Secondary Logo

Journal Logo

Clinical Science Aspects

Left-Sided Ventricular–arterial Coupling and Volume Responsiveness in Septic Shock Patients

Li, Suwei; Wan, Xianyao; Laudanski, Krzysztof; He, Ping; Yang, Lihui

Author Information
doi: 10.1097/SHK.0000000000001327
  • Free
  • Editor's Choice



Improving tissue perfusion and oxygenation are considered the targets in shock resuscitation strategies, but guiding fluid therapy effectively remains a significant challenge (1, 2). While fluid is the necessary medium for the heart to work as a pump, fluid overload carries significant side effects (3, 4). Different indices of fluid responsiveness, such as central venous pressure (CVP), pulmonary artery occlusion pressure (PAOP), global end-diastolic volume (GEDV), stroke volume variation (SVV), pulse pressure variation (PPV), a variation of inferior vena cava diameter, passive leg raising (PLR), and the rapid fluid challenge, were investigated to provide just optimal expansion of intravascular space in patients with septic shock (5). Still, up to 50% of septic patients are not fluid responsive, or fluid loading does not lead to the adequate tissue perfusion (6, 7). Ineffectiveness of fluid loading may be due to increased capillary leakage, increased capacitance in blood vessels, pulmonary hypertension, diastolic heart dysfunction, or discordance between the heart and compliance of the arterial system (8–11). Abnormal ventricular–arterial coupling (VAC) is often an overlooked reason for lack of hemodynamic improvement after fluid loading due to the fact that the interaction between the left ventricle and the arterial system determines global cardiovascular performance (12–14). VAC is most commonly assessed by analysis of the relationship between cardiac output and arterial elastance (15, 16). If cardiac preload is adequate and fixed, VAC will largely determine the stroke volume, but the determination itself is an interplay between left ventricle and arterial elastance (15, 16). Subsequently, the poor performance of the left ventricle due to inappropriate VAC may lead to diminished cardiac output and inadequate tissue perfusion even when fluid loading is optimal. It has been demonstrated that ventricular–arterial decoupling occurs in septic patients, but the study did not account for variation in preload (17, 18).

Here, we investigated VAC in septic shock patients after successful fluid resuscitation determined by global end-diastolic volume. Specifically, we hypothesized that in some septic shock patients, decoupling of VA will result in an apparent lack of fluid responsiveness. Moreover, we hoped to find further clues on how to address key factors leading to VAC decoupling.


Study sample

We enrolled a convenience sample of 35 patients admitted to the Department of Critical Care Medicine, the First Affiliated Hospital of Dalian Medical University in the period of June 2014 to December 2015. The inclusion criteria were: septic shock as defined by Society of Critical Care Medicine/European Society of Critical Care (19); age ≥ 18 years; monitoring of continuous cardiac output with pulse indicator continuous cardiac output (PiCCO) ≥ 48 h; during the first 48 h of PiCCO monitoring, within any interval ≤ 6 h, variation of global end-diastolic volume index (ΔGEDVI) increase above or equal to 10% regardless of reason; no alterations in dosage of vasopressors, inotropic, sedative and analgesic drugs within 6 h of the first 48 h after PiCCO monitoring. Also, patients were enrolled only when several clinical variables were met as specified in the section below. The entry exclusion criteria were: acute myocardial infarction; pulmonary embolism; acute central nervous system emergencies (cerebral infarction or hemorrhage, cerebral herniation); patient or family refusal for treatment; non-compliance with treatment protocol as specified in the manuscript.

The Institutional Research and Ethics Committee of our hospital approved this study for human subjects. Informed consent was obtained from all patients or next of kin before the data were included in the study.

