Prevalence, incidence, persistence and resolution of anaemia in patients with heart failure and left ventricular systolic dysfunction and the relationship to outcome K. Goode, JG. Cleland, AL. Clark, K. Wong, P. Pellicori, A. Yassin, D. Cullington Department of Cardiology, Castle Hill Hospital, Kingston-upon-Hull, UK Background: Most reports of the epidemiology of anaemia and its relationship to prognosis in patients with heart failure (HF) come from multi-centre clinical trials (that may be unrepresentative of the epidemiological problem) or hospital discharge populations, but few from unselected hospital out-patients in clinical practice. Methods: Consenting patients with suspected HF referred to a local clinic underwent a systematic evaluation including prior medical history, symptoms, signs, electro- and echocardiograms, standard haematology and biochemistry profiles and aminoterminal pro-brain natriuretic peptide (NT-proBNP). Left ventricular systolic dysfunction (LVSD) was defined as an ejection fraction of <40%. Patients were followed clinically and through electronic record systems. Data are presented for baseline (BL) and 12-month follow-up (FU) visits. The WHO definition for anaemia was used (haemoglobin (Hb) <13 for men; <12 for women). Estimated glomerular filtration rate (eGFR) was derived using the 4-variable MDRD equation. Continuous variables were compared using the Mann- Whitney-U and the Wilcoxon signed rank test; categorical variables using the chi-squared and McNemar test. Prevalence estimates for anaemia according to age, sex and renal function were derived using a logistic regression model. Results (cont): Patients who resolved their anaemia had better 12-month survival than those with persistent or incident anaemia. However, by 24-months, patients with resolved anaemia were doing no better than those with incident anaemia (see Figure 5). Those with persistent anaemia had the worst short-term and long-term outcomes. Conclusions: The prevalence and incidence of anaemia are both high in patients with heart failure and LVSD and usually persists. Despite having high baseline NT-proBNP, resolution of anaemia is associated with a favourable 1- year survival but this benefit does not persist in the long- term. Further analysis is necessary to determine if this is due to a re-occurrence of anaemia. Anaemia is a useful prognostic indicator in heart failure and may be a target for therapy. Analysis funded by Amgen UK Results: Of 4456 patients presenting with suspected HF, 1791 (40%) had LVSD; see Table 1 for characteristics of LVSD cohort. At BL, 597 patients (33%) were anaemic according to the WHO definition (459 men; 138 women). Patients with anaemia were older, more likely to have iscaemic heart disease, more symptomatic with lower FEV 1 and shorter 6-minute walk test distance, worse renal function, higher hsCRP and NT-proBNP and more likely to be taking a diuretic. However, there were no differences in ejection fraction, echocardiographic indices or ECG rhythm (see Table 1). Of those with anaemia at BL, 185 (31%) were 1.1-2.0 g/dL below, 80 (13.4%) were 2.1-3.0 g/dL below and 33 (6%) were >3 g/dL below the WHO definition. A 1 g/dL difference in the definition of anaemia