Lapaquistat AcetateClinical Perspective
Development of a Squalene Synthase Inhibitor for the Treatment of Hypercholesterolemia
Background—Lapaquistat acetate is a squalene synthase inhibitor investigated for the treatment of hypercholesterolemia.
Methods and Results—This report summarizes the phase 2 and 3 results from the lapaquistat clinical program, which was halted at an advanced stage as a result of potential hepatic safety issues. Efficacy and safety data were pooled from 12 studies (n=6151). These were 6- to 96-week randomized, double-blind, parallel, placebo- or active-controlled trials with lapaquistat monotherapy or coadministration with other lipid-altering drugs in dyslipidemic patients, including a large (n=2121) 96-week safety study. All studies included lapaquistat 100 mg daily; 5 included 50 mg; and 1 included 25 mg. The main outcome measures were the percent change in low-density lipoprotein cholesterol, secondary lipid/metabolic parameters, and overall safety. Lapaquistat 100 mg significantly decreased low-density lipoprotein cholesterol by 21.6% in monotherapy and by 18.0% in combination with a statin. It also reduced other cardiovascular risk markers, such as C-reactive protein. Total adverse events were higher for lapaquistat than placebo, although individual events were generally similar. At 100 mg, there was an increase in alanine aminotransferase value ≥3 times the upper limit of normal on ≥2 consecutive visits (2.0% versus 0.3% for placebo in the pooled efficacy studies; 2.7% versus 0.7% for low-dose atorvastatin in the long-term study). Two patients receiving lapaquistat 100 mg met the Hy Law criteria of alanine aminotransferase elevation plus increased total bilirubin.
Conclusions—Squalene synthase inhibition with lapaquistat acetate, alone or in combination with statins, effectively lowered low-density lipoprotein cholesterol in a dose-dependent manner. Elevations in alanine aminotransferase, combined with a rare increase in bilirubin, presented potential hepatic safety issues, resulting in termination of development. The lapaquistat experience illustrates the current challenges in lipid-altering drug development.
Squalene synthase catalyzes the conversion of farnesyl diphosphate to squalene downstream from HMG CoA reductase in the cholesterol biosynthesis pathway (Figure 1), thus avoiding reduction of important intermediate metabolites such as ubiquinone, thought to be involved in development of myopathy,1 but still providing a similar mechanism of action as statins to upregulate low-density lipoprotein (LDL) receptors and to reduce LDL cholesterol (LDL-C). Since the 1970s, several squalene synthase inhibitors have undergone early-stage development as potential lipid-altering drugs.2
Editorial see p 1925
Clinical Perspective on p 1985
Lapaquistat acetate (Takeda Pharmaceutical Co. Ltd, Japan) was the only squalene synthase inhibitor to reach advanced clinical testing. In trials involving >6000 patients, lapaquistat produced a significant and consistent reduction in LDL-C levels at the 50- and 100-mg daily doses; however, 100 mg was associated with liver enzyme elevations that suggested potential hepatic toxicity. The 50-mg dose did not appear to have a similar risk of liver enzyme elevation, but the LDL-C reduction was deemed insufficient for commercial viability, and development was terminated in 2007.
This article summarizes the phase 2 and 3 clinical trial efficacy and safety of lapaquistat. The reasons for publishing an ultimately unsuccessful drug development program are 3-fold: to fulfill an ethical obligation to the many patients and investigators who participated in these trials; to contribute to the body of knowledge regarding this novel therapeutic class; and to provide insight into the regulatory environment within which lipid-altering agents are currently being developed, which has evolved considerably since the statins were first introduced in the 1980s.
The lapaquistat acetate development program involved 39 clinical studies from phases 1 through 3. This analysis includes the 12 phase 2 and 3 trials conducted at sites in the United States, Canada, Europe, Russia, South America, and South Africa between 2002 and 2007. Studies ranged from 6 to 96 weeks, and the data from all trials except the 96-week safety study, which is described separately, were pooled. All studies were randomized, double-blind, parallel, placebo- or active-controlled trials (except for 1 double-dummy trial) conducted in patients with hypercholesterolemia or mixed dyslipidemia. Participants had a 4- to 7-week diet run-in period that was maintained throughout the study. Four studies included monotherapy arms (3 compared lapaquistat with placebo; 2 included atorvastatin and simvastatin arms; and the fourth included ezetimibe and lapaquistat/ezetimibe arms without a placebo arm). All other studies assessed lapaquistat coadministered with statins, fenofibrate, or ezetimibe. All trials included lapaquistat 100 mg; 5 included a 50 mg/d; and 1 included 25 mg/d. On the basis of study design, atorvastatin, simvastatin, and rosuvastatin at various doses were the companion/comparator treatments. Fenofibrate was used in 1 study, and ezetimibe was permitted in 2 studies in addition to statin treatment.
