Mechanisms of exercise limitation and prevalence of pulmonary hypertension in hypersensitivity pneumonitis

Abstract

Background

Hypersensitivity pneumonitis (HP) is an immune-mediated condition triggered by inhaled antigens in susceptible individuals. Patients with HP often present with exercise limitation due to multiple mechanisms, including ventilatory and cardiocirculatory impairments, particularly pulmonary hypertension (PH).

Methods

In this cross-sectional study, patients with HP of varying severity underwent clinical assessment, including quality-of-life questionnaires, incremental treadmill cardiopulmonary exercise testing (CPET), pulmonary function testing, 6-min walk test and transthoracic echocardiography (TTE). Patients with a diffusing capacity of the lung for carbon monoxide (D_LCO) below 40% of predicted and/or a tricuspid regurgitation velocity >2.5 m·s⁻¹ or pulmonary artery systolic pressure >35 mmHg on TTE were referred for right heart catheterisation.

Results

Among 90 patients, 65% were female, with a median age of 65.5 years, and 85% had fibrotic HP. CPET was performed in 51 patients (56%), with a mean peak oxygen consumption of 58±13% of predicted. Exercise capacity was reduced in 98% of patients, with multifactorial limitation in 88%, primarily due to cardiocirculatory impairment (76%) and impaired gas exchange (66%). The mean 6-min walk distance was 334±116 m. Mean forced vital capacity and D_LCO were 56±20% and 42±22% predicted, respectively. PH was identified in 59% of patients (95% CI 48–72%), predominantly with combined post- and pre-capillary involvement.

Conclusions

Reduced exercise capacity is highly prevalent in patients with HP and is typically multifactorial. PH is also common, with a predominance of combined post- and pre-capillary patterns. These findings underscore the need for a comprehensive evaluation strategy in symptomatic HP patients.

Shareable abstract

Hypersensitivity pneumonitis is a heterogeneous disease that presents both fibrotic and inflammatory components. The prevalence of cardiovascular limitations is high, mainly due to pulmonary hypertension in individuals with fibrotic disease. https://bit.ly/4nnz77T

Introduction

Hypersensitivity pneumonitis (HP) is a heterogeneous immune-mediated disease that presents with varying degrees of pulmonary inflammation and fibrosis in susceptible individuals after exposure to environmental antigens [1, 2]. The disease can be classified as non-fibrotic HP (non-fHP), which primarily shows inflammatory changes or fibrotic HP (fHP), which is characterised by fibrotic and inflammatory changes of different degrees on chest computed tomography [1–3]. Patients with fHP typically have shorter survival and faster progression, especially when the triggering antigen is not identified [1, 2, 4].

HP can occur at various ages, including in children and the elderly, with an average age of 50–60 years [1, 2, 4]. The prevalence of HP varies, reflecting differences in intensity of exposure and geographical and social factors [5]. In the United States, the prevalence in the general population ranges from 1.67 to 2.71 per 100 000, increasing to 11.2 per 100 000 among those aged ≥65 years [6]. In Europe, the incidence of interstitial lung disease (ILD) is estimated to be 7.6 per 100 000 cases per year, and HP is identified in ∼6.6% of cases [7]. HP is estimated to represent 23% of all ILD in Brazil [8].

Pulmonary hypertension (PH) is a common complication of ILDs. The pooled prevalence of PH in the general population with ILD is ∼36% and can occur in 19−28% of cases of fHP, in which it is associated with a significantly poorer prognosis [9, 10]. PH is associated with reduced forced vital capacity (FVC), diffusing capacity of the lung for carbon monoxide (D_LCO), lower oxygen saturation in the 6-min walk test (6MWT) and reduced exercise capacity in patients with HP [11]. Beyond PH, several mechanisms of impairment can manifest or worsen during exercise in chronic pulmonary disease, including ventilatory, cardiovascular and gas-exchange limitations [12–16].

