1. INTRODUCTION

Chronic infection with hepatitis C virus (HCV) is a well-established risk factor for hepatocellular carcinoma (HCC), which ranks among the most prevalent and lethal malignancies worldwide.1 HCV triggers liver inflammation, leading to tissue necrosis, fibrosis, and cellular regeneration, thus fostering genetic mutations that drive neoplastic transformation. Moreover, HCV has direct carcinogenic potential, inducing pro-oncogenic effects within infected cells through oxidative stress, host cell checkpoint dysregulation, and deoxyribonucleic acid (DNA) damage.2,3 In addition, an estimated 10% to 20% of patients develop liver cirrhosis over a span of 20 to 30 years following HCV infection, with an annual hepatic decompensation risk of 3% to 6%, and an annual HCC risk of 1% to 5%.4

Detecting the onset of HCC is often hampered by the absence of clinical symptoms and physical indicators, resulting in late-stage diagnosis and unfavorable prognosis. Despite the availability of many therapeutic avenues for HCC, the primary curative approaches continue to be surgical liver resection, local radiofrequency ablation, and liver transplantation. Nevertheless, even after successful curative interventions, the HCC recurrence rate can be as high as 70%, thereby contributing to significant long-term mortality.5–7 To address this issue, effective antiviral therapy has emerged as the most promising strategy to prevent both the onset and recurrence of HCC in patients with chronic hepatitis C (CHC). Notably, studies from the interferon (IFN) era revealed that achieving a sustained virological response (SVR) could significantly reduce the risk of HCC occurrence and recurrence compared to not achieving an SVR.8

The landscape of CHC treatment was revolutionized with the advent of IFN-free, all-oral direct-acting antivirals (DAAs). DAAs have exhibited superior safety profiles and shorter treatment duration, even for advanced liver disease stages, including decompensated cirrhosis, yielding a remarkable SVR rate exceeding 95%.9–12 Although two studies published in 2016 raised concerns about the high risk of HCC occurrence and recurrence after DAA therapies,13,14 most recent studies have not supported that DAA therapy increases the risk of HCC recurrence.15–22 In addition, a meta-analysis of six studies encompassing 1105 patients who received DAAs vs 1912 controls either untreated or undergoing peg-IFN-α-based regimens with follow-up ranging from 1.25 to 4 years concluded that DAA therapy was correlated with a substantial 64% decrease in HCC recurrence compared to untreated controls.23

With regards to prognosticators for HCC recurrence among DAA-treated patients, previous studies have identified factors such as DAA treatment response, duration between HCC treatment and initiating DAA treatment, tumor characteristics including size and number of nodules, history of multiple HCC treatments, alpha-fetoprotein (AFP) level, and Child-Pugh score.16,17,21,24–27 However, limited real-world data are currently available on the risk factors, particularly tumor characteristics, associated with tumor recurrence in CHC-HCC patients post all-oral DAA treatment in Taiwan. Therefore, this study aimed to evaluate the therapeutic efficacy of DAAs in Taiwanese patients with prior HCV-induced HCC and investigate the independent risk factors of HCC recurrence following DAA therapy. To ensure homogeneity, we exclusively included patients with early HCC (Barcelona clinic liver cancer [BCLC] stages 0 and A) who were successfully treated without radiological evidence of residual tumors before starting DAA treatment. In addition, we developed a risk stratification system based on the significant independent factors identified in our analysis to estimate the cumulative HCC recurrence rate after DAA treatment.

