Discordance between lactic acidemia and hemodynamics in patients with advanced heart failure

Abstract Background Elevated lactic acid (LA) levels carry a poor prognosis in patients with shock. Data are lacking on the significance of LA levels in patients with acute decompensated heart failure (ADHF). Hypothesis This study assessed the relationship between LA levels, hemodynamics and clinical outcomes. Methods This was a retrospective analysis of registry data of 100 advanced heart failure patients presenting for right heart catheterization (RHC) for concern of ADHF. LA levels (normal ≤2.1 mmol/L) were obtained prior to RHC; no significant changes in therapy were made between LA collection and RHC. Results Median age was 58 (47.3, 64.8) years; 57% were receiving inotropes prior to RHC. Median pulmonary capillary wedge pressure (PCWP) and cardiac index (CI) were 28 (21, 35) mmHg and 2.0 (1.7, 2.5) L/min/m2, respectively. Eighty patients had normal LA prior to RHC. There was no correlation between LA levels and PCWP (R = 0.09, p = .38); 63% of the normal LA group had a PCWP >24 mmHg. There was a moderate inverse correlation between LA and CI (R = − 0.40; p < .001); 58% of the normal LA group had a CI <2.2 L/min/m2. Thirty‐day survival free of death/hospice, inotrope dependence, progression to heart transplant/left‐ventricular assist device implant was comparable between the normal and elevated LA groups (28% vs. 20%; p = .17). Conclusion In patients presenting with ADHF, normal LA levels do not exclude the presence of depressed CI (a hemodynamic criteria for cardiogenic shock) and may not offer accurate risk stratification. Invasive hemodynamics should not be delayed based on normal LA levels alone.


| INTRODUCTION
Elevated lactic acid (LA, >2.1 mmol/L) levels have long been used as a reliable marker of end-organ hypoperfusion and clinical deterioration and serves as a useful clinical prognostication biomarker in patients presenting with sepsis. [1][2][3] Type A lactic acidemia occurs during anaerobic glycolysis resulting in peripheral tissue hypoxia, 4 while Type B lactic acidemia occurs in disorders where tissue hypoxia is absent including malignancy, metformin overdose, seizures and B 2 -agonist use. In the cardiac intensive care setting, LA levels are often used as surrogate biomarker to monitor response to inotropes or mechanical circulatory support (MCS) in patients presenting with cardiogenic shock. 5 LA monitoring is also utilized as a clinical metric (when >4 mmol/L) to upgrade status for heart transplantation when hemodynamics are unavailable in patients with worsening chronic heart failure (HF). 6 Cardiogenic shock is conventionally defined by sustained hypotension (<90 mmHg) and elevated intracardiac filling pressures manifesting with signs of end-organ hypoperfusion. Seemingly, it would be plausible to conclude a high prevalence of elevated LA levels in the setting of overt cardiogenic shock. However, cardiogenic shock may present in many phenotypes, including vasodilatory, euvolemic, normotensive, right ventricular, and biventricular variants for which accepted definitions do not apply and treatment strategies differ. 7,8 In a single-center study, Adamo et al. showed a low prevalence of elevated LA in advanced HF patients prior to left-ventricular assist device (LVAD) implantation. 9 This poor correlation between LA and decompensated hemodynamics in advanced HF patients questions the prognostic ability of LA. With the broad spectrum of clinical presentations encompassing decompensated heart failure (DHF) and cardiogenic shock, a re-examination of objective measurements of end-organ perfusion and invasive hemodynamics is needed. We therefore sought to assess the relationship between LA levels, hemodynamics and clinical outcomes in patients with advanced heart failure presenting with clinical decline.

