Review of Paclitaxel Technologies in PAD Treatment: Where Do We Stand?

Craig M. Walker, MD, FACC, FACP Chairman, New Cardiovascular Horizons Founder, President, and Medical Director Clinical Professor of Medicine Cardiovascular Institute of the South Tulane University School of Medicine Houma, LA New Orleans, LA Clinical Editor Clinical Professor of Medicine Vascular Disease Management LSU School of Medicine New Orleans, LA Global Vascular Digest Disclosures

Speaker’s Bureau: Stockholder: • Abbott • Cardiva • Bard • Bristol-Myers-Squibb/Sanofi Consultant/Medical/Scientific Boards: • Cardiva • Abbott • Gore • Boston Scientific • Philips Volcano • Cardiva • Jannsen • CR Bard • Lake Regional Medical PVD Training: • Medtronic • Abbott • Philips Volcano • Bard • Boston Scientific • Philips Volcano PTX Background

The chemistry and pharmacological properties of paclitaxel have been well characterized and studied

• 20 carbon atoms, a molecular weight of 853.91 g/mol. • Highly lipophilic, and is freely soluble in common organic solvents • Medical uses: Cancer / Chemotherapy, Restenosis / Anti-restenosis treatment • Mechanism of action: stabilization of microtubules during the final G2/M phase of cell division (apoptosis) IV administration requires solubilization of drug with Cremophor® • Injected drug is immediately bioavailable in bolus or infusion • Usually administered in multiple cycles

IN.PACT Admiral IFU, Taxol IFU, Bristol-Myers Squibb PTX Mechanism of Action

• Paclitaxel (Cytotoxic) – Interferes with cell division

• Cytotoxic drugs halt cell division, inducing apoptosis Paclitaxel Pharmacokinetics

•How does IV use of PTX compare to peripheral devices?

•How does intravenous PTX exposure compare to DCB and DES use?

•Intravenous infusion of paclitaxel for chemotherapy exhibits maximum concentration (Cmax) and area under the curve (AUC) orders of magnitude higher than what is measured systemically in DCB and DES use.

1. Ohstu T, et al. Clinical Cancer Research 1995;1:599-606. 2. Krishnan P, et al. Circ 136:1102-1113. Presented by Granada J at ISET, Hollywood, FL, USA 2019. 3. MDT IN.PACT Admiral DCB IFU. DCB Paclitaxel Exposure is Hundreds-fold Lower than a Single 3h Infusion for Cancer Treatment

• Area under curve (AUC) is the drug concentration in systemic circulation over time • For DCBs, the concentration of bioavailable paclitaxel is 2-3 orders of magnitude lower than in oncological application

1. Clinical Cancer Research 1: 599-606, 1995, Ohstu T et al. 2. Circulation 136: 1102-1113, Krishnan P et al; 3. IN.PACT Admiral IFU, Medtronic PTX Dosing in Cancer vs PAD Therapy

In comparison to PTX doses for oncological treatment, maximum plasma concentration for DCB is under 1/100th (orders of magnitude lower)

1. TAXOL Package Insert BMS Princeton, NJ Rev April 2011 2. IN.PACT Admiral 6x120mm (data on file at Medtronic) 3. Based upon animal studies (data on file at Medtronic)

ǂ administered q 3 wks. with a median of six courses DES and DCB RCT 12M Primary Patency

In general, DES and DCB have shown similar results in SFA RCTs 100% 10 89.0% 88.5% 87.5% 86.4% 9 83.1% 82.3% 80% 79.5% 8 70.9% 73.5% 65.0% 67.0% 7 60% 56.5% 56.8% 55.0% 6 52.4% * 5

40% 32.8% 4 3

Primary Patency (KM) (KM) Patency Primary 20% 2 Mean Lesion Length (cm) (cm) Length Lesion Mean 1 0% 0

Eluvia DCB / Zilver PTX Control

Results from different clinical investigations are not directly comparable. Information provided for educational purposes only.