In the 96-week study, patients received lapaquistat 100 mg/d, atorvastatin 10 mg/d, or both.
Patients in all trials had elevated LDL-C, >100 or >130 mg/dL (>2.6 or >3.4 mmol/L), depending on cardiovascular disease risk and concomitant statin therapy, and triglyceride levels <400 mg/dL (4.6 mmol/L). Trials also included heterozygous and homozygous familial hypercholesterolemia populations. Subjects with alanine aminotransferase (ALT) or aspartate aminotransferase (AST) levels >1.5 times the upper limit of normal (ULN), creatinine >1.5 mg/dL (133 μmol/L), or a creatine phosphokinase level >3 times the ULN were excluded.
The 96-week study enrolled patients with LDL-C levels of 130 to 190 mg/dL (3.4 to 4.9 mmol/L) with or without coronary heart disease and triglyceride levels <400 mg/dL (<4.6 mmol/L).
Studies were conducted under World Medical Association Declaration of Helsinki and good clinical practice guidelines and received ethics committee approval. All patients gave written informed consent before undergoing any study-related procedures.
Outcome Measures and Statistical Analysis
The primary efficacy end point was the percent change from baseline to the end of the study in LDL-C (direct) measured by preparative ultracentrifugation,3 with the exception of 1 study that used LDL-C calculated by the Friedewald formula.4 Secondary efficacy measures included non–high-density lipoprotein cholesterol (non-HDL-C), total cholesterol, apolipoprotein A1, apolipoprotein B, triglycerides, very-low-density lipoprotein cholesterol, HDL-C, and high-sensitivity (hs) C-reactive protein (CRP). Results were compared among treatment groups through the use of ANCOVA with terms for treatment, study, and baseline value as covariates. For lipid and apolipoprotein measurements, the baseline value was the mean of the available observations from the last 3 predose visits (within 3 weeks before randomization). For triglycerides and hs-CRP, baseline values and posttreatment values were expressed as medians, and a nonparametric analysis was used to compare the treatment groups. Missing values were extrapolated with the last observation carried forward method. P values for treatment comparisons and interactions were assessed at the 0.05 significance level.
Safety analyses included all randomized subjects who received at least 1 dose of study medication (the safety analysis set). Efficacy analyses included all patients who received at least 1 dose of double-blind study medication and in whom there was both a baseline value and at least 1 treatment value for that variable (the full analysis set).
To make appropriate comparisons, we report separately the efficacy for 3 monotherapy and 5 statin coadministration studies. The data from placebo groups in coadministration studies include patients in placebo-, active-, and statin-control groups. Safety analyses include all 12 of the placebo-controlled phase studies, with data pooled by treatment group. Because of the longer treatment period and the absence of a placebo group, the 96-week safety study is presented separately.
Exposure and Baseline Characteristics
A total of 6151 patients received double-blind study medication: 2462 received placebo/control therapy and 3689 received lapaquistat at a dose of 100 mg (n=2646), 50 mg (n=981), or 25 mg (n=62). Median duration of exposure was similar between placebo/control and lapaquistat (23.6 and 23.0 weeks, respectively). Patients on the 100-mg dose received the longest exposure (median, 23.5 weeks compared with 12.6 and 8.0 weeks for 50 and 25 mg, respectively). Overall completion rate was 89.5% for placebo and 88.1% for lapaquistat. Principal reasons for withdrawal in both placebo/control and lapaquistat groups were adverse events (AEs; 3.1% and 5.7%, respectively) and voluntary withdrawal of consent (3.8% and 3.3%, respectively).
Table 1 shows that the baseline demographic and clinical characteristics were mostly similar in the treatment groups. In all treatment groups (except for the small lapaquistat 25 mg group), a slight majority of patients had coronary heart disease or equivalent or at least 2 risk factors for coronary heart disease.
