Role of IPF genetic risk loci in post-COVID-19 lung abnormalities: a cohort study
•,,,,,,,,,.
...
Abstract
Introduction Persistent lung abnormalities following COVID-19 infection are common. Similar parenchymal changes are observed in idiopathic pulmonary fibrosis (IPF). We investigated whether common genetic risk factors in IPF are associated with developing lung parenchymal abnormalities following severe COVID-19 disease.
Methods Consecutive adults hospitalised for laboratory-confirmed COVID-19 infection were prospectively recruited from March to May 2020. Three single-nucleotide polymorphisms (SNPs) conferring risk for IPF were genotyped (MUC5B rs35705950, ATP11A rs1278769 and DPP9 rs12610495). High-resolution CT and pulmonary function tests were performed at 3 months postdischarge from hospital. Ground glass opacities and reticulation on imaging were visually quantified by two expert thoracic radiologists. Linear regression was used to evaluate the association between risk alleles at each of the three SNPs and (a) lung parenchymal abnormalities as well as (b) pulmonary function, adjusted for age, sex, smoking history and days spent on supplemental oxygen during acute illness.
Results 71 patients were included. Mean age was 63±16 years, 62% were male, 31% were ever-smokers and median hospital length of stay was 9±11 days, with 23% requiring mechanical ventilation. The MUC5B risk allele was associated with a significant decrease in ground glass (β=−0.8, 95% CI −1.5 to –0.1, p=0.02) at 3 months, and this finding was paralleled by a concurrent but non-significant trend towards increased diffusion capacity for carbon monoxide (DLCO) (β=8.8, 95% CI −1.2 to 18.8, p=0.08) compared with patients without this risk allele. None of the risk alleles were significantly associated with reticulation at 3 months.
Conclusion In an adjusted analysis controlling for severity of infection, MUC5B was associated with reduced ground glass and a trend towards concordant higher DLCO at 3 months after severe COVID-19 illness. This hypothesis-generating result suggests a possible protective effect of MUC5B in postinfectious lung abnormalities as compared with fibrosis in IPF, highlighting a plausible trade-off between its role in immune defence and epithelial cell function.
What is already known on this topic
Previous genome-wide association studies suggest a correlation between idiopathic pulmonary fibrosis (IPF) susceptibility loci and severity of COVID-19 infection, with the common risk polymorphism MUC5B conferring protection from COVID-19-related hospitalisation.
What this study adds
This study represents the first investigation of the relationship between genetic risk factors for IPF and radiologic abnormalities on imaging following hospitalisation for COVID-19. While a risk factor for IPF, we suggest that MUC5B may protect from development of post-COVID-19 lung abnormalities. This hypothesis-generating finding may indicate a trade-off between immune defence and epithelial cell function, where increased mucin protects from severity of infection and postinfectious lung changes while increasing the lifetime risk of recurrent, aberrant alveolar repair thought to drive IPF.
How this study might affect research, practice or policy
This study prompts further research into the role of IPF genetic risk loci in postinfectious and postinflammatory lung abnormalities.
Introduction
Approximately half of patients hospitalised for COVID-19 demonstrate persistent lung parenchymal abnormalities in the year following their illness.1 Similar inflammatory and fibrotic changes are also observed in interstitial lung diseases (ILDs) such as idiopathic pulmonary fibrosis (IPF), a disorder characterised by progressive lung scarring and early mortality. While our understanding of the aetiology of IPF remains incomplete, it likely arises from a complex interplay of genetic, environmental and ageing-related risk factors that ultimately give rise to fibroblast proliferation and deposition of extracellular matrix.2
Large genome-wide association studies have uncovered 20 single-nucleotide polymorphisms (SNPs) that are correlated with susceptibility to IPF including the common variant MUC5B,3 and data gathered during the COVID-19 pandemic have shown an association between some of these genetic loci and severity of COVID-19 disease.4 5 Subsequent work has demonstrated the genetic overlap between IPF and COVID-19 severity at MUC5B, ATP11A and DPP9 are due to the same causal variants, although MUC5B and ATP11A may portend opposite directions of effect in the two conditions, increasing risk for IPF while acting in a protective manner in COVID-19.6 7 SNPs at these loci can each be linked to host defence and are involved in mucin production (MUC5B) as well as regulation of the inflammatory response (ATP11A, DPP9).8 9
While a link between the genetics of IPF and COVID-19 severity is established, no studies have examined how these loci influence lung changes on imaging following COVID-19 infection. We tested whether genetic risk factors for IPF are associated with persistent lung parenchymal abnormalities following severe COVID-19.
