Bayesian inference for prediction of survival probability in prime-boost vaccination regimes.

We focus on Bayesian inference for survival probabilities in a prime-boost vaccination regime in the development of an Ebola vaccine. We are interested in the heterologous prime-boost regimen (unmatched vaccine deliverys using the same antigen) due to its demonstrated durable immunity, well-tolerated safety profile, and suitability as a population vaccination strategy. Our research is motivated by the need to estimate the survival probability given the administered dosage. To do so, we establish two key relationships. Firstly, we model the connection between the designed dose concentration and the induced antibody count using a Bayesian response surface model. Secondly, we model the association between the antibody count and the probability of survival when experimental subjects are exposed to the Ebola virus in a controlled setting using a Bayesian probability of survival model. Finally, we employ a combination of the two models with dose concentration as the predictor of the survival probability for a future vaccinated population. We implement our two-level Bayesian model in Stan, and illustrate its use with simulated and real-world data. Performance of this model is evaluated via simulation. Our work offers a new application of drug synergy models to examine prime-boost vaccine efficacy, and does so using a hierarchical Bayesian framework that allows us to use dose concentration to predict survival probability.

Dose-response modeling in high-throughput cancer drug screenings: an end-to-end approach.

Personalized cancer treatments based on the molecular profile of a patient's tumor are an emerging and exciting class of treatments in oncology. As genomic tumor profiling is becoming more common, targeted treatments for specific molecular alterations are gaining traction. To discover new potential therapeutics that may apply to broad classes of tumors matching some molecular pattern, experimentalists and pharmacologists rely on high-throughput, in vitro screens of many compounds against many different cell lines. We propose a hierarchical Bayesian model of how cancer cell lines respond to drugs in these experiments and develop a method for fitting the model to real-world high-throughput screening data. Through a case study, the model is shown to capture nontrivial associations between molecular features and drug response, such as requiring both wild type TP53 and overexpression of MDM2 to be sensitive to Nutlin-3(a). In quantitative benchmarks, the model outperforms a standard approach in biology, with $\approx20\%$ lower predictive error on held out data. When combined with a conditional randomization testing procedure, the model discovers markers of therapeutic response that recapitulate known biology and suggest new avenues for investigation. All code for the article is publicly available at https://github.com/tansey/deep-dose-response.

Prediction of the dose range for adverse neurological effects of amiodarone in patients from an in vitro toxicity test by in vitro-in vivo extrapolation.

Amiodarone is an antiarrhythmic agent inducing adverse effects on the nervous system, among others. We applied physiologically based pharmacokinetic (PBPK) modeling combined with benchmark dose modeling to predict, based on published in vitro data, the in vivo dose of amiodarone which may lead to adverse neurological effects in patients. We performed in vitro-in vivo extrapolation (IVIVE) from concentrations measured in the cell lysate of a rat brain 3D cell model using a validated human PBPK model. Among the observed in vitro effects, inhibition of choline acetyl transferase (ChAT) was selected as a marker for neurotoxicity. By reverse dosimetry, we transformed the in vitro concentration-effect relationship into in vivo effective human doses, using the calculated in vitro area under the curve (AUC) of amiodarone as the pharmacokinetic metric. The upper benchmark dose (BMDU) was calculated and compared with clinical doses eliciting neurological adverse effects in patients. The AUCs in the in vitro brain cell culture after 14-day repeated dosing of nominal concentration equal to 1.25 and 2.5 µM amiodarone were 1.00 and 1.99 µg*h/mL, respectively. The BMDU was 385.4 mg for intravenous converted to 593 mg for oral application using the bioavailability factor of 0.65 as reported in the literature. The predicted dose compares well with neurotoxic doses in patients supporting the hypothesis that impaired ChAT activity may be related to the molecular/cellular mechanisms of amiodarone neurotoxicity. Our study shows that predicting effects from in vitro data together with IVIVE can be used at the initial stage for the evaluation of potential adverse drug reactions and safety assessment in humans.

Dose-Response Modeling: Extrapolating From Experimental Data to Real-World Populations.

