Role of exercise stress echocardiography in systemic sclerosis: pathophysiological and prognostic significance of the systemic sclerosis with a heart failure and preserved ejection fraction phenotype

Abstract

Aims

Left ventricular (LV) diastolic dysfunction and heart failure with preserved ejection fraction (HFpEF) are common cardiac complications of patients with systemic sclerosis (SSc). Exercise stress echocardiography is often used in symptomatic patients with SSc to detect abnormal increases in pulmonary pressures during exercise, but the pathophysiologic and prognostic significance of exercise stress echocardiography to assess the presence of HFpEF in these patients is unclear.

Methods and results

Patients with SSc (n = 140) underwent ergometry exercise stress echocardiography with simultaneous expired gas analysis. The HFA-PEFF score ≥ 5 points was used to diagnose HFpEF. Thirty-five patients met the HFpEF criteria (prevalence 25%). Compared with patients with SSc-non-HFpEF, those with SSc-HFpEF were older and had a higher prevalence of coronary artery disease, more severe LV diastolic dysfunction (by definition), depressed right ventricular systolic function, reduced exercise capacity (lower peak oxygen consumption), and poorer ventilatory efficiency. Exercise right heart catheterization was performed in 25 patients and it confirmed elevated pulmonary capillary wedge pressure during peak exercise in patients with SSc-HFpEF. Participants were followed up to assess the primary composite endpoint: all-cause mortality, HF hospitalization, unplanned hospital visits requiring intravenous diuretics, or oral diuretic intensification. Compared with SSc-non-HFpEF, SSc-HFpEF had a 5.3-fold increased risk of the composite outcomes (hazard ratio 5.29, confidence intervals 2.06–13.5, P = 0.0005).

Conclusion

In addition to pulmonary haemodynamics, exercise stress echocardiography may be useful to identify the HFpEF phenotype that has different pathophysiology and clinical outcomes in patients with SSc.

Graphical Abstract

Open in new tabDownload slide

Introduction

Pulmonary hypertension (PH) is a serious complication in patients with systemic sclerosis (SSc), with a relatively high frequency and increased morbidity and mortality.¹ Because of the progressive nature of the disease, the importance of identifying PH at an early stage has been increasingly recognized.^2–4 Pulmonary vascular resistance and pulmonary pressures can remain normal in the early stage but often become abnormal only during physiologic stress, such as exercise.^5,6 Being non-invasive, exercise stress echocardiography is often used in symptomatic patients with SSc to detect such abnormalities in pulmonary haemodynamics during exercise.^2,4,7–9

Exercise stress echocardiography also provides valuable information on left ventricular (LV) diastolic dysfunction and its reserve during exercise, namely the assessment of heart failure (HF) with preserved ejection fraction (HFpEF).^8,10–12 Given that LV diastolic dysfunction or HFpEF is common in patients with SSc,^13–17 the assessment of HFpEF using exercise stress echocardiography may provide pathophysiologic and prognostic implications. Accordingly, the present study performed exercise stress echocardiography with simultaneous expired gas analysis to characterize the prevalence, pathophysiology, and clinical outcomes of SSc patients with HFpEF compared with those without.

