LDL goal after new-generation stents

 A secondary analysis of the CONNECT trial, that stands for (a bit of a stretch...)  “A Randomized Comparison of Long-Term Vascular Healing Between Biodegradable Polymer vs Durable Polymer Everolimus Eluting Stents in Acute ST-Elevation Myocardial Infarction” found that a lower target LDL level was associated with decreased neoatherosclerosis, which is associated with subsequent stent failure (see LDL goal after stent JAMAcardiol2025 in dropbox, or doi:10.1001/jamacardio.2025.4723)

 

Details:

-- the CONNECT trial was a prospective, multicenter, open-label, assessor-blind, randomized clinical trial to investigate the association between achieving guideline-endorsed target LDL levels with long-term neoatherosclerosis formation after current-generation drug-eluting stent implantation. This trial was conducted in 7 centers in 2 countries (Switzerland and Japan)

    -- patients at least 18 years old were eligible if they presented with an ST-segment elevated myocardial infarction (STEMI) and had a percutaneous coronary intervention; 239 patients had either biodegradable- or durable-polymer everolimus-eluting stents between June 2017 and June 2020.

-- the goal of this study was to determine the prevalence of neoatherosclerosis (defined as the presence of atherosclerotic change, comprising  fibroatheroma, fibrocalcific plaques, or macrophage accumulations within the neointima of a stented segment and having a longitudinal extension of >1mm at 3 years after stent implantation

    -- patients had repeat coronary catheterization at 3 years with optical coherence tomography (OCT) to evaluate the stented area

-- patients received guideline-endorsed targeted LDL levels per their country:

    -- the European Society of Cardiology promoted a goal of <55 mg/dL for Switzerland

    -- the Japan Atherosclerosis Society had a goal of <70 mg/dL (same as in the US at that time)

 

-- mean age 64, 15% female, BMI 26

-- diabetes 12%, hypertension 50%, dyslipidemia 45%, current smoker 40%, family history CAD 2%, eGFR <60 in 12%, previous PCI 5%

-- meds: aspirin 7%, P2Y12 inhibitor 2%

-- statins: baseline on high intensity 3%, on low or moderate intensity 10% 

-- baseline labs: creatinine 0.87 mg/dL; total cholesterol 197 mg/dL, HDL 46 mg/dL, LDL 124 mg/dL, triglycerides 133 mg/dL

-- Hgb A1C 6.08%, peak creatine kinase 2271 U/L

-- LVEF 50%, angiographic left main disease 1%, LAD 45%, LCX 12%, RCA 42%

-- thrombolysis (TIMI) flow grade before PCI: 0-1 in 59%, 2 in 32%, 3 in 9%; after PCI, flow grade 3 in 99%

-- stents: BP-EES, biodegradable-polymer everolimus-eluting stent in 88 (49%); and DP-EES, durable-polymer everolimus-eluting stent in 90 (51%)

 

-- Main outcomes and measures: prevalence in neoatherosclerosis 3 years after the primary PCI and stent placement, comparing patients with vs without achievement of their guideline-endorsed LDL target

 

Results:

--178 patients underwent OCT at 3 years:

    -- 98 patients (55%) achieved the target LDL level: 39 of 91 (43%) in Switzerland and 59 of 87 (68%) in Japan

        -- on-treatment LDL-C levels: 48 mg/dL in those achieving their target

    -- 80 patients (45%) did not achieve the target LDL level

        -- on-treatment LDL-C levels: 87 mg/dL

    -- differences in baseline assessments: in those who achieved the LDL target, LVEF was greater (52% vs 47%, p=0.01); there was a higher frequency of durable-polymer everolimus-eluting stent use in those who achieved the target LDL than in those who did not (58% vs 41%, p=0.04)

-- meds used at the 3-year mark:

    -- statin dosing: at 3 years, high-intensity statins were used in 87 (89%) who achieved target LDL levels vs 55 (69%) of those who did not

    -- no difference in ezetimibe use; other lipid lowering meds were used rarely

    -- aspirin use was lower in those achieving the target LDL vs not, 47 (48%) vs 51 (64%), p=0.049, but P2Y12 inhibitors were higher in Japan, especially as monotherapy 

