Glycoprotein VI (GPVI) and Abacavir

Dr Patrick Mallon University College Dublin PLATELET FUNCTION UPON SWITCHING TO TAF VS CONTINUING ABC: A RANDOMISED SUBSTUDY

Patrick W. Mallon1, Alan Winston2, Frank Post3, Dermot Kenny4, Colm Bergin5, Robert T. Maughan1, Elena Alvarez-Barco1, Willard Tinago1, Eimear Dunne4, Mingjin Yan6, Moupali Das6, Martin Rhee6

1University College Dublin, Ireland; 2Imperial College London, UK; 3King's College Hospital NHS Foundation Trust, London, UK; 4Royal College of Surgeons in Ireland, Dublin, Ireland; 5St. James's Hospital, Dublin, Ireland; 6Gilead Sciences, Inc., Foster City, CA, USA Disclosures

Speaker Bureau / Honoraria: • ViiV Healthcare, Merck Sharpe and Dohme, Gilead, Janssen Cilag (Tibotec), Bristol Myers Squibb Research funding / educational grants: – Science Foundation Ireland – Health Research Board (Ireland) – Wellcome Trust – GlaxoSmithKline / ViiV Healthcare – Gilead Sciences – Bristol Myers Squibb – Janssen Cilag (Tibotec) – Merck Sharpe and Dohme

3 Background

• PLWH (People living with HIV) are at increased risk of myocardial infarction1 • Cardiovascular (CV) disease in PLWH has multifactorial etiology2 – Traditional risk factors – HIV infection – Antiretroviral agents • Many studies (eg, D:A:D) show association of abacavir (ABC) with CV events3, 4 – Others did not, including meta-analysis of randomised studies5 – No CV signal in registrational studies, underlying mechanistic effect likely subtle – The association also appears reversible, pointing to platelet dysfunction as a potential mechanism

D:A:D, data collection on adverse events of anti-HIV drugs. 1. Freiberg MS, et al. JAMA Int Med 2013;173:614-22;2. JA O’Halloran. Future Virology 2013;8:1021-34; 3. Worm SW, et al. J Infect Dis 2010;201:318-30; 4. Alvarez A, et al. AIDS 2017;31:1781-95. 5. Ding X et al. JAIDS 2012;61(4):441-7. 4 Platelet Activation, Thrombosis and M.I.

Changes in systemic environment…. Liver Inflammation (HIV) Acute Coronary Syndrome Drugs

Collagen

sGPVI GPVI

P2Y1 ADP P2Y12 Thromboxane TPα ACTIVATION Infarction Epinephrine

PAR1 Thrombin Platelet PAR 4 Aggregation Thrombosis Platelet Reactivity - Aggregation • Increased platelet aggregation associated with CVD events1 • Platelet reactivity measured by aggregometry2 • ABC associated with more reactive platelets (cross-sectional study)

1 0 0 A B C N o A B C

5 0 PlateletAggregation, % p=0.008

0 - 1 . 0 - 0 . 5 0 . 0 0 . 5 1 . 0 1 . 5 Log [ADP] Concentration, μM

1. Trip MD et al. NEJM 1990; 322(22):1549-54. 2. Satchell CS et al. JID 2011; 204(8):1202-10 6 Platelet Reactivity - Aggregation • Increased platelet aggregation associated with CVD events1 • Platelet reactivity measured by aggregometry2 • ABC associated with more reactive platelets (cross-sectional study)

1 0 0 A B C N o A B C

5 0 PlateletAggregation, % EC50 p=0.008 0 - 1 . 0 - 0 . 5 0 . 0 0 . 5 1 . 0 1 . 5 Log [ADP] Concentration, μM

1. Trip MD et al. NEJM 1990; 322(22):1549-54. 2. Satchell CS et al. JID 2011; 204(8):1202-10 7 Platelet Reactivity - Aggregation • Increased platelet aggregation associated with CVD events1 • Platelet reactivity measured by aggregometry2 • ABC associated with more reactive platelets (cross-sectional study)

1 0 0 A B C N o A B C

Lower EC50 More reactive 5 0 Higher EC50 Less reactive PlateletAggregation, % EC50 p=0.008 0 - 1 . 0 - 0 . 5 0 . 0 0 . 5 1 . 0 1 . 5 Log [ADP] Concentration, μM

