A Systematic Review of PCSK9 Inhibitors Alirocumab and Evolocumab

BACKGROUND: The proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors are a new class of cholesterol-lowering medications that provide significant reductions in lipids but at a large cost relative to statins. With 2 such drugs now on the market, alirocumab and evolocumab, comparing the evidence base for these drugs is necessary for informed decision making. OBJECTIVE: To compare the benefits and harms of the PCSK9 inhibitors alirocumab and evolocumab. METHODS: The databases Ovid MEDLINE, Cochrane Library, SCOPUS, and ClinicalTrials.gov were used to search for randomized controlled trials of alirocumab or evolocumab with any relevant comparator reporting health outcomes, lipid outcomes, or harms through September 2015, and information was requested from manufacturers. Results were reviewed according to standard review methods. RESULTS: The database searches revealed 17 fair- and good-quality trials; however, none had primary health outcomes or directly compared PCSK9 inhibitors. Alirocumab (75 mg to 150 mg subcutaneously every 2 weeks) resulted in significantly greater reductions in low-density lipoprotein cholesterol (LDL-C; -8% to -67%) at 12-24 weeks in patients with (a) heterozygous familial hypercholesterolemia and (b) patients at high or varied cardiovascular (CV) risk who were not at LDL-C goals with statin therapy. The highest strength evidence was for patients with high CV risk not at LDL-C goals. Alirocumab also resulted in high-density lipoprotein cholesterol (HDL-C) increases of 6%-12%. Low- and moderate-strength evidence for adjudicated CV events at 52-78 weeks for a priori analyses indicated no benefit. Low- and moderate-strength evidence also found no differences in harms except possibly slightly more injection-site reactions. Evolocumab (120 mg subcutaneously every 2 weeks to 420 mg every 4 weeks) resulted in significantly greater reductions in LDL-C (-32% to -71%) at 12-52 weeks in patients with heterozygous or homozygous familial hypercholesterolemia, patients intolerant of statins, and patients with varied CV risk not at LDL-C goal with statin therapy. The highest strength evidence was for heterozygous familial hypercholesterolemia and patients not at LDL-C goals. Moderate-strength evidence showed HDL-C increases in the range of 4.5%-6.8%. Harms were not different between groups, except possibly slightly greater overall adverse event reporting. Evidence on adjudicated CV outcomes was insufficient to draw conclusions because of sparseness of events, study limitations, and inability to assess consistency of findings. CONCLUSIONS: Alirocumab and evolocumab have evidence of large improvements in lipid levels. The strength of the evidence is greater for alirocumab than evolocumab in patients with high CV risk who were not at LDL-C target goals, while evidence for evolocumab is stronger in patients with heterogeneous familial hypercholesterolemia and patients with varied CV risk who were not at LDL-C target goals. Evidence on adjudicated CV outcomes for a priori analyses is unable to show benefit for alirocumab and is insufficient to draw conclusions for evolocumab. Important questions remain about the comparative effects on long-term health outcomes.

• The proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors alirocumab and evolocumab produce large reductions in low-density lipoprotein cholesterol (LDL-C) when given in addition to statins in a variety of populations with hypercholesterolemia or alone or in combination with ezetimibe in patients intolerant to statins. • Cardiovascular benefits are as yet not clearly established, with some analyses of incidentally reported or secondary outcomes indicating benefit. • Concerns raised about the evidence include the methodology behind cardiovascular endpoint analyses, long-term effects of large reductions in lipids, whether comparison groups were optimal, and cost of the drugs in light of these concerns.