The Ranger Drug Coated balloon is an investigational device, not available for sale in the United States

Received October 20, 2018 Accepted November 7, 2018 All cause death

• 1 year 2.3% vs 2.3% • 3 years 7.2% vs 3.8% • 5 years 14.7% vs 8.1%

• Drug has short life—why is effect so delayed? Who were the patients studied?

• 28 RCTs – N=4663 patients (4133 IC and 530 CLI) – 4 RCTs with DES – 24 RCTs with DCB • 16 RCTs with DCB alone • 4 RCTs with DCB/BMS • 3 RCTs with ISR • Pooled weighted data Potential Confounders

• Dose/Duration Effect (Devices different) • Adjudication of Cause of Death • Inherent biases and withdrawal of consent • Mechanism? Drug devices different • ( 40% PTA arm had drug device) • (Intention to treat analysis) • Poor follow-up in PTA arm DCB’s are different • Dose of drug • Form of drug • Excipient • Balloon materials DES’s are different • Dose of drug • Form of drug • Polymer or not No Causal Association

“As Sir Bradford Hill noted, knowledge of the mechanism may be limited by current knowledge.”

http://ahajournals.org on April 1, 2019 I’m confused

Caddyshack FDA letter to Physicians – March 15, 2019

• FDA finds 50% higher risk of 5-year mortality with paclitaxel PAD products. The agency suggested alternative treatment options should be used in most cases until more data is available.

https://www.cardiovascularbusiness.com/topics/vascular-endovascular/fda-higher-risk-mortality-paclitaxel-pad FDA’s own preliminary analysis included three trials with at least 5 years of follow-up:

• Impact Admiral DCB • Zilver PTX DES • Lutonix DCB

https://www.cardiovascularbusiness.com/topics/vascular-endovascular/fda-higher-risk-mortality-paclitaxel-pad “While the analysis are ongoing, our preliminary review of this data has identified a potentially concerning signal of increased long-term mortality in study subjects treated with paclitaxel-coated products compared to patients with uncoated devices.”

“In total amount the 975 subjects in these 3 trials, there was approximately 50% increased risk of mortality.” (20.1% versus 13.4% crude risk of death at 5 years.)

https://www.cardiovascularbusiness.com/topics/vascular-endovascular/fda-higher-risk-mortality-paclitaxel-pad FDA acknowledged pooling the studies may lead to a “greater uncertainty in the results” and acknowledged that the specific cause and mechanism of the higher mortality is unknown.

https://www.cardiovascularbusiness.com/topics/vascular-endovascular/fda-higher-risk-mortality-paclitaxel-pad FDA will convene an advisory committee meeting with its Circulatory System Devices Panel to discuss these potential mechanisms, facilitate a discussion on long-term mortality signal, consider modifications to US trials of paclitaxel products and re-examine the risk benefits profile.

https://www.cardiovascularbusiness.com/topics/vascular-endovascular/fda-higher-risk-mortality-paclitaxel-pad In the meantime, FDA recommends providers continue to monitor patients and discuss the risks and benefits of these devices.

https://www.cardiovascularbusiness.com/topics/vascular-endovascular/fda-higher-risk-mortality-paclitaxel-pad “For some individual patients at particularly high risk for restenosis, clinicians may determine that the benefits of using a paclitaxel-coated product may outweigh the risks.”

https://www.cardiovascularbusiness.com/topics/vascular-endovascular/fda-higher-risk-mortality-paclitaxel-pad I’m Even More Confused!

I’m more confused?

Mr. Bean Problems • Pooled weighted data • Assumes all devices are the same? • Poor follow-up (numbers may be wrong) • Crossover (Many patients in PTA arm got paclitaxel) • No causal correlation (drug/excipient/polymer/ poor follow-up) • Why are deaths increasing as time passes with short drug duration????

