Excellent hepatitis C virus cure rates despite increasing complexity of people who use drugs: Integrated‐Test‐stage Treat study final outcomes

Achieving hepatitic C virus (HCV) elimination requires linking people who use drugs (PWUD) into care. We report final direct‐acting antivirals (DAAs)‐based outcomes from the Integrated‐Test‐stage ‐Treat (ITTREAT) study. Project ITTREAT (2013–2021), based at an addiction centre, was a ‘one‐stop’ service with innovative linkage to care strategies. Primary outcome was sustained virological response (SVR12) (intention to treat ITT) including whether individuals were recruited in first (period 1) versus last four (period 2 included the COVID‐19 pandemic) years of the study. Number recruited were n = 765, mean age 40.9 ± 10.1 years, 78% males, history of current/past injecting drug use (IDU) and alcohol use being 77% and 90%, respectively. Prevalence of a positive HCV PCR was 84% with 19% having cirrhosis. Comparing those recruited in period 2 versus period 1, there was increasing prevalence of IDU, 90% versus 72% (p < .001); homelessness, 67% versus 50% (p < .001); psychiatric diagnosis, 84% versus 50% (p < .001); overdose history 71% versus 31% (p < .001), receiving opioid agonist treatment (OAT) 75% versus 52% (p < .001) and comorbidity 44% versus 25% (p < .001). Of those treated with DAAs (n = 272), ITT SVR rates were 86% (95% CI: 81%–90%), being similar in period 2 versus period 1. Predictors of non‐SVR were receiving OAT (OR 0.33, 95% CI: 0.12–0.87, p = .025) and ≥80% adherence (OR 0.01, 95% CI: 0.003–0.041, p < .001). Reinfection rates period 2 versus period 1 (per 100 person‐years) were 1.84 versus 1.70, respectively. In the treated cohort, mortality was 15%, being mostly drug‐related. Despite increasing complexity of PWUD, high SVR12 rates are achievable with use of OAT and good adherence.


| INTRODUC TI ON
Globally in 2019, an estimated 58 million people had chronic hepatitis C virus (HCV) infection, resulting in around 290,000 deaths. 1 HCV results in considerable morbidity in people who use drugs (PWUD), the main at risk group for HCV infection.An Australian study reported viral hepatitis in 76% of underlying liver deaths in PWUD. 2 PWUD do not access traditional models of care, and this might explain why globally only about 21% of those infected with HCV have been diagnosed of whom only 62% have received DAA treatment. 1In the UK, in 2021, there were around 90,000 individuals living with chronic HCV infection, almost all being PWUD.In England, only about 50% of PWUD were aware of their chronic HCV infection in 2021. 3Additionally, between 2019 and 2020 only 25% of individuals with severe HCV-related liver disease received DAAs prior to hospital admission. 4Thus, there still remains an unmet need for timely HCV diagnosis and treatment in PWUD in the UK.Those currently undiagnosed are likely to represent the hardest to reach PWUD, necessitating innovative strategies for linkage to care.The importance of developing decentralised models of care for PWUD cannot be overemphasised, especially since excellent HCV cure rates can be achieved (~90%) 5 with an ~80% reduction in all-cause, liver and drug-related mortality. 6ough the advent of direct-acting antivirals (DAA) 7 has revolutionised HCV management and led to the WHO mandate to eliminate HCV by 2030, 8 this will not be possible without moving away from the traditional hospital-based delivery of HCV care.In fact, at current levels of diagnosis and treatment, only 11 high-income countries are on track to eliminate HCV by 2030, 60% being off track by at least 20 years. 9The COVID pandemic further threatens our ability to meet WHO elimination targets 10 and could further negatively impact engagement.Between 2019 and 2021, there was a 40% reduction in the number of individuals accessing HCV treatment in the UK. 3 To simplify the care pathway, we developed an integrated, decentralised model of care for PWUD with HCV infection based at a large addiction centre in southeast England (ITTREAT service).We have earlier published patient reported, health economic, qualitative and interim clinical outcomes (until Mar 2018) 11,12 (Figure S1).
Here we present final clinical outcomes in those receiving DAAonly regimens during the entire study duration (Dec 2013-Dec 2021).

