Changing the Paradigm for Post-MI Cholesterol Lowering From Intensive Statin Monotherapy Towards Intensive Lipid-lowering Regimens and Individualized Care

Kausik K. Ray


Eur Heart J. 2021;42(3):253-256. 

Patients with acute coronary syndromes (ACS) including myocardial infarction (MI) are the very definition of individuals in whom gene–environment interaction over their lifetime will result in the most serious manifestations of atherosclerosis and who remain at greatest risk of future cardiovascular events. Lowering LDL-cholesterol (LDL-C) with high-intensity lipid-lowering therapies (LLTs) initiated within 10 days of an ACS reduces risk more than lower intensity regimens.[1,2] Lowering elevated LDL-C through the addition of PCSK9 (proprotein convertase subtilisin/kexin type 9) monoclonal antibodies to high-intensity statins between 2 and 3 months after an ACS is superior to high-intensity statins alone.[3] European guidelines have therefore updated recommendations for patients with atherosclerotic cardiovascular disease (ASCVD) including recent ACS, advocating that both a 50% lowering and an LDL-C below 1.4 mmol/L should be achieved, in a step-wise fashion, starting with statins and then through addition of non-statin LLTs if needed.[4] However, these guidelines do not explicitly recommend a time frame in which to achieve these reductions. Furthermore, during the past decade, secondary prevention has not moved beyond a statin-centric monotherapy approach, with use of combination therapy with ezetimibe or a PCSK9 inhibitor low at 9% and 1%, respectively, with only 18% of patients with ASCVD achieving an LDL-C <1.4mmol/L in the Da Vinci registry.[5]

It is against this backdrop that the study by Schubert et al. from the SWEDEHEART registry in this issue of the European Heart Journal provides a welcome reminder of the importance of lowering LDL-C early after MI.[6] The authors assessed the relationship between changes in LDL-C between admission for an MI and at weeks 6–10 post-MI during cardiac rehabilitation (CR) with the subsequent risk of various cardiovascular events and mortality. Among 40 607 patients over a median follow-up of 3.78 years, those achieving the largest reductions in LDL-C (categorized by quartiles) or percentage reduction in LDL-C had the lowest risk of a range of cardiovascular events and all-cause mortality. In part, those achieving the greatest percentage reductions and thus the greatest absolute reductions are likely to be more adherent to medications and more healthy (healthy user bias), which are unmeasured variables and thus may contribute to the observations. Whilst these observational data[6] do not demonstrate causality, they reinforce the well-established benefits of LDL-C lowering from randomized trials and meta-analyses. Beyond the headlines of lower is better, larger percentage reductions are better, and earlier may be better, a closer look at the characteristics of this cohort provides important insights into the woeful gap between current clinical practice and the totality of evidence. Standards of cardiovascular care are high in Sweden but, even in this bastion of good clinical practice, there are important lessons to be learnt that could inform future practice.

First, ~77% were statin naïve at the time of the index event, which are similar to proportions observed in clinical trials. Given the use of statins in primary prevention, the obvious conclusion is that the 77% of patients at risk of MI either do not take or are not offered statins, or are intolerant, and suitable replacements have not been offered or considered. Hence we must improve identification and treatment of those most at risk of MI in the primary prevention setting.

Secondly, patients who had a MI but were statin naïve were very different from those on statins prior to their MI in terms of clinical characteristics and their LDL-C management, which probably impact outcomes. Comparing statin-naïve patients with those on statins, LDL-C levels were higher (3.4 mmol/L vs. 2.2 mmol/L) and on-treatment LDL-C levels at CR visit were lower (1.9 mmol/L vs. 2.1 mmol/L). Using multivariable models, a 1 mmol/L lowering of LDL-C was associated with the broadly one-fifth lowering in risk of six individual endpoints plus major adverse cardiac events (MACE) among statin-naïve patients as would be predicted from trials. In contrast, using the same multivariable adjustment, only one of the six endpoints was lower (stroke), with no difference in MACE per unit LDL-C reduction among those on statins prior to MI. As unconfounded trials of statins demonstrate that the relative benefit per mmol/L lowering is constant and trials of addition of MAbs to statins show identical benefit, the apparent lack of benefit among those on prior statins suggests that this part of the data are likely to be confounded.[7,8] Furthermore, very few patients in this group would have demonstrated a 1 mmol/L decrease as the average on-treatment LDL-C fell by 0.1 mmol/L. Perhaps the biggest take-home message is that prior statin use probably identifies sicker patients with multiple comorbidities, with a higher likelihood of prior ASCVD, and who we know derive greater absolute benefits from add-on non-statin LLT.

Thirdly, in analyses by quartiles of LDL-C reduction, it follows that those in quartile 1 with an average LDL-C reduction <0.36 mmol/L consist largely of those receiving prior statins (58%) vs. those in quartile 4 (4%). Among quartile 1 vs. quartile 4 (on average > 1.85 mmol/L LDL-C reduction), there was a higher prevalence of diabetes (32% vs. 12%), prior MI (27% vs. 4%), prior revascularization (32% vs. 4%), prior heart failure (6% vs. 1%), and prior stroke (7% vs. 2%), suggesting that those in quartile 1 were much sicker. Moreover, comparing mean LDL-C levels on treatment, LDL-C levels rose from 2.1 mmol/L to 2.3 mmol/L in quartile 1 but fell from 4.3 mmol/L to 1.8 mmol/L in quartile 4. Overall, of patients on low- or moderate-intensity statins at admission, if this was changed to high-intensity statins, LDL-C reductions of 0.6 mmol/L were observed, whereas if there was no change in therapy, levels only fell by 0.1 mmol/L. Use of ezetimibe at the index event was 1%, increasing to the dizzy heights of 7% after 10 weeks and use of MAbs non-existent. Taken together, these data suggest that there is insufficient treatment intensification of those on prior statins therapy. Among those initiated on therapy but experiencing a less than predicted reduction in LDL-C (42% of the cohort), adherence to oral daily medications or side effects from medications may contribute to the observed lack of change.[9]

