Article Text
PDF
Abstract
Background Fabry disease (FD) causes multiorgan sphingolipid accumulation, with cardiac involvement responsible for the largest burden of morbidity and mortality. Exercise intolerance in FD is prevalent, yet the mechanisms of this are poorly understood. The aim of this study was to assess exercise intolerance in FD and identify whether this correlates with the phase of cardiomyopathy.
Methods This was a retrospective observational study of adults with FD undergoing cardiopulmonary exercise testing (CPEX) between September 2011 and September 2023 at a national referral centre in the UK. The primary outcome measure was peak oxygen uptake (V̇O2peak), with forced expiratory volume in 1 s (FEV1) used to quantify respiratory impairment. Age-normalised/sex-normalised values were additionally calculated, based on published normal ranges for subgroups of age and sex. The cardiomyopathy phase was classified on a 4-point scale by two FD experts using contemporaneous imaging and biochemistry results.
Results CPEX was completed by 42 patients, with a median age of 54 years and of whom 62% were male. Patients were approximately equally distributed across the four cardiomyopathy phases. At phase I, the mean (±SD) V̇O2peak was 28.7±7.7 mL/kg/min, which represented a significant underperformance of 23%, relative to age-normalised and sex-normalised values (expected mean: 37.3±3.2 mL/kg/min, p=0.006). V̇O2peak declined significantly across the cardiomyopathy phases (p=0.010), reaching a mean of 21.2±6.1 mL/kg/min at phase IV. Normalised FEV1 was not found to show a corresponding significant change with cardiomyopathy phase (p=0.683). Impaired left atrial global longitudinal strain as well as biochemical markers of inflammation were associated with impaired V̇O2peak.
Conclusions This study identifies significantly impaired aerobic capacity in FD, even in those without phenotypic cardiomyopathy. No corresponding changes in respiratory impairment were observed, suggesting that exercise intolerance may be due to early cardiac sphingolipid accumulation and subsequent atrial and ventricular dysfunction, which increases as cardiomyopathy progresses. As such, peak V̇O2peak holds promise as a therapeutic marker of response to FD-specific therapy.
- cardiomyopathy
- cardiomyopathy, restrictive
Data availability statement
Data are available on reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Exercise intolerance in Fabry disease is common and begins early in the disease course.
Given the multisystemic nature of the disease, it is important to understand mechanisms underpinning exercise intolerance.
WHAT THIS STUDY ADDS
This study provides insights into the aetiology of exercise tolerance in Fabry disease.
Exercise intolerance occurs in all cardiac phases of Fabry disease, including in those without phenotypic cardiomyopathy, and worsens as cardiomyopathy progresses.
This study also highlights impaired atrial function as a potential contributor to exercise intolerance.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Exercise intolerance in Fabry disease is common, begins early and appears to be cardiac in aetiology, due to early myocardial sphingolipid deposition.
Peak oxygen uptake measured via cardiopulmonary exercise testing may be a promising marker of response to therapy in patients with Fabry disease.
Introduction
Fabry disease (FD) is an X linked lysosomal storage disorder in which deficiency of the enzyme α-galactosidase A1 leads to progressive multiorgan accumulation of sphingolipid. The resultant phenotype, which involves cardiac, renal and cerebrovascular disease, is associated with premature morbidity and mortality.2 Treatment includes enzyme replacement therapy (ERT) and oral chaperone therapy (OCT), and there is promise from both substrate reduction and gene therapy. Therapy is most effective when delivered early in the course of disease.3 However, current guidelines recommend initiation of therapy only when left ventricular hypertrophy (LVH) or late gadolinium enhancement (LGE) has been identified on cardiac magnetic resonance imaging (CMR).4 Imaging with CMR has given rise to the concept of four phases of FD cardiomyopathy, beginning with an early accumulation phase during which there is early wall thickening, impaired global longitudinal strain (GLS) and biomarker release. In the later disease phases, in which therapy is probably ineffective, there is the development of hypertrophy and fibrosis with persistent troponin release, reflecting myocardial inflammation and injury, with an increasing risk of heart failure and sudden cardiac death.5
We hypothesised that impaired aerobic capacity is detectable in the early accumulation phase in FD and becomes progressively more severe during the later phases of inflammation, hypertrophy and fibrosis.5 We have used routinely collected data from our specialist FD clinic to examine the relationship between exercise tolerance measured by peak oxygen uptake (V̇O2peak) and the severity of cardiomyopathy in patients with established FD.
