The Chronological Demographics of Ventricular-Arterial Decoupling in Patients with Sepsis and Septic Shock: A Prospective Observational Study

Background Ventricular-arterial coupling (V-A coupling) recently gathers attention from clinicians to evaluate the interaction between afterload and left ventricular systolic function. We aimed to describe the chronological demographics of V-A decoupling in patients with sepsis and septic shock through the clinical course. Method We conducted a single-center prospective observational study comprising adult patients with sepsis and septic shock admitted to the tertiary care hospital between 04/2017 and 03/2019. Patients’ characteristics, lab data on admission, and echocardiographic parameters including Ea and Ees on the day- 1, 2, 3, 7, and 14–28 were collected. V-A decoupling was defined as Ea/Ees ≥ 1.36. Results Seventy-one patients with sepsis or septic shock were enrolled. The prevalence of V-A decoupling was as follows; day-1: 25.4%, day-2: 23.8%, day-3: 13.3%, day-7: 18.5%, day-14–28: 30.3%, respectively. Ea was higher in patients with V-A decoupling than those without throughout the clinical course (day1; 2.8 vs. 1.8, p < 0.01, day2; 2.7 vs. 1.9, p < 0.01, day3; 2.8 vs. 2.1, p = 0.06, day7; 2.7 vs. 1.9, p = 0.02, day14-28; 2.4 vs. 1.8, p = 0.08). This increase in Ea was mainly induced by reduced stroke volume (SV) as well as high systolic blood pressure (SBP) in the earlier course of sepsis but only by increased SBP in the later course of sepsis. Ees was lower in patients with V-A decoupling than those without throughout the clinical course (day1; 1.3 vs. 2.1, p < 0.01, day2; 1.5 vs. 2.3, p < 0.01, day3; 1.6 vs. 2.3, p = 0.02, day7; 1.8 vs. 2.3, p = 0.01, day14-28; 1.2 vs. 1.9, p = 0.07). Conclusion We reported that V-A decoupling was commonly seen in patients with sepsis and septic shock. In patients with V-A decoupling, both Ea and Ees were significantly altered, but the causes of these alterations appeared to be changing over the clinical course of sepsis.


Introduction
Sepsis is a dysregulated systemic inflammatory response to an infection 1 and can result in end-organ dysfunction such as acute kidney injury, encephalopathy, and myocardial dysfunction. 2 Sepsis-induced myocardial dysfunction was first described as an acute reversible myocardial depression in patients with sepsis and septic shock by Parker et al in 1984. 3 Despite being described over three decades ago, sepsis-induced cardiomyopathy (SICM) has been underrecognized and the reported prevalence varies from 10-70%. 4 A lack of clear and consistent diagnostic criteria and the dynamic changes in hemodynamics are likely to be responsible for the under-recognition of this entity.
Left ventricular ejection fraction (LVEF) has been most commonly used as an indicator of left ventricular systolic function. LVEF could be affected by afterload that is a resultant of resuscitation with fluid and vasopressors and this makes it difficult to assess accurate left ventricular function in patients with sepsis and septic shock. Therefore, there has been a necessity for taking afterload into account.
Ventricular-arterial coupling (V-A coupling) recently gathers attention from clinicians as such an indicator. 5 V-A coupling is the ratio between arterial elastance (Ea) and left ventricular endsystolic elastance (Ees). Ea is the slope of the end-systolic pressure-stroke volume relationship, which represents how much pressure (mm Hg) is necessary to create stroke volume per unit (ml). Ees is the slope of the end-systolic pressure-volume relationship and therefore it represents how much pressure (mm Hg) is necessary to expand left ventricular volume per unit (ml). Ea/Ees represents the interaction between afterload and ventricular systolic function. When Ea/Ees is near unity, the efficiency of the cardiovascular system is optimal and when it is higher than the normal range, it is called V-A decoupling. Guarracino et al reported that septic patients often showed significant V-A decoupling. 6 While V-A decoupling suggests an altered cardiovascular performance, 5 it has not been well-described whether V-A decoupling in septic patients is due to increased Ea, reduced Ees, or both, and whether the etiologies for alteration of Ea/Ees vary through clinical course or not.
In this study, we aimed to describe chronological demographics of Ea and Ees, as well as other echocardiographic parameters in patients with sepsis and septic shock with or without V-A decoupling.

