Introducing New GRAPPA Projects.

Every year at the Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA) annual meeting, new project ideas are presented and discussed with a view to obtaining feedback and support. Arising from previous work, a project proposal was presented at the 2023 meeting; the project aims to improve early diagnosis of psoriatic arthritis (PsA) by comparing a physician-based vs a patient questionnaire-based approach. This project has received the backing of the GRAPPA research committee, but additional funding will be required. A second project, approved by GRAPPA, was presented on delivering an epidemiology training module before the GRAPPA annual meeting in 2024, which will target both established GRAPPA clinicians and trainees. Attendance at such a module would enhance the quality of research in psoriatic disease.

COVID-19 olfactory dysfunction: associations between coping, quality of life, and mental health.

BACKGROUND: Persistent olfactory dysfunction (OD) is a common symptom following SARS-CoV-2 infection that can greatly impact quality of life (QoL). Because coping strategies have been shown to moderate the effect of disease symptoms on functional and affective outcomes, this study aims to determine whether specific coping strategies are associated with and moderate QoL outcomes. METHODOLOGY: Participants with prior SARS-CoV-2 infection underwent psychophysical olfactory testing with Sniffin' Sticks and completed questionnaires to elicit subjective olfactory function, coping strategies, olfactory-specific QoL, general QoL, and mental health. RESULTS: There were 93 participants included in the study. Olfactory specific QoL scores were significantly worse among individuals with subjective and psychophysically measured OD compared to those with subjective and psychophysically confirmed normosmia. Olfactory-specific QoL, general QoL, and anxiety symptom scores were positively correlated with avoidant and disengagement coping among individuals with subjective and psychophysically measured OD. Depression symptom scores were positively correlated with avoidant and disengagement coping and negatively correlated with approach and engagement coping. There were no significant moderating effects on the association between olfactory performance and QoL or mental health screening assessment. CONCLUSIONS: Approach and engagement coping mechanisms are associated with improved depression, whereas avoidant and disengagement coping tracks with worse QoL and mental health screening assessment, offering an opportunity to counsel patients accordingly.

Single-Cell Synchro-Subtractive-Additive Nanoscale Surgery with Femtosecond Lasers.

Herein, an in situ "synchro-subtractive-additive" technique of femtosecond laser single-cell surgery (FLSS) is presented to address the inadequacies of existing surgical methods for single-cell manipulation. This process is enabled by synchronized nanoscale three-dimensional (3D) subtractive and additive manufacturing with ultrahigh precision on various parts of the cells, in that the precise removal and modification of a single-cell structure are realized by nonthermal ablation, with synchronously ultrafast solidification of the specially designed hydrogel by two photopolymerizations. FLSS is a minimally invasive technique with a post-operative survival rate of 70% and stable proliferation. It opens avenues for bottom-up synthetic biology, offering new methods for artificially synthesizing organelle-like 3D structures and modifying the physiological activities of cells.

Lactoperoxidase catalytically oxidize hydrogen sulfide via intermediate formation of sulfheme derivatives.

The biological chemistry of hydrogen sulfide (H2S) with physiologically important heme proteins is in the focus of redox biology research. In this study, we investigated the interactions of lactoperoxidase (LPO) with H2S in the presence and absence of molecular dioxygen (O2) or hydrogen peroxide (H2O2). Under anaerobic conditions, native LPO forms no heme-H2S complex upon sulfide exposure. However, under aerobic conditions or in the presence of H2O2 the formation of both ferrous and ferric sulfheme (sulfLPO) derivatives was observed based on the appearances of their characteristic optical absorptions at 638 nm and 727 nm, respectively. Interestingly, we demonstrate that LPO can catalytically oxidize H2S by H2O2 via intermediate formation of relatively short-lived ferrous and ferric sulfLPO derivatives. Pilot product analyses suggested that the turnover process generates oxidized sulfide species, which include sulfate S O 4 2 - and inorganic polysulfides ( H S x - ; x = 2-5). These results indicated that H2S can serve as a non-classical LPO substrate by inducing a reversible sulfheme-like modification of the heme porphyrin ring during turnover. Furthermore, electron paramagnetic resonance data suggest that H2S can act as a scavenger of H2O2 in the presence of LPO without detectable formation of any carbon-centered protein radical species, suggesting that H2S might be capable of protecting the enzyme from radical-mediated damage. We propose possible mechanisms, which explain our results as well as contrasting observations with other heme proteins, where either no sulfheme formation was observed or the generation of sulfheme derivatives provided a dead end for enzyme functions.

