Study rates heart scan techniques

“Magnetic resonance imaging (MRI) scans should be used to assess patients with suspected heart disease, rather than standard checks,” reports BBC News today.

This story is based on a large, well-designed study comparing a new technique called cardiovascular magnetic resonance (CMR) imaging against the commonly used alternative test, single-photon emission computed tomography (SPECT). The researchers tested the scan’s ability to diagnose significant coronary heart disease, also looking at how they compared against standard angiography, where dye is introduced into the blood vessels to highlight any blockage or narrowing. The study found that CMR performed as well or better than SPECT on a number of key diagnostic measures. Together with the fact that CMR does not expose patients to ionising radiation, the researchers say the results show that CMR should be more widely adopted.

However, CMR will not be suitable for all patients, including some with medical implants and those who could experience claustrophobia inside the scanner. Further research will also be needed to demonstrate that improved diagnosis through techniques such as CMR actually improves patient outcomes. That said, these results do suggest the technique has merit.

Where did the story come from?

The study was carried out by researchers from the University of Leeds and was funded by the British Heart Foundation. The study was published in the peer-reviewed journal The Lancet.

The coverage by the BBC on this story was accurate, and featured comments from independent experts and explanations of the need for confirmation in other centres, population groups and for an assessment of cost.

What kind of research was this?

This was a randomised trial comparing how well two types of non-invasive scanning techniques could diagnose coronary heart disease: a newer test called cardiovascular magnetic resonance (CMR) versus the widely used technique of single-photon emission computed tomography (SPECT).

CMR uses magnetic fields and radio waves to produce images of the inside of the body. It does not use ionising radiation. SPECT requires a radiation-emitting chemical (a radioisotope) to be injected into the bloodstream. The radioactive emissions are detected and used to create an image. This technique exposes patients to small amounts of ionising radiation. Both methods provide functional tests for angina, in that heart function or perfusion is detected after an injection of a chemical that stresses the heart.

Both of these techniques were compared with another imaging technique called X-ray coronary angiography, which acted as the reference standard. In this way, the researchers could directly compare the results of the two scans in a single patient and then turn to X-ray scanning to confirm which was most accurate.

In X-ray coronary angiography, a contrast agent is introduced into the coronary artery and X-ray images are taken. Again, the patient is exposed to ionising radiation and in addition the technique used to introduce the contrast agent is invasive. This is an anatomical test that shows where any narrowed arteries might be.

Patients with suspected disease were randomised into two groups, receiving either CMR before SPECT (prior to confirmation with angiography) or SPECT before CMR (prior to confirmation with angiography). Offering CMR and SPECT in a random order reduces the chances of the results being influenced by bias: for example, the process of performing a scan might potentially influence the results seen in any subsequent scan, and therefore always performing a particular type of scan first may skew the results.

This study design provides an appropriate way to test the diagnostic accuracy of a new technique, as it compares CMR with both the widely used SPECT and the ‘gold standard’ X-ray angiography.

What did the research involve?

The trial enrolled 752 patients with angina (chest pain due to lack of blood to the heart) that needed further investigation and at least one other risk factor for coronary heart disease. Patients were excluded if they had previously undergone heart bypass surgery.

All patients were scheduled to receive all three tests. The patients’ hearts were imaged using CMR, SPECT and X-ray angiography and the images analysed by people experienced in interpreting the results, to make a diagnosis. The order of CMR and SPECT imaging was randomised and those reading the results of the tests were unaware of the results of previous testing, except at the end when the results could be revealed to the treating clinician to determine treatment.

What were the basic results?

The overall results suggested that 39% of the recruited patients had significant coronary heart disease identified using X-ray angiography.

The researchers found the following for CMR:

• A sensitivity of 86.5% [95% confidence interval (CI) 81.8% to 90.1%]. This means that 86.5% of patients with the disease identified using X-ray angiography had a positive result on CMR. Therefore, these people were correctly identified as having coronary heart disease.
• A specificity of 83.4% [95% CI 79.5% to 86.7%]. This means that 83.4% of patients without coronary heart disease during X-ray angiography correctly received a negative result using CMR. These people were correctly identified as not having coronary heart disease.
• A positive predictive value of 77.2% [95% CI 72.1% to 81.6%]. This means that 77.2% of patients who were diagnosed as having coronary heart disease by CMR did in fact have coronary heart disease. But, conversely, 22.8% of patients would be incorrectly identified.
• A negative predictive value of 90.5% [95% CI 87.1% to 93.0%]. This means that 90.5% of patients who had a negative result by CMR did not have coronary heart disease. But, conversely 9.5% of patients would be incorrectly reassured.

The sensitivity and negative predictive value for CMR was significantly better than those for the widely used SPECT technique. The specificity and positive predictive value of the two techniques were similar.

How did the researchers interpret the results?

The researchers conclude that this trial has shown “CMR’s high diagnostic accuracy in coronary heart disease and CMR’s superiority over SPECT”. They say it should be adopted more widely for the investigation of coronary heart disease.

Conclusion

This research has demonstrated the diagnostic accuracy of CMR in diagnosing coronary heart disease. CMR also has the advantage that it is a non-invasive technique that does not expose patients to ionising radiation. However, CMR will not be suitable for all patients, as due to the high magnetic fields involved, patients with some medical implants will not be able to use it. Due to the confined nature of many scanners it is also not suitable for patients who suffer from claustrophobia (although this is also the case with many SPECT scanners).

Some points to note:

• The tests were done in a group of patients at relatively high risk of coronary heart disease, with almost 40% having the disease. The accuracy of the test in a community sample of lower-risk patients will need to be tested.
• The research was conducted by skilled, experienced operators, meaning that its accuracy might not be the same in units where fewer procedures are performed.
• X-ray angiography itself is not itself a perfect test and therefore might not have been ideal as a reference standard.

Further research will be needed to test if improved diagnosis, using techniques such as CMR, actually improves patient outcomes. Cost, cost-effectiveness and the availability of scanners will also need to be evaluated.