Methods
We conducted a single-center prospective observational study comprising adult patients with sepsis and septic shock who were admitted to the intensive care unit (ICU) from the emergency department (ED) in Urasoe General Hospital (UGH) between April 2017 and March 2019. UGH is a tertiary-care teaching hospital with 334 acute-care beds. Among those who were admitted to UGH through ED, we enrolled patients with sepsis and septic shock who is older or equal to 18 years old. Patients were screened in the ED and assessed the eligibility to this study by investigators.
Sepsis is defined as a life-threatening organ dysfunction caused by the dysregulated host response to infection. Organ dysfunction was defined as an acute change in total sequential organ failure assessment (SOFA) score ≥ 2 points consequent to the infection. 7 The baseline SOFA score was assumed to be zero in patients whose preexisting organ dysfunction was unclear. 8 Of these, septic shock was defined as a requirement of vasopressors to maintain a mean arterial pressure of ≥ 65 mm Hg and a serum lactate level >2 mmol/L in the absence of hypovolemia.
Exclusion criteria included patients who did not wish to receive the intensive care, patients with congenital heart diseases, pregnant patients, patients younger than 18 years old, patients with a past history of cardiac surgery, patients with sepsis that is second or more in the same hospitalization, patients underwent extracorporeal membrane oxygenation.
All patients enrolled in this study underwent comprehensive transthoracic echocardiograms (TTEs) within 24 h after meeting the criteria of sepsis (day1), as well as day2, and day3. To assess the reversibility of echocardiographic parameters, TTE was also performed on day7 and day14-28. Therefore, comprehensive TTE was performed up to five times in total throughout the hospitalization. TTE was performed using GE Vivid 9E (GE Healthcare, Horten, Norway) by registered medical sonographers certificated by The Japanese Society of Ultrasonics in Medicine who had experienced at least three hundred comprehensive TTEs under supervision. LVEF was assessed using the modified Simpson method. E/A ratio, E/e" ratio, and deceleration time of mitral E-wave velocity (DcT) were also evaluated. To assess right ventricular function, tricuspid annular plane systolic excursion (TAPSE), and peak tricuspid annular velocity (S') were also obtained.
Ea was calculated by the following formula: where SBP is systolic blood pressure and SV represents stroke volume. 9 Ees was calculated using the single-beat method as follows: where DBP is diastolic blood pressure and End(est) is the estimated normalized ventricular elastance. 10 End(est) was calculated by the following formula: where tNd is the ratio of pre-ejection period (R-wave to flow onset) to total systolic period (R-wave to end-flow). 6 V-A coupling was calculated as Ea/Ees and V-A decoupling was defined as Ea/Ees ≥ 1.36. Systemic vascular resistance (SVR) was calculated as follows: [mean arterial pressure (MAP)central venous pressure (CVP)] × 80 / [cardiac output (CO)], where CVP was estimated based on the American Society of Echocardiography recommendations. 11 The patients' characteristics were summarized by a median with a 25th and 75th percentiles for continuous variables. Frequencies and percentages were shown for categorical variables. The patients' characteristics were analyzed with Wilcoxon signed-rank test for continuous variables and the Chi-squared test for categorical variables. We also performed the Friedman test to evaluate if there is any change in Ea/Ees through the clinical course.
All statistical analyses were performed using R (version 3.6.1; R Foundation for Statistical Computing, Vienna, Austria).
Investigators explained the above methodology of this study to the screened patients and obtained both verbal and written consent. All authors had full access to all study data and analyses. This study was approved by the Research and Institutional Review Committee of the UGH (IRB number: 2017B004).

Results
In total, 359 patients were diagnosed with sepsis (267 patients) or septic shock (92 patients) in the study period. Of these, 71 patients with sepsis or septic shock were enrolled.  Table 2 and Supplemental Table 1. The scheme of pressure -volume loop relationship with Ea and Ees on day-1 was shown in Figure 1. Subgroup analyses in patients with sepsis only and septic shock only were shown in Supplemental Table 2 and 3. The comparison between patients with sepsis and septic shock were shown in Supplemental Table 4.
In the Friedman test, Ea/Ees did not significantly change throughout the clinical course (p = 0.618). V-A decoupling was not associated with significantly higher in-hospital mortality compared to those without V-A decoupling (18.8% vs. 12.8%, p = 0.68). ICU length of stay was significantly longer in patients with V-A decoupling (4 days vs. 2 days, p = 0.01), as shown in Table 3.