What is already known about this subject
• This study compared the strength of the evidence base for each of the PCSK9 inhibitors alirocumab and evolocumab. • The evidence base for alirocumab is stronger than evolocumab for patients at high cardiovascular risk who have not reached LDL-C targets with statin therapy. Although there is strong evidence that alirocumab lowers lipids significantly, there is currently no evidence of a benefit in cardiovascular outcomes when limiting to adjudicated cardiovascular events and a priori analyses. • Evidence for evolocumab is stronger than alirocumab for patients with heterogeneous familial hypercholesterolemia and patients with varied cardiovascular risk not at LDL-C target, but evidence on cardiovascular events is insufficient or absent. Evolocumab also has evidence in these other populations: homozygous familial hypercholesterolemia, populations of patients with varied cardiovascular risk, and follow-up studies of varied populations.
Additionally, there has been criticism that even the LDL-C benefits of PCSK9 inhibitors may have been overestimated because of underutilization of high-intensity statin therapy. 11,12 In the largest PCSK9 inhibitor trial to date, the ODYSSEY LONG-TERM trial of alirocumab, 20 which alone accounts for 45% of mortality events in the Navarese 2015 meta-analysis, 7 only 47% of patients were receiving high-dose statins, despite the trial's focus on high-risk patients. In their published response to this criticism, the ODYSSEY LONG-TERM authors agreed with this limitation and promised that the ongoing ODYSSEY Outcomes study (NCT01663402) will address this uncertainty. 21 To support the work of the participating organizations of the Drug Effectiveness Review Project, the Pacific Northwest Evidence-based Practice Center conducted a systematic review of the comparative effectiveness of alirocumab and evolocumab versus each another, other lipid-lowering regimens, or placebo. 22 The purpose of this review was to evaluate the body of evidence for each drug based on the population and comparators studied to better stratify the evidence for decision making with regard for their place in therapy and for selecting a specific PCSK9 inhibitor.

■■ Methods
For this systematic review, randomized controlled trials (RCTs) of alirocumab or evolocumab were included that reported 1 or more health outcomes (e.g., cardiovascular events), lipid outcomes (e.g., LDL-C), or harms (e.g., withdrawal from study because of adverse events). For cardiovascular events, only those that were adjudicated through a formal process were included; those events that were reported as incidental findings were excluded.
To identify relevant studies, Ovid MEDLINE (through week 2, September 2015), the Cochrane Library databases (2009 through 2015), and Scopus (2010 through 2015) were searched, using drug names as search terms. Additional studies were located by hand searching reference lists of included studies, ClinicalTrials.gov, and requested information from drug manufacturers. Two reviewers, using consensus to resolve disagreements, assessed studies for eligibility. Study characteristics, including baseline patient characteristics, and study results were abstracted by 1 reviewer and checked by a second reviewer for accuracy.
The internal validity (quality) of individual studies were assessed based on predefined criteria developed for the Drug Effectiveness Review Project (see Appendix A for individual study ratings, available in online article). 23 Individual studies were assigned ratings of good, fair, or poor. Studies with major flaws were rated as poor; those with essentially no flaws were rated as good; and the remainder were rated as fair. Data in studies that were clinically and methodologically similar were pooled using random effects models, and statistical heterogeneity was explored using the I-squared statistic. 24 drugs such as ezetimibe alone are also inadequate. Monoclonal antibodies that inhibit proprotein convertase subtilisin/kexin type 9 (PCSK9) have been under development as novel therapies, potentially filling these gaps in current standard therapies for hypercholesterolemia. The first 2 drugs in this new class of antihyperlipidemic treatments received U.S. Food and Drug Administration (FDA) approval in mid-2015. In combination with other treatments-typically maximally tolerated statinsalirocumab was approved for use in heterozygous familial hypercholesterolemia or clinical atherosclerotic cardiovascular disease who required additional low-density lipoprotein cholesterol (LDL-C) lowering, and evolocumab was approved for patients with clinical atherosclerotic cardiovascular disease or heterozygous and homozygous familial hypercholesterolemia who required additional LDL-C lowering.
