Magnesium? Winnie asked, keeping her tone as loose as an unwinding coil of slack line. Magnesium?
Yup. Magnesium, I repeated. One gram iv.
I caught her lingering look, but Winnie and I go back six years. Once she was sure I meant it, she produced the vial. One gram, coming up. Mrs. Wu, our elderly patient, had come in gasping for air. She had been shopping with her daughter when suddenly her legs lost power and her lungs grew tight. As soon as Mrs. Wu arrived, Winnie had put her on oxygen; now she seemed a little better. But she still clutched her daughter’s hand and kept up a steady singsong of Cantonese: Maybe Chinese medicine . . . I imagined her saying, then the daughter soothing, It’s the hospital. They know.
Mrs. Wu’s electrocardiogram—the tracing of electrical activity in the heart—looked like a Jackson Pollock. Instead of a smooth tracing bringing forth 80 beautifully spaced spikes a minute, it was spitting out a rapid-fire and asynchronous 140 spikes, or heartbeats, a minute. Mrs. Wu’s heart was creating a painting that already had a title: atrial fibrillation.
Our hearts have four chambers to handle the intricate job of replenishing oxygen-poor blood and pumping it out. While the right atrium and right ventricle pump oxygen-poor blood from the body out to the lungs, the left atrium and left ventricle receive oxygen-rich blood from the lungs and pump it to the rest of the body. The atria and ventricles must work together to achieve the split-second timing that allows the chambers of the heart to pump blood in and out some 60 to 100 times a minute, day in, day out. The atria set the rhythm for the muscular contractions we call the heartbeat, and the ventricles supply the horsepower.
A heartbeat begins when the natural pacemaker in the right atrium fires. The electric discharge then ripples through the atria, which contract to propel blood into the ventricles. To set off a contraction in the ventricles, the electric signal then passes through a cluster of cells—called the atrioventricular node—in the fibrous tissue separating the atria from the ventricles below. The signal then streaks down specialized trunk cables called Purkinje fibers that fan out into the thick ventricular muscle and bring forth a life-giving contraction.
In Mrs. Wu’s case, each atrial cell was blaring at will like an orchestra in revolt. Millions of clashing electric wavelets reduced her atrial ripple to chaos. There are many conditions, from infections to heart attacks, that can cause atrial fibrillation. But it can also mysteriously pounce on healthy young people. In Mrs. Wu’s case, her 72 years were probably to blame, though we would still hunt for the usual suspects.
I glanced at the overhead monitor. Mrs. Wu watched me intently. She raised her right hand and fluttered it over her chest. Yes, I know, I nodded. Her heartbeats came in a rush of fluorescent green spikes. The squiggly chaos in between represented her atria shuddering—500 times a minute. Luckily, the atrioventricular node is a responsible gatekeeper; it doesn’t respond to every erratic atrial discharge, or the ventricles would quiver into life-ending impotence. But the atrioventricular node can’t always hold down the atrial cacophony. In some cases, when it fails, the ventricles beat too fast and fill inefficiently. Blood pressure then drops and blood starts eddying back into the lungs, cutting off oxygen flow to the body.
The solution is theoretically simple: boost the atrioventricular node’s blocking power and slow down the ventricles. Resynchronizing the atria would be even better, but that’s a harder trick.
Atrial fibrillation has been recognized for centuries. And so, curiously, has the mainstay of treatment, digitalis. In 1785, William Withering described the foxglove plant’s power over the motion of the heart. He had his patients chew the foxglove leaf for the desired effect; we modern hotshots now understand that the active principle in foxglove is digitalis, and we use a digitalis analogue—digoxin—in pills and intravenous solution. The compound works, through a complicated mechanism, by boosting the atrioventricular node’s damping action and thus slowing the ventricular rate. There are newer drugs, to be sure, but each has its drawbacks. Among the most widely used is diltiazem, a calcium channel blocker. But that brings us to ions, and magnesium, and ancient seas.
