The Brain's Recovery from Stroke


by Guy McKhann, M.D.

March 6, 2013

One of the mysteries in clinical neurology is how the brain recovers—or doesn’t recover—from injury. Two people may have seemingly similar strokes involving the motor area of the right cerebral hemisphere, resulting in the inability to use the left arm and hand. Over the next three months, Subject A makes a slow but consistent recovery, and within six months is using his right hand almost normally. The other, Subject B, has made very little recovery, and is still quite disabled at six months.

There are several possibilities for why “A” recovered. Perhaps the damage to the motor area was not as great, and the part of the brain that controls the arm and hand was only lightly damaged. Another possibility is that some other area of the brain takes over at least some of the functions of the motor area. Perhaps the motor area in the other side of the brain takes over. Or perhaps some area on the same side of the brain, one we don’t normally associate with motor function, steps up.

Dr. Steven Zeiler and his colleagues at Johns Hopkins have taken a step toward solving this mystery, as outlined in the article “Stroke Damage in Mice Overcome by Training That ‘Rewires’ Brain Centers.” The steps in the experiment are a little complicated, so bear with me.

The researchers first trained mice to use one paw to grasp food pellets through a narrow slit. That training took 9-10 sessions. The next step was to create an artificial stroke by damaging the motor area controlling the trained paw. As expected, the mice lost their ability to get the food pellets.

They next gave the mice “physical therapy,” retraining them to get the pellets starting two days after the “stroke.” That worked, and after a similar period of training the mice were back to where they had started, just like Subject A mentioned above.

In the human, one can use imaging to try to see what part of the brain is active, and get some clue how the brain might have reorganized. Such studies have been done in people with strokes, but not with consistent results. In the mouse one can look at the brain and see what might be going on. Zeiler first found that the damaged neurons in the motor area had not recovered, ruling out that region as the mechanism. However, the medial premotor cortex, an area in front of the usual motor area, had become active. In healthy mice, damaging this medial premotor cortex did not cause any paralysis—normally this is not a motor-control area; it becomes one as a response to injury. To put the icing on the cake, the researchers damaged the medial premotor cortex in the mice with a damaged motor area, and saw the original motor deficit return.

That’s only part of the story. What’s going on in the medial premotor cortex? The medial premotor cortex contains inhibitory neurons that keep other neurons inactive. After the injury to the motor cortex, these inhibitory neurons become quiet, allowing the other neurons to be more active and to take over motor functions.

What about Subject B? Why didn’t he recover? Perhaps his lesion was larger, knocking out both the primary motor area and the equivalent of the premotor area. Or maybe the inhibitory neurons in the premotor area didn’t get turned off. Maybe he didn’t receive adequate physical therapy. Addressing the timing, dose, and use of medicines are the next steps.

Attempts to go from the results of therapy in a mouse model of stroke to an application in the human have a dismal record. However, most previous studies have been different: a large stroke is induced in the mouse and some agents are given to the mouse to try to make the stroke smaller. Many agents do that in the mouse, but not a single one has had any positive effect in the human. The studies I report here are different. They are getting at the cellular mechanisms of recovery.

I must acknowledge that Dr. Zeiler and his colleagues are also colleagues of mine at Hopkins. But I make no apology—they did a good job!