Semi-mechanistic Model Research Articles

Size-Optimized Ultrasmall Porous Silica Nanoparticles Depict Vasculature-Based Differential Targeting in Triple Negative Breast Cancer.

Rapid sequestration and prolonged retention of intravenously injected nanoparticles by the liver and spleen (reticuloendothelial system (RES)) presents a major barrier to effective delivery to the target site and hampers clinical translation of nanomedicine. Inspired by biological macromolecular drugs, synthesis of ultrasmall (diameter ≈12-15 nm) porous silica nanoparticles (UPSNs), capable of prolonged plasma half-life, attenuated RES sequestration, and accelerated hepatobiliary clearance, is reported. The study further investigates the effect of tumor vascularization on uptake and retention of UPSNs in two mouse models of triple negative breast cancer with distinctly different microenvironments. A semimechanistic mathematical model is developed to gain mechanistic insights into the interactions between the UPSNs and the biological entities of interest, specifically the RES. Despite similar systemic pharmacokinetic profiles, UPSNs demonstrate strikingly different tumor responses in the two models. Histopathology confirms the differences in vasculature and stromal status of the two models, and corresponding differences in the microscopic distribution of UPSNs within the tumors. The studies demonstrate the successful application of multidisciplinary and complementary approaches, based on laboratory experimentation and mathematical modeling, to concurrently design optimized nanomaterials, and investigate their complex biological interactions, in order to drive innovation and translation.

Nuclei size distribution modelling in wet granulation

Nucleation is an important wet granulation rate process that sometimes has a profound effect on granule attributes and which needs to be captured in process modelling studies. However, existing models fail to predict nuclei size distribution of a range of spray conditions typically used in industry. In this paper, the dimensionless nucleation number Ψn is used to develop two new nuclei size distribution models, one empirical and one semi-mechanistic. The empirical model assumes a log-normal distribution (LND), and the semi-mechanistic model is based on a approach proposed by Hapgood et al. (2009), which applies the Poisson distribution (PD) function. Modelling parameters are estimated using Monte Carlo simulations (MCS) data. From the models, the nuclei size distribution can be easily determined using analytical equations, which simplifies the inclusion in a population balance modelling (PBM) framework. The results of both models are assessed using MCS data as well as experimental data from literature. The empirical LND model is able to capture the MCS results accurately, and the predictions agree reasonably well with the experimental results over a wide range of dimensionless nucleation number (0<Ψn<3). The predictions of the semi-mechanistic modified Poisson distribution (MPD) model do not agree qualitatively with the MCS or experimental results. A sensitivity analysis shows that the MCS modelling assumptions need to capture the spatial drop distribution in the spray accurately, while the drop size distribution can be assumed to be uniform. Overall, we recommend that the LND model with the parameter values estimated be used in PBM frameworks to determine nuclei size distribution for a wide range of experimental conditions in mixers and fluidised granulators.

Real-time Epidemic Forecasting: Challenges and Opportunities.

Infectious disease outbreaks play an important role in global morbidity and mortality. Real-time epidemic forecasting provides an opportunity to predict geographic disease spread as well as case counts to better inform public health interventions when outbreaks occur. Challenges and recent advances in predictive modeling are discussed here. We identified data needs in the areas of epidemic surveillance, mobility, host and environmental susceptibility, pathogen transmissibility, population density, and healthcare capacity. Constraints in standardized case definitions and timely data sharing can limit the precision of predictive models. Resource-limited settings present particular challenges for accurate epidemic forecasting due to the lack of granular data available. Incorporating novel data streams into modeling efforts is an important consideration for the future as technology penetration continues to improve on a global level. Recent advances in machine-learning, increased collaboration between modelers, the use of stochastic semi-mechanistic models, real-time digital disease surveillance data, and open data sharing provide opportunities for refining forecasts for future epidemics. Epidemic forecasting using predictive modeling is an important tool for outbreak preparedness and response efforts. Despite the presence of some data gaps at present, opportunities and advancements in innovative data streams provide additional support for modeling future epidemics.

Simultaneous Semi-Mechanistic Population Pharmacokinetic Modeling Analysis of Enalapril and Enalaprilat Serum and Urine Concentrations From Child Appropriate Orodispersible Minitablets.