Cardiopulmonary exercise testing (CPET) is a valuable tool for diagnosing and monitoring patients with ILD. Patients with ILD have decreased exercise capacity, characterised by reduced peak oxygen uptake (V′_O2), reduced ventilatory reserve and hyperventilation, compared with patients without ILD [13, 15]. Mechanisms of exercise limitation in patients with ILD can be multifactorial and influenced by a combination of alterations such as impaired pulmonary gas exchange and pulmonary cardiovascular and peripheral musculoskeletal changes [11]. However, these mechanisms are still not fully understood and may vary depending on the population studied and the progression of the disease.

Few studies have evaluated exercise performance in patients with fHP and non-fHP [12, 13, 14, 17, 18], and none have been performed in a population with different functional severities. Furthermore, information on the pulmonary haemodynamic patterns in HP patients remains scarce, directly influencing the reported prevalence of PH in this condition – an aspect that has likely evolved following the recent updates to the diagnostic criteria for PH [10, 15, 16]. Therefore, the present study aimed to assess a broad population of patients with fibrotic and non-fHP, their exercise capacity and mechanisms of limitation as the prevalence of PH and its haemodynamic profiles.

Methods

Study population

All patients diagnosed with fibrotic or non-fHP with moderate or high diagnostic confidence according to the American Thoracic Society, the Japanese Respiratory Society and the Latin American Thoracic Association (ATS/JRS/ALAT) 2020 or CHEST 2021 guidelines followed at our centre were invited to participate in this cross-sectional study [1, 2]. Patients with low diagnostic confidence for HP – defined as those with nonsuggestive radiological patterns, no identifiable exposure or unavailable biopsy – were excluded. The distinction between fibrotic and non-fHP was made by two researchers using the ATS/JRS/ALAT criteria [1], primarily considering the presence or absence of radiological and/or histopathological evidence of fibrosis. In cases of uncertainty or diagnostic disagreement, a multidisciplinary team was responsible for case review. The protocol was approved by the local research ethics committee (process number 50165521.0.0000.0068). Patients were excluded if they had other lung disease or significant cardiovascular disease. This study was financed by the São Paulo Research Foundation (FAPESP), Brasil, process number 2021/06920-0, and the National Council for Scientific and Technological Development (CNPq - 307595/2021-8). Informed consent was obtained from all participants included in the study.

Objectives

The primary objective of this study was to describe the mechanisms underlying exercise limitations in patients with HP as well as to determine the prevalence of PH in HP and haemodynamic characteristics based on right heart catheterisation (RHC).

Study protocol

Visit 1

Clinical and demographic characteristics, as well as quality of life, were evaluated using the validated Portuguese versions of the Short Form Health Survey-36 (SF-36) and the King's Brief Interstitial Lung Disease (K-BILD) questionnaires [19–21]. Pulmonary function tests were carried out according to the ATS guidelines [22, 23] and performed using a pneumotachograph and body plethysmograph (Medical Graphics Corporation, St. Paul, MN, USA, and Elite Dx). The predicted values for the Brazilian population were used [24, 25]. The following functional parameters were assessed: FVC, forced expiratory volume in 1 s (FEV₁), total lung capacity, residual volume and D_LCO. The prevalence of restrictive and nonspecific patterns, and reduced D_LCO were determined as recommended [26]. The maximal voluntary ventilation (MVV) was calculated using the equation FEV₁×20.1+15.4, as proposed by Fox et al. [27], for those with fHP.

6MWT was performed following standard procedures, with heart rate and oxygen saturation (S_pO2) measured at rest and after exercise [28]. S_pO2 was measured by pulse oximetry (Onyx, model 9500; Nonin, Plymouth, MN, USA). Breathlessness was evaluated using the modified Borg scale before and after exercise [29]. Predictive values for the Brazilian population were used to assess the distance walked [30].

Visit 2

All patients underwent transthoracic echocardiography (TTE) (IE 33 equipment; Philips Medical Systems, Bothell, USA) to assess tricuspid regurgitation velocity (TRV) and/or pulmonary artery systolic pressure (sPAP). CPET was performed using a cycle ergometer (Ergostik; Geratherm Respiratory GmbH, Bad Kissingen, Germany) with a progressively increasing workload, according to the individual performance of each patient. CPET variables were measured breath-by-breath (Vmax Encore CPET; VIASYS CareFusion, San Diego, CA, USA) and included V′_O2, carbon dioxide output (V′_CO2), heart rate, respiratory rate, tidal volume, minute ventilation (V′_E), end-tidal carbon dioxide tension (P_ETCO2), oxygen pulse and respiratory exchange ratio. Patients who required supplemental oxygen were excluded from performing the CPET.

