Tutorial: Understanding the transport, deposition, and translocation of particles in human respiratory systems using Computational Fluid-Particle Dynamics and Physiologically Based Toxicokinetic models

Highlights

• A step-by-step tutorial on how to establish the CFPD-PBPK/TK model is presented.

• Key procedures to guarantee the accuracy of the CFPD-PBPK/TK simulations are emphasized.

• Challenges and research directions to build the next-generation virtual lung model are discussed.

Abstract

Dynamic modeling of how particulate matter (PM) transport, deposit, and translocate from human respiratory systems to systemic regions subject to indoor and outdoor exposures are essential for case-specific lung dosimetry predictions and occupational health risk assessments. Because of the invasive nature and imaging resolution limitations of existing in vitro and in vivo methods, Computational Fluid-Particle Dynamics plus Physiologically Based Pharmacokinetic/Toxicokinetic (CFPD-PBPK/TK) models have been employed to predict the fate of the respirable aerosols for decades. This paper presents a guide on how to use the multiscale CFPD-PBPK/TK models to predict lung dosimetry and systemic translocations quantitatively with 3D subject-specific human respiratory systems. The tutorial aims to clarify possibly ambiguous concepts. The step-by-step modeling procedure should help researchers set up the CFPD-PBPK/TK model accurately, following the standard model validation and verification (V&V) processes, and to bring the lung dosimetry predictions to health endpoints. Starting from the fundamentals of CFPD and PBPK/TK governing equations, the tutorial covers the problem identification, pre-processing, solving, and post-processing steps to perform a computational lung aerosol dynamics simulations, emphasizing on (a) the importance of correct reconstruction and mesh generation of the pulmonary airways; (b) the significance of choosing the appropriate turbulence model to predict the laminar-to-turbulence pulmonary airflow regimes; and (c) the standard (V&V) procedures of submodels in the CFPD-PBPK/TK modeling framework. The tutorial also highlights the deficiencies of current CFPD-PBPK/TK models, clarifies the missing biomechanisms and aerosol dynamics in the respiratory systems that need to be considered to build the next-generation virtual human whole-lung models.

Introduction

Airborne particulate matter (PM), especially the inhalable particles or droplets with diameters less than 2.5 μm (PM_2.5) have been a health concern for many decades (Chuang, 2019). Specifically, exposures to fine and ultrafine PM may have adversely affected human health, causing respiratory diseases, neurological ailments, and ultimately cancer (Bakand & Hayes, 2016; Chuang, 2019). It is widely accepted that the respiratory pathological responses to the inhaled PM in humans and mammals are mainly induced by their deposition and retention in the airway. An association has been demonstrated between inhaled ultrafine particles and various adverse health effects, including increased morbidity and mortality (Bakand & Hayes, 2016). Inhaled nanoparticle deposition in lung airways is more uniformly distributed in human lung airways compared to microparticles (Feng et al., 2018). Due to their small size, nanoparticles can cross biological barriers, such as the air-blood barrier, and therefore can reach cells and tissues normally protected (Som, Wick, Krug, & Nowack, 2011).

Although measurements have been done on particle deposition in human patient tissues as well as in vivo animal studies, there are many ethical constraints that significantly limit the operational flexibilities of experimental investigations. Besides, it used to be difficult to provide the high-resolution data for the researcher to understand the particle dynamics in human lung airways quantitatively and to evaluate the connections among realistic exposure levels, lung uptakes, and health effects. Additionally, the complexity of pre-existing lung diseases and breathing patterns make the personalized exposure health risk assessment even more challenging. Thus, there are knowledge and data gaps of accurate information on how the PM transmits and deposits in the human respiratory system, which is critically needed for more precise health risk studies.

To overcome the drawbacks of those conventional investigating methods, the high-fidelity Computational Fluid-Particle Dynamics (CFPD) models are promising and capable of providing informative high-resolution deposition data based on the natural laws of physics in a noninvasive manner (Dong, Ma, et al., 2018; Feng, Marchal, Sperry, & Yi, 2020; Frederick et al., 2002; Inthavong, Mouritz, Dong, & Tu, 2013; Kleinstreuer & Feng, 2013; Kuga, Ito, Chen, Wang, & Kumagai, 2020; Kuga et al., 2018; Longest & Oldham, 2006; Shimazaki, Okubo, Yamamoto, & Yoshida, 2009; Tian, Ahmadi, Wang, & Hopke, 2012; Tian, Inthavong, Lidén, Shang, & Tu, 2016; Xi & Longest, 2007; Zhao, Feng, Bezerra, Wang, & Sperry, 2019). Indeed, the development of fluid dynamics, computational science, and medical imaging disciplines has spawned a new flourishing interdisciplinary research field in modeling the transport, deposition, and translocation of airborne PM from the indoor environment to human respiratory systems. Complementing bench, in vivo and clinical methods, CFPD models are developed based on first principles, which have the unique potential to unveil the underlying physics and chemistry of airflow and PM in the far-field and the respiratory systems with the “x-ray” visions of how individual PM transport and deposit.

