Conditional borrowing external data to establish a hybrid control arm in randomized clinical trials.

Utilizing external data from the real world, including data from historical clinical trials, has received increasing interest in drug development. The use of external data to support drug evaluation in clinical trials has mainly been through using various matching methods for baseline characteristics to form external control arms in single-arm trials or to augment control arms of randomized controlled trials in hybrid approaches. However, matching the baseline characteristics between the trial and the external subjects can only guarantee comparability on the level of baseline characteristics. Differences in outcomes between the two data sources may still exist due to contemporaneous and operational characteristics. Similarity between the outcomes in the trial control and the external subjects with similar baseline characteristics can be critical in leveraging the external subjects in the clinical trials. In this paper, a resampling method for augmenting control arms in randomized controlled trials is proposed under the conditional borrowing framework. The new method establishes empirical distributions for the hazard ratio in outcomes between the external and trial control subjects. The borrowing decision is then derived from this empirical distribution using a measure of similarity. Once the borrowing decision is established, the borrowing weights for the external subjects, based on the similarity measure, are incorporated in the weighted partial likelihood to evaluate the treatment effect. The operating characteristics of the hybrid control arm, under both the conditional borrowing and unconditional borrowing frameworks, are evaluated. Simulation is conducted to evaluate Type I error, bias, and power. An illustrative example using simulated data is also presented.

Discordant responses of plasma low-density lipoprotein cholesterol and lipoprotein(a) to alirocumab: A pooled analysis from 10 ODYSSEY Phase 3 studies.

AIMS: Proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors consistently reduce low-density lipoprotein cholesterol (LDL-C) by 50-60% and lipoprotein(a) (Lp(a)) by 20-30%, but the mechanism of Lp(a) lowering remains unclear. If Lp(a) is cleared by the LDL receptor, similar to LDL-C, then one would expect PCSK9 inhibition to induce a concordant LDL-C/Lp(a) response in an approximately 2:1 ratio. We aim to determine the prevalence of discordant plasma LDL-C/Lp(a) response to the PCSK9 inhibitor alirocumab. METHODS: This is a post hoc, pooled analysis of 10 randomized controlled trials from the ODYSSEY Phase 3 clinical trial program for alirocumab. Patients enrolled in the trials were high cardiovascular risk and/or with heterozygous familial hypercholesterolemia. The primary end point was prevalence of discordant LDL-C/Lp(a) response to alirocumab at 24 weeks. Discordant response was defined as LDL-C reduction >35% and Lp(a) reduction ≤10%, or LDL-C reduction ≤35% and Lp(a) reduction >10%. RESULTS: Of the 1709 patients in the pooled study cohort, 62.4% were male, and the mean age was 59.2 (SD: 11.0) years. Baseline mean LDL-C was 126.5 (SD: 46.3) mg/dL and baseline median Lp(a) was 46.9 (interquartile range: 21.8-89.0) mg/dL. Total prevalence of discordant LDL-C/Lp(a) response was 21.5% (12.6% with LDL-C >35% reduction and Lp(a) ≤10% reduction; 8.9% with LDL-C ≤35% reduction and Lp(a) >10% reduction). Baseline Lp(a) and familial hypercholesterolemia status did not affect discordance. CONCLUSION: A high prevalence of discordant LDL-C/Lp(a) response was observed with alirocumab, further suggesting that PCSK9 inhibitor therapy with alirocumab reduces plasma Lp(a) through alternative pathways to LDL receptor clearance.

Proof of Concept: Drug Selection? Or Dose Selection? Thoughts on Multiplicity Issues.

In new drug development process, one of the most important milestones for a drug candidate is to establish Proof of Concept (PoC) at early Phase II stage. Among many challenges in PoC clinical trial design and analysis, the application of multiplicity comparison procedures (MCP) is frequently discussed when multiple doses or drugs are included in one PoC study. In such discussion, one fundamental question of applying multiplicity adjustment is which error one should consider to control and at what level. Should it be the experiment-wise error or the compound-wise error? In this paper, the multiplicity issues in two cases of PoC studies are used as examples to discuss the concept of different types of error and the level of the error rate control. With a clear understanding of the type of error and error rate control, the debate of applications of the multiplicity adjustment procedures in the PoC studies can be reconciled.