Exercise capacity was considered reduced if the peak V′_O2 was <85% of the predicted value [31]. The mechanisms of exercise limitation were defined as follows – 1) ventilatory limitation: ventilatory reserve (V′_E/MVV) lower than 15%; 2) cardiocirculatory limitation: anaerobic threshold (AT) lower than 40% of the predicted peak V′_O2, and/or oxygen pulse lower than 81% of the predicted value or with an early plateau; 3) cardiocirculatory limitation suggestive of PH: P_ETCO2 at AT lower than 40 mmHg and a V′_E/V′_CO2 slope >34; and 4) impaired gas exchange: oxygen desaturation (delta between baseline and end of exercise) >4% [31]. The reference values used in the CPET were those established by Wasserman and Hansen [32].

Visit 3

Patients with a TRV of ≥2.5 m·s⁻¹ and/or a sPAP of ≥35 mmHg on TTE and/or a D_LCO <40% of the predicted value underwent RHC, including those oxygen-dependent who met these indications [33–36]. The RHC was performed using a 7 F pulmonary artery catheter inserted through the jugular vein to assess haemodynamic parameters, including cardiac output, which was measured using the thermodilution technique. Mean pulmonary artery pressure (mPAP) >20 mmHg confirmed the diagnosis of PH. PH was classified according to pulmonary artery wedge pressure (PAWP) as pre-capillary (PAWP ≤15 mmHg, pulmonary vascular resistance (PVR) >2 Wood units (WU)), post-capillary (PAWP >15 mmHg, PVR <2 WU) or combined (mPAP >20 mmHg, PAWP >15 mmHg and PVR >2 WU) [33]. The haemodynamic measurements were performed in accordance with the 2022 European Society of Cardiology (ESC)/European Respiratory Society (ERS) guidelines [33]. Oxygen-dependent patients underwent RHC using oxygen via nasal cannula to maintain adequate oxygen saturation above 92%.

Hypoxaemic patients underwent all assessments except for CPET.

Statistical analysis

Quantitative variables are reported as n (%), mean±sd for variables with a normal distribution and median (25th–75th percentiles) for variables with non-normal distribution. The normality of the variables was assessed using the Shapiro–Wilk test. Qualitative variables were described using absolute and relative frequencies for all patients. The unpaired t-tests and the Mann–Whitney U-test were used to compare continuous variables, and the Fisher exact test and the chi-square test were used to compare categorical variables. The prevalences of reduced exercise capacity and PH were reported as proportions with 95% confidence interval (95% CI). The associations between variables were evaluated with the Pearson or Spearman correlation coefficients, as appropriate. A linear regression model was developed to assess variables associated with peak oxygen consumption and the presence of PH. Variables with a p-value <0.1 in univariable analyses were considered for inclusion in the multivariable linear regression model. Statistical significance was set at p<0.05. Data were analysed using IBM SPSS Statistics version 24 (SPSS Inc., Chicago, IL, USA).

Results

A total of 131 patients with a diagnosis of HP followed at the outpatient clinic of our institution were assessed between April 2022 and July 2023 to participate in this study. Of the initial sample, 90 patients were included in this study. 41 patients were excluded: 29 had low diagnostic confidence for HP (according to the ATS 2020/CHEST 2021 guidelines [1, 2]), six had severe heart disease and six refused to participate. Among the 90 patients, 76 (85%) had fHP, 29 were oxygen-dependent, eight had musculoskeletal limitation and the remaining 39 underwent CPET. Only 14 patients (15%) had non-fHP: one was oxygen-dependent, one had musculoskeletal limitation and 12 underwent CPET (figure 1).

FIGURE 1.

