Assumptions and the CAST Trial

Not long ago a heart attack was much tougher to treat

In 1982 the National Heart, Lung and Blood Institute (NHLBI) launched the Cardiac Arrhythmia Suppression Trial (CAST) which was completed in 1986 and published in the New England Journal of Medicine in 1989.  I learned about it the year it came out from my medical school professors since it caused quite a stir in the cardiology community.  While I don’t think medical students hear about it anymore (it is 20 years out of date, after all) I think they could still glean a valuable lesson from it.

Let’s go back in time.  If you had a heart attack in 1982 there really wasn’t much that could be done.  Angioplasty, developed in 1977, hadn’t made its way to mainstream hospitals until the mid-80s, and streptokinase, the first thrombolytic, wasn’t in common use until the GISSI trial was published in 1986.  Mostly, you would come into the hospital and receive morphine and nitroglycerin until the affected heart tissue expired and the pain subsided.  Long ago doctors noticed that people who exhibited ventricular electrical irritability in the form of frequent premature ventricular complexes (PVCs) and ventricular tachycardia were at much higher risk of dying from ventricular fibrillation (VF) in the early period after a heart attack.

Several rhythm control drugs were available at that time and appeared to do a good job of decreasing the frequency of PVCs.  It became common practice to empirically start patients on medications such as lidocaine in an effort to decrease the risk of death from ventricular arrhythmia.

It was during that time that several other antiarrhythmic drugs became available.  Encainide, flecainide, and moricizine were considerably more effective than lidocaine at extinguishing PVCs and became popular among cardiologists.  If lidocaine is good—so the thinking went—these others must be better.

So the logic went like this:

1. More PVCs equals more death from VF
2. Lidocaine decreases the rate of PVCs
3. Ergo, using medications even better than lidocaine will prevent death from VF

Here’s where the lesson comes in.  In medicine we are taught to implement a hypothesized therapy into general use only after it has been rigorously validated.  This last part was bypassed, as lidocaine, encainide, flecainide and moricizine were all used in thousands of patients before any significant research was done to validate the line of logic.

Enter the CAST trial.  When the NHLBI proposed the design of this study (prospective, placebo controlled) some prominent members of the cardiology community were aghast at the possibility of withholding an established therapy from the patients who were randomized to the placebo arm.  They claimed that it was medical malpractice to not give antiarrhythmics to patients with heart attacks and frequent PVCs.

What is good for some might be bad for everyone

Skip forward to 1989.  The results of the CAST trial settled the debate and retired the routine use of antiarrhythmics in the setting of heart attacks.  It turns out that the empiric use of these medications actually led to a higher rate of death among these patients than the placebo pills.  Apparently this class of medications actually stimulates a more lethal type of ventricular arrhythmia in patients with recently-injured heart muscle.

In the ensuing years encainide and moricizine were pulled from the market, and flecainide is now used only cautiously in younger patients with atrial fibrillation.  We no longer attempt to suppress asymptomatic ventricular activity in people with recent heart attacks and instead focus on restoring blood flow.

The lesson can’t be forgotten.  Even if a course of therapy seems logical we can’t deploy it into general use until the hypothesis is thoroughly tested.  Assumptions and anecdotal evidence are not enough.