related to systemic stress are employed for one or more such AI/ML or other computational models.
  • GLP-1 receptor agonist diabetes therapies have been employed for some period of time and a body of clinical evidence has been developed regarding their use. In general, it is not believed within the medical field that general use of GLP-1 receptor agonist therapy causes increased mortality based on various cardiac conditions.
  • GLP-1 receptor agonist therapies (and certainly for diabetes) produce an overall general health benefits and mortality benefits.
  • GLP-1 receptor agonist therapy may result in systemic stress and/or premature aging in a certain sub-population(s) of patients and thereby negatively impact the health and/or mortality of these patient sub-populations.
  • the causes of the adverse effects may be directly mediated by the GLP-1 receptor agonists alone and/or in combination with other factors that affect systemic stress such as weight loss, genetic or lifestyle related factors.
  • GLP-1 receptor agonist therapies may cause premature cardiac aging which will only become apparent to patients multiple years or more after initiation of such therapies.
  • systemic (32736-2072) effects may be a causative factor.
  • loss of lean muscle mass may result in negative long-term effects on cardiac tissue.
  • Biomarkers indicative of adverse effects on cardiac function may include both biochemical biomarkers and biomarkers detected from electronic wearable devices such as smartwatches and/or fitness tracking devices.
  • heart rate data is used by one or more AI/ML or other computational models to manage GLP-1 receptor agonist therapies for weight-loss patients.
  • the heart rate is used for analysis relative to other factors such as heart rate data before initiation of therapy, heart rate data for patients who have undergone weight-loss therapies without GLP-1 receptor agonist therapies, heart rate data for patients who have undergone GLP-1 receptor agonist therapies for weight-loss with positive outcomes, heart rate data for patients who underwent GLP-1 receptor agonist therapies for weight-loss with negative outcomes, and/or the like.
  • elevation of average heart rate may indicate some level of systemic stress. Elevated (e.g., average) resting heart rate as detected by a smartwatch or fitness tracking device may provide more discriminate value to determining whether such patients are experiencing systemic stress while undergoing such therapies.
  • HRV daily heart rate variability
  • Cardiac rhythm variations may also serve as a biomarker for one or more AI/ML or other computational models to manage GLP-1 therapies for patients for weight-loss according to some embodiments.
  • blood pressure may also serve as a biomarker for such models.
  • cardiac health is related to endothelial function. Accordingly, biomarkers of endothelial function can also serve as measures of cardiac health.
  • Endothelial cells of blood vessels perform functions important for cardiovascular homeostasis by regulating blood fluidity and fibrinolysis, vascular tone, angiogenesis, monocyte/leukocyte adhesion, and platelet aggregation.
  • the vascular endothelium maintains cardiovascular health and vascular endothelium dysfunction is a significant causative factor for many cardiovascular ailments, such as atherosclerosis, aging, hypertension, obesity, and diabetes.
  • Endothelial dysfunction is characterized by imbalanced vasodilation and vasoconstriction, elevated reactive oxygen species (ROS), proinflammatory factors, as well as deficiency of nitric oxide (NO) bioavailability.
  • ROS reactive oxygen species
  • NO nitric oxide
  • Systemic stress can include an environment of oxidative stress that can cause negative effects on the health and function of endothelial cells.
  • oxidative stress can cause acute and delayed expression of leukocyte adhesion molecules by endothelial cells.
  • systemic stress can also trigger the synthesis and secretion of acute phase proteins in the liver, including Serum Amyloid A, C - reactive protein, haptoglobin and fibrinogen, all of which favor a systemic pro-inflammatory, pro- thrombotic state. The persistence of this state may be a major contributor to the development of cardiovascular disease.
  • gluconeogenesis such as increased release of free fatty acids and promotion of insulin resistance (IR)
  • IR insulin resistance
  • GLP-1 receptor agonist therapies may already exhibit IR
  • the long-term effect of GLP-1 receptor agonist therapies on metabolic function may contribute to and/or worsen metabolic function.
  • endothelial nitric oxide synthase eNOS
  • eNOS protein has a role in the modulation, regulation, and control of the cardiovascular system in a normal physiological state and in cardiovascular diseases are discussed.
  • nitric oxide (NO) production and activity have been shown to impairment in nitric oxide (NO) production and activity. Modifications in bioavailability of NO have been found to cause endothelial dysfunction, increasing susceptibility to hypertension, progression of atherosclerosis, hypercholesterolemia, thrombosis, stroke, DM and its complications associated with DM. High levels of NOx are associated with adverse clinical events observed in diabetic patients, such as endothelial dysfunction, insulin resistance and pancreatic beta-cell dysfunction. (see Ta ⁇ s S. Assmann et al., Nitric oxide levels in patients with diabetes mellitus: A systematic review and meta-analysis, Nitric Oxide Volume 61, 30 December 2016, Pages 1-9).
  • one or more biomarkers relevant to cardiac health, endothelial function, systemic stress, and/or premature aging are employed by one or more AI/ML or other computational models to manage patients undergoing GLP-1 receptor agonist therapies.
  • GLP-1 receptor agonist therapy may also result in subacute effects on kidney function that manifests in the long-term.
  • Embodiments described herein are directed to detecting and analyzing subacute effects that may have serious long-term consequences and that would otherwise not be addressed by conventional medical practices.
  • endothelial dysfunction may cause systemic stress involving multiple organs. It has been reported that “[e]ndothelial dysfunction occurs in chronic kidney disease (CKD) and increases the risk for cardiovascular disease. The mechanisms of endothelial dysfunction seem to evolve throughout kidney disease progression culminating in reduced L-arginine transport and impaired nitric oxide bioavailability in advanced disease. (see The Vascular Endothelium in Chronic Kidney Disease: A Novel Target For Aerobic Exercise, Exerc Sport Sci Rev.2016 Jan; 44(1): 12–19., Christopher R.
  • Detection of subacute kidney injury or dysfunction through one or more suitable biomarkers may thus indicate systemic stress and/or premature aging in patients undergoing GLP-1 receptor agonist therapies.
  • biomarkers are used for one or more AI/ML or other computational models for management of patients undergoing such therapies.
  • the measurement of creatine and other related analyte levels using a wearable biochemical sensor may also be useful as a biomarker related to kidney function.
  • biomarkers and other indicators of subacute impaired kidney function may include variations in pH levels, fluid retention, causing swelling in legs, ankles or feet; shortness of breath; fatigue; nausea; weakness; irregular heartbeat; chest pain or pressure.
  • biomarkers/indicators may be obtained using patient reported outcomes via a suitable patient health app, obtained via smartwatch sensor data, and/or via other medical testing.
  • one or more biomarkers (as discussed herein) relevant to subacute kidney dysfunction are employed by one or more AI/ML or other computational models to manage patients undergoing GLP-1 receptor agonist therapies.
  • Applicant discusses biomarkers related to pancreas function in a section specific to glucose and insulin levels, Applicant shall not repeat that analysis in this subsection. While glucose and insulin are related to pancreas function, they are also biomarkers for diabetes and obesity and would not be specific biomarkers of changes in pancreas function in patients receiving GLP-1 receptor agonist therapies. Accordingly, Applicant will discuss in this subsection certain serious acute complications of GLP-1 receptor agonist therapies and how such acute complications may be viewed based on the hypotheses related to embodiments discussed herein.
  • GLP-1 receptor agonist therapies may be associated with pancreatitis, although a complete understanding of the relationship between the therapies and pancreatitis is not yet clear (see Association of Pancreatitis with Glucagon-Like Peptide-1 Agonist Use, Annals of Pharmacotherapy, Volume 44, Issue 5, 2010, Sarah L Anderson et al). Certain GLP-1 receptor agonist products have been reported to increase risk of pancreatitis (see Glucagonlike Peptide 1–Based Therapies and (32736-2072) Risk of Hospitalization for Acute Pancreatitis in Type 2 Diabetes Mellitus: A Population- Based Matched Case-Control Study, Sonal Singh et al., JAMA Intern Med. 2013;173(7):534-539).