2. METHODS 2.1. Study design and patient population

This retrospective, single-center study enrolled a total of 208 consecutive adult (≥18 years) patients with CHC-HCC who underwent all-oral DAA treatment at our hospital between January 2017 and August 2021. CHC was defined as the presence of detectable HCV antibodies (anti-HCV) (Abbott HCV EIA 2.0; Abbott Laboratories, Abbott Park, IL) and quantifiable serum HCV ribonucleic acid (RNA) (Cobas TaqMan HCV Test version 2.0; Roche Diagnostics GmbH, Mannheim, Germany; lower limit of quantification: 15 IU/mL) for a minimum of 6 months. The inclusion criteria were (1) HCC diagnosis at first occurrence, established either through pathology or noninvasive criteria according to guidelines from the American Association for the Study of Liver Disease (AASLD)28; (2) HCC before DAA exposure classified as BCLA stage 0 or A29; (3) attainment of a complete radiological response (CRR) via any form of anti-HCC therapy, in line with the modified Response Evaluation Criteria in Solid Tumor (mRECIST) criteria, as evidenced by the absence of residual tumors or complete necrosis assessed with dynamic computed tomography (CT) or magnetic resonance imaging (MRI) 1 to 2 months after HCC treatment, and subsequent 3 to 6-month follow-up30; and (4) verification of CRR status by at least one dynamic CT/MRI assessment conducted within 3 months before or after initiating DAA treatment. The exclusion criteria were (1) undergoing liver transplantation; (2) a follow-up duration of less than 6 months; (3) advanced stage HCC (BCLC stage B, C, D); and (4) active HCC, treated HCC lesions lacking CRR, or the presence of uncharacterized nodules within 3 months before or after commencing DAA treatment. Finally, 94 patients were included for statistical analysis. This study was granted approval by the Institutional Review Board of Taipei Veterans General Hospital and adhered to the principles of the Declaration of Helsinki and the International Conference on Harmonization for Good Clinical Practice. All patients provided written informed consent before participation. The case collection process is depicted in Fig. 1.

Fig. 1:

Flowchart of case collection. CHC = chronic hepatitis C; DAA = direct-acting antiviral; HCC = hepatocellular carcinoma.

For each enrolled patient, the following demographic characteristics and clinical parameters were recorded before initiating DAA treatment: sex, body mass index, age, history of previous antiviral treatment, liver fibrosis status, signs of portal hypertension, details and stage of HCC, previous HCC treatment modality, and recurrence history. Standard laboratory techniques were used to collect pre- and post-DAA therapy laboratory data for subsequent statistical analysis.

2.2. DAA treatment protocol

In this study, selection of the DAA regimen was based on clinical decisions by physicians, taking into account factors such as HCV genotype/subtype, viral load, patient attributes, drug-drug interaction profiles, and evolving criteria for reimbursement by the National Health Insurance Administration of the Ministry of Health and Welfare.31,32 Given the hepatotoxicity risks associated with some NS3/4A protease inhibitors, guidelines recommend pairing a nucleotide NS5B polymerase inhibitor with an NS5A inhibitor as the standard regimen for patients with decompensated cirrhosis.31,32 The HCV genotype was determined using a commercially available assay (Cobas HCV GT; Roche Diagnostics GmbH). Before the initiation of DAA therapy, a comprehensive review of the patients’ regular medications was performed to assess potential drug-drug interactions. Medications with the potential for drug-drug interactions were either discontinued, substituted with alternative drugs, or initiated at the lowest dose with regular monitoring by physicians.33 During the treatment period, the patients were evaluated by physicians at weeks 1 and 2, followed by bi-weekly assessments (or more frequent evaluations in cases of adverse events) until completion of therapy. SVR at 12 weeks (SVR12) was defined as undetectable HCV RNA (≤15 IU/mL) at the end of DAA treatment and again 12 weeks after completing therapy. Patients without SVR12 data were considered not to have achieved SVR12.

2.3. Follow-up post-DAA treatment and HCC recurrence surveillance

All of the included patients underwent periodic imaging follow-up to detect HCC recurrence, in accordance with international guidelines. AFP levels were measured every 3 months, and dynamic CT/MRI imaging was generally performed every 3 to 6 months for patients achieving CRR within 2 years. Patients who achieved CRR after more than 2 years of prior HCC treatment underwent ultrasonography or dynamic CT/MRI every 4 to 6 months. When a suspected HCC lesion was detected in the liver through ultrasonography, further imaging studies such as dynamic CT/MRI and/or biopsy were arranged for confirmation. HCC recurrence was diagnosed by histology or characteristic imaging findings observed in contrast-enhanced CT/MRI, showing hyper-enhancement in the arterial phase and a washout pattern in the portal or delayed phases.

