Plaque Regression and Endothelial Progenitor Cell Mobilization With Intensive Lipid Elimination Regimen (PREMIER)

Patients surviving an acute coronary syndrome (ACS) are at a high risk of recurrent ACS, especially early after the index event.¹ Early and intensive lipid-lowering therapy (ILLT) with statins is the current standard of care for patients with ACS, based on both clinical trial data and studies showing coronary atherosclerotic plaque regression and stabilization.² Despite statin therapy, however, the incidence of recurrent cardiovascular events remains high after an ACS.³ In addition, such maximization of early lipid-lowering therapy occurs only in a minority (≈23%) of eligible patients.^4,5 Progression or rupture of lipid-rich nonculprit atherosclerotic plaque leads to most recurrent cardiovascular events and has prompted a continued effort toward lowering of total cholesterol and specifically LDL (low-density lipoprotein) levels post-ACS. However, it remains unknown whether the most intensive and acute LDL lowering post-ACS with LDL apheresis can lead to a rapid and detectable reduction in nonculprit coronary plaque volume. LDL apheresis acutely removes LDL particles that transport cholesterol in the plasma portion of blood by extracorporeal filtration. The terms LDL apheresis and lipid apheresis can be used interchangeably. It also has the potential to mobilize endothelial progenitor cells (EPCs) known to play a crucial role in vascular healing by promoting endothelialization of atherosclerotic plaque and neovascularization of ischemic myocardium.⁶ We designed a pilot study of nonfamilial hyperlipidemia (FH) ACS patients undergoing percutaneous coronary intervention (PCI) to examine the impact of LDL apheresis on LCL levels, coronary atherosclerotic burden, and EPC mobilization.

The authors declare that all supporting data are available within the article.

A total of 160 participants were enrolled from 329 screened during phase I and phase II of the trial. Baseline characteristics of participants in each phase and combined are provided in Table 1. Mean age was 65.0±8.6 years (64.0±8.8 in ILLT group versus 66.1±8.4 in SMT group), and 99% were men for each treatment group and overall. The prevalence of diabetes mellitus (51%), hypercholesterolemia (93%), prior MI (42%), and coronary revascularization (47%) was similar in the 2 study groups, whereas cancer was higher in the SMT group (18% versus 6%; P=0.0255). Tendon xanthoma or corneal arcus was not observed in any patient during a comprehensive preenrollment physical. Lipid-lowering therapy inclusive of a statin was used by 76% of participants before enrollment in the ILLT and 80% in the SMT arms (P=0.5703). Over 80% of the participants both in phase I and phase II received intensive statin treatment with either atorvastatin ≥40 mg or rosuvastatin ≥20 mg. An admitting diagnosis of a non–ST-segment–elevation MI or unstable angina was present in 49% and 50%, respectively, with no significant differences between the ILLT and SMT arms (non–ST-segment–elevation MI, 48% versus 51%; P=0.7516 and unstable angina, 51% versus 49%; P=0.8743). Average baseline values of total cholesterol and LDL were 181.5±42.1 and 113.6±33.5 mg/dL, respectively, with no significant differences across the study arms.

PCI procedure was performed on 94.4% native and 5.6% saphenous venous bypass graft lesions (Table 2). On average 1.36±0.64 lesions were treated per patient involving 1.59±0.98 stents, 93% of which were drug eluting. Unfractionated heparin was used for anticoagulation during the procedures in 81%, bivalirudin in 26%, and glycoprotein IIb/IIIa–inhibiting agents in 8% patients. Aspirin and a P2Y12 receptor–inhibiting drug were prescribed to all patients. Dual antiplatelet therapy with aspirin and P2Y12 inhibitor was used in 96% of patients post-PCI; 89% and 88% of patients continued this therapy at 3 and 6 months, respectively, with no significant difference between the study groups. All baseline IVUS assessments were performed during the index PCI procedure. Follow-up coronary angiography and IVUS interrogation of the target coronary artery segments were performed in 89% of the participants after a mean interval of 90.7±9.8 days in the ILLT and 94.0±18.3 days in the SMT arms (P=0.1880).

