Hospitalized Acute Exacerbation of COPD Impairs Flow and Nitroglycerin-Mediated Peripheral Vascular Dilation

Abstract

Vascular function, as measured by flow mediated dilation (FMD) and nitroglycerin mediated dilation (NMD), is impaired in COPD. Increases in systemic inflammatory mediators during acute exacerbations of COPD (AECOPD) may further impair vascular function and may account for the increased prevalence of cardiovascular disease in COPD patients. Similarly it may account for the increased morbidity and mortality in COPD patients hospitalized with acute exacerbations. We hypothesized that FMD and NMD would be impaired during AECOPD requiring hospitalization and that vascular function would improve upon AECOPD resolution. We used FMD and NMD to evaluate vascular function in 19 patients hospitalized with AECOPD. FMD and NMD were repeated approximately three months later in 8 of these patients. In these eight patients significant improvements were observed in FMD (2.6 ± 1.5% vs 5.1 ± 2.4%, p = 0.04) and NMD (5.0 ± 2.6% vs 13.3 ± 4.5, p = 0.02) after resolution of their exacerbation. We conclude that endothelial and vascular smooth muscle function is markedly impaired during AECOPD requiring hospitalization and improves following resolution. The systemic vascular impairment that occurs during AECOPD may partially explain the observed increased in cardiac morbidity and mortality that occur in this population.

Keywords::

• Chronic Obstructive Pulmonary Disease
• acute exacerbation of COPD
• Vascular Dysfunction

INTRODUCTION

Cardiac morbidity and mortality occurs at rates 2–3-fold higher in stable COPD patients as compared to age matched non-COPD patients (1–3). Additionally, cardiovascular diseases are a common cause of death in patients hospitalized for severe acute exacerbations of COPD (AECOPD) (4–6). Cardiovascular disease and COPD are both associated with elevated markers of systemic inflammation, but the mechanisms behind the relationship of COPD and heightened systemic inflammation are not well understood (7). It is possible that the inflammatory mediators produced in the lung during AECOPD simply “spillover” into the systemic circulation, or that the cause of pulmonary inflammation (usually tobacco smoke) simultaneously produces systemic inflammation. Alternatively COPD may incite biological processes in the lung that kindle systemic inflammation via oxidative stress and tissue hypoxia (8).

AECOPD is known to increase markers of systemic inflammation such as IL-6 and C-reactive protein (CRP) (9–12). Inflammation and oxidative stress are key elements in the development and progression of atherosclerotic disease via their deleterious effects on endothelial cells (EC) and vascular smooth muscle cells (VSMC) (13, 14). Therefore, alterations in vascular function represent a potential pathophysiologic link between COPD and atherosclerotic cardiovascular disease.

Vascular smooth muscle and the vascular endothelium both play a role in the maintenance of normal arterial function. Each can be assessed by flow-mediated dilation (FMD) and nitroglycerin mediated dilation (NMD) of the brachial artery. These parameters have been shown to be reliable indices of cardiovascular risk (15–17). FMD is a nitric oxide (NO) mediated, endothelial dependent process that correlates with the presence and degree of coronary disease and coronary artery endothelial dysfunction, findings which have been confirmed by left heart catheterization (17, 18).

Impaired NMD suggests that the vascular dysfunction is due to non-endothelial factors such as atherosclerosis or smooth muscle dysfunction (19). Both FMD and NMD have been used to evaluate the impact of inflammatory diseases on endothelial function (16, 20–22). Furthermore, transient increases in systemic inflammation such as that associated with immunization have been shown to temporarily impair both FMD and NMD (22). Vascular dysfunction measured by FMD has been shown to exist in patients with stable COPD (23, 24). The level of impairment was associated with elevated markers of systemic inflammation, worse lung function, and more emphysema present on radiographic imaging (23, 24).

AECOPD is associated with an increased systemic inflammatory state. It might be expected that during an AECOPD vascular function would worsen and might be related to the cardiac and prothrombotic events that have been described in the clinical presentation of patients with AECOPD (4, 25). We hypothesized that the acute inflammatory state induced by AECOPD impairs vascular function measured by FMD and NMD and improves following exacerbation resolution. Some of this data was presented previously as an abstract (26).

METHODS

Study patients

Patients were identified from those hospitalized for AECOPD, which was defined using previously published guidelines (27). Patients were 40–80 years old and had ≥20 pack-year history of smoking. Patients were excluded if there was pneumonia or pulmonary edema on chest radiograph, acute coronary syndrome, or history of peripheral vascular disease or aortic aneurysm. Healthy nonsmokers were studied to ensure that FMD could be performed reproducibly compared to published normal control data, and were men and women age 40–80, who did not have coronary artery disease, diabetes, uncontrolled hypertension, COPD, lung cancer or interstitial lung disease. The protocol (#4593) was approved by the Temple University Institutional Review Board. All patients provided written informed consent.

