Standard pulmonary function tests and respiratory oscillometry patterns in hypersensitivity pneumonitis and idiopathic pulmonary fibrosis

Abstract

Background

Hypersensitivity pneumonitis (HP) is an interstitial lung disease (ILD) caused by repeated exposure to inhaled antigens, leading to small airway and parenchymal inflammation. Diagnosis is based on a detailed clinical history, chest imaging and invasive tests such as bronchoalveolar lavage. Distinguishing HP from other ILDs is challenging. Respiratory oscillometry, a novel pulmonary function test (PFT), is highly sensitive to small airway abnormalities. Oscillometry measurement of reactance is strongly correlated with gender-age-physiology score, a prognostic tool used to predict mortality and disease severity in idiopathic pulmonary fibrosis (IPF).

Objective

To determine if oscillometry and standard PFT patterns are different in HP and IPF.

Methods

39 HP (79.5% with fibrotic HP) were enrolled from October 2022 to December 2023 for oscillometry before clinically-indicated standard PFTs and compared with 39 age-matched and sex-matched patients with IPF who also had same day oscillometry and standard PFTs. The main oscillometry metrics of interest were R5-19 (the difference in resistance from 5 to 19 Hz, a metric of small airway function and ventilatory inhomogeneity that increases with worsening respiratory mechanics), X5 (reactance at 5 Hz) which primarily reflects respiratory elastance and AX (area of reactance), a summative measure of the respiratory system stiffness across a range of frequencies, that behaves similarly but in opposite direction to X5.

Results

Patients with HP exhibited higher residual volume/total lung capacity (RV/TLC), lower per cent predicted (%) forced expiratory volume in 1 s (FEV₁) and % predicted forced vital capacity (FVC) than IPF (p<0.05) while FEV₁/FVC and %TLC were similar. Oscillometry showed higher R5-19 in HP. RV/TLC ratio correlated with AX (r²=0.72), X5 (r²=0.66) and R5-19 (r²=0.64).

Conclusion

Gas trapping (RV/TLC>0.40) is a feature of HP not observed in IPF. The strong correlations of RV/TLC with AX, X5 and R5-19 suggest that oscillometry can provide non-invasive markers of small airway obstruction in HP that can differentiate it from IPF.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Differentiation of hypersensitivity pneumonitis from idiopathic pulmonary fibrosis is important for effective therapeutic interventions but often difficult due to overlapping clinical features.

WHAT THIS STUDY ADDS

• Oscillometry is a non-invasive novel effort-independent pulmonary function test that provides metrics of ventilatory heterogeneity and small airway obstruction that can help distinguish hypersensitivity pneumonitis from idiopathic pulmonary fibrosis.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• The addition of oscillometry to standard pulmonary function tests as routine assessment of interstitial lung disease will offer external validation of our findings that could lead to the adoption of oscillometry as part of routine clinical practice and improved patient outcomes.

Introduction

Hypersensitivity pneumonitis (HP) is an immune-mediated interstitial lung disease (ILD) that results from repeated exposure to inhaled environmental antigens leading to inflammation and fibrosis in the small airways and lung parenchyma.¹ The clinical presentation of HP varies widely from acute, reversible inflammation to chronic fibrotic disease.¹ Current diagnostic guidelines emphasize a multidisciplinary approach that includes a detailed history to identify the inciting antigen, high-resolution CT (HRCT) of the chest and supportive evidence such as lymphocytosis in the bronchoalveolar lavage or histopathological features of small airway inflammation and parenchymal damage consistent with HP on lung biopsy.¹ While histology is the gold standard, surgical lung biopsy is associated with significant morbidity and mortality and usually reserved for cases where clinical, radiologic and other non-invasive diagnostic criteria are inconclusive.^{1 2} Transbronchial cryobiopsy has emerged as a less invasive alternative to surgical biopsy but is associated with morbidity and not widely available.^{1 3 4} Accurate diagnosis remains challenging as an inciting antigen is not found in 30–50% of cases.^{5 6} Differentiating HP from idiopathic pulmonary fibrosis (IPF) is challenging, particularly in fibrotic HP (fHP), due to overlapping clinical and radiological presentations. Misclassification of HP as IPF delays appropriate intervention, potentially leading to worse outcomes or inappropriate treatment strategies.⁷ Physiological markers to differentiate HP from other ILDs have not been well described. An early study using detailed analysis of flow-volume indices, body plethysmography, oesophageal balloon and closing volume identified mechanical inhomogeneity of ventilation as a key feature of HP when compared with a group of patients exposed to allergens without respiratory symptoms.⁸ Others had reported that small airway dysfunction in HP that was not discernible by spirometry could be detected using multiple-breath washout⁹ and respiratory oscillometry.^{10 11}