Methods
Study population
This was an observational retrospective study of adults with genetically proven FD attending a ‘one-stop’ clinic in the Centre for Rare Disease at the Queen Elizabeth Hospital, Birmingham, UK (QEHB). Patients attending this clinic underwent routine assessments including blood sampling, transthoracic echocardiography (TTE) and CMR. Patients reporting effort intolerance and fatigue were routinely referred for exercise testing using cardiopulmonary exercise testing (CPEX). In some cases, the 6 min walk test (6MWT) was used as an alternative test, either as an initial screening assessment since it could be performed easily in clinic without requiring equipment, or in patients not able to undergo CPEX.
Patients were included if they underwent CPEX between September 2011 (introduction of the ‘one-stop’ clinic at QEHB) and September 2023. Where patients completed multiple CPEXs during the study period, only the most recent tests were included. CPEXs were additionally excluded where these occurred >2 years before or after a CMR or TTE that could be used to classify the cardiomyopathy phase. Data relating to patient demographics, comorbidities and FD-specific therapy history at the time of the assessment of the cardiomyopathy phase were then extracted from the electronic health record.
Cardiopulmonary exercise testing
CPEX was performed using a treadmill (h/p cosmos quasar) according to individualised ramp protocols, which aimed to bring patients to peak capacity after between 8 and 10 min. Patients received encouragement to exercise to peak capacity, with a ‘best effort’ CPEX being defined as achieving a respiratory exchange ratio (RER) >1.0.6–8 Effort was also quantified using rating of perceived exertion (RPE; on a 6–20 scale),9 peak blood pressure and heart rate, which was additionally age-normalised by calculating the percentage of the expected peak heart rate, defined as 220 minus age (in years).10
Performance on the CPEX was primarily quantified by the V̇O2peak using a breath-by-breath analyser (Ultima CardiO2), which was calibrated according to the manufacturer’s instructions prior to use. The breath-by-breath data were averaged using a moving average algorithm whereby an average of five breaths are taken, excluding the highest and lowest. To account for the effects of age and sex, normalised values were additionally assessed, which were calculated by dividing the observed value by the expected value for each patient to produce a percentage. The expected V̇O2peak values for each patient were calculated using equations 5–8 reported by Graves et al.11 These represented regression models, derived from data published by the Cooper Institute,12 which were evaluated to estimate the 50th percentile of V̇O2peak for a given age and sex.
Secondary CPEX outcomes assessed were the V̇O2 at the ventilatory threshold (VT1, defined as the lactate threshold); peak end-tidal CO2 pressure (PET CO2); peak O2 pulse and ventilation/CO2 production (VE/VCO2) slope. VT1 was determined manually using the V-slope method; when there was uncertainty, the ventilatory equivalent method was used to confirm VT1.13 Prior to CPEX, patients also underwent a spirometry assessment following a published protocol14 to assess lung function, from which the forced expiratory volume in 1 s (FEV1) was recorded. Age-normalised/sex-normalised/height-normalised FEV1 values were produced using the expected values from the ‘race-neutral’ reference equation described by Bowerman et al.15
CMR and TTE imaging
Contrast-enhanced CMR (1.5 T Avanto, Siemens Healthcare, Erlangen, Germany) was performed in line with standard protocols.7 TTE data were collected by an accredited sonographer (AMA) using the iE33 and EPIC ultrasound systems (Philips), according to the British Society of Echocardiography minimum dataset.12 Data were extracted for the CMR and TTE performed closest to the date of the CPEX assessment used in the analysis, where available.
