ORIGINAL RESEARCH Association Between Norepinephrine Levels and Abnormal Iron Status in Patients With Chronic Heart Failure: Is Iron Deficiency More Than a Comorbidity? Pedro Moliner, MD; Cristina Enjuanes, MD, PhD; Marta Tajes, PhD; Miguel Cainzos-Achirica, MD, MPH; Josep Lup on, MD, PhD; Alberto Garay, MD, PhD; Santiago Jimenez-Marrero, MD; Sergi Yun, MD; Nuria Farre, MD, PhD; Merce Cladellas, MD, PhD; Carles Dıez, MD; Jose Gonzalez-Costello, MD; Josep Comin-Colet, MD, PhD Background-—Mechanisms underlying iron homeostasis dysregulation in patients with chronic heart failure remain unsettled. In cardiomyocyte models, norepinephrine may lead to intracellular iron depletion, but the potential association between catecholamines (sympathetic activation markers) and iron metabolism biomarkers in chronic heart failure is unknown. Methods and Results-—In this cross-sectional analysis, we studied the association between plasma norepinephrine levels and serum iron status biomarkers indicating iron storage (ferritin), iron transport (transferrin saturation), and iron demand (soluble transferrin receptor) in a prospective cohort of 742 chronic heart failure patients (mean age, 7211 years; 56% male). Impaired iron status was defined as ferritin <100 lg/L or transferrin saturation <20%. Impaired iron status was observed in 69% of patients. In multivariate models, greater norepinephrine levels were associated with impaired iron transport (transferrin saturation <20%, odds ratio=2.28; 95% CI [1.19–4.35]; P=0.013), but not with impaired iron storage (ferritin <100 lg/L, odds ratio=1.25; 95% CI [0.73–2.16]; P=0.415). Norepinephrine was a significant predictor of increased iron demand (soluble transferrin receptor, standardized b-coefficient=0.12; P=0.006) and low transferrin saturation (standardized b-coefficient= 0.12; P=0.003). However, norepinephrine levels were not associated with iron or ferritin levels (P>0.05). Adjusted norepinephrine marginal means were significantly higher in patients with impaired iron status compared with those with normal iron status (528 pg/mL [505–551] versus 482 pg/mL [448–518], respectively; P=0.038). Conclusions-—In chronic heart failure patients, increased sympathetic activation estimated with norepinephrine levels is associated with impaired iron status and, particularly, dysregulation of biomarkers suggesting impaired iron transport and increased iron demand. Whether the relationship between norepinephrine and iron metabolism is bidirectional and entails causality need to be elucidated in future research. ( J Am Heart Assoc. 2019;8:e010887. DOI: 10.1161/JAHA.118.010887.) Key Words: anemia • chronic heart failure • iron • iron deficiency • norepinephrine • sympathetic nervous system I ron deficiency is associated with worse quality of life, functional capacity, and prognosis in patients with chronic heart failure (CHF), regardless of the presence of anemia.1–6 Consequently, current guidelines recommend assessing the iron profile of patients with CHF and treating them with intravenous iron, when appropriate.7–10 Nevertheless, the From the Heart Failure Unit, Department of Cardiology, Hospital Universitari Germans Trias i Pujol, Badalona, Spain (P.M., J.L.); Department of Medicine, Universitat Autonoma de Barcelona, Spain (P.M., J.L., N.F., M.C.); Community Heart Failure Program, Cardiology Department, Bellvitge University Hospital (C.E., M.C.-A., A.G., S.J.-M., S.Y., J.C.-C.), Cardiovascular Diseases Research Group, Bellvitge Biomedical Research Institute (IDIBELL; C.E., M.C.-A., A.G., S.J.-M., S.Y., C.D., J.G.-C., J.C.-C.), and Advanced heart Failure and Heart Transplant Unit, Department of Cardiology, Bellvitge University Hospital (C.D., J.G.-C.), Hospitalet de Llobregat, Barcelona, Spain; Heart Diseases Biomedical Research Group, Program of Research in Inflammatory and Cardiovascular Disorders, Hospital del Mar Biomedical Research Institute (IMIM), Barcelona, Spain (C.E., M.T., N.F., J.C.-C.); Johns Hopkins Ciccarone Center for the Prevention of Heart Disease, Department of Cardiology, Johns Hopkins Medical Institutions, Baltimore, MD (M.C.-A.); School of Medicine and Medical Sciences, Universitat Internacional de Catalunya, Sant Cugat del Valles, Barcelona, Spain (M.C.-A.); Heart Failure Program, Department of Cardiology, Hospital del Mar, Parc de Salut Mar, Barcelona, Spain (N.F.); Department of Cardiology, Hospital del Mar, Parc de Salut Mar, Barcelona, Spain (M.C.); Department of Clinical Sciences, School of Medicine, Universitat de Barcelona, Spain (J.C.-C.). Correspondence to: Josep Comin-Colet, MD, PhD, Hospital Universitari de Bellvitge, Department of Cardiology, 19th Floor, Feixa Llarga s/n, 08907 Hospitalet de Llobregat, Barcelona, Spain. E-mail: jcomin@bellvitgehospital.cat Received October 8, 2018; accepted January 9, 2019. ª 2019 The Authors. Published on behalf of the American Heart Association, Inc., by Wiley. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non- commercial and no modifications or adaptations are made. DOI: 10.1161/JAHA.118.010887 Journal of the American Heart Association 1 Sympathetic Activation and Iron Status in CHF Moliner et al ORIGINAL RESEARCH Study Design and Study Population Clinical Perspective The DAMOCLES (Definition of the neuro-hormonal activation, What Is New? myocardial function, genomic expression and clinical out- comes in heart failure patients) study was a single-center, • The exact mechanisms by which chronic heart failure patients develop iron deficiency are still unknown. observational, prospective cohort study of 1236 consecutive • To the best of our knowledge, we demonstrate, for the first patients diagnosed with CHF. The methodology of the time, an interplay between raised sympathetic nervous DAMOCLES has been published previously by our group.2,17 system activity and systemic iron deficiency in patients with Briefly, for inclusion, patients had to be diagnosed with CHF chronic heart failure and, particularly, with those