This cross-sectional study included 123 consecutive IPF patients. The American Thoracic Society (ATS) and European Respiratory Society criteria were used for diagnosis [18]. These patients were diagnosed and underwent treatment for interstitial lung diseases at the referral center of a university tertiary care institution.
We included patients meeting the following criteria: SpO2 at rest and in ambient air ≥92%; no history of oxygen therapy; Gender-Age-Physiology Index score ≤5, which corresponds to mild or moderate disease [19]; and modified Medical Research Council (mMRC) dyspnea scale score ≤3 [20]. We excluded patients presenting with left cardiomyopathy, locomotor disease, another severe comorbidity (e.g., stroke, lung cancer, pulmonary thromboembolism), emphysema, or other pulmonary diseases related to IPF. All patients signed an informed consent form. The study was approved by the Ethics Committee of the University of Brasília.
Of the 123 patients evaluated, 38 were selected for baseline TTE acquired according to the protocol recommended by the American Society of Echocardiography [9]. Detailed diastolic function measurements were recorded. Five patients were excluded due to the presence of the following comorbidities: endocardial alterations caused by coronary disease (n = 2); dynamic left ventricular (LV) outflow tract obstruction (OTO; n = 1); and dilated cardiomyopathy with LV dysfunction, left atrial enlargement, and signs of increased LV filling pressure (n = 2). In addition to the five patients excluded following TTE, six refused to participate in the study. Thus, 27 patients completed the study and underwent CPT (Fig. 1).
Transthoracic Doppler echocardiography at rest
The examination was performed at rest using Vivid I equipment (GE Healthcare, Milwaukee, WI, USA) with a 2.5–3.5 MHz variable frequency transducer. Doppler analysis was performed in real time, and measurements were obtained from a mean of 3 to 5 beats. Conventional echocardiographic examination was performed to evaluate RV function based on the FAC% (abnormal, <35%), TAPSE (abnormal, <17 mm), and tricuspid annulus (S′) (abnormal, <9.5 cm/s) measured by TD [10]. The STE of the RV was considered normal when values were less than −20% [9].
The modified Bernoulli equation was applied to calculate the systolic pulmonary artery pressure (sPAP) from tricuspid regurgitation. A non-invasive estimate of the mean pulmonary artery pressure (mPAP) was calculated using the following formula: 0.61 × sPAP + 2 mmHg [21]. sPAP was assumed to be equal to the RV systolic pressure in the absence of RV OTO or pulmonary stenosis [22]. SPAP values <37 mmHg were considered normal [23].
After measuring the pressure gradient between the RV and right atrium, the estimated right atrial pressure, based on the inferior vena cava inspiratory collapsibility index, was added to the parameter [9, 24]. Images were obtained to optimize the best RV visualization position (i.e., apical four-chamber view, the same for RV diameter measurement), grayscale for STE analysis (50–90 frames/s), and sector depth for better resolution.
Two-dimensional echocardiography using the STE technique to evaluate RV function
Images were analyzed offline with Echo PAC software version 201; (GE Vingmed, Horten, Norway), and the STE technique was used to measure two-dimensional strain [25].
The endocardial region of the RV free wall was manually outlined at the end of the systole and automatically adjusted to include the entire myocardium. Systole in the cardiac cycle was defined using the event-timing feature of the software, which allowed the beginning and end of the RV systole to be delimited.
RV global longitudinal strain (GLS) was evaluated at the basal, medial, and apical portions of the free wall (Fig. 2), and the final result was expressed as the mean value of the three segments. GLS is defined as the percentage of myocardial thickening versus the original fiber length and is conventionally represented as a negative value [25].
Cycle ergometer CPT associated with dynamic TTE measurements
Using a cycle ergometer (CG-04; Inbramed®, Porto Alegre, RS, Brazil), all patients underwent an incremental ramp protocol with progressively increasing load. After 1-min rest to allow patients to adapt to the cycle ergometer, CPT was started at 60 rotations/min without load (0 W). Subsequently, the load was steadily increased at 5–30 W/min. Thus, the calculated value of the maximum predicted V’O2 (O2 consumed) and patient age, dyspnea level during exercise, and exercise capacity were used to provide an incremental period of 8–12 min, according to ATS recommendations [26]. Expiratory gas measurements were recorded from all patients at each breath for the analysis of ventilatory and cardiovascular variables (Quark PFT; COSMED, Rome, Italy).
The V’O2, V’CO2, V’E, and final expiratory O2 and CO2 pressures were recorded as means over 15 s. Continuous heart rate monitoring and electrocardiography were performed, and pulse oximetry (iPod®; Nonin Medical Inc., Plymouth, MN, USA) measurements were continuously recorded. Systolic and diastolic blood pressures were measured using the auscultatory method at each load increase. The anaerobic threshold was non-invasively estimated using the gas exchange method. The slope of V’O2 versus V’CO2 graph was analyzed on equal scales (V-slope technique) and confirmed using the ventilatory technique, which assessed the behavior of ventilatory equivalents (V’E/V’CO2 and V’E/V’O2) and their respective final expiratory pressures. Dyspnea and muscle fatigue sensations were also evaluated throughout the CPT using the Borg effort perception scale [27].
Echocardiographic images were acquired during CPT in the region of the ventricular apex along the left midclavicular line, which allowed for better tricuspid regurgitation image acquisition. During the test period, images were saved each minute for offline analysis to estimate the sPAP value throughout the test. During the recovery period, tricuspid regurgitation spectral traces were recorded to obtain measurements for 3 min after exercise.
Statistical analysis
Normally distributed continuous variables are expressed as mean ± standard deviation or as range. Categorical variables are expressed as percentages. Means were compared using t-test for independent samples. Results were considered statistically significant when p < 0.05. Data were analyzed using SPSS for Mac OS X© (version 25.0.0; SPSS Inc., Chicago, IL, USA).