Skip to main content
1004-2012, 1002-2004, 1005-2015
class com.aha.ucm.component.cis.TagListPageData=[,docNativeURL=null,docName=UCM_464788,docStatus=RELEASED,dOutDate=null,docSSFileName=UCM_464788_Effect-of-n-3-Polyunsaturated-Fatty-Acids-or-Rosuvastatin-in-Patients-with-Chron.jsp,docTitle=Effect of n-3 Polyunsaturated Fatty Acids or Rosuvastatin in Patients with Chron,xWebsites=professional,dDocAuthor=amy.stefanos,xNextReviewDate=05/07/2016 4:10 PM,xTier1=36,xFeaturedItem=No,xElectronicRegistration=No,UserLocale=null,xSubCategory=,dpEvent=null,xComments=Commentary from Jeffrey Anderson, MD, FAHA, on the GISSI-HF trial,NoHttpHeaders=null,UserTimeZone=null,xRegionDefinition=GENERIC_RD_COL_1,xVideoRenditions=,xSnippetItem=,xNotes=autoconverted |,UserDateFormat=null,encodeDocUrl=null,isDocProfileDone=null,xKeywords=science news, commentary, aerobic exercise; exercise, physical; physical activity; rehabilitation; strength training,xTier2=,refreshSubMonikers=null,xEditorStepReassignedUsers=null,xLinkTextToDisplay=,dDocAccount=WCM/SOP/SMD,xEndDateTime=null,xClbraAliasList=null,ClientEncoding=null,xCpdIsLocked=0,xUsageRightsDate=null,xModifyDate=01/24/2020 9:59 PM,xTier3=,xEventDate=null,dSubscriptionType=null,xCopyright=No,xPackagedConversions=,dSubscriptionAlias=null,xStorageRule=,dpName=null,xDepartment=Science Operations,dStatus=RELEASED,dPublishType=,xCopyrightDetails=,xSubType=64,isDocProfileUsed=null,xWebsiteObjectType=Data File,xWebFlag=,xSeeAlsoLinks=,xClbraUserList=null,xPartitionId=,xCpdIsTemplateEnabled=0,xLinkWebAddress=,xDontShowInListsForWebsites=,xStartDateTime=null,dInDate=11/05/2019 4:03 PM,xWebsiteSection=professional:1463,dDocName=UCM_464788,dpAction=null,dRevLabel=8,dSecurityGroup=AHAMAH-Public,xCategory=,refreshMonikers=null,xDamConversionType=,dDocFormats=null,xAssociatedImage=,dDocType=SingleColumn,xBusinessOwner=Business Owner,xUploadDate=null,xDiscussionCount=0,xMainFlowEntryCriteria=True,xItemInformation=,xUsageRights=,xDiscussionType=N/A,xRecipeTaxonomy=,dSubscriptionID=null,dOriginalName=UCM_464788.xml,xProfileTrigger=SingleColumn,dLocation=,dRevisionID=7,dPublishState=,dReleaseState=Y,xTrashDeleter=null,dMessage=,dWebExtension=xml,dExtension=xml,dProcessingState=Y,xTrashDeleteName=null,dIsCheckedOut=0,xForceFolderSecurity=null,dRevClassID=464788,dIsPrimary=1,dFileSize=21301,dIndexerState=,dFlag1=,xviaAddNewContentService=,dIsWebFormat=0,xCollectionID=null,dRevRank=0,xReadOnly=null,dCheckoutUser=,dFormat=Application/xml,dWorkflowState=,dDocID=1909450,dRendition2=,dRendition1=,xInhibitUpdate=null,dReleaseDate=01/24/2020 9:59 PM,xTrashDeleteLoc=null,dCreateDate=11/05/2019 4:07 PM,xHidden=null,labelTier1=ScienceNews,labelTier2=,labelTier3=,labelTier4=,mobileNavURL=DEFAULT2_VALUE_FROM_getDataForAdvanceSearch,xContactPhoneNumber=,xContactEmailAddress=,xContactName=,xATGRolesDisciplines=,xPublishDate=03/06/2009 6:00 PM,xRobotParameter=,xCommunities=1004-2012, 1002-2004, 1005-2015,xMembershipLevel=,rsCalories=null,rsSodium=null,rsRecipeTaxonomy=null,rsServings=null,rsTotalTime=null,rsTotalFat=null,rsTotalCarbs=null,rsFeaturedImage=null,xDisplayComments=

