Machine learning algorithms for population-specific risk score in coronary artery bypass grafting.

BACKGROUND: The aim of this study was to develop a new risk prediction score (NH Score) for patients undergoing coronary artery bypass grafting (CABG) specific to the Indian population and compare it to the Society of Thoracic Surgeon (STS) Score and the EuroSCORE II. METHOD: The baseline features of adult patients who underwent CABG between the years 2015 and 2021 (n = 6703) were taken and split into training data (2015-2020; n = 5561) and validation data (2020-2021; n = 1142). The CatBoost algorithm was trained to predict risk scores (NH score), and the performance was tested on the validation set by Precision-Recall Curve and F1 Score. Model calibration was measured by the Brier Score, Expected Calibration Error and Maximum Calibration Error. RESULTS: The NH score outperformed both the STS and EuroSCORE II for all outcomes. For mortality, the PR AUC for NH Score was (0.463 [95% confidence interval [CI], 0.28-0.64]) compared to 0.113 [95% CI, 0.04-0.22] for the STS score and 0.146 [95% CI, 0.06-0.31] for the EuroSCORE II (p ≪ 0.0001). With respect to morbidity NH Score was superior to the STS score (0.43 [95% CI, 0.33-0.50]) vs. (0.229 [95% CI, 0.18-0.3, p < 0.0001). The observed to the predicted ratio for NH score was superior to the STS Score and similar to EuroSCORE II. NH Score was also more accurate at predicting the risk of prolonged ventilation compared to the STS Score. CONCLUSION: NH score shows an excellent improvement over the performance of STS score and EuroSCORE II for modelling risk predictions for patients undergoing CABG in Indian population. It warrants further validation for larger datasets.

Effect of preoperative statins on respiratory complications after coronary artery bypass grafting.

OBJECTIVES: Limited data exist on the effect of preoperative statin therapy on postoperative respiratory complications. Machine learning algorithms (MLA) can process large, heterogenous data, and have immensely improved the ability for risk prediction. In this study, we sought to examine the role of preoperative statins on respiratory complications in patients undergoing coronary artery bypass grafting (CABG) using MLA. METHODS: The study population contained the data of patients who underwent CABG between the years 2015 and 2019 (n = 5638). Three hundred and thirty-seven independent variables were recorded and the data was randomly split with stratified sampling into training and testing data with 20% of the data (1113 records) reserved for model testing. Various models including linear models, Random forest, SVM, and XGboost were trained to predict the incidence of postoperative respiratory complications. Forty-seven important features were found to impact model prediction (p ≤ .05) using the global surrogate model method. A conventional multivariable linear regression model was then used to identify predictors of respiratory complications. RESULTS: One thousand three hundred sixty-two (24.5%) patients developed a respiratory complication in our series. The respiratory complication was seen in 561 (29.7%) of the patients who were not on statin compared to only 801 (21.8%) who were on a statin, p < .0001. The area under the curve for receiver operating characteristic curve using statins and respiratory complications was 0.706. Statins showed positive feature importance in all the MLA models. CONCLUSIONS: MLA showed that statins impacted the prediction of respiratory complications in all the models studied. The study confirmed that preoperative statins reduced the risk of respiratory complications by 21%.

Joint Disposition Properties and Comprehensive Pharmacokinetic Characterization of Antibody-Drug Conjugates.

Antibody-drug conjugates (ADCs) comprise 3 distinct parts: a specific antibody carrier (mAb), a linker, and a cytotoxic payload. Typical pharmacokinetic (PK) characterization of ADCs remains fragmented using separate noncompartmental analyses (NCA) of individual analytes, offering little insight into the dynamic relationships among the ADC components, and the safety and efficacy implications. As a result, it is exceedingly difficult to compare ADCs in terms of favorable PK characteristics. Therefore, there is a need for characterizing ADCs using the joint disposition properties critical for understanding the fate of an ADC complex and clinical implications. In this communication, we describe 3 joint disposition metrics (JDMs) for integrated NCA of ADCs based on a combination of common analytes of ADC, payload, conjugated payload, and total mAb. These JDMs were derived, each in a simple form of a ratio between appropriate PK parameters of two analytes, from the presumed drug delivery scheme behind typical ADC designs, in terms of (1) linker stability, (2) therapeutic exposure ratio, and (3) effective drug-to-antibody ratio in vivo. The validity of the JDM-based PK characterization was examined against model-based analyses via their applications to 3 clinical candidates: PF-06650808, PF-06647020, and PF-06664178. For instance, the linker stability estimates for PF-06650808, PF-06647020, and PF-06664178 were 0.31, 0.14, and 0.096, respectively, from the JDM-based analyses vs. 0.23, 0.11, and 0.086 by the model-based approach. Additionally, the JDMs were estimated for a number of FDA-approved or otherwise well-documented ADCs, showing their utilities in comparing ADCs in terms of favorable PK characteristics.