The actual median durations of exposure in the long-term study were between 95 and 96 weeks in all 3 groups, with 77.2%, 72.3%, and 73.7% in the atorvastatin, lapaquistat, and coadministration groups, respectively, completing at least 84 weeks of treatment. Principal reasons for withdrawal were AEs (8.8%, 13.4%, and 12.8%, respectively) and voluntary withdrawal (8.0%, 8.6%, and 7.7%, respectively).
Because of study designs, diet or stable statin treatment, and differing lipid entry criteria, the mean LDL-C at baseline was higher in the monotherapy compared with coadministration trials. Figure 2 illustrates the changes in LDL-C values from baseline to week 12 (monotherapy) or week 24 (coadministration) studies. LDL-C reductions for lapaquistat 50- or 100-mg monotherapy or combination therapy were highly statistically significant compared with placebo (P<0.001). Both 50- and 100-mg doses reduced LDL-C levels more in monotherapy than in coadministration trials.
Table 2 summarizes the changes in other lipids for monotherapy studies. Lapaquistat consistently reduced mean non–HDL-C, total cholesterol, apolipoprotein B, and very-low-density lipoprotein cholesterol levels and median triglycerides relative to placebo (P≤0.01). Although both doses of lapaquistat significantly reduced hs-CRP (P<0.001), reductions were larger with lapaquistat 100 mg than with lapaquistat 50 mg. Small, statistically nonsignificant increases in HDL-C levels occurred in all 3 dose groups; changes in apolipoprotein A1, although statistically significant between lapaquistat 100 mg and placebo, were negligible.
Table 3 summarizes the results for other parameters in the coadministration studies. When combined with other lipid-altering drugs, lapaquistat significantly reduced non–HDL-C, total cholesterol, apolipoprotein B, triglycerides, and very-low-density lipoprotein cholesterol compared with placebo (P≤0.006). Lapaquistat 100 mg reduced hs-CRP relative to placebo (P<0.001), whereas the reduction with 50 mg trended toward significance (P=0.059). HDL-C and apolipoprotein A1 were not statistically significantly altered.
Familial Hypercholesterolemia Populations
In the 12-week homozygous familial hypercholesterolemia (n=44) study, 21 patients received placebo and 23 received lapaquistat 50 or 100 mg. Baseline mean LDL-C levels on existing lipid-lowering therapy were 329 and 350 mg/dL (8.53 and 9.06 mmol/L), respectively. Lapaquistat reduced LDL-C levels by 10.66% compared with 3.59% for placebo (P=0.203). A large (650 subjects), 24-week coadministration study included patients with severe hypercholesterolemia, defined as LDL-C ≥100 mg/dL on stable treatment with simvastatin or atorvastatin 80 mg or rosuvastatin 40 mg. Genetically confirmed heterozygous familial hypercholesterolemia was present in 44.6%, and the majority of the others met the clinical definition for heterozygous familial hypercholesterolemia. Lapaquistat 100 mg reduced LDL-C an additional 16.4% relative to placebo, with no difference in those heterozygous familial hypercholesterolemia patients with or without confirmed mutations of the LDL receptor (no statistical analysis performed).
Table 4 summarizes the AEs. Data for lapaquistat monotherapy and coadministration with statins or other lipid-lowering agents are combined for this analysis and presented separately for the 25-, 50-, and 100-mg doses. Only 62 patients received lapaquistat 25 mg, so limited conclusions can be drawn from the data.
Among all groups, treatment-emergent AEs at any time during study participation ranged from 54.5% to 62.0%. Incidences of individual AEs were generally similar between the placebo and lapaquistat groups; the most common were headache, nasopharyngitis, myalgia, upper respiratory tract infection, diarrhea, and back pain, with no trend in any body system.
Lapaquistat did not appear to increase the incidence of serious AEs or deaths. A total of 9 patients, 5 receiving placebo and 4 receiving lapaquistat, died. All were judged by the investigator to be unrelated to study drug. No deaths were reportedly due to hepatic toxicity.
Table 5 displays AE findings in the long-term safety study. Overall AEs, events leading to withdrawal, and related events did not differ substantially among the 3 treatment groups, but were slightly lower overall in the atorvastatin monotherapy group. Serious AEs were similar across the treatment groups. Five deaths occurred: 1 each in the atorvastatin monotherapy and lapaquistat monotherapy groups and 3 in the lapaquistat acetate/atorvastatin coadministration group. None of the deaths suggested a causal relationship with study drug.
