Pulmonary hypertension (PH) is a common complication of chronic obstructive pulmonary disease (COPD).1 The prevalence of PH in COPD patients it is not negligible, it has been reported from 23% to 91%, depending on the severity of the disease and the diagnostic criteria used to define it.2–5 Clinically, PH increases the risk of hospitalization for COPD exacerbation and entails a lower survival rate.3,6

Patients with COPD typically show reduced exercise tolerance compared with healthy controls with similar anthropometric characteristics. Several causal factors have been proposed. On one hand, increased ventilatory requirements and abnormal dynamic ventilatory mechanics7 are thought to be the main limiting factor, although peripheral muscle weakness8 and right ventricular diastolic function9 have also shown to limit exercise capacity in COPD patients. Moreover, the presence of PH has been hypothesized to further decrease exercise tolerance.3,10 It is necessary to characterize pulmonary hemodynamics in patients with moderate to severe COPD-PH in order to improve their management and to choose appropriate rehabilitation strategies.

Several studies have attempted to clarify whether the exercise limitation in COPD patients with PH is primarily due to the ventilatory impairment or to the vascular component with contradictory results. Sims et al.10 found that in patients with severe COPD the presence of PH was associated with lower distance covered in the six-minute walk test (6MWT), regardless of pulmonary function, anthropometric or demographic characteristics. Also, they showed that higher pulmonary arterial pressure (PAP) was associated with lower exercise capacity. More recently, the presence of PH was found to worsen maximal exercise capacity in a group of severe COPD patients.11 However, there are also opposite studies suggesting that exercise impairment is due, basically, to the ventilatory limitation.12–15 The majority of previous studies present some limitations, such as a selective sample only including severe COPD patients candidates for lung transplantation,12 small sample size14 or inaccurate definition of PH,15 that might complicate getting clear conclusions.

The 6th World Symposium on Pulmonary Hypertension (WSPH) “encourage cardiopulmonary exercise test (CPET) for a more elaborate distinction between pulmonary versus circulatory limitation” with the aim to discriminate between different groups of PH. Specifically between the group 1 (pulmonary arterial hypertension, PAH) with a predominant cardiovascular profile (exhausted circulatory reserve) versus the group 3 with a predominant obstructive/restrictive profile (exhausted ventilatory reserve).16 This distinction arise from the study by Boerrigter et al.17 who showed that COPD patients with severe PH (mean PAP ≥40mmHg) had an exhausted circulatory reserve at the end of exercise with a reduced slope of the cardiac output/oxygen consumption ratio, low mixed venous oxygen saturation and low arterial partial pressure of CO2 (PaCO2), whilst the breathing reserve was preserved. On the other hand, COPD patients with moderate PH or without PH showed increased PaCO2 and exhaustion of the breathing reserve at the end of the exercise, whereas the circulatory reserve was maintained.17

Therefore, we aimed to determine whether in a cohort of stable COPD patients the presence of PH is associated with reduced exercise tolerance, as assessed by maximal and submaximal exercise tests, and worse quality of life.

The main finding of this study is that the presence of PH was associated with impaired exercise capacity in COPD patients adjusted for age, BMI and airflow obstruction. Patients with increased PAP reached lower maximal work-load while cycling, presented lower VO2peak, covered less distance in the 6MWT, developed greater O2 desaturation during exercise, and had more dyspnea and worse quality of life. Although the ventilatory impairment and the reduced PaO2 are important factors limiting exercise in COPD patients, the multivariate analysis shows that PH per se has a detrimental effect on exercise capacity for a given degree of ventilatory impairment.

