Skip main navigation

G-CSF for Extensive STEMI

Results From the STEM-AMI OUTCOME CMR Substudy
Originally published Research. 2019;125:295–306



In the exploratory Phase II STEM-AMI (Stem Cells Mobilization in Acute Myocardial Infarction) trial, we reported that early administration of G-CSF (granulocyte colony-stimulating factor), in patients with anterior ST-segment–elevation myocardial infarction and left ventricular (LV) dysfunction after successful percutaneous coronary intervention, had the potential to significantly attenuate LV adverse remodeling in the long-term.


The STEM-AMI OUTCOME CMR (Stem Cells Mobilization in Acute Myocardial Infarction Outcome Cardiac Magnetic Resonance) Substudy was adequately powered to evaluate, in a population showing LV ejection fraction ≤45% after percutaneous coronary intervention for extensive ST-segment–elevation myocardial infarction, the effects of early administration of G-CSF in terms of LV remodeling and function, infarct size assessed by late gadolinium enhancement, and myocardial strain.

Methods and Results:

Within the Italian, multicenter, prospective, randomized, Phase III STEM-AMI OUTCOME trial, 161 ST-segment–elevation myocardial infarction patients were enrolled in the CMR Substudy and assigned to standard of care (SOC) plus G-CSF or SOC alone. In 119 patients (61 G-CSF and 58 SOC, respectively), CMR was available at baseline and 6-month follow-up. Paired imaging data were independently analyzed by 2 blinded experts in a core CMR lab. The 2 groups were similar for clinical characteristics, cardiovascular risk factors, and pharmacological treatment, except for a trend towards a larger infarct size and longer symptom-to-balloon time in G-CSF patients. ANCOVA showed that the improvement of LV ejection fraction from baseline to 6 months was 5.1% higher in G-CSF patients versus SOC (P=0.01); concurrently, there was a significant between-group difference of 6.7 mL/m2 in the change of indexed LV end-systolic volume in favor of G-CSF group (P=0.02). Indexed late gadolinium enhancement significantly decreased in G-CSF group only (P=0.04). Moreover, over time improvement of global longitudinal strain was 2.4% higher in G-CSF patients versus SOC (P=0.04). Global circumferential strain significantly improved in G-CSF group only (P=0.006).


Early administration of G-CSF exerted a beneficial effect on top of SOC in patients with LV dysfunction after extensive ST-segment–elevation myocardial infarction in terms of global systolic function, adverse remodeling, scar size, and myocardial strain.

Clinical Trial Registration:

URL: Unique identifier: NCT01969890.


Despite groundbreaking progresses in the treatment of ST-segment–elevation myocardial infarction (STEMI) achieved by early reperfusion with primary percutaneous coronary interventions (PCI), the occurrence of adverse left ventricular (LV) remodeling remains associated with worse clinical outcome.1–4 Adjunctive etiologic therapies aimed to reduce the amount of cell death after STEMI are at present still lacking.5

Editorial, see p 307

In This Issue, see p 259

Meet the First Author, see p 260

Regenerative therapies have been investigated on top of standard of care (SOC) for their potential to restore the damaged myocardium subsequent to ischemic injury.6,7 Based on a consistent body of evidence showing that the bone marrow (BM) is a source of cells which interplays with post-infarction processes leading to scar formation and maturation,8,9 clinical trials of intracoronary delivery10–13 or cytokine mobilization14–23 of BM-cells have been performed with the goal to boost such a physiological response to ischemia.

G-CSF (granulocyte colony-stimulating factor) is an endogenous hematopoietic cytokine, produced by monocytes, fibroblasts, and endothelial cells, that is involved in the mobilization of granulocytes, stem, and progenitor cells from the BM into the blood circulation.24 In addition, G-CSF, binding with its receptor expressed by cardiomyocytes after an ischemic injury, directly activates prosurvival signaling.25 Although G-CSF has been consistently confirmed in experimental animal models to act as a regenerative and cardioprotective drug after STEMI,26,27 however, clinical studies testing this hypothesis in humans have generated conflicting results.28,29

In the Phase II exploratory STEM-AMI trial (Stem Cells Mobilization in Acute Myocardial Infarction),30 we have previously showed by cardiac magnetic resonance (CMR) that, 6 months after STEMI, indexed LV end-diastolic volume (LVEDVI) and infarct size were significantly reduced in patients receiving G-CSF as compared with placebo. Such an attenuation of adverse LV remodeling was further confirmed at 3-year follow-up.31

Prompted by these evidences, the nationwide, multicenter, randomized, controlled, Phase III STEM-AMI OUTCOME trial (Stem Cells Mobilization in Acute Myocardial Infarction Outcome)32 was designed to test the premise that early G-CSF therapy (Filgrastim, Hexal, Holzkirchen, Germany) in high-risk patients with STEMI may favorably impact their long-term outcome, by decreasing mortality and cardiac-related morbidity. Within such a pivotal trial, aimed at enrolling 1530 patients to G-CSF versus SOC randomized 1:1 ratio, the STEM-AMI OUTCOME CMR Substudy has been carved out to consistently test, by means of an adequate sample size, the hypothesis that G-CSF can favorably impact on LV remodeling and function, infarct size, and myocardial strain.


The data that support the findings of this study are available from the corresponding author on reasonable request.

Study Design

The STEM-AMI OUTCOME CMR Substudy trial was a substudy of the STEM-AMI OUTCOME trial.32 The study complied with the Declaration of Helsinki and was approved by the Italian Medicines Agency (Agenzia Italiana del Farmaco, AIFA) and the local Ethical Committees of participating Institutes.

The STEM-AMI OUTCOME trial was designed to randomize to G-CSF or SOC (in a 1:1 ratio) patients with a first anterior STEMI undergoing a successful primary PCI or PCI rescue between 2 and 12 hours after symptom onset (or within 24 hours if symptoms persist) and with evidence of LV dysfunction within 24 hours after revascularization (LV ejection fraction [LVEF] ≤45% on 2D transthoracic echocardiography).

Patients randomized to treatment received, within 24 hours after reperfusion, 5 µg/kg G-CSF subcutaneously twice a day for 6 consecutive days (from day 0 to day 5) and prematurely interrupted if the threshold value of 50.000 white blood cells was reached. This threshold was set as hyperleukocytosis stopping rule in the previous STEM-AMI and similar G-CSF studies in STEMI.16,23

The exclusion criteria were (1) previous anterior myocardial infarction (MI); (2) known previous LV dysfunction (LVEF <45%); (3) angiographic evidence of coronary anatomy not suitable for PCI, or needing coronary artery bypass grafting (CABG); and (4) previous history of coronary artery bypass grafting or PCI on left anterior descending artery within 6 months. Additional exclusion criteria for the CMR Substudy was a known contraindication to CMR.

The primary composite end point of the main study was a reduced occurrence of any of the following in the G-CSF group compared with the control group over 2-year of follow-up: all-cause death, recurrence of MI, or hospitalization due to heart failure.

In February 2015, the STEM-AMI OUTCOME trial was amended to increase recruitment rates. The amendment widened patient enrollment to all-comer first STEMI with evidence of LV dysfunction after successful primary PCI having symptoms-to-balloon time within 12 or 24 hours if symptoms persisted. The study was stopped in February 2016 for sponsor’s decision after Data Safety Monitoring Board recommendation due to low recruitment rate. Overall, 532 patients were recruited.

Twelve out of 44 participating Italian centers in the STEM-AMI OUTCOME trial joined the CMR Substudy. A complete list of CMR Substudy participating centers and investigators is available in Online Table I as well as the Steering Committee member list (Online Table II). Patients included in the STEM-AMI OUTCOME CMR Substudy were scheduled to perform CMR at 7 and 180 days after enrollment to assess LVEF, LVEDV, LV end-systolic volume (LVESV), infarct size, presence of microvascular obstruction (MVO) and myocardial strain.

Based on the results of the phase II STEM-AMI trial,30,31 the primary end point of the CMR Substudy was a LVEDV change of ≥10 mL/m2 between baseline (day 7; ie, T0) and after 6 months (day 180; ie, T6). The secondary end points were a LVEF change >5% and infarct size reduction.32 Safety end points were survival, major adverse cardiac and cerebrovascular events (MACCE), and rate of restenosis. MACCE were defined as death, reinfarction, heart failure, urgent revascularization, and stroke. Clinical events and MACCE were also recorded over the 6-month follow-up.