-- prevalence of neoatherosclerosis:

    -- patients who achieved the target LDL: 7 patients [7%]

    -- patients who did not: 15 patients [19%]

        -- odds ratio for those who did not achieve the LDL target, 3.00 (1.19-8.24), P=.02

-- prevalence of neoatherosclerosis by achieved LDL goal:

    -- LDL <55 mg/dL: 5 of 64 patients [8%]

    -- LDL 55-60 mg/dL 7 of 61 patients [11%]

    -- LDL >70 mg/dL: 10 of 53 patients [19%]

        -- assessing these buckets of LDL range, there was a non-significant p=0.22 difference [but the buckets are large, and those with an LDL of 54 vs 55 would be in different buckets, although the actual difference in LDL is insignificant]

-- their figure 2 noted an almost linear relationship between the LDL level and the probability of neoatherosclerosis at the 3-year end of the study. and the increased risk extends to very low LDL levels;

-- the proportion of patients with in-stent fibroatheroma was significantly lower in those who achieved the target LDL compared to those who did not: 5 patients [5%] vs 13 patients [16%], p=0.02

    -- but the proportion with in-stent macrophages was comparable and no significant differences in minimum lumen area or mean neointimal area

-- intensity of statin therapy and prevalence of neoatherosclerosis at 3-year follow-up:

    -- overall, 142 patients (80%) were on high intensity statins, with mean LDL of 59 mg/dL

        -- 81% of Swiss patients and 78% of Japanese patients

    --22 patients (12%) were on low- or moderate-intensity statins, with mean LDL of 66 mg/dL

    -- 14 (8%) of patients had no statins, with mean LDL of 128 mg/dL

-- prevalence of neoatherosclerosis:

    -- if on high-intensity statin: 12 of 142 patients (8%)

    -- if on low- or  moderate-intensity statin: 3 of 22 patients (14%)

    -- if not on statins: 7 of 14 patients (50%)

        -- p<0.001 for the trend

-- Determinants of neoatherosclerosis:

    -- comparing those with vs without neoatherosclerosis:

        -- baseline A1c higher (6.52% vs 6.01%), p=0.048

        -- no difference by baseline clinical characteristics and biochemical findings, or stent type

        -- mean on-treatment LDL at 3-years was significantly higher: 83 mg/dL if present and 63 mg/dL if no neoatherosclerosis, p=0.005

            -- those with neoatherosclerosis were less likely to be on a high intensity statin, 55% vs 83% on the high-intensity statin

        -- in multivariable logistic regression model, only the LDL level at 3 years was an independent determinant of neoatherosclerosis formation, per a 25-mg/dL increase: odds ratio 1.46 (1.09-2.01), p=0.01

            -- diabetes status, kidney failure and stent type were not significant 

 

Commentary:

-- early stent failures happen at a rate of about 1-2%/year, without decreasing over time, and this risk seems to be largely from neoatherosclerosis, induced by rupture of a neointimal fibroadenoma or restenosis of an atheroma inside the stent

    -- a pretty recent larger trial of 512 patients having optical coherence tomography, as in the above study, found that the presence of neoatherosclerosis increased from the time of stent implantation to 20% at 1-3 years, 30% at 3-7 years and 75% when >7 years: https://www.ahajournals.org/doi/10.1161/CIRCINTERVENTIONS.121.011693 . much more aggressively so than in the current (smaller) study

-- this study added a few things to the knowledge base regarding stent thrombosis and failure:

    -- LDL was an independent predictor of neoatherosclerosis 3 years after a modern drug-eluting stent was implanted in patients having a STEMI

        -- and it did not seem to matter if the everolimus-eluting stent was biodegradable or durable-polymer

    -- the prevalence of neoatherosclerosis was lower in those who achieved guideline-endorsed target LDL levels

    -- those treated with high intensity statins did better

        -- though i would argue that this last finding is really not so useful: they did not show that high-intensity statins by themselves were associated with the improved outcomes. was it just the fact that these high-intensity statins lowered the LDL more, or was it the pleiotropic effects of these high dose statins (though older studies did not find increased pleiotropic effects with the higher dose). to determine the role of high intensity statins, the researchers would need to compare neoatherosclerosis outcomes stratified by the level of achieved LDL levels in those on high vs lower intensity statins. And i have certainly seen many patients achieve these guideline-endorsed LDL targets with moderate intensity statins. is there a reason for using high-intensity statins in these patients?