Methods
Study overview
Consecutive adults aged over 18 years old and hospitalised for laboratory-confirmed COVID-19 infection from March to May 2020 were prospectively recruited from two academic hospitals in Vancouver.10 Patients with ILD predating COVID-19 infection were excluded. Consenting patients underwent standardised follow-up and blood sample collection as well as investigations 3 months after COVID-19 symptom onset, including high-resolution CT (HRCT) and pulmonary function testing (PFTs). Individuals who did not complete these investigations or were deceased prior to 3-month follow-up were excluded.
Patient and public involvement
Patients and the public were not involved in the design, conduct, or reporting of this study. The development of the research question and outcome measures was informed by patient experience of breathlessness and exercise limitation following COVID-19 infection.
Genotyping
Blood samples were collected and DNA was extracted from buffy coat (Qiagen DNeasy Blood and Tissue Kit, Qiagen, Germany). Primers were constructed for each SNP of interest (MUC5B rs35705950, ATP11A rs1278769, DPP9 rs12610495) and Sanger sequencing was used for genotyping (Sequencing and Bioinformatics Consortium, University of British Columbia).11 These SNPs were selected on the basis of their reported OR in association with IPF and ILA in prior genome-wide association studies,3 12 the anticipated relative frequencies of the risk alleles in similar populations, and an early signal of association in the COVID-19 Host Genetics Initiative between these SNPs and COVID-19 severity.13
HRCT assessment of fibrosis
The percentage of the lung with ground glass opacities and reticulation was visually measured in 5% increments for three zones of each lung by two expert thoracic radiologists with extensive experience with semiquantitative scoring of fibrosis.14 The percentage of total lung involved by each abnormality was calculated, and the mean of the two radiologists’ scores was taken for subsequent analyses.
Statistical analysis
The primary outcome was the association between risk SNPs, each dichotomised as no risk allele versus at least one risk allele present, and percentage of ground glass and reticulation present on HRCT at 3 months postdischarge. This was assessed by linear regression for each of ground glass and reticulation while controlling for age, sex, smoking status and days spent on supplemental oxygen during acute illness. In this model, both ground glass and reticulation outcomes required Box-Cox transformation to ensure normal distribution of residuals, as assessed by QQ plot. Linear regression was repeated for forced vital capacity (FVC) and diffusion capacity for carbon monoxide (DLCO) as secondary outcomes.
Results
Cohort characteristics
A total of 119 patients were eligible for recruitment (figure 1), of whom 71 patients were included in subsequent analyses (table 1). The mean age was 63±16 years, 62% were male and 31% were ever-smokers. Patients demonstrated a variety of comorbidities, of which hypertension (49%), dyslipidaemia (39%), and diabetes (28%) were most common. The median hospital length of stay and duration of supplemental oxygen were 9±11 and 5±12 days, respectively, and 23% required mechanical ventilation. Patients were not vaccinated against COVID-19 and did not receive pharmacotherapy for COVID-19.
Study cohort flow diagram. ILD, interstitial lung disease; PFT, pulmonary function testing; SNP, single-nucleotide polymorphism.
Table 1
•
Cohort characteristics
Risk SNPs and post-COVID-19 HRCT abnormalities
Risk allele frequencies for MUC5B (rs35705950, T allele), ATP11A (rs1278769, G allele), and DPP9 (rs12610495, G allele) were 9.2%, 64.1% and 23.9%, respectively. This is similar to large genome-wide association studies showing frequencies of 14.9%, 77.9% and 30.5%.3 15 In adjusted analysis, MUC5B was associated with reduced ground glass at 3 months after symptom onset compared with patients without such a risk allele (β=−0.8, 95% CI −1.5 to –0.1, p=0.02) (table 2A). This association persisted in a sensitivity analysis additionally controlling for comorbidities of hypertension, diabetes, coronary artery disease and chronic kidney disease. None of the risk alleles were associated with reticulation (table 2B).
Table 2
•
Linear regression examining the association between each risk SNP (MUC5B, ATP11A and DPP9) and extent of (A) Box-Cox transformed ground glass and (B) Box-Cox transformed reticulation on HRCT (% of lung involved) as well as (C) FVC and (D) DLCO (% predicted) on PFT at 3 months following discharge from hospital
Risk SNPs and post-COVID-19 pulmonary physiology
Risk alleles were not associated with FVC or DLCO at 3 months postdischarge (table 2C,D). However, paralleling the ground glass finding, the MUC5B risk genotype was related to a statistically non-significant increase in DLCO compared with patients without this risk allele (β=8.8, 95% CI −1.2 to 18.8, p=0.08).
