Emerging views of statin pleiotropy and cholesterol lowering

Abstract

Over the past four decades, no class of drugs has had more impact on cardiovascular health than the 3-hydroxy-methylglutaryl coenzyme A reductase inhibitors or statins. Developed as potent lipid-lowering agents, statins were later shown to reduce morbidity and mortality of patients who are at risk for cardiovascular disease. However, retrospective analyses of some of these clinical trials have uncovered some aspects of their clinical benefits that may be additional to their lipid-lowering effects. Such ‘pleiotropic’ effects of statins garnered intense interest and debate over its contribution to cardiovascular risk reduction. This review will provide a brief background of statin pleiotropy, assess the available clinical evidence for and against their non-lipid-lowering benefits, and propose future research directions in this field.

1. Introduction

Coronary heart disease (CHD) continues to be the leading cause of death in adults, accounting for one-third of all deaths in 2015.¹ The 3-hydroxy-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors or statins play a crucial role in preventing and reducing cardiovascular disease. The emergence of statins followed decades of accumulating evidence that established the causal link between cholesterol and cardiovascular mortality, including epidemiological data from the Seven Countries Study, Framingham Study, and MRFIT trial^2–4 that recognized low-density lipoprotein cholesterol (LDL-c) as the primary risk factor for CHD.

The first statin, mevastatin (Compactin; ML-236B), was isolated from fungal species Penicillium citrinum by Endo et al.⁵ in 1976. This was followed by isolation of mevinolin from Aspergillus terreus⁶ and its clinical and commercial success as lovastatin. As of 2020, at least nine statins with a range of structures and pharmacologic parameters have been developed, of which seven are currently approved in the United States (Figure 1⁷) while one (cerivastatin) was withdrawn from market.⁸ All statins competitively bind to HMG-CoA reductase’s enzymatic site, and the Ki of statins are generally in the nanomolar range.⁹ This inhibits the rate-limiting step of cholesterol biosynthesis, the conversion of HMG-CoA to mevalonate (Figure 2), and leads to decreased hepatic cholesterol production, upregulation of LDL receptor density on hepatocyte, and increased serum LDL-C clearance. In subsequent clinical trials, the LDL-lowering effects of statins have been shown to reduce cardiovascular events.

Figure 1.

Chemical structures of statins of major historical and clinical significance and their classifications by marketing generation and key chemical characteristics.

Figure 2.

Flow chart of cholesterol biosynthesis and key enzymatic steps that are targeted by key classes of lipid-lowering drugs, with the statins’ target sites broken down by canonical lipid-lowering and pleiotropic effects.

2. Early clinical evidence suggesting statin pleiotropy

Several interesting observations from statin trials led to speculations that statins might act beyond LDL-lowering in achieving their clinical efficacy, a phenomenon later termed ‘statin pleiotropy’. First, the magnitude of CHD reduction (30–40% in all observed MIs) was somewhat greater than what could be anticipated by statins’ LDL lowering (25% in WOSCOPS and AFCAPS/Tex-CAPS trials). The time to benefit was also noted to be more rapid compared to other classes of lipid-lowering agents available at the time (e.g. fibrates in VA-HIT trial,¹⁰ niacin in CDP trial,¹¹ and cholestyramine in LRC trial¹²), with the Kaplan–Meier curves starting to diverge by 1 year in the WOSCOPS, AFCAPS/Tex-CAPS, HPS, and TNT trials.^13–19 In PROVE-IT TIMI-22 trial, the benefit of high-dose atorvastatin emerged as early as 30 days²⁰ and in the MIRACL trial that enrolled 3086 ACS patients, aggressive atorvastatin therapy led to measurable clinical benefit as early as at 16 weeks.²¹ In addition, statins were consistently shown to be effective in the primary prevention of strokes across all major trials (37% in 4S trial,¹⁷ 31% in CARE trial,¹⁸ 25% in TNT trial,¹⁶ and 25% in HPS trial¹⁵). This is a notable finding given that large observational studies failed to show convincing association between ischemic stroke and elevated LDL-C levels.^22–24 Finally, in the secondary prevention SPARCL trial, which enrolled 4731 patients with recent strokes and TIAs but no known CHD, atorvastatin 80 mg daily decreased LDL-C by 53% and resulted in a 16% decrease in total stroke at 5 years.²⁵ Statin’s quicker and disproportionally greater cardiovascular effect, coupled with its unexpectedly strong effect on cerebrovascular outcomes, provided the initial clinical suggestion of statin pleiotropy.

