Abstract
Objective
To identify environmental factors that may precipitate angioedema.
Study Design
Case series with chart review.
Setting
An urban tertiary care medical center.
Subjects and Methods
After institutional review board approval was obtained, a chart review of all patients who presented to Temple University Hospital with angioedema from January 2012 to December 2014 was performed. Patient demographics and hospital course were gathered. Environmental data on the dates of presentation, including precipitation, humidity, and air pollution, were obtained from regional Environmental Protection Agency online data banks and used for statistical analysis.
Results
In total, 408 patients, representing 450 episodes of angioedema, met the inclusion criteria for the study. Most patients were female (58%), African American (74%), and on an angiotensin-converting enzyme inhibitor (ACE-I) upon presentation (61%). Days with unhealthy levels of air pollution were associated with an increased likelihood of angioedema presentation (odds ratio [OR], 1.82; confidence interval [CI], 1.01-3.30; P = .046). Subgroup analysis revealed that elevated ground-level ozone was the primary air pollutant significantly associated with angioedema presentations (OR, 4.95; CI 1.92-12.76; P = .0009). Higher ground-level ozone was also associated with ACE-I angioedema presentations (P = .017) but not with non-ACE-I angioedema presentations (P = .86). Air quality was not predictive of angioedema severity or need for intubation.
Conclusion
Angioedema is a complex, multifactorial disease resulting in potentially life-threatening complications. This is the first study to demonstrate that higher levels of air pollution, specifically ground-level ozone, are associated with significantly increased rates of angioedema episodes, although not severity.
Background
The modern medical understanding of angioedema originated in the 19th century with the work of Quincke, who described a series of patients presenting with a swelling disorder.1,2 Angioedema is a condition that is characterized by rapid accumulation of fluid in subepithelial or submucosal tissues due to leakage from blood vessels. 3 While this condition can affect any part of the body, it most commonly occurs in the face and perioral region. 4 Currently, angioedema is broadly categorized based upon the presence or absence of an inheritance pattern. 5 Most cases in the United States are acquired angioedema and most commonly associated with angiotensin-converting enzyme inhibitors (ACE-I). 6
Despite our growing understanding of the molecular pathways that are involved in angioedema, the specific factors that precipitate an episode are still poorly understood. Even in cases of ACE-I and C1 esterase deficiency-associated angioedema, the current evidence available provides no definitive explanation as to why patients develop symptoms on any particular day.
Few studies examine angioedema triggers. For patients with hereditary angioedema (HAE), patient surveys have demonstrated that certain factors may provoke an episode such as physical exertion, mental stress, or mechanical trauma before an episode. 7 Allergic responses to food have also been proposed as triggers for HAE. 8
Similarly, there is also a paucity of information regarding the timing of an episode of angioedema. It has been well documented that patients who take ACE-I may develop angioedema any time during their treatment course, with cases presenting as late as 20 years after initiation of therapy.9-12 Although, there is some evidence suggesting a contribution of elevated pollen counts to angioedema presentations, confounding variables make the results inconclusive. 13
While it is clear that external factors are involved in the development of angioedema, little can be concluded from the available literature as to how the environment affects this disease. This study aims to investigate the role of air pollution in triggering episodes of angioedema.
Methods
After institutional review board approval from Temple University Hospital was obtained (protocol 24322), a chart review of all consecutive patients who presented with angioedema (International Classification of Diseases, Ninth Revision [ICD-9] code 995.1 and International Classification of Diseases, Tenth Revision [ICD-10] code T78.3) to Temple University Hospital, an urban tertiary care center, between January 2012 and December 2014 was performed. Patients with documented angioedema either in the emergency department or during hospitalization were included. Patients who had episodes that were incorrectly coded or attributed to another cause such as anaphylaxis were excluded. Anaphylaxis was determined by using the criteria outlined by Sampson et al. 14 Specifically, patients were excluded if airway edema was accompanied by documented systemic signs and symptoms such as bronchospasm, hives, or hypotension.
The date of presentation to the hospital, sex, ethnicity, tobacco use, comorbid medical conditions, family history of HAE, history of atopic disease, and ACE-I use were recorded. Furthermore, the clinical details of each episode of angioedema, including symptoms at presentation, anatomic site(s) involved, laryngoscopic findings, timing of airway intervention if necessary, treatment medications, length of stay, and disposition were collected. Laboratory testing results including C1-INH (quantitative), C3, and C4 levels were documented. Any identifiable trigger for the angioedema was recorded as well.
