Usefulness of Granulocyte Colony-Stimulating Factor in Patients With a Large Anterior Wall Acute Myocardial Infarction to Prevent Left Ventricular Remodeling (The Rigenera Study)

Usefulness of Granulocyte Colony-Stimulating Factor in Patients With a Large Anterior Wall Acute Myocardial Infarction to Prevent Left Ventricular Remodeling (The Rigenera Study)
of 7
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Usefulness of Granulocyte Colony-Stimulating Factor in PatientsWith a Large Anterior Wall Acute Myocardial Infarction toPrevent Left Ventricular Remodeling (The Rigenera Study) Antonio Maria Leone, MD a, *, Leonarda Galiuto, MD a , Barbara Garramone, MD a ,Sergio Rutella, MD, PhD b , Maria Benedetta Giannico, MD a , Salvatore Brugaletta, MD a ,Matteo Perfetti, MD a , Giovanna Liuzzo, MD a , Italo Porto, MD, PhD a ,Francesco Burzotta, MD, PhD a , Giampaolo Niccoli, MD, PhD a , Luigi Marzio Biasucci, MD a ,Giuseppe Leone, MD b , Antonio Giuseppe Rebuzzi, MD a , and Filippo Crea, MD a Intracoronary injection of bone marrow stem cells seems to improve left ventricular (LV)function after acute myocardial infarction (AMI). Granulocyte colony-stimulating factor(G-CSF) could improve myocardial function and perfusion noninvasively through mobili-zation of stem cells into peripheral blood, although previous clinical trials have producedcontroversial results. Forty-one patients with large anterior wall AMI at high risk of unfavorable remodeling were randomized 1:2 to G-CSF (10   g/kg/day for 5 days) or toconventional therapy. All patients underwent successful primary or rescue percutaneouscoronary intervention. LV function was assessed by echocardiography before G-CSF ad-ministration,  > 5 days after AMI, and at follow-up. Only patients with a LV ejectionfraction < 50% at baseline were enrolled in the study. After a median follow-up of 5 months(range 4 to 6) patients treated with G-CSF exhibited improvement in LV ejection fraction,from 40    6% to 45    6% (p    0.068) in the absence of LV dilation (LV end-diastolicvolume from 147  33 to 144  46 ml at follow-up, p  0.77). In contrast, patients treatedconventionally exhibited significant LV dilation (LV end-diastolic volume from 141  35 to168  41 ml, p  0.002) in the absence of change in LV ejection fraction (from 38  6%to 38    8%, p    0.95). However, when comparing patients treated with G-CSF withcontrols, variations in these parameters were significantly different at 2-way analysis of variance (p    0.04 for LV end-diastolic volume, p    0.02 for LV ejection fraction). Inconclusion, G-CSF prevents unfavorable LV remodeling and improves LV function inpatients with large anterior wall AMI and decreased LV ejection fraction after successfulpercutaneous coronary intervention. © 2007 Elsevier Inc. All rights reserved. (Am JCardiol 2007;100:397–403) Intracoronary administration of bone marrow stem cells hasbeen shown to improve left ventricular (LV) function afteracute myocardial infarction (AMI) 1 ; however, despite aninvasive approach, the amount of stem cells delivered to thedamaged tissue is relatively small. Further, several studieshave shown that function of stem cells is impaired in pa-tients with risk factors. 2 Several basic 3 and clinical 4,5 studieshave supported the notion that mobilization of stem cells bygrowth factors might improve LV function by increasingtheir homing in the damaged myocardium. Despite a safeclinical profile of granulocyte colony-stimulating factor(G-CSF) even in patients with AMI, 6,7 recent trials of G-CSF in these patients have produced conflicting re-sults. 8–13 A major limitation of previous studies has beeninclusion of patients with substantially preserved LV func-tion. Interestingly, Schachinger et al 14 found that the largestbeneficial effect of intracoronary injection of bone marrowmononuclear cells was present only in a subgroup of pa-tients with a baseline LV ejection fraction   49%. There-fore, we focused on G-CSF therapy inn patients with largeanterior AMI and impaired LV ejection fraction who are athigh risk of unfavorable LV remodeling. MethodsPatients and protocol:  The present study assessed thepotential efficacy of G-CSF administration on cardiac func-tion in patients with a first large anterior AMI and a LVejection fraction  50% despite successful percutaneous re-vascularization of the infarct-related artery. Exclusion cri-teria were cardiogenic shock, uncontrolled myocardial isch-emia or arrhythmias, malignancies, severe infections, Institutes of   a Cardiology and of   b Haematology, Catholic University of the Sacred Heart, Rome, Italy. Manuscript received January 29, 2007;revised manuscript received and accepted March 6, 2007.This study was supported by a grant from the Fondazione Cassa diRisparmio di Roma, Rome, Italy, to UNICATT Cord Blood Bank of theCatholic University of the Sacred Heart of Rome and by FondazioneInternazionale Ricerche Per il Cuore, ONLUS  ,  Rome, Italy. Lenograstim(Myelostim 34) was supplied by Italfarmaco S.p.A., Milan, Italy,which hadno role in the collection, analysis, and interpretation of the data.