Ion Mobility: Question and Answer Session with the Experts

Krauss, Ron M., MD
Senior Scientist and Director of Atherosclerosis Research
Children’s Hospital and Research Center Oakland
Oakland, CA
Adjunct Professor of Medicine
University of California San Francisco
Also by this Author 

Superko, H. Robert, MD
Chief Medical Officer
Celera Corporation
Alameda, CA
Also by this Author 

Question:  What is the relationship between lipid regions separated by Ion Mobility and those on traditional gradient gel electrophoresis?

Robert Superko, MD:  Well, let me ask Ron to comment because he’s the individual who developed both of those techniques.

Ronald Krauss, MD:  Well, as everyone knows, the gradient gel method was developed to analyze LDL and HDL subfractions in a way that separated these individual subspecies that we defined some years ago very well and give us an estimate of their distribution.  So when we developed the Ion Mobility method, we used those reference ranges for the particle size for those LDL and HDL subclasses to calibrate the size distribution that we were seeing by Ion Mobility.  So what we’re getting with Ion Mobility is a direct measurement of those same subfractions within LDL and HDL, but it’s actually an improved analytic procedure because it measures the absolute concentrations rather than just the relative distribution.

We still get the A/B phenotyping definition.  But the size units are not the same actually.  The concentrations are nanomoles per liter and the particle size measurements are actually, if anything, more accurate, more properly measured than gradient gel because it’s a procedure that’s based on physical principles that directly measures particle size and concentration.  So all of these are advances beyond the gradient gel and, in addition, we get measurements for VLDL and IDL, which were not traditionally available using the gradient gel method.  So for those interested in the LDL in particular and the HDL as well, it should be a very smooth transition to the Ion Mobility.

Dr. Superko:  You know, Ron, I wonder, I think there might be people on the call who aren’t all that familiar with the past history of lipoprotein subclasses; and if I just might comment that throughout the past 20 plus years, studies like the Boston Area Heart study, the Stanford Five City Project, the Harvard Physicians Health Study , and the Quebec Cardiovascular Study have all shown a powerful relationship of these small particles to cardiovascular events.  And I remember all along the Holy Grail has been to have a quantitative test compared to size and percent distribution by GGE.  And that essentially is now what is provided with the Ion Mobility method.  Is that a fair statement?

Dr. Krauss:  That’s right.  Yes, this is a method that is unique in that it directly measures the concentrations of these particles in a way that is really an improvement on all of our previous methods.

Question:   How does this method compare to the NMR or the VAP test?

Dr. Krauss:  I can comment on that as well.  There really is not anything yet in the literature that’s a direct comparison of the particle concentrations.  There is a report showing that the particle phenotypes, the LDL subclass phenotypes, match up reasonably well and we have some data that we’ve developed that’s not yet published that shows the correspondence with NMR in particular to be reasonably good.  It’s not going to be perfect.  There are differences in the procedures involved.  And as I had mentioned, the Ion Mobility measures particles directly.  The NMR goes through a calculation.  These measurements are both quite good, but they’re not identical.

The VAP uses a completely different system.  It measures the cholesterol in various fractions separated in an ultracentrifuge procedure.  And the comparisons that we’ve seen so far have not been quite as strong, but they do tend to go in the same directions.  They’re just not quantitatively as strong as with NMR.

Dr. Superko:  This is Robert again.  One of the interesting things I think is that the vast history that has used gradient gel electrophoresis and another one of Ron’s techniques, analytic ultracentrifugation, are not only in these studies like the Quebec Cardiovascular Study and the Stanford Five City Project but are also utilized in the arteriographic regression studies.  And if you want to compare the new Ion Mobility’s ability to determine a pattern A or pattern B with those other methods such as NMR and VAP, then John Sninsky has a publication that compares these different laboratory methods..  And there is, if I recall, about a 95% correlation between Ion Mobility and gradient gel.  So in terms of moving from one to the other for phenotyping pattern A and pattern B, it looks like Ion Mobility is the best of the group.

Question:  Is the ATP Report due anytime soon?  If so, will it discuss the clinical use of lipoprotein subclass testing?