To evaluate the environmental impact on the incidence of angioedema, peak 1-day air pollution measurements (Air Quality Index [AQI]), rainfall, humidity, temperature, windspeed, and smog levels on the day of presentation were abstracted from a regional air monitoring database. 15 Air pollution data were available for the local county. Air pollution values are reported as unitless indices ranging from 1 to 300 and are stratified based on Environmental Protection Agency (EPA) definitions of good (1-50), moderate (51-100), and unhealthy (101-150). 15 The 5 components of air pollution (ground-level ozone, particulate matter, carbon monoxide, sulfur dioxide, and nitrogen dioxide) were independently analyzed as well. Rainfall and weather data were obtained from the local regional station. 16
Statistical Analysis
Statistical analysis was performed using JMP Pro version 14.0.0 (SAS, Cary, North Carolina). The number of angioedema presentations for each individual calendar day during the study period was recorded and stratified into 3 groups: days with 0 presentations, days with 1 presentation, and days with multiple presentations (defined as ≥2). A univariate logistic regression was performed for each environmental factor with the number of angioedema presentations per day. In addition, 2 multivariate logistic regression models were constructed to identify factors independently associated with angioedema episodes per day (1 compared to 0 and >2 compared to 0). Covariates significant on univariate analysis, AQI, and season of presentation were selected for multivariate analysis.
In addition, a subgroup analysis was performed for AQI and season of presentation. The 5 individual pollutants in AQI were stratified into healthy and unhealthy levels to determine their individual correlation with angioedema episodes. A subgroup analysis of individual summer months compared to other months of the year was performed to determine an association with frequency of angioedema episodes.
Last, to estimate angioedema severity, a separate analysis was performed in a similar fashion to assess the risk of intubation. Demographic data, clinical histories, and significant environmental factors found in the previous analysis were analyzed with a univariate model and as an aggregate on a multivariate logistic regression model. Associations were reported as odds ratio (OR) with 95% confidence intervals (CIs). P values less than .05 were considered statistically significant.
Results
Study Group
During the study period, 469 consecutive patients presented to the hospital with a diagnosis of angioedema. Of these patients, 61 were miscoded and as such excluded. This resulted in a total of 408 patients who presented with angioedema between 2012 and 2014 and were included in the analysis.
There were 257 women (58%) and 193 men. The mean age was 54.9 years (range, 14-90 years). Most patients were African American (74%), followed by Hispanics (15%) and whites (10%). The general emergency department population at our institution consists of 68% African Americans, 16% Hispanics, and 9% whites. Most patients reported being either current (38%) or former (38%) smokers, and 24% endorsed alcohol use. Sixty-one percent of cases were attributed to ACE-I use. A total of 450 angioedema episodes were captured during the study period, including 42 recurrent episodes. Nineteen percent of patients required intubation, and 57% were admitted ( Table 1 ). There were 295 days with 1 angioedema episode, 74 days with 2 episodes, and 15 days with 3 episodes. Ninety patients (20%) had angioedema-specific lab testing performed, 9 of whom had abnormal testing and went on to have a further workup. None of these 9 patients had HAE. Two patients in the cohort were already known to have HAE.
Demographics. a
Abbreviations: ACE-I, angiotensin-converting enzyme inhibitor; ICU, intensive care unit.
Values are presented as number (%) unless otherwise indicated.
Predictors of Angioedema Episodes
On univariate analysis, AQI, total rainfall, average humidity, average temperature, maximum windspeed, average smog levels, year of presentation, and season of presentation were not significantly different between days with 0 and 1 angioedema episodes. However, days with multiple episodes of angioedema were significantly associated with season and AQI levels. Specifically, summer days (June 21 to September 22) were significantly associated with days with multiple angioedema episodes presenting (odds ratio [OR], 2.00; confidence interval [CI], 1.02-3.94; P = .04). Furthermore, days with unhealthy levels of air quality (100-150) were significantly associated with angioedema episodes compared to days with moderate levels (50-100) (OR, 3.24; CI, 1.38-7.60; P = .007) ( Table 2 ). Moderate AQI levels were chosen as the reference point, as most days in the 3 year study period were considered moderate. Other aforementioned environmental factors failed to demonstrate a significant association with multiple angioedema episodes. Season of presentation and AQI were included in the multivariate analysis, which demonstrated that AQI remained statistically associated with angioedema episodes (OR, 2.95; CI, 1.25-7.00; P = .014), whereas the summer months were no longer independently significant (OR, 1.98; CI, 0.99-3.94; P = .053).
Selected Predictors for Days with Multiple Angioedema Episodes. a
Abbreviations: CI, confidence interval; NA, not applicable; OR, odds ratio; —, not applicable, there is no P value for references.
P values in bold are statistically significant.
Moderate AQI used as reference since the majority of days are in this category.
Average windspeed divided into high and low categories by median value.