Although these results are expected, it is necessary to characterize pulmonary hemodynamics in patients with moderate to severe COPD and to analyze their response to exercise to choose better therapeutic strategies, especially for those who are candidates for pulmonary rehabilitation. In keeping with this, previous studies have shown that pulmonary rehabilitation improve exercise capacity in patients with COPD-PH but concomitant treatment with pulmonary vasodilators do not improve the results of the pulmonary rehabilitation program in exercise tolerance.27 Additionally the most recent Task Force on Pulmonary Vascular Diseases also recommended pulmonary rehabilitation in patients with COPD and moderate PH.16

Patients with COPD-PH showed a profile of exhausted circulatory reserve during exercise. The oxygen pulse, that provides an estimate of left ventricle stroke-volume changes during exercise,28 was significantly lower in the COPD-PH group (54% predicted) than in COPD patients without PH (73% predicted). These results are in line with the study by Boerrigter et al.17 who showed in their patients with severe PH an exhausted circulatory reserve at the end of the exercise while the breathing reserve was maintained.17

Exercise limitation in patients with COPD was initially associated as a direct consequence of airway obstruction, but this association has shown only a weak relation.29 The dynamic hyperinflation leads by the limitation of expiratory flows has been associated with functional weakness of the inspiratory muscles and with the restriction of thoracic expansion of normal tidal volume during exercise.30 On the other hand, changes in skeletal muscle are associated with exercise limitation.31 In the case of the presence of PH in patients with COPD, the literature describes the limited exercise capacity in patients with COPD and abnormal pulmonary artery pressure.15 Our group showed differences between lung function, but when correcting by this factor, it still appears this difference in exercise capacity, that can be explained by the increase of PAP.

We observed that the presence of PH decreased oxygen uptake and oxygen pulse, in addition to the ventilatory impairment typically observed in COPD patients. Based on the Fick principle, O2 pulse depends on both stroke volume and arteriovenous oxygen difference (Ca-vO2). For the cardiovascular nature of PH, these patients have demonstrated alterations in stroke volume.32 On the other hand, abnormalities on arterial blood gases have been described as playing a role in exercise intolerance.33 In patients with PH, hypoxemia accelerates the early occurrence of lactic acidosis increasing the ventilatory drive (by stimulation of the carotid bodies) contributing to an excessive increase in VE during exercise.34 This is another contributing factor for the reduction in exercise tolerance in our patients, reproduced at both exercise tests.

Cuttica et al., in a retrospective analysis of almost 5000 patients with advanced COPD, candidates for lung transplantation, reported that PH of any cause resulted in significant functional impairment measured by 6MWT and this result was associated with an increased risk of mortality.3 Likewise, Sims et al., in 362 patients with severe COPD who were evaluated for lung transplantation, demonstrated that higher values of mPAP were associated with impaired exercise tolerance measured by 6MWT, independently of the severity of the airflow obstruction.10 Our results are in line with these previous reports in a population of patients less severely impaired; we also found that the 6min walk distance (6MWD) in our group with COPD-PH was about 100m lower than in the group without PH (p<0.001). Contrasting with these findings, the 6MWD did not differ in other similar studies although evaluating a smaller sample of patients.11,12,14

To overcome these discrepancies, in our study we assessed comprehensively exercise tolerance by using maximal and submaximal exercise tests. We confirmed the impairment in exercise capacity in the COPD-PH group also in the CPET, the gold standard measurement for exercise capacity.

In our population, the VO2peak at the CPET was 33% lower in the group with PH, which is in line with previous studies.11,15 Although this finding could be due to lower FEV1 in the COPD-PH group, in the model adjusted by the FEV1 value, the probability of having a VO2peak<60% predicted was 3.5 higher in the COPD-PH group than in the non-PH COPD group.

Similarly, Vonbank et al., showed a significant decrease of 25% in VO2peak in COPD patients with PH, as compared with patients without PH.15 Thirapatarapong et al. also showed a reduction of 16% in VO2peak in COPD-PH patients.11 Of note, in both studies, there were no differences in the severity of airflow obstruction between groups. In contrast, other studies did not find a significant difference in VO2peak when comparing COPD-PH with non-PH COPD patients.12,35

In our study, the maximum workload achieved at the CPET was significantly lower in the COPD-PH group than in the non-PH group. This is in line with Thirapatarapong et al., who found significant difference in severe COPD patients (21(15)% predicted for COPD vs. 15(9)% predicted COPD-PH) and Skjorten et al., who studied 98 patients with a broad range of airway obstruction (72(31) watts for COPD vs 40(21) watts for COPD-PH).11,35 However, Holverda et al., Vonbank et al., Pynnaert et al., and Adir et al. did not find any difference.12–15 Furthermore, our results are more in line with the mechanistic explanation for PH because this loss of lung vasculature and distensibility is responsible for limiting exercise capacity.