Consequences • All patients will die (lawyers are already creating ads) So where does that leave us? • In large RCT data sets, there DOES appear to be a mortality signal in terms of all-cause mortality. (FDA) • Dose does not seem to be the key discriminant (FDA) – DCB doses (up to 8mg/balloon) vs Stents (Eluvia – 0.4mg/stent) and arterial concentration over time have not yet been properly considered – FDA viewing this as a class effect for now • Patient level industry-wide meta-analysis still being planned • FDA panel to be convened in June • Clinical practitioners are left to use devices selectively The problem with focused meta-analysis with bias

My cath lab nurse Joe Brunet commonly cites ‘that Boudin (a very high fat pork and rice sausage made in Louisiana) must be a healthier food than lettuce as there has never been a Boudin recall yet there are frequent lettuce recalls.” The problem with focused weighted meta-analysis is that we can draw erroneous conclusions FDA AND KATSONIS INTERPRETATION HAVE HUGE IMPACT

• Hospitals are questioning whether they should stock drug-eluting devices. • New device protocols will be prohibitively expensive as at least 5 year F/U needed. • There will be law suits • I’ve changed my consent forms. • THIS IS A BIG PROBLEM Review of Paclitaxel Technologies in PAD Treatment: Where Do We Stand?

One Year Data Two Year Data: Mortality RR 1.68 [1.15, 2.47] Five Year Data: Mortality RR 1.97 [1.27, 2.93] Sensitivity Analysis

Dose related effects: DCB<3 ug/mm2 – No clear effect DES and DCB>3ug/mm2 – Possible effect

Exposure = (Dose)(휋)(푅푉퐷)(퐿푒푛𝑔푡ℎ)(푡) Drug-Coated Devices and Mortality IN.PACT Clinical Program: Patient-Level Meta-Analysis Mortality Through 5Years Freedom From All-Cause Mortality Through 5 Years

5-years DCB PTA P- (n=183 (n=14 value* 7) 3) All-cause 9.3% 11.2% 0.399 Mortality (140) (12) *P-value was from frailty model with study as random effect

Courtesy: Peter Schneider, MD IN.PACT Clinical Program: Patient-Level Meta-Analysis Distribution of Paclitaxel in DCB Group byTercile No significant difference in mortality betweengroups. No direct impact of paclitaxel dose exposure and survivalstatus.

Highest Dose DCB=Best Survival

1. Taxol (Paclitaxel) Injection Warnings and Safety Information. Bristol Myers-Squibb,2011

Courtesy: Peter Schneider, MD Levant 2 Studies - All Cause Death

p = 0.264 All DCB (N=1029) Subjects Survival wit Censore Subject Time %[95% CI] d s at h Subjects Risk Eve nt 60 months 82.7% 147 863 19 [79.7, 85.3] PTA (N=160) 60 months 87.8% 17 141 2 [81.1, 92.3]

Courtesy: Dierk Scheinert, MD Levant 2 Quartile Analysis 5 Year All Cause Death by Dose Binary Kaplan-Meier Dose AnalysisDCB PTA P-value DoseAnalysisDCB PTA P-value Overall(max 147/10 17/16 0.2111 Overall 82.7% 87.8% 0.2641 dose) 29 0 (max dose) (14.2%) (10.6 %)

Dose DCB PTA P-value Dose DCB PTA P-value

< 1.88mg 37/28 NA <1.88mg 86.5% NA 2 (13.1 %) 1.88-3.14mg 85.1% NA 1.88-3.14mg 39/30 NA 7 (12.7 3.14-5.03mg 84.0% NA %) 3.14-5.03 mg 37/256 NA (14.5%) >5.03mg 79.3% NA 0.2542

>5.03mg 32/18 NA 0.1822 4 1 1Homogeneity test (17.4 Homogeneity test 2Log-Rank test for homogeneity 2Two-sided Cochran%)-Armitage Trend test BPV/LTNX/0816/0086c

Courtesy: Dierk Scheinert, MD Levant 2 All Cause Mortality by Time

Mortality Rate Mortality Rate

P-value (Likelihood General Time Ti Ratio Chi- Peripheral Artery LEVANT 2 LEVANT 2 me Square Test: Point Disease DCB PTA Subjects DCB vs. Population Po Subjects (n=160) PTA) (n=487) int (n=1029)

