This article refers to ‘Contributions of cardiac dysfunction and volume status to central haemodynamics in chronic heart failure’ by W.L. Miller et al., published in this issue on pages 1097–1105. Fluid overload with central and/or peripheral congestion is the cardinal feature characterizing the syndrome of heart failure (HF). The pathophysiology underlying fluid retention is complex and multifactorial but it seems that this process develops at the early stage of the disease, being related to augmented reflex sympatho-excitation (in response to transient elevations in intracardiac pressures) with an impairment in renal sodium and water control.1 In parallel, myocardial dysfunction with resulted reduced cardiac output causes arterial underfilling, sensed as a reduction in effective circulating blood volume (BV) with baroreceptors unloading and reflex neurohormonal activation, which also leads to further deterioration (retention) in renal sodium and water handling.2 Although the mechanisms underlying the link between congestion and spiraling progression of HF are not clearly understood, the association between presence of congestion and more severe symptoms and higher risk of adverse clinical outcomes is documented.3 Congestion remains the main reason for hospital admission with acute HF and its persistence at discharge contributes to readmission and predicts poor outcomes.4, 5 Thus, effective and safe decongestion remains the major goal and challenge for optimal management of patients with HF in the chronic, stable settings and even more for those admitted to hospital with signs and symptoms of acute HF. Additionally, prevention of fluid overload, identification of patients at risk, early detection and monitoring of congestion during the entire natural course of HF are essential to improve patients' quality of life and outcomes. Fluid accumulation starts in the intravascular compartment which comprises arterial and venous systems – the latter being much more complaint to store an increasing volume of circulating blood.5, 6 In particular, splanchnic veins with low vascular resistance and high capacitance receive ∼25% of the cardiac output and serve as major functional venous reservoir able to accumulate as much as 20–50% of the total BV.6, 7 With time, hydrostatic pressure in the vessels increases and fluid accumulation expands to the extravascular, interstitial space which is more compliant with fluid capacity 3–4 times bigger than the intravascular compartment.5, 6 Clinically, it manifests as multi-organ congestion. The interstitial and intravascular compartments are separated by highly permeable membranes and the amounts of fluid distributed between them are mainly related to hydrostatic and colloid osmotic pressures across the capillary membranes. Typically, stable settings of chronic, symptomatic HF, despite an overall increase in body fluid in the majority of patients, can be characterized by an equilibrium in fluid distribution across the extracellular compartments (interstitial and intravascular) (Figure 1A). The magnitude of fluid accumulation is one of the major determinants of HF symptom severity and remains the target of decongestive therapy.5 The pathophysiology underlying deterioration of this equilibrium is complex with multiple triggering factors, clinically manifesting as worsening signs and symptoms of HF leading to clinical decompensation, often with hospital admission with a diagnosis of acute HF. Patterns of fluid overload development (i.e. gradual vs. rapid), localization and magnitude of congestion, and its response to therapy all are important elements for clinical evaluation of patients admitted with acute HF. Additionally, patterns of congestion in decompensated HF patients are associated with different treatment response, and short-term and long-term outcomes.8 Such clinical heterogeneity seems to be the result of the relative predominance of one type of congestion – either in the intravascular or in the interstitial compartments (Figure 1B and 1C, respectively). Typically, the former presents as rapid involvement of venous blood reservoir with an increase in intracardiac filling pressure (haemodynamic congestion) and sudden development of symptoms (Figure 1B). This pattern is often referred to as vascular congestion5, 9 of which pulmonary oedema due to acute hypertensive emergency is most illustrative and results from sympathetically-driven fluid redistribution mainly from the splanchnic circulation, rather than from fluid accumulation.7 These patients may not present excessive volume overload, which explains absence of any detectable changes in body weight and/or BV preceding decompensation.7 In contrast, the pattern with predominance of interstitial congestion (referred to as cardiac congestion5, 9) mainly results from constant accumulation of fluid in the interstitial tissues, develops gradually with much slower clinical deterioration, an increase in BV and body weight (often of multi-kilogram magnitude), and gradual increase in intracardiac pressures (Figure 1C). Of note, fluid balance at the level of the interstitium is strongly dependent on the lymphatic system able to increase fluid removal by 10- to 50-fold when hydrostatic pressure rises.5 However, the lymphatic drainage may be deteriorated in case of significantly elevated right heart pressure due to decreased perfusion gradient at the level of the thoracic duct – which may lead to lymphatic congestion. In relation to the above considerations, Miller et al.10 provide an interesting and clinically relevant piece of information. The authors aimed to study simultaneously: intravascular volume status (using radiolabel indicator-dilution technique to assess total BV and plasma volume), haemodynamics with right heart catheterization (RHC), and measures of cardiac function with detailed echocardiographic assessment, to explore the relative contributions of each component to congestion in HF. Patients with stable chronic HF were