Hot on the heels of that episode of Catalyst, research into the effects of cholesterol in humans is fast becoming a lot more complicated than just ‘good’ and ‘bad’.
While these umbrella terms are used to broadly assess cardiovascular health, they in fact comprise hundreds of chemically distinct species of fats (lipids), and do not provide researchers with any clear cut information about what these individual chemicals do in the circulation.
VBU researcher Joe Moxon, studies peripheral arterial disease (PAD), a group of narrowing and weakening (aneurysmal) conditions affecting the arteries outside the heart. While a definitive link has been proven between atherosclerosis and cholesterol, no such link has been definitively demonstrated with abdominal aortic aneurysm (AAA), a disease which causes the body’s main blood vessel to weaken, dilate and eventually rupture.
Dr Moxon and colleagues recently conducted the first in-depth study to determine whether cholesterol components were associated with (AAA) presence, the results of which are published in Circulation: Cardiovascular Genetics journal. In conjunction with researchers at the BakerIDI, Dr Moxon used mass spectrometry and statistical modelling to compare the chemical composition of cholesterol carried in the blood of patients with atherothrombosis and abdominal aortic aneurysm.
“We used data and blood samples from approximately 400 patients – half with lower limb atherosclerosis and half with AAA. These groups were deliberately chosen since their conditions have similar risk factors, but the mechanisms driving the complaints appear distinct and treatments for one disease is not necessarily appropriate for the other. It’s therefore important to be able to distinguish the patients with AAAs from those with similar non-aneurysmal complaints,” Dr Moxon said.
“Currently, the only definitive way to diagnose AAA is using imaging techniques such as ultrasound and CT, followed by repeated imaging assessments to monitor AAA progression. This may involve exposing patients to radiation, and incurs travel and infrastructure costs.”
Dr Moxon’s analysis measured the concentrations of individual lipid species in blood samples taken from the recruited participants. The results were subtle – of over 330 assessed lipids, only ten were present at significantly higher blood concentrations in the aneurysmal patients. However, examination of the chemical structure of these lipids revealed that all ten contained linoleic acid, a common dietary acid.
“The data showed that patients with elevated concentrations of these lipids were 60 to 90% more likely to have an aneurysm than atherosclerosis,” Dr Moxon said. “We did some further bioinformatics with BakerIDI and basically said, ‘is this clinically useful – can these lipids be used to identify patients who have AAA?’
To assess the diagnostic potential of these lipids for AAA Dr Moxon and colleagues used statistical modelling to build two predictive models.
“Basically, we gave the computer a panel of patient specific features and told it to build us the best model it could to distinguish the AAA patients from those with atherosclerosis. For the first model, the panel comprised typical clinical data such as age, sex, diabetes etc. The second panel provided all of these clinical features but also included all assessed blood-borne lipids.”
“The results were encouraging. In the second model, we started to see that the lipids we identified were being incorporated in preference to more traditional measures of dyslipidaemia such as total cholesterol and triglyceride concentration. This suggests that our lipid species may have greater discriminant ability than commonly used parameters.”
Side-by-side comparisons of the performance of the two models provided further encouraging data.
“Statistical comparisons suggested that the second model had a more powerful diagnostic ability than the first, suggesting that the incorporation of lipid species significantly improved our ability to distinguish the AAA patients from those with atherosclerosis,” Dr Moxon said.
While this model seems promising, Dr Moxon stresses its diagnostic ability is still fairly low. “However, the novel features we identified may be useful in conjunction with other biomarkers to build a more significant model, although further work is required to assess this,” he said.
The next phase of Dr Moxon’s work involves more modelling with BakerIDI, this time to see whether having linoleic acid-containing lipids in the system is ‘good’ or ‘bad’.
“We know that this can be modified through dietary assistance, such as increasing the intake of oil supplements – safflower oil in particular,” Dr Moxon said. “But we wanted to determine if they were a good or bad thing.”
Secondary analyses focussed on a sub-set of 265 patients with detailed follow-up history leading to an unexpected but welcome discovery, which will become an ongoing side project in 2014.
“We thought about what a relevant end point would be. Both AAA and atherosclerosis patients are at increased risk of heart attack. We therefore decided to investigate whether any of the lipids we identified were associated with heart attack risk” said Dr Moxon.
“The lipids which showed an association with AAA did not appear to affect heart attack risk, however we were surprised to identify a separate lipid that appears strongly predictive of heart attacks in these patients,” he said.
“Patients with lower concentrations of this lipid were at a higher risk of heart attack than those with concentrations above median – five times more likely to have a heart attack, in fact.”
While Dr Moxon is hopeful this discovery will have prognostic potential, study of a larger group of patients is required.
“This study was done with a small number of patients, so we need to study more patients from around Australia and also conduct pre-clinical analyses to determine the mechanisms underpinning this relationship,” Dr Moxon said.