Linear risk scores needed in HCM

Cardiovascular News

Linear risk scores needed in HCM

30 May 2007

MedWire News: Experts in hypertrophic cardiomyopathy (HCM) are conducting population-based studies to develop improved risk stratification methods for management of the disease.
Speaking at the UK Cardiomyopathy Association Conference, held in London, UK, Perry Elliott (The Heart Hospital, University College London) noted that current risk stratification relies on relatively crude ways of assessing risk, based on a small number of simple, non-linear risk factors.

“We need to start modeling risk in a slightly more sophisticated way,” he said, emphasizing that there is a continuum of risk, with complex interactions between risk factors and variations over time.

Elliott explained that determining risk in HCM is particularly challenging because, although HCM affects 1 in 500 of the general population, it remains relatively uncommon in everyday clinical practice.

Another problem is that although sudden unexplained death (SUD) is the best known complication of HCM, it is actually quite rare, now occurring in less than 1% of the HCM population per year. Finally, there is a great degree of heterogeneity in symptoms, even within families with the same underlying genetic defect.

Elliott outlined the main risk factors on which risk stratification in HCM is currently based, namely unexplained syncope, a family history of sudden unexplained death (SUD), the severity of left ventricular hypertrophy (LVH), presence of an abnormal BP response on exercise, and the presence of arrhythmia on 24-hour ECG monitoring.

The problem with each of these risk factors is that there caveats to their interpretation.

For example, Elliott noted, a family history of SUD is emotively the most powerful risk factor. Yet if the logic of this is followed too stringently, it might lead to implanting defibrillators in multiple family members even though, statistically, a family history of SUD is no more or less predictive than other risk factors. There are some families in whom such a history is very bad, and so taking this into account may be enough. But for other families – perhaps where a more distant relative has been affected and the cause remains uncertain – it is not so powerful.

Similarly, caveats apply to all the other risk factors. Fainting does not necessarily carry an adverse prognostic significance, and measures of LVH can only be semi-quantitative - while the threshold of ≥30 mmHg in heart muscle thickness is considered a risk factor, there is always a gradation of risk around this measure.

Elliott noted that left ventricular outflow tract obstruction (LVTOTO), a defining feature of HCM, has recently been established as a risk factor.

Unlike most of the other risk factors it is modifiable, since the outflow tract obstruction can be relieved by surgery or alcohol ablation. But owing to the risk associated with any type of surgery, the decision to operate has to be taken on the basis of gradient severity and in conjunction with other risk factors.

In practice, clinicians have to look at the global risk burden, since the more severe the disease, the more likely patient is to have multiple risk factors.

Speaking to MedWire News, Elliott said that developments in two key areas will lead to improved risk stratification that will help to aid decision-making in borderline cases of HCM.
The first involves doing a much larger study, “looking at other variables that tell you about the severity of the disease,” he said.

“When you’re trying to construct risk scores it’s always much better to have thousands of patients because then the scores have more predictive power,” he explained. “So for cholesterol and blood pressure there are now some very well trusted risk algorithms.

“By contrast, historically we’ve only had relatively small numbers of HCM patients to study.”
But over the past decade this has changed quite rapidly and most are centers now following several thousand patients, he said.

“I think if we study 10,000 patients this should be able to generate more meaningful risk scores.”

Elliott noted that such studies are now underway at The Heart Hospital.

“The second thing is that we’re always looking out for new techniques that tell you about the vulnerability of the heart to rhythm disturbance,” he added.

These include non-invasive techniques, such as the T-wave alternans test, that in conjunction with clinical markers will help to focus on those individuals who are at the greatest risk.

“So it’s a case of taking a more population-based approach as well as technical advances,” Elliott summarized.

Link: UK Cardiomyopathy Association Annual Conference; London, UK: 25 May 2007