Methods

Study population

Consecutive patients with SSc who underwent exercise stress echocardiography for the evaluation of PH or dyspnoea in Gunma University Hospital, Maebashi, Japan between October 2019 and November 2023 were retrospectively identified. Dermatologists or rheumatologists made the diagnosis of SSc according to the 2013 classification criteria from the American College of Rheumatology/European League against Rheumatism.¹⁸ LV diastolic dysfunction and HFpEF are different entities. HFpEF is a clinical syndrome of HF and requires the presence of symptoms or signs of HF (dyspnoea, fatigue, or peripheral oedema) and objective evidence of increased LV filling pressures either at rest or with exertion. In the current study, the presence of elevated LV filling pressures (i.e. diagnosis of HFpEF) was made by the Heart Failure Association Pre-test assessment, Echocardiography, and natriuretic peptide, Functional testing, and Final aetiology (HFA-PEFF) algorithm steps 2–3 (see Supplementary data online, Figure S1).¹⁹ In brief, the HFA-PEFF score was calculated as the sum of echocardiographic functional [age-specific cut-offs for early diastolic mitral annular velocity (e’) velocity, early transmitral flow velocity (E)/e’ ratio, tricuspid regurgitation (TR) velocity, and longitudinal strain: maximum 2 points], morphological [rhythm-specific left atrial (LA) volume, relative wall thickness, and sex-specific measures of LV mass: maximum 2 points], and natriuretic peptide (maximum 2 points) domains (step 2). Subsequently, two or three points were added depending on the E/e’ ratio and TR velocity during exercise stress echocardiography (step 3). Symptomatic patients with the combined score from Steps 2 and 3 ≥ 5 points were classified as SSc-HFpEF. The original HFA-PEFF algorithm recommends exercise RHC to differentiate HFpEF in patients with equivocal data.¹⁹ In the present study, the results of exercise RHC were not used because our goal was to identify the HFpEF phenotype based solely on non-invasive echocardiographic assessment. Patients with total HFA-PEFF score <5 points were labeled as SSc-non-HFpEF.¹⁹ The H₂FPEF score was also calculated based on 6 clinical and echocardiographic variables.²⁰

We excluded patients with EF <50%, left-sided valvular heart disease (greater than moderate regurgitation, greater than mild stenosis), and infiltrative, restrictive, or hypertrophic cardiomyopathy. The study was approved by our institutional review board with the waiver of consent and was performed in accordance with the Declaration of Helsinki. All authors have read and agreed to the manuscript as written.

Exercise stress echocardiography

Transthoracic echocardiography was performed by experienced sonographers using a commercially available ultrasound system (Vivid E95; GE Healthcare, Horten, Norway). The EF was determined by method of disks in four-chamber views. Systolic mitral annular tissue velocity at the septal annulus (mitral s’) was also measured to assess LV systolic function. Septal E/e’ ratio was determined to estimate LV filling pressure. Stroke volume was calculated from the LV outflow dimension and pulse Doppler, and cardiac output (CO) was determined from the product of the heart rate and stroke volume (SV). Systolic tissue velocity at the lateral tricuspid annulus (TV s’) and tricuspid annular plane systolic excursion (TAPSE) were measured to assess right ventricular (RV) systolic function. The RV basal, mid-cavity, and longitudinal dimensions were measured at end-diastole with RV-focused views. Right atrial (RA) maximum volume was measured in the apical four-chamber views.²¹ The RA pressure (RAP) was estimated from the diameter of the inferior vena cava and its respiratory change at rest and during exercise, coded as 3 mmHg, 8 mmHg, and 15 mmHg.⁵ The estimated pulmonary artery systolic pressure (PASP) was calculated as 4 × (peak TR velocity)² + estimated RAP. Right ventricular-pulmonary artery coupling was assessed by the ratio of TAPSE to PASP.²² The mean pulmonary artery pressure (mPAP) was calculated as 0.61 × PASP + 2. The slope of mPAP to CO (mPAP/CO slope) was calculated as the changes of peak exercise and baseline estimated mPAP divided by the changes of peak and baseline CO.²³ The severity of pericardial effusion was semi-quantitatively assessed on the size of the echo-free space: none-trivial, small (<10 mm), moderate (10–20 mm), or large (>20 mm).²⁴ Lung ultrasound was performed to assess sonographic signs of lung congestion (ultrasound B-lines) using the same secta probe.^25,26 B-lines are hyperechoic lines that originate from the pleural line to the bottom of the ultrasound screen. The lung ultrasound was scanned at four intercostal spaces at rest and during peak exercise in the right hemithorax (right third and fourth spaces along the mid-axillary and mid-clavicular lines).²⁷ All Doppler measurements represent the mean of respiratory-averaged 2 beats in sinus rhythm and ≥ 3 beats in atrial fibrillation (AF).