Study design

Figure 1 shows the study design and patient flow throughout the study. Patients underwent bedside spirometry, history and physical exam. All patients had FMD and NMD measured within 72 hours of hospitalization (FMD-1 and NMD-1). Patients were asked to return for repeat measurement of FMD and NMD (FMD-2 and NMD-2) 3 months following hospitalization.

Figure 1  Study design: 19 patients were enrolled in the study at time of AECOPD admission. All 19 had FMD but 1 patient was unable to receive sublingual nitroglycerin secondary to systolic blood pressure below 100 mmHg. Then, 11 patients did not return for repeat FMD or NMD because 7 were lost to follow-up, 2 had repeat AECOPD, 1 was transplanted and 1 died. Eight of the remaining patients had FMD. One patient was unable to have NMD at follow up due to systolic blood pressure <100 mm Hg. Definition of abbreviations: AECOPD (acute exacerbation of COPD), FMD (Flow-mediated dilation), NMD (Nitroglycerin-mediated dilation).

Flow-mediated dilation

FMD and NMD of the brachial artery were performed according to established guidelines and previously published studies (28, 29). Prior to FMD or NMD patients fasted for 12 hours, and did not receive bronchodilators or venipuncture within 4 hours. Patients were asked to abstain from smoking for 24 hours and all reported they had done so prior to FMD and NMD. All vascular studies were performed in a quiet room with the subject lying down. The nitroglycerin phase was omitted if the systolic blood pressure was <100 mmHg. FMD and NMD were calculated as the percent change in brachial artery diameter from baseline to flow or nitroglycerin mediated dilation.

Demographic data

Demographic data was collected from patient interview and review of the medical record. Cardiovascular risk was calculated using the ATP (Adult Treatment Panel) III calculator according to the Third Report of the National Cholesterol Education Program Expert Panel. Cardiomyopathy was defined as left ventricular ejection fraction <50%, and hypercholesterolemia was defined as a total cholesterol >200 mg/dL. Hypertension, diabetes, and depression were self reported, and the patient's outpatient medication list was documented at time of enrollment.

Spirometry

Spirometry performed by the patient within the previous year following ATS guidelines (30) was used as baseline spirometry. Spirometry data are presented as percent predicted using prediction equations from NHANES III (31).

Statistical methods

The t-test, paired t-test, and Pearson correlation were used where appropriate. Univariate analysis was performed to identify the factors influencing FMD or NMD. Small sample size prevented a multivariate analysis from being performed. A Wilcoxon signed rank test was used to compare measurements of FMD and NMD at AECOPD and following resolution of AECOPD because the data was not normally distributed.

Table 1  Comparison of patients at admission and those that had follow up FMD

	AECOPD (n = 19)	COPD Follow-up (n = 8)	P value	Healthy non-smokers (n = 9)
Age, yrs	61 ± 7	61 ± 8	1.0	53 ± 6
Caucasian, #(%)	9 (47.4)	5 (62.5)	0.77	9 (100)
Men, #(%)	12 (63.1)	4 (50)	0.84	6 (66)
FEV1,% predicted (n = 16)	26 ± 16	33 ± 22	0.36	n/a
pH (n = 11)	7.37 ±.07	n/a	n/a	n/a
PaCO2 (mmHg) (n = 11)	54.5 ± 17.4	n/a	n/a	n/a
Smoking burden, pack yrs	52 ± 25	51 ± 22	0.92	0
BMI (Kg/m^2)	27 ± 7	29. ± 7	0.66	n/a
ATP risk (n = 12)	7.4 ± 3.3	9.1 ± 3.3	0.23	4.6 ± 1.4
Current smoking, #(%)	3 (15.8)	1 (12.5)	0.71	0
Diabetes, #(%)	2 (10.5)	1 (12.5)	0.59	0
Hypertension, #(%)	10 (52.6)	4 (50)	0.76	1 (11)
Cardiomyopathy, #(%)	1 (5.3)	0	0.65	0
Hypercholesterolemia, #(%)	10 (52.6)	5 (62.5)	0.96	1 (11)
Depression, #(%)	6 (31.6)	2 (25)	0.91	1 (11)
Coronary artery disease, #(%)	2 (10.5)	0	0.88	0
ACE inhibitor, #(%)	3 (15.8)	2 (25)	0.88	0
Beta-blocker, #(%)	2 (10.5)	0 (0)	0.88	0
Calcium channel blocker, #(%)	3 (15.8)	1 (12.5)	0.71	0
Statin, #(%)	7 (36.8)	2 (25)	0.88	1 (11)
BP (systolic) mmHg	133 ± 19	133 ± 20	1.0	127 ± 14
BP (diastolic) mmHg	76 ± 10	72 ± 10	0.35	72 ± 12
BP (mean) mmHg	95± 10	92 ± 9	0.47	89 ± 10
HR, bpm	92 ± 16	93 ± 19	0.89	75 ± 8
Oxygen requirement, LPM	2.8 ± 1.4	1.8 ± 1.2	0.09	0

