Introduction

Direct-acting antiviral (DAA) therapies for hepatitis C virus (HCV) infection are associated with very high rates (>95%) of sustained virological response (SVR) at 12 weeks.^1-3 Nevertheless, non-attendance for SVR testing remains a challenging aspect in accurately documenting SVR and ensuring continuity of care.^4-7 A concern in marginalized populations is treatment adherence to maximize chance of cure while also engaging to ensure patients can be treated again, should treatment failure occur. Other patients with greater risk for advanced liver disease or at risk of progression may need regular review and/or ongoing surveillance for liver cancer. While the rate of loss to follow-up (LTFU) has been reported to be small in randomized clinical trials,^{8, 9} for various reasons, trial patients may differ from real-world settings, and data are conflicting with LTFU rates varying with patient characteristics (e.g. people who inject drugs, gender, genotype, prior HCV treatment).^{4, 5, 7, 10}

The rapid uptake of DAA therapy in Australia has led to a 30% reduction in liver-related deaths due to HCV since 2015 and a plateauing of HCV-related hepatocellular carcinoma (HCC) cases.¹¹ Having data on the rate of LTFU and the factors that may affect engagement with treatment and follow-up are crucial to inform further improvement of HCV service delivery and treatment. Moreover, continued engagement after HCV therapy is vital to monitor for progressive liver disease to help manage comorbid risks and follow-up for potential HCV reinfection. Using data from a large multicenter prospective study of people with HCV infection who received DAA therapy in Australia, we report the rate of LTFU from HCV treatment and examine factors that affect LTFU in a real-world setting.

Methods

Details of the OPERA-C study have been previously described.¹² Briefly, the OPERA-C is a multicenter prospective study of adults with HCV who received DAA therapy during February 2016 to July 2021 in one of 26 centers across Australia (six states and one territory).

Results

OPERA-C study recruited 3729 patients, and after exclusions (n = 110 patients; see Fig. 1), 3619 were included in the analysis. Most patients (86.5%) were recruited from gastroenterology/hepatology clinics at tertiary centers. At recruitment, patients were 52.0 years (SD = 10.5), 33.6% had cirrhosis (69.4% Child–Pugh class B/C), and 19.3% had HCV treatment prior to DAA era. Over half of the cohort had genotype 1 (53.7%) and 37.6% genotype 3.

Rate of LTFU according to sociodemographic and clinical factors

The rate of LTFU was highest among First Nations Peoples (28.5%, 95% CI 20.2–36.7) and patients with a history of IDU and/or ORT (23.2%, 95% CI 20.0–26.4), anxiety (19.2%, 95% CI 16.3–22.0), and depression (18.4%, 95% CI 16.0–20.9) (Fig. 2). Compared to patients aged ≥60 years who had an LTFR of 6.0% (95% CI 4.4–7.6), the rate of LTFU was over threefold higher for patients 18–49 years (22.4%, 95% CI 20–2-24.7) and nearly double for 50–59 years (11.8%, 95% CI 10.1–13.4).

Discussion

In this real-world experience of over 3600 Australians with HCV infection treated with DAAs over a 4.5-year period, we assessed factors associated with being lost to follow-up. Younger age and identifying as First Nations Peoples more than doubled a person's risk of being lost to follow-up after HCV treatment. People with mental health comorbidity were more likely to be lost to follow-up from HCV treatment services. Interestingly, these people did continue to utilize mental health services, providing an important opportunity for continuity of care. Report of current IDU and receipt of opioid replacement therapy at the time of DAA initiation were also important determinants of LTFU. As might be anticipated, younger people and those with a shorter duration of infection and less severe liver disease were more likely to be LTFU. Taken together, these data point to a number of vulnerable groups who might be prioritized for more intense efforts to follow-up to reduce the risk of being lost to care. Furthermore, the data reflect the diverse nature of the HCV-affected community and that developing equivalent diversity in the health services treating HCV and following up post-treatment are critical. Key opportunities for expanding treatment and follow-up are in First Nations primary health care, addiction treatment, and mental health services.¹⁹

Unlike many patient HCV registries, patients in the OPERA-C cohort were well characterized for liver disease and cirrhosis. In this cohort, people with HCV cirrhosis were three times less likely to be LTFU, confirming previous findings of the REACH-C study.²⁰ The Australian experience differs from previously published data showing either no difference in LTFU rates or an increased risk of LTFU from HCV treatment among patients with advanced liver disease. A German study including 7747 patients reported no association between LTFU from HCV treatment and presence of cirrhosis.⁷ In the OPERA-C study, participants were mostly recruited from gastroenterology/hepatology clinics located in tertiary hospitals, while half the centers included in the German study were clinics specialized in ORT treatment. In Germany and in Australia, ORT and cirrhosis treatment rarely occur in the same site, leading to a disjunction in care. Therefore, in Australia, as patients with cirrhosis received HCV treatment in the same clinic as their cirrhosis treatment, they were less likely to be LFTU for HCV treatment. Moreover, novel data from the OPERA-C study showed that patients with less liver fibrosis (higher LSM) appeared to be more likely to be LTFU.

Importantly, patients with a longer history of infection or prior treatment were nearly three times more likely to remain engaged in follow-up than patients whose first treatment was with DAAs. The study population who initiated HCV treatment in 2017 or later, which included younger patients and more vulnerable groups, were 2.82 more likely to be LTFU compared to those treated early, when DAA medication for HCV became funded through Australia's universal health-care system.