3. Potential mechanisms of statin pleiotropy: Plaque stability and vascular inflammation

Several potential mechanisms were initially proposed to explain statins’ apparent pleiotropic effects. One early hypothesis was that statins stabilized atherosclerotic plaque and disproportionally reduced the incidence of clinical ischemic events. This was first investigated by direct imaging of coronary vessels with intravascular ultrasound (IVUS) in the REVERSAL, ASTEROID, and SATURN trials, which demonstrated slowing and regression of atherosclerotic plaques.^26–28 The salutary changes in plaque burden occurred by 18–24 months, in line with the observed time to efficacy in most statin outcome trials, but the degree of regression was rather small at around 1% as measured by percent atheroma volume on IVUS. This suggests that in the high-risk patient cohort with the most derivable benefit, plaque burden alone was inadequate to account for changes in the incidence of clinical cardiovascular events. Subsequent studies using newer imaging technologies confirmed that atherosclerotic plaque composition likely play a much bigger role in determining plaque vulnerability. For example, the EASY-FIT study employed optical coherence tomography to show that patients on higher intensity atorvastatin led to thicker fibrous cap in coronary plaques,²⁹ while the much larger multinational PARADIGM study followed 1255 patients longitudinally with serial coronary computed tomography angiography and showed that statin therapy resulted in not only slower progression of atherosclerosis volume but also concomitant increased plaque calcification and reduction in high-risk plaque features.³⁰ Such findings have been coupled with animal studies that statin can alter smooth muscle and collagen content of atherosclerotic plaques,³¹ increase plaque calcification,³² and reduce matrix metalloproteinase production and cap degradation^33,34 by mechanisms that are independent of cholesterol lowering.

As it became increasingly evident that statin therapy led to plaque stabilization and regression,^35,36 vascular inflammation became the next focus of investigation. Hyperactivation of the cellular and humoral immune systems with increased reactive oxygen species generation can lead to a pro-inflammatory cascade, including cytokine release and T lymphocyte and macrophage recruitment and activation, which together may predispose to accelerated atherosclerosis and plaque vulnerability.³⁷ A number of studies showed association between the cardinal inflammatory markers C-reactive protein (CRP) and increased cardiovascular risk.^38,39 Interestingly, statins reduced both short-term and long-term CRP levels by 14% in the PRINCE trial at 12 weeks,⁴⁰ by 34% in the MIRACL trial at 16 weeks,⁴¹ and by 38% in the CARE trial at 5 years.⁴² The significance of modulating vascular inflammation was highlighted in the JUPITER trial, in which rosuvastatin 20 mg daily reduced LDL by 50% and CRP by 37% at 1.9 years of follow-up and resulted in a decrease of 44% in primary endpoints (occurrence of a first major cardiovascular event), 54% in all MI, and 20% in all-cause mortality.⁴³ The degree of cardiovascular benefit exceeded projections from earlier trials based on LDL-C lowering alone,⁴⁴ suggesting that statin’s anti-inflammatory effects might account for the difference in efficacy. This hypothesis is corroborated by studies showing that statins can inhibit in vitro activation of several pro-inflammatory transcription factors including NFκB, AP-1, and HIF-1α⁴⁵ and to alter the balance of T-cell differentiation by blunting proinflammatory IL-17 helper T cells while promoting the FoxP3-expressing regulatory T cells (Tregs) that induce immune tolerance.⁴⁶ Furthermore, several adhesion molecules including integrins, selectins, PECAM-1, ICAM-1, and VCAM-1 were shown to be affected by statin treatment in mediating leukocyte-endothelium adhesion and transmigration.^47,48