High-Throughput SFC-MS/MS Method to Measure EPSA and Predict Human Permeability.

Permeability is a key factor driving the absorption of orally administered drugs. In early discovery, the efficient evaluation of permeability, particularly for compounds violating Lipinski's Rule of 5, remains challenging. Addressing this, we established a high-throughput method to measure the experimental polar surface area (HT-EPSA) as an in vitro surrogate to measure permeability. Compared to earlier methods, HT-EPSA significantly reduces data acquisition time with enhanced sensitivity, selectivity, and data quality. In the effort of translating EPSA to human in vitro and in vivo passive permeability, we demonstrated the application of EPSA for predicting Caco-2 cell and human intestinal permeability, showing improvements over topological polar surface area and the parallel artificial membrane permeability assay for rank-ordering permeability in a proteolysis targeting chimera case study. The HT-EPSA method is expected to be highly beneficial in guiding early stage compound rank-ordering, faster decision-making, and in predicting in vitro and/or in vivo human intestinal permeability.

Predicting Collision Cross-Section Values for Small Molecules through Chemical Class-Based Multimodal Graph Attention Network.

Libraries of collision cross-section (CCS) values have the potential to facilitate compound identification in metabolomics. Although computational methods provide an opportunity to increase library size rapidly, accurate prediction of CCS values remains challenging due to the structural diversity of small molecules. Here, we developed a machine learning (ML) model that integrates graph attention networks and multimodal molecular representations to predict CCS values on the basis of chemical class. Our approach, referred to as MGAT-CCS, had superior performance in comparison to other ML models in CCS prediction. MGAT-CCS achieved a median relative error of 0.47%/1.14% (positive/negative mode) and 1.40%/1.63% (positive/negative mode) for lipids and metabolites, respectively. When MGAT-CCS was applied to real-world metabolomics data, it reduced the number of false metabolite candidates by roughly 25% across multiple sample types ranging from plasma and urine to cells. To facilitate its application, we developed a user-friendly stand-alone web server for MGAT-CCS that is freely available at https://mgat-ccs-web.onrender.com. This work represents a step forward in predicting CCS values and can potentially facilitate the identification of small molecules when using ion mobility spectrometry coupled with mass spectrometry.

Multiparametric MRI-based radiomic models for early prediction of response to neoadjuvant systemic therapy in triple-negative breast cancer.

Triple-negative breast cancer (TNBC) is often treated with neoadjuvant systemic therapy (NAST). We investigated if radiomic models based on multiparametric Magnetic Resonance Imaging (MRI) obtained early during NAST predict pathologic complete response (pCR). We included 163 patients with stage I-III TNBC with multiparametric MRI at baseline and after 2 (C2) and 4 cycles of NAST. Seventy-eight patients (48%) had pCR, and 85 (52%) had non-pCR. Thirty-six multivariate models combining radiomic features from dynamic contrast-enhanced MRI and diffusion-weighted imaging had an area under the receiver operating characteristics curve (AUC) > 0.7. The top-performing model combined 35 radiomic features of relative difference between C2 and baseline; had an AUC = 0.905 in the training and AUC = 0.802 in the testing set. There was high inter-reader agreement and very similar AUC values of the pCR prediction models for the 2 readers. Our data supports multiparametric MRI-based radiomic models for early prediction of NAST response in TNBC.