1 Yr 32/487 (6.6%) 1 Yr 18/1029 (1.7%) 4/160 (2.5%) 0.530

2 Yrs 53/487 (10.9%) 2 Yrs 46/1029 (4.5%) 8/160 (5.0%) 0.768

3 Yrs 78/487 (16.0%) 3 Yrs 82/1029 (8.0%) 10/160 (6.3%) 0.437

4 Yrs 99/487 (20.3%) 4 Yrs 121/1 14/160 (8.8%) 0.250 029 5 Yrs 114/487 (23.4%) (11.8 %) 5 Yrs 145/1 17/1 0.221 029 60 (14.1 (10.6 %) %)

Mueller T et al, ‘Mortality Rates and Mortality Predictors in Patients with Symptomatic Peripheral Artery Disease Stratified According to Age and Diabetes’, J Vasc Surg 2014; 59:1291-9.

Courtesy: Dierk Scheinert, MD Levant 2 Cause of Death at 5 Years

Levant 2 highest cause PTA All DCB Subjects Subjects of death at 5 years: Cause of Death (N=160) (N=102 9) Cardiovascular 5 (3.1%) 50 (4.9%) Gastrointestinal 2 (1.3%) 2 (0.2%) •Cardiovascular Infectious 0 (0.0%) 3 (0.3%) Neurological / NervousSystem 1 (0.6%) 10 (1.0%) •Respiratory Respiratory 3 (1.9%) 30 (2.9%) Systematic complication 0 (0.0%) 1 (0.1%) Genito-urinary system 2(1.3%) 2(0.2%) Various 4 (2.5%) 49 (4.8%) Carcinoma (not specified 2 (1.3%) 19 (1.8%) elsewhere) Death (non-cardiac or 1 (0.6%) 23 (2.2%) neurological) Sepsis 1 (0.6%) 0 (0.0%) Other 0 (0.0%) 7 (0.7%) Total 17 (10.6%) 147 (14.3%)

Courtesy: Dierk Scheinert, MD Mortality for Patients Treated with StellarexDCB (ATK)

PooledRCTs All (Pooled RCTs andnon-RCTs)

Endpoint DCB KM PTAKM DCB KM Estimate Estimate(SE)1 Estimate(SE) N=419 N=170 (SE) N=2351 All-CauseDeath

1-year 1.8%(0.7%) 1.3%(0.9%) 2.0%(0.4%)

2-year 6.5%(1.3%) 5.9%(1.9%) 5.5%(0.7%)

3-year 9.3%(1.5%) 9.9%(2.4%) 7.9%(0.9%)

CardiovascularDeath

1-year 0.5%(0.4%) 0% 0.1%(0.2%)

2-year 1.0%(0.5%) 1.4%(1.0%) 1.3%(0.3%)

3-year 1.9%(0.7%) 2.8%(1.4%) 1.9%(0.5%)

Non-Cardiovascular Death2

1-year 1.3%(0.6%) 1.3%(0.9%) 1.5%(0.4%)

2-year 5.6%(1.1%) 4.6%(1.7%) 4.2%(0.6%)

3-year 7.5%(1.4%) 7.3%(2.1%) 6.1%(0.8%)

1 KM: Kaplan-Meier; SE: Standard Error. 2 P values calculated using the log-rank test. 3 Unknowns were included as non-cardiovasculardeaths. ©2019Koninklijke Philips N.V. All rights reserved. Some or all products manufactured by Spectranetics, a Philips company. Approved for external distribution. D048417-00 012019

Courtesy: Sean Lyden, MD Survival: Pooled RCT’s All-Cause Mortality through ThreeYears

DCB

PTA

P =0.93

Survival Probability

Days After Procedure No difference in all-cause mortality between Stellarex DCB and PTA through 3 years

Courtesy: Sean Lyden, MD Survival: PooledRCT’s CV Mortality through ThreeYears

DCB PTA

P =0.33

Survival Probability

Days After Procedure No difference in CV-related mortality between Stellarex DCB and PTA through 3 years

Courtesy: Sean Lyden, MD Survival: PooledRCT’s Non-CV Mortality through ThreeYears

