• transthyretin cardiac amyloidosis
• cardiopulmonary exercise test
• stress echocardiography
• functional capacity

Aims
Transthyretin cardiac amyloidosis (ATTR-CA) is characterized by progressive left ventricular diastolic and systolic dysfunction associated with the autonomic nervous system and ventilatory impairment, responsible for reduced effort tolerance. Cardiopulmonary exercise testing (CPET) is the gold standard to assess exercise capacity. Its combination with exercise echocardiography (eCPET) allows us to understand the mechanisms of exercise limitation in detail. This study aims to define the primary determinants of functional impairment in ATTR-CA patients and the main predictors of peak VO2, investigating possible peculiarities and differences as compared to a sex and age-matched population of patients with heart failure with preserved ejection fraction (HFpEF) and hypertrophic phenotype. In this paper, a preliminary analysis of 24 patients is presented.

Methods and results
Comprehensive assessment, including eCPET, was performed in 15 ATTR-CA patients and 9 HFpEF patients with suspected ATTR-CA, in whom updated diagnostic investigations did not confirm the diagnosis. A significantly worse diastolic dysfunction was observed in the study group, as proven by lower median values of peak lateral e' and medial e' (11 [9.5, 13] cm/s vs. 18 [11, 19] cm/s, p 0.033 and 10 [7, 11.5] cm/s vs 16 [11, 18] cm/s, p 0.017 respectively) and higher median values of peak E/e' (10.9 [9.4, 13.8] vs 7.4 [5.9, 10.5], p 0.043). There were no statistically significant differences in most of the variables analyzed, including peak VO2, cardiac output (CO) reserve, chronotropic reserve, ventilatory reserve, ventilatory efficiency (measured through VE/VCO2 slope), maximal peripheral oxygen extraction (C(a-v)O2) and right ventricle – pulmonary artery coupling (expressed as TAPSE/PASP ratio). ATTR-CA patients showed a lower increase in stroke volume from rest to peak effort (∆SV: 29 [19, 36.5]% vs. 37 [33, 45]%, p 0.073), lower ∆CO (112.12 [94.8, 148.86]% vs. 130 [108.57, 157.41]% p 0.591) and lower ∆TAPSE (22.22 [18.82, 31.22]% vs. 32.14 [15, 44.44]% p 0.591), although the pre-specified level of significance was not reached. The correlation between peak VO2 and ΔCO was stronger in ATTR-CA patients than in controls (ρs = 0.796, p 0.000384 vs ρs = 0.215, p 0.578). ATTR-CA patients with peak VO2 values below the median value of 12.3 ml/min/kg had lower ΔCO as compared to the corresponding matched population (82.93 [71.95, 106.99]% vs 108.57 [68, 150]%) and higher ∆C(a-v)O2 (101.58 [83.75, 186.97]% vs 78.11 [68.34, 98.7]%), despite this differences were not statistically significant, likely for the limited sample size. Among patients with peak VO2 ≥12.3 ml/min/kg, the study group showed significantly lower peak lateral e' (12.5 [11, 14.25] cm/s vs. 19 [19, 20] cm/s, p 0.045), lower peak medial e' (10.5 [9.75, 12] cm/s vs 18 [17.5, 18.5] cm/s, p 0.049) and higher peak E/e' (9.8 [9.1, 11.15] vs 5.65 [5.23, 5.9], p 0.050). History of AF and β-blocker therapy were the main clinical predictors of limited functional capacity in ATTR-CA patients. Median values of peak VO2 were 9.35 [8.6, 10.7] ml/min/kg in patients with a history of AF vs 18.6 [11.95, 19.15] ml/min/kg in the patients without history of AF (p 0.014) and 10.7 [9.48, 12.55] ml/min/kg in patients on β-blockers vs 18.6 [13, 18.95] ml/min/kg in patients not receiving β-blockers (p 0.082). Finally, a significant reverse correlation between circulating levels of TnHS at baseline and peak VO2 was observed in ATTR-CA patients (ρs = -0.649, p 0.0163).

Conclusions
In this study, the functional limitation of ATTR-CA patients seemed mainly driven by lower CO during effort rather than lower C(a-v)O2, as compared to controls. A worse diastolic function during exercise and lower indexes of biventricular systolic function, such as SV and TAPSE, were also found, although the latter was without statistical significance. Among possible predictors of exercise limitation, we observed a detrimental effect of a history of AF and β-blockers and a significant correlation between circulatory levels of TnHS and peak VO2, suggesting biomarkers' potential role in predicting functional capacity in addition to their well-known prognostication use. Further studies with larger sample sizes and more adjustments for confounding factors are needed to achieve greater statistical power.