Impact of MAFLD on HBV-Related Stage 0/A Hepatocellular Carcinoma after Curative Resection

Backgrounds and Aim: Metabolic-associated fatty liver dis-ease (MAFLD) is a novel term proposed in 2020 to avoid the exclusion of certain subpopulations, though the application of this term in the real world is very limited. Here, we aimed to evaluate the impact of MAFLD on hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC) after curative resection. Methods: Patients with chronic hepatitis B (CHB)-related HCC who received hepatectomy between January 2010 and December 2019 were consecutively selected. The association between histologically proven concurrent MAFLD and clinical outcomes were retrospectively analyzed. Results: Among the 812 eligible patients with CHB-related HCC, 369 (45.4%) were diagnosed with concurrent MAFLD. After a mean follow-up of 65 months, 303 patients (37.3%) developed HCC recurrence, 111 (13.7%) died, and 12 (1.5%) received liver transplantation. Although no differences in the incidences of HCC recurrence (HR: 0.902, 95% CI: 0.719–1.131, p = 0.370) and death or liver transplantation (HR: 0.743, 95% CI: 0.518–1.006, p = 0.107) were observed between patients with and without MAFLD in multivariate analysis, the patients with MAFLD tended to achieve better recurrent-free survival compared to patients without MAFLD. Notably, lean MAFLD (BMI < 23 kg/m2) was a relative risk factor for tumor recurrence (HR: 2.030, 95% CI: 1.117–3.690, p = 0.020) among patients with MAFLD. Conclusions: The overall prognosis in HBV-related early-stage HCC, in terms of HCC recurrence and death or liver transplantation, was not significantly different between patients with and without MAFLD. Among patients with MALFD, lean-MAFLD was a risk factor for HCC recurrence. Further studies are warranted to validate these results.


Introduction
Hepatocellular carcinoma (HCC) is the sixth most commonly diagnosed cancer worldwide and ranks fourth in terms of cancer mortality [1,2]. The high incidence of HCC in Asia

Study Design and Ethics
This study was designed as a multicenter cross-sectional retrospective study in Taiwan. The Institutional Review Board of Kaohsiung Chang Gung Memorial Hospital approved this study (IRB number: 201701632A3), with a waiver of the requirement for informed consent owing to the retrospective design of the study with minimal risk to the participants.

Study Population
The data were obtained from the Chang Gung Research Database (CGRD), which is derived from the largest private hospital system in Taiwan, Chang Gung Memorial Hospital (CGMH); the database is systematically updated annually to include new data generated at CGMH. The CGRD data is obtained from two medical centers and two regional hospitals: Keelung, Linkou, Chiayi, and Kaohsiung CGMH. We retrospectively reviewed the CGRD database and retrieved data for patients with HCC treated between January 2010 to December 2019.
The inclusion criteria were as follows: (1) patients diagnosed with only HBV infection based on the presence of hepatitis B surface and negative hepatitis C antibody; (2) Barcelona Clinic Liver Cancer (BCLC) stage 0 and A (early-stage) HCC; (3) who received curative liver resection and for whom a pathologic hepatic steatosis report was available.

Data Collection
All data were collected retrospectively from medical record at the time of surgery, including age, gender, presence of T2DM, hypertension, alcohol consumption, smoking history, serum biochemistry, and hepatitis B markers, and HBV DNA (detection limit of 20 IU/mL, Roche COBAS TaqMan; Roche Molecular System, Branchburg, NJ, USA). The histological features of the resected tumor, including satellite nodules, capsule invasion, microvascular invasion, tumor differentiation, histologic grade, and the cirrhosis were recorded.

Study Outcomes
The primary outcome was recurrence-free survival (RFS), defined as the interval between surgery and the date of diagnosis of the first HCC recurrence. The secondary outcome was overall survival (OS), defined as the interval between the date of surgery and death or liver transplantation, or date of last follow-up. The end date of follow-up was 31 December 2020.

