The semantics of COPD exacerbations

)Frits Franssen Frits Franssen: the clinician’s perspective

12 July, 2017

COPD exacerbation as a result of pneumonia?
During a recent morning report, a resident presented the case of a 67 year old COPD patient who was referred to the emergency department in the previous evening because of increased shortness of breath, non-productive cough and fever. Symptoms had started in the morning and were progressive throughout the day. On presentation, the patient was dyspneic and tachypneic, oxygen saturation was 86% on room air and temperature was 38.6°C. Chest auscultation revealed slight wheezing and discrete crackling over the right frontal thorax. Chest radiograph showed consolidation in the middle lobe. Arterial blood gas analysis indicated hypoxemia without respiratory acidosis. Finally, peripheral blood leukocytes and high-sensitivity C-reactive protein were elevated.

The resident concluded that the patient had a COPD exacerbation as a result of a pneumonia in the middle lobe. The patient was hospitalised, oxygen supplementation, nebulised bronchodilation therapy, broad-spectrum antibiotics and a course of oral glucocorticosteroids were initiated. Overnight, the patient had been clinically stable and was already less symptomatic. While the case of another patient was presented, I was still reflecting on the established diagnosis and initiated treatment by the resident. Doesn’t a diagnosis of pneumonia rule out a diagnosis of COPD exacerbation? Doesn’t a diagnosis of exacerbation require symptoms lasting for several days? Should a COPD patient with pneumonia routinely be treated with systemic glucocorticosteroids?

What defines an exacerbation?
In the 1987 study by Anthonisen and colleagues on the effects of broad-spectrum antibiotics, exacerbations of COPD were defined and classified based on the presence of symptoms [1]; the occurrence of increased dyspnea, sputum production and sputum purulence was defined as a Type I exacerbation. When only two of these symptoms were present, the event was defined as a Type II exacerbation. Type III exacerbations required the presence of one of the major symptoms, in combination with upper respiratory tract infection within the past five days, fever without other cause, increased wheezing, increased cough or more than 20 percent increase in respiratory rate or heart rate as compared with baseline [1]. While the authors mentioned that the classification of exacerbations was done for analytical purpose only, as treatment was the same for all exacerbations, I’m overlooking the criterium ‘fever without other cause’. I agree that fever without other cause contributes to the diagnosis of exacerbation, but does this also imply that fever with an identified focus, such as pneumonia, excludes the diagnosis of exacerbation? The answer was not provided in the paper.

In the first Global initiative for Chronic Obstructive Lung Disease (GOLD) report for COPD, increased breathlessness was mentioned as the main symptom of an exacerbation, often accompanied by wheezing and chest tightness, increased cough and sputum, change of the color or consistency of sputum, and fever [2]. Also, a number of accompanying nonspecific complaints were described. Importantly, the GOLD scientific committee mentioned that among other criteria, new radiologic anomalies suggestive of pulmonary disease may herald a COPD exacerbation. While details of these abnormalities were not provided, this suggested that consolidation on chest X-ray may indeed prelude an exacerbation. However, in the same section, chest radiography was highlighted as a useful tool in identifying alternative diagnoses that can mimic the symptoms of an exacerbation [2], suggesting that the presence of consolidation would result in a diagnosis of pneumonia instead of exacerbation.

The required duration of increased symptoms for diagnosis of exacerbation was not defined by that time. Seemungal and colleagues incorporated the time factor, by defining exacerbations as increase in symptoms according to criteria modified from Anthonisen and colleagues for at least two consecutive days [3]. In the same year, an operational definition of exacerbation was proposed by a working group of respiratory physicians from Europe and the United States [4]. The definition reached by the groups not only included a time factor but also introduced healthcare use as a defining factor: a sustained worsening of the patient’s condition, from the stable state and beyond normal day-to-day variations, that is acute in onset and necessitates a change in regular medication in a patient with underlying COPD. The group described ‘sustained’ as a worsened condition for at least 24 hours, but also acknowledged that severe deterioration of a patient’s condition over a shorter period of time should not limit the definition [4]. Interestingly, regarding the change in regular medication, the experts stated that this implies that exacerbations do not respond to short-acting bronchodilators.