  • Applicant’s opinion of such clinical data related to pancreatitis reinforces Applicant’s hypothesis that certain subpopulations of patients receiving GLP-1 receptor agonist therapies for weight loss experience negative systemic effects that are yet not fully understood. Applicant thus posits that the use of AI/ML or other computational models to classify such patients during therapy will lead to improvements in patient outcomes. [0239] Applicant’s hypothesis of systemic stress in patients undergoing GLP-1 agonist therapies leads to the conclusion that changes in physiological processes as modulated by neural activity should be considered and may (but not required for all embodiments) be useful for one or more AI/ML or other computational models for patient management as discussed herein.
  • the neuroinflammation mediated response to systemic stress involves an inflammatory process that may occur by release of neuropeptides (e.g., Substance P (SP)) or other inflammatory mediators, from sensory nerves and the activation of mast cells or other inflammatory cells.
  • neuropeptides e.g., Substance P (SP)
  • SP Substance P
  • CPF Corticosteroid releasing factor
  • SP and other neuropeptides initiate a systemic stress response by activation of neuroendocrinological pathways such as the sympathetic nervous system, hypothalamic pituitary axis, and the renin angiotensin system, with the concomitant release of stress hormones (i.e., catecholamines, corticosteroids, growth hormone, (32736-2072) glucagons, and renin).
  • stress hormones i.e., catecholamines, corticosteroids, growth hormone, (32736-2072) glucagons, and renin.
  • Stress hormones can initiate the acute phase response (APR) and induce acute phase proteins which are essential mediators of inflammation. Changes in neural activity such as changes in sympathetic/parasympathetic activity may also have effects on hormone levels, which may contribute to systemic stress and/or premature aging.
  • Systemic stress may be characterized by release of stress mediators, including corticotropin-releasing hormone (CRH), adrenocorticotropin (ACTH), glucocorticoids (GCs) and the catecholamines, adrenaline and noradrenaline.
  • Inflammatory stress may be characterized by proinflammatory cytokines, such as tumor necrosis factor ⁇ (TNF- ⁇ ), interleukin 1 (IL-1) and IL-6.
  • Mitochondria perform a critical role in cellular homeostasis. Mitochondria contain oxidative phosphorylation machinery to enable aerobic ATP generation, and multiple metabolic pathways, such as ⁇ -oxidation of fatty acids and the tricarboxylic acid and urea cycles. Indeed, over 90% of cellular energy generation takes place in the mitochondria. In addition, mitochondria have important biosynthetic activities, control intracellular Ca2+ metabolism and signaling, regulate thermogenesis, generate most cellular reactive oxygen species (ROS) and serve as the gatekeeper of the cell for programmed cell death (apoptosis).
  • ROS reactive oxygen species
  • Lactate is an important metabolic substrate and an intercellular and inter-tissue redox signaling molecule that provides energy for oxidative metabolism and helps maintain redox homeostasis.
  • Energy metabolism is based on continuous redox reactions which, when occurring in balance, maintain redox homeostasis in cells.
  • disruptions in homeostasis are detrimental to the body, regardless of whether the change leads to excessive oxidation or reduction. For example, when cells enter an oxidative state, the cells produce more active substances, undergo accelerating aging, and potentially cause cardiovascular disease.
  • one or more biomarkers related to systemic oxidation are employed for one or more AI/ML or other computational models to manage patients undergoing GLP-1 receptor agonist therapy for weight loss.
  • one or more biomarkers related to metabolic rate and/or function are employed for one or more AI/ML or other computational models to manage patients undergoing GLP-1 receptor agonist therapy for weight loss.
  • biomarkers may provide insight into systemic stress and/or premature aging of patients undergoing GLP-1 receptor agonist therapy for weight loss.
  • one or more biochemical sensors and/or other sensors are employed to sense biomarkers indicative of cortisol and/or norepinephrine (NE) levels in patients undergoing GLP-1 receptor agonist therapy.
  • NE norepinephrine
  • Cortisol is a steroid hormone which plays a major part in the body's metabolic reaction to psychological and physical stress.
  • Cortisol is secreted by the adrenal glands which are on top of each kidney. Cortisol is a product of the complex (32736-2072) interaction between the hypothalamus and pituitary glands in the brain and the adrenal glands, known as the hypothalamic-pituitary-adrenal (HPA) axis.
  • HPA hypothalamic-pituitary-adrenal
  • Cortisol secretion generally follows a natural 24-hr cycle. In healthy individuals, peak levels are reached about 30 min after waking—this early peak is known as the cortisol awakening response (CAR). Levels decline throughout the day, with lowest levels occurring during the early sleeping phase. However, prolonged exposure to stressors can lead to the overstimulation of the HPA axis and result in fluctuating cortisol levels.
  • cortisol levels at any specific time may not provide accurate insight into a patient’s response to GLP-1 receptor agonist therapy. Accordingly, a conventional blood test measuring the cortisol level of a patient at a specific time will provide a clinician with little to no guidance regarding whether the patient’s GLP-1 receptor agonist therapy should be continued, discontinued, or modified or whether some other intervention should occur.
  • cortisol levels may provide insight into the patient’s response to therapy including, but not limited to, whether the patient is experiencing undue levels of systemic stress and/or experiencing premature aging.
  • the measurement and analysis of cortisol levels may predict possible vascular injury. It has been proposed that overexposure to cortisol in vascular tissue results in cells becoming desensitized to cortisol and thereby resulting in vascular inflammation.
  • Norepinephrine is a neurotransmitter that is synthesized from dopamine. Norepinephrine is a signaling molecule within the sympathetic nervous system as part of the body’s response to danger and significant stress (e.g., systemic stress). As relevant to (32736-2072) embodiments herein, stress triggers the release of norepinephrine from adrenal glands.
  • norepinephrine generates a number of physiological responses including pupil dilation; changes in skin tone as blood and oxygen is diverted; increase heart rate and blood pressure; conversion of glycogen in the liver to glucose; and increase depth and rate of respiration. Additionally, elevated norepinephrine levels may contribute to disruptions in sleep cycle/patterns.
  • one or more biochemical sensors are employed to sense creatine levels in patient undergoing GLP-1 receptor agonist therapy (e.g., as detected in ISF).
  • Creatine is a naturally-occurring substance that is absorbable from nutritional sources and is made by the human body in the liver, kidneys, and pancreas.
  • Creatine, creatine phosphate, or phosphocreatine are stored in the muscles where used for energy. In healthy individuals, the kidneys maintain creatine levels within normal ranges. Measurement of creatine levels provide a fairly reliable indicator of kidney function. Elevated Creatine level generally signifies impaired kidney function absent external causes. [0256] Elevated creatinine levels may also indicate muscle damage. GLP-1 receptor agonist therapy has been correlated with a loss of lean muscle mass in patients (although it has not been definitively shown to correlate to more lean muscle loss than non-pharmaceutically based weight-loss regimens). Applicant posits that the mechanism(s) of action causing the loss of lean muscle mass may also affect cardiac muscle tissue.
  • creatine levels are measured and used as input to and/or used as training data for one or more AI/ML or other computational models for patients undergoing GLP-1 receptor agonist therapy. Although most patients undergoing such therapy will not develop acute, serious effects to their kidneys, long-term systemic stress may impact kidney function. Accordingly, elevated creatine levels at subacute levels may also be an indicator (e.g., in combination with other biomarkers or other factors) that a clinician should consider modification or cessation of the patient’s therapy.
  • Lactate is a byproduct of glucose metabolism in glycolysis.
  • the glycolysis pathway is activated to compensate for a lack of ATP production when hypoxia inhibits the tricarboxylic acid (TCA) cycle.
  • TCA tricarboxylic acid
