Send letter to journal:
Re: Digitalis in atrial fibrillation: not the best friend nor the worst foe.

Dear Editor,

We have carefully read the article by Gjesdal et al [1] showing the result of an exploratory analysis pooling data from the SPORTIF III and V trials [2,3]. By means of a Cox proportional hazard analysis the authors concluded that the use of digitalis may increase mortality.
Although we acknowledge the merit of this manuscript, in our opinion some points deserve further elucidation.
In a previous publication [4] we showed that anticoagulation therapy was inversely associated while digoxin (digitalis) and antiarrhythmic drugs were directly associated with increased mortality after adjustment for other covariates in patients treated with a “rate control” approach compared to patients treated with a “rhythm control” approach. That conclusion did not apply to patients with advanced heart failure (HF) as the studies available for the meta-analysis did not include patients with NYHA class IV HF.
Gjesdal et al suggested that the possible ominous effect of the digitalis may be concealed in patients with HF. However, no data are available about the possible progression to the end stage disease nor any mention about concomitant drug therapy, severity of symptoms, mechanism of HF and indication for digitalis use. They also reported a percentage of digoxin users up to 10% with levels in the range of toxicity, but they conclude that “there is little reason” to think that this is a serious problem as it was not clinically relevant.
Moreover, in a previous independent non–funded publication we showed that long-term ximelgatran therapy was associated with a significantly reduced risk of major life-threatening bleeding in the prevention of atrial fibrillation-related stroke (OR 0.71 [0.55-0.92], p=0.009, p for H=0.83, I2 0%) [5].
How did the authors address this point?
Of note, the authors did not mention the hepatotoxicity related to ximelagatran prolonged administration but they include ximelagatran therapy in the multivariate analysis. We showed that ximelagratan treatment > 3 months was associated with an odds ratio for hepatotoxicity of 6.73 (95% confidence interval: 5.01-9.05) compared to warfarin (p<0.001). In absolute terms, for prolonged treatments, the incidence of hepatotoxicity rose from 1.1% to 7.1%, with a number needed to harm of 17 [5-7]. This risk led the U.S. Food and Drug Administration to deny approval for ximelagratan in the USA, the European Agency for the Evaluation of Medicinal Products (EMEA) to allow only short term administration and the manufacturer (Astrazeneca), after a further fatality in a post-marketing study, to withdraw the drug [8-10]
We wonder if this point has been adequately addressed.
Although hepatic damage was clinically silent in most cases (even without discontinuation of therapy), 2 fatal cases of hepatic injury have been reported in the SPORTIF trials and no data are available about the long term hepatic function of enrolled patients. Although a possible pharmacodinamic/pharmacokynetic interaction between digoxin and ximelgatran has been not demonstrated as both metabolisms do not involve the liver cytochrome P-450 system [11], the “clinical interaction”, i.e. the impact of hepatotoxicity on mortality cannot be excluded. Thus, the “background noise” of hepatotoxicity should have been included in the analysis.
In conclusion, we wonder if a multi-drug context in which one of the two “study drugs” has been definitely withdrawn because of its potential fatal effect is really the appropriate context to assess the potentially fatal effect of a third drug.

Reference

1. K Gjesdal, J Feyzi and S B Olsson.
Digitalis: a dangerous drug in atrial fibrillation? An analysis of the SPORTIF III and V data.
Heart 2008:94;191-196.

2. Olsson SB, The Executive Steering Committee on behalf of the SPORTIF III Investigators.
Stroke prevention with the oral direct thrombin inhibitor ximelagatran compared with warfarin in patients with non- valvular atrial fibrillation (SPORTIF III): randomised controlled trial.
Lancet 2003;362:1691–8.

3. Albers GW, Diener HC, Frison L, et al.
Ximelagatran vs warfarin for stroke prevention in patients with nonvalvular atrial fibrillation: a randomized trial.
JAMA 2005;293:690–8.

4. Testa L, Biondi-Zoccai GG, Dello Russo A, Bellocci F, Andreotti F, Crea F.
Rate-control vs. rhythm-control in patients with atrial fibrillation: a meta-analysis.
Eur Heart J. 2005;26:2000-6.

5. Testa L, Andreotti F, Trotta G, Biondi Zoccai GGL, Burzotta F, Bellocci F, Crea F.
Ximelagatran/melagatran versus conventional anticoagulant therapy: meta-analysis of 13 randomised controlled trials enrolling 22639 patients. European Society of Cardiology/World congress of Cardiology 2007. Oral Presentation for the Young Investigator Award in Thrombosis.
Published in Int J Cardiol. 2007 Nov 15;122(2):117-24.

6. Testa L, Bhindi R, Agostoni P, Abbate A, Zoccai GG, van Gaal WJ.
The direct thrombin inhibitor Ximelagatran/melagatran: a systematic review on clinical applications and an evidence based assessment of risk benefit profile.
Expert Opin Drug Saf. 2007;6:397-406.

7. Testa L, Van Gaal W, Agostoni P, Abbate A, Trotta G, Biondi-Zoccai GG.
Ximelagatran versus warfarin in the prevention of atrial fibrillation- related stroke: both sides of the story.
Stroke. 2007 ;38(7):e57.

8. AstraZeneca receives action letter from FDA for Exanta™ (ximelagatran).
Available at http://fdaadvisorycommittee.com

9. Successful outcome of the mutual recognition procedure for Exanta™ (ximelagatran) in Europe.
Press release from European Agency for the Evaluation of Medicinal Products (EMEA).
Available at http://www.emea.eu.int

10. AstraZeneca Decides to Withdraw Exanta™ (ximelagatran).
Press release from AstraZeneca international, February 14, 2006.
(Accessed February 15, 2006, at http://www.astrazeneca.com/pressrelease).