4
Assessment of Controlled Human Inhalation Exposure Studies at EPA and Associated Adverse Events
INTRODUCTION
Controlled human inhalation exposure (CHIE) studies at the U.S. Environmental Protection Agency’s (EPA’s) Human Studies Facility in Chapel Hill, North Carolina, are focused on gaining an improved understanding of short-term physiologic and biomarker responses to criteria pollutant exposures, with an emphasis on ozone (O3) and airborne particulate matter (PM). In particular, EPA indicates that a major benefit of these studies is that they provide important information that will inform future reviews of the O3 and PM National Ambient Air Quality Standards (NAAQS).
Nearly all of the EPA CHIE studies were conducted either in healthy young adults or in carefully selected subjects over a wider age range with predispositions to measurable transient and reversible physiologic or biomarker responses, while attempting to minimize the likelihood of adverse events.1,2 Therefore, the objective of EPA CHIE studies has been to produce transient and reversible biomarker or physiologic responses that inform about biologic mechanisms of pollutant effects but do not cause clinical effects. The experimental results of transient outcomes (such as temporary changes in lung function) in response to controlled human exposures have contributed to EPA’s Integrated Science Assessments to support reviews and revisions of the NAAQS, in conjunction with epidemiologic evidence of significant associations between ambient air concentrations of O3 and PM2.5 and adverse health effects, as well as with physiologic or biomarker responses reported in the epidemiologic studies (see Chapter 3). In this regard, the biomarker responses to both short- and long-term exposures to air pollutants have been critical factors in the interpretation of the roles of short-term responses in the initiation and progression of chronic effects.
Some of EPA’s CHIE studies in Chapel Hill involve diluted diesel-engine exhaust (DE) that contains ambient air ultrafine particles (UFPs) composed primarily of elemental carbon (EC). The results of these CHIE studies could be more useful for informing regulatory approaches that focus on source emissions (such as National Emission Standards for Hazardous Air Pollutants for diesel-engine exhaust) as well as for revising the PM2.5 NAAQS. Likewise, wood smoke CHIE study results could be more useful for regulatory approaches that focus on source emissions as well as for affecting the PM2.5 or PM10 NAAQS. Wood smoke is a minor mass component in ambient air in most heavily populated regions of the United States, and the chemical composition of the PM within wood smoke is very different from the ambient air PM in these regions.
__________________
1 As indicated in Chapter 2, the U.S. Department of Health and Human Services defines an adverse event is any untoward or unfavorable medical occurrence in a human subject, including any abnormal sign (for example, abnormal physical exam or laboratory finding), symptom, or disease, temporally associated with the subject’s participation in the research, whether or not considered related to the subject’s participation in the research.
2 As indicated in Chapter 1, EPA CHIE studies cannot be conducted on children or pregnant or nursing women.
THE EIGHT STUDIES IDENTIFIED BY EPA FOR CONSIDERATION BY THE COMMITTEE
Twenty-one CHIE studies had been active at EPA’s Human Studies Facility at some point time from January 2009 to October 2016 (see Table C-1 in Appendix C). Eight CHIE studies were identified by EPA for consideration by the committee (see Table 4-1).3 The committee’s review of those studies (see Appendix C) included considerations of the testability of the hypotheses, appropriateness of the study design and outcome measures, and potential value of the results. The reviews are summarized below.
The pollutants included in the eight studies were O3 alone in two of them, sequential exposures to O3 alone and nitrogen dioxide (NO2) alone in one, a mixture of O3 plus DE in one, concentrated PM2.5 in Chapel Hill ambient air in two, concentrated Chapel Hill ambient UFPs in one, and wood smoke in one. For the recent studies involving DE and wood smoke, the basic characterization of the exposures to air pollution mixtures in the DE and wood smoke studies was in terms of PM mass concentrations, with some data on the particle number concentration (primarily attributable to EC), as well as some characterization of the other hazardous air pollutants that were also present within the mixtures.
A common objective of the studies is to contribute to a body of knowledge about the potential health effects from exposure to air pollutants, which would add to the results of toxicologic and epidemiologic studies (see Chapter 3). It is important to note that these studies were not intended to reflect the variability of ambient pollutant concentrations in the real world, and that the relatively small number of subjects involved in each study tends to limit the generalizability of the results, by themselves, to a broader population. However, the CHIE study findings could add relevant new knowledge to future NAAQS reviews. For example, the findings of CHIE studies of PM2.5 extracted from Chapel Hill ambient air could add new insights concerning human responses to PM2.5 in Chapel Hill ambient air as well as possible insights to observed responses to ambient air exposures in other U.S. regions having very different PM2.5 chemical compositions.
Cardiopulmonary Responses to Exposure to Ozone and Diesel Engine Exhaust with Moderate Exercise in Healthy Adults (DEPOZ)
Background: Assessing the components in air responsible for particular health effects is difficult because ambient air pollution is a complex mixture of gases and PM. O3 and DE are often important components of those complex mixtures. It is stated in the Institutional Review Board (IRB) application that it is not known whether coexposure to both O3 and DE, as would occur when humans are exposed to polluted ambient air, can induce additive or synergistic effects, and also whether exposure to DE, or DE with O3, can alter a subsequent exposure to O3. This study was designed to examine whether coexposures to O3 and DE, at concentrations in the upper range of those encountered in urban settings, can induce additive or synergistic effects, and whether a previous DE exposure can alter a response to subsequent O3 exposure.
Hypothesis: There were three specific hypotheses for this study.
- Healthy adults exposed either to DE and O3 or DE alone on one day will not experience a significant decrement in pulmonary function in response to an O3 exposure on the second day, relative to exposure to O3 alone;
- A coexposure to DE and O3 on one day would cause significant cardiopulmonary responses to an O3 exposure on the next day; and
- Two consecutive days of O3 exposure would affect cardiovascular responses.
__________________
3 An application for institutional review board approval and a consent to participate in a research study for each CHIE study was provided by EPA on November 19, 2014.
