Asthma in HIV-Infected Population: A Review of Respiratory Symptoms, Pulmonary Function Abnormalities and Pathophysiology

Several studies have investigated prevalence of respiratory symptoms in the HIV-infected population (Table 1). Before anti-retroviral therapy (ART), there was a high prevalence of dyspnea and cough in this population [2]. These studies showed low CD4 cell counts, history of smoking and intravenous drug use, and previous history of pneumonia were associated with increased risk of dyspnea and cough symptoms. More recent studies in the post-ART era show that respiratory symptoms continue to be common in HIV-infected individuals, with the prevalence of cough ranging from 23-37% and prevalence of dyspnea ranging from 16-44% [3,4]. In these studies, significant predictors of respiratory symptoms include age, history of intravenous drug use, smoking history, higher HIV RNA levels and history of pneumonia. In addition George et al.found that forced expiratory volume in 1 second (FEV1)/forced vital capacity (FVC) was significantly lower in participants with respiratory symptoms [3]. In conclusion, respiratory symptoms are common in the HIV-infected population, and individuals with intravenous drug use, history of smoking and pneumonia appear to be at high risk.

Table 1: Studies of respiratory symptoms in HIV-infected individuals in the post-ART era

In addition to respiratory symptoms in HIV-infected individuals, investigators have also long noted an association between HIV and obstructive lung disease (Table 2). Two recent studies found an association between low FEV1/FVC and age, pack year history of smoking, history of bacterial pneumonia, and use of ART [3,4]. Although these studies find that smoking and ART use is associated with worse obstruction, there is also evidence that indicates an association of obstructive lung disease with poor HIV control. In a cohort of urban intravenous drug users, HIV-infected participants with HIV viral loads ≥ 200,000 copies/ml were at significantly higher risk of obstructive lung disease (OR 3.41, 95% CI 1.24 to 9.39, p=0.02) compared to HIV-uninfected participants [5]. Analysis of pulmonary function over time in this cohort showed that participants with an HIV viral load >75,000 copies/ml had a greater annual decline in FEV1 and FVC compared to HIV-uninfected participants and HIV participants with viral load ≤ 75,000 copies/ml [6]. A similar association was found between CD4 cell count and pulmonary function, and participants with CD4 counts< 100 cells/mm3were found to have the most rapid decline in FEV1 and FVC compared to HIV-uninfected participants. Overall, there is a high prevalence of obstructive lung disease in the HIV-infected population, and risk factors for obstructive lung disease in this population include intravenous drug use, history of smoking, history of pneumonia, and poor HIV control.

Table 2: Studies of pulmonary function testing in HIV-infected cohorts during the post-ART eraSGRQ = St. George Respiratory Questionnaire.

The association between asthma and HIV has not been studied as well as the association between HIV and airway obstruction in general. In the pre-ART era, studies investigated an association between HIV and bronchial hyper-responsiveness, an important feature of the pathophysiology of asthma (Table 3). A small study of 66 HIV-infected and matched HIV-uninfected controls showed that airway hyper-responsiveness was not more common in HIV-infected subjects compared to HIV-uninfected controls (19.3% vs. 12.9%, p>0.1) [7]. This study used methacholine challenge testing as the means to detect airway hyper-responsiveness, which was defined as a decline of ≥ 20% in FEV1 from the post-diluent value after inhalation of ≤125 cumulative breath units. In contrast, another early study with 105 individuals with Acquired Immunodeficiency Syndrome (AIDS) found that 33% of participants had decreased airflow rates and 31% had significant response to bronchodilators [8]. Poirier et al. did the largest study of bronchial hyper-responsiveness in 248 male HIV-infected and 236 HIV-uninfected participants [9].

Table 3: Overview of asthma studies in HIV-infected cohorts

There was a statistically significant greater prevalence of current wheezing in the HIV-infected participants compared to HIV-uninfected participants (54.4%, vs. 21.2%, p<0.001). The prevalence of bronchial hyper-responsiveness was only statistically significant in HIV-infected smokers vs. HIV-uninfected smokers (30.1% vs. 13.3%, p<0.05). HIV-infected participants with bronchial hyper-responsiveness were more likely have a reduced FEV1/FVC (0.76 vs. 0.81, p=0.001) and elevated IgE levels > 100IU/ml (43.1 vs. 23.5, p=0.005), and they tended to be more likely to have a positive skin allergy test (63.1 vs. 49.7, p=0.06). Overall, in the pre-ART era, bronchial hyper-responsiveness appeared to be common in HIV-infected individuals, especially those with a history of smoking.