Intensity of statin treatment after acute coronary syndrome, residual risk, and its modification by alirocumab: insights from the ODYSSEY OUTCOMES trial.

AIMS: Statins are pivotal to the secondary prevention of major adverse cardiovascular events, but some patients are statin-intolerant. We examined the effects of the proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitor alirocumab on the risk of major adverse cardiovascular events according to the intensity of background statin treatment. METHODS AND RESULTS: The ODYSSEY OUTCOMES trial compared alirocumab with placebo in 18,924 patients with acute coronary syndrome and dyslipidaemia despite intensive or maximum-tolerated statin treatment (including no statin if intolerance was documented). The primary outcome (major adverse cardiovascular events) comprised coronary heart disease death, non-fatal myocardial infarction, ischaemic stroke, or unstable angina. Median follow-up was 2.8 years. Baseline statin treatment was high-intensity (88.8%), low/moderate-intensity (8.7%) or none (2.4%). Median baseline low-density lipoprotein cholesterol was 86, 89 and 139 mg/dL (P < 0.001) in these statin treatment categories, respectively. Alirocumab produced similar relative reductions in low-density lipoprotein cholesterol from baseline across statin treatment subgroups, but the mean absolute reductions differed (52.9, 56.7 and 86.1 mg/dL, respectively; P < 0.001). With placebo, the incidence of major adverse cardiovascular events was highest in the no statin subgroup (10.8%, 10.7% and 26.0% respectively). Alirocumab reduced major adverse cardiovascular events in each statin subgroup (hazard ratio 0.88, 95% confidence interval (CI) 0.80-0.96; 0.68, 0.49-0.94; and 0.65, 0.44-0.97, respectively; Pinteraction = 0.14) with a gradient of absolute risk reduction: 1.25%, 95% CI 0.34-2.16; 3.16%, 0.38-5.94; 7.97%, 0.42-15.51; Pinteraction = 0.106). CONCLUSIONS: PCSK9 inhibition with alirocumab reduces the relative risk of major adverse cardiovascular events after acute coronary syndrome irrespective of background statin treatment. However, patients on no statin are at high absolute risk for recurrent major adverse cardiovascular events; alirocumab substantially reduces that risk. PCSK9 inhibition may be an important therapeutic strategy for statin-intolerant patients with acute coronary syndrome.

Clinical Efficacy and Safety of Alirocumab After Acute Coronary Syndrome According to Achieved Level of Low-Density Lipoprotein Cholesterol: A Propensity Score-Matched Analysis of the ODYSSEY OUTCOMES Trial.

BACKGROUND: Recent international guidelines have lowered recommended target levels of low-density lipoprotein cholesterol (LDL-C) for patients at very high risk for major adverse cardiovascular events (MACE). However, uncertainty persists whether additional benefit results from achieved LDL-C levels below the conventional targets. Inferences from previous analyses are limited because patients who achieve lower versus higher LDL-C on lipid-lowering therapy differ in other characteristics prognostic for MACE and because few achieved very low LDL-C levels. To overcome these limitations, we performed a propensity score-matching analysis of the ODYSSEY OUTCOMES trial (Evaluation of Cardiovascular Outcomes After an Acute Coronary Syndrome During Treatment With Alirocumab) which compared alirocumab with placebo in 18 924 patients with recent acute coronary syndrome receiving intensive or maximum-tolerated statin treatment. METHODS: Patients on alirocumab were classified in prespecified strata of LDL-C achieved at 4 months of treatment: <25 (n=3357), 25 to 50 (n=3692), or >50 mg/dL (n=2197). For each stratum, MACE (coronary heart disease death, nonfatal myocardial infarction, ischemic stroke, or hospitalization for unstable angina) after month 4 was compared in patients receiving placebo with similar baseline characteristics and adherence by using 1:1 propensity score matching. RESULTS: Across achieved LDL-C strata of the alirocumab group, patients differed by baseline LDL-C, lipoprotein(a), use of intensive statin therapy, study medication adherence, and other demographic, medical history, biometric, and laboratory criteria. After propensity score matching, characteristics were similar in corresponding patients of the alirocumab and placebo groups. Treatment hazard ratio, 95% CI, and absolute risk reduction (number per 100 patient-years) for MACE were similar in those with achieved LDL-C <25 mg/dL (hazard ratio, 0.74 [95% CI, 0.62-0.89]; absolute risk reduction, 0.92) or 25 to 50 mg/dL (hazard ratio, 0.74 [95% CI, 0.64-0.87]; absolute risk reduction, 1.05). Patients with achieved LDL-C >50 mg/dL had poorer adherence and derived less benefit (hazard ratio, 0.87 [95% CI, 0.73-1.04]; absolute risk reduction, 0.62). No safety concerns were associated with a limited period of LDL-C levels <15 mg/dL. CONCLUSIONS: After accounting for differences in baseline characteristics and adherence, patients treated with alirocumab who achieved LDL-C levels <25 mg/dL had a reduction in the risk of MACE that was similar to that of patients who achieved LDL-C levels of 25 to 50 mg/dL. Registration: URL: https://www.clinicaltrials.gov; Unique identifier: NCT01663402.