  • lactate accumulates during strenuous exercise and is rapidly metabolized by pyruvate dehydrogenase (PDH).
  • PDH pyruvate dehydrogenase
  • the balance between glycolysis and PDH flux may be a key determinant of lactate levels.
  • Circulating lactate may also be a supplementary source of glucose that satisfies excitatory brain activities when blood glucose levels are insufficient.
  • Lactate can provide valuable insight into patient response to GLP-1 receptor agonist therapy according to some representative embodiments.
  • lactate levels For example, inflammatory environments exhibit elevated lactate levels (Pucino, V., Bombardieri, M., Pitzalis, C. & Mauro, C. Lactate at the crossroads of metabolism, inflammation, and autoimmunity. Eur. J. Immunol.47, 14–21 (2017)). Additionally, evidence indicates aerobic glycolysis produces lactate under stressful conditions, such as trauma, infection, myocardial infarction, and heart failure. Under conditions of stress, lactate has been suggested to act as a biofuel that minimizes or eliminates blood glucose use and provides additional glucose. Therefore, increased lactate levels (hyperlactatemia) may indicate a protective response to stress. Hyperlactatemia can also result from increased or accelerated aerobic glycolysis during the stress response.
  • lactate levels are correlated to physical activity
  • correlation of lactate levels and physical activity data from a wearable device may be useful to discern whether increased lactate levels are due to the effects of GLP-1 receptor agonist therapy, including systemic stress and premature aging.
  • elevated lactate levels occurring at times of little or no physical activity may provide higher probative value of AI/ML or other computational models than average lactate levels or lactate levels during higher levels of physical exertion.
  • one or more biochemical sensors are employed to sense ketone levels in patients undergoing GLP-1 receptor agonist therapy (e.g., as detected in ISF).
  • ketones The body can use ketones as a source of energy in the absence of a carbohydrate source.
  • Many diets recommended for weight-loss patients involve significant restriction on carbohydrate intake.
  • the liver converts fatty acids into ketone bodies for use as an energy source. This process is especially important when an individual's blood glucose has decreased, and they must maintain an energy source for organs such as the brain.
  • Ketone metabolism consists of the oxidation and utilization of ketone bodies by mitochondria, especially in organs with high energy demand (such as the brain and the heart).
  • Prolonged fasting may lead to an excess of ketones and cause ketosis. Excess ketones in combination with reduced insulin levels can induce tissue damage in patients.
  • primary ketone body, ⁇ -hydroxybutyrate is known to modulate inflammation and, specifically, inflammation of cardiac tissue. Ketone-related analyte levels may provide a suitable biomarker (individually and in combination with one or more other biomarkers) of the possibility of systemic stress, premature aging, and/or other possible sequelae for patients undergoing GLP-1 receptor agonist therapy.
  • Histone acetylation is important for the global transcription and specific changes in gene expression. Because ⁇ OHB is catabolized to acetyl-CoA in target tissues, metabolism of ⁇ OHB into acetyl-CoA should raise intracellular acetyl-CoA levels, and increase the acetylation of histones and non-histone proteins which are involved in several cellular processes controlling anabolic and catabolic reactions during fasting (32736-2072) response. In line with this notion, ⁇ OHB may have far-reaching effect on overall metabolic health.” [0265] In some embodiments, one or more biochemical sensors are employed to sense glutamine levels in patients undergoing GLP-1 receptor agonist therapy (e.g., as detected in ISF).
  • Glutamine is a non-essential amino acid and is fundamental to intermediary metabolism, interorgan nitrogen exchange via ammonia (NH3) transport, and pH homeostasis. Glutamine is used as a substrate for nucleotide synthesis (purines, pyrimidines, and amino sugars), nicotinamide adenine dinucleotide phosphate (NADPH), antioxidants, and many other biosynthetic pathways involved in the maintenance of cellular integrity and function. [0267] Glutamine plays an important role in the modulation of inflammation. Glutamine’s cytoprotective and antioxidant properties may be particularly important in environments of high catabolism induced by weight-loss therapies.
  • IBD inflammatory bowel disease
  • blood glucose levels are elevated when patients experience psychological and physical stress. It is not entirely understood whether blood glucose is a cause or an effect in the state of patients subjected to systemic stress.
  • the internal physiological processes involving hyperglycemia possibly function in an interrelated feedback loop with multiple processes reinforcing negative outcomes within the loop. (32736-2072)
  • elevated blood glucose can be a physiologic response to hormones (e.g., epinephrine or cortisol) that are released under high systemic stress and, hence, may indicate greater overall illness severity. Additionally, hyperglycemia may be indicative of systemic and organ-specific metabolic dysfunction. This is most likely in patients with impaired insulin signaling.
  • glucose levels in patients undergoing GLP-1 receptor agonist therapies are monitored using a suitable sensor system and the patient glucose data is provided to one or more AI/ML or other computational models for patient management as discussed herein.
  • the glucose levels may be a relevant biomarker to distinguish patients who may experience positive response to therapy from patients who may experience long-term negative outcomes (e.g., systemic stress indicative of premature aging), because increased insulin resistance is associated with elevated oxidative stress in patients (see Boris Hansel, Philippe Giral, Estelle Nobecourt, Sandrine Chantepie, Eric Bruckert, M.
  • elevated glucose levels in combination with other biomarkers may improve patient-classification of AI/ML or computational model(s). For example, elevated glucose levels in the absence of external activities (e.g., as detected using the patient activity tracking methods described herein) may indicate systemic stress occurring in response to GLP-1 receptor agonist therapy as opposed to dietary intake.
  • Elevated levels of inflammation may be employed to detect systemic stress in patients undergoing GLP-1 receptor agonist therapy using one or more AI/ML or other (32736-2072) computational models as discussed herein. Although it is reported that GLP-1 receptor agonist therapy generally reduces inflammatory processes in patients, certain subsets of patients (e.g., those likely to experience negative long-term outcomes) may not exhibit reduced levels of inflammation or may exhibit increased levels of inflammation. Elevated levels of inflammation may be indicative of systemic effects related to premature aging.
  • systemic inflammation and pro-thrombotic environments promote disfunction and cellular injury cardiac tissue, endothelial tissue, renal tissue, pancreatic tissue, hepatic tissue, among others. Further, systemic inflammation also contributes to metabolic dysfunction which, in turn, generates other negative outcomes including systemic stress.
  • one or more biochemical sensors or other medical equipment are employed to sense cytokines and pro-inflammatory enzymes levels in patients undergoing GLP-1 receptor agonist therapy.
  • the detection of one or more pro-inflammatory- related analytes may occur using a wearable biosensor, an implantable biosensor, and/or through lab testing of blood, and/or the like.
  • An inflammatory cytokine or proinflammatory cytokine is a type of signaling molecule (a cytokine) that is secreted from immune cells like helper T cells (Th) and macrophages, and certain other cell types that promote inflammation. They include interleukin-1 (IL-1), IL-6, IL-12, and IL-18, tumor necrosis factor alpha (TNF- ⁇ ), interferon gamma (IFN ⁇ ), and granulocyte-macrophage colony stimulating factor (GM-CSF) and play an important role in mediating the innate immune response. Inflammatory cytokines are predominantly produced by and involved in the upregulation of inflammatory reactions.
  • IL-6 interleukin-6
  • PKC protein kinase C
  • endothelial dysfunction or damage may be a contributing factor of premature aging in patients undergoing GLP-1 receptor agonist therapies and may contribute to damage to one or more organs including the heart and kidneys.
  • certain biomarkers for use by one or more AI/ML or other computational models for managing patients undergoing GLP-1 receptor therapies may be obtained using one or more sensors of smartwatches and/or fitness tracking devices.
  • prolonged systemic stress in a patient may induce fatigue in the patient through one or more mechanisms.
