The Burnout PRedictiOn Using Wearable aNd ArtIficial IntelligEnce (BROWNIE) study: a decentralized digital health protocol to predict burnout in registered nurses

BACKGROUND

When job demand exceeds job resources, burnout occurs. Burnout in healthcare workers extends beyond negatively affecting their functioning and physical and mental health; it also has been associated with poor medical outcomes for patients. Data-driven technology holds promise for the prediction of occupational burnout before it occurs. Early warning signs of burnout would facilitate preemptive institutional responses for preventing individual, organizational, and public health consequences of occupational burnout. This protocol describes the design and methodology for the decentralized Burnout PRedictiOn Using Wearable aNd ArtIficial IntelligEnce (BROWNIE) Study. This study aims to develop predictive models of occupational burnout and estimate burnout-associated costs using consumer-grade wearable smartwatches and systems-level data.

METHODS

A total of 360 registered nurses (RNs) will be recruited in 3 cohorts. These cohorts will serve as training, testing, and validation datasets for developing predictive models. Subjects will consent to one year of participation, including the daily use of a commodity smartwatch that collects heart rate, step count, and sleep data. Subjects will also complete online baseline and quarterly surveys assessing psychological, workplace, and sociodemographic factors. Routine administrative systems-level data on nursing care outcomes will be abstracted weekly.

DISCUSSION

The BROWNIE study was designed to be decentralized and asynchronous to minimize any additional burden on RNs and to ensure that night shift RNs would have equal accessibility to study resources and procedures. The protocol employs novel engagement strategies with participants to maintain compliance and reduce attrition to address the historical challenges of research using wearable devices.

TRIAL REGISTRATION

NCT05481138.

Machine Learning Predicts Conversion from Normal Aging to Mild Cognitive Impairment Using Medical History, APOE Genotype, and Neuropsychological Assessment

Background

Identifying individuals at risk for mild cognitive impairment (MCI) is of urgent clinical need.

Objective

This study aimed to determine whether machine learning approaches could harness longitudinal neuropsychology measures, medical data, and APOEɛ4 genotype to identify individuals at risk of MCI 1 to 2 years prior to diagnosis.

Methods

Data from 676 individuals who participated in the 'APOE in the Predisposition to, Protection from and Prevention of Alzheimer's Disease' longitudinal study (N = 66 who converted to MCI) were utilized in supervised machine learning algorithms to predict conversion to MCI.

Results

A random forest algorithm predicted conversion 1-2 years prior to diagnosis with 97% accuracy (p = 0.0026). The global minima (each individual's lowest score) of memory measures from the 'Rey Auditory Verbal Learning Test' and the 'Selective Reminding Test' were the strongest predictors.

Conclusions

This study demonstrates the feasibility of using machine learning to identify individuals likely to convert from normal cognition to MCI.

Impact of Various Estimated Glomerular Filtration Rate Equations on the Pharmacokinetics of Meropenem in Critically Ill Adults

IMPORTANCE

Meropenem dosing is typically guided by creatinine-based estimated glomerular filtration rate (eGFR), but creatinine is a suboptimal GFR marker in the critically ill.

OBJECTIVES

This study aimed to develop and qualify a population pharmacokinetic model for meropenem in critically ill adults and to determine which eGFR equation based on creatinine, cystatin C, or both biomarkers best improves model performance.

DESIGN SETTING AND PARTICIPANTS

This single-center study evaluated adults hospitalized in an ICU who received IV meropenem from 2018 to 2022. Patients were excluded if they had acute kidney injury, were on kidney replacement therapy, or were treated with extracorporeal membrane oxygenation. Two cohorts were used for population pharmacokinetic modeling: a richly sampled development cohort (n = 19) and an opportunistically sampled qualification cohort (n = 32).

MAIN OUTCOMES AND MEASURES

A nonlinear mixed-effects model was developed using parametric methods to estimate meropenem serum concentrations.

RESULTS

The best-fit structural model in the richly sampled development cohort was a two-compartment model with first-order elimination. The final model included time-dependent weight normalized to a 70-kg adult as a covariate for volume of distribution (Vd) and time-dependent eGFR for clearance. Among the eGFR equations evaluated, eGFR based on creatinine and cystatin C expressed in mL/min best-predicted meropenem clearance. The mean (se) Vd in the final model was 18.2 (3.5) liters and clearance was 11.5 (1.3) L/hr. Using the development cohort as the Bayesian prior, the opportunistically sampled cohort demonstrated good accuracy and low bias.

CONCLUSIONS AND RELEVANCE

Contemporary eGFR equations that use both creatinine and cystatin C improved meropenem population pharmacokinetic model performance compared with creatinine-only or cystatin C-only eGFR equations in adult critically ill patients.

Population pharmacokinetic model of cefepime for critically ill adults: a comparative assessment of eGFR equations

Cefepime exhibits highly variable pharmacokinetics in critically ill patients. The purpose of this study was to develop and qualify a population pharmacokinetic model for use in the critically ill and investigate the impact of various estimated glomerular filtration rate (eGFR) equations using creatinine, cystatin C, or both on model parameters. This was a prospective study of critically ill adults hospitalized at an academic medical center treated with intravenous cefepime. Individuals with acute kidney injury or on kidney replacement therapy or extracorporeal membrane oxygenation were excluded. A nonlinear mixed-effects population pharmacokinetic model was developed using data collected from 2018 to 2022. The 120 included individuals contributed 379 serum samples for analysis. A two-compartment pharmacokinetic model with first-order elimination best described the data. The population mean parameters (standard error) in the final model were 7.84 (0.24) L/h for CL1 and 15.6 (1.45) L for V1. Q was fixed at 7.09 L/h and V2 was fixed at 10.6 L, due to low observed interindividual variation in these parameters. The final model included weight as a covariate for volume of distribution and the eGFRcr-cysC (mL/min) as a predictor of drug clearance. In summary, a population pharmacokinetic model for cefepime was created for critically ill adults. The study demonstrated the importance of cystatin C to prediction of cefepime clearance. Cefepime dosing models which use an eGFR equation inclusive of cystatin C are likely to exhibit improved accuracy and precision compared to dosing models which incorporate an eGFR equation with only creatinine.

