Magnitude of Low-density Lipoprotein Reduction and Impact on Major Cardiovascular Outcomes.

Less is more (?) Ludwig Mies van der Rohe Atherosclerotic cardiovascular (CV) disease is a primary worldwide cause of morbidity and mortality.1 Its association with elevated levels of low-density lipoprotein (LDL) cholesterol (LDL-C) has been broadly established. Therefore, cholesterol-lowering drugs (eg, statins, ezetimibe, and proprotein convertase subtilisin/kexin 9 [PCSK9]-inhibiting monoclonal antibodies) are massively used in primary and secondary prevention of cardiovascular events.2 Systematic reviews and metanalysis on LDL-C–lowering therapies have found an association between the absolute reduction of LDL levels and the reduction of major vascular events (composite of CV death, acute coronary syndromes, revascularization, and stroke). In particular, the 5-year incidence of major coronary events, revascularization, and stroke is reduced by one-fifth per mmol/L of LDL-C reduction. However, achieved LDL level seems to be associated with a reduction of the risk of coronary death and myocardial infarction (MI).3,4 Accordingly, the latest European guidelines recommend reaching low levels of LDL-C, especially for very high-risk (LDL-C <55 mg/dL) and high-risk (LDL-C <70 mg/dL) patients, together with a reduction of at least 50% of LDL-C from the baseline.5,6 Absolute reduction of LDL levels and achieved LDL level may not be the only variables to consider when LDL-C–lowering therapies are used because other parameters may help in predicting the clinical benefit of the treatment. Notably, Navarese et al7 recently published a meta-analysis suggesting that the greatest benefit from LDL-C–lowering therapy may occur in patients with baseline LDL-C levels of 100 mg/dL or greater. In the current issue of the Journal, Ennezat et al8 aimed at appraising, in an interesting and extensive meta-analysis, the association between changes in LDL with lipid-lowering agents and clinical outcomes, using refined meta-regression methods (Fig. 1). The authors tried to appraise whether the percentage of LDL reduction is associated with the effect size of all cause and CV mortality (primary endpoints). The effect size on MI, stroke, and non-CV death (secondary endpoints) were also assessed. The impact of other variables such as absolute LDL reduction, baseline LDL levels, achieved LDL levels, and CV mortality rate of the population studies on the aforementioned endpoints was also assessed. The analysis was performed on 60 selected randomized controlled trials with a mean follow-up of at least 52 weeks. A total of 323,950 subjects were included. Patients enrolled in the studies presented unknown or known atherosclerotic CV disease. Furthermore, some of them showed other comorbidities that were not consistently represented in all the trials (ie, left ventricular dysfunction, kidney disease or aortic stenosis, rheumatoid arthritis, or chronic obstructive pulmonary disease). Hence, those variables were not included in the analysis, representing a potential limitation. The comparison focused on lipid-modifying agents (statins, ezetimibe, and PCSK9-inhibiting monoclonal antibodies) versus placebo/standard/usual care and between intensive versus less intensive LDL-C–lowering therapies. The effect of hypolipemic therapy on the endpoints has been assessed in absolute rate differences (ARD), number needed to treat (NNT), and rate ratios as secondary analyses. NNT and ARD allow a better appreciation of the clinical relevance of the treatment benefit, as the greater the absolute benefit, the smaller the NNT. Rate ratios, like hazard ratios, however, may overstate the benefit of the treatment in low event rates population while NNT increases. In general, all the analyses showed a high annual NNT. Mortality (all-cause and CV mortality) was overall reduced by lipid-lowering therapy, although it seems like it was neither related to the relative (expressed in percentage) or absolute LDL reduction nor to baseline and achieved LDL-C levels. The effect on mortality was not evident in trials where more than 50% reduction of LDL-C occurred and in the ones with low baseline LDL-C levels. MI risk was overall reduced, mainly in trials with high baseline LDL levels and with an important magnitude of LDL-C reduction (relative: 50%; absolute: 65 mg/dL). Indeed, those were the characteristics present in the studies that presented the lowest NNT. Conversely, the benefit on stroke risk was not evident with such important reduction in LDL. Interestingly, it was found that achieving LDL <55 mg/dL was not associated with further significant reduction in the primary endpoints. However, a prespecified secondary analysis of 25,982 patients from the randomized FOURIER trial reported a monotonic relationship between LDL levels and major CV outcomes.9 It should be considered that the heterogeneity of the clinical setting of the trials, as the unknown presence or absence of other comorbidities in all the studies, represents a noteworthy limitation: these confounding factors may affect the outcomes. Moreover, this systematic review and meta-analysis used studies where hypolipemic drugs were used for primary and secondary prevention, gathering 2 populations with different baseline risk. It is difficult to establish only one parameter (baseline LDL, achieved LDL, percentage reduction, absolute reduction, or CV mortality rate of the population studies) for predicting the reduction of all major events, and it must be borne in mind that residual individual risk still plays an important role in causing future events, even under hypolipemic treatment.4 In this regard, beside the burden of non-LDL atherogenic lipoproteins (eg, apoB-containing particles, triglyceride rich lipoproteins, and lipoprotein (a)), inflammation has been demonstrated, with the CANTOS trial, to play a key role in this context.10,11 Indeed, this phase III study, proved that canakinumab (an anti-interleukin-1β antibody) was effective for the secondary prevention of CV events in subjects who had a previous MI and who had persistent pro-inflammatory response (ie, high-sensitivity C-reactive protein levels of ≥2 mg/L) despite the use of other aggressive secondary preventive strategies, without affecting LDL levels.11 Nevertheless, it seems to be not a universally cost-effective solution.12 On top of that, other trials investigating the benefit of colchicine for secondary prevention confirmed the hypothesis of an inflammatory-related CV risk burden.13,14 Differently from canakinumab, its use has been included in the latest guidelines of cardiovascular prevention.15FIGURE 1.: Impact of LDL cholesterol (LDL-C) lowering on CV outcomes. Pt, patients; RCTs, randomized controlled trials.Interestingly, the reduction of MI seems to be associated with a more intensive treatment, hence with a lower achieved LDL and a greater magnitude of the reduction. It may be a consequence of plaque stabilization induced by these drugs, which, by reducing the lipidic content and increasing the thickness of the fibrous cap, reduce the risk of plaque rupture and subsequent thrombosis.16,17 In an era of precision medicine, the need to identify which subgroup of patients would benefit the most from LDL-C levels reduction is clinically and economically relevant. Indeed, we could design an intensive and strict therapeutical path for these patients and spear the administration of costly drugs to subjects who would not be aided by them.18