The impact of hiatus hernia in hypersensitivity pneumonitis

Around 40% of patients with idiopathic pulmonary fibrosis (IPF) have a hiatus hernia (HH), the presence of which has been linked with greater lung function decline and increased mortality.1-3 HH has been reported in approximately one-third of patients with hypersensitivity pneumonitis (HP), but data are sparse.3 There are parallels between fibrotic HP and IPF suggesting HH may have a similar impact in both conditions.4 A causal relationship between HH and lung fibrosis has not been established. HH and chronic microaspiration (CMA) might promote or potentiate lung fibrosis,1 and it is hypothesised that increased negative intrathoracic pressure caused by lung fibrosis can lead to retraction of the gastro-oesophageal junction into the thoracic cavity.5 In this study, we explored the prevalence and impact of HH in patients with HP. One hundred and forty-one consecutive patients with a multidisciplinary diagnosis of HP,6, 7 referred to the Royal Brompton Hospital between January 2010 and December 2014, were retrospectively evaluated (Table 1). The presence and size of HH (graded 0–4) was determined on computed tomography (CT) scans by two consultants.3, 5 The extent (nearest 5%) and pattern of fibrosis, were recorded as were CT features potentially indicative of CMA (consolidation, oesophageal dilatation and tree-in-bud pattern). In a control population of 175 patients without interstitial lung disease (ILD) undergoing CT coronary angiography, the presence and size of HH was also assessed. Associations between HH, transplant-free survival, and lung function trajectory were assessed (change in forced vital capacity [FVC] and diffusing capacity of the lungs for carbon monoxide [DLco] at 12 (±6) months from baseline) with censoring at 01/02/2022. Statistical analyses were conducted with STATA (StataCorp, College Station, TX, USA), with significance tested by Kruskal–Wallis, Mann–Whitney U, Chi-squared or Fisher's exact tests. Survival analyses were calculated by Cox proportional hazards. Lung function trajectory was analysed using linear mixed effects models with subject as a random effect and time from baseline, baseline lung function, and antacid use as fixed effects. This study had ethical approval under IRAS ID 296357. The prevalence of HH was significantly higher in HP than in controls (60% vs. 25%; p < 0.001), with the HP cohort having larger HH graded as 2–4 (32% vs. 3%; p < 0.001). Patients with HP were older (mean age = 64.2 ± 9.6 years vs. 60.7 ± 13.1 years; p = 0.007), but were matched for sex (F = 78 (55%) vs. 100 (57%)). One hundred thirty-one patients with HP (93%) had evidence of fibrosis on CT. The mean extent of fibrosis as assessed by consensus evaluation of CT findings was 53% (±27.4%) in patients with HH, and 43.9% (±29.5%) in patients without. Presence and size of HH were not associated with BMI. There was no relationship between the presence of HH and visual extent of fibrosis on CT (with HH, mean fibrosis extent (±SD) on CT 53.0% (±24.7%) versus without HH, 43.9% (±29.5%); p > 0.05). HH was not associated with baseline lung function (with HH, mean(%) FVC 2.0 L (68.9%) ± 0.9 L and DLco 39.5 ± 13.6% versus mean(%) FVC 2.2 L (69.2%) ± 1.1 L and DLco 41.1 ± 17.1% without HH (all p > 0.05)). HH was not associated with mortality (Figure 1A). Median survival in those with HH was 5.48 years (2.47–10.25 years) versus 5.43 years (2.27–8.81 years) in those without, HR:1.09 (95% CI: 0.72–1.66, p = 0.69). HH was not associated with lung function decline; with HH median 12-month FVC change −10 mL (IQR: −120 to +120 mL) and DLco change −1.6% (IQR: −15.0 to +10.6%) and without HH 12-month FVC change −40 mL (−210 to +150 mL) and 12-month DLco change −2.3% (−18.8 to +9.1%) (all p > 0.05). In patients with fibrotic HP with a usual interstitial pneumonia (UIP) pattern on CT (n = 23) there was a non-statistically significant trend towards poorer survival associated with HH and lung function decline. Median survival with HH was 4.13 years (2.59–6.72 years) vs. 8.44 years (5.48–n/a years) in those without, (HR = 2.86, 95% CI: 0.91–8.97, p = 0.07) (Figure 1B) but there was no such trend in those with non-UIP patterns of ILD (Figure 1C). Patients with UIP and HH experienced median 12-month FVC decline of −20 mL (−100 mL to +100 mL) and DLco decline of −2.99% (−4.67% to +4.35%) and those with UIP and no HH experienced mean FVC rise of +160 mL (−40 mL to +210 mL) and DLco rise of +0.83% (−4.27% to +5.69%) (all p > 0.05). In patients with a UIP pattern, there was a significant relationship between presence of HH and FVC decline (coefficient: −1.08 (95% CI: −1.74 to −0.42), p = 0.001) and DLco (coefficient: −1.84 (95% CI: −2.89 to −0.78), p = 0.001). However, significance was lost when adjusting for baseline lung function. Consolidation was more prevalent in HP patients with larger HH of grade 2–4 than those with smaller or absent grade 0–1 HH (13% vs. 3%; p = 0.021), with an exposure-response relationship (Figure 1D). We report that the prevalence of HH in patients with HP is more than twice that of a control population and higher than in equivalent IPF studies with over half being moderate-to-large hernias.1-3 The prevalence of HH in our control group (25%) was similar to the 23% prevalence in a study of nearly 5000 non-ILD patients.5 There was no association between HH and lung function decline or mortality. This contrasts with data in IPF where there is an established association between HH, disease progression and survival.1, 3 In the subgroup of HP patients with a UIP pattern, there was a trend to an association between HH and mortality and a similar direction of lung function decline. Numbers in this group were small and so this observation needs further exploration, but is plausible given similar previous findings in patients with IPF. We attempted to establish the directionality to the relationship between HH and HP. There was no association between increased fibrosis (on CT or lung function) and larger HH, suggesting that negative intrathoracic pressure does not promote the development of HH. However, consolidation, a non-specific CT marker of aspiration, was increasingly more apparent in HP patients with