Machine Learning Identifies Smartwatch-Based Physiological Biomarker for Predicting Disruptive Behavior in Children: A Feasibility Study

Objective: Parents frequently purchase and inquire about smartwatch devices to monitor child behaviors and functioning. This pilot study examined the feasibility and accuracy of using smartwatch monitoring for the prediction of disruptive behaviors. Methods: The study enrolled children (N = 10) aged 7-10 years hospitalized for the treatment of disruptive behaviors. The study team completed continuous behavioral phenotyping during study participation. The machine learning protocol examined severe behavioral outbursts (operationalized as episodes that preceded physical restraint) for preparing the training data. Supervised machine learning methods were trained with cross-validation to predict three behavior states-calm, playful, and disruptive. Results: The participants had a 90% adherence rate for per protocol smartwatch use. Decision trees derived conditional dependencies of heart rate, sleep, and motor activity to predict behavior. A cross-validation demonstrated 80.89% accuracy of predicting the child's behavior state using these conditional dependencies. Conclusion: This study demonstrated the feasibility of 7-day continuous smartwatch monitoring for children with severe disruptive behaviors. A machine learning approach characterized predictive biomarkers of impending disruptive behaviors. Future validation studies will examine smartwatch physiological biomarkers to enhance behavioral interventions, increase parental engagement in treatment, and demonstrate target engagement in clinical trials of pharmacological agents for young children.

Efficacy and safety of lamotrigine in pediatric mood disorders: A systematic review

AIM

To appraise the current evidence on the efficacy and safety of lamotrigine (LAM) in the treatment of pediatric mood disorders (PMD) (i.e., Major Depressive disorder [MDD], bipolar disorder [BD]).

METHODS

Major databases were searched for randomized controlled trials (RCTs), open-label trials, and observational studies reporting on pediatric (age < 18 years) patients treated with LAM for mood disorders.

RESULTS

A total of 3061 abstracts were screened and seven articles were selected for inclusion. Seven studies (319 BD and 43 MDD patients), including one RCT (n = 173), three prospective (n = 105), and three retrospective (n = 84) studies, met the study criteria with a study duration range from 8 to 60.9 weeks. The mean age of this pooled data is 14.6 ± 2.0 years. LAM daily dosage varied from 12.5 to 391.3 mg/day among the studies. In an important finding, the RCT reported favorable outcomes with LAM (HR = 0.46; p = 0.02) in 13- to 17-year-old age group as compared with 10- to 12-year-old age group (HR = 0.93; p = 0.88). In addition, time to occurrence of a bipolar event trended toward favoring LAM over placebo. All the studies identified LAM as an effective and safe drug in PMDs especially, BDs. Overall, LAM was well tolerated with no major significant side effects and no cases of Stevens-Johnson syndrome.

CONCLUSIONS

Most studies suggested that LAM was safe and effective in pediatric patients with mood disorders. However, the data regarding the therapeutic range for LAM are lacking. Based on the data, there is inconsistent evidence to make conclusive recommendations on therapeutic LAM dosage for mood improvement in the pediatric population. Further studies including larger sample sizes are required to address this relevant clinical question.

Abstract P403: In Situ Physiological and Behavioral Monitoring With Digital Sensors for Cerebrovascular Disease: A Scoping Review

Background: Cerebrovascular disease is a life-threatening neurological event and a leading cause of long-term disability and death worldwide. Early detection of characteristic behavioral and physiological changes associated with cerebrovascular disease is critical to improving patient outcomes and quality of life measures. The growing prevalence of remote monitoring tools, from wearable devices to smartphone applications, that facilitate in situ observation of patients and the environments of daily life holds promise for more timely recognition and possible prevention of cerebrovascular accidents (CVA) like stroke.

Objective: The goal of this scoping review is to examine and establish categories of innovation with digital sensors that monitor physiological and behavioral variables in situ to augment the current screening and diagnostic processes for patients with cerebrovascular disease.

Methods: Guided by the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist, a robust search strategy for spanning multiple databases from 2012 to September 30, 2022, excluding review articles, articles including interventions, and articles not published in English, was implemented. Among the databases searched were Web of Science; Scopus; Ovid Embase; Ovid Cochrane Central Register of Controlled Trials; and Ovid MEDLINE and Epub ahead of print, in-process and other nonindexed citations, and daily.

Results: This search strategy aggregated 689 articles, of which 101 (14.7%) articles met the inclusion criteria for this scoping review. Articles were divided into two categories based on their focus: physiological and behavioral. Articles with a physiological focus were sorted into one of nine subcategories according to the signal(s) monitored: motor function, heart rhythm, heart rate, kinematic analysis, physical activity, blood pressure, sensory deficit, electrodermal activity, and intracranial pressure. Articles focusing on behavioral variables were sorted into two subcategories: mood and fatigue. Most studies used an ECG-enabled smartwatch, like an Apple Watch 3, or passive smartphone sensors.

Conclusions: This scoping review identified disparate methods and conclusions associated with the use of digital sensors for in situ physiological and behavioral monitoring of cerebrovascular disease patients. While most articles evaluated pilot validation and feasibility trials, the lack of randomized controlled trials is a critical literature gap specific to this evolving research area.

PISTACHIo (PreemptIon of diSrupTive behAvior in CHIldren): real-time monitoring of sleep and behavior of children 3-7 years old receiving parent-child interaction therapy augment with artificial intelligence - the study protocol, pilot study

BACKGROUND

Emotional behavior problems (EBP) are the most common and persistent mental health issues in early childhood. Early intervention programs are crucial in helping children with EBP. Parent-child interaction therapy (PCIT) is an evidence-based therapy designed to address personal difficulties of parent-child dyads as well as reduce externalizing behaviors. In clinical practice, parents consistently struggle to provide accurate characterizations of EBP symptoms (number, timing of tantrums, precipitating events) even from the week before in their young children. The main aim of the study is to evaluate feasibility of the use of smartwatches in children aged 3-7 years with EBP.

