Heartflow, Inc. · NASDAQ:HTFL
Heartflow — Transcatheter Cardiovascular Therapeutics Conference 2019
20 slides from Heartflow's conference presentation, Sep 28, 2019
Slide 1: Cost-effectiveness analysis of non-invasive FFRCT for stable chest pain evaluation
Cost-effectiveness analysis of non-invasive FFRCT for stable chest pain evaluation A comparison to coronary CTA and functional testing based on the PROMISE Trial Udo Hoffmann, MD, MPH, Júlia Karády, MD, Borek Foldyna, MD, Michael T. Lu, MD, MPH, Maros Ferencik, MD, PhD, MCR, Amit Pursnani MD, Alexan
Slide 2: Disclosures
Disclosures Research support on behalf of the institution: KOWA, MedImmune, HeartFlow, Duke University (Abbott), Oregon Health & Science University (AHA, 13FTF16450001), Columbia University (NIH, 5R01-HL109711), NIH/NHLBI 5U01HL123339 Consulting Fees: Abbott, Duke University (NIH), Recor Medical
Slide 3: Low Dx Yield of Fx Testing in Elective ICA
Low Dx Yield of Fx Testing in Elective ICA u NCDR - 376,430 patients with de novo chest pain 2005-2007 Positive test Equivocal Negative test Index test All patients No stress test result test result result All patients 376,430 16% 68% 4% 12% No CAD 233,515 (62%) 65% 59% 73% 71% CAD 142,912 (38%) 35%
Slide 4: Can FFRCT improve patient selection for ICA and outcomes?
FFRCT • Remote calculation of lesion specific FFR based on resting False positive anatomic assessment standard coronary CTA 28 Can FFRCT improve patient selection for ICA and outcomes?
Slide 5: Objective
Objective To compare cost effectiveness of 3 strategies of noninvasive diagnostic testing in patients with stable chest pain. Functional testing Coronary CTA Coronary CTA+CT FFR
Slide 6: Markov Decision Model of Stable Chest Pain
Markov Decision Model of Stable Chest Pain Lifetime Simulation Testing Health States Model Outputs Strategies Model population: PROMISE Each Patient enters the Non- Downstream simulation 100 times obstructive Testing CAD Functional Testing Obstructive Treatment No CAD CAD Health Outcomes Coronary CT
Slide 7: What drives decision making?
What drives decision making ? 1. In whom to calculate FFRCT: intermediate CAD (30-69% stenosis) - 31% of the model population 2. Who goes to ICA? I. Functional testing: reversible myocardial ischemia; II. Coronary CTA: 70% in >1 vessel or 50% CAD in LM; III. FFRCT : <0.80 3. MACE risk: HR based on a
Slide 8: Stable Chest Pain Population referred for noninvasive dx testing
Stable Chest Pain Population referred for noninvasive dx testing* Variables N=1,000,300 Mean age (years), median (IQR) 60.0 (54.4-65.9) Female sex, n (%) 5,270 (52.7) Mean Number of risk factors 3 Atypical angina 7,773 (77.7) Mean Probability of obstructive CAD (Diamond and Forrester) 53.3% Normal T
Slide 9: Test distribution and downstream testing at 60 days: Functional testing
Test distribution and downstream testing at 60 days Functional testing ICA ICA 9.4% 10.0% SPECT SPECT 67.5% No further testing 67.2% No further testing 91.6% 90.0% Patients with Patients with stable chest pain stable chest pain and without ICA ICA and without known CAD 6.5% 5.8% known CAD STECHO STE
Slide 10: Test results and downstream testing at 60 days: Coronary CTA
Test results and downstream testing at 60 days Coronary CTA Severely Severely Patients with ICA Patients with ICA abnormal abnormal stable chest stable chest 5.9% 4.4% pain and pain and without known without known CAD ICA CAD ICA CTA CTA 22.0% 25.0% Moderately Moderately abnormal abnormal 31.6% 31.4
Slide 11: Markov Model accurately simulates downstream testing, intervention, cost, and MACE
Markov Model accurately simulates downstream testing, intervention, cost, and MACE CTA Strategy Functional Strategy Observed Simulated Observed Simulated N=4,996 N=499,600 N=5,007 N=500,700 60 days ICA (%) 12.2 12.3 8.1 8.2 Coronary Revascularization 6.2 6.4 3.2 3.3 Cost in US $ $2,49421 $2,546 $2,2
Slide 12: Simulation of the Effect of FFRCT
Simulation of the Effect of FFRCT 12
Slide 13: Patient Selection for ICA
Patient Selection for ICA 100% 8.1% 12.3% 10.5% 80% 60% 91.9% 87.7% 89.5% 40% 20% 0% Functionaly Strategy CTA Strategy CTA+FFRCT Strategy No ICA ICA Applied to PROMISE, CTA+ FFRCT would have saved 14.6% (n=90/615) of dx ICA
Slide 14: Yield of Dx Testing
Yield of Dx Testing 90 80 70 60 50 40 85 30 72 48 20 38 10 0 NCDR 2004-8 PROMISE FxnI PROMISE CTA PROMISE CEA % CTA+FFRCT Strategy % of patients with obstructive CAD (>50%) in ICA
Slide 15: ICA Yield
ICA Yield 15% 12% 9% 5.7% 46.3% 4.2% 40.5% 6% 4.8% 59.3% 3% 6.6% 53.6% 6.3% 59.5% 3.3% 40.7% 0% Functional Strategy CTA Strategy CTA+FFRCT Strategy Revascularization No revascularization % of patients undergoing coronary Revascularization
Slide 16: Five Year Results
Five Year Results 2 years 5 years Index test Coronary CTA Functional CTA+FFRCT Coronary CTA Functional CTA+FFRCT Revascular 6.6 3.6 6.3 6.8 4.2 6.5 ization, % PCI 4.6 2.9 4.3 4.7 3.4 4.4 CABG 2.0 0.7 2.0 2.1 0.8 2.1 MACE*, % 1.3 1.3 1.3 3.5 3.6 3.5 *CHD death and myocardial infarction Similar health
Slide 17: Lifetime Cost-effectiveness: Quality Adjusted Life Years
Lifetime Cost-effectiveness* 24.8 • FFRCT less costly and more Quality Adjusted Life Years (QALY) 24.6 effective than functional CTA+FFRCT CTA testing strategy 24.4 Functional (dominant/cost effective) 24.2 • FFRCT less costly than 24.0 coronary CTA and marginally 23.8 5,000 5,500 6,000 6,500 7,000
Slide 18: Limitations
Limitations u Scientific evidence limited for some modeling assumptions (gold standard QCA, treatment of CAD) u Model only considers effect of statin treatment as being different between strategies u No modeling of adherence to medical therapies 18
Slide 19: Conclusions
Conclusions Stable Markov Model with accurate simulation of the results of the PROMISE Trial Compared to functional testing and anatomic assessment alone, supplementing coronary CTA with FFRCT results in 1. Highest yield of dx testing for obstructive CAD (86%) 2. Highest yield of ICA (revascularizat
Slide 20: Thank you!
ACHUSETTS GENERAL !
Heartflow — Transcatheter Cardiovascular Therapeutics Conference 2019 on 6ix