In the era of ART, only a few studies have examined asthma in HIV. A large study focused on pulmonary disease in patients in the Veterans Health system and used a cohort of HIV-infected (n=33,420) and HIV-uninfected (n=66,840) individuals in the Veterans Aging Cohort [10]. The median age of the cohort was 45 years with most participants being male and smokers. In both groups asthma was the second most common ICD-9 diagnosis with the incidence being similar in both groups (5.6 per 1000 person years in HIV-infected cohort vs. 5.6 per 1000 person years in HIV-uninfected cohort, p=0.6).

In this study, HIV was not an independent risk factor for asthma, but pulmonary testing was not performed. Incident pulmonary diseases, including asthma, COPD and pulmonary infections were less likely in participants with lower HIV RNA levels and use of ART at baseline. In another study with a cohort of 223 HIV-infected participants who performed spirometry, 9% had a bronchodilator response [11]. Participants with a bronchodilator response were more likely to have abnormal spirometry and wheezing. This study also found the prevalence of doctor-diagnosed asthma was 20.6%, much higher compared to a recent estimate of 7.3% in the general population, although it is difficult to make direct comparisons [1].

Risk factors for doctor-diagnosed asthma in the cohort including parental history of asthma, female sex, obesity, current smoking, history of Pneumocystis pneumonia, and not being on ART. Another recent study compared prevalence of comorbidities in an HIV-infected population (n=14,005) and their age and sex-matched controls using administrative data from the Canadian health care system in Ontario [12]. The study population in this analysis was mostly males living in an urban setting with an average age of 45.36 years. HIV-infected participants had a high prevalence of asthma (12.7%, 95% CI 12.2% to 13.3%), and the prevalence ratio of asthma in HIV-infected vs. HIV-uninfected individuals was 1.31 (95% CI 1.20 to 1.43). In the same study, chronic obstructive pulmonary disease (COPD) was also a common comorbidity with a prevalence in HIV-infected persons of 7.9% (95% CI 7.5% to 8.3%), which was about 38% less prevalent than asthma. In conclusion, HIV-infected individuals appear to have a higher prevalence of asthma and bronchial hyper-responsiveness compared to the general population, but large studies specifically testing for asthma in the HIV-infected population and appropriate control groups are lacking.

Specific mechanisms resulting in asthma in the HIV-infected population are not well-understood. Investigators have theorized about potential roles of ART, inflammation, and obesity in the pathogenesis of asthma in this population (Figure 1), but definitive studies are lacking. Causes of asthma may be similar to those in the HIV-uninfected population, but there may be mechanisms unique to HIV infection as well. Potential HIV-related mechanisms leading to asthma include damage to the lung from immune activation and inflammation, direct effect of HIV proteins, or activation of memory CD4 T cells [3]. Mechanisms may also vary by asthma “phenotype” as is seen in the HIV-uninfected population. Recent research shows there are different phenotypes of asthma based on the severity of symptoms, key inflammatory biomarkers and corticosteroid response [13]. Whether such phenotypes exist in HIV-infected individuals is not known.

Figure 1: Proposed mechanisms involved in the pathogenesis of asthma in HIV-infected population;

One factor to consider in the pathogenesis of asthma in HIV-infected individuals is the role of HIV virus. As noted above, some studies indicate that higher viral loads and poor HIV control are associated with worse lung function [5,6]. In addition, some studies have shown that higher HIV RNA levels are predictors of respiratory symptoms [3]. These studies would indicate that the virus itself may be important in pathways that can lead to asthma. Unfortunately, research studying the mechanisms involving the virus itself leading to asthma is lacking.