METHODS

This randomized double-blind controlled study aims to recruit a total of 100 participants, consisting of 50 children aged 3-7 years with an EBP measure rated above the clinically significant range (T-score ≥ 60) (Eyberg Child Behavior Inventory-ECBI; Eyberg & Pincus, 1999) and their parents who are at least 18 years old. Participants are randomly assigned to the artificial intelligence-PCIT group (AI-PCIT) or the PCIT-sham biometric group. Outcome parameters include weekly ECBI and Pediatric Sleep Questionnaire (PSQ) as well as Child Behavior Checklist (CBCL) obtained weeks 1, 6, and 12 of the study. Two smartphone applications (Garmin connect and mEMA) and a wearable Garmin smartwatch are used collect the data to monitor step count, sleep, heart rate, and activity intensity. In the AI-PCIT group, the mEMA application will allow for the ecological momentary assessment (EMA) and will send behavioral alerts to the parent.

DISCUSSION

Real-time predictive technologies to engage patients rely on daily commitment on behalf of the participant and recurrent frequent smartphone notifications. Ecological momentary assessment (EMA) provides a way to digitally phenotype in-the-moment behavior and functioning of the parent-child dyad. One of the study's goals is to determine if AI-PCIT outcomes are superior in comparison with standard PCIT. Overall, we believe that the PISTACHIo study will also be able to determine tolerability of smartwatches in children aged 3-7 with EBP and could participate in a fundamental shift from the traditional way of assessing and treating EBP to a more individualized treatment plan based on real-time information about the child's behavior.

TRIAL REGISTRATION

The ongoing clinical trial study protocol conforms to the international Consolidated Standards of Reporting Trials (CONSORT) guidelines and is registered in clinicaltrials.gov (ID: NCT05077722), an international clinical trial registry.

Environmental chemicals and endogenous metabolites in bile of USA and Norway patients with primary sclerosing cholangitis

Primary sclerosing cholangitis (PSC) is a complex bile duct disorder. Its etiology is incompletely understood, but environmental chemicals likely contribute to risk. Patients with PSC have an altered bile metabolome, which may be influenced by environmental chemicals. This novel study utilized state-of-the-art high-resolution mass spectrometry (HRMS) with bile samples to provide the first characterization of environmental chemicals and metabolomics (collectively, the exposome) in PSC patients located in the United States of America (USA) (n = 24) and Norway (n = 30). First, environmental chemical- and metabolome-wide association studies were conducted to assess geographic-based similarities and differences in the bile of PSC patients. Nine environmental chemicals (false discovery rate, FDR < 0.20) and 3143 metabolic features (FDR < 0.05) differed by site. Next, pathway analysis was performed to identify metabolomic pathways that were similarly and differentially enriched by the site. Fifteen pathways were differentially enriched (P < .05) in the categories of amino acid, glycan, carbohydrate, energy, and vitamin/cofactor metabolism. Finally, chemicals and pathways were integrated to derive exposure-effect correlation networks by site. These networks demonstrate the shared and differential chemical-metabolome associations by site and highlight important pathways that are likely relevant to PSC. The USA patients demonstrated higher environmental chemical bile content and increased associations between chemicals and metabolic pathways than those in Norway. Polychlorinated biphenyl (PCB)-118 and PCB-101 were identified as chemicals of interest for additional investigation in PSC given broad associations with metabolomic pathways in both the USA and Norway patients. Associated pathways include glycan degradation pathways, which play a key role in microbiome regulation and thus may be implicated in PSC pathophysiology.

Evidence for machine learning guided early prediction of acute outcomes in the treatment of depressed children and adolescents with antidepressants

BACKGROUND

The treatment of depression in children and adolescents is a substantial public health challenge. This study examined artificial intelligence tools for the prediction of early outcomes in depressed children and adolescents treated with fluoxetine, duloxetine, or placebo.

METHODS

The study samples included training datasets (N = 271) from patients with major depressive disorder (MDD) treated with fluoxetine and testing datasets from patients with MDD treated with duloxetine (N = 255) or placebo (N = 265). Treatment trajectories were generated using probabilistic graphical models (PGMs). Unsupervised machine learning identified specific depressive symptom profiles and related thresholds of improvement during acute treatment.

RESULTS

Variation in six depressive symptoms (difficulty having fun, social withdrawal, excessive fatigue, irritability, low self-esteem, and depressed feelings) assessed with the Children's Depression Rating Scale-Revised at 4-6 weeks predicted treatment outcomes with fluoxetine at 10-12 weeks with an average accuracy of 73% in the training dataset. The same six symptoms predicted 10-12 week outcomes at 4-6 weeks in (a) duloxetine testing datasets with an average accuracy of 76% and (b) placebo-treated patients with accuracies of 67%. In placebo-treated patients, the accuracies of predicting response and remission were similar to antidepressants. Accuracies for predicting nonresponse to placebo treatment were significantly lower than antidepressants.

CONCLUSIONS

PGMs provided clinically meaningful predictions in samples of depressed children and adolescents treated with fluoxetine or duloxetine. Future work should augment PGMs with biological data for refined predictions to guide the selection of pharmacological and psychotherapeutic treatment in children and adolescents with depression.

Systematic review: Wearable remote monitoring to detect nonalcohol/nonnicotine-related substance use disorder symptoms

BACKGROUND AND OBJECTIVES

Substance use disorders (SUDs) are chronic relapsing diseases characterized by significant morbidity and mortality. Phenomenologically, patients with SUDs present with a repeating cycle of intoxication, withdrawal, and craving, significantly impacting their diagnosis and treatment. There is a need for better identification and monitoring of these disease states. Remote monitoring chronic illness with wearable devices offers a passive, unobtrusive, constant physiological data assessment. We evaluate the current evidence base for remote monitoring of nonalcohol, nonnicotine SUDs.