  • Biomarkers related to fatigue may be obtained directly by processing movement data using one or more accelerometer components of smartwatches and/or fitness tracking devices.
  • steps taken, average walking speed, and/or changes to gait may be analyzed from movement data to generate information indicative of a fatigue in patients.
  • Cardiac activity data, respiration data, sleep pattern data, blood oxygenation levels, blood pressure levels, EMG data, and/or any other suitable patient data may be obtained using one or more sensors of smartwatches and/or fitness tracking devices.
  • Certain biomarkers have been discussed herein in detail. Other known biomarkers related to systemic stress, premature aging, cardiac function, endothelial function, kidney function, metabolic function, systemic oxidation may be employed for one or more AI/ML or other computational models.
  • FGF19 and/or FGF21 may be measured or detected for use with one or more AI/ML or other computational models according to some embodiments.
  • FGF21 is a regulator of several metabolic pathways, including ketogenesis, gluconeogenesis, and lipolysis, and the protein is to a high degree expressed by the liver and also by skeletal muscles that releases this hormone into circulation in response to insulin stimulation.
  • biomarkers generally identifies whether a given biomarker is significantly out of range of expected values.
  • significant systemic changes can occur without an individual biomarker becoming significantly out of range.
  • a system employing one or more AI/ML or other computational models as described herein enable the development of a large data set of biomarkers that analyzed together may provide a more accurate and reliable indication of adverse events associated with GLP-1 receptor agonist therapy. This combination of data sets and the ability to perform comparisons to other large data sets while considering unexpected medical outcomes allows an analysis that is not known by Applicant to be possible using conventional medical practices.
  • one or more AI/ML or other computational models are trained using a classification based on identified instances of so-called “Ozempic face” (“criterion 1”) in patients. Further data may be employed to further refine the classification according to some embodiments herein. For example, lower weights/BMI at therapy commencement, lower levels of weight loss, percentage of lean muscle loss versus percentage of body fat loss, and/or similar measures may be used as an additional factor for classification (“criteria 2”).
  • cardiac activity e.g., elevated resting heart rate, negative changes in HRV, intermittent cardiac rhythm irregularities
  • criteria 3 may be used as one or more criteria (referred to as “criteria 3”).
  • these criteria may be calculated as differences relative to pre-therapy states and/or differences relative to sets of patients (e.g., patients who experienced non- pharmaceutically assisted weight-loss).
  • the various captured biomarkers e.g., one or more biochemical biomarkers, one or more physiological biomarkers, one or more activity biomarkers, etc.
  • the various captured biomarkers can then be used to train one or more AI/ML or other computational models to identify whether a prospective patient belongs to this group or not (or a probability or likelihood thereof or a metric related thereto).
  • one or more AI/ML or other computational models are trained using a classification based on patients who successfully lost weight following a non-pharmaceutically assisted regimen with beneficial outcomes.
  • one or more AI/ML or other computational models are trained using a classification based on patients who successfully lost meaningful amounts of weight and maintained health outcomes for multiple years.
  • the various captured biomarkers e.g., one or more biochemical biomarkers, one or more physiological biomarkers, one or more activity biomarkers, etc.
  • the various captured biomarkers can then be used to train one or more AI/ML or other computational models to identify whether a prospective patient undergoing GLP-1 receptor agonist therapy exhibits an effect that correlates with this group or not (or a probability or likelihood thereof or a metric related thereto).
  • the patient response to GLP-1 receptor agonist therapy may be expected to deviate, to some extent, from patients who experienced weight-loss from non-pharmaceutical methods (“non-pharma patients”) or other methods.
  • the AI/ML may classify GLP-1 receptor agonist therapy patients based on different degrees of deviation from such patients. Classification of the GLP-1 patent using one or more such AI/ML or other computational models, may determine the probability of long-term sequelae from systemic stress and/or premature aging depending upon how “close” or “far” the GLP-1 patient is to the model(s) of respective group of patients used for training of the one or more models.
  • EXAMPLE 3 [0287]
  • blood testing is employed as a methodology to classify patients for one or more AI/ML or other computational models to manage patients undergoing GLP-1 receptor agonist therapies.
  • a number of blood tests exists that attempt to quantify “biological age.” For example, such biological age calculation methods are described in Biological age estimate using circulating blood biomarkers, Communication Biology 6, Article Number: 1089 (2023) by Jordan Bortz et al. Such methods use existing clinical assay panels to obtain relevant biomarkers to estimate the age of an individual under test as the equivalent age within the same-sex population of the general population, which corresponds to an individual’s mortality risk. In experience with some methods, values ranged between 20-years younger and 20-years older than individuals’ chronological age, exposing the magnitude of ageing signals contained in blood markers.
  • a commercial service that calculates a biological age component is the “personal health analytics dashboard” service of InsideTracker (Segterra, Inc., Cambridge, MA).
  • An initial patient classification stage for AI/ML training may occur by subjecting patients undergoing GLP-1 receptor agonist therapies for weight-loss to biological age analysis before and after therapy. Patients may be classified in one or more classes, such as positive improvement in biological age, no improvement in biological age, and one or more levels of negative change in biological age.
  • the non-blood test biomarkers of such patients e.g., obtained via wearable or implantable biosensors, smartwatches/fitness tracking devices, patient reported outcomes
  • these groups of patients may then be used to train one or more AI/ML or other computational models. These one or more trained AI/ML or other computational models may then be used to manage patients undergoing GLP-1 receptor agonist therapies.
  • one or more trained AI/ML or other computational models as described herein may predict later occurrence of sequelae, thereby allowing intervention to prevent long-term negative outcomes for patients undergoing GLP-1 therapies.
  • EXAMPLE 4 [0289]
  • one or more AI/ML or other computational models are trained using a classification based on patients who experienced sustained negative cardiac outcomes after GLP-1 receptor agonist therapy as measured using cardiac rhythm metrics relative to pre-therapy levels. For example, negative changes in resting heart rate, HRV, cardiac rhythm changes (changes in QT interval) may be evaluated.
  • Patients may be classified into respective categories based on one or more defined levels of negative changes in cardiac health.
  • the various captured biomarkers e.g., one or more biochemical biomarkers, one or more physiological biomarkers, one or more activity biomarkers, etc.
  • the various captured biomarkers can then be used to train one or more AI/ML or other computational models to identify whether a prospective patient belongs to a respective group or not (or a probability or likelihood thereof or a metric related thereto).
  • one or more AI/ML or other computational models are trained using a classification based on patients who experienced failure of GLP-1 receptor agonist therapy to achieve improvements in metabolic function and optionally in combination with one or more other negative outcomes.
  • GLP-1 receptor agonists As reported in Long-term effects of GLP-1 receptor agonists in type 2 diabetic patients: A retrospective real-life study in 131 patients (Diabetes Metab Syndr.2019 Jan-Feb;13(1):332-336, A Hemmer et al.), GLP-1 receptor agonist therapies for diabetes exhibit therapy failure in a number of patients.
  • the various captured biomarkers e.g., one or more biochemical biomarkers, one or more physiological biomarkers, one or more activity biomarkers, etc.
  • the various captured biomarkers can then be used to train one or more AI/ML or other computational models to identify whether a prospective patient belongs to a respective group or not (or a probability or likelihood thereof or a metric related thereto).