Dr. Krauss:  So there has been a revisiting of the ATP Guideline approach.  And the current version of ATP IV is going to be released as a systematic review of the literature addressing the evidence to support the efficacy and safety of targeting LDL cholesterol and non-HDL cholesterol goals.  It’s going to be very limited to those questions and there is going to be a second extension of that at some later point where the AHA and the ACC are going to come in with some more extended guidelines.  But the initial phase of the process is expected to be released in November, the systematic review.  And based on everything that I’ve heard and seen, it’s not going to address subclass testing at all.  It is going to touch on non-HDL cholesterol and to some extent Apo B, but it’s very limited in terms of its scope and it’s something we just have to wait for other groups—  and other groups have stepped in with guidelines related to subclass testing, but it’s not going to come anytime soon from the ATP process.

Queston:  Has Ion Mobility testing been shown to reduce mortality by changing management of dyslipidemia?

Dr. Krauss:  Well the IM is still a new procedure.  We’ve been working on it in the lab for over ten years; but in terms of its clinical application, it’s really just getting started.  And we’ve had some publications, one in particular that has addressed the cardiovascular risk predictive value for IM and, not surprisingly, it’s shown that the smaller LDL particles and the overall phenotype associated with smaller LDL, higher triglyceride, lower levels of large HDL, all of those are associated with cardiovascular events; specifically myocardial infarction and death from myocardial infarction.  So that’s been shown with Ion Mobility.

There hasn’t been yet a report of its use in clinical trials aimed at testing, for example, statin effects on CVD.  Those analyses are under way.  There are some data coming forth on that, but there certainly is evidence that the Ion Mobility is very consistent with previous reports using other methods that show a strong relationship of LDL particles with cardiovascular outcomes.  And there hasn’t been enough power to look at total mortality, but cardiovascular events are definitely predicted by this profile with small LDL particles.

Dr. Superko:  And Dr Krauss has done a lot of work with diet changes in subclasses and medications.  And I think it’s fair to say that the studies that you’ve done with Ion Mobility are quite consistent with the results you receive from gradient gel studies.

Dr. Krauss:  That’s right, Robert.  We’ve published now several reports on diet interventions, weight loss interventions, interventions with statin, with a CETP inhibitor, and with a PPAR active agent.  And all these results are consistent with earlier evidence.  We found some new things as well, but the major changes are what one would expect.  For example, statin treatment selectively reduces levels of larger and medium-sized LDL and that does contribute to the CVD benefit.  But we and others have shown that the smaller particles show less reduction with statin and Ion Mobility demonstrated that very nicely.

Dr. Superko:  So I think that’s why it’s important that the information derived from older gradient gel studies can be applied using the information obtained from the newer and quantitative IM.  There are 11 arteriographic studies that have looked at these subclasses and they all consistently show that individuals with this small pattern B have more rapid progression arteriographically when in the control group.

Paul Williams has a very nice paper just published in 2013 from the HATS trial — that was one of Greg Brown’s studies in Seattle — that used simvastatin and nicotinic acid.  And Dr Williams sort of quantitated subclasses by multiplying the percent distribution by LDL cholesterol.  And in that study it turned out that the small LDL IIIB was an independent predictor not only of arteriographic change but clinical events as well.  So I think the clinical event and arteriographic studies are quite consistent and allows us to move into the new realm of Ion Mobility and real quantitation.

Question:  How does a subfraction profile correspond to genetic risk?

Dr. Krauss:  Well there’s some research going on in this area very actively but we don’t yet use genetic testing for lipoprotein profiles at least in clinical practice or other genetic tests that are out there offered by Quest Diagnostics and others.  But the genetic association with lipoprotein particles, particularly LDL, has taught us some very interesting lessons about what we can and cannot achieve with drug treatment.

So, for example, there is a very recently discovered gene sortilin that is very strongly associated with LDL cholesterol in multiple large studies.  And using Ion Mobility, we’ve been able to show that the effect of the sortilin gene is specific to the small and very small LDL particles.  So it does contribute to risk for heart disease.  And because statins have minimal effect on these smaller particles, this genetic information may help us understand why statins may not be effective particularly in individuals who have the variant of this trait.  And it’s actually the common variant that causes the higher risk.  But that’s really right now a research level question and there certainly are other important genetic factors that contribute to family history.

When we look at families and see premature heart disease, I think those of us who have been doing this testing for a while, know that there’s a very high prevalence of the small LDL profile, the pattern B profile that does have a genetic basis.  So genetics are very important in all of this but the information we get from the testing itself, the lipoprotein testing is really sufficient.  And at this point the genetic tests per se, the DNA analyses, are probably not going to add very much for the foreseeable future, although this research is moving very rapidly.  We may eventually have a panel of genetic tests that can predict risk that relates to lipoproteins and can be applied, for example, to children to help identify early susceptibility and start intervening earlier on.  But this is all really for the future.