Enalapril is recommended as the first line of therapy and is proven to improve survival rates for treatment of Pediatric Heart Failure; however, an approved drug and child appropriate dosage formulation is still absent. The present analysis was conducted to perform a detailed model informed population pharmacokinetic analysis of prodrug enalapril and its active metabolite enalaprilat in serum and urine. Further, a model informed dosage form population-pharmacokinetic analysis was conducted to evaluate differences in pharmacokinetics of enalapril and its active metabolite enalaprilat when prodrug was administered to 24 healthy adults in a crossover, two periods, two treatments, phase I clinical trial using child-appropriate orodispersible mini-tablets (ODMT) and reference (Renitec®) dosage formulation. A simultaneous semi-mechanistic population-pharmacokinetic model was developed using NONMEM software, which predicted full profile serum and urine concentrations of enalapril and enalaprilat. First-order conditional estimation with interaction was used for parameter estimation. Transit compartments added using Erlang distribution method to predicted enalapril absorption and enalaprilat formation phases. Normalized body weight was identified as covariate related to enalapril volume of distribution. Visual predictive check (VPC) plots and conducted bootstrap analysis validated the model. The data from the two formulations were pooled for population-pharmacokinetic analysis and covariate effect of the formulation was found on mean transit time (MTT1) of enalapril absorption. In addition, data of each formulation were modeled separately and the estimated parameters of each individual administered both formulations were correlated using paired samples Wilcoxon rank test (p < 0.05 = significant) which also showed only a significant difference (p = 0.03) in MTT1 i.e., 5 min early appearance of enalapril from ODMT compared to reference tablets. No difference in the pharmacokinetics of active enalaprilat was found from the ODMT compared to the reference formulation. The population pharmacokinetic analysis provided detailed information about the pharmacokinetics of enalapril and enalaprilat, which showed that the ODMT formulation might have similar pharmacodynamic response compared to the reference formulation.

Predicting dry matter intake using conceptual models for cattle kept under tropical and subtropical conditions1.

Using empirical models to predict voluntary dry matter intake (VDMI) of cattle across production systems in the (Sub-)Tropics often yields VDMI estimates with low adequacy (i.e., accuracy and precision). Thus, we investigated whether semimechanistic conceptual mathematical models (CMM) developed for cattle in temperate areas could be adopted and adjusted to adequately predict VDMI of stall-fed cattle in the (Sub-)Tropics. The CMM of Conrad et al. (1964) (C1) and Mertens (1987) (M1) were identified and adopted for their simplicity in reflecting physicophysiological VDMI regulation. Both CMM use 2 equations that estimate the physiologically and physically regulated VDMI and retain the lower VDMI prediction as actual VDMI. Furthermore, C1 was modified by increasing the daily average fecal dry matter output from 0.0107 to 0.0116 kg/kg body weight, yielding the modified model C2. For M1, the daily neutral detergent fiber intake capacity was increased from 0.012 to 0.0135 kg/kg body weight and the daily metabolizable energy requirements for maintenance from 0.419 to 0.631 MJ/kg0.75 body weight, whereas the metabolizable energy requirements for gain was reduced from 32.5 to 24.3 MJ/kg body weight gain, yielding the modified model M2. Last, also the mean of the physically and physiologically regulated VDMI rather than the lower of both estimates was retained as actual VDMI to generate the models C3 (from C1), C4 (from C2), M3 (from M1), and M4 (from M2). The 8 CMM were then evaluated using a data set summarizing results from 52 studies conducted under (sub)tropical conditions. The mean bias, root mean square error of prediction (RMSEP) and concordance correlation coefficient (CCC) were used to evaluate adequacy and robustness of all CMM. The M4, C2, and C1 were the most adequate CMM [i.e., lowest mean biases (0.07, -0.22, and 0.14 kg/animal and day, respectively), RMSEP (1.62, 1.93, and 2.0 kg/animal and day, respectively), and CCC (0.91, 0.86, and 0.85, respectively)] and robust of the 8 CMM. Hence, CMM can adequately predict VDMI across diverse stall-fed cattle systems in the (Sub-)Tropics. Adjusting CMM to reflect the differences in feed quality and animal physiology under typical husbandry conditions in the (Sub-)Tropics and those in temperate areas improves the adequacy of their VDMI predictions.

Abstract 684: A semi-mechanistic platform model to capture individual animal responses to checkpoint inhibitors in a syngeneic mouse model