A main potential source of bias might have stemmed from the open design of this randomized controlled study. Adopted limitation strategies included: (1) CMR studies performed by experts blinded to treatment allocation; (2) imaging interpretation in a core laboratory, with blinded expert readers; and (3) adjudication of clinical events by an Event Validation Committee.

The reporting of this study was in agreement to the STROBE statement (Strengthening The Reporting of OBservational Studies in Epidemiology; Online Table III).33

Written informed consent was obtained for both the STEM-AMI OUTCOME trial and the CMR Substudy.

CMR Protocol and Analysis

A detailed description of the method is available in the Online Data Supplement.

Statistical Analysis

Details of statistical analysis are described in the Online Data Supplement.


Baseline and Infarction-Related Characteristics

Centers participating in the CMR Substudy recruited and randomly assigned to G-CSF or SOC a total of 199 consecutive patients between November 2013 and December 2015. The flow of participants throughout the trial is detailed in Figure 1. Among the 161 patients who gave a written informed consent to the CMR Substudy, baseline CMR imaging was obtained for 135 patients (n=69 G-CSF-treated subjects and n=66 controls). Fifteen patients (8 in the G-CSF and 7 in the control group) were lost at follow-up or did not perform the 6-month CMR, while one 6-month CMR imaging set in the G-CSF group was technically insufficient for analysis. Thus, paired imaging data sets were available for 119 patients (n=61 G-CSF-treated subjects and n=58 controls). Mean follow-up time was 6.3±1.4 months.

Figure 1.

Figure 1. Flow chart of the STEM-AMI OUTCOME CMR (Stem Cells Mobilization in Acute Myocardial Infarction Outcome Cardiac Magnetic Resonance) Substudy. Patients were randomized 1:1 to G-CSF (granulocyte colony-stimulating factor) or to standard of care (SOC) and invited to participate in the CMR Substudy. Patients enrolled in the substudy, undergoing baseline CMR, were followed for 6 mo. CMR was repeated at follow-up. AMI indicates acute myocardial infarction; EF, ejection fraction; and PCI, primary coronary intervention.

The demographic and clinical baseline characteristics and MI-related characteristics are listed in Table 1 and Table 2, respectively.

Table 1. Demographic, Clinical, and Biochemical Patients’ Characteristics

VariableG-CSF (n=61)SOC (n=58)P Value
Age, y60±1062±100.44
Male sex, n (%)50 (82.0)49 (84.5)0.81
Body mass index, kg/m226.7±4.925.7±3.00.16
Cardiovascular risk factors
 Family history of CAD, n (%)18 (29.5)15 (25.9)0.69
 Current smokers, n (%)30 (49.2)26 (44.8)0.71
 Diabetes mellitus, n (%)10 (16.4)13 (22.4)0.49
 Hypertension, n (%)17 (27.9)21 (36.2)0.43
 Dyslipidemia, n (%)19 (31.2)24 (41.4)0.26
Medical history
 Prior MI, n (%)4 (6.6)6 (10.3)0.52
 Prior TIA, n (%)1 (1.6)0 (0.0)1.00
 Angina, n (%)4 (6.6)4 (6.9)1.00
 CKD, n (%)4 (6.6)1 (1.7)0.37
 COPD, n (%)3 (4.9)1 (1.7)0.62
 PAD, n (%)4 (6.6)1 (1.7)0.37
Blood analysis on admission
 Hemoglobin, g/dL14.5±1.414.8±1.40.31
 WBC count (×103/mm3)11.8±3.311.5±4.40.69
 Platelet count (×103/mm3)235±57236±590.96
 Creatinine, mg/dL0.92±0.330.93±0.230.88
 Glycemia, mg/dL149±45159±570.27
 C-reactive protein, mg/dL*1.6±3.42.1±7.60.74
 NT-proBNP, pg/mL1812±20821321±13470.27
Medication at discharge, n (%)
 Aspirin60 (98.4)57 (98.3)1.00
 Clopidogrel4 (6.6)5 (8.6)0.74
 Prasugrel30 (49.2)29 (50.0)1.00
 Ticagrelor27 (44.3)24 (41.4)0.85
 Statins60 (98.4)56 (96.6)0.61
 ACE inhibitors56 (91.8)51 (87.9)0.55
 Angiotensin receptor blockers2 (3.3)4 (6.9)0.43
 β-blockers58 (95.1)56 (96.6)1.00
 Diuretics23 (37.7)25 (43.1)0.58

Categorical variables are presented as counts (n) and proportions (%); quantitative variables are expressed as mean±SD. ACE indicates angiotensin-converting enzyme; BNP, natriuretic peptide type B; CAD, coronary artery disease; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; G-CSF, granulocyte colony-stimulating factor; MI, myocardial infarction; PAD, peripheral artery disease; SOC, standard of care; TIA, transient ischemic attack; and WBC, white blood cell.

*n=48 and n=40 for the G-CSF and SOC groups, respectively.

†n=31 and n=28 for the G-CSF and SOC groups, respectively.

Table 2. Myocardial Infarction Related and Angiographic Characteristics

VariableG-CSF (n=61)SOC (n=58)P Value
Cardiovascular condition
 SBP, mm Hg136.5±26.5136.1±24.10.93
 DBP, mm Hg81.7±14.682.9±14.80.65
 Killip class ≥3, n (%)4 (6.6)2 (3.4)0.68
 IABP, n (%)3 (4.9)2 (3.4)1.00
 ECG involved leads, n (%) Pts >5 leads21 (34.4)22 (37.9)0.71
 Echo LVEF, %38.5±5.638.8±4.80.71
 CK-MB peak, mg/L254.5±178.2201.0±155.60.08
 Troponin peak, quintiles3.1±1.52.9±1.40.40
 MI site, n (%)
  Anterior46 (75.4)46 (79.3)0.67
  Inferior4 (6.6)1 (1.7)0.37
  Lateral1 (1.6)2 (3.4)0.61
  Inferior/lateral2 (3.3)2 (3.4)1.00
  Anterior/lateral8 (13.1)6 (10.3)0.79
  Anterior/inferior0 (0.0)1 (1.7)0.49
Angiographic data
 Onset of MI to PCI, h5.8±4.34.5±3.60.08
 TIMI flow 0 pre-PCI, n (%)44 (72.1)40 (70.0)0.84
 TIMI flow 3 post-PCI, n (%)58 (95.1)56 (97.0)1.00
 PCI plus stent, n (%)60 (98.4)58 (100)1.00
 Drug eluting stent, n (%)59 (96.7)54 (93.1)0.43
 PCI rescue, n (%)4 (6.5)4 (6.8)1.00
 Multivessel disease, n (%)32 (52.5)31 (53.4)1.00
 PCI to G-CSF administration, min784.6±501.7
Antithrombotic treatments, n (%)
 Aspirin59 (96.7)54 (93.1)0.43
 IIb/IIIa inhibitors23 (37.7)26 (44.8)0.46
 Clopidogrel3 (4.9)3 (5.2)1.00
 Prasugrel29 (47.5)29 (50.0)0.86
 Ticagrelor26 (42.6)24 (41.4)1.00

Categorical variables are presented as counts (n) and proportions (%); quantitative variables are expressed as mean±SD. CK-MB indicates creatine kinase-myocardial band; DBP, diastolic blood pressure; G-CSF, granulocyte colony-stimulating factor; HR, heart rate; IABP, intraaortic balloon pump; LVEF, left ventricular ejection fraction; MI, myocardial infarction; PCI, primary coronary intervention; Pts, patients; SBP; systolic blood pressure; SOC, standard of care; and TIMI, thrombolysis in myocardial infarction.

Ninety-two out of 119 patients (78%) had anterior MI. The 2 groups were similar for clinical features, cardiovascular risk factors, and medical therapy at discharge (Table 1). Moreover, as shown in Table 2, the 2D echocardiographic LVEF at enrollment was comparable in the 2 groups (38.5%±5.6 versus 38.8%±4.8 in the G-CSF and SOC groups, respectively). The G-CSF group exhibited a not-significant trend for a larger MI than the SOC group, as indexed by CK-MB (creatine kinase-myocardial band) peak levels, a worse Killip score, and longer symptom-to-balloon time.