Discussion
In this study, we reported that V-A decoupling is commonly seen in patients with sepsis and septic shock. Ea/Ees did not significantly change throughout the clinical course. This study also demonstrated the main drivers of V-A decoupling. While low Ees was the main driver of V-A decoupling in patients with sepsis, both low Ees and high Ea were main drivers of V-A decoupling in patients with septic shock. In patients with early septic shock and V-A decoupling, both low SV and high SBP contributed to high Ea. Unlike patients with sepsis and V-A decoupling, SV stayed lower in patients with septic shock and V-A decoupling than those without V-A decoupling throughout the clinical course. This suggests that persistent V-A decoupling in septic shock may be due to persistent reduced SV. Furthermore, in patients with septic shock, high Ea was also a main driver of V-A decoupling especially in the early clinical course, which was also consistent with significantly higher SVR in V-A decoupling group. These results may suggest the need for investigation of the optimal blood pressure target in this population.
It was previously reported that V-A coupling has a greater dependence on Ea than Ees 12 and increased Ea was the main reason for V-A decoupling in patients with sepsis and septic shock. 5 However, this study suggests that both Ea and Ees are responsible for V-A decoupling. In this study, abnormality in the value of Ea and Ees was already seen on day 1 but the cause of these alterations appeared to be changing through the clinical course. Without question, earlier hemodynamic stabilization will improve outcomes, and knowing Ea and Ees is important to personalize and optimize the therapeutic strategy for each patients' condition. This study adds important knowledge regarding how V-A decoupling occurs and which components are commonly causing V-A decoupling in each time phase.
In this study, high SBP, unlike DBP, contributed to both increased Ea and decreased Ees, especially in the early phase. Previously it was suggested that the increases in Ea are generally induced by vasopressors such as norepinephrine and the consequent increase in arterial tone. 13 In this study, patients with V-A decoupling received a significantly higher dose of norepinephrine while the target of mean arterial pressure was set at 65 mm Hg for all patients regardless of cardiac function. This may suggest that personalization of target blood pressure may be necessary to optimize the interaction between cardiac function and afterload. In addition, E/e" was higher in patients with V-A decoupling suggesting an elevated left ventricular filling pressure. On the other hand, a study suggested V-A decoupling defined as Ea/Ees > 1.4 predicted the fluid responsiveness and fluid challenge reduced Ea. 14 While we need to be careful about fluid overload in patients with elevated E/e", the combination of V-A decoupling, especially high Ea, and normal E/e" may be useful to identify those who can potentially optimize the cardiovascular interaction by fluid administration.
Given that Ea can be reduced with fluid administration, dynamic Ea, which is defined as the ratio of pulse pressure variation and stroke volume variation may also be a useful parameter to predict the fluid responsiveness. 15 Ees generally depends on the reduction of myocardial contractility and a recent study reported that norepinephrine increased Ees more than Ea, especially in those with higher Ea/Ees, resulting in the normalization  of Ea/Ees. On the other hand, those who did not increase stroke volume with norepinephrine had significantly lower Ea/Ees. 16 In this previous study, Ees was significantly lower and Ea/Ees was significantly higher in responders to norepinephrine while Ea was similar between responders and non-responders. This suggests that norepinephrine is a reasonable choice for those with V-A decoupling due to low Ees. While the current Surviving Sepsis Campaign Guidelines suggest the fluid administration and norepinephrine to resuscitate hemodynamically unstable septic patients, 17 there is no specific guidance regarding who can benefit from fluid or vasopressors more. Based on the above discussion, those with V-A decoupling mainly due to high Ea may respond fluid and those with V-A decoupling mainly due to low Ees may respond to norepinephrine. Although inodilators have been traditionally used to increase Ees, the issue with inotropes is an increase in tachyarrhythmia, 18 which is known to be associated with V-A decoupling 13 and worse outcomes. 19 Therefore, careful selection of patients who would benefit from inodilators is required.
Morelli et al reported that the use of esmolol was associated  with significantly increased stroke work index, 20 which may lead to increased Ees. Another prospective study by Morelli et al suggested that the use of esmolol was also associated with significantly reduced Ea. 21 These results suggest that patients with V-A decoupling with increased Ea and even decreased Ees may benefit from beta-blockers. Furthermore, the recent meta-analysis suggested that the use of ultra-short acting beta-blocker including esmolol and landiolol was associated with lower mortality in patients with sepsis. 22 On the other hand, Levy et al suggested that fast esmolol titration in the very early phase of septic shock was associated with increased hemodynamic instability. 23 A recent study suggested the heterogeneous cardiovascular profile and responses to the intervention and reported how interventions including fluid administration, vasopressors, and inotropes affect Ea and Ees in patients with septic shock. 24 In this study, only 31.3% of patients with V-A decoupling had past history of heart failure suggesting that majority of cardiovascular alteration is due to sepsis. Therefore, the assessment of cardiovascular interaction needs to be taken place in the early course of sepsis regardless of previous cardiac function. Recently, Guarracino released a mobile application (iElastance: https://apps.apple.com/us/app/ ielastance/id556528864) to calculate Ea and Ees and this will make the calculation of these indices easier at the bedside. Further research should focus on establishing the precise algorism of hemodynamic resuscitation. Previous meta-analyses reported that reduced LVEF was reported not to be significantly associated with higher mortality. 