Recently, three large trials (CANTOS, COLCOT, and CIRT) studied the clinical outcome of reducing inflammation without altering LDL-C level. In CANTOS, treatment with canakinumab (an interleukin-1β antibody) reduced CRP by 26–41% and the composite primary endpoint by 15%, MI by 24%, and stroke by 26%. In COLCOT, colchicine lowered the primary endpoint by 23% and cardiovascular death by 16%.⁴⁹ In CIRT, treatment with low-dose methotrexate failed to lower either CRP or the composite primary cardiovascular endpoint.⁵⁰ While these trials suggest that inflammation modulation influenced cardiovascular outcomes and might theoretically contribute to statin pleiotropy, it should be noted that Mendelian randomization models showed CRP not to be a direct causal factor,^51,52 and a meta-analysis of 24 trials failed to show correlation of magnitude of CRP reduction to cardiovascular risk reduction.⁵³ Therefore, it remains unclear whether statins’ anti-inflammatory effects are independent mediators vs. confounding factors in cardiovascular event reduction. In addition, even assuming vascular inflammation to be an independent causal factor of cardiovascular events, it remains debatable whether statins reduce inflammation independent of LDL-C, which itself could lower oxidized LDL (oxLDL) in the atherosclerotic plaque, reduce macrophage and platelet activation,⁵⁴ and ultimately contribute to reduction in inflammation.

4. Bedside to bench side: Statins’ effect on prenylation and beyond

An important reason for the growing interest in statin pleiotropy is the experimental data supporting statins’ lipid-independent cellular effects. One proposed mechanism is thought to be due to statin’s effect on modulating protein prenylation. Mevalonic acid, the intermediate metabolite targeted by statins, is the shared precursor for biosynthesis of isoprenoids, which are essential for the production of 15-carbon farnesyl pyrophosphate (FFP) and 20-carbon geranylgeranyl pyrophosphate (GGPP) (Figure 2). FFP and GGPP in turn are substrates for post-translational modification of Rho superfamily of small GTPases (e.g. Ras, Rho, Rab, and Cdc42), which facilitate their cell membrane trafficking, localization, and signalling.⁵⁵ Statins have been shown to inhibit Rho-family protein signalling in animal models and humans.^56–61 In particular, inhibition of Rho by statins leads to the upregulation of eNOS,⁶² inhibition of vascular reactivity,⁶³ attenuation of leukocyte adhesion,⁶⁴ mobilization of endothelial progenitor cells from bone marrow, and reendothelialization after vascular injury.^65–67 In addition to endothelial cells, statins can target other cell lineages by promoting cell-cycle arrest in fibroblast,⁶⁸ attenuating vascular smooth muscle cell proliferation,^69,70 and decreasing platelet reactivity.^71,72 The proposed molecular mechanisms of these effects include both Rho/ROCK signalling and a variety of additional pathways such as inhibition of thromboxane biosynthesis,^73,74 modulation of cytosolic calcium concentration,⁷⁵ PECAM-1-mediated PI3K signalling,⁷⁶ and PPAR signalling.^77,78 The observed regulatory effects on smooth-muscle proliferation by various statins^79,80 are of particular importance in preventing cardiac allograft vasculopathy in post-transplant patients.^81,82

5. Challenging the statin pleiotropy hypothesis with ezetimibe and PCSK9 inhibitor trials

Despite the myriad of experimental data supporting statin pleiotropy, it remains difficult to definitively prove that these in vitro findings translate to clinical significance in humans. Importantly, it is nearly impossible to prove that statin’s observed effects on plaque stability, endothelial dysfunction, and vascular inflammation occur independently of lipid-lowering instead of simply being its secondary downstream effects. One historic argument pointed to the relatively poor cardiovascular efficacy displayed by older, non-statin lipid-lowering agents to highlight the mechanistic uniqueness of statins. This paradigm was first challenged by the IMPROVE-IT trial, in which ezetimibe 10 mg on top of simvastatin 40 mg resulted in 17% additional LDL-C lowering compared to simvastatin 40 mg plus placebo and resulted in 6.4% reduction in the composite endpoint and 13% reduction in all MI at a follow-up of 7 years.⁸³ While the lipid-lowering effect of ezetimibe was modest, it nonetheless illustrated that improvement in cardiovascular endpoints was achievable by a non-statin lipid-lowering agent.