G-CSF was administered an average of 13.1±8.4 hours after PCI. Patients received on average 7.7±3.3 administrations. According to the study protocol, G-CSF treatment was prematurely stopped in 41 patients who reached 50.000 white blood cell count. This latter group of patients received on average 7.1±3.0 G-CSF doses. No differences were observed in drug treatment at discharge between the 2 groups.

Clinical Outcome

Online Table V summarizes clinical events registered in hospital and during the 6-month follow-up in both treatment groups. We did not observe any MACCE or malignant arrhythmias during hospitalization, except for one urgent revascularization in the SOC group. Overall, a low incidence of nonfatal cardiac and noncardiovascular events over the 6-month observation period was recorded, with no significant differences between groups. Cumulatively, MACCE did not significantly change between the 2 groups at follow-up, even if their incidence was slightly higher among SOC with respect to G-CSF patients (10.3% versus 3.3%, respectively).

CMR Findings

Raw data of LVEF, LVEDV, and LVESV at baseline and follow-up for each G-CSF, and SOC patient are reported in Online Figure I.

Baseline and 6-month CMR quantitative measurements are listed in Online Table VI. Consistent with the CK-MB peak data, G-CSF patients showed on average a significantly larger infarct size than SOC subjects, as assessed by LGE mass, that is, the mean extent of MLGE/MLV was 4.5% greater in the G-CSF than in the SOC group (P=0.0224). This trend was maintained at 6-month follow-up (P=0.0515). Moreover, baseline MVO was more present, although not significantly, in G-CSF patients.

Functional Outcome

Changes over time of CMR-measured parameters are summarized in Table 3.

Table 3. Changes Over Time of CMR-Measured Parameters Unadjusted and Adjusted for Baseline Variables

ΔT6–T0 in CMR FindingsAnalysisG-CSFSOCP Value
LVEF, %Model 14.21 (1.45 to 6.97)0.22 (−2.61 to 3.06)0.0484
Model 24.72 (1.94 to 7.50)−0.32 (−3.17 to 2.53)0.0153
Model 34.46 (1.71 to 7.21)−0.04 (−2.86 to 2.78)0.0270
Full model4.75 (1.99 to 7.51)−0.35 (−3.18 to 2.49)0.0142
LVEDVI, mL/m2Model 10.93 (−3.11 to 4.97)0.34 (−3.80 to 4.49)0.8410
Model 2−0.10 (−4.11 to 3.91)1.42 (−2.69 to 5.54)0.6071
Model 3−0.11 (−4.04 to 3.82)1.44 (−2.60 to 5.47)0.5920
Full model−0.63 (−4.59 to 3.33)1.98 (−2.08 to 6.05)0.3747
LVESVI, mL/m2Model 1−2.53 (−6.53 to 1.46)0.92 (−3.18 to 5.01)0.2347
Model 2−3.68 (−7.57 to 0.21)2.12 (−1.88 to 6.11)0.0456
Model 3−3.54 (−7.38 to 0.30)1.97 (−1.97 to 5.91)0.0522
Full model−4.12 (−7.94 to −0.29)2.58 (−1.35 to 6.51)0.0198
LGE (% of LV myocardial mass)Model 1−2.76 (−5.06 to −0.46)−2.21 (−4.54 to 0.11)0.7407
Model 2−3.07 (−5.42 to −0.73)−1.90 (−4.27 to 0.47)0.4945
Model 3*−2.87 (−5.19 to −0.54)−2.10 (−4.45 to 0.24)0.6487
Full model*−3.10 (−5.48 to −0.73)−1.50 (−4.09 to 1.10)0.3755
2D-GCS, %Model 1−2.93 (−4.85 to −1.00)−1.53 (−3.50 to 0.45)0.3161
Model 2−3.12 (−5.10 to −1.13)−1.33 (−3.37 to 0.71)0.2253
Model 3−3.06 (−5.05 to −1.06)−1.51 (−3.68 to 0.65)0.3032
Full model−3.31 (−5.36 to −1.26)−1.34 (−3.55 to 0.86)0.2071
2D-GLS, %Model 1−2.52 (−3.96 to −1.08)−0.65 (−2.13 to 0.83)0.0757
Model 2−2.62 (−4.11 to −1.13)−0.54 (−2.07 to 0.99)0.0616
Model 3−2.60 (−4.09 to −1.11)−0.28 (−1.90 to 1.34)0.0392
Full model−2.67 (−4.21 to −1.12)−0.23 (−1.90 to 1.44)0.0401

Data are expressed as mean (95% CI). Model 1 was the unadjusted analysis. Model 2 was adjusted for symptom-to-balloon time, CK-MB peak, and Killip score. Model 3 was adjusted for baseline MLGE/MLV and MVO. The full model was adjusted for symptom-to-balloon time, CK-MB peak, Killip score, baseline MLGE/MLV, and MVO. CK-MB indicates creatine kinase-MB; CMR, cardiac magnetic resonance; 2D-GCS, 2D global circumferential strain; 2D-GLS, 2D global longitudinal strain; LVEDVI, left ventricular end-diastolic volume indexed; LVEF, left ventricular ejection fraction; LVESVI, left ventricular end-systolic volume indexed; MLGE, mass of the segments with late gadolinium enhancement; MLV, left ventricular mass; and MVO, microvascular obstruction.

*Baseline MLGE/MLV was not included in these models.

Data are presented unadjusted and adjusted for baseline variables that are independent predictors of worse outcome and show an imbalance between G-CSF and SOC groups. Model 1 was the univariable unadjusted analysis. Model 2 was adjusted for clinical variables related to STEMI prognosis (symptom-to-balloon time, CK-MB peak, and Killip score). Model 3 was adjusted for CMR baseline variables associated with adverse LV remodeling and worse outcome (LGE expressed as percentage of LV myocardial mass, and MVO). The full model was adjusted for all of the abovementioned baseline variables.

The G-CSF group showed a significant increase in LVEF from baseline to 6 months (a 4.2% gain on average, P=0.0063; Figure 2A), whereas no appreciable change was observed in the SOC group, with a significant interaction between time and treatment (P=0.0484), that is, suggesting that the magnitude of the effect of time on LVEF recovery was dependent on treatment. Most importantly, a significant difference in the change over time of the LVEF between the G-CSF and the SOC group was confirmed in all Models of the ANCOVA (Table 3), stepwise adjusting for baseline clinical and CMR variables related to MI extent and severity (symptom-to-balloon time, CK-MB peak, Killip score, MLGE/MLV, and MVO). Indeed, ΔLVEFT6–T0 was 5.1% higher on average in the G-CSF than in the SOC group (P=0.0142; Figure 2B). Consistently, 6-month LVEF was significantly higher in the G-CSF versus SOC group, after adjustment for the same baseline covariates (P=0.0446; Online Figure II). Finally, an analysis of the relationship between the symptom-to-balloon time and the change over time in LVEF showed that treatment with G-CSF resulted in a trend increase (P=0.075) towards the increase of the proportion of patients with LVEF improvement above the median value (2.5% in ΔLVEFT6–T0) when reperfused beyond 300 minutes (Online Figure III).

Figure 2.

Figure 2. Changes in the left ventricular ejection fraction (LVEF) from baseline (T0) to 6 months (T6). A, LVEF increased significantly in the G-CSF (granulocyte colony-stimulating factor) group and not in the standard of care (SOC) group. Data are presented as mean±SEM. B, There was a significant difference in the Δ increment of LVEF over time (from T0–T6) in G-CSF vs SOC-treated patients, when adjusted for symptom-to-balloon time, creatine kinase-myocardial band (CK-MB) peak, Killip score, baseline MLGE/MLV, and microvascular obstruction (MVO). Data are presented as estimated marginal means±SEM and plotted with 95% CI. MLGE indicates mass of the segments with late gadolinium enhancement; and MLV, left ventricular mass.