Respiratory oscillometry is a novel pulmonary function test (PFT) modality that measures respiratory system impedance.^{12 13} Conducted during tidal breathing, it is easier, faster and better tolerated by patients compared with spirometry which requires repeated forced expiratory manoeuvres.¹⁴ Studies show that oscillometry is also more sensitive to changes in the lung periphery and small airways.^{15 16} The oscillometry metric of reactance at 5 Hz (X5) has been shown to distinguish ILD from obstructive lung diseases with the difference between the mean inspiratory X5 (X5.in) and mean expiratory X5 occurring in the opposite direction for ILD compared with that observed in obstructive lung diseases.17,20 In HP, low X5, high area of reactance (AX, an integrated area under the curve from X5 to the resonance frequency, Fres), a metric of ventilatory inhomogeneity and high R5-20 (difference of resistance from 5 to 20 Hz, a metric of small airway resistance and ventilatory inhomogeneity) has been reported.^{10 11} However, the correlation of these oscillometry metrics with standard PFTs has not been assessed to the best of our knowledge.

We previously showed in a well-characterised IPF cohort that X5.in and the end-inspiratory reactance (XeI) were strongly correlated with % predicted forced vital capacity (FVC) and the gender-age-physiology score.²¹ The XeI, in particular, was highly correlated with the pulmonary vascular volume,²¹ a quantitative CT marker associated with disease prognosis.²² The patterns of oscillometry and standard PFT that include plethysmography among the different ILD subtypes have not been well characterised.

In this study, we aim to characterise the oscillometry and standard PFT profiles of patients with HP, their relationships with each other and compare them to those found in IPF. This study seeks to determine whether oscillometry can provide additional, non-invasive biomarkers of HP that can improve the diagnostic ability to differentiate HP from IPF.

Methods

Participant recruitment

This was a prospective observational study that recruited from the ILD Clinic at University Health Network (UHN) for respiratory oscillometry before each clinically indicated standard PFT.

Patient and public involvement

Patients and the public were not involved in the study design.

For the current analysis, we included patients who met the American Thoracic Society (ATS), Japanese Respiratory Society (JRS), and Asociación Latinoamericana del Tórax (ALAT) criteria for HP¹ and were enrolled between October 2022 and December 2023. Age and sex-matched patients, diagnosed with IPF according to the latest ATS/European Respiratory Society (ERS)/JRS/ALAT guidelines²³ and already enrolled in an ongoing prospective IPF study,²¹ were included for comparison. Only patients diagnosed as high probability or definite HP or IPF after multidisciplinary review were included.²⁴

Spectral (5–37 Hz)^{12 25} and 10 Hz monofrequency²⁶ oscillometry were performed according to ERS guidelines and established quality control standards using the tremoflo C-100 device (Thorasys, Montreal, Canada) followed by same day standard PFTs at the Toronto General Hospital Pulmonary Function Laboratory. Spirometry, body plethysmography, diffusing capacity for carbon monoxide (DLCO) using the BodyBox 5500 Plethysmography (Medisoft, Sorinnes, Belgium) and 6-minute walk test (6MWT) were conducted following international guidelines.27,29 Reference values for the standard PFT were derived from Gutierrez et al³⁰ and from Oostveen et al³¹ for oscillometry. The key spectral oscillometry metrics are R5 (resistance at 5 Hz), a measure of the total respiratory resistance, R5-19 (difference of resistance from 5 Hz to 19 Hz) which reflects small airway resistance and ventilatory inhomogeneity, X5 and AX (figure 1A,C). X5, an accepted measure of respiratory elastance, is affected also by the heterogeneity of lung periphery and is a key determinant of AX.^{13 32} X5 and AX move in opposite directions. Worsening respiratory mechanics are reflected by an increase in resistance, higher Fres and AX and more negative reactance values.^{13 15} The 10 Hz measurements are a novel intrabreath oscillometry technique that tracks changes in respiratory mechanics continuously during tidal breaths and focuses on the zero-flow instants of breathing, such as end expiration and end inspiration, where the upper airway non-linearities are minimal.²⁶ Changes in impedance for intrabreath oscillometry are plotted against flow and volume to identify changes during the inspiration and expiration phases of tidal breathing (figure 1B,D).