results in a halving of the reported prevalence to 17% (see Figure 1). The prevalence of anaemia was higher in men, in those that were older and in those with worse renal function (see Figure 2). The prevalence of iron depletion was lowest in those with normal Hb, increasing with decreasing Hb. There were no differences in the prevalence of macrocytic disorders in those with or without anaemia (see Figure 3). At 12 months, 1401 patients were reassessed (1329 with a FU Hb), 235 had died and 155 did not re-attend. Of those reassessed, the point prevalence of anaemia increased from 32% to 38%. Figure 1: Cumulative prevalence of anaemia relative to the WHO definition. So 0 on the x-axis represents Hb<13 g/dL for men and Hb<12 g/dL for women; at -1 this represents Hb<12 g/dL for men and Hb<11 g/dL for women. Results (cont): Of 910 patients with out anaemia at BL, 186 (20%) developed anaemia at FU. Of 419 patients with anaemia at BL, 105 (25%) resolved their anaemia at FU. The persistence of anaemia was greater in men, in those with worse renal function and in older patients. The incidence of anaemia was also higher in those with worse renal function and in older patients, but not between men and women (see Figure 4). Patients who resolved their anaemia at FU, improved their serum iron, transferrin saturation but not their ferritin levels (see Table 2). Those with incident anaemia had falling serum iron but a modest improvement in ferritin. Those who remained anaemic or remained non-anaemic at FU had no worsening or improvement in iron depletion measures. The proportion of patients meeting the Fair-HF inclusion criteria did not change for any group. Only patients with no anaemia at BL or FU had a modest improvement in vitamin B12 levels. Patients in the resolved group had the highest BL NT-proBNP levels. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% Level 1 Level 2 Level 3 Level 4 CKD Level Prevalence (%) >80 yrs 71-80 yrs 61-70 yrs <=60 yrs Men Ischaemic Heart Disease 1174 (67) 750 (64.1) 424 (72.9) <0.0005 QRS Width (msec) 112 (98.0 - 142) 112 (98.0 - 142) 116 (98.0 - 144) 0.37 Sinus Rhythm 1095 (65.2) 730 (64.8) 365 (65.9) 0.67 Previous MI (on ECG) 723 (49.3) 471 (48.6) 252 (50.8) 0.42 LVEDD Height indexed (cm/m) 3.71 (3.35 - 4.05) 3.71 (3.35 - 4.05) 3.71 (3.35 - 4.06) 0.97 LVESD Height indexed (cm/m) 3.14 (2.75 - 3.54) 3.13 (2.75 - 3.54) 3.16 (2.73 - 3.54) 0.89 LV mass BSA indexed (g/m 2 ) 151 (121 - 193) 150 (122 - 193) 153 (119 - 194) 0.89 Left Atrium Height indexed (cm/m) 2.58 (2.31 - 2.90) 2.57 (2.27 - 2.88) 2.60 (2.36 - 2.90) 0.071 EF Simpson (%) 32.0 (26.0 - 37.0) 32.0 (26.3 - 37.0) 32.0 (26.0 - 37.0) 0.79 Biochemistry NT-proBNP (ng/L) 1759 (726 - 3964) 1400 (615 - 3246) 2816 (992 - 6026) <0.0005 GFR (4-var) (ml/min/1.73m 2 ) 58.3 (44.6 - 72.2) 61.4 (48.8 - 74.8) 49.6 (36.7 - 65.6) <0.0005 hsCRP (mg/L) 4.70 (1.90 - 9.15) 3.90 (1.70 - 7.50) 6.70 (2.68 - 18.0) <0.0005 All patients (n=1791) No Anaemia (n=1194) Anaemia (n=597) p-value Age (yrs) 72.0 (64.0 - 78.0) 70.0 (63.0 - 77.0) 74.0 (68.0 - 80.0) <0.0005 Male 1330 (74.3) 871 (72.9) 459 (76.9) 0.072 Exam / History COPD 187 (10.7) 126 (10.8) 61 (10.5) 0.85 Peripheral Vascular Disease 121 (6.9) 64 (5.5) 57 (9.8) 0.001 CVA / TIA 