TABLE 4-1 Descriptive Information on Eight CHIE Studies
| Study Name (Pollutant) | Health Condition of Subjects | No. of Subjects Exposed (# male; # female) | No. of Subjects Planned to be Exposed (# male; # female) | Age of Subjects | IRB-Approved Exposure Duration | IRB-Approved Maximum Exposure Conc. | Actual Exposure Conc. |
|---|---|---|---|---|---|---|---|
| DEPOZ Diesel-engine exhaust (DE) and ozone (O3) |
Healthy | M: 11 F: 8 |
Study Completed | Avg: 27.8 Range: 22-53 |
Two 2-hour exposures: Day1: DE only, O3 only, DE+O3, or clean air Day 2: O3 only |
DE, 300 μg/m3 O3, 300 ppb |
DE (µg/m3) Avg: 295 Range: 244-326 O3 (ppb) Avg: 300 Range: 300-300 |
| ENDZONE [Ozone (O3) and Nitrogen dioxide (NO2)] |
Healthy | M: 17 F: 14 |
Study completed | Avg: 28.9 Range: 8-41 |
Four 2-hour sessions: - Clean air + O3 - NO2 + O3 - Clean air + NO2 - O3 + NO2 |
O3, 300 ppb NO2, 500 ppb |
O3 (ppb) Avg: 300 Range: 299-300 NO2 (ppb) Avg: 500 Range: 500-525 |
| GEMINOZ [Ozone (O3)] |
Healthy | M: 8 F: 4 |
50 total | Avg: 26.7 Range: 20-36 |
Two 2-hour sessions: 1 for ozone and 1 for clean air | 300 ppb | O3 (ppb) Avg: 300 Range: 299-300 |
| KINGCON [Particulate matter (PM) < 2.5 µm] |
Mild asthma | M: 2 F: 14 |
Study Completed | Avg: 49.8 Range: 45-58 |
Two 2-hour exposures: 1 PM and 1 clean air exposure | Up to 600 μg/m3 | PM2.5 (µg/m3) Avg: 236 Range: 38-579 |
| OMEGACON (PM < 2.5 µm) |
Healthy | M: 8 F: 22 |
Study Completed | Avg: 57.9 Range: 51-72 |
Two 2-hour exposures: 1 clean air (Day 1) and 1 PM (Day 2) fish oil, olive oil, or no oil 4 wks prior |
Up to 600 μg/m3 | PM2.5 (µg/m3) Avg: 278 Range: 83-470 |
| SOZIAL (O3) |
Healthy; 4-point perceived stress symptom score <2 or >6 | M: 13 F: 20 |
40 total | Avg: 27.0 Range: 21-33 |
Two 2-hour sessions 1 ozone and 1 clean air | 300 ppb | O3 (ppb): Avg: 300 Range: 300-300 |
| WOODSIE [wood smoke particles (WSPs)] |
Healthy | M: 17 F: 22 |
Study Completed | Avg: 27.5 Range: 18-38 |
One 2-hour session: - WSP + Live attenuated influenza virus (LAIV), or - Clean Air + LAIV |
WSP, 500 µg/m3 LAIV,1 ml FluMist |
WSP (µg/m3) Avg: 488 Range: 435-526 |
| XCON [Ultrafine particles (UFPs)] |
Metabolic syndrome | M: 13 F: 22 |
Study Completed | Avg: 47.3 Range: 26-70 |
Two 2-hour exposures: 1 (UFPs) and 1 (clean air) | Up to 600,000 UFP/cm3 | # UFP/cm3: Avg: 211,462 Range: 17,295-563,912 UFP (µg/m3) Avg: 118 Range: 35-359 |
Study Design and Outcome Measures: A randomized crossover single-blind study design was used, involving healthy subjects between the ages of 18 and 55 in four exposure regimes:
Regime 1. Combined exposure to DE and O3 (Day 1); exposure to O3 alone (Day 2).
Regime 2. Exposure to O3 alone (Day 1); exposure to O3 alone (Day 2).
Regime 3. Exposure to DE (Day 1); exposure to O3 alone (Day 2).
Regime 4. Exposure to clean air (Day 1); exposure to O3 alone (Day 2).
Each subject is assigned randomly to a sequence of all four exposure regimes. Each regime is separated by at least 13 days. Subjects are exposed while undergoing moderate intermittent exercise. The DE exposure concentration was 300 µg/m3, and O3 exposure concentration was 300 ppb. There is one follow-up visit approximately 18 hours after the last exposure.
The principal outcomes measured for pulmonary function are FEV1 (forced expiratory volume of air that can be forcibly blown out in one second) and FVC (forced vital capacity, the volume of air that can forcibly be blown out after full inspiration). Other primary measured end points are heart rate variability (HRV), blood inflammatory factors (such as IL-6), blood clotting factors (such as fibrinogen), and susceptibility factors [such as the genotype glutathione-S-transferase M l (GSTM1) null].
Results: According to Madden et al. (2014), the study results suggest that the combination of O3 and DE exposure can alter respiratory responses in a greater than additive manner, and O3-induced pulmonary function decrements are greater with a prior exposure to DE compared to a prior exposure to filtered air.
In addition, Stiegel et al. (2015) reported that samples of blood, exhaled breath condensate, and urine collected from DEPOZ study subjects were used to develop a method for characterizing and interpreting changes in the expression of cytokines in biological media. Such changes are considered to be indicative of an inflammatory response to external stressors.
Discussion: In general, the study was designed adequately for its stated goals. The protocol described in EPA’s application for IRB approval of the DEPOZ study described how the hypothesis on pulmonary responses could be tested by analyzing data collected on pulmonary function. However, the protocol to test the hypotheses concerned with cardiovascular responses is not as specific as those involving pulmonary function, and EPA’s application does not state how data on cardiovascular effects would be analyzed. The sample size calculated based on statistical power considerations focused on changes in pulmonary function responses. Power for testing other responses is not clear. Thus, it is not clear whether the hypothesis involving cardiovascular responses was testable. Given the size of the expected changes in GSTM1 and the variability of responses, the sample size was probably too small to reach definitive conclusions about the effect of the GSTM1 null genotype on the effects of O3 exposure on pulmonary function.
The comparative exposures to DE in ambient air described in EPA’s application for IRB approval either were not documented or were from a simulation study, in which the exposures were much briefer than exposures of the DEPOZ study subjects.
Like all studies of limited and prescribed exposures to subjects, many questions remain. Effects upon the very young and old and upon those with existing health conditions were not investigated. The study involved a limited number of exposure conditions, none of which are typical of real-world exposures, so using the results of the study, by themselves, to predict responses in real-world situations would be limited.
