Interleukin-6 and Tumor Necrosis Factor-α Are Associated with Quality of Life–Related Symptoms in Pulmonary Arterial Hypertension

Abstract

Rationale: Inflammation is associated with symptoms in many chronic illnesses; however, this link has not been established in pulmonary arterial hypertension.

Objectives: The objective of this study was to investigate the association between inflammatory markers and quality of life–related symptoms in patients with pulmonary arterial hypertension. We hypothesized that higher circulating IL-6 and tumor necrosis factor-α levels would be associated with worse quality of life–related symptoms.

Methods: We performed a secondary analysis using baseline and 3-month assessments of 62 subjects in a clinical trial of aspirin and simvastatin to determine the association between plasma IL-6 and tumor necrosis factor-α levels and the Medical Outcomes Study Short Form-36 subscales (pain, vitality, mental health).

Measurements and Main Results: The mean age was 49.7 ± 13.4 years; 87% were female. Higher IL-6 levels were significantly associated with lower Medical Outcomes Study Short Form-36 subscale scores, indicating worse bodily pain, vitality, and mental health (all P < 0.01). Higher tumor necrosis factor-α levels were significantly associated with increased bodily pain, but better mental health scores.

Conclusions: IL-6 and tumor necrosis factor-α levels are associated with certain quality of life domains in patients with pulmonary arterial hypertension.

Clinical trial registered with www.clinicaltrials.gov (NCT00384865).

Pulmonary arterial hypertension (PAH) is a chronic, debilitating disease affecting primarily young to middle-aged women; the mean age is 50 years at diagnosis (1). Patients are limited in their daily activities and the risk of mortality is high (2). Elevated right ventricular afterload leads to right heart failure and, ultimately, premature death (3). Although patients with PAH report multiple symptoms, dyspnea, fatigue, anxiety, and depression are some of the most prominent and severe manifestations (4–6). These symptoms impair function (7) and negatively impact quality of life (QOL) (4–6, 8, 9). To date, the biologic mechanisms of the symptomatology and patient-reported outcomes of PAH have not been well studied.

Investigators have found a link between inflammation and symptoms in other chronic conditions. Inflammatory cytokines such as IL-6 and tumor necrosis factor (TNF)-α can produce malaise, fatigue, muscle aches, and loss of appetite (10). Inflammation has been linked to fatigue (11), sleep disturbance, pain (12–15), depression (16, 17), and anxiety (18–20). Studies have shown an increase in cytokines (21) in idiopathic and heritable PAH, and higher IL-6 levels are associated with an increased risk of death in PAH (22). Despite these observations, there are few data regarding the associations between cytokine levels and QOL in PAH.

Patient-reported outcomes have gained importance in clinical research (23). Patients are able to relay information that cannot be captured via physical or physiological data. A commonly used patient-reported outcome is the Medical Outcomes Study Short Form-36 (SF-36). The SF-36 contains eight subscales that measure distinct health concepts including physical and mental health components. Using a study sample from a randomized clinical trial of aspirin and simvastatin (24), we investigated the association of IL-6 and TNF-α levels with QOL-related symptoms on the SF-36 subscales for bodily pain, vitality (energy/fatigue), and general mental health (psychological distress and psychological well-being) in PAH. Some of the results of this study have been previously reported in the form of an abstract (25).

Methods

Study Design and Sample

This study was approved by the University of Pennsylvania (#807771) Institutional Review Board. Patients meeting inclusion criteria signed an informed consent form. The patients were participants in the ASA-STAT study, a randomized clinical trial of the safety and efficacy of aspirin and simvastatin in PAH (24, 26). The parent study was a double-blind, placebo-controlled (enteric coated aspirin, 81 mg once daily and/or simvastatin, 40 mg), 2 × 2 factorial multicenter trial that included patients at least 18 years of age with PAH (mean pulmonary artery pressure of >25 mm Hg at rest with a pulmonary capillary wedge pressure < 16 mm Hg).The following clinical subtypes of PAH were included: idiopathic, heritable, or associated PAH with connective tissue disease, HIV infection, congenital systemic-to-pulmonary shunt, or former anorexigen use. Exclusion criteria included absolute indication for aspirin therapy (e.g., coronary artery disease), current statin therapy, peptic or duodenal ulcer, and bleeding (detailed criteria are published) (24, 26).

Procedures

After informed consent was obtained, subjects in the parent study were randomized to enteric coated aspirin (81 mg once daily), simvastatin (40 mg daily), both, or neither (placebo). The primary outcome of the parent study was 6-minute walk distance (6MWD) at 6 months. Secondary outcomes included N-terminal pro–brain natriuretic peptide (NT-proBNP) levels, C-reactive protein (CRP) levels, New York Heart Association (NYHA) functional class, and the SF-36 at baseline, 6 weeks, 3 months, and 6 months (24). Detailed methods have been published and the protocol is available (26).