*Corresponding author: Tel: 39-06-3015-4187; fax: 39-06-305-5535.  E-mail address: (A.M. Leone).0002-9149/07/$ – see front matter © 2007 Elsevier Inc. All rights reserved. www.AJConline.orgdoi:10.1016/j.amjcard.2007.03.036  hematologic diseases, splenomegaly on abdominal echocar-diography, and age  80 years.From June 2003 to May 2006, 72 patients with largeanterior ST-segment elevation AMI and impaired LV func-tion were hospitalized in our department. At the end of May2006, after several published clinical trials suggested thatG-CSF might not improve postinfarction LV function, 13 enrollment was stopped to perform the present interim anal-ysis. At that point, 41 of the 60 patients srcinally includedgave their written informed consent to be randomized 1:2 toG-CSF therapy (n  14) or conventional therapy (n  27).Patients randomized to G-CSF therapy were treated withlenograstim (recombinant human G-CSF; Myelostim 34,Italfarmaco S.p.A., Milan, Italy) at a dose of 10   g/kg/dayfor 5 days starting   5 days after AMI and/or a completecoronary stenting. Clinical characteristics are listed inTable 1. Originally, all patients should have had undergoneevaluation by both conventional and myocardial contrastechocardiography (MCE) before starting therapy; unfortu-nately, in May 2004 the European Agency for the Evalua-tion of Medicinal Products declared that the ultrasoundcontrast agent used in our trial was no longer suitable forechocardiography after some adverse events were reportedin patients with ischemic heart disease. Collectively, at thattime 17 patients (6 in G-CSF group and 11 in control group)underwent MCE during hospitalization and at follow-up andtheir data are currently available; in 2 other patients receiv-ing G-CSF therapy, MCE was performed during hospital-ization but not at follow-up.Standard therapy for patients using G-CSF therapy andcontrols at discharge consisted of aspirin (100 mg), clopi-dogrel, carvedilol, ramipril, and atorvastatin. To monitorstem/progenitor cell mobilization, CD34  cells were as-sessed by flow cytometry using anti-CD34 antibody(Caltag, Burlingame, California). The study was approvedby the ethics committee of the Catholic University of Rome.All patients gave their informed consent according to theDeclaration of Helsinki. Echocardiography:  In all patients   5 days after ST-segment elevation AMI and/or completion of percutane-ous revascularization (i.e., without a residual stenosis  50% in infarct-related and noninfarct–related arteries)and then within 6 months, conventional echocardiogra-phy and MCE were performed. LV end-diastolic andend-systolic volumes, LV ejection fraction, and wall mo-tion score index were independently evaluated with echo-cardiography by 2 experienced blinded observers (LGand BG) according to recommendations of the AmericanSociety of Echocardiography. Accordingly, regional wallmotion was semiquantitatively scored as appropriate(1    normal, 2    hypokinesia, 3    akinesia, 4    dys-kinesia) and a wall motion score index was calculated asthe sum of the score of all segments divided by the totalnumber of segments. LV volumes were calculated by themodified Simpson biplane method. Temporal changes inLV volumes were calculated as percent changes at 6months compared with baseline. LV remodeling was con-sidered an increase  20% in LV end-diastolic volume at6 months compared with baseline. Ejection fraction wascalculated from the formula (end-diastolic volume   end-systolic volume)/end-diastolic volume. Myocardial contrast echocardiography:  Details havebeen described elsewhere. 