| Patients and methods
The ITTREAT study was an 8-year prospective cohort study (Dec 2013-Dec 2021) and is one of the largest addiction centre-based HCV treatment programmes in England.Details of how the ITTREAT service was conceived and set up have previously been described 11,13 but are summarised again below.Figure 1A shows the participant pathway.

| Details of the HCV care model and services provided
The ITTREAT was an integrated and decentralised service that was run by a nurse who was based full time at the addiction centre and dually trained in viral hepatitis and mental health.The following services were provided: BBV testing, HCV treatment, liver fibrosis assessment, peer mentor, social and psychiatric support, a needle syringe programme (until 2017 due to funding issues) and opioid agonist treatment (OAT).
The BBV testing was done by either the hepatitis nurse or the key workers who have been trained by the nurse; this was by either venous blood or dry blood spot test (DBST) (Alere toxicology/Abbott, lower limit of detection (LLD) for HCV RNA ~1000 IU/mL; Public Health England Manchester DBS service LLD 660 IU/mL).DBST allowed testing of HCV antibody (if positive reflex HCV qualitative RNA), hepatitis B surface antigen, hepatitis B core antibody and HIV antibody.
Those with a positive HCV qualitative RNA were recalled and offered HCV quantitative RNA/genotype; FibroScan® (402 Echosens); clinical bloods; and liver ultrasound and gastroscopy if cirrhotic (both performed at the local hospital).From 2020 onwards, there was restricted availability of point-of-care HCV quantitative RNA/genotype (St George's Capillary South West London Pathology, LLD ~200 IU/mL).
A dedicated hepatologist attended the addiction centre about once a month to provide input into the more complex cases.These included individuals with advanced cirrhosis and or and where there were complex psychosocial issues.

| Provision of HCV treatment
All those with a positive HCV RNA were deemed treatment eligible (irrespective of ongoing drug and alcohol use), unless not stable enough to engage/other reasons (deferred candidates). 11After initial assessments, the hepatitis nurse presented individuals at the weekly regional multidisciplinary meeting (MDM) for HCV treatment decisions.Once DAA regimen was approved, HCV treatment was delivered at the addiction centre by the hepatitis nurse under   total numbers that could be treated/year though these were relaxed in the second half of the study.There was also the need to genotype at least 75% of the individuals, but this has just been relaxed recently as has the need for weekly discussion of patients at regional MDM, especially for the less complex individuals.Finally, changes in the addiction centre providers in 2018 and 2020 and the COVID pandemic also affected study recruitment and HCV treatment.

| Data collection
The hepatitis nurse prospectively collected all demographic and clinical data.Though all individuals were offered retesting for reinfection, this was not always feasible due to the nature of the study cohort.

| Linkage to care strategies
Due to the increasing complex nature of the cohort as the study progressed we adopted novel linkage to engagement strategies including 1.An integrated and decentralised service run by an experienced nurse trained both in viral hepatitis and mental health.She worked closely with the addiction staff, peer mentors and social workers and trained them in performing BBV testing 'to make every contact count'.
2. Increasing use of peer mentors who encouraged BBV testing and treatment, accompanied individuals for appointments and helped reduce stigma and negative myths around HCV treatment.
4. Use of ultrasound for venepuncture if poor venous access.
5. Minimal monitoring approach dependent on clinical need.
7. Depending on the individual needs (as some were reluctant to attend the addiction centre unless absolutely necessary and deemed stable for treatment), we did dispense the entire DAA course at one time, drugs being delivered to the individual's home/homeless shelter.In the homeless, shelter we provided lockers for individuals to store the DAAs, and if needed, hostel managers and peer mentors supervised DAA administration.In those instances where there were psychosocial concerns, individuals were asked to attend the addiction centre two-four weekly to receive the DAAs.