Computational fluid-particle dynamics (CFPD) was developed based on the conservation laws of mass, momentum, and energy of both continuous and discrete phases (Kleinstreuer, 2018). CFPD models have been widely used in aircraft, automobile, and machinery design for many decades, complementing wind-tunnel or other experiments. CFPD models play a critical role in exploring alternate study designs and provide high-resolution data in the noninvasive, cost-effective, and time-saving manner.

In the last several decades, CFPD models have been extensively employed to investigate biofluid mechanics and dynamics problems (Bui, Moon, Chae, Park, & Lee, 2020; Inthavong, 2020; Schroeter, Asgharian, & Kimbell, 2019; Tian & Ahmadi, 2020; Tu, Inthavong, & Ahmadi, 2012). Indeed, fluid dynamics, dissolved species transport, and particle dynamics play significant roles in most of the processes in human respiratory systems. The CFPD methodologies can fill the knowledge gap due to the deficiency of traditional in vitro and in vivo methods, as well as make breakthroughs to pave the way to establish a reliable and efficient numerical investigation framework for occupational exposure risk assessment on a subject-specific level. The employment of CFPD models will be also beneficial in:

• Getting insightful views and deep understandings by visualizing the fundamental aerosol dynamics using different variables, and identifying key parameters that can influence the exposure risks.

• Accelerating the research cycle significantly by spotting the key parameters and ranges for more effective in vitro and in vivo experimental designs.

• Boosting the innovation possibilities in a broader domain by using time-saving and cost-effective numerical efforts.

Integrating the Physiologically based Pharmacokinetic/Toxicokinetics (PBPK/TK) models (Corley et al., 2015; Corley et al., 2012; Frederick et al., 2002; A. ; Haghnegahdar, Feng, Chen, & Lin, 2018), the multiscale CFPD-PBPK/TK models are able to predict further the health endpoints of the inhaled PM and vapors associated with the delivered lung dose, i.e., the translocation from the respiratory system to systemic regions. The lung deposition data and translocation data are direct evidence to evaluate potential health risks as a function of different parameters such as ambient ventilation conditions, relative humidity, breathing conditions, etc (Feng, Marchal, Sperry, & Yi, 2020). It can contribute significantly to solving the problem of both dosimetry and the health effects of inhaled toxic particulate matter and optimal therapeutic particle delivery to predetermined lung sites. Therefore, the CFPD-PBPK/TK models are beneficial to quantify the full process of generation, dispersion, and inhalation of airborne PM to ensure a comprehensive characterization of the relationship of potential health risks to individuals with different ventilation and breathing conditions.

This paper serves as both a tutorial and a review on how to employ CFPD-PBPK/TK models to predict the transport, deposition, and translocation of inhaled aerosols from their generation sources to human respiratory systems and the connected systemic regions (i.e., health endpoints) in the whole body. Specifically, key definitions and governing equations are introduced and discussed in Section 2, focusing on the most widely used CFPD and PBPK/TK modeling framework for computational lung aerosol dynamics simulations. For the tutorial purpose, Section 3 provides a step-by-step guide on how to perform computational lung aerosol dynamics simulations using CFPD and PBPK/TK methods, with details on (a) the importance of corrected reconstruction of surface smoothness of the pulmonary airways from computed tomography/magnetic resonance imaging (CT/MRI) scan data; (b) the significance of choosing the appropriate turbulence model to predict the laminar-to-turbulence pulmonary airflow regimes; (c) the standard validation and verification procedures of submodels in the CFPD-PBPK/TK modeling framework; and (d) the available different multiphase flow models and their distinct advantages on modeling different types of aerosols. As the challenges of the current modeling procedure, the key assumptions and simplifications used in most of the popular computational lung aerosol dynamics models are listed in Section 4, with several on-going research efforts on developing the next-generation computational lung aerosol dynamics models by encompassing more physiologically realistic mechanisms.