Figure 1. Impedance is plotted as resistance (black) and reactance (red) for a patient with IPF (A,B) and HP (C,D). Spectral oscillograms (A,C): the R5-19, Fres and AX are illustrated. Note the increased R5-19 in the patient with HP (C) compared with the patient with IPF (A). Intrabreath oscillograms (B,D): arrows indicate inspiratory and expiratory directions wherever looping is obvious. Asterisk marks the drop in reactance during the latter half of expiration, indicating tidal expiratory flow limitation in this HP subject. Note the opposite looping directions as well as the different patterns of the reactance versus flow loops in the HP and IPF examples. AX, area of reactance; Fres, resonance frequency; HP, hypersensitivity pneumonitis; IPF, idiopathic pulmonary fibrosis; R5-19, difference in resistance from 5 to 19 Hz.

FVC, forced expiratory volume in 1 s (FEV₁), residual volume (RV) and total lung capacity (TLC) were compared with the following key oscillometry metrics of interest: R5, R19 (resistance at 19 Hz), R5-19, X5 and AX. An RV/TLC ratio of ≥ 40% was used as an indicator of gas trapping, based on population and chronic obstructive pulmonary disease (COPD) cohort studies that identified this cut-off to be predictive of progression to COPD in smokers with preserved spirometry,³³ and associated with increased mortality and worse clinical outcomes in patients with COPD.³⁴

Chest HRCT within 6 months of paired oscillometry-standard PFT testing was reviewed by an experienced chest radiologist following the ATS/JRS/ALAT guidelines.1,3 In brief, the fHP pattern is characterised by irregular linear opacities/coarse reticulation with lung distortion (ie, fibrosis), random axial and craniocaudal distribution frequently with relative sparing in the lower lung zones and findings suggestive of small airway disease including ill-defined centrilobular nodules, mosaic attenuation, a three-density pattern and/or expiratory gas trapping.³⁵ The HP group was categorised as fHP and non-fibrotic HP (nfHP) phenotypes based on the CT findings and extensive chart review of the electronic medical records of the clinical history, multidisciplinary discussions and other laboratory findings.

Statistical analyses were conducted using RStudio (The R Foundation, Boston, Massachusetts, USA). Absolute and percent predicted values were reported as mean±SD or median with IQR, as appropriate depending on data distribution. Relationships among PFTs and oscillometry parameters were explored using linear, second-degree polynomial and exponential regression models, accounting for age, sex, height and weight. We used the coefficient of determination (r²) to assess and compare the goodness-of-fit across these models. Comparisons between HP and IPF groups, and fHP versus nfHP groups were conducted using independent t-tests or Mann-Whitney U tests for continuous variables, Fisher’s exact test or χ² tests for categorical variables.

Results

From October 2022 to December 2023, 39 of 40 patients with HP identified from the UHN ILD clinic consented to the study. The HP (mean age 69±9) and IPF (mean age 69±10) groups were older with almost equal sex distribution and similar smoking history (table 1). Most patients with HP (31/39; 79.5%) had fibrotic disease. Both groups showed a mild restrictive pattern on standard PFTs with mildly reduced percent predicted (%) TLC and mildly reduced % DLCO but the HP group had lower %FVC and %FEV₁ compared with the IPF group (p<0.05 for both, table 2). FEV₁/FVC was normal and similar in both groups. Notably, the RV/TLC ratio in the HP group was higher than IPF (p=0.04). Functional exercise capacity, assessed with 6MWT, was normal and similar in both groups (p>0.10, table 1). The Borg scales were similar except for the pre-6MWT fatigue scores (p=0.03, table 1).

Table 1. Demographic, clinical characteristics and standard pulmonary function tests of HP and age-matched and sex-matched patients with IPF.