Flowchart with patient disposition. HP: hypersensitivity pneumonitis; CPET: cardiopulmonary exercise testing; PH: pulmonary hypertension; RHC: right heart catheterisation.

Clinical and functional data are presented in table 1. The median age of all patients was 65.5 years (58–73), and 66% were females. Patients with fHP had lower pulmonary function values and a shorter distance walked in 6MWT compared to patients with non-fHP (supplementary e-Table 1). Exposure was identified in 88 (98%) patients, with mould being the most common source (72%). The most concomitant comorbidities included systemic arterial hypertension (50%), gastroesophageal reflux disease (47%) and diabetes (39%). 85 (94%) of the patients reported dyspnoea. A restrictive pattern was found in 79 (87%) of the patients and D_LCO was reduced in 84 (93%). The distance walked in the 6MWT was 334±116 m (67±11% predicted).

TABLE 1.

Clinical characteristics, pulmonary function tests, 6-min walk test results and quality of life data in patients with hypersensitivity pneumonitis (HP) (n=90)

Age years	65.5 (58–73)
Female sex	59 (66%)
BMI kg·m−2	27.9 (25.2–31.6)
Former or current smoker	42 (47%)
Supplemental oxygen therapy	30 (33%)
Time from diagnosis to assessment months	36 (8.5–60)
Dyspnoea	85 (94%)
BDI	6 (3–7)
mMRC
0–1	23 (25%)
2	36 (40%)
3	26 (29%)
4	5 (6%)
Comorbidities
Systemic arterial hypertension	45 (50%)
GORD	42 (47%)
Diabetes	35 (39%)
Obesity	32 (36%)
Exposure	88 (98%)
Mould	65 (72%)
Birds	60 (67%)
Feather pillow and similar	13 (14%)
Pulmonary function tests
FVC L	1.71 (1.23–2.22)
% predicted	56±20
FEV1 L	1.40 (1.04–1.91)
% predicted	58±20
FEV1/FVC	84.7 (80.3–88.7)
TLC L	2.92 (2.46–3.73)
% predicted	60 (49–72)
RV L	1.22 (0.98–1.52)
% predicted	64 (49–83)
RV/TLC	38.8 (33.8–48.6)
DLCO mL·min−1·mmHg−1	7.97 (5.14–11.40)
% predicted	42.5 (27–54.3)
6MWT
Distance m	334±116
% predicted	67±11
O2 desaturation %	81
Borg dyspnoea score	4 (3–5)
Borg leg discomfort score	2(0–4.75)
Quality of life
SF-36 score
Physical functioning	25 (15–40)
Role limitations due to physical health	0 (0–50)
Role limitations due to emotional health	33 (0–100)
Vitality	50 (31.3–65)
Mental health	68 (48–84)
Social functioning	50 (25–87)
Bodily pain	57 (37–87)
General health	40 (25–65)
K-BILD score	57 (45–69)

Values are expressed as mean±sd, median (interquartile range) or n (%). BMI: body mass index; BDI: baseline dyspnoea index; mMRC: modified Medical Research Council; GORD: gastro-oesophageal reflux disease; FVC: forced vital capacity; FEV₁: forced expiratory volume in 1 s; TLC: total lung capacity; RV: residual volume; D_LCO: diffusing capacity of the lungs for carbon monoxide; 6MWT: 6-min walk test; SF-36: short-form health survey-36; K-BILD: King's Brief Interstitial Lung Disease.

51 patients underwent maximum CPET. Peak V′_O2 was 11.5 (9.9–15.3) mL·kg⁻¹·min⁻¹ (58±13% predicted). The minimum oxygen saturation reached was 86±8% in the overall population. The CPET variables are listed in table 2. Patients with fHP exhibited lower peak V′_O2 (% predicted), reduced ventilatory reserve and greater desaturation compared to patients with non-fHP (supplementary e-Table 2).

TABLE 2.