Abstract Syngeneic mouse models have been widely employed in preclinical discovery of checkpoint inhibitors as they enable study of drug impact on the intact immune system (Lechner, 2013; Murphy, 2015). However, the interpretation of such studies remains challenging partly due to the large variability in individual animal responses to drug treatment. In this work, we describe the generation of a model platform that captures essential aspects of the pharmacokinetics, cellular and tumor growth effects of murine surrogates of two checkpoint therapeutic antibodies, anti-PD1 and anti-CTLA4, in the CT26 syngeneic tumor model. The model describes individual animal responses with regard to drug exposure, key intra-tumoral cell kinetics and tumor volume changes and provides biologically plausible explanations for the observed differences between good and poor responders to treatment with anti-PD1 or anti-CTLA4. We used the model to predict the antibody dose-response relationships for individual animals and to identify dose thresholds above which complete tumor elimination can be achieved in good responders. In contrast, our models predict that poor responders would not achieve complete response even with much higher drug doses. The parameters in our model that impact the response in poor responders are not drug-related. This finding suggests that immune-cell related barriers have to be crossed in order to achieve a therapeutic response in these animals - possibly via combination therapy. In addition, we identified the net tumor cell doubling rate, one potential parameter that contributes to individual variability in response to treatment, as the most sensitive biological parameter determining tumor volume changes upon treatment with anti-PD1 or anti-CTLA4. Measuring individual animal tumor cell growth characteristics may help with the experimental design and qualification of animals for studies (in addition to absolute tumor volume), and thereby reduce inter-animal variability and enhance the interpretability of study results, especially in combination with a model such as the one presented here. This model platform can be adapted to capture and compare checkpoint drug effects in different syngeneic tumor models. Moreover, it can be expanded to add additional drug mechanisms and can serve as a tool to inform the experimental design of mouse studies. Citation Format: Alison M. Betts, Lin Lin, Carissa L. Young, Wenlian Qiao, Jatin Narula, Peter O'Brien, Derek Bartlett, Andrea Hooper, Jason Williams, John M. Burke, Joshua Apgar, Fei Hua, Lore Gruenbaum. A semi-mechanistic platform model to capture individual animal responses to checkpoint inhibitors in a syngeneic mouse model [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 684.

Abstract 687: Implementation of inter-patient variability in semi-mechanistic model of melanoma treatment with vemurafenib

Abstract Introduction: The number of parameters in quantitative systems pharmacology (QSP), mechanistic and semi-mechanistic models is greater than in conventional PK/PD models. That is why it is not possible to implement inter-patient variability using only clinical data. The objective of this study was to describe inter-patient variability in the semi-mechanistic model of melanoma treatment with vemurafenib, a BRAF inhibitor, on the basis of published human in vitro, in vivo and clinical data, and to predict overall response from phase 2 clinical trial. Methods: Model described melanoma progression and its treatment with vemurafenib. It includes (i) proliferation, apoptosis, and transition between states (c-MET negative/positive and sensitive/resistant to BRAF inhibitor) of cancer cells; (ii) HGF effect on proliferation of melanoma cells; (iii) PK/PD of vemurafenib. Various types of data were used to extract inter-patient variability, for example: (a) in vitro data on effect of vemurafenib on melanoma cell lines [1]; (b) in vivo data on tumor growth in melanoma patients without treatment [2]; (c) in vivo data on HGF level in blood plasma of melanoma patients [3]. Results: Different virtual patients developed by the model reproduce all types of tumor response to vemurafenib monotherapy according to RECIST criteria: complete response, partial response, stable disease and progressive disease. To evaluate the predictive ability of the model in a quantitative manner we compared results of simulation of virtual patients population with clinical outcomes. An overall response rate was 70% in the model versus 53% (95% CI, 44 to 62) observed in phase 2 clinical trial [4]. Overestimation can be explained by intermittent administration and dose reduction in case of adverse events in clinical trials which were not taken into account in the model. Conclusions: Our work establishes the benefit of using of in vitro and in vivo data in addition to clinical data to predict the variability in treatment outcomes of cancer patients. [1] Stones et al. Front Genet. 2013 May 8;4:66 [2] Hartung et al. PLoS One. 2017 May 4;12(5):e0176080 [3] Hügel et al. Melanoma Res. 2016 Aug;26(4):354-60 [4] Sosman et al. N Engl J Med. 2012 Feb 23;366(8):707-14 Citation Format: Dmitry Shchelokov, Oleg Demin Jr, Oleg Demin. Implementation of inter-patient variability in semi-mechanistic model of melanoma treatment with vemurafenib [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 687.

Nonlinear pharmacokinetics of rituximab in non-Hodgkin lymphomas: A pilot study.

Rituximab is an anti-CD20 monoclonal antibody approved in non-Hodgkin lymphoma (NHL). This study aimed to assess the relationship between antigen mass and nonlinear pharmacokinetics of rituximab in NHL patients. In a retrospective cohort of 25 NHL patients treated with rituximab, antigen mass was assessed at baseline by measuring metabolic tumour volume (MTV) by positron emission tomography. Rituximab pharmacokinetics was described using a semimechanistic 2-compartment model including a latent target antigen. Rituximab target-mediated elimination was described as irreversible binding between rituximab and it target. Histology (follicular or diffuse large B-cell lymphomas), initial MTV and body weight were tested as covariates on pharmacokinetic parameters. The model allowed a satisfactory description of rituximab serum concentrations. Target-mediated elimination was maximum at the beginning of treatment and became negligible towards the end of follow-up. The second-order elimination of rituximab due to target binding and complex elimination increased with baseline MTV. Central volume of distribution increased with body weight (P = .022) and baseline MTV (P = .005). This study quantified for the first time the target-mediated elimination of rituximab in NHL patients and confirmed rituximab retention by antigen mass.