Cardiovascular AEs were recorded but not adjudicated, and were infrequent (<0.2% in all treatment groups) and showed no difference between the placebo and any lapaquistat group either in the safety data analysis or in the long-term trial.
Table 6 summarizes liver-associated enzymes and bilirubin. Overall, incidences of markedly abnormal values (≥3 times the ULN on ≥2 consecutive visits, ≥5 times the ULN, or ≥10 times the ULN) were low, with lapaquistat 50 mg daily being similar to placebo. Lapaquistat 100 mg had a 2% incidence of ALT ≥3 times the ULN on ≥2 consecutive visits compared with 0.3% for placebo. The incidence of ALT ≥5 times the ULN was 1.1% for lapaquistat 100 mg versus 0.2% for placebo; for ALT ≥10 times the ULN, the incidences were 0.3% and 0, respectively. The majority of the consecutive elevations occurred at 12 weeks or later (Figure 3).
Increases in AST paralleled ALT, but at a reduced rate; for lapaquistat 100 mg values ≥3 times the ULN on at least 2 consecutive visits, the rate was 0.8% versus 0.2% for placebo; for values ≥5 times the ULN, the rates were 0.8% and 0.1%, respectively; and for ≥10 times the ULN, the rates were 0.2% and 0, respectively. One patient (<0.1%) on lapaquistat experienced both sustained ALT >3 times the ULN and an increase in total bilirubin >2 times the ULN.
In the long-term study (Table 6), transaminase elevations in the lapaquistat 100 mg monotherapy group had incidences similar to the other trials: 0.3%, 1.7%, and 2.7% for ≥10, ≥5, and ≥3 times the ULN sustained, respectively. In the 10-mg atorvastatin group, the incidences were 0%, 0.4%, and 0.7% for ≥10, ≥5, and ≥3 times the ULN sustained, respectively. The incidences in the combination group were slightly lower than those seen on lapaquistat monotherapy. One patient receiving combination therapy had a sustained increase in ALT ≥3 times the ULN with ≥2 times the ULN in total bilirubin. In all treatment groups, the majority of the consecutive elevations occurred after at least 10 weeks of treatment.
Table 7 shows complete liver function data for the 2 patients meeting Food and Drug Administration–defined Hy Law criteria of ALT elevation accompanied by increases in total bilirubin.5 Because these cases were central to the decision to discontinue lapaquistat development, they are described in greater detail.
The first case was a 53-year-old man in a coadministration study receiving lapaquistat 100 mg and simvastatin 20 mg. Sixty days after initiation of study drug, the patient had an ALT of 103 U/L (reference range, 5 to 25 U/L), an AST of 140 U/L (reference range, 8 to 22 U/L), and a total bilirubin of 3.2 mg/dL (55.2 μmol/L; reference range, 0.1 to 1.1 mg/dL [1.7 to 18.8 μmol/L]). Bilirubin values were mildly elevated before randomization, but the bilirubin level at week 8 was >2 times the baseline level. The patient reported pruritus several days before the abnormal laboratory findings but no other symptoms. Lapaquistat and simvastatin were discontinued on day 62, and by the next day, the ALT, AST, and bilirubin values declined. At the next follow-up visit 16 days later, the pruritus, ALT, AST, and total bilirubin levels had fully resolved. The subject was withdrawn from the study and did not undergo any assessment, such as viral studies of other potential causes of liver enzyme elevation.
The other patient was a 55-year-old man randomized to lapaquistat 100 mg/atorvastatin 10 mg coadministration in the long-term study. On day 34, his ALT increased to 229 U/L (ULN, 25 U/L), AST to 392 U/L (ULN, 22 U/L), and total bilirubin to 9.75 mg/dL (ULN, 1.1 mg/dL), and his alkaline phosphatase was 225 U/L (ULN, 72 U/L). The patient reported nausea, vomiting, right midquadrant pain, erythema, rash, weakness, skin peeling, dark urine, and yellow sclera. Study medication was discontinued immediately, and within 5 days his symptoms had resolved or were improving. The laboratory abnormalities were partially resolved 17 days later and returned to baseline levels 41 days after discontinuation. Further investigation, including non–drug-related causes of hepatitis, such as viral hepatitis and autoimmune disorders, was negative.
