This is my car, Milano. He's oldish, high milage, quite stylish and is somewhat ambivalent about whether he is a boy car or a girl car. He's also some sort of Italian derivative which means that spares can be an issue. I like him a lot and am dreading what the Glasgow winters will to to his incipient rust...
Anyway, this post is really about another Milano, the magical apoA1 Milano.
We all know that low HDL is associated with an increased risk of CHD. We know this from an enormous amount of observational evidence. In general, people with low HDL are much more likely to suffer a heart attack. Totally convincing association. This has lead to the hypothesis that HDL protects against heart attacks. To the point where researchers are now developing HDL-like particles as drugs. Pour 'em in and suck out that arteriosclerosis. Sooooo cool.
I suppose the first spanner in the works was apoA1 Milano. Even before torcetrapib.
The original study was published back in 1985 and is available as a pdf download here.
The original authors, Franceschini's group, concluded:
"The AIM [apoA1 Milano] variant originated in a community with a low prevalence of cardiovascular disease, which is certainly not attributable to the modest number of AIM carriers. We put forward the hypothesis that this may be the consequence of a general environmental condition in which the small community lived and which did not allow, up to now, the expression of the selective effect of the mutation."
It's worth emphasising that apoA1 Milano was found in a community which had a low prevalence of heart disease and, within that community, it was not considered to be the explanation for that low incidence of CVD. Everyone in the community was protected. An hypothesis I might suggest, derived from the lipid hypothesis patch of "HDL is good", is that under high CVD risk conditions that apoA1 Milano might be bad news. There was no selection effect to eliminate the gene in Limone sul Gardo because the population here was already doing something very right. Probably living on Lardo.
The impression I get from Franceschini is that he is proposing the opposite hypothesis. Because apoA1 Milano results in low HDL, people with it should have been sick, so there is magic in the single amino acid change which stops people with low HDL getting heart disease, so long as it's the Milano type HDL they have. Magic. Cranks up their trigs too, but still no problems. Super magic. This was back in 1985 and I don't see that Franceschini would be knocking the "HDL is good" hypothesis at that stage. The fact that Milano is different to normal apoA1 allowed this hypothesis to float briefly.
There's another genetic error resulting in low HDL, a problem with the gene for LCAT (Lecithin Cholesterol Acyltransferase). LCAT takes free cholesterol from tissues/lipoproteins and esterifies it ready for placing in to HDL particles. The mutation leads to cholesterol build up in some tissues and very low HDL levels in the blood. If you can't put cholesterol in to HDL, there won't be much HDL cholesterol. And what little HDL is present will have a normal apoA1 on it's surface. Nothing Milanesque about it. It's a dribble of that same HDL that keeps anyone with high HDL free of CVD, if you believe that. You would have thought this was an absolute cert of a recipe for atherosclerosis. It's not.
Franceschini's group did the work and were again good enough to publish in a journal with free access. The abstract is here. Just click on the PubMed link for full text.
It turns out that if you have a double dose of the defective LCAT gene you will be walking around with an HDL of around 9mg/dl. The standard deviation around this mean is almost 5mg/dl. Go figure how low some of these people must be on the HDL front. People with the LCAT mutation are rare, but are not clumped in anywhere particularly pleasant like Limone sul Gardo. They live in a nasty world of pasta and sugar based ice cream.
So what about arteriosclerosis?
The study used carotid intima-media thickness as a surrogate for arteriosclerosis. This seems reasonable to me. This is what was found:
"The average and maximum IMT values in the carriers were 0.07 [average] and 0.21 mm [maximum] smaller than in controls (P=0.0003 and P=0.0027), respectively. Moreover, the inheritance of a mutated LCAT genotype had a remarkable gene-dose-dependent effect in reducing carotid IMT (P=0.0003 for average IMT; P=0.001 for maximum IMT)."
I'll translate that. The lower the HDL the less the arteriosclerosis, p<0.0003. Count the zeros.
I think it's also worth mentioning that HDL transfers both apoprotein C2 and apoprotein E to VLDLs. No HDL means poor processing of VLDLs, exactly as for the Milano carriers. Do these victims also have high trigs, as well as low HDL? You bet they do. The lower the HDL and the higher the trigs, the less arteriosclerosis. Even though their HDL:trig ratio is appalling. Go figure. And you thought trigs were sticky and caused...
