Machine Learning-Based Quantification of Patient Factors Impacting Remission in Patients With Ulcerative Colitis: Insights from Etrolizumab Phase III Clinical Trials.

Etrolizumab, an investigational anti-ß7 integrin monoclonal antibody, has undergone evaluation for safety and efficacy in phase III clinical trials on patients with moderate to severe ulcerative colitis (UC). Etrolizumab was terminated because mixed efficacy results were shown in the induction and maintenance phase in patients with UC. In this post hoc analysis, we characterized the impact of explanatory variables on the probability of remission using XGBoost machine learning (ML) models alongside with the SHapley Additive exPlanations framework for explainability. We used patient-level data encompassing demographics, physiology, disease history, clinical questionnaires, histology, serum biomarkers, and etrolizumab drug exposure to develop ML models aimed at predicting remission. Baseline covariates and early etrolizumab exposure at week 4 in the induction phase were utilized to develop an induction ML model, whereas covariates from the end of the induction phase and early etrolizumab exposure at week 4 in the maintenance phase were used to develop a maintenance ML model. Both the induction and maintenance ML models exhibited good predictive performance, achieving an area under the receiver operating characteristic curve (AUROC) of 0.74 ± 0.03 and 0.75 ± 0.06 (mean ± SD), respectively. Compared with placebo, the highest tertile of etrolizumab exposure contributed to 15.0% (95% confidence interval (CI): 9.7-19.9) and 17.0% (95% CI: 8.1-26.4) increases in remission probability in the induction and maintenance phases, respectively. Additionally, the key covariates that predicted remission were CRP, MAdCAM-1, and stool frequency for the induction phase and white blood cells, fecal calprotectin and age for the maintenance phase. These findings hold significant implications for establishing stratification factors in the design of future clinical trials.

Machine Learning for Exposure-Response Analysis: Methodological Considerations and Confirmation of Their Importance via Computational Experimentations.

Exposure-response (E-R) is a key aspect of pharmacometrics analysis that supports drug dose selection. Currently, there is a lack of understanding of the technical considerations necessary for drawing unbiased estimates from data. Due to recent advances in machine learning (ML) explainability methods, ML has garnered significant interest for causal inference. To this end, we used simulated datasets with known E-R "ground truth" to generate a set of good practices for the development of ML models required to avoid introducing biases when performing causal inference. These practices include the use of causal diagrams to enable the careful consideration of model variables by which to obtain desired E-R relationship insights, keeping a strict separation of data for model-training and for inference generation to avoid biases, hyperparameter tuning to improve the reliability of models, and estimating proper confidence intervals around inferences using a bootstrap sampling with replacement strategy. We computationally confirm the benefits of the proposed ML workflow by using a simulated dataset with nonlinear and non-monotonic exposure-response relationships.

Risk Factors of Hyperglycemia After Treatment With the AKT Inhibitor Ipatasertib in the Prostate Cancer Setting: A Machine Learning-Based Investigation.

PURPOSE: Hyperglycemia is a major adverse event of phosphatidylinositol 3-kinase/AKT inhibitor class of cancer therapeutics. Machine learning (ML) methodologies can identify and highlight how explanatory variables affect hyperglycemia risk. METHODS: Using data from clinical trials of the AKT inhibitor ipatasertib (IPAT) in the metastatic castrate-resistant prostate cancer setting, we trained an XGBoost ML model to predict the incidence of grade ≥2 hyperglycemia (HGLY ≥ 2). Of the 1,364 patients included in our analysis, 19.4% (n = 265) of patients had HGLY ≥2 events with a median time of first onset of 28 days (range, 0-753 days), and 30.0% (n = 221) of patients on an IPAT regimen had at least one HGLY ≥2 event compared with 7.0% (n = 44) of patients on placebo. RESULTS: An 11-variable XGBoost model predicted HGLY ≥2 events well with an AUROC of 0.83 ± 0.02 (mean ± standard deviation). Using SHapley Additive exPlanations analysis, we found IPAT exposure and baseline HbA1c levels to be the strongest predictors of HGLY ≥2, with additional predictivity of baseline measurements of fasting glucose, magnesium, and high-density lipoproteins. CONCLUSION: The findings support using patients' prediabetic status as a key factor for hyperglycemia monitoring and/or trial exclusion criteria. Additionally, the model and relationships between explanatory variables and HGLY ≥2 described herein can help identify patients at high risk for hyperglycemia and develop rational risk mitigation strategies.