  • one or more AI/ML or other computational models are trained using a classification based on patients who exhibit negative changes in endothelial function.
  • endothelial function is invasively evaluated by analyzing the dose–response curves of coronary arteries where endothelial dysfunction is evidenced by a decreased vasodilatory response or even vasoconstriction of the coronary artery to acetylcholine. Endothelial health may be measured before and after therapy at suitable times and patients exhibiting reductions in endothelial function may be employed for classification purposes.
  • the various captured biomarkers e.g., one or more biochemical biomarkers, one or more physiological biomarkers, one or more activity biomarkers, etc.
  • the various captured biomarkers can then be used to train one or more AI/ML or other computational models to identify whether a prospective patient belongs to a respective group or not (or a probability or likelihood thereof or a metric related thereto).
  • FIG.9 depicts biomarker data set 901 according to some embodiments. Applicant has described respective biomarkers in this application and their probative value for the various patient states useful for determining patient response or predicted patient response to GLP-1 therapy.
  • a respective biomarker data set 901 may be constructed by selecting one or more biomarkers (e.g., biomarkers 1-N) described herein to model one or more of suitable patient states.
  • a respective set of biomarkers can be selected for each of the patient states shown in patient output screen 801 of FIG.8 (e.g., Systemic Stress, Immune-Response, Endothelial Metric, Kidney Metric, Glucose Levels Metric, Cardiac Changes, Sleep Changes Metric, and Patient Activity Metric) and further biomarker sets may be defined according to the disclosure of this application.
  • the value for the blood glucose biomarker for a given instance of biomarker data set 901 for a patient at a respective time may be calculated by processing raw sensor data over a week or other suitable period of time to generate an average blood glucose level.
  • the values for the ⁇ -hydroxybutyrate and cortisol levels may be calculated by processing raw sensor data over a week or other suitable period of time to generate respective average biomarker levels.
  • the value for the lactate biomarker may be calculated by processing raw sensor data over a week or other suitable period of time to generate an average lactate level only at times when the patient is exhibiting low activity levels as discussed herein.
  • a value related to elevated resting heart rate may calculated by processing raw sensor data over a week or other suitable period of time to generate a value that represents an average excess of the patient’s heart rate (when the patient is at rest) over a benchmark value (e.g., population selected value or a pre-therapy value previously measured for the patient). Similar processing may be performed to calculate an average deviation in HRV. [0296] Now, as previously discussed, the respective devices and system in FIGS. 1-3 measure, process, and/or store such instances of biomarker data for patients in various stages of therapy.
  • biomarker data set 901 can be generated for each patient in a non-pharmaceutically assisted weight-loss regime, each patient in a pre-GLP-1 therapy state, each patient in an ongoing, GLP-1 therapy state, and each patient in a post- GLP-1 therapy state. Biomarker data set 901 for such patients can be generated for each of these patients at suitable intervals.
  • training of one or more computation models e.g., AI/ML models
  • the training process may begin by selecting biomarker data sets from the aggregated sets of biomarker data of patients that represent respective relevant classes.
  • FIG.10 depicts convolutional neural network (CNN) 1001 according to some embodiments.
  • CNN 1001 includes input layer to receive respective values from biomarker set 901 (e.g., biomarkers 1-N).
  • CNN 1001 may be a single layer network or may include one or more hidden layers.
  • CNN 1001 includes output layer with one or more outputs. The outputs of CNN 1001 may be used to perform classification operations according to embodiments discussed herein.
  • a given output of CNN 1001 may represent whether a given data set belongs to a given classification as discussed herein. Other outputs may be additionally or alternatively defined as discussed herein. For example, a numerical output or metric with a defined range may be created that represents a level of systemic stress or any other patient state discussed herein related to negative and/or positive outcomes for GLP-1 patients.
  • Training of CNN 1001 may occur by applying known CNN coefficient (node weights) refinement/error minimization operations during a training phase of operation. As shown in FIG.11, for example, multiple groups of training data (e.g., respective groups of biomarker data sets 901) may be provided for training purposes in system 1101.
  • one group of biomarker data sets 901 are selected as representing patients who experienced “Ozempic face” and experienced high systemic stress biomarkers during GLP-1 weight loss therapy.
  • Another group of biomarker data sets 901 are selected as representing patients who experienced healthy outcomes with non-pharmaceutically assisted weight loss regimens.
  • these two groups biomarker data sets 901 are from classes of patients representing different extremes of patient outcomes.
  • Other groups of biomarkers data sets 901 (not shown) (32736-2072) could be likewise provided for intermediate patient classes.
  • Training of CNN 1001 occurs by iteratively applying each respective data set 901 to CNN 1001, generating the output(s) using CNN 1001, calculating error, and refining the coefficients (node weights) of the CNN.
  • new patients e.g., patients undergoing GLP-1 therapy for weight loss
  • Biomarker data set for a new patient is provided to trained CNN 1001 and the output(s) of trained CNN 1001 may be used to classify the new patient based on the respective biomarker data obtained for that patient.
  • the output(s) may be provided to the patient, the patient’s physician, and/or any other suitable medical professional.
  • the GLP-1 therapy for the patient can be terminated, modified, or continued using, in part, the patient classification process.
  • different variations to computational models may be applied for management of GLP-1 patients. For example, instead of using one biomarker data set for a given time or time period to represent a patient state, multiple biomarker data sets can be employed each representing different times or time periods such as respective biomarker data sets representing the patient state after 3 months of therapy, 6 months of therapy, 9 months of therapy, and 12 months of therapy. To accommodate different times and/or time periods, modifications of the computational module may be applied. For example, multiple CNNs may be defined. A respective CNN could be defined separately for each of the 3, 6, 9, and 12 month biomarker data sets.
  • RNNs Recurrent Neural Networks
  • a CNN could be defined to process the “delta” or change between the 3 month and 6 month biomarker data sets.
  • a CNN could be defined to receive both the 3 month and 6 month biomarker data sets as inputs as another example of how to process different instances of biomarker data sets from different time periods.
  • multiple CNNs could be defined for various combinations of such biomarker data sets from different time periods.
  • FIG.13 depicts a flowchart for training a computational model according to some representative embodiments.
  • biomarker data is obtained from large set of patients (e.g., using the devices and systems shown in FIGS.1-3).
  • classes are defined to represent patient states/outcomes.
  • patients are identified (32736-2072) corresponding to patient states/outcomes of defined classes.
  • sets of biomarker data for respective times or time periods for identified patients are assembled.
  • one or more outputs for the computational model are defined.
  • expected computational model output(s) for assembled biomarker data sets are assigned.
  • the computational model is trained using the assembled biomarker data sets and defined expected output(s). Upon training, the trained computation model may be employed as shown in FIG.12 to classify new patient biomarker data sets for GLP-1 therapy patients as discussed herein.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Public Health (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Epidemiology (AREA)
  • Data Mining & Analysis (AREA)
  • Primary Health Care (AREA)
  • Databases & Information Systems (AREA)
  • Pathology (AREA)
  • Chemical & Material Sciences (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Medicinal Chemistry (AREA)
  • Investigating Or Analysing Biological Materials (AREA)