Effect of alirocumab on major adverse cardiovascular events according to renal function in patients with a recent acute coronary syndrome: prespecified analysis from the ODYSSEY OUTCOMES randomized clinical trial.

AIMS: Statins reduce cardiovascular risk in patients with acute coronary syndrome (ACS) and normal-to-moderately impaired renal function. It is not known whether proprotein convertase subtilisin-kexin type 9 (PCSK9) inhibitors provide similar benefit across a range of renal function. We determined whether effects of the PCSK9 inhibitor alirocumab to reduce cardiovascular events and death after ACS are influenced by renal function. METHODS AND RESULTS: ODYSSEY OUTCOMES compared alirocumab with placebo in patients with recent ACS and dyslipidaemia despite intensive statin treatment. Estimated glomerular filtration rate (eGFR) <30 mL/min/1.73 m2 was exclusionary. In 18 918 patients, baseline eGFR was 82.8 ± 17.6 mL/min/1.73 m2, and low-density lipoprotein cholesterol (LDL-C) was 92 ± 31 mg/dL. At 36 months, alirocumab decreased LDL-C by 48.5% vs. placebo but did not affect eGFR (P = 0.65). Overall, alirocumab reduced risk of the primary outcome (coronary heart disease death, non-fatal myocardial infarction, ischaemic stroke, or unstable angina requiring hospitalization) with fewer deaths. There was no interaction between continuous eGFR and treatment on the primary outcome or death (P = 0.14 and 0.59, respectively). Alirocumab reduced primary outcomes in patients with eGFR ≥90 mL/min/1.73 m2 (n = 7470; hazard ratio 0.784, 95% confidence interval 0.670-0.919; P = 0.003) and 60 to <90 (n = 9326; 0.833, 0.731-0.949; P = 0.006), but not in those with eGFR < 60 (n = 2122; 0.974, 0.805-1.178; P = 0.784). Adverse events other than local injection-site reactions were similar in both groups across all categories of eGFR. CONCLUSIONS: In patients with recent ACS, alirocumab was associated with fewer cardiovascular events and deaths across the range of renal function studied, with larger relative risk reductions in those with eGFR > 60 mL/min/1.73 m2.

Cost-Effectiveness of Alirocumab in Patients With Acute Coronary Syndromes: The ODYSSEY OUTCOMES Trial.

BACKGROUND: Cholesterol reduction with proprotein convertase subtilisin-kexin type 9 inhibitors reduces ischemic events; however, the cost-effectiveness in statin-treated patients with recent acute coronary syndrome remains uncertain. OBJECTIVES: This study sought to determine whether further cholesterol reduction with alirocumab would be cost-effective in patients with a recent acute coronary syndrome on optimal statin therapy. METHODS: A cost-effectiveness model leveraging patient-level data from ODYSSEY OUTCOMES (Evaluation of Cardiovascular Outcomes After an Acute Coronary Syndrome During Treatment With Alirocumab) was developed to estimate costs and outcomes over a lifetime horizon. Patients (n = 18,924) had a recent acute coronary syndrome and were on high-intensity or maximum-tolerated statin therapy, with a baseline low-density lipoprotein cholesterol (LDL-C) level ≥70 mg/dl, non-high-density lipoprotein cholesterol ≥100 mg/dl, or apolipoprotein B ≥80 mg/l. Alirocumab 75 mg or placebo was administered subcutaneously every 2 weeks. Alirocumab was blindly titrated to 150 mg if LDL-C remained ≥50 mg/dl or switched to placebo if 2 consecutive LDL-C levels were <15 mg/dl. Incremental cost per quality-adjusted life-year (QALY) was determined with the addition of alirocumab versus placebo and, based on clinical efficacy findings from the trial, was stratified by baseline LDL-C levels ≥100 mg/dl and <100 mg/dl. RESULTS: Across the overall population recruited to the ODYSSEY OUTCOMES trial, using an annual treatment cost of US$5,850, the mean overall incremental cost-effectiveness ratio was US$92,200 per QALY (base case). The cost was US$41,800 per QALY in patients with baseline LDL-C ≥100 mg/dl, whereas in those with LDL-C <100 mg/dl the cost per QALY was US$299,400. Among patients with LDL-C ≥100 mg/dl, incremental cost-effectiveness ratios remained below US$100,000 per QALY across a wide variety of sensitivity analyses. CONCLUSIONS: In patients with a recent acute coronary syndrome on optimal statin therapy, alirocumab improves cardiovascular outcomes at costs considered intermediate value, with good value in patients with baseline LDL-C ≥100 mg/dl but less economic value with LDL-C <100 mg/dl. (Evaluation of Cardiovascular Outcomes After an Acute Coronary Syndrome During Treatment With Alirocumab [ODYSSEY OUTCOMES]; NCT01663402).

Risk Categorization Using New American College of Cardiology/American Heart Association Guidelines for Cholesterol Management and Its Relation to Alirocumab Treatment Following Acute Coronary Syndromes.

BACKGROUND: The 2018 US cholesterol management guidelines recommend additional lipid-lowering therapies for secondary prevention in patients with low-density lipoprotein cholesterol ≥70 mg/dL or non-high-density lipoprotein cholesterol ≥100 mg/dL despite maximum tolerated statin therapy. Such patients are considered at very high risk (VHR) based on a history of >1 major atherosclerotic cardiovascular disease (ASCVD) event or a single ASCVD event and multiple high-risk conditions. We investigated the association of US guideline-defined risk categories with the occurrence of ischemic events after acute coronary syndrome and reduction of those events by alirocumab, a PCSK9 (proprotein convertase subtilisin/kexin type 9) inhibitor. METHODS: In the ODYSSEY OUTCOMES trial (Evaluation of Cardiovascular Outcomes After an Acute Coronary Syndrome During Treatment With Alirocumab), patients with recent acute coronary syndrome and residual dyslipidemia despite optimal statin therapy were randomly assigned to alirocumab or placebo. The primary trial outcome (major adverse cardiovascular events, ie, coronary heart disease death, nonfatal myocardial infarction, ischemic stroke, or hospitalization for unstable angina) was examined according to American College of Cardiology/American Heart Association risk category. RESULTS: Of 18 924 participants followed for a median of 2.8 years, 11 935 (63.1%) were classified as VHR: 4450 (37.3%) had multiple prior ASCVD events and 7485 (62.7%) had 1 major ASCVD event and multiple high-risk conditions. Major adverse cardiovascular events occurred in 14.4% of placebo-treated patients at VHR versus 5.6% of those not at VHR. In the VHR category, major adverse cardiovascular events occurred in 20.4% with multiple prior ASCVD events versus 10.7% with 1 ASCVD event and multiple high-risk conditions. Alirocumab was associated with consistent relative risk reductions in both risk categories (hazard ratio=0.84 for VHR; hazard ratio=0.86 for not VHR; Pinteraction=0.820) and by stratification within the VHR group (hazard ratio=0.86 for multiple prior ASCVD events; hazard ratio=0.82 for 1 major ASCVD event and multiple high-risk conditions; Pinteraction=0.672). The absolute risk reduction for major adverse cardiovascular events with alirocumab was numerically greater (but not statistically different) in the VHR group versus those not at VHR (2.1% versus 0.8%; Pinteraction=0.095) and among patients at VHR with multiple prior ASCVD events versus a single prior ASCVD event (2.4% versus 1.8%; Pinteraction=0.661). CONCLUSIONS: The US guideline criteria identify patients with recent acute coronary syndrome and dyslipidemia who are at VHR for recurrent ischemic events and who may derive a larger absolute benefit from treatment with alirocumab. CLINICAL TRIAL REGISTRATION: URL: https://www.clinicaltrials.gov. Unique identifier: NCT01663402.