1198-P: The M-EASE Studies: A Modeling and Simulation Approach to Further Characterize the Efficacy of Low-Dose Empagliflozin as Adjunctive to Insulin Therapy in Type 1 Diabetes (T1DM)

Among the two phase 3 trials of empagliflozin (EMPA) in T1DM, one (EASE-3), demonstrated diabetic ketoacidosis risk minimization by including a lower dose (2.5 mg) than approved for use in T2DM while maintaining HbA1c benefit of -0.28%. We aimed to determine the effect of HbA1c lowering of EMPA 2.5 mg by modelling and simulation in the phase 3 trial (EASE-2) that did not investigate this dose. The effect of insulin adjustment on HbA1c was also characterized. Independent of empiric data from EASE-3 (that evaluated EMPA 2.5 mg), M-EASE-2, an empirical model, evaluated the direct effect of EMPA exposure on changes in HbA1c after simulated 26- and 52-week treatment durations in the EASE-2 study. M-EASE-2 model development was informed by EASE-2 and EASE-1 (phase 2 study, included EMPA 2.5 mg) data. M-EASE-1, a semi-mechanistic model (informed by EASE-1 and EASE-2 data), simulated changes in HbA1c after a 26-week treatment period (with/without insulin dose adjustment) driven by changes in EMPA exposure, insulin dose and mean daily glucose. Trial simulations were conducted in 239 and 500 patients per dose group in M-EASE-2/-1, respectively. The models were validated using EASE-3 trial data. In M-EASE-2, the simulated mean (median; 95% CI) HbA1c change for EMPA 2.5 mg was -0.29% (-0.29; -0.39, -0.19) at Week 26 and -0.29% (-0.29; -0.40, -0.19) at Week 52. The simulations performed were consistent with EASE-3 results. In M-EASE-1, the 26-week HbA1c change for EMPA 2.5 mg was -0.28% (-0.28; -0.40, -0.09) and -0.41% (-0.40; -0.53, -0.20) with adjusted and stable insulin therapy, respectively. The HbA1c benefit from low-dose EMPA 2.5 mg that was directly observed in a phase 3 clinical trial was confirmed using two modelling approaches from independent studies. Efficacy was sustained over 52 weeks, and the approach predicted that greater efficacy of EMPA as adjunct to insulin in T1DM can be achieved on a stable vs. variable insulin dose. Disclosure B.A. Perkins: Advisory Panel; Self; Abbott, Boehringer Ingelheim International GmbH, Boehringer Ingelheim International GmbH, Insulet Corporation. Research Support; Self; Boehringer Ingelheim International GmbH. Other Relationship; Self; Abbott, Boehringer Ingelheim International GmbH, Lilly Diabetes, Medtronic, Novo Nordisk Inc., Sanofi. N. Soleymanlou: Employee; Self; Boehringer Ingelheim Canada Ltd. J. Rosenstock: Research Support; Self; AstraZeneca, Bristol-Myers Squibb Company, Genentech, Inc., GlaxoSmithKline plc., Lexicon Pharmaceuticals, Inc., Melior Pharmaceuticals, Inc., Bukwang Pharm. Co., Ltd., Merck &amp; Co., Inc., Oramed Pharmaceuticals, PegBio Co., Ltd., Pfizer Inc. Other Relationship; Self; Boehringer Ingelheim International GmbH, Eli Lilly and Company, Intarcia Therapeutics, Inc., Janssen Pharmaceuticals, Inc., Novo Nordisk Inc., Sanofi. J.S. Skyler: Advisory Panel; Self; ADOCIA, Applied Therapeutics, Dance Biopharm Holdings Inc., Orgenesis Ltd., Tolerion, Inc., Viacyte, Inc. Board Member; Self; Dexcom, Inc., Intarcia Therapeutics, Inc., Moerae Matrix, Inc. Consultant; Self; AstraZeneca, Boehringer Ingelheim Pharmaceuticals, Inc., Dalcor, Dialogics, Elcelyx Therapeutics, Inc., Esperion, GeNeuro Innovation, Ideal Life, Immunomolecular Therapeutics, Intrexon, Kamada, Nestlé, Sanofi, Valeritas, Inc., Zafgen, Inc. Stock/Shareholder; Self; Dexcom, Inc., Ideal Life, Intarcia Therapeutics, Inc., Intrexon, Moerae Matrix, Inc. L.M. Laffel: Advisory Panel; Self; Lilly Diabetes, Novo Nordisk A/S, Roche Diabetes Care, Sanofi. Consultant; Self; AstraZeneca, Boehringer Ingelheim International GmbH, Dexcom, Inc., Janssen Pharmaceuticals, Inc., UpToDate. K. Liesenfeld: Employee; Self; Boehringer Ingelheim International GmbH. D. Neubacher: Employee; Self; Boehringer Ingelheim Pharma GmbH &amp; Co. KG. M. Riggs: Consultant; Self; Boehringer Ingelheim International GmbH. C.K. Johnston: Consultant; Self; Boehringer Ingelheim International GmbH. R.J. Eudy-Byrne: Consultant; Self; Boehringer Ingelheim International GmbH. Other Relationship; Spouse/Partner; Eli Lilly and Company, Medtronic, Tandem Diabetes Care. A. Elmokadem: None. J. George: Employee; Self; Boehringer Ingelheim International GmbH. J. Marquard: Employee; Self; Boehringer Ingelheim International GmbH. V.C. Nock: Employee; Self; Boehringer Ingelheim Pharma GmbH &amp; Co. KG. Funding Boehringer Ingelheim; Eli Lilly and Company