Dose-response modeling of biological agents has traditionally focused on describing laboratory-derived experimental data. Limited consideration has been given to understanding those factors that are controlled in a laboratory, but are likely to occur in real-world scenarios. In this study, a probabilistic framework is developed that extends Brookmeyer's competing-risks dose-response model to allow for variation in factors such as dose-dispersion, dose-deposition, and other within-host parameters. With data sets drawn from dose-response experiments of inhalational anthrax, plague, and tularemia, we illustrate how for certain cases, there is the potential for overestimation of infection numbers arising from models that consider only the experimental data in isolation.

Estrogenic activity of food contact materials-evaluation of 20 chemicals using a yeast estrogen screen on HPTLC or 96-well plates.

Food contact materials (FCM) may contain complex mixtures of estrogenic chemicals. A yeast estrogen screen performed on high performance thin-layer chromatography plates (planar-YES, P-YES) is promising for analysis of such mixtures, as it could allow for better elucidation of effects compared with established methods in microtiter plates. However, the P-YES has not been directly compared with established methods. We compared the performance of a microtiter plate YES (lyticase-YES, L-YES) to P-YES on silica gel HPTLC plates using 17ß-estradiol (E2), 20 chemicals representative of migrants from plastic FCM, and three migrates of coated metal food cans. Effective doses (ED10, ED50) and estradiol equivalencies were calculated for each chemical. Thirteen chemicals had calculable EDs in the L-YES or P-YES, with average EDs 13-fold (range 0.63-36) more potent in P-YES than in the L-YES. Normalized to E2, the median estrogenicity was within 1.5-fold (0.43-8.8) between the assays. Therefore, P-YES was as or more sensitive than L-YES but potencies relative to E2 were comparable between assays. With chromatography, the P-YES detected estrogenicity in coated metal cans, effects that were unmeasurable in L-YES. With the sample preparation methods used in this study, both YES assays are sufficiently sensitive to detect bisphenol A below the specific migration limit for plastic packaging (0.05 mg/kg food). This study demonstrates that P-YES outperforms L-YES because it is more sensitive, provides comparable estradiol equivalents, and circumvents confounding mixture effects. The P-YES will be useful for routine monitoring of FCM and toxicant identification in problematic materials. Graphical abstract.

Implications of nonlinearity, confounding, and interactions for estimating exposure concentration-response functions in quantitative risk analysis.

Recent advances in understanding of biological mechanisms and adverse outcome pathways for many exposure-related diseases show that certain common mechanisms involve thresholds and nonlinearities in biological exposure concentration-response (C-R) functions. These range from ultrasensitive molecular switches in signaling pathways, to assembly and activation of inflammasomes, to rupture of lysosomes and pyroptosis of cells. Realistic dose-response modeling and risk analysis must confront the reality of nonlinear C-R functions. This paper reviews several challenges for traditional statistical regression modeling of C-R functions with thresholds and nonlinearities, together with methods for overcoming them. Statistically significantly positive exposure-response regression coefficients can arise from many non-causal sources such as model specification errors, incompletely controlled confounding, exposure estimation errors, attribution of interactions to factors, associations among explanatory variables, or coincident historical trends. If so, the unadjusted regression coefficients do not necessarily predict how or whether reducing exposure would reduce risk. We discuss statistical options for controlling for such threats, and advocate causal Bayesian networks and dynamic simulation models as potentially valuable complements to nonparametric regression modeling for assessing causally interpretable nonlinear C-R functions and understanding how time patterns of exposures affect risk. We conclude that these approaches are promising for extending the great advances made in statistical C-R modeling methods in recent decades to clarify how to design regulations that are more causally effective in protecting human health.

Model-based estimation of lowest observed effect concentration from replicate experiments to identify potential biomarkers of in vitro neurotoxicity.