METHODS

We performed a systematic, comprehensive literature review and screened 1942 papers.

RESULTS

We found 15 studies that focused mainly on the intoxication stage of SUD. These studies used wearable sensors measuring several physiological parameters (ECG, HR, O2 , Accelerometer, EDA, temperature) and implemented study-specific algorithms to evaluate the data.

DISCUSSION AND CONCLUSIONS

Studies were extracted, organized, and analyzed based on the three SUD disease states. The sample sizes were relatively small, focused primarily on the intoxication stage, had low monitoring compliance, and required significant computational power preventing "real-time" results. Cardiovascular data was the most consistently valuable data in the predictive algorithms. This review demonstrates that there is currently insufficient evidence to support remote monitoring of SUDs through wearable devices.

SCIENTIFIC SIGNIFICANCE

This is the first systematic review to show the available data on wearable remote monitoring of SUD symptoms in each stage of the disease cycle. This clinically relevant approach demonstrates what we know and do not know about the remote monitoring of SUDs within disease states.

Network science approach elucidates integrative genomic-metabolomic signature of antidepressant response and lifetime history of attempted suicide in adults with major depressive disorder

Background: Individuals with major depressive disorder (MDD) and a lifetime history of attempted suicide demonstrate lower antidepressant response rates than those without a prior suicide attempt. Identifying biomarkers of antidepressant response and lifetime history of attempted suicide may help augment pharmacotherapy selection and improve the objectivity of suicide risk assessments. Towards this goal, this study sought to use network science approaches to establish a multi-omics (genomic and metabolomic) signature of antidepressant response and lifetime history of attempted suicide in adults with MDD. Methods: Single nucleotide variants (SNVs) which associated with suicide attempt(s) in the literature were identified and then integrated with a) p180-assayed metabolites collected prior to antidepressant pharmacotherapy and b) a binary measure of antidepressant response at 8 weeks of treatment using penalized regression-based networks in 245 'Pharmacogenomics Research Network Antidepressant Medication Study (PGRN-AMPS)' and 103 'Combining Medications to Enhance Depression Outcomes (CO-MED)' patients with major depressive disorder. This approach enabled characterization and comparison of biological profiles and associated antidepressant treatment outcomes of those with (N = 46) and without (N = 302) a self-reported lifetime history of suicide attempt. Results: 351 SNVs were associated with suicide attempt(s) in the literature. Intronic SNVs in the circadian genes CLOCK and ARNTL (encoding the CLOCK:BMAL1 heterodimer) were amongst the top network analysis features to differentiate patients with and without a prior suicide attempt. CLOCK and ARNTL differed in their correlations with plasma phosphatidylcholines, kynurenine, amino acids, and carnitines between groups. CLOCK and ARNTL-associated phosphatidylcholines showed a positive correlation with antidepressant response in individuals without a prior suicide attempt which was not observed in the group with a prior suicide attempt. Conclusion: Results provide evidence for a disturbance between CLOCK:BMAL1 circadian processes and circulating phosphatidylcholines, kynurenine, amino acids, and carnitines in individuals with MDD who have attempted suicide. This disturbance may provide mechanistic insights for differential antidepressant pharmacotherapy outcomes between patients with MDD with versus without a lifetime history of attempted suicide. Future investigations of CLOCK:BMAL1 metabolic regulation in the context of suicide attempts may help move towards biologically-augmented pharmacotherapy selection and stratification of suicide risk for subgroups of patients with MDD and a lifetime history of attempted suicide.

ACE2 and TMPRSS2 SARS-CoV-2 infectivity genes: deep mutational scanning and characterization of missense variants

The human angiotensin-converting enzyme 2 (ACE2) and transmembrane serine protease 2 (TMPRSS2) proteins play key roles in the cellular internalization of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the coronavirus responsible for the coronavirus disease of 2019 (COVID-19) pandemic. We set out to functionally characterize the ACE2 and TMPRSS2 protein abundance for variant alleles encoding these proteins that contained non-synonymous single-nucleotide polymorphisms (nsSNPs) in their open reading frames (ORFs). Specifically, a high-throughput assay, deep mutational scanning (DMS), was employed to test the functional implications of nsSNPs, which are variants of uncertain significance in these two genes. Specifically, we used a 'landing pad' system designed to quantify the protein expression for 433 nsSNPs that have been observed in the ACE2 and TMPRSS2 ORFs and found that 8 of 127 ACE2, 19 of 157 TMPRSS2 isoform 1 and 13 of 149 TMPRSS2 isoform 2 variant proteins displayed less than ~25% of the wild-type protein expression, whereas 4 ACE2 variants displayed 25% or greater increases in protein expression. As a result, we concluded that nsSNPs in genes encoding ACE2 and TMPRSS2 might potentially influence SARS-CoV-2 infectivity. These results can now be applied to DNA sequence data for patients infected with SARS-CoV-2 to determine the possible impact of patient-based DNA sequence variation on the clinical course of SARS-CoV-2 infection.

Toward a Definition of "No Meaningful Benefit" From Antidepressant Treatment: An Equipercentile Analysis With Cross-Trial Validation Across Multiple Rating Scales

Background: Many patients with major depressive disorder (MDD) who experience no meaningful benefit (NMB) from antidepressive treatment go undetected. However, there is a lack of consensus on the definition of NMB from antidepressants. Methods: Equipercentile linking was used to identify a threshold for percent change in 17-item Hamilton Depression Rating Scale (HDRS-17) scores that equated with a Clinical Global Impressions-Improvement (CGI-I) score of 3 (minimally improved), a proxy for NMB, after 4 and 8 weeks of citalopram or escitalopram treatment, using data from the Pharmacogenomic Research Network Antidepressant Medication Pharmacogenomic Study (PGRN-AMPS). The NMB threshold for the HDRS-17 was validated by equating a CGI-I rating of 3 with percent change values from the clinician- and patient-rated versions of the Quick Inventory of Depressive Symptomatology (QIDS-C and QIDS-SR) using data from PGRN-AMPS and phase 1 of the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial. This study was conducted between June 2021 and September 2021. Results: In PGRN-AMPS, a 30% improvement in HDRS-17 score corresponded to a CGI-I rating of 3 at 4 and 8 weeks. The 30% improvement threshold was also observed for QIDS-C and QIDS-SR scores in both PGRN-AMPS and STAR*D. Similar results were observed for percent change in HDRS-17 and QIDS-based measures in lower- and higher-severity groups based on a median split of baseline total scores. Conclusions: Improvement in depressive severity of ≤ 30%, as assessed using the HDRS-17, QIDS-C, and QIDS-SR, may validly define NMB from antidepressants during short-term treatment.