Alirocumab (Praluent) and evolocumab (Repatha) have shown striking LDL-C reductions over 10 to 78 weeks (26%-67%). But, as with many novel specialty drugs, the promise of better results with PCSK9 inhibitors comes with a high price tag-in this case, the cost is estimated at more than $14,000 per patient per year. [3][4][5] In this cost-conscious era of health care, clinicians, payers, and the public are already questioning how to best manage this costly new drug class. 6 While there have been reviews of the evidence published, 7,8 there has been concerning criticism of the analysis of the evidence, as well as the evidence itself. These concerns center around long-term health outcomes (including handling of data in meta-analyses) 9,10 and comparison groups in trials (e.g., high-intensity statins). [11][12][13][14] To date, no clinical trial of a PCSK9 inhibitor designed a priori, with adequate size and duration, to assess cardiovascular outcomes as the primary outcomes has yet been completed. However, 4 such trials are underway and are expected to be completed in 2018. [15][16][17][18] In the meantime, Navarese et al. (2015) pooled data from the 24 available smaller studies with follow-up ranging from 2 months to 2 years and reported that PCSK9 inhibitors as a group reduce all-cause mortality (odds ratio [OR] = 0.45; 95% confidence interval [CI] = 0.23-0.86) and rates of myocardial infarction (OR = 0.49; 95% CI = 0.26-0.93). 7 However, these findings have been questioned based on concerns that the statistical approach used by Navarese et al. is biased in how study arms with zero events were handled. 9,10 Critics demonstrated that use of a few different less-biased conventional methods produced results of consistently smaller magnitude and in 1 case lost statistical significance. Because of the instability and imprecision of these findings, critics urge us to wait for the results of the adequately designed cardiovascular endpoint trials now underway before promoting use of these drugs. Given the increased demand by payers such as Medicare for definitive evidence that interventions improve health outcomes, and not merely intermediate endpoints, 19 this debate has important implications for coverage policy determinations as well.
For studies with multiple arms of differing doses, the doses within the FDA-approved dosing range were analyzed. Body of evidence strength was graded for a given key outcome based on the guidance established for the Evidence-based Practice Center Program of the Agency for Healthcare Research and Quality. 25 This grading assessed the limitations, directness, consistency, and precision of the body of evidence for each outcome and resulted in evidence rating of high strength (★★★★), moderate strength (★★★), low strength (★★), or insufficient (★) to draw conclusions. In the tables presented in this review, these ratings are shown as stars. Individual study quality and the strength of evidence ratings were dually reviewed with disagreements resolved using consensus.

Heterozygous and Homozygous Familial Hypercholesterolemia
Compared with placebo, there was low-strength evidence that alirocumab achieved a higher LDL-C reduction in patients with heterozygous familial hypercholesterolemia who took a maximum statin dose plus ezetimibe, based on 1 fair-quality trial (N = 77). Similar effects were shown for high-density lipoprotein cholesterol (HDL-C), but there was insufficient evidence to draw conclusions about harms. 38 Evidence for alirocumab compared with placebo in patients with heterozygous familial hypercholesterolemia who took a low-to moderate-intensity statin was insufficient (1 fair-quality RCT, N = 22). 39 There were no studies of alirocumab in patients with homozygous familial hypercholesterolemia.
Compared with placebo, there is high-strength evidence from two 12-week RCTs (fair-and good-quality, N = 499) that evolocumab (140 mg every 2 weeks to 420 mg every 4 weeks) achieved a greater LDL-C reduction in patients with heterozygous familial hypercholesterolemia who were largely taking a high-intensity statin plus ezetimibe, with a greater improvement in HDL-C and no differences in harms (moderatestrength evidence). 32,34 In patients with homozygous familial hypercholesterolemia who took a maximum statin dose and ezetimibe (Table 3), low-strength evidence based on 1 small, good-quality trial (N = 50) suggested that evolocumab (420 mg every 4 weeks) reduced LDL-C significantly more than placebo at 12 weeks. 33 There was no difference in HDL-C change, and there was no difference in the percentage of patients with serious adverse events, neurocognitive events, or those withdrawing because of treatment-emergent adverse events.