Interestingly, ART itself has been associated with the pathogenesis of asthma in HIV-infected individuals, but data are conflicting. Evidence suggesting a detrimental relationship with ART comes from pediatric literature. A large study compared asthma medication use in HIV-infected children on ART to medication use in HIV-infected children not on ART [14]. Based on a time dependent Cox model, researchers found that HIV-infected participants on ART were more likely to be on asthma medication compared to HIV-infected participants not on ART (hazard ratio=3.34, p=0.01). In addition, the study analyzed asthma medication use at follow up in participants born in pre-ART era. The prevalence of asthma medication use at 11 years of age in this group was higher in HIV-infected participants on ART compared to those not on ART (OR 3.38, p=0.02). The researchers hypothesized that loss of CD4 T cells in untreated HIV-infected children is protective against asthma, and ART acts as a potential risk factor for asthma due to immune-reconstitution of CD4 T cells. Similar results were seen in an earlier smaller study of 136 HIV-infected adult participants in a community clinic showed that a recent CD4 count of ≥ 200 cells/dL was positively associated with current asthma (p=0.01) [15]. In addition, the mean CD4 count in participants with asthma was significantly higher than other participants (log CD4 cells/μ l5.63±0.73 vs. 4.92±1.38, p=0.01). Potential mechanisms involving ART in the pathogenesis of asthma include auto-immune reaction or an immune reconstitution-like reaction to antigens in lungs and direct effects of ART. Overall, the role of ART in asthma pathogenesis in the HIV-infected population is unclear and may differ in children and adults, possibly due to varying underlying causes of asthma.

Certain inflammatory pathways and cytokines may be important in the pathogenesis of asthma in the HIV-infected population. Immunoglobulin E (IgE) is a well-known mediator of allergy and asthma. Early studies showed significantly higher IgE levels in the HIV-infected population with CD4 cell count ≤ 200 cells/mm3, and there was an inverse relationship between IgE level and both helper T cell and suppressor/cytotoxic T cell numbers [16]. Another study showed IgE levels were elevated in HIV-infected vs. HIV-uninfected participants (log IgE IU/ml 4.40±0.48 vs. 3.60±0.18, p<0.05), with the difference being especially strong when comparing participants with AIDS (log IgE IU/ml 4.96±0.23 vs. 3.60±0.18, p<0.005) [17]. The study also showed that IgE levels were important markers of disease progression and survival. The exact etiology of elevated IgE levels is not well-understood, but T helper cells may play a role. HIV leads to decreases and alterations in T helper cells, which are critical in IgE synthesis [2]. More recently, IgE levels have been shown to be positively associated with sputum eosinophil count in a cohort of 223 HIV-infected participants [11]. An older study found elevated IgE levels to be independent predictors of bronchial hyper-responsiveness based on methacholine challenge (p=0.0035) [9]. Thus, the clinical significance of elevated IgE levels in HIV-infected individuals needs to be further investigated, along with the relationship between elevated IgE levels and asthma in HIV. Future research focusing on correlation between elevated IgE levels and pulmonary function, and interventions to decrease the IgE levels in HIV-infected individuals could provide further insight into the role of IgE in this population.

Recent studies also provide evidence for involvement of other inflammatory mediators in the development of asthma in HIV. In the cohort of 223 HIV-infected participants discussed above, doctor-diagnosed asthma appeared to be more common in participants with high sputum interleukin (IL-4) (27% with asthma if high IL-4 vs. 10.5% with asthma if low IL-4, p=0.02), and high regulated on activation, normal T cell expressed and secreted (RANTES) (26% vs. 9.8%, p=0.02) [11]. The association between IL-4 and RANTES with doctor-diagnosed asthma indicates a role of type 2 T helper (Th2) cells in the pathogenesis of asthma in the HIV-infected population. Bronchodilator response was associated with high plasma macrophage inflammatory protein (MIP)-1α (26.3% with asthma if high MIP-1α vs. 4.8% with asthma if low MIP-1α, p=0.002) and high sputum MIP-1β (4.6% vs. 1.0%, p<0.001). RANTES, MIP-1α and MIP-1β are chemokines that suppress HIV and are released by CD8 T cells during chronic HIV infection [18]. These studies suggest that there is a complex role of inflammation in the pathogenesis of asthma in HIV-infected individuals.