Modeling Subcutaneous Absorption of Long-Acting Insulin Glargine in Type 1 Diabetes.

Subcutaneous (sc) administration of long-acting insulin analogs is often employed in multiple daily injection (MDI) therapy of type 1 diabetes (T1D) to cover patient's basal insulin needs. Among these, insulin glargine 100U/mL (Gla-100) and 300 U/mL (Gla-300) are formulations indicated for once daily sc administration in MDI therapy of T1D. A few semi-mechanistic models of sc absorption of insulin glargine have been proposed in the literature, but were not quantitatively assessed on a large dataset. The aim of this paper is to propose a model of sc absorption of insulin glargine able to describe the data and provide precise model parameters estimates with a clear physiological interpretation. Three candidate models were identified on a total of 47 and 77 insulin profiles of T1D subjects receiving a single or repeated sc administration of Gla-100 or Gla-300, respectively. Model comparison and selection were performed on the basis of their ability to describe the data and numerical identifiability. The most parsimonious model is linear two-compartment and accounts for the insulin distribution between the two compartments after sc administration through parameter k. Between the two formulations, we report a lower fraction of insulin in the first versus second compartment (k = 86% versus 94% in Gla-100 versus Gla-300, p < 0.05), a lower dissolution rate from the first to the second compartment ([Formula: see text] versus 0.0008 min-1 in Gla-100 versus Gla-300, p << 0.001), and a similar rate of insulin absorption from the second compartment to plasma ([Formula: see text] versus 0.0016 min-1 in Gla-100 versus Gla-300, p = NS), in accordance with the mechanisms of insulin glargine protraction. The proposed model is able to both accurately describe plasma insulin data after sc administration and precisely estimate physiologically plausible parameters. The model can be incorporated in simulation platforms potentially usable for optimizing basal insulin treatment strategies.

Modeled CA-125 kinetics during neoadjuvant chemotherapy for predicting the likelihood of optimal interval debulking surgery in ovarian cancer patients: Data from CHIVA trial (a GINECO study).

5546 Background: A pre-operative predictive biomarker of CC0 interval debulking surgery (IDS) likelihood would be helpful. The modeled CA125 elimination rate constant KELIM predicts OS in 1st line setting (You et al. Clin Cancer Res 2019). The predictive/prognostic values of KELIM regarding CC scores at IDS, and survivals, during neo-adjuvant chemotherapy were assessed. Methods: The data of the CHIVA randomized phase II trial, comparing carboplatin-paclitaxel +/- nintedanib before IDS (NCT01583322), were used. A semi-mechanistic model was built to describe CA125 longitudinal kinetics during the first 100 treatment days. The relationships between KELIM and IDS CC scores, PFS &amp; OS, were assessed with other major prognostic factors (grade, histology, GCIG CA125 response, FIGO stage, and arm) using multivariate logistic regression (logit), C-index &amp; survival tests. Results: The longitudinal kinetics of 529 CA125 values, assessed every 3 weeks during neo-adj chemotherapy, were modeled in 133 patients (out of 188). KELIM (as a continuous covariate) was the only significant predictive factor of CC0 IDS likelihood using multivariate analyses (OR = 12.37, 95% CI [4.32-39.67]). CC0 IDS probability can be estimated with patient KELIM: ≥ 90 % if standardized KELIM ≥ 0.12. Non-parametric survival models confirmed the independent predictive values of KELIM categorized by terciles regarding PFS &amp; OS (Table). The parametric model linking KELIM (as a continuous covariate) with OS allows to predict the patient survivals (months) based on their estimated KELIM (HR = 0.20, [0.10-0.39]). Conclusions: The prognostic &amp; predictive values of the modeled CA125 KELIM are also confirmed regarding CC0 IDS likelihood, PFS and OS with neo-adjuvant chemotherapy. Patient KELIM is calculable online, based on observed CA125 values, on http://www.biomarker-kinetics.org/ . Clinical trial information: 2011-006288-23. [Table: see text]

