Doc: Incidental MRI findings don’t always warrant worry

Keith Roach
To Your Health

Dear Dr. Roach: A recent MRI of my internal auditory canals revealed an incidental finding of “microangiographic changes” in the area of the pons. I am 65 years old, do not smoke or drink, exercise rather vigorously for an hour almost every day, eat sensibly, am not diabetic and have normal cholesterol and blood pressure without medication. Can you comment?


Dear D.D.: I see this result frequently. The changes seen on your MRI scan are not specific, but they can be associated with damage to blood vessels from many of the conditions you have avoided, especially smoking and high blood pressure. The pons, in the deep brain, is susceptible to damage from high blood pressure. However, some people with no risk factors will have these changes. My practice when I see these is to re-evaluate whether there are any risk factors that could be better managed, consider the use of aspirin if indicated and advise on diet and exercise; then, if all is as it should be, tell my patient not to worry too much, as these findings on MRI are not by any means a guarantee of developing brain disease.

Dear Dr. Roach: In a recent column, you mention that apixaban has a lower risk of intracranial hemorrhage than warfarin. But shouldn’t you note that the absolute risk is extremely low? It drives me crazy to hear the advertisements for NOACs claim a 60 percent reduction (which is true) but not mention that absolute risk is very low. I can understand the use of NOACs in people who have difficulty achieving stable anticoagulation levels or who don’t have their level checked regularly. But lowering the cost of health care should be considered when prescribing an anticoagulant.


Dear P.W.: I share P.W.’s concern for the difference between absolute risk reduction and relative risk reduction, but the concept is one that some people have a hard time with.

In the current case, the use of a new oral anticoagulant (“NOAC” for short), such as apixaban, has a lower risk of major bleeding. How much lower? In an analysis of the major studies, about 5.4 percent of the NOAC group had a major bleed, while 6.2 percent of the warfarin group did.

That can be expressed as an absolute difference of 0.8 percent (6.2 percent minus 5.4 percent), or also as a 13 percent reduction in risk (100 percent minus 5.4 percent/6.2 percent).

While both a 0.8 percent absolute risk difference and 13 percent relative risk difference are correct, I agree with J.W. that the relative risk reduction can overstate the case, especially for low-risk events. (The 60 percent reduction, by the way, comes from a decreased risk for hemorrhagic stroke: a 58 percent relative risk reduction and a 0.7 percent absolute risk reduction.)

The issue with the cost of health care is also complex. In the case of J.F., whose insurance did not cover apixaban, the cost to the patient is much higher for a NOAC.

However, because the total cost to the health care system for a major bleed is so high, NOACs were found to save the system money — between $100 and $500 per person per year, even though the drugs themselves cost more than warfarin (the analysis including the costs of monitoring).

Whose money should the doctor save? The patient’s or the system’s? Is it worth the extra drug costs to the patient to have about a 1 percent per year lower risk of stroke and major bleeding?

These are hard questions, and most of the time the doctor makes them. Some patients want to be involved in making them, and I personally prefer to make these decisions with the patient.

Email questions to