	HP (n=39)	IPF (n=39)	P value
Age (years)	69.3 (9.0)	68.5 (10.0)	0.77
Sex=M (%)	18 (46.2)	19 (48.7)	1.00
Height (cm)	166.1 (9.5)	165.6 (9.9)	0.83
Weight (kg)	80.3 (18.3)	75.8 (14.2)	0.24
BMI (kg/m2)	28.9 (5.2)	27.5 (3.7)	0.19
Smoking history (%)
People who have never smoked	25 (64.1)	17 (43.6)	–
People who have quit smoking ≤20 pk years	8 (20.5)	13 (33.3)	–
People who have quit smoking >20 pk years	6 (15.4)	9 (23.1)	–
Nintedanib	8	23	–
Pirfenidone	0	8	–
FVC (L)	2.3 (0.9)	2.60 (0.9)	0.18
% FVC	63.7 (16.24)	73.7 (21.5)	0.02
FEV1 (L)	1.9 (0.7)	2.1 (0.7)	0.25
%FEV1	66.0 (18.7)	77.1 (21.0)	0.02
FEV1/FVC	82.8 (6.7)	81.4 (6.3)	0.35
FEF25-75%	2.3 (1.1)	2.68 (1.3)	0.21
%FEF25-75	109.0 (44.3)	127.4 (59.2)	0.13
TLC (L)	4.1 (1.2)	4.2 (0.9)	0.68
%TLC	67.3 (17.7)	73.8 (15.9)	0.11
RV (L)	1.7 (0.6)	1.5 (0.3)	0.14
%RV	75.1 (27.3)	71.6 (15.5)	0.51
RV/TLC	42.1 (11.9)	36.9 (7.9)	0.04
DLCO (mL/min/mm Hg)	12.5 (4.3)	12.2 (3.8)	0.77
%DLCO	69.9 (16.9)	69.6 (19.3)	0.94
6MWT distance (m)	399.1 (127.5)	440.1 (143.9)	0.20
% 6MWT distance	86.8 (24.4)	99.0 (30.1)	0.06
Borg Dyspnoea Score (pre)	1.0 (1.1)	0.8 (1.1)	0.28
Borg Dyspnoea Score (post)	2.4 (2.1)	3.0 (1.6)	0.21
Borg Fatigue Score (pre)	1.51 (1.70)	0.78 (1.15)	0.03
Borg Fatigue Score (post)	2.78 (1.94)	2.42 (2.12)	0.45

Data are shown as mean (SD).

Bold values are statistically significant.

BMI, body mass index; DLCO, diffusing capacity for carbon monoxide; FEF_25-75%, forced expiratory flow at 25–75% of FVC; FEV₁, forced expiratory volume in 1 s; FVC, forced vital capacity; HP, hypersensitivity pneumonitis; IPF, idiopathic pulmonary fibrosis; 6MWT, 6-minute walk test; RV, residual volume; TLC, total lung capacity.

Table 2. Spectral and intrabreath oscillometry in HP and IPF groups.

	HP (n=39)	IPF (n=39)	P value
R5 (cm H2O s/L)	4.06 (3.19 to 5.15)	3.51 (2.93 to 4.57)	0.13
%R5	123.96 (93.59 to 147.29)	114.12 (102.92 to 133.18)	0.57
Z-Score R5 (mean (SD))	0.71 (1.22)	0.50 (0.91)	0.75
R19 (cm H2O s/L)	2.94 (2.47 to 3.55)	2.93 (2.46 to 3.61)	0.88
R5-19 (cm H2O s/L)	1.05 (0.61 to 1.38)	0.60 (0.30 to 1.04)	0.01
X5 (cm H2O s/L)	−2.36 (–3.83 to –1.72)	−2.09 (–2.89 to –1.48)	0.12
Z-Score X5 (mean (SD))	−2.68 (2.64)	−1.82 (1.73)	0.10
X5.in (cm H2O s/L)	−2.72 (–3.48 to –1.80)	−2.22 (–3.18 to –1.54)	0.13
X5.ex (cm H2O s/L)	−2.14 (–3.78 to –1.59)	−1.93 (–2.84 to –1.32)	0.10
AX (cm H2O/L)	12.97 (8.98 to 23.98)	12.28 (6.99, to 20.55)	0.24
%AX	356.76 (200.80 to 688.33)	343.75 (209.57 to 482.54)	0.56
Fres (Hz)	19.80 (17.17 to 24.23)	19.60 (16.12 to 23.33)	0.53
ReE (cm H2O s/L)	3.53 (2.87 to 4.07)	3.20 (2.70 to 3.78)	0.34
ReI (cm H2O s/L)	2.76 (2.25 to 3.16)	2.31 (2.08 to 2.79)	0.05
ΔR (ReE-ReI, cm H2O s/L)	0.57 (0.37 to 1.22)	0.85 (0.38 to 1.24)	<0.01
XeE (cm H2O s/L)	−0.56 (–1.30, to –0.07)	−0.22 (–0.69 to 0.20)	0.05
XeI (cm H2O s/L)	−0.58 (–1.06 to –0.40)	−0.58 (–1.04 to –0.23)	0.28
ΔX (XeE-XeI, cm H2O s/L)	0.40 (–0.31 to 0.70)	0.34 (0.11 to 0.69)	0.64