-- the identification of coronary artery neoatherosclerosis is a surrogate marker for an actual clinical outcome, and there is a potential danger of conflating a surrogate marker with an actual clinical event; there have been several surrogate markers that have not panned out to be useful over the years: https://gmodestmedblogs.blogspot.com/2024/06/using-surrogate-markers-for-disease-are.html)

    -- that being said, neoatherosclerosis is very likely a major mechanism leading to late stent failure after drug-eluting stent implantation, including stent thrombosis and in-stent restenosis, and it does make sense that this marker for the accumulating risk of a cardiac event is highly likely to manifest itself in future stent failure

-- studies shave found that the formation of atheroma in stented segments is a rapid process, occurring 10-20 times more rapidly than the native atherosclerotic process, perhaps related to the endothelial dysfunction by stent-related endothelial denudation and slow recovery

--I would argue that LDL is also just a surrogate marker, and that there are, in fact, more useful markers that better reflect the atherosclerotic process:  

    -- lipoprotein (a), Lp(a), is a strong atherogen/thombogen and is associated with earlier atherosclerotic clinical events as well as increased recurrent cardiovascular events. As per https://gmodestmedblogs.blogspot.com/2025/06/high-lpa-increases-risk-of-recurrent.html, Lp(a) does not have a clear direct medication that lowers it, but this blog pretty clearly shows that PCSK9 inhibitors are effective in lowering Lp(a) levels and decreasing recurrent cardiovascular events independent of LDL levels. In the absence of documented clinical benefit from the many direct Lp(a) meds in development, both the US and European cardiology societies recommend more aggressive LDL lowering in individuals with high Lp(a) levels. I personally believe that the recent study outcomes for PCSK9 inhibitors are strong enough to recommend them now in high risk patients with high Lp(a) levels, pending the results of the ongoing studies, though insurance companies don't seem to buy that argument.

    -- apolipoprotein B: the small LDLs in the electrophoretic assessment of lipid moieties are associated with high apoB levels; these are associated with the more oxidizable and atherogenic LDLs , with up to a 3-fold rate of atherosclerosis  (https://gmodestmedblogs.blogspot.com/2023/05/cad-presumed-mechanism-and-apob-was.html ), and, again, I think it makes sense to be more aggressive in treating LDL levels to a lower target

        -- i would add that Lp(a) is also oxidized as is LDL, and both oxidized entities enter the tissue macrophages much more so than the unoxidized variants do.

        -- and the small LDLs also are oxidized more readily and are more atherogenic than the larger LDL particles

        -- and entry into the macrophages is what leads to foam cells and the later cascade to atherosclerotic plaques

        --  a systematic review of low LDL levels found that the lower the LDL, the better in patients at high cardiovascular risk, down to an LDL of 21 mg/dL (https://gmodestmedblogs.blogspot.com/2018/08/very-low-ldl-levels-benefit-without-harm.html ), and the Fourier study more recently found benefit to LDL <20 mg/dL: cad high risk LDL less than 20 fourier studyCirc2023 in dropbox, or DOI: 10.1161/CIRCULATIONAHA.122.063399. A prior analysis of the Fourier study, which used a PCSK9 inhibitor, found that the PCSK9 lowered the Lp(a), and the greater the lowering, the better the outcome (Lp(a) Fourier study PCSK9 dec events Circ2019 in dropbox, or doi:10.1161/CIRCULATIONAHA.118.037184)

        -- would a lower target LDL than the guidelines have, perhaps down to LDL<40, improve the neoatherosclerotic outcomes more in this very high-risk group who got the stents?? the figure 2 above from the article does suggest that the lower the LDL, the better

-- one concern about stents is patient perceptions, wherein studies have shown that putting in the stents creates the impression in many people that they don't have to worry about their hearts anymore. sort of like the patients in the last blog who are smokers but have no nodules on their CT scans so they think it is ok to continue smoking (see  https://gmodestmedblogs.blogspot.com/2026/01/age-based-low-dose-lung-ct-but-more.html). and part of the reason that patients' adherence to taking statins is suboptimal in secondary prevention is likely the misperception that they are out of the woods after having a stent...