Self-help groups and opioid use disorder treatment: An investigation using a machine learning-assisted robust causal inference framework.

OBJECTIVES: This study investigates the impact of participation in self-help groups on treatment completion among individuals undergoing medication for opioid use disorder (MOUD) treatment. Given the suboptimal adherence and retention rates for MOUD, this research seeks to examine the association between treatment completion and patient-level factors. Specifically, we evaluated the causal relationship between self-help group participation and treatment completion for patients undergoing MOUD. METHODS: We used the Substance Abuse and Mental Health Services Administration's (SAMHSA) Treatment Episode Data Set: Discharges (TEDS-D) from 2015 to 2019. The data are filtered by the patient's opioid use history, demographics, treatment modality, and other relevant information. In this observational study, machine learning models (Lasso Regression, Decision Trees, Random Forest, and XGBoost) were developed to predict treatment completion. Outcome Adaptive Elastic Net (OAENet) was used to select confounders and outcome predictors, and the robust McNemars test was used to evaluate the causal relationship between self-help group participation and MOUD treatment completion. RESULTS: The machine-learning models showed a strong association between participation in self-help groups and treatment completion. Our causal analysis demonstrated an average treatment effect on treated (ATT) of 0.260 and a p-value < 0.0001 for the robust McNemars test. CONCLUSIONS: Our study demonstrates the importance of participation in self-help groups for MOUD treatment recipients. We found that participation in MOUD along with self-help groups caused higher chances of treatment completion than MOUD alone. This suggests that policymakers should consider further integrating self-help groups into the treatment for OUD to improve the adherence and completion rate.

Quantitative Measurement of Rate of Targeted Protein Degradation.

Targeted protein degradation (TPD) is a therapeutic approach that leverages the cell's natural machinery to degrade targets instead of inhibiting them. This is accomplished by using mono- or bifunctional small molecules designed to induce the proximity of target proteins and E3 ubiquitin ligases, leading to ubiquitination and subsequent proteasome-dependent degradation of the target. One of the most significant attributes of the TPD approach is its proposed catalytic mechanism of action, which permits substoichiometric exposure to achieve the desired pharmacological effects. However, apart from one in vitro study, studies supporting the catalytic mechanism of degraders are largely inferred based on potency. A more comprehensive understanding of the degrader catalytic mechanism of action can help aspects of compound development. To address this knowledge gap, we developed a workflow for the quantitative measurement of the catalytic rate of degraders in cells. Comparing a selective and promiscuous BTK degrader, we demonstrate that both compounds function as efficient catalysts of BTK degradation, with the promiscuous degrader exhibiting faster rates due to its ability to induce more favorable ternary complexes. By leveraging computational modeling, we show that the catalytic rate is highly dynamic as the target is depleted from cells. Further investigation of the promiscuous kinase degrader revealed that the catalytic rate is a better predictor of optimal degrader activity toward a specific target compared to degradation magnitude alone. In summary, we present a versatile method for mapping the catalytic activity of any degrader for TPD in cells.

Ag Intercalation in Layered Cs3Bi2Br9 Perovskite for Enhanced Light Emission with Bound Interlayer Excitons.

Cesium bismuth bromide (CBB) has garnered considerable attention as a vacancy-ordered layered perovskite with notable optoelectronic applications. However, its use as a light source has been limited due to its weak photoluminescence (PL). Here, we demonstrate metal intercalation as a novel approach to engineer the room-temperature PL of CBB using experimental and computational methods. Ag, when introduced into CBB, occupies vacant sites in the spacer region, forming octahedral coordination with surrounding Br anions. First-principles density functional theory calculations reveal that intercalated Ag represents the most energetically stable Ag species compared to other potential forms, such as Ag substituting Bi. The intercalated Ag forms a strong polaronic trap state close to the conduction band minimum and quickly captures photoexcited electrons with holes remaining in CBB layers, leading to the formation of a bound interlayer exciton, or BIE. The radiative recombination of this BIE exhibits bright room-temperature PL at 600 nm and a decay time of 38.6 ns, 35 times greater than that of free excitons, originating from the spatial separation of photocarriers by half a unit cell separation distance. The BIE as a new form of interlayer exciton is expected to inspire new research directions for vacancy-ordered perovskites.