Table 8 shows creatine phosphokinase values >5 and >10 times the ULN in both sets of studies, demonstrating a low frequency of events across the treatment groups. One subject receiving 100 mg lapaquistat in the long-term study experienced a serious AE (non–exercise-related muscle pain, tenderness, and creatine phosphokinase >10 times the ULN with an increase in serum creatinine). The investigator judged this to be unrelated to study drug, but due to hypokalemia, and the rhabdomyolysis resolved after potassium therapy. The subject was not rechallenged with lapaquistat. One subject on placebo also experienced an elevated creatine phosphokinase >10 times the ULN associated with muscle symptoms.
This report summarizes the safety and efficacy data from >6000 participants in the phase 2 and 3 studies for lapaquistat (see Table 9 for information on trial registration). In general, lapaquistat demonstrated significant improvements in lipid parameters and was well tolerated. Although theoretical reasons exist as to why squalene synthase inhibitors may help to avoid the muscle-related side effects associated with statins,1 the ability to assess whether lapaquistat afforded any muscle-related side effect benefit, either compared with statins or in combination with statins, is limited, because the evaluation occurred through spontaneous reports from study participants, not through a validated muscle symptom assessment tool. Furthermore, creatine phosphokinase data provide little to suggest that lapaquistat, alone or in combination with a statin, had a reduced incidence of myalgia. Liver safety signals prompted the most concern, and were the reason that this promising lipid-altering drug was terminated from further development. During the early preclinical development of lapaquistat, some hepatocellular changes were noted at high exposures in a multiple-dose canine model. However, multiple other nonclinical studies supported good hepatic safety and the decision to proceed with clinical development of lapaquistat.6
Hepatotoxicity is the single most common safety reason for drugs failing to gain regulatory approval or being withdrawn from the market.7 Definitive hepatic safety evaluations of investigational agents are challenging, because severe drug-induced liver injury is often rare, even with drugs known to incur an unacceptably high risk of hepatotoxicity. Other confounders include alternative causes of liver enzyme elevation, such as possible contributions from concurrent drug therapies and other medical conditions. Mild and moderate isolated transaminase elevations are often unreliable as surrogate markers of long-term injury owing to both the excess capacity of the liver to adapt to injury and the poor specificity of transaminase elevations, especially AST.8 When coupled with increases in bilirubin, however, transaminase elevations are a more reliable indicator of hepatocellular damage and predictor of serious outcomes.
The Hy Law is a term used to signify the high risk associated with persistent and significant ALT elevations in the presence of hyperbilirubinemia.5,9 It is named after Hyman Zimmerman, who, in the 1970s, described how drug-induced liver injury resulted in a high rate (10% to 50%) of mortality caused by acute liver failure in the pretransplantation era. The Hy Law is defined as elevations ≥3 times the ULN of ALT or AST, with increases in total bilirubin to >2 times the ULN with no other demonstrable cause, such as viral hepatitis A, B, or C, preexisting or acute hepatobiliary disease, or another drug capable of causing the observed injury. Hypothetically, for every 10 patients meeting the Hy Law, 1 patient will progress to acute liver failure.5 Furthermore, the Temple9 corollary to the Hy Law implies that for every 10 patients who experience significant elevations of ALT (eg, >5 to >10 times the ULN), 1 patient will meet the Hy Law. In the present analysis, 13 of 5107 subjects receiving any dose of lapaquistat experienced an ALT >10 times the ULN; thus, approximately one of these subjects would be expected to meet the Hy Law.9,10 Given that 1 patient (possibly 2 patients) did meet the Hy Law, this would appear consistent.
Regulatory agencies, such as the US Food and Drug Administration, interpret the Hy Law cases as representing hepatocellular injury sufficient to impair bilirubin excretion, and use such cases as a way to identify drugs likely to cause severe liver injury.10,11 With lapaquistat 100 mg, the frequency of ALT elevations >10 times the ULN was 0.3% (12 of 4064), and of the Hy Law cases was 0.05% (2 of 4064). The 2 Hy Law cases raised the most concern about potential severe liver toxicity. The case that occurred in the 24-week study had total bilirubin elevation before the start of the study drug, suggesting that the patient may have had a benign impairment of bilirubin metabolism, such as the Gilbert syndrome. This is important because, without the increase in bilirubin in this case, which may not have been study drug related, the mild elevations to 3 to 4 times the ULN in ALT and AST would not have met the criteria for the Hy Law, even if the transaminase elevations were drug related. This is complicated by the lack of additional diagnostic testing for alternative causes of the laboratory and clinical presentation, and complicates the objective certainty of this being a true case of drug-induced criteria fulfilling the Hy Law.
