Hepatocyte growth factor (HGF) plays a role in the improvement of cardiac function and remodeling. Their serum levels are strongly related with mortality in chronic systolic heart failure (HF). The aim of this study was to study prognostic value of HGF in acute HF, interaction with ejection fraction, renal function, and natriuretic peptides. We included 373 patients (age 76 ± 10 years, left ventricular ejection fraction [LVEF] 46 ± 14%, 48% men) consecutively admitted for acute HF. Blood samples were obtained at admission. All patients were followed up until death or close of study (>1 year, median 371 days). HGF concentrations were determined using a commercial enzyme-linked immunosorbent assay (human HGF immunoassay). The predictive power of HGF was estimated by Cox regression with calculation of Harrell C-statistic. HGF had a median of 1,942 pg/ml (interquartile rank 1,354). According to HGF quartiles, mortality rates (per 1,000 patients/year) were 98, 183, 375, and 393, respectively (p <0.001). In Cox regression analysis, HGF (hazard ratio 1SD = 1.5, 95% confidence interval 1.1 to 2.1, p = 0.002) and N-terminal pro b-type natriuretic peptide (NT-proBNP; hazard ratio 1SD = 1.8, 95% confidence interval 1.2 to 2.6, p = 0.002) were independent predictors of mortality. Interaction between HGF and LVEF, origin, and renal function was nonsignificant. The addition of HGF improved the predictive ability of the models (C-statistic 0.768 vs 0.741, p = 0.016). HGF showed a complementary value over NT-proBNP (p = 0.001): mortality rate was 490 with both above the median versus 72 with both below. In conclusion, in patients with acute HF, serum HGF concentrations are elevated and identify patients at higher risk of mortality, regardless of LVEF, ischemic origin, or renal function. HGF had independent and additive information over NT-proBNP.
Hepatocyte growth factor (HGF) is a disulfide-linked heterodimeric molecule comprising a 69-kDa kringle-containing α chain and a 34-kDa β chain. The HGF system (HGF and its receptor c- Met ) has been found in various tissues and integrates complex biologic processes. HGF has been described as a potent mitogenic growth factor for hepatocytes thought to be liver specific but also reported to possess mitogenic, motogenic, morphogenic, and antiapoptotic activities in different cell types. Binding of HGF to its receptor c-Met leads to the activation of a signal cascade whose ultimate in vivo effects include organ development, angiogenesis, and tissue regeneration. Plasma concentrations of HGF are increased in response to the damage of the liver and kidney.
In the setting of cardiovascular diseases, an increase of HGF has been observed in hypertension, atherosclerosis, acute myocardial infarction, and chronic heart failure (HF). In patients with advanced HF and depressed left ventricular ejection fraction (LVEF), HGF has been reported to be predictive of cardiovascular and all-cause mortality, especially in patients with ischemic HF. This is possibly because of that ischemia is a strong inducer of HGF synthesis. However, the value of HGF concentrations in the setting of acutely decompensated HF has not been evaluated, and data in HF patients with preserved LVEF are scarce.
Therefore, this study was aimed to analyze the prognostic information yielded by serum HGF concentrations in patients admitted for acute HF and to evaluate the interaction with EF, HF origin, renal function, and natriuretic peptides (NP).
Methods
From September 2009 to February 2013, we prospectively recruited a total of 373 patients consecutively admitted with acute HF. Local ethics committees approved the study, and informed consent was obtained from each patient. Blood samples were collected at admission and stored at −80°C until processed. Echocardiography was performed on all patients. A final diagnosis of acute HF was established following the criteria of contemporary guidelines. In addition to these criteria, a concentration of N-terminal pro b-type natriuretic peptide (NT-proBNP) >900 pg/ml at admission was an inclusion criteria. Reference values from the N-terminal Pro-BNP investigation of dyspnea in the emergency department (PRIDE) study were followed. Patients with acute coronary syndrome, significant valvular heart disease, chronic obstructive pulmonary disease as source of dyspnea, pulmonary embolism, ventricular arrhythmias, stage 5 chronic kidney disease, liver cirrhosis, hyperthyroidism, Cushing’s syndrome, and life expectancy <1 year were not included. Clinical, echocardiographic, and biochemical characteristics were prospectively recorded. Laboratory workup was performed according to manufacturer instructions.