2.4. Primary and secondary endpoints

The primary endpoint was the cumulative incidence of HCC recurrence after initiating DAA treatment. The date of HCC diagnosis, determined through typical imaging such as CT/MRI, was used for cumulative failure curve analysis. To explore the influence of tumor characteristics on post-DAA HCC recurrence, patients with BCLC stage A HCC were categorized into three subgroups: (A1) a single tumor sized 2 to 5 cm; (A2) a single tumor larger than 5 cm; and (A3) two or three nodules, none exceeding 3 cm in size. The secondary endpoints encompassed the efficacy of DAA treatment in the included patients and identification of independent baseline and 12-week post-DAA treatment variables capable of predicting HCC recurrence, including the DAA treatment response (SVR12).

2.5. Definitions

The fibrosis-4 (FIB-4) score is a noninvasive scoring system used to estimate the degree of hepatic fibrosis through laboratory test data. An FIB-4 score >3.25 has been shown to have a positive predictive value of 82.1% for confirming advanced fibrosis (F3-F4) with a specificity of 98.2%.34 In this study, advanced fibrosis (F3-4) was defined as the presence of cirrhosis-related clinical or radiological manifestations, along with evidence of portal hypertension such as esophageal or gastric varices, or a FIB-4 score >3.25.

The albumin-bilirubin (ALBI) score/grade serves as a measure of liver function. It was originally devised as a prognostic factor for HCC patients, and has since gained widespread recognition. The ALBI score is calculated as: (log10 bilirubin [µmol/L] × 0.66) + (albumin [g/L] × −0.0852). An ALBI score ≤−2.60 is defined as ALBI grade 1; a score >−2.60 to ≤−1.39 as ALBI grade 2; and a score >−1.39 as ALBI grade 3. The ALBI score/grade offers an objective measure, capable of detecting subtle changes in liver function compared to the Child-Pugh or model for end-stage liver disease (MELD) scores. The ALBI score/grade has been shown to be a good predictor of survival in patients with various etiologies of nonmalignant liver diseases.35

2.6. Statistical analysis

All statistical analyses were conducted using SPSS version 23.0 (SPSS Inc., Chicago, IL). Demographic and clinical parameters before initiating DAA treatment were presented as means, SDs, or percentages as appropriate. Categorical variables were analyzed using the chi-squared test or Fisher exact test. Continuous variables were analyzed using independent t tests or the Mann-Whitney U test. Paired t tests were used to compare sequential changes in laboratory data before and after DAA therapy. Kaplan-Meier survival and cumulative incidence curves were compared using the log-rank test. Multivariable Cox regression analysis (hazard ratios [HRs] and 95% CI estimates) was used to assess the risk of HCC recurrence at any point during follow-up, incorporating baseline and 12-week post-DAA treatment variables. A statistically significant result was defined as p < 0.05. The duration of follow-up was calculated from the initiation of DAA treatment to death, the last visit, or the final abdominal image follow-up, as applicable.

3. RESULTS 3.1. Clinical characteristics of the enrolled patients

A total of 208 patients with CHC and a history of HCC who underwent DAA treatment were initially enrolled. Of these patients, 114 were subsequently excluded according to the exclusion criteria, and the remaining 94 patients were included in the final analysis (Fig. 1). The mean age of the included patients was 75.9 ± 8.9 years, and 42 (44.7%) were male. Eighteen (19.1%) patients had failed previous pegylated IFN therapy. The median baseline HCV RNA level was 5.92 log10 IU/mL (interquartile range [IQR] = 5.09-6.52 log10 IU/mL). The distribution of HCV genotypes was as follows: 1.1% genotype 1a, 67.0% genotype 1b, 26.6% genotype 2, 1.1% genotype 3, 2.1% genotype 6, and 2.1% unclassified genotype. According to the predefined criteria, 66 (70.2%) patients were classified as having advanced (F3-4) fibrosis, and five (5.3%) were diagnosed with decompensated (CTP class B-C) cirrhosis. About the type of DAAs used, 61.7% of the patients were treated with sofosbuvir-based regimens. Detailed information about the DAA regimens is summarized in Table 1. The median follow-up period after initiating DAA treatment was 22.1 months (IQR = 8.6-35.9 months).