Toward Individualized Prediction of Response to Methotrexate in Early Rheumatoid Arthritis: A Pharmacogenomics-Driven Machine Learning Approach

OBJECTIVE

To test the ability of machine learning (ML) approaches with clinical and genomic biomarkers to predict methotrexate treatment response in patients with early rheumatoid arthritis (RA).

METHODS

Demographic, clinical, and genomic data from 643 patients of European ancestry with early RA (mean age 54 years; 70% female) subdivided into a training (n = 336) and validation cohort (n = 307) were used. The genomic data comprised 160 single-nucleotide polymorphisms (SNPs) previously associated with RA or methotrexate metabolism. Response to methotrexate monotherapy was defined as good or moderate by the European Alliance of Associations for Rheumatology (EULAR) response criteria at the 3-month follow-up. Supervised ML methods were trained with 5 repeats and 10-fold cross-validation using the training cohort. Prediction performance was validated in the independent validation cohort.

RESULTS

Supervised ML methods combining age, sex, smoking, rheumatoid factor, baseline Disease Activity Score in 28 joints (DAS28) scores and 160 SNPs predicted EULAR response at 3 months with the area under the receiver operating curve of 0.84 (P = 0.05) in the training cohort and achieved a prediction accuracy of 76% (P = 0.05) in the validation cohort (sensitivity 72%, specificity 77%). Intergenic SNPs rs12446816, rs13385025, rs113798271, and ATIC (rs2372536) had variable importance above 60.0 and along with baseline DAS28 scores were among the top predictors of methotrexate response.

CONCLUSION

Pharmacogenomic biomarkers combined with baseline DAS28 scores can be useful in predicting response to methotrexate in patients with early RA. Applying ML to predict treatment response holds promise for guiding effective RA treatment choices, including timely escalation of RA therapies.

A Characterization of the Clinical Global Impression Scale Thresholds in the Treatment of Adolescent Depression Across Multiple Rating Scales

Introduction: The Clinical Global Impressions-Improvement (CGI-I) scale is widely used in clinical research to assess symptoms and functioning in the context of treatment. The correlates of the CGI-I with efficacy scales for adolescent major depressive disorder are poorly understood. This study focused on benchmarking CGI-I scores with changes in the Children's Depression Rating Scale-Revised (CDRS-R) and the Quick Inventory of Depressive Symptomatology-Adolescent (17-item) Self-Report (QIDS-A17-SR). Methods: We examined three datasets with the clinician-rated CDRS-R to ascertain equivalent percent changes in total scores and CGI-I ratings. Exploratory analyses examined corresponding percentage changes in the QIDS-A17-SR and the CGI-I ratings. The CGI-I was the reference scale for nonparametric equipercentile linking with the Equate package in R. Results: CGI-I scores of 1 mapped to ≥78%-95% change in CDRS-R scores at 4-6 weeks across three datasets. CGI-I scores of 2 mapped to 56%-94% change in CDRS-R scores at 4-6 weeks across three studies. CGI-I scores of 3 mapped to 30%-68% changes in CDRS-R scores at 4-6 weeks across three studies. CGI-I scores of 4 mapped to a range of 29%-44% at 4-6 weeks across three studies. There was no significant difference (p ≥ 0.6) between treatment groups in both the Treatment of Adolescents with Depression and Treatment of Resistant Depression in Adolescents studies, for each CGI-I score ( = 1, or = 2 or = 3, or ≥4), associated mapping of total depression severity score, or associated percent change from baseline for corresponding follow-up visits. There was no significant sex difference (p > 0.2) in CGI-I linkages to CDRS-R total or percentage changes. Conclusions: These findings establish clear relationships among CGI-I scores and the CDRS-R and the QIDS-A17-SR. These benchmarks have utility for clinical trial study design, inter-rater reliability training, and clinical implementation.

Multi-Omics Characterization of Early- and Adult-Onset Major Depressive Disorder

Age at depressive onset (AAO) corresponds to unique symptomatology and clinical outcomes. Integration of genome-wide association study (GWAS) results with additional “omic” measures to evaluate AAO has not been reported and may reveal novel markers of susceptibility and/or resistance to major depressive disorder (MDD). To address this gap, we integrated genomics with metabolomics using data-driven network analysis to characterize and differentiate MDD based on AAO. This study first performed two GWAS for AAO as a continuous trait in (a) 486 adults from the Pharmacogenomic Research Network-Antidepressant Medication Pharmacogenomic Study (PGRN-AMPS), and (b) 295 adults from the Combining Medications to Enhance Depression Outcomes (CO-MED) study. Variants from top signals were integrated with 153 p180-assayed metabolites to establish multi-omics network characterizations of early (<age 18) and adult-onset depression. The most significant variant (p = 8.77 × 10−8) localized to an intron of SAMD3. In silico functional annotation of top signals (p < 1 × 10−5) demonstrated gene expression enrichment in the brain and during embryonic development. Network analysis identified differential associations between four variants (in/near INTU, FAT1, CNTN6, and TM9SF2) and plasma metabolites (phosphatidylcholines, carnitines, biogenic amines, and amino acids) in early- compared with adult-onset MDD. Multi-omics integration identified differential biosignatures of early- and adult-onset MDD. These biosignatures call for future studies to follow participants from childhood through adulthood and collect repeated -omics and neuroimaging measures to validate and deeply characterize the biomarkers of susceptibility and/or resistance to MDD development.