Dose schedule optimization and the pharmacokinetic driver of neutropenia.

Toxicity often limits the utility of oncology drugs, and optimization of dose schedule represents one option for mitigation of this toxicity. Here we explore the schedule-dependency of neutropenia, a common dose-limiting toxicity. To this end, we analyze previously published mathematical models of neutropenia to identify a pharmacokinetic (PK) predictor of the neutrophil nadir, and confirm this PK predictor in an in vivo experimental system. Specifically, we find total AUC and Cmax are poor predictors of the neutrophil nadir, while a PK measure based on the moving average of the drug concentration correlates highly with neutropenia. Further, we confirm this PK parameter for its ability to predict neutropenia in vivo following treatment with different doses and schedules. This work represents an attempt at mechanistically deriving a fundamental understanding of the underlying pharmacokinetic drivers of neutropenia, and provides insights that can be leveraged in a translational setting during schedule selection.

Preclinical pharmacokinetic/pharmacodynamic/efficacy relationships for alisertib, an investigational small-molecule inhibitor of Aurora A kinase.

PURPOSE: Alisertib (MLN8237) is an investigational inhibitor of Aurora A kinase (AAK). Aurora A plays an essential role in the regulation of spindle assembly and chromosome alignment during mitosis. Inhibition of Aurora A by alisertib in tissue culture has previously been demonstrated to lead to improper chromosomal alignment and disruption of spindle organization, resulting in a transient mitotic delay. The spindle organization defects induced by alisertib have been used to develop a pharmacodynamic (PD) assay for Aurora A inhibition based on the percentage of mitotic cells with proper chromosomal alignment at the metaphase plate (% aligned spindles, abbreviated as AS). The transient mitotic delay that occurs with AAK inhibition permits the use of the mitotic index (the fraction of cells in the population currently undergoing mitosis, abbreviated as MI) as an additional PD assay. When the two PD assays were used in Phase I clinical trials, the reduction in AS was strongly correlated with dose levels and exposures in patients from single time point PD measurements; however, MI failed to show any correlation. To further understand this clinical finding, we constructed PK/PD/efficacy models for AS and MI that can precisely capture the temporal dynamics of the PD markers from in vivo xenograft studies. METHODS: A PK/PD study was conducted using a single oral dose of alisertib at 3, 10, and 20 mg/kg in HCT-116 xenografts implanted subcutaneously in mice. An extravascular, two-compartmental pharmacokinetic (PK) model was used to describe the drug kinetics. Consistent with the mechanistic hypothesis for AAK inhibition, the PD biomarkers such as AS and MI were fitted to PK using a direct response inhibitory sigmoid model and an indirect response turnover model, respectively. The antitumor activity of alisertib dosed orally for 21 days with different dose levels and schedules was evaluated. RESULTS: The PK/PD models showed a fast, sustained response for AS after alisertib administration, whereas MI exhibited a slow, transient response. The PK/efficacy relationship for alisertib in HCT-116 xenografts closely corresponds to the PK/PD relationship for the PD markers, with all three IC50s in close agreement (303, 270, and 280 nM, respectively). CONCLUSION: The PK/PD and PK/efficacy models show that both AS and MI are equally relevant as mechanism-based PD markers to capture drug activity. However, of the two PD markers, the fast, sustained response of AS makes it the only clinically viable PD marker for defining a dose-response relationship, as its maximal effect can be captured from a wider time window with a single PD sampling; while the window to capture dose-related MI response is narrower.

CSR-1 RNAi pathway positively regulates histone expression in C. elegans.