A subgroup analysis for risk of angioedema presentations with the 5 individual components of AQI was performed. Days with unhealthy level of the pollutants were compared to days with healthy levels. Only days with unhealthy levels of ground-level ozone were significantly associated with multiple angioedema episodes (OR, 4.95; CI, 1.92-12.76; P = .0009) ( Table 3 ). In addition, mean ground-level ozone values were not significantly different on days with 1 angioedema episode compared to days with none (mean, 34.3 vs 35.7; P = .40). However, days with 2 or more angioedema episodes were associated with significantly higher ground-level ozone values compared to days without angioedema (43.1 vs 34.3; P = .026) ( Figure 1 ). In addition, subgroup analysis for the individual summer months was conducted. August 2012 had a significantly higher number of angioedema episodes compared to all other months in 2012 (P = .007) ( Figures 2 and 3 ).
Association of Unhealthy Levels of Air Pollution Components on Days with Multiple Angioedema Episodes: A Multivariate Model. a
Abbreviations: CI, confidence interval; NA, not applicable; OR, odds ratio.
P values in bold are statistically significant.
Sulfur dioxide and carbon monoxide have insufficient days with unhealthy air pollution levels for adequate analysis.
Ozone levels stratified by angioedema episodes. Bar chart representing ground-level ozone, recorded by regional Environmental Protection Agency centers. Ozone levels on days with >2 angioedema episodes (*) were significantly higher than days without episodes (P = .029).
Episodes of angioedema by season. Stratification of angioedema episodes by season overlaid with average ground-level ozone.
Episodes of angioedema by month. Stratification of angioedema episodes by month overlaid with average ground-level ozone. August 2012 (*) had a significantly higher number of angioedema episodes and ozone levels (P = .007).
ACE-I Use
ACE-I angioedema was not associated with AQI but was associated with ground-level ozone. No significant difference was found in AQI levels when comparing days with at least 1 presentation of ACE-I angioedema to days with non-ACE-I angioedema (P = .73) or days without angioedema (P = .23). However, ground-level ozone values were higher on days with ACE-I angioedema compared to days with non-ACE-I angioedema (39.0 vs 34.0; P = .029) and days without angioedema (P = .017). In addition, ground-level ozone values were no different when comparing days with non-ACE-I angioedema to those with no angioedema (P = .86).
Predictors of Intubation
On univariate analysis, age greater than 55 years, ACE-I use, lack of diagnosis of atopy, the presence of an identifiable angioedema trigger, and increased anatomic subsites involved were all significantly associated with the need for endotracheal intubation. The Hispanic race was associated with a decreased likelihood of endotracheal intubation compared to African Americans ( Table 4 ).
Predictors of Intubation. a
Abbreviations: ACE-I, angiotensin-converting enzyme inhibitor; AQI, Air Quality Index; CI, confidence interval; NA, not applicable; OR, odds ratio; —, not applicable, there is no P value for references.
P values in bold are statistically significant.
All covariates that were significant on univariate analysis were included in a multivariate logistic regression model. The only significant predictors of intubation were age greater than 55 years, ACE-I use, and greater number of subsites involved (P = .04, P = .04, P = .0001, respectively). On subgroup analysis of involved anatomic subsites, tongue (OR, 6.70; CI, 3.55-13.04; P < .0001) and laryngeal (OR, 6.20; CI, 3.15-12.47; P < .0001) involvement were most associated with need for endotracheal intubation.
Discussion
The atmospheric AQI is a measure calculated by the EPA that represents an aggregate of the 5 main pollutants regulated by the Clean Air Act: ground-level ozone, particulate matter, carbon monoxide, sulfur dioxide, and nitrogen dioxide. 17 Measures of these pollutants are converted to indices such that the EPA’s threshold for AQI and each individual pollutant is below 100. Although only few days per year average above this threshold, there is still a strong correlation between AQI and detrimental health effects even at lower levels.18,19 While most of these studies demonstrate correlation, they fail to prove causality. 20
To our knowledge, this is the first study demonstrating an association between air pollution, ground-level ozone values, and angioedema. Overall, days with AQI above 100 correlated with an increased number of angioedema presentations as compared to days with lower AQI. This relationship held true on multivariate analysis.
Of the 5 components of AQI, ground-level ozone and particulate matter are considered the main pollutants. 15 In this study, ground-level ozone was the only component of AQI that was significantly associated with multiple angioedema episodes. Ground-level ozone has been implicated to contribute to various respiratory conditions, including asthma, chronic obstructive pulmonary disease, and acute respiratory distress syndrome.21,22 It is formed by the interaction between air pollutants, such as volatile organic compounds, and nitrogen oxide in the presence of sunlight. 23 Ambient ground-level ozone values usually vary between 0.020 and 0.040 ppm. Moderate elevations in levels from 0.070 to 0.120 ppm are usually noted during the peak ground-level ozone months of May through September.
