10/6/2016

1

Hepatitis C virus reinfection after successful treatment among PWID: Clinical and Public Health Implications

Håvard Midgard, MD PhD candidateAkershus University Hospital and Oslo University Hospital, Norway

Disclosures

• No conflicts of interest

10/6/2016

2

Starting point

• New DAA treatment provides unique opportunities for prevention of liver disease burden, epidemic control and HCV elimination

• Ongoing injecting risk behaviours can lead to reinfection after successful treatment

• High levels of reinfection might challenge – Individual- and population-level treatment benefits– Cost-benefit of expensive DAAs– Existing HCV prevention strategies

Overview

• Reinfection after interferon-based treatment

• Reinfection after DAA treatment

• Risk factors for reinfection

• Individual- and population-level implications

• Strategies to address reinfection

10/6/2016

3

Reinfection estimates: IDU ever (n=795)

0 5 10 15 20

Weir 2016 (n=277)

Midgard 2016 (n=94)

Pineda 2015 (n=84)

Marco 2013 (n=119)

Hildsen 2013 (n=23)

Grebely 2012 (n=67)

Grady 2012 (n=42)

Grebely 2010 (n=35)

Currie 2008 (n=9)

Backmund 2004 (n=18)

Dalgard 2002 (n=27)

Reinfection incidence per 100 PY (95%CI)

2.1

Modified from Midgard et al. J Hepatology 2016 (In Press)

Reinfection estimates: IDU post-treatment (n=153)

0 10 20 30

Weir 2016 (n=30)

Midgard 2016 (n=37)

Pineda 2015 (n=13)

Grebely 2012 (n=26)

Grady 2012 (n=11)

Grebely 2010 (n=16)

Currie 2008 (n=2)

Backmund 2004 (n=9)

Dalgard 2002 (n=9)

Reinfection incidence per 100 PY (95%CI)

5.6

Modified from Midgard et al. J Hepatology 2016 (In Press)

10/6/2016

4

Differences in reinfection estimates reflect

1. Heterogeneity in study populations– Risk behaviours (former vs. recent PWID, acute vs. chronic HCV)– Harm reduction coverage – Background viremic prevalence

2. Variations in study designs– Prospective vs. retrospective designs – Small sample sizes and short longitudinal follow-up– Insufficient risk factor assessment

3. Virological methods– Testing intervals: ”The more often you look”– Sequencing methods: “The closer you look”

EOT SVRTW0

Time

HC

V R

NA

LOD

A. Late relapse of majority variant

EOT SVRTW0

Time

HC

V R

NA

LOD

B. Persistance of minority variant

Virus 1

Virus 2

EOT SVRTW0

Time

HC

V R

NA

LOD

C. Reinfection

Virus 1

Virus 3

EOT SVRTW0

Time

HC

V R

NA

LOD

A. Late relapse of majority variant

EOT SVRTW0

Time

HC

V R

NA

LOD

B. Persistance of minority variant

Virus 1

Virus 2

EOT SVRTW0

Time

HC

V R

NA

LOD

C. Reinfection

Virus 1

Virus 3

EOT SVRTW0

Time

HC

V R

NA

LOD

A. Late relapse of majority variant

EOT SVRTW0

Time

HC

V R

NA

LOD

B. Persistance of minority variant

Virus 1

Virus 2

EOT SVRTW0

Time

HC

V R

NA

LOD

C. Reinfection

Virus 1

Virus 3

Scenarios for viral recurrence post-SVR

10/6/2016

5

Pooled reinfection incidence from 11 studies

IDU ever IDU post treatment0

5

10

15

Po

ole

d r

ein

fec

tio

n in

cid

en

ce

(pe

r 1

00

PY

)

5.6/100 PY

9 studies153 patients29 cases522 PY

2.1/100 PY

11 studies795 patients43 cases2082 PY

Modified from Midgard et al. J Hepatology 2016 (In Press)

Long-term reinfection risk: Little is known

• Existing reinfection estimates are– mainly based on small studies with short follow-up time– including cases with spontaneous clearance– probably lower than reported rates of primary infection

• Even low rates could be a concern over time – Particularly if constant rates, no re-treatment, no scale-up– Rates may be declining due to a “saturation effect”

• Projected 5-year risk (“worst case scenario”)– IDU ever: 10.5%– IDU post-treatment: 28%

10/6/2016

6

Long-term reinfection risk: 7-year follow-up

Pe

rsis

ten

t H

CV

re

infe

ctio

n

inc

ide

nc

e r

ate

(p

er

10

0 P

Y)

IDU ever(n=94)

IDU after treatment(n=37)

0

2

4

6

8

10

Time at risk after SVR (PY) 593 206

Persistent reinfections 10 10

Incidence per 100 PY 1.7 4.9

95% CI 0.8–3.1 2.3–8.9

Midgard et al. J Hepatology 2016

11%

27%

0%

20%

40%

60%

80%

100%

Cu

mu

lati

ve r

isk

Reinfection risk after DAA treatment

• Are current estimates generalizable for the DAA era?