One proposed mechanism that could link ART and asthma involves the peroxisome-proliferator activated receptor γ (PPARγ), which has been shown to be involved in pathogenesis of lipodystrophy in HIV-infected individuals likely secondary to decreased PPARγ expression due to ART [19]. PPARγ is produced by adipocytes and is regulates genes associated with growth, insulin sensitivity and immunity. PPARγ also regulates production and secretion of adipokines such as leptin and adiponectin, which have been shown to be associated with asthma, as discussed below. Given the metabolic effects of PPARγ in the general population, the role of PPARγ in HIV-uninfected asthmatics is also being studied. In a study of 34 non-smoker asthmatic participants, Benayoun et al.showed that PPARγ expression was increased in bronchial submucosa, bronchial epithelium and airway smooth muscle in steroid-untreated asthmatics [20]. In comparison, steroid-treated asthmatics had significantly lower PPARγ levels, especially in bronchial submucosa and airway smooth muscle. The researchers also noted that degree of PPARγ expression negatively correlated with FEV1. In contrast, Spears et al.showed that PPARγ agonist rosiglitazone improved FEV1 (183 ml, p=0.051) in asthmatic smokers [21]. Since PPARγ plays a role in asthma pathogenesis and HIV-related lipodystrophy, it may be an important factor in obesity related HIV-associated asthma.

In the general population, there are emerging links between obesity and asthma. Several studies of adults and children reported an association between elevated body mass index (BMI) and prevalence of asthma. A recent meta-analysis showed that BMI in the overweight and obese range increased risk of 1 year incidence asthma diagnosis (OR 1.51, 95% CI 1.27 to 1.80) [22]. Researchers also noted a dose-response effect of high BMI on asthma diagnosis. Associations between obesity and asthma in HIV-infected population have recently been investigated in a cohort of 1833 HIV-infected participants researchers, finding that obesity was a risk factor for multi-morbidity compared to HIV-infected individuals with normal BMI (OR 1.52, 95% CI 1.15 to 2.00) [23]. In HIV-infected individuals, the odds of doctor-diagnosed asthma was 2.3 times greater in obese compared to non-obese participants [11]. These studies indicate that obesity phenotype asthma is prevalent in HIV-infected population. One potential mechanism linking obesity and asthma is inflammation, and studies in this area have focused on adipokines produced by adipocytes [24-26]. Leptin is a proinflammatory adipokine that increases production of inflammatory cytokines. In contrast, adiponectin is an adipokine that downregulates the production of inflammatory cytokines. Both leptin and adiponectin have been associated with asthma in mouse models, but data on the effects of these adipokines in human samples are lacking. The mechanisms behind the effect of role of these adipokines in development of asthma are not well understood. Evidence does show that these adipokines affect eosinophil chemotaxis in lung tissue. It is also hypothesized that increased production of inflammatory cytokines such has leptin, IL-4 and TNF-α leads to decreased activity of regulatory T-lymphocytes and increased production of IL-10 by macrophages and adipocytes, thereby shifting the balance of immune status towards a Th2 phenotype and increasing the risk of allergic disease, but there is no clear evidence to support this hypothesis thus far [27]. Therefore, the role of inflammatory cytokines and adipokines in obese HIV-infected asthmatics needs to be further investigated.

Another link to obesity and asthma is oxidative stress, and the role of oxidized low-density lipoprotein (ox-LDL) in pulmonary disease has gathered attention. Scichilone et al.studied LDL subclasses in 24 asthmatics and 24 healthy matched controls [28]. The results showed that in asthmatics, the least pro-inflammatory LDL (LDL-1 and LDL-2) appear to be lower and the most pro-inflammatory LDL (LDL-3 and 4) appear to be higher. Also, LDL-1 (least pro-inflammatory) positively correlated with lung function. LDL-3 (most pro-inflammatory) negatively correlated with lung function suggesting that LDL is a potent trigger for lung inflammation. Thus, based on early evidence, LDL-ox might be an important mediator in the pathogenesis of inflammatory lung diseases like asthma. It is important to note that several studies have shown that obese asthmatics are more prone have poor therapeutic response to traditional asthmatic treatments like glucocorticoids [26], and the treatment options for these patients are limited. Early studies on small samples show that weight loss does result in improvement in lung function. Thus obesity may be linked to asthma through several pathways but there is still limited understanding of these mechanisms. Further investigations need to focus on the inflammatory pathways related to adiponectins and oxidized LDL to understand these mechanisms in order to develop therapeutic options, especially in HIV-infected population where obesity associated asthma phenotype is prevalent.