Statin-Intolerant Patients
There was no evidence for alirocumab in patients who were intolerant to statins. In statin-intolerant patients, there was low-strength evidence based on 2 fair-quality, 12-week RCTs (GAUSS and GAUSS-2; N = 434) that evolocumab (280 mg every 4 weeks or 140 mg every 2 weeks) led to a greater reduction in LDL-C than placebo, with similar effects in HDL-C and harms. 40,41 There is also low-strength evidence from the  GAUSS study (N = 62) that the combination of evolocumab (420 mg every 4 weeks) plus ezetimibe (10 mg) led to a greater percentage of LDL-C reduction than ezetimibe (10 mg) alone, but there was insufficient evidence to draw conclusions on other outcomes. 41

Secondary Treatment in Patients at High Cardiovascular Risk (Not Achieving Study Targets with First-Line Therapies)
Several studies evaluated a PCSK9 inhibitor in patients with varying levels of cardiovascular risk, who had not achieved an LDL-C of ≤ 100 mg/dL or ≤ 70 or 75 mg/dL using mainly statin-based therapy before randomization. Very few patients had familial forms of hypercholesterolemia in these studies. In high-risk patients, moderate-strength evidence based on a good-quality trial (ODYSSEY COMBO II, N = 720) demonstrated that alirocumab (75 mg to 150 mg given every 2 weeks) resulted in a higher proportion of patients with high cardiovascular risk reaching the study goal of LDL-C < 70 mg/dL at 24 weeks (relative risk [RR] = 1.70, 95% CI = 1.46-1.95) than ezetimibe (10 mg). 27 Similarly, the difference in the percentage of change in LDL-C and HDL-C at 24 weeks was greater with alirocumab (-29.8% and 8.1%, P < 0.001). Moderate-strength evidence found that there was no difference in overall adverse event reporting between alirocumab and ezetimibe and that the rate for reporting any adverse event was high (69%). Lowstrength evidence found that withdrawal because of adverse events and serious adverse events were not different between the groups. Low-strength evidence suggested a higher rate of injection-site reactions with alirocumab than with placebo injection plus ezetimibe, but this estimate was imprecise (wide confidence interval, inadequate sample size to evaluate this outcome). Neurocognitive events were rare.
Based on 2 trials (ODYSSEY COMBO I and ODYSSEY Long-Term; N = 2,656), there was high-strength evidence that alirocumab (75 mg to 150 mg given every 2 weeks) resulted in a higher proportion of patients with high cardiovascular risk reaching the study goal of LDL-C < 70 mg/dL at 24 weeks than placebo (pooled RR = 9.65, 95% CI = 7.7-12.0). 20,30 The difference in the percentage of reduction in LDL-C and HDL-C was also greater. High-strength evidence found no statistical difference in overall adverse event reporting and that the rate for reporting any adverse event was high (75%). Moderate-strength evidence found that withdrawal because of adverse events and serious adverse events were also not different between groups. Low-strength evidence suggested that the risk of injection-site reactions (pooled RR = 1.4, 95% CI = 0.98-2.1) or neurocognitive events were not increased, although the estimates were imprecise because of few events.
In a comparison of evolocumab and ezetimibe (both with statin therapy) in patients with high cardiovascular risk, the LAPLACE-2 study (N = 329 for this comparison) provided low-strength evidence that when added to either atorvastatin (10 mg or 80 mg), compared with ezetimibe (10 mg), evolocumab (420 mg monthly) resulted in higher rates of meeting an LDL-C target of < 70 mg/dL at 12 weeks, with similar rates of patients with overall adverse events. 35 This study provided insufficient evidence to draw conclusions about HDL-C, serious adverse events, or withdrawal because of adverse events because of the small magnitude of change or event rates. In patients with high cardiovascular risk, a small 12-week (N = 104 for placebo comparison) study provided low-strength evidence that evolocumab (420 mg monthly) resulted in higher rates of meeting LDL-C targets of < 100 mg/dL and < 70 mg/dL when added to statins in Japanese patients, compared with placebo. 29 There was greater mean change in LDL-C but insufficient evidence to draw conclusions about other outcomes.
There was moderate-and low-strength evidence of no difference in adjudicated cardiovascular events between alirocumab and ezetimibe at 52 weeks or between alirocumab and placebo at 52-78 weeks (Table 4). 20,27,30 For evolocumab, adjudicated cardiovascular events were reported as secondary or post hoc outcomes, but evidence was insufficient to draw conclusions because of too few events (imprecise estimates), study limitations, and lack of ability to verify findings across multiple studies (Table 4). 29,37

Secondary Treatment in Patients at Varied Cardiovascular Risk (Not Achieving Study Targets with First-Line Therapies)
Low-strength evidence from 2 small (N = 154 total) fair-quality RCTs indicated that in patients stabilized on statins who have not achieved an LDL-C of < 100 mg/dL, alirocumab (150 mg subcutaneously every 2 weeks for 8-10 weeks) resulted in significantly more patients achieving study goal (LDL-C < 100 mg/ dL) and a greater percentage of reductions (49%-67% more) than statins alone. 31,36 Evidence for adverse events and in subgroups, compared with statins alone, was insufficient because of small sample sizes.
In short-term comparisons of evolocumab and placebo, LAPLACE-TIMI 57 and LAPLACE-2 (N = 996 for placebo comparison) provided high-strength evidence that, at 12 weeks in patients with varying risk levels and not meeting LDL targets, significantly more patients taking evolocumab (420 mg monthly) than taking placebo (both with statin therapy) achieved an LDL-C of < 70 mg/dL and had a greater percentage of reduction in LDL-C. 28,35 There is also moderate-strength evidence of modest HDL-C increases with evolocumab (420 mg monthly) and moderate-to high-strength evidence of no differences in adverse events. Based on 1 good-quality longer-term trial (N = 901; 52 weeks), there is moderate-strength evidence that evolocumab (420 mg given every 4 weeks) also results in significantly more patients achieving a goal of LDL-C < 70 mg/ dL, compared with placebo; low-strength evidence of a modest increase in HDL-C; and evidence of no difference in harms (low-and moderate-strength evidence depending on outcome).  , that evolocumab (420 mg monthly or 140 mg every 2 weeks plus standard care -primarily statins) reduced LDL-C by 61% more than standard care alone at 12 weeks. 37 This reduction was largely sustained at 48 weeks (58.4% more than usual care at week 48). This analysis of the 2 studies also provided low-strength evidence of a greater proportion of patients meeting an LDL-C goal of < 100 mg/dL or < 70 mg/dL and a greater increase in HDL-C at 12 weeks than with standard therapy alone.
There is low-strength evidence that slightly more patients on evolocumab experienced any adverse event at 12 weeks, compared with statins alone, without differences in serious adverse events but insufficient evidence to draw conclusions about other adverse event outcomes.