In this study, we included data from the baseline and 3-month assessments. We incorporated both assessments to provide more power while minimizing the presence of missing data (which occurred for later assessments). Age, sex, race/ethnicity, PAH type, and medication use were collected via subject interviews and chart reviews. Phlebotomy was performed in the morning, using a standardized technique after an overnight fast (except water). Blood was centrifuged and plasma samples were immediately stored at –70°C until analysis.

SF-36.

While the SF-36 overall measures generic QOL, the subscales cover areas relating to physical and mental health concepts. The eight subscales of the SF-36 are combined to form the mental and physical health composite summary scores. These subscales include QOL-related symptoms. A priori, we selected three subscales of the SF-36 as our primary outcomes. The “pain” subscale asks patients how much pain they have had over the past 4 weeks and how much the pain has interfered with their lives. The “vitality” subscale asks questions regarding how much energy patients have and how “worn out” or tired they feel. The “mental health” subscale asks questions about whether patients feel nervous or depressed. Subscales and the composite summary scores are reported on a scale of 0–100 (normalized to the U.S. population), with higher scores indicating better QOL. For normative data, scores higher than 50 are considered to indicate QOL above average (27, 28). Psychometric properties have been established (29); Cronbach α ranges between 0.73 and 0.96.

Laboratory Analyses

Human plasma IL-6 was measured with a high-sensitivity solid-phase quantitative sandwich ELISA provided by R&D Systems, Inc. (Minneapolis, MN) (catalog no. HS600B). The assay has a detection range of 0.156–10.0 pg/ml. The interassay coefficient of variation in the laboratory is 14%. TNF-α plasma was measured with a sandwich ELISA provided by Millipore Corporation (Billerica, MA) (catalog no. EZHTNFA). The assay has a dynamic range of 15.6–1,000 pg/ml. The interassay coefficient of variation is 15%. Detailed analytic methods for other biomarkers have been published (24).

Data Analysis

Continuous variables are shown as means ± SD or medians (interquartile range) and categorical variables as n (percentage). Comparisons between SF-36 baseline and 3-month data points were analyzed using paired t tests. Wilcoxon signed-rank tests were used for IL-6 and TNF-α levels. Generalized estimating equations were used to assess the association between IL-6 and TNF-α (independent variables) and the SF-36 subscales (bodily pain, vitality, and mental health; dependent variables) at baseline and 3 months. Combining baseline and 3-month data points increases the power of our analyses. Generalized estimating equation models were adjusted for those variables thought potentially related to inflammation and QOL: age, sex, race, body mass index (BMI) (30), PAH type, NYHA functional class, 6MWD, NT-proBNP, and randomization to study drug (31). We also adjusted for CRP to account for nonspecific inflammation and subgroup analyses based on the use of medications (prostacyclin analogs or endothelin receptor antagonists). P values less than 0.05 were considered statistically significant. SAS version 9.3 (Cary, NC) (32) was used for all analyses.

Results

ASA-STAT included 65 subjects. Sixty-two subjects had available biomarker analyses and composed the study sample for this analysis (Table 1). The mean age of the participants was 49.7 ± 13.4 years. The majority were female (87%), white (68%), and carried a diagnosis of idiopathic/heritable PAH (56%). More than half were taking phosphodiesterase type 5 inhibitors (63%) and/or endothelin receptor antagonists (56%); 48% received a prostacyclin analog (inhaled, continuous intravenous or subcutaneous); and 81% were taking warfarin. Thirty-nine (63%) of the subjects were NYHA functional class II. The mean 6MWD at baseline was 434.5 ± 113.1 m. NYHA functional class and 6MWD did not change over the 3-month follow-up period. Plasma IL-6 and TNF-α levels did not differ between baseline (3.3 and 3.6 pg/ml, respectively) and 3 months (3.3 and 3.6 pg/ml; P ≥ 0.59 for both). NT-proBNP and CRP levels were also unchanged over 3 months (P ≥ 0.70 for both). SF-36 scores on both sub- and summary scores also remained stable over 3 months (Table 2).

Table 1.

