Effect of Fish Oil on Ventricular Tachyarrhythmia and Death in Patients With Implantable Cardioverter Defibrillators: The Study on Omega-3 Fatty Acids and Ventricular Arrhythmia (SOFA) Randomized Trial

Context Very-long-chain n-3 polyunsaturated fatty acids (omega-3 PUFAs) from fish are thought to reduce risk of sudden death, possibly by reducing susceptibility to cardiac arrhythmia.

Objective To study the effect of supplemental fish oil vs placebo on ventricular tachyarrhythmia or death.

Design, Setting, and Patients The Study on Omega-3 Fatty acids and ventricular Arrhythmia (SOFA) was a randomized, parallel, placebo-controlled, double-blind trial conducted at 26 cardiology clinics across Europe. A total of 546 patients with implantable cardioverter-defibrillators (ICDs) and prior documented malignant ventricular tachycardia (VT) or ventricular fibrillation (VF) were enrolled between October 2001 and August 2004. Patients were randomly assigned to receive 2 g/d of fish oil (n = 273) or placebo (n = 273) for a median period of 356 days (range, 14-379 days).

Main Outcome Measure Appropriate ICD intervention for VT or VF, or all-cause death.

Results The primary end point occurred in 81 (30%) patients taking fish oil vs 90 (33%) patients taking placebo (hazard ratio [HR], 0.86; 95% confidence interval [CI], 0.64-1.16; P = .33). In prespecified subgroup analyses, the HR was 0.91 (95% CI, 0.66-1.26) for fish oil vs placebo in the 411 patients who had experienced VT in the year before the study, and 0.76 (95% CI, 0.52-1.11) for 332 patients with prior myocardial infarctions.

Conclusion Our findings do not indicate evidence of a strong protective effect of intake of omega-3 PUFAs from fish oil against ventricular arrhythmia in patients with ICDs.

Trial Registration clinicaltrials.gov Identifier: NCT00110838

Sudden cardiac death is one of the most common and often the first manifestation of coronary heart disease. It is responsible for approximately 50% of all mortality from cardiovascular disease in the Western world.¹ The majority of sudden deaths are directly caused by acute ventricular arrhythmia.² Epidemiological studies indicate that intake of very-long-chain n-3 polyunsaturated fatty acids (omega-3 PUFAs) as present in fish or fish oil is associated with a reduction in cardiovascular mortality.^3-7 These observational studies showed a strong relationship between omega-3 PUFAs and sudden death, but not between omega-3 PUFAs and nonfatal heart disease.^5,8,9 This is confirmed by the outcome of some clinical trials; in the Diet and Reinfarction Trial (DART),¹⁰ consumption of fish or fish oil reduced fatal heart disease by more than 30%, and the open-label Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto miocardico (GISSI) trial⁴ found a 45% reduction in sudden death in patients consuming omega-3 PUFAs.

Animal and in vitro studies also suggest an effect of omega-3 PUFAs through reduced propensity for arrhythmia.^11-18 Infusion of omega-3 PUFAs also prevented inducible sustained ventricular tachycardia (VT) in some patients.¹⁹ In contrast, Burr et al did not show a protective effect of intake of fish or fish oil on cardiac death in a trial of 3114 patients with stable angina.²⁰ Furthermore, 2 recent trials on fish oil and ventricular arrhythmia in patients with implantable cardioverter-defibrillators (ICDs) yielded inconclusive results.^21,22 Thus, the potential of omega-3 PUFAs for reducing risk of life-threatening arrhythmia in patients with ICDs is unclear.

We report the effect of omega-3 PUFAs from fish on the incidence of recurrent ventricular arrhythmia and all-cause mortality in a large double-blind, randomized trial of patients with ICDs.

Corresponding Author: Ingeborg A. Brouwer, PhD, Wageningen Centre for Food Sciences, Human Nutrition, Bomenweg 2, 6703 HD Wageningen, the Netherlands (ingeborg.brouwer@wur.nl).

Author Contributions: Dr Brouwer had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Brouwer, Zock, Camm, Böcker, Hauer, Wever, Ronden, Katan, Schouten.

Acquisition of data: Brouwer, Zock, Böcker, Hauer, Wever, Dullemeijer, Ronden, Lubinski, Buschler, Schouten.

Analysis and interpretation of data: Brouwer, Zock, Camm, Dullemeijer, Katan, Lubinski, Schouten.

Drafting of the manuscript: Brouwer, Zock, Camm, Hauer, Schouten.

Critical revision of the manuscript for important intellectual content: Brouwer, Zock, Camm, Böcker, Hauer, Wever, Dullemeijer, Ronden, Katan, Lubinski, Buschler.

Statistical analysis: Brouwer, Zock, Dullemeijer, Schouten.

Obtained funding: Brouwer, Zock, Katan, Schouten.

Administrative, technical, or material support: Brouwer, Böcker, Hauer, Dullemeijer, Ronden, Buschler.

Study supervision: Brouwer, Zock, Camm, Böcker, Hauer, Katan, Schouten.