The findings of the DE exposures might have marginal relevance to future reviews of the PM2.5 NAAQS, as EC and organic carbon (OC) can represent substantial fractions of PM2.5 mass. However, interpretations of the results are limited by the absence of knowledge of the contributions of NO2, EC, OC, and the numerous other hazardous air pollutants within the DE mixture to the responses attributable to the exposures. Furthermore, the exhausts from newer diesel engines produce only small fractions of the hazardous air pollutants, EC, and OC compared to those emitted from diesel engines of older models (Hes
terberg et al., 2011). Between 1999 and 2010, total carbon (OC and EC) generally decreased in both urban and rural areas, with the strongest trends in the western States (Hand et al., 2013). OC decreased by 3.3% to 6.5% per year, while EC, which is almost all attributable to diesel engine emissions, declined by 3.2% to 7.8% per year (Blanchard et al., 2013).
Effects of Sequential Exposure to Nitrogen Dioxide and Ozone in Healthy Adult Human Volunteers (ENDZONE)
Background: As different pollutants reach peak ambient concentrations at different times during the day, it is important to consider whether exposure to one pollutant sensitizes an individual so that a response to a subsequent exposure is augmented. The purpose of this study is to determine whether exposure to O3 or NO2 enhances cardiopulmonary effects of healthy adults in response to a subsequent exposure to the other pollutant, relative to exposure to either pollutant without a subsequent exposure.
Hypothesis: The study is designed to test two general hypotheses.
- Preexposure to a relatively low concentration of NO2 will sensitize individuals to a subsequent O3 exposure and lead to greater changes in cardiopulmonary function compared to O3 exposure preceded by clean air exposure, and
- Preexposure to O3, at a concentration that has been previously associated with small changes in cardiopulmonary function, will prime individuals to have a greater response to NO2 compared to preexposure to clean air.
Study Design and Outcome Measures: Healthy study subjects are involved in four exposure regimens:
Regimen 1: Exposure to clean air followed by O3.
Regimen 2: Exposure to NO2 followed by O3.
Regimen 3: Exposure to O3 followed by NO2.
Regimen 4: Exposure to clean air followed by NO2.
Each regimen involves exposures and intermittent, moderate exercise on two consecutive days with a third follow-up day. Each study participant is exposed randomly to all four exposure regimes, and each regimen is separated by at least 13 days. Primary outcome measures are cardiac electrophysiology, pulmonary function, and pulse-wave analysis to measure arterial stiffness. Secondary measures include analysis of blood clotting/coagulation factors and other soluble factors present in plasma.
Results: Study results were not available when this report was being prepared.
Discussion: The study was designed appropriately to meet the stated goals of the study. However, the inappropriate temporal sequence of O3 and NO2 exposures precludes the likelihood of effectively addressing O3 NAAQS issues. As noted in EPA’s submission for IRB approval, the real-world sequential exposures involve peak morning exposures to NO2 followed by peak early afternoon exposures to O3. Why then select the second of the sequential 2-hour controlled inhalation exposures 24 hours later? Exposure to NO2, followed a day later by O3, could be informative, although less so than would be a 2-hour delay between the two exposures on the same day, as they most often occur in ambient air. The O3, followed a day later by NO2, is not likely to be very informative with regard to responses to real-world exposures. Both O3 and NO2 exposures have been included in previous CHIE studies without clinically adverse effects. For a study of the physiologic effects of sequential exposures of inhalation exposures to O3 and NO2, there was little justification provided for the temporal sequences.
Epigenetic Effect Modifications with Ozone Exposure on Healthy Volunteers (GEMINOZ)
Background: Epigenetics refers to mechanisms not involving changes in DNA sequence that influence gene expression. Researchers have explored how changes in the epigenome might affect a person’s susceptibility to effects caused by air pollution exposures. However, separating the role of genetics from the effect of epigenetic factors presents a substantial challenge. One approach is to study monozygote (MZ) twins, which have identical genetic sequences and different epigenomes. By involving MZ twins as study subjects, effects attributable to epigenetics can be explored separately from the effects of genetics. The study is intended to determine whether differences in baseline epigenetic profiles between subjects are associated with responsiveness to O3 exposure and whether O3 exposure itself causes acute changes in a subject’s epigenome.
Hypothesis: Epigenetic factors in healthy individuals or individuals with the same genetic makeup (that is, identical twins) affect the responsiveness to inflammation following ozone exposure.
Study Design and Outcome Measures: Healthy MZ twins and healthy nontwin subjects are exposed during two sessions separated by an interval of about 14 days, to clean air on one day and O3 on the other day, and involving intermittent exercise. Primary outcome measures include pulmonary function, lung inflammation, and epigenetic changes as indicated by bronchoalveoalar lavage.
Results: Study results were not available when this report was being prepared.
Discussion: This study has strong biologic justification, outcome measures are validated, and the sample size estimate is adequate for the intended power. However, the use of MZ twin studies adds another factor that may limit the applicability of the results to the general population.
Mechanisms by which Air Pollution Particles Exacerbate Asthma in Older Adults with Mild Asthma (KINGCON)
Background: As discussed in EPA’s application for IRB approval of this study design, previous observational studies have indicated that asthmatics with the null genotype for GSTM1 have increased susceptibility to O3 and DE exposures. As those previous observational studies focused on children, there remains a lack of evidence on the effects of inhaled pollutants on older adults with asthma in relation to the GSTM1 genotype.
Hypothesis: Older adults (45-65 years old) with mild asthma who have a GSTM1-null genotype will have a greater inflammatory response to PM exposure than do older adults with mild asthma who are GSTM1 sufficient.
Study Design and Outcome Measures: This study compares the response of older adults with mild asthma that are GSTM1-null and GSTM1-positive to fine particulate matter (PM2.5) and UFPs. Subjects are randomly exposed to both clean air and concentrated PM2.5 and UFPs, with exposures separated by a minimum of 2 weeks. Responses of primary interest include changes in FVC and FEV1 immediately after the exposure and acute increases in airway neutrophils (as reflected in recovered bronchoalveolar lavage samples) 24 hours after exposure. Subjects are allowed to participate in the exposure study even if they decline or are excluded from bronchoscopy.