Evaluating co-primary endpoints collectively in clinical trials.

Often a treatment is assessed by co-primary endpoints so that a comprehensive picture of the treatment effect can be obtained. Co-primary endpoints can be different medical assessments angled at different aspects of a disease, therefore, are used collectively to strengthen evidence for the treatment effect. It is common sense that if a treatment is ineffective, the chance to show that the treatment is effective in all co-primary endpoints should be small. Therefore, it may not be necessary to require all the co-primary endpoints to be statistically significant at the 1-sided 0.025 level to control the error rate of wrongly approving an ineffective treatment. Rather it is reasonable to allow certain variation for the p -values within a range close to 0.025. In this paper, statistical methods are developed to derive decision rules to evaluate co-primary endpoints collectively. The decision rules control the error rate of wrongly accepting an ineffective treatment at the level of 0.025 for a study and the error rate at a slightly higher level for a treatment that works for all the co-primary endpoints except perhaps one. The decision rules also control the error rates for individual endpoints. Potential applications in clinical trials are presented.

Bioequivalence approaches for highly variable drugs and drug products.

Over the past decade, concerns have been expressed increasingly regarding the difficulty for highly variable drugs and drug products (%CV greater than 30) to meet the standard bioequivalence (BE) criteria using a reasonable number of study subjects. The topic has been discussed on numerous occasions at national and international meetings. Despite the lack of a universally accepted solution for the issue, regulatory agencies generally agree that an adjustment of the traditional BE limits for these drugs or products may be warranted to alleviate the resource burden of studying relatively large numbers of subjects in bioequivalence trials. This report summarizes a careful examination of all the statistical methods available and extensive simulations for BE assessment of highly variable drugs/products. Herein, the authors present an approach of scaling an average BE criterion to the within-subject variability of the reference product in a crossover BE study, together with a point-estimate constraint imposed on the geometric mean ratio between the test and reference products. The use of a reference-scaling approach involves the determination of variability of the reference product, which requires replication of the reference treatment in each individual. A partial replicated-treatment design with this new data analysis methodology will thus provide a more efficient design for BE studies with highly variable drugs and drug products.

Comparisons of test statistics arising from marginal analyses of multivariate survival data.

We investigate the properties of several statistical tests for comparing treatment groups with respect to multivariate survival data, based on the marginal analysis approach introduced by Wei, Lin and Weissfeld ["Regression Analysis of multivariate incomplete failure time data by modelling marginal distributions," JASA vol. 84 pp. 1065-1073]. We consider two types of directional tests, based on a constrained maximization and on linear combinations of the unconstrained maximizer of the working likelihood function, and the omnibus test arising from the same working likelihood. The directional tests are members of a larger class of tests, from which an asymptotically optimal test can be found. We compare the asymptotic powers of the tests under general contiguous alternatives for a variety of settings, and also consider the choice of the number of survival times to include in the multivariate outcome. We illustrate the results with simulations and with the results from a clinical trial examining recurring opportunistic infections in persons with HIV.

A decision rule for evaluating several independent clinical trials collectively.

Usually, in applying for market approval of a new drug, more than one similarly designed clinical trial is conducted to support efficacy claims. How to evaluate these trials collectively and assess the overall type I error of a decision rule can be a challenging statistical issue. In this paper, we propose a decision rule to collectively evaluate p-values from several similarly designed and independently conducted trials. A concept of overall hypotheses, which is essentially union or intersection combinations of individual trials' hypotheses, is used so that the overall type I error can be controlled at desired levels. We also discuss some important properties of the approach, including the selection of the overall type I error rates and power. Examples are presented.