All patients underwent semi-supine ergometry exercise echocardiography, starting at 20 watts (W) for 5 min, with 20 W increments in 3 min to participant-reported exhaustion (Angio imaging, Lode B.V., Groningen, the Netherlands), as previously reported.^23,27,28 We obtained echocardiographic images at baseline and during all stages of exercise. Expired gas analysis was performed simultaneously throughout the study to measure obtain breath-by-breath oxygen consumption (VO₂), carbon dioxide production (VCO₂), tidal volume, respiratory rate, and minute ventilation (V_E) (AE-100i, MINATO Medical Science, Osaka, Japan). The ventilatory efficiency was assessed by the V_E vs. VCO₂ slope.

Exercise right heart catheterization protocol

Exercise RHC was performed at the discretion of individual physicians based on clinical indications. A 8-Fr sheath was inserted through the right internal jugular vein in a subset of participants, as reported in our previous studies.^21,28,29 Transducers were zeroed at the midaxilla, as measured using calipers in all participants. Right atrial pressure, pulmonary artery pressures, and pulmonary artery wedge pressure (PAWP) were measured at end-expiration, and the mean measurement of three heartbeats was recorded. Pulmonary artery wedge pressure position was verified by typical waveforms and appearance on fluoroscopy. After assessing the resting haemodynamics, patients underwent the same supine cycle ergometry protocol as that in exercise stress echocardiography, wherein the exercise was started at a 20-W workload for 5 min and increased by 20-W increments in 3-min stages up to the patient's exhaustion point. Pressures were measured at all stages of the exercise. A 5-Fr radial arterial cannula was used to obtain arterial blood gases. Arterial-venous O₂ difference was measured directly as the difference between systemic arterial and PA oxygen content (=saturation × haemoglobin × 1.34).⁶ CO was calculated by the direct Fick method (CO = VO₂/arterial-venous O₂ difference) at baseline and peak exercise. SV was determined from the quotient of CO and heart rate. Pulmonary vascular function was assessed by pulmonary vascular resistance [PVR (mean PAP—PAWP)/CO] and PA compliance (PAC, SV/PA pulse pressure)].

Follow-up and outcome analysis

Patient follow-up was initiated on the day of exercise stress testing. The primary endpoint was a composite of all-cause mortality and worsening HF defined by hospitalization for HF, unplanned visits requiring intravenous diuretic treatment, or intensification of oral diuretics (started or increased for worsening HF after 1 month of exercise testing). Prognostic data were collected through medical records, death certificates, and phone calls to the patient or family members.

Statistical analysis

Data are reported as mean (standard deviation), median (interquartile range: IQR), or number (%) unless otherwise specified. Between-group differences were compared using unpaired t-test, Wilcoxon rank sum test, or χ² test, as appropriate. Correlations were assessed using Pearson's correlation coefficient. Logistic regression analysis was performed to assess association with the presence of HFpEF. Event rates were assessed using Kaplan–Meier curve analysis and univariable and multivariable Cox proportional hazards models were used to assess the independent prognostic power. All tests were two-sided, with statistical significance set at P < 0.05. All statistical analyses were performed using JMP 16.2.0 (SAS Institute, Cary, NC, USA).

Results

Clinical characteristics according to the HFA-PEFF algorithm

We identified 140 patients with SSc who underwent exercise stress echocardiography, of whom 35 (25%) met the criteria of SSc-HFpEF and 105 did not. Of the 35 patients with SSc-HFpEF, 24 were diagnosed based on the assessment at rest (HFA-PEFF score step 2 ≥ 5 points), and the remaining 11 were diagnosed with HFpEF after exercise stress echocardiography (HFA-PEFF score steps 2–3 ≥ 5 points).

Compared with SSc-non-HFpEF, patients with SSc-HFpEF were older and had a higher H₂FPEF score and a higher prevalence of systemic hypertension, coronary artery disease, and severe NYHA functional class (Table 1). Sex, body mass index, disease duration of SSc, and prevalence of limited cutaneous type and other comorbidities were similar between groups. Patients with SSc-HFpEF were treated with calcium channel blockers more frequently than SSc-non-HFpEF, but other medication use was similar between groups. By definition, patients with SSc-HFpEF had higher natriuretic peptide levels, lower renal function, greater LV mass index, and larger LA volume index than those with SSc-non-HFpEF, consistent with LV diastolic dysfunction. There were no differences in RV dimensions, RA volume, and the severity of pericardial effusion between groups. No differences were observed in pulmonary function test parameters.