Abstract

Des modes de réalisation de la présente invention comprennent des méthodes et des systèmes de gestion de résultats pour des thérapies agonistes du récepteur GLP-1 de perte de poids à l'aide de modèles entraînés par des données biochimiques et/ou d'une autre analyse centrée sur le patient. Dans certains modes de réalisation, un ou plusieurs biomarqueurs ou d'autres données de patient sont utilisés pour entraîner l'IA/ML ou d'autres modèles de calcul et/ou en tant qu'entrées de l'IA/ML ou d'autres modèles de calcul pour l'évaluation de la réponse du patient à une thérapie par agoniste du récepteur GLP-1. Par exemple, les modèles de calcul peuvent être configurés pour effectuer des classifications par rapport au succès de la thérapie par agoniste du récepteur GLP-1 et/ou à des effets indésirables pendant une thérapie par agoniste du récepteur GLP-1.
PCT/US2025/024402 2024-04-12 2025-04-11 Méthodes et systèmes de gestion de résultats pour des thérapies agonistes du récepteur glp-1 de perte de poids à l'aide de modèles entraînés par des données biochimiques et/ou d'une autre analyse centrée sur le patient Pending WO2025217600A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US202463633360P 2024-04-12 2024-04-12
US63/633,360 2024-04-12

Publications (1)

Publication Number Publication Date
WO2025217600A1 true WO2025217600A1 (fr) 2025-10-16

Family

ID=95651198

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2025/024402 Pending WO2025217600A1 (fr) 2024-04-12 2025-04-11 Méthodes et systèmes de gestion de résultats pour des thérapies agonistes du récepteur glp-1 de perte de poids à l'aide de modèles entraînés par des données biochimiques et/ou d'une autre analyse centrée sur le patient

Country Status (1)

Country Link
WO (1) WO2025217600A1 (fr)