| METHODS
This was a retrospective analysis of an invasive hemodynamics registry of advanced heart failure patients where providers were tasked to predict hemodynamics prior to right heart catheterization (RHC). 10 Data for this current analysis were abstracted from this registry. The study took place at the University of Chicago Medical Center with patient data was collected between 2016 and 2018. Patients were eligible for inclusion if they had a history of heart failure with reduced ejection fraction, were >18 years of age, had a recorded admission LA level, and were presenting in either the outpatient or inpatient setting with apparent clinical decline or decompensation, as defined by worsening symptoms or objective markers consistent with worsening congestion or perfusion. Patients were excluded if they were mechanically ventilated or presented following cardiac arrest. Patients presenting with an acute coronary syndrome were excluded. Acute HF was defined as new onset (<1 month) HF with left ventricular (LV) cavity end-diastolic dimensions 11 <5.3 cm for women and <5.9 cm for men. 11 Outpatients were screened for study inclusion if they were evaluated in the clinic setting and sent directly for RHC if deemed necessary by the treatment team, specifically for precision-guided management of HF. Pre-RHC laboratory measurements were collected in the pre-procedure holding area for patients admitted from the clinic. Inpatients were eligible for study inclusion if they presented to the emergency room or were transferred from an outside institution to either the inpatient floor or cardiac care unit for man-  area (m 2 )], stratified above and below the median. 16 Pulmonary artery pulsatility index ([pulmonary artery systolic pressure -pulmonary artery diastolic pressure]/[right atrial pressure]; PAPi) was calculated, stratified above and below the median, and correlated to LA levels. 17,18 A combined 30-day composite endpoint of freedom from death or hospice discharge, inotrope dependence, heart transplantation or LVAD implantation was analyzed by Kaplan-Meier analyses and compared between normal and elevated LA groups using the log-rank test. Outcomes at 30 days for patients discharged were adjudicated given uniform follow-up clinical visits for all patients (non-hospice) within the registry. Univariate Cox proportional hazard ratio analyses were performed to assess for significant predictors associated with 30-day survival with medical therapy alone. Statistical significance was set at 0.05, and all tests were two-tailed.