Effect of n-3 Polyunsaturated Fatty Acids or Rosuvastatin in Patients with Chronic Heart Failure

Disclosure: None
Pub Date: Friday, March 6, 2009
Authors: Jeffrey Anderson, MD, FAHA 


  1. Tavazzi L, Maggioni AP, Marchioli R, Barlera S, Franzosi MG, Latini R, Lucci D, Nicolosi GL, Porcu M, Tognoni G.,  Effect of n-3 polyunsaturated fatty acids in patients with chronic heart failure (the GISSI-HF trial): a randomised, double-blind, placebo-controlled trial.,  Lancet,  372 (9645) 1223-30. View in PubMed
  2. Gissi-HF Investigators, Tavazzi L, Maggioni AP, Marchioli R, Barlera S, Franzosi MG, Latini R, Lucci D, Nicolosi GL, Porcu M, Tognoni G.,  Effect of rosuvastatin in patients with chronic heart failure (the GISSI-HF trial): a randomised, double-blind, placebo-controlled trial.,  Lancet,  372 (9645) 1231-9. View in PubMed

Clinical Question

Heart failure (HF) represents a major component of the burden of cardiovascular (CV) morbidity and mortality, and its relative impact continues to increase. Contemporary medical therapies have only modestly impacted this burden. Thus, discovering additional therapies that innovatively target HF pathways is a major goal of medical research.

The Gruppo Italiano per lo Studio della Sopravvivenza nell'Infarto miocardico (GISSI)-Prevenzione trial investigators previously showed that n-3 polyunsaturated fatty acids (PUFA) lowered total and especially sudden death rates after myocardial infarction (Lancet 1999;354:447). Currently, PUFA is the only dietary supplement recommended for CV prevention (Circulation 2007;115:1481). PUFA also exerts important lipid lowering and vascular antiinflammatory properties and is well tolerated, but it had not been tested for efficacy in a HF population.

Hydroxymethylglutaryl-coenzyme A reductase inhibitors (statins) have experienced enormous success for both primary and secondary prevention of atherothrombosis. In addition to lipid-lowering, they have been postulated to exert beneficial pleiotropic effects. Promising but inconclusive data from mostly observational sources suggested that statins might reduce CV mortality in patients with HF of various etiologies (Int J Cardiol 2007;119:48). Rosuvastatin is a potent, well-tolerated statin recently shown to decrease CV events in subjects with normal low-density lipoprotein (LDL) but elevated C-reactive protein (NEJM 2008;359:2195).

The GISSI-investigators thus undertook a randomized, factorial trial of PUFA and rosuvastatin in chronic HF.


The investigators performed a randomized, double-blind, placebo-controlled, factorial design study in 7,046 patients from 347 Italian centers. Adults with chronic, class II-IV HF of any cause and ejection fraction (EF) were eligible. Advanced comorbidities (e.g., cancer, liver disease), acute coronary syndrome or revascularization within 1 month, planned cardiac surgery within 3 months, and pregnancy or child-bearing potential were exclusions. Subjects were randomly assigned to 1g PUFA daily or to matching placebo. A second (factorial) randomization to rosuvastatin, 10 mg daily or matching placebo was performed in 4,631 eligible patients (i.e., statin not given, not indicated, and not contraindicated). The PUFA study cohort averaged 67 years of age; EF averaged 33% (only 10% had EF >40%); HF class was II in 63%, III in 34%, and IV in 3%. HF etiologies were ischemic in 50%, dilative in 29%, and hypertensive in 15%. Patients received appropriate HF therapies. The statin cohort characteristics were similar except that etiology was ischemic in 40%.