■■ Discussion
This systematic review of PCSK9 inhibitor drugs finds low-to high-strength evidence of moderate-to-large magnitude reduction in LDL-C when added to statin therapy in patients with familial forms of hypercholesterolemia and alone or added to ezetimibe in those who are intolerant of statins. Although there are no direct comparison studies, and the lipid outcomes are fairly similar across the trials, the strength of this evidence varies by population and by drug. In patients with heterozygous hypercholesterolemia, the trials of evolocumab were larger (499 vs. 98), and the strength of the evidence was consistently stronger than for alirocumab (moderate to high vs. low or insufficient; Table 2). In contrast, the evidence for patients with high cardiovascular risk is stronger for alirocumab. Studies comparing alirocumab with ezetimibe were somewhat larger (720 vs. 329), twice as long (24 weeks vs. 12 weeks), and the strength of evidence was better (low to moderate vs. low and insufficient) than those of evolocumab. Studies comparing alirocumab with placebo in this population were much larger (2,656 vs. 310), twice as long (24 weeks vs. 12 weeks), and had much better

Lipid and Adverse Event Outcomes for Alirocumab and Evolocumab (continued)
low-strength evidence in patients with homozygous familial hypercholesterolemia and patients deemed intolerant to statin therapy and low-to moderate-strength evidence in broadly varied populations of all of the studied populations but with longer follow-up (Table 4). Homozygous familial hypercholesterolemia is a difficult to treat form of familial hypercholesterolemia, but with evolocumab, patients achieved reductions of 32% from a mean baseline LDL-C of 348 mg/dL; however, this finding is based on only 50 patients. Unfortunately, the evidence for the long-term health benefits and harms of the PCSK9 inhibitors alirocumab and evolocumab is insufficient to draw conclusions at this time, mainly because of limited or no information on the long-term impact of the large reductions in LDL-C. The current evidence strength of evidence (high for lipid outcomes, low to moderate for harms vs. low and insufficient). Additionally, evidence for adjudicated cardiovascular outcomes is low to moderate for alirocumab, showing no differences compared with ezetimibe or placebo, while evidence for evolocumab is insufficient ( Table 4). The final population in which both drugs were being studied was a varied cardiovascular risk group that had not met studydefined LDL-C targets while taking risk-appropriate treatment.
Here, evolocumab had a much larger study group (1,375 vs. 124), longer trials (52 weeks vs. 10 weeks), and much better strength of evidence (low to high vs. low and insufficient). Of the populations studied with both drugs, this is the one that is least clear on how it fits into step therapy with numerous other antihyperlipidemic drugs available. Evolocumab also has  comparison with continuing statins alone, they clearly exceed the typical threshold for cost-effectiveness of $50,000 to $100,000 per QALY. 44 The analyses using this model suggest that in order for the drugs to be cost-effective, the price range should be closer to $3,600 per year than the current price range of approximately $14,000 per year.
Another important issue raised about this evidence is the comparison groups, including the concomitant antihyperlipidemic therapy given. Clearly, in a step-therapy approach, the choices in the face of inadequate treatment include continuing to maximize the current therapy, add a second treatment, or even switch treatments. Pharmacologically, the PCSK9 inhibitors are likely to work better when used alongside a statin, and most of the studies continue statin therapy in PCSK9 inhibitor-treated patients. A fair comparison, then, would be maximization of the statin therapy (i.e., using high-intensity statins and/or maximally tolerated doses) in the control group in particular. Most of these studies do not adhere to this standard (Tables 2 and 3); consequently, future studies are required to determine the true incremental benefits of PCSK9 inhibitors.
The bulk of the evidence, then, is from patients currently receiving varying doses of statins who have not achieved studyspecified target LDL-C levels. While this less than optimal use of statins is not an ideal comparator as an alternative strategy to PCSK9 inhibitors, it may reflect real-world statin use. When limited to populations of patients who are at high cardiovascular risk and whose baseline LDL-C on statin therapy are on health outcomes such as cardiovascular events is limited because these studies reported them as secondary outcomes and were not designed to assess cardiovascular events. This review differs from others because it was limited to the assessment of cardiovascular outcomes that were adjudicated, rather than including cases reported incidentally (Table 4). For evolocumab, the evidence available was insufficient to make any conclusions. For alirocumab, the findings, although of low and moderate strength, did not identify statistically significant differences in high-risk patients compared with placebo or ezetimibe with 52-78 weeks of follow-up. Although these findings are not causational, as hypothesis-generating outcomes, they raise important questions about the long-term benefits of these drugs. As previously noted, we must await the outcome of ongoing trials with predefined cardiovascular endpoints as primary outcomes.