LVEDVI did not change in both groups from baseline to 6 months (Figure 3A). Consistently, the change over time of LVEDVI was not significantly different between G-CSF and SOC patients, even after adjustment for the abovementioned baseline variables (P=0.3747; Figure 3B), although a slight LVEDVI increase was present in SOC patients only. Similarly, LVESV indexed (LVESVI) did not significantly change in both groups from baseline to 6 months (Figure 3C), even if a decrease was observed in G-CSF patients only. However, covariance analysis showed that, when adjusting for baseline MI-related variables, there was a significant difference in the change of LVESVI from baseline to 6-month follow-up between the 2 groups: G-CSF patients tended to reduce and SOC patients to increase mean LVESVI, which resulted in a mean difference of 6.7 mL/m2 in ΔLVESVIT6–T0 (P=0.0198; Figure 3D).

Figure 3.

Figure 3. Changes in the indexed left ventricular end-diastolic (LVEDVI) and indexed left ventricular end-systolic (LVESVI) volumes from baseline (T0) to 6 months (T6). LVEDVI did not change over time in both the G-CSF (granulocyte colony-stimulating factor) and the standard of care (SOC) groups (A), whereas LVESVI showed a decreasing trend in the G-CSF group only (C). Data are presented as mean±SEM. When adjusted for symptom-to-balloon time, CK-MB (creatine kinase-MB) peak, Killip score, baseline MLGE/MLV, and microvascular obstruction (MVO), LVEDVI change over time (ΔT6–T0) showed an increasing trend in the SOC group (B). On the contrary, there was a significant decrease in the ΔT6–T0 of LVESVI in G-CSF vs SOC-treated patients (D). Data are presented as estimated marginal means±SEM and plotted with 95% CI. MLGE indicates mass of the segments with late gadolinium enhancement; and MLV, left ventricular mass.

The analysis of the change of LGE mass normalized to LV mass (MLGE/MLV) from baseline to follow-up in the 2 groups resulted in a significant reduction in MLGE/MLV over time in the G-CSF group only (P=0.0383; Figure 4A), while the SOC group did not change, with a significant effect of the treatment factor (P=0.0423). The change over time of indexed LGE mass was not significantly different between G-CSF and SOC patients, even after adjustment for baseline MI-related variables (P=0.3755; Figure 4B), although a reduction was evident in the G-CSF group only.

Figure 4.

Figure 4. Changes in the mass of the segments with late gadolinium enhancement (MLGE), normalized to the left ventricular mass (MLV), from baseline (T0) to 6 months (T6). A, MLGE/MLV decreased significantly in the G-CSF (granulocyte colony-stimulating factor) group and not in the standard of care (SOC) group. Data are presented as mean±SEM. B, The Δ change over time in LGE showed a decreasing trend in G-CSF vs SOC-treated patients, when adjusted for symptom-to-balloon time, CK-MB (creatine kinase-MB) peak, Killip score, and baseline microvascular obstruction (MVO). Data are presented as estimated marginal means±SEM and plotted with 95% CI.

Myocardial Strain Outcome

We also compared temporal changes of 2D longitudinal and circumferential strains in G-CSF versus SOC patients. 2D-GLS significantly improved in both groups, with a more pronounced amelioration in G-CSF patients (P<0.0001 and P=0.0129 in the G-CSF and SOC groups, respectively; Figure 5A). Consistently, when adjusted for the aforementioned baseline MI-related variables, the Δ improvement over time was significantly higher in the G-CSF group (−2.44% on average in G-CSF versus SOC group, P=0.0401; Figure 5B). In analogy, 2D-GCS improved significantly in the G-CSF group only (P=0.0063; Figure 5C), although the difference in the change over time between the 2 groups did not reach statistical significance after adjustment for baseline MI-related features (P=0.2071; Figure 5D).

Figure 5.

Figure 5. Changes in the 2-dimensional global longitudinal (2D-GLS) and 2-dimensional global circumferential (2D-GCS) strains from baseline (T0) to 6 months (T6).A–C, Both 2D-GLS and 2D-GCS decreased significantly more in the G-CSF (granulocyte colony-stimulating factor) with respect to the standard of care (SOC) group. Data are presented as mean±SEM. B, There was a significant difference between the Δ decrease over time of 2D-GLS in G-CSF (granulocyte colony-stimulating factor) vs SOC-treated patients, when adjusted for symptom-to-balloon time, CK-MB (creatine kinase-MB) peak, Killip score, baseline MLGE/MLV, and microvascular obstruction (MVO), whereas (D) only a trend in 2D-GCS. Data are presented as estimated marginal means±SEM and plotted with 95% CI. MLGE indicates mass of the segments with late gadolinium enhancement; and MLV, left ventricular mass.


The main finding of this study is that, in patients with LV dysfunction after successfully reperfused STEMI, G-CSF treatment is effective in improving LV performance, as shown by a significant increase of LVEF and 2D-GLS due to reduced scar size, as well as in attenuating adverse LV remodeling as demonstrated by the significant reduction in LVESVI. To the best of our knowledge, we report here the largest study in STEMI investigating, by CMR, the cardioprotective and antiremodeling benefits of early administration of G-CSF. These results corroborate the previous findings of the STEM-AMI trial, in which we have shown, in patients with a large anterior STEMI conditioning LV dysfunction receiving early G-CSF treatment versus placebo, an attenuation of LV remodeling in terms of adjusted LVEDVI at 6-month30 and 3-year follow-up.31

The STEM-AMI OUTCOME CMR Substudy was conducted in the context of the larger STEM-AMI OUTCOME prospective, nationwide, multicenter, randomized, open-label, Phase III trial,32 which was designed to randomize (1:1) 1530 STEMI patients to G-CSF versus SOC with the primary end point to assess whether G-CSF could reduce the occurrence of all-cause death, recurrence of MI or hospitalization due to heart failure. The sample size of the CMR Substudy (120 patients) was calculated according to the results of the previous STEM-AMI trial. Specifically, the prespecified primary end point was the change from baseline to 6 months of ≥10 mL/m2 difference in adjusted LVEDV and as secondary end points a change of 5% in LVEF as well as infarct size reduction. Notably, the STEM-AMI trial was conceived more than a decade ago with an unique study design in the context of the contemporary published randomized trials testing G-CSF in STEMI.30 At that time, the information available suggested that G-CSF appeared to be more salutary in STEMI in the presence of LV dysfunction and when initiated early34 and that patients enrolled in previous negative trials with G-CSF were mainly at low-risk, with preserved LVEF (>50%), small volumes, and low percent anterior localization, and without definite symptom-to-balloon times.20 No other study except the STEM-AMI trial had enrolled patients affected by anterior STEMI with a LVEF ≤45% after reperfusion with symptom-to-balloon time ≥2 and ≤12 hours.

Accordingly, the intention-to-treat (ITT) populations of the STEM-AMI OUTCOME trial and CMR Substudy have included patients with the same aforementioned features. Both studies have been, however, subsequently amended with the aim to increase accrual rate, thus shifting the initial ITT to a modified-ITT population to include all presenting STEMI patients showing LVEF <45% regardless of symptom-to-balloon time. The resulting changes have produced evenly differences between the ITT and modified-ITT populations in both G-CSF and control patients in terms of anterior STEMI frequency (from 100% to 75% and 79%, respectively). Although we do not have a clear explanation why in this study G-CSF counteracted adverse remodeling by acting on LVESV rather than, as prespecified, on LVEDV, however, modified-ITT changes may at least partly be taken into account for this difference. Nevertheless, since CMR is considered the gold standard technique for LV volumes and function evaluation, the significant 6.7 mL intergroup difference between G-CSF and SOC in adjusted LVESVI, which is a well-known independent prognosticator after STEMI,35 is a thorough confirmation of the anti-remodeling effect exerted by G-CSF. Furthermore, the secondary study end points of LVEF improvement and infarct size reduction when adjusted for LV mass36,37 have been matched. In fact, our findings showed that as compared with SOC, only G-CSF administration was associated with a significant reduction of LGE amount between baseline and follow-up timeline. As a result, G-CSF-treated patients showed an improvement over SOC of LV performance as demonstrated by a significant increase of unadjusted and adjusted LVEF, which consistently matched scar and LVESVI changes when adjusted for main clinical and CMR baseline covariates. It is well known that infarct healing and attenuation of myocardial remodeling play a fundamental role in the functional recovery of the LV.38 Notably, these findings accounted for the presence of worse infarct-related profile in G-CSF patients at baseline which, if uncontrolled-for, may represent a potential bias for result’s interpretation. Interestingly, the salutary functional effect on LVEF appears to be more evident at tardive reperfusion times (>5 hours), paralleling previous similar findings we reported for LVEDV.30 Taken together, these findings suggest that G-CSF-mediated effects on damaged myocardium after extensive STEMI has the potential to translate into an improved clinical outcome. It is worth noting that in previous pivotal pharmacological studies dramatically impacting clinical practice after MI, apparently modest changes in LVEF (<5%) have been associated to improved prognosis.39–42