| Study outcomes
Our primary outcome was intention to treat (ITT) SVR12.As the study evolved, we were left with an increasingly complex cohort to treat.We therefore assessed SVR12 in the whole group and then stratified SVR12 by whether individuals were recruited the first four (period 1) and last four (period 2) years of the study.Period 2 included the COVID-19 pandemic.Since only a small number of individuals (n = 57) were recruited after the national COVID lockdown (March 2020), they were not analysed separately.Secondary outcomes included service uptake (BBV screening, fibroscan, HCV treatment), HCV treatment adherence and HCV reinfection.

| Study definitions
These were as previously reported 11 but are indicated below again.
PWUD: current or past drug use (injecting or non-injecting) and or those currently receiving OAT.
Current injecting and non-injecting drug use or alcohol use: use at time of HCV treatment initiation and or during HCV treatment.
Overdose: taking large quantities of a single or multiple drugs either intentionally or unintentionally.
Suitability for HCV treatment: willingness and motivation to engage and be adherent with HCV treatment.
SVR12: absence of detectable virus (at any level) 12 weeks after end of treatment (EOT).
Reinfection: any level of detectable virus after achievement of SVR12.
Homelessness: unstable housing including rough sleepers, temporary accommodation and 'sofa-surfing' at initial assessment.
Modified ITT (mITT) analysis: Excluded those who did not attend for SVR bloods.
True virological failure: positive HCV RNA at time of SVR12 bloods.ribavirin.There was no particular reason for older individuals receiving paritaprevir/ombitasvir and dasabuvir though it is possible that they were considered more stable and therefore likely to be complaint with a greater pill burden.
Figure 1E shows SVR12 rates dependent on patient, viral and DAA regimen-related factors.SVR rates were significantly higher if

| Details of non-SVR12
Of the n = 39 not achieving SVR12, there were only n = 10 (26%) true virological failures (Figure 1B Table S1 summarises the outcomes in those who did not achieve SVR12 with their first course of DAA.Of the 39 not achieving SVR12, n = 11 were retreated with DAAs, treatment being successful in n = 10.

| Deaths
Data on mortality were available for 745/765 individuals, 86/745 (12%) dying during the study period.We were able to ascertain the cause of death in about 70% of the individuals (n = 56).This included drug overdose in n = 39 (45%), liver-related in seven (8%), both drug and liver-related in one (1%) and other causes in n = 12 (14%).Despite our best efforts, we were unable to ascertain the cause of death in the remaining 30% (n = 27).One of the reasons was that PWUD and the homeless tend to be a migratory population, and if death occurred outside our region, the exact cause could not be established.