Data are shown as median (IQR).

Bold values are statistically significant.

AX, reactance area between 5 Hz and Fres; Fres, resonance frequency; HP, hypersensitivity pneumonitis; IPF, idiopathic pulmonary fibrosis; R5, resistance at 5 Hz; R19, resistance at 19 Hz; R5-19, difference between 5 and 19 Hz; ReE, resistance at end-expiration; ReI, resistance at end-inspiration; X5, reactance at 5 Hz; XeE, reactance at end-expiration; XeI, reactance at end-inspiration; X5.ex, mean expiratory X5; X5.in, mean inspiratory X5.

The spectral oscillometry patterns in HP and IPF were similar with the notable exception of the elevated R5-19 in the HP group when compared with the patients with IPF (p=0.01; table 2, figure 1A,C). The pattern of normal R5 and R19, abnormally low X5, high Fres and high AX was similar in the two groups (p>0.05 for all, table 2). Intrabreath oscillometry revealed higher resistance at end inspiration, resistance at end-inspiration (ReI) (p=0.05) but lower ΔR (resistance at end-expiration (ReE)-ReI, p<0.01, table 2) and lower reactance at end expiration, reactance at end-expiration (XeE) (p=0.05) when compared with the IPF group. Tidal changes in resistance, ReI-ReE, were lower in the HP group compared with IPF (p<0.01) with no difference in the tidal changes in reactance, XeI-XeE (p=0.64, table 2). There are markedly different looping patterns and opposite directions of flow in the reactance versus flow diagrams in the intrabreath oscillograms of the patient with IPF and patient with HP (figure 1B,D) which were not reflected in the corresponding spectral oscillograms of the mean reactance values versus frequency curves (figure 1A,C).

Lung function patterns in fibrotic and non-fibrotic HP

The 31 patients with fHP who comprised 79.5% of the cohort had similar demographic and smoking profiles to the 8 patients with nfHP (table 3). Of the 31 patients with fHP, 8 were on antifibrotic therapy, while there were none in the nfHP group. The fHP group had lower %TLC (p=0.02), DLCO (p=0.04) and a trend to lower % DLCO (p=0.08) when compared with the nfHP (table 4). No differences were observed in the %FEV^₁, %FVC, %RV, RV/TLC or the 6MWT absolute or % walk distance. Spectral and intrabreath oscillometry results were similar in the fHP and nfHP subgroups with the exception of a lower XeI in the fHP group (p=0.07, table 3).

Table 3. Demographic, clinical characteristics and standard PFTs in fHP and nfHP.

