current controversies in lipids:LDL, HDL, TG

a series of 3 articles in Lancet on where we're at and future possibilities about lipids: one on LDL, one HDL, and other Triglycerides (TG). brief review of points.

1. LDL (see lipids LDL controversies lancet 2014 in dropbox, or Lancet 2014; 384: 607–17)

--reinforces that calculated LDL (by Friedewald equation) is usually adequate to evaluate LDL (though underestimates true LDL if high TGs, esp if LDL is very low either naturally or by using statins), or when HDL is very high. No need to get direct LDL, unless unable to calculate (usually because of very high TG).

--promotes the use of non-fasting lipid measurement. also, there are somewhat stronger data that the non-HDL lipid fraction (total chol minus HDL) is a better predictor of events than LDL, both in those on statins and not on meds (see below as well).

--there are data from several studies that the relative risk reduction of statin use is higher in those with lower absolute risk, supporting the concept that it is more useful to start statins early than late (ie, preventing more advanced atherosclerotic lesions is better than treating ones already there). [the counter-argument is that the absolute risk reduction in those initially at lower risk is much lower than in those with more advanced disease -- i would argue however that the intervention studies just aren't long enough. if we did primary prevention studies and followed patients 30+ years, my guess is that their absolute risk reduction would be quite high. and this is consistent with the observation that statins don't do much in patients with end-stage renal failure or heart failure, who likely have very advanced atherosclerotic disease – further suggesting that earlier treatment is better. this also reinforces, in my opinion, lipid testing early -- as in teens -- and aggressive lipid reduction, especially by lifestyle changes]

--reinforces that statins should be prescribed along with appropriate lifestyle changes [it is quite concerning to me that a few studies have shown that patients on statins often start eating indiscriminately and decreasing exercise, since the statins so dramatically decrease the lipids -- the thing they and their providers often focus on.  this all negates the plethora of beneficial effects of healthy diet and exercise for the whole body and mind, and could potentially make things worse]

--if you still follow LDL levels in patients on statins (treating to goal, which is discouraged by latest am heart assn guidelines, but encouraged elsewhere and by me), and you are unable to achieve the target you desire by just increasing the potency of the statin, then DO NOT suggest adding any other LDL lowering agent. there is no benefit and potential harm to adding niacin, fibrate, or ezetimibe (ezetimibe, by the way, has never been shown to decrease clinical events and, until it is shown to do so, i would strongly argue never to use it)

    ​--some potential new advances (awaiting studies clearly showing clinical benefit) include new meds to prolong the life of LDL receptors, PCSK9 inhibitors (PCSK9 being a very promising target, and the inhibitors interfere with binding of the PCSK9 protein to LDL receptors and less degradation of the receptor, with dramatic decreases in LDL), and a few agents approved for patients with homozygous hypercholestrerolemia (mipomersen, lomitapide)

to me, there is some issue about whether LDL is the bad actor or a surrogate marker for something else (eg oxidized LDL): it seems that LDL crosses the endothelial wall, initiates an inflammatory response and is oxidized, and activates endothelial cells which express adhesion molecules and attract platelets. the adhesion molecules lead to WBC adhesion, migration into subendothelial space, release of cytokines that cause differentiation of monocytes to macrophages and up-regulate their scavenger receptors. and it is the oxidized LDL (not native LDL) which is taken up by the tissue macrophages through the scavenger receptor, creating foam cells. diabetes leads to increased LDL oxidation (which justifies treating LDL more aggressively in diabetics).

2. HDL (see lipids HDL controversies lancet 2014 in dropbox, or Lancet 2014; 384: 618–625)

--environmental factors that affect HDL: lowered by obesity, diabetes, inflammation, smoking; raised by estrogen, thyroid hormone, exercise, alcohol

--the data on the HDL decreasing cardiovasc risk is consistent in different populations, ethnic/racial groups, and worldwide. and most studies of patients on statins confirm the protective association of HDL

--the questions of late dealing with the real role of HDL revolve around: disorders associated with low HDL (mutations of apo-A1, ABCA-1,LCAT) do not consistently predict high cardiovasc risk (though the question is raised whether findings from these and other genetic HDL variants are applicable overall to the general population), and the fact that HDL raising drugs do not reduce cardiac events (old study on niacin, the Coronary Drug Project, did show benefit but only after 6 years and mortality benefit only 9 years later; adding niacin to statins doesn't do much. the CETP inhibitors increase HDL but torcetrapib led to increase cardiac events). so, concept has evolved of "HDL function", ie that it is not the HDL cholesterol concentration itself but its function that is important. studies which have looked at function (eg promoting cholesterol efflux from macrophages) show a not-so-great correlation with HDL quantity, and function is a strong inverse predictor of cardiac events. there are a couple of CETP inhibitors being tested which seem to improve cholesterol efflux (though there are concerns about the methods used to measure efflux). there are also other therapies in the works which might support reverse cholesterol transport without raising HDL levels (eg, by raising apo-A1 levels)