1. Trip MD et al. NEJM 1990; 322(22):1549-54. 2. Satchell CS et al. JID 2011; 204(8):1202-10 8 Platelet Function: Glycoprotein VI (GPVI) and Abacavir

sGPVI Metalloproteinase

Collagen GPVI

GPVI Collagen Related Platelet Peptide sGPVI

• Collagen receptor – expressed on platelets1 • Conventional wisdom- ‘increases’ in soluble GPVI associated with cardiovascular events – acute ischaemic stroke2

1. Nieswandt B et al. Blood 2003; 102(2):449-61; 2. Al-Tamimi M, et al. Stroke 2011; 42(2):498-50 9 Platelet Function: Glycoprotein VI (GPVI) and Abacavir

• ‘Lower’ sGPVI levels in PLWH prior to CAD 3 sGPVI (Case-control study) 4 Metalloproteinase • Persistently ‘lower’ sGPVI in those remaining on ABC SWIFT Trial4 Collagen (virologically suppressed, switching from ABC to TDF vs GPVI remaining on ABC)

5 .0 T D F /F T C A B C /3 T C

GPVI L m

/ 4 .5

Collagen g n

Related Platelet , I

Peptide V P 4 .0 G

sGPVI e l p=0.003 b u l

1 o 3 .5 • Collagen receptor – expressed on platelets S • Conventional wisdom- ‘increases’ in soluble 3 .0 GPVI associated with cardiovascular events – 0 1 2 2 4 3 6 4 8 acute ischaemic stroke2 Week 1. Nieswandt B et al. Blood 2003; 102(2):449-61; 2. Al-Tamimi M, et al. Stroke 2011; 42(2):498-50; 3. Trevillyan et al, Platelets, 2017,28(3): 301-4. 4. O’Halloran J et al. AIDS 2018. Feb 12. [Epub ahead of print]. 10 Study 1717 Platelet Substudy Aims

To determine changes in GPVI function and associated platelet reactivity in a group of virologically-suppressed PLWH switching away from ABC to TAF (Study 1717)

11 Study 1717 Platelet Substudy Aims

To determine changes in GPVI function and associated platelet reactivity in a group of virologically-suppressed PLWH switching away from ABC to TAF (Study 1717)

Hypothesis

Switch from ABC to TAF would result in: 1. Decreases in platelet reactivity measured by aggregometry 2. Increases in soluble GPVI (based on findings from the SWIFT study1)

1. O’Halloran J et al. AIDS 2018. Feb 12 [Epub ahead of print] 12 Study 1717 Platelet Substudy Study Design: Switch from ABC/3TC to TAF/FTC

Phase 3, randomised, double-blind, active-controlled study in US and EU (Study 1717) (Primary endpoint at Week 48)

ABC/3TC + Third Agent n=280 TAF/FTC OD N=556 Continue Third Agent • HIV-1 RNA <50 c/mL for ≥6 mo 1:1 • No CD4 criteria ABC/3TC OD • Estimated CrCL ≥50 mL/min • No single tablet regimen allowed n=276 Continue Third Agent

Week 0 12 48 96

3TC, lamivudine; FTC, emtricitabine Winston A et al. Lancet HIV 2018 [Epub ahead of print] 13 Study 1717 Platelet Substudy Study Design: Switch from ABC/3TC to TAF/FTC

Phase 3, randomized, double-blind, active-controlled study in US and EU (Study 1717) (Primary endpoint at Week 48)

ABC/3TC + Third Agent n=280 TAF/FTC QD N=556 Continue Third Agent • HIV-1 RNA <50 c/mL for ≥6 mo 1:1 • No CD4 criteria ABC/3TC QD • Estimated CrCL ≥50 mL/min • No single tablet regimen allowed n=276 Continue Third Agent

Week 0 12 48 96

Platelet Substudy n=61 From four clinical sites in Dublin and London 0 4 12

3TC, lamivudine; FTC, emtricitabine Winston A et al. Lancet HIV 2018 [Epub ahead of print] 14 Study 1717 Platelet Substudy Methods Platelet reactivity measured using aggregometry at baseline, Week 4 and Week 12 • Five platelet agonists* Collagen

• Between-group comparison of population EC50 by F-test GPVI sGPVI

PAR1 TRAP* PAR4

P2Y1 ADP* ACTIVATION P2Y12 Epinephrine

Arachidonic Acid TPα

Platelet

*Adenosine diphosphate (ADP), collagen, epinephrine, and thrombin receptor-activating peptide (TRAP) and arachidonic acid; 15 Study 1717 Platelet Substudy Methods Platelet reactivity measured using aggregometry at baseline, Week 4 and Week 12 • Five platelet agonists* CD42b • Between-group comparison of population EC50 by F-test Collagen Platelet surface markers at baseline and week 12 • GPVI, CD42b [GP1bA] and P-selectin [CD62P]† P-sel • Single-colour flow cytometry • GPVI shedding induced by collagen-related peptide • Between-group comparison by Wilcoxon rank sum test GPVI

