Respiratory variation in inferior vena cava diameter: surrogate of central venous pressure or parameter of fluid responsiveness? Let the physiology reply

In the previous issue of Critical Care, Muller and colleagues investigated whether respiratory variation in inferior vena cava diameter (ΔIVC) could be a useful predictor of fluid responsiveness in spontaneously breathing patients. The study concludes that accuracy was not very good and therefore that this parameter should be used with caution in these patients. There is still confusion about the meaning of IVC respiratory variations, whether the patient is spontaneously breathing or mechanically ventilated. In this brief commentary, we try to summarize as clearly as possible the significance of IVC variation in different clinical settings.

In the previous issue of Critical Care, Muller and colleagues [1] investigated whether the respiratory variations of inferior vena cava diameter (ΔIVC) could be an accurate predictor of fl uid responsiveness in spontaneously breathing patients with acute circulatory failure. Th e main result of this study is that ΔIVC predicted fl uid responsiveness moderately well: a ΔIVC value above 40% was associated slightly with fl uid responsiveness, whereas values under 40% were in conclusive. Th e main conclusion is that ΔIVC should be interpreted with caution in spontaneously breathing patients.
Fluid responsiveness is crucial in the management of a patient in shock. Hypovolemia is associated with a worse outcome, but fl uid overload is also associated with increased mortality [2], rendering somewhat hazardous the management of fl uids in the most critically ill patients by using fl uid challenges alone. In the last 10 years, many clinical and experimental studies on this subject have been carried out, leading to the validation in mechanically ventilated patients of a few 'dynamic' parameters based on heart-lung interactions. Notably, a signifi cant dilation of the IVC during tidal ventilation accurately predicts fl uid responsiveness [3,4].
However, there is still some confusion regarding the study of the IVC, and of its respiratory changes, in the intensive care unit. Th e paper by Muller and colleagues [1] off ers an opportunity to try to re-emphasize the basic physiology of the IVC (Figure 1). ΔIVC depends on few factors: the intrathoracic and abdominal pressures, the central venous pressure (CVP), and the compliance of the vessel. In mechanically ventilated patients, the objective of studying the IVC is to assess its ability to dilate during tidal ventilation, when intrathoracic pressure is increasing more than abdominal pressure. Th is dilation actually refl ects the ability of the IVC to receive more volume (preload reserve), like a preserved compliance. Th e IVC is then on the steep part of the relationship between IVC diameter and CVP ( Figure 2). As reported by Barbier and colleagues [5], such a relationship is curvilinear. In contrast, the absence of signifi cant dilation refl ects the inability of the IVC to receive more fl uid (no preload reserve), owing to low compliance. Th e IVC is then on the fl at part of its relationship with CVP ( Figure 2).
In spontaneously breathing patients, the situation is completely diff erent. Now, the objective of studying the IVC is not to evaluate its ability to dilate but its ability to collapse in response to a decrease in intrathoracic pressure and an increase in abdominal pressure. In such a situation, changes in IVC diameter refl ect simply the inter action between CVP and the range of gradient

Abstract
In the previous issue of Critical Care, Muller and colleagues investigated whether respiratory variation in inferior vena cava diameter (ΔIVC) could be a useful predictor of fl uid responsiveness in spontaneously breathing patients. The study concludes that accuracy was not very good and therefore that this parameter should be used with caution in these patients. There is still confusion about the meaning of IVC respiratory variations, whether the patient is spontaneously breathing or mechanically ventilated. In this brief commentary, we try to summarize as clearly as possible the signifi cance of IVC variation in diff erent clinical settings. between intrathoracic and abdominal pressures. In other words, the vein may collapse either because the CVP is very low or because the intrathoracic pressure becomes markedly negative. Th e latter situation occurs in severe acute asthma [6], exacerbation of acute chronic obstructive pulmonary disease, or any marked respiratory failure. In 1981, Mintz and colleagues [7] described a positive relationship between IVC diameter during expiration and right atrial pressure (RAP). In 1990, Kircher and colleagues [8] reported the value of IVC collapsibility in predicting RAP in patients who breathed quietly and hence did not develop any important variation of intrathoracic pressure. A collapsibility of greater than 50% indicated an RAP of below 10 mm Hg, and a collapsibility of less than 50% indicated an RAP of above 10 mm Hg [8]. In 2007, Osman and colleagues [9] found that a CVP of less than 10 mm Hg could not reliably distinguish responders from non-responders to fl uids but that a CVP of greater than 15 mm Hg correctly identifi ed nonresponders. Th erefore, one can hypothesize that the absence of respiratory changes in IVC diameter, refl ecting elevated CVP, should be associated with fl uid unresponsiveness but that respiratory changes in IVC diameter, refl ecting a normal or low CVP, are incon clusive. Surprisingly, the fi ndings of Muller and colleagues [1] indicate the opposite: a signifi cant IVC diameter variation was associated with fl uid responsiveness, whereas the absence of variation was inconclusive. Th is discrepancy may be explained by the study population. As reported in Table 2 of the article [1], 40% of the patients had hemorrhagic or hypovolemic shock, with a ΔIVC expected to be high (many patients actually had a complete collapse).   [5]. The pressure/diameter relationship shows an initial steep part where a minimal increase in CVP, in response to increased intrathoracic pressure, is associated with a large increase in IVC diameter and a fl at part where the compliance of the IVC decreases, resulting in less IVC dilation and a larger increase in CVP. Dark arrow: eff ect of increased intrathoracic pressure in a preload-responsive patient. Gray arrow: eff ect of increased intrathoracic pressure in a preload-unresponsive patient.
In conclusion, as re-emphasized by Muller and colleagues [1], it seems hazardous to manage fl uids in a spontaneously breathing patient by using IVC respiratory variations only, until further data are published. Alter native methods, such as passive leg raising [10] and a minifl uid challenge [11], could be used.