Clinical management

All patients in this study were treated according to the two-step protocol. In the first step, we aimed to achieve the following goals: fluid infusion would be given if 15 min infusion of 300 to 500 mL bolus of crystalloid solution resulted in an increase of CO more than 10%; once CO remained stable despite fluid infusion, fluid resuscitation would be done to maintain central venous pressure (CVP) 8 to 12 mm Hg; vasopressors would be used to maintain mean arterial pressure (MAP) ≥ 65 mm Hg if necessary; if saturation of oxygen in central vein (ScvO2) < 70% or lactate (Lac) > 2 mmol/L, red blood cells would be infused to maintain hematocrit ≥ 30%, and/or inotropic drugs would be added. We escalated the therapy by employing PiCCO monitoring if ScvO2 remains < 70%, or lactate level remained above 2 mmol/L, or urine output < 0.5 mL·kg−1·h−1. Also, if the patient required a high amount of vasopressors: dopamine > 5 ug·kg−1·min−1, or dobutamine > 5 ug·kg−1·min−1, or epinephrine > 0.2 ug·kg−1·min−1, or norepinephrine > 0.2 ug·kg−1·min−1; and MAP < 65 mm Hg (hypertension patient MAP < 75 mm Hg) PiCCO monitoring would be placed as well.

Hemodynamic measurements

CVP was measured via central venous catheter indwelled into the internal jugular or subclavian vein and zeroed to an axillary line with the fourth intercostals level in supine position. Cardiac index (CI), stroke volume index (SVI), global end diastolic ventricular volume index (GEDVI), systemic vascular resistance index (SVRI), arterial systolic blood pressure (ASP), mean arterial blood pressure (MAP), and heart rate (HR) were recorded with the PiCCO monitor (Pulsion Medical Systems, Munich, Germany) via femoral access in accordance with the manufacturer's instruction. Triplicate transpulmonary thermodilution measurements of 15 mL∼4°C iced 0.9% sodium chloride bolus were injected. Lac and ScvO2 were measured by a blood gas analyzer (GEM Premier 4000, Instrumentation Laboratory Co, Ann Arbor, Mich).

The initial point of the hemodynamic record interval was T1, and the end point was T2. The hemodynamic indicators were named X1 and X2, respectively. So, the difference in hemodynamic parameters (ΔX) was equal to X2 − X1.

The evaluation of ventricular–arterial coupling in the pressure–volume curve was calculated as Ees = ESP/(ESV − V0), Ea = ESP/SV (Fig. 1). Ees represents left ventricular end-systolic elastance, ESP represents the end-systolic pressure of ventricle, ESV represents the end-systolic volume of the ventricle, and V0 represents the cross point of the ventricular end-systolic pressure–volume curve and horizontal axis. V0 is assumed to be negligible according to the most common strategy to evaluate VAC (17, 20, 21). We used (GEDV/4 − SV) to represent ESV and ASP of the femoral artery for ESP. V0 was considered negligible. Ea represents effective arterial elastance. Finally, to keep uniform the comparisons, all formulas were indexed as follows: EesI = ASP/(GEDVI/4 − SVI), EaI = ASP/SVI, EaI/EesI = (GEDVI/4-SVI)/SVI (20, 21). Finally, the dimensionless product of formula– EaI/EesI defined ventricular arterial coupling (VAC). However, one has to mind that the ventricular arterial coupling can be adequately assessed by concomitant evaluation of EesI, EaI and EaI/EesI (15).

Fig. 1
Fig. 1:
Ventricular pressure–volume diagram from which effective arterial elastance (Ea) and left ventricular (LV) end-systolic elastance (Ees) are derived.

According to the cut-off value of 10% in a variation of CI (ΔCI), we divided the participants into a ventricular volume responsive group (VVr group; n = 16) and ventricular volume unresponsive group (VVur group; n = 19) after the completion of the initial resuscitation.


Results for continuous variables with normal distributions are given as mean ± standard deviation (SD). Student t test and analysis of variance were used to compare means between the two groups. Single-model ANOVA with LSD post-hoc analysis was used in multiple group contrasts. Results for qualitative variables were expressed as percentages and compared between groups using Fisher exact test. Pearson linear regression analysis was used to analyze correlation. P values less than 0.05 were considered statistically significant. The statistical analysis was performed with SPSS version 19.0 (IBM, Armonk, NY).


The demographic and clinical characteristics of the included patients at the origin of PiCCO monitoring were summarized in Table 1.