Definition
The proposed criteria for diagnosis of MAFLD are based on evidence of hepatic steatosis (>5%) plus one of the following three criteria, namely overweight/obesity (BMI ≥ 23 kg/m 2 in Asians), presence of type 2 diabetes mellitus (T2DM), or evidence of metabolic dysregulation (≥2 of the following metabolic risk abnormalities: waist circumference ≥ 90/80 cm in Asian men and women, blood pressure ≥ 130/85 mm Hg or specific drug treatment, plasma triglycerides ≥ 150 mg/dL or specific drug treatment, plasma high-density lipoproteincholesterol < 40 mg/dL for men and <50 mg/dL for women or specific drug treatment, prediabetes, a homeostasis model assessment of insulin resistance score (HOMA-IR index) ≥ 2.5, and a plasma high-sensitivity C-reactive protein (hs-CRP) level > 2 mg/L) [17].
The diagnosis of HCC was defined according the histopathology reports for surgically resected tumor tissues and based on the criteria of the practice guidelines of the EASL or AASLD [20,21]. HCC was staged according to the BCLC guide-lines [22]. Histologic grade of tumor differentiation was scored using the modified nuclear grading scheme outlined by the Edmondson and Steiner, with tumor grade categorized as well, moderately, and poorly differentiated [20]. Liver cirrhosis was defined as Ishak fibrosis score 5-6 from non-tumor part [21]. T2DM was defined based on the World Health Organization (WHO) National diabetic group criteria [23]. Lean-MAFLD was subclassified as a BMI < 23 kg/m 2 in Asians [24].

Statistical Analysis
Statistical analyses were performed using SPSS Version 23.0. (IBM Corp., Armonk, NY, USA) for Windows. Continuous variables were expressed as means ± standard deviations, while categorical variables were summarized as frequencies and relative percentages. The relationship between RFS and OS was analyzed using Kaplan-Meier survival curves, and comparisons were determined using the log-rank test. Cox proportional hazards regression models was employed for univariate and multivariate analysis of the hazard ratio (HR) of RFS and OS. p-values < 0.05 were considered statistically significant.

Patient Characteristics
A total of 952 patients with BCLC stage 0 or A HCC received primary curative hepatectomy between January 2010 and December 2019. We excluded 93 patients without pathological hepatic steatosis reports and 47 patients whose clinical profiles were not available or undetermined in terms of the diagnostic criteria for MAFLD. Ultimately, the remaining 812 patients were eligible for this analysis ( Figure 1); 443 patients had HCC with MAFLD (the MAFLD group) and the other 369 patients had HCC without MAFLD (the non-MAFLD group).

Patient Characteristics
A total of 952 patients with BCLC stage 0 or A HCC received primary curative tectomy between January 2010 and December 2019. We excluded 93 patients w pathological hepatic steatosis reports and 47 patients whose clinical profiles we available or undetermined in terms of the diagnostic criteria for MAFLD. Ultimate remaining 812 patients were eligible for this analysis ( Figure 1); 443 patients had with MAFLD (the MAFLD group) and the other 369 patients had HCC without M (the non-MAFLD group).  Table 1 presents the baseline clinicopathological characteristics of the study c The mean age of the study population was 56.2 years, and the majority of patients male (n = 693, 85.3%). The median tumor diameter was 2.7 cm and all patients had stage 0 (n = 189, 23.3%) or A (n = 623, 76.7%) HCC. Moreover, 199 patients (24.6% diabetic before surgery, and 410 patients (50.5%) were diagnosed with cirrhosis. pared with the non-MAFLD group, patients in the MAFLD group had a signifi higher body mass index (BMI; p < 0.001), more frequently had T2DM (p < 0.001), h tension (p < 0.001), and used statins (p = 0.008), had higher serum ALT (p = 0.001) higher platelet count (p < 0.001), and were more likely to have well-differentiated logic grade (p < 0.001), but were less likely to have microvascular invasion (p < 0.00 had lower serum AFP (p = 0.006).   Table 1 presents the baseline clinicopathological characteristics of the study cohort. The mean age of the study population was 56.2 years, and the majority of patients were male (n = 693, 85.3%). The median tumor diameter was 2.7 cm and all patients had BCLC stage 0 (n = 189, 23.3%) or A (n = 623, 76.7%) HCC. Moreover, 199 patients (24.6%) were diabetic before surgery, and 410 patients (50.5%) were diagnosed with cirrhosis. Compared with the non-MAFLD group, patients in the MAFLD group had a significantly higher body mass index (BMI; p < 0.001), more frequently had T2DM (p < 0.001), hypertension (p < 0.001), and used statins (p = 0.008), had higher serum ALT (p = 0.001) and a higher platelet count (p < 0.001), and were more likely to have well-differentiated histologic grade (p < 0.001), but were less likely to have microvascular invasion (p < 0.001) and had lower serum AFP (p = 0.006).

