LOWERING LDL LEVELS WITH REPATHA (EVOLOCUMAB) DID NOT ADVERSELY affect COGNITIve Function IN LANDMARK PHASE 3 Study

• Repatha Cognitive Function Trial Evaluated Subset of Patients From Cardiovascular Outcomes Trial
  
• No Signal for Worsening Cognitive Function Observed Even in Patients Who Reached LDL Cholesterol Below 25 mg/dL
  
• Detailed Results From One of the Largest Randomized, Controlled Studies of Cognitive Function Presented at a Late-Breaking Clinical Trials Session at ACC.17

THOUSAND OAKS, Calif. (March 18, 2017) – Amgen (NASDAQ:AMGN) today announced detailed results from the Repatha^® (evolocumab) cognitive function trial (EBBINGHAUS) evaluating the impact on cognitive function in 1,974 patients enrolled in the Repatha cardiovascular outcomes study (FOURIER). The study demonstrated that the effect of Repatha on the primary endpoint of executive function was non-inferior to placebo. In addition, there was no statistical difference between Repatha and placebo on the other cognitive domains tested: working memory, memory function and psychomotor speed (secondary endpoints). The detailed results from EBBINGHAUS were presented at a Late-Breaking Clinical Trials Session at the American College of Cardiology 66^th Annual Scientific Session (ACC.17) inWashington, D.C.

“There has long been a debate that low LDL cholesterol levels could lead to negative effects on memory or other cognitive functions,” said Robert P. Giugliano, M.D., S.M., Brigham and Women’s Hospital, Boston and lead study investigator. “We did not find evidence for a decline in neurocognitive function after nearly two years of treatment with evolocumab using a dedicated series of neuropsychologic tests. We also asked patients and their physicians to provide their assessments and found no differences between evolocumab and placebo. These findings provide strong support for the safety of reducing LDL with evolocumab to levels well below current treatment targets.”

In the primary cohort of 1,204 patients, followed for a median of 19 months, the change from baseline raw score of spatial working memory strategy index of executive function was similar in the Repatha and placebo groups (mean baseline score 17.8; mean change from baseline ‑0.2 versus -0.3, respectively). The primary endpoint was below the pre-specified margin, demonstrating non-inferiority. The primary endpoint was assessed by the Cambridge Neuropsychological Test Automated Battery (CANTAB) Spatial Working Memory strategy index of executive function, an established language- and culture-independent computerized, tablet-based cognitive assessment tool.

“These results, from one of the largest randomized controlled trials on cognitive function, clearly demonstrated that lowering LDL-C to unprecedented levels with Repatha did not negatively impact cognition,” said Sean E. Harper, M.D., executive vice president of Research and Development at Amgen. “These results further support the safety of Repatha as an effective LDL-lowering therapy with no observed negative effects on memory or other cognitive functions.”

Secondary endpoint results in the three cognitive domains of working memory, memory function and psychomotor speed were consistent with the primary endpoint result. Patients treated with Repatha experienced change from baseline similar to placebo in all three cognitive domains tested: spatial working memory between-errors score (baseline 21.0, Repatha -0.5 versus placebo -0.9), paired associates learning total errors adjusted (baseline 25.8, change with Repatha -1.5 versus -1.5 with placebo), RTI median five-choice reaction time (baseline 355.9, 5.2 milliseconds change with Repatha versus 0.9 milliseconds with placebo). Changes from baseline in the global composite score were similar between treatment arms (baseline -0.008, changes with Repatha 0.03 versus 0.06 with placebo). Results in the full cohort were consistent with those in the primary cohort.

In an exploratory analysis, these results were consistent regardless of achieved low-density lipoprotein cholesterol (LDL-C) levels, including 661 patients with the lowest LDL-C level (<25 mg/dL).

In the EBBINGHAUS study, neurocognitive adverse event rates were similar between treatment arms. In the full cohort, 19 (1.9 percent) neurocognitive adverse events were reported in the Repatha group compared to 16 (1.6 percent) events in the placebo group. In the 27,564-patient Repatha cardiovascular outcomes trial (FOURIER), neurocognitive adverse events were reported in 1.6 percent in the Repatha group compared to 1.5 percent in the placebo group. The adverse events identified in EBBINGHAUS were consistent with the adverse events identified in FOURIER.

Repatha Cognitive Function (EBBINGHAUS) Study Design

EBBINGHAUS (Evaluating PCSK9 Binding antiBody Influence oN coGnitive HeAlth in high cardiovascUlar risk Subjects) is a double-blind, placebo-controlled randomized non-inferiority trial involving of 1,974 patients with clinically evident atherosclerotic cardiovascular disease enrolled in the Repatha cardiovascular outcomes trial (FOURIER). The primary non-inferiority assessment of the primary endpoint spatial working memory strategy index of executive function (assessing executive function, or high-level thinking and decision making) was performed on the primary cohort of 1,204 patients who enrolled on or before the first dose of investigational product and had at least one post-baseline CANTAB assessment. The full cohort (1,974 patients) included all randomized patients. The primary endpoint was assessed by comparing the 95 percent confidence interval (CI) with the pre-specified non-inferiority margin for the treatment difference between Repatha and placebo. Secondary endpoints were the CANTAB Spatial Working Memory between-errors score (assessing working memory, or the ability to hold material in mind while that material is being actively processed); the CANTAB Paired Associates Learning Total Errors Adjusted (assessing memory function, or the ability to store and retrieve information by associating an event with a time and place); and the CANTAB Reaction Time Five-Choice Median Reaction Time (assessing psychomotor speed, which is responsible for detecting and responding to a stimulus). For all three secondary endpoints, the 95 percent CI for the estimated treatment difference between Repatha and placebo spanned equivalence.

Primary and secondary endpoints were assessed using a tablet-based tool at baseline, weeks 24, yearly, and at study end. The primary analysis compared the mean change from baseline in patients who had a baseline cognitive assessment on or prior to the first day of study drug.