The advent of PCSK9 inhibitors, by far, the most potent non-statin lipid-lowering drug, provided another opportunity to test statin pleiotropy. To date, four large outcome trials using PCSK9 monoclonal antibodies have been published. The pair of studies using bococizumab (SPIRE-1 and SPIRE-2, with different LDL-C entry levels) was plagued by the development of high titres of anti-drug antibodies, ultimately leading to premature study termination. Nonetheless, they still showed 56% LDL-C reduction at 14 weeks on the background of 93% statin use and resulted in decrease of 21% in composite primary endpoints and 24% in non-fatal MI at 12 months in the higher-LDL, longer-duration SPIRE-2 (LDL-C ≥ 100 mg/dL).⁸⁴ In the FOURIER trial that enrolled 27 564 patients with CAD and LDL-C ≥ 70 mg/dL on maximal dose of statin, evolocumab led to 59% lower LDL-C compared to placebo and resulted in a decrease of 15% in composite primary endpoint and 27% in myocardial infarction at 2.2 years.⁸⁵ In the ODYSSEY OUTCOMES trial that enrolled 18 924 patients who had ACS within the preceding 1–12 months and one of several elevated lipid parameters while on maximal dose of statin, alirocumab led to 63% initial relative lowering of LDL-C vs. placebo and resulted in a decrease of 15% in the composite primary endpoint, 12% in any coronary events, and 15% in all-cause mortality at 2.8 years.⁸⁶

The results from the PCSK9 inhibitor trials rekindled debates over the importance of statin pleiotropy: substantial LDL-lowering beyond that of maximal intensity statin was shown to be achievable and correlated with further cardiovascular event reduction. Additionally, several early features of statin efficacy underlying the conception of statin pleiotropy (time to benefit, plaque regression, and stroke reduction) were replicated by PCSK9 inhibitors. The time to benefit was about 1 year in the FOURIER trial and 2 years in the ODYSSEY OUTCOMES trial, comparable to these observed in the WOSCOP and AFCAPS/Tex-CAPS trials. The burden of atherosclerotic plaques was assessed in the GLAGOV trial, in which evolocumab lowered LDL-C by 61% and induced greater reduction of atheroma volume than statin alone on serial IVUS.⁸⁷ Finally, the incidence of ischemic stroke was reduced by 21 and 27% with PCSK9 inhibition in the FOURIER trial and the ODYSSEY Outcomes trials, respectively, comparable to the degree of stroke reduction in most statin trials. Together, these findings demonstrate that the several clinical characteristics suggestive of pleiotropy are no longer unique to statins. One meta-regression analysis with data from statin and non-statin trials including PCSK9 inhibitors was able to model changes in stroke incidence entirely from total cholesterol changes, leading the authors to claim that there is ‘no longer room for pleiotropic effects of statin’.⁸⁸ However, it should be pointed out that many of the outcome results of PCSK9 inhibitors were achieved on top of statin therapy, and it is possible that PCSK9 inhibitors or further lipid lowering could potentiate or augment statin pleiotropy.

6. Pleiotropic effects or confounders?

A major challenge in assessing statin pleiotropy is the potential presence of multiple confounding factors. Mendelian randomization studies using genome-wide lipid-associated single-nucleotide polymorphisms allowed re-analysis of composite large randomized control trial data for direct association between lipid-lowering and cardiovascular outcomes. Using an adapted Egger technique, Labos et al.⁸⁹ showed that each 1 mmol/L of LDL-C change from statin therapy was associated with a hazard ratio of 0.77 in cardiovascular endpoints, with an intercept indistinguishable from zero, which suggested that statins’ cardiovascular benefits were entirely derived from LDL-C lowering. Similar analysis on stroke outcomes revealed more heterogenous findings, with genetically mediated LDL-elevation associated with increased risk of ischemic and large artery atherosclerotic strokes but not with small artery occlusion or cardioembolic strokes.⁹⁰ Another strike against statin pleiotropy is a Mendelian randomization analysis that suggested that CRP was not a direct causal factor in cardiovascular risk reduction.⁵¹

7. Comparative statin pleiotropy beyond traditional cardiovascular events

The well-established epidemiologic association between LDL-C and atherosclerotic cardiovascular events makes it difficult to untangle statin’s LDL-lowering and pleiotropic effects. Venous thromboembolism (VTE), a non-classic cardiovascular event, provides an intriguing vantage point. In the JUPITER trial, the incidence of VTE was a pre-specified secondary endpoint and shown to be reduced by 43% in the rosuvastatin arm compared to placebo.⁹¹ In the FOURIER trial, treatment with evolocumab also reduced incidence of VTE by 29%, and the reduction remains statistically significant at 31% when results from FOURIER and ODYSSEY Outcomes trials are combined.⁹² These findings are somewhat surprising, given the lack of clear epidemiologic associations between VTE incidence and traditional lipid parameters such as LDL-C and HDL-C.^93–96 Analysis of secondary biomarkers revealed several interesting distinctions: while rosuvastatin lowered inflammatory marker CRP by 37% in the JUPITER trial, PCSK9 inhibitors had no effect on CRP in the trials to date;^87,97,98 on the other hand, in the FOURIER trial, evolocumab ameliorated several non-traditional lipid parameters including lipoprotein(a) (Lp(a)),^85,99 while in the JUPITER trial, rosuvastatin had no effect on the median Lp(a) level.¹⁰⁰ These findings suggest that while both statins and PCSK9 inhibitors can achieve profound LDL-C lowering, they exhibit differential effects towards the spectrum of secondary pro-atherosclerotic, pro-thrombotic serum markers. While nuanced, such differences may nonetheless carry physiologic significance and determine pleiotropic effects of statins vs. PCSK9 inhibitors. The exact mechanism behind this divergence is unclear but may be related to their distinct sites of action: statin targets cholesterol synthesis and could directly deplete intracellular cholesterol storage, while PCSK9 inhibitors increase cell-surface LDL receptor density and clear serum LDL particles without disrupting intracellular cholesterol synthesis or sterol flux, a process that has been linked to regulation of inflammatory and immune responses.¹⁰¹ Inclisiran, an siRNA-based drug that inhibits PCSK9 synthesis,¹⁰² and bempedoic acid, a small molecule inhibitor of ATP-citrate lyase (ACL) upstream of HMG-CoA synthetase in the cholesterol biosynthetic pathway,^103,104 are novel players in the lipid-lowering arena; results from their anticipated outcome trials could in turn shed important insight on whether VTE reduction and CRP/Lp(a) alterations are drug-specific effects vs. class effects of lipid-lowering.

8. Comparison of CV event reduction between statins and non-statins

The percentage outcome reduction (∼15%) achieved by PCSK9 inhibitors was relatively modest compared to their potent LDL-C reduction (∼40–60%), while the early statin trials showed at least proportional, if not supra-proportional reduction in cardiovascular events relative to their LDL-C lowering effects. Given difficulty in head-to-head comparison between trials due to different and evolving primary endpoints, an attempt is made here to calculate the ratio of % difference in major coronary event rate per absolute LDL-C difference in major secondary prevention trials (Table 1). Using this mathematical estimate, the average reduction of events for statins is 23% per 40 mg/dL LDL-C lowering, higher than ezetimibe (19%) and PCSK9 inhibitors (12%). A summary of statin vs. non-statin trials is further summarized in Table 2. In addition, when the percent reduction in major coronary event rate is plotted against percent LDL-C lowering, the statin versus non-statin trials do not fall along the same line, but instead form two distinct lines (Figure 3). Indeed, the ezetimibe/PCSK9 trials fall on a much less sharp incline than statin trials due to relatively less event reduction compared to their lipid-lowering potency. Again, these estimates are crude at best, require multiple assumptions, and do not account for difference in follow-up time (∼5 years for statins, 7 years for ezetimibe, and 2–3 years for PCSK9 inhibitors). Furthermore, the interpretation should be cautioned that PCSK9 inhibitor therapy was instituted on the background of maximally tolerated statin therapy, and it is possible that a different linear relationship occurs in the hyper-LDL-C depletion range or with combination therapy. Nonetheless, such analysis suggests that statins may improve clinical outcomes more substantially than non-statin agents per unit of LDL-C lowering. The pending longer-term outcome results of PCSK9 inhibitor trials will provide a better comparison with respect to the same unit of time.

Table 1.

Comparison of major coronary event rates across secondary prevention trials with statins vs. non-statins

	Statin vs. placebo		High vs. low intensity statin			Statin/ ezetimibe vs. statin	Statin/PCSK9 inhibitor vs. statin
	4S	CARE	TNT	IDEAL	PROVE-IT	IMPROVE-IT	FOURIER	ODYSSEY OUTCOMES
Follow-up period (year)	5.4	5.0	4.9	4.8	2.0	7.0	2.2	2.8
Mean LDL-C differencea (mg/dL)	65	38	24	21	33	17	56	46 d
Mean % LDL -C difference	35%	28%	24%	20%	35%	24%	59%	48%
Absolute reduction in major coronary eventsb (%)	10%	3%	1.6%	1.1%	1.1%	1.7%	1.0% c	1.1%
% difference in major coronary events	33%	24%	20%	11%	13%	8%	18%	12%
% reduction in CV event/40 mg/dL of LDL-C reduction	20%	25%	33%	21%	16%	19%	13%	10%
Average by category	23% (statin trials)					19% (ezetimibe trials)	12% (PCSK9 trials)

^aMeasures difference in LDL between treatment and placebo arms at follow-up, NOT reduction at follow-up vs. baseline.

^bAs directly reported by trial, or if not directly reported, calculated as sum of CHD death and non-fatal MI unless otherwise described.

^cNot directly reported. Calculated retrospectively as sum of (cardiovascular death—stroke death) + (total MI—fatal MI).

^dAveraged from reported values at 4, 12, and 48 months (LDL-C values at other time points not reported).

Table 2.

Comparison of major cardiovascular clinical trials by therapy category

	Statins	Ezetimibe	PCSK9 monoclonal Ab	Inflammatory modulators	Bempedoic acid
LDL-C lowering	↓↓ to ↓↓↓ (depending on intensity)	↓	↓↓↓	–	↓
Primary vs. secondary prevention	Primary and secondary	Secondary	Secondary	Secondary	Pendinga
Compared vs. no statin?	Yes available	No	No	No	No
CRP reduction?	Yes	Yes	No	Yesb	Yes
Plaque reduction on IVUS?	Yes	N/A	Yes	N/A	N/A
Endothelial function improvement?	Yes	No	N/A	N/A	N/A
Stroke reduction?	Yes	Yes	Yes	Yes/Noc	Pending
Years to first CV endpoint improvement	1–2	7	2–3	∼2	Pending

^aOutcome trial (CLEAR Outcome) pending.

^bReduced in CANTOS and COLCOT trials but not in CIRT trial.

^cReduced in COLCOT but not CANTOS or CIRT trials.

Figure 3.

Graphic correlation of percent reduction in major coronary events to percent LDL-C lowering based on published data from major secondary prevention trials. Notably, statin trials are distributed along a steeper line than non-statin trials, suggesting contribution from statins’ pleiotropic effects on CV outcomes beyond pure LDL-C lowering.

