LDL: more data supporting treat-to-target

A recent article looked at the relationship between achieved LDL cholesterol levels and changes in coronary artery atherosclerosis progression, as determined by coronary CT angiography, CTA (see lipids intensive LDL lowering dec CAD progression JACC2016 in dropbox, or doi.org/10.1016/j.jcmg.2016.04.013). 

Details:

-- 147 patients with evident atherosclerotic plaques on CTA had quantitative plaque size measured both at baseline and at follow-up CTA two years later (median 3.2 years)

-- baseline characteristics: mean age 62, 57% male, 65% with hypertension, 33% diabetes, 20% active smoking, 27% dyslipidemia (defined as total cholesterol >240 mg/dL, LDL >130, HDL <40, triglycerides >150, and/or treatment with lipid-lowering agents), cardiac risk score by both the ATP III and Framingham risks: 60% low risk (<10% over 10 years), 30% intermediate (10 to 20%), and 10% high risk (>20%). Number risk factors = 1.6

-- multi-center, observational study, assessing those who had LDL <70, versus >70, to assess changes in quantitative measurement of plaque volume, using the modified 17-segment American Heart Association model for coronary segment classification

-- this was an industry-supported study done in 4 Korean  centers

Results:

-- of note, those that had a follow-up LDL <70 had a higher prevalence of diabetes (p=0.002) [ie, were likely at even higher risk of progressive disease]

-- those with an LDL <70 had significant attenuation in plaque progression: 12.7 mm³ versus 44.2 mm³, p=0.014.

-- Multivariate analysis found that the only factor influencing plaque progression was the follow-up LDL level (p=0.021), controlling for age, hypertension, active smoking, and follow-up LDL <70.

-- all patients who achieved an LDL <70 were on a statin (n=37), whereas 63% (n= 70) of those with follow-up LDL >70 were on a statin; those actually taking a statin had similar plaque progression as the overall group, with plaque progression of 12.7 versus 41.8 mm³​

-- subgroup analysis showed that annual plaque volume could be attenuated with aggressive LDL control (4.6 mm³ versus 14.5 mm³)

Commentary:

-- coronary atherosclerosis remains a major cause of global morbidity and mortality (though decreasing in many resource-rich countries, but increasing in resource poor countries), with an estimated global burden of 17 million deaths annually.

-- Prior studies looking at intravascular ultrasound (IVUS) have shown that intensive LDL lowering can halt the progression of atherosclerosis and even promote some regression. CTA provides a noninvasive assessment of CAD with high reproducibility and a diagnostic performance correlating well with that of the invasive IVUS.

-- As noted by the editorialists (see lipids intensive LDL lowering dec CAD review JACC2016 in dropbox or doi:10.1016/j.jcmg.2016.08.002), several coronary angiography and IVUS studies have found that LDL levels need to be brought down to the 70-80 mm/dL range in order to decrease plaque progression, and even lower LDLs were needed for plaque volume reduction. The above study found almost a 70% decrease in the annual rate of plaque progression in those with LDL <70.

-- Plaque progression is in fact a reasonable surrogate marker for future cardiac events. One study found that plaque progression over time was associated with a 28% likelihood of a cardiac event versus 10% in those without plaque progression. Other studies have shown that plaques with rapid progression were much more likely to be the culprit lesions in cardiac events (ie, the lesions that rupture and cause an acute cardiac event)

-- a few additional comments:

    -- this was not an intervention study. As with other studies looking at achieved LDL and cardiac events, these patients were not randomized to specific LDL targets, so one cannot definitively conclude that it is beneficial to treat-to-target. For example, perhaps those who more easily achieve a lower LDL target have a less malignant atherosclerotic course. However, there are a few lines of argument which suggest that treating to lower LDL targets is highly likely to be beneficial:

        -- the Treating to New Targets trial (Barter P. N Engl J Med 2007; 357:1301-1310​), which compared the effects of atorvastatin 10 versus atorvastatin 80 mg, found that though the number of clinical cardiac events was lower overall in the group on 80 mg, there was no difference between the two groups if one looked at achieved LDL levels (by the way, this trial also showed that the number of cardiac events was also related to the HDL level: those with LDL >100 but HDL >55 had the same 5-year risk of a cardiovascular event as those with achieved LDL<70, but HDL <38. and statins had essentially no effect on HDL, independent of their dose). see top graph (graph A) in first graph below

        --data from the constellation of many lipid trials, looking at both the achieved LDL on statins and the natural LDL of patients, show a straight line relationship: those with lower achieved or natural LDL down to 70 had fewer cardiac events. see second graph below from LaRosa JC. N engl J Med 2005; 353:1425. ( this also suggests, as seen in several studies, that it is the on-statin achieved LDL that matters, and the "pleotrophic" non-lipid effects don't seem to matter so much. 