| RESULTS
Patient characteristics for the overall cohort are shown in Table 1.
Out of 128 patients within registry, 100 patients meet criteria for inclusion ( Figure 1). Median patient age was 58 (47.3, 64.8) years, 70% were male, 39% were Caucasian. Median left ventricular ejection fraction was 22% 18,28 ; mean left ventricular end-diastolic dimension was 6.4 (1.3) cm. Median LA prior to RHC was 1.6 (1, 2.1) mmol/L, with a range of 0.6-7.4 mmol/L (Appendix Figure 1). 57% of patients were on either inotropes and/or vasopressors prior to RHC. 7% (N = 7) of patients in total had percutaneous mechanical support devices (N = 6, Intra-aortic balloon pump; N = 1, Impella) with inotropes and/or vasopressors prior to RHC. Median RA pressure was 15.5 (10,22) ; PCWP (21,35)      LA is predominantly produced in skeletal muscles and under aerobic conditions, is cleared by the liver. Under aerobic conditions, pyruvate is the end produce of glycolysis, while under anaerobic conditions, LA is an end-product of glycolysis and a substrate for gluconeogenesis. 19 During maximal exercise where there is a shift from aerobic to anaerobic metabolism, LA production rises significantly.
Macrocirculatory and microcirculatory dysfunction leading to tissue hypoxia results in excess LA production, and decreased clearance due liver dysfunction from systemic hypoperfusion. 4,20 Several studies have demonstrated the association of elevated LA levels or poor clearance of LA with mortality in both ICU populations and those presenting with sepsis. 1,21,22 However, only limited data have been published on the association of LA levels and cardiogenic shock or DHF, both of which are associated with high short-term mortality. 18,[23][24][25] Cardiogenic shock is conventionally defined by a systolic blood pressure less than 90 mmHg for greater than 30 min with elevated left ventricular filling pressures and clinical and serologic signs end-organ hypoperfusion, including elevations in serum LA. [26][27][28] The findings of clinical extremis are most likely to be present following acute myocardial infarction but may not be present in patients with acute on chronic heart failure presentations. Nonetheless, these patients may have grossly abnormal hemodynamics including a depressed CI and elevated PCWP while having normal serum LA. In chronic, advanced HF patients, physiologic adaptation to grossly elevated filling pressures may occur without any acute perturbations in end-organ function. This was confirmed in a prior study of, nearly uniformly inotrope-dependent end-stage HF patients undergoing LVAD implantation who had RHC prior to surgery with corresponding LA levels measured; only 28% of the subgroup with CI ≤2.2 L/min/m 2 had elevated LA levels 9 and also in another study of patients with DHF where a modest inverse correlation between LA and Fick CI was observed but not with PCWP. 29 Furthermore, the authors observed that elevated LA levels were associated with other markers of abnormal end-organ perfusion and worsened INTERMACS classification, supporting the theory that acute on chronic HF presentations with elevated LA may be a late event. Our study is unique from prior evaluations of LA levels in acute decompensated heart failure (ADHF) given the heterogeneous nature of presentations with varying clinical trajectories,~60% on inotropes and detailed correlative evaluation of LA with several hemodynamic variables and surrogate indices. Despite a depressed CI in these subgroups, oxygen delivery to peripheral tissues in resting states is likely maintained. However, these patients are unlikely to meet metabolic demands in moderate and above physical activity as reflected in their poor functional class.
We observed that normal LA levels are common and often discordant with abnormal hemodynamics in a cohort of predominantly chronic, advanced heart failure patients presenting with DHF. Nearly 50% of patients had normal LA levels in the setting of PCWP >24 mmHg and CI ≤2.2 L/min/m 2 . In the current study, we examined three surrogate hemodynamic indexes: Ratio of SBP to PCWP, CPI and proportional pulse pressure. SBP/PCWP<4 is associated with inhospital mortality and escalation to mechanical support in patients with acute coronary syndromes. 15 CPI is associated with progression T A B L E 2 Cox proportional hazard ratio analyses for 30-day risk of death, hospice, inotrope dependence, heart transplantation or LVAD implantation to heart transplant, LVAD and increased all-cause mortality in advanced heart failure patients, 16  This cohort was a highly-selected, convenience sample comprised predominantly of advanced HF patients, and the results may not be applicable to other populations. Accurate comparison of baseline covariates and clinical outcomes between the normal and elevated LA groups is limited by an unbalanced sample size. Furthermore, subclassification of outcomes including progression to heart transplantation and LVAD implantation is center and provider dependent, and therefore may not be generalizable. The majority of the patients in the advanced heart failure cohort met the primary endpoint which clearly affected the ability for assessment of clinical risk predictors which would have had an association with the composite outcome of interest. This therefore undoubtedly affected the ability to perform multivariate modeling. We calculated an estimated Fick CI 12 to establish the cut-off of 2.2 L/min/m 2 which can be flawed when compared to direct measurement, 38 though may be more favorable compared to thermodilution-derived CI due to the high prevalence of tricuspid regurgitation in this population.

| CONCLUSIONS
In patients with advanced HF presenting with DHF, a normal LA on presentation was remarkably common and may indicate that the basic metabolic demands of the body are still being met despite severely reduced cardiac function. However, normal LA levels may not accurately discriminate the clinical severity of presentation in advanced HF patients as defined by invasive hemodynamic. Furthermore, patients of this type may be in overt cardiogenic shock with a normal LA, along with having a high short-term mortality risk and need for advanced heart failure therapies. Elevated LA levels in these patients, however, carry more certainty of immediate high clinical acuity and should be triaged appropriately to better mitigate adverse outcomes.

CONFLICT OF INTEREST
Nir Uriel receives grant support and consulting fees from Abbott Healthcare, Leviticus Cardio, and Medtronic. Daniel Burkhoff has received an unrestricted educational grant from Abiomed. Valluvan Jeevanandam receives consulting fees from Abbott Healthcare.