Impact of MAFLD on the Outcomes of HBV-HCC
After a mean follow-up of 65 months, 303 patients (37.3%) had developed recurrent HCC, 111 (13.7%) died, and 12 patients (1.5%) had received liver transplantation. The rates of RFS and OS after LR in HBV-related HCC among the patients with and without MAFLD are shown in Figure 2. Although there were no statistically significant differences between patients with MAFLD and without MAFLD, patients with MAFLD had favorable RFS and OS compared to patients without MAFLD (p = 0.37 and p = 0.106, respectively). In subgroup analysis based on various clinical characteristics ( Figure 3). RFS and OS were not significantly different between the patients with MAFLD and without MAFLD in the BCLC stage 0 ( Figure 3A), BCLC stage A ( Figure 3B), no liver cirrhosis ( Figure 3C), and liver cirrhosis ( Figure 3D) subgroups.

MAFLD Subgroup Analysis: Lean-MAFLD Is a Risk Factor
NAFLD is increasingly being recognized in non-obese or lean individuals, especially in Asia, and lean individuals with NAFLD may even have poorer outcomes than obese individuals with NAFLD. To explore this issue, we further stratified our HCC-MAFLD cohort into a lean subgroup (BMI < 23 kg/m 2 ) and non-lean subgroup (BMI ≥ 23 kg/m 2 ). Among the 368 patients with MAFLD and HCC, 28 (7.6%) patients were classified as lean-MAFLD, and the other 340 were classified as non-lean-MAFLD. As shown in Figure 4A, patients with lean-MAFLD had significantly poorer RFS compared to patients with nonlean-MAFLD (p = 0.021). In contrast, the rate of death or liver transplantation did not significantly differ be-tween these two groups (p = 0.784, Figure 4B). In the multivariate analysis that included the entire MAFLD cohort of 369 patients, lean-MAFLD was associated with a significantly higher risk of HCC recurrence than non-lean MAFLD (HR, 2.030; 95% CI, 1.117-3.690, p = 0.020), independently of other predictive factors (   CI, 1.117-3.690, p = 0.020), independently of other predictive factors (  The clinicopathological features of the patients with HCC in the lean-MAFLD (BMI < 23 kg/m 2 ) and non-lean-MAFLD (BMI ≥ 23 kg/m 2 ) subgroups are summarized in Table 5. Lean-MAFLD was associated with older age (p = 0.015), diabetes (p < 0.001), lower serum ALT (p = 0.048), and lower BMI (p < 0.001), but not with other characteristics, such as platelet count, serum AFP, microvascular invasion, or histological stage, that were significantly different between patients with and without MAFLD.