The study concluded after a median of 3.9 years of follow-up. PUFA reduced total mortality from 29.1% to 27.3% [adjusted hazard ratio (HR) 0.91, p=0.04; number needed to treat (NNT) = 56] and death or CV hospitalization (the coprimary endpoint) from 59%to 57% (HR = 0.92, p = 0.009; NNT = 44). In treatment compliant patients, the HR for all-cause death was 0.86 (p = 0.004). Secondary outcomes of CV death, sudden death, CV admissions, and MI were numerically lower. No heterogeneity was observed in major subgroups. As expected, PUFA decreased triglycerides (slightly) and was well tolerated, with similar total (29%) and adverse reaction-related rates of discontinuation (3%) as placebo.

In contrast, rosuvastatin had no impact on all-cause mortality (29% vs. 28%, adj. HR 1.00, p = 0.94) and on death or CV hospitalization (57% vs. 56%, adj. HR 1.01, p = 0.90). Secondary outcomes also were not affected. These neutral results occurred despite LDL cholesterol reductions of 32% at 1 year and 27% at 3 years and C-reactive protein reductions of 17%. Also, there was no difference in results by degree of LDL reduction or by compliance. Rosuvastatin was well tolerated, with similar total (35%) and adverse reaction-related rates of discontinuation (4.6%) as placebo except for a small increase in discontinuations from rises in liver enzymes (26 vs. 12 patients). No interaction was noted between rosuvastatin and PUFA. 

Clinical Implication/Application

These study results are intriguing and instructive. A priori, one might have assumed that statin therapy would be more likely to show benefit than PUFA, but the opposite was noted. Thus, GISSI-HF emphasizes both the need for large randomized trials to establish therapeutic efficacy and the limitations of small or observational databases, even when intuitively appealing. For PUFA, the study showed that a 1 g daily supplement given on top of conventional HF therapies can safely provide a modest additional reduction in morbidity and mortality (both HF-related and arrhythmic) of 9% to perhaps 14% (with compliance). In contrast, rosuvastatin showed absolutely no effect of CV outcomes despite being well tolerated and reducing LDL cholesterol (by about 30%) as well as C-reactive protein. The GISSI-HF study stands in contrast to the earlier, mostly observational database, but it is consistent with the large, parallel controlled trial, CORONA (NEJM 2007;357:2248), which also tested rosuvastatin 10 mg. CORONA exclusively enrolled patients with ischemic HF. CORONA patients were somewhat older and more symptomatic than in GISSI-HF. Despite this, CORONA found only a nonsignificant 8% reduction in CV death, MI, or stroke (although fewer CV- or HF-related admissions occurred). Results of CORONA and GISSI-HF are still consistent with reductions in acute ischemic episodes in HF patients, but we learn from these studies that the rate of these episodes is surprisingly low in this population (e.g., only 0.7% fatal MI rate over 4 years in GISSI-HF). The modest impact of ischemic events in advanced HF is poorly recognized but not new: the large BEST HF study found only an 0.8% rate of fatal MI and a 1.6% rate of nonfatal MI over 4.1 years (Am J Cardiol 2005;95:558).

The clinical implications of GISSI-HF may be summarized as follows: For PUFA, supplementation may be recommended as safe, modestly effective preventive therapy in chronic HF, consistent with its use in other primary and secondary CV prevention settings. For statin therapy, GISSI-HF assists in decision making as follows: (a) Statins are not indicated in patients with non-ischemic HF; (b) Statins are not efficacious in patients with ischemic HF who do not have a guidelines-supported indication; (c) However, in HF patients with an indication for lipid-lowering, including those at higher than average ischemic risk, there is reassurance that statins can be given safely. 

-- The opinions expressed in this commentary are not necessarily those of the editors or of the American Heart Association.