Definition of abbreviations: BMI (body mass index), ATP (adult treatment panel), ACE (angiotensin converting enzyme), BP (blood pressure), HR (heart rate), bpm (beats per minute), LPM (liters per minute), n/a (not available).

There was no statistical difference between the AECOPD and COPD follow-up groups.

RESULTS

Demographics

Nineteen patients (12 men and 7 women) age 61.1 ± 7.2 yrs, FEV₁ 26.1 ± 16% predicted, admitted for AECOPD had FMD and NMD measured. Eight patients had measurements at follow-up, 117 ± 33 days post discharge from the hospital. The 11 patients who were lost to follow-up were unable to be contacted (7), were in exacerbation at time of follow-up (2), had died (1), or underwent lung transplantation (1). The one death was related to respiratory failure during a subsequent AECOPD at another institution. Demographic and physiologic characteristics of all 19 patients studied during hospitalization for AECOPD and the subgroup of 8 patients who had measurements at follow-up were not different and are summarized in Table 1.

Vascular function in AECOPD and resolution

The healthy nonsmokers we studied demonstrated an FMD of 10.8 ± 4.7%, which is consistent with values reported in the literature for normal, non-smoking adults and confirms our ability to perform and interpret FMD. In AECOPD FMD was markedly reduced compared to normals. (2.8 ± 1.7% vs.10.8 ± 4.7%, p < 0.001) There was a significant improvement in FMD following resolution of AECOPD at 3 months (2.6 ± 1.5 vs 5.1 ± 2.4%; p = 0.04) as shown in Figure 2. One patient had a systolic blood pressure below 100 mmHg, and did not undergo NMD measurement during the acute exacerbation. NMD during AECOPD was also markedly impaired compared to healthy non-smoking controls (8.0 ± 4.3% vs 21.4 ± 6.0%; p < 0.001). Figure 3 shows a significant improvement of NMD from exacerbation (NMD-1) to follow-up NMD (NMD-2) (5.0 ± 2.6% vs 13.3 ± 4.5%; p = 0.02) in the 7 patients who completed NMD during the exacerbation and after resolution. There was no difference in baseline brachial artery diameter at time of AECOPD and following resolution of AECOPD 3 months later (0.415 ± 0.068 cm vs 0.414 ± 0.080 cm; p = 0.9).

Figure 2  Individual FMD in the 8 patients that had measurements at admission (FMD 1) and follow-up (FMD 2). The horizontal bar represents the mean FMD at that time point. FMD improved from 2.6 ± 1.5% at admission to the hospital to 5.1 ± 2.4% on follow-up (p = 0.04).

Figure 3  Individual NMD in the 7 patients that had measurements at admission (NMD 1) and follow-up (NMD 2). The horizontal bar represents the mean NMD at that time point. NMD improved from 5.0 ± 2.6% to 13.3 ± 4.5% on follow-up (p = 0.02).

Univariate analysis of FMD

Univariate analysis was performed to determine factors that correlated with FMD or NMD in AECOPD. Age and NMD were positively correlated with FMD (r = 0.55, p = 0.02 and r = 0.52, p = 0.02, respectively). Neither FMD nor NMD were affected by FEV₁, gender, race, or the use of beta-blockers, calcium channel blockers, or angiotensin converting enzyme inhibitors.

DISCUSSION

Patients hospitalized for AECOPD had severe vascular dysfunction as evidenced by markedly reduced FMD and NMD. Upon recovery, both FMD and NMD improved significantly. Not only were both FMD and NMD dramatically lower than normal individuals (32) but FMD in our cohort was significantly lower than that reported for other systemic diseases with well-known vascular complications (16, 21). To our knowledge this is the first report showing impaired vascular function in patients with hospitalized AECOPD that improves following resolution of the acute exacerbation.

Vascular dysfunction has previously been shown in stable COPD (23, 24, 33). Sabit et al. showed that COPD patients had evidence of vascular stiffness as measured by aortic pulse wave velocity (34), and that the severity of airflow obstruction, age, IL-6 level, and arterial blood pressure were independent predictors of vascular stiffness. Barr et al. found that FMD was impaired in a group of former smokers and that the magnitude of impairment was more severe in those COPD patients with lower FEV₁ and more emphysema evident on chest CT imaging (24). Two other studies have shown that COPD patients had lower FMD and NMD when compared to normal controls and that NMD was lower in COPD patients compared to non-smoking controls but not to smokers without COPD (23, 33). Our data is consistent with previously reported impairments in both FMD and NMD in COPD, but our data highlights the effect of an acute exacerbation on vascular function.