15 In brief, real-time contrastpulse sequencing was performed with a Sequoia ultra-sound system (Siemens, Malvern, Pennsylvania). A sec-ond-generation ultrasound contrast agent (SonoVue,Bracco, Milan, Italy) was administered intravenously (5ml at 1 ml/min). Myocardial opacification with MCE wasvisually assessed in each of the 17 myocardial segments andsemiquantitatively scored. A single perfusion score wasassigned on the basis of both the change in myocardialsignal intensity throughout the replenishment curve anddegree of opacification at peak contrast effect. Scores weregraded as 1 (normal opacification), 2 (decreased opacifica-tion), or 3 (no opacification). A contrast score index wascalculated as the sum of myocardial contrast echocardio-graphic scores in each segment divided by the total numberof segments. Endocardial lengths of severe wall motionabnormality (wall motion score 3) and transmural contrastdefect (contrast defect score 3) were calculated in eachapical view, averaged, and expressed as percent whole LVlength or wall motion abnormality length. Statistical analysis:  Normal distribution was tested bythe Kolmogorov-Smirnov test. Continuous variableswere compared by  t   test and 2-way analysis of variance(ANOVA) or by their corresponding nonparametric al-ternatives (Mann-Whitney U test or Wilcoxon test), asappropriate (Statistica 5.5, Stat Soft, Inc., Tulsa, Okla-homa). The Rigenera clinical trial was designed to ran-domize 60 patients 1:2 to G-CSF therapy or standard ther-apy to obtain, with 95% power and an SD of 5% in the 2groups, a possible significant difference of 5% in LV ejec-tion fraction. However, in the present interim analysis, weacknowledge limited power to discriminate a potential sig-nificant difference between patients receiving G-CSF ther- Table 1Clinical characteristicsVariable G-CSF Group(n  14)Control Group(n  27)pValueAge (yrs) 53  11 56  11 0.41Men 13 (93%) 27 (100%) 0.16Primary percutaneouscoronary intervention3 (29%) 12 (44%) 0.14Thrombolysis 10 (71%) 15 (56%) 0.32Previous MI 0 (0%) 0 (0%) 1Multivessel coronarydisease4 (29%) 7 (26%) 0.86Diabetes mellitus 2 (14%) 5 (19%) 0.73Hypertension 8 (57%) 16 (59%) 0.89Smokers 10 (71%) 19 (70%) 0.94Hypercholesterolemia(LDL  130 mg/dl)9 (64%) 16 (60%) 0.75Creatine phosphokinase(UI/L)4,592  3,717 5,396  2,840 0.67Troponin T (ng/ml) 11.7  7.9 15.8  9.67 0.18Baseline CD34  cells/   l 4.81  3.31 7.84  6.22 0.10  5-day CD34  cells/   l 50.33  35.20 2.13  2.17   0.0001LDL  low-density lipoprotein.398  The American Journal of Cardiology (  apy and controls. Continuous data are presented as mean  SD. Differences with a p value   0.05 were consideredstatistically significant. ResultsClinical and laboratory findings:  All patients under-went successful percutaneous coronary intervention of theinfarct-related artery and 11 (26.8%) of a noninfarct–relatedartery (4 in the G-CSF group, 28.5%, and 7 in the controlgroup, 25.9%). However, complete revascularization wasachieved in all patients (Table 1). All patients at dischargeand then at follow-up were on standard therapy consistingof aspirin, clopidogrel, carvedilol, ramipril, and atorvasta-tin. No patient exhibited major adverse cardiac events orsevere G-CSF–related side effects during hospitalizationand at follow-up. At follow-up, 3 patients in the G-CSFgroup and 7 in the control group underwent repeat percu-taneous coronary intervention for asymptomatic binary in-stent restenosis at coronary angiography. Three patientsdisclosed mild bone pain during G-CSF administration. Onepatient randomized to G-CSF therapy refused follow-upechocardiography after completion of G-CSF administra-tion and reported no event at telephone follow-up.CD34  cell count significantly increased after G-CSFadministration. CD34  cell levels in controls showed a mildincrease 5 days after AMI according to our previous data 4 (Table 1). LV volumes and function at echocardiography:  Atbaseline, all parameters of LV function were similar in theG-CSF and control groups. Interestingly, however, LV end-diastolic and end-systolic volumes were significantly in-creased at follow-up in patients treated conventionally butnot in patients treated with G-CSF (Table 2 and Figures1 to 2). Therefore, LV ejection fraction was virtually un-changed in patients treated conventionally compared withthose treated with G-CSF in whom a mild improvement was present (Table 2 and Figure 3). However, when comparing patients treated with G-CSF with controls, variations of these parameters were significantly different (Table 2), even at 2-way ANOVA (LV end-diastolic volume, p  0.04; LVend-systolic volume, p    0.02; LV ejection fraction, p   0.02; Figures 1 to 4). Wall motion score index improved more in the G-CSFgroup compared with the control group (Table 2 and Figure 4). Collectively, at follow-up, adverse remodelingwas present in 14 of 27 patients in the control group versus4 of 13 in the G-CSF group (p  0.31). Myocardial contrast