A pilot spectroscopy study of adversity in adolescents

Background

Childhood adversity is a global health problem affecting 25-50% of children worldwide. Few prior studies have examined the underlying neurochemistry of adversity in adolescents. This cross-sectional study examined spectroscopic markers of trauma in a cohort of adolescents with major depressive disorder (MDD) and healthy controls. We hypothesized that historical adversity would have a negative relationship with spectroscopic measures of glutamate metabolites in anterior cingulate cortex.

Methods

Adolescent participants (aged 13-21) underwent a semi-structured diagnostic interview and clinical assessment, which included the self-report Childhood Trauma Questionnaire (CTQ), a 28-item assessment of childhood adversity. Proton magnetic resonance spectroscopy (1H-MRS) scans at 3 Tesla of an anterior cingulate cortex (ACC) voxel (8 cm3) encompassing both hemispheres were collected using a 2-dimensional J-averaged sequence to assess N-acetylaspartate (NAA), Glx (glutamate+glutamine) and [NAA]/[Glx] concentrations. Generalized linear models assessed the relationships between CTQ scores and metabolite levels in ACC.

Results

Thirty-nine participants (17 healthy controls, 22 depressed participants) underwent 1H-MRS and completed the CTQ measures. There were decrements in [NAA]/[Glx] ratio in the ACC of participants with childhood adversity while no significant relationship between CTQ total score and any of the ACC metabolites was found in the combined sample. Exploratory results revealed a positive association between Glx levels and CTQ scores in depressed participants. Conversely the [NAA]/[Glx] ratio had a negative association with total CTQ scores in the depressed participants. Emotional Abuse Scale showed a significant negative relationship with [NAA]/[Glx] ratio in the combined sample when adjusted for depression severity.

Conclusions

Our findings suggest that childhood adversity may impact brain neurochemical profiles. Further longitudinal studies should examine neurochemical correlates of childhood adversity throughout development and in populations with other psychiatric disorders.

A Retrospective Examination of the Impact of Pharmacotherapy on Parent-Child Interaction Therapy

Objective: Parent-child interaction therapy (PCIT) is an evidence-based approach for children aged 2-7 years with disruptive behavior problems. This study examined the effectiveness of PCIT with and without concurrent pharmacotherapy. Methods: A convenience sample was collected from a retrospective chart review of preschool-aged children treated with PCIT at the Mayo Clinic Young Child Clinic between 2016 and 2020. Quantitative and qualitative data were abstracted from all patients. The sample was divided into two groups based on psychotropic medications status (medicated and unmedicated) at the initiation of PCIT. Effectiveness of treatment was assessed with the change in Eyberg Child Behavior Inventory (ECBI) score. The change over time in ECBI score was compared between the two PCIT groups with and without concurrent pharmacotherapy using a linear mixed model. Results: Of the 62 youth, 38.71% were females. Mean age was 4.71 ± 1.17 years. The mean baseline ECBI score was 148.74 ± 30.86, indicating clinically significant disruptive behaviors. The mean number of PCIT sessions was 6.59 ± 3.82. There was no statistically significant difference in ECBI scores between the two groups at pre-PCIT (medication group: 149.68, standard error [SE] = 11.61 vs. unmedicated group: 147.92, SE = 10.93, p = 0.8904) and at post-PCIT (medication group: 116.27 [SE = 11.89] vs. unmedicated group: 128.86 [SE = 11.57], p = 0.3464). There was a statistically significant improvement in ECBI scores for both groups after completing therapy (medication group = -33.41 [-22.32%], SE = 6.27, p < 0.0001; d = 1.144; unmedicated group = -19.06 [-12.88%], SE = 5.78, p = 0.0022; d = 1.078). Conclusions: PCIT reduced disruptive behaviors in this sample of young children regardless of concurrent pharmacotherapy. Future prospective studies should consider one particular pharmacological agent and long-term outcomes of treatment. PCIT and certain pharmacological treatments could have complex and important bidirectional priming effects for both treatments.

Multi-omics driven predictions of response to acute phase combination antidepressant therapy: a machine learning approach with cross-trial replication

Combination antidepressant pharmacotherapies are frequently used to treat major depressive disorder (MDD). However, there is no evidence that machine learning approaches combining multi-omics measures (e.g., genomics and plasma metabolomics) can achieve clinically meaningful predictions of outcomes to combination pharmacotherapy. This study examined data from 264 MDD outpatients treated with citalopram or escitalopram in the Mayo Clinic Pharmacogenomics Research Network Antidepressant Medication Pharmacogenomic Study (PGRN-AMPS) and 111 MDD outpatients treated with combination pharmacotherapies in the Combined Medications to Enhance Outcomes of Antidepressant Therapy (CO-MED) study to predict response to combination antidepressant therapies. To assess whether metabolomics with functionally validated single-nucleotide polymorphisms (SNPs) improves predictability over metabolomics alone, models were trained/tested with and without SNPs. Models trained with PGRN-AMPS' and CO-MED's escitalopram/citalopram patients predicted response in CO-MED's combination pharmacotherapy patients with accuracies of 76.6% (p < 0.01; AUC: 0.85) without and 77.5% (p < 0.01; AUC: 0.86) with SNPs. Then, models trained solely with PGRN-AMPS' escitalopram/citalopram patients predicted response in CO-MED's combination pharmacotherapy patients with accuracies of 75.3% (p < 0.05; AUC: 0.84) without and 77.5% (p < 0.01; AUC: 0.86) with SNPs, demonstrating cross-trial replication of predictions. Plasma hydroxylated sphingomyelins were prominent predictors of treatment outcomes. To explore the relationship between SNPs and hydroxylated sphingomyelins, we conducted multi-omics integration network analysis. Sphingomyelins clustered with SNPs and metabolites related to monoamine neurotransmission, suggesting a potential functional relationship. These results suggest that integrating specific metabolites and SNPs achieves accurate predictions of treatment response across classes of antidepressants. Finally, these results motivate functional investigation into how sphingomyelins might influence MDD pathophysiology, antidepressant response, or both.