Atorvastatin was the predominant statin used, and its mean daily dose was similar in the ILLT and SMT arms of the study from hospital discharge to 30 days (74.4±14.0 versus 69.2±20.4 mg; P=0.1377) and at 90 days (73.3±15.1 versus 70.2±19.5 mg; P=0.3845). LDL apheresis was successfully completed in 93% of patients assigned to the ILLT arm. Mean apheresis duration was 143.19±79.67 minutes, and 2,255±855 mL of plasma was exchanged through the Liposorber filters. Mean hospital days in the ILLT and SMT arms were 2.54±1.87 and 1.75±1.49 days, respectively (P=0.0037).

Mean LDL reduction at discharge was 53% in ILLT and 17% in SMT groups compared with baseline (P<0.0001 for both; Figure 3). Significantly lower LDL levels were sustained in the ILLT and SMT groups at 30 days (66.4±19.6 and 73.4±30.8 mg/dL; P<0.0001 for both) and 90 days (68.8±29.5 and 73.2±26.3 mg/dL; P<0.0001 for both). There was no significant difference in LDL levels between study groups at 30 days (P=0.10) and 90 days (P=0.34). Trends of lipoprotein levels are shown in Figure I in the Data Supplement.

The primary efficacy end point (intention-to-treat analysis without and with adjustment) is illustrated in Figure 4. The raw change in total plaque volume on average decreased more in the ILLT group than that in the SMT group (−6.01 versus −0.95 mm³; difference of means, −5.06 [95% CI, −11.61 to 1.48]; P=0.1286), while the percentage change in total plaque volume on average decreased by 4.81% in the ILLT group but increased by 2.31% in the SMT group, with a difference of −7.13% between the two treatment groups ([95% CI, −14.59 to 0.34] P=0.0611) without adjustment for covariates (Figure 4A). After adjusting for participating sites, age, preexisting coronary artery disease, diabetes mellitus, baseline LDL levels, and baseline plaque burden, similar results were obtained for raw change or percentage change in total plaque volume on average (Figure 4B). There was no difference in the necrotic core or other plaque components. Corresponding values of vessel, lumen volumes, and necrotic core component in both treatment groups are shown in Table I in the Data Supplement. Sensitivity analyses were performed for intent-to-treat population comparing methods of either assigning zero to missing change score or using multiple imputation approach for missing follow-up assessment. This difference between the ILLT versus SMT groups was consistent. Subgroup analyses and within-group comparisons for treatment effects together with the interactions of treatment and group are depicted in Figure 5.

Figure 6 depicts EPC mobilization in the study. There was robust increase in EPC-CFU/mL from baseline to post-PCI and 30-day follow-up in both groups; however, a sustained mobilization of EPCs at 90 days was present in the ILLT group (P=0.0015) compared with SMT (P=0.0844).

There were 3 major peri-PCI AEs in 2 ILLT participants and none in the SMT group that met the definition of primary safety end point (Table 3). The total number (67 in ILLT versus 59 in SMT) and percentage of patients (45.2% in ILLT versus 42.1% in SMT; P=0.7505) with serious AEs were not significantly different between the ILLT and SMT arms of the study. Statin-related abnormal liver function tests and muscle injury events were also similar in both study arms. A complete safety report inclusive of all serious AEs and AEs is included as the Data Supplement. Overall, 7 patients experienced a MACE event; there were 2 deaths, both in the ILLT group, 3 nonfatal MIs (2 in ILLT and 1 in SMT), and 2 ischemic strokes (1 in ILLT and 1 in SMT). Six-month MACE rates were 6% and 3% in the ILLT and SMT groups, respectively (P=0.4469). One patient died ≈2 months post-enrollment from respiratory failure related to his chronic obstructive lung disease and another at ≈4 months secondary to an ischemic stroke.

PREMIER is the first trial to demonstrate safety and a strong trend for an early (90 days) coronary plaque regression with a single LDL apheresis with statin therapy in ACS patients treated with PCI compared with SMT comprising of statins alone. There was a significantly greater LDL lowering at discharge and a more sustained EPC mobilization in the LDL apheresis arm.

These data from the PREMIER trial may support early and intensive lowering of LDL cholesterol in patients with ACS. The early changes in nonculprit coronary plaque volumes evaluated in our trial are comparable to other ACS trials with IVUS follow-up. Endo et al¹⁵ recently analyzed 173 ACS patients from 4 prospective clinical trials who underwent serial IVUS of nonculprit lesions post-PCI on statin treatment at baseline and at 6 or 8 months of follow-up. The overall change in coronary plaque volume was −1.5%. Over a mean clinical follow-up of 3.5 years, achievement of plaque regression was independently associated with MACE (hazard ratio, 0.42 [95% CI, 0.19–0.88]; P=0.02). In PREMIER, although the −4.81% change in total plaque volume observed early in the ILLT arm is promising, no claim regarding the sustainability of coronary atheroma regression or its impact on future clinical events can be made. In the LACMART (Low Density Lipoprotein-Apheresis Coronary Morphology and Reserve Trial), LDL apheresis treatment over 1 year lowered the LDL of statin-treated FH patients with a significant reduction in the coronary plaque area and an increase in the minimum luminal diameter.¹⁴

Given the novelty and the potential risk to human subjects associated with this trial, PREMIER was a rigorously conducted clinical trial as per US FDA Investigational Device Exemption guidance. It collected comprehensive AE data that were reviewed and monitored by an independent Data Monitoring Committee and the VA Cooperative Studies Program Clinical Research Pharmacy Coordinating Center. Major AEs in the peri-PCI period (primary safety end point) were rare, and overall serious AEs were similar across both treatment groups. Peri-PCI AEs in the ILLT arm included those that occurred during LDL apheresis and were predominantly nonserious with complete resolution.

Circulating EPC-CFUs increased in both groups compared with baseline levels, with the greatest increase observed at 30 days post-PCI. As patients presenting with an ACS are known to have limited ability or a failure to mobilize EPCs, these data from the PREMIER trial provide evidence regarding EPC mobilization in ACS patients.¹³ As increased circulating levels of EPCs after ACS have been associated with better cardiovascular outcomes, the more durable EPC mobilization seen with LDL apheresis further supports acute LDL lowering in ACS.⁶

We believe that the lack of statistically significant reductions in coronary plaque volumes between ILLT and SMT groups at 90 days on unadjusted analysis and the marginal significance on an adjusted analysis could most likely be a result of more intensive statin therapy in both groups during phase 2 of the study. Further clinical studies with a larger sample size may help account for the larger-than-expected variations in coronary plaque volumes observed in this trial.

Statins constitute the backbone of the current guidelines endorsed on intensive LDL lowering for secondary prevention in ACS.¹⁶ However, the clinical evidence for an early and intensive lipid lowering in ACS has not been consistent. The MIRACL (Myocardial Ischemia Reduction With Aggressive Cholesterol Lowering) RCT compared atorvastatin 80 mg daily with placebo initiated within 1 to 4 days of an ACS event.¹⁷ It demonstrated a significant reduction in composite recurrent ischemic events in the first 16 weeks, driven primarily by a reduction in ischemic events requiring rehospitalization. Despite near identical reductions in LDL levels in the active study arms (62 mg/dL), the A to Z trial showed no risk reduction during the first 4 months compared with MIRACL with a 16% reduction in similar ischemic end points.^18–20 In the PROVE IT-TIMI22 trial (Pravastatin or Atorvastatin Evaluation and Infection Therapy), early intensive statin treatment in ACS was shown to favorably impact long-term mortality and subsequent cardiac events.² Lower LDL levels in the active arm (33 mg/dL) were achieved in the PROVE IT-TIMI22 trial compared with MIRACL or A to Z at 4 months. However, a more recent Cochrane systematic review assessing the effects of early statins in 18 RCTs and >14 000 ACS patients concluded that it did not significantly lower MACE.²¹ In a contemporary ACS study of patients (ALPS-AMI [Assessment of Lipophilic vs Hydrophilic Statin Therapy in Acute Myocardial Infarction]) with baseline LDL ≥70 mg/dL treated with PCI, LDL lowering by 30% within 4 weeks of an acute MI significantly reduced MACE (9.4% versus 3.4%; P=0.013).^22,23 For the first time, this study defined the optimal period to achieve a targeted reduction of LDL in ACS—an important clarification that could have accounted for the heterogeneity of outcomes observed in prior ACS statin trials. Further, the early reduction group had significantly lower mortality and rate of cardiac death than the late reduction group. The inconsistent clinical effect of LDL lowering in ACS could also be related to the intensity and celerity of LDL lowering. Compared with the MIRACL and A to Z ACS trials, in PREMIER, a mean LDL of 53 mg/dL (a 53% reduction from baseline) was achieved at discharge and sustained <70 mg/dL at 90 days in the LDL apheresis arm. In our study, the robust LDL lowering with a single LDL apheresis faded over time, and at 30 days, the LDL levels were similar to the statin only–treated arm.

To date, the use of PCSK9 (proprotein convertase subtilisin kexin type 9) monoclonal antibodies (PCSK9 inhibitors) has mainly been studied in ACS patients following a run-in period of 2 to 16 weeks on statins.^24,25 A recent study has assessed the use of PCSK9 inhibitors in the early in-hospital management of an ACS, predominantly in patients not on prior statin therapy.²⁶ Evolocumab in this study significantly lowered LDL levels at 8 weeks from baseline compared with placebo (77.1±15.8% versus 35.4±26.6%; P<0.001). Moreover, LDL apheresis has been shown to dramatically lower circulating PCSK9 and Lp(a) levels,²⁷ upregulate target tissue LDL receptors,²⁸ and restore sensitivity to statins that increase PCSK9 levels.²⁹ These effects may help explain a more sustainable lowering of LDL in non-FH patients compared with FH and demonstrate the feasibility of an initial LDL apheresis followed by longer term treatment with statins and PCSK9 inhibitors in patients with ACS.

Since LDL apheresis was first introduced in the mid-1970s, it has predominantly been used for treating FH homozygotes with LDL >500 mg/dL, FH heterozygotes with LDL ≥300 mg/dL or ≥100 mg/dL, and documented coronary heart disease or peripheral artery disease.³⁰ No patient in this ACS trial had baseline LDL ≥160 mg/dL or peripheral stigmata consistent with FH. The US FDA-approved Liposorber LA-15 system was used for this trial. As described earlier, it isolates the plasma component that flows through 1 or 2 dextran sulfate cellulose (Liposorber) columns, which binds ApoB-100 proteins including LDL, VLDL, and Lp(a). Plasma and red blood cells are pooled and returned to the patient.¹¹ There is minimal lowering on VLDL and no appreciable change in blood HDL or triglyceride levels in our study. Other methods of LDL apheresis include immunoadsorption, heparin extracorporeal LDL precipitation, and direct adsorption of lipoprotein using hemoperfusion and a Liposorber D whole blood filtration technique without isolating plasma. A single LDL apheresis with any of the above methods can lower LDL levels by ≈50% to 80%, and Lp(a) levels from 65% to 70%. A more modest reduction in fibrinogen levels, use of a lower heparin dose, a highly compact design with superior portability, and ease of use made the Liposorber LDL apheresis system more suitable for this trial of ACS patients treated with PCI.

Our study is limited by its small sample size, short follow-up, evaluation of a surrogate end point, absence of Lp(a) and other inflammatory marker data, and not being powered to assess clinical outcomes. ILLT patients also had a longer hospital stay due to the apheresis procedure. All study sites were Veterans Affairs medical centers and, therefore, enrolled a disproportionate number of men, limiting the generalizability of its findings. Therefore, these results need to be validated by a larger clinical outcomes trial involving a more diverse patient population. Additionally, no PCSK9 inhibitors were used in this trial for medical management of hypercholesterolemia during follow-up, and genetic testing to exclude patients with FH was not undertaken.

Despite these limitations, the PREMIER trial tested a novel treatment paradigm and provides new and relevant data on safety and early coronary plaque regression with acute lowering of LDL with a single LDL apheresis treatment in ACS patients maintained on statin therapy.

We acknowledge the members of the Veterans Affairs North Texas Health Care System Office of Research and Development for providing crucial support for the trial, study coordinators, and LDL (low-density lipoprotein) apheresis technicians who significantly contributed to the success of this study.