Blood samples for HGF measurement were centralized at the immunology laboratory of one center. HGF concentrations in human serum were determined by enzyme-linked immunosorbent assay, according to manufacturer instructions (human HGF immunoassay, Quantikine ELISA; R&D Systems Europe, Ltd, Abingdon, UK). The intra-assay coefficient of variation for HGF was 7.1% for 339 pg/ml and 4.1% for 3,759 pg/ml. The interassay coefficient of variation for HGF was 7.1% for 363 pg/ml and 5.4% for 3,790 pg/ml. Glomerular filtration rate (GFR) was estimated through several formulae: Modification of Diet in Renal Disease Study equation, Chronic Kidney Disease Epidemiology Collaboration equations based on creatinine, and Hoek formula based on cystatin C.
During hospitalization, the physician responsible for patients was unaware of HGF levels. Study end point was all-cause mortality. All patients were followed up until death or close of study in August 2014 (371 days, range 2 to 1,782). Quantitative variables are described as mean values (±SD) or median values (interquartile rank). Qualitative variables are shown as frequency distribution. For comparisons between quantitative variables, either Student’s t test, ANOVA, or nonparametric Mann-Whitney or Kruskal-Wallis tests were performed. For comparison of qualitative variables, chi-square tests were used with evaluation of linear trend in variables with ordered categories. Primary variable of results was all-cause mortality. Mortality rates are expressed as events per 1,000 patients-year. For comparison between mortality rates among HGF quartiles, Kaplan-Meier and log-rank test were used. Additionally Cox regression analysis was performed to estimate hazard ratio (HR) and 95% confidence intervals (CIs) of mortality. The main predictive variable was HGF levels. It was stratified in 4 categories using quartiles as cut-off points and then merged into 2 categories according to median value. By orthogonal polynomials, it was found that the linear trend of risk produced by HGF was significant; hence, it was also examined as a quantitative variable. For this purpose, HGF was standardized to a mean of 0 and SD of 1 to enable the comparison of the size effect with NT-proBNP, which was also standardized.
The independent predictive power of mortality associated with HGF was performed with adjustment for potential confounders. Model 1 included HFG concentration and age. Model 2 included model 1 plus estimated GFR, hemoglobin, albumin, diabetes, atrial fibrillation, and ischemic HF. Model 3 included model 2 plus LVEF and NT-proBNP concentrations. Model 4 included model 3 plus diuretic doses and therapies for HF. In model 4, a sequential exclusion was performed to evaluate independent predictors of mortality. The predictive ability of HGF was tested by comparing the C Harrell statistical applied to variables included in model 4 both with and without introduction of HGF and by calculating the Integrated Discrimination Improvement.
Finally, patients were classified into 4 categories according to a combination of HGF and NT-proBNP values above or below median: category 1: both below median; category 2: only NT-proBNP above median; category 3: only HGF above median; category 4: both above median. A Kaplan-Meier analysis and a multivariate Cox regression analysis were also conducted using this combined variable to assess prognosis. Statistical results were considered significant when the value of p was <0.05. Statistical procedures were performed with IBM Statistical Package for the Social Sciences, version 20.0 (IBM Corp. Released 2011. IBM SPSS Statistics for Windows, Version 20.0., IBM Corp, Armonk, New York).