A paradigm shift is occurring in toxicology following the report of the National Research Council of the USA National Academies entitled "Toxicity testing in the 21st Century: a vision and strategy". This new vision encourages the use of in vitro and in silico models for toxicity testing. In the goal to identify new reliable markers of toxicity, the responsiveness of different genes to various drugs (amiodarone: 0.312-2.5 [Formula: see text]; cyclosporine A: 0.25-2 [Formula: see text]; chlorpromazine: 0.625-10 [Formula: see text]; diazepam: 1-8 [Formula: see text]; carbamazepine: 6.25-50 [Formula: see text]) is studied in 3D aggregate brain cell cultures. Genes' responsiveness is quantified and ranked according to the Lowest Observed Effect Concentration (LOEC), which is estimated by reverse regression under a log-logistic model assumption. In contrast to approaches where LOEC is identified by the first observed concentration level at which the response is significantly different from a control, the model-based approach allows a principled estimation of the LOEC and of its uncertainty. The Box-Cox transform both sides approach is adopted to deal with heteroscedastic and/or non-normal residuals, while estimates from repeated experiments are summarized by a meta-analytic approach. Different inferential procedures to estimate the Box-Cox coefficient, and to obtain confidence intervals for the log-logistic curve parameters and the LOEC, are explored. A simulation study is performed to compare coverage properties and estimation errors for each approach. Application to the toxicological data identifies the genes Cort, Bdnf, and Nov as good candidates for in vitro biomarkers of toxicity.

Evaluation of the risk of perchlorate exposure in a population of late-gestation pregnant women in the United States: Application of probabilistic biologically-based dose response modeling.

The risk of ubiquitous perchlorate exposure and the dose-response on thyroid hormone levels in pregnant women in the United States (U.S.) have yet to be characterized. In the current work, we integrated a previously developed perchlorate submodel into a recently developed population-based pregnancy model to predict reductions in maternal serum free thyroxine (fT4) levels for late-gestation pregnant women in the U.S. Our findings indicated no significant difference in geometric mean estimates of fT4 when perchlorate exposure from food only was compared to no perchlorate exposure. The reduction in maternal fT4 levels reached statistical significance when an added contribution from drinking water (i.e., 15µg/L, 20µg/L, or 24.5µg/L) was assumed in addition to the 90th percentile of food intake for pregnant women (0.198µg/kg/day). We determined that a daily intake of 0.45 to 0.50µg/kg/day of perchlorate was necessary to produce results that were significantly different than those obtained from no perchlorate exposure. Adjusting for this food intake dose, the relative source contribution of perchlorate from drinking water (or other non-dietary sources) was estimated to range from 0.25-0.3µg/kg/day. Assuming a drinking water intake rate of 0.033L/kg/day, the drinking water concentration allowance for perchlorate equates to 7.6-9.2µg/L. In summary, we have demonstrated the utility of a probabilistic biologically-based dose-response model for perchlorate risk assessment in a sensitive life-stage at a population level; however, there is a need for continued monitoring in regions of the U.S. where perchlorate exposure may be higher.

Estimation of iodine nutrition and thyroid function status in late-gestation pregnant women in the United States: Development and application of a population-based pregnancy model.

Previously, a deterministic biologically-based dose-response (BBDR) pregnancy model was developed to evaluate moderate thyroid axis disturbances with and without thyroid-active chemical exposure in a near-term pregnant woman and fetus. In the current study, the existing BBDR model was adapted to include a wider functional range of iodine nutrition, including more severe iodine deficiency conditions, and to incorporate empirically the effects of homeostatic mechanisms. The extended model was further developed into a population-based model and was constructed using a Monte Carlo-based probabilistic framework. In order to characterize total (T4) and free (fT4) thyroxine levels for a given iodine status at the population-level, the distribution of iodine intake for late-gestation pregnant women in the U.S was reconstructed using various reverse dosimetry methods and available biomonitoring data. The range of median (mean) iodine intake values resulting from three different methods of reverse dosimetry tested was 196.5-219.9µg of iodine/day (228.2-392.9µg of iodine/day). There was minimal variation in model-predicted maternal serum T4 and ft4 thyroxine levels from use of the three reconstructed distributions of iodine intake; the range of geometric mean for T4 and fT4, was 138-151.7nmol/L and 7.9-8.7pmol/L, respectively. The average value of the ratio of the 97.5th percentile to the 2.5th percentile equaled 3.1 and agreed well with similar estimates from recent observations in third-trimester pregnant women in the U.S. In addition, the reconstructed distributions of iodine intake allowed us to estimate nutrient inadequacy for late-gestation pregnant women in the U.S. via the probability approach. The prevalence of iodine inadequacy for third-trimester pregnant women in the U.S. was estimated to be between 21% and 44%. Taken together, the current work provides an improved tool for evaluating iodine nutritional status and the corresponding thyroid function status in pregnant women in the U.S. This model enables future assessments of the relevant risk of thyroid hormone level perturbations due to exposure to thyroid-active chemicals at the population-level.