Table 1
Table 1:
Demographic and clinical characteristics

Sixteen patients were in the VVr group while 19 patients were classified as VVur. Fluid responsive individuals had expected changes in CI, SVI, SVRI, Lac, and ScvO2 over the study interval (Table 2). VVur exhibited diminished VAC (EaI/EesI) significantly at T1 as compared with VVr patients (P = 0.039). Interestingly, VAC (EaI/EesI) increased in VVur, while it was decreased in the VVr group (P = 0.04). The significant differences of EaI/EesI trends were concomitant with substantial variation of arterial elastance index between the two time points (P<0.001) (Table 2).

Table 2
Table 2:
Hemodynamic indicators in fluid responsive (VVr) and fluid unresponsive (VVur) group

In order to assess the effect of VAC determinants on fluid responsiveness in more depth, we divided all patients into a ΔEaI/EesI > 0 group (EaI highly variable while ventricular elastance rather low) and ΔEaI/EesI ≤ 0 (EesI much more variable than artrial elastance) group (Table 3). There were 26 patients in the ΔEaI/EesI > 0 group, and 9 in the ΔEaI/EesI ≤ 0 group. ΔCI, ΔSVI, and ΔEesI were significantly elevated in the ΔEaI/EesI ≤ 0 group as compared with individuals in the ΔEaI/EesI > 0 group (Table 3). On the other hand, ΔEaI and ΔSVRI were significantly diminished in the ΔEaI/EesI ≤ 0 group. The significantly higher proportion of patients in the ΔEaI/EesI ≤ 0 group were categorized as ventricular volume responsive (88.9%) while only 26.9% of patients with ΔEaI/EesI > 0 were defined as ventricular volume responsive (P = 0.01). Furthermore, we found that change in cardiac index (ΔCI) correlated negatively with ΔEaI/EesI (Fig. 2A), ΔEaI (Fig. 2B) and ΔSVRI (Fig. 2C). Additionally, VAC correlated with the serum level of lactic acid (r = −0.464; P = 0.005). Changes in other parameters (CVP, GEDVI, EesI, ScvO2) had no correlation with ΔCI (data not shown). None of these measurements taken at the beginning of the study (T1) correlated with ΔCI either (data not shown).

Table 3
Table 3:
Hemodynamic indicators in ΔEaI/EesI≤0 group and ΔEaI/EesI>0 group
Fig. 2
Fig. 2:
Change in cardiac index (ΔCI) inversely correlated to ΔEaI/EesI (A), ΔEaI (B), and ΔSVRI (C).

To explore the role of the VAC component on global tissue perfusion (determined as lactate production or ScvO2), we split the studied individuals in groups based on ΔCI versus ΔEaI/EesI variable as described in Table 3. We found nine subjects ΔEaI/EesI ≤ 0. Only one exhibited ΔCI below 10% versus the remaining eight had a ΔCI above 10%. However, in the remaining 26 patients with ΔEaI/EesI > 0, 8 had an increase in ΔCI over 10% while 18 show a small ΔCI (below 10%). We found no difference in tissue perfusion indices at the beginning or end of the observation period (data not shown). In contrast, an analysis of change demonstrated that when compared in patients with ΔCI<10%, change in CI was significantly less in the ΔEaI/EesI > 0 group versus ΔEaI/EesI ≤ 0. Concomitantly, changes in ventricular arterial compliance (EaI/EesI) were significant but there was no change in lactate clearance if CI was unchanged (Table 4). The groups that least effectively cleared lactate were those with that in ΔEaI/EesI > 0 and ΔCI < 10% (Table 4). Both groups with ΔCI > 10% had good clearance irrespective of their ventricular arterial coupling (Table 4). ScvO2 was similar across all groups (Table 4).