Table 1 - Baseline clinical characteristics of the study population Characteristics Patients (n = 94) Age, y, mean ± SD 75.9 ± 8.9 Male sex, n (%) 42 (44.7%) Genotype 1, n (%)  1a 1 (1.1%)  1b 63 (67%)  2 25 (26.6%)  3 1 (1.1%)  6 2 (2.1%)  Unclassified 2 (2.1%) HCV RNA, IU/mL  ≥1,000,000, IU/mL, n (%) 49 (52.1%)  <1,000,000, IU/mL, n (%) 45 (47.9%) Decompensated (CTP class B-C) cirrhosis (%) 5 (5.3%) Previous IFN failure, n (%) 18 (19.1%) Diabetes mellitus, n (%) 28 (29.8%) Body mass index, mean ± SD 24.78 ± 3.69 Coinfected with HBV, n (%) 3 (3.2%) Advanced fibrosis (F3-F4), n (%) 66 (70.2%) Presence of EV/GV 16 (17%) DAAs regimen  PrOD 16 (17.0%)  DCV + ASV 9 (9.6%)  SOF + RBV 13 (13.8%)  SOF + DCV 2 (2.1%)  SOF + LDV 31 (33.0%)  SOF + VEL 12 (12.8%)  EBR + GZR 2 (2.1%)  GLE + PIB 9 (9.6%)

ASV = asunaprevir; CTP = Child-Turcotte-Pugh; DAAs = direct-acting antivirals; DCV = daclatasvir; EBR = elbasvir; EV = esophageal varices; GLE = glecaprevir; GV = gastric varices; GZR = grazoprevir; HBV = hepatitis B virus; IFN = interferon; LDV = ledipasvir; PIB = pibrentasvir; PrOD = paritaprevir/ritonavir, ombitasvir, and dasabuvir; RBV = ribavirin; SOF = sofosbuvir; VEL = velpatasvir.

3.2. HCC characteristics before initiating DAA treatment and previous treatment history

Before the commencement of DAA therapy, the HCC characteristics of the 94 patients were as follows: 17.0% were classified as BCLC stage 0 (ie, single tumor ≤2 cm); 46.8% as BCLC stage A with a single tumor of 2 to 5 cm in diameter (A1); 13.8% as BCLC stage A with a single tumor ≥5 cm in diameter (A2); 11.7% as BCLC stage A with two or three nodules (A3); and 10.7% could not be categorized due to insufficient data. Before initiating DAA treatment, 31.9% of the patients had a history of HCC recurrence after initial treatment. The distribution of the first HCC treatment was as follows: 55.3% surgical treatment, 31.9% radiofrequency ablation, and 12.8% transcatheter arterial chemoembolization (TACE). About the last HCC treatment to achieve CRR, 48.9% underwent surgery, 36.2% received radiofrequency ablation, 13.8% underwent TACE, and 1.1% underwent stereotactic body radiation therapy (Table 2). The median time from the initial HCC diagnosis to DAA treatment was 23.4 months (IQR = 3.5-67.4 months), and the median time from last HCC treatment with CRR to initiating DAA treatment (treatment durability) was 7.8 months (IQR = 2.0-31.3 months).