In this type of patients, it is very critical the evaluation of the quality of life or multidimensional index like BODE index.36 Both measurements include the assessment of exercise tolerance and, we found a significant difference in both outcomes in the COPD-PH group (Table 1). Although these patients should be characterized with objective evaluations, we should not forget that patients are concerned about their symptoms and how these affect their quality of life, so they should be incorporated in the evaluation of the exercise capacity to explore the interaction between physical performance and quality of life.

Furthermore, patient's characteristics showed differences in pulmonary function, dyspnea and quality of life between COPD-PH and non-PH COPD group. These differences might explain, in part, the difference in maximum workload or VO2peak between the groups. However, we applied a logistic regression model with the objective to adjust important variables that influence during exercise, such as age, BMI, and very important in respiratory patients, FEV1%pred. These models were applied in submaximal (6MWD<450m) and maximal (VO2peak<60% pred) exercise tests, and in all these models, we found differences in exercise capacity between both groups. All these findings may suggest that the presence of PH plays an important role in limiting exercise capacity.

The characterization of pulmonary hemodynamics and exercise tolerance in patients with COPD-PH can contribute to improving their management, especially in the rehabilitation strategies. Recently, the European Respiratory Society (ERS) and the American College of Chest Physicians had developed guidelines for exercise in PH37,38; particularly, the ERS declares that the supervised exercise training may improve right ventricular function and pulmonary hemodynamics in patients with stable PH. Improved hemodynamics may contribute to an increase in exercise capacity and quality of life of patients. Unlike other chronic respiratory diseases, PH requires strict supervision of exercise in specialized centers, so the characterization of patients is essential in order to establish the most beneficial and least risky programs for each one.

This study has some limitations that ought to be disclosed. The first limitation of the study refers to the inclusion rates, which were different by the presence/absence of PH – this is, of course, important in assessing the risk of bias. Secondly, PH was assessed mostly through echocardiography. Although RHC is an invasive test not recommended for routine assessment in COPD patients, non-invasive measurement of PAP has shown good correlation with RHC39,40 and echocardiography is an acceptable screen for PH. Thirdly, in the context of COPD, pulmonary hypertension may involve the left heart (diastolic dysfunction possibly secondary to hyperinflation, associated systolic dysfunction) but, unfortunately, in our study, we could not exclude diastolic dysfunction in all patients (only in the 21% that had a RHC).

Finally, the decision to lower 10% the load in the constant-work test for PH patients precluded us to compare both groups statistically. Still, COPD-PH group reached shorter endurance time, and we can argue that with a 10% more of work-load the difference might have been significantly greater. We consider that endurance time is a very useful outcome to compare the results of an intervention such as a pulmonary rehabilitation program, especially in patients with PH because it reflects the resistance of the general musculature.

In conclusion, PH in COPD patients impairs their exercise tolerance in maximal and submaximal tests and demonstrated a detrimental effect per se, irrespective of age, BMI, FEV1%pred and PaO2. Pulmonary hemodynamics should be considered to better understand the exercise intolerance in patients with moderate to severe COPD and associated PH to improve their management.

The study was supported by grants from the Fondo de Investigación Sanitaria, Instituto de Salud Carlos III (PI17/1515), Fondo Europeo de Desarrollo Regional (FEDER), Unión Europea. “Una manera de hacer Europa” Sociedad Española de Neumología y Cirugía Torácica (SEPAR) and Societat Catalana de Pneumologia (SOCAP). Elena Gimeno-Santos had personal funding from Instituto de Salud Carlos III, Sara Borrell contract (Acción Estratégica en Salud 2013–2016).

The authors declare no conflict of interest.