At rest, heart rate, systolic blood pressure (BP), and oxygen saturation were similar between patients with and without SSc-HFpEF (Table 2). Compared with SSc-non-HFpEF, patients with SSc-HFpEF demonstrated lower mitral e’ velocity and higher E/e’ ratio whereas mitral E-wave, s’ velocity, CO, ePASP, TAPSE/PASP, TR severity, and the number of ultrasound B-lines did not differ between groups.

Cardiac and pulmonary response to exercise according to the HFA-PEFF algorithm

Compared with SSc-non-HFpEF, patients with SSc-HFpEF demonstrated poorer exercise capacity, evidenced by lower peak exercise intensity, shorter exercise duration, and lower peak VO₂ (Figure 1A–C, Table 2). These differences remained significant after adjusting for age (all P < 0.05). While systolic BP and oxygen saturation were similar between groups, patients with SSc-HFpEF demonstrated more severe chronotropic incompetence during exercise than those with SSc-non-HFpEF. During peak exercise, mitral e’ and s’ velocities were lower, E/e’ ratio and mPAP/CO slope were higher, and TAPSE, TV s’, and TAPSE/PASP were lower in patients with SSc-HFpEF than those in patients with SSc-non-HFpEF, with marginal differences in exercise ePASP, TR severity, and ultrasound B-lines. Simultaneous expired gas analysis showed poorer ventilatory efficiency (higher V_E vs. VCO₂ slope) and a trend toward lower tidal volume in patients with SSc-HFpEF than in those with SSc-non-HFpEF (Figure 1D).

We performed comparisons between SSc-HFpEF (n = 35) and age-matched SSc-non-HFpEF (n = 70). Key differences in exercise capacity and echocardiographic and expired gas data during peak exercise remained significant between the groups (see Supplementary data online, Table S3).

Exercise right heart catheterization

Ten (10%) of SSc-non-HFpEF and 15 (43%) of SSc-HFpEF underwent clinically indicated exercise RHC. Of the 10 patients with SSc-non-HFpEF, 3 met the diagnosis of pre-capillary PH (defined by mean PAP > 20 mmHg and PVR >2 WU). Consistent with results obtained in exercise stress echocardiography, SSc patients with HFpEF had higher natriuretic peptide levels and larger LV mass index and LA volume index (see Supplementary data online, Table S1). Table 3 shows comparisons of RHC findings at rest and during supine exercise between SSc-non-HFpEF (n = 10) and SSc-HFpEF (n = 15). At rest, there were no differences in central haemodynamics, pulmonary vascular function, and CO between groups. Of the 25 patients who underwent exercise RHC, the majority of patients (n = 20, 80%) developed exercise-induced PH defined by mPAP/CO slope >3.0 mmHg/L/min and 3 patients showed a negative value due to a paradoxical reduction in CO from rest to exercise (Figure 2). Of the 20 patients with exercise-induced PH, 13 (65%) were SSc-HFpEF and the remaining 7 were SSc-non-HFpEF. Patients with SSc-HFpEF demonstrated a marked increase in PAWP and a trend towards smaller increase in CO compared with those with SSc-non-HFpEF, with no difference in PASP elevation (Figures 3A–C). The PVR and PA compliance during exercise did not differ between groups. Peak E/e’ ratio was moderately correlated with invasively measured PAWP during exercise (r = 0.42, P = 0.04).

There were no significant differences between the patients who underwent exercise RHC (n = 25) and those who did not (n = 115), except for beta-blocker and calcium channel blocker use, NT-proBNP levels, and LV mass index (see Supplementary data online, Table S2).

Outcome analysis

Follow-up data were available in 130 patients. During a median follow-up time of 2.0 (interquartile range: 1.2–3.0) years, the composite outcome of all-cause mortality or worsening HF events occurred in 19 patients (14%), including 4 all-cause deaths, 2 HF hospitalization, 2 unplanned hospital visits requiring intravenous diuretics, and 11 oral diuretic intensification. Compared with patients with SSc-non-HFpEF, those with SSc-HFpEF had a 5.2-fold increased risk of the composite outcomes [hazard ratio (HR) 5.29, 95% confidence intervals (CI) 2.06–13.5, P = 0.0005, Figure 4]. This association remained significant after adjusting for either age or ePASP (adjusted HR 6.57, 95% CI 2.29–18.8, P = 0.0005 and adjusted HR 4.43, 95% CI 1.68–11.7, P = 0.003).