• Increased treatment uptake among people with ongoing risk behaviours

• Less fear of treatment adverse effects• Less interaction with health care providers

• Less potential for behavioural change?• Increasing reinfection rates?

10/6/2016

7

Reinfections in SOF Phase 3 trials (n=3004)

• 7 reinfections after 3 months (SVR12 - SVR24)• 750 person-years of follow-up• Reinfection incidence 0.9/100 PY

Sarrazin et al. EASL 2015

C-EDGE CO-STAR: Reinfection incidence

Immediate and deferred treatment groups (EOT - FW24)

• 6 reinfections out of 296 total patients• 130.6 person-years of follow-up• 4.6 reinfections per 100 person years

• 5 of 6 cases tested positive for opioids other than OST• 3 of 6 cases cleared spontaneously

Dore et al. Ann Int Med 2016, Dalgard et al. INHSU 2016

• Grazoprevir/elbasvir for patients on stable OST (n=301)• High SVR rates and high adherence• High proportion with positive urine drug screen

10/6/2016

8

Risk factors for reinfection

• Identifying those at highest risk for reinfection could aid post-treatment HCV care (“secondary prevention”)

• Predictors for reinfection have not been clearly identified– Low statistical power– Lack of behavioural data

• Factors associated with reinfection/superinfection1

– Poorer social functioning at enrolment (AOR 5.85)– Methamphetamine injecting during follow-up (AOR 7.29)

• OST protective against reinfection2

1 Grebely et al. Hepatology 20122 Bruneau et al. INHSU 2016

Implications at the individual level

• Reinfections after spontaneous clearance have a benign course1

– Lower viral loads than in primary infection– High rates of spontaneous clearance (30-100%)– Evidence of a partial protective immunity against persistent

reinfection with the same viral strain

• Spontaneous clearance of reinfections after treatment can occur2

• Early reinfections may be easy to treat (acute, no virological failure)

• Reinfection in a cirrhotic patient is more concerning than in a non-cirrhotic patient

1 Grebely et al. Lancet Infect Dis 20122 Dore et al. Ann Int Med 2016

10/6/2016

9

The “prevention benefit” hypothesis

• Good theoretical evidence from dynamic models1,2

1. Scaled-up DAA treatment + OST can reduce viremic prevalence2. Treating active PWID could be more cost-effective than treating those

with no ongoing transmission risk3. More future infections and HCV-related morbidity/mortality will be

averted than lost through reinfections

• No empirical evidence (yet) showing that HCV treatment for PWID reduces HCV transmission

• Little empirical evidence showing that achieving SVR could result in behavioural change– Models assume reinfection risk = primary infection risk– Alternation between high/low risk states

1 Martin et al. Hepatology 20132 Hickman et al. Curr Opin Infect Dis 2015

A slow treatment scale-up could create an increasing pool of susceptible individuals

0

500

1,000

1,500

2,000

2,500

3,000

3,500

0

50

100

150

200

250

An

nu

al N

um

ber T

reate

dSeco

nd

ary

In

fecti

on

s

Treat 1% Treat 2% Treat 4% Treat 8% Treat 10%

Treat 1% Treat 2% Treat 4% Treat 8% Treat 10%

Razavi et al. INHSU 2015

10/6/2016

10

Reduction of reinfection probability could increase impact of scale-up

Razavi-Sherarer et al. INHSU 2016

Model inputs, aggressive treatment strategy in Norway: HCV RNA prevalence 48%, harm reduction 87%, PWID mortality 2%

Addressing reinfection: Potential strategies

1. Acknowledgement without stigma and discrimination

2. Education and counselling including peer support

3. Harm reduction optimization

4. Post-treatment surveillance and rapid re-treatment

1. Scaled-up DAA treatment among PWID

2. Targeted treatment of high-risk transmitters and injecting networks (“bring your friends” strategy)1

1 Hellard et al. Int J Drug Policy 2015

10/6/2016

11

Future research priorities

• Monitor incidence of reinfection following DAA treatment among individuals with ongoing risk behaviours

• Identify risk factors for reinfection

• Explore patient attitudes towards reinfection and risk avoidance following treatment

• Evaluate novel prevention and re-treatment strategies (post-treatment HCV care)

Conclusions

• Pooled incidence from 11 studies of reinfection following interferon-based treatment among PWID– 2.1/100 PY among those with IDU ever– 5.6/100 PY among those with post-treatment IDU

• Strategies to address reinfection– Acknowledgement, education, counselling, peer support– Harm reduction optimization– Post-treatment surveillance and re-treatment– Scaled-up DAA treatment– Targeted treatment of high-risk transmitters and injecting networks

