Hepatocellular Carcinoma Risk Declines but Remains High Enough for Screening in the First 7 Years After Hepatitis C Virus Cure With Direct-Acting Antivirals in Patients With Cirrhosis or High Fibrosis-4 Score

Hepatitis C virus (HCV) eradication by direct-acting antiviral (DAA) medications has been found to reduce hepatocellular carcinoma (HCC) risk, but a substantial risk persists in patients with cirrhosis and those with Fibrosis-4 (FIB-4) score ≥3.25.^1,2 In a previous study, we found that, during the first 4 years after DAA-induced sustained virologic response (SVR), annual HCC incidence remained well above 2% in those with pretreatment cirrhosis and above 1% in those without cirrhosis but with pretreatment FIB-4 score ≥3.25.¹ In that study, annual HCC risk declined during the first 4 years of follow-up, but the follow-up time was not long enough to determine whether HCC risk continues to decline to levels low enough that HCC screening may no longer be warranted. In this study, we extended the follow-up of this DAA-cured cohort up to 7 years to determine whether annual HCC risk, stratified by cirrhosis and FIB-4 categories, declines to levels low enough to preclude the requirement for HCC screening.

Using electronic health records (Supplementary Methods),¹ we identified 29,033 patients infected with HCV enrolled in the Veterans Affairs (VA) health care system who achieved DAA-induced SVR from January 2013 to December 2015. The FIB-4 score was calculated using data recorded within 6 months preceding treatment as: (age × aspartate aminotransferase ) / (platelet count × √ alanine aminotransferase). Cirrhosis was diagnosed based on the recording of at least 2 International Classification of Disease (ICD), 9^th or 10^th revision codes for cirrhosis or related complications before treatment initiation.¹ Patients were followed for incident HCC until December 2021, which was defined by the documentation of at least 2 ICD-9 155.0 or ICD-10 C22.0 codes more than 180 days after treatment initiation. We calculated the annual HCC incidence using both competing risk and censored analysis during each year of follow-up after treatment, stratified by pretreatment cirrhosis status and FIB-4 score and change in FIB-4 score within 1 year after treatment initiation.

Most patients were male (96.6%) and non-Hispanic White (52.2%). Mean age was 61.1 years, and diabetes (28.9%), alcohol use disorder (43.7%), and substance use disorders (37.7%) were common (Supplementary Table 1).

Among the 7533 patients with pretreatment cirrhosis, 948 (12.6%) developed HCC during a mean follow-up period of 4.9 years (2.6 per 100 patient-years). In patients with FIB-4 score ≥3.25, the annual HCC incidence decreased from 3.8% in year 1 to 1.4% in year 7 (test of trends, P < .001), but remained substantial even up to 7 years post SVR (Figure 1). In patients with cirrhosis and FIB-4 score <3.25, the annual HCC incidence ranged from 0.7% to 1.3% and did not change significantly over time (P = .53).

Figure 1.

Annual HCC incidence in 29,033 patients with HCV who achieved DAA-induced SVR during 2013–2015, followed until December 2021, according to pretreatment cirrhosis status and FIB-4 score. ^aHCC risk calculated by censoring patients at the time of death or liver transplantation. ^bHCC risk calculated by performing a competing risk proportional subhazards analysis with death and liver transplantation as competing risks.

Among the 21,500 patients without pretreatment cirrhosis, 541 (2.5%) developed HCC during a mean follow-up period of 5.4 years (0.5 per 100 patient-years). Patients with FIB-4 score ≥3.25 had a significantly higher HCC incidence (1.2 per 100 patient-years) compared with those with FIB-4 score <3.25 (0.2 per 100 patients-years) (Figure 1). In patients without cirrhosis and with FIB-4 score ≥3.25, the annual HCC incidence remained stable, but substantial for up to 7 years post SVR (0.8%–1.3%; P = .06). For those without cirrhosis and with FIB-4 score <3.25, the annual HCC incidence varied between 0.2% and 0.4% per year (P = .04).

In a subgroup analysis examining HCC incidence according to change in pre-/post-treatment FIB-4 score, those with cirrhosis continued to have a high HCC incidence rate regardless of change in FIB-4 score (1.0–4.5 per 100 patient-years) (Supplementary Table 2). Among those without cirrhosis with persistently elevated FIB-4 score ≥3.25, HCC incidence was high (2.2 per 100 patient-years). In those without cirrhosis and a drop in FIB-4 score from ≥3.25 pretreatment to <3.25 post-treatment, HCC incidence was lower but substantial (1.0 per 100 patient-years).

As the number of patients achieving DAA-induced SVR continues to increase, understanding HCC risk over time is of utmost importance, as it can inform HCC screening strategies during follow-up. In this study, we provide HCC incidence by cirrhosis and FIB-4 categories up to 7 years post SVR. Similar to prior studies with up to 3–4 years of follow-up,^1,3,4 HCC incidence was highest in patients with cirrhosis and FIB-4 score ≥3.25, followed by patients with cirrhosis and FIB-4 <3.25 and patients without cirrhosis with FIB-4 score ≥3.25. HCC risk in these 3 groups met or exceeded the approximately 1% per year threshold considered to be cost-effective with current modalities⁵ for each year and up to 7 years post SVR. In patients without cirrhosis with FIB-4 score <3.25, the risk of HCC remained below the cost-effective threshold throughout the follow-up time period. HCC incidence estimates were higher when obtained using the censored approach than the competing risk analysis (Figure 1). Our findings support continuing HCC screening in patients with cirrhosis and in patients without cirrhosis with a pretreatment FIB-4 score ≥3.25.

In patients with cirrhosis and FIB-4 score ≥3.25, HCC incidence was the lowest during year 7 of follow-up and in the calendar year 2020. One reason for this nadir may be the COVID-19 pandemic, which began in the year 2020, and has had detrimental impacts on HCC screening and early detection.⁶ Our group, among others, previously described that HCC screening and diagnosis rates within the VA declined 44% and 13%, respectively, after the pandemic.⁷ As current guidelines recommend HCC screening only in patients with cirrhosis,⁸ it is not surprising that this drop in HCC incidence was only observed in patients with cirrhosis (Figure 1). The confounding effects of the COVID-19 pandemic on HCC screening and diagnosis preclude firm conclusions on the trends in HCC risk post SVR in this highest risk group and necessitate longer follow-up.

In summary, we found that all patients with DAA-induced SVR, except those with low FIB-4 scores and no pretreatment cirrhosis, had high enough HCC incidence rates, even up to 7 years post SVR, to warrant continued HCC screening. In patients with cirrhosis and FIB-4 score ≥3.25, who have the highest absolute risk, HCC incidence appeared to decline progressively each year and up to 7 years post SVR, although it remained well above thresholds that warrant screening. These early results require validation in non-veteran cohorts and with follow-up that extends beyond the COVID-19 pandemic when, hopefully, HCC screening and diagnosis practices return to normal.

Abbreviation used in this paper:

DAA

direct-acting antiviral

FIB-4

Fibrosis-4 score

HCC

hepatocellular carcinoma

HCV

hepatitis C virus

ICD

International Classification of Diseases

SVR

sustained virologic response

VA

Veterans Affairs

Data Availability

Individual data will not be shared. More information on the methods and analysis are available by contacting the corresponding author.