Physiological principles
Pulmonary rehabilitation (PR) is ‘an evidence-based, multidisciplinary and comprehensive intervention for patients with COPD that is designed to reduce symptoms, optimize functional status, increase patient participation and reduce healthcare costs through stabilizing or reversing systemic manifestations of the disease’ [
31]. Exercise training is a major component of PR, and aims to modify skeletal muscle function to enhance exercise capacity [
32,
33]. Improved skeletal muscle performance and subsequent reduced lactate production throughout exercise enhances the relationship between respiratory muscle load and respiratory muscle capacity. This is achieved through modifications in breathing pattern, in the context of airflow limitation with optimization of pulmonary mechanics, which reduces exertion-related dyspnea, with a resultant improvement in exercise capacity leading to further improvement in skeletal muscle performance [
34,
35].
In addition to impaired exercise capacity and disuse in the stable state, several other mechanisms have been implicated in the muscle wasting and weakness associated with exacerbation [
36]. Systemic inflammation, confirmed by an elevation in serum interleukin (IL)-6 and IL-8 levels during acute illness, has been shown to have an inverse relationship with quadriceps muscle strength [
37]. In addition, oxidative stress is prominent in the peripheral skeletal muscle of patients during an acute exacerbation, which adversely affects mitochondrial function and the contractile properties of the skeletal muscle [
38]. Furthermore, blood gas abnormalities affect skeletal muscle function, with hypoxemia being associated with muscle weakness, inhibition of protein synthesis, and activation of proteolysis [
39], while hypercapnic acidosis worsens skeletal muscle fatigability [
40] and the endurance properties of the diaphragm [
41]. Muscle wasting as a consequence of systemic corticosteroid treatment through the inhibition of protein synthesis, downregulation of the anabolic insulin-like growth factor-1 pathway, and activation of catabolic pathways, combined with appetite suppression and reduced dietary intake due to systemic inflammation during acute illness, drives the energy imbalance between supply and demand [
42].
Although exercise training plays an important role in improving patient outcomes in COPD, other components contribute significantly to the benefits of PR. PR also addresses the nutritional deficits that are common in stable COPD and during acute exacerbations [
43]. Low body mass index is associated with poor prognosis in patients with COPD [
44], and caloric supplementation may help to maintain or restore body weight and fat mass, and ensure adequate protein intake. The patient education component of PR is aimed at self-management and enhanced autonomy, in order to encourage early self-identification and treatment of exacerbations. Education programs may also include breathing strategies to control dyspnea as well as bronchial hygiene techniques [
43]. Psychosocial support may help to address the anxiety, depression, and other mental health problems that are often associated with chronic respiratory disease [
45,
46].
Pulmonary rehabilitation delivered in the post-acute exacerbation recovery stage
Although well established as part of chronic care in stable COPD [
43,
47], there is mounting evidence for the utility of PR in the early recovery period following an exacerbation [
48]. Furthermore, data support the role of PR in preventing exacerbations and in reducing acute healthcare utilization, including unscheduled physician visits, ED attendances, and hospital admissions [
49]. There are specific features of acute exacerbations that make them an important target for PR. Skeletal muscle dysfunction is evident, with a decline in quadriceps muscle strength of 5% between day 3 and 8 of hospital admission [
37]. In the absence of any intervention, quadriceps force continues to decline for up to 3 months after hospital discharge [
50]. Immobility and reduced physical activity are major contributors to muscle wasting and weakness, with hospitalized patients spending less than 10 minutes per day walking [
51]. Furthermore, these patients remain inactive for up to 1 month after discharge compared with patients with stable COPD and similar disease severity.