Endogenous small interfering RNAs (endo-siRNAs) have been discovered in many organisms, including mammals. In C. elegans, depletion of germline-enriched endo-siRNAs found in complex with the CSR-1 Argonaute protein causes sterility and defects in chromosome segregation in early embryos. We discovered that knockdown of either csr-1, the RNA-dependent RNA polymerase (RdRP) ego-1, or the dicer-related helicase drh-3, leads to defects in histone mRNA processing, resulting in severe depletion of core histone proteins. The maturation of replication-dependent histone mRNAs, unlike that of other mRNAs, requires processing of their 3'UTRs through an endonucleolytic cleavage guided by the U7 snRNA, which is lacking in C. elegans. We found that CSR-1-bound antisense endo-siRNAs match histone mRNAs and mRNA precursors. Consistently, we demonstrate that CSR-1 directly binds to histone mRNA in an ego-1-dependent manner using biotinylated 2'-O-methyl RNA oligonucleotides. Moreover, we demonstrate that increasing the dosage of histone genes rescues the lethality associated with depletion of CSR-1 and EGO-1. These results support a positive and direct effect of RNAi on histone gene expression.

Transient noise amplification and gene expression synchronization in a bistable mammalian cell-fate switch.

Progenitor cells within a clonal population show variable proclivity toward lineage commitment and differentiation. This cell-to-cell variability has been attributed to transcriptome-wide gene expression noise generated by fluctuations in the amount of cellular machinery and stochasticity in the biochemical reactions involved in protein synthesis. It therefore remains unclear how a signaling network, in the presence of such noise, can execute unequivocal cell-fate decisions from external cues. Here, we use mathematical modeling and model-guided experiments to reveal functional interplay between instructive signaling and noise in erythropoiesis. We present evidence that positive transcriptional feedback loops in a lineage-specific receptor signaling pathway can generate ligand-induced memory to engender robust, switch-like responses. These same feedback loops can also transiently amplify gene expression noise in the signaling network, suggesting that external cues can actually bias seemingly stochastic decisions during cell-fate specification. Gene expression levels among key effectors in the signaling pathway are uncorrelated in the initial population of progenitor cells but become synchronized after addition of ligand, which activates the transcriptional feedback loops. Finally, we show that this transient noise amplification and gene expression synchronization induced by ligand can directly influence cell survival and differentiation kinetics within the population.

Synthetic conversion of a graded receptor signal into a tunable, reversible switch.

The ability to engineer an all-or-none cellular response to a given signaling ligand is important in applications ranging from biosensing to tissue engineering. However, synthetic gene network 'switches' have been limited in their applicability and tunability due to their reliance on specific components to function. Here, we present a strategy for reversible switch design that instead relies only on a robust, easily constructed network topology with two positive feedback loops and we apply the method to create highly ultrasensitive (n(H)>20), bistable cellular responses to a synthetic ligand/receptor complex. Independent modulation of the two feedback strengths enables rational tuning and some decoupling of steady-state (ultrasensitivity, signal amplitude, switching threshold, and bistability) and kinetic (rates of system activation and deactivation) response properties. Our integrated computational and synthetic biology approach elucidates design rules for building cellular switches with desired properties, which may be of utility in engineering signal-transduction pathways.

Tunable signal processing in synthetic MAP kinase cascades.

The flexibility of MAPK cascade responses enables regulation of a vast array of cell fate decisions, but elucidating the mechanisms underlying this plasticity is difficult in endogenous signaling networks. We constructed insulated mammalian MAPK cascades in yeast to explore how intrinsic and extrinsic perturbations affect the flexibility of these synthetic signaling modules. Contrary to biphasic dependence on scaffold concentration, we observe monotonic decreases in signal strength as scaffold concentration increases. We find that augmenting the concentration of sequential kinases can enhance ultrasensitivity and lower the activation threshold. Further, integrating negative regulation and concentration variation can decouple ultrasensitivity and threshold from the strength of the response. Computational analyses show that cascading can generate ultrasensitivity and that natural cascades with different kinase concentrations are innately biased toward their distinct activation profiles. This work demonstrates that tunable signal processing is inherent to minimal MAPK modules and elucidates principles for rational design of synthetic signaling systems.

Integrating extrinsic and intrinsic cues into a minimal model of lineage commitment for hematopoietic progenitors.