| DISCUSS ION
The SVR12 rates were excellent (86%) and no different in the two study periods.Those receiving ribavirin and OAT and having ≥80% selfreported DAA adherence were significantly more likely to achieve SVR12.Our reinfection rates remained low, consistent with a recent international study (3.1/100 person-years) 17 and were comparable across the two study periods.Almost all who were reinfected were likely injecting within similar networks.
Our SVR12 rates were excellent despite the COVID-19 pandemic, 10 the increased levels of at-home drinking 18 and drug use 19 and the increasingly complexity of PWUD as our study progressed.In the second half of our study, almost all of our recruited cohort had a history of current or past IDU, more than two-thirds were homeless, more than 70% had history of an overdose and a psychiatric illness, with over 40% having additional comorbidity.We attribute the high SVR12 rates to our decentralised and integrated model of care with all services being provided at one site to include BBV testing, HCV treatment, assessment of liver fibrosis, peer mentor and psychiatric support and OAT provision.Additionally, our hepatitis nurse was based full time at the addiction centre, hence easily accessible, and dually trained in both viral hepatitis and mental health.She provided a personalised and non-judgmental service.Prior qualitative data from our study indicates positive impact of an integrated and personalised community-based service delivered by a dedicated hepatitis nurse. 12Additional factors for our success included involvement of peer mentors, contingency management, 20,21 a minimal monitoring approach 22 and a test-and-treat plan. 23Our easy-to-replicate model has been nationally adopted and also acknowledged by the European Centre for Disease Control and Prevention as one of the 15 internationally recognised models of good practice. 24spite the cohort becoming increasingly complex over the last 8 years, our SVR12 rates (86%) and treatment adherence (91%) are consistent with two recent meta-analyses. 5,25As previously reported in PWUD, 25 there were very few true virological failures (26% of all non-SVRs) in our study.In fact, our mITT SVR12 rates (96%) are similar to hospital-based studies. 26Loss to follow-up (4%) and non-adherence (7%) were the main reasons for non-SVR12, being higher than that reported in non-PWUD real-world studies (<1%). 26e of ribavirin was associated with significantly higher SVR12 rates in our study.This is not surprising as ribavirin was used predominantly with the older DAA regimens and in those with cirrhosis.
Directly observed treatment has been shown to enhance SVR12 rates in vulnerable adults. 27Our 7% non-adherence supports such a strategy, especially since non-adherence to DAA was more likely to result in non-SVR12 in our study.9][30] However, since adherence was self-reported in our study, we accept that this may have underestimated the severity of non-adherence.Use of OAT was also a predictor of SVR12 in our study.A recent meta-analysis showed current/recent OAT was associated with an increased odds of recent HCV antibody/RNA testing, DAA treatment uptake though there was insufficient evidence of an association with treatment completion or SVR12 following DAA therapy. 31Our encouraging results nonetheless support upscaling OAT to improve SVR12 rates in PWUD.
Currently, we are treating about two-thirds of HCV-positive individuals in a community setting (Figure S2).Nationally however the picture is very different.As of December 2020 about 72% with chronic HCV in England were treated in secondary care with

1. 1 . 1 |
Inclusion criteria 1.All adults aged ≥18 years who self-referred themselves to the addiction centre.2. Able and willing to give informed consent.1.1.2| Exclusion criteria 1. Unwilling to give informed consent.Such individuals were still offered blood-borne virus (BBV) testing and HCV treatment, but their data were not collected.

F
I G U R E 1 (A) Participant flow chart.(B) Summary of hepatitis C virus (HCV) treatment data.(C) Intention to treat (ITT) and modified (mITT) sustained virological (SVR12) rates (95% confidence intervals) in whole cohort and in those recruited in period 1 and period 2. (D) HCV care cascade.(E) Intention to treat (ITT) sustained virological response (SVR12) rates (95% confidence intervals) based on patient, viral and direct-acting antiviral regimens factors.

Table 4
shows the demographic, clinical and virological data in those receiving HCV treatment stratified by DAA regimen used.Overall, sofosbuvir-based regimes were used in 174/272 (64%) and 63/272 (23%) received ribavirin consistent with national prescribing policy.Those receiving glecaprevir/pibrentasvir were younger, less likely to receive ribavirin and more likely to have a psychiatric diagnosis and receive 8 weeks treatment.This was the preferred option in younger more chaotic individuals.Those receiving paritaprevir/ombitasvir and dasabuvir were older, less likely to have psychiatric diagnoses and history of overdoses and more likely to have genotype 1 disease and receive 12 weeks of treatment with Baseline data in the whole cohort and stratified by whether recruited in period 1 versus period 2.Note:Brackets [] indicate number with data available.Normal values: bilirubin 0-21 μmol/L, ALT 0-41 IU/L, albumin 35-52 g/L, INR 0.8-1.2,platelets 150-450 × 10 9 /L.Of the 400 that underwent genotyping, 6 in period 1 and 10 in period 2 had indeterminate results.Only those with positive HCV RNA were offered Fibroscan, unsuccessful in four (three in period 1 and one in period 2).Baseline data in those that received versus those in whom hepatitis C virus treatment was deferred.
Univariate and multivariate predictors of receiving hepatitis C virus treatment.Bases for all categorical variables are the 'no' category.
TA B L E 5