Results: Study results were not available when this report was being prepared.
Discussion: The design and methods are appropriate for the stated goals of the study. However, consideration of the relevance of the study results to all mild asthmatics needs to take into account the variability of exposures to ambient PM and ambient gaseous components (such as O3 and NO2).
The description of the study protocol is too long and too complicated with regard to informing potential study subjects. Because the subjects have mild asthma, discussion of risks for normal (non asthmatic) subjects is not entirely relevant. The background section of the EPA application for IRB approval reported that 736 normal nonasthmatics had a rate of <0.1% complications from bronchoscopy. The likelihood of complications for the group under study is expected to be greater. Providing the complication rates of a subpopulation with reactive airways disease would be more realistic and helpful to the participants.
Cardio-protective effects of Omega-3 Fatty Acids Supplementation in Healthy Older Subjects Exposed to Air Pollution Particles (OMEGACON)
Background: As indicated in EPA’s application for IRB approval for this study, short-term exposures to ambient PM at elevated concentrations can lead to cardiac arrhythmias, worsening heart failure, and acute atherosclerotic/ischemic cardiovascular complications, particularly in certain at-risk groups. Reactive oxygen species produced in humans after exposure to PM have been implicated as a potential mechanism for adverse effects of air pollutants, and genetic polymorphisms of glutathione S-transferases (GSTs) have been shown to participate in the antioxidant defenses to air pollutants. Also, studies show omega-3 fatty acids have the potential to reduce cardiovascular (CV) effects, including arrhythmias, through a reduction in oxidative stress. The goal of this study was to determine if fish oils/omega-3 fatty acids would reduce or mitigate the respiratory and CV effects of PM.
Hypothesis: The study is designed to test these hypotheses.
- PM exposures cause adverse CV effects and omega-3 fatty acid supplementation pretreatment would attenuate the adverse CV effects.
- Healthy older subjects with a GSTM1-positive genotype have lower CV risk than subjects with GSTM1-null genotype when exposed to PM.
Study Design and Outcome Measures: In a randomized, double-blind study, involving older subjects (age 50-75 years), subjects are given either fish oil (containing omega-3 fatty acids) or olive oil supplements for 4 weeks. After that treatment, each subject is involved in a 2-day exposure sequence: clean air on the first day, and PM2.5 and UFPs on the second day.
Primary outcome measures are heart rate variability measurement and peripheral venous blood markers for specific and nonspecific immune responses. Secondary measures are endothelial cell function (as measured by flow-mediated dilation of the brachial artery) and pulmonary function measurements.
The sample size is based on the potential to detect a change of 0.13 units in brachial artery diameter measured by ultrasound, as observed in an earlier pilot study conducted at the EPA facility in Chapel Hill, North Carolina.
Results: Fish oil containing omega-3 fatty acids blunted the changes in heart rate variability and QT-interval prolongation on electrocardiograms associated with PM exposure (Tong et al., 2012). Also, dietary supplementation with olive oil, but not fish oil containing omega-3 fatty acids, blunted the negative impact of PM exposure on endothelial cell function as measured by flow-mediated dilation of the brachial artery. In addition, olive oil treatment was associated with increased levels of a fibrinolysis marker (tissue-type plasminogen activator) after PM exposure.
Discussion: The strengths of this study include the biologic significance of the question of whether the omega-3 fatty acid treatment could potentially reduce the cardiac-related effects of PM exposure, and certain biologic markers of inflammation. The weaknesses include the small sample size, only one dose of
omega-3 fatty acids, and only one PM exposure, each of which can limit the generalizability to more realistic situations.
This study reinforced the biological plausibility of effects of PM on CV; thrombolytic systems observed in toxicologic and epidemiologic studies and antioxidant treatment appear to modify PM response. Additional studies would be needed to better define the antioxidant effects of omega-3 fatty acids in mitigating CV effects in sensitive individuals exposed to PM.
The Interaction of Social Factors with Air Pollution (SOZIAL)
Background: As discussed in EPA’s request for IRB approval to modify the previously approved study protocol,4 acute and chronic exposures to ambient concentrations of O3 are associated with asthma and other health effects. Also, social factors such as psychologic stress are considered to be important contributors to asthma outcomes. A greater understanding of the effects of psychosocial stress on health responses to air-pollutant exposures would help to understand which groups and individuals are at increased risk from air pollution.
Hypothesis: Social factors such as psychologic stress modify how people respond to air pollution.
Study Design and Outcome Measures: A randomized double-blind crossover study design is used to compare the cardiopulmonary responses of two groups of healthy adults with different levels of perceived chronic stress to O3 and clean air. Subjects who score less than 2 on the 4-point Perceived Stress Scale (PSS4) and subjects with PSS4 values greater than 6 are randomly exposed to clean air and on a separate visit to O3, with exposures separated by a minimum of 13 days. All exposures are performed while subjects perform moderate intermittent exercise.
HRV is the primary outcome measure. Possible secondary measures include pulmonary function, analysis of blood clotting/coagulation factors, biomarkers of stress, cognitive function, pulse-wave analysis, and analysis of soluble factors present in plasma.
Results: Study results were not available when this report was being prepared.
Discussion: The design of the study is appropriate, except for some lack of clarity in the randomization procedure. Two regimes are described: one with low PSS4 and the other with high PSS4. However, it was not clear whether the randomization to ozone exposure or to clean air (placebo) was carried out within arms of PSS4 or for all volunteers together.
This study has the potential to contribute novel scientific information, largely because there have been few, if any, CHIE studies that have examined the effect of psychologic stress on O3 exposure and biologic responses. However, other studies will have to be performed for conformation, and to explore more exactly the causal pathways involved.
Effects of Wood Smoke Particles on Influenza-Induced Nasal Inflammation in Normal Volunteers (WOODSIE)
Background: Wood smoke (from sources such as wildfires) is an important source of ambient PM. Wood fires used for indoor heating, ambience, or cooking contribute to indoor air pollution. Influenza virus infections are an important cause of morbidity and mortality in the United States and worldwide. The effects of WSP exposure on subsequent responses to infectious agents including live attenuated influenza virus (LAIV) has not been previously studied in a controlled setting. A finding that exposure to
__________________
4 The request proposed six changes, including the addition of study personnel and an increase in the venous blood sampling amount from 25 ml to 30 ml for the study to accommodate a change in one of the proposed assays.