9. Re-purposing of statin for other systemic inflammatory diseases

While it remains under debate whether part of statin’s cardiovascular benefit is derived from its non-lipid-lowering actions, substantial evidence has emerged to support its re-purposed use in certain non-cardiovascular diseases with no apparent link to hypercholesterolemia. In particular, given statins’ documented anti-inflammatory actions both in vitro and in clinical trials, they were explored as immune modulatory agents in systemic inflammatory diseases.¹⁰⁵ Two meta-analyses showed statins to attenuate disease activity of rheumatoid arthritis (RA) by lowering of serum inflammatory markers and symptom improvement,^106,107 while a large population-based nested case–control study showed reduced risk of RA with statin use.¹⁰⁸ Similarly, statins have been reported to confer beneficial effects in systemic lupus erythematosus,¹⁰⁹ periodontitis,^110,111 primary sclerosing cholangitis,¹¹² inflammatory bowel diseases,¹¹³ cognitive function/dementia,^114,115 and psychological well-being.¹¹⁶ The successful secondary application of statins in non-cardiovascular fields lends credence from a different dimension to their pleiotropic effects.

A perhaps timely re-application of statin therapy is in pulmonary medicine, in particular for re-purposed use during the current COVID-19 pandemic, which has infected over 48 million people globally and resulted in 1.2 million deaths as of 1 November 2020. In patients infected with SARS-CoV-2 virus, critical illness developed as a result of acute respiratory distress syndrome coupled with cytokine storm, leading to unchecked systemic hyperactivation of immune response that often prove fatal. In this context, statins are actively being explored for re-purposed application as an anti-inflammatory, cardiopulmonary protective agent in the fight against COVID-19.^117–119 We can extrapolate statin’s potential therapeutic effects against SARS-CoV-2 by prior reports of its clinical benefit in asthma¹²⁰ and inflammatory lung diseases¹²¹ and of statin’s direct interaction with SAR-CoV2 main protease in silico.¹²² While randomized clinical trials are needed to establish statin’s efficacy in this arena, given the present lack of approved targeted treatment, statins may be a safe, readily available added option in the globally fight against COVID-19.

10. Current and future perspectives regarding statin pleiotropy

Studies into statin’s expanded clinical use outside of cardiovascular diseases provided fresh evidence for statin pleiotropy. Nevertheless, the debate over its legitimacy in the realm of cardiovascular diseases still continues given conflicting data and competing interpretations from existing clinical trials. To help settle this debate, we propose three investigative approaches: (1) developing a ‘neo-statin’ that inhibits cholesterol synthesis without affecting its other cellular pathways such as prenylation; (2) comparing head-to-head benefits between statin and ezetimibe and/or PCSK9 inhibitors in a large clinical trial; and (3) creating tissue-specific HMG CoA reductase (HMGCR) knockout animal models and studying their cardiovascular outcomes.

Regarding the first approach, we would need a molecule that acts at later steps in the cholesterol biosynthetic pathways such as squalene synthetase inhibitor that spares the isoprenoid synthesis branch,⁵⁵ thereby dissecting out reduced prenylation from the drug effect. The squalene synthase inhibitors fit such requirements (see Figure 2), and one member lapaquistat acetate progressed as far as phase III studies but unfortunately was abandoned due to mediocre potency in LDL-C lowering (18–23%) and concerns over hepatotoxicity.¹²³ While other enzyme targets exist at more distal steps of cholesterol synthesis pathway, the effort at developing their inhibitors has remained largely academic, and there is currently no known industrial effort at their pharmaceutical adaptation. Of note, bempedoic acid is an upstream inhibitor in the cholesterol biosynthetic pathway upon which hepatic activation can inhibit ACL, the enzyme that generates acetyl-CoA (Figure 2). Four clinical trials (CLEAR Serenity, CLEAR Tranquility, CLEAR Harmony, and CLEAR Wisdom) have shown its efficacy in further reduction of LDL-C by 15–20% on the background of maximally tolerated statin treatment while also reducing hs-CRP level.^{103,124–126} The bempedoic acid/ezetimibe combination (Nexlizet) was just approved by FDA for use in adults with heterozygous familial hypercholesterolemia or established ASCVD requiring further LDL-C lowering, and the CLEAR Outcomes trial is expected to report its cardiovascular outcome results in 2022.¹⁰⁴ While bempedoic acid acts above the FFP bifurcation point and therefore unable to mechanistically distinguish cholesterol synthesis from prenylation, it would be very interesting to observe whether it replicates some of statins’ efficacy on cardiovascular risk reduction and is more superior than ezetimibe when added to statins with equivalent LDL lowering.