Discussion
This is the first investigation specifically examining the relationship between genetic risk factors for IPF and lung abnormalities on HRCT following hospitalisation for COVID-19. In a well-characterised cohort predating vaccination and specific therapies, we found no significant association between IPF risk alleles and reticulation on imaging 3 months post-COVID-19. However, MUC5B, the strongest common genetic risk factor for IPF, showed an association with reduced ground glass, a finding independently reinforced by a congruent though non-significant trend with higher DLCO. This hypothesis-generating result reinforces previous data suggesting differential effects of risk SNPs in COVID-19 compared with IPF.6 7
Previous genome-wide association studies suggest a correlation between IPF risk loci and severity of COVID-19 infection as defined by critical illness and hospitalisation.6 16 While most IPF risk SNPs, including DPP9, have been associated with greater disease severity,5 16MUC5B and ATP11A are associated with opposite effects.6 7 A study using administrative data similarly confirmed that MUC5B was associated with fewer COVID-19-related hospitalisations and post-COVID-19 pneumonias.7 While highlighting the complex biology that influences COVID-19 severity, these large population studies do not examine the role of genetics in post-COVID-19 interstitial change, an outcome that is difficult to study since disease severity also influences development of fibrosis.17 Our study mitigates this confounding by restricting the population to hospitalised patients and controlling for days spent on supplemental oxygen as a surrogate marker of severity, with results suggesting that MUC5B may not contribute to lung abnormalities in the post-COVID-19 state as it seems to in IPF.
This discrepant role MUC5B may play in post-COVID-19 lung abnormalities is not well understood. The rs35705950 polymorphism in IPF encodes a gain-of-function mutation leading to increased production of a gel-forming mucin.18 This protein is essential in airway clearance and defence, and its deficiency causes increased susceptibility to bacterial infection and inflammation.19 At the same time, its abundant expression is observed in distal airways of patients with IPF,20 suggesting it may interfere with alveolar repair processes ultimately culminating in fibrosis.18 It may be that the MUC5B rs35705950 polymorphism represents a trade-off between immune defence and epithelial cell function, where increased mucin protects from severity of infection and subsequent postinfectious lung changes while simultaneously increasing the lifetime risk of recurrent, aberrant alveolar repair that gives rise to IPF.
We evaluated how IPF risk loci influence post-COVID-19 lung abnormalities at 3 months following severe disease, though it remains unclear if these SNPs influence persistence of abnormalities over longer periods of time, which has important implications for postinfectious lung fibrosis. In a long-term study following SARS-CoV, pulmonary interstitial changes largely improved over time, although some chronic ground glass opacity or band-like consolidation persisted in 38% of patients at 15 years following infection.21 A study of Middle East respiratory syndrome CoV also showed persistent findings in 33% of patients following recovery, although follow-up time was notably shorter ranging from 32 to 230 days after hospital discharge.22 Although ground glass opacities are commonly thought to represent reversible inflammation,23 persistence of such findings and bands years after infection would presumably suggest these represent true structural damage and fibrosis. It is unclear what role IPF risk alleles may play in persistent lung abnormalities and ventilatory defects in the long term following COVID-19 or other viral pneumonias, and future studies are needed to better understand pathogenesis of postinfectious lung changes and identify those at greatest risk.
Our small sample size, reduced power, and hospitalised cohort represent important limitations. Patients who died in hospital were necessarily excluded from the cohort and introduced significant survivorship bias. We lacked data on genetic ancestry, and any effects of tested SNPs may be more evident in subgroups with higher frequencies of the risk alleles in question. While MUC5B was associated with reduced post-COVID-19 ground glass and this plausible finding is supported by a similar trend in PFTs and the broader literature, this should be interpreted as hypothesis-generating and requires confirmation in larger population-based studies. In addition, although individuals with pre-existing ILD were excluded, our patients did not undergo systematic pre-COVID-19 HRCT to confirm that observed lung findings were definitively due to COVID-19, which is relevant given SNPs like MUC5B have been associated with interstitial lung abnormalities.24
Conclusion
Despite these limitations, our study is the first to examine how genetic risk factors for IPF influence post-COVID-19 abnormalities on imaging. The common MUC5B polymorphism rs35705950 was associated with a decrease in ground glass following severe COVID-19, paralleled independently by a non-significant trend towards an increase in DLCO. These results lend strength to a hypothesis that the role of MUC5B in postinfectious fibrosis may be opposite that in IPF, which requires further study.