Remarkably, in the present study, we further demonstrated for the first time a benefit of G-CSF on 2D-GLS, which is a well-established prognostic parameter in post-STEMI patients.43 Feature tracking-CMR allows, with high spatial resolution, a retrospective tracking of LV myocardium based on cine images in both long- and short-axis views. Several studies demonstrated that abnormalities in future tracking-CMR-derived strain are linked to adverse remodeling.44 Notably, in the largest study to date of patients with MI undergoing CMR with deformation imaging, Eitel et al45 showed both 2D-GLS and 2D-GCS are associated with increased cardiac event rates after MI, whereas 2D-GLS had the strongest impact for the prediction of adverse prognosis at 1-year follow-up. Moreover, 2D-GLS is an independent predictor of clinical outcome after MI, even after adjustment for traditional cardiac risk factors including LVEF and infarct size.45,46 Our data demonstrated, once adjusted for MI-related variables including LGE extent and presence of MVO, a significant amelioration of 2D-GLS in G-CSF-treated patients with respect to patients receiving SOC, as well as a significant 2D-GCS improvement in the G-CSF group only.

Previous studies with G-CSF after STEMI have generated mixed and confounding results. In particular, the STEMMI (Stem Cells in Myocardial Infarction) and REVIVAL-2 (Regenerate Vital Myocardium by Vigorous Activation of Bone Marrow Stem Cells) trials, the only 2 studies evaluating post-STEMI effects of G-CSF using CMR as the imaging tool, have reported negative outcomes in terms of LV remodeling and function.19,22 We believe, however, that some relevant differences have to be taken into an account to explain discrepancies with our data. Specifically, in these studies, patients were mainly at low-risk, with preserved LVEF, small volumes, and low prevalence of anterior MI and without unknown symptom-to-balloon time. In addition, in both trials, baseline LVEF at CMR was assessed before PCI, as reported in the majority of published cell therapy studies.47 Conversely, in both the STEM-AMI trial and STEM-AMI OUTCOME CMR Substudy, we assessed baseline EF by CMR at day 7 post-PCI, when the myocardial stunning is partially resolved.48 This aspect may partly explain the lack of EF improvement in the control group. We believe this feature as relevant for discriminating the effect of a new therapeutic agent on top of the well-known early benefit on ventricular function provided by PCI.4 Furthermore, the STEMMI trial19 was not powered to measure LVEF, and LV volumetric changes because the sample size was calculated on a prespecified end point change in systolic wall thickening. Moreover, this study did not enroll patients with LV dysfunction. Similarly, in the REVIVAL-2 study,22 the mean LVEF as detected by CMR was 50%, and infarct size as measured by nuclear test was the primary prespecified end point. It is also worth underlying that the greatest benefits of cardioprotective mechanisms are reached when LV dysfunction is present.34,49,50 Accordingly, in a porcine STEMI model of ischemia-reperfusion, it has been shown that a LVEF cutoff <45% is important for G-CSF therapy to result in significantly lower detriment of LVEF compared with controls.51 Interestingly, within published G-CSF trials in STEMI, only the FIRSTLINE study (The Front-Integrated Revascularization and Stem Cell Liberation in Evolving Acute Myocardial Infarction by Granulocyte Colony-Stimulating Factor), which shares with our study early G-CSF administration (within 1 day after PCI) and baseline LVEF cutoff ≤45%, showed positive results on EF, although by means of 2D echocardiography assessment.18

Several mechanisms have been proposed for the salutary effect of G-CSF in the infarcted heart, including angiogenesis, direct protection of cardiomyocytes from apoptosis, and reduction of myocardial fibrosis.26 In this regard, it is worth mentioning that our results are in agreement with the REPAIR-AMI (Reinfusion of Enriched Progenitor Cells and Infarct Remodeling in Acute Myocardial Infarction) Magnetic Resonance Imaging substudy as obtained after intracoronary BM-cell delivery.52 This evidence, however, has not been confirmed afterwards.53,54 The 2 approaches substantially differ in timing of cell activity, assuming that the G-CSF-driven BM-cell mobilization effect into the peripheral blood, which last 5 days, can be viewed as a repeated dose strategy versus the single shot administration of intracoronary cell delivery.

Furthermore, beside cell-dependent mechanisms, G-CSF was consistently shown to exert an adjunctive direct protective action on ischemic cardiomyocytes by activating prosurvival pathways.25 Importantly, this effect is mediated by the expression of the G-CSF receptor by injured cardiomyocytes, which has been shown to peak into the infarcted area within 24 hours from coronary ligation and decline over 1 week in a mouse model of MI.55 Moreover, in a pig ischemia-reperfusion model, Beohar et al56 have investigated early and delayed G-CSF administration, showing that early treatment immediately after MI decreased ventricular dilatation, while delayed treatment (5 days) had a deleterious effect on LV remodeling. Accordingly, we believe that the early G-CSF administration scheme (within 24 hours from PCI) we followed in both the STEM-AMI and STEM-AMI OUTCOME studies has been a critical factor driving the success of G-CSF therapy.


As previously mentioned, the CMR Substudy, as part of the Phase III STEM-AMI OUTCOME trial, was not designed with a placebo arm. However, we believe the single-blind randomization design, taking advantage of 2 experts CMR readers showing high accordance, ensures the reliability of the results given the functional and not patient-related nature of end points measured. In addition, although a modified-ITT population has been included in the statistical analysis, we believe that the resulting noncritical deviations from the ITT population have not affected the original study design, which was conceived to recruit patients with reperfused extensive STEMI and moderate-to-severe LV dysfunction.


Overall, the results of the CMR Substudy confirmed the protective mode of action exerted by G-CSF as indicated by measurable and relevant functional readouts in the post-infarction myocardium. Thus, suggesting G-CSF may be considered a viable therapy in STEMI patients with a high likelihood of adverse LV remodeling. Although cell-dependent and independent mechanisms by which G-CSF interplays with myocardial ischemic injury have to be fully elucidated, our data support the concept that G-CSF can confer robust post-conditioning cardioprotection on top of gold standard therapy in terms of preservation of LV global function, scar size shrinkage, as well as the counteraction of LV adverse remodeling.

Nonstandard Abbreviations and Acronyms


2D global circumferential strain


2D global longitudinal strain


bone marrow


coronary artery bypass grafting


Creatine kinase-MB


cardiac magnetic resonance


granulocyte colony-stimulating factor


heart rate




late gadolinium enhancement


left ventricular


left ventricular end-diastolic volume


left ventricular end-diastolic volume indexed


left ventricular ejection fraction


left ventricular end-systolic volume


left ventricular end-systolic volume indexed


major adverse cardiac and cerebrovascular event


myocardial infarction


mass of the segments with late gadolinium enhancement


left ventricular mass


microvascular obstruction


primary coronary intervention


standard of care


ST-segment–elevation myocardial infarction


We thank Maurizio C. Capogrossi, MD for the relentless support, Dr Silvia Pica for CMR technical assistance, and Dr Aoife Gowran for critical comments and editing assistance.


*F.A. and G. Pontone are joint first authors.

†G.I.C. and G. Pompilio are joint senior authors.