Classifying cardiomyopathy phase
Two FD experts (RPS and AR) independently reviewed the blood biochemistry, ECGs, CMR and TTE imaging of all patients to classify the cardiomyopathy phase on a four-category ordinal scale. Both FD experts were blinded to patient identifiers and to results of exercise testing when classifying patients. The criteria used as a guideline for these classifications are visualised in figure 1; however, the experts were permitted to deviate from these for borderline cases, or for patients who had missing data for some of the criteria. Discrepancies between the experts were then resolved on consensus (n=4); these were predominantly patients with incomplete data, such as those who did not undergo CMR due to having an existing cardiac device or claustrophobia.
Phases of fabry cardiomyopathy. *Abnormalities in biochemistry markers are not always observed at cardiomyopathy phase II, with borderline increases often being observed. †Thresholds for abnormality were 14 ng/L for troponin-T and 16 ng/L for troponin-I. CMR, cardiac magnetic resonance imaging; EF, ejection fraction; GLS, global longitudinal strain; HS, high sensitivity; LGE, late gadolinium enhancement; LV(H), left ventricular (hypertrophy); NT-proBNP, N-terminal pro B-type natriuretic peptide; TTE, transthoracic echocardiography.
Haematology and biochemistry
Data for the haematology and biochemistry tests performed closest to the cardiomyopathy phase classification were additionally extracted. The laboratory transitioned from assessing high-sensitivity (HS) troponin-I to troponin-T (ELISA, Roche Diagnostics) during the study period; hence, these were treated as separate markers for analysis. N-terminal pro-B type natriuretic peptide (NT-proBNP) was measured by sandwich immunoassay with magnetic particle separation and chemiluminescent detection on an E170 analyser (Roche Diagnostics, Burgess Hill, United Kingdom). The laboratory truncated NT-proBNP levels at a lower limit of 5 ng/L and C reactive protein (CRP) levels at a lower limit of 3 mg/L; values below these thresholds were assigned values of 4 ng/L and 2 mg/L, respectively, for analysis.
Patient and public involvement
Patients were consulted and involved in the research question and design for this study.
Statistical methods
Comparisons between observed and expected V̇O2peak values were performed using paired t-tests. Patient characteristics were compared across the four cardiomyopathy phases using Jonckheere-Terpstra tests for continuous variables, to account for the fact that the cardiomyopathy phases were ordinal. For binary variables, associations with cardiomyopathy phases were assessed using Mann-Whitney U tests (ie, comparing the ‘average’ cardiomyopathy phase between categories). The associations between cardiomyopathy phase and both V̇O2peak and FEV1 were additionally assessed using linear regression models, which treated the cardiomyopathy phase as a continuous covariate, to estimate the change in V̇O2peak or FEV1 associated with an increase of one cardiomyopathy phase. To further interrogate the association between cardiomyopathy and exercise tolerance, associations between V̇O2peak and both biochemistry and CMR/TTE parameters were assessed using Spearman’s rank correlation coefficients (rho).
Continuous variables are reported as ‘mean±SD’ when approximately normally distributed, or as ‘median (IQR)’ otherwise. Cases with missing data were excluded from the analysis of the affected variable. All analyses were performed using IBM SPSS V.29 (IBM, Armonk, New York, USA).Continuous variables are reported as ‘mean±SD’ where approximately normally distributed, or as ‘median (IQR)’ otherwise
Results
Study cohort
Of the 157 patients followed up at the FD clinic during the study period, six were transferred to another hospital for follow-up after their initial assessment and 23 did not report symptoms of effort intolerance or fatigue during the follow-up period. The 128 patients with these symptoms most commonly reported breathlessness and fatigue; 30 were deemed too unfit to complete exercise testing (due to pain or mobility issues); five were referred for testing, but did not attend; and 39 only completed a 6MWT. Of the remainder, 54 completed a CPEX test, of whom 42 performed with best effort (ie, RER >1) and were included in the analysis (figure 2).
Study flow chart. 6MWT, 6 min walk test; CPEX, cardiopulmonary exercise testing; FD, Fabry disease; RER, respiratory exchange ratio.