Discussion
This study aimed to evaluate the clinical impact of concurrent MAFLD on the prognosis of HCC patients with CHB. In this large multicenter study, we analyzed 812 consecutive patients who were classified into the MAFLD and non-MAFLD groups after curative resection for HBV-related early-stage HCC (BCLC stage 0 or A). The main finding was that MAFLD was not associated with either RFS or OS. Furthermore, we found that the lean-MAFLD subtype was an independent risk factor for RFS among HCC patients with MAFLD. To our knowledge, this is the first study to assess the associations between the new criteria for MAFLD and the outcomes of HBV-related HCC after curative resection.
Due to the increasing global prevalence of metabolic diseases, which may frequently co-exist with other etiologies that contribute to hepatic steatosis, the new concept of concomitant MAFLD and other liver diseases (dual etiology fatty liver disease) was first proposed in 2020 [17,18]. In contrast to NAFLD, for which the criteria are based on 'negative' assumptions, the newly proposed definition of MAFLD is based on 'positive' assumptions and emphasize the contribution of metabolic dysfunction. A recent review summarizing the newest studies that compared the clinical and prognostic characteristics of subjects with NAFLD and MAFLD was reported [25]. In Taiwan, Huang et al. examined and compared the clinical and histologic features of MAFLD versus NAFLD in patients with biopsy-proven hepatic steatosis. They reported that the novel diagnostic criteria for MAFLD include an additional 38.9% of patients with hepatic steatosis and help to better identify patients with a high degree of disease severity for early intervention than the previous NAFLD criteria [26]. A study conducted by Wang et al. in China in 2021 compared patients with MAFLD and HBV-MAFLD in a large biopsy proven cohort [27]. They reported that patients with HBV-MAFLD had similar metabolic features as patients with pure MAFLD, and the presence of HBV infection was associated with a lower grade of steatosis but higher grades of inflammation and fibrosis in MAFLD. Even though several studies have investigated and validated the clinical significance of the MAFLD criteria, huge controversies remain in regard to the change in definition from NAFLD to MAFLD for the description of fatty liver disease; notably, the consequences of this change on recurrence in HCC remain unclear.
In contrast to a previous study, our study was conducted in a HBV-endemic area and is the first evaluation of the impact of 'pathologically-proven' MAFLD on HBV-related HCC after curative resection. Hepatic steatosis can be detected using se-rum biomarkers, imaging techniques or histology; however, pathologic diagnosis remains the gold standard and is more reliable. We evaluated hepatic steatosis based on the pathological assessment of the resected non-tumor tissues, which is more accurate than evaluations of core biopsies, which was used from most published studies [26,27]. In the present study, we observed no significant differences between the RFS and OS of patients with and without MAFLD. This association remained consistent regardless of BCLC stage or the presence or absence of cirrhosis, which further confirms no significant differences among HCC patients with or without MAFLD. This result is similar, but not exactly the same, recently published Korean study [28]. Yoon et al. compared 196 patients with NAFLD and 142 without NAFLD, and concurrent NAFLD was not associated with either RFS and OS after resection in CHB-HCC [28]. In fact, according to the diagnostic criteria for NAFLD, CHB, and NAFLD cannot be concurrently diagnosed. Therefore, this unsynergistic result was unexpected. In fact, assessment of a large population cohort over 25 years revealed an increasing number of patients with HCC develop liver metabolic disorders, including NAFLD or MAFLD [29], and these diseases are emerging as new precancerous conditions in addition to the traditional well-characterized risk factor, viral-induced cirrhosis. Theoretically, we assumed patients with HBV-MAFLD-HCC would have poorer RFS and OS due to the 'double-risk' condition. In contrast, these patients seemed to have favorable RFS and OS compared to patients without MAFLD (Figure 2). Furthermore, the patients in the MAFLD group had a lower frequency of thrombocytopenia, lower serum AFP, a higher proportion of well differentiation tumors, and a lower proportion of microvascular invasion, which are so-called 'positive' factors for better outcomes in HCC. These pathological differences were also noted in the study by Yoon et al. [28], in which well-differentiated tumor stage and microvascular invasion were less frequent in the NAFLD group. Hence, we postulate that MAFLD exerts a protective effect against HCC recurrence in CHB-HCC. The possible reasons and relationships by which MAFLD affects tumor differentiation and vascular invasion remain still unclear. Further experimental studies are needed to explore the underlying mechanisms of HCC development in animal models of NAFLD/MAFLD. We further stratified the patients with MAFLD into lean and non-lean-MAFLD subgroups. Notably, the lean-MAFLD patients had significantly poorer RFS than the non-lean-MAFLD patients (Figure 4). Currently, it is recognized that between 5% and 45% of patients with NAFLD in