Timescales of Evidence Evaluation for Decision Making and Associated Confidence Judgments Are Adapted to Task Demands.

Decision making often involves choosing actions based on relevant evidence. This can benefit from focussing evidence evaluation on the timescale of greatest relevance based on the situation. Here, we use an auditory change detection task to determine how people adjust their timescale of evidence evaluation depending on task demands for detecting changes in their environment and assessing their internal confidence in those decisions. We confirm previous results that people adopt shorter timescales of evidence evaluation for detecting changes in contexts with shorter signal durations, while bolstering those results with model-free analyses not previously used and extending the results to the auditory domain. We also extend these results to show that in contexts with shorter signal durations, people also adopt correspondingly shorter timescales of evidence evaluation for assessing confidence in their decision about detecting a change. These results provide important insights into adaptability and flexible control of evidence evaluation for decision making.

Flexibility of Timescales of Evidence Evaluation for Decision Making.

To understand the neural mechanisms that support decision making, it is critical to characterize the timescale of evidence evaluation. Recent work has shown that subjects can adaptively adjust the timescale of evidence evaluation across blocks of trials depending on context [1]. However, it's currently unknown if adjustments to evidence evaluation occur online during deliberations based on a single stream of evidence. To examine this question, we employed a change-detection task in which subjects report their level of confidence in judging whether there has been a change in a stochastic auditory stimulus. Using a combination of psychophysical reverse correlation analyses and single-trial behavioral modeling, we compared the time period over which sensory information has leverage on detection report choices versus confidence. We demonstrate that the length of this period differs on separate sets of trials based on what's being reported. Surprisingly, confidence judgments on trials with no detection report are influenced by evidence occurring earlier than the time period of influence for detection reports. Our findings call into question models of decision formation involving static parameters that yield a singular timescale of evidence evaluation and instead suggest that the brain represents and utilizes multiple timescales of evidence evaluation during deliberation.

Modeling Fast-scan Cyclic Voltammetry Data from Electrically Stimulated Dopamine Neurotransmission Data Using QNsim1.0.

Central dopaminergic (DAergic) pathways have an important role in a wide range of functions, such as attention, motivation, and movement. Dopamine (DA) is implicated in diseases and disorders including attention deficit hyperactivity disorder, Parkinson's disease, and traumatic brain injury. Thus, DA neurotransmission and the methods to study it are of intense scientific interest. In vivo fast-scan cyclic voltammetry (FSCV) is a method that allows for selectively monitoring DA concentration changes with fine temporal and spatial resolution. This technique is commonly used in conjunction with electrical stimulations of ascending DAergic pathways to control the impulse flow of dopamine neurotransmission. Although the stimulated DA neurotransmission paradigm can produce robust DA responses with clear morphologies, making them amenable for kinetic analysis, there is still much debate on how to interpret the responses in terms of their DA release and clearance components. To address this concern, a quantitative neurobiological (QN) framework of stimulated DA neurotransmission was recently developed to realistically model the dynamics of DA release and reuptake over the course of a stimulated DA response. The foundations of this model are based on experimental data from stimulated DA neurotransmission and on principles of neurotransmission adopted from various lines of research. The QN model implements 12 parameters related to stimulated DA release and reuptake dynamics to model DA responses. This work describes how to simulate DA responses using QNsim1.0 and also details principles that have been implemented to systematically discern alterations in the stimulated dopamine release and reuptake dynamics.