Large randomised controlled studies also started using healthcare utilisation to define the severity of exacerbations; moderate, if courses of glucocorticosteroids or antibiotics were needed, and severe if the patient required hospital admission. Some studies included a minimal duration of symptoms [5, 6], while others did not [7, 8]. None of them included chest radiography to rule out alternative diagnoses. In the 2007 GOLD strategy document, the required change in medication was not specified, the time factor was not incorporated, and chest X-ray was again suggested to identify alternative diagnoses with comparable symptoms [9]. The heterogeneity of exacerbations was increasingly recognised, and the identification of distinct biological exacerbation clusters [10]. Despite these advances, recent large international COPD cohort studies defined and classified exacerbations based on need for glucocorticosteroids/antibiotics or hospital admission only [11-13].

Wedzicha took the definition of exacerbation forward by emphasising the clinical importance of mild exacerbations, defined as those events requiring increases in regular inhaled medication [14]. Thus, self-managed events not leading to treatment with systemic glucocorticoids or antibiotics are increasingly considered as an outcome of clinical trials [15]. Subsequently, the 2017 GOLD report classifies exacerbations as mild (treated with short acting bronchodilators only), moderate (treated with short acting bronchodilators plus antibiotics and/or oral glucocorticosteroids) or severe (requiring hospitalization or emergency room visit)[16]. Whether a history of two mild exacerbations in the past year justifies stratification of a COPD patient in GOLD groups C or D compared to a patient with two moderate exacerbations or at least one hospitalisation is not addressed in the report. Also, there is no mention of the minimal duration of symptoms required for the diagnosis of exacerbation. Importantly, GOLD 2017 emphasises that exacerbations must be differentiated from cardiopulmonary comorbidities, including pneumonia. This suggests that exacerbation and pneumonia are indeed mutually exclusive diagnoses.

Following the lack of a standardised definition, several studies in recent years suggested a differentiation between pneumonic and non-pneumonic exacerbations. An analysis of the Danish healthcare service registry reported increased length-of-stay, intensive care unit admissions and short-term mortality in exacerbating COPD patients with a primary or secondary diag¬nosis of pneumonia [17]. Worse outcomes of pneumonic exacerbations were confirmed in another study in which the presence of consolidation on chest radiograph was associated with increased in-hospital mortality [18], although this was not confirmed by others [19]. Without denying the clinical relevance of pneumonia in COPD, discriminating between exacerbations with and without evidence of pneumonia, in my opinion, only complicates terminology and increases confusion. Also, this would require differentiation between exacerbating and non-exacerbating pneumonias in COPD, as not all pneumonias in COPD are accompanied by increased bronchoconstriction/wheezing or need for systemic glucocorticosteroids.

Pneumonia with underlying COPD
Following the Anthonisen definition [1], the patient presented by the resident meets the criteria for a Type III exacerbation, based on increased dyspnea and cough and elevated breathing frequency. However, chest radiograph revealed an alternative diagnosis for fever and respiratory symptoms and the duration of symptoms did not fulfill the criteria suggested by some of the publications highlighted above [3, 5, 6].

My personal opinion is that this patient should be labelled with a primary diagnosis of pneumonia, instead of COPD exacerbation as a result of pneumonia. Although readers may consider this labeling a matter of semantics, I hope that my historic perspective on the evolving components and criteria for COPD exacerbation indicates that there still is no general agreement on its definition and classification. Since exacerbations play a central role in the refined ABCD assessment and treatment algorithm of COPD [16], a standardised and stable definition is essential. Also, well-defined distinction between exacerbation and pneumonia is required, as pneumonia is considered an adverse event of the treatment with inhaled glucocorticosteroids which are prescribed to prevent exacerbations [20]. Moreover, this will drive back unnecessary treatment of all flare-ups of respiratory symptoms in COPD patients with systemic corticosteroids.

Finally, a more detailed and standardised definition of exacerbation, including the criterion of exclusion of alternative diagnoses (such as pneumonia, acute cardiovascular events and pulmonary embolism) will decrease the clinical heterogeneity of exacerbations, allow studies to increase our understanding of these COPD-related events, enable the development biomarkers for early-detection and prognostication and stimulate innovative interventions. I welcome your opinion on this important topic.