Discovery and Identification of Three Homocysteine Metabolites by Chemical Derivatization and Mass Spectrometry Fragmentation.

Discovery and identification of a new endogenous metabolite are typically hindered by requirements of large sample volumes and multistage purifications to guide synthesis of the standard. Presented here is a metabolomics platform that uses chemical tagging and tandem mass spectrometry to determine structure, direct synthesis, and confirm identity. Three new homocysteine metabolites are reported: N-succinyl homocysteine, 2-methyl-1,3-thiazinane-4-carboxylic acid (MTCA), and homolanthinone.

Efficacy and safety of once weekly selinexor 40 mg versus 60 mg with pomalidomide and dexamethasone in relapsed and/or refractory multiple myeloma.

Objective: To identify the optimal dose of selinexor in combination with pomalidomide and dexamethasone (SPd). Methods: An analysis of efficacy and safety of 2 once-weekly selinexor regimens (60 mg and 40 mg) with pomalidomide and dexamethasone (SPd-60 and SPd-40, respectively) given to patients with relapsed/refractory multiple myeloma (RRMM) in the STOMP (NCT02343042) and XPORT-MM-028 (NCT04414475) trials. Results: Twenty-eight patients (60.7% males, median age 67.5 years) and 20 patients (35.0% males, median age 65.5 years) were analyzed in the SPd-40 and SPd-60 cohorts, respectively. Overall response rate was 50% (95% confidence interval [CI] 30.6-69.4%) and 65% (95% CI 40.8-84.6%), respectively. Very good partial response or better was reported in 28.6% (95% CI 13.2-48.7%) and 30.0% (95% CI 11.9-54.3%) of patients, respectively. Among 27 responders in both cohorts, the 12-month sustained response rate was 83.3% (95% CI 64.7-100.0%) for SPd-40 and 28.1% (95% CI 8.9-88.8%) for SPd-60. Median progression-free survival was 18.4 months (95% CI 6.5 months, not evaluable [NE]) and 9.5 months (95% CI 7.6 months-NE) for SPd-40 and SPd-60, respectively. Twenty-four-month survival rates were 64.2% (95% CI 47.7-86.3%) for SPd-40 and 51.1% (95% CI 29.9-87.5%) for SPd-60. Treatment-emergent adverse events (TEAEs) included neutropenia (all grades: SPd-40 64.3% versus SPd-60 75.0%), anemia (46.4% versus 65.0%), thrombocytopenia (42.9% versus 45.0%), fatigue (46.4% versus 75.0%), nausea (32.1% versus 70.0%) and diarrhea (28.6% versus 35.0%). Conclusion: The all-oral combination of SPd exhibited preliminary signs of efficacy and was generally tolerable in patients with RRMM. The overall risk-benefit profile favored the SPd-40 regimen.

Receptor-Targeted Peptide Conjugates Based on Diphosphines Enable Preparation of 99mTc and 188Re Theranostic Agents for Prostate Cancer.