Table 4
Table 4:
Changes in perfusion markers between three groups with different VAC and CI


Our study demonstrated that VAC plays a role in cardiovascular performance in individuals with septic shock. In patients with no difference of ΔGEDVI, a subgroup of volume responsive showed decreased ΔEaI/EesI while those ventricular volume unresponsive had increased ΔEaI/EesI. Concomitantly, ΔCI was significantly higher in the ΔEaI/EesI ≤ 0 group than that in the ΔEaI/EesI > 0 group. This suggests that in clinical situations the lack of apparent improvement in hemodynamics could be due to VA decoupling, and VAC affects ventricular volume responsiveness or fluid responsiveness. The data suggests that the primary component may be loss of arterial elastance. Similar suggestions have been made by others but never shown so definitely. This opens a theoretical therapeutic opportunity for patients with “unresponsive shock.” Most significantly, our study suggests that increasing vasopressors may, in fact, worsen VAC by increasing arterial stiffness (22). These findings are aligned with the decrease of EaI/EesI in the VVr group demonstrating improved fluid responsiveness, CI increase, and more favorable lactic acid clearance as compared with the VVur group. Furthermore, in the ΔEaI/EesI ≤ 0 group, CI increased more than in the ΔEaI/EesI > 0 group, and Lac decreased more. These relationships are not straightforward as we demonstrated that lactate production is more directly aligned with CI than with VAC.

Optimal VAC depends on mechanical properties of the heart (EesI) and effective arterial elastance (EaI). The interaction between them determines VAC. Determining which of these two components has a greater effect may offer a clue as to the most appropriate therapeutic strategy. Theoretically, increasing EesI is beneficial if VAC is suboptimal. Although we can somewhat influence EesI, there are no straightforward ways to modify EaI. In the decompensated stage of septic shock, decoupling of VA was characterized by more severe decreased left myocardial contractility compared with left ventricular afterload in a rabbit model of septic shock (16). Concomitantly, Zhou et al. (23) demonstrated that in elderly (range 61–80 years old)septic shock patients, the lower cardiac contraction was the primary cause of VA decoupling. This is likely secondary to decreased age-related elastance of the arterial system, but a definitive study should include patients with a wider spectrum of ages. Our study did not include a sufficient age range among different individuals to conduct such analysis. Nevertheless, increasing EesI in order to optimize VAC remains a very attractive theoretical option to optimize tissue perfusion in septic patients (24, 25).

Alternatively, VAC optimization via EaI may be achieved by decreasing cardiac afterload (17). Our study demonstrated that ΔEaI and ΔEaI/EesI in the VVr group were significantly lower than those in the VVur group while ΔEesI was not different between these two groups. So, decreasing cardiac afterload was also a proper way to improve VAC and fluid responsiveness.

Several considerations must be taken into account while analyzing the results of the study. First, this was a retrospective study with a short interval for observation (≤ 6 h) of ΔGEDVI equal or above 10%. Several other factors were variable during data collection affecting fluid responsiveness (mechanical ventilation, ongoing fluid resuscitation). We restricted changes in vasopressor, inotropic, analgesic, and sedative drugs to minimize their interference of arterial elastance and cardiac contractility, two pivotal components of VAC. However, such a scenario is rarely seen in clinical situations considering the dynamic nature of sepsis. Most of the time, vasopressors are frequently adjusted in clinical settings. Furthermore, we fixed the dose, but we had no way to assess the biological effect of the agents. The dynamic nature of sepsis may render their effects on elastance and heart contractility variable despite fixed doses due to the changes in pH, cortisol effect, and other factors (26, 27). The evaluation of VAC was done not by left ventricular–pressure–volume loops although these methods have several limitations (28). We did not perform a power analysis due to the lack of high-quality pre-existing studies. However, we found that 25% of our subjects had ΔEaI/EesI ≤ 0 suggesting high prevalence of the VAC decoupling. Several other subanalyses found subgroups well represented. Finally, our variables were measured using a PiCCO monitor. Despite the sophistication of the monitor, it is highly invasive and not always readily available. However, ultrasound may provide a more noninvasive and rapid way to assess several components of VAC. Chen et al. (29) had developed and validated a method to estimate left ventricular end-systolic elastance in humans from noninvasive single-beat parameters based on cardiac ultrasound, such as SV and EF. Finally, we did not establish a clear relationship between lactate clearance and changes in VAC components defined by ΔEaI/EesI. Partially, this was because of negligible representation of patients with ΔEaI/EesI ≤ 0 and significant change in cardiac index rendering a complete statistical analysis impossible. Second, the degree of VAC decoupling was not enough to impair changes in cardiac performance and subsequent tissue perfusion. Only when decoupling was affected and cardiac index depressed did we see significant depression in lactate clearance. It remains to be seen if there is a threshold of VAC impairment causing clinically significant changes. Or, it is likely that in our study patients were not sick enough. The value of ScvO2 across studied groups suggests this. Additionally, our window of measurements was relatively short and the level of lactate was not very high.