Bayesian penalized log-likelihood ratio approach for dose response clinical trial studies.

In literature, there are a few unified approaches to test proof of concept and estimate a target dose, including the multiple comparison procedure using modeling approach, and the permutation approach proposed by Klingenberg. We discuss and compare the operating characteristics of these unified approaches and further develop an alternative approach in a Bayesian framework based on the posterior distribution of a penalized log-likelihood ratio test statistic. Our Bayesian approach is much more flexible to handle linear or nonlinear dose-response relationships and is more efficient than the permutation approach. The operating characteristics of our Bayesian approach are comparable to and sometimes better than both approaches in a wide range of dose-response relationships. It yields credible intervals as well as predictive distribution for the response rate at a specific dose level for the target dose estimation. Our Bayesian approach can be easily extended to continuous, categorical, and time-to-event responses. We illustrate the performance of our proposed method with extensive simulations and Phase II clinical trial data examples.

Probabilistic derivation of the interspecies assessment factor for skin sensitization.

An interspecies sensitization assessment factor (SAF) is used in the quantitative risk assessment (QRA) for skin sensitization when a murine-based NESIL (No Expected Sensitization Induction Level) is taken as point of departure. Several studies showed that, on average, the murine sensitization threshold is in good correspondence with that determined in humans. However, on an individual level, the murine and human sensitization thresholds may differ considerably. In this study, the interspecies SAF was quantified by using a probabilistic approach, to be able to take these cases into account. As expected, the geometric means of the probability distributions of murine and human sensitization threshold ratios were close to one, but taking the 95 th percentile of these distributions resulted in an interspecies SAF of 15. By using this value, one is sure that with 95% probability, the sensitization threshold determined in mice does not underestimate the human threshold. It can be concluded that a murine-based NESIL requires the use of an interspecies SAF (of 15) in the QRA for skin sensitization, to correct for the differences between mice and humans. This empirically derived interspecies SAF contributes to refinement of the risk assessment methodology.

A flexible Bayesian approach for modeling monotonic dose-response relationships in drug development trials.

Clinical trials often involve comparing 2-4 doses or regimens of an experimental therapy with a control treatment. These studies might occur early in a drug development process, where the aim might be to demonstrate a basic level of proof (the so-called proof of concept (PoC) studies), at a later stage, to help establish a dose or doses that should be used in phase III trials (dose-finding), or even in confirmatory studies, where the registration of several doses might be considered. When a small number of doses are examined, the ability to implement parametric modeling is somewhat limited. As an alternative, in this paper, a flexible Bayesian model is suggested. In particular, we draw on the idea of using Bayesian model averaging (BMA) to exploit an assumed monotonic dose-response relationship, without using strong parametric assumptions. The approach is exemplified by assessing operating characteristics in the design of a PoC study examining a new treatment for psoriatic arthritis and a post hoc data analysis involving three confirmatory clinical trials, which examined an adjunctive treatment for partial epilepsy. Key difficulties, such as prior specification and computation, are discussed. A further extension, based on combining the flexible modeling with a classical multiple comparisons procedure, known as MCP-MOD, is examined. The benefit of this extension is a potential reduction in the number of simulations that might be needed to investigate operating characteristics of the statistical analysis.