A composite outcome of all-cause mortality, HF hospitalization, and unplanned hospital visits requiring intravenous diuretics was also higher in patients with SSc-HFpEF compared with SSc-non-HFpEF (HR 6.97, 95% CI 2.06–23.5, P = 0.002).

Discussion

This study provides a detailed characterization of the prevalence, clinical characteristics, cardiac and pulmonary response to exercise, and clinical outcomes of patients with SSc and HFpEF, namely the SSc-HFpEF phenotype. After evaluation of exercise stress echocardiography, 25% of patients with SSc met the HFpEF diagnosis based on the HFA-PEFF algorithm. In addition to LV diastolic abnormalities at rest and during exercise (by definition), SSc patients with HFpEF were elderly and had more severe symptom burden and multiple comorbidities, consistent with typical clinical features of HFpEF. Compared with patients with SSc-non-HFpEF, those with SSc-HFpEF demonstrated reduced exercise capacity, and ventilatory inefficiency on simultaneous expired gas analysis. Exercise RHC confirmed elevations in invasively measured LV filling pressures during exercise, albeit in a subset of patients with SSc-HFpEF. Importantly, clinical outcomes were poorer in patients with SSc-HFpEF than in those with SSc-non-HFpEF. These data show that the SSc-HFpEF phenotype may be considered in the distinct entity that may require a specific treatment strategy. The pathophysiologic and prognostic significance of the HFpEF phenotype in SSc underscores the importance of assessing LV diastolic function in addition to PH on exercise stress echocardiography.

Identification of occult HFpEF in patients with SSc

Patients with SSc are at high risk of developing PH. A meta-analysis showed that PH occurs in approximately 9% of patients.¹⁶ Afflicted patients have increased mortality and hospitalization rates, with the poorest survival of any cause of PH (3-year survival: 52%).¹ Due to the progressive nature of the disease, early detection and diagnosis are critical in the management of SSc.^30,31 In the early stage of the disease, PA pressure is often normal at rest but develops abnormally only during exercise. Exercise stress echocardiography can identify this exercise-induced increase in PA pressure to assess the risk of PH.⁷ LV diastolic dysfunction or HFpEF is another important cardiac manifestation in SSc patients, observed in 20–45% of patients.^15,32 The presence of HFpEF is associated with poor exercise tolerance, haemodynamic abnormalities, and poor clinical outcomes.^23,33 Like PH, LV diastolic dysfunction or HFpEF may become apparent only during exercise.^10,23,28 The coexistence of HFpEF in SSc is noteworthy because it may worsen clinical outcomes and influence therapeutic strategy.³² Pulmonary vasodilators for SSc-related pre-capillary PH would have the potential to induce pulmonary congestion by increasing pulmonary blood flow to the non-compliant left heart, further emphasizing the importance of identifying HFpEF in practice.³⁴ Accordingly, we hypothesized that exercise echocardiography would be useful not only to detect exercise-induced PH but also to identify occult HFpEF in patients with SSc.

The current study examined consecutive patients with SSc who underwent exercise stress echocardiography for the evaluation of PH or dyspnoea. We found that 35 patients with SSc (25%) met the HFpEF diagnosis based on the HFA-PEFF algorithm, of whom 11 could not be diagnosed at rest and were identified by exercise stress echocardiography. While the prevalence of HFpEF in SSc patients depends on individual studies, an important finding in our study is that 30% of SSc-HFpEF could not be diagnosed without exercise stress echocardiography. This suggests that many patients with SSc-HFpEF will be missed in clinical practice if the diagnosis relies on resting assessments.