In conclusion, we identified the relationship between VAC and the ventricular volume responsiveness or fluid responsiveness in septic patients. Subsequent decoupling cannot be addressed by more fluid resuscitation but rather by other means. Further research should focus on a new strategy bringing in the VAC as one target of septic shock resuscitation strategy, especially since several components of VAC may be differentially affected in sepsis.


The authors thank all nurses, residents, and other personnel of the participating department of Critical Care Medicine for their generous cooperation.


1. Malbrain MLNG, VanRegenmortel N, Saugel B, De Tavernier B, Van Gaal PJ, Joannes-Boyau O, Teboul JL, Rice TW, Mythen M, Monnet X. Principles of fluid management and stewardship in septic shock: it is time to consider the four D's and the four phases of fluid therapy. Ann Intensive Care 8 1:66, 2018.
2. Loflin R, Winters ME. Fluid resuscitation in severe sepsis. Emerg Med Clin North Am 35 1:59–74, 2017.
3. Elwan MH, Roshdy A, Elsharkawy EM, Eltahan SM, Coats TJ. The haemodynamic dilemma in emergency care: is fluid responsiveness the answer? a systematic review. Scand J Trauma Resusc Emerq Med 25 1:25, 2017.
4. Byrne L, Van Haren F. Fluid resuscitation in human sepsis: time to rewrite history? Ann Intensive Care 7 1:4, 2017.
5. Marik PE, Lemson J. Fluid responsiveness: an evolution for our understanding. Br J Anaesth 112 4:617–620, 2014.
6. Hou PC, Filbin MR, Napoli A, Feldman J, Pang PS, Sankoff J, Lo BM, Dickey-White H, Birkhahn RH, Shapiro NI. Cardiac output monitoring managing intravenous therapy (COMMIT) to treat emergency department patients with sepsis. Shock 46 2:132–138, 2016.
7. Feissel M, Kalakhy R, Banwarth P, Badie J, Pavon A, Faller JP, Quenot JP. Plethysmographic variation index predicts fluid responsiveness in ventilated patients in the early phase of septic shock in the emergency department: a pilot study. J Crit Care 28 5:634–639, 2013.
8. Toscani L, Aya HD, Antonakaki D, Bastoni D, Watson X, Arulkumaran N, Rhodes A, Cecconi M. What is the impact of the fluid challenge technique on diagnosis of fluid responsiveness? a systematic review and meta-analysis. Crit Care 21 1:207, 2017.
9. Semler MW, Rice TW. Sepsis resuscitation: fluid choice and dose. Clin Chest Med 37 2:241–250, 2016.
10. Miranda M, Balarini M, Caixeta D, Bouskela E. Microcirculatory dysfunction in sepsis: pathophysiology, clinical monitoring, and potential therapies. Am J Physiol Heart Circ Physiol 311 1:H24–H35, 2016.
11. Lv X, Wang H. Pathophysiology of sepsis-induced myocardial dysfunction. Mil Med Res 3:30, 2016.
12. Antonini-Canterin F, Carerj S, Di Bello V, Di Salvo G, La Carrubba S, Vriz O, Pavan D, Balbarini A, Nicolosi GL. Research Group of the Italian Society of Cardiovascular Echography (SIEC). Arterial stiffness and ventricular stiffness: a couple of disease or a coupling disease? a review from the cardiologist's point of view. Eur J Echocardiogr 10 1:36–43, 2009.
13. Starling MR. Left ventricular-arterial coupling relations in the normal human heart. Am Heart J 125 6:1659–1666, 1993.
14. Little WC, Pu M. Left ventricular-arterial coupling. J Am Soc Echocardiogr 22 11:1246–1248, 2009.
15. Shim CY. Arterial-cardiac interaction: the concept and implications. J Cardiovasc Ultrasound 19 2:62–66, 2011.
