The Saga of COVID-19 in Chronic Lung Disease: History Repeats Itself

SARS-CoV-2 led to a pandemic similar to the influenza pandemic of 1918, with over 700 million COVID-19 cases and 7 million deaths recorded. Respiratory involvement has resulted in numerous hospitalizations and deaths (1). Elderly men with chronic comorbid conditions such as heart, lung, and renal diseases have faced the worst outcomes. Chronic respiratory conditions are considered a major vulnerability in SARS-CoV-2 infection. As the pandemic expanded, studies reported on this detrimental relationship between the acute viral infection and various chronic lung diseases (CLDs). Most studies relied either on limited sample sizes or, in the case of large samples, on administrative data or meta-analyses that lacked detailed data (2–4).

Prior epidemiological studies have shown that respiratory viral infections such as SARS and Middle East respiratory syndrome led to a higher burden in patients with CLD (5). The COVID-19 pandemic provides a unique opportunity to explore many questions about this relationship that were not possible with previous pandemics. We are now better equipped to observe, measure, and analyze; we also collaborate and communicate findings more effectively.

In this issue of AnnalsATS, Crothers and colleagues (pp. 1034–1043) take us one step closer to understanding by clarifying the health outcomes of SARS-CoV-2 infection in patients with CLDs such as asthma, chronic obstructive pulmonary disease (COPD), interstitial lung disease (ILD), and sarcoidosis (6). The authors employed a rigorously validated methodology using International Classification of Diseases, 10th Revision, codes using data from the Veterans Health Administration (VHA) COVID-19 Shared Data Resource database. The VHA established this resource in April 2020 to compile pertinent COVID-19 information swiftly and consistently for VHA research and operational communities. It includes details on test timing and results, interventions, patient outcomes, and preexisting conditions and medications. The database also records molecular variants of SARS-CoV-2 and their impact on chronic lung diseases.

The authors, using a validated algorithm, uniquely grouped study outcomes as mild, moderate, critical, and fatal. The authors determined the risk of hospitalization, indicators of critical illness, and 30-day mortality and further characterized and compared certain clinical and laboratory abnormalities in hospitalized COVID-19 patients with and without CLD.

The basic demographics of the study population were similar to those of prior studies, except for an overwhelming majority being male. As expected, COPD and asthma were the most prevalent CLDs in this large cohort, and slightly less than one-third of them had severe CLD. The veterans with SARS-CoV-2 and CLD (except for asthma) showed higher hospitalization and worse physiologic and laboratory measures and required higher degrees of support than those with no CLD.

The patients with CLD did worse in severity outcomes than those with no CLD. The study, through its larger sample size, allowed a more detailed investigation of subtypes of CLD. It further showed worse outcomes for idiopathic pulmonary fibrosis (IPF) and no increased risk for asthma, with the rest of CLDs in between. Interestingly, the risk difference for fatal outcomes in CLD and each subtype was smaller than other severity outcomes, with asthma and sarcoidosis showing no increased risk of death. Younger veterans (aged <60 yr) and those who received immunosuppressive medications before the infection did worse.

Prior studies addressed the issue of CLD in SARS-CoV-2 infection outcomes. The studies clearly showed a higher rate of ICU admission and mortality in those with ILD and COPD (2, 3). The present study, however, is unique in several aspects. The shear sample size provides an excellent opportunity to answer questions. Furthermore, because there is no restriction from race/ethnicity and socioeconomic standpoints to access the services, it provides more inclusive and more comprehensive data on related groups (7).

Pathophysiologic studies of severe SARS-CoV-2 infection point to a type I, neutrophilic hyperinflammatory response with elevated C-reactive protein, neutrophilia, and lymphopenia associated with worse outcomes (8). In contrast, a recent meta-analysis showed that neutropenia and lymphopenia, together with increased C-reactive protein and D-dimer, were associated with worse outcomes (9). Our team as well as others showed that the newer variants, such as SARS-CoV-2 B.1.1.529 (Omicron variant), produce a lower inflammatory response and thus less severe clinical outcomes (10). The present study identified a hypoinflammatory profile in the CLD group. Thus, the worse outcome in CLD may be a result of combined effects of an exhausted immune response and a lower threshold for admission. Worse outcome in patients with CLD with prior use of immunosuppression points to the role of inflammation. Along the inflammatory pathways in reaction to SARS-CoV-2, lack of worse fatal outcomes in individuals with asthma in this study as well as others may also point to the protective effects of type 2 inflammation against this virus-related acute neutrophilic dominant inflammation (11).

The landscape of having four severity categories of outcomes has several implications and may help to decipher the findings in this paper. First of all, and as pointed by the authors, in CLD compared with no CLD, the COVID-19 outcome severity risk difference diminished with increasing severity group. This may in part relate to a lower admission threshold for those with CLD. Interestingly, the widest risk difference was for mortality in IPF (16.9%) compared with no CLD (4.6%).

There are some methodological aspects that could improve trust in the findings. Laboratory reports were constructed to reflect the most severe outcomes, potentially skewing the data toward more critical conditions rather than stable states. It may be beneficial for analyses to use median values to better represent the general condition of the patients. Age stratification should be considered in future work; it is recommended to stratify into more detailed age groups beyond the current categories of <65 and ⩾65 years. Research indicates that each additional decade of age among veterans with COVID-19 significantly increases the mortality rate (12, 13). In terms of exposure groups, the authors’ sensitivity analysis could be enhanced by employing inverse propensity score weighting or propensity score matching (14, 15). The current sample size of the non-CLD group is approximately five times larger than that of the CLD group. Using these methods could help mitigate potential sample bias in the data that may affect the outcomes (14, 15).

How can the data on interaction between SARS-CoV-2 infection and CLD help our patients? First and most important will be resource allocation in preventative measures for patients with certain CLDs (e.g., IPF, other ILDs, COPD), including vaccination and exposure precautions. Earlier detection with SARS-CoV-2 testing, together with other respiratory viral infections, and early pharmacotherapy added to a lower threshold for hospitalization may be considered to improve the outcomes in these patients.

There are some unanswered questions that this current database may help to answer. A comparison between those who had the worst outcomes (severity levels “critical” and “fatal”) within CLD as a whole (IPF, other ILD, and COPD) and in each group of CLD may provide information on the most vulnerable for earlier admission and intervention. Using machine learning on laboratory profiles of the infected patients in various severity categories may produce a better predictive model for more severe outcomes (16). Furthermore, the differential effects of SARS-CoV-2 variants on CLD will help us to understand the underlying mechanisms that increase the risk of worse outcomes (17). Last but not the least, the effect of wider vaccination in the latter stages of the pandemic may have changed the risk differences and need further reevaluation.