Dr. Superko:  Ron, there was an interesting paper series that Paul Williams did about HDL subclasses in family members and whether or not each family member exercised or not.  And I think what he showed was a very powerful similarity between those family members so that the phenotypic genetic inheritance might be different than the DNA type of inheritance.

Dr. Krauss:  Well, it does reflect the underlying DNA effects, but, you’re right, the study I think you may be referring to was done in twins.

Dr. Superko:  Yes, Twins.

Dr. Krauss:  And that’s the best kind of family data to get is when you have twins that you can look at, for example, twins who are exercising and twins who are not exercising.  And it turns out that the genetic influence on HDL and also smaller LDL is quite strong in those twin studies.  And so, again, the measurement of the lipoproteins in these families, as you say, does tell us about the genetic underpinnings but we are far from knowing what all the specific genes are that are responsible for that.

Question:  How do we compare previous lipid studies with Ion Mobility?  Are the numbers the same for total cholesterol, HDL, triglycerides, LDL, VLDL?

Dr. Superko:  Well, if I understand the question correctly, the numbers for total cholesterol and LDL cholesterol and HDL cholesterol would be the same.  If they’re asking about these subfractions, the quantitative amount of subfractions, then that’s what’s new about the Ion Mobility; and we don’t have those measurements except perhaps with ANUC (analytic ultracentrifugation) in some of the older studies.  Did I interpret that question correctly?

Dr. Krauss:  I think you probably did, Robert.  I mean, I think the important thing is that the Ion Mobility is measuring particles, the total particle.  And the cholesterol and triglyceride are in the particle, but the standard lipid measurements are really focused on lipid composition, and they’re correlated with lipoprotein particles.  The VLDL concentrations, for example, are correlated with triglycerides since that’s where most of the triglyceride is transported.  But after that the correlations are pretty poor with cholesterol because, as you know, we can have a normal LDL cholesterol and can still have very abnormal lipoprotein particles.

Dr. Krauss:  There is a well established concept that I think has now become widely appreciated by clinicians largely through what NMR has been developing as their message it is looking at total LDL particles which they feel is an adequate measure of risk.  We think it’s actually more important to measure the individual subfractions.  But the point is that even with total LDL particles, there are people who are at risk who have normal or even low LDL cholesterol but have high levels of LDL particles and that’s what’s so important about these particle measurements.

Dr. Superko:  So perhaps one of the things is that the LDL number that Ion Mobility gives you is the true LDL region or the combination of the LDL subfractions.  And I believe the NMR particle number also includes IDL.

Dr. Krauss:  Yes, they do include IDL.  And I think one of the reasons that they probably do that is that the LDL cholesterol, the standard LDL cholesterol measurement that you get on all the standard reports is actually the sum of true LDL plus IDL.  So the NMR, I think, is using that definition whereas technically the LDL measured separately is going to be a somewhat lower number because the IDL is really a completely different type of particle.

Question:  Will you clarify the current debate on whether niacin is effective in treating small dense LDL versus the use of statins.

Dr. Superko:  Ron, I’ll step in quick and then ask for your comment.  I don’t think there’s a controversy about niacin’s effect on small LDL.  That’s been looked at in a number of studies, some of which Ron and I have done.  And, clearly, nicotinic acid given in adequate doses reduces small LDL significantly and at the same time can increase the large region and, thus, the absolute amount of reduction in total LDL might be a bit deceiving.  But that has been published a number of times.

Perhaps the question has more to do with is there clinical utility in using nicotinic acid?  And my belief, there is and there are some issues and problems with some of the negative studies.  But I’ll leave that to another day because it might take too long.  So, Ron, do you want to comment on that?

Dr. Krauss:  Yes.  I think there has been controversy around niacin.  But just to briefly describe, one of the issues here is that the recent trials that failed to show a benefit of niacin only tested that benefit in patients who were already being very aggressively treated with statins, and, as needed, ezetimibe, to get LDL cholesterol levels very low into the 60s.  And we and others have reported, as I think you might have mentioned, that the risk of cardiovascular disease in patients on statin plus niacin is really related to the small LDL particles.  And if you get these levels very low with aggressive statin treatment, it may be very hard to show an incremental benefit from adding niacin, whereas most of the patients that we treat in practice don’t get down to LDL cholesterols that low.  These studies were selected for people that were able to achieve a very low LDL and, at least in my experience, you have to give a high dose of statins in many patients,  and even then, you usually don’t get down into that range.  And so niacin can have the additional benefit of lowering the small particles on top of what statins are doing, but that wasn’t really tested in these trials.