	Fibrotic HP (n=31)	Non-fibrotic HP (n=8)	P value
Age (years)	69.7 (8.5)	66.8 (9.9)	0.45
Sex=M (%)	15 (48.4)	3 (37.5)	0.70
Height (cm)	167.0 (9.9)	162.8 (7.1)	0.19
Weight (kg)	81.3 (18.2)	76.3 (19.6)	0.53
BMI (kg/m2)	28.9 (4.9)	28.7 (6.8)	0.95
Smoking history (%)
People who have never smoked	20 (64.5)	4 (50)	–
People who have quit smoking ≤20 pk years	7 (22.6)	4 (50)	–
People who have quit smoking >20 pk years	4 (12.9)	0	–
Nintedanib	8	0	–
Pirfenidone	0	0	–
Corticosteroids	11 (35.5)	4 (50)	0.47
Mycophenolate mofetil	16 (51.6)	5 (62.5)	0.59
FVC (L)	2.3 (0.9)	2.5 (0.8)	0.43
%FVC	60.8 (14.7)	74.8 (18.0)	0.07
FEV1 (L)	1.9 (0.7)	2.0 (0.6)	0.55
%FEV1	63.3 (18.8)	76.8 (15.2)	0.05
FEV1/FVC	83.1 (6.7)	81.8 (6.9)	0.64
FEF25-75%	2.4 (1.1)	2.1 (0.7)	0.44
%FEF25-75	110.9 (48.1)	101.8 (25.1)	0.47
TLC (L)	4.0 (1.3)	4.3 (0.9)	0.48
%TLC	64.2 (17.5)	79.2 (13.5)	0.02
RV (L)	1.7 (0.6)	1.7 (0.5)	0.83
%RV	72.6 (27.8)	84.3 (24.9)	0.27
RV/TLC	42.4 (11.4)	40.8 (14.5)	0.78
DLCO (mL/min/mm Hg)	11.6 (4.0)	15.8 (4.3)	0.04
%DLCO	66.8 (15.7)	81.1 (17.4)	0.08
6MWT distance (m)	386.1 (110.4)	452.7 (183.9)	0.11
% 6MWT distance	84.5 (22.3)	96.2 (31.7)	0.13
Borg Dyspnoea Score (pre)	1.0 (1.2)	1.07 (0.9)	0.93
Borg Dyspnoea Score (post)	2.6 (2.2)	1.9 (1.5)	0.38
Borg Fatigue Score (pre)	1.29 (1.65)	2.43 (1.69)	0.14
Borg Fatigue Score (post)	2.79 (2.16)	2.71 (0.49)	0.86

Data are shown as mean (SD).

Bold values are statistically significant.

BMI, body mass index; DLCO, diffusing capacity for carbon monoxide; FEF_25-75%, forced expiratory flow at 25–75% of FVC; FEV₁, forced expiratory volume in 1 s; fHP, fibrotic HP; FVC, forced vital capacity; HP, hypersensitivity pneumonitis; 6MWT, 6-minute walk test; nfHP, non-fibrotic HP; PFT, pulmonary function test; RV, residual volume; TLC, total lung capacity.

Table 4. Spectral and intrabreath oscillometry in fHP and nfHP.

	Fibrotic HP (n=31)	Non-fibrotic HP (n=8)	P value
R5 (cm H2O s/L)	4.0 (3.16 to 5.07)	4.43 (3.88 to 5.48)	0.30
%R5	120.13 (91.57 to 144.72)	133.67 (104.85 to 148.73)	0.57
Z-Score R5 (mean (SD))	0.51 (1.25)	0.81 (0.85)	0.43
R19 (cm H2O s/L)	2.79 (2.38 to 3.52)	3.30 (2.95 to 3.69)	0.60
R5-19 (cm H2O s/L)	1.05 (0.69 to 1.38)	1.09 (0.50 to 1.47)	0.98
X5 (cm H2O s/L)	−2.37 (–3.83 to –1.67)	−2.14 (–3.59 to –1.86)	0.91
Z-Score X5 (mean (SD))	−2.58 (2.27)	−2.72 (2.77)	0.89
X5.in (cm H2O s/L)	−2.81 (–3.55 to –1.76)	−2.52 (–3.20 to –1.97)	0.97
X5.ex (cm H2O s/L)	−2.16 (–3.78 to –1.56)	−1.89 (–3.93 to –1.74)	0.91
AX (cm H2O/L)	14.50 (8.98 to 22.67)	11.89 (10.0 to 24.60)	0.93
%AX	356.76 (199.23 to 736.91)	337.9 (239.4 to 466.3)	0.68
Fres (Hz)	19.80 (17.22 to 23.26)	21.29 (16.87 to 24.71)	0.91
ReE (cm H2O s/L)	3.32 (2.82 to 4.07)	3.88 (3.63 to 4.06)	0.15
ReI (cm H2O s/L)	2.66 (2.12 to 3.0)	2.95 (2.73 to 3.28)	0.19
ΔR (ReI-ReE, cm H2O s/L)	0.50 (0.32 to 1.22)	0.82 (0.50 to 1.21)	0.18
XeE (cm H2O s/L)	−0.40 (–1.18 to –0.08)	−0.93 (–1.63, –0.13)	0.63
XeI (cm H2O s/L)	−0.80 (–1.10 to –0.47)	−0.41 (–0.50 to –0.33)	0.07
ΔX (XeI-XeE, cm H2O s/L)	0.42 (–0.01 to 0.70)	−0.14 (–0.82 to 0.56)	0.19