Financial Disclosures: None reported.

Funding/Support: Funding for this study was provided by Wageningen Centre for Food Sciences, a non-profit alliance of major Dutch food industries, TNO Nutrition and Food Research, Maastricht University and Wageningen University and Research Centre, the Netherlands, with financial support by the Dutch government. An additional grant was received from the European Union (SEAFOODplus integrated project: No. 506359).

Role of the Sponsor: The SOFA trial was designed by the members of the executive and steering committees and then approved by the member organizations of the funder, Wageningen Centre for Food Sciences (WCFS). No industry members of WCFS were involved in the conduct of the study; in the collection, management, analysis, and intrepretation of the data; or in the preparation, review, and approval of the manuscript. Dr Peter Zock left Wageningen University on February 1, 2005, to join Unilever, which is a member organization of the trial's funder WCFS. Since February 2005 he has contributed to writing the manuscript. However, this was not in his capacity as a Unilever employee, and Dr Zock's contributions to the manuscript since February 2005 were neither submitted to, nor seen, revised, or approved by his employer, Unilever.

The SOFA Study Group:Executive Committee/Data analyses: I. A. Brouwer, C. Dullemeijer; Steering Committee: E. G. Schouten (chair), A.J. Camm (co-chair), P. L. Zock, D. Böcker, R. N. W. Hauer, E. F. D. Wever, J. E. Ronden; Investigators: A. Zienciuk, A. Lubinski (Medical University of Gdansk, Gdansk, Poland), J. Vogt, H. Buschler (Herz-und Diabeteszentrum Bad Oeynhausen, Bad Oeynhausen, Germany), J. Köbe, D. Böcker (Universitätsklinikum Münster, Münster, Germany), Z. Kornacewicz-Jach, R. Rzeuski (Pomeranian Academy of Medicine, Szczecin, Poland), A.R. Ramdat Misier (Isala Clinics, Zwolle, the Netherlands), E. F. D. Wever (St. Antonius Hospital, Nieuwegein, the Netherlands), H. Szwed, A. Przybylski (Instytut Kardiologii, Warzawa, Poland), R. N. W. Hauer (UMC Utrecht, Utrecht, the Netherlands), M. Sepsi, B. Semrad (FN Brno Interni Kardiologicka Klinika, Brno, the Czech Republic), A. Meijer (Catharina Hospital, Eindhoven, the Netherlands), L. J. Jordaens (Erasmus Medical Centre Rotterdam, Rotterdam, the Netherlands), A. J. Camm, I. Savelieva (St George's Hospital, London, England), R. Tavernier (University Hospital Gent, Gent, Belgium), H. R. Figulla, C. Kalmbach (Klinikum der Friedrich-Schiller-Universität, Jena, Germany), A. A. M. Wilde (Academic Medical Centre, Amsterdam, the Netherlands), M. Trusz-Gluza, A. Filipecki (I Klinika Kardiologii, Katowice, Poland), W. Haverkamp (Medizinische Klinik m. S. Kardiologie, Berlin, Germany), M. Zabel (Klinikum Benjamin Franklin Berlin, Berlin, Germany), J. M. Morgan (Southampton General Hospital, Southampton, England), H. Pitschner (Kerckhoff-Klinik GmbH, Bad Nauheim, Germany), M. J. Griffith (Queen Elizabeth Hospital, Birmingham, England), A. Podczeck, K. Steinbach (Wilhelminenspital Wien, Vienna, Austria), M. Block (Stiftsklinik Augustinum, Munchen, Germany), P. Kulakowski (Grochowski Hospital, Warszawa, Poland), J. Kautzner (Institute of Clinical and Experimental Medicine, Prague, the Czech Republic), M. Fromer (CHUV, University Hospital Lausanne, Lausanne, Switzerland).

Core Laboratory: D. A. M. J. Theuns (Cardialysis, Rotterdam, the Netherlands).

Endpoint Adjudication Committee Members: I. Savelieva, C. Wolpert (University Hospital Mannheim, Mannheim, Germany), A. A. M. Wilde, J. O. Schwab (Universitätsklinikum Bonn, Bonn, Germany), M. Fromer, M. J. Griffith, L. J. Jordaens (Erasmus University Rotterdam, Rotterdam, the Netherlands).

Data and Safety Monitoring Board Members: E. Boersma (Erasmus Medical Centre Rotterdam, Rotterdam, the Netherlands), P. J. Schwartz (University of Pavia, Pavia, Italy), K. Swedberg (Göteborg University, Göteberg, Sweden).

Acknowledgment: The data analyses were checked by E. Boersma, MSc, PhD, statistician at the Erasmus Medical Centre Rotterdam, Department of Cardiology, the Netherlands. Dr Boersma received compensation for his work as the chair of the data and safety monitoring board, but did not receive any compensation for checking the data analyses. We thank all the research nurses, co-investigators, research assistants at Wageningen University, the personnel of the Core Laboratory in Rotterdam and the Laboratory of the Division of Human Nutrition in Wageningen for their valuable contributions to the study.