The rate of LTFU reported here were in line with previous reports from comparable countries such as Canada⁵ (13%), Italy⁴ (10.5%), and Germany^{7, 21} (2.2–13%), as well as data from Australia (6.8–15%).^{1, 20} These studies also reported that rates of LTFU from HCV treatment varied according to patient group. The high rate of LTFU from HCV treatment among younger patients included in the OPERA-C study was consistent with that of another Australian²⁰ and an Italian study.⁴ In contrast, a Canadian⁵ study reported no association with age at treatment initiation and LTFU. While male sex was associated with LTFU in some studies in Europe,^{4, 7} this has not been shown previously in our or prior Australian data.²⁰ Many studies have reported differences in health-seeking behavior according to gender, with adult men consistently consulting less than adult women in primary care, after adjusting for consultations for reproductive reasons.^{22, 23} The lower health-seeking behavior in men may partly explain our findings of poor engagement in HCV follow-up compared to women.

First Nations Peoples in our study were twice as likely to be LTFU from HCV treatment compared to non-Indigenous Australians. HCV prevalence and incidence is higher for First Nations Peoples,²⁴ and liver disease is a major contributor to the mortality gap between First Nations Peoples and non-Indigenous Australian adults.²⁵ In 2015, the rate of newly acquired HCV was 13-fold higher among First Nations Peoples compared to non-Indigenous Australians.²⁴ Understanding the sociodemographic and clinical characteristics of First Nations Peoples with HCV who initiate HCV treatment is key to inform potential interventions which may improve delivery of HCV care for this patient group. In the OPERA-C study, prior analysis focused on First Nations patients identified younger age, less advanced liver disease (as measured by FIB4), and a recency of treatment as factors associated with higher LTFU.¹⁸

The characteristics of the HCV population who initiated treatment early was significantly different from those treated after 2016. The “early HCV treatment cohort” were older and had more advanced liver disease (cirrhosis, HCC) and possibly patients engaged in care for other conditions. In contrast, the “later HCV treatment cohort” comprised younger patients and more vulnerable groups, namely, First Nations Peoples, patients of lower socioeconomic status, IDU, and/or ORT and patients on medication for depression and psychotic illness. This might suggest that patients with less liver disease defined care needs may predominate. In the REACH-C study,²⁰ year of initiation of DAA treatment was also the strongest predictor of LTFU; patients who initiated HCV treatment in 2019 were 4.5 times (95% CI 3.57–5.64) more likely to be LTFU compared to those initiating treatment in 2016. Our study did not directly assess models of HCV treatment, but does point to opportunities to optimize engagement for vulnerable groups.

Our study found higher LTFU in patients on ORT. Other studies have shown patients who received ORT in a different clinic than their HCV treatment were over three times more likely to be LTFU from HCV treatment (adj-OR = 3.35 95% CI 2.35–4.78).⁷ In a Canadian study,⁵ participants receiving ORT in the same clinic as HCV treatment were less likely to be LTFU (adj-OR = 0.09 95% CI 0.02–0.46).⁵ In our study, patients who were current IDU or were on ORT were 1.66 times more likely to be LTFU for HCV treatment. In the Australian clinical environment, these data highlight the need to expand treatment into ORT centers and practitioners.¹⁹

HCV infection often occurs concurrently with mental health disorders. Prior to DAA era, mental health comorbidity often precluded interferon-based treatment due to the risk of neuropsychiatric side effects and worsening of existing mental health symptoms.²⁶ With the uptake of DAA therapy for HCV, mental health illness is no longer a barrier to treatment,²⁷ but appears to be an ongoing barrier to continuing care after treatment.

Strengths of this study are the prospective and multicenter design, linked data from national databases (MBS/PBS), the use of validated methods to assess fibrosis, cirrhosis, and liver disease severity. The use of medications as a surrogate to identify people with comorbidities may underestimate comorbidity burden as non-pharmacological approaches are often used (e.g. diet-controlled diabetes mellitus). Notwithstanding these limitations, the RxRisk-V has been validated and used in Australian and international cohorts.^{18, 28, 29} While patients without follow-up data for >12 months were LTFU from the treating center, it is possible other clinical follow-up occurred elsewhere, and such data were not captured. Lastly, there may be other factors that may explain LTFU that were not examined in this study such as history of incarceration or the use of non-MBS-related services.

Sociodemographic and clinical factors were associated with LTFU. Broadening and strengthening the network of health providers caring for patients with HCV may help to address this. Targeting vulnerable groups such as First Nations Peoples, youth health, and those with addiction and mental health disorders should be enabled to care for patient assessment, treatment, and follow-up as well as screen for advanced liver disease. For patients who remain at risk of reinfection or of progressive liver disease, engagement is critical. Expanding the role of GPs and other health providers in relation to HCV care will require sufficient support from specialists to enable this. It will require formal education and training programs focused on HCV screening, assessment, treatment, follow up, and the importance of continuity of care. Our findings emphasize the need to tailor patient-centered HCV care and may inform strategies to improve continuity after HCV treatment. Provision of telehealth, outreach nursing from the tertiary clinic, access to point-of-care RNA testing, and more flexible clinic locations and appointment availability could also encourage priority populations to engage with HCV care.

Informed consent

Informed consent was obtained from all subjects who participated in the study. The Human Ethics Committees of the Royal Brisbane and Women's Hospital (HREC/15/QRBW/183), participating hospitals, and QIMR Berghofer Medical Research Institute (P2126) approved this study. All procedures followed were in accordance with the ethical standards of the abovementioned committees and with the Helsinki Declaration of 1975, as revised in 2008.