WSP alters influenza infection would have broad public health implications. Also, physiologic and biomarker changes, such as those considered in this study, could potentially be used for population-based studies.
This study is focused on the pathophysiology underlying the association between exposure to PM and the likelihood of a viral infection and the response to that infection. The study also is designed to test novel assays of granulocyte activation and lipid mediator activation which have not previously been used in this type of research.
Hypothesis: Exposure to WSP enhances influenza virus–induced granulocyte and natural killer cell activation, via hyaluronic acid-mediated effects on interferon gamma production. Oxidant stress and viral replication may also be affected.
Study Design and Outcome Measures: A randomized, placebo-controlled study compares nasal lavage fluid granulocyte responses to LAIV administered after either WSP or clean air, in normal healthy volunteers. Subjects receive either WSP or placebo (clean air), followed by a standardized dose of LAIV and serial postinfection sampling of nasal lavage fluids, nasal biopsy, and blood.
Results: Study results were not available when this report was being prepared.
Discussion: Strengths of the study are the public health importance of the research question, and the experimental design with a robust array of end points that is appropriate for the pathways under investigation. A strength of the design is the inclusion of repeated time points, and sampling of nasal lavage fluids, nasal biopsy material, and blood. Time points are appropriate (0, 1, 2, 7-10, and 21-28 days), covering the anticipated duration of infection and assessing acute and subacute time points. There is a clean air comparator, allowing each subject to serve as his or her own control. The study results might point to mechanisms that explain the association between PM exposure and increased risk of respiratory infection. Another strength of the study is the partnering with other laboratories, leveraging the study results to address additional end points (granulocyte activation and lipid mediator activation) that are highly relevant to the toxicologic pathways involved in this response.
Physiologic Changes in Adults with Metabolic Syndrome Exposed to Concentrated Ultrafine Chapel Hill Air Particles (XCON)
Background: Metabolic syndrome (MeS) refers to a collection of risk factors (such as high blood pressure) that increase the likelihood of developing cardiovascular disease (CVD) or type-2 diabetes mellitus (DM). As discussed by Devlin et al. (2014), clinical CVD and DM have been shown to increase the susceptibility to clinical health effects of ambient PM exposure. Also, some studies suggest that MeS might increase susceptibility to inflammatory or physiologic effects of pollution. The rationale for this study was to examine biologic responses to concentrated ambient UFP exposure in patients with MeS. Also, there is considerable interest in the potential role of ultrafine particles in causing adverse health effects, relative to PM2.5.
Hypothesis: UFP exposure to individuals with MeS will result in changes in endothelial response as indicated by flow-mediated dilation of the brachial artery and various heart rate variability and blood biomarkers.
Study Design and Outcome Measures: This study is focused on evaluating subjects with MeS between the ages of 25 and 70 years. A double-blind study is used in which each participant is exposed to clean filtered air and air containing concentrated UFPs, in randomized order. A crossover design is used, comprising two treatments, two sequences, and two periods. Each exposure is separated by 2 weeks. Repeated measurements were taken over a 24-hour period. Outcome measures include flow-meditated dilation (brachial artery ultrasound) and heart rate variability, peripheral venous blood samples, specific and
nonspecific immune responses (cytokines and C-reactive protein), coagulation factors (von Willibrand factor, factor IX, fibrinogen, thrombin, vasoactive factors), and soluble components of PM (transitional metals).
Results: Results of this study indicate that exposure to UFPs did not cause measureable changes in brachial artery diameter or blood pressure. However, UFP exposure caused changes in cardiac electrophysiologic repolarization, heart rate variability, and vascular biomarkers of inflammation and fibrinolysis (Devlin et al., 2014). It is not clear if the findings are related to UFP size or number of particles per unit volume of air.
Discussion: The research question is well focused and the hypothesis is clear and testable. The subgroup chosen for this study is considered to be at a higher risk than the general population for CVD and DM, as well as considered to have increased risk of health problems from exposure to ambient particle pollution. Although the sample size was low, power was sufficient to test primary end points. The observed biologic and physiologic responses associated with the UFP exposures could lend biologic plausibility to observational studies finding associations of UFP with clinical outcomes. The time points for measurements following exposure, 1 and 20 hours, are sufficient to see acute and prolonged effects on CV, inflammatory, and endothelial responses.
UFP exposure caused changes in the vascular markers of inflammation and fibrinolysis and UFP exposure might affect some biologic pathways through oxidative stress processes. Most changes were observed in individuals with the enzyme GSTM1. Because the study age range of study subjects is 25-70 years and given that MeS is more prevalent in individuals older than 50 years, studies involving more subjects older than 50 years would be needed to confirm those results.
The findings of the UFP controlled-exposure studies could add new knowledge for addressing whether to establish a future NAAQS for UFPs. Such an evaluation would also need to be informed by data from studies of responses to specific component concentrations of the UFPs. Another key consideration is whether any future UFP NAAQS should rely on particle number concentration as an index of exposure, instead of a mass concentration.
EVALUATION OF EVIDENCE FOR ADVERSE EVENTS RESULTING FROM PARTICIPATION IN A CHIE STUDY
As mentioned previously, EPA’s CHIE studies are not intended to induce adverse effects that would require medical intervention in study subjects. The agency strives to establish and maintain experimental conditions and involve human subjects with characteristics that reflect the potential to identify and evaluate physiologic or biologic response to pollutant exposures. The kinds of biologic responses (such as inflammation) or biomarkers considered in those studies are transient and expected to dissipate within a few days.
Consideration of risks associated with participation in CHIE studies focuses on the probability of the occurrence of a serious adverse event (such as asthma attack, myocardial infarction, or death), not the transient and reversible biomarker or physiologic responses that these short-term inhalation exposure studies are designed to examine (such as a small decrement in pulmonary function). Risks of serious adverse events temporally associated with the subject’s participation in a CHIE study might be affected by one or more of the following:
- Air-pollutant exposures occurring independently from the CHIE study, several days prior to or during the multiday experimental protocols,
- Intended pollutant exposures during the experiments,
- Preexisting medical conditions or sensitivities of subjects to the CHIE study pollutant(s),
- Other experimental procedures during the CHIE study (such as blood sampling or bronchoscopy), and
- Chance occurrences of pathophysiologic events (such as a serious adverse cardiac or pulmonary event), although unrelated to air-pollutant exposures, that might happen to subjects during the CHIE study.