Assays performed on fresh whole blood (citrate) Collagen Platelet Related Peptide sGPVI

*Adenosine diphosphate (ADP), collagen, epinephrine, and thrombin receptor-activating peptide (TRAP) and arachidonic acid; †GPVI kind gift from Liz Gardiner and Rob Andrews from Monash University, Melbourne. 16 Study 1717 Platelet Substudy Methods Platelet reactivity measured using aggregometry at baseline, Week 4 and Week 12 • Five platelet agonists* CD42b • Between-group comparison of population EC50 by F-test Collagen Platelet surface markers at baseline and week 12 • GPVI, CD42b [GP1bA] and P-selectin [CD62P]† P-sel • Single-colour flow cytometry • GPVI shedding induced by collagen-related peptide • Between-group comparison by Wilcoxon rank sum test GPVI

Assays performed on fresh whole blood (citrate) Collagen Platelet Related Sample size: 40 per arm Peptide • 80% power to determine between-group difference of 15% sGPVI in platelet aggregation (p <0.05)

Main Study Platelet Substudy TAF/FTC ABC/3TC TAF/FTC ABC/3TC n=280 n=276 n=29 n=32 Age, y (range) 52 (20, 79) 52 (24, 74) 50 (43, 53) 49 (38, 54) Female, % 14 22 28 31 White,% 73 72 52 60 CD4 count, cells/mm3 654 (489, 849) 700 (546, 891) 659 (503, 833) 616 (512, 774) Duration on ABC/3TC, yrs 8 (3, 11) 8 (4, 11) 8 (3, 10) 8 (4, 10) Platelet count, x109/µl 220 (182, 254) 218 (181, 259) 203 (186, 235) 218 (190, 265) Current smoker, % 27 23 17 22 Hyperlipidemia, % 47 51 31 41 Hypertension, % 39 40 28 34 Diabetes, % 12 13 3 3 Cardiovascular disease, % 6 6 3 6

Data are median (IQR) or %, unless specified otherwise. 18 Study 1717 Platelet Substudy Platelet Reactivity in Response to Collagen

TAF/FTC n=29 ABC/3TC n=32 0 .0 4 Collagen p = 0 .2 9 p = 0 .0 0 5 p = 0 .0 2

L GPVI sGPVI m /

g 0 .0 3 m

, 0

5 PAR1

C TRAP* E PAR4 n 0 .0 2 e g

a P2Y1 l l ADP*

o ACTIVATION

C P2Y

12 n

a 0 .0 1

e Epinephrine M Arachidonic Acid TPα

0 .0 0 B a s e lin e W e e k 4 W e e k 1 2 Platelet

• Higher collagen EC50 (i.e., less reactive platelets) in TAF/FTC group at both Weeks 4 and 12

19 Study 1717 Platelet Substudy Platelet Reactivity in Response to Collagen

TAF/FTC n=29 ABC/3TC n=32 0 .0 4 Collagen p = 0 .2 9 p = 0 .0 0 5 p = 0 .0 2

L GPVI sGPVI m /

g 0 .0 3 m

, 0

5 PAR1

C TRAP* E PAR4 n 0 .0 2 e g

a P2Y1 l l ADP*

o ACTIVATION

C P2Y

12 n

a 0 .0 1

e Epinephrine M Arachidonic Acid TPα

0 .0 0 B a s e lin e W e e k 4 W e e k 1 2 Platelet

• Higher collagen EC50 (i.e., less reactive platelets) in TAF/FTC group at both Weeks 4 and 12 • Similar results seen with TRAP and ADP but not with Epinephrine or Arachidonic Acid

20 Study 1717 Platelet Substudy Platelet Reactivity in Response to TRAP

TAF/FTC n=29 ABC/3TC n=32 3 p = 0 .3 3 p = 0 .0 0 4 p = 0 .5 3 Collagen L / l GPVI sGPVI o m 2 , 0 5 PAR1 C