A Dose-Response Study on the Relationship between White Meat Intake and Metabolic Dysfunction-Associated Steatotic Liver Disease (MASLD) in Southern Italy: Results from the Nutrihep Study.

(1) Background: Metabolic dysfunction-associated liver disease (MASLD) is one of the most important causes of liver disease worldwide. Meat consumption is a growing trend and white meat has been shown to have beneficial effects on cardiometabolic risk factors. The aim of this study was to investigate the dose-response relationship between white meat intake and MASLD at survey level in a Southern Italy setting. (2) Methods: This cross-sectional study encompassed 1192 subjects (509 males, 42.7%) without missing data from the second wave of the NUTRIHEP cohort (2014-2016). Adjusted dose-response modeling was employed for statistical analysis; (3) Results: There were 587 subjects with MASLD (49.2%), i.e., 278 males (54.6%) and 309 females (45.2%). By increasing the intake, an unfavorable influence of white meat on MASLD was significantly revealed in females, whereas a protective effect of white meat was detectable in males. Male sex was shown to be involved in other associations in this study, such as influencing the preference for specific foods such as poultry and chicken skin. (4) Conclusions: Our data suggest that white meat does not have a clear-cut independent dose-response effect on MASLD, but sex may be a trigger moderator for age and BMI, with an increasing unfavorable effect of white meat in women, and a favorable effect in men.

Reproductive Capacity, but not Food Consumption, is Reduced by Continuous Exposure to Typical Genotoxic Stressor γ-Rays in the sentinel species Gammarus fossarum.

The long-term impacts of radiocontaminants (and the associated risks) for ecosystems are still subject to vast societal and scientific debate while wildlife is chronically exposed to various sources and levels of either environmental or anthropogenic ionizing radiation from the use of nuclear energy. The present study aimed to assess induced phenotypical responses in both male and female gammarids after short-term continuous γ-irradiation, acting as a typical well-characterized genotoxic stressor that can interact directly with living matter. In particular, we started characterizing the effects using standardized measurements for biological effects on few biological functions for this species, especially feeding inhibition tests, molting, and reproductive ability, which have already been proven for chemical substances and are likely to be disturbed by ionizing radiation. The results show no significant differences in terms of the survival of organisms (males and females), of their short-term food consumption which is linked to the general health status (males and females), and of the molting cycle (females). In contrast, exposure significantly affected fecundity (number of embryos produced) at the highest dose rates for irradiated females (51 mGy h-1) and males (5 and 51 mGy h-1). These results showed that, in gammarids, reproduction, which is a critical endpoint for population dynamics, is the most radiosensitive phenotypic endpoint, with significant effects recorded on male reproductive capacity, which is more sensitive than in females. Environ Toxicol Chem 2024;43:2071-2079. © 2024 SETAC.

The association between histopathologic effects and liver weight changes induced in mice and rats by chemical exposures: an analysis of the data from Toxicity Reference Database (ToxRefDB).

Absolute (ALW) and relative (RLW) liver weight changes are sensitive endpoints in repeat-dose rodent toxicity studies, and their changes are often used for quantitative assessment of health effects induced by hepatotoxic chemicals using the benchmark dose-response modeling (BMD) approach. To find biologically relevant liver weight changes to chemical exposures, we evaluated all data available for liver weight changes and associated liver histopathologic findings from the Toxicity Reference Database (ToxRefDB). Our analysis of 389 subchronic mouse and rat studies for 273 chemicals found significant differences in treatment-related ALW and RLW changes between dose groups with and without liver histopathologic changes. In addition, we demonstrate that chemical treatment-induced ALW and RLW changes can predict the presence of histopathologic findings and inform the selection of biologically relevant liver weight changes for BMD modeling and derivation of toxicity values.

Identifying and Predicting Delayed Mortality with Toxicokinetic-Toxicodynamic Models.

