Skeletal muscle weakness is a common and serious finding in patients with advanced chronic obstructive pulmonary disease (COPD) and contributes to their morbidity and mortality, increasing the risk of exacerbations, hospitalisations and death by 3–4-fold.1 2 The treatment for muscle dysfunction of COPD is extremely limited, and the multitude of interventions to address poor muscle performance has not been fully explored in these patients. Over the past decade, there has been an explosion of interest and research on this topic. Despite that, the pathophysiological mechanisms linking the lung disease of COPD with skeletal muscle dysfunction remains largely unknown. Identifying a link between lung disease and muscle performance might indeed be a daunting task because of the influence of other comorbid conditions (and medications), previous musculoskeletal injury and the history of physical activity that might influence the current status of skeletal muscle in this condition. There are, however, several observations that are widely known and accepted. Firstly, skeletal muscle weakness increases with progression of disease. Secondly, histologically, biopsies of large limb muscles consistently demonstrate a reduction in muscle mass, especially of the anaerobic type-IIx fibres, a shift of fibre type from type 1 fibres to a predominance of type 2 fibres and the depletion of mitochondrial oxidative enzymes leading to uncoupling of oxidative phosphorylation and reduced aerobic capacity.3 Interestingly, these muscles also demonstrate increased oxidative stress and accelerated apoptosis and several recent reports have linked systemic inflammation with the physiological and histological changes observed in the skeletal muscles of COPD patients, raising the possibility that inflammation may be the principal “driver” of the muscle dysfunction in COPD.4 5 The latter concept has been fuelled by animal experiments, which showed that lung inflammation via systemic inflammation can induce local inflammatory responses in skeletal muscles, leading to muscle atrophy and inhibition of muscle regeneration.6

The study by Barreiro and colleagues7 in this issue of Thorax reports some provocative data that challenges the “inflammatory theory” of skeletal muscle dysfunction in COPD (see page 10.1136/thx.2007.078030).7 In this carefully conducted cross sectional study, the authors found that there were less inflammatory changes in the quadriceps of patients with COPD than in the control subjects. In fact, protein expression in the muscles for tumour necrosis factor (TNF)α in COPD subjects was only about 60% of that in control subjects. TNFα expression correlated positively with muscle function (ie, higher levels were associated with increased quadriceps strength). Other biomarkers of inflammation including vascular endothelial growth factor and TNF receptors showed similar trends. On the other hand, markers of oxidative stress were more prominent in COPD muscles. There was increased expression for protein carbonyl products in COPD muscles, which was associated with reduced quadriceps strengths. Collectively, these data raise the provocative hypothesis that local inflammation may not be detrimental (but may even be helpful) in maintaining skeletal muscle integrity and function of COPD patients.

Under the traditional paradigm, proinflammatory molecules such as TNFα are considered to be catabolic, leading to protein and muscle breakdown. However, there is emerging evidence that they may be anabolic in certain cases. For instance, TNFα may play a critical regulatory role in muscle regeneration, possibly by activating the p38 mitogen activated protein kinase pathway, which is essential for myogenesis,8 and exogenous systemic administration of TNFα has been shown to promote skeletal muscle protein synthesis.9 Moreover, following acute injury to peripheral muscles, inflammatory cytokines are essential to muscle recovery and regeneration.10 Inflammatory cytokines such as TNFα can function both as a catabolic and an anabolic promoter, depending on its local concentration. At low concentrations (eg, 1 U/ml or less), TNFα is catabolic but at high concentrations, it becomes anabolic (eg, 100 U/ml or more).11

It is plausible then that in mild to moderate COPD, there may be increased local expression of inflammatory mediators in response to increased oxidative load (which may be helpful in maintaining muscle function) but with progression of disease resulting in progressive atrophy of peripheral muscles from disuse and medications (eg, systemic corticosteroids) and with additional increases in local oxidative stress, the local inflammatory responses may attenuate. This, in turn, may promote muscle breakdown and exacerbate skeletal muscle dysfunction. This concept may explain some of the discrepancies between the findings by Barreiro and colleagues7 and those by Montes de Oca et al,12 who studied a group of patients with COPD with milder disease (mean forced expiratory volume in 1 s, 43% of predicted versus 33% of predicted for Barreiro’s study). Under this paradigm, reducing local or even systemic inflammation is unlikely to be helpful; it may even exacerbate muscle dysfunction.

Obviously, based on this cross sectional study, it is impossible to ascribe causation, and the hypothesis that local inflammation is a beneficial response to increased oxidative stress is still very speculative. Nevertheless, the findings of Barreiro et al suggest that we need to look at inflammation (at least in local tissues) in a new light. This is already happening in congestive heart failure (CHF). In CHF, where skeletal muscle changes are strikingly similar to those observed in COPD,13 several recent randomised controlled trials of anti-TNF therapies in CHF have failed to demonstrate any measurable benefits for the patient. In fact, two studies were terminated prematurely because of futility14 and the other alarmingly showed increased risk of death and hospitalisation for heart failure in patients randomised to receive infliximab.15 Previous studies examining the use of anti-inflammatory agents for soft tissue injury have demonstrated parallel findings. Although non-steroidal anti-inflammatory drugs can reduce a particular marker of inflammation, their overall benefit towards healing of tendon and muscle injury is questionable.16

Skeletal muscle dysfunction is a major contributor of morbidity and mortality in COPD. Its pathogenesis, although multifactorial, is not clearly defined. The findings of Barreiro and colleagues7 challenge the prevailing notion that local inflammation contributes to the decline in muscle function. Like any good research, they have made us think outside the box and to approach this growing problem in a new and refreshing way. Only well designed clinical and animal studies in the future can answer the question: is inflammation good, bad or irrelevant for skeletal muscles in COPD? Or perhaps the more pressing question should be: what aspects of the inflammatory response should be adjusted during the various degenerative and regenerative phases of skeletal muscle as this tissue responds to physical activity, exercise and disuse in COPD? Inflammation is a complicated tangled web of intertwining cascades that are carefully controlled by numerous positive and negative feedback and feed-forward mechanisms. Until we develop a better understanding of the influence of inflammation on muscle injury, regeneration and healing in COPD, clinicians should strongly encourage their COPD patients to stop smoking, exercise regularly and engage in pulmonary rehabilitation, which are proven ways to enhance muscle performance in COPD patients.