Electrical impulses—the life-giving current between nerve and muscle cells—arose when our primordial ancestors fashioned cell membranes that strictly control the flow of different ions in and out of the cell. (An ion is a molecule or atom that has lost or gained an electron; the main players in heart tissue are potassium, sodium, and calcium ions.) Cells in the heart suck potassium ions in and push sodium ions out. This inside-outside difference is achieved by pumps in the cell membrane that work like submicroscopic waterwheels, exchanging sodium ions for potassium ions. The resulting segregation of ions creates a small electrical potential within the cell.
But electrical impulses cannot travel from cell to cell without another mechanism: ion channels embedded in the cell membrane. When jolted by an electric discharge from a neighboring cell, certain ion channels in the cell membrane open, allowing sodium ions to pour in. When the influx of sodium reaches a threshold, it triggers the release of electrons from the cell—this is the electrical impulse. Ion channels in a neighboring cell sense the electron discharge and open, allowing sodium ions to rush in. This, in turn, changes the charge within that cell, causing it to unleash a storm of electrons to its neighbor. It’s a domino effect that happens so fast that the electrical impulse is propagated nearly instantaneously across the cardiac muscle. The result: The coordinated contraction we call the heartbeat. Meanwhile, the first cell’s pumps drive sodium ions out and suck potassium ions back in, thus restoring the cell’s original state of affairs.
The bandleaders of this intricate sequence are the pacemaker cells. They and the cells in the atrioventricular node depend not on sodium but on a third ion, calcium, to slowly leak in through its channels and trigger a discharge. This system of channels and pumps makes the most intricate watchwork look like child’s play. But its intricacy is also its weak point. When the timing mechanism is off, normally obedient cells start firing on their own, sparking all sorts of mayhem.
Which is where magnesium—call it the damage-control ion—comes in. Magnesium sulfate, the stuff of Epsom salts and an essential nutrient, has been recognized for more than 40 years as having antiarrhythmic properties. I stumbled across this tidbit—a small paragraph buried in a large text—while recently studying for my emergency-medicine board review. But I couldn’t find any large, blinded trials testing its efficacy in treating atrial fibrillation. I asked several friends in cardiology, Magnesium for A-fib?
Never heard of it, came the uniform reply.
How magnesium might work remains a mystery. Maybe it blocks calcium channels in cardiac cells to make the atrioventricular node more resistant. Maybe by helping the sodium-potassium pump and boosting intracellular potassium, it makes firing less random. Attempts to fiddle with the heart’s firing mechanism (atrial fibrillation is but one of dozens of arrhythmias) led to many new wonder drugs that went bust. Tragically, some even killed more than they cured.
But magnesium is generally innocuous, I reasoned, and it was said to work rapidly. Mrs. Wu’s eyes were closed now, but her hand kept fluttering over her chest like an anxious sparrow. The faster I slowed Mrs. Wu’s rate the better. Digitalis would work, but it takes three hours for its effects to be felt. While diltiazem, the calcium channel blocker, works more quickly, it can drop a patient’s blood pressure. And Mrs. Wu’s was 100 systolic, which was a bit borderline.
I watched the magnesium drip into Mrs. Wu’s iv. No change. Her monitor still pattered as before.
Okay, okay, give her the digoxin, I sighed to Winnie. She smiled consolingly. I left the cardiac room to see other patients. Ten minutes later, she was tugging at my sleeve.
She converted, Winnie announced.
Already? Puzzled, I said, Boy, that digoxin worked fast. Though I knew very well that it usually only slows the rate without making it stable.
No, Winnie insisted. The magnesium. Then she held up a syringe filled with a clear solution.
Digoxin. I haven’t given it yet. It was the magnesium.
Whoa was the most intelligent comment I could produce.
We rushed back together. Mrs. Wu’s monitor beeped contentedly at a regular 84 beats a minute. Her daughter smiled at us and placed a hand over her own heart and said, Much better. Not so fast now.
And how, I thought. Now Mrs. Wu smiled and nodded quickly. A cure.
I’ll be darned, I said to Winnie.
At that moment, in a flash, as my brain shouted, Eureka, I heard the siren call of unscientific thinking. I had scooped even the cardiologists. I would try magnesium on all my A-fibbers from now on.
But immediately I knew that all the successes would stick in my brain, with the humdrum failures thrown away. A little selective memory and next thing you know you’re selling snake oil.
Because there is another side to newly diagnosed atrial fibrillation: a certain percentage of A-fibs will convert to a regular rhythm even if nothing is done. For years it was thought that digoxin not only slowed an aberrant rhythm but stabilized it too. Then researchers found that it was no better than a placebo for stabilizing the heartbeat.
Later that day I hit the medical library. The largest study I could find on magnesium in new-onset atrial fibrillation included a paltry and utterly inconclusive 15 patients. Why hadn’t anyone studied it more definitively? Magnesium at my hospital costs a dollar a dose—no profit there—but I couldn’t blame the drug companies; doctors shouldn’t have to wait for drug companies to tell them which drugs to study. I polled more cardiologists. They’d never heard of it being used for atrial fibrillation. I mentioned it to a colleague in Boston.
Oh, we use it all the time, he informed me with a shrug.
Know of any good studies? I asked.
None, he answered just as casually. But it works.
In short, Boston doctors treat atrial fibrillation, the most common arrhythmia, differently from their New York colleagues. And no one had bothered to determine what works best.
I wondered what more I didn’t know that everyone else did. Atrial fibrillation has been pondered for more than 200 years, intravenous magnesium for most of the century. Eurekas like this shouldn’t happen anymore in cardiac medicine.
Four days later Mr. Pak came in. He had been in trouble since the night before. Wizened and frail, he sat upright and restless in the bed. Every muscle in his neck stood out ropy and quivering from the strain of breathing. The rhythm on the monitor looked awful: rapid—180 beats a minute—and irregular.
The long irregular waves meant potential big trouble. And they had three possible causes. The first, and most dangerous, was ventricular tachycardia. This disordered contraction of the ventricles can cause blood pressure to plummet and frequently leads to ventricular fibrillation—a completely chaotic rhythm more crisply known as sudden death.
The second possibility was a bypass tract, an anomalous ribbon of heart tissue that can conduct some of the chaotic atrial impulses directly into the ventricles without passing through the atrioventricular node. The dicey part is that digoxin and diltiazem—standard treatment for atrial fibrillation—both accelerate conduction through the aberrant pathway, thus allowing atrial chaos to pass unhindered into the ventricles. This, too, can trigger ventricular fibrillation and sudden death.
The third option was plain atrial fibrillation (though a ventricular rate of 180 pretty much disqualified it as plain) with aberrant conduction through Mr. Pak’s aged Purkinje fibers.
While I stared at Mr. Pak’s monitor, Winnie hurriedly inserted two iv lines. Finally, I asked Rob, my second-year resident, Now what do we do?
Not a bad thought, I agreed. A strong electric current delivered to the chest through metal paddles stills all unruly cells and allows the natural pacemaker to regain control. But the shock is painful. To mute the pain, we give sedatives, which can drop blood pressure or suppress breathing.
What else? I asked him.
Call a doctor?
We both laughed. Then I caught Winnie’s eye.
Again? her glance asked.
Two grams this time. Over five minutes.
The magnesium went in. We all stared at the monitor. For us, Mr. Pak’s heart had become that luminous green tracing. Suddenly the rate seemed to ebb: 170 . . . 165, and then poof! The wide sinuous waves disappeared. With the slower rate, the Purkinje fibers had had time to reset and could conduct each beat with their usual blinding speed. In five more minutes, the rate came down to a very gratifying 110. And Mr. Pak, even more gratified, could now breathe. Magnesium had quieted his atrial cacophony to a slow hum. The resident stepped back.
Who was that masked man? he whistled.
Epsom salts, I explained, magnesium sulfate. I guess now we’re going to have to do that study after all—look at 50, 100 cases. Got to answer the key question: Who do you trust, your own eyes?
Or? Rob asked.
Or the real data.