The limitations of noninvasive CPET in the identification of cardiocirculatory causes of exercise intolerance Ventilatory and gas exchange variables (in addition to ECG, HR and systemic BP) are not sensitive for uncovering mild-to-moderate cardiocirculatory abnormalities on exercise [58]. For instance, flattening of the V′O2/WR relationship [48] might only be evident in patients with moderate-to-severe cardiac output impairment [50]. Even in these patients, it is not uncommon to observe a late-exercise rise in V′O 2 due to the increased metabolic cost of ventilation in morbidly obese subjects [87] and those with respiratory comorbidity [20]. Thus, a plateau in the V′O2/WR plateau might not be clearly discernible, particularly in a short test. Interpretation of the O2 pulse has also become more complex in the past decade because of the growing prevalence of patients whose exertional HR is under pharmacological (β-blocker, ivabradine) or non-pharmacological (pacemaker) control. Anaemia and/or hypoxaemia may severely decrease arterial O2 content, making O2 pulse interpretation particularly challenging in a physically unfit patient. In some patients with unexplained exercise intolerance, an abnormally low cardiac output is not mechanistically linked to intrinsic cardiopulmonary disease but a failure to increase right atrial pressure, i.e. reduced pre-load [88]. It is also important to recognise that there are “central cardiovascular” causes of dyspnoea that are not associated with substantial impairment in stroke volume and/or cardiac output, e.g. heart failure with preserved ejection fraction [89, 90], exercise-induced PH [91] and right ventricle-to-pulmonary circulation uncoupling [58]. Thus, CPET variables reflecting “downstream” consequences of impaired O2 delivery might be insensitive to these common abnormalities. In this context, measurement of pulmonary haemodynamics, left ventricular filling pressures, Fick cardiac output and arterial–venous O2 content difference might be useful in selected patients with unexplained exercise intolerance (“invasive CPET”) [92, 93]. A less complex approach involves CPET in association with exercise echocardiography to assess stroke volume, diastolic function, mitral function, left ventricular outflow and dynamic pulmonary arterial pressures (“CPET imaging”) [94]. For instance, right ventricle-to-pulmonary circulation uncoupling showed an important role in determining flattening of the V′O2/WR relationship in patients with different cardiovascular diseases referred for exertional dyspnoea [95]. This specific haemodynamic abnormality was associated with high V′E–V′CO 2 indices [94] and left atrial dynamic impairment in heart failure [84] and hypertrophic cardiomyopathy [96]. Moreover, an echocardiographic E/e′ ratio of 15 at peak exercise in association with high V′E– V′CO 2 indices has been found useful in the diagnosis of heart failure with preserved ejection fraction in hypertensive patients with unexplained dyspnoea [97]. Risk assessment Limitations in the prognostic assessment of cardiopulmonary diseases The value of peak V′O2 as a prognostic index in heart failure with reduced ejection fraction has been constantly re-examined there is therefore ongoing controversy regarding the “best” peak V′O 2 threshold (if any) for each individual patient [58]. As outlined in a chapter by AGOSTONI and CATTADORI [49] later in this Monograph, it seems unlikely that CPET variables alone can predict outcome in such a heterogenous disease. The effects of interventions on the submaximal ventilatory gas exchange variables (e.g. V′E–V′CO 2 indices, PETCO2) have not been prospectively investigated in a large number of patients their predictive role may therefore change over time depending on the specific treatment approach. Categories of progressive risk for PH patients have also been proposed based on ventilatory gas exchange variables [62]. However, the supporting evidence for the chosen cut-off remains rather limited compared with heart failure with reduced ejection fraction. xvi https://doi.org/10.1183/2312508X.10015318