That's really all I wanted to say about papers. Having a high HDL might or might not be good or bad. If you are a sort of average person with sort of average genes the level of HDL in you bloodstream reflects the amount of fat, particularly saturated fat, in your diet. It's a surrogate. It seems a perfectly reasonable point of view that it is the saturated fat that is good and the HDL merely reflects this, an epiphenomenon. Get a gene for low HDL but do what, in anyone else, should raise HDL and perhaps you will still get the benefits, without the "marker" of high HDL. This seems to be a perfectly reasonable hypothesis to me.
Other than eating saturated fat, accessing saturated fat by weight loss or mimicking ketones by taking niacin, you don't have many levers left to raise HDL anyway. Torcetrapib excepted, and we all know about poor old torcetrapib.
This was all floating around the desktop but the post was triggered by Lynn's comments on the previous post. She has an elevated sdLDL number, the evil incarnate sort of LDL, despite a very sensible eating pattern. So you have to ask whether what matters is the eating pattern or the lipoprotein pattern. Does it matter what your lipids are, if you are eating real food such as eaten by Lynn or even the Kitavans?
Personally I'm a little envious of Lynn's calcium score of zero despite the sdLDL. Only an n=1 anecdote, but I like it. I feel this calcium score might matter more than a sdLDL number. After all, people try and lower their sdLDL to achieve what she already possesses...
Never forget that originally total cholesterol was evil incarnate, then LDL was bad, now sdLDL is bad, what will be the next bodge? I tend to favour purple spotted sdLDL as the problem. Anyone with the pink tinged sdLDL will be fine. Until the next ad hoc is floated.
Oh, and just an addendum on apoA1 Milano and my favourite cardiologist Dr Nissen:
Nissen is using apoA1 Milano as "Drano" [©Kendrick] to unblock coronary arteries. His logic, as far as logic goes, appears to have been that the low HDL in a few members of a generally heart disease free population must have been due to some magic in its structure to produce protection using just this miniscule amount of special HDL. The fact that the whole population was particularly heart healthy seems to have escaped him. And of course, if he'd been interested enough to find out why that population was healthy he might have succeeded in preventing heart disease. But that's not a lot of use when you can make money from a cure. Nissen can get atherosclerosis regression with intravenous apoA1 Milano, a cure!
So how does apoA1 Milano Drano work? You could say it doesn't, and Nissen is a crook, but I would never say anything like that. Obviously apoA1 Milano really does hoover cholesterol out of atheroma and transport it back to the liver. Err, like yeah.
What have people other than Nissen found out about apoA1 Milano? How about this:
"In fact, in some systems, acceptors containing the Milano variant of apoA-I promoted significantly less efflux than the acceptors containing wild-type apoA-I (apoA-I(wt)). Additionally, intracellular cholesteryl ester hydrolysis in macrophage foam cells was not different in the presence of either apoA-I(Milano) or apoA-I(wt). CONCLUSION: Collectively these studies suggest that if the Milano variant of apoA-I offers greater atheroprotection than wild-type apoA-I, it is not attributable to greater cellular lipid mobilization."
Translation: However the Drano works (oops, I mean IF the Drano works... sceptics, these folks), it's not by reverse cholesterol transport. Poof, what's that I hear, another patch on the lipid hypothesis bursting?
So how does apoA1 Milano work to alter the thickness of abnormal arteries on ultrasound scan?
These people seem to have some sort of an idea:
"In in vitro studies, incubation of platelets with apoA1 Milano reduced ADP-induced platelet aggregation by about 50%, but apoA1 Milano had no direct effect on vasoreactivity. This study provides further evidence for critical role of platelets in thrombosis. Use of apoA1 Milano offers a novel approach to inhibit arterial thrombosis."
Translation: apoA1 Milano is an antithrombotic agent. A bit like statins really. Or, dare I mention anything this cheap, aspirin. The thrombosis theory of arteriosclerosis got a bit sidelined by the cholesterol fed rabbit. Amazing what people believe.