313 e-Letters

  • Author response: Eosinophils as covariates

    We thank James R Camp for his response and interest in our study. To answer the question posed directly, we did not use blood eosinophils as a covariate in the model, since leukocyte differential count is not routinely made at every outpatient visit for COPD patients in Denmark.

    The relation between blood eosinophils in COPD and pulmonary infections is not a trivial one. As mentioned by James R Camp, mouse models indicate that eosinophils have antibacterial properties in vitro (1). However, few clinical studies have included blood eosinophil counts as a risk factor of pneumonia in COPD, mostly showing either a weak or no association (2,3).

    Eosinophils from human blood have been demonstrated to have bactericidal effects against S. aureus and E. coli, but noteworthy, this effect was not as potent as the neutrophils (4). Additionally, severe acute bacterial infection like sepsis almost uniformly causes eosinopenia (5,6) and experimental lipopolysaccharide injection in healthy humans and diabetic humans cause profound and long-lasting eosinopenia (7). This is not easily comprehensible if the eosinophils are a needed part of the innate host immune response to bacterial infection.

    An alternative explanation for a possible association could be that eosinophils and neutrophils act in bacterial infection in a complex interplay, while regulating and adjusting the response of each other. To support this, it has been demonstrated that integrin β chain-2 (CD18),...

    Show More
  • Eosinophils as covariates

    We recently read the recent publication by Elköf and colleagues in the recent issue of Thorax titled ‘Use of inhaled corticosteroids and risk of acquiring Pseudomonas aeruginosa in patients with chronic obstructive pulmonary disease’(1) with great interest. The paper highlights an important clinical observation in a well-defined cohort.

    We were interested that Elköf and colleagues, tentatively discuss that biological mechanisms resulting from ICS alterations on the immune system may be an explanation for a change in the microbial composition in the airways(1). As the authors discussed, eosinophilic inflammation in COPD identifies a group of patients with ICS responsiveness(2). In the mouse model, there are data examining that eosinophils have anti-microbial properties(3). Access to eosinophil counts from this cohort may be invaluable in unravelling the relationship of eosinophils and COPD and could provide insight into the impact of steroids in bacterial infection. Did the authors investigate the peripheral blood eosinophil count as a covariate in their main analyses?


    1. Eklöf J, Ingebrigtsen TS, Sørensen R, Saeed MI, Alispahic IA, Sivapalan P, et al. Use of inhaled corticosteroids and risk of acquiring <em>Pseudomonas aeruginosa</em> in patients with chronic obstructive pulmonary disease. Thorax. 2021:thoraxjnl-2021-217160.
    2. Bafadhel M, Peterson S, De Blas MA, Calverley PM, Rennard SI, Richter K, et al....

    Show More
  • Response to Hursoy and colleagues

    To the editor,

    We thank N. Hürsoy and colleagues for their interest in our study of patients four months after severe COVID-19 [1]. We agree that there needs to be continued development of terms describing the radiographic appearance of post-COVID fibrotic-like patterns. We acknowledge that without the benefit of histopathology or serial imaging, our ability to define pulmonary fibrosis is limited.

    The authors posit that parenchymal bands, irregular densities, and ground glass opacities, may be considered fibrotic-like patterns. We have included irregular densities, characterized as reticulations or traction bronchiectasis, as fibrotic-like changes. We did not include parenchymal bands [2], as these can be associated with atelectasis, which is common in COVID and can disappear over time [3]. Similarly, we did not include isolated ground glass opacities as fibrotic-like changes, as these have been found to decrease over time in CT lung cancer screening cohorts [4] and in other post COVID-19 cohorts [5, 6].

    A priori, we evaluated for both previously established interstitial lung abnormality categories [7], as well as categories of radiographic abnormalities reported in Acute Respiratory Distress Syndrome (ARDS) survivors using an established scoring system [8]. This inclusive approach should facilitate meta-analyses and comparisons with future studies of COVID-19 survivors, interstitial lung disease studies, and studies of non-COVID ARDS survivors. Fu...

    Show More
  • Different Views About Post-Covid Fibrotic-Like Patterns

    Dear Editor,

    We have read with great interest the article investigating the relationship between computed tomography (CT) findings of the patients with fibrotic-like patterns and telomere length after four months of acute COVID-19 infection. According to the literature and our experience, post-COVID interstitial lung disease is a potential public health problem. Thus, we aimed to share our concerns about the fibrotic-like patterns in this group of patients.