Results
Three hundred seventy-three patients admitted to the hospital with acute HF were recruited. Characteristics for the population are presented in Table 1 . Preserved ejection fraction (LVEF ≥50%) represented 51% of patients. HGF values were available for 338 of patients (90.6%). Patients with missing values did not differ in any of the variables analyzed (data not shown). Characteristics of patients according to HGF quartiles are listed in Table 2 . Concentrations of HGF were associated with higher NT-proBNP and cystatin C values and less satisfactory renal function values.
| Variable | Total population (N=373) | Alive (N=279) (74.8%) | Deaths (N=94) (25.2%) | p |
|---|---|---|---|---|
| Age (years) | 76.32 (10.02) | 75.25 (10.37) | 79.50 (8.17) | <0.001 |
| Male/Female | 48.2%/51.8% | 47.5%/52.5% | 50.5%/49.5% | 0.61 |
| LVEF | 46.43 (14.15) | 47.05 (13.97) | 44.51 (14.57) | 0.14 |
| LVEF >50% | 51.1% | 52.6% | 46.6% | 0.32 |
| LA (mm) | 47.31 (7.84) | 46.62 (7.69) | 49.77 (7.94) | 0.007 |
| LVMI (g/m 2 ) | 129.37 (46.39) | 130.70 (45.97) | 124.75 (48.12) | 0.49 |
| BMI (Kg/m 2 ) | 29.51 (5.55) | 29.73 (5.76) | 28.86 (4.84) | 0.19 |
| Atrial Fibrillation | 51.5% | 51.6% | 51.1% | 0.93 |
| Hypertension | 76.9% | 76.7% | 77.7% | 0.85 |
| IHD | 36.7% | 37.6% | 34% | 0.53 |
| Valvular disease | 39.2% | 37.3% | 45.5% | 0.23 |
| Diabetes mellitus | 47.2% | 45.2% | 53.2% | 0.17 |
| Hyperlipidemia | 43.3% | 49.3% | 26.1% | 0.001 |
| Stroke | 16.3% | 14.1% | 23.2% | 0.075 |
| Peripheral vascular disease | 10.9% | 10.6% | 12.2% | 0.77 |
| Hemoglobin (g/dL) | 12.32 (2.16) | 12.41 (2.15) | 12.02 (2.15) | 0.13 |
| Uric acid (mg/dL) | 7.71 (3.22) | 7.53 (3.41) | 8.31 (2.39) | 0.32 |
| Blood urea (mg/dL) | 58.16 (47.42) | 52.40 (42.99) | 75.20 (55.42) | <0.001 |
| Creatinine (mg/dL) | 1.45 (0.96) | 1.39 (0.98) | 1.61 (0.86) | 0.06 |
| Sodium (mEq/l) | 138.88 (4.72) | 139.15 (4.50) | 138.08 (5.26) | 0.057 |
| Potassium (mEq/l) | 4.25 (0.646) | 4.21 (0.64) | 4.36 (0.65) | 0.044 |
| Albumin < 3.5 g/dl | 34.2% | 30.3% | 47.5% | 0.015 |
| Hyperlipidemia | 36.7% | 49.3% | 26.1% | 0.001 |
| NT-proBNP * (pg/mL) | 4233 (7133) | 3556 (5461) | 7489 (15077) | <0.001 |
| HGF * (pg/mL) | 1942 (1354) | 1759 (1262) | 2299 (1439) | <0.001 |
| Cystatin * (mg/L) | 1.45 (0.86) | 1.37 (0.85) | 1.67 (0.81) | <0.001 |
| eGFR MDRD (mL/min/1.72m 2 ) | 54.97 (24.54) | 56.94 (24.79) | 49.07 (22.89) | 0.007 |
| eGFR CKD-EPI (mL/min/1.72m 2 ) | 50.98 (21.70) | 53.02 (21.70) | 44.86 (20.59) | 0.002 |
| eGFR Hoek13 (mL/min/1.72m 2 ) | 52.22 (21.48) | 54.99 (22.30) | 43.85 (16.23) | 0.000 |