Variables of the cardiopulmonary exercise test (overall population n=51)

Peak V′O2, mL·kg−1·min−1	11.5 (9.9–15.3)
Peak V′O2, % predicted	58±13
V′O2 at AT, mL·kg−1·min−1	7.9 (6.8–10.2)
RER at peak	1.1±0.1
HR, beats·min−1	129±23
HR % predicted	85±19
RR, breaths·min−1	46 (40–52)
V′O2/HR % predicted	67 (59.5–78.5)
PETCO2, at AT mmHg	31±4
Ventilatory reserve %	25±16.4
Slope V′E/V′CO2	41±8
Minimum SpO2 %	86±8
Borg dyspnoea score	5 (4–7)
Borg leg discomfort score	5 (3–7)

Values are expressed as mean±sd or median (interquartile range). V′_O2: oxygen uptake; AT: anaerobic threshold; RER: respiratory exchange ratio; HR: heart rate; RR: respiratory rate; P_ETCO2: end-tidal carbon dioxide tension; V′_E: minute ventilation; V′_CO2: carbon dioxide production; HP: hypersensitivity pneumonitis; S_pO2: oxygen saturation.

Exercise capacity was reduced in 50 (98%) patients (95% CI 89−100%), and 88% of patients exhibited at least two mechanisms of limitation. Ventilatory limitation, cardiocirculatory limitation, limitation suggestive of PH and impaired gas exchange occurred in 16 (32%), 38 (76%), 24 (48%) and 33 (66%) patients, respectively. Eight patients did not meet the interpretation criteria and two exhibited a dysfunctional breathing pattern (figure 2). Among patients with ventilatory limitation, cardiocirculatory limitation, limitations suggestive of PH and impaired gas exchange, the peak V′_O2 (% predicted) was 58% (36–72), 60% (37–77), 58.5% (36–78) and 60% (36–78), respectively, expressed as median (range). In the univariable analysis, variables with p<0.1 included classification, 6-min walk distance, sex, FVC and the presence of PH. In the multivariate analysis, only FVC showed statistical significance on peak V′_O2 (supplementary e-Table 3).

FIGURE 2.

Venn diagram showing the distribution of limitation mechanisms during the cardiopulmonary exercise test in patients with hypersensitivity pneumonitis (n=51). CPET: cardiopulmonary exercise testing; PH: pulmonary hypertension.

Of the 90 patients evaluated, 61 patients had indication for RHC: 21 (34.5%) based on TTE, 16 (26%) based on D_LCO <40% and 24 (39.5%) based on both factors. Seven patients refused the procedure, four were lost to follow-up, three underwent lung transplantation and three died. Therefore, 44 patients underwent RHC in this study. Excluding the patients that had indication for RHC but did not perform (17 patients), the prevalence of PH was found to be 59% (43 out of 73 patients) (95% CI 48−72%). Table 3 presents the RHC and TTE variables. The mPAP was 28 (24–36.3) mmHg. 13 patients (30%) had a mPAP of between 20 and 25 mmHg. Eight patients (18.6%) had a PVR above 5 WU. Among those who underwent RHC, 43 patients (97%) had confirmed PH, 27 (62.7%) had a combined pattern (pre- and post-capillary), 11 (25.5%) had a pre-capillary pattern, four (9.3%) had a post-capillary pattern and one (2.5%) was nonclassifiable. Among the variables assessed during RHC, including mPAP, PAWP, PVR, cardiac index and mixed venous oxygen saturation, no differences were observed between the fHP and non-fHP groups (supplementary e-Table 4). PH was confirmed in 53% of patients with fHP and in 36% of non-fibrotic patients, with no statistically significant difference between groups (p=0.11). Among the predictors evaluated, P_ETCO2 showed the best performance for diagnosing PH (supplementary e-table 5). Patients with PH presented lower FVC, D_LCO and P_ETCO2 at AT, greater oxygen desaturation and a shorter distance walked in 6MWT than those without PH. There were no significant differences in peak V′_O2 between the groups (table 4).

TABLE 3.