To characterize risks to study participants in the eight CHIE studies reviewed by the committee, EPA compared the exposures to the various pollutants used in these eight CHIE studies to exposure scenarios associated with these pollutants that might be experienced by various groups of people living in the United States. However, those comparison scenarios are mostly unsupported by ambient monitoring data. The possibility that, in some cases, the risks in the CHIE studies could be greater than those in the comparison scenario are discussed in detail in Chapter 6.
The potential adverse outcomes are described in the protocols of the eight studies in a consistent manner, with the possible exception of DE in DEPOZ, where potential adverse outcomes are described physiologically (slight alterations in blood clotting, oxygen diffusion capacity, and changes in heart rate variability), whereas other studies describe such outcomes in terms of symptoms (such as irritation to the nose, eyes, throat, and airways; pain on deep inspiration and cough) (see Table 4-2). In all of the protocols, the expected effects are considered to be transient and reversible.
In addition to the hazards from the inhalation exposure to the pollutant, volunteers face risks posed by the bronchoscopy that is used to measure some of the effects of the exposure. Transnasal fiberoptic bronchoscopy was used in the GEMINOZ and KINGCON studies. Medical screening is designed to exclude subjects who might be at greater risk from the procedure. Also, at least one physician and several nurses are on site at all times during a CHIE study to provide emergency medical care. Subjects are removed from studies if their cardiac or lung functions deviate from expected patterns. Standard blood chemistry panels are run at various time points before, during, and after most studies. As a result EPA has experienced an overall complication rate of less than 0.1% related to bronchoscopies at the EPA Human Studies Facility Laboratory. Stahl et al. (2015) report that since the introduction of fiberoptic bronchoscopy in the 1960s, published rates of complication (for example, airway trauma, bleeding, and vomiting) have ranged from <0.1 to 11%. Mortality rates were between 0 and 0.1%.
TABLE 4-2 Potential Health Outcomes Described in EPA CHIE Study Protocols
| Exposure Agent | CHIE Study | Potential Health Outcome from Exposure |
|---|---|---|
| Particulate Matter (Ambient Air and Diesel Engine Exhaust) | KINGCON, OMEGACON, XCON | Chest pain, mild dyspnea, headache, cough, wheeze, and decrements in pulmonary function. All of these effects are expected to resolve a few hours after exposure. |
| Ozone | GEMINOZ, ENDZONE, SOZIAL | Decrements in pulmonary function, irritation to nose, eyes, throat, and airways, chest pain, and cough, all of which resolve a few hours after exposure. There might be an inflammatory reaction lasting 24 hours after exposure and participants may have an increased chance of getting a respiratory infection. |
| Nitrogen Dioxide | ENDZONE | Decrements in pulmonary function, mild irritation to nose, eyes, throat, and airways, chest pain, and cough, all of which resolve a few hours after exposure. |
| Diesel Engine Exhaust | DEPOZ | Decrements in pulmonary function, slight alterations in blood clotting, pulmonary function, and changes in heart rate variability (HRV). |
| Wood Smoke | WOODSIE | Mild mucosal irritation to eyes and nose.a |
aThe consent form, dated September 24, 2014, for WOODSIE indicates that “No adverse effect on lung function or cardiovascular stability has been reported during experimental WSP exposures in humans” (page 6).
The events listed in Table 4-3 include adverse events reported to the UNC IRB for all CHIE studies conducted at the EPA Human Studies Facility Laboratory from January 2009 to February 2015 (not only the eight CHIE studies identified by EPA for consideration by the committee). The table also indicates whether the occurrence of an event led to a change in EPA’s CHIE study protocol. The adverse events reports provided by EPA from four CHIE studies (ENDZONE, DEPOZ, OMEGACON, and XCON), and from an additional study (CAPTAIN), support the low short-term risk characterization for these studies. Most of the time, there were no reported incidences of the potential adverse events listed in the protocols. One subject, exposed to O3, experienced chest discomfort on deep inspiration, and the subject was retained for monitoring until pulmonary function returned within 5% of the preexposure measurement. Another subject who had developed a persistent cough was followed up over 3 months. The follow-up activities during that period included the subject being seen by an EPA physician, receiving medication for 1 week, receiving emails and phone calls from an EPA nurse, and being scheduled for an appointment at the UNC Ambulatory Care Center Pulmonary Clinic (EPA, 2014a).
A different subject experienced an episode of bradycardia during a clean air exposure in OMEGACON. Another subject, exposed to O3, was found to have a cardiac arrhythmia on the follow-up day.
The reported unexpected serious adverse event of paroxysmal atrial fibrillation experienced by a study subject a very short time after being exposed to concentrated ambient particles during the XCON CHIE study was appropriately noted on monitoring and, as reported in the published case report, the subject’s response reverted to normal sinus rhythm spontaneously without clinical sequelae approximately 2 hours after cessation of the controlled exposure (Ghio et al., 2012). The subject was observed overnight in the hospital following the observed event.
For studies involving elderly subjects or subjects that have existing conditions, like asthma, risks might become more substantial compared to risks for younger, healthier participants. For example, in the studies involving O3, the researchers disclose to subjects, without elaboration, that epidemiologic reports have shown that elderly people may get sick or even die in high-O3 environments. Similarly, certain genetic characteristics might put individuals at higher or lower risk. There is a possibility that exposure to some of these pollutants could cause an asthma attack in previously undiagnosed persons or exacerbate a known asthmatic condition. Likewise, those exposures might uncover an unidentified preexisting cardiac condition or exacerbate a known condition. Both KINGCON (subjects with mild asthma) and XCON (subjects with metabolic syndrome) involve subpopulations that may already be particularly at risk. But if studies were to include more at-risk subpopulations, where EPA’s experience is less likely to be predictive, the likelihood of adverse outcomes could increase. This might complicate the risk–benefit calculus in determining whether such studies, or studies that involve subpopulations with even greater risk, should be conducted. However, on the other hand, these at-risk subpopulations are likely to receive the greatest benefit from the knowledge that EPA gleans from such studies, when this knowledge is used in revising air-quality standards for the United States.