Citations (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110213225A1 (en) 2009-08-31 2011-09-01 Abbott Diabetes Care Inc. Medical devices and methods
US8346335B2 (en) 2008-03-28 2013-01-01 Abbott Diabetes Care Inc. Analyte sensor calibration management
US20170127989A1 (en) 2008-03-31 2017-05-11 Abbott Diabetes Care Inc. Shallow implantable analyte sensor with rapid physiological response
US10108781B2 (en) 2012-09-28 2018-10-23 Abbott Diabetes Care Inc. Glucose or ketone body analyte monitoring in vivo
US20200205706A1 (en) 2010-07-22 2020-07-02 Abbott Diabetes Care Inc. Systems and Methods for Improved In Vivo Analyte Sensor Function
US20200402656A1 (en) 2019-06-22 2020-12-24 Advanced Neuromodulation Systems, Inc. Ui design for patient and clinician controller devices operative in a remote care architecture
US20200398062A1 (en) 2019-06-21 2020-12-24 Advanced Neuromodulation Systems, Inc. System, method and architecture for facilitating remote patient care
US11006871B2 (en) 2009-02-03 2021-05-18 Abbott Diabetes Care Inc. Analyte sensor and apparatus for insertion of the sensor
US20210158923A1 (en) * 2017-05-04 2021-05-27 Changing Health Limited Distributed user monitoring system
US20210251530A1 (en) 2010-12-09 2021-08-19 Abbott Diabetes Care Inc. Analyte sensors with a sensing surface having small sensing spots
US11264133B2 (en) 2007-06-21 2022-03-01 Abbott Diabetes Care Inc. Health management devices and methods
US20220105350A1 (en) 2020-10-07 2022-04-07 Advanced Neuromodulation Systems, Inc. Systems and methods for managing remote therapy sessions
US20220142519A1 (en) 2009-07-31 2022-05-12 Abbott Diabetes Care Inc. Method and devices for analyte monitoring calibration
US20220189626A1 (en) 2020-12-11 2022-06-16 Advanced Neuromodulation Systems, Inc. Systems and methods for detecting and addressing quality issues in remote therapy sessions
US11577081B2 (en) 2020-12-31 2023-02-14 Advanced Neuromodulation Systems, Inc. System and method for controlling neurostimulation according to user activity detected through patient use of icon driven user interface
US20230100246A1 (en) 2019-06-21 2023-03-30 Advanced Neuromodulation Systems, Inc. System and method for modulating therapy in a remote care architecture
US20230271019A1 (en) 2022-02-26 2023-08-31 Advanced Neuromodulation Systems, Inc. Digital health architecture including a virtual clinic for facilitating remote programming
US20230414136A1 (en) 2007-04-14 2023-12-28 Abbott Diabetes Care Inc. Method and apparatus for providing data processing and control in medical communication system
US20240033427A1 (en) 2022-07-27 2024-02-01 Abbott Diabetes Care Inc. Systems and methods for diabetes management

Patent Citations (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20230414136A1 (en) 2007-04-14 2023-12-28 Abbott Diabetes Care Inc. Method and apparatus for providing data processing and control in medical communication system
US11264133B2 (en) 2007-06-21 2022-03-01 Abbott Diabetes Care Inc. Health management devices and methods
US8346335B2 (en) 2008-03-28 2013-01-01 Abbott Diabetes Care Inc. Analyte sensor calibration management
US20170127989A1 (en) 2008-03-31 2017-05-11 Abbott Diabetes Care Inc. Shallow implantable analyte sensor with rapid physiological response
US11006871B2 (en) 2009-02-03 2021-05-18 Abbott Diabetes Care Inc. Analyte sensor and apparatus for insertion of the sensor
US20220142519A1 (en) 2009-07-31 2022-05-12 Abbott Diabetes Care Inc. Method and devices for analyte monitoring calibration
US20110213225A1 (en) 2009-08-31 2011-09-01 Abbott Diabetes Care Inc. Medical devices and methods
US20200205706A1 (en) 2010-07-22 2020-07-02 Abbott Diabetes Care Inc. Systems and Methods for Improved In Vivo Analyte Sensor Function
US20210251530A1 (en) 2010-12-09 2021-08-19 Abbott Diabetes Care Inc. Analyte sensors with a sensing surface having small sensing spots
US10108781B2 (en) 2012-09-28 2018-10-23 Abbott Diabetes Care Inc. Glucose or ketone body analyte monitoring in vivo
US20210158923A1 (en) * 2017-05-04 2021-05-27 Changing Health Limited Distributed user monitoring system
US20230100246A1 (en) 2019-06-21 2023-03-30 Advanced Neuromodulation Systems, Inc. System and method for modulating therapy in a remote care architecture
US20200398062A1 (en) 2019-06-21 2020-12-24 Advanced Neuromodulation Systems, Inc. System, method and architecture for facilitating remote patient care
US20230317303A1 (en) 2019-06-22 2023-10-05 Advanced Neuromodulation Systems, Inc. Data labeling system and method operative with patient and clinician controller devices disposed in a remote care architecture
US20200402656A1 (en) 2019-06-22 2020-12-24 Advanced Neuromodulation Systems, Inc. Ui design for patient and clinician controller devices operative in a remote care architecture
US20220105350A1 (en) 2020-10-07 2022-04-07 Advanced Neuromodulation Systems, Inc. Systems and methods for managing remote therapy sessions
US20220189626A1 (en) 2020-12-11 2022-06-16 Advanced Neuromodulation Systems, Inc. Systems and methods for detecting and addressing quality issues in remote therapy sessions
US11577081B2 (en) 2020-12-31 2023-02-14 Advanced Neuromodulation Systems, Inc. System and method for controlling neurostimulation according to user activity detected through patient use of icon driven user interface
US20230271019A1 (en) 2022-02-26 2023-08-31 Advanced Neuromodulation Systems, Inc. Digital health architecture including a virtual clinic for facilitating remote programming
US20240033427A1 (en) 2022-07-27 2024-02-01 Abbott Diabetes Care Inc. Systems and methods for diabetes management

Non-Patent Citations (20)