Table 2 - HCC characteristics before the initiation of DAAs Characteristics Patients (n = 94) HCC morphology before DAA initiation by BCLC stage, n (%)  Single tumor ≤2 cm (0) 16 (17.0%)  Single tumor 2-5 cm (A1) 44 (46.8%)  Single tumor >5 cm (A2) 13 (13.8%)  Two or three nodules, each ≤3 cm (A3) 11 (11.7%)  Insufficient information to classify 10 (10.7%) Previous HCC recurrence, n (%) 30 (31.9%) HCC treatment required to achieve CRR, n (%)  1 64 (68.1%)  2 12 (12.8%)  ≥3 18 (19.1%) First HCC treatment, n (%)  Surgery 52 (55.3%)  RFA 30 (31.9%)  TACE 12 (12.8%) Time interval from initial HCC diagnosis to DAA treatment, n (%)  ≤12 mo 42 (44.7%)  12-24 mo 5 (5.3%)  24-48 mo 16 (17.0%)  48-72 mo 9 (9.6%)  >72 mo 22 (23.4%) Last HCC treatment, n (%)  Surgery 46 (48.9%)  RFA 34 (36.2%)  TACE 13 (13.8%)  Other (SBRT) 1 (1.1%) Time interval from HCC CRR to DAA treatment, n (%)  ≤6 mo 45 (47.9%)  6-12 mo 10 (10.6%)  12-24 mo 9 (9.6%)  24-36 mo 11 (11.7%)  >36 mo 19 (20.2%)

BCLC = Barcelona clinic liver cancer; CRR = complete radiological response; DAAs = direct-acting antivirals; HCC = hepatocellular carcinoma; RFA = radiofrequency ablation; SBRT = stereotactic body radiation therapy; TACE = transcatheter arterial chemoembolization.

3.3. Virological response

Following the initiation of DAA therapy, only one patient did not complete the full course due to personal reasons, and consequently did not achieve SVR12. All patients who completed DAA treatment achieved viral clearance (HCV RNA <15 IU/mL) at the end of the treatment. Posttreatment follow-up revealed that three patients had virologic relapse at 12 weeks after completing DAA treatment. Intent-to-treat analysis demonstrated an SVR12 rate of 95.7% (90/94 patients). Among the four patients who did not achieve SVR12, one received the daclatasvir + asunaprevir regimen, and the other three received the sofosbuvir + ribavirin regimen. The SVR12 rate was 100% (21/21) in the pan-genotype regimen and 94.5% (69/73) in the genotype-specific regimen (p = 0.572).

3.4. Laboratory data before and 12 weeks after DAA treatment

Laboratory data of the patients before and 12 weeks after DAA treatment are summarized in Table 3. The median alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels at 12 weeks after DAA treatment (EOT12) were 19 and 25 IU/L, respectively, and the median levels of albumin, total bilirubin, and platelet count at 12 weeks after DAA were 4.1 g/dL, 0.83 mg/dL, and 131 × 1000/μL, respectively. Based on these data, the distribution of ALBI grades was as follows: grade 1 (56.4%), grade 2 (41.5%), and grade 3 (2.1%). Compared to pretreatment values, significant declines in AST (p < 0.001), ALT (p < 0.001), and FIB-4 score (p < 0.001) were observed after DAA therapy. In addition, significant increases in albumin (p < 0.001) and hemoglobin (p = 0.002) were noted after DAA therapy. Due to the potential effect of hepatitis on laboratory values, EOT12 laboratory data were used to investigate the risk of future HCC recurrence in the analysis.