Oxygen and Hydrogen Isotopes in Human Hair and Tap Water: Modeling Relationships in a Modern Mexican Population

This study investigated the relationship between 18O and 2H isotopes in samples of Mexican hair and drinking water. The purpose of this study was twofold: to quantify the relationship between isotopes in Mexican hair and tap water, in order to understand the impact of water stress and differing socioeconomic status on accurate predictions of drinking water; and to determine whether currently existing semimechanistic models can accurately represent the relationship between hair and tap water. This study used a subset of paired samples of human hair (n = 62) and tap water (n = 76). Isotope values in tap water ranged from –11.4‰ to –4.3‰ and –79.1‰ to –22.5‰, and in hair from +9.5‰ to +16.1‰ and –90.8‰ to –53.7‰, for δ18O and δ2H, respectively. The most depleted δ18O and δ2H hair values came from individuals in the state of Morelos. For modern Mexican populations, positive correlations between isotopes in hair and water were not significant, with correlation coefficients r = 0.61 (p = 0.05) and r = 0.60 (p = 0.06) for 18O and 2H, respectively. Error-in-variables regression yielded linear fits that were somewhat better for 2H relative to 18O: δ18Oh = 0.183 [±0.132] δ18Otw + 15.7 [±0.9]‰ (r2 = 0.23); δ2Hh = 0.181 [±0.076] δ2Htw – 64.0 [±3.0]‰ (r2 = 0.34). In short, data from this Mexican population did not exhibit the strong relationships between isotope values of 18O and 2H in tap water and hair that have been characteristic of other populations studied to date. Given the economic stratification of this region and the poor correlation between hair and water samples, the authors considered the possibility that l, the fraction of the diet derived from local sources, and fs, the fraction of nonexchangeable H in keratin that was fixed in vivo, are local rather than global parameters for this population. The authors estimated different values of l and fs for each location. Given the anticipated importance of the nonlocal dietary contribution, they treated the isotopic content of nonlocal food and the offset parameters for predicting isotopes in locally derived food as tuning parameters and compared the results with parameters based on the American supermarket diet. They found that, although O and H isotopes in water and hair maintained similar geographic distributions, O and H isotopes in tap water explained only a small part of the variation observed in hair samples. Compared to the standard American supermarket diet, the Mexican estimates for nonlocal diet and local diet offsets predicted regional distributions of l and fs that cleanly segregated urban areas from rural towns.

Development and validation of a model for whole course aging of nickel added to a wide range of soils using a complementary error function

After water-soluble nickel (Ni) is added to soils, its bioavailability/toxicity, isotopic exchangeability, and extractability decline over time. Two separate semi-mechanistic models (Sqrt-model and Ln-model), have been developed to predict short- and long-term aging of Ni added to soils. To continuously predict Ni aging in one model, a semi-mechanistic model (Erfc-model) was developed in the present study where the description of the diffusion process was based on a complementary error function. The Erfc-model for whole course aging predicted short- and long-term Ni aging continuously with regression coefficients (R2) of 0.87 and 0.94, which were comparable with the Sqrt-model (0.91) and Ln-model (0.89). Based on the Erfc-model, soil pH and aging time were two important factors influencing the Ni aging processes. An additional 45 soils covering short and long terms of aging were applied to validate the model. A strong correlation (R2 = 0.85) was found between the measured and the predicted lability of Ni in soils with root-mean-square-error (RMSE) of 8.97%, illustrating that the Erfc-model could continuously predict short- and long-term aging processes of Ni added to soils. The model could be used as an alternative to normalize ecotoxicological data for driving soil environmental quality standards.