The online-only Data Supplement is available with this article at

Correspondence to Giulio Pompilio, MD, PhD, Vascular Biology and Regenerative Medicine Unit, Centro Cardiologico Monzino IRCCS, Via Carlo Parea 4, 20138 Milan, Italy. Email


  • 1. St John Sutton M, Pfeffer MA, Moye L, Plappert T, Rouleau JL, Lamas G, Rouleau J, Parker JO, Arnold MO, Sussex B, Braunwald E. Cardiovascular death and left ventricular remodeling two years after myocardial infarction: baseline predictors and impact of long-term use of captopril: information from the Survival and Ventricular Enlargement (SAVE) trial.Circulation. 1997; 96:3294–3299.LinkGoogle Scholar
  • 2. Kelly DJ, Gershlick T, Witzenbichler B, Guagliumi G, Fahy M, Dangas G, Mehran R, Stone GW. Incidence and predictors of heart failure following percutaneous coronary intervention in ST-segment elevation myocardial infarction: the HORIZONS-AMI trial.Am Heart J. 2011; 162:663–670. doi: 10.1016/j.ahj.2011.08.002CrossrefMedlineGoogle Scholar
  • 3. Ng VG, Lansky AJ, Meller S, Witzenbichler B, Guagliumi G, Peruga JZ, Brodie B, Shah R, Mehran R, Stone GW. The prognostic importance of left ventricular function in patients with ST-segment elevation myocardial infarction: the HORIZONS-AMI trial.Eur Heart J Acute Cardiovasc Care. 2014; 3:67–77. doi: 10.1177/2048872613507149CrossrefMedlineGoogle Scholar
  • 4. Mamas MA, Anderson SG, O’Kane PD, Keavney B, Nolan J, Oldroyd KG, Perera D, Redwood S, Zaman A, Ludman PF, de Belder MA; British Cardiovascular Intervention Society and the National Institute for Cardiovascular Outcomes Research. Impact of left ventricular function in relation to procedural outcomes following percutaneous coronary intervention: insights from the British Cardiovascular Intervention Society.Eur Heart J. 2014; 35:3004–3012a. doi: 10.1093/eurheartj/ehu303CrossrefMedlineGoogle Scholar
  • 5. Gerczuk PZ, Kloner RA. An update on cardioprotection: a review of the latest adjunctive therapies to limit myocardial infarction size in clinical trials.J Am Coll Cardiol. 2012; 59:969–978. doi: 10.1016/j.jacc.2011.07.054CrossrefMedlineGoogle Scholar
  • 6. Banerjee MN, Bolli R, Hare JM. Clinical studies of cell therapy in cardiovascular medicine: recent developments and future directions.Circ Res. 2018; 123:266–287. doi: 10.1161/CIRCRESAHA.118.311217LinkGoogle Scholar
  • 7. Braunwald E. Cell-Based therapy in cardiac regeneration: an overview.Circ Res. 2018; 123:132–137. doi: 10.1161/CIRCRESAHA.118.313484LinkGoogle Scholar
  • 8. Laflamme MA, Zbinden S, Epstein SE, Murry CE. Cell-based therapy for myocardial ischemia and infarction: pathophysiological mechanisms.Annu Rev Pathol. 2007; 2:307–339. doi: 10.1146/annurev.pathol.2.010506.092038CrossrefMedlineGoogle Scholar
  • 9. Frangogiannis NG. Cell biological mechanisms in regulation of the post-infarction inflammatory response.Curr Opin Physiol. 2018; 1:7–13. doi: 10.1016/j.cophys.2017.09.001CrossrefMedlineGoogle Scholar
  • 10. Schächinger V, Erbs S, Elsässer A, et al; REPAIR-AMI Investigators. Improved clinical outcome after intracoronary administration of bone-marrow-derived progenitor cells in acute myocardial infarction: final 1-year results of the REPAIR-AMI trial.Eur Heart J. 2006; 27:2775–2783. doi: 10.1093/eurheartj/ehl388CrossrefMedlineGoogle Scholar
  • 11. Hirsch A, Nijveldt R, van der Vleuten PA, Tijssen JG, van der Giessen WJ, Tio RA, Waltenberger J, ten Berg JM, Doevendans PA, Aengevaeren WR, Zwaginga JJ, Biemond BJ, van Rossum AC, Piek JJ, Zijlstra F; HEBE Investigators. Intracoronary infusion of mononuclear cells from bone marrow or peripheral blood compared with standard therapy in patients after acute myocardial infarction treated by primary percutaneous coronary intervention: results of the randomized controlled HEBE trial.Eur Heart J. 2011; 32:1736–1747. doi: 10.1093/eurheartj/ehq449CrossrefMedlineGoogle Scholar
  • 12. Traverse JH, Henry TD, Pepine CJ, et al; Cardiovascular Cell Therapy Research Network (CCTRN). Effect of the use and timing of bone marrow mononuclear cell delivery on left ventricular function after acute myocardial infarction: the TIME randomized trial.JAMA. 2012; 308:2380–2389. doi: 10.1001/jama.2012.28726CrossrefMedlineGoogle Scholar
  • 13. Choudry F, Hamshere S, Saunders N, et al. A randomized double-blind control study of early intra-coronary autologous bone marrow cell infusion in acute myocardial infarction: the REGENERATE-AMI clinical trial†.Eur Heart J. 2016; 37:256–263. doi: 10.1093/eurheartj/ehv493CrossrefMedlineGoogle Scholar
  • 14. Ellis SG, Penn MS, Bolwell B, Garcia M, Chacko M, Wang T, Brezina KJ, McConnell G, Topol EJ. Granulocyte colony stimulating factor in patients with large acute myocardial infarction: results of a pilot dose-escalation randomized trial.Am Heart J. 2006; 152:1051.e9–1051.e14. doi: 10.1016/j.ahj.2006.09.003CrossrefGoogle Scholar
  • 15. Engelmann MG, Theiss HD, Hennig-Theiss C, Huber A, Wintersperger BJ, Werle-Ruedinger AE, Schoenberg SO, Steinbeck G, Franz WM. Autologous bone marrow stem cell mobilization induced by granulocyte colony-stimulating factor after subacute ST-segment elevation myocardial infarction undergoing late revascularization: final results from the G-CSF-STEMI (Granulocyte Colony-Stimulating Factor ST-Segment Elevation Myocardial Infarction) trial.J Am Coll Cardiol. 2006; 48:1712–1721. doi: 10.1016/j.jacc.2006.07.044MedlineGoogle Scholar
  • 16. Kuethe F, Figulla HR, Herzau M, Voth M, Fritzenwanger M, Opfermann T, Pachmann K, Krack A, Sayer HG, Gottschild D, Werner GS. Treatment with granulocyte colony-stimulating factor for mobilization of bone marrow cells in patients with acute myocardial infarction.Am Heart J. 2005; 150:115. doi: 10.1016/j.ahj.2005.04.030CrossrefMedlineGoogle Scholar
  • 17. Deng Z, Yang C, Deng H, Yang A, Geng T, Chen X, Ma A, Liu Z. Effects of GM-CSF on the stem cells mobilization and plasma C-reactive protein levels in patients with acute myocardial infarction.Int J Cardiol. 2006; 113:92–96. doi: 10.1016/j.ijcard.2006.06.014CrossrefMedlineGoogle Scholar
  • 18. Ince H, Petzsch M, Kleine HD, Eckard H, Rehders T, Burska D, Kische S, Freund M, Nienaber CA. Prevention of left ventricular remodeling with granulocyte colony-stimulating factor after acute myocardial infarction: Final 1-year results of the front-integrated revascularization and stem cell liberation in evolving acute myocardial infarction by gran.Circulation. 2005; 112:I73–I80.MedlineGoogle Scholar
  • 19. Ripa RS, Jørgensen E, Wang Y, Thune JJ, Nilsson JC, Søndergaard L, Johnsen HE, Køber L, Grande P, Kastrup J. Stem cell mobilization induced by subcutaneous granulocyte-colony stimulating factor to improve cardiac regeneration after acute ST-elevation myocardial infarction: result of the double-blind, randomized, placebo-controlled stem cells in myocardial infarction (STEMMI) trial.Circulation. 2006; 113:1983–1992. doi: 10.1161/CIRCULATIONAHA.105.610469LinkGoogle Scholar
  • 20. Leone AM, Galiuto L, Garramone B, Rutella S, Giannico MB, Brugaletta S, Perfetti M, Liuzzo G, Porto I, Burzotta F, Niccoli G, Biasucci LM, Leone G, Rebuzzi AG, Crea F. Usefulness of granulocyte colony-stimulating factor in patients with a large anterior wall acute myocardial infarction to prevent left ventricular remodeling (the rigenera study).Am J Cardiol. 2007; 100:397–403. doi: 10.1016/j.amjcard.2007.03.036CrossrefMedlineGoogle Scholar
  • 21. Kang HJ, Lee HY, Na SH, et al. Differential effect of intracoronary infusion of mobilized peripheral blood stem cells by granulocyte colony-stimulating factor on left ventricular function and remodeling in patients with acute myocardial infarction versus old myocardial infarction: the MAGIC Cell-3-DES randomized, controlled trial.Circulation. 2006; 114:I145–I151. doi: 10.1161/CIRCULATIONAHA.105.001107MedlineGoogle Scholar
  • 22. Zohlnhöfer D, Ott I, Mehilli J, Schömig K, Michalk F, Ibrahim T, Meisetschläger G, von Wedel J, Bollwein H, Seyfarth M, Dirschinger J, Schmitt C, Schwaiger M, Kastrati A, Schömig A; REVIVAL-2 Investigators. Stem cell mobilization by granulocyte colony-stimulating factor in patients with acute myocardial infarction: a randomized controlled trial.JAMA. 2006; 295:1003–1010. doi: 10.1001/jama.295.9.1003CrossrefMedlineGoogle Scholar
  • 23. Takano H, Hasegawa H, Kuwabara Y, et al. Feasibility and safety of granulocyte colony-stimulating factor treatment in patients with acute myocardial infarction.Int J Cardiol. 2007; 122:41–47. doi: 10.1016/j.ijcard.2006.11.016CrossrefMedlineGoogle Scholar
  • 24. Anderlini P, Donato M, Chan KW, Huh YO, Gee AP, Lauppe MJ, Champlin RE, Körbling M. Allogeneic blood progenitor cell collection in normal donors after mobilization with filgrastim: the M.D. anderson cancer center experience.Transfusion. 1999; 39:555–560.CrossrefMedlineGoogle Scholar
  • 25. Harada M, Qin Y, Takano H, et al. G-CSF prevents cardiac remodeling after myocardial infarction by activating the Jak-Stat pathway in cardiomyocytes.Nat Med. 2005; 11:305–311. doi: 10.1038/nm1199CrossrefMedlineGoogle Scholar
  • 26. Shim W, Mehta A, Lim SY, Zhang G, Lim CH, Chua T, Wong P. G-CSF for stem cell therapy in acute myocardial infarction: friend or foe?Cardiovasc Res. 2011; 89:20–30. doi: 10.1093/cvr/cvq301CrossrefMedlineGoogle Scholar