Asian populations are lean [30]. To the best of our knowledge, this is the first study to indicate the term 'lean-MAFLD' in the HCC study. The pathophysiology of lean MAFLD is puzzling and poorly understood. Lean-NAFLD patients are reported to have better metabolic and histologic profiles than individuals with obese-NAFLD, but may experience accelerated disease progression and poorer out-comes [31,32]. Genetic variations in PNPLA3, transmembrane 6 superfamily 2 (TM6SF2), membrane-bound Oacyltransferase domain-containing 7 (MBOAT7), or other genes associated with hepatic steatosis may possibly explain the relationship between lean NAFLD and the increased future risk of developing severe liver disease or RFS in HCC. In this study, there were no significant differences in any pathologic characteristics, including microvascular invasion or histological stage, between the lean-and non-lean-MAFLD groups; however, the lean-MAFLD group was older than the non-lean-MAFLD group, which was consistent with a recent study published from Taiwan in which the prevalence of lean MAFLD was higher in elder age [33], which may explain the higher risk of HCC recurrence in the lean-MAFLD group. Another possible reason may be that lean-MAFLD (BMI < 23 kg/m 2 ) could be due to cancer-related malnutrition, which could also lead to poorer RFS in the lean-MAFLD group. In addition, the complex interaction among multiple factors including genetic, dietary, sugar intake, enterohepatic circulation, and gut microbiota is likely to modify individual metabolic health status between lean-and non-lean MAFLD patients. However, due to the retrospective design and the limited number of patients with lean-MAFLD in our study cohort, further studies of larger cohorts with comprehensive metabolic profiles, and gut microbiota analysis before and after HCC resection are required to evaluate the characteristics and prognosis of this special population.
This study has several strengths. Firstly, a large sample size of recently treated (from 2010 to 2019) patients was collected, which means the baseline characteristics and prevalence of metabolic disease are close to the rates of disease in the current population. Second, since referral is not required in Taiwan, there was no referral bias in the study. Even though some patients were lost to follow-up, we contacted these patients by phone and/or checked their vital status using the Cancer Screening and Tracing Information Integrated System for Taiwan. Therefore, we could confirm the vital status of each patient. Third, to reduce issues related to the heterogeneity of these diseases in Western populations (including environmental and genetic factors), we recruited a homogenous study population, which may help to draw more precise conclusions for Asian patients. Most importantly, this is the first study to investigate the im-pact of pathologically proven MAFLD on HBV-HCC after curative liver resection, which can provide more accurate steatosis evaluation than those by core biopsies or image.
Our approach has some inherent potential limitations. First of all, this was not a prospective study. However, we believe that the risk of bias was small, because most patients were followed by the same physicians throughout the course of their disease, with clinical and laboratory assessments and HCC screening using ultrasonography every 3-6 months. Second, we only enrolled patients with HBV-related early-stage HCC with or without MAFLD. Whether MAFLD has a similar impact on the prognosis of HCC caused by different etiologies (such as hepatitis C virus) or even advanced HCC remains to be investigated, and our results need to be validated in other cohorts and in Western countries. Third, not all data could be obtained from the electronic medical records, such as the HOMA-IR index, lipid profile, or waist circumference, resulting in the exclusion of some patients from the analysis, especially in the lean-MAFLD group. Future prospective studies of larger numbers of patients with HCC may help clarify this point.

Conclusions
In conclusion, the overall prognosis in terms of HCC recurrence and death or liver transplantation in CHB-related HCC were not significantly different between patients with and without MAFLD. However, concurrent MAFLD showed a mild survival benefit, which may be due to the higher frequency of well-differentiated tumors and lower frequency of microvascular invasion compared to patients without MAFLD. Moreover, patients in the lean-MAFLD group had a higher rate of tumor recurrence than patients with non-lean-MAFLD.

Institutional Review Board Statement:
The clinical data was acquired with the approval and permission of the Institutional Review Board of Kaohsiung Chang Gung Memorial Hospital. The study protocol was approved by the Institutional Review Board of Kaohsiung Chang Gung Memorial Hospital (IRB number: 201701632A3). The requirement for written informed consent was waived by the Institutional Review Board due to the retrospective design of the study and absence of relevant to human biological ethic issues.

Informed Consent Statement: Not applicable.
Data Availability Statement: All analyzed data are included in this published article. The original data are not suitable for publication owing to confidentiality issues, and only available upon reasonable request to the corresponding author.