divided according to the value of total BV into those with hypervolaemia (defined as total BV >+8% above referenced normal volume) and euvolaemia (total BV ≤+8%). The essential conclusion was that only combined analysis of volume (intravascular), pressure (intracardiac), and cardiac function provides a comprehensive picture of HF status with consequences that may guide individualized therapy.10 We believe that there are some additional findings of particular interest. Firstly, despite including patients with moderate-severe HF symptoms [New York Heart Association (NYHA) class III–IV], less than 40% presented with normal intravascular volume (euvolaemia). Additionally, the authors carefully evaluated resting haemodynamics and reported that 20% of patients had normal right ventricular and 24% normal left ventricular filling pressures.10 It reassures us that in even in patients with advanced HF (NYHA class III–IV), reported symptoms are often neither related to excess of intravascular volume nor to elevated intracardiac pressures. Here, one may want to refer to the seminal work of Prof. Phillip Poole-Wilson, who more than three decades ago was teaching us that in acute HF severe dyspnoea is typically related to elevated left atrial pressure, which reduction almost always leads to an improvement in symptoms.11 However, patients with chronic HF tend to report symptoms even when fluid overload is corrected with diuretics and shortness of breath under these circumstances is not related simply to central haemodynamics but is determined more by the interaction of changes in respiratory pattern and the metabolic consequences of reduced perfusion of exercising skeletal muscle.11 Secondly, the assessment of intravascular volume (BV) even using sophisticated and accurate methods (not readily available in clinical practice) does not tell us everything about fluid status and congestion in HF. As depicted in Figure 1, measures of only one compartment may be even misleading in the proper interpretation of the presence or absence of fluid overload (e.g. intravascular euvolaemia with fluid overload in the interstitial compartment). Total intravascular volume rather poorly correlates with physical signs and symptoms of congestion, assessment of natriuretic peptide levels, and does not provide meaningful information about fluid redistribution from the splanchnic reservoir.7 It has been documented that stimulation of splanchnic nerves with resultant involvement of splanchnic blood pool leads to significant increase in cardiac preload and cardiac output within minutes, all being entirely unrelated to any change in total BV.12 It forms the strong background for the management of congestion based on pattern of its development. For predominant ‘cardiac type’ of congestion, diuretics, aquaretics or ultrafiltration (in resistant forms) are preferable and recommended (Figure 1D).5, 9 Other forms of therapy (although less thoroughly investigated) may include hypertonic saline infusion or different forms of compression therapy.5 Interestingly, there are reports that sodium–glucose co-transporter 2 (SGLT2) inhibitors tend to facilitate decongestion in volume-overloaded HF patients via different mechanisms, one of which may be associated with stimulation of osmotic diuresis13 which potentially places SGLT2 inhibitors as effective decongestive therapy. For predominant vascular type of congestion, vasodilators (with diuretics) are recommended5, 9 and recent reports have shown that neural control of the splanchnic system may also be a novel form of management (Figure 1E).7, 9, 14 Last, but not least, in HF, elevated (cardiac) pressures should not be equated with excessive (intravascular) volumes with further therapeutic decisions. In this study, in the euvolaemic patients, 70% displayed elevated right heart and 63% left heart filling pressures; conversely, among those with hypervolaemia, around 15% had normal cardiac filling pressures.10 It confirms that although in decompensated HF RHC-derived indices are the gold standard used to guide up/down-titration of diuretics/vasodilators to a haemodynamic goal, in stable settings it may not be applicable. In the ESCAPE trial, RHC used in addition to routine HF management was associated with better exercise tolerance and quality of life, but not with improvement in mortality and hospitalizations.15 To achieve euvolaemia in patients with HF (both in acute and chronic settings) remains challenging (one my even say – almost impossible) for numerous reasons. A reliable, reproducible, universal and simple tool for assessment of congestion in our diagnostic armamentarium is still lacking.16, 17 As described above, the underlying mechanisms and triggering factors are complex, in some cases unidentifiable. An assessment of the central haemodynamics through the prism of pressures does not provide information about the excessive volume. Dynamic interplay between the intravascular and the interstitial compartments determines the magnitude and clinical picture of congestion. Fluid can move from the extravascular space into the central veins to maintain appropriate level of right heart filling and required level of cardiac output. The rate at which this volume is exchanged is called the plasma refilling rate (PRR). However, if fluid is removed too fast from the intravascular space (a rate greater than the PRR), it can lead to reduced cardiac output/renal hypoperfusion with subsequent activations of neurohormonal systems and intrarenal mechanisms that cause further sodium and water retention with detrimental consequences. The study by Miller et al.10 confirms that what really matters for better characterization of HF patients is to evaluate all three components – pressure, volume and cardiac function. Unfortunately, the authors do not tell us how (and if) it translates into better management of congestion and so we remain in our continuous search for elusive object of euvolaemia in HF. Conflict of interest: none declared.