In addition to the uncertainty around long-term outcomes, and with an estimated price of almost $14,000 per year, 3 determining the place these drugs have in the therapy for hypercholesterolemia is critical. Economic modeling may be helpful, but it is limited without evidence of the health outcomes of these drug treatments. A recent economic model, using assumed cardiovascular benefits, found that the cost-effectiveness ratio (cost per quality adjusted life-year [QALY]) was $681,000 for those with familial hypercholesterolemia (either type) and $557,000 for patients with established cardiovascular disease and LDL-C > 70 mg/dL. 44 Although these are estimates for

DISCLOSURES
This project was funded by The Drug Effectiveness Review Project. Project participants reviewed the manuscript but had no role in conducting the work or writing the manuscript. Any comments received from the participants during the course of the review were taken at the discretion of the authors independently. All authors had access to the data and a role in writing the manuscript. McDonagh, Peterson, and Holzhammer declare no conflict of interest or financial interest in any therapy discussed in this article. Fazio declares receiving compensation from Sanofi for a presentation on his science to a group of their advisors and has served as a consultant to MSD, BASF, NHP, Sanofi, Ionis Pharmaceuticals, and Kowa.
Study concept and design were primarily contributed by McDonagh, along with Peterson and Holzhammer, with assistance from Fazio. Holzhammer took the lead in data collection, with assistance from McDonagh and Peterson. Data interpretation was performed by McDonagh, Peterson, and Fazio. The manuscript was written by McDonagh, Peterson, and Fazio, with assistance from Holzhammer, and revised by all the authors. evolocumab in patients with high cardiovascular risk not at LDL-C target, while evidence for evolocumab is stronger in patients with heterogeneous familial hypercholesterolemia and patients with varied cardiovascular risk not at LDL-C target. Important questions remain about the comparative effects of either drug on long-term health outcomes. above targets of 70 mg/dL to 75 mg/dL or < 100 mg/dL, we find large absolute reductions in LDL-C and higher proportions of patients achieving these study goals. However, with recent guidelines suggesting that therapy be directed by risk level and not specific LDL-C targets, 45 and without good-quality evidence on cardiovascular outcomes for the PCSK9 inhibitors, it is less clear how to interpret and apply the findings of these studies. Using the percentage of reduction in LDL-C from baseline is not a better marker, since we do not yet understand the implications of 30%-60% above and beyond any initial reductions with a statin. Adding to this lack of clarity, are the several studies in varied-or average-risk populations. In a recent editorial, Shrank et al. (2015) discussed the challenges and implications of determining the best use of PCSK9 inhibitors in light of the recent changes in recommendations for treating hyperlipidemia but recommends a step-approach that promotes maximizing the use of low-cost statin medications rather than PCSK9 inhibitors. 14 With 2 PCSK9 inhibitors approved and others in the pipeline, there will not only be questions about the role of this new class of drugs in the treatment of hypercholesterolemia, but also comparative questions about whether there are important differences among the drugs. Because these drugs will be taken for many years and are expensive, and because there are still questions regarding long-term benefit and harm, studies in the highest risk populations that are designed to assess cardiovascular and potential harm outcomes with appropriate comparison groups (guideline-recommended statin-intensity level or maximally tolerated high-intensity statins) are needed in order to shed light on these important questions.

Limitations
The main limitations of this review are those related to the evidence itself, as previously described (e.g., lack of true optimization of statin therapy in control groups). Additional potential limitations of this review include the search strategy and the exclusion of studies published in languages other than English. It was attempted to locate all studies relevant to the scope of this review, but it is possible that evidence was missed. There was limited ability to assess potential publication and reporting bias because of the few opportunities to pool studies and the lack of availability of study protocols. Finally, the scope of this review limited the inclusion of some studies included in other reviews (i.e., comparisons to no treatment) and within those incidentally reported cardiovascular outcomes.