echocardiography:  Some base-line characteristics were unbalanced in the 2 groups, inparticular the baseline contrast score index was significantlyworse in the G-CSF group compared with the control group(Table 3). However, the contrast score index improved nonsignificantly at follow-up in the 2 groups. Contrast de-fect length decreased in the G-CSF and control groups. LVlength at baseline was similar in the 2 groups but tended toincrease in controls and decrease in patients receivingG-CSF therapy (p    0.07 by ANOVA). Contrast defectlength/LV length significantly decreased in the controlgroup only (p    0.03), although this decrease was moreapparent in the G-CSF group (p    0.24). Baseline wallmotion abnormality length was slightly shorter in theG-CSF group, in which a further mild decrease was ob-served at follow-up (p  0.17); this was not the case in thecontrol group in which it remained substantially unchangedat follow-up (p    0.59). Accordingly, a decrease in wallmotion abnormality length/LV length was slightly moreapparent in the G-CSF group (p  0.17) than in the controlgroup (p  0.18). At follow-up variation in the percentageof the left ventricle with severe wall motion and consensualperfusion defect was similar in the 2 groups. Discussion In our study, G-CSF was associated with a significantimprovement in LV function similar to that observedafter intracoronary injection of stem cells in patients withAMI and, more interestingly, a lack of LV dilation com-pared with patients treated conventionally, in whom, incontrast, we found no improvement in LV ejection frac-tion and a significant dilation of LV volumes in additionto currently recommended medical therapy. With regardto the a lack of an improvement in LV function inpatients treated conventionally, these data are not com-pletely surprising if we consider that patients enrolled inour study were selected to be highly prone to unfavorableremodeling with large anterior AMI 16 evaluated by echo-cardiography  5 days after the acute event, when most of the spontaneous recovery of LV function had alreadytaken place. 17 Accordingly, our data are similar to thoseof Ince et al 8 and Takano et al 11 who demonstratedsignificant improvement in LV ejection fraction and a Table 2Echocardiographic findingsG-CSF Group(n  13)Control Group(n  27)pValueBaselineLV end-diastolicvolume (ml)147  33 141  35 0.62LV end-systolicvolume (ml)89  24 89  24 0.99LV ejectionfraction (%)40  6 38  6 0.27Wall motion score index 2.00  0.16 2.12  0.32 0.22Follow-upLV end-diastolicvolume (ml)144  46 168  41 0.10LV end-systolicvolume (ml)82  34 106  34 0.04LV ejection fraction (%) 45  6 38  8 0.005Wall motion score index 1.74  0.22 2.09  0.38 0.005VariationLV end-diastolicvolume (ml)  3  36 27  41 0.03Percent variation   2  24 24  34 0.02LV end systolicvolume (ml)  7  29 17  30 0.02Percent variation   7  31 23  36 0.01LV ejection fraction (%) 5  9 0  6 0.04Percent variation 16  29 0  16 0.03Wall motion score index   0.26  0.24   0.04  0.49 0.07Percent variation   13  12   1  18 0.04399 Coronary Artery Disease/G-CSF in Anterior Myocardial Infarction  preservation toward LV remodeling in patients treatedsoon after percutaneous coronary intervention withG-CSF. Even in these reports, LV function in controlsfailed to recover at follow-up, with a trend toward LVdilation. In contrast, in more recent and larger trials, suchas the STEM cells in Myocardial Infarction (STEMMI) 9 and REgenerate VItal myocardium by Vigorous Activityof bone marrow stem cells (REVIVAL-2) 10 trial, G-CSFfailed to guarantee an improvement in spontaneous re-covery observed in control patients. In particular in theREVIVAL-2, G-CSF therapy started 5 days after AMIfailed to decrease infarcts and ameliorate LV function asmeasured by scintigraphy and magnetic resonance imag-ing compared with standard therapy; however, it has to bestressed that patients enrolled in these trials comparedwith patients enrolled in our study were at low risk withpreserved LV ejection fraction (  50% in the 2 trials),small volumes, and low percent anterior localization(about 50% of all patients). Considering these points,demonstrating further improvement by G-CSF therapy isunlikely. Moreover, a key point is represented by time of baseline assessment of LV function. The spontaneousrecovery reported to be the explanation of improved LVejection fraction observed in the first nonrandomizedstudies 18 is particularly apparent when evaluation of LVejection fraction at 6-month follow-up is compared withacute or subacute