The second case was consistent with drug-induced liver injury in that it occurred within ≈1 month of exposure to combination therapy with lapaquistat 100 mg/atorvastatin 10 mg, involved severe increases in both ALT and bilirubin, and was associated with classic symptoms of hepatitis. Further diagnostic testing demonstrated no nondrug explanation for these abnormal laboratory and clinical findings. Both patients had rapid reduction and return to baseline values in bilirubin and transaminases on withdrawal of lapaquistat.
Although the data on lapaquistat 50 mg are far more limited in terms of both numbers of patients and duration of exposure, it appears that the lower dose had fewer ALT elevations and thus may be safer from a hepatic standpoint. However, the lapaquistat 50-mg dose was determined by Takeda to demonstrate insufficient lipid-lowering efficacy to justify the large long-term trials required to support a marketing application, even though the LDL-C reduction was as good as or better than current Food and Drug Administration–approved and marketed drugs, such as niacin and bile acid sequestrants.
To provide clinical and regulatory historical context to these 2 cases found in >6000 study participants, it is useful to review the hepatic safety data of the most widely used and evaluated statin over the last decade, atorvastatin. The incidence of sustained ALT elevations to >3 times the ULN at the atorvastatin 80-mg dose at the time of marketing approval was ≈2% to 3% on the basis of only ≈320 subjects treated for at least 1 year, ≈3 times that seen with 40 mg.12 The 80-mg dose provides an additional 5% to 6% additional LDL-C reduction over the 40-mg dose. One of the more carefully performed larger studies assessing the hepatic effects of high-dose atorvastatin was a double-blind, comparative trial against simvastatin in 826 patients.13 The study titrated atorvastatin from 20 to 80 mg over a 12-week period, and maintained the 80-mg dose for another 24 weeks. Of the 391 patients treated with atorvastatin 80 mg, 3.8% (15 of 391) had ALT levels >3 times the ULN, 2.1% (8 of 391) had ALT levels >5 times the ULN, and 1.2% (5 of 391) had ALT levels >10 times the ULN. One subject (0.25%) met the Hy Law criteria with ALT >3 times the ULN and bilirubin >2.0 mg/dL. Three additional subjects had ALT >3 times the ULN and exhibited increases in bilirubin levels above the ULN (normal range, <1.10 mg/dL). All of these patients were reported to have had normal baseline bilirubin levels. Two of the patients reported asthenia and fatigue, and 3 reported pruritus. The onset of ALT elevations occurred within 6 to 12 weeks after the start of the 80-mg dose, and elevated levels usually declined to normal within 5 weeks after discontinuing treatment.
This is not to suggest that atorvastatin 80 mg is unsafe, but rather that significant ALT elevations, even with bilirubin increases, may occur with other commonly used lipid-lowering drugs at doses generally presumed to be safe. Although case reports suggest that such occurrences represent drug-induced hepatitis or idiosyncratic reactions, others suggest that such liver findings may be related to inhibition of cholesterol synthesis or cholesterol reduction, and may be a theoretical exception to the Hy Law.13 If the latter were universally true, then statins that lowered cholesterol the most, such as rosuvastatin, would be associated with the greatest elevations in liver enzymes, which is not the finding from comparative clinical trials.13 Furthermore, despite widespread clinical use for >20 years, the reports of drug-induced liver injury with statins are limited to rare case reports or reports from clinical trials in which the causality of the severe hepatotoxicity was difficult to establish.14
The LDL-C lowering of squalene synthase inhibition is now well validated by the lapaquistat program in which 50 and 100 mg reduced LDL-C levels 18% and 23%, respectively, in monotherapy studies and 14% and 19%, respectively, in coadministration studies. The additional LDL-C decrease when lapaquistat was added to the highest dose of the 3 most widely used statins would be consistent with a further 8-fold increase in their doses based on an ≈5% to 7% additional LDL-C reduction for every doubling of statin dose.15 Lapaquistat 50 and 100 mg also produced statistically significant reductions compared with placebo in non–HDL-C, total cholesterol, apolipoprotein B, triglycerides, and very-low-density lipoprotein cholesterol, with none of the AEs on HDL-C seen with other agents in development that