180 (10.3) 107 (9.1) 73 (12.5) 0.027 Systolic BP (mmHg) 128 (112 - 144) 130 (115 - 146) 122 (108 - 140) <0.0005 Diastolic BP (mmHg) 75.0 (66.0 - 85.0) 78.0 (68.0 - 87.0) 70.0 (62.0 - 79.0) <0.0005 BMI (kg/m 2 ) 27.2 (24.1 - 30.7) 27.4 (24.1 - 30.9) 26.9 (24.0 - 30.3) 0.084 NYHA Class III/IV 628 (37.7) 368 (32.8) 260 (47.7) <0.0005 Oedema (Ankles or worse) 382 (25.4) 210 (20.9) 172 (34.4) <0.0005 Raised JVP >1cm 310 (21.0) 167 (16.9) 143 (29.6) <0.0005 FEV1 % Predicted 67.1 (51.6 - 84.1) 70.0 (53.2 - 86.0) 62.0 (48.3 - 77.9) <0.0005 6MWT Distance (m) 300 (165 - 375) 330 (195 - 390) 210 (105 - 315) <0.0005 ECG / Echo Medication Loop Diuretic 1313 (77.4) 829 (72.8) 484 (86.7) <0.0005 Aldosterone RA 456 (26.9) 274 (24.1) 182 (32.6) <0.0005 Betablocker 952 (56.1) 637 (55.9) 315 (56.5) 0.84 ACE Inhibiter 1237 (72.9) 845 (74.2) 392 (70.3) 0.087 Table 1: Characteristics of patients with LVSD split by presence and absence of anaemia. Categorical variables presented as n (%); continuous variables presented as median (IQR) 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 Haemoglobin relative to WHO definition (g/dL) Cumulative Prevalence (%) . 1 g/dL Change 1/2 the prevalence 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% Level 1 Level 2 Level 3 Level 4 CKD Level Prevalence (%) >80 yrs 71-80 yrs 61-70 yrs <=60 yrs Women Figure 2: Estimated prevalence of anaemia using the WHO criteria according to age, renal dysfunction and sex. Chronic kidney disease level 1 = eGFR ≥90; level 2 = eGFR 60-89 ; level 3 = eGFR 30-59; level 4 = eGFR <30 ml/min/1.73m 2 55% 53% 58% 65% 65% 48% 22% 74% 59% 47% 42% 8% 7% 10% 6% 8% 8% 10% 4% 9% 10% 5% 14% 14% 14% 14% 11% 15% 18% 9% 13% 16% 16% 24% 26% 18% 15% 15% 29% 51% 12% 19% 27% 37% 0% 20% 40% 60% 80% 100% All Patients (n=1329) Men (n=1000) Women (n=329) CKD Level 1 (n=81) CKD Level 2 (n=523) CKD Level 3 (n=592) CKD Level 4 (n=79) Age ≤60 (n=223) Age 61-70 (n=396) Age 71-80 (n=513) Age >80 (n=197) No-anaemia at BL or 1-year Resolved Anaemia Incident Anaemia Persistent Anaemia Figure 3: Baseline prevalence of hematininc deficiencies in patients with LVSD according to anaemia status. Fair-HF inclusion criteria was ferritin<100 μg/L or ferritin 100-300 μg/L with tranferrin saturation <20%. Figure 4: Incidence, resolution and persistence of anaemia in patients with LVSD at 12-months after BL; overall and split by sex, renal function and age. 2% 2% 3% 10% 27% 7% 51% 57% 8% 9% 4% 14% 6% 3% 14% 25% 47% 11% 57% 66% 8% 9% 6% 14% 23% 13% 34% 46% 68% 19% 52% 70% 6% 7% 5% 12% 0% 10% 20% 30% 40% 50% 60% 70% 80% IRON DEPLETION Hypochromic (MCHC <26 pg) Microcytic (MCV < 80 fl) Iron <8 μmol/L TSat <15% TSat <20% Ferritin <30 μg/L Ferritin <100 μg/L Fair-HF OTHER Macrocytic (MCV >100 fl) Vitamin B12 <180 ng/L Red Cell Folate <147 μg/L Low Vitamin B12 OR Low RCF Prevalence ≥1g/dL above WHO (n=795) +/- 1g/dL from WHO (n=698) ≥1g/dL below WHO (n=298) Haematinic Variable Normal Range MCHC 26 – 34 pg MCV 80 – 100 fl Iron 8 – 34 μmol/L Ferritin 15 – 300 μg/L Transferrin Saturation 15 – 50 % Vitamin B12 180 – 1130 ng/L Red Cell Folate 147 – 650 μg/L No Anaemia Both (n=724) Incident Anaemia (n=186) Resolved Anaemia (n=105) Persistent Anaemia (n=314) Age (yrs) 69.0 (61.0 - 76.0) 