Benchtop 99Mo/99mTc and 188W/188Re generators enable economical production of molecular theranostic 99mTc and 188Re radiopharmaceuticals, provided that simple, kit-based chemistry exists to radiolabel targeting vectors with these radionuclides. We have previously described a diphosphine platform that efficiently incorporates 99mTc into receptor-targeted peptides. Here, we report its application to label a prostate-specific membrane antigen (PSMA)-targeted peptide with 99mTc and 188Re for diagnostic imaging and systemic radiotherapy of prostate cancer. Methods: Two diphosphine-dipeptide bioconjugates, DP1-PSMAt and DP2-PSMAt, were formulated into kits for radiolabeling with 99mTc and 188Re. The resulting radiotracers were studied in vitro, in prostate cancer cells, and in vivo in mouse xenograft models, to assess similarity of uptake and biodistribution for each 99mTc/188Re pair of agents. Results: Both DP1-PSMAt and DP2-PSMAt could be efficiently radiolabeled with 99mTc and 188Re using kit-based methods to furnish the isostructural compounds M-DP1-PSMAt and M-DP2-PSMAt (M = [99mTc]Tc, [188Re]Re). All 99mTc/188Re radiotracers demonstrated specific uptake in PSMA-expressing prostate cancer cells, with negligible uptake in prostate cancer cells that did not express PSMA or in which PSMA uptake was blocked. M-DP1-PSMAt and M-DP2-PSMAt also exhibited high tumor uptake (18-30 percentage injected dose per gram at 2 h after injection), low retention in nontarget organs, fast blood clearance, and excretion predominantly via a renal pathway. Importantly, each pair of 99mTc/188Re radiotracers showed near-identical biologic behavior in these experiments. Conclusion: We have prepared and developed novel pairs of isostructural PSMA-targeting 99mTc/188Re theranostic agents. These generator-based theranostic agents have potential to provide access to the benefits of PSMA-targeted diagnostic imaging and systemic radiotherapy in health care settings that do not routinely have access to either reactor-produced 177Lu radiopharmaceuticals or PET/CT infrastructure.

Development of novel tools for dissection of central versus peripheral dopamine D2-like receptor signaling in dysglycemia.

Dopamine (DA) D2-like receptors in both the central nervous system (CNS) and the periphery are key modulators of metabolism. Moreover, disruption of D2-like receptor signaling is implicated in dysglycemia. Yet, the respective metabolic contributions of CNS versus peripheral D2-like receptors including D2 (D2R) and D3 (D3R) receptors remain poorly understood. To address this, we developed new pharmacological tools, D2-like receptor agonists with diminished and delayed blood-brain barrier capability, to selectively manipulate D2R/D3R signaling in the periphery. We designated bromocriptine methiodide (BrMeI), a quaternary methiodide analogue of D2R/D3R agonist and diabetes drug bromocriptine, as our lead compound based on preservation of D2R/D3R binding and functional efficacy. We then used BrMeI and unmodified bromocriptine to dissect relative contributions of CNS versus peripheral D2R/D3R signaling in treating dysglycemia. Systemic administration of bromocriptine, with unrestricted access to CNS and peripheral targets, significantly improved both insulin sensitivity and glucose tolerance in obese, dysglycemic mice in vivo. In contrast, metabolic improvements were attenuated when access to bromocriptine was restricted either to the CNS through intracerebroventricular administration or delayed access to the CNS via BrMeI. Our findings demonstrate that the coordinated actions of both CNS and peripheral D2-like receptors are required for correcting dysglycemia. Ultimately, the development of a first-generation of drugs designed to selectively target the periphery provides a blueprint for dissecting mechanisms of central versus peripheral DA signaling and paves the way for novel strategies to treat dysglycemia.

Loss of mitochondrial pyruvate transport initiates cardiac glycogen accumulation and heart failure.