25,26 One of the challenges of LVEF is that LVEF is greatly affected by afterload, especially in patients with reduced cardiac function. On the other hand, Yan et al previously reported a significant association between V-A decoupling and higher mortality. 5 Abnormalities of other parameters that takes an interaction between ventricular function and afterload such as Afterload-related cardiac performance, and myocardial performance index (also known as Tei index) are also shown to be associated with higher mortality. 27,28 Therefore, the interaction between ventricular function and afterload may better to be taken into account when discussing the association between cardiac function and clinical outcomes. In this study, in-hospital mortality was not significantly different between patients with V-A decoupling and without. This is likely because the number of patients who died was small in this study. While we could not assess whether V-A decoupling is independently associated with higher in-hospital mortality due to the small number, it appeared that patients with V-A decoupling had higher severity score as well as higher vasopressor doses. Therefore, the future study needs to clarify the independent association between V-A decoupling and mortality in patients with sepsis. Further, ICU-length of stay was significantly longer in those with V-A decoupling than those without. While this emphasizes the importance of those with V-A decoupling, this may also explain the reason why the prevalence of V-A decoupling increased in the late clinical course (day-14 to −28) since those without V-A decoupling were likely to be discharged earlier than those with V-A decoupling.
The strength of this study is that we followed Ea/Ees on day1, day2, day3, and thereafter. To our best knowledge, this is the first study that chronologically assessed Ea/Ees in patients with sepsis and septic shock. While SICM defined by reduced LVEF was reported to recover in 7-10 days, 29 it has been unclear whether Ea/Ees can be reversible in the same timeframe or not. In fact, another parameter, the global longitudinal strain of the left ventricular wall, was previously reported to be irreversible even after recovery from sepsis, unlike LVEF. 30 Therefore, there has been a possibility that V-A decoupling could remain even after recovery from sepsis. On the other hand, while a longer ICU length of stay with V-A decoupling may explain the higher prevalence of V-A decoupling, those with a history of heart failure or myocardial infarction had significantly higher prevalence of V-A decoupling on day1. These patients are likely to continue to have cardiovascular dysfunction even after the recovery from sepsis. Further, this study demonstrated no significant changes in Ea/Ees throughout the clinical course. Therefore, it is also possible that these persistent V-A decoupling seen in this study was due to patient's underlying cardiovascular conditions. Nonetheless, since sepsis is known to be associated with an increased risk of cardiovascular events even after the recovery from sepsis, 31 these factors may need to be monitored in the follow-up.
There are some limitations in this study. First, the number of included patients was small. This might have affected the statistical significance of the difference between patients with Ea/Ees ≥1.36 and <1.36. Therefore, a further larger study is warranted. Second, the accuracy of echocardiography greatly depends on the skill of an operator. Therefore, inter-operator variability might have affected the study result. However, all the operators in this study were very well-trained certified sonographers and we believe such risk was minimized. Third, consistent Ea and Ees cannot be measured when patients have an arrhythmia. As previously reported, it is not uncommon for patients with sepsis to have tachyarrhythmia. 32 Therefore, the utility of V-A coupling is limited in such patients. Fourthly, we used an estimated CVP based on inferior vena cava (IVC) diameter and variation to calculate SVR. However, in patients who are mechanically ventilated, the IVC diameter does not necessarily correlate CVP. 33 While only one-third of included patients underwent mechanical ventilation in the entire population, about half of patients with V-A decoupling underwent mechanical ventilation. Therefore, this might have affected the results. Furthermore, in this study, it was difficult to obtain previous V-A coupling. Thus, some patients might have had just baseline cardiovascular function even if they presented abnormal cardiac function and it was impossible to note whether it was new-onset sepsis-related cardiovascular dysfunction or not.
In conclusion, we reported that V-A decoupling is commonly seen in patients with sepsis and septic shock (25.4% on day 1). In patients with V-A decoupling, both Ea and Ees were significantly altered. Future studies need to focus on establishing the precise algorism to optimize the interaction between cardiac function and afterload.

Ethics Approval and Consent to Participate
This study was approved by Urasoe General Hospital Institutional Review Board (IRB number: 2017B004). This prospective study complied with the Declaration of Helsinki. Written informed consent was obtained from patients or their next-of-kin or surrogate if designated in advance.

Consent for Publication
Consent for publication was obtained by the investigators (The information about publication was explained in the consent form).

Availability of Data and Materials
The datasets used and analyzed during the current study are available from the corresponding author on reasonable requests.

Author's Contributions
M.N. is responsible for the conception of the study design, data collection, interpretation of the analysis, writing of the draft, and critical revision of the manuscript. R.S. contributed to analysis, interpretation of the analysis, writing of the draft, K.T., T.N., and AM contributed substantially to data collection. S.U. supervised the conception of the study design, interpretation of the analysis drafting and critical revision of the manuscript. All authors approved the submission of the final manuscript.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.