Main variables of transthoracic echocardiography and right heart catheterisation (n=86)

Transthoracic echocardiography#	n=86
sPAP, mmHg	41±12
TRV, m·s−1	3.02±0.50
EF, %	64 (63–68)
Right heart catheterisation	n=44
RAP, mmHg	7±3
mPAP, mmHg	28 (24–36.3)
PAWP, mmHg	13.6±4.7
PVR, WU	3.5 (2.67–4.54)
Cardiac index	2.83±0.63
SvO2, %	70 (66–74)
Haemodynamic classification of PH	n=43
Pre-capillary	11 (25.5 %)
Combined post- and pre-capillary	27 (62.7 %)
Post-capillary	4 (9.3 %)
Nonclassifiable	1 (2.3 %)

Values are expressed as mean±sd, median (interquartile range) or n (%). sPAP: pulmonary artery systolic pressure; TRV: tricuspid regurgitation velocity; EF: ejection fraction; RAP: right atrial pressure; mPAP: mean pulmonary artery pressure; PAWP: pulmonary artery wedge pressure; PVR: pulmonary vascular resistance; WU: Wood units; S_vO2: mixed venous oxygen saturation; PH: pulmonary hypertension. ^#: four patients did not undergo transthoracic echocardiography.

TABLE 4.

Comparison of clinical, functional, echocardiographic, haemodynamic and exercise data between patients with and without pulmonary hypertension (PH)

	Overall population#	Non-PH	PH	p-value
Patients, n	73	30	43
Age, years	63 (61–74)	61 (55–67)	68 (61–74)	0.23
Female	47 (65%)	19 (65%)	28 (65%)	0.97
BMI, kg·m−2	27 (25–32)	27 (25–32)	28.5 (25–32)	0.72
Pulmonary function test
FVC, L	1.93 (1.2–2.3)	2.06 (1.59–2.67)	1.76 (1.21–2.07)	0.013
% predicted	58±20	64±19	53±20	0.023
TLC, L	3.01 (1.85–3.5)	3.33 (2.75–4.34)	2.97 (2.45–3.50)	0.047
% predicted	64±19	69±23	56±22	0.451
DLCO, mL·min−1·mmHg−1	7.97 (5.14–11.4)	10.3 (6.84–15)	7.27 (5.26–9.12)	0.008
% predicted	42.5±28	53±24	36±15	0.006
FVC/DLCO	1.46 (1.18–188)	1.19 (1.03–1.45)	1.52 (1.33–1.84)	0.054
Cardiopulmonary exercise testing	n=45	n=21	n=24
Peak V′O2, % predicted	58±12	60±14	56±11	0.338
Oxygen pulse, % predicted	67 (59–78)	67 (62–78)	66 (58–74)	0.507
PETCO2, at AT mmHg	31±4	33±4	30±3	0.008
V′E/V′CO2, slope	40±8	38±8	43±8	0.069
Minimum SpO2, %	86±7	90±7	84±7	0.007
6-min walk test
Distance, m	346±114	387±117	318±104	0.011
% predicted	70±11	76±10	65±11	0.186
Echocardiogram
sPAP, mmHg	40±12	34±12	43.8±12	0.727
TRV, m·s−1	3.00±0.51	2.75±0.55	3.13±0.45	0.045
EF, %	64 (63–68)	66 (64–68)	64 (62.5–68)	0.101
K-BILD score	57 (45–69)	59.5 (52–74.5)	57 (46.5–65.5)	0.314

Values are expressed as mean±sd, median (interquartile range) or n (%). BMI: body mass index; FVC: forced vital capacity; TLC: total lung capacity; D_LCO: diffusing capacity of the lung for carbon monoxide; V′_O2: oxygen uptake; P_ETCO2: end-tidal carbon dioxide tension; AT: anaerobic threshold; V′_E: minute ventilation; V′_CO2: carbon dioxide production; S_pO2: oxygen saturation; sPAP: pulmonary artery systolic pressure; TRV: tricuspid regurgitation velocity; EF: ejection fraction; K-BILD: King's Brief Interstitial Lung Disease; RHC: right heart catheterisation. ^# : seven patients with suspected PH refused RHC, four were lost to follow-up, three underwent lung transplantation and three died.