my concern is that either the drugs used to target HDL are mixed bags (eg, niacin also increases glucose/causes insulin resistance, increases CRP levels, etc), and that there are non-protective and even pro-inflammatory HDL particles (eg jama march 5, 2014 had article on dysfunctional HDL: pts with atherosclerotic arteries have their apo A1 oxidized by myeloperoxidase and lose its cardioprotective properties. also there are data that HDL with apo-C3 are pro-inflammatory and atherogenic -- see lipids inflam hdl apoC3 jaha 2012 in dropbox, or doi: 10.1161/JAHA.111.000232. I am personally not convinced that because the few drugs tried so far do not work that we can then conclude that HDL is not an appropriate therapeutic target. for example, if we did not have statins and only had ezetimibe as an LDL-lowering agent, we might conclude that LDL is not a target since ezetimibe doesn't seem to help clinically.

3. Trigylcerides (see lipids Triglyc controversies lancet 2014 in dropbox, or Lancet 2014; 384: 626–635​)

--confusing if this is a real risk factor. people with genetic hypertriglyceridemia with very high levels (chylomicronemia syndrome) do not seem to develop atherosclerotic disease.  (however, in people with mild to moderate hypertriglyceridemia, the lipoprotein particle size is smaller and could enter the arterial wall and cause atherosclerosis.) in addition,  most people with high triglycerides have concomitant low HDLs, which explain much of the prior association. But, there is still a residual risk of high triglycerides in meta-analyses. the trials mixed as to whether people on statins who have residual high TG are at increased risk of CAD event. [this analysis does not include a couple of studies I have seen over the years: eg the Quebec Study has found that triglyceride levels >176 (vs<132) predicted patients' having small, dense, and highly oxidizable LDL particles (and those particles are 3x as atherogenic, in some studies). presumably, this explains the strong relationship between high apo-B levels and atherosclerosis, higher than just the LDL level: each LDL particle has one apo-B, so for a given LDL concentration, those with small LDLs have higher concentrations of apo-B.]

--another area of confusion is whether fasting or nonfasting triglycerides is more useful in risk prediction. data has shown that endothelial dysfunction occurs soon after a meal which raises triglycerides. data from three trials in JAMA found that nonfasting TG was more predictive of clinical events [see nonfasting TG and cad events jama2007 in dropbox, or JAMA. 2007;298(3):299-308; nonfasting TG and risk of stroke jama2008 in dropbox, or JAMA. 2008;300(18):2142-2152; nonfasting vs fasting TG jama2007 in dropbox, or JAMA. 2007;298(3):309-316)]. and one concern about the use of non-fasting triglycerides is that the Friedewald equation was based on fasting TG,  but there is still a high correlation between calculated and measured LDL in the nonfasting state

--there is a long and interesting discussion of remnant cholesterol, which i will leave to you (though there is argument that remnant cholesterol, which is  total chol minus LDL and minus HDL, has been assoc with cardiovasc disease, can lead to endothelial injury and inflammation, and, unlike native LDL, is taken up by macrophages leading to foam cells. this is also the argument about oxidized LDL, as above). also mendelian randomization studies have found genetically raised remnant chol levels is assoc with about 2-fold inc in cardiovasc disease.

--treatment includes weight loss, exercise, decreasing dietary fructose (though I would add that a recent systematic review did not show a clear association – see Atherosclerosis 2014; 232: 125-133.), dec alcohol, statins (esp intensified regimens), fish oils/fibrates (though data lacking/mixed that this helps cardiac risk). i would also add that low glycemic index diets consistently lower TGs (and improve the cholesterol/HDL ratio). hard to assess the risk of lowering TG on cardiac events, since the lipid trials (including most statin trials) excluded people with very high TG (eg >4.5 mmol/L, or 400 mg/dl), but still the meta-analyses do show some benefit from lowering TG, in trials with fibrates, fibrates and niacin, fibrates and statins, but the data overall is much less impressive than statin-lowering of LDL.

--guidelines for treatment vary considerably (US endocrine society and 3 European ones) recommend treating mild-to-mod increased TGs, but Am Heart Assn and ACC plus a European one does not. all agree to treat very high levels to avoid pancreatitis (though variability of lower limits of when to start meds from 5.7 - 11.4 mmol/L, or 500-1000 mg/dl)