  • 27. Deindl E, Zaruba MM, Brunner S, Huber B, Mehl U, Assmann G, Hoefer IE, Mueller-Hoecker J, Franz WM. G-CSF administration after myocardial infarction in mice attenuates late ischemic cardiomyopathy by enhanced arteriogenesis.FASEB J. 2006; 20:956–958. doi: 10.1096/fj.05-4763fjeCrossrefMedlineGoogle Scholar
  • 28. Kasra M, Aria R, Bobak M. Granulocyte colony stimulating factor therapy for acute myocardial infarction.Cochrane Database Syst Rev. 2013; 5:CD008844.Google Scholar
  • 29. Zimmet H, Porapakkham P, Porapakkham P, Sata Y, Haas SJ, Itescu S, Forbes A, Krum H. Short- and long-term outcomes of intracoronary and endogenously mobilized bone marrow stem cells in the treatment of ST-segment elevation myocardial infarction: a meta-analysis of randomized control trials.Eur J Heart Fail. 2012; 14:91–105. doi: 10.1093/eurjhf/hfr148CrossrefMedlineGoogle Scholar
  • 30. Achilli F, Malafronte C, Lenatti L, et al; STEM-AMI Investigators. Granulocyte colony-stimulating factor attenuates left ventricular remodelling after acute anterior STEMI: results of the single-blind, randomized, placebo-controlled multicentre STem cEll Mobilization in Acute Myocardial Infarction (STEM-AMI) Trial.Eur J Heart Fail. 2010; 12:1111–1121. doi: 10.1093/eurjhf/hfq150CrossrefMedlineGoogle Scholar
  • 31. Achilli F, Malafronte C, Maggiolini S, et al; STEM-AMI trial Investigators. G-CSF treatment for STEMI: final 3-year follow-up of the randomised placebo-controlled STEM-AMI trial.Heart. 2014; 100:574–581. doi: 10.1136/heartjnl-2013-304955CrossrefMedlineGoogle Scholar
  • 32. Achilli F, Malafronte C, Cesana F, et al; STEM-AMI OUTCOME Trial Investigators. Granulocyte-colony stimulating factor for large anterior ST-elevation myocardial infarction: rationale and design of the prospective randomized phase III STEM-AMI OUTCOME trial.Am Heart J. 2015; 170:652.e7–658.e7. doi: 10.1016/j.ahj.2015.07.005CrossrefGoogle Scholar
  • 33. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies.J Clin Epidemiol. 2008; 61:344–349. doi: 10.1016/j.jclinepi.2007.11.008CrossrefMedlineGoogle Scholar
  • 34. Abdel-Latif A, Bolli R, Zuba-Surma EK, Tleyjeh IM, Hornung CA, Dawn B. Granulocyte colony-stimulating factor therapy for cardiac repair after acute myocardial infarction: a systematic review and meta-analysis of randomized controlled trials.Am Heart J. 2008; 156:216.e9–226.e9. doi: 10.1016/j.ahj.2008.03.024CrossrefGoogle Scholar
  • 35. White HD, Norris RM, Brown MA, Brandt PW, Whitlock RM, Wild CJ. Left ventricular end-systolic volume as the major determinant of survival after recovery from myocardial infarction.Circulation. 1987; 76:44–51.LinkGoogle Scholar
  • 36. Pontone G, Guaricci AI, Andreini D, et al. Prognostic stratification of patients with ST-segment-elevation myocardial infarction (PROSPECT): a cardiac magnetic resonance study.Circ Cardiovasc Imaging. 2017; 10:1–11.LinkGoogle Scholar
  • 37. Lemkes JS, Janssens GN, van der Hoeven NW, et al. Timing of revascularization in patients with transient ST-segment elevation myocardial infarction: a randomized clinical trial.Eur Heart J. 2019; 40:283–291. doi: 10.1093/eurheartj/ehy651CrossrefMedlineGoogle Scholar
  • 38. Hammer-Hansen S, Bandettini WP, Hsu LY, Leung SW, Shanbhag S, Mancini C, Greve AM, Køber L, Thune JJ, Kellman P, Arai AE. Mechanisms for overestimating acute myocardial infarct size with gadolinium-enhanced cardiovascular magnetic resonance imaging in humans: a quantitative and kinetic study.Eur Heart J Cardiovasc Imaging. 2016; 17:76–84. doi: 10.1093/ehjci/jev123MedlineGoogle Scholar
  • 39. Dargie HJ. Effect of carvedilol on outcome after myocardial infarction in patients with left-ventricular dysfunction: the CAPRICORN randomised trial.Lancet. 2001; 357:1385–1390.CrossrefMedlineGoogle Scholar
  • 40. Doughty RN, Whalley GA, Walsh HA, Gamble GD, López-Sendón J, Sharpe N; CAPRICORN Echo Substudy Investigators. Effects of carvedilol on left ventricular remodeling after acute myocardial infarction: the CAPRICORN Echo Substudy.Circulation. 2004; 109:201–206. doi: 10.1161/01.CIR.0000108928.25690.94LinkGoogle Scholar
  • 41. Pfeffer MA, Greaves SC, Arnold JM, Glynn RJ, LaMotte FS, Lee RT, Menapace FJ, Rapaport E, Ridker PM, Rouleau JL, Solomon SD, Hennekens CH. Early versus delayed angiotensin-converting enzyme inhibition therapy in acute myocardial infarction. The healing and early afterload reducing therapy trial.Circulation. 1997; 95:2643–2651.LinkGoogle Scholar
  • 42. Rutherford JD, Pfeffer MA, Moyé LA, Davis BR, Flaker GC, Kowey PR, Lamas GA, Miller HS, Packer M, Rouleau JL. Effects of captopril on ischemic events after myocardial infarction. Results of the survival and ventricular enlargement trial. SAVE Investigators.Circulation. 1994; 90:1731–1738.LinkGoogle Scholar
  • 43. Vo HQ, Marwick TH, Negishi K. MRI-Derived myocardial strain measures in normal subjects.JACC Cardiovasc Imaging. 2018; 11:196–205. doi: 10.1016/j.jcmg.2016.12.025.CrossrefMedlineGoogle Scholar
  • 44. Mangion K, McComb C, Auger DA, Epstein FH, Berry C. Magnetic resonance imaging of myocardial strain after acute ST-segment-elevation myocardial infarction a systematic review.Circ Cardiovasc Imaging. 2017; 10:1–10.LinkGoogle Scholar
  • 45. Eitel I, Stiermaier T, Lange T, Rommel KP, Koschalka A, Kowallick JT, Lotz J, Kutty S, Gutberlet M, Hasenfuß G, Thiele H, Schuster A. Cardiac magnetic resonance myocardial feature tracking for optimized prediction of cardiovascular events following myocardial infarction.JACC Cardiovasc Imaging. 2018; 11:1433–1444. doi: 10.1016/j.jcmg.2017.11.034CrossrefMedlineGoogle Scholar
  • 46. Palazzuoli A, Beltrami M, Gennari L, Dastidar AG, Nuti R, McAlindon E, Angelini GD, Bucciarelli-Ducci C. The impact of infarct size on regional and global left ventricular systolic function: a cardiac magnetic resonance imaging study.Int J Cardiovasc Imaging. 2015; 31:1037–1044. doi: 10.1007/s10554-015-0657-3CrossrefMedlineGoogle Scholar
  • 47. Reffelmann T, Könemann S, Kloner RA. Promise of blood- and bone marrow-derived stem cell transplantation for functional cardiac repair: putting it in perspective with existing therapy.J Am Coll Cardiol. 2009; 53:305–308. doi: 10.1016/j.jacc.2008.10.018CrossrefMedlineGoogle Scholar
  • 48. Engblom H, Hedström E, Heiberg E, Wagner GS, Pahlm O, Arheden H. Rapid initial reduction of hyperenhanced myocardium after reperfused first myocardial infarction suggests recovery of the peri-infarction zone: one-year follow-up by MRI.Circ Cardiovasc Imaging. 2009; 2:47–55. doi: 10.1161/CIRCIMAGING.108.802199LinkGoogle Scholar
  • 49. Jeevanantham V, Butler M, Saad A, Abdel-Latif A, Zuba-Surma EK, Dawn B. Adult bone marrow cell therapy improves survival and induces long-term improvement in cardiac parameters: a systematic review and meta-analysis.Circulation. 2012; 126:551–568. doi: 10.1161/CIRCULATIONAHA.111.086074LinkGoogle Scholar
  • 50. Pompilio G, Nigro P, Bassetti B, Capogrossi MC. Bone marrow cell therapy for ischemic heart disease: the never ending story.Circ Res. 2015; 117:490–493. doi: 10.1161/CIRCRESAHA.115.307184LinkGoogle Scholar
  • 51. Angeli FS, Amabile N, Shapiro M, Mirsky R, Bartlett L, Zhang Y, Virmani R, Chatterjee K, Boyle A, Grossman W, Yeghiazarians Y. Cytokine combination therapy with erythropoietin and granulocyte colony stimulating factor in a porcine model of acute myocardial infarction.Cardiovasc Drugs Ther. 2010; 24:409–420. doi: 10.1007/s10557-010-6263-7CrossrefMedlineGoogle Scholar
  • 52. Dill T, Schächinger V, Rolf A, Möllmann S, Thiele H, Tillmanns H, Assmus B, Dimmeler S, Zeiher AM, Hamm C. Intracoronary administration of bone marrow-derived progenitor cells improves left ventricular function in patients at risk for adverse remodeling after acute ST-segment elevation myocardial infarction: results of the reinfusion of enriched progenitor cells and infarct remodeling in acute myocardial infarction study (REPAIR-AMI) cardiac magnetic resonance imaging substudy.Am Heart J. 2009; 157:541–547. doi: 10.1016/j.ahj.2008.11.011CrossrefMedlineGoogle Scholar
  • 53. Gyöngyösi M, Wojakowski W, Lemarchand P, Lunde K, Bartunek J, Marban E, Assmus B, Henry TD, Jay H. Meta-analysis of Cell-based CaRdiac stUdiEs (ACCRUE) in patients with acute myocardial infarction based on individual patient data.Circ Res. 2016; 118:1254–1263.AbstractGoogle Scholar
  • 54. Wollert KC, Meyer GP, Lotz J, Ringes-Lichtenberg S, Lippolt P, Breidenbach C, Fichtner S, Korte T, Hornig B, Messinger D, Arseniev L, Hertenstein B, Ganser A, Drexler H. Intracoronary autologous bone-marrow cell transfer after myocardial infarction: the BOOST randomised controlled clinical trial.Lancet. 2004; 364:141–148. doi: 10.1016/S0140-6736(04)16626-9CrossrefMedlineGoogle Scholar
  • 55. Kuhlmann MT, Kirchhof P, Klocke R, Hasib L, Stypmann J, Fabritz L, Stelljes M, Tian W, Zwiener M, Mueller M, Kienast J, Breithardt G, Nikol S. G-CSF/SCF reduces inducible arrhythmias in the infarcted heart potentially via increased connexin43 expression and arteriogenesis.J Exp Med. 2006; 203:87–97. doi: 10.1084/jem.20051151CrossrefMedlineGoogle Scholar
  • 56. Beohar N, Flaherty JD, Davidson CJ, Vidovich M, Singhal S, Rapp JA, Erdogan A, Lee DC, Rammohan C, Brodsky A, Wu E, Pieper K, Virmani R, Bonow RO, Mehta J. Granulocyte-colony stimulating factor administration after myocardial infarction in a porcine ischemia-reperfusion model: functional and pathological effects of dose timing.Catheter Cardiovasc Interv. 2007; 69:257–266. doi: 10.1002/ccd.20925CrossrefMedlineGoogle Scholar