ADVERSE EVENT REPORTING
EPA defines and reports adverse events according to 2007 guidance from the Office for Human Research Protections of the Department of Health and Human Services (OHRP, 2007). EPA investigators are expected to follow the definitions and reporting time frames for adverse events and unanticipated problems provided by the IRB of record for their project (EPA, unpublished material, April 27, 2015).
UNC provides the IRB of record for CHIE studies conducted at EPA’s Human Studies Facility, located on the UNC campus. UNC’s Office of Human Research Ethics (OHRE) is responsible for ethical and regulatory oversight of research conducted at the university that involves human subjects, regardless of funding source. OHRE administers, supports, and guides the work of the IRBs and all related activities (UNC, 2014, pp. 31, 76).
TABLE 4-3 Events Reported to the UNC IRB for All CHIE Studies from January 2009 to February 2015
| CHIE Study for Reported Event | Event Description | Reportable per IRB Policy?a | Corrective Action |
|---|---|---|---|
| CAPTAIN | Six subjects consented to the study using forms containing an error in the heading of a table. The heading referred to O3 instead of PM. | Yes | Corrected and approved consent forms will be used to re-consent the six subjects. Correct forms will be provided to any new subjects. |
| CAPTAIN | Six subjects were quoted a higher amount of reimbursement by the recruitment office. They will be paid the quoted amount. | Yes | Given the fact that these subjects were quoted the higher amount, they will be paid $1,857 and $1,757 as specified in the consent form. |
| CAPTAIN | A subject was disqualified prior to scheduled exposure to PM due to an unacceptable number of preventricular contractions on overnight Holter recordings. | Yes | None needed |
| CAPTAIN | Subject was disqualified due to unstable blood pressure and heart rate. Subject was not exposed to PM. | No | None needed |
| CAPTAIN | Enrolled subject was disqualified after overnight Holter following exposure to clean air revealed cardiac rhythm findings. | No | None needed |
| CAPTAIN | Subject showed a 12-beat run of ectopic atrial tachycardia about 2 hours following exposure to PM. | No | None needed |
| CAPTAIN | Subject was disqualified following clean air exposure based on overnight Holter findings. | No | None needed |
| CAPTAIN | Disqualification of study subject due to illness and rescheduling difficulty. | No | None needed |
| DEPOZ | After 2 consecutive days of O3 exposure, subject had a 43% and 58% decrement in FVC and FEV1, respectively, but returned to normal by next day. This decrement normally occurs in ~3% to 5% for this age group. Chest discomfort on deep inspiration. | No | Subject retained for additional time until pulmonary function returned within 5% of preexposure number. |
| DEPOZ | Cardiac arrhythmia noticed on follow-up day. | Yes | Removed from study. |
| DEPOZ | A study participant experienced a persistent cough possibly related to participation in a research protocol at EPA’s Human Studies Facility. | Yes | Subject removed from study.b Future subjects who present a cough within the first 15 minutes of exposure will be removed from the exposure room. |
| ENDZONE | Study participant received unexpected concentration of pollutant exposure. There were no adverse sequalae. | Yes | Apparently, this is the first time that an event of this nature had occurred at the EPA Human Studies Facility during several decades of CHIE studies. To prevent a similar event from occurring in the future, EPA will provide procedural reminders to exposure room operators to ensure that the NO2 delivery system is shut down with the proper protocol. |
| ENDZONE | Subject removed from study after having premature ventricular contractions during exercise. | Yes | None needed |
| ENDZONE | Subject experienced nonsustained ventricular tachycardia during rest approximately 21 hours following clean air exposure. | No | None needed |
| ENDZONE | Subject felt faint during exam prior to first study-related exposure. | No | None needed |
| ENDZONE | Subject placed on hold due to change in blood chemistry. | Yes | None needed |
| OMEGACON | A study subject experienced bradycardia during clean air exposure. | No | Though IRB did not require, subject was removed from study. |
| OMEGACON | One study subject had high blood pressure during and 1 hour after clean air exposure. | No | Though IRB did not require, subject was removed from study. |
| OMEGACON | A female study subject with a history of migraine headaches had migraine after exposure to concentrated airborne particulate matter. Symptom disappeared by next day. | Yes | Protocol was modified to exclude people with history of migraine headaches.c |
| OMEGACON | A study subject had cardiac arrhythmia during clean air exposure. | Yes | Though IRB did not require, subject was removed from study.d |
| OMEGACON | Subject had a 4-beat run of ventricular tachycardia after clean air exposure. This brief arrhythmia was noted on 24-hour Holter monitor. Subject denied any symptoms. | No | Though IRB did not require, subject was removed from study. |
| XCON | Subject experienced atrial fibrillation/atrial flutter. | Yes | The safety protocols in place for this study worked as planned, and testing was halted early. Subject was removed from study and referred for medical follow-up of underlying condition. Incident was reported in a case report by Ghio et al. (2012). |
| XCON | Short episode of elevated heart rate. Subject denied any symptoms. | Yes | Subject removed from exposure room and study. Subject was provided with copies of EKG and Holter recording and referred for medical follow-up. |
aAccording to EPA this table includes adverse events and other events. Some of the events were not reportable to the UNC IRB on the basis of its policy requirements.
b“Follow-up provided for 3 months after event including (1) being seen by an NHEERL physician; (2) receiving medication for 1 week; (3) emails and phone calls by an NHEERL nurse; and (4) scheduling an appointment for the study subject at the UNC. Ambulatory Care Center Pulmonary Clinic” (EPA, 2014a, p. 28).
c“Two days of follow-up including (1) giving the subject medicine for pain relief and a visit by NHEERL’s on duty physician on the first day and (2) a follow-up conversation with the principal investigator on the second day” (EPA, 2014a, p. 28).
d“Two days of follow-up including review of subject’s holter monitor recording a by one doctor and two nurses. On the second day, the EPA medical staff advised the study subject to see a private physician because the principal investigator believed that the study subject had an underlying medical condition. The EPA provided the study subject with a copy of medical test results” (EPA, 2014a, p. 28).
Source: EPA, unpublished material, November 23, 2015.