Peak V′_O2 (% predicted) was most strongly correlated with FVC (r=0.54, p<0.001) and D_LCO (r=0.40, p=0.007). The Borg dyspnoea score showed a significant inverse correlation with peak V′_O2 (% predicted) (r= −0.29, p=0.039). sPAP was negatively correlated with FVC (% of predicted) (r= −0.288, p=0.031), D_LCO (% of predicted) (r= −0.30, p=0.039) and the distance in the 6MWT (r= −0.28, p=0.037).

Discussion

To the best of our knowledge, this is the first study to comprehensively assess exercise limitations and haemodynamic patterns in patients with both fHP and non-fHP. Exercise limitations were found to be multifactorial in most cases (88% of patients), particularly among those with fHP, with ventilatory and cardiocirculatory limitations, including those suggestive of PH, as the most frequently observed mechanisms. The prevalence of PH was high, presenting mainly with combined pre- and post-capillary profiles and was associated with greater functional impairment, shorter distance walked in the 6MWT and greater desaturation.

Fell et al. [37] conducted a study involving patients with idiopathic pulmonary fibrosis (IPF), reporting a mean peak V′_O2 of ∼14±4.2 mL·kg⁻¹·min⁻¹. In our study, the peak V′_O2 was 13.6±7.5  mL·kg⁻¹·min⁻¹, with lower FVC values and greater desaturation observed during the 6MWT. In another study, Faisal et al. [38] compared 16 patients with mild-to-moderate ILD to 16 sex- and age-matched patients with mild-to-moderate COPD and 16 healthy subjects. They found a peak V′_O2 of 68±17% predicted in patients with ILD, compared to 75±18% predicted in patients with COPD, in contrast to our findings, which showed a peak V′_O2 of 58±13%. Few studies have assessed patients with HP during exercise [12, 17]. In our study, aerobic capacity was reduced in almost all patients, with peak V′_O2 lower than that reported by Dias et al. [17] (58 versus 83% predicted) and similar to the findings of Oliveira et al. [12] in patients with HP and PH who underwent CPET (58 versus 68% predicted). It is important to note that, compared to these studies, our patients with HP showed lower FVC and D_LCO [12, 17], mainly those with fHP. In particular, patients in our study showed a significant reduction in peak V′_O2, which was also associated with variables of pulmonary function tests and a higher degree of dyspnoea.

In fact, various mechanisms can collectively contribute to exercise limitations in patients with ILD, as these conditions induce numerous changes in physiological and sensory responses to exercise. These patients often show reduced peak V′_O2, hyperventilation, decreased ventilatory reserve and hypoxaemia [13, 15, 38]. In patients with fHP, previously described mechanisms include ventilatory limitation with increased ventilatory inefficiency in those with PH [13, 17]. In this study, the mechanisms of exercise limitation were multifactorial in most cases, with most patients demonstrating at least two reasons for exercise impairment, including cardiocirculatory mechanisms and those indicative of PH in a significant proportion of cases, while ventilatory limitation was not the predominant factor.

The prevalence of PH was high, confirmed by RHC in 59% of the patients, and most of them (62.7%) showed a combined haemodynamic pattern. The prevalence of PH-ILD has increased in recent years, along with a better understanding of the PH phenotypes in chronic lung diseases, inflammation and vascular remodeling; increased patient life expectancy; chronic hypoxaemia; availability of new therapies; advances in diagnostic methods; and recent updates to PH diagnostic criteria [33, 39, 40–42]. Currently, PH can be diagnosed based on mPAP values >20 mmHg [33]. However, many published studies have not used this new criterion. A study involving 8991 patients with IPF found that the prevalence of PH was higher after the last definition (73.6%) compared to the previous version (47.6%). Among patients with IPF, 36.8% exhibited a pre-capillary phenotype. Similarly, the prevalence of PH associated with COPD increased from 52.4% to 82.4%, with a significant proportion having post-capillary PH (17.6%). In IPF, the prevalence was 7.3% [43], but little is known about the prevalence of post-capillary PH-ILD.