assessment, as in the STEMMI trial, 9 but was not evident in controls enrolled in the REVIVAL-2 50100150200250300 50100150200250300    L   V   E   D   V   (  m   l   ) Baseline BaselineFollow up Follow up147+32 144+ +14154 35  168+41G-CSF Control groupp=0.002p=0.772-way ANOVAp=0.03      L   V  e  n   d   d   i  a  s   t  o   l   i  c  v  o   l  u  m  e   (  m   l   ) Figure 1. LV end-diastolic volumes at baseline and follow-up in patients treated with G-CSF and in controls. In the control group, there was a significantincrease in LV end-diastolic volume from baseline to follow-up (p  0.002). This dilation did not take place in the G-CSF group (p  0.77). This differenttrend was significant by 2-way ANOVA for repeated measurements (p  0.03). 20406080100120140160180 20406080100120140160180 20406080100120140160180 20406080100120140160180    L   V   E   S   V   (  m   l   ) Baseline BaselineFollow up Follow up89+24 82+ +9843 24  106+34G-CSF Control groupp=0.006p=0.412-way ANOVAp=0.021      L   V  e  n   d  s  y  s   t  o   l   i  c  v  o   l  u  m  e Figure 2. LV end-systolic volumes at baseline and follow-up in patients treated with G-CSF and in controls. In the control group, there was a significantincrease in LV end-systolic volume from baseline to follow-up (p  0.006). This dilation did not take place in the G-CSF group (p  0.41). This differenttrend was significant by 2-way ANOVA for repeated measurements (p  0.021).400  The American Journal of Cardiology (  trial, 10 so recovery seems strictly dependent on selection of patients and timing of baseline evaluation. Our group foundthat the near totality of spontaneous improvement after MI inpatients after primary or rescue percutaneous coronary inter-vention is observed during hospitalization without subsequentimprovements from discharge to 3-month follow-up. 15 The prevalent effect of G-CSF in our study is preven-tion of LV dilation that was observed in the controlgroup. In particular, we can speculate that G-CSF limitsthe infarct area, thus preventing infarct expansion, de-pending on myocardial regions bordering the infarction.With the strong limitation of the small sample of myo-cardial contrast echocardiographic subanalysis and thepurpose of generating hypotheses to test in future studies,myocardial contrast echocardiographic findings in pa-tients treated with G-CSF showed decreases in LV lengthand its percentage affected by myocardial akinesia, sug-gesting a kind of shrinking or retraction of the infarctzone despite an infarct expansion. This result could bedue in part to the well-known mobilization of stem/ progenitor cells from bone marrow and potential differ-entiation into cardiomyocytes, 19 even to the recently dis-covered properties of G-CSF directly on the myocardium.Experimentally G-CSF was proved to improve healing of the infarcted myocardium through increased expressionof tumor growth factor-   and type I collagen. 20 More-over, Harada et al 21 demonstrated a protective effect of G-CSF on ischemia/reperfusion injury through the Jak/ Stat pathway. G-CSF, through its G-CSF receptor that isnaturally expressed on cardiomyocytes and by activating 2530354045505560 2530354045505560    L   V   E   F   (   %   ) Baseline BaselineFollow up Follow up40+6 45+ +836 6  38+8G-CSF Control groupp=0.95p=0.0682-way ANOVAp=0.04      L   V  e   j  e  c   t   i  o  n   f  r  a  c   t   i  o  n   (   %   ) Figure 3. LV ejection fraction at baseline and follow-up in patients treated with G-CSF and in controls. In the G-CSF group, there was an increase in LVejection fraction from baseline to follow-up (p    0.068). This improvement did not take place in the control group (p    0.95). This different trend wassignificant by 2-way ANOVA for repeated measurements (p  0.04). 11,21,41,61,822,22,42,62,83 11,21,41,61,822,22,42,62,83    W   M   S   I Baseline BaselineFollow up Follow up2.00+0.16 2.12+0.321.74+0.22 2.09+0.38G-CSF Control groupp=0.63p=0.0032-way ANOVAp=0.07      W  a   l   l   M  o   t   i  o  n   S  c  o  r  e   I  n   d  e  x Figure 4. Wall motion score index at baseline and follow-up in patients treated with G-CSF and in controls. In the G-CSF group there was a significantdecrease in wall motion score index from baseline to follow-up (p    0.003). This improvement was not significant in the control group (p    0.63). Thisdifferent trend had borderline significance by 2-way ANOVA for repeated measurements (p  0.07).401 Coronary Artery Disease/G-CSF in Anterior Myocardial Infarction
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks