Hepatocellular carcinoma (HCC) is the sixth most commonly diagnosed cancer worldwide and the third leading cause of cancer mortality.¹ The diagnosis of early-stage HCC has considerably improved with the advancement of surveillance and diagnostic techniques for HCC,² and the treatment of early-stage HCC is determined by an assessment of tumor characteristics, liver function, and the patient's general condition.³ To address the complexity of early-stage HCC management, the American Association for the Study of Liver Diseases (AASLD) has established an algorithm for surgical treatment of early-stage HCC based on tumor burden and liver function.⁴

Historically, surgical resection has been considered the gold standard for patients with a single lesion and preserved liver function.^3,5,6 Although surgical resection has been proven helpful in achieving tumor control and improving overall survival (OS),⁷ the issue remains whether it is superior to other treatments in terms of tumor size, degree of liver dysfunction, and patient performance.^3,5,6 In the recently announced AASLD algorithm for surgical treatment, resection was recommended for patients with a single lesion of 2 to 5 cm, Child-Turcotte-Pugh (CTP) class A, with no clinically significant portal hypertension (CSPH).⁴ We examined the proportion of surgical resection in real-life clinical practice among these patients and assessed whether it resulted in improved survival outcomes compared to alternate treatments.

Liver transplantation (LT) is the treatment of choice for HCC within Milan criteria, if unsuitable for resection.^3,8 Accordingly, the recently proposed AASLD algorithm recommends LT for the following cases: (1) a solitary lesion (up to 5 cm) with impaired liver function (with CSPH or CTP B); (2) multiple lesions (up to 3 nodules, up to 3 cm) independent of liver dysfunction. However, LT may not always be feasible due to high procedure costs and organ shortages. Hepatic resection, ablation, and transarterial therapy were recommended as alternative treatment options for patients with early-stage HCC.^3,5,6 Nevertheless, the optimum modality is constantly conflicting and controversial.^9,10 The purpose of this study is to identify effective alternative treatments for individuals with early-stage HCC, per the newly proposed AASLD algorithm, are recommended for LT but have not received it.

1. Study design and population

This is a single-center, retrospective cohort study based on the prospective HCC registry, which records clinical characteristics, tumor characteristics, and initial treatments of newly diagnosed HCC patients at Samsung Medical Center, Seoul, South Korea. HCC was diagnosed histologically or clinically using dynamic computed tomography and/or magnetic resonance imaging (nodule >1 cm with arterial hypervascularity and portal-/delayed-phase washout), based on the guidelines of the Korean Liver Cancer Association-National Cancer Center.³ We screened a total of 1,473 consecutive treatment-naïve, newly diagnosed, early-stage HCC (single tumor, 2–5 cm or 2–3 lesions, ≤3 cm without vascular, bile duct, lymph node or distant metastasis) patients registered at Samsung Medical Center HCC registry between January 2013 and December 2018 (Fig. 1). Among them, we excluded patients who met the following exclusion criteria: those with a liver function CTP class of C (n=14), an Eastern Cooperative Oncology Group performance status (ECOG PS) of 3–4 (n=5), those who only received best supportive care (n=10), or those with unknown α-fetoprotein (AFP) levels (n = 2). Finally, 1,442 HCC patients with CTP class A or B, ECOG PS of 0–2 and had received HCC-specific treatment were analyzed. This study was conducted after the approval of the Institutional Review Board of Samsung Medical Center (approval number: 2024-024-001). Informed consent was waived for this study as it is a retrospective analysis, and all patient data were anonymized in accordance with institutional and ethical guidelines.

Figure 1. Study population diagram. HCC, hepatocellular carcinoma; CTP, Child-Turcotte-Pugh; ECOG PS, Eastern Cooperative Oncology Group performance status; AFP, α-fetoprotein; CSPH, clinically significant portal hypertension.

2. Variables and study outcomes

We categorized patients into two groups, according to the newly proposed algorithm for surgical treatment by AASLD, based on the number of tumors, CTP class, and presence of CSPH.⁴ The resection recommended group was characterized as having a single lesion, CTP class A, and no CSPH. For these patients, the presence of cirrhosis was assessed through a review of histology (if available) or radiological features. The LT-recommended group was characterized as having multiple tumors (2 to 3 lesions), or a single lesion without preserved liver function (CTP class B or with CSPH). CSPH was evaluated by the presence of any of the following clinical indicators: splenomegaly with thrombocytopenia (platelets <100,000/μL), portosystemic varices (abnormally enlarged veins, detected by upper endoscopy or cross-sectional images) and/or ascites confirmed in imaging studies.⁴

In the LT-recommended group, treatment options were categorized into LT, resection, ablation, and transarterial therapy. Patients who received either LT as an initial treatment or a subsequent therapy following bridging therapy (e.g., transarterial chemoembolization [TACE]) within 3 months of diagnosis were included in the LT group. Radiofrequency ablation, microwave ablation, cryotherapy, laparoscopic radiofrequency ablation, and a combination of radiofrequency ablation and TACE were all included in the ablation group. Transarterial therapy included TACE, drug-eluting bead TACE, and transarterial radioembolization.

Exposure refers to the initial treatment that the patient received. The primary outcome was OS. Patients were monitored from the initial diagnosis day to the last follow-up or death, whichever comes first. The following variables were also collected from the HCC registry: age at diagnosis, sex, ECOG PS, underlying comorbidities (diabetes mellitus, hypertension), etiology of chronic liver disease, CTP score, albumin-bilirubin (ALBI) grade, tumor size, number, platelet count, total bilirubin, aspartate aminotransferase, alanine aminotransferase,⁸ albumin, prothrombin time (international normalized ratio), creatinine, AFP and protein induced by vitamin K antagonist-II (PIVKA-II) levels.

3. Statistical analysis

Values are expressed as median (interquartile range) or number (%). Differences between baseline characteristics were compared using the t-test, Mann-Whitney U test, analysis of variance, chi-square test, or Fisher exact test, as appropriate. Survival was estimated using the Kaplan-Meier method and compared with the log-rank test to analyze differences in OS between groups. For risk analysis, we calculated hazard ratios (HRs) with 95% confidence intervals (CIs) for all-cause mortality using Cox proportional hazard regression models. In the study, the following factors were adjusted: age (≤65 years vs >65 years), sex (male vs female), diabetes (yes vs no), ECOG PS (0 vs ≥1), etiology of chronic liver disease (viral vs non-viral), presence of cirrhosis (yes vs no), ALBI grade (1 vs ≥2), tumor number (single vs multiple), tumor size, AFP levels (≤200 ng/mL vs >200 ng/mL) and PIVKA -II levels (≤100 ng/mL vs >100 ng/mL). The cutoff values of AFP and PIVKA-II were chosen according to published cutoff points.¹¹ Analyses were performed separately for resection-recommended patients and LT-recommended patients. In the analysis of resection recommend patients, patients were categorized into those who received resection and those who received treatments other than resection. In the analysis of LT-recommended patients, risk analyses were performed on subsets of the three most applied treatments (resection, ablation, and transarterial therapy) due to the limited number of patients who received LT or radiation therapy. For the LT-recommended group, additional subset analyses were performed between the single tumor without preserved liver function and the multiple tumor subgroups. In addition, we performed survival analysis and multivariable Cox analysis to evaluate the impact of CSPH on the clinical outcomes of patients who received surgical resection for early-stage HCC. All analyses were performed using R software (4.1.2; R Foundation for Statistical Computing, Vienna, Austria) and SPSS version 27.0 (IBM Corp., Armonk, NY, USA). A two-tailed p<0.05 was considered significant.

1. Surgical resection recommend group (single lesion with CTP class A, without CSPH)

In these patients, 85.8% of patients (679/791 patients) received surgical resection and 14.2% of patients (112/791 patients) did not. Surgical resection was not performed due to several considerations, including insufficient future liver remnant, the presence of an intrahepatic indeterminate nodule, or a high surgical risk due to comorbidities or advanced age (80 years or beyond). They were subjected to alternative treatment modalities, such as LT (n=2), local ablative therapy (n=56), transarterial therapies (n=53), or radiation therapy (n=1).

The comparison of baseline characteristics between patients who received resection and patients who received treatment other than resection are shown in Table 1. Those who received resection were younger, had better ECOG PS, better ALBI grades and had larger tumor sizes.

Table 1 . Baseline Characteristics of Patients with a Single Lesion and Classified as CTP class A without Portal Hypertension (Resection vs Others).

Characteristic	Resection (n=679)	Others (n=112)*	p-value
Age, yr	58 (51–66)	64 (56–72)	<0.001
Male sex	552 (81.3)	89 (79.5)	0.743
ECOG performance status			<0.001
0	439 (97.1)	75 (89.3)
≥1	24 (3.5)	13 (11.6)
Diabetes mellitus	151 (22.2)	34 (30.4)	0.078
Hypertension	248 (36.5)	48 (42.9)	0.239
Etiology			0.034
Viral	548 (80.7)	80 (71.4)
Non-viral	131 (19.3)	32 (28.6)
Cirrhosis	297 (43.7)	46 (41.1)	0.671
ALBI grade			<0.001
1	658 (96.9)	97 (86.6)
2	21 (3.1)	15 (13.4)
Tumor size, cm	3.0 (2.4–3.7)	2.6 (2.2–3.5)	0.005
Tumor size			0.031
2.0–3.0 cm	334 (49.2)	68 (60.7)
3.1–5.0 cm	345 (50.8)	44 (39.3)
Lab
Platelet, ×103/μL	176 (148–208)	165 (133–208)	0.042
Total bilirubin, mg/dL	0.6 (0.5–0.9)	0.7 (0.5–0.9)	0.135
Aspartate transaminase, U/L	27 (22–36)	28 (24–40)	0.102
Alanine transaminase, U/L	27 (19–38)	23 (18–38)	0.361
Albumin, g/dL	4.5 (4.3–4.7)	4.4 (4.1–4.6)	<0.001
Prothrombin time, INR	1.03 (0.99–1.08)	1.05 (1.00–1.10)	0.003
Creatinine, mg/dL	0.90 (0.79–1.00)	0.90 (0.78–1.02)	0.701
α-Fetoprotein, ng/mL	11 (3–107)	6 (3–25)	0.016
≤200	537 (79.1)	97 (86.6)	0.085
>200	142 (20.9)	15 (13.4)
PIVKA-II, mAU/mL†	47 (27–189)	38 (25–187)	0.151
≤100	448 (66.8)	74 (67.3)	>0.999
>100	223 (33.2)	36 (32.7)

Data are presented as median (interquartile range) or number (%)..

CTP, Child-Turcotte-Pugh; ECOG, Eastern Cooperative Oncology Group; ALBI, albumin-bilirubin; INR, international normalized ratio; PIVKA-II, protein induced by vitamin K antagonist-II..

*Others included liver transplantation (n=2), local ablative therapy (n=56), transarterial therapies (n=53), or radiation therapy (n=1); ^†10 Patients had missing data..

The 5-year survival rate of the whole surgical resection recommend group was 86.8% and those who received surgical resection had a significantly better rate than those who received other treatments (89.4% vs 72.3%, p<0.001) (Fig. 2). Those who received other treatments were at a higher risk of mortality compared to those who received resection (adjusted HR [aHR], 2.73; 95% CI, 1.87 to 3.98) (Table 2). The result remained consistent even when assessing individuals who achieved a complete response after therapy independently (Supplementary Table 1, Supplementary Fig. 1A).

Figure 2. Overall survival in patients classified as Child-Turcotte-Pugh class A without clinically significant portal hypertension (resection vs others).

Table 2 . Risk of Mortality by Treatment Modality.

	No. of patient	Survival rate at 5 yr (%)	Unadjusted HR (95% CI)	p-value	Adjusted HR (95% CI)	p-value
Single tumor with preserved liver function*
Resection	679	89.4	Reference		Reference
Others†	110	75.9	2.73 (1.87–3.98)	<0.001	2.42 (1.61–3.64)	<0.001
Single tumor with impaired liver function or multiple tumors‡
Resection	188	79.0	Reference		Reference
Ablation	181	76.0	1.38 (0.94–2.04)	0.109	1.36 (0.88–2.10)	0.184
Transarterial therapy	256	57.8	2.58 (1.82–3.64)	<0.001	2.32 (1.58–3.44)	<0.001

HR, hazard ratio; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; ALBI, albumin-bilirubin; AFP, α-fetoprotein; PIVKA-II, protein induced by vitamin K antagonist-II; CSPH, clinically significant portal hypertension..

*Adjusted for age (≤65 years vs >65 years), sex (male vs female), ECOG PS (0 vs ≥1), diabetes (yes vs no), etiology (viral vs non-viral), cirrhosis (yes vs no), AIBI grade (1 vs ≥2), tumor size (<3 cm vs ≥3 cm), AFP (≤200 ng/mL vs >200 ng/mL), and PIVKA-II (≤100 mAU/mL vs >100 mAU/mL); ^†Others included local ablative therapy, transarterial therapies, or radiation therapy. Two patients who received liver transplantation were not included; ^‡age (≤65 years vs >65 years), sex (male vs female), ECOG PS (0 vs ≥1), diabetes (yes vs no), etiology (viral vs non-viral), CSPH (yes vs no), AIBI grade (1 vs ≥2), number of tumors (single vs multiple), tumor size (<3 cm vs ≥3 cm), AFP (≤200 ng/mL vs >200 ng/mL), and PIVKA -II (≤100 mAU/mL vs >100 mAU/mL);.

2. LT-recommended group (single lesion without preserved liver function or multiple lesions)

Among 651 LT-recommended patients, only 22 patients (3.4%) underwent LT, either as initial treatment or a subsequent therapy following bridging therapy (e.g., TACE), within 3 months of diagnosis. Supplementary Fig. 2 illustrates the discrepancies in real-world treatment from the surgical recommendations provided by AASLD. LT was only administered to 1.7% of patients with a single lesion, CTP class A, and CSPH. All LTs were performed using grafts from living donors. Among alternative treatments other than LT, the most common treatment modality was transarterial therapy (256/651 patients, 39.3%) followed by resection (188/651, 28.9%) and ablation (181/651, 27.8%). The comparison of baseline characteristics of patients between who received LT, surgical resection, ablation or TACE are shown in Table 3. The selection of alternative treatments was determined by considering the degree of liver dysfunction and number of tumors (Supplementary Fig. 3). Surgical resection was the mainstay of treatment for patients with single tumor HCC and CSPH. Regarding the extent of resection, 26 patients (13.8%) underwent wedge resection, 73 patients (38.8%) underwent segmentectomy, 34 patients (18.2%) underwent bi-segmentectomy, and 55 patients (29.2%) underwent resection of three or more segments. In contrast, transarterial therapy was the primary treatment approach for patients who had single HCC and were classified as CTP class B, or for those with multiple HCC and impaired liver function (Supplementary Fig. 3).

Table 3 . Comparison of Characteristics of Patients Recommended for LT (LT vs Resection vs Ablation vs Transarterial Therapy).