Ventricular fibrillation cardiac arrest produces a chronic striatal hyperdopaminergic state that is worsened by methylphenidate treatment.

Cardiac arrest survival rates have improved with modern resuscitation techniques, but many survivors experience impairments associated with hypoxic-ischemic brain injury (HIBI). Currently, little is understood about chronic changes in striatal dopamine (DA) systems after HIBI. Given the common empiric clinical use of DA enhancing agents in neurorehabilitation, investigation evaluating dopaminergic alterations after cardiac arrest (CA) is necessary to optimize rehabilitation approaches. We hypothesized that striatal DA neurotransmission would be altered chronically after ventricular fibrillation cardiac arrest (VF-CA). Fast-scan cyclic voltammetry was used with median forebrain bundle (MFB) maximal electrical stimulations (60Hz, 10s) in rats to characterize presynaptic components of DA neurotransmission in the dorsal striatum (D-Str) and nucleus accumbens 14 days after a 5-min VF-CA when compared to Sham or Naïve. VF-CA increased D-Str-evoked overflow [DA], total [DA] released, and initial DA release rate versus controls, despite also increasing maximal velocity of DA reuptake (Vmax ). Methylphenidate (10 mg/kg), a DA transporter inhibitor, was administered to VF-CA and Shams after establishing a baseline, pre-drug 60 Hz, 5 s stimulation response. Methylphenidate increased initial evoked overflow [DA] more-so in VF-CA versus Sham and reduced D-Str Vmax in VF-CA but not Shams; these findings are consistent with upregulated striatal DA transporter in VF-CA versus Sham. Our work demonstrates that 5-min VF-CA increases electrically stimulated DA release with concomitant upregulation of DA reuptake 2 weeks after brief VF-CA insult. Future work should elucidate how CA insult duration, time after insult, and insult type influence striatal DA neurotransmission and related cognitive and motor functions.

Fast-scan cyclic voltammetry demonstrates that L-DOPA produces dose-dependent, regionally selective bimodal effects on striatal dopamine kinetics in vivo.

Parkinson's disease (PD) is a debilitating condition that is caused by a relatively specific degeneration of dopaminergic (DAergic) neurons of the substantia nigra pars compacta. L-DOPA was introduced as a viable treatment option for PD over 40 years ago and still remains the most common and effective therapy for PD. Though the effects of L-DOPA to augment striatal DA production are well known, little is actually known about how L-DOPA alters the kinetics of DA neurotransmission that contribute to its beneficial and adverse effects. In this study, we examined the effects of L-DOPA administration (50 mg/kg carbidopa + 0, 100, and 250 mg/kg L-DOPA) on regional electrically stimulated DA response kinetics using fast-scan cyclic voltammetry in anesthetized rats. We demonstrate that L-DOPA enhances DA release in both the dorsal striatum (D-STR) and nucleus accumbens (NAc), but surprisingly causes a delayed inhibition of release in the D-STR. In both regions, L-DOPA progressively attenuated reuptake kinetics, predominantly through a decrease in Vmax . These findings have important implications on understanding the pharmacodynamics of L-DOPA, which may be informative for understanding its therapeutic effects and also common side effects like L-DOPA-induced dyskinesias (LID). L-DOPA is commonly used to treat Parkinsonian symptoms, but little is known about how it affects presynaptic DA neurotransmission. Using in vivo fast-scan cyclic voltammetry, we show L-DOPA inhibits DA reuptake in a region-specific and dose-dependent manner, and L-DOPA has paradoxical effects on release. These findings may be important when considering mechanisms for L-DOPA's therapeutic benefits and adverse side-effects.

Neurobiological model of stimulated dopamine neurotransmission to interpret fast-scan cyclic voltammetry data.

Fast-scan cyclic voltammetry (FSCV) is an electrochemical method that can assess real-time in vivo dopamine (DA) concentration changes to study the kinetics of DA neurotransmission. Electrical stimulation of dopaminergic (DAergic) pathways can elicit FSCV DA responses that largely reflect a balance of DA release and reuptake. Interpretation of these evoked DA responses requires a framework to discern the contribution of DA release and reuptake. The current, widely implemented interpretive framework for doing so is the Michaelis-Menten (M-M) model, which is grounded on two assumptions- (1) DA release rate is constant during stimulation, and (2) DA reuptake occurs through dopamine transporters (DAT) in a manner consistent with M-M enzyme kinetics. Though the M-M model can simulate evoked DA responses that rise convexly, response types that predominate in the ventral striatum, the M-M model cannot simulate dorsal striatal responses that rise concavely. Based on current neurotransmission principles and experimental FSCV data, we developed a novel, quantitative, neurobiological framework to interpret DA responses that assumes DA release decreases exponentially during stimulation and continues post-stimulation at a diminishing rate. Our model also incorporates dynamic M-M kinetics to describe DA reuptake as a process of decreasing reuptake efficiency. We demonstrate that this quantitative, neurobiological model is an extension of the traditional M-M model that can simulate heterogeneous regional DA responses following manipulation of stimulation duration, frequency, and DA pharmacology. The proposed model can advance our interpretive framework for future in vivo FSCV studies examining regional DA kinetics and their alteration by disease and DA pharmacology.

Controlled cortical impact injury influences methylphenidate-induced changes in striatal dopamine neurotransmission.

Traumatic brain injury features deficits are often ameliorated by dopamine (DA) agonists. We have previously shown deficits in striatal DA neurotransmission using fast scan cyclic voltammetry after controlled cortical impact (CCI) injury that are reversed after daily treatment with the DA uptake inhibitor methylphenidate (MPH). The goal of this study was to determine how a single dose of MPH (5 mg/kg) induces changes in basal DA and metabolite levels and with electrically evoked overflow (EO) DA in the striatum of CCI rats. MPH-induced changes in EO DA after a 2-week daily pre-treatment regime with MPH was also assessed. There were no baseline differences in basal DA or metabolite levels. MPH injection significantly increased basal [DA] output in dialysates for control but not injured rats. Also, MPH injection increased striatal peak EO [DA] to a lesser degree in CCI (176% of baseline) versus control rats (233% of baseline). However, daily pre-treatment with MPH resulted in CCI rats having a comparable increase in EO [DA] after MPH injection when compared with controls. The findings further support the concept that daily MPH therapy restores striatal DA neurotransmission after CCI.

Chronic methylphenidate treatment enhances striatal dopamine neurotransmission after experimental traumatic brain injury.

Traumatic brain injury (TBI) results in functional deficits that often are effectively treated clinically with the neurostimulant, methylphenidate (MPH). We hypothesized that daily MPH administration would reverse striatal neurotransmission deficits observed in the controlled cortical impact (CCI) model of TBI. CCI or naïve rats received daily injections of MPH (5 mg/kg) or saline for 14 days and were assessed on day 15 using fast scan cyclic voltammetry. Dopamine (DA) transporter (DAT) localization, DA-related proteins, and transcription factor (c-fos) expression were also assessed. CCI resulted in reduced electrically evoked overflow of DA and maximal velocity of DA clearance (V(max)). In contrast, CCI was associated with a decrease in the apparent K(M) of DAT. Daily dose of MPH after CCI resulted in robust increases in evoked DA overflow and V(max) as well as increased apparent K(M). Reductions in total striatal DAT expression occurred after CCI and were not further affected by MPH. In contrast, membrane-bound striatal DAT levels were increased in both CCI groups. MPH post-CCI significantly increased striatal c-fos levels compared with saline. These results support the hypothesis that daily MPH improves striatal DA neurotransmission after CCI. DAT expression and transcriptional changes affecting DA protein function may underlie the injury and MPH-induced alterations in neurotransmission observed.