DCB PTA

P =0.67

Survival Probability

Days After Procedure No difference in non-CV related mortality between Stellarex DCB and PTA through 3 years

Courtesy: Sean Lyden, MD Zilver PTX RCT Final 5-year MortalityAnalysis

PTA / BMS ZILVER PTX n =143 n =336 Died =17 Died =48 KM =17.6% KM =18.7% p = 0.53

No significant difference between Zilver PTX and PTA / BMS

Courtesy: Michael Dake, MD Zilver PTX Covariate Analysis Multivariate Covariate p-value • Cox proportional hazards model Age 0.0002 • Included comorbidities that may be PTX vs. PTA/BMS 0.54 related to mortality as well as other Hypertension 0.46 Hypercholesterolemia 0.63 factors of interest Pulmonary disease 0.58 • No significant difference between Previous MI 0.94 Zilver PTX and PTA / BMS Diabetes 0.11 Gender 0.47 Carotid disease 0.14 Congestive heart failure 0.08 Cardiac arrhythmia 0.21 Claudication/CLI 0.15 Country 0.56 Smoking 0.17 Lesion length 0.12

Courtesy: Michael Dake, MD Zilver PTX Dose Analysis

5-year Mortality Rate Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 11.5% 13.6% 13.4% 20.0% 13.2% p=0.72

~0.3 mg Increasing Total Paclitaxel Dose ~3 mg

No impact of Zilver PTX paclitaxel dose on mortality rate

The amount of paclitaxel on a Zilver PTX stent is approximately 10% to 20% of the amount on a DCB for the same device size and dose density

Courtesy: Michael Dake, MD Causes of Death Through 5 Years – RCT

Cause RCT – PTX RCT – PTA / BMS (n=336) (n=143) Cardiovascular 4.8% 5.6% Cancer 4.8% 1.4% Pulmonary 1.8% 1.4% Stroke 0.6% 0.7% Trauma 0.0% 1.4% GI 0.3% 0.0% Multiple/Unknown 2.1% 1.4% No significant difference in cause of death, p=0.56

Courtesy: Michael Dake, MD Causes of Death Through 5 Years – RCT and BMS

Cause RCT – PTX RCT – PTA / BMS Zilver BMS Study* (n=336) (n=143) (n=110) Cardiovascular 4.8% 5.6% 4.5% Cancer 4.8% 1.4% 6.4% Pulmonary 1.8% 1.4% 1.8% Stroke 0.6% 0.7% 0 Trauma 0.0% 1.4% 0 GI 0.3% 0.0% 0.9% Multiple/Unknown 2.1% 1.4% 0.9% No increased rate of cardiovascular, cancer, or other cause of death for Zilver PTX compared to PTA or BMS

*The Zilver BMS study enrolled 110 patients with femoropopliteal artery disease for 5-year follow-up 10

Courtesy: Michael Dake, MD Additional Bare Metal Stent Study

• 110 Zilver BMS patients • 5-year follow-up 5-year Mortality • Comparable mortality to No Paclitaxel 16.9%* Zilver PTX studies * 2 additional deaths beyond 5 years not included in rate

Courtesy: Michael Dake, MD IMPERIAL Summary of Deaths:RCT (2:1 randomization)

CEC-Adjudicated Cause of Death* 1-year Rate Eluvia Zilver PTX All 2.0%(6/301) 3.9%(6/152) Cardiac 1.0%(3/301) 3.3%(5/152) Vascular 0.0%(0/301) 0.7%(1/152) Non-Cardiovascular 1.0%(3/301) 0.0%(0/152) Days fromIndex DeviceType Site-Reported Cause of Death CECAdjudication* Procedure Eluvia Cardiac arrest Cardiac 89 Eluvia Unknown cause of death Cardiac 195 Eluvia Congestive heart failure diastolic dysfunction acute on chronic Cardiac 275 Eluvia Multi organfailure Non-cardiovascular 140 Eluvia Giant cell B-cell non-Hodgkinlymphoma Non-cardiovascular 192 Eluvia Respiratory insufficiency type 1 Non-cardiovascular 329 Zilver PTX Worsening heart failure Cardiac 31 Zilver PTX Cardiac arrest Cardiac 78 Zilver PTX Unknown cause of death Cardiac 171 Zilver PTX Cardiopulmonary failure Cardiac 327