Patients are at high risk of re-exacerbation and re-admission in the early recovery phase. Therefore, there is a potential role for an intervention in the post-exacerbation period after an acute episode to reduce the re-admission risk. A recent Cochrane systematic review of five randomized controlled trials (RCTs) of early PR post-acute exacerbation [
48] concluded that there was a significant reduction in hospital admissions in patients enrolled in PR programs following an exacerbation (odds ratio 0.22, 95% confidence interval 0.08 to 0.58). More importantly, these data showed that only four patients need to receive PR in the post-acute phase in order to prevent one re-admission, with an overall reduction in mortality observed also (OR 0.28; CI 0.10 to 0.84). There were no serious adverse events in any of the five studies reviewed. Although Eaton
et al. showed only a trend towards a reduction in 90-day re-admission in the PR group compared with the usual care (UC) group (23% versus 32%, respectively), the adherence to the PR program was only 40% [
52]. However, in the trial of Seymour
et al., the effect of early outpatient PR on 3-month re-admission rate was investigated in patients enrolled within 1 week of hospital discharge [
50]. The intervention group in this study received 16 exercise training sessions over a 3 month period rather than the standard PR approach of 2 sessions per week for 8 weeks. This ensured that the effect of the treatment was tested, whereas the results of the trial by Eaton
et al.[
52] were, in part, a consequence of the failure of delivery of the treatment rather than necessarily a failure of the treatment itself. In the Seymour study, re-admission rate at 3 months was lower in the early PR group compared with the UC group (7% versus 33%, respectively;
P = 0.02) [
50]. Interestingly, Seymour
et al. reported that the rate of ED attendances not requiring admission were similar between the two groups, but the rate of hospital attendance of any type was lower in the early PR group (27% versus 57%;
P = 0.02). The post-discharge frequency of exacerbations was lower in the early PR group (0.27 versus 1.1;
P < 0.01). Although informative, these trials were limited by a relatively short-term follow-up period. By contrast, Ko
et al. investigated the effect on healthcare utilization at 12 months of an 8-week program of supervised outpatient PR in patients enrolled up to 3 weeks following hospital discharge [
53]. Although the PR group showed improvement in health status at 3 and 6 months, this effect did not persist at 12 months, and there was no reduction in healthcare utilization at 12 months. Similarly, Puhan
et al. reported, albeit in an underpowered study, that early PR failed to improve exacerbation rate at 18 months [
54]. This is not surprising as the patients enrolled in such trials have severe and very severe COPD, and the interventions that are applied are unlikely to have effect on long-term benefit as the disease process progresses. Despite this, the short-term gains to the patient and acute healthcare providers are clear. In the future, we may target these patients during the exacerbation as inpatients. Acknowledging that these patients have very high levels of dyspnea during this period, which prevents exercise, novel use of technologies that accommodate for or modify dyspnea, such as neuromuscular electrical stimulation [
55] and non-invasive ventilation [
56], have been used as adjuncts to exercise training in pilot studies, but further work is required.
Non-invasive ventilation
Non-invasive ventilation (NIV) is well established as the treatment of choice for patients with COPD with acute decompensated hypercapnic respiratory failure (AHRF) who fail to respond to standard medical therapy [
65,
66]. Importantly, and in addition to a reduction in mortality, hospital length of stay is reduced compared with standard treatment. Although NIV remains controversial as a domiciliary treatment to reduce hospital admission and improve survival in patients with COPD with stable chronic respiratory failure, the physiological mechanisms by which long-term NIV results in clinical improvement in patients with severe COPD and hypercapnic respiratory failure are well-described [
67]. Indeed, Nickol
et al. showed that 3 months of NIV enhanced gas exchange through alterations in pulmonary mechanics (shown as reduced gas trapping), and also increased ventilatory sensitivity to carbon dioxide. However, there was limited effect on non-volitional muscle strength [
68]. Clinical manifestations of these physiological changes are reflected as reduced dyspnea and improved HRQL, and it is hypothesized that there will be an associated reduction in acute exacerbations and hospitalization, with a potential for improved survival. However, trial data are as yet inconclusive for this high-risk group of patients with severe COPD.
An observational study from Tuggey
et al. showed that, following initiation of domiciliary NIV in a cohort of patients with COPD who were prone to recurrent admissions, there was a significant reduction in total hospital days and days spent in the intensive care unit. This was, not unexpectedly, associated with substantial cost savings [
69]. This is in contrast to several RCTs that have failed to show a convincing benefit in terms of acute healthcare utilization. In the trial by Casanova
et al., there was no difference in survival between patients with stable COPD randomized to NIV or UC [
70], although the proportion of patients who required hospital admission at 3 months was reduced in the intervention group (5% versus 15%;
P < 0.05, respectively). Unusually, ventilator set-up in this trial was aimed at reducing accessory muscle use and reducing dyspnea, which explains, in part, the low inspiratory positive airway pressure (IPAP) of 12 cm H
2O applied. Clini
et al. randomized 90 patients to long-term oxygen therapy (LTOT) alone or home NIV (IPAP 14 cm H
2O) with LTOT, as part of a multicentre trial [
71]. Adherence to NIV was high in this study at 9 hours per day, but there was only a trend to a reduction in hospital admission comparing admission rate before and after enrolment (45% decrease in hospital admissions in the intervention group versus 27% increase in the UC group). More recently, McEvoy
et al. found a significant improvement in survival in a combined NIV and LTOT group in both intention-to-treat and per-protocol (>4 hours NIV use per night) analyses (HR 0.63, 95% CI 0.40 to 0.99,
P = 0.045 and HR 0.57, CI 0.33 to 0.96,
P = 0.036, respectively) [
72]. This was achieved with an adherence of 4.3 hours per night. Despite this beneficial effect, NIV in addition to LTOT treatment conferred no benefit in terms of HRQL or hospital admission, albeit the IPAP in this trial was again low, at 12.9 cm H
2O.