    Post-COVID fibrosis is not as the same as the other interstitial lung diseases. In the article, the authors describe CT findings of fibrotic-like patterns as limited to reticulation, honeycomb cysts, and traction bronchiectasis. However, post-COVID fibrosis CT findings were shown to be more varied and may include parenchymal bands, irregular densities, and ground-glass areas (1–3). As we move towards the future, all of us need to create a common language, a lingua franca in the definition of post-COVID fibrosis. To achieve this, we need brainstorming and close cooperation.

    It will also be helpful to elaborate the characteristics of the non-fibrotic pattern in the table. The clinical importance of the ground glass areas, which persist four months after active infection but not defined as fibrotic, is unknown. We consider that these patterns cannot be separated from fibrotic-like patterns precisely. Additionally, we can also classify parenchymal bands as fibrosis-like appearance. In our experience...

    Show More
  • Reply to: Referral to pulmonary rehabilitation (PR) by a current PR practitioner has no effect on PR completion rate

    We thank Dr Abdulqawi for interest in our work (1). He comments that the referral, uptake and completion rates for pulmonary rehabilitation in the current study were lower than in a previous study by Jones and colleagues (2). We would caution against retrospective comparison with unmatched historical controls due to confounding factors such as differences in patient characteristics and practice pathways that may contribute to inaccurate point estimates.

    We hypothesised that the COPD discharge bundle would impact on referral rates. Strengths of the current work include the prospective real-world nature of the study, with the research team having no involvement in treatment allocation. The clinical team delivering the bundle were blinded to the study objectives, thus minimising any Hawthorne effect.

    Dr Abdulqawi raises the point that pulmonary rehabilitation completion rates were low in the current study (albeit based on a low denominator). The reasons for non-completion of PR are often complex and multi-factorial (3) and may not be directly related to referral source. However, what is clear is that without a referral for pulmonary rehabilitation, uptake and completion rates are zero.

    1. Barker RE BL, Maddocks M, Nolan CM, Patel S, Walsh JA, Polgar O, Wenneberg J, Kon SSC, Wedzicha JA, Man WDC, Farquhar M. Integrating Home-Based Exercise Training with a Hospital at Home Service for Patients Hospitalised with Acute Exacerbations of COPD: Developing the M...

    Show More
  • Dog bites man - COVID-19 and smoking.

    The idea that smoking might have a protective effect against COVID-19 is an intriguing, man bites dog type of story, which gives it a certain attraction. Happily, it appears to be false and the assumption of harm has turned out to be correct[1-5].
    Our data show clearly that in the 2.4 million Zoe COVID Symptom Study App users, people who smoked were at increased risk of symptomatic COVID-19[2] and were at risk of more severe disease, which is consistent with a systematic review of patients hospitalized with COVID-19[4]. Our findings are also consistent with The UCL COVID-19 Social Study3 which found increased risk of test confirmed COVID-19 (OR=2.14 (1.49–3.08)) and with the COVIDENCE study where smokers had an OR of1.42 (0.99-2.05) for test-confirmed COVID-19[1].
    The OpenSafely dataset based on data from the primary care records of 17.3 million adults in the UK found that, adjusted for age and sex, also identifies smoking as a risk factor - current smoking was associated with a hazard ratio for COVID-19-related death of 1.14 (1.05–1.23)5. The apparently protective effect in the “fully adjusted” model is due to over-correction producing collider bias.
    Since any protective effect of smoking in COVID-19 appears to be illusory, pursuing a mechanism for it is unlikely to be productive.

    1 Holt H, Talaei M, Greenig M, et al. Risk factors for developing COVID-19: a population-based longitudinal study (COVIDENCE UK). medRxiv 2021:2021.2003...

    Show More
  • Referral to pulmonary rehabilitation (PR) by a current PR practitioner has no effect on PR completion rate

    We have read the paper by Barker et al. (1) with interest. We congratulate the authors for conducting and publishing their prospective cohort study evaluating the effect of COPD discharge bundle on pulmonary rehabilitation (PR) referral and uptake following hospitalisation for acute exacerbation of COPD (AECOPD).

    The authors have shown that the COPD discharge bundle had a positive effect on PR referral compared with a no bundle (17.5% (40 of 228) referral rate vs 0%(0 of 63)). This figure is lower than the expected 30% referral rate to PR following AECOPD (2). However, the paper offers no potential reasons for the lower referral rate.