HH of increasing size. This suggests that HH and subsequent CMA, could potentiate ILD in HP.8 With up to 60% of HP cases having an unidentified causative antigen, and only 5%–15% of those exposed to causative antigens developing HP, we speculate that CMA could drive the development of HP in genetically predisposed individuals.9, 10 This theory is supported by a recent Mendelian randomization study that showed GORD increased the risk of IPF, yet not vice versa.11 With this potential multi-hit hypothesis involving CMA, further research in HP patients could involve objective measures of aspiration with gastrointestinal physiology studies to unlock novel pathomechanistic understandings of HP.12 There were several study limitations. Firstly, our data was acquired retrospectively from a single centre. Our HP-UIP sub-group was small and so future studies could expand this. While most HP patients had 12-month FVC and DLco data, future studies could assess longer-term lung function trajectory. HRCT is not the gold-standard investigation for detecting HH; high-resolution manometry, barium swallow radiography or endoscopy are more routinely used.3 However, HRCT is accurate in detecting HH and all scans were reviewed by two experienced consultants.3 Other strengths include complete survival data and the inclusion of a relevant control group.2 Our study has demonstrated that 60% of HP patients have a HH, over half of which are moderate-to-large. There was no association with poorer outcomes, yet their prevalence and association with CT signs of aspiration may further elucidate HP aetiopathogenesis. David A. Heriot: Conceptualization (supporting); data curation (equal); formal analysis (lead); investigation (supporting); methodology (supporting); project administration (equal); resources (supporting); software (equal); visualization (lead); writing – original draft (lead); writing – review and editing (supporting). Carmel J. W. Stock: Conceptualization (supporting); data curation (equal); formal analysis (supporting); methodology (supporting); project administration (supporting); resources (supporting); software (equal); supervision (supporting); visualization (supporting); writing – original draft (supporting); writing – review and editing (supporting). Zain-Ul-Abideen Mumtaz: Data curation (equal); investigation (supporting); writing – review and editing (supporting). R. Gisli Jenkins: Methodology (supporting); writing – review and editing (supporting). Felix Chua: Methodology (supporting); writing – review and editing (supporting). Phillip L. Molyneaux: Methodology (supporting); writing – review and editing (supporting). Anand Devaraj: Methodology (supporting); writing – review and editing (supporting). Vasilis Kouranos: Methodology (supporting); writing – review and editing (supporting). Athol U. Wells: Conceptualization (supporting); methodology (supporting); visualization (supporting); writing – review and editing (supporting). Elizabetta A. Renzoni: Methodology (supporting); resources (supporting); writing – review and editing (supporting). Simon P. G. Padley: Conceptualization (supporting); data curation (equal); investigation (supporting); methodology (supporting); project administration (supporting); resources (supporting); supervision (supporting); writing – review and editing (supporting). Sujal R. Desai: Conceptualization (lead); data curation (equal); formal analysis (supporting); investigation (lead); methodology (lead); project administration (equal); resources (lead); supervision (lead); visualization (supporting); writing – original draft (supporting); writing – review and editing (lead). Peter M. George: Conceptualization (lead); data curation (equal); formal analysis (supporting); investigation (lead); methodology (lead); project administration (equal); resources (lead); supervision (lead); visualization (supporting); writing – original draft (supporting); writing – review and editing (lead). Elizabetta A. Renzoni is an Editorial Board member of Respirology and a co-author of this article. She was excluded from all editorial decision-making related to the acceptance of this article for publication. She reports lecture fees paid to her institution from Chiesi; advisory board work from Roche; and grants, lecture fees paid to her institution, support for attending meetings and advisory board work from Boehringer Ingelheim, outside of submitted work. R. Gisli Jenkins Jenkins reports grants from Biogen, Galecto, Nordic Biosciences and RedX; personal fees from Brainomix, Bristol Myers Squibb, Chiesi, Cohbar, Daewoong, Vercayte, Resolution Therapeutics, Roche, PatientMPower, Pinsent Masons LLP, Boehringer Ingelheim, Galapagos, Vicore and NuMedii; grants and personal fees from AstraZeneca, GlaxoSmithKline and Pliant; and is President for Action for Pulmonary Fibrosis, outside of submitted work. Felix Chua reports personal fees from Boehringer Ingelheim, outside of submitted work. Phillip L. Molyneaux reports advisory board fees from Trevi and Qureight; advisory board and speaker fees from Boehringer Ingelheim and Hoffman-La Roche; and grants and advisory board fees from AstraZeneca, outside of submitted work. Anand Devaraj reports personal fees from Boehringer Ingelheim, Roche, Vicore, Galectto and Brainomix; and stock options from Brainomix, outside of submitted work. Vasilis Kouranos reports lecture fees from Novartis and Boehringer Ingelheim, outside of submitted work. Athol U. Wells reports personal and lecture fees from Boehringer Ingelheim and Roche; and advisory board work from Veracyte, outside of submitted work. Simon P.G. Padley has nothing to disclose. Sujal R. Desai has nothing to disclose. Peter M. George reports personal fees from Teva, Cipla, AstraZeneca, GSK and Avalyn; personal fees and support for attending conferences from Roche; grants, personal fees and support for attending conferences from Boehringer Ingelheim; and personal fees and stock options from Brainomix, outside of submitted work. The other authors had nothing to disclose. This retrospective analysis was ethically approved and did not require individual patient consent as per IRAS ID 296357. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to the privacy or ethical restrictions.