        --the editorialists had a very similar graph to the second one below (see reference above), showing that achieved LDL targets were associated both with changes in percent of atheroma volume as well as angiographic mean luminal diameter/extent of stenosis

       ​ -- the data from the ezetimibe trial (see http://gmodestmedblogs.blogspot.com/2015/06/improve-it-trial-ezetimibe.html ) showed that adding ezetimibe to simvastatin decreased cardiac events, more so than simvastatin itself, again suggesting further decreases in LDL were beneficial. The data on PCSK9 inhibitors is more preliminary, but have shown a dramatic decrease in LDL over just using  statins, and a decreased cardiac event rate (Robinson JG. N Engl J Med 2015; 372: 1489 studied high-risk patients on max dose statin but LDL >70, then randomized to PCSK9 inhibitor plus statin vs continued statin, and found a 62% further decrease in LDL by the PCSK9 inhibitor and a 48% decrease in cardiac events over 78 weeks). These studies further support a low LDL target

    -- so, despite the lack of rock-solid studies, there is a remarkable convergence of data from different types of studies suggesting that targeting a lower LDL is the appropriate approach. The graph in the editorialists' article shows that the achieved LDL correlates very well with changes in atheroma volume as well as the residual luminal diameter of coronary arteries. The second graph below that shows that clinical events track with the achieved LDL. And the Targeting New Targets trial showed that it was the achieved LDL and not the dosage of atorvastatin (10 vs 80mg, which correspond to "moderate" and "high" intensity statins by the 2013 ACC guidelines) that was associated with clinical outcomes (though the lower achieved LDL was more common with the 80 mg dosage)

-- So why did the new AHA guidelines in 2013 make a point of eliminating the treating-to-target approach? They state that this was because there was not enough rigorous data to show that treating-to-target really mattered (as noted above: no specific trials looked at this). However, many of their other recommendations had much less data to support them, including looking at the individual’s 30-year risk profile or treating a 21 year-old with LDL >190 aggressively (both of which I support, it is just that there really are no data to support these by RCTs, as opposed to the above data which i think pretty strongly support a targeted LDL). My best guess is that they feel that statins are used insufficiently overall, so they wanted to develop the easiest algorithm to reinforce using statins in people at increased cardiovascular risk (in fact, in the above Korean study only 56% of really high-risk patients who were not on a statin initially were actually put on a statin by their clinicians). But there are several studies showing that applying the 2013 recommendations would lead to dramatic overtreatment (eg, the AHA/ACC 2013 guidelines would recommend statins for nearly all men and two thirds of women >55 years old). In addition to exposing lots of people probably unnecessarily to statins, the approach of stratifying patients into needing either intensive- vs moderate-dose statin therapy does potentially create unfortunate clinical consequences in those who clearly need statins:

    -- I have seen several patients with clinical CAD who, when put on atorvastatin 10 mg, achieve an LDL in the 40s. Should I really be putting them on high intensity statin? There are more adverse events taking higher dose statins, and what is likely to be the benefit?? (and there are animal data, and old human data on lipid-lowering therapies, as well as the recent PCSK9 studies, finding that there may be neurocognitive issues with too-low an LDL.... Cholesterol is part of the cell membranes of neurons, which is integral to neural transmission)

    ​-- I have also seen many very high-risk patients on atorvastatin 80 mg who do not achieve an acceptable LDL level (specifically, <70). In many cases, by switching them to rosuvastatin 40 mg they have done much better. I have not seen this as a tested clinical strategy, but given our understanding of statins and the above data on the benefit of LDL <70, I think this is a reasonable strategy. And I would try this before adding ezetimibe, or PCSK9 inhibitors...

--so, I am concerned about the clinical implications of adopting a simple algorithm of "high" vs "moderate" intensity statins: we may be overtreating some patients, as well as undertreating others, in a disease with potentially very bad outcomes and with tried-and-true meds which are quite effective in lowering the likelihood of these events, and with minimal adverse effects. I do in general support an aggressive approach to LDL management, especially in those at higher risk of clinical cardiac events, and I strongly suspect that the clinical risk of undertreating with statins is much higher than that of overtreating. However, I also think we should be tracking and following lipid levels, both to make sure that the patients are taking their statins appropriately, and also to titrate the statin dose to the individual.

For prior critiques of the 2013 ACC/AHA guidelines, see http://gmodestmedblogs.blogspot.com/2014/04/aha-lipid-guidelines-again.html  , http://gmodestmedblogs.blogspot.com/2014/10/yet-another-analysis-that-lipid.html  and,  http://gmodestmedblogs.blogspot.com/2015/07/comparison-of-2013-accaha-lipid.html