The Nation's Health + [Heart scans]

Heart scan tomfoolery

Heart Scan Blog reader, Steve, sent these interesting questions about his heart scan experience. (I sometimes forget that this blog is called "The Heart Scan Blog" and was originally--several years ago--meant to discuss heart scans. It has evolved to become a much broader conversation.)

The answers are a bit lengthy, so I'll tackle Steve's questions in two parts, the second in another blog post.

Dr. Davis,

I had a heart scan last year. The score was 96. While not a horrible score, it
was a wake up call, and I changed my lifestyle.

I had another scan this year and the heart scan score went up to 105, but the
volume score went down from 141 to 136.

The report I received said this:

'The calcium volume score is less in the current study as compared with the
original or reference study. This is an excellent coronary result and indicates
that there has been a net decrease in coronary plaque burden. The current
prevention program is very effective and should be continued.'

This is all well and good, but I have two questions:

1. Am I really going in the right direction even though the heart scan score
went up 9%?

2. Here are results that make no sense to me:
- Left Main volume went up from 22.4 to 35.6
- LAD went down from 95.2 to 91.3
- LCX volume went down from 23.2 to 0
- RCA volume went up from 0 to 9.3

Why would there be so much variation from year to year, and why would the plaque
move from site to site?


Questions like Steve's come up with some frequency, so I thought it would be worthwhile to discuss in a blog post.

First of all, the conventional heart scan score, or "calcium score" or "Agatston score" (after Dr. Arthur Agatston, developer of the simple algorithm for calcium scoring, as well as South Beach Diet fame), is the product of the area of the plaque in a single CT "slice" image
multiplied by a density coefficient, i.e., a number ranging from 1 to 4 that grades the x-ray density of the plaque. (1 is least dense; 4 is most dense.) A density coefficient of 1 therefore signifies some calcium within plaque, with higher density coefficients signifying increasing calcium content and density. Incidentally, "soft" plaque, i.e., non-calcified, would fall in the less than 1 range, even the negative range (fatty tissue within plaque).

The volume, or "volumetric," score is the brainchild of Drs. Paulo Raggi and Traci Callister, who expressed concern that, if we cause plaque to shrink in volume, the density coefficient used to calculate the calcium score would increase (since they believed that calcium could not be reduced, contrary to our Track Your Plaque experience, thereby leading to misleading results. They therefore developed an algorithm that did not rely on density coefficients, but used the same two-dimensional area obtained in the standard heart scan score, but replaced the density coefficient with a (mathematically interpolated) vertical axis (z-axis) measure of plaque "height." This 3-dimensional volumetric value therefore provided a method to generate a measure of calcium volume. In their original publication, the volume score proved more reproducible than the standard calcium score. This way, any reduction in plaque volume would not be influenced by the misleading effects of calcium density, but reflect a real reduction in volume.

Callister and Raggi's study also highlighted that calcium scoring in any form is subject to variability. Back in 1998 (when their study was published), there was a bit more variation than today due to the image acquisition methods used. But, even today, there is about 9% variation in scoring even if performed repeatedly (with less percentage variation the higher the score).

Unfortunately, volume scoring never caught on and the calcium score has been the most commonly used value by most heart scan centers and in most clinical studies. And, in all practicality, the two values nearly always track together: When calcium score increases, volume score increases in tandem; when calcium score decreases, volume score decreases in tandem.

Steve is therefore an exception to the general observation that calcium score and volume score travel together. Steve's calcium score increased, while his volume score decreased. From the above discussion, you can surmise a few things about Steve's experience:"

1) In all likelihood, the changes in both calcium score and volume score could simply be due to variability, i.e., variation in the placement of his body on the scan table, variation in position of the heart, variation in data acquisition, etc. There is a high likelihood that neither value changed; both are essentially unchanged.

2) If the changes are not due to scan variability, but are real, then it could be that the calcified plaque is reduced in volume but increased in density. If true, this is probably still a favorable phenomenon, since plaque volume is a powerful predictor of coronary "events" and an increase in plaque density is likely a benign phenomenon. It would also raise questions about the adequacy of vitamin D and vitamin K2 status, both major control factors over calcium deposition and metabolism.

So, in all likelihood, Steve's apparent discrepant results are modest good news, especially since calcium scores can ordinarily be expected to increase at the rate of 30% per year if no action is taken. Experiencing no change in score, calcium or volumetric, carries a very excellent prognosis, with risk for heart attack approaching zero. (I'm impressed that Steve accomplished this on his own, something the majority of my colleagues haven't the least bit of interest doing.)

Part 2 of Steve's question will be tackled in a separate post.