With the occurrence of an event deemed reportable by the UNC IRB, EPA investigators are required to report the event to various offices of EPA in addition to the IRB (EPA, unpublished material, April 27, 2015). The IRB must report to the UNC Vice Chancellor for Research unanticipated problems involving risks to subjects and others. The chancellor is responsible for all required reporting of unanticipated problems involving risks to subjects or others and the resulting IRB actions to the appropriate federal agencies. That reporting would generally be coordinated through the OHRE (UNC, 2014, p. 13).
RELATIONSHIP OF SHORT-TERM CHIE STUDY EXPOSURE TO CHRONIC DISEASE RISKS
Ambient air fine particulate matter (PM2.5) is an important criteria pollutant because of its well-documented epidemiologic evidence for association with lifespan shortening, especially via ischemic heart disease (IHD) and for lung cancer. The epidemiologic evidence for these effects, along with evidence from other sources, was instrumental in the lowering of the PM2.5 annual average concentration limit to 12 µg/m3 in 2014. DE contains ultrafine EC particles and larger sized aggregate EC particles, as well as surface coatings of OC, and miners exposed to DE underground have excess lung cancer that has been associated with its EC mass concentration in some analyses (Attfield et al., 2012; Silverman et al., 2012) but to a lesser extent or not at all in other analyses (Moolgavkar et al. 2015, Crump et al. 2015, 2016). Therefore, it may be prudent to consider whether the small increments to long-term cumulative PM exposure resulting from the short-term PM exposures of the volunteer subjects in the CHIE studies present significant risk increments of chronic effects such as cancer.
One way to consider the possible magnitude of significant incremental risk is to perform a classical risk calculation. However, such a calculation would necessarily be based on results from populations exposed for variable and much longer times (years, decades) than the short exposures in CHIE studies (hours). The Committee concluded that a risk calculation for CHIE exposures based on such disparate data would be so uncertain as to be virtually meaningless, as well as potentially misleading.
Another approach is to base the risk estimate on knowledge gained in national studies of the effects of exposures to ambient air PM2.5 and its source-related components in multiple U.S. cities. The general population is exposed to ambient air PM2.5 that includes, as a small fraction, diesel exhaust particles (DEPs) generated by traffic sources. In the study of excess lung cancer in 100 U.S. cities associated with ambient air PM2.5, and which involved two large national cohorts, Thurston et al. (2013, 2016a) found statistically significant associations for chronic exposures to PM2.5 attributable to coal combustion, but only marginal associations for PM2.5 attributable to traffic sources, and there were no such associations for any other PM2.5 source category. The cumulative exposures to DEP in these large cohort studies were many orders of magnitude higher than those in the short-term CHIE exposures. Thus, any incremental risks for lung cancer and IHD mortality resulting from CHIE study exposures to DEP. are likely to be either zero or extremely small and undetectable.
Risks of chronic diseases, such as lung cancer and ischemic heart disease, have been shown to correlate with long-term cumulative exposure to PM2.5 (Crouse et al. 2012; Hamra et al. 2014). For several CHIE studies listed in Table 4-1, the IRB-approved maximum exposure concentrations of PM2.5 were 600 µg/m3 for two hours. Exposures at those concentrations would add a very small increment to the cumulative long-term ambient PM2.5 exposures of many people in the United States. For example, the average ambient PM2.5 concentration in the western United Sates between 2000 and 2015 (average of 58 monitoring sites) was 11.6 µg/m3 (EPA, 2016f). Although that concentration results from a decreasing trend during those years, consider a scenario in which a person is exposed at that ambient concentration for 8 hours per day, 300 days per year, for 40 years. That would result in a cumulative exposure of 1,113,600 µg/m3-hours, which is about 925 times greater than the potential cumulative exposure of 1,200 µg/m3-hours for subjects in several CHIE studies in Table 4-1. This suggests that any increment of chronic-disease risk resulting from PM2.5 CHIE exposures in the studies considered by the committee would be vanishingly small relative to real-world exposures.
The objective of EPA CHIE studies has been to produce transient and reversible biomarker or physiologic responses that inform about biologic mechanisms of pollutant effects but do not cause clinical effects. Rather, these perturbations are expected to abate over time and the experimental design typically involves monitoring subjects for a sufficiently long observation period (see Chapter 5). If the perturbations induced by the exposures were to sustain, it is uncertain how they might influence subsequent responses to ambient exposures encountered in daily life. It is the committee’s judgment, however, that the types of perturbations identified in the CHIE study protocols reviewed by the committee would not persist.
CONCLUSIONS
EPA CHIE studies have been, by design, limited to exposures to subjects that are highly unlikely to exhibit responses of adverse clinical significance through
- Screening of potential subjects and
- Selection of pollutants (or mixtures), and concentrations thereof, that are not expected to produce adverse short-term responses, usually based on known associations reported in observational epidemiologic studies in larger populations, or in laboratory animals at comparable concentrations.
EPA CHIE studies are, by design, limited to the characterization only of those outcomes that reflect transient and reversible biomarker or physiologic effects, which can be used for
- Developing biomarkers of exposure, or for identifying early indicators of disease initiation and progression, and
- Studying the joint effects of different pollutants.
For the study participants in the eight EPA CHIE studies reviewed by the committee, it is the committee’s judgment that any risks of a serious adverse event with long-term sequelae were unlikely large enough to be of concern, realizing it is not possible to ever conclude that there was no risk.
Specific concerns have been expressed about CHIE-study risks of chronic diseases, such as lung cancer and ischemic heart disease, which are correlated with long-term cumulative exposure to PM2.5 (particles with an aerodynamic diameter less than or equal to 2.5 μm). However, because those diseases are considered to be associated with cumulative effects that develop over long periods, PM2.5 exposures in CHIE studies (for example, ≤ 600 µg/m3 over 2 hours) would add an extremely small increment to the cumulative lifetime PM2.5 exposures of most people in the United States. This suggests that any increment of chronic disease risk resulting from CHIE exposures would be vanishingly small.
At this time, there is insufficient information for the committee to formulate overall determinations of the success or potential utilities of the eight CHIE studies that were provided to the committee. The committee has some concern about the adequacy of the process for ensuring the most important topics are selected for CHIE studies, and whether there was sufficient senior scientific input into that process (see Chapter 5).