Misidentification Subtype of Alzheimer's Disease Psychosis Predicts a Faster Cognitive Decline

The presence of psychosis is associated with a more rapid decline in Alzheimer's disease (AD), but the impact of paranoid (persecutory delusions) and misidentification (misperceptions and/or hallucinations) subtypes of psychosis on the speed of decline in AD is still unclear. We analyzed data on Alzheimer's Disease Neuroimaging Initiative 2 participants with late mild cognitive impairment or AD, and we described individual trajectories of Alzheimer's Disease Assessment Scale–Cognitive Subscale scores using a semimechanistic logistic model with a mixed effects–based approach, which accounted for dropout and adjusted for baseline Mini Mental State Examination scores. The covariate model included psychosis subtypes, age, gender, education, medications, and Apolipoprotein E epsilon 4 (Apo‐e ε4 genotype). We found that the Alzheimer's Disease Assessment Scale–Cognitive Subscale rate of increase was doubled in misidentification (β r,misid_subtype = 0.63, P = 0.031) and mixed (both subtypes; β r,mixed_subtype = 0.70, P = 0.003) when compared with nonpsychotic (or paranoid) patients, suggesting that the misidentification subtype may represent a distinct AD sub‐phenotype associated with an accelerated pathological process.

Inter-annual variation in seasonal dengue epidemics driven by multiple interacting factors in Guangzhou, China

Vector-borne diseases display wide inter-annual variation in seasonal epidemic size due to their complex dependence on temporally variable environmental conditions and other factors. In 2014, Guangzhou, China experienced its worst dengue epidemic on record, with incidence exceeding the historical average by two orders of magnitude. To disentangle contributions from multiple factors to inter-annual variation in epidemic size, we fitted a semi-mechanistic model to time series data from 2005–2015 and performed a series of factorial simulation experiments in which seasonal epidemics were simulated under all combinations of year-specific patterns of four time-varying factors: imported cases, mosquito density, temperature, and residual variation in local conditions not explicitly represented in the model. Our results indicate that while epidemics in most years were limited by unfavorable conditions with respect to one or more factors, the epidemic in 2014 was made possible by the combination of favorable conditions for all factors considered in our analysis.

Estimating summer sea surface pCO2 on a river-dominated continental shelf using a satellite-based semi-mechanistic model

A satellite-based semi-mechanistic model was developed to estimate the sea surface partial pressure of CO2 (pCO2) on the river-dominated Louisiana Continental Shelf (LCS) in summer months using satellite-derived products including chlorophyll-a concentration (Chl-a_sat), sea surface salinity (SSS_sat), and sea surface temperature (SST_sat). The model analytically expresses pCO2 as the sum of physical (pCO2,mix) and biological (ΔpCO2,bio) contributions which are the primary controlling factors associated with the pCO2 variability on the LCS. The SSS_sat, derived from a locally calibrated retrieval algorithm, was used in a two-end member river-ocean mixing model to estimate the concentrations of dissolved inorganic carbon (DIC) and total alkalinity (TA) resulting from physical mixing. The mixing-induced DIC and TA were used together with SST_sat to calculate the pCO2,mix resulting from the mixing processes at the in situ temperature. The biological uptake of pCO2 through net community production (ΔpCO2,bio) was derived from a previously validated local Chl-a_sat product along with the difference between the pCO2,mix and the extensive in situ underway pCO2 data. The satellite-derived pCO2 agreed with in situ measurements with a median absolute difference of 37 μatm, mean difference of −5 μatm, and a root mean square difference of 58 μatm for five summer cruises spanning 2006–2009. The semi-mechanistic remote sensing model differentiates physical and biological influences on the pCO2 variability, which will help to better investigate the variability and the underlying controlling mechanisms of pCO2 and CO2 flux on the LCS. This study demonstrates that the use of locally calibrated satellite products (Chl-a_sat, SSS_sat) should be promoted to improve the satellite estimation of pCO2 in coastal waters. This method can potentially be applied in other river-dominated continental shelves with local tuning.

Dynamic mathematical modeling of heat and mass transfer incorporating with the local nutrient and biomass limitation of growth in a packed-bed solid-state bioreactor

This research develops on our previous semi-mechanistic model that describes the dynamic physical and biochemical processes taking place in a packed-bed bioreactor to analyze the relationship of nutrient limitation, biomass accumulation, metabolic heat generation, and mathematical description of packed-bed porous media. The experimental and simulation data proved that glucose concentration gradients in the biofilm could be neglected due to small biofilm thickness and high diffusivity of glucose in the biofilm. The prediction results also showed that an increase in the initial substrate concentration leads to a rise in the temperature gradient in the bed. The model proposes that if the diameter of substrate particle is too large (r > 0.1 cm), the growth rate will decrease significantly due to the high biomass accumulation in the biofilm, and temperature gradients decrease in the column. This can be used as a strategy to control the overheating problem in the bed.

Assessing the performance of real-time epidemic forecasts: A case study of Ebola in the Western Area region of Sierra Leone, 2014-15.