biomarkers according to the European Society of Cardiology diagnostic that suggest impaired iron transport (transferrin saturation criteria, had at least 1 recent acute decompensation of CHF <20%) and increased iron demand (raised soluble transferrin requiring intravenous diuretic therapy (either hospitalized or in receptor levels). the day care hospital), and had to be in stable condition at the time of study entry. Exclusion criteria were: significant What Are the Clinical Implications? primary valvular disease, clinical signs of fluid overload, • Our results may support the hypothesis that iron deficiency pericardial disease, restrictive cardiomyopathy, hypertrophic might not just be a comorbidity, but may also be a key cardiomyopathy, hemoglobin levels <8.5 g/dL, active malig- element in the pathophysiological sequence leading to, and nancy, and chronic liver disease. The study was approved by promoting the progression of, chronic heart failure. the local committee of ethics for clinical research and was conducted in accord with the principles of the Declaration of Helsinki. All patients gave written informed consent before mechanisms underlying dysregulation of iron homeostasis in study entry. patients with CHF are still not well established. For the present analysis, norepinephrine levels were Sympathetic activation, which entails increased serum evaluated in all consecutive participants included in the levels of norepinephrine, is one of the main therapeutic targets DAMOCLES study between June 2004 and January 2011. in CHF because of its deleterious long-term effects,7,8,11–15 Patients included in the DAMOCLES study after this period did particularly in patients with reduced left ventricle ejection not have norepinephrine levels measured and thus were fraction (LVEF). In this context, recent studies have shown that excluded from this analysis. in cardiomyocyte models, norepinephrine leads to intracellular iron depletion.16 These results in cellular models suggest a role of sympathetic activation in the development of iron home- Baseline Evaluation: Catecholamines, Iron ostasis dysregulation. However, the association between Parameters, and Other Laboratory Measurements sympathetic activity and iron status has not been evaluated Methods of blood collection and management, as well as the so far at the patient level. quantification of norepinephrine levels performed in the The aim of this study was thus to assess the potential baseline DAMOCLES study visit, have been previously association between sympathetic activation, as measured reported.17 In summary, patients were resting in a supine using plasma norepinephrine levels, and several biomarkers of position in a quiet room for 30 to 60 minutes after venous serum iron homeostasis, including ferritin as a maker of iron cannulation. Tubes with blood samples were immersed in availability, transferrin saturation (TSAT) as a marker of iron melting ice and frozen until they were processed. Levels of supply, and the soluble transferrin receptor (sTfR) as a marker norepinephrine were measured from 1.5 mL of plasma by of iron demand, in a large population of consecutive patients high-resolution liquid chromatography. Norepinephrine analy- with CHF followed in a specialized, multidisciplinary CHF unit, sis had a coefficient of variation of 8.7%. Serum N-terminal in which all these measurements were available. pro-b-type natriuretic peptide levels were measured in pg/mL using an immunoassay based on chemiluminescence using the Elecsys System (Roche, Indianapolis, IN). Serum iron (mg/ Methods dL) was measured using spectrophotometry; serum ferritin Data are available upon request from a third party. Because of (ng/mL) and transferrin (mg/dL) were measured using the sensitive nature of the data collected for this study, immunoturbidimetry. TSAT was estimated using the formula: reasonable requests to access the data set from qualified TSAT=serum iron (mg/dL)/[serum transferrin (mg/dL) researchers trained in human subject confidentiality protocols 91.25]. Iron status was also assessed by measuring serum may be sent to Bellvitge Biomedical Research Institute at the soluble transferrin receptor (sTfR; in mg/L) levels using an corresponding author. enzyme immunoassay. Hemoglobin (g/dL) was measured with DOI: 10.1161/JAHA.118.010887 Journal of the American Heart Association 2 Sympathetic Activation and Iron Status in CHF Moliner et al ORIGINAL RESEARCH impedance laser colorimetry. Glomerular filtration rate was higher levels of norepinephrine as an independent variable estimated from serum creatinine using the formula of the and levels of each of the iron status biomarkers as the Modification of Diet in Renal Disease Study Group equation.18 dependent variables. Multivariable models used step-wise forward conditional methods and were adjusted for the predictors that showed a significant association with impaired Study Definitions iron status in the bivariate logistic regression analyses. Impaired iron status was defined as ferritin <100 ng/mL or Finally, general linear models were used to calculate TSAT <20%. To explore the association between nore- adjusted marginal means and 95% CIs of norepinephrine pinephrine and different components of the iron pathway, according to different iron and anemia states. All general we also defined impaired iron storage as ferritin <100 ng/ linear models were adjusted for factors associated with mL and impaired iron transport as TSAT <20%. Increased increased norepinephrine levels. Factors associated with iron demand (high sTfR) was defined as levels of sTfR raised norepinephrine levels have been described previously >75th percentile of its distribution in the study population. by our group.17 Anemia was defined using the World Health Organization All statistical tests and CIs were constructed with a type 1 criteria (cut-off values of 13 g/dL in men and 12 g/dL in error alpha level of 5% with no adjustments for multiplicity, women).19 and P<0.05 were considered statistically significant. SPSS (version 22.0; IBM, Armonk, NY) and R software (version 3.0.1; R Foundation for Statistical Computing, Vienna, Austria) Statistical Analyses were used for statistical analyses. This was a cross-sectional analysis using the baseline data from the DAMOCLES study. Demographic and clinical char- acteristics and laboratory test results were summarized using Results basic descriptive statistics, both in the overall cohort as well as stratified by norepinephrine tertiles. For quantitative Study Population variables, the arithmetic mean (and SD) or median (and Of the 1236 CHF patients included in the DAMOCLES study, interquartile range) were reported as appropriate, and P 494 did not have baseline levels of norepinephrine measured values were obtained using ANOVA and Kruskal–Wallis tests, and were excluded from the present analyses. This defined a respectively. For qualitative variables, number and percent- final study population of 742 patients. Serum iron status ages within specified groups were calculated, and P values parameters were available in all DAMOCLES study were derived using v2 tests. participants. When necessary, natural logarithm transformation was used to fit skewed continuous variables into normal distribu- tions. Specifically, norepinephrine was modeled in 2 ways for Baseline Characteristics of the Study Participants the analyses: as a continuous exposure (log-transformed), as Baseline characteristics of the study participants, overall well as a dichotomous exposure (“high norepinephrine” levels and by norepinephrine tertiles, are shown in Table 1. defined as norepinephrine >90th percentile [1050 pg/mL] of Overall, mean age was 72 years, and 56% participants were the distribution in the study population, as compared with male. Mean LVEF was 44%, median N-terminal pro-b-type ≤90th percentile). natriuretic peptide levels were 1547 pg/mL, and 45% of To assess the association between sympathetic activation, patients had a New York Heart Association functional class as measured using norepinephrine levels, and the different of III or IV. Median levels of norepinephrine were 523 iron status biomarkers, we first used logistic and linear (351–730), and 515 (69%) patients had impaired iron regression to assess the crude (unadjusted) associations status. between norepinephrine and each of the relevant iron status By norepinephrine tertiles, patients in the upper tertile of biomarkers. Also, generalized additive models were used to norepinephrine (higher sympathetic activation) were older, graphically display the relationship between log nore- more frequently male, and had a worse CHF clinical profile, pinephrine serum levels and log TSAT, log sTfR, and log with higher N-terminal pro-b-type natriuretic peptide levels, ferritin, respectively. Unadjusted logistic and linear regression lower body mass index, and worse New York Heart Association analyses were also used to assess the bivariate associations functional class than their lower norepinephrine tertile coun- between each of the predictors included in Table 1 and each terparts. Also, the higher the norepinephrine tertile, the higher of the iron status biomarkers. the prevalence of all iron abnormalities. Interestingly, LVEF, Multivariable-adjusted linear and logistic regression mod- systolic blood pressure, and heart rate did not differ between els were used to evaluate the adjusted associations between categories of norepinephrine levels. DOI: 10.1161/JAHA.118.010887 Journal of the American Heart Association 3 Sympathetic Activation and Iron Status in CHF Moliner et al ORIGINAL RESEARCH Table 1. Baseline Characteristics of the Study Population (N=742), Overall and by Norepinephrine Tertiles Overall T1 (<414 pg/mL) T2 (414–654 pg/mL) T3 (>654 pg/mL) n=742 n=248 n=248 n=246 P Value Age, y 7211 6911 7311 7511 <0.001 Sex, male, n (%) 418 (56) 137 (55) 134 (54) 147 (60) 0.40 BMI, kg/m 2 286 306 285 275 <0.001 Systolic BP, mm Hg 12523 12724 12424 12322 0.13 Heart rate, bpm 7415 7315 7515 7416 0.62 NT-proBNP, pg/mL 1547 [662–3942] 1124 [489–2460] 1672 [647–3495] 2391 [923–6436] <0.001 2 eGFR, mL/min per 1.73 m 5924 6124 5923 5626 0.08 LVEF, % 4417 4517 4417 4318 0.46 HFpEF, n (%) 278 (38) 91 (37) 95 (38) 92 (37) 0.933 HFmrEF, n (%) 116 (16) 45 (18) 38 (15) 33 (13) 0.346 HFrEF, n (%) 348 (47) 112 (45) 115 (47) 121 (49) 0.655 Etiology of CHF, ischemic, n (%) 303 (41) 112 (45) 96 (39) 95 (39) 0.24 Atrial fibrillation, n (%) 234 (32) 72 (29) 85 (34) 77 (31) 0.45 NYHA class III or IV, n (%) 330 (45) 98 (40) 104 (42) 128 (52) 0.012 Norepinephrine, median (IQR) 523 [351–730] 313 [252–352] 524 [467–592] 858 [730–1111] <0.001 Comorbidities, n (%) Hypertension 582 (78) 190 (77) 195 (79) 197 (80) 0.64 Diabetes mellitus 345 (47) 130 (52) 110 (44) 105 (43) 0.67 CKD (eGFR <60 mL/min/1.73 m ) 2 418 (56) 126 (51) 140 (57) 152 (62) 0.048 Anemia 365 (49) 124 (50) 115 (46) 126 (51) 0.53 COPD 163 (22) 57 (23) 49 (20) 57 (23) 0.59 Obesity 247 (33) 101 (41) 84 (34) 62 (25) 0.001 Impaired iron status 515 (69) 172 (69) 164 (66) 179 (73) 0.278 Impaired iron transport 438 (59) 137 (55) 139 (56) 162 (66) 0.028 Impaired iron storage 280 (38) 104 (42) 88 (36) 280 (38) 0.247 Treatments, n (%) ACEI or ARBs 578 (78) 205 (83) 196 (79) 177 (72) 0.014 Beta-blockers 656 (88) 214 (86) 224 (90) 218 (89) 0.37 Aldosterone antagonists 308 (42) 110 (44) 101 (41) 97 (39) 0.52 Digoxine 96 (13) 34 (14) 38 (15) 24 (10) 0.166 Statins 426 (57) 147 (59) 139 (56) 140 (57) 0.754 Loop diuretics 655 (88) 219 (88) 222 (90) 214 (87) 0.684 Anticoagulants 356 (48) 107 (43) 118 (48) 131 (53) 0.079 ICD 22 (3) 11 (5) 2 (1) 9 (4) 0.046 CRT 10 (1) 4 (2) 4 (2) 2 (1) 0.68 Serum levels Hemoglobin, g/dL 12.41.8 12.51.8 12.61.8 12.51.9 0.87 Ferritin, lg/L [IQR] 142 [73–272] 141 [75–262] 132 [71–251] 155 [70–289] 0.229* TSAT, % 17.8 [12.2–24.5] 18.4 [12.8–25.5] 19.1 [13.1–25.2] 16.1 [11.2–23.3] <0.001* Transferrin 24847 24341 24946 25151 0.116 Raised sTfR, n (%) 133 (25) 38 (24) 38 (20) 57 (30) 0.058* Continued DOI: 10.1161/JAHA.118.010887 Journal of the American Heart Association 4 Sympathetic Activation and Iron Status in CHF Moliner et al ORIGINAL RESEARCH Table 1. Continued Overall T1 (<414 pg/mL) T2 (414–654 pg/mL) T3 (>654 pg/mL) n=742 n=248 n=248 n=246 P Value Ferritin index, [IQR] 0.76 [0.54–1.14] 0.72 [0.53–1.14] 0.72 [0.55–1.06] 0.8 [0.56–1.20] 0.348* Serum iron, mg/dL [IQR] 57 [42–79] 56 [41–77] 61 [43–84] 52 [40–74] 0.007* Endogenous erythropoietin, U/L [IQR] 15 [9–25] 14.1 [8–22] 15 [9–25] 17 [9–26] 0.267* Data are presented as meansSD, medians (25th–75th percentile), or numbers (with percentages), where appropriate. Anemia is defined as hemoglobin level <12 g/dL in women and <13 g/dL in men. Impaired iron transport is defined as TSAT <20%; impaired iron storage is defined as ferritin <100 lg/L; and impaired iron status is defined as ferritin <100 lg/L or TSAT <20%. Raised sTfR = soluble transferrin receptor >75th percentile. ACEI indicates angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; BMI, body mass index; BP, blood pressure; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; CRT, cardiac resincronization therapy; ferritin index, sTfR/log10[ferritin]; HFmrEF, heart failure with mid-range ejection fraction (defined as LVEF 40–49%); HFpEF, heart failure with preserved ejection fraction (defined as LVEF >50%); HFrEF, heart failure with reduced ejection fraction (defined as LVEF <40%); ICD, implantable cardioverter-defibrillator; LVEDD, left ventricular end diastolic diameter; LVEF, left ventricular ejection fraction; NT-proBNP, N-terminal pro-brain-type natriuretic peptide; NYHA, New York Heart Association functional class; sTfR, soluble transferrin receptor; TSAT, transferrin saturation; T1, first tertile; T2, second tertile; T3, third tertile. *P value from nonparametric tests (Kruskal–Wallis). Unadjusted Associations Between levels of norepinephrine (defined as norepinephrine >90th Norepinephrine and Serum Iron Status percentile of the distribution, as compared to norepinephrine Biomarkers ≤90th percentile) were significantly associated with impaired iron status (odds ratio, 2.21; 95% CI, 1.11–4.41), impaired Figure 1 displays crude (unadjusted) associations between iron transport (odds ratio, 2.28; 95% CI, 1.19–4.35), and norepinephrine levels (log transformed) and levels of each iron increased iron demand (odds ratio, 2.23; 95% CI, 1.24–4.01), deficiency biomarker. There was evidence of an inverse, whereas it was not associated with impaired iron storage (low nonlinear association between higher norepinephrine levels ferritin alone) or with presence of anemia (Table 3). Interest- and lower TSAT (P value for linear component, 0.001; P value ingly, the association between norepinephrine levels and the for the nonlinear component, 0.046) and of a positive, linear different biomarkers suggesting iron deficiency was indepen- association between norepinephrine levels and sTfR (P value dent of LVEF. We examined the interaction between LVEF for the linear component, 0.007). In contrast, there was no categories and this association, and it was not significant evidence of a crude association between norepinephrine (P=0.865). levels and ferritin. To analyze the association between iron deficiency and In univariate logistic regression analyses (Table 2), both a norepinephrine, which was the main objective of the study, we log-unit increase in norepinephrine levels as well as a performed bi- and multivariable-adjusted linear regression norepinephrine level >90th percentile (as compared with analyses between increasing levels of log-transformed nore- <90th percentile) was significantly associated with higher pinephrine and levels of several serum iron status biomarkers. odds of iron deficiency. Results are presented in Table 4. Nnorepinephrine levels were inversely associated with TSAT (b= 0.124; P=0.003) and were positively associated with sTfR levels (b=0.115; Other Factors Associated With Serum Iron Status P=0.006). On the other hand, there was no statistically Biomarkers in Unadjusted Analyses significant association of norepinephrine levels with either Table 2 presents the baseline variables that showed statis- ferritin or hemoglobin. Multivariate models analyzing iron tically significant associations with abnormal iron status in status were adjusted by factors associated with abnormal iron bivariate logistic regression analyses. Hyponatremia, anemia, status represented in Table 2 and also body mass index, and norepinephrine levels were the strongest, statistically because it remained at the limit of significance in the significant clinical predictors. univariate analysis (odds ratio, 1.027 [0.997–1.057]; P=0.075). Multivariable-Adjusted Associations Between Norepinephrine and Serum Iron Status Multivariable-Adjusted Associations Between Biomarkers Serum Iron Status and Norepinephrine Levels In line with the main objectives of the study, multivariable Figure 2 displays the adjusted marginal means of serum logistic regression analyses adjusting for variables signifi- norepinephrine according to different iron deficient and cantly associated with iron deficiency were performed. High anemia states, obtained using fitted general linear models. DOI: 10.1161/JAHA.118.010887 Journal of the American Heart Association 5 Sympathetic Activation and Iron Status in CHF Moliner et al ORIGINAL RESEARCH Table 2. Baseline Variables With Statistically Significant Associations With Abnormal Iron Status in Bivariate Logistic Regression Analyses OR (95% CI) P Value Age (per y) 1.026 (1.012–1.040) <0.001 Sex (male vs female) 1.654 (1.198–2.284) 0.002 DM (yes vs no) 1.533 (1.116–2.105) 0.008 SBP (per mm Hg) 1.008 (1.001–1.015) 0.021 LVEF (per 1%) 1.012 (1.003–1.021) 0.012 2 eGFR (per 1 mL/min/1.73 cm ) 0.991 (0.985–0.997) 0.006 logNT-proBNP (per 1 pg/mL) 1.553 (1.187–2.031) 0.001 Albumin (per 1 g/dL) 0.555 (0.388–0.792) 0.001 hs-CRP (per 1 mg/L) 1.183 (1.076–1.301) 0.001 Hemoglobin (per 1 g/dL) 0.727 (0.662–0.797) <0.001 Anemia (yes vs no) 2.707 (1.912–3.833) <0.001 ACEI (yes vs no) 0.486 (0.345–0.684) <0.001 ARB (yes vs no) 2.066 (1.322–3.229) 0.001 HDZ+NTG (yes vs no) 2.060 (1.335–3.181) 0.001 MRA (yes vs no) 0.718 (0.524–0.984) 0.039 Na <135 (yes vs no) 2.844 (1.091–7.411) 0.032 NYHA III to IV (yes vs no) 1.438 (1.045–1.977) 0.026 logNE (per log 1 pg/mL) 2.026 (1.091–3.762) 0.025 NE>p90 (yes vs no) 2.737 (1.414–5.299) 0.003 Anemia is defined as hemoglobin level <12 g/dL in women and <13 g/dL in men. Impaired iron status is defined as ferritin <100 lg/L or TSAT <20%. ACEI indicates angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; HDZ+NTG, hydralazine and nitrates; hs-CRP, high-sensitivity C-reactive protein; logBNP, log N-terminal pro-brain- type natriuretic peptide; logNE, log norepinephrine concentration; LVEF, left ventricular ejection fraction; MRA, mineralcorticoid receptor antagonists; NE>p90, norepinephrine serum concentration >90th percentile (1050 pg/mL); NYHA, New York Heart Association functional class; SBP, systolic blood pressure. 0.038); in patients with impaired iron transport than in those with normal iron transport (537 [511–566] versus 482 pg/ mL [452–513]; P value, 0.012); and in patients with increased iron demand defined as levels of sTfR >75th percentile than in those with normal iron demand (590 [540–646] versus 518 pg/mL [492–545]; P value, 0.014). Figure 1. Unadjusted associations between norepinephrine levels (log-transformed) and levels of serum biomarkers of iron Also, there was a progressive increase in adjusted status. The associations were calculated using generalized norepinephrine marginal means in each consecutive sTfR additive models (GAM). Association between log norepinephrine quartile. In contrast, adjusted norepinephrine levels were not serum levels and log TSAT (A), log NE serum levels and log sTfR significantly different according to iron storage or anemia (B), and log norepinephrine serum levels and log ferritin (C). sTfR status (all P>0.05). indicates soluble transferrin receptor; TSAT, transferrin saturation. Results were consistent with the previous analyses: Nore- Discussion pinephrine levels were significantly higher in patients with In a sample of 742 patients with CHF, raised norepinephrine impaired iron status than in those with normal iron status levels were associated with impaired iron status and, (528 [505–551] versus 482 pg/mL [448–518]; P value, particularly, with those biomarkers that suggest impaired DOI: 10.1161/JAHA.118.010887 Journal of the American Heart Association 6 Sympathetic Activation and Iron Status in CHF Moliner et al ORIGINAL RESEARCH Table 3. Bi- and Multivariate-Adjusted Associations Between High Norepinephrine Levels (>90th Percentile) and Impaired Iron Status States Raised Norepinephrine Unadjusted Models Adjusted Models OR 95% CI P Value OR 95% CI P Value Impaired iron status 2.589 1.334 to 5.025 0.005 2.206 1.105 to 4.405 0.025 Impaired iron transport 2.824 1.566 to 5.095 0.001 2.276 1.192 to 4.346 0.013 Impaired iron storage 1.083 0.656 to 1.787 0.756 1.253 0.728 to 2.155 0.415 Increased iron demand 2.376 1.339 to 4.215 0.003 2.229 1.239 to 4.009 0.007 Anemia 1.606 0.983 to 2.622 0.058 1.149 0.657 to 2.009 0.627 Anemia is defined as hemoglobin level <12 g/dL in women and <13 g/dL in men. Impaired iron transport is defined as TSAT <20%; impaired iron storage is defined as ferritin <100 lg/L; impaired iron status is defined as ferritin <100 lg/L or TSAT <20%; and increased iron demand is defined as soluble transferrin receptor >75th percentile. Raised norepinephrine = norepinephrine serum concentration >90th percentile. OR indicates odds ratio. iron transport (TSAT <20%) and increased iron demand (raised higher risk of developing heart failure (HF) over time may give levels of sTfR). This is particularly important given that these 2 further support to the hypothesis that iron may be involved in iron-deficient states have been shown to strongly correlate the pathophysiological sequence leading to HF.30 with worse clinical profile and prognosis in patients with Disturbances in energy metabolism at the cellular level and CHF.20 Given the proven role of neurohormonal activation impairment of the contractile function of cardiomyocytes and and, particularly, sympathetic activation in the pathophysio- skeletal muscle cells as a result of iron deficiency may affect logical cascade leading to, and promoting the progression of, patients across all LVEF categories and may promote disease CHF, the new data emerging from our study are particularly progression in all of them. relevant given that they may support the hypothesis that iron A previous study of our group showed that the particular deficiency might not just be a comorbidity, but could also be a iron states that are associated with a worse clinical profile key element in the pathophysiological sequence leading to, and prognosis are those that entail abnormalities in the and promoting the progression of, CHF. To the best of our transport of iron or an increased demand of iron at the cellular knowledge, this is the first clinical study suggesting a level. However, in the above-mentioned study, isolated potential interplay between norepinephrine levels as a abnormalities at the storage level, represented by normal surrogate of activation of the sympathetic system and the TSAT and low ferritin, were associated with a clinical profile frequent dysregulation of iron metabolism observed in and prognosis undistinguishable from patients with CHF and patients with CHF. normal iron status.20 Interestingly, in the present study, a It is well known that iron deficiency is common in patients raised sympathetic nervous system (SNS) activity, repre- with CHF, and it is associated with worse quality of life, sented by raised norepinephrine levels, was particularly functional capacity, and prognosis.1–6 The negative impact of associated with biomarkers suggesting abnormalities in iron iron deficiency may not only be explained by the role of iron in erythropoiesis, but may also be driven by the effects of iron in Table 4. Bivariate and Multivariate-Adjusted Associations cardiomyocyte function and, particularly, in mitochondrial Between Increasing Levels of NE and Levels of Iron Status respiration. These effects have been confirmed in cellular Biomarkers (All Log-Transformed) models where iron deficiency in the cardiomyocyte leads to log NE mitochondrial dysfunction, hypertrophy, impaired cardiomy- ocite contractility, and ventricular dysfunction21–28 and where Unadjusted Models Adjusted Models restoration of intracellular iron levels reversed these R 2 Β P Value R2 Β P Value effects.27–29 Haddad et al showed, in an animal model of log iron 0.014 0.120 0.001 0.379 0.071 0.019 cardiac iron deficiency, that the myocardium of these animals log ferritin 0.000 0.017 0.642 0.084 0.024 0.503 showed impaired mitochondrial respiration at the cardiomy- log TSAT 0.022 0.148 <0.001 0.255 0.124 0.003 ocyte level and a blunted response in ventricular contractility log sTfR 0.014 0.116 0.007 0.076 0.115 0.006 after administration of dobutamine. Interestingly, these abnor- malities were reversed after administration of intravenous log Hb 0.002 0.039 0.288 0.245 0.005 0.878 iron.27 At a clinical level, cohort studies conducted in healthy logHb indicates log haemoglobin; log NE, log norepinephrine; logsTfR, log soluble individuals in whom impaired iron status was associated with transferrin receptor; logTSAT, log transferrin saturation. DOI: 10.1161/JAHA.118.010887 Journal of the American Heart Association 7 Sympathetic Activation and Iron Status in CHF Moliner et al ORIGINAL RESEARCH Figure 2. Predicted adjusted marginal means of serum norepinephrine levels by iron status categories. Norepinephrine adjusted marginal means calculated with general linear models according to (A) iron status, (B) iron transport, and (C) iron demand: sTfR classified in quartiles (Q1 ≤1.21 mg/L; Q2 1.22–1.60 mg/L; Q3 1.61–2.24 mg/L; Q4 ≥2.25 mg/L). Impaired iron status was defined as TSAT <20% or ferritin <100 ng/mL. Impaired iron transport was defined as TSAT <20%. sTfR indicates soluble transferrin receptor; TSAT, transferrin saturation. transport and increased iron demand at the cellular level. This with beta-blockers. This association may give further support association was independent of relevant factors such as body to the hypothesis that the modulation of the neurohormonal mass index and inflammatory status.31,32 drive of the sympathetic system by using beta-blockers may It has been suggested that systemic iron depletion may interplay with the development of iron abnormalities at the precede iron deficiency at the myocardial level. Occult blood myocardial level in patients with HF. loss attributed to use of antiplatelet agents or anticoagulants, Taking all into account, we may hypothesize that iron poor intake of iron in diet, and reduced intestinal absorption deficiency might be a relevant factor involved in the complex attributed to inflammation, among other factors, have been pathophysiological cascade of HF given that it may be suggested as potential causes of systemic iron deficiency in involved in the initiation and progression of the functional these patients.33 However, despite these assumptions, the abnormalities observed in these patients leading to clinical exact mechanisms by which CHF patients develop iron events and functional impairments. According to this hypoth- deficiency at a systemic and myocardial level are still esis, iron deficiency may be not just an “unpleasant innocent unknown. In consequence, it is important to investigate bystander,” but a key factor actively involved in the mecha- alternative mechanisms that may lead to iron deficiency at the nisms involved in onset and progression of CHF. cardiomyocyte level on top of systemic iron deficiency. From this point of view, the interplay between iron metabolism disturbances and neurohormonal activation represented by Study Limitations raised levels of norepinephrine in patients with CHF observed Our study has several limitations that must be discussed. First in our study needs to be taken into account. Activation of the of all, in this study, we conducted a cross-sectional analysis; SNS, which involves an elevation of serum norepinephrine therefore, causality may not be inferred. The deleterious levels, has long-term deleterious effects on CHF and is one of influence between the 2 terms of the equation may be the main therapeutic targets in HF.7,8 From our data, we bidirectional. Raised norepinephrine levels as an expression suggest that the SNS could play a role in dysregulation of iron of sympathetic activation may promote iron depletion at a homeostasis in patients with CHF. Our observations are cellular level, particularly in patients with systemic iron consistent with those from basic science models. Exposure of deficiency; on the other hand, it is also possible that systemic adult cardiomyocyte to norepinephrine leads to a downreg- iron deficiency may lead to a greater impairment in cardiomy- ulation of type 1 transferrin receptor expression.16 These ocyte function and mitochondrial function that would lead to molecular changes may lead to intracellular iron depletion. In more-pronounced oxidative stress, cardiomyocyte death, pump the same study, exposure of cardiomyocytes to the beta failure, and an enhanced (compensatory) neurohormonal agonist, isoprenaline, also entails downregulation of type 1 activation represented by raised norepinephrine levels. This transferrin receptor.16 In another study, Melenovsky et al will promote HF progression and clinical adverse events. analyzed cardiomyocytes from hearts of patients transplanted Second, because this was a single-center study, our because of advanced HF.29 Compared with controls, car- findings may not be representative of other CHF patient