    The study had two bundle groups:
    • COPD discharge bundle delivered by a current PR practitioner
    • COPD discharge bundle delivered by a practitioner with no involvement in PR

    Compared to delivery by a practitioner with no PR involvement, completion of the bundle delivery by a current PR practitioner resulted in higher referral and pick-up rates (60% vs 12% and 40% vs 32%, respectively). These results support the concept of integrating PR and hospital services.

    Unfortunately, the completion rate (number of subjects who completed PR divided by the number of referrals) was disappointingly low. Also, there was no difference between the two bundle groups (13% (2 of 15) vs 12% (3 of 25)), as stated in the supplementary data.

    It seems that patients' willingness or ability to complete PR is not af...

    Show More
  • Current smoking as a risk factor for COVID-19

    There is no question that the harms of smoking hugely outweigh any potential health benefits. Many people, ourselves included, assumed at the beginning of the pandemic that greater susceptibility to COVID-19 would be another harm of tobacco smoking to be added to the long list. Surprisingly, most of the epidemiological data published over the last year do not support this claim. Indeed whereas ex-smokers are consistently found to be at increased risk of both SARS-CoV-2 infection and severe COVID-19, current smokers are consistently at lower risk than ex-smokers and in many studies they appear to be at a lower risk than never smokers. The lower infection rate in smokers compared to non-smokers and ex-smokers has been found across 62 studies (1, 2), including now a full cohort with a dose-response pattern (3).

    The authors’ response does not counter the observation that among nearly 27,000 individuals who had a SARS-CoV-2 test in their study, smoking prevalence was lower in those who tested positive than in those who tested negative.

    In the OpenSAFELY study (4) too, the direction of the association between smoking and death from COVID-19 depends critically on what adjustments are made. The primary analysis appears to be based on a fully adjusted Cox regression model in which the hazard ratio for current smokers relative to never smokers was 0.89 (95% CI 0.82-0.97). The value (1.14; 1.05-1.23) cited by Hopkinson and colleagues is after adjusting for age and sex...

    Show More
  • Response to correspondence (Jackson et al. Thorax 2021)

    The influence of obesity on both asthma and T2 biomarkers remains poorly understood and we fully agree this requires further investigation, as does the relationship between obesity, depression and persistent symptoms of breathlessness. However, the data correlating obesity and FeNO is conflicting and the reported weak positive associations have often not been adjusted for corticosteroid dose and may simply reflect higher doses of corticosteroid therapy in more breathless obese patients than by those of normal weight, rather than a specific mechanistic relationship.
    Moreover, the UKSAR population appears very different from the cohorts described in some of these reports. For example, the average FeNO was only 25ppb in the Komakula study, whilst in the study by Lugogo subjects were predominantly T2-low across all BMI categories: the upper quartile value of blood eosinophils in both lean and obese groups was <300 cells/µL, whilst the upper quartile of FeNO in both lean and obese groups was <30ppb. In contrast, even in the UKSAR T2 high cohort, the mean BMI was in the obese range.
    The nature and veracity of the ‘T2-low’ phenotype remains unclear, particularly in severe asthma. What is increasingly apparent is that patients are frequently prescribed high dose inhaled and systemic corticosteroids for respiratory symptoms, which suppresses T2 inflammation in the process. In the context of obesity and other co-morbidities known to be associated with increased re...

    Show More
  • Erector spinae muscle area is not associated with mortality in the COPDGene cohort

    We thank Tanimura and colleagues for their thoughtful commentary on our recent manuscript, “Respiratory exacerbations are associated with muscle loss in current and former smokers” and read their analysis of erector spinae muscle area (ESMA) with interest (1). In their commentary, they note that muscle loss can occur heterogeneously, with the greatest expected impact on the muscles of ambulation. They suggest that erector spinae muscles, due to their fiber composition and anti-gravity role, are a better reflection of inactivity-related muscle loss and posit that changes in pectoralis muscle area (PMA) may only reflect changes in nutrition (as measured by body mass index, BMI).

    We agree that muscle loss is unlikely to be uniform; however, a disconnect has been reported between the postural muscles of the trunk and ambulatory muscle (e.g. quadriceps) weakness, despite similar fiber types (2). Few studies measure both groups of muscles simultaneously, but there is evidence that inspiratory force is more affected than peripheral muscle force in patients with COPD; implying that deconditioning is not the sole driver of muscle dysfunction (3). While the pectoralis muscle potentially underestimates inactivity-related atrophy, these studies suggest its role as an accessory muscle of inspiration makes it a reasonable target for capturing any underlying systemic process.

    In contrast to Tanimura et al’s findings, in the COPDGene participants (n=8,603) BMI was more stro...

    Show More