The ozone molecule consists of 3 oxygen atoms arranged in a cyclic structure. Normally present as a gas, ozone is a potent oxidizing agent. It enters the body by dissolving into the bloodstream and immediately undergoes oxidation with biological molecules. 24 Most important, ozone reacts with polyunsaturated fatty acids, which leads to the formation of lipid oxidation products (LOPs) and reactive oxygen species (ROS), such as hydrogen peroxide. LOPs and ROS have been purported to increase inflammation, through the 5-lipooxygenase pathway, and cause vasodilation, through the nitric oxide (NO) pathway.24-26 At the same time, activation of the bradykinin pathway has been shown to lead to increased levels of vascular permeability factors and activation of NO synthase.27,28 Thus, patients with a hereditary or acquired angioedema, who already have elevated levels of these factors that increase blood circulation, may be pushed out of equilibrium and into an angioedema episode by the presence of elevated ozone levels in the atmosphere via LOPs and ROS.
Ground-level ozone values rise in the spring and peak in the summer. These waxing and waning levels have been demonstrated to predict asthma outbreaks in select populations. 29 While this clear correlation exists, a mechanism has not yet been determined. In this study, we noted an association with increased incidence of angioedema episodes during the summer months on univariate analysis, most notably in August 2012, when the mean ground-level ozone values were higher than average and there was a significantly higher frequency of angioedema episodes compared to other months in 2012.
While higher levels of air pollution and ozone had an increased likelihood of multiple angioedema presentations, most patients presented on days with good to moderate AQI. This inconsistent pattern may occur because triggers do not always cause an episode, suggesting a more complex relationship between angioedema and the environment. 30 Straka et al 13 examined environmental allergens and similarly noted that the incidence of ACE-I angioedema was higher during spring and summer months when there are elevated tree and ragweed pollen counts. This study, however, lacked confirmatory allergy testing, and a significant portion of the patients also presented outside of peak allergy season. Nevertheless, pollen counts may be a major confounding factor in our study and require further investigation.
This relationship is further complicated by the presence of multiple subtypes of angioedema. Current literature suggests angioedema presentation, its triggers, and treatment may vary depending on the mechanism of action of the specific subtype.3,31 Perhaps only certain subtypes of angioedema are subject to the effects of air pollution. For patients with HAE, Zotter et al 7 attempted to determine the presence of trigger factors by using patient surveys and diaries. Patients most frequently reported physical exertion, mental stress, or mechanical trauma before an episode. 7 While these data suggest that there may be an environmental or emotional contribution, the nature of the study introduces significant room for recall bias due to the lack of objective testing. Steiner et al 8 attempted to address this issue by testing for elevated allergic responses to self-reported food triggers in HAE patients who experience episodes of abdominal angioedema. However, there was no confirmatory evidence that self-reported food triggers caused these episodes as neither skin nor serum testing returned positive. 8 Although our study did not separately examine HAE patients due to low numbers, we did have a significant portion of patients with ACE-I use. We found that ACE-I angioedema was more strongly associated with days with higher ground-level ozone values compared to non-ACE-I angioedema. However, this association was not demonstrated with AQI, likely due to the relatively few days in the year with high AQI.
Furthermore, this study failed to show that elevated AQI levels correlated with increased frequency of intubation. Instead, other factors were found to be significant, such as age, number of anatomic subsites involved, and ACE-I use, suggesting that once an angioedema episode is initiated, progression is dependent on other patient characteristics. As such, treatment recommendations cannot be made at this time for angioedema episodes associated with increased air pollution levels.
Our findings suggest an association with angioedema and air pollution at a single institution. A multi-institutional or national database review would be crucial for confirmation. Inclusion of institutions with a larger percentage of hereditary angioedema would further elucidate the effect of air pollution on the various subtypes. The addition of a cohort from a rural setting would allow for evaluation of the role of drastically different air pollutant levels and allergens in triggering episodes. Additional studies examining the association of pollen levels with angioedema are under way and will help clarify this relationship.
Limitations of this study include its retrospective design. Patient charts were identified for review using only the ICD codes for angioedema, which may have missed patients who were coded differently. In addition, the recorded time of onset of angioedema may be skewed, as the captured time and date were based on presentation to the hospital rather than initiation of symptoms. Finally, this study only demonstrates association at a single location, albeit typical of large urban cities in the United States.
Conclusion
Angioedema is a complex multifactorial disease resulting in potentially life-threatening complications. This study is the first to demonstrate an association between air pollution and increased episodes of angioedema. Specifically, ground-level ozone was the main pollutant associated with multiple episodes of angioedema presenting on the same day. There was an association between ground-level ozone and ACE-I angioedema but no association with angioedema severity, and most patients presented with angioedema on days with good air quality, suggesting other contributing factors.
Author Contributions
Disclosures
Footnotes
No sponsorships or competing interests have been disclosed for this article.
This article was presented at the AAO-HNSF 2018 Annual Meeting & OTO Experience; October 7-10, 2018; Atlanta, Georgia.