Our study is unique in that we studied FMD and NMD in subjects during an acute severe exacerbation that resulted in hospitalization. This may have important clinical implications since patients hospitalized with AECOPD have been reported to have high incidences of pulmonary embolism or cardiac events. Venous thromboembolism (VTE) occurs 5 times more commonly in COPD and can be an unexpected cause of AECOPD in up to 25% of these patients (25). In a recent autopsy study of 46 patients who died within 24 hours of hospitalization for AECOPD, patients were found to have died from congestive heart failure (CHF) in 37% and VTE in 21% (4).

Systemic inflammation has been shown to be greater during AECOPD (9, 12) and the severity of FMD and NMD impairment correlates with inflammatory markers even in stable COPD (23, 34). This led to our premise that vascular function as measured by FMD and NMD would be further impaired during an exacerbation. The magnitude of vascular impairment in our patients both during exacerbation and at resolution was worse than most of the previously published reports on FMD in COPD (23, 24). This may be due to differences in technique (forearm occlusion vs. upper arm occlusion), our patient population being more severely obstructed, or our patients not achieving full resolution of their exacerbation. Acute hypercapnia can result in vasodilatation which may impair the capacity of the vessel to respond to stimuli such as FMD or NMD. However, in our patients there was no significant difference in the baseline brachial artery diameter during AECOPD compared to AECOPD resolution, thus making it unlikely that the impaired FMD and NMD was due to arterial vasodilatation from hypercapnia.

Flow mediated dilation is a response to nitric oxide released by the vascular endothelium resulting from distention of the vessel following sudden return of flow to the previously obstructed vessel. FMD reflects endothelial function; while NMD reflects endothelium independent function (i.e., vascular smooth muscle), which could be due to atherosclerosis or impaired vascular smooth muscle function (28). In our study FMD and NMD, both of which are risk factors for cardiovascular disease, were markedly impaired suggesting abnormalities in both endothelial and vascular smooth muscle function (19). Others have also shown that there is impairment in both FMD and NMD in patients with stable COPD (23, 24). Cardiovascular disease occurs at a 2–3-fold increased risk in COPD patients (1–3, 35), and both cardiovascular disease and COPD are associated with increased markers of inflammation (7).

It is possible that the increased systemic inflammatory state associated with COPD leads to an accelerated development of cardiovascular disease in this patient population. Patients hospitalized for AECOPD are at increased risk for mortality during and after the admission (36–38). The increased inflammatory state associated with AECOPD may even further accelerate cardiovascular diseases and place patients at an increased risk of death due to cardiac events. This heightened risk is theoretically temporary because once the exacerbation is treated; vascular function returns to baseline, although still impaired compared to the normal state.

Limitations of this study include a small sample size, which limited statistical power and prevented multivariate analysis. FMD can be impaired in advanced age and two of our subjects (both had FMD during hospitalization and one during follow-up) were above age 70. However, there still was an improvement in the FMD following resolution of the AECOPD suggesting that the exacerbation altered FMD even in those of advanced age. Our patients were not free from cardiovascular risk factors, which may have lowered their FMD. However, because the FMD improved following resolution of AECOPD we feel that the exacerbation worsened any pre-existing vascular dysfunction that was present.

All of our patients were treated with corticosteroids which may have affected FMD and NMD. The effect of acute administration of glucocorticoids on vascular function is not entirely understood and conflicting evidence exists. One study reported that corticosteroids alter vascular function while a second reported that the acute administration of corticosteroids has no effect (39, 40). Although all of our patients were treated acutely with corticosteroids none of them were on chronic therapy prior to AECOPD hospitalization.

Another limitation was the loss of patient follow-up for repeat FMD and NMD. This was largely due to the severity of the underlying disease of patients hospitalized for AECOPD. A larger study with a greater rate of follow-up will be needed to definitely characterize the changes in FMD and NMD that occur during exacerbation and following resolution of patients hospitalized with AECOPD.

In conclusion, this study demonstrates that both the endothelium and vascular smooth muscle contribute to vascular dysfunction COPD. The acute and chronic systemic vascular effects of COPD were shown by impairment in both FMD and NMD during an acute exacerbation, which improved but remain below normal at resolution. Further research is needed on a larger scale to elaborate the mechanisms of vascular impairment during AECOPD and to explore potential treatments.