CEC,Zilver ClinicalPTX Events Committee. Coronary heart disease Cardiac 353 *TheZilver CEC PTXconsidered all deaths cardiacIntra unless-cerebral an unequivocalhaemorrhage non-cardiac cause could be established.Vascular 271

Courtesy: William Gray, MD IMPERIAL Pharmacokinetics (N=13) and Long Lesion (N=50)Sub-studies

• All patients treated with Eluvia

• Plasma paclitaxel unquantifiable (<1 ng/mL) in 11/13 PK patients at 10 min, and in all patients at all subsequent timepoints (30 min, and 1, 2, 3, 4, 6, 12, 24,48 hours post-implant)

• 1-year mortality 0% in both sub-studies

PK, pharmacokinetics Courtesy: William Gray, MD RANGER SFA Summary of Deaths CEC-Adjudicated Cause of Death 3-year Rate (2:1 randomization)RangerDCB PTA

All 13.8%(9/65) 10.7%(3/28)

Cardiac 4.6%(3/65) 7.1%(2/28)

Vascular 0.0%(0/65) 0.0%(0/28)

Non-Cardiovascular 9.2%(6/65) 3.6%(1/28) Days from DeviceType Site-Reported Cause of Death CECAdjudication Index Procedure RangerDCB acute heart failure Cardiac 611 RangerDCB acute myocardialinfarction Cardiac 383 RangerDCB cardiac decompensation Cardiac 265 RangerDCB worsening of general health and medicalhistory Non-Cardiovascular 1003 RangerDCB pancreascarcinoma Non-Cardiovascular 477 RangerDCB multifactorial diseases Non-Cardiovascular 859 RangerDCB multiple organfailure Non-Cardiovascular 256 RangerDCB carcinoma tonsillaris left Non-Cardiovascular 843 RangerDCB breast cancer Non-Cardiovascular 450 Control PTA cardiac decompensation Cardiac 114 Cardiovascular failure with knowncoronary Control PTA artery disease and Cardiac 526 cardiomyopathy multiorgan failure under hepatorenalsyndrome, Courtesy: Control PTA thrombocytopenia and septic shock Non-Cardiovascular 397 William Gray, MD Analysis #1

First Published Online February 12, 2019 The Medicare Experience

• 16,560 Medicare beneficiaries admitted for femoropopliteal artery revascularization from 01/1/2016 to 12/31/2016 • Drug-coated devices (DES/DCB) compared with non– drug-coated devices (BMS/PTA) • All-cause mortality analyzed through 09/30/2017 • Median 389 days, longest 600 days The Medicare Experience

Dilation of Left Popliteal Artery using Drug-Coated Balloon, Percutaneous Approach

Dilation of Right Femoral Artery with Intraluminal Dilation of Left Femoral Device, using Drug- Artery with Drug-eluting Coated Balloon, Intraluminal Device, Percutaneous Approach using Drug-Coated Balloon, Percutaneous Approach

Dilation of Right Femoral Artery, Bifurcation, with Drug-eluting Intraluminal Device, Percutaneous Approach Dilation of Left Popliteal Artery, Bifurcation, with Intraluminal Device, Percutaneous Approach The Medicare Experience

• Kaplan-Meier methods to estimate long-term survival between devices; log-rank tests to evaluate for differences • Cox proportional hazards regression to calculate adjusted hazard ratios • Model included age, sex, comorbidities, atherectomy use, and hospital characteristics • Analyses stratified by: • Treatment with a balloon only (DCB vs PTA) or stent (DES± DCB vs BMS) • Patients with or without critical limb ischemia. Baseline Patient Characteristics

v Baseline Patient Characteristics Hospital Characteristics All Patients, All Devices