Question:  Are LDL particles a better measure of cardiovascular risk than Apo B?  Why?

Dr. Superko:  I think they’re asking about LDL particle number versus the Apo B value.  And both are legitimate and there’s strong support for Apo B measurements since there’s one Apo B on every LDL or atherogenic particle.

Part of the issue is that the standard Apo B measurement measures whole plasma Apo B.  So it measures Apo B throughout the entire spectrum.  And some studies have actually isolated the LDL region and then measured Apo B in the LDL region on the ultracentrifuge and called that LDL Apo B.  So there’s two ways of talking about Apo B, B100 and B48.  B48 also cross-reacts with the standard laboratory test so you’re not sure if you’re measuring the Apo B actually on the LDL particle or Apo B48 on other particles.

Nevertheless, total Apo B in studies like the Quebec Cardiovascular Study are useful in predicting cardiovascular risk.  You can fine tune that now down to the numbers of the specific particles and subparticles you want.  And that’s the beauty of Ion Mobility.  You can now say, “How many total LDLs or how many LDLs in the small or the very small region do I have?” and look at the clinical utility of that based on studies like Malmo.  So if I had to pick one or the other today for clinical utility, I’d pick the Ion Mobility number, but there is validity in Apo B.

Ron, do you have comments?

Dr. Krauss:  No, I think you’ve described that quite well, Robert.

Question:  In the era of evidence-based medicine, third-party payers look to professional society guidelines when determinging coverage.  Does the AHA ACC have any recommendations for guidelines regarding the use of lipoprotein subclass testing for patient evaluation or management?

Dr. Krauss:  Certainly, the official guidelines that I’m aware of don’t specifically talk about subclass measurements as being part of standard of care.  I think there has been some concern about the various methods that are out there and issues of standardization.  And as was pointed out, evidence-based is rather difficult because these particle measurements are statistically correlated with some of the standard measurements such as Apo B and non-HDL cholesterol.  So even though the measurement is superior in terms of directly measuring the causal particles, the particles that are causing the atherosclerosis, it’s very difficult to prove that statistically.

But there are studies that I’m aware of in the pipeline that are beginning to support the value of the particle measurements above and beyond the standard measurements.  And my guess is that the guidelines are going to start reflecting those at some point, but not just yet.

Dr. Superko:  Some of the guidelines, like the European Consensus Conference in 2011, are somewhat supportive in that they say that the LDL size is an independent predictor of the rate of coronary disease progression, but they do not recommend it for widespread public screening.  And I think this has been a confusing point in that some people suggest this test be used in shopping centers for widespread screening.  And that’s nothing that I think Ron or I would ever say, but it would be used in patients in which it gives you additional information that’s clinically useful.

And even things like the Agency for Healthcare Research and Quality, say that both the LDL particle concentration and size are important predictors.  However, the difference in laboratory methods out there in the community complicates the interpretation and use of that knowledge.  So I think that’s where Ron’s work becomes really important because if you look at the body of evidence, the body of knowledge, it’s tremendous in regards to the importance of subclasses and this is largely based on Dr. Krauss’ work.  And now we have a laboratory method that, as Ron says, is based on the first principles of physics and you do away with a lot of the laboratory problems we’ve had in the past.

Question:  Does an increased number of bigger HDL particles mean that they function better to enhance reverse cholesterol transport?

Dr. Krauss:  The bigger HDL particles are a marker for reduced heart disease risk.  Part of the reason that it’s associated with reduced risk is because these particles can remove cholesterol from cholesterol-enriched macrophages in the artery wall.  That’s been shown.  But it’s not the only particle that does that.  So it turns out there are other particles within the HDL that do this and some of them are in the small HDL.  And so this is a field that’s under active research because it would really be helpful to know what all the particles of HDL might be that are important in this regard.

But the large HDL per se has not been shown to have a direct causal relationship with heart disease risk.  It’s really a marker, in part, because it’s associated with these other lipoprotein changes that we know are related to risk and that’s levels of small LDL and triglyceride rich lipoprotein particles which are often elevated when the large HDL is low.  And conversely when the large HDL is high, it’s associated with lower levels of these atherogenic particles.  But at this point, it is not at all clear that these particles, these larger HDL, are directly affecting risk.  And this has important implications for some treatments that are designed to raise HDL and large HDL, in particular, with the hope of reducing heart disease risk.  The challenge is to show that raising of these large HDL is going to be beneficial.