E TRAP*

P PAR4 A R P2Y1 T 1 ADP*

n ACTIVATION

a P2Y12 e

M Epinephrine

Arachidonic Acid TPα 0 B a s e lin e W e e k 4 W e e k 1 2 Platelet

• Higher TRAP-EC50 (i.e., less reactive platelets) in TAF/FTC group at Weeks 4

21 Study 1717 Platelet Substudy Platelet Reactivity in Response to ADP

TAF/FTC n=29 ABC/3TC n=32 2 .5 p = 0 .0 7 p = 0 .0 3 p = 0 .0 9 Collagen

2 .0 L

/ GPVI sGPVI l o m ,

0 1 .5 5 PAR1 C TRAP* E PAR4 P

D 1 .0

A P2Y

n ADP* ACTIVATION a P2Y e 12 M 0 .5 Epinephrine

Arachidonic Acid TPα

0 .0 B a s e lin e W e e k 4 W e e k 1 2 Platelet

• Higher ADP EC50 (i.e. less reactive platelets) in TAF/FTC group at Week 4

22 Study 1717 Platelet Substudy Platelet Reactivity in Response to ADP

• No evaluable between-group differences with epinephrine or arachidonic acid Collagen

GPVI sGPVI

PAR1 TRAP* PAR4

P2Y1 ADP* ACTIVATION P2Y12 Epinephrine

Arachidonic Acid TPα

Platelet

23 Study 1717 Platelet Substudy Platelet Surface GPVI Expression

TAF/FTC n=29 ABC/3TC n=32

Platelet Surface GPVI CD42b 8 0 0 0 p = 0 .7 0 p = 0 .0 3 t Collagen e l e t a l 6 0 0 0 P-sel P / I V P G

e 4 0 0 0 c a f r u S

n GPVI

a 2 0 0 0 i d e M Collagen Platelet 0 Related B a s e lin e W e e k 1 2 Peptide sGPVI

Higher platelet surface GPVI expression in the TAF/FTC group at week 12

24 Study 1717 Platelet Substudy Soluble GPVI Expression

TAF/FTC n=29 ABC/3TC n=32

Platelet Surface GPVI 5 0 CD42b 8 0 0 0 p = 0 .7 0 p = 0 .0 3 t Collagen e 4 0 l e t a l 6 0 0 0 P-sel P / I 3 0 V P G

e 4 0 0 0 c 2 0 a f r u S

n GPVI a 2 0 0 0 1 0 i d e

M Platelet 0 Collagen 0 0 4 1 2 2 4 4 8 Related B a s e lin e W e e k 1 2 Week Peptide TA F/FTC , n= 268 261 259 252 229 ABC/3TC, n= 267 259 259 255 243 sGPVI Higher platelet surface GPVI expression in the TAF/FTC group at week 12 Greater increases in sGPVI expression in the TAF/FTC group to week 48

25 Study 1717 Platelet Substudy Platelet Surface GPVI Expression

TAF/FTC n=29 ABC/3TC n=32 GPVI Shedding Induced by Platelet Surface GPVI Collagen-Related Peptide CD42b 8 0 0 0 p = 0 .7 0 p = 0 .0 3 t B a s e lin e W e e k 1 2 Collagen e l 0 e t a l 6 0 0 0 P-sel P / I V g P n G i

d -2 0 e

4 0 0 0 d c e a f h r S u

I S

n GPVI P a 2 0 0 0 i G d -4 0 e % M Collagen Platelet 0 Related B a s e lin e W e e k 1 2 p = 0 .5 3 p = 0 .4 7 -6 0 Peptide sGPVI Higher platelet surface GPVI expression in the TAF/FTC group at week 12 Not mediated through changes in GPVI shedding

26 Study 1717 Platelet Substudy – Poster Abstract 677LB Soluble GPVI

TAF/FTC ABC/3TC 5 0

4 0 CD42b , %

Collagen

sGPVI 3 0 P-sel

2 0

1 0 Mean Change in in Change Mean GPVI 0 0 4 1 2 2 4 4 8 Week Collagen Platelet TAF/FTC, n= 268 261 259 252 229 Related Peptide ABC/3TC, n= 267 259 259 255 243 sGPVI

1. Mallon PW et al. CROI 2018. Abstract 677LB. 27 Study 1717 Platelet Substudy Platelet surface CD42b and P-selectin expression

TAF/FTC n=29 ABC/3TC n=32

Platelet CD42b Platelet P-selectin CD42b 5 0 0 0 0 8 0 0 p = 0 .2 2 p = 0 .1 0 p = 0 .2 2 p = 0 .8 3 t Collagen t e l e l e t e t a

4 0 0 0 0 l a P-sel l p /

p 6 0 0 / n i b t c 2 e 4 3 0 0 0 0 l D e s C -

P e 4 0 0 c e a c f r 2 0 0 0 0 a f u

r GPVI S u

s n a n 2 0 0 i a d 1 0 0 0 0 i e d Platelet e

M Collagen M Related 0 0 Peptide B a s e lin e W e e k 1 2 B a s e lin e W e e k 1 2 sGPVI No difference in platelet CD42b or P-selectin expression at baseline or week 12