* Cited by examiner, † Cited by third party
Title
A HEMMER ET AL., DIABETES METAB SYNDR., vol. 13, no. 1, January 2019 (2019-01-01), pages 332 - 336
ABDUIREDA, M. H. ET AL.: "Liraglutide compromises pancreatic β cell function in a humanized mouse model", CELL METAB., 2016
ANUM S. MINHAS ET AL.: "Imaging Assessment of Endothelial Function: An Index of Cardiovascular Health", FRONT, CARDIOVASC. MED., vol. 9, 15 April 2022 (2022-04-15)
BORIS HANSELPHILIPPE GIRALESTELLE NOBECOURTSANDRINE CHANTEPIEERIC BRUCKERTM. JOHN CHAPMANANATOL KONTUSH: "Metabolic Syndrome is Associated with Elevated Oxidative Stress and Dysfunctional Dense High-Density Lipoprotein Particles Displaying Impaired Antioxidative Activity", THE JOURNAL OF CLINICAL ENDOCRINOLOGY & METABOLISM, vol. 89, 1 October 2004 (2004-10-01), pages 4963 - 4971
CHRISTINA H. WEISHELDON E. LITWIN: "Hypergiycemia and Adverse Outcomes in Acute Coronary Syndromes: Is Serum Glucose the Provocateur or lnnocent Bystander?", DIABETES, vol. 63, no. 7, 2014, pages 2209 - 2212
CHRISTOPHER R. MARTENSDANIELLE L. KIRKMANDAVID G. EDWARDS: "The Vascular Endothelium in Chronic Kidney Disease: A Novel Target For Aerobic Exercise", EXERC SPORT SCI REV., vol. 44, no. 1, January 2016 (2016-01-01), pages 12 - 19
CLINTON D. HUMPHREY, MDANNA C. LAWRENCE, MD: "Implications of Ozempic and Other Semaglutide Medications for Facial Plastic Surgeons", FACIAL PLAST. SURG., vol. 39, 2023, pages 719 - 721
DAVID J. LEEHEY ET AL.: "Acute Kidney Injury Associated With Semaglutide", KIDNEY MED., vol. 3, no. 2, March 2021 (2021-03-01), pages 282 - 285
FRONT. NEUROSCI., 20 March 2015 (2015-03-20)
JORDAN BORTZ ET AL.: "Biological age estimate using circulating blood biomarkers", COMMUNICATION BIOLOGY, vol. 5, 2023
MOHAMMAD SHEHAB ET AL., COMPUTERS IN BIOLOGY AND MEDICINE, vol. 145, June 2022 (2022-06-01), pages 105458
N TRANT GARCIAM ANIQAS ALIA ALLYSM NAULI: "Endothelial Nitric Oxide Synthase (eNOS) and the Cardiovascular System", PHYSIOLOGY AND IN DISEASE STATES AM J BIOMED SCI RES., vol. 15, no. 2, 2022, pages 153 - 177
NEUROENDOCRINE SCIENCE, vol. 9, 2015
NISSEN SEWOLSKI K.: "Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes", N ENGL J MED, vol. 356, 2007, pages 2457 - 2471, XP002676098, DOI: 10.1056/NEJMOA072761
PUCINO, V.BOMBARDIERI, M.PITZALIS, C.MAURO, C.: "Lactate at the crossroads of metabolism, inflammation, and autoimmunity", EUR. J. IMMUNOL., vol. 47, 2017, pages 14 - 21, XP071228070, DOI: 10.1002/eji.201646477
RYO KAWAKAMI: "Ketone body and FGF21 coordinately regulate fasting-induced oxidative stress response in the heart", SCIENTIFIC REPORTS, vol. 12, 2022
SARAH L ANDERSON ET AL.: "Association of Pancreatitis with Glucagon-Like Peptide-1 Agonist Use", ANNALS OF PHARMACOTHERAPY, vol. 44, 2010
SODHI MREZAEIANZADEH RKEZOUH AETMINAN M: "Risk of gastrointestinal adverse events associated with glucagon-like peptide-1 receptor agonists for weight loss", JAMA, vol. 330, no. 18, 2023, pages 1795 - 1797
SONAL SINGH ET AL.: "Glucagonlike Peptide 1-Based Therapies and Risk of Hospitalization for Acute Pancreatitis in Type 2 Diabetes Mellitus: A Population-Based Matched Case-Control Study", JAMA INTERN MED., vol. 173, no. 7, 2013, pages 534 - 539
TAÍS S. ASSMANN ET AL.: "Nitric oxide levels in patients with diabetes mellitus: A systematic review and meta-analysis", NITRIC OXIDE, vol. 61, 30 December 2016 (2016-12-30), pages 1 - 9

Similar Documents

Publication Publication Date Title
Reddy et al. Prediction of hypoglycemia during aerobic exercise in adults with type 1 diabetes
US20250054621A1 (en) Health management
Hieda et al. One-year committed exercise training reverses abnormal left ventricular myocardial stiffness in patients with stage B heart failure with preserved ejection fraction
US20210020294A1 (en) Methods, devices and systems for holistic integrated healthcare patient management
Riddell et al. Individual glucose responses to prolonged moderate intensity aerobic exercise in adolescents with type 1 diabetes: the higher they start, the harder they fall
Krinsley Glycemic control, diabetic status, and mortality in a heterogeneous population of critically ill patients before and during the era of intensive glycemic management: six and one-half years experience at a university-affiliated community hospital
Menezes et al. Cardiac rehabilitation and exercise therapy in the elderly: Should we invest in the aged?
Pandey et al. Body mass index and cardiorespiratory fitness in mid-life and risk of heart failure hospitalization in older age: findings from the Cooper Center Longitudinal Study
Jacobs et al. Incorporating an exercise detection, grading, and hormone dosing algorithm into the artificial pancreas using accelerometry and heart rate
Bilchick et al. Prediction of mortality in clinical practice for medicare patients undergoing defibrillator implantation for primary prevention of sudden cardiac death
US7914449B2 (en) Diagnostic support system for diabetes and storage medium
Pinna et al. Different estimation methods of spontaneous baroreflex sensitivity have different predictive value in heart failure patients
Georga et al. Data mining for blood glucose prediction and knowledge discovery in diabetic patients: The METABO diabetes modeling and management system
Zhang et al. Data-based modeling for hypoglycemia prediction: Importance, trends, and implications for clinical practice
CN113679348B (zh) 血糖预测方法、血糖预测装置、设备及存储介质
EP4199820A1 (fr) Surveillance virtuelle des taux de glucose chez un patient faisant appel à une technologie d'apprentissage machine et à une technologie de jumeau numérique
Hawkins et al. The impact of change in physical activity on change in arterial stiffness in overweight or obese sedentary young adults
Kraal et al. Energy expenditure estimation in beta-blocker-medicated cardiac patients by combining heart rate and body movement data
US20230005587A1 (en) Determining whether adjustments of insulin therapy recommendations are being taken into account
Morrison et al. Late afternoon vigorous exercise increases postmeal but not overnight hypoglycemia in adults with type 1 diabetes managed with automated insulin delivery
Jensen et al. Evaluation of an algorithm for retrospective hypoglycemia detection using professional continuous glucose monitoring data
Na et al. Effects of chronotropic incompetence on exercise capacity in people with heart failure versus age-matched controls
Vonderau et al. Use of continuous glucose monitoring for evaluation of hypoglycemia pre‐and post‐operatively for patients undergoing total pancreatectomy with autologous islet cell transplant
US20230005586A1 (en) Determining total daily basal dose mismatch
EP3864662B1 (fr) Outil de traitement personnalisé

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 25723561

Country of ref document: EP

Kind code of ref document: A1