Patients with COPD are at greatest risk of death and re-admission immediately after an episode of AHRF. Indeed, the reported re-admission rate is 79.9% with a 1-year mortality rate of 49.1% [
73]. Two recent trials have focused on this high-risk group [
74,
75]. In the trial by Cheung
et al., patients who had required NIV for AHRF were randomized to domiciliary nocturnal NIV or continuous positive airway pressure (CPAP) of 5 cm H
2O. The intervention was shown to have a lower rate of recurrent AHRF compared with the control group (38.5% versus 60.2%;
P = 0.039, respectively) and a longer median time to first re-admission (71 days versus 56 days;
P = 0.048, respectively) [
74]. CPAP as an appropriate control arm in patients with COPD is interesting. The methodological aim was to balance the possible negative physiological effects of CPAP in patients with severe COPD against the concerns about using a control group that were not exposed to a mask interface. The use of interface with minimal pressure delivery allowed testing of the hypothesis that NIV is beneficial in COPD patients with post-acute hypercapnic respiratory failure. In a separate trial, Funk
et al. enrolled patients who had required NIV for AHRF, but randomized the patients, after a run-in period on NIV of 6 months, to either continuation or withdrawal of NIV. The primary endpoint was escalation of ventilation. They found that the rate of ventilation escalation was lower in the NIV continuation group compared with the withdrawal group (15% versus 77%;
P = 0.0048, respectively) [
75]. These studies suggest a benefit of using domiciliary NIV in patients who are recovering from a recent acute exacerbation complicated by acute hypercapnic respiratory failure.
At present, there is controversy about the use of domiciliary NIV, and there are currently no widely accepted criteria for commencing domiciliary NIV in stable COPD, despite the practice being widespread [
76]. The available data indicate that patient selection is important. Specifically, the patients most likely to benefit from long-term domiciliary NIV are those who exhibit symptomatic chronic hypercapnic respiratory failure and those with severe episodes of acute exacerbation requiring acute NIV during hospital admission [
77]. Because the prognosis is better for patients who have hypercapnia that is reversible during the post-exacerbation recovery phase [
78,
79], it is important to target long-term NIV to patients who remain hypercapnic following their acute episode, as shown by the studies of Cheung
et al. and Funk
et al.[
74,
75]. Preliminary screening data in 25 patients from a UK RCT of post-exacerbation domiciliary NIV suggest a prevalence of persistent severe hypercapnia (arterial partial pressure of carbon dioxide > 7 kPa 2 weeks after an episode of AHRF) of over 40% [
80]. However, further studies are needed to elucidate the trajectory of hypercapnia in a large cohort of patients with COPD treated with acute NIV.
Two RCTs are ongoing in the UK [
81] and the Netherlands [
82] to establish the effect of domiciliary NIV in reducing mortality and hospital admission for patients with COPD who are hypercapnic. The UK trial is focused on patients following an acute hospital admission requiring NIV, and the trial from the Netherlands is focused on patients with stable COPD who are hypercapnic. There are, however, several challenges in conducting such studies. Firstly, the absence of a true placebo for NIV makes it difficult to have a robust control group for comparison. Most studies to date have compared NIV with UC, with or without LTOT [
70‐
72,
75], but a limitation of this approach is that it does not take into account the placebo effects of being given a mask interface. Cheung
et al. attempted to address this by administering nasal CPAP at 5 cm H
2O to patients in the control group [
74]. However, as the authors acknowledged, the possibility remained that the CPAP had a beneficial physiological effect on the control group, and could not therefore be considered to be a true placebo [
83,
84]. Secondly, the interpretation of the potential benefits of NIV are hampered by relatively short follow-up periods in the trials published to date; only two studies [
71,
72] have followed patients up for 2 years or more. Clearly, as patients established on domiciliary NIV are likely to remain on it for several years, it would seem advantageous for future studies to assess its benefits over the longer term.