Preliminary Evidence for Anhedonia as a Marker of Sexual Trauma in Female Adolescents

Introduction

Major depressive disorder (MDD) is a common condition with heterogeneous presentations that often include predominant anhedonia. Previous studies have revealed that childhood trauma is a potent risk factor for the development of MDD; however, the clinical implications of this finding are not fully understood.

Methods

Participants were adolescents (age 13–21 years) with a diagnosis of moderate-to-severe major depressive disorder and healthy controls. We used generalized linear models to assess the relationship between anhedonia severity and trauma severity in a cross-sectional dataset.

Results

This cross-sectional analysis of an adolescent sample that underwent clinical evaluations and a trauma assessment, suggested that anhedonia was associated with historical trauma severity. The association between anhedonia and sexual abuse was greater in female participants compared to male participants.

Discussion

Our results were partially in line with the reported literature in adult samples. Future studies aiming to characterize the trauma–anhedonia relationship in adolescents should utilize scales designed specifically to measure these constructs in young populations, and scales that assess specific subtypes of anhedonia.

Prediction of short-term antidepressant response using probabilistic graphical models with replication across multiple drugs and treatment settings

Heterogeneity in the clinical presentation of major depressive disorder and response to antidepressants limits clinicians' ability to accurately predict a specific patient's eventual response to therapy. Validated depressive symptom profiles may be an important tool for identifying poor outcomes early in the course of treatment. To derive these symptom profiles, we first examined data from 947 depressed subjects treated with selective serotonin reuptake inhibitors (SSRIs) to delineate the heterogeneity of antidepressant response using probabilistic graphical models (PGMs). We then used unsupervised machine learning to identify specific depressive symptoms and thresholds of improvement that were predictive of antidepressant response by 4 weeks for a patient to achieve remission, response, or nonresponse by 8 weeks. Four depressive symptoms (depressed mood, guilt feelings and delusion, work and activities and psychic anxiety) and specific thresholds of change in each at 4 weeks predicted eventual outcome at 8 weeks to SSRI therapy with an average accuracy of 77% (p = 5.5E-08). The same four symptoms and prognostic thresholds derived from patients treated with SSRIs correctly predicted outcomes in 72% (p = 1.25E-05) of 1996 patients treated with other antidepressants in both inpatient and outpatient settings in independent publicly-available datasets. These predictive accuracies were higher than the accuracy of 53% for predicting SSRI response achieved using approaches that (i) incorporated only baseline clinical and sociodemographic factors, or (ii) used 4-week nonresponse status to predict likely outcomes at 8 weeks. The present findings suggest that PGMs providing interpretable predictions have the potential to enhance clinical treatment of depression and reduce the time burden associated with trials of ineffective antidepressants. Prospective trials examining this approach are forthcoming.

Selective Serotonin Reuptake Inhibitor Pharmaco-Omics: Mechanisms and Prediction

Selective serotonin reuptake inhibitors (SSRIs) are a standard of care for the pharmacotherapy of patients suffering from Major Depressive Disorder (MDD). However, only one-half to two-thirds of MDD patients respond to SSRI therapy. Recently, a "multiple omics" research strategy was applied to identify genetic differences between patients who did and did not respond to SSRI therapy. As a first step, plasma metabolites were assayed using samples from the 803 patients in the PGRN-AMPS SSRI MDD trial. The metabolomics data were then used to "inform" genomics by performing a genome-wide association study (GWAS) for plasma concentrations of the metabolite most highly associated with clinical response, serotonin (5-HT). Two genome-wide or near genome-wide significant single nucleotide polymorphism (SNP) signals were identified, one that mapped near the TSPAN5 gene and another across the ERICH3 gene, both genes that are highly expressed in the brain. Knocking down TSPAN5 and ERICH3 resulted in decreased 5-HT concentrations in neuroblastoma cell culture media and decreased expression of enzymes involved in 5-HT biosynthesis and metabolism. Functional genomic studies demonstrated that ERICH3 was involved in clathrin-mediated vesicle formation and TSPAN5 was an ethanol-responsive gene that may be a marker for response to acamprosate pharmacotherapy of alcohol use disorder (AUD), a neuropsychiatric disorder highly co-morbid with MDD. In parallel studies, kynurenine was the plasma metabolite most highly associated with MDD symptom severity and application of a metabolomics-informed pharmacogenomics approach identified DEFB1 and AHR as genes associated with variation in plasma kynurenine levels. Both genes also contributed to kynurenine-related inflammatory pathways. Finally, a multiply replicated predictive algorithm for SSRI clinical response with a balanced predictive accuracy of 76% (compared with 56% for clinical data alone) was developed by including the SNPs in TSPAN5, ERICH3, DEFB1 and AHR. In summary, application of a multiple omics research strategy that used metabolomics to inform genomics, followed by functional genomic studies, identified novel genes that influenced monoamine biology and made it possible to develop a predictive algorithm for SSRI clinical outcomes in MDD. A similar pharmaco-omic research strategy might be broadly applicable for the study of other neuropsychiatric diseases and their drug therapy.

Real-world experiences with acupuncture among breast cancer survivors: a cross-sectional survey study

PURPOSE

The purpose of this study was to evaluate acupuncture use among breast cancer survivors, including perceived symptom improvements and referral patterns.

METHODS

Breast cancer survivors who had used acupuncture for cancer- or treatment-related symptoms were identified using an ongoing prospective Mayo Clinic Breast Disease Registry (MCBDR). Additionally, Mayo Clinic electronic health records (MCEHR) were queried to identify eligible participants. All received a mailed consent form and survey including acupuncture-related questions about acupuncture referrals, delivery, and costs. Respondents were also asked to recall symptom severity before and after acupuncture treatment and time to benefit on Likert scales.