34.3% *No difference in 32.5% survival in adjusted analysis • Adjusted HR 0.97; 95% CI, 0.91- 1.04; P = .43 Balloon Angioplasty

34.8% *No difference in 33.3% survival in adjusted analysis • Adjusted HR 0.94; 95% CI, 0.86- 1.03; P = .17 Stenting

33.5% *No difference in 33.5% survival in adjusted analysis • Adjusted HR 0.97; 95% CI, 0.89- 1.06; P = .48 With Critical Limb Ischemia

40.1%

38.1% *No difference in survival in adjusted analysis • Adjusted HR 0.93; 95% CI, 0.85- 1.01; P = .09 Without Critical Limb Ischemia

29.0% *No difference in survival in adjusted 26.5% analysis • Adjusted HR 0.94; 95% CI, 0.85- 1.03; P = .20 Conclusions

• No difference in survival between drug-coated and uncoated devices • Including by device type and among those with or without CLI Limitations • Longest follow-up 600 days • No signal that survival trend changing up to this time point All Patients, All Devices

34.3% *No difference in 32.5% survival in adjusted analysis • Adjusted HR 0.97; 95% CI, 0.91- 1.04; P = .43 Conclusions

March 1, 2019 Drug-Eluting Stent Implantation and Long-Term Survival Following Peripheral Artery Revascularization Eric A. Secemsky, Harun Kundi, Ido Weinberg, Marc Schermerhorn, Joshua A. Beckman, Sahil A. Parikh, Michael R. Jaff, Jihad Mustapha, Kenneth Rosenfield and Robert W. Yeh http://www.onlinejacc.org/content/early/recent Long-Term Survival after Peripheral DES

• 51,456 Medicare beneficiaries admitted for peripheral artery revascularization with DES or BMS from 12/01/2012 through 11/30/15 • FDA approval of Zilver PTX 11/2012 • All-cause mortality was analyzed through 12/31/2016 • Median follow-up 2.0 years, longest 4.1 years • Required ICD-9-CM code for PAD AND ICD-9-PCS code for stenting of non-coronary peripheral vessel Baseline Demographics Characteristics N=51,456 Age, mean (SD) 72.8 ± 10.5 Male, n(%) 27,886 (54.2%) Chronic limb ischemia, n(%) 30,708 (59.7%) Acute limb ischemia, n(%) 3,654 (7.1%) Congestive heart failure, n(%) 10,273 (20.0%) Valvular heart disease, n(%) 4,650 (9.0%) Hypertension, n(%) 43,231 (84.0%) Diabetes mellitus, n(%) 20,220 (39.3%) Cerebrovascular disease, n(%) 5,647 (11.0%) Chronic obstructive pulmonary disease, n(%) 14,644 (28.5%) Renal failure, n(%) 15,887 (30.9%) Cancer, n(%) 1,882 (3.7%) Hypothyroidism, n(%) 6,361 (12.4%) Coagulopathy, n(%) 1,746 (3.4%) Obesity, n(%) 4,969 (9.7%) Weight loss, n(%) 2,824 (5.5%) Fluid and electrolyte disorders, n(%) 10,047 (19.5%) Anemia, n(%) 1,996 (3.9%) Depression, n(%) 5,015 (9.7%) Temporal Trends in Use of DES for PAD

Q4 Long-Term Survival after Peripheral DES

51.7% *No difference in survival in adjusted 50.1% analysis • Adjusted HR 0.98; 95%CI, 0.93-1.03; P = .53 Long-Term Survival after Peripheral DES

*No difference in survival in adjusted CLI analyses • CLI: Adjusted HR 0.97; 95%CI, 0.92-1.03; P = .32 Non- • Non-CLI: CLI Adjusted HR 1.01; 95%CI, 0.91-1.13; P = .80 Overall Conclusions

• Among Medicare beneficiaries, no difference in mid-term or long-term survival between patients treated with drug-coated and uncoated devices • Relationship persisted between device types (balloon or stent) and among patients with and without CLI Zilver PTX Key Points