| Q1 ≤ 1350 Mean ± SD | Q2 1350-1900 Mean ± SD | Q3 1901-2700 Mean ± SD | Q4 ≥ 2700 Mean ± SD | P | |
|---|---|---|---|---|---|
| HGF (pg/mL) | 1127 (226) | 1642 (254) | 2299 (313) | 3466 (1699) | <0.001 |
| LVEF | 45.72 (14.92) | 45.81 (13.52) | 48.26 (13.51) | 45.55 (14.44) | 0.56 |
| LVEF >50% | 49.4% | 53% | 58% | 43.9% | 0.32 |
| LA (mm) | 45.86 (9.42) | 47.71 (8.96) | 48.92 (6.71) | 46.15 (6.45) | 0.12 |
| LVMI (g/m2) | 151.17 (71.16) | 129.18 (37.09) | 124.15 (39.58) | 125.79 (45.91) | 0.17 |
| Age (years) | 74.19 (11.42) | 76.38 (9.73) | 77.32 (9.31) | 77.06 (10.08) | 0.18 |
| Male/Female | 60.7%/39.3% | 50.6%/49.4% | 43.5%/56.5% | 42.4%/57.6% | 0.06 |
| BMI (Kg/m 2 ) | 29.02 (5.47) | 28.70 (4.74) | 29.81 (6.26) | 30.52 (5.64) | 0.14 |
| Albumin < 3.5 mg/dL | 25.6% | 32.1% | 43.5% | 38.2% | 0.11 |
| Hemoglobin (g/dL) | 12.43 (2) | 12.43 (2.23) | 12.14 (2.06) | 12.15 (2.31) | 0.69 |
| Uric acid (mg/dL) | 7.49 (4.70) | 6.78 (1.59) | 8.17 (2.48) | 7.78 (2.48) | 0.66 |
| Sodium (mEq/L) | 137.84 (4.63) | 139.35 (4.55) | 139.58 (4.45) | 138.84 (5.19) | 0.08 |
| Potasium (mEq/L) | 4.22 (0.58) | 4.24 (0.70) | 4.28 (0.55) | 4.27 (0.70) | 0.93 |
| Blood urea (mg/dL) | 48.58 (31.87) | 56.18 (60.61) | 56.17 (46.47) | 69.29 (48.15) | 0.04 |
| Creatinine (mg/dL) | 1.21 (0.38) | 1.52 (1.51) | 1.46 (0.81) | 1.63 (0.87) | 0.04 |
| Atrial fibrillation | 48.8% | 58.8% | 58.8% | 38.8% | 0.21 |
| Hypertension | 75% | 72.9% | 76.5% | 82.4% | 0.21 |
| IHD | 38.1% | 35.3% | 41.2% | 37.6% | 0.84 |
| Valvular disease | 34% | 54.7% | 40.6% | 29.2% | 0.16 |
| Diabetes mellitus | 41.7% | 44.7% | 49.4% | 49.4% | 0.25 |
| Hyperlipidemia | 47.9% | 38.2% | 51.8% | 41.4% | 0.78 |
| Stroke | 16.7% | 17.9% | 15.5% | 17.2% | 0.98 |
| Peripheral vascular disease | 5.9% | 5.7% | 22.5% | 13.3% | 0.12 |
| eGFR MDRD (mL/min/1.72m 2 ) | 62.13 (21.82) | 57.75 (29.04) | 52.5 (22.96) | 47.46 (22.36) | 0.001 |
| eGFR CKD-EPI (mL/min/1.72m 2 ) | 58.01 (19.80) | 53.09 (23.26) | 48.61 (20.84) | 44.12 (20.97) | <0.001 |
| eGFR Hoek 13 (mL/min/1.72m 2 ) | 58.49 (23.45) | 51.90 (19.39) | 50.48 (21.48) | 46.82 (18.94) | 0.004 |
| NT-proBNP * (pg/mL) | 3195 (5041) | 4513 (7079) | 4661 (6715) | 5408 (12346) | 0.004 |
| Cystatin C * (mg/L) | 1.32 (0.59) | 1.44 (0.82) | 1.52 (1.07) | 1.67 (0.90) | 0.008 |
| Loop diuretics | 75.6% | 83.1% | 88.7% | 90.5% | 0.007 |
| ACE inhibitors | 62.7% | 55.3% | 46.9% | 50.6% | 0.016 |
| ARB | 24.1% | 28.2% | 30.9% | 22.6% | 0.8 |
| MRA | 45.5% | 33.3% | 50% | 43.3% | 0.9 |
| BB | 60.2% | 57.6% | 66.7% | 57.1% | 0.9 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