RESULTS

Acupuncture use was reported among 415 participants (12.3%) of the MCBDR. Among MCBDR and MCEHR eligible participants, 241 women returned surveys. A total of 193 (82.1%) participants reported a symptomatic benefit from acupuncture, and 57 (24.1% of participants) reported a "substantial benefit" or "totally resolved my symptoms" (corresponding to 4 and 5 on the 5-point Likert scale). The mean symptom severity decreased by at least 1 point of the 5-point scale for each symptom; the percentage of patients who reported an improvement in symptoms ranged from 56% (lymphedema) to 79% (headache). The majority of patients reported time to benefit as "immediate" (34%) or "after a few treatments" (40.4%). Over half of the participants self-referred for treatment; 24.1% were referred by their oncologist. Acupuncture delivery was more frequent in private offices (61.0%) than in hospital or medical settings (42.3%). Twelve participants (5.1%) reported negative side effects, such as discomfort.

CONCLUSIONS

Acupuncture is commonly utilized by patients for a variety of breast cancer-related symptoms. However, patients frequently self-refer for acupuncture treatments, and most acupuncture care is completed at private offices, rather than medical clinic or hospital settings.

Pharmacogenomics-Driven Prediction of Antidepressant Treatment Outcomes: A Machine-Learning Approach With Multi-trial Replication

We set out to determine whether machine learning–based algorithms that included functionally validated pharmacogenomic biomarkers joined with clinical measures could predict selective serotonin reuptake inhibitor (SSRI) remission/response in patients with major depressive disorder (MDD). We studied 1,030 white outpatients with MDD treated with citalopram/escitalopram in the Mayo Clinic Pharmacogenomics Research Network Antidepressant Medication Pharmacogenomic Study (PGRN‐AMPS; n = 398), Sequenced Treatment Alternatives to Relieve Depression (STAR*D; n = 467), and International SSRI Pharmacogenomics Consortium (ISPC; n = 165) trials. A genomewide association study for PGRN‐AMPS plasma metabolites associated with SSRI response (serotonin) and baseline MDD severity (kynurenine) identified single nucleotide polymorphisms (SNPs) in DEFB1,ERICH3,AHR, and TSPAN5 that we tested as predictors. Supervised machine‐learning methods trained using SNPs and total baseline depression scores predicted remission and response at 8 weeks with area under the receiver operating curve (AUC) > 0.7 (P < 0.04) in PGRN‐AMPS patients, with comparable prediction accuracies > 69% (P ≤ 0.07) in STAR*D and ISPC. These results demonstrate that machine learning can achieve accurate and, importantly, replicable prediction of SSRI therapy response using total baseline depression severity combined with pharmacogenomic biomarkers.

Machine Learning Helps Identify New Drug Mechanisms in Triple-Negative Breast Cancer

This paper demonstrates the ability of mach- ine learning approaches to identify a few genes among the 23,398 genes of the human genome to experiment on in the laboratory to establish new drug mechanisms. As a case study, this paper uses MDA-MB-231 breast cancer single-cells treated with the antidiabetic drug metformin. We show that mixture-model-based unsupervised methods with validation from hierarchical clustering can identify single-cell subpopulations (clusters). These clusters are characterized by a small set of genes (1% of the genome) that have significant differential expression across the clusters and are also highly correlated with pathways with anticancer effects driven by metformin. Among the identified small set of genes associated with reduced breast cancer incidence, laboratory experiments on one of the genes, CDC42, showed that its downregulation by metformin inhibited cancer cell migration and proliferation, thus validating the ability of machine learning approaches to identify biologically relevant candidates for laboratory experiments. Given the large size of the human genome and limitations in cost and skilled resources, the broader impact of this work in identifying a small set of differentially expressed genes after drug treatment lies in augmenting the drug-disease knowledge of pharmacogenomics experts in laboratory investigations, which could help establish novel biological mechanisms associated with drug response in diseases beyond breast cancer.

Prediction of adenocarcinoma development using game theory

Recent research shows that gene expression changes appear to correlate well with the progression of many types of cancers. Using changes in gene expression as a basis, this paper proposes a data-driven 2-player game-theoretic model to predict the risk of adenocarcinoma based on Nash equilibrium. A key innovation in this work is the pay-off function which is a weighted composite of the expression of a cohort of tumor-suppressor genes (as one player) and an analogous cohort of oncogenes (as the other player). Another novelty of the model is its ability to predict the risk that a healthy sample will develop adenocarcinoma, if its associated gene expression is comparable to that of early-stage tumor samples. The model is validated using two of the largest publicly available adenocarcinoma datasets. The results show that i) the model is able to distinguish between healthy and cancerous samples with an accuracy of 93%, and ii) 95% of the healthy samples said to be at risk had gene expressions comparable to those of samples with stage I or stage II tumors, thereby predicting the imminent onset of adenocarcinoma.

Model-based unsupervised learning informs metformin-induced cell-migration inhibition through an AMPK-independent mechanism in breast cancer

We demonstrate that model-based unsupervised learning can uniquely discriminate single-cell subpopulations by their gene expression distributions, which in turn allow us to identify specific genes for focused functional studies. This method was applied to MDA-MB-231 breast cancer cells treated with the antidiabetic drug metformin, which is being repurposed for treatment of triple-negative breast cancer. Unsupervised learning identified a cluster of metformin-treated cells characterized by a significant suppression of 230 genes (p-value < 2E-16). This analysis corroborates known studies of metformin action: a) pathway analysis indicated known mechanisms related to metformin action, including the citric acid (TCA) cycle, oxidative phosphorylation, and mitochondrial dysfunction (p-value < 1E-9); b) 70% of these 230 genes were functionally implicated in metformin response; c) among remaining lesser functionally-studied genes for metformin-response was CDC42, down-regulated in breast cancer treated with metformin. However, CDC42's mechanisms in metformin response remained unclear. Our functional studies showed that CDC42 was involved in metformin-induced inhibition of cell proliferation and cell migration mediated through an AMPK-independent mechanism. Our results points to 230 genes that might serve as metformin response signatures, which needs to be tested in patients treated with metformin and, further investigation of CDC42 and AMPK-independence's role in metformin's anticancer mechanisms.