Most of the treatments that are out there, and niacin is a great example, that raise the large HDL also reduce levels of the atherogenic smaller LDL; and we think that’s the more important effect.  So measuring HDL can be useful in assessing risk, but at this point it’s not considered a target for treatment because it’s really not been shown that this is directly beneficial in terms of pulling cholesterol out of the arteries.

Question:    How should clinicians prioritize subfractionation findings versus other cardiovascular test results with regard to diagnosis and treatment outcome?

Dr. Superko:  Well I’ll jump into that for a second.  And I think what I said before, that if the information from the test provides additional information beyond what you already have, then it’s useful.  And an example of that would be whether or not you want to determine cardiovascular risk in an individual and whether or not they have this small pattern B or the less dangerous pattern A.  And one way to assess that is simply fasting triglycerides.

If the triglycerides are elevated above approximately 200 to 250 mg/dl, the likelihood of having small LDL is very, very high.  If the triglycerides are relatively low, less than about 70 mg/dl, the likelihood of having the bad pattern B is very, very low.  But within that range of approximately 70 to 200 mg/dl, making a guesstimate on whether or not they have the pattern A or B simply based on fasting triglycerides, you can make a mistake about 50% of the time.

So if my job for the patient is trying to determine cardiovascular risk, and I know that if you’re a pattern B, that increases risk independently, particularly within that range, and it amplifies other risk factors, then that would be a very useful piece of information for me in terms of treatment.

Now in terms of treatment, one of the most effective treatments is the least expensive.  And it is avoidance of simple carbohydrates in the diet, weight loss and exercise.  So you’re promoting more healthy lifestyle in the person with the pattern B.  And if then you turn to medications, as we’ve discussed, getting the total amount of atherogenic LDLs down would be important.  And then, if necessary, the use of nicotinic acids.

In patients with established coronary disease, those arteriographic regression studies I mentioned all support the fact that if you don’t do anything in your pattern B patient, then the arteriographic rate of progression is about twice as fast as the pattern As with coronary disease.  So there’s something about the rate of progression.  And in Paul Williams’ most recent paper in 2013, the arteriographic change is directly related to the amount of small LDLs as he showed in his paper.  So I think if it allows you to do something different, then there’s clinical utility in the test.

Question:  This generates a lot of additional information.  What do we do with it?

Dr. Superko:  The method generates a lot of information, and I think it’s reported in two ways, sort of a brief summary and then on page 3 or 4 there’s all the different subfractions.  And so it can be used for individuals who just want to know what’s the small LDL number or the particle number, if you will, and the lipid values and maybe phenotype.  And then people who are more interested and dig into the data and want to know what amount of a very small LDL you have or medium size LDL or large LDL, then there’s a lot more detail on I think the third or fourth page.

So the primary important information is, in my opinion, what’s the phenotype, or your pattern A or pattern B; and what’s the number of the dangerous LDL particles?  And I can help predict risk with that and then I can help determine therapy and follow the response to my therapy, whether it be weight loss or exercise or drugs.  And, number two, if I want to really look into the detail of it, I can go into that highly detailed page.

So I’d be interested in what Ron’s feelings about the clinical use are.

Dr. Krauss:  Well, when I follow patients on treatment in particular, I look really, first and foremost, at the concentration of the medium and small LDL particles because that’s where the risk is associated and I want to be sure that we’re getting the levels down into the lower risk range.  So that’s where the principal value I see in terms of patient management.

It can certainly also help in assessing risk, as you were saying, in people who have borderline lipid triglyceride in the 150 plus or minus range, and other risk factors, family history, etc., in helping to decide how aggressive to be in managing those patients.  And, again, it’s looking at the whole profile but starting with the small and medium-sized LDL.  And that generally goes along with the total LDL particles, but not always.  So I find the Ion Mobility report to be most useful in identifying the risk associated with the smaller LDL.

And as I said a moment ago, this is often associated with lower levels of the larger HDL, so that information also helps to reinforce the overall risk assessment.  But the important use of Ion Mobility is, I think, for managing and following patients.  Just lowering LDL cholesterol doesn’t tell the whole story.