Regarding the second approach, an ideal ‘match’ would be to randomize patients to statin vs. PCSK9 inhibitors to achieve identical LDL-C lowering and assess if there is a difference in clinical outcomes. While conceptually simple, blinding would be difficult due to different routes of drug administration; furthermore, the cost of PCSK9 inhibitor and the need to withhold a guideline-indicated treatment in statin pose financial and ethical dilemmas. One unique scenario would be to study patients with statin intolerance. Indeed, at least five trials have been conducted (GAUSS 1–4 and ODYSSEY ALTERNATIVE) to that effect, and in fact, one trial (ODYSSEY ALTERNATIVE) was able to re-challenge one study arm with atorvastatin under blinded conditions with high adherence rate.¹²⁷ Unfortunately, the duration of follow-up was short (24 weeks), and the majority of the enrolled patients was switched to alirocumab during the open-label follow-up phase, precluding a precious chance at comparing cardiovascular outcomes between alirocumab and atorvastatin monotherapies at otherwise grossly comparable LDL-C-lowering efficacies (45% vs. 32%). Nonetheless, a repeat effort with a similar trial design but large enrolment size and longer follow-up to allow outcome comparison would hold immense value. Another trial design would be to use low intensity statin/PCSK9 inhibitor combination vs. high intensity statin to lower LDL-C by identical degrees and investigate any differences in clinical outcomes. On this aspect, we may derive some insight from a small prospective study in Taiwan of 60 patients that compared simvastatin 40 mg to simvastatin 10 mg/ezetimibe 10 mg and showed improved vasoreactivity at 28 days despite identical lipid-lowering by both treatment groups when compared to placebo.¹²⁸

Regarding the third approach, the aim is to fundamentally dissect out the role of LDL-C lowering on different cell lineages and tissue components of the cardiovascular system. For instance, if the statins’ effect on endothelial functions (e.g. via small GTPase prenylation and eNOS modulation) carries physiologic significance independent of atherosclerotic plaque formation, an endothelial-cell-specific HMGCR knockout animal model (homozygous or heterozygous) should in theory translate into improved cardiovascular outcomes without perturbing the systemic lipid profiles. HMGCR has been successfully knocked out in the liver, skeletal muscle cells, and myeloid cells, which have served as powerful genetic tools in study statin-associated hepatotoxicity, myopathy, and effects of macrophage migration on atherosclerosis, respectively.^129–131 Similar targeted deletions in endothelial cells, smooth muscle cells, and platelets could yield important insight into statin pleiotropy. Given embryonic lethality of homozygous HMGCR knockout in mice,¹³² an inducible model of genetic ablation may be preferable and better reproduce physiologic effect of adult-stage statin therapy.

11. Conclusions

In an era of rising cardiovascular disease burden, statins served as a textbook example of successful hypothesis-driven drug development. To fully account for statin’s clinical impact on cardiovascular disease, the concept of statin pleiotropy emerged and has led to a substantial body of research evaluating its significance. While growing scientific work has provided provocative rationale for statin’s pleiotropic effects and encouraged statin’s expanded application in non-cardiovascular fields, an equally expanding body of clinical data pointed to a more nuanced interpretation of the statin pleiotropy. Indeed, results from ezetimibe and PCSK9 inhibitor trials, coupled with Mendelian association analyses, have provided strong arguments as to whether statin pleiotropy is clinically meaningful. Taken together, statins likely confer the majority of cardiovascular benefits through LDL-C lowering, which itself may affect multiple molecular pathways beyond atherosclerotic plaque formation. We expect that newer studies with novel lipid-lowering agents, targeted experimental approaches, and creative trial designs will more effectively address the questions of statin pleiotropy, which, despite a substantial body of literature, still remains unanswered.