Background: Heart failure involves metabolic alterations including increased glycolysis despite unchanged or decreased glucose oxidation. The mitochondrial pyruvate carrier (MPC) regulates pyruvate entry into the mitochondrial matrix, and cardiac deletion of the MPC in mice causes heart failure. How MPC deletion results in heart failure is unknown. Methods: We performed targeted metabolomics and isotope tracing in wildtype (fl/fl) and cardiac-specific Mpc2-/- (CS-Mpc2-/-) hearts after in vivo injection of U-13C-glucose. Cardiac glycogen was assessed biochemically and by transmission electron microscopy. Cardiac uptake of 2-deoxyglucose was measured and western blotting performed to analyze insulin signaling and enzymatic regulators of glycogen synthesis and degradation. Isotope tracing and glycogen analysis was also performed in hearts from mice fed either low-fat diet or a ketogenic diet previously shown to reverse the CS-Mpc2-/- heart failure. Cardiac glycogen was also assessed in mice infused with angiotensin-II that were fed low-fat or ketogenic diet. Results: Failing CS-Mpc2-/- hearts contained normal levels of ATP and phosphocreatine, yet these hearts displayed increased enrichment from U-13C-glucose and increased glycolytic metabolite pool sizes. 13C enrichment and pool size was also increased for the glycogen intermediate UDP-glucose, as well as increased enrichment of the glycogen pool. Glycogen levels were increased ~6-fold in the failing CS-Mpc2-/- hearts, and glycogen granules were easily detected by electron microscopy. This increased glycogen synthesis occurred despite enhanced inhibitory phosphorylation of glycogen synthase and reduced expression of glycogenin-1. In young, non-failing CS-Mpc2-/- hearts, increased glycolytic 13C enrichment occurred, but glycogen levels remained low and unchanged compared to fl/fl hearts. Feeding a ketogenic diet to CS-Mpc2-/- mice reversed the heart failure and normalized the cardiac glycogen and glycolytic metabolite accumulation. Cardiac glycogen levels were also elevated in mice infused with angiotensin-II, and both the cardiac hypertrophy and glycogen levels were improved by ketogenic diet. Conclusions: Our results indicate that loss of MPC in the heart causes glycogen accumulation and heart failure, while a ketogenic diet can reverse both the glycogen accumulation and heart failure. We conclude that maintaining mitochondrial pyruvate import and metabolism is critical for the heart, unless cardiac pyruvate metabolism is reduced by consumption of a ketogenic diet.

Implementing circulating tumor DNA as a prognostic biomarker in resectable non-small cell lung cancer.

Systemic treatment of resectable non-small cell lung cancer (NSCLC) is evolving with emerging neoadjuvant, perioperative, and adjuvant immunotherapy approaches. Circulating tumor DNA (ctDNA) detection at clinical diagnosis, during neoadjuvant therapy, or after resection may discern high-risk patients who might benefit from therapy escalation or switch. This Review summarizes translational implications of data supporting ctDNA-based risk determination in NSCLC and outstanding questions regarding ctDNA validity/utility as a prognostic biomarker. We discuss emerging ctDNA capabilities to refine clinical tumor-node-metastasis (TNM) staging in lung adenocarcinoma, ctDNA dynamics during neoadjuvant therapy for identifying patients deriving suboptimal benefit, and postoperative molecular residual disease (MRD) detection to escalate systemic therapy. Considering differential relapse characteristics in landmark MRD-negative/MRD-positive patients, we propose how ctDNA might integrate with pathological response data for optimal postoperative risk stratification.

SMIM1 absence is associated with reduced energy expenditure and excess weight.

BACKGROUND: Obesity rates have nearly tripled in the past 50 years, and by 2030 more than 1 billion individuals worldwide are projected to be obese. This creates a significant economic strain due to the associated non-communicable diseases. The root cause is an energy expenditure imbalance, owing to an interplay of lifestyle, environmental, and genetic factors. Obesity has a polygenic genetic architecture; however, single genetic variants with large effect size are etiological in a minority of cases. These variants allowed the discovery of novel genes and biology relevant to weight regulation and ultimately led to the development of novel specific treatments. METHODS: We used a case-control approach to determine metabolic differences between individuals homozygous for a loss-of-function genetic variant in the small integral membrane protein 1 (SMIM1) and the general population, leveraging data from five cohorts. Metabolic characterization of SMIM1-/- individuals was performed using plasma biochemistry, calorimetric chamber, and DXA scan. FINDINGS: We found that individuals homozygous for a loss-of-function genetic variant in SMIM1 gene, underlying the blood group Vel, display excess body weight, dyslipidemia, altered leptin to adiponectin ratio, increased liver enzymes, and lower thyroid hormone levels. This was accompanied by a reduction in resting energy expenditure. CONCLUSION: This research identified a novel genetic predisposition to being overweight or obese. It highlights the need to investigate the genetic causes of obesity to select the most appropriate treatment given the large cost disparity between them. FUNDING: This work was funded by the National Institute of Health Research, British Heart Foundation, and NHS Blood and Transplant.