Finally, our current approach to MI care is procedure driven, treating disease instead of preserving health. CR is greatly undervalued, and the present study demonstrates the important role CR can play in the long-term management of the post-MI patient through assessment and reinforcement of lifestyle, medication adherence, and personalizing care. Current approaches to LLT are very much focused on small molecules which require daily dosing. It is imperative, therefore, that patients understand the importance of adherence, which tracks with long-term LDL-C exposure and future risk.[10] When non-adherence and suboptimal LLT prescriptions are coupled at a population level, this represents a significant health burden.[10] Assuming modest between-person biological variability in LDL-C percentage reduction, the greatest contributor to interindividual variability is likely to be non-adherence.[9] In analyses stratified by low-/moderate-intensity statins vs. high-intensity stains, the highest event rates were observed among those with an LDL-C increase irrespective of statin regimen, with a marked higher risk difference compared with the two other groups (<50% or >50% reduction), which were generally more similar. This reinforces the notion that the medications only work if you take them. The greatest strength of real-world data is perhaps their ability to highlight at scale the shortfalls in public health/implementation, helping to shape future practice, by forcing us to ask ourselves the basic question: is what we are doing sensible and good enough?

Cardiovascular risk reduction is agnostic to how LDL-C is lowered.[7,11] What matters is the absolute reduction and the duration of that reduction, hence the concept of cumulative exposure to the time-averaged LDL-C,[12] which are functions of adherence (patients) and the treatment intensity prescribed (doctors). Statins, ezetimibe, and PCSK9 inhibitors reduce LDL-C and cardiovascular events with apparent differences in results reconciled through one unifying paradigm; namely that a 1 mmol/L absolute reduction in LDL-C provides approximately a one-fifth reduction in the risk of cardiovascular events with about half the benefit in year 1 of treatment irrespective of approach.[7] New therapies such as bempedoic acid[13] and inclisiran[14] have been or are undergoing the process of approval for LDL-C-lowering indications with ongoing outcomes trials. These treatment options offer a unique opportunity to improve implementation of guidelines, taking into account current costs and the unique characteristics of each therapy to bring the population-mean LDL-C closer to 1.4 mmol/L (Graphical abstract). Among those with a MI and with prior ASCVD or on oral LLT (probably LDL-C >2 mmol/L), add MAbs pre-discharge (also maximize statins). For those with statin intolerance, add ezetimibe and MAbs. For all others, maximize statins and add ezetimibe pre-discharge, thus increasing the absolute LDL-C reduction in year 1, with a resultant greater relative risk reduction during the highest risk period (as the benefit per 1 mmol/L lowering is half of that in later years). At CR visit for patients with LDL-C reductions <50% and LDL-C >1.4 mmol/L, reinforce adherence to LLTs (probably poorly adherent), with inclisiran a potential add-on to overcome non-adherence and achieve goals. For those with >50% lowering but LDL-C >1.4 mmol/L, potential options based on cost, patient preference, and availability, include MAbs, bempedoic acid, or inclisiran. A pragmatic approach such as this would distribute costs of medications appropriately to those at the highest risk[15] and address unmet needs, and could lead to better attainment of guideline recommendations (Graphical abstract).

Graphical abstract

Improving population level LDL-C goal attainment through a personalised approach to intensive LDL-C lowering. Panel (A) implications for the physician. Pathways for optimising LDL-C goal attainment. Step 1 is to stratify patients based on easily identifiable drivers of residual risk by prior history of ASCVD or prior oral LLT at admission. For these patients optimise statin dose (intensity) and add MAb pre-discharge. Step 2 is for those with a history of statin intolerance and not on any oral LLT at admission add ezetimibe and a Mab pre discharge. Steps 1–2 account for about one quarter of patients. Step 3 all other patients (those statin naive prior to admission) receive high intensity statins plus ezetimibe. Step 4 personalise additional add on LLT based during cardiac rehabilitation based on response to statins plus ezetimibe. For those where non-adherence is the biggest driver of failure of goal attainment consider inclisiran. For the remainder consider between bempedoic acid, MAbs and inclisiran based on cost effectiveness, availability and patient preference. Panel (B) implications for the payer. The population distribution of risk post MI. Step 1, move from a high-intensity generic statin based approach for all to a high-intensity LLT based approach first through use of generic high-intensity statins plus generic ezetimibe for all. Step 2, personalise use of novel therapies with cost differentials ranging from $$$ high cost to $ lower cost for MAbs, inclisiran and bemepedoic acid based on drivers of residual risk, distributing cost and matching need. Panel (C) implications for public health. Moving from a high-intensity statin approach for all to a high-intensity statin plus ezetimibe approach for all increases the proportion of patients attaining goals and reduces residual risk compared to current approaches. For the remainder personalising LDL-C as suggested in Panel (A), results in additional patients attaining LDL-C goals and reducing population residual risk further. ASCVD, atherosclerotic cardiovascular disease; Eze, ezetimibe; HI, high-intensity; LDL, low density lipoprotein; MAb, monoclonal antibody; MI, myocardial infarction.