What We Do
More Efficient
Trial Enrollment
Inform Policy
Numbers
Label
Expansion
Estimate Patient Numbers
Putting Trial Results
in Context
More Efficient Trial Enrollment – Perisphere has developed an algorithm to identify patients fast and efficiently for trial enrollment
Context
A trial may screen 10-20 patients to identify one participant
Objective: Develop an algorithm to reduce the number needed to screen
Approach
- Identified five data sets with biomarker measurements
- Conducted machine learning to predict high biomarker
- Validated the algorithm in UK Biobank
Key Findings
Using a standard screening approach
1600
screened to identify 100 eligible patients
Using Perisphere’s algorithm
536
screened to identify 100 eligible patients
Impact
Screening can be >50% more efficient
Label Expansion: Perisphere’s analysis expanded Denosumab approval to include men
Context
Denosumab was approved for osteoporosis in women
Objective: Provide data to justify label expansion to men and support a Phase IV (safety) study using RWD
Approach
- Identified five data sets with biomarker measurements
- Conducted machine learning to predict high biomarker
- Validated the algorithm in UK Biobank
Key Findings
- Medicare has >1 million men with osteoporosis
- Adverse events can be identified in Medicare and provide control rates for a phase IV study
Impact
FDA expanded Denosumab for men
Phase IV study approved using Medicare claims
Phase IV study approved using Medicare claims
Putting trial results in context: Perisphere addresses diversity in trials as it relates to real world data
Context
- Trials enroll select populations
- Clinicians, regulatory authorities, patients want to know if trial results apply to people like them
Questions
- Are FOURIER trial participants similar to real-world patients with CVD?
- Are results similar for Black and white US adults
Approach
- Apply trial eligibility criteria to large RWD (REGARDS study)
- Compare trial participants and RWD
– Population characteristics
– CVD event rates
Key Findings
Patients RWD look similar to trial participants
Age CVD risk factors Medication use
Impact
- FOURIER participants look like patients in the real-world
- Repatha sales have increased year over year
Estimate population size: Perisphere determines the number of patients who may benefit from a trial
Context
- SPRINT trial showed intensive blood pressure goal prevents CVD and death
- Objective: Determine the number of US adults eligible for intensive blood pressure lowering
Approach
- Analyzed data from a large government survey
- Conducted Monte-Carlo Simulation
- Estimated events prevented with population-wide treatment
Key Findings
Impact
- 108,000 deaths could be prevented annually by population-wide adoption of SPRINT trial intervention
Inform policy: Perisphere provides data that translates RWD into policy
Context
- Lipid medications are under-used after MI
- Cholesterol testing guides therapy use
- Objective: Determine lipid testing following myocardial infarction among Medicare beneficiaries by state
Approach
- Analyzed data from all Medicare beneficiaries hospitalized for myocardial infarction in 2019
- Reviewed pharmacy records for medication fills
- Mapped rates by region
Key Findings
Less than 30% of patients receive lipid testing following a myocardial infarction- Substantial variation was present by state
Impact
- US Senators asked Government Accountability Office to investigate low rate of lipid testing