Table 3 - Laboratory data before and 12 wk after DAA treatment of the enrolled patients (n = 94) Characteristics Before DAA treatment 12 wk post-DAA treatment p (paired) Mean white cell count (×1000/μL) 5.1 ± 1.7 5.4 ± 1.8 0.084 Mean hemoglobin, g/dL 12.4 ± 1.9 12.8 ± 1.9 0.002 Mean platelet count (×1000/μL) 140 ± 59 134 ± 51 0.173 Mean prothrombin time, s 11.9 ± 2.0 12.0 ± 1.9 0.810 Mean albumin, g/dL 3.8 ± 0.5 4.0 ± 0.5 <0.001 Mean AST, U/L 79 ± 74 34 ± 43 <0.001 Mean ALT, U/L 75 ± 67 28 ± 48 <0.001 Mean total bilirubin, mg/dL 0.82 ± 0.42 0.86 ± 0.42 0.244 Mean creatinine, mg/dL 1.12 ± 1.06 1.21 ± 1.25 0.138 Mean alpha-fetoprotein, ng/mL 51.5 ± 153.3 241.6 ± 1513.9 0.219 Mean FIB-4 score 5.96 ± 4.66 4.43 ± 2.96 <0.001 ALBI grade 1, n (%) 46 (48.9%) 53 (56.4%) 0.179

ALBI = albumin-bilirubin; ALT = alanine aminotransferase; AST = aspartate aminotransferase; DAA = direct-acting antiviral; FIB-4 = fibrosis-4.

3.5. Cumulative incidence of HCC recurrence and comparison of characteristics

Over a median follow-up period of 22.1 months (IQR = 8.6-35.9 months) after DAA treatment, HCC recurred in 51 (54.3%) of the study patients. Recurrence occurred adjacent to the site of the previously treated (ablated, operated, or embolized) tumors in 18 (35.3%) of these 51 patients. Seven of these 18 (38.9%) patients had recurrence within 1 year after the last HCC treatment, whereas 11 (61.1%) had recurrence more than 1 year after the last treatment. The cumulative incidence rates of HCC recurrence after initiating DAA treatment were 31.1% at 1 year, 42.5% at 2 years, 57.3% at 3 years, 60.6% at 4 years, and 68.5% at 5.7 years (Fig. 2). The median time to HCC recurrence from initiating DAA treatment was 9.6 months (IQR = 5.8-22.6 months). Comparing the characteristics of the patients with and without recurrent HCC, those who developed HCC recurrence were significantly more likely to have an EOT12 AFP level >10 ng/mL (19.6% vs 4.7%, p = 0.03) (Table 4). Furthermore, the patients with a history of recurrent HCC before initiating DAA treatment (p = 0.003) and those with BCLC A3 lesions (two or three nodules without any tumor exceeding 3 cm, p = 0.044) were more likely to experience HCC recurrence after DAA treatment (Table 4).

Table 4 - Comparison of the characteristics between the patients with or without HCC recurrence post-DAA treatment Characteristics No HCC recurrence (n = 43) HCC recurrence (n = 51) p Mean age, y 75.6 ± 8.4 76.3 ± 9.4 0.709 Male sex, % 18 (41.9%) 24 (47.1%) 0.614 Mean BMI, kg/m2 24.6 ± 4.0 25.0 ± 3.4 0.611 HCV RNA ≥1,000,000, IU/mL (%) 21 (48.8%) 24 (47.1%) 0.863 HCV genotype 1, % 31 (73.8%) 33 (66.0%) 0.417 Previous IFN failure, % 8 (18.6%) 10 (19.6%) 0.902 Diabetes mellitus, % 11 (25.6%) 17 (33.3%) 0.413 SOF-based DAAs, % 27 (62.8%) 31 (60.8%) 0.842 SVR12 rate by DAAs, % 43 (100%) 47 (92.2%) 0.082 Advanced fibrosis (F3-F4), % 28 (65.1%) 38 (74.5%) 0.321 Decompensated (CTP class B-C) cirrhosis, % 3 (7.0%) 2 (3.9%) 0.511 Mean EOT12 albumin, g/dL 4.04 ± 0.44 3.98 ± 0.47 0.500 Mean EOT12 AST, U/L 35.2 ± 60.9 33.1 ± 18.1 0.815 Mean EOT12 ALT, U/L 28.8 ± 67.9 27.5 ± 19.6 0.893 Mean EOT12 total bilirubin, mg/dL 0.82 ± 0.43 0.90 ± 0.40 0.336 EOT12 AFP >10 ng/mL (%) 2 (4.7%) 10 (19.6%) 0.03 Mean EOT12 WBC (×1000/μL) 5.49 ± 1.72 2.29 ± 1.84 0.600 Mean EOT12 Hgb, g/dL 12.7 ± 1.7 12.9 ± 2.1 0.484 Mean EOT12 platelet count (×1000/μL) 140 ± 59 129 ± 43 0.324 Mean EOT12 prothrombin time, s 11.8 ± 1.5 15.2 ± 22.2 0.329 EOT12 ALBI grade, % 0.250  1 27 (62.8%) 26 (51.0%)  2/3 16 (37.2%) 25 (49.0%) BCLC stage before DAAs, n (%) 0.044  0/A1/A2 37 (86.0%) 36 (70.6%)  A3 2 (4.7%) 9 (17.6%)  Unclassified 4 (9.3%) 6 (11.8%) Prior HCC recurrence, n (%) 7 (16.3%) 23 (45.1%) 0.003 Last HCC treatment, n (%) 0.113  Surgery 26 (60.5%) 20 (39.2%)  RFA 14 (32.6%) 20 (39.2%)  TACE 3 (7.0%) 10 (19.6%)  Other (SBRT) 0 (0%) 1 (2.0%) The mean time interval from HCC CRR to DAA treatment, mo 25.2 ± 36.1 21.3 ± 34.8 0.838