The prevalence of standardized toxicity testing in ecotoxicology has largely obscured the notion that toxicity is a function of time as well. The necessity of considering time is vividly demonstrated by observations of delayed mortality, that is, deaths continue to occur even when animals are no longer exposed to a toxicant. In this contribution, I explore to what extent toxicokinetic-toxicodynamic (TKTD) models from the framework of the General Unified Threshold model for Survival (GUTS) can capture delayed mortality, and to what extent this phenomenon can be predicted from short-term standard tests. I use a previously published data set for fluoroquinolones in Daphnia magna that shows strongly delayed mortality (using immobilization as a proxy for death). The model analysis shows that the GUTS stochastic death models can capture delayed mortality in the complete data set with a long recovery phase, but that the delayed effects would not have been predicted from a 2-day standard test. The study underlines the limited information content of standard acute test designs. Toxicokinetic-toxicodynamic modeling offers a handle on the time aspects of toxicity but cannot always be relied on to provide accurate extrapolations based on severely limited standard tests. The phenomenon of delayed toxicity requires more structured study to clarify its prevalence and impact; I discuss several avenues for further investigation. Environ Toxicol Chem 2024;43:1030-1035. © 2024 SETAC.

On determining the BMD from multiple outcomes in developmental toxicity studies when one outcome is intentionally missing.

Public health concerns over the occurrence of developmental abnormalities that can occur as a result of prenatal exposure to drugs, chemicals, and other environmental factors has led to a number of developmental toxicity studies and the use of the benchmark dose (BMD) for risk assessment. To characterize risk from multiple sources, more recent analytic methods involve a joint modeling approach, accounting for multiple dichotomous and continuous outcomes. For some continuous outcomes, evaluating all subjects may not be feasible, and only a subset may be evaluated due to limited resources. The subset can be selected according to a prespecified probability model and the unobserved data can be viewed as intentionally missing in the sense that subset selection results in missingness that is experimentally planned. We describe a subset selection model that allows for sampling pups with malformations and healthy pups at different rates, and includes the well-known simple random sample (SRS) as a special case. We were interested in understanding how sampling rates that are selected beforehand influence the precision of the BMD. Using simulations we show how improvements over the SRS can be obtained by oversampling malformations, and how some sampling rates can yield precision that is substantially worse than the SRS. We also illustrate the potential for cost saving with oversampling. Simulations are based on a joint mixed effects model, and to account for subset selection, use of case weights to obtain valid dose-response estimates.

Information-theoretic model-averaged benchmark dose analysis in environmental risk assessment.

An important objective in environmental risk assessment is estimation of minimum exposure levels, called Benchmark Doses (BMDs), that induce a pre-specified Benchmark Response (BMR) in a dose-response experiment. In such settings, representations of the risk are traditionally based on a specified parametric model. It is a well-known concern, however, that existing parametric estimation techniques are sensitive to the form employed for modeling the dose response. If the chosen parametric model is in fact misspecified, this can lead to inaccurate low-dose inferences. Indeed, avoiding the impact of model selection was one early motivating issue behind development of the BMD technology. Here, we apply a frequentist model averaging approach for estimating benchmark doses, based on information-theoretic weights. We explore how the strategy can be used to build one-sided lower confidence limits on the BMD, and we study the confidence limits' small-sample properties via a simulation study. An example from environmental carcinogenicity testing illustrates the calculations. It is seen that application of this information-theoretic, model averaging methodology to benchmark analysis can improve environmental health planning and risk regulation when dealing with low-level exposures to hazardous agents.

Pathway network-based quantitative modeling of the time-dependent and dose-response anti-inflammatory effect of Reduning Injection.