Data are shown as median (IQR).

AX, area of reactance; fHP, fibrotic HP; Fres, resonance frequency; HP, hypersensitivity pneumonitis; nfHP, non-fibrotic HP; R5, resistance at 5 Hz; R19, resistance at 19 Hz; R5-19, difference between 5 and 19 Hz; ReE, reactance at end-expiration; ReI, reactance at end-inspiration; X5, reactance at 5 Hz; XeE, resistance at end-expiration; XeI, resistance at end-inspiration; X5.ex, mean expiratory X5; X5.in, mean inspiratory X5.

Overall, the standard PFT and oscillometry findings reflect greater physiological impairment in the patients with fHP with lower %TLC and DLCO and increased lung stiffness as indicated by the lower XeI. The patients with fHP and patients without nfHP have similar abnormalities in gas trapping as measured by the RV/TLC ratio, along with small airway obstruction and ventilatory inhomogeneity as reflected by R5-19 and AX.

Correlation between oscillometry, PFTs and 6MWT

We assessed the correlations between oscillometry with standard PFTs with a focus on the standard PFT parameters that were different between HP and IPF, namely RV/TLC, %FEV₁ and %FVC. among the different regression models, the best fit was observed using polynomial regression model (online supplemental table 1). The highest correlations were observed between RV/TLC with multiple oscillometry metrics of ventilatory inhomogeneity and small airway obstruction, namely AX (r²=0.72). X5 (r²=0.66), R5-19 (r²=0.64). The correlations of %FEV₁ and % FVC to the oscillometry metrics were generally poor, with the highest observed between %FVC and AX (r²=0.41). No correlations were found between DLCO, %DLCO, 6MWT distance or % walk distance with any of the oscillometry metrics (data not shown).

Discussion

Small airway involvement and ventilatory inhomogeneity are features of HP that are poorly detected by standard PFTs. In this prospective study, we conducted same day spectral and intrabreath monofrequency oscillometry, followed by spirometry, plethysmography, DLCO and 6MWT to provide comprehensive assessment of the respiratory physiology patterns in patients with HP and patients with IPF. Notably, the majority of the patients with HP in the cohort had fHP where differentiation from IPF is of clinical relevance. The single distinguishing feature on standard PFT that differentiated HP from IPF was an increase in RV/TLC, indicating evidence of gas trapping not observed in IPF. The RV/TLC ratio in our cohort (0.42) was similar to the RV/TLC measured in previous studies of HP where values of 0.42 and 0.39 were reported.^{10 11} Increased RV/TLC in the presence of normal FEV₁/FVC likely represents small airway obstruction as FEV₁ remains unchanged until ~70% of all small airways are obstructed.^{36 37}

Same-day spectral and intrabreath oscillometry confirmed the findings of small airway obstruction and ventilatory inhomogeneity in the patients with HP, with increase in R5-19 and AX. Increased R5-20, a measure similar to R5-19, and AX were also reported in 20 patients with HP evaluated with impulse oscillometry with values that were comparable to ours (R5−20=1.4 cm H₂O s/L; AX=18.7 cm H₂O/L).¹¹ R5-20 was reported to be normal in the 28 patients with HP studied by Dias et al.¹⁰ Direct comparison of our data with this group is limited as their patients were evaluated with respiratory oscillometry after spirometry and plethysmography, that is, at a lung volume that is likely higher than functional residual capacity. Oscillometry measures the mechanical impedance of the respiratory system, which is volume-dependent and should be measured prior to spirometry as recommended by the recent ERS guidelines.¹²