Cohort characteristics
The cohort had a median age of 54 (IQR 39–62) years, and 62% were male (table 1). Patients were approximately equally distributed across cardiomyopathy phase I (19%; n=8), phase II (26%; n=11), phase III (21%; n=9) and phase IV (33%; n=14). As would be expected, comparisons between cardiomyopathy phases found significant trends in age and sex distributions, with the median age increasing from 35 to 62 years (p<0.001) and male proportion from 25% to 86% (p=0.005) between cardiomyopathy phases I and IV. Increasing cardiomyopathy phase was also associated with higher rates of chronic kidney disease (p=0.007), and of treatment with statins (p=0.006) and beta-blockers (p=0.029). Of the biochemistry markers, creatinine (p=0.003), HS troponin-I (p<0.001), NT-proBNP (p<0.001) and haemoglobin (p<0.001) were all found to increase significantly with cardiomyopathy phase. Significant changes in CMR and TTE parameters were also observed (table 2).
Cohort characteristics by cardiomyopathy phase
Cohort characteristics by cardiomyopathy phase
CPEX outcomes by cardiomyopathy phase
Patients achieved a mean RER of 1.15±0.08, which was found to be similar across the four cardiomyopathy phases (p=0.701, table 3). Of the other markers of effort considered, only the peak heart rate was found to differ significantly with cardiomyopathy phase, declining from a mean of 166±23 bpm in phase I to 135±36 bpm in phase IV (p=0.006). However, this was largely a reflection of the increasing age across the cardiomyopathy phases, with the age-normalised peak heart rate not significantly different across the four phases (p=0.294).
CPEX outcomes by cardiomyopathy phase
The mean V̇O2peak for the cohort was 24.8±9.0 mL/kg/min, and decreased significantly with cardiomyopathy phase, from 28.7±7.7 mL/kg/min in phase I to 21.2±6.1 mL/kg/min in phase IV (p=0.010), with regression modelling estimating a reduction of 3.3 mL/kg/min (95% CI 1.0 to 5.6) per phase (figure 3A). However, significant trends in age and male sex prevalence with cardiomyopathy phases likely acted as considerable confounders in this analysis. As such, normative V̇O2peak values were calculated for each patient, to account for the potential effects of these factors. Patients in all four cardiomyopathy phases were found to significantly underperform, relative to these expected values (all p<0.05). For example, in cardiomyopathy phase I, while the mean expected V̇O2peak was 37.3±3.2 mL/kg/min, patients achieved a mean V̇O2peak of only 28.7±7.7 mL/kg/min, hence underperforming by an average of 23% (±17, p=0.006). The magnitude of this underperformance increased significantly across the cardiomyopathy phases (p=0.018), with regression modelling estimating a change of 7 percentage points (95% CI 2 to 13) per phase (figure 3B).
Association between cardiomyopathy phase and V̇O2peak/FEV1 grey points represent the values for individual patients and are plotted with jitter. Diamonds represent the mean values within each cardiomyopathy phase subgroup, with whiskers representing 95% CIs. Broken lines are from linear regression models, with the cardiomyopathy phase treated as a continuous covariate. Normalisation of V̇O2peak and FEV1 was performed as described in table 3, with values reported as a percentage of the expected value for a patient’s age and sex (and height for FEV1). 6MWT: 6 min walk test; CPEX, cardiopulmonary exercise testing; FEV1, forced expiratory volume in 1 s; RER, respiratory exchange ratio; V̇O2peak, peak oxygen uptake.
Of the secondary CPEX outcomes considered, only the peak PET CO2 was found to vary significantly with cardiomyopathy phase, declining from a mean of 37.7±4.1 mm Hg to 34.8±6.9 mm Hg between phases I and IV (p=0.047, table 3). The V̇O2 at VT1 showed a downwards trend with cardiomyopathy phase, but this did not reach statistical significance (p=0.063). No significant trends were observed for the peak O2 pulse (p=0.718) or the VE/VCO2 slope (p=0.215).