also cause significant transaminase elevations in short-term trials.16 Another potentially favorable, and perhaps unexpected, finding was the robust reduction in hs-CRP in the monotherapy (Table 2) and when added to existing statin (Table 3) compared with placebo. Lapaquistat 50 and 100 mg daily lowered median hs-CRP 16.7% and 25%, respectively, compared with a 9.52% increase on placebo (both P<0.001). When added to existing stable statin therapy, similar reductions of 14.29% (P=0.059) and 25% (P=0.001) were seen with lapaquistat 50 and 100 mg versus a decrease of 9.88% seen in the placebo-treated group. Although many lipid-lowering agents, including fibrates,17 niacin,18 and ezetimibe19 (when added to statins), reduce CRP, none of these agents has a mechanism similar to statins; thus, it is likely that CRP reduction is mediated via a reduction in either the plasma atherogenic apolipoprotein B–containing lipoproteins, very-low-density lipoprotein and LDL, or intrahepatic cholesterol. This is in contrast to the widely disseminated view of statin-related CRP reduction being caused by some as-yet unproven and independent pleiotrophic effect.
There are also abundant data linking different statins,20 and the same statins at different doses,21 that produce greater LDL-C reductions with greater decreases in CRP. In this large development program of lapaquistat, the effects on CRP were very similar to those seen for statins, especially for monotherapy, and contrast with ezetimibe monotherapy in which a similar reduction in LDL-C produced by ezetimibe is not associated with CRP reduction.22 This would suggest that inhibition of intracellular, especially intrahepatic, cholesterol may be a more significant driver of CRP reduction than circulating LDL levels. However, as a consequence, it is currently impossible with lipid-lowering agents, including lapaquistat, to separate the lipid- and CRP-lowering effects or their possible independent effect on cardiovascular disease. Furthermore, no non–lipid-lowering agent, such as aspirin, has yet been described that independently lowers CRP that does not also have a known direct effect on cardiovascular disease risk. Therefore, it has been impossible so far to test the hypothesis that lowering CRP elevation alone will have any impact on cardiovascular risk. Despite these encouraging efficacy results, lapaquistat acetate development was terminated owing to the uncertainty regarding hepatic safety in 2 (or arguably 1) lapaquistat-treated subjects of 5107 exposed to the drug. It is possible that 2 decades ago, such hepatic findings in phase 2/3 may not have prevented the continued development and potential approval of lapaquistat. Answering the question regarding hepatic safety definitively for lapaquistat in the current regulatory environment would have required much larger and longer trials before marketing. Although it is true that similar hepatic safety trials were required and conducted for the early statins, they were done only after the drugs were approved and marketed.23
Lapaquistat acetate is the first, and so far only, squalene synthase inhibitor to undergo extensive clinical development. It was efficacious in lowering LDL-C, but 2 cases of severe liver enzyme elevations in >5100 study participants exposed to the drug resulted in termination of the development program. It is hoped that the publication of the results and experiences of this clinical trial development program is beneficial to future researchers of squalene synthase inhibitors and illustrative to clinicians in terms of the challenges faced in the development of lipid-altering drug therapies.
Sources of Funding
The studies were sponsored by Takeda Global Research and Development Center, Inc., Deerfield, IL. The authors are grateful to them for access to and permission to publish the development program results.
Dr Stein has received research grants related to trials of lapaquistat and consulting fees from Takeda. He has also received grants for trials of numerous lipid-modifying agents, consulting fees, and honoraria for professional input regarding lipid-altering agents and/or has delivered lectures for AACC, Abbott, AstraZeneca, Food and Drug Administration, F. Hoffmann–La Roche, GSK, ISIS, Merck & Co, National Lipid Association, Novartis, Reliant, Sankyo, Sanofi-aventis, Schering-Plough, and Wyeth. Dr Bays has received research grants from and has served as a consultant for Takeda, as well as numerous other pharmaceutical companies. J. Pedicano and Drs O'Brien, Piper, and Spezzi are employees of Takeda.