Similarly, recent evidence also implicates the role of insulin resistance in the pathogenesis of asthma with Theusen et al.finding an association between insulin resistance wheezing (OR 1.87, 95% CI 1.38 to 2.54, p=0.002) and asthma-like symptoms (OR 1.61, 95% CI 1.23 to 2.10, p=0.001) [29]. Further evidence for a link between insulin resistance and asthma comes from a study in which clinical measures of insulin resistance and obesity were correlated with airway hyper-responsiveness (AHR) based on methacholine challenge test [30]. Researchers found that in men, FEV1 significantly correlated with fasting insulin level and insulin resistance based on homeostasis model of assessment of insulin resistance (HOMA-IR). After adjustment for age, HOMA-IR was a statistically significant risk factor for AHR, but became non-significant after adjustment for age and BMI. Similarly, in women, the HOMA-IR and BMI were statistically significant risk factors for AHR after adjustment for age. The relationship between HOMA-IR and AHR again became insignificant after adjustment for age and BMI. Insulin resistance is important in the HIV-infected population because of the high prevalence of metabolic changes in this population. Recent estimates of diabetes mellitus is ~3% in ART-naïve HIV-infected individuals and ~10% in individuals on highly active ART [31]. In addition, metabolic changes that occur in HIV infection have been known to cause lipohypertrophy likely secondary to chronic inflammation [32]. Biron et al.found a prevalence of metabolic syndrome of 18.2% in cross-sectional study of 269 HIV-infected individuals started on ART with a median duration of 30 months of treatment [33]. The researchers concluded that HIV-infected individuals on ART are at high risk of complications like cardiovascular disease. Thus, HIV-infected individuals with obesity and insulin resistance may be at risk for asthma. Future studies in this area need to explore the relationship between insulin resistance, and asthma in HIV-infected individuals, including investigating prevalence of AHR in HIV-infected population with insulin response and their response to treatment.

Current asthma treatments used in the general population have not been tested in the HIV-infected population. If HIV-infected individuals have different mechanisms of asthma than the general population, then current therapies may be less effective, particularly if HIV or ART are involved in pathogenesis. In addition, there are several interactions of concern in the HIV-infected population including increased complications of inhaled corticosteroids (ICS) such as candidiasis, tuberculosis, and bacterial pneumonia [34]. The TORCH (towards a revolution in COPD health) study in COPD treatment studied adverse effects of combination therapy with salmeterol and fluticasone compared to the placebo group [35]. The study found that probability of developing bacterial pneumonia during the 3 year follow-up was greater in combination therapy group than the placebo group (19.6% vs. 12.3%, p<0.001). Tuberculosis risk may also be increased. In a study of 853,439 participants with inhaled respiratory medicine use, the use of ICS was associated with an increased rate of TB diagnosis (adjusted OR 1.20. 95% CI 1.08-1.34, p<0.001) [36]. The HIV-infected population is likely at increased risk of such complications with ICS therapy. Furthermore, ART can lead to systemic side effects of ICS. Pharmacologic studies and case reports provide evidence that combination of ritonavir and fluticasone can lead to symptoms of Cushing’s syndrome and adrenal insufficiency, and the combination treatment is contraindicated in HIV-infected individuals [37]. The interaction between these medications is likely secondary to the CYP3A4 system, and other inhaled corticosteroids also are substrates of CYP3A4, which is inhibited by ritonavir. It is possible that there are interactions between ritonavir and other inhaled corticosteroids at higher doses. These important treatment considerations should be remembered in the HIV-infected population and further research needs to determine efficacy and safety of current asthma treatments in this population.

In summary, pulmonary diseases including asthma appear to be highly prevalent in the HIV population, and they are increasingly becoming an important cause of morbidity and mortality. Asthma is one of the most common pulmonary diseases in the HIV-infected population, and recent evidence suggests a higher rate of bronchial hyper-responsiveness in HIV-infected individuals. Smoking, in particular, appears to be a strong risk factor in development of respiratory symptoms and obstructive pathophysiology. Newer research points to several inflammatory cytokines, metabolic diseases such as obesity, and ART in the pathogenesis of asthma in HIV patients. Future research needs to focus on understanding the mechanisms that lead to development of asthma in HIV to improve prevention and treatment.