• ITT data available to Katsanos K, et al. did not identify all patients who were treated with a Zilver PTX stent • Patient-level data were not used in the analysis • 40% of patients in the PTA group were treated with a Zilver PTX stent within the first year due to protocol-specified cross-over

• As-treated patient level analysis demonstrates no difference in mortality rate for Zilver PTX compared to PTA/BMS • Causes of death for Zilver PTX are similar to PTA/BMS PTA Group Composed of Zilver PTX Patients

Zilver PTX Randomized Trial

Primary Randomization PTA n = 237 Zilver PTX n = 242 (41)

PTA / BMS n = 143 (17) Zilver PTX n = 94 (7)

•40% of PTA group = Zilver PTX Zilver PTX n = 336 (48) •70% of patients in study = Zilver PTX Zilver PTX RCT

5-Year Mortality Analysis PTA / BMS Zilver PTX n = 143 n = 242 Died = 17 Died = 41 KM = 17.6% KM = 22.1%

Zilver PTX n = 94 Died = 7 KM = 9.4% Zilver PTX RCT – As-treated

Final 5-Year Mortality Analysis

PTA / BMS Zilver PTX n = 143 n = 336 Died = 17 Died = 48 KM = 17.6% KM = 18.7%

No significant difference between Zilver PTX and PTA / BMS Covariate Analysis – RCT

• Cox proportional hazards model • Included comorbidities that may be related to mortality as well as other factors of interest • No significant difference between Zilver PTX and PTA / BMS (p=0.51) Dose Analysis

Meta-analysis from Katsanos incorrectly identified Zilver PTX as a high dose device, led to incorrect dose calculation.

Total amount of paclitaxel on a Zilver PTX stent is approximately 10% to 20% of the amount on a DCB Zilver PTX has similar total amount of paclitaxel compared to Eluvia with no polymer and a shorter paclitaxel exposure Dose Analysis Dose Exposure Analysis -RCT

• 0.3 mg Increasing Total Paclitaxel Dose – 3mg • 30 mm Increasing Lesion Length – 300 mm

No impact of Zilver PTX paclitaxel dose on mortality rate Causes of Death Through 5 Years

RCT and BMS

No increased rate of cardiovascular, cancer, or other causes of death for Zilver PTX compared to PTA or BMS

* The Zilver BMS study enrolled 110 patients with femoropopliteal artery disease for 5-year follow-up, ClinicalTrials.gov Identifier: NCT00827619 No Increases Long-term Mortality with DES • Medicare CMS population • Similar mortality for BMS and • 51,456 patients DES through 4.1 years • Overall adjusted p=0.53 – 47,351 BMS • Without CLI adjusted p=0.95 – 4,105 DES (Zilver PTX) • With CLI adjusted p=0.32

Secemsky E, et al. J Am Coll Cardiol. E-pub ahead of print 01March2019. doi https://doi.org/10.1016/j.jacc.2019.02.020 Conclusions

• Conclusion of Katsanos K, et al. was not based on patient-level data and miscalculated the dose exposure • As-treated patient-level analysis of RTC data shows no increased long-term mortality risk with Zilver PTX compared to PTA and BMS – Covariate analysis supports no significant difference – No impact of Zilver PTX paclitaxel dose on mortality rate – No significant differences in causes of death • Mortality rates for the Zilver PTX stent are consistent with rates reported in literature for PAD patients • Japan data confirm RCT findings showing no increased long-term mortality risk with Zilver PTX compared to BMS • Cook will continue to work with global regulatory authorities and independent physician led groups to evaluate safety using patient-level data SFA Stenting Clinical Trials at 12M • What vessel beds are we most comfortable stenting in? • What is an acceptable SFA TLR rate?

30%

21% 20%

14% 12%

10% 10% 9% 9% 7% 4.5% 4%

Target Lesions Revascularization Revascularization Lesions Target 4% 4%

BMS ZilverPTX Eluvia Covered Stent SFA Coronary and Iliac benchmarks

Results from different clinical investigations are not directly comparable. Information provided for educational purposes only. Review of Paclitaxel Technologies in PAD Treatment: Where Do We Stand?