As a radiologist, this is my coffee-mug aphorism for the epistemology of medical imaging. It reminds me that the picture or the test result is not the patient or even the disease.

Here is one of the hierarchies I use when thinking about coronary disease. It's a rough ordering from top to bottom of how far away you are from "the thing itself" when looking for evidence of a previous myocardial infarct.

From useless to definitive:

Framingham risk score - modern blood tests ( HDL , LDL-P, oxidized LDL, etc.)

Calcium score (Heart scan for coronary calcium)

Intravascular Coronary Ultrasound

Abnormal coronary angiogram

Abnormal coronary CTA

EKG evidence of infarct

Stress SPECT Nuclear Medicine perfusion

Cardiac perfusion MRI with dobutamine

Late Gadolinium enhancement cardiac MRI (LGE)

Death with autopsy evidence of MI.

I have ordered these roughly according to how definitive they are in establishing myocardial injury - whether you have actually had a heart attack sometime in the past (not acutely, as that can often be shown with serum enzymes). 

Think of Lindeberg's studies of the Kitavans. He used EKGs (electrocardiograms) to look for evidence of prior MI (myocardial infarct or heart attack). Comparing to the known prevalence of EKG abnormalities in western subjects, he concluded that there was no evidence of heart disease in Kitava.

The Kitavans are fascinating because they have high carb consumption and appear to be much healthier than those on the SAD - the standard american diet. They also have some "paradoxical" lipid markers that (rightfully) call into question the very idea that the lipid markers that cardiologists love to measure mean anything outside the context of a particular diet. In my opinion, they help cast doubt on the very idea that either good or bad blood lipids "cause" anything. Rather, they are just markers for the diet - the diet effects the actual changes.

EKG is not that sensitive and can be often be false positive.

What if we wanted the most definitive evidence that we could get, short of sacrificing our poor subjects in the interest of science and having a pathologist look at their hearts?

How about invasive coronary angiograms? Unfortunately, a conventional angiogram only shows stenosis or narrowing of the coronary arteries and does not show a heart attack unless there is gross wall motion abnormality, and that is an indirect sign anyway.

How about a nuclear medicine stress test? - this 20-year-old procedure is the most traditional non-invasive way to look for heart disease - unfortunately it also requires a fairly large infarcted (damaged) area to reliably detect if there has ever been a heart attack. If looking just for physiologic evidence of stenosis, it requires a 75% or greater arterial narrowing to reliably detect disease. It is not very sensitive.

It would surprise even a lot of my fellow physicians to know that the most sensitive way to tell if there is any coronary atherosclerosis whatsoever is not nuclear medicine or an invasive coronary angiogram - it is CCTA or coronary CTA.

CCTA is not "calcium scoring" (I avoid calling calcium scoring a "heart scan" as at my center there are a half-dozen procedures that could be called that). CCTA is a CT scan done with a 64- slice multi-detector CT scanner that rotates fast enough to image the coronary arteries. Although the spatial resolution is not as high as conventional catheter angiography, CCTA can directly image early "soft" plaque that is totally invisible on conventional angiography or on simple calcium scoring. The only test that is better - the absolute gold standard for detection of coronary plaque in the living - is called intravascular ultrasound or IVUS. But IVUS requires invasive insertion of an ultrasound probe into the coronary arteries in a procedure similar to conventional angiography. It is mainly a research tool. Not so good for screening asymptomatic subjects.

So to detect any atherosclerosis of the coronary arteries in a population, the best practical test would be CCTA.

But what if we wanted to know, with as great a degree of certainty as possible without waiting for the death of our subjects, whether they had ever had a heart attack or any damage to the heart muscle? After all, the important thing is if you have an actual heart attack or die - that is "the thing itself" in this case.

Then the choice would be cardiac MRI with late gadolinium enhancement or LGE.

Here is how LGE works. When heart muscle is damaged, there are pathologic changes in the tissue that we characterize, in our highly precise way as medical scientists, as "scar tissue". This scar tissue will have a pattern of blood pooling or stasis, where, unlike normal heart muscle, the blood seeps in slowly, and then does not wash out as fast with time.

LGE exploits this phenomenon. We give an agent intravenously that shows up bright on MRI images called gadolinium or Gd (It's actually a compound or chelate as pure Gadolinium is poisonous).

Firstly, we can first get heart images during the early part of the Gd administration and use this to look for ischemia or temporary insufficiency of blood flow with stress. (This is a good way to tell if there is a narrowing of one of the arteries supplying the heart - the coronary arteries - and is similar to the mechanism used in a nuclear medicine stress test - only much more sensitive.) Then we get delayed images. On these delayed images, normal heart muscle will return to its non- bright appearance, but the scar or damaged heart muscle will stay bright. The important thing to know is that we can detect such areas with much higher resolution than nuclear medicine techniques, and we can see non- transmural (less severe -not all the way through the heart wall) infarcts and infarcts so small they have no effect on wall motion.

This LGE technique has been validated with animal and human pathologic studies. It truly is the best way other than autopsy to tell if there has been permanent damage to the heart muscle.

Mainstream thinking still maintains that lots of sustained aerobic exercise is good for cardiovascular health - to the point where everyone uses the word "cardio" as a synonym for long sessions of aerobic effort. Like the diet-heart and lipid hypotheses, this idea is actually only about 40 years old. In the first half of the 20th century, it would have been thought as absurd as the idea that butter clogs your arteries.

So let's say you want to see just how protective "cardio" is against the number one killer of Americans -coronary heart disease. Instead of just looking for atherosclerosis, though, we want "the thing itself" of actual myocardial damage. Who knows, maybe runners get more plaque but have fewer plaque ruptures or a less thrombogenic blood profile?

So in the Breuckmann study, they recruited 102 active marathon runners. To be a marathon runner (and perhaps to maximize their power to show how healthy "cardio" is) required at least 5 marathons in the past 3 years. Many had run dozens or more in their lifetime. Anyone with a known history of heart disease or diabetes was excluded. The average age was 57 with age 50-72. The median number of marathons was 20. Weekly mileage was 35 (55km). Mean work was 4700 METs per week.

There were 102 totally asymptomatic age-matched controls, also with no history of diabetes, who had no significant history of vigorous exercise.