Teramachi et al. [44] reported a 20% prevalence of post-capillary PH-ILD, including IPF and non-IPF cases. In particular, one-third of these patients had cardiovascular disease and heart failure with preserved ejection fraction (HFpEF). Patients with post-capillary PH had significantly less dyspnoea, higher D_LCO, higher resting arterial oxygen tension, lower PVR and higher lowest oxygen saturation during the 6MWT than those with pre-capillary PH [44]. Oliveira et al. [45] investigated pulmonary vascular and right ventricular loads during exercise. Patients with PH-ILD at rest and during exercise exhibited increased right ventricular workload during exercise, reduced pulmonary vascular reserve and decreased maximal oxygen consumption. Although the main focus was pre-capillary PH, the findings highlight the complexity of the interaction between pulmonary and cardiac functions in these patients, suggesting that post-capillary PH may be underdiagnosed [45]. In another study involving 42 patients with ILD, Wasserman et al. [46] demonstrated that peak V′_O2 values were not strongly correlated with the degree of ventilatory impairment, but were closely associated with the severity of gas exchange and cardiocirculatory dysfunction. This finding reinforces the critical role of the cardiocirculatory component in exercise limitation among patients with ILD, as observed in our study [46]. Furthermore, particularly in fibrotic diseases, the transforming growth factor-β signalling pathway, which is considered a key mechanism in the fibrogenic process in ILD, also plays a pivotal role in the pathogenesis of myocardial fibrosis in heart failure with HFpEF [44].

Chronic inflammation and fibrosis can cause structural changes in the lungs and increased resistance in the pulmonary capillary bed. This results in elevated pressure in the left atrium and, subsequently, in the pulmonary circulation [43]. Over time, chronic PH can place additional strain on the left ventricle, exacerbating post-capillary hypertension. Despite this, the relatively high prevalence of patients with post-capillary PH raises concerns about the reliability and validity of PAWP measurements in individuals with advanced lung disease [40]. Recently, some authors have proposed precautions when measuring PAWP, as it can vary depending on the phase of the respiratory cycle. However, the uncertainty zone includes values between the limits of 12 and 18 mmHg for PAWP [47]. In this study, we assessed haemodynamic measurements at end-expiration, averaged over three respiratory cycles. Additionally, oxygen-dependent patients also underwent RHC with supplemental oxygen via nasal cannula to maintain adequate oxygen saturation above 92%.

Persistent exposure or unidentifiable causative antigen are factors associated with worse survival [48]. Previous articles from Brazil demonstrated that exposures can be identified in up to 95% of cases, as reported by Gimenez et al. [49]. In our study, antigens were identified in ∼98% of patients, likely reflecting the high environmental antigen burden in the population and the current diagnostic criteria, in which the presence of a relevant exposure combined with a typical high-resolution computed tomography pattern for HP is sufficient for diagnosis.

This study had some limitations that must be addressed. It was developed in a single reference centre in our country, which may explain the higher proportion of patients with more severe diseases. However, this probably reflects the characterisation of patients from Brazil as a whole because patients from all regions were referred to our centre. Another limitation is that most of the patients had fibrotic disease, and several (30 of 90, 33%) required supplemental oxygen, which precluded CPET in the system we used. Arterial blood analysis at peak exercise to assess the alveolar-arterial gradient and the assessment of dynamic hyperinflation were not performed. The simplified exercise test interpretation algorithm employed in this study (figure 2), while enhancing reproducibility through the use of well-defined categorical variables, does not fully capture the integrative and multifactorial nature of exercise physiology. Exercise responses involve complex interactions among cardiovascular, respiratory, muscular and metabolic systems, which may be underrepresented by simplified models. Furthermore, some patients refused RHC, which may have an impact on the reported prevalence of PH.

Conclusions

Reduction in exercise capacity is common in patients with HP, and the underlying mechanism is largely multifactorial, mainly involving cardiocirculatory limitations suggestive of PH, which is frequent, presenting combined pre- and post-capillary patterns as the most prevalent profile, particularly in the fibrotic phenotype, and is associated with worse lung function, oxygen desaturation and exercise capacity. These findings highlight the need for a more comprehensive approach to the evaluation of symptomatic HP patients.

Supplementary material

Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.

Supplementary material

00656-2025.SUPPLEMENT