• Novel prevention and re-treatment strategies should be evaluated

10/6/2016

12

Backup slides

Meta-analysis: Projected 5-year risk

Modified from Simmons et al. Clin Infect Dis 2016

Low-risk

43 studies

n=7969

High-risk

14 studies

n=771

HIV co-infected

4 studies

n=309

0

5

10

15

20

5-y

ea

r re

infe

ctio

n r

isk

po

st-

SV

R (%

)

10%1.9/100 PY

15%3.2/100 PY

10/6/2016

13

Narrow intervals: All episodes are captured

Grebely et al. Hepatology 2012

Wide intervals: Persistent cases are captured

GT 3a GT 1a GT 2a GT 2b Not genotyped HCV RNA neg

ID 435

ID 430

ID 406

ID 379

ID 370

ID 330

ID 269

ID 144

ID 116

ID 109

ID 026

ID 345

SVR24 1 2 3 4 5 6 7 8 BL

Time after SVR (years)

Midgard et al. J Hepatol 2016

10/6/2016

14

C-EDGE CO-STAR: Urine drug screening

Dore et al. Ann Int Med 2016

Immediate Treatment Arm; EBR/GZR Treatment Phase

Deferred Treatment Arm; Placebo Phase

* 8 drug classes: amphetamines, barbiturates, benzodiazepines, cannabinoids, cocaine, opiates, phencyclidine, propoxyphene

0

10

20

30

40

50

60

70

Day 1 TW1 TW2 TW4 TW6 TW8 TW10 TW12

% o

f P

atie

nts

wit

h P

osi

tive

Uri

ne

Dru

g Sc

reen

Time Point

0

10

20

30

40

50

60

70

% o

f P

atie

nts

wit

h P

osi

tive

Uri

ne

Dru

g Sc

reen

Time Point

Any drug use of 8classes*

Any drug use of 7classes (excl.cannabinoids)Cannabinoids

Benzodiazepines

Opiates

Cocaine

Amphetamines

ENR(n=93)

BL(n=93)

W4(n=86)

EOT(n=81)

SVR12(n=73)

SVR24(n=67)

0

20

40

60

80

100

Time point

Pro

po

rtio

n (%

)

A

Non-injecting drug use

Hazardous alcohol use

OST

Injecting drug use

ENR(n=55)

BL(n=53)

W4(n=44)

EOT(n=35)

SVR12(n=38)

SVR24(n=34)

0

10

20

30

40

50

Time point

Pro

po

rtio

n (%

)

B Daily injecting

Paraphernalia sharing

Use of non-sterile needles

ENR(n=93)

BL(n=93)

W4(n=86)

EOT(n=81)

SVR12(n=73)

SVR24(n=67)

0

20

40

60

80

100

Time point

Pro

po

rtio

n (%

)

A

Non-injecting drug use

Hazardous alcohol use

OST

Injecting drug use

ENR(n=55)

BL(n=53)

W4(n=44)

EOT(n=35)

SVR12(n=38)

SVR24(n=34)

0

10

20

30

40

50

Time point

Pro

po

rtio

n (%

)

B Daily injecting

Paraphernalia sharing

Use of non-sterile needles

ENR(n=93)

BL(n=93)

W4(n=86)

EOT(n=81)

SVR12(n=73)

SVR24(n=67)

0

20

40

60

80

100

Time point

Pro

po

rtio

n (%

)

A

Non-injecting drug use

Hazardous alcohol use

OST

Injecting drug use

ENR(n=55)

BL(n=53)

W4(n=44)

EOT(n=35)

SVR12(n=38)

SVR24(n=34)

0

10

20

30

40

50

Time point

Pro

po

rtio

n (%

)

B Daily injecting

Paraphernalia sharing

Use of non-sterile needles

ENR(n=93)

BL(n=93)

W4(n=86)

EOT(n=81)

SVR12(n=73)

SVR24(n=67)

0

20

40

60

80

100

Time point

Pro

po

rtio

n (%

)

A

Non-injecting drug use

Hazardous alcohol use

OST

Injecting drug use

ENR(n=55)

BL(n=53)

W4(n=44)

EOT(n=35)

SVR12(n=38)

SVR24(n=34)

0

10

20

30

40

50

Time point

Pro

po

rtio

n (%

)

B Daily injecting

Paraphernalia sharing

Use of non-sterile needles

ACTIVATE: Risk behaviours during and following IFN-based treatment

Midgard et al. INHSU 2016

10/6/2016

15

Simulation of HCV incidence by number of network partners and injecting frequency

Rolls et al. Journal of Theoretical Biology 2012

Impact of network-based strategies

Random No treatment

Hellard et al. Int J Drug Policy 2015