From Llama to Nanobody: A Streamlined Workflow for the Generation of Functionalised VHHs.

Nanobodies are recombinant antigen-specific single domain antibodies (VHHs) derived from the heavy chain-only subset of camelid immunoglobulins. Their small molecular size, facile expression, high affinity, and stability have combined to make them unique targeting reagents with numerous applications in the biomedical sciences. From our work in producing nanobodies to over sixty different proteins, we present a standardised workflow for nanobody discovery from llama immunisation, library building, panning, and small-scale expression for prioritisation of binding clones. In addition, we introduce our suites of mammalian and bacterial vectors, which can be used to functionalise selected nanobodies for various applications such as in imaging and purification. Key features • Standardise the process of building nanobody libraries and finding nanobody binders so that it can be repeated in any lab with reasonable equipment. • Introduce two suites of vectors to functionalise nanobodies for production in either bacterial or mammalian cells.

An explainable machine learning framework for predicting the risk of buprenorphine treatment discontinuation for opioid use disorder among commercially insured individuals.

OBJECTIVES: Buprenorphine is an effective evidence-based medication for opioid use disorder (OUD). Yet premature discontinuation undermines treatment effectiveness, increasing the risk of mortality and overdose. We developed and evaluated a machine learning (ML) framework for predicting buprenorphine care discontinuity within 12 months following treatment initiation. METHODS: This retrospective study used United States (US) 2018-2021 MarketScan commercial claims data of insured individuals aged 18-64 who initiated buprenorphine between July 2018 and December 2020 with no buprenorphine prescriptions in the previous six months. We measured buprenorphine prescription discontinuation gaps of ≥30 days within 12 months of initiating treatment. We developed predictive models employing logistic regression, decision tree classifier, random forest, extreme gradient boosting, Adaboost, and random forest-extreme gradient boosting ensemble. We applied recursive feature elimination with cross-validation to reduce dimensionality and identify the most predictive features while maintaining model robustness. For model validation, we used several statistics to evaluate performance, such as C-statistics and precision-recall curves. We focused on two distinct treatment stages: at the time of treatment initiation and one and three months after treatment initiation. We employed SHapley Additive exPlanations (SHAP) analysis that helped us explain the contributions of different features in predicting buprenorphine discontinuation. We stratified patients into risk subgroups based on their predicted likelihood of treatment discontinuation, dividing them into decile subgroups. Additionally, we used a calibration plot to analyze the reliability of the models. RESULTS: A total of 30,373 patients initiated buprenorphine and 14.98% (4551) discontinued treatment. C-statistic varied between 0.56 and 0.76 for the first-stage models including patient-level demographic and clinical variables. Inclusion of proportion of days covered (PDC) measured after one month and three months following treatment initiation significantly increased the models' discriminative power (C-statistics: 0.60 to 0.82). Random forest (C-statistics: 0.76, 0.79 and 0.82 with baseline predictors, one-month PDC and three-months PDC, respectively) outperformed other ML models in discriminative performance in all stages (C-statistics: 0.56 to 0.77). Most influential risk factors of discontinuation included early stage medication adherence, age, and initial days of supply. CONCLUSION: ML algorithms demonstrated a good discriminative power in identifying patients at higher risk of buprenorphine care discontinuity. The proposed framework may help healthcare providers optimize treatment strategies and deliver targeted interventions to improve buprenorphine care continuity.