diomyocytes of patients with HF showed reduced levels of populations. Third, SNS activity is a complex concept that intracellular iron. In this study, iron deficiency at the entails several biomarkers and measures. Using nore- myocardial level was associated with a lower treatment rate pinephrine levels as a marker of sympathetic activation may DOI: 10.1161/JAHA.118.010887 Journal of the American Heart Association 8 Sympathetic Activation and Iron Status in CHF Moliner et al ORIGINAL RESEARCH be seen as a simplification of the SNS. Third, SNS activity is a 3. Enjuanes C, Klip IT, Bruguera J, Cladellas M, Ponikowski P, Banasiak W, van Veldhuisen DJ, van der Meer P, Jankowska EA, Comın-Colet J. Iron deficiency complex concept that may be defined by combined informa- and health-related quality of life in chronic heart failure: results from a tion obtained from several biomarkers and functional mea- multicenter European study. Int J Cardiol. 2014;174:268–275. 4. Enjuanes C, Bruguera J, Grau M, Cladellas M, Gonzalez G, Mero~ no O, Moliner- sures. According to this, estimation of SNS activity based on Borja P, Verdu JM, Farre N, Comın-Colet J. Iron status in chronic heart failure: norepinephrine levels alone may be considered an oversim- impact on symptoms, functional class and submaximal exercise capacity. Rev Esp Cardiol (Engl Ed). 2016;69:247–255. plification. Nevertheless, this parameter was used as a 5. Bojarczuk J, Josiak K, Kasztura M, Kustrzycka-Kratochwil D, Nowak K, Jagielski surrogate of the SNS when the neurohormonal hypothesis D, Banasiak W, Jankowska EA, Ponikowski P. Iron deficiency in heart failure: impact on response to cardiac resynchronization therapy. Int J Cardiol. was developed and guided the development of key therapeu- 2016;222:133–134. tic strategies for treatment of CHF patients Although sample 6. Martens P, Verbrugge F, Nijst P, Dupont M, Tang WH, Mullens W. Impact of iron deficiency on response to and remodeling after cardiac resynchronization processing and the laboratory method of analysis of nore- therapy. Am J Cardiol. 2017;119:65–70. pinephrine are complex and may be seen as limitations of our 7. Yancy CW, Jessup M, Bozkurt B, Butler J, Casey DE Jr, Colvin MM, Drazner MH, study, we want to highlight that the procedures followed for Filippatos GS, Fonarow GC, Givertz MM, Hollenberg SM, Lindenfeld J, Masoudi FA, McBride PE, Peterson PN, Stevenson LW, Westlake C. 2017 ACC/AHA/ obtaining and processing the samples and for measuring HFSA Focused Update of the 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology/American norepinephrine levels were accurate and based on standard- Heart Association Task Force on Clinical Practice Guidelines and the Heart ized laboratory methods. Failure Society of America. J Am Coll Cardiol. 2017;70:776–803. And, fourth, we used biomarkers that indirectly estimate 8. Ponikowski P, Voors AA, Anker SD, Bueno H, Cleland JGF, Coats AJS, Falk V, Gonzalez-Juanatey JR, Harjola VP, Jankowska EA, Jessup M, Linde C, the true iron status; however, these biomarkers are those Nihoyannopoulos P, Parissis JT, Pieske B, Riley JP, Rosano GMC, Ruilope LM, Ruschitzka F, Rutten FH, van der Meer P; ESC Scientific Document Group. used in daily clinical practice.29,34 2016 ESC guidelines for the diagnosis and treatment of acute and chronic heart failure: the Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Eur Heart J. 2016;37:2129–2200. Conclusions 9. Anker SD, Comin Colet J, Filippatos G, Willenheimer R, Dickstein K, Drexler H, L€uscher TF, Bart B, Banasiak W, Niegowska J, Kirwan BA, Mori C, von Eisenhart Rothe B, Pocock SJ, Poole-Wilson PA, Ponikowski P; FAIR-HF Trial Investiga- In patients with CHF, an increased sympathetic activation tors. Ferric carboxymaltose in patients with heart failure and iron deficiency. N estimated with raised norepinephrine levels is associated with Engl J Med. 2009;361:2436–2448. impaired iron status. Specifically, we observed robust, inde- 10. Ponikowski P, van Veldhuisen DJ, Comin-Colet J, Ertl G, Komajda M, Mareev V, McDonagh T, Parkhomenko A, Tavazzi L, Levesque V, Mori C, Roubert B, pendent associations between norepinephrine levels and Filippatos G, Ruschitzka F, Anker SD; CONFIRM-HF Investigators. Beneficial effects of long-term intravenous iron therapy with ferric carboxymaltose in dysregulation of biomarkers suggesting impaired iron trans- patients with symptomatic heart failure and iron deficiency. Eur Heart J. port (low TSAT) and increased iron demand (raised sTfR), 2015;36:657–668. whereas no associations were identified with iron storage 11. Hjalmarson A, Goldstein S, Fagerberg B, Wedel H, Waagstein F, Kjekshus J, Wikstrand J, ElAllaf D, Vıtovec J, Aldershvile J, Halinen M, Dietz R, Neuhaus KL, (ferritin) or presence of anemia. Longitudinal studies are Janosi A, Thorgeirsson G, Dunselman PH, Gullestad L, Kuch J, Herlitz J, Rickenbacher P, Ball S, Gottlieb S, Deedwania P; MERIT-HF Study Group. needed to better understand the directionality and potential Effects of controlled-release metoprolol on total mortality, hospitalizations, causal relationship underlying the observed associations. and wellbeing in patients with heart failure: the Metoprolol CR/XL Randomized Intervention Trial in congestive Heart Failure (MERIT-HF). JAMA. 2000;283:1295–1302. 12. Packer M, Coats AJ, Fowler MB, Katus HA, Krum H, Mohacsi P, Rouleau JL, Tendera M, Castaigne A, Roecker EB, Schultz MK, DeMets DL. Effect of Sources of Funding carvedilol on survival in severe chronic heart failure. N Engl J Med. 2001;344:1651–1658. This work was supported by an unrestricted grant from Vifor 13. Packer M, Bristow MR, Cohn JN, Colucci WS, Fowler MB, Gilbert EM, Shusterman NH. The effect of carvedilol on morbidity and mortality in patients Pharma for editorial support and open access of the with chronic heart failure. N Engl J Med. 1996;334:1349–1355. publication. 