RV/TLC and R5-19, the metrics that were different between the HP and IPF groups, were highly correlated in the HP group. Correlation analysis also revealed strong associations of RV/TLC with AX and X5, metrics of ventilatory inhomogeneity and lung elastance.¹³ While correlation analysis between the oscillometry and standard PFT measurements was not done in 20 patients with HP studied by Guerrero et al, RV/TLC improved (p=0.002) with a trend of improvement in R5-20 (p=0.08) after 4 weeks of corticosteroid therapy.¹¹ This finding provides tempting evidence that RV/TLC and R5-20 (or R5-19) could be used as potential markers to follow response to therapy or disease progression. We intend to investigate this posit with increased enrolment and longer follow-up assessment of our cohort.

Comparison with an age-matched and sex-matched IPF cohort revealed that although both groups shared similar degrees of diffusion impairment, patients with HP showed significantly greater restriction (%FVC) and higher RV/TLC ratios. The spectral and intrabreath oscillometry measurements in the 39 patients with IPF were similar to values reported in a larger cohort of the patients with IPF followed at our centre,²¹ a site of a Canada-wide ILD registry.³⁸ The observed differences between HP and IPF align with the distinct pathophysiology of HP, which involves airway-centred inflammation and fibrosis, compared with the primarily parenchymal fibrosis characteristic of IPF. Oscillometry parameters further distinguished the two diseases with increased R5-19 that was notably absent in patients with IPF. Additionally, the AX demonstrated strong correlations with RV/TLC, reinforcing the presence of subtle mechanical abnormalities associated with peripheral airway involvement. Intrabreath oscillometry revealed elevated ReI, where homogenisation of flow in the lung periphery is expected to occur,³⁹ suggesting the general increase in resistance in the whole bronchial tree in HP. These findings highlight the sensitivity of oscillometry for detecting changes in respiratory mechanics that may not be apparent with conventional PFTs, providing further data to the growing literature about the potential of oscillometry for early detection of lung disease.^{15 16}

Stratification of patients with HP into fibrotic and non-fibrotic phenotypes showed no statistically significant differences in oscillometry measures between the groups with the exception of a trend to lower XeI in fHP, which reflects stiffer lungs in the patients with fHP compared with the nfHP group. Of the different oscillometry parameters, XeI was found to be mostly highly correlated with IPF severity in an earlier study.²¹ The current findings suggest that oscillometry abnormalities are present regardless of fibrotic status in HP and likely reflect a common pathophysiologic substrate involving small airways irrespective of fibrotic progression.

This study has several limitations. The single-centre cohort limits the generalisability of the findings. The small sample size, especially in the nfHP subgroup, limits the statistical power to detect differences between the phenotypes. We conducted a preliminary analysis using receiver operating characteristics (ROC) curves to explore the performance of R5-19 and RV/TLC in discriminating HP from IPF. We found values for the area under the curve to be 0.67 and 0.61, respectively. We derived a cut-off value for R5-19 of 1.02 cm H₂O s/L with a specificity of 74% and sensitivity of 56% to distinguish HP from IPF. The poor performance of the ROC curves likely reflects the small sample size of our overall group. While the HP group had lower %FVC, the %TLC and %DLCO were similar to the IPF group. Moreover, the key indices of restrictive respiratory mechanics, namely AX, X5 and Fres were similar, suggesting that the degree of respirology physiological impairment was similar between the groups. Performance of the ROC curve analysis is expected to improve with increased enrolment as our study continues. A larger HP sample size would also permit a robust comparison of the performance of oscillometry and standard PFT between the patients with fHP and patients with IPF to identify key metrics and their cut-off values to help improve the diagnostic certainty between these two populations and other ILDs.

In summary, respiratory oscillometry reveals important physiological distinctions in patients with HP, including evidence of small airway inhomogeneity and gas trapping that differentiate it from IPF. These features were present across both fHP and nfHP phenotypes, suggesting a shared component of airway dysfunction. Oscillometry may serve as a valuable, non-invasive adjunct to conventional PFTs in assessing small airway function in patients with HP. Further studies with a larger number of patients are required to estimate the extent of small airway involvement in ILD and incorporate respiratory oscillometry in the diagnostic work-up of these patients.

Data availability statement

Data are available upon reasonable request.