Although the definition of HFpEF using the HFA-PEFF algorithm in the current study resulted in more severe LV diastolic dysfunction in patients with SSc-HFpEF than in SSc-non-HFpEF, exercise RHC confirmed increases in invasively measured PAWP during exercise in the patients with SSc-HFpEF. This suggests the utility of the HFA-PEFF algorithm for non-invasive identification of HFpEF among patients with SSc. Importantly, patients with SSc-HFpEF demonstrated reduced exercise capacity, worse ventilatory efficiency, and poorer clinical outcomes compared with those with SSc-non-HFpEF. These data indicate that SSc-HFpEF is a unique phenotype of SSc that has a different pathophysiology and clinical course compared with garden-variety SSc. Despite a similar increase in PASP (+30 mmHg), exercise RHC showed a marked increase in PAWP during exercise in patients with SSc-HFpEF compared with SSc-non-HFpEF, suggesting that exercise PH was primarily driven by the post-capillary component. With the emerging evidence-based pharmacological treatment for HFpEF, such as sodium-glucose cotransporter 2 inhibitors, semaglutide, or angiotensin-neprilysin inhibitors,^35–37 further studies are warranted to test the efficacy of these drugs in SSc-HFpEF.

Clinical implications

The 2022 European Society of Cardiology (ESC) and the European Respiratory Society (ERS) guidelines for the management of PH recommend exercise stress echocardiography to aid the decision to perform RHC in symptomatic patients with SSc.⁷ The ESC/ERS guidelines also suggest the potential role of exercise RHC in identifying occult HFpEF in patients with PH.⁷ The current results show that exercise stress echocardiography is useful not only for assessing the risk of developing PH but also for identifying HFpEF, emphasizing its importance in the evaluation of SSc in clinical practice. Further studies are needed to determine how best to treat SSc-HFpEF.

Limitations

The present study had several limitations. This was a retrospective study conducted at a tertiary referral centre and all patients underwent exercise stress echocardiography, introducing referral and selection bias. The number of participants was modest. In particular, exercise RHC was performed on a subset of participants. Sodium-glucose co-transporter 2 (SGLT2) inhibitor use was low in our HFpEF population. This is likely due to the fact that the diagnosis of HFpEF was not made at the time of exercise echocardiography and that SGLT2 inhibitors were not available for HFpEF at the beginning of patient enrollment. Ultrasound B-lines are not specific to pulmonary congestion. B-lines observed in patients with SSc were more likely to be related to interstitial lung diseases than pulmonary congestion.²⁷ Because of the time constraints for imaging during exercise, lung ultrasound was performed at four intercostal spaces in the right hemithorax (right third and fourth spaces along the mid-axillary and mid-clavicular lines) rather than the full 28-region assessment. This might limit the detection of B-lines in patients with SSc, where interstitial lung disease often occurs from the lung bases.³⁸

Conclusion

Exercise stress echocardiography identified coexisting HFpEF in patients with SSc. The SSc-HFpEF phenotype demonstrated elevated LV filling pressures during exercise, poorer exercise capacity, worse ventilatory efficiency, and increased risk of adverse outcomes than patients with SSc without HFpEF, suggesting this phenotype is a distinct entity that may require specific treatment strategy. These data shed light on managing coexisting HFpEF in patients with SSc.

Supplementary data

Supplementary data are available at European Heart Journal - Cardiovascular Imaging online.

Acknowledgements

M.O. received research grants from the Fukuda Foundation for Medical Technology, Mochida Memorial Foundation for Medical and Pharmaceutical Research, Nippon Shinyaku, Takeda Science Foundation, Japanese Circulation Society, Japanese College of Cardiology, AMI Inc., Nippon Boehringer-Ingelheim, JSPS KAKENHI (21K16078), and AMED (23jm0210104h0002). T.H. received research grants from the Bayer Academic Support. H.I. received scholarship funds or Donations from Abbott Medical Japan, Boehringer Ingelheim Japan, Bristol-Myers Squibb Inc., and Pfizer Japan Inc. S.M. received research grants from Sun Pharma, Taiho Pharmaceutical, Novartis, Eli Lilly, and Maruho Co.

Funding

None.

Data availability

The data underlying this article cannot be shared publicly due to the privacy of individuals that participated in the study.

References

Author notes