73.0 (66.0 - 78.3) 72.0 (66.0 - 77.0) 74.0 (68.0 - 80.0) Male 532 (73.5) 140 (75.3) 73 (69.5) 255 (81.2) BL NT-proBNP (ng/L) 1197 (510 - 2750) 1755 (833 - 3268) 2973 (1068 - 5820) 2405 (870 - 5069) FU NT-proBNP (ng/L) 875 (354 - 1926) 1522 (522 - 2930) 1651 (668 - 3144) 1548 (642 - 3483) Wilcoxon Signed Rank Test P<0.0005 P=0.24 P=0.002 P=0.002 BL EF Simpson (%) 32.0 (26.0 - 37.0) 31.0 (26.0 - 36.0) 30.0 (25.8 - 37.0) 33.0 (26.8 - 37.3) FU EF Simpson (%) 36.0 (30.0 - 44.0) 35.0 (29.0 - 40.0) 32.5 (27.3 - 40.0) 34.5 (29.0 - 42.0) Wilcoxon Signed Rank Test P=<0.0005 P=<0.0005 P=0.012 P=<0.0005 BL GFR (4-var) (ml/min/1.73m 2 ) 63.1 (51.2 - 76.7) 55.5 (42.8 - 66.9) 54.9 (39.1 - 68.5) 50.0 (37.8 - 64.2) FU GFR (4-var) (ml/min/1.73m 2 ) 61.8 (48.5 – 75.1) 48.5 (37.2 – 63.2) 50.8 (38.2 – 68.2) 42.8 (30.1 – 56.8) Wilcoxon Signed Rank Test p<0.0005 p<0.0005 p=0.073 p<0.0005 BL Iron (umol/L) 16.0 (12.0 - 20.0) 15.0 (13.0 - 19.0) 9.50 (6.00 - 14.8) 10.0 (8.00 - 14.0) FU Iron (umol/L) 16.0 (13.0 - 20.0) 13.0 (11.0 - 16.0) 16.0 (12.0 - 19.0) 12.0 (9.00 - 16.0) Wilcoxon Signed Rank Test P=0.63 P=0.001 P<0.0005 P=0.10 BL Transferrin Sat. (%) 25.0 (18.0 - 33.0) 24.0 (18.0 - 31.0) 18.0 (11.8 - 26.0) 17.0 (12.0 - 23.0) FU Transferrin Sat. (%) 25.0 (20.0 - 32.0) 21.0 (17.0 - 28.0) 24.0 (16.8 - 31.3) 20.5 (13.0 - 27.0) Wilcoxon Signed Rank Test P=0.38 P=0.18 P=0.002 P=0.18 BL Ferritin (ug/L) 93.5 (52.0 - 162) 92.0 (51.0 - 143) 76.0 (34.5 - 196) 82.0 (42.0 - 177) FU Ferritin (ug/L) 95.0 (54.8 - 171) 112 (60.0 - 195) 81.0 (43.0 - 146) 107 (50.0 - 203) Wilcoxon Signed Rank Test P=0.48 P=0.012 P=0.24 P=0.39 BL FAIR-HF, n (%) 236 (59.4) 49 (58.3) 30 (66.7) 120 (70.6) FU FAIR-HF, n (%) 260 (57.4) 70 (61.4) 37 (63.8) 103 (63.2) McNemar Test P=0.9 P=1.0 P=1.0 P=1.0 BL Red Cell Folate (ug/L) 373 (265 - 531) 360 (232 - 500) 357 (244 - 534) 397 (274 - 555) FU Red Cell Folate (ug/L) 396 (288 - 575) 376 (270 - 591) 392 (259 - 467) 429 (305 - 610) Wilcoxon Signed Rank Test P<0.0005 P=0.042 P=0.49 P=0.027 BL Vitamin B12 (ng/L) 321 (248 - 426) 340 (253 - 458) 337 (250 - 423) 335 (246 - 472) FU Vitamin B12 (ng/L) 337 (250 - 435) 317 (247 - 458) 355 (258 - 479) 353 (251 - 499) Wilcoxon Signed Rank Test P=0.05 P=0.26 P=0.60 P=0.38 Alive after Follow-up Visit (months) Cumulative Survival Incidence Groups No anaemia both visits Resolved anaemia Incident anaemia Persistent anaemia Incidence Groups No anaemia both visits Resolved anaemia Incident anaemia Persistent anaemia 0.012 0.12 <0.0005 Persistent - 0.62 <0.0005 Incident - - <0.0005 Resolved Incident Resolved None Pair-wise p-values 0.012 0.12 <0.0005 Persistent - 0.62 <0.0005 Incident - - <0.0005 Resolved Incident Resolved None Pair-wise p-values 10% 22% 18% 8% 16% 24% 27% 35% 47% 36% 33% 24% 56% 45% 42% 30% 0% 10% 20% 30% 40% 50% 60% 70% No Anaemia Incident Anaemia Resolved Anaemia Persistant Anaemia Dead 12m (%) Dead 24m (%) Dead 36m (%) Dead 48m (%) Figure 5: Mortality at 12, 24, 36 and 48-months after FU (left panel); Kaplan-Meier curve of long-term outcome (right panel) for patients with no, incident, resolved and persistent anaemia. Table 1: Changes in NT-proBNP, ejection fraction, renal function, iron status and vitamin B12 from BL to FU according to anaemia status.