Limitations:

-- using creatinine based eGFR as the marker of renal function is fraught, as more studies show that eGFR based on cystatin-C predicts future bad clinical outcomes much, much more than eGFR based on creatinine (as in the above study), including in cardiovascular outcomes: https://gmodestmedblogs.blogspot.com/2025/12/test.html

-- the current study was limited in that it had only a 3-year horizon, and as we know, the atherosclerotic is a progressive, cumulative process. it would have been helpful to have intermediate assessments at 1 and 2 years after the stent implantation to quantify the neoatherosclerotic progression (although this is invasive and we do have that larger study mentioned above showing a pretty rapid neoatherosclerotic progression)

-- as per above, there are several concerns about surrogate markers instead of having actual disease outcomes, which make it more difficult to have a truly accurate and granular understanding (ie, is LDL is that the appropriate target? should we incorporate Lp(a) and apolipoprotein B as likely being more important clinical markers? other studies have found that the total cholesterol/HDL ratio is a more accurate predictor of future atherosclerotic events, as well as the calculation of non-HDL lipids)

-- this study was a secondary analysis of the CONNECT study, and secondary analyses in general do not have the statistical rigor of planned randomized controlled trials

-- it is pretty remarkable how non-aggressive the clinicians were is treating the lipids in these patients before their STEMI: the baseline LDL was 124 mg, and only about 10% on statins (rarely the intensive ones). this is pretty different from current clinical practice

-- also, there were really few patients with diabetes (12%), hypertension 50%, and dyslipidemia 45% (though there was no specific definition). this might limit generalizability to other areas (especially in the US, where numbers of these medical problems are significantly higher)

so, an interesting study, which highlights the very probable situation that progressive neoatherosclerosis is the major issue in stent failure over time, despite using the most modern stents. this brings up a few issues:

-- no doubt, some patients with stents do not continue in cardiology specialist care and come to primary care for the ongoing medical issues

-- we should make sure that the pretty clear conclusions above are enforced:

    -- there should be aggressive lowering of the LDL levels. not sure what the exact number is, but given the very high risk of these patients, probably the lower the better (per https://gmodestmedblogs.blogspot.com/2018/08/very-low-ldl-levels-benefit-without-harm.html ), with a target of <40 being a good starter

        -- lowering the target to a very low one is currently the only thing we can do to cover for the patients with high Lp(a) levels and apoB levels

        -- the PCSK9 meds seem to be very effective in lowering Lp(a) levels, and a few studies have suggested benefit for those with high Lp(a) levels (more definitive studies on them and new direct Lp(a) inhibitors are on the horizon: see https://gmodestmedblogs.blogspot.com/2025/06/high-lpa-increases-risk-of-recurrent.html )

-- the most recent American Heart Association guidelines suggest checking lipids in patients on statins only once to make sure the lipid levels are okay, and more times if we suspect issues on nonadherence to meds. there are several studies that statin adherence is really terrible (around the 50% mark), with less adherence in primary prevention and more in those who had a cardiovascular event (but still not a whole lot better than 50-60% in the latter, even just 1 year after an MI). But given this level of nonadherence, medication fatigue, etc, i would argue that we check lipids in almost everyone every year (sooner in some) just to make sure that these individuals really benefit from the meds we have for lipid lowering are being taken

    -- the most recent AHA guideline update in 2025 does comment that the lower the LDL the better but has a target of 55 mg/dL (as in the Swiss guidelines above), commenting that checking Lp(a) and apoB may be useful to "refine cardiovascular risk assessment", but no comment on regularly checking lipid levels in most patients, nor recommendations regarding LDL targets in those with high Lp(a) or apolipoprotein B levels: https://www.jacc.org/doi/10.1016/j.jacc.2025.04.024

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them