,Contributors: D-CM and CJR conceived the study. D-CM, AWW, AS, JJ, JL, CC and CJR contributed to patient clinical data acquisition. JY performed genotyping, while CJH and DM analysed and interpreted chest imaging. D-CM conducted statistical analysis, interpreted the data, and drafted the manuscript. All authors reviewed and edited the final draft of the manuscript. D-CM is the guarantor of the study.
Funding: This study was supported by personal donation from Angus Reid. DCM’s salary was supported by the Academic Enhancement Fund (University of British Columbia’s Divison of Respiratory Medicine) and the Michael Smith Research Trainee Award. The funders had no role in study design, data analysis, or preparation of the manuscript.
Competing interests: None declared.
Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review: Not commissioned; externally peer reviewed.
Data availability statement
Data are available on reasonable request. The datasets used and analysed during the current study are available from the corresponding author on reasonable request.
Ethics statements
Patient consent for publication:
Not applicable.
Ethics approval:
This study involves human participants and was approved by University of British Columbia’s Research Ethics Board (H20-02744). Participants gave informed consent to participate in the study before taking part.
Fabbri L, Moss S, Khan FA, et al. Parenchymal lung abnormalities following hospitalisation for COVID-19 and viral pneumonitis: a systematic review and meta-analysis. Thorax2023; 78:191–201.
Hecker L, Thannickal VJ. Nonresolving fibrotic disorders: idiopathic pulmonary fibrosis as a paradigm of impaired tissue regeneration. Am J Med Sci2011; 341:431–4.
Allen RJ, Guillen-Guio B, Oldham JM, et al. Genome-Wide Association Study of Susceptibility to Idiopathic Pulmonary Fibrosis. Am J Respir Crit Care Med2020; 201:564–74.
van Moorsel CHM, van der Vis JJ, Duckworth A, et al. The MUC5B Promoter Polymorphism Associates With Severe COVID-19 in the European Population. Front Med8:668024
Verma A, Minnier J, Wan ES, et al. A MUC5B Gene Polymorphism, rs35705950-T, Confers Protective Effects Against COVID-19 Hospitalization but Not Severe Disease or Mortality. Am J Respir Crit Care Med2022; 206:1220–9.
van der Mark VA, Ghiboub M, Marsman C, et al. Erratum to: Phospholipid flippases attenuate LPS-induced TLR4 signaling by mediating endocytic retrieval of Toll-like receptor 4. Cell Mol Life Sci2017; 74:1365.
Hollingsworth LR, Sharif H, Griswold AR, et al. DPP9 sequesters the C terminus of NLRP1 to repress inflammasome activation. Nature New Biol2021; 592:778–83.
Hobbs BD, Putman RK, Araki T, et al. Overlap of Genetic Risk between Interstitial Lung Abnormalities and Idiopathic Pulmonary Fibrosis. Am J Respir Crit Care Med2019; 200:1402–13.
Initiative C-HG. The COVID-19 Host Genetics Initiative, a global initiative to elucidate the role of host genetic factors in susceptibility and severity of the SARS-CoV-2 virus pandemic. Eur J Hum Genet2020; 28:715–8.
Marinescu D-C, Hague CJ, Muller NL, et al. Integration and Application of Radiologic Patterns From Clinical Practice Guidelines on Idiopathic Pulmonary Fibrosis and Fibrotic Hypersensitivity Pneumonitis. Chest2023; 164:1466–75.
Fingerlin TE, Murphy E, Zhang W, et al. Genome-wide association study identifies multiple susceptibility loci for pulmonary fibrosis. Nat Genet2013; 45:613–20.
Nakano Y, Yang IV, Walts AD, et al. MUC5B Promoter Variant rs35705950 Affects MUC5B Expression in the Distal Airways in Idiopathic Pulmonary Fibrosis. Am J Respir Crit Care Med2016; 193:464–6.
Zhang P, Li J, Liu H, et al. Long-term bone and lung consequences associated with hospital-acquired severe acute respiratory syndrome: a 15-year follow-up from a prospective cohort study. Bone Res2020; 8:8.