28 Study 1717 Platelet Substudy Limitations

• Did not attain full recruitment targets • Study not designed to measure clinical cardiovascular events • Did not recruit participants with pre-existing renal dysfunction

• Aggregometry assay produces population level EC50 which are not amenable to normal adjustment for covariates

29 Study 1717 Platelet Substudy Conclusions

Switching from ABC/3TC to TAF/FTC was

associated with: sGPVI • Early (week 4 and week 12) decreases in platelet Metalloproteinase reactivity induced by collagen Collagen GPVI

GPVI Collagen Related Platelet Peptide sGPVI

1. Mallon PW et al. CROI 2018. Abstract 677LB 30 Study 1717 Platelet Substudy Conclusions

Switching from ABC/3TC to TAF/FTC was

GPVI Collagen Related Platelet Peptide sGPVI

1. Mallon PW et al. CROI 2018. Abstract 677LB 31 Study 1717 Platelet Substudy Conclusions

Switching from ABC/3TC to TAF/FTC was

associated with: sGPVI • Early (week 4 and week 12) decreases in platelet Metalloproteinase reactivity induced by collagen Collagen • Increases in expression of collagen receptor - GPVI GPVI – on platelets at week 12 • Increases in soluble GPVI that persisted to week GPVI 1 48 Collagen Related Platelet Peptide sGPVI

1. Mallon PW et al. CROI 2018. Abstract 677LB 32 Study 1717 Platelet Substudy Conclusions

Switching from ABC/3TC to TAF/FTC was

1. Mallon PW et al. CROI 2018. Abstract 677LB 33 Conclusions (2) Results suggest an inherent platelet defect in participants on ABC: - increased platelet reactivity - decreased expression of both platelet surface GPVI and soluble GPVI - reversed with switch from ABC

These data implicate platelet dysfunction as a viable, robust and consistent mechanism to explain how ABC contributes to a reversible, increased risk of myocardial infarction

34 Acknowledgements

Study participants

Platelet sub-study team: Gilead Sciences (Martin Rhee, Moupali Das), Imperial College London (Alan Winston, Andrew Lovell, Maryam Khan, Myra McClure, Ken Legg, Michael Wood, Jasmini Alagaratnam, Tommy Pasvol), King’s College Hospital NHS Foundation Trust (Frank Post, Priya Bhagwandin, Leigh McQueen, Oluwayomi Adegbaju, Lucy Campbell, Emily Wandolo), Royal College of Surgeons in Ireland (Dermot Kenny, Eimear), St James’ Hospital, Dublin (Colm Bergin, Maria Gannon, Miriam Moriarty), University College Dublin (Patrick Mallon, Elena Alvarez-Barco, Willard Tinago, Robert Maughan, Alejandro Abner Garcia Leon, Jane O’Halloran).

UCD School of Medicine

35 Acknowledgements

We extend our thanks to: The participants, their families and all participating study investigators and staff: A Antinori, T Benfield, Y Bennani, P Benson, J Berenguer, C Bergin, L Bernard, J Bogner, C Brinson, F Castelli, B Clotet, B Conway, C Creticos, G Crofoot, J de Wet, O Degen, E DeJesus, G Di Perri, C Dietz, M Dupon, H Edelstein, S Esser, V Estrada Perez, G Fätkenheuer, J Gallant, J Gathe, D Goldstein, R Grossberg, P Hay, W Henry, M Johnson, M Karris, S Kegg, A LaMarca, CS Larsen, A Lazzarin, C Lucasti, P Mallon, M Marquez Solero, C McDonald, A Mills, J Morales Ramirez, P Morlat, K Mounzer, M Moutschen, G Moyle, C Mussini, C Orkin, O Osiyemi, P Peyrani, AD Podzamczer Palter, FA Post, T Prazuck, D Prelutsky, F Raffi, M Ramgopal, B Rashbaum, E Ribera, G Richmond, G Rizzardini, J Rockstroh, P Ruane, L Santiago Colon, A Scarsella, G Schembri, P Shalit, D Shamblaw, J Slim, G Smith, H-S Stellbrink, C Stephan, A Stoehr, J Szabo, S Taylor, A Thalme, L Vandekerckhove, T Vanig, G Voskuhl, D Ward, A Winston, and C Zurawski. This study was funded by Gilead Sciences, Inc.