Initial development of the manuscript concept and content was done by the authors, as were subsequent revisions; assistance with the process was provided by Daniel Knight and Patrick Covernton and was supported by Takeda Global Research and Development Center, Inc.
Continuing medical education (CME) credit is available for this article. Go to http://cme.ahajournals.org to take the quiz.
- Received June 30, 2010.
- Accepted February 25, 2011.
- © 2011 American Heart Association, Inc.
Lipid Research Clinics. Manual of Laboratory Operations: Lipid and Lipoprotein Analysis (Revised). Washington, DC: US Government Printing Office; 1982. US Department of Health and Human Services publication No. (NIH) 75-678.
- Friedewald WT,
- Levy RI,
- Fredrickson DS
- Temple R
US Food and Drug Administration Guidance for Industry. Drug-Induced Liver Injury: Premarketing Clinical Evaluation. July 2009. www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM174090.pdf Accessed October 30, 2009.
Lipitor (atorvastatin calcium tablets) [US prescribing information[. Morris Plains, NJ: Parke-Davis Division of Warner-Lambert Co; January 1997.
- Illingworth DR,
- Crouse JR 3rd.,
- Hunninghake DB,
- Davidson MH,
- Escobar ID,
- Stalenhoef AF,
- Paragh G,
- Ma PT,
- Liu M,
- Melino MR,
- O'Grady L,
- Mercuri M,
- Mitchel YB
- Grundy SM,
- Vega GL,
- McGovern ME,
- Tulloch BR,
- Kendall DM,
- Fritz-Patrick D,
- Ganda OP,
- Rosenson RS,
- Buse JB,
- Robertson DD,
- Sheehan JP
- Despres JP,
- Lemieux I,
- Pascot A,
- Alméras N,
- Dumont M,
- Nadeau A,
- Bergeron J,
- Prud'homme D
- Ballantyne CM,
- Houri J,
- Notarbartolo A,
- Melani L,
- Lipka LJ,
- Suresh R,
- Sun S,
- LeBeaut AP,
- Sager PT,
- Veltri EP
- Pearson TA,
- Ballantyne CM,
- Veltri E,
- Shah A,
- Bird S,
- Lin J,
- Rosenberg E,
- Tershakovec AM
- Bradford RH,
- Shear CL,
- Chremos AN,
- Dujovne C,
- Downton M,
- Franklin FA,
- Gould AL,
- Hesney M,
- Higgins J,
- Hurley DP,
- Langendorfer A,
- Nash DT,
- Pool JL,
- Schnaper H
Elevated low-density lipoprotein cholesterol is the cornerstone of coronary artery disease prevention, and although statins have provided enormous advances, there remains an unmet need for new effective, well-tolerated, safe low-density lipoprotein cholesterol–lowering drugs. This report summarizes the phase 2 and 3 results from the lapaquistat clinical program, which was halted at an advanced stage as a result of potential hepatic safety issues. Lapaquistat acetate is a squalene synthase inhibitor, a step after the statins, investigated for the treatment of hypercholesterolemia. Data were pooled from 12 studies (n=6151) lasting 6 to 96 weeks. Trials were randomized, double-blind, parallel, and placebo or active controlled with lapaquistat monotherapy or coadministration with other lipid-altering drugs. All studies included lapaquistat 100 mg daily; 5 included 50 mg; and 1 included 25 mg. The main outcome measure was low-density lipoprotein cholesterol, secondary lipid/metabolic parameters, and overall safety. Lapaquistat 100 mg significantly decreased low-density lipoprotein cholesterol by 21.6% in monotherapy and 18.0% in combination with a statin. It also significantly reduced C-reactive protein. Total adverse events were higher for lapaquistat than placebo, although individual events were similar. At 100 mg, alanine aminotransferase values ≥3 times the upper limit of normal on ≥2 consecutive visits increased (2.0% to 2.7% versus 0.3% to 0.7% for placebo or low-dose atorvastatin). Two patients receiving lapaquistat 100 mg met the Hy Law criteria of alanine aminotransferase elevation plus increased total bilirubin. Squalene synthase inhibition with lapaquistat, alone or in combination with statins, effectively lowered low-density lipoprotein cholesterol and C-reactive protein. Elevations in alanine aminotransferase combined with a rare increase in bilirubin presented potential hepatic safety issues, resulting in termination of development. The lapaquistat experience illustrates the current challenges in lipid-altering drug development.