Novelty and Significance

What Is Known?

  • Cytokine release in the acute postinfarction period is crucial for the response to injury and modulates myocardial tissue repair. Cytokine response is particularly evident in the injured myocardium after ST-segment-elevation myocardial infarction (STEMI), contributing to cardiac remodeling.

  • At preclinical level, the bone marrow leukocytes-mobilizing cytokine G-CSF (granulocyte colony-stimulating factor) has been demonstrated to ameliorate cardiac function and scar size after myocardial infarction through both cell-mediated and cell-independent mechanisms.

  • Results from randomized clinical trials were, however, inconclusive.

What New Information Does This Article Contribute?

  • This article reports the largest prospective randomized study with cardiac magnetic resonance testing the cardioprotective effects of G-CSF in STEMI patients with left ventricular dysfunction (LV) following primary reperfusion at high-risk of LV remodeling.

  • Early subcutaneous administration of 5 µg/kg G-CSF twice a day for 6 consecutive days exerted beneficial effects in terms of global systolic function, adverse remodeling, scar size, and myocardial strain.

The therapeutic modulation of cytokines in STEMI has the potential to improve infarct healing and LV remodeling. Previous randomized clinical trials in STEMI using the bone marrow-mobilizer G-CSF have enrolled patients mainly at low-risk, with preserved LV function, small volumes, low prevalence of anterior myocardial infarction, and unknown symptom-to-balloon time. No other study except the STEM-AMI OUTCOME CMR (Stem Cells Mobilization in Acute Myocardial Infarction Outcome Cardiac Magnetic Resonance) Substudy trial have included patients affected by extensive STEMI with symptom-to-balloon time ≥2 and ≤12 hours, LV ejection fraction ≤45% after successful revascularization, and receiving G-CSF within 12 hours after reperfusion. The main results of this study are that early administration of G-CSF in STEMI patients at high risk of LV adverse remodeling is safe and can confer robust postconditioning cardioprotection on top of gold standard therapy in terms of LV global function amelioration, LV dilation counteraction, scar size shrinkage, and myocardial strain improvement. This drug-repositioning study suggests that G-CSF may be considered a viable therapy in STEMI patients with high likelihood of LV remodeling. These results may also have important implications for translational research aimed to better elucidate mechanisms by which G-CSF exerts a salutary effect onto the ischemic myocardium.


eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.