Autoregulation of transcription factors and cross-antagonism between lineage-specific transcription factors are a recurrent theme in cell differentiation. An equally prevalent event that is frequently overlooked in lineage commitment models is the upregulation of lineage-specific receptors, often through lineage-specific transcription factors. Here, we use a minimal model that combines cell-extrinsic and cell-intrinsic elements of regulation in order to understand how both instructive and stochastic events can inform cell commitment decisions in hematopoiesis. Our results suggest that cytokine-mediated positive receptor feedback can induce a "switch-like" response to external stimuli during multilineage differentiation by providing robustness to both bipotent and committed states while protecting progenitors from noise-induced differentiation or decommitment. Our model provides support to both the instructive and stochastic theories of commitment: cell fates are ultimately driven by lineage-specific transcription factors, but cytokine signaling can strongly bias lineage commitment by regulating these inherently noisy cell-fate decisions with complex, pertinent behaviors such as ligand-mediated ultrasensitivity and robust multistability. The simulations further suggest that the kinetics of differentiation to a mature cell state can depend on the starting progenitor state as well as on the route of commitment that is chosen. Lastly, our model shows good agreement with lineage-specific receptor expression kinetics from microarray experiments and provides a computational framework that can integrate both classical and alternative commitment paths in hematopoiesis that have been observed experimentally.

Positive receptor feedback during lineage commitment can generate ultrasensitivity to ligand and confer robustness to a bistable switch.

Cytokines and lineage-specific transcription factors are critical molecular effectors for terminal differentiation during hematopoiesis. Intrinsic transcription factor activity is often believed to drive commitment and differentiation, whereas cytokine receptor signals have been implicated in the regulation of cell proliferation, survival, and differentiation. In erythropoiesis, recent experimental findings provide direct evidence that erythropoietin (Epo) can generate commitment cues via the erythropoietin receptor (EpoR); specifically, EpoR signaling leads to activation of the transcription factor GATA-1, which then triggers transcription of erythrocyte-specific genes. In particular, activated GATA-1 induces two positive feedback loops in the system through the enhanced expression of both inactive GATA-1 and EpoR, the latter of which is externally regulatable by Epo. Based upon this network architecture, we present a mathematical model of GATA-1 activation by EpoR, which bidirectionally links a lineage-specific receptor and transcription factor. Our deterministic model offers insight into stimulus-response relationships between Epo and several downstream effectors. In addition to the survival signals that EpoR provides, steady-state analysis of our model suggests that receptor upregulation during lineage commitment can also generate ultrasensitivity to Epo and bistability in GATA-1 activity. These system-level properties can induce a switch-like characteristic during differentiation and provide robustness to the mature state. The topology also suggests a novel mechanism for achieving robust bistability in a purely deterministic manner without molecular cooperativity. The analytical solution of a generalized, minimal model is provided and the significance of each of the two positive feedback loops is elucidated through bifurcation analysis. This network topology, or variations thereof, may link other receptor-transcription factor pairs and may therefore be of general relevance in cellular decision-making.

Effects of forced uncoupling protein 1 expression in 3T3-L1 cells on mitochondrial function and lipid metabolism.

Obesity-related increase in body fat mass is a risk factor for many diseases, including type 2 diabetes. Controlling adiposity by targeted modulation of adipocyte enzymes could offer an attractive alternative to current dietary approaches. Brown adipose tissue, which is present in rodents but not in adult humans, expresses the mitochondrial uncoupling protein 1 (UCP1) that promotes cellular energy dissipation as heat. Here, we report on the direct metabolic effects of forced UCP1 expression in white adipocytes derived from a murine (3T3-L1) preadipocyte cell line. After stable integration, the ucp1 gene product was continuously expressed during differentiation and reduced the total lipid accumulation by approximately 30% without affecting other adipocyte markers, such as cytosolic glycerol-3-phosphate dehydrogenase activity and leptin production. The expression of UCP1 also decreased glycerol output and increased glucose uptake, lactate output, and the sensitivity of cellular ATP content to nutrient removal. However, oxygen consumption and beta-oxidation were minimally affected. Together, our results suggest that the reduction in intracellular lipid by constitutive expression of UCP1 reflects a downregulation of fat synthesis rather than an upregulation of fatty acid oxidation.