The bold value merely indicates that the p-value is statistically significant, but it is not necessary; it can also be presented without bold.

ALBI = albumin-bilirubin; ALT = alanine aminotransferase; AST = aspartate aminotransferase; BCLC = Barcelona clinic liver cancer; BMI = body mass index; CRR = complete radiological response; CTP = Child-Turcotte-Pugh; DAA = direct-acting antiviral; EOT12 = 12 wk post end of DAAs treatment; HCC = hepatocellular carcinoma; HCV = hepatitis C virus; Hgb = Hemoglobin; IFN = interferon; RFA = radiofrequency ablation; SBRT = stereotactic body radiation therapy; SOF = sofosbuvir; TACE = transcatheter arterial chemoembolization; WBC = white blood cell.

Fig. 2:

The overall cumulative incidence of HCC recurrence after initiating DAA treatment was 31.1% at 1 y, 42.5% at 2 y, 57.3% at 3 y, 60.6% at 4 y, and 68.5% at 5.7 y, respectively. DAA = direct-acting antiviral; HCC = hepatocellular carcinoma.

3.6. Independent factors associated with HCC recurrence

Building on the differences in characteristics between the patients with and without HCC recurrence, continuous variables were transformed into categorical data for Cox regression analysis. In univariate analysis, the variables associated with HCC recurrence were failure to achieve SVR12 after DAA treatment, EOT12 AFP >10 ng/mL, BCLC A3 lesions, last HCC treatment with CRR being TACE, and history of previous recurrence before DAA treatment (Table 5). Because TACE is the standard treatment option for BCLC A3 lesions, to avoid a dilutional effect, only the variable of BCLC A3 lesions was selected into multivariate analysis. A multivariable Cox model was then used to identify variables independently associated with the primary endpoint, and the results were as follows: previous tumor recurrence before DAA treatment (HR = 3.15, 95% CI, 1.63-6.11, p = 0.001), failure to achieve SVR12 after DAA treatment (HR = 6.829, 95% CI, 1.42-32.83, p = 0.016), EOT12 AFP >10 ng/mL (HR = 2.34, 95% CI, 1.06-5.17, p = 0.036), and multiple tumor nodules (BCLC A3 lesions) compared to single nodules (BCLC 0, A1, A2) (HR = 2.31, 95% CI, 1.05-5.09, p = 0.039) (Table 5).

Table 5 - Cox regression analysis for independent predictors of HCC recurrence Parameter Univariate Multivariate HR 95% CI p HR 95% CI p Age, y  ≥65 vs <65 1.244 0.490-3.160 0.646 Sex