Association between disease markers and V̇O2peak
Of the biochemistry markers assessed, elevated HS troponin-I, NT-proBNP, CRP, white cell count, neutrophil count and the neutrophil-lymphocyte ratio were all associated with significant lowering of age-normalised/sex-normalised V̇O2peak (table 4). Of the CMR/TTE parameters, increasing TR velocity and LA end-diastolic volume (EDV) were significantly associated with reductions in the age-normalised/sex-normalised V̇O2peak, while improvements in LA GLS, LA ejection fraction (EF) and fractional area change (FAC) were associated with significantly higher age-normalised/sex-normalised V̇O2peak. Analysis of the observed V̇O2peak (ie, without age-normalisation/sex-normalisation) returned similar results.
Correlations between V̇O2peak and biochemistry/imaging parameters
FEV1 by cardiomyopathy phase
The mean FEV1 for the cohort was 2.61±1.00 L, which was significantly lower than the mean expected value (3.18±0.71 L, p<0.001), yielding a mean underperformance of 18% (± 25, table 3). FEV1 was found to be correlated with V̇O2peak (rho: 0.468, p=0.002, table 4). However, unlike V̇O2peak, normalised FEV1 values were not found to differ significantly with cardiomyopathy phase (p=0.683), with mean underperformance of 10% vs 17% for phase I versus phase IV (table 3 and figure 3C,D).
Discussion
This study has demonstrated impaired peak aerobic exercise capacity in patients with FD, the magnitude of which increased progressively with cardiomyopathy severity. The results provide evidence of impaired aerobic capacity in patients with gene-positive FD but no detectable cardiomyopathy on advanced imaging techniques, including deformation echocardiography and CMR (ie, the cardiomyopathy phase I subgroup). These results are consistent with cellular data showing that cardiac sphingolipid accumulation begins early in FD, and raise the prospect of clinical benefit of commencing therapy in early phase disease.16 Breathlessness,17 exhaustion and fatigue18 are common in FD but, due to the multisystem nature of FD, it can be a challenge to link symptoms to a specific mechanism. The early impairments in V̇O2peak exhibited in this study suggest that these symptoms may, in part, be explained by myocardial disease, which may not be detectable using conventional imaging. To put the change in exercise capacity into context, the normative ranges reported11 show the expected V̇O2peak decline to be approximately 2.5 mL/kg/min per decade in the general population. In this study, the estimated reduction in age-normalised/sex-normalised V̇O2peak was 7 percentage points per cardiomyopathy phase, equivalent to approximately 2.1 mL/kg/min per phase. As such, the reduction in exercise tolerance resulting from a progression of one cardiomyopathy phase is equivalent to almost a decade of ageing.
Chronotropic incompetence is well documented in FD and its contribution to exercise capacity impairment has been described.19 However, we demonstrate no significant change in age-normalised peak heart rate with cardiomyopathy phase, suggesting alternative mechanisms. We demonstrate an association between atrial structural and functional parameters, specifically increased LA EDV and impaired LA GLS, LA EF and LA FAC with V̇O2peak impairment. Impaired LA function with increased stiffness has been associated with impaired exercise tolerance in adults with heart failure with preserved EF.20 However, this is a novel finding in FD, highlighting atrial myopathy as an important component of exercise intolerance and cardiomyopathy pathogenesis in FD.16
Given FD is multisystemic,21 there is likely to be an element of physical deconditioning, pulmonary involvement and skeletal muscle dysfunction that contributes to impaired exercise capacity in patients with FD. While not measured in this study, endothelial dysfunction or abnormal vascular remodelling in patients with FD, as a result of increase oxidative stress or systemic inflammation, may also impact oxygen delivery and exercise intolerance. Interestingly, our findings demonstrate higher immune cell counts and CRP were associated with lower age-normalised/sex-normalised V̇O2peak. A high neutrophil-lymphocyte ratio is an established marker of ongoing systemic inflammation. These findings are of relevance given the impression of cardiac involvement in FD reflecting an inflammatory cardiomyopathy, evidenced by troponin elevation, and T2 elevation and T1 pseudonormalisation on CMR. The exact cause of exercise intolerance in FD remains to be determined.16 However, since worsening of cardiac-specific biomarkers and impairment of atrial function were associated with a reduction in normalised V̇O2peak, our data suggest that the major limiting factor is myocardial dysfunction. This study offers the hope that formal exercise testing could be a target to identify patients with early phase myocardial involvement who might benefit from early therapy. Furthermore, exercise testing might allow a more sensitive measurement of treatment response in FD.