16. Chirinos JA. Ventricular-arterial coupling: invasive and non-invasive assessment. Artery Res 2013; 7 1: doi: 10.1016/j.artres.2012.12.002. [Epub ahead of print].
17. Li SW, Liu DW, Wang XT, Chai WZ, Chen XK, Zhang HM, Zhang H. Left ventricular-arterial coupling in the resuscitation of refractory septic shock. Zhonghua Yi Xue Za Zhi 92 15:1012–1016, 2012.
18. Li SW, Liu DW, Wang XT, Zhang B, Chen XK, Chai WZ, Zhang HM, Yang QH, He HW. Variations of left ventricular-arterial coupling in septic shock. Zhonghua Yi Xue Za Zhi 93 35:2782–2785, 2013.
19. Singer M, Deutschman CS, Seymour CW, Shankar-Hari M, Annane D, Bauer M, Bellomo R, Bernard GR, Chiche JD, Coopersmith CM, et al. The third international consensus definitions for sepsis and septic shock (Sepsis-3). JAMA 315 8:801–810, 2016.
20. Chantler PD, Lakatta EG, Najjar SS. Arterial-ventricular coupling: mechanistic insights into cardiovascular performance at rest and during exercise. J Appl Physiol 105 4:1342–1351, 2008.
21. Sunagawa K, Maughan WL, Burkhoff D, Sagawa K. Left ventricular interaction with arterial load studied in isolated canine ventricle. Am J Physiol 245 (5 pt 1):H773–H780, 1983.
22. Lesur O, Delile E, Asfar P, Radermacher P. Hemodynamic support in the early phase of septic shock: a review of challenges and unanswered questions. Ann Intensive Care 8 1:102, 2018.
23. Zhou XY, Li L, Gong SJ, Yu YH, Dai HW, Yan J. The influence of left ventricular-arterial coupling on the prognosis of elderly patients with septic shock. Zhongha Nei Ke Za Zhi 55 6:435–439, 2016.
24. Tsolaki V, Makris D, Mantzarlis K, Zakynthinos E. Sepsis-induced cardiomyopathy: oxidative implications in the initiation and resolution of the damage. Oxid Med Cell Longev 2017:7393525, 2017.
25. Suzuki T, Suzuki Y, Okuda J, Kurazumi T, Suhara T, Ueda T, Nagata H, Morisaki H. Sepsis-induced cardiac dysfunction and β-adrenergic blockade therapy for sepsis. J intensive Care 5:22, 2017.
26. Dubin A, Lattanzio B, Gatti L. The spectrum of cardiovascular effects of dobutamine-from healthy subjects to septic shock patients. Rev Bras Ter Intensiva 29 4:490–498, 2017.
27. Zimmerman JJ. Adjunctive steroid therapy for treatment of pediatric septic shock. Pediatr Clin North Am 64 5:1133–1146, 2017.
28. Chirinos JA, Sweitzer N. Ventricular-arterial coupling in chronic heart failure. Card Fail Rev 3 1:12–18, 2017.
29. Chen CH, Fetics B, Nevo E, Rochitte CE, Chiou KR, Ding PA, Kawaguchi M, Kass DA. Noninvasive single-beat determination of left ventricular end-systolic elastance in humans. J Am Coll Cardiol 38 7:2028–2034, 2001.

Fluid responsiveness; resuscitation; septic shock; ventricular–arterial coupling; ASP; arterial systolic pressure; CI; cardiac index; CVP; central venous pressure; EaI; effective arterial elastance; EesI; left ventricular end-systolic elastance; GEDVI; global end-diastolic volume index; HR; heart rate; Lac; lactate; MAP; mean arterial pressure; PiCCO; pulse indicator continuous cardiac output; ScvO2; central venous oxygen saturation; SVI; stroke volume index; SVRI; systemic vascular resistant index; VAC; ventricular–arterial coupling; VVr group; ventricular volume responsive group; VVur group; ventricular volume unresponsive group

Copyright © 2019 by the Shock Society