ETHNOPHARMACOLOGICAL RELEVANCE: Traditional Chinese medicine (TCM) has extensive healing effects on inflammatory diseases with few side effects. Reduning injection (RDNI), a TCM prescription composed of Lonicera japonica Thunb., Gardenia jasminoides Ellis. and Artemisia annua L., is wildly used for treating inflammatory diseases. However, the mechanism of action of RDNI, like most TCM prescriptions, is unclear due to the complexity of relationships between components and their curative effects. AIM OF THE STUDY: To develop a universal systems pharmacology protocol for mechanism modeling of TCM and apply it to reveal the real-time anti-inflammatory effect of Reduning Injection. MATERIALS AND METHODS: Combined with database mining and references, a regulatory mechanism network of inflammation was constructed. A quantitative model was established afterwards by integrating pharmacokinetic data and two network parameters, namely Network Efficiency and Network Flux. The time-dependent and dose-response relationship of RDNI on the regulation of inflammation was then quantitatively evaluated. ELISA tests were performed to verify the reliability of the model. RESULTS: Three cytokines, namely IL-1ß, IL-6 and TNF-α were screened out to be markers for evaluation of the anti-inflammatory effect of RDNI. An HPLC method for the simultaneous determination of 10 RDNI compounds in SD rat plasma was established and then applied to pharmacokinetic studies. Based on compound activity and pharmacokinetic data, the time-dependent effect of RDNI were quantitatively predicted by the pathway network-based modeling procedure. CONCLUSIONS: The quantitative model established in this work was successfully applied to predict a TCM prescription's real-time dynamic healing effect after administration. It is qualified to provide novel insights into the time-dependent and dose-effect relationship of drugs in an intricate biological system and new strategies for investigating the detailed molecular mechanisms of TCM.

Impact of delineation errors on the estimated organ at risk dose and of dose errors on the normal tissue complication probability model.

BACKGROUND: Normal tissue complication probability (NTCP) models are often based on doses retrieved from delineated volumes. For retrospective dose-response studies focusing on organs that have not been delineated historically, automatic segmentation might be considered. However, automatic segmentation risks generating considerable delineation errors and knowledge regarding how these errors impact the estimated organ dose is important. Furthermore, organ-at-risk (OAR) dose uncertainties cannot be eliminated and might affect the resulting NTCP model. Therefore, it is also of interest to study how OAR dose errors impact the NTCP modeling results. PURPOSE: To investigate how random delineation errors of the proximal bronchial tree, heart, and esophagus impact the estimated OAR dose, and to investigate how random errors in the doses used for dose-response modeling affect the estimated NTCPs. METHODS: We investigated the impact of random delineation errors on the estimated OAR dose using the treatment plans of 39 patients treated with conventionally fractionated radiation therapy of non-small-cell lung cancer. Study-specific reference structures were defined by manually contouring the proximal bronchial tree, heart and esophagus. For each patient and organ, 120 reshaped structures were created by introducing random shifts and margins to the entire reference structure. The mean and near-maximum dose to the reference and reshaped structures were compared. In a separate investigation, the impact of random dose errors on the NTCP model was studied performing dose-response modeling with study sets containing treatment outcomes and OAR doses with and without introduced errors. Universal patient populations with defined population risks, dose-response relationships and distributions of OAR doses were used as ground truth. From such a universal population, we randomly sampled data sets consisting of OAR dose and treatment outcome into reference populations. Study sets of different sizes were created by repeatedly introducing errors to the OAR doses of each reference population. The NTCP models generated with dose errors were compared to the reference NTCP model of the corresponding reference population. RESULTS: A total of 14 040 reshaped structures with random delineation errors were created. The delineation errors resulted in systematic mean dose errors of less than 1% of the prescribed dose (PD). Mean dose differences above 15% of PD and near-maximum doses differences above 25% of PD were observed for 211 and 457 reshaped structures, respectively. Introducing random errors to OAR doses used for dose-response modeling resulted in systematic underestimations of the median NTCP. For all investigated scenarios, the median differences in NTCP were within 0.1 percentage points (p.p.) when comparing different study sizes. CONCLUSIONS: Introducing random delineation errors to the proximal bronchial tree, heart and esophagus resulted in mean dose and near-maximum dose differences above 15% and 25% of PD, respectively. We did not observe an association between the dose level and the magnitude of the dose errors. For the scenarios investigated in this study, introducing random errors to OAR doses used for dose-response modeling resulted in systematic underestimations of the median NTCP for reference risks higher than the universal population risk. The median NTCP underestimation was similar for different study sizes, all within 0.1 p.p.