14. Packer M. Effect of carvedilol on the morbidity of patients with severe chronic heart failure: results of the Carvedilol Prospective Randomized Cumulative Survival (COPERNICUS) Study. Circulation. 2002;106:2194–2199. 15. Flather MD, Shibata MC, Coats AJS, Van Veldhuisen DJ, Parkhomenko A, Disclosures Borbola J, Cohen-Solal A, Dumitrascu D, Ferrari R, Lechat P, Soler-Soler J, Tavazzi L, Spinarova L, Toman J, B€ ohm M, Anker SD, Thompson SG, Poole- Comin-Colet has received speaker fees from Vifor Pharma. Wilson PA; SENIORS Investigators. Randomized trial to determine the effect of nebivolol on mortality and cardiovascular hospital admission in elderly patients The remaining authors have no disclosures to report. with heart failure (SENIORS). Eur Heart J. 2005;26:215–225. 16. Maeder MT, Khammy O, dos Remedios C, Kaye DM. Myocardial and systemic iron depletion in heart failure implications for anemia accompanying heart failure. J Am Coll Cardiol. 2011;58:474–480. References 1. Klip IT, Comin-Colet J, Voors AA, Ponikowski P, Enjuanes C, Banasiak W, Lok 17. Farre N, Arany o J, Enjuanes C, Verdu-Rotellar JM, Ruiz S, Gonzalez-Robledo G, DJ, Rosentryt P, Torrens A, Polonski L, van Veldhuisen DJ, van der Meer P, Mero~ no O, de Ramon M, Moliner P, Bruguera J, Comin-Colet J. Differences in Jankowska EA. Iron deficiency in chronic heart failure: an international pooled neurohormonal activity partially explain the obesity paradox in patients with heart analysis. Am Heart J. 2013;165:575–582. failure: the role of sympathetic activation. Int J Cardiol. 2015;181:120–126. 2. Comın-Colet J, Enjuanes C, Gonzalez G, Torrens A, Cladellas M, Mero~ no O, 18. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate Ribas N, Ruiz S, G omez M, Verdu JM, Bruguera J. Iron deficiency is a key method to estimate glomerular filtration rate from serum creatinine: a new determinant of health-related quality of life in patients with chronic prediction equation. Ann Intern Med. 1999;130:461–470. heart failure regardless of anaemia status. Eur J Heart Fail. 2013;15:1164– 19. Nutritional anaemias. Report of a WHO scientific group. World Health Organ 1172. Tech Rep Ser. 1968;405:5–37. DOI: 10.1161/JAHA.118.010887 Journal of the American Heart Association 9 Sympathetic Activation and Iron Status in CHF Moliner et al ORIGINAL RESEARCH 20. Moliner P, Jankowska EA, van Veldhuisen DJ, Farre N, Rozentryt P, Enjuanes C, deficiency impairs contractility of human cardiomyocytes through decreased Polonski L, Mero~no O, Voors AA, Ponikowski P, Van der Meer P, Comin-Colet J. mitochondrial function. Eur J Heart Fail. 2018;20:910–919. Clinical correlates and prognostic impact of impaired iron storage versus 29. Melenovsky V, Petrak J, Mracek T, Benes J, Borlaug BA, Nuskova H, Pluhacek T, impaired iron transport in an international cohort of 1821 patients with Spatenka J, Kovalcikova J, Drahota Z, Kautzner J, Pirk J, Houstek J. Myocardial chronic heart failure. Int J Cardiol. 2017;243:360–366. iron content and mitochondrial function in human heart failure: a direct tissue 21. Chen C, Paw BH. Cellular and mitochondrial iron homeostasis in vertebrates. analysis. Eur J Heart Fail. 2017;19:522–530. Biochim Biophys Acta. 2012;1823:1459–1467. 30. Silvestre OM, Goncalves A, Nadruz W Jr, Claggett B, Couper D, Eckfeldt JH, 22. Anderson GJ, Vulpe CD. Mammalian iron transport. Cell Mol Life Sci. Pankow JS, Anker SD, Solomon SD. Ferritin levels and risk of heart failure— 2009;66:3241–3261. the Atherosclerosis Risk in Communities Study. Eur J Heart Fail. 2017;19: 23. Richardson DR, Lane DJ, Becker EM, Huang ML, Whitnall M, Suryo Rahmanto 340–347. Y, Sheftel AD, Ponka P. Mitochondrial iron trafficking and the integration of 31. Tkaczyszyn M, Drozd M, Wez grzynowska-Teodorczyk K, Flinta I, Kobak K, iron metabolism between the mitochondrion and cytosol. Proc Natl Acad Sci Banasiak W, Ponikowski P, Jankowska EA. Depleted iron stores are associated USA. 2010;107:10775–10782. with inspiratory muscle weakness independently of skeletal muscle mass in 24. Bayeva M, Ardehali H. Mitochondrial dysfunction and oxidative damage to men with systolic chronic heart failure. J Cachexia Sarcopenia Muscle. sarcomeric proteins. Curr Hypertens Rep. 2010;12:426–432. 2018;9:547–556. 32. Van Aelst LNL, Abraham M, Sadoune M, Lefebvre T, Manivet P, Logeart D, 25. Bayeva M, Gheorghiade M, Ardehali H. Mitochondria as a therapeutic target in Launay JM, Karim Z, Puy H, Cohen-Solal A. Iron status and inflammatory heart failure. J Am Coll Cardiol. 2013;61:599–610. biomarkers in patients with acutely decompensated heart failure: early in- 26. Naito Y, Tsujino T, Matsumoto M, Sakoda T, Ohyanagi M, Masuyama T. hospital phase and 30-day follow-up. Eur J Heart Fail. 2017;19:1075–1076. Adaptive response of the heart to long-term anemia induced by iron 33. Jankowska EA, Rozentryt P, Witkowska A, Nowak J, Hartmann O, Ponikowska deficiency. Am J Physiol Heart Circ Physiol. 2009;296:H585–H593. B, Borodulin-Nadzieja L, Banasiak W, Polonski L, Filippatos G, McMurray JJ, 27. Haddad S, Wang Y, Galy B, Korf-Klingebiel M, Hirsch V, Baru AM, Rostami F, Anker SD, Ponikowski P. Iron deficiency: an ominous sign in patients with Reboll MR, Heineke J, Fl€ogel U, Groos S, Renner A, Toischer K, Zimmermann F, systolic chronic heart failure. Eur Heart J. 2010;31:1872–1880. Engeli S, Jordan J, Bauersachs J, Hentze MW, Wollert KC, Kempf T. Iron- 34. Grote Beverborg N, Klip IT, Meijers WC, Voors AA, Vegter EL, van der Wal HH, regulatory proteins secure iron availability in cardiomyocytes to prevent heart Swinkels DW, van Pelt J, Mulder AB, Bulstra SK, Vellenga E, Mariani MA, de failure. Eur Heart J. 2017;38:362–372. Boer RA, van Veldhuisen DJ, van der Meer P. Definition of iron deficiency based 28. Hoes MF, Grote Beverborg N, Kijlstra JD, Kuipers J, Swinkels DW, Giepmans on the gold standard of bone marrow iron staining in heart failure patients. BN, Rodenburg RJ, van Veldhuisen DJ, de Boer RA, van der Meer P. Iron Circ Heart Fail. 2018;11:e004519. DOI: 10.1161/JAHA.118.010887 Journal of the American Heart Association 10