Baseline characteristics of 62 patients with pulmonary arterial hypertension

Characteristic	Value
Age, yr	49.7 ± 13.4
Female sex	54 (87)
Body mass index, kg/m2	27.9 ± 6.9
Race/ethnicity
White (non-Hispanic)	42 (68)
Black	13 (21)
Other	7 (11)
PAH etiology
Idiopathic	32 (51)
Heritable	3 (5)
Congenital heart disease	6 (10)
Connective tissue disease	18 (29)
Drugs/toxins	2 (3)
Overlap syndrome	1 (2)
Concomitant medications
Sildenafil	39 (63)
Ambrisentan	18 (29)
Bosentan	17 (27)
Epoprostenol	15 (24)
Iloprost (inhaled)	10 (16)
Treprostinil (intravenous)	5 (8)
Warfarin	50 (81)
NYHA functional class
Class I	5 (8)
Class II	39 (63)
Class III	18 (29)
6MWD, m	434.5 ± 113.1
NT-proBNP, pg/ml	1,100.3 ± 1,919.6
CRP, mg/L	5.4 ± 9.0

Definition of abbreviations: 6MWD = 6-minute-walk distance; CRP = C-reactive protein; NT-proBNP = N-terminal pro–brain natriuretic peptide; NYHA = New York Heart Association; PAH = pulmonary arterial hypertension.

Data are shown as means ± SD or as n (%).

Table 2.

Medical Outcomes Study Short Form-36 at baseline and 3 months: norm-based scoring

	Baseline	3 Months	P Value
Physical functioning	37.3 ± 10.2	37.5 ± 10.0	0.94
Role-physical	42.6 ± 9.9	42.1 ± 10.6	0.78
Bodily pain	50.2 ± 8.2	49.0 ± 9.0	0.44
General health	41.0 ± 11.2	39.6 ± 9.7	0.46
Vitality	49.4 ± 8.9	48.4 ± 9.2	0.58
Social functioning	47.3 ± 9.7	46.9 ± 9.9	0.78
Role-emotional	48.4 ± 10.2	48.5 ± 11.0	0.94
Mental health	51.3 ± 8.8	51.6 ± 8.9	0.86
Physical component score	42.8 ± 7.5	42.1 ± 7.7	0.59
Mental component score	49.1 ± 7.6	48.9 ± 7.8	0.86

P value for difference between baseline and 3-month assessments; mean ± SD; paired t test.

Analyses using generalized estimating equation models showed that higher IL-6 levels were significantly associated with lower SF-36 subscale scores, indicating worse bodily pain, vitality, and mental health after adjustment for age, sex, race, BMI, PAH type, NYHA functional class, 6MWD, CRP, NT-proBNP, and randomization to treatment group (Table 3). Higher TNF-α levels were significantly associated with lower bodily pain scores, consistent with worse pain-related QOL. Unexpectedly, higher TNF-α levels were associated with better mental health on the SF-36 subscales.

Table 3.

Adjusted general estimating equation models for IL-6 and tumor necrosis factor-α and symptoms

	Estimate	SE	P Value
IL-6 (per SD)
Bodily pain	−1.63	0.01	<0.001
Vitality (fatigue)	−1.90	0.29	<0.001
Mental health (anxiety/depression)	−2.07	0.72	0.004
TNF-α (per SD)
Bodily pain	−0.93	0.04	<0.001
Vitality (fatigue)	1.99	1.08	0.07
Mental health (anxiety/depression)	1.18	0.49	0.02

Definition of abbreviation: TNF-α = tumor necrosis factor-α.

Adjusted for age, sex, race, body mass index, pulmonary arterial hypertension (PAH) type, functional class, 6-minute-walk distance, C-reactive protein, N-terminal pro–brain natriuretic peptide, and randomized PAH treatment group.

Adjusted models stratified by diagnosis showed that limiting the analysis to those with idiopathic/heritable PAH also demonstrated a significant association between IL-6 and bodily pain, vitality, and mental health (Table 4). In contrast, only TNF-α and vitality levels were significantly associated. Unexpectedly, as TNF-α levels increased vitality was improved. Analyses for connective tissue disease were similar (data not shown). There was no significant interaction between sex or diagnosis and IL-6 or TNF-α. The results of the subgroup analyses based on the use of medications (prostacyclin analogs or endothelin receptor antagonists) were consistent with the overall results.

Table 4.

Adjusted general estimating equation models for IL-6 and tumor necrosis factor-α and symptoms in idiopathic and heritable pulmonary arterial hypertension

	Estimate	SE	P Value
IL-6 (per SD)
Bodily pain	−2.06	0.38	<0.001
Vitality (fatigue)	−2.26	0.27	<0.001
Mental health (anxiety/depression)	−2.85	0.98	0.004
TNF-α (per SD)
Bodily pain	1.67	1.61	0.30
Vitality (fatigue)	2.24	0.23	<0.001
Mental health (anxiety/depression)	0.75	0.63	0.24

Definition of abbreviation: TNF-α = tumor necrosis factor-α.

Adjusted for age, sex, race, body mass index, functional class, 6-minute-walk distance, C-reactive protein, N-terminal pro–brain natriuretic peptide, and randomized pulmonary arterial hypertension treatment group.

