Related Resources • Heart failure review • Beta blockers in heart failure • Ace inhibitors in heart failure • CRT in heart failure

By DrRich

Dateline: November 4, 2002

In an article published in the October 31 issue of the New England Journal of Medicine, researchers report that treatment with digoxin may increase the rate of death in women with heart failure.   This finding is regarded as extremely surprising, as digoxin (and similar drugs in the digitalis group) are some of the oldest and most revered drugs in common usage for patients with heart failure, and it has already stirred up strong feelings among cardiologists.

The authors, who are from Yale University, reached this startling conclusion after examining data from a previously-published clinical trial: the Digitalis Investigation Group (DIG) trial.  In the DIG trial, men and women with heart failure were randomized to receive either digoxin or a placebo. The DIG trial was conducted in the first place because, while digitalis has been in common use for many decades, there was little data from randomized trials to document its efficacy in treating heart failure.  The DIG trial was designed to put the issue to rest once and for all.

The main findings of the DIG trial were that digoxin does not improve mortality in patients with heart failure, but it does seem to reduce the need for hospitalization in those patients.  In other words, digoxin was not as effective as more modern drugs currently used in heart failure, but it seemed to help.  After the DIG trial, it became scientifically OK (but not mandatory) to treat patients who have heart failure with digoxin.  Case closed - or so every one thought.

But then the Yale investigators applied to the NIH (which "owns" the DIG trial) to re-evaluate the data from this trial in order to study possible gender-differences in the response to heart failure therapy. What they found surprised them: while digoxin reduced the mortality rate from cardiovascular causes in men, the women in the DIG trial who received digoxin had a 4% increase in cardiovascular death, as compared to women taking placebo. This translates to one excess death for every 25 women taking digoxin. (At the same time, the rate of hospitalization in women receiving digoxin was reduced.)

What could account for the excess of death in women taking digoxin?

At least 3 possible explanations for the excess deaths in women have been proposed:

1) From a statistical standpoint, any post hoc (i.e., after the fact) analysis of data from a randomized trial is subject to question.  The DIG study was not designed to examine gender differences in the response to digoxin, so any conclusions drawn regarding gender differences have less-than-optimal statistical value.  The finding of increased mortality in women may thus represent a statistical artifact rather than a real phenomenon.

2) Digoxin levels in women, at least in women who had digoxin levels recorded in the DIG study, seemed to be higher than in men, so the differences in outcome could possibly be related to the digoxin level rather than to gender.  If this were the case, perhaps the excess mortality could be mitigated by using fairly low doses of digoxin. However, investigators were not required to record digoxin levels in the DIG trial.  Since relatively few patients had digoxin levels recorded, it is not possible to know if these levels were actually higher in women.

3) Perhaps digoxin is simply more dangerous in women than in men; that is, perhaps women are just more sensitive to digoxin.

While the true explanation for this disturbing finding is unknown, DrRich finds explanation #3 to be at least a plausible one. One of the well-documented toxic effects of digitalis drugs is that they can cause life-threatening heart arrhythmias. These arrhythmias, more prominent among digitalis-toxic patients who also have certain metabolic abnormalities such as low potassium levels, can produce sudden death.  It is intriguing that the increase in mortality among women taking digoxin is NOT accompanied by an increase in hospitalization - this finding suggests sudden death as a mechanism of the increase in mortality.  All these findings could be explained by an increase among women in sensitivity to digoxin, or in the metabolic disturbances that provoke digitalis-induced arrhythmias.

What are doctors going to do about this new finding?

Who knows? Quite possibly nothing, since the "experts" aren't going to agree on its meaning.

The scientific bodies that devise official guidelines for the treatment of heart failure, for instance, will simply not be able to make a strong statement about using digoxin in women, since they do not pay much attention to post hoc data analyses.  (There are sound statistical reasons for them to adopt this policy.)  So it is unlikely that the guidelines, or physician habits, will be changed on the basis of this new finding. 

What should women with heart failure do about this new finding?

Women with heart failure do not labor under the same statistical constraints as scientific deliberative bodies.  What they need to know is: what is likely to be the best thing for ME?  These women need to keep in mind that only 20% of the controversy among doctors about this new finding is due to the fact that digoxin may be useful in some patients with heart failure. The other 80% is due to the fact that doctors are used to prescribing digoxin for heart failure, whether it does any good or not, and along comes a study messing with tradition.  A controversy is to be expected. 

In any case, here is what women with heart failure need to know about the use of digoxin:

1. At best, the benefits from using digoxin in heart failure are relatively marginal; that is, digoxin may provide a modest reduction in the need for hospitalization.
2. At worst, digoxin might substantially increase the risk of death.
3. In the modern era, even taking the results of the DIG trial at face value, there are far more effective treatments for heart failure than digoxin.  If beta blockers, ACE inhibitors and diuretics are used to their full potential, it is likely that digoxin will not be needed (and may not be additionally effective) at all.
4. According to every survey that has been done, beta blockers and ACE inhibitors (both of which not only improve symptoms but also increase survival) are grossly underutilized by doctors treating heart failure.
5. Using digoxin in women with heart failure before beta blockers and ACE inhibitors have been added, and their doses maximized, probably ought to be considered bad medicine (since at the least it suggests the doctor is not well-versed in the modern treatment of heart failure.)
6. If the use of beta blockers, ACE inhibitors and diuretics has been maximized, and full consideration of cardiac resynchronization therapy has been given, then (and only then) is it reasonable to consider adding digoxin in a woman with heart failure who has continuing symptoms.  Only when you get to this point do you actually need to worry about the possible increase in mortality caused by digoxin. 

By following this logic tree, women with heart failure will only rarely need to be concerned about a digoxin-induced mortality risk.  Much more commonly, they will need to concern themselves with finding a doctor who actually knows what he/she is doing in the management of heart failure.