Real-time forecasts based on mathematical models can inform critical decision-making during infectious disease outbreaks. Yet, epidemic forecasts are rarely evaluated during or after the event, and there is little guidance on the best metrics for assessment. Here, we propose an evaluation approach that disentangles different components of forecasting ability using metrics that separately assess the calibration, sharpness and bias of forecasts. This makes it possible to assess not just how close a forecast was to reality but also how well uncertainty has been quantified. We used this approach to analyse the performance of weekly forecasts we generated in real time for Western Area, Sierra Leone, during the 2013–16 Ebola epidemic in West Africa. We investigated a range of forecast model variants based on the model fits generated at the time with a semi-mechanistic model, and found that good probabilistic calibration was achievable at short time horizons of one or two weeks ahead but model predictions were increasingly unreliable at longer forecasting horizons. This suggests that forecasts may have been of good enough quality to inform decision making based on predictions a few weeks ahead of time but not longer, reflecting the high level of uncertainty in the processes driving the trajectory of the epidemic. Comparing forecasts based on the semi-mechanistic model to simpler null models showed that the best semi-mechanistic model variant performed better than the null models with respect to probabilistic calibration, and that this would have been identified from the earliest stages of the outbreak. As forecasts become a routine part of the toolkit in public health, standards for evaluation of performance will be important for assessing quality and improving credibility of mathematical models, and for elucidating difficulties and trade-offs when aiming to make the most useful and reliable forecasts.

Development of a Semimechanistic Pharmacokinetic-Pharmacodynamic Model Describing Dextroamphetamine Exposure and Striatal Dopamine Response in Rats and Nonhuman Primates following a Single Dose of Dextroamphetamine.

Acute central nervous system exposure to dextroamphetamine (d-amphetamine) elicits a multitude of effects, including dual action on the dopamine transporter (DAT) to increase extracellular dopamine, and induction of a negative feedback response to limit the dopamine increase. A semimechanistic pharmacokinetic and pharmacodynamic (PK/PD) model with consideration of these multiple effects as a basis was developed. Integrated pharmacokinetics of d-amphetamine in plasma, brain extracellular fluid (ECF) via microdialysis, and cerebrospinal fluid were characterized using a population approach. This PK model was then linked to an indirect-response pharmacodynamic model using as a basis the measurement of extracellular striatal dopamine, also via microdialysis. In both rats and nonhuman primates (NHPs), d-amphetamine stimulation of dopamine outflow (reverse transport) through DAT was primarily responsible for the dose-linear increase in dopamine. As well, in both species a moderator function was needed to account for loss of the dopamine response in the presence of a relatively sustained d-amphetamine ECF exposure, presumptive of an acute tolerance response. PK/PD model structure was consistent between species; however, there was a 10-fold faster return to baseline dopamine in NHPs in response to an acute d-amphetamine challenge. These results suggest preservation from rodents to NHPs regarding the mechanism by which amphetamine increases extracellular dopamine, but a faster system response in NHPs to tolerate this increase. This microdialysis-based PK/PD model suggests greater value in directing preclinical discovery of novel approaches that modify reverse transport stimulation to treat amphetamine abuse. General value regarding insertion of an NHP model in paradigm rodent-to-human translational research is also suggested.

Multiobjective Identification of a Feedback Synthetic Gene Circuit

Kinetic (i.e., dynamic) semimechanistic models based on the first principles are particularly important in systems and synthetic biology since they can explain and predict the functional behavior that emerges from the time-varying concentrations in cellular components. However, gene circuit models are nonlinear higher order ones and have a large number of parameters. In addition, experimental measurements are often scarce, and enough signal excitability for identification cannot always be achieved. These characteristics render the identification problem ill-posed, so most gene circuit models present incomplete parameter identifiability. Thus, parameter identification of typical biological models still appears as an open problem, where ensemble modeling approaches and multiobjective optimization arise as natural options. We address the problem of identifying the stochastic model of a closed-loop synthetic genetic circuit designed to minimize the gene expression noise. The model results from the feedback interaction between two subsystems. Besides incomplete parameter identifiability, the closed-loop dynamics cannot be directly identified due to the lack of enough input signal excitability. We apply a two-stage approach. First, the open-loop averaged time-course experimental data are used to identify a reduced-order stochastic model of the system direct chain. Then, closed-loop steady-state stochastic distributions are used to identify the remaining parameters in the feedback configuration. In both cases, multiobjective optimization is used to address the parameter identifiability, providing sets of parameters valid for different state-space regions. The methodology gives good identification results, provides clear guidelines on the effect of the parameters under different scenarios, and it is particularly useful for easily combining time-course population averaged and steady-state single-cell distribution experimental data.