Discussion

This study aimed to investigate the association between circulating inflammatory cytokines and QOL-related symptoms (bodily pain, vitality, and mental health) in patients with PAH enrolled in a clinical trial. To our knowledge, this is the first study to demonstrate an association between IL-6 and TNF-α levels and certain QOL domains in PAH. As IL-6 levels increased, reported levels of bodily pain, vitality (fatigue), and mental (anxiety/depressions) symptoms also increased. As TNF-α levels increased bodily pain increased. In contrast, as TNF-α levels increased mental health QOL-related symptoms actually decreased.

Higher circulating IL-6 levels are associated with greater self-reported pain affecting QOL in other conditions. Patients reporting increased depressive symptoms associated with greater pain intensity have higher IL-6 levels (13). In older adults, higher IL-6 levels are associated with more pain over a 6-year period after controlling for depression, age, sex, BMI, medication use, and health behaviors (sleep, exercise, and smoking) (15).

IL-6 was also associated with the vitality subscale in our study. Similar to our findings, patients with rheumatoid arthritis with higher levels of IL-6 levels also report increased levels of fatigue (33). In the Whitehall II study, investigators found an association between higher IL-6 and CRP levels and an increase in the incidence of self-reported fatigue (34). IL-6 was also significantly associated with all dimensions of fatigue (general fatigue, physical fatigue, reduced activity, mental fatigue, reduced motivation) in patients with diabetes mellitus type 2 (35). Our analyses were adjusted for CRP, suggesting that IL-6 has a specific biologic effect, rather than just indicating greater inflammation. Our study supports the possible mechanistic role of IL-6 in worsening symptoms and QOL in PAH. In contrast to our findings, a study including patients with heart failure with reduced ejection fraction found no association between IL-6 and self-reported fatigue levels (36).

We demonstrated an association between higher levels of IL-6 and lower SF-36 mental health scores reflecting anxiety and depressive symptoms. Anxiety has been linked to higher IL-6 levels independent of age, sex, and depressive symptoms (37). Higher IL-6 levels were also associated with anxiety and depression in patients with cancer and in those experiencing trauma (38, 39). Omega-3 supplementation significantly reduced IL-6 and TNF-α levels in addition to anxiety levels in a clinical trial (40). A paradigm for this link between inflammation and anxiety has been established in studies of patients with cancer (38).

Higher TNF-α levels were also associated with worse QOL-related pain levels, an association that has been made in patients with rheumatoid arthritis (41). Although unexpected, our study also demonstrated that higher TNF-α levels were associated with lower levels of anxiety and depression. However, others have reported similar findings (42), and TNF-α was not associated with depressive symptoms (43) or anxiety (44) in other studies. We also found that higher TNF-α levels were associated with higher vitality levels in this cohort of patients with PAH. This contradicts literature on other diseases (45, 46). It is possible that the treatments used in PAH might affect TNF levels, accounting for these unexpected results. Current PAH therapies targeting the underlying pathobiology of PAH such as endothelin receptor antagonists (bosentan) have been shown to decrease IL-6 in systemic sclerosis (47) and TNF-α in a murine model (48). Prostacyclin (epoprostenol) has also been shown to reduce IL-6 in individuals with traumatic brain injury (49) and TNF-α (50). A phosphodiesterase type 5 inhibitor (sildenafil) is also associated with lower IL-6 levels in men with diabetes type 2 (51). Physical activity can also lower TNF-α levels in chronic heart failure (52). We did not collect data on physical activity or exercise patterns in the current study.

There are several limitations to this study. First, the small sample size derived from a clinical trial could limit the generalizability of these observations. Despite this, the cohort is demographically and therapeutically comparable to the larger population with PAH (7). Second, although the SF-36 provides insight into how patients feel regarding their physical and mental health including QOL-related symptoms (pain, vitality, and mental health), it does not account for the multidimensionality of symptoms such as fatigue, anxiety, and depression. Another QOL measure, the U.S. Cambridge Pulmonary Hypertension Outcome Review, was not available before the initiation of the original ASA-STAT trial (53). Third, there may be other factors influencing inflammatory biomarkers that were not collected, such as exercise and alcohol intake. Additional inflammatory biomarkers may reveal other mechanisms of symptoms in PAH. Furthermore, the addition of other symptoms that are important, severe, and interfere with the lives of patients with PAH need to be included in future studies.

Conclusions

Ours is the first study to describe the association between the inflammatory biomarkers IL-6 and TNF-α and QOL-related symptoms of pain, fatigue, anxiety, and depression in PAH measured by the SF-36 subscales (bodily pain, vitality, mental health). Although this gives us preliminary evidence of the association between inflammation and symptoms in PAH, more research is needed to further define those possible mechanisms that may lead to targeted interventions to improve symptoms and QOL.