Our results extend the findings of previous smaller studies of exercise testing in FD. A previous study assessed exercise capacity and demonstrated a general decline in V̇O2peak, which improved with ERT.22 However, this study was of 15 patients and underlying mechanisms were not explored. A study of 29 patients demonstrated impairment in V̇O2peak, in FD compared with healthy controls.8 A positive correlation between aerobic capacity impairment and greater right ventricular volumes were also observed in patients with myocardial fibrosis without LVH.
Our data are consistent with the development of a known restrictive cardiomyopathy with advancing disease.5 23 Pseudonormalisation of T1 times24 were also observed with a lowering up to phase III and ‘pseudonormalisation’ in phase IV. Elevation of HS troponin and NT-proBNP were also in keeping with published literature,25 highlighting the progressive nature of FD cardiomyopathy, with males generally exhibiting a more severe and earlier-onset phenotype.17 14% of those in phase IV were female, demonstrating the relevance of X linked inactivation, highlighting the need for equally close monitoring in males and females.26
Strengths and limitations
The primary strength of this study was the large sample size of deeply phenotyped patients incorporating exercise, imaging and biochemical data. However, there are limitations that need to be considered. While the sample size is one of the largest CPEX studies in FD, it is small compared with non-FD studies; hence, only relatively large effect sizes were detectable. CPEX testing was also not available in all patients, which may have introduced selection bias. Cardiomyopathy phase also strongly correlated with both age and sex, with both factors being predictive of exercise test performance. To adjust for these, normalised values of V̇O2peak and FEV1 were calculated, based on published normative ranges for healthy control cohorts. However, we cannot exclude a degree of residual confounding by age and sex. Finally, resting FEV1 measurements may not correlate with peak exercise responses. Future studies should incorporate postexercise pulmonary testing to account for this.
Conclusion
We demonstrate significant impairments in aerobic capacity in FD across all phases of cardiomyopathy, which appear to be primarily due to cardiac limitation. V̇O₂peak was significantly impaired, even in patients without phenotypic evidence of cardiomyopathy, respiratory limitation, or biochemical imbalance, suggesting that cardiac sphingolipid accumulation occurs early in the course of FD, leading to fatigue and exercise intoleranceWe demonstrate significant impairments in aerobic capacity in FD across all phases of cardiomyopathy, which appear to be primarily due to cardiac limitation. V̇O2peak was significantly impaired, even in patients without phenotypic evidence of cardiomyopathy, respiratory limitation or biochemical imbalance, suggesting that cardiac sphingolipid accumulation takes place early in the course of FD, leading to fatigue and exercise intolerance.
Data availability statement
Data are available on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study was approved by West Midlands South Birmingham Research Ethics Committee (23/WM/0180 IRAS 325613); was registered locally at QEHB (CARMS-18469) and was conducted in accordance with local legislation and institutional requirements. The requirement for written informed consent from participants was waived since data were acquired from a research database using routinely collected clinical data with the secondary purpose of research.
Acknowledgments
We would like to acknowledge the support and work of the Department of Echocardiography and Cardiac Magnetic Resonance Imaging at the Queen Elizabeth Hospital, Birmingham, UK in facilitating this study. We would also like to thank the Cardiology Department for conducting the cardiopulmonary exercise testing.
References
Footnotes
X @RichardSteeds
Contributors RPS, TG, PFC, TEN, JT: idea and conceptualisation. AR, JH, SET, CO'S, JvV, RV: planning, data acquisition, analysis, interpretation and writing up of manuscript. NC, AMA, JvV: data acquisition. RPS, TEN, JT: review and editing of manuscript. All authors read and approved the final manuscript. RPS is the guarantor.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.