Statistics from Altmetric.com
More than 40 years ago Benjamin Burrows and his colleagues1 described the distinctive clinical, functional, radiological and pathological characteristics of the chronic obstructive pulmonary disease (COPD) phenotypes that they called emphysematous and bronchial types of chronic airways obstruction. They identified a subgroup of patients who were “thin” and had evidence of emphysema on chest x ray, while another subgroup was found to be of “stocky build” and had chest x ray changes suggestive of previous pulmonary inflammatory disease. Postmortem anatomical emphysema severity was positively related to emphysema grade on chest x ray, and with total lung capacity and age. Anatomical emphysema was, on the other hand, inversely related to chronic inflammatory changes on chest x ray, and with sputum volume, carbon dioxide tension and diffusing capacity.1 All patients had severe irreversible airways obstruction and died of respiratory failure but, noticeably, reduction in forced expiratory volume in 1 s (FEV1) was not significantly related to anatomical emphysema extent and severity. Burrows et al introduced the terms type A and type B to empirically differentiate patients with the emphysematous type from those with the bronchial type of chronic airways obstruction.1 This phenotypic differentiation of patients with COPD, without being not even seriously questioned or challenged, has never been widely accepted and in more recent times it has been almost totally neglected by the pulmonary scientific community.
In this issue of Thorax, investigators from the Universities of Kyoto, Shiga, and Vancouver,2 continuing a long history of research activity in the field of the in vivo anatomical study of COPD by high resolution computed tomography (HRCT), elegantly bring back to life and strengthen Burrows’ findings of 1966 (see page 20).2 Ogawa and colleagues2 found, indeed, that patients with COPD who, at HRCT, had reduced x ray attenuation values, compatible with emphysematous destruction of lung parenchyma, were thinner than those who had bronchial wall thickening compatible with chronic inflammatory changes of the conductive airways. A significant inverse relationship was found between body mass index (BMI) and emphysema extent at HRCT, whereas no correlation was present between BMI and thickness of the bronchial wall. FEV1 was inversely correlated with BMI, HRCT emphysema extent and bronchial wall thickness, but there was no significant difference in mean FEV1 between the two HRCT phenotypes of COPD.2
These findings point to the fact that patients with COPD may have different systemic clinical manifestations, reflecting different pathophysiological mechanisms of expiratory airflow limitation. Several other papers3–6 have shown a reduction in BMI in patients with HRCT findings of predominant emphysema but the study of Ogawa and colleagues2 is the only one in which HRCT quantitative parameters related to both destruction of lung parenchyma7 and airways remodelling8 have been measured in a large series of patients with COPD.
Spirometric detection of not fully reversible airflow limitation by FEV1 is the integral result of a spectrum of different underlying pathological conditions that are unified under the acronym COPD. The use of this term has permitted the pulmonary community to speak a common language and to increase the awareness in the general public and the health system authorities of one of the present and future world leading causes of chronic morbidity and mortality. Nonetheless, diagnosing, categorising and measuring disease progression by spirometric parameters such as FEV1 and forced vital capacity (FVC), as suggested by recent guidelines, cannot provide a panoramic view of the complexity of COPD.9–11 Like many other common diseases, COPD is a heterogeneous disorder.12–16 However, no other specialists would diagnose and classify specific diseases by rudimentary standards as pulmonologists do with COPD.17 As highlighted by Burge18 commenting on the results of clinical trials with inhaled corticosteroids over the past decade, it is likely that the different pathological changes underlying COPD may respond differently to inhaled steroids. The results of these studies could have been different if patients had been subdivided into subgroups according to their clinical presentation, instead of being enrolled by the level of airflow limitation alone.18 The same explanation may hold true for the deceiving results of more recent pharmacological trials.19 20
There is growing evidence that HRCT could provide in vivo information about the various pathological changes occurring in patients with COPD and may permit differentiation of those with predominant airway obstruction from those with predominant emphysematous destruction. The same group of investigators who report in this issue of Thorax have previously shown that pulmonary function abnormalities are more accurately predicted by multivariate regression with HRCT measurements of both extent of low attenuation areas and airway wall thickening than by univariate regression with low attenuation areas.21 Furthermore, they have also shown that thickening of large airways, assessed from the percentage wall area of the right apical segmental bronchus, as in the present paper, is related to thickening of the airways with an internal diameter smaller than 2 mm, as assessed by histological evaluation.8 In this line of evidence it can be said that wall thickness of the large airways at HRCT can predict the anatomical status of small airways, the major site of airway obstruction in COPD.22 Together with the extent of low attenuation areas, wall thickness of the large airways on HRCT can then be used to help classify patients with airflow limitation as having either a predominant phenotype of increased airway resistance or a predominant phenotype of increased lung compliance. Orlandi and colleagues23 measured both airway wall thickness and lung parenchyma x ray attenuation by HRCT and showed that patients with chronic productive cough had significantly thicker bronchial walls than patients without chronic productive cough who, on the other hand, had a significant increase in the percentage lung area with reduced x ray attenuation and a lower lung diffusing capacity. Accordingly, O’Donnell and colleagues24 found that sputum neutrophil counts in patients with COPD was closely related to airway dysfunction, but not to the severity of emphysema, as assessed by HRCT and lung diffusing capacity.
The paper by Ogawa and colleagues2 offers a demonstration of how the overall picture of COPD could be complex and shows to what extent the unifying spirometric assessment of expiratory airflow limitation could be of limited value in order to identify different pathophysiological mechanisms among individuals. Using HRCT as a criterion standard to identify the two more relevant phenotypes of COPD and extending the observation, apart from BMI, towards several other clinical, functional and instrumental variables (ie, symptoms, physical signs, functional evaluation at rest and during exercise, quality of life, frequency of exacerbations, chest x ray findings, biological markers in expired air, sputum and blood) and, eventually, to genetic differences, could facilitate recognition of patients with COPD whose responsiveness to a specific therapeutic approach may be different. The development of a standardised method for classifying COPD phenotypes in clinical practice may have a great impact in understanding the results of pharmacological trials, on the clinical approach to patient treatment and our knowledge of the natural history of the disease. Such a methodology is strongly required.25 Attempts to define the COPD clinical phenotypes have been published in the past year by several groups of investigators.4–6 26 More relevant data are expected from ongoing prospective large scale clinical trials, such as the ECLIPSE11 and the NIH funded COPDGene Project.
It is not essential to solve the dilemma of whether the term COPD expresses correctly the protean clinical presentations of the same disease or that it combines different clinical entities, but it is necessary to realise that there are needs and research opportunities in an area that is largely unknown. The words “expiratory airflow limitation” expresses our present inaccuracy in differentiating increased airway resistance from increased lung compliance.27 HRCT studies have shown that at least two radiological patterns exist in which either airway obstruction or emphysematous destruction predominate. Ogawa and colleagues2 have convincingly demonstrated that, regardless of expiratory airflow limitation, the different pathological changes seen in vivo by HRCT are brought by people with different body habits. Let us jump over the hindering barrier of airflow limitation and explore the COPD world beyond!
Competing interests: None.