Characteristic	LT (n=22)	Resection (n=188)	Ablation (n=181)	Transarterial therapy (n=256)	p-value
Age, yr	56 (53–59)	59 (53–66)	62 (56–69)	62 (56–69)	<0.001
Male sex	19 (86.4)	149 (79.3)	140 (77.4)	201 (78.5)	0.800
ECOG performance status ≥1	0	5 (2.7)	13 (7.2)	24 (9.4)	0.021
Diabetes mellitus	4 (18.2)	51 (27.1)	74 (40.9)	90 (35.2)	0.015
Hypertension	4 (18.2)	53 (28.2)	68 (37.6)	93 (36.3)	0.074
Etiology					0.021
Viral	21 (95.5)	161 (85.6)	137 (75.7)	200 (78.1)
Non-viral	1 (4.5)	27 (14.4)	44 (24.3)	56 (21.9)
Child-Pugh class					<0.001
A	14 (63.6)	183 (97.3)	159 (87.8)	213 (83.2)
B	8 (36.4)	5 (2.7)	22 (12.2)	43 (16.8)
ALBI grade					<0.001
1	9 (40.9)	164 (87.2)	126 (69.6)	141 (55.1)
2	9 (40.9)	24 (12.8)	52 (28.7)	114 (44.5)
3	4 (18.2)	0	3 (1.7)	1 (0.4)
Tumor number					<0.001
1	7 (41.8)	126 (67.0)	68 (37.6)	75 (29.3)
2	10 (45.5)	58 (30.9)	106 (58.6)	117 (45.7)
3	5 (22.7)	4 (2.1)	7 (3.9)	64 (25.0)
Tumor size, cm	2.4 (2.0–3.0)	2.6 (2.2–3.0)	2.0 (1.6–2.3)	2.2 (1.7–2.8)	<0.001
Lab
Platelet, ×103/μL	79 (59–116)	129 (89–167)	104 (70–144)	95 (68–130)	<0.001
Total bilirubin, mg/dL	1.2 (0.6–1.7)	0.7 (0.5–1.1)	0.8 (0.6–1.3)	0.9 (0.6–1.4)	<0.001
Aspartate transaminase, U/L	41 (30–58)	31 (24–44)	37 (28–53)	38 (30–51)	<0.001
Alanine transaminase, U/L	29 (26–45)	27 (19–41)	28 (21–45)	29 (22–42)	0.666
Albumin, g/dL	3.6 (3.0–4.4)	4.4 (4.1–4.7)	4.3 (3.8–4.5)	4.1 (3.6–4.4)	<0.001
Prothrombin time, INR	1.19 (1.10–1.40)	1.06 (1.01–1.11)	1.09 (1.04–1.19)	1.12 (1.05–1.24)	<0.001
Creatinine, mg/dL	0.92 (0.81–0.97)	0.88 (0.77–0.99)	0.88 (0.74–1.04)	0.84 (0.74–0.98)	0.211
α-Fetoprotein, ng/mL	9 (3–60)	10 (4–72)	9 (4–28)	11 (5–38)	0.442
≤200	17 (77.3)	157 (83.5)	164 (90.6)	236 (92.2)	0.009
>200	5 (22.7)	31 (16.5)	17 (9.4)	20 (7.8)
PIVKA-II, mAU/mL*	57 (30–157)	52 (27–145)	25 (20–37)	35 (20–86)	<0.001
≤100	14 (63.6)	127 (67.6)	159 (92.4)	186 (78.2)	<0.001
>100	8 (36.4)	61 (32.5)	13 (7.6)	52 (21.9)

Data are presented as median (interquartile range) or number (%)..

LT, liver transplantation; ECOG, Eastern Cooperative Oncology Group; ALBI, albumin-bilirubin; INR, international normalized ratio; PIVKA-II, protein induced by vitamin K antagonist-II..

*18 Patients had missing data..

When comparing the survival rate of patients who underwent LT to those who underwent alternative treatments, the 5-year survival rate was 81.8% for the LT group and 69.6% for the non-LT group (Fig. 3A). The 5-year survival rates for resection (n=188), ablation (n=181), and transarterial therapy (n=256) in the non-LT group were 79.0 0%, 76.0%, and 57.8%, respectively (Fig. 3B). Given the limited sample size of patients who underwent LT, a multivariable analysis for mortality was conducted on the subset of patients who underwent resection, ablation, and transarterial therapy. After the adjustment of baseline characteristics, which showed significant differences among the three groups (Table 3), the risk of mortality was comparable between patients who underwent ablation and surgical resection (aHR, 1.36; 95% CI, 0.88 to 2.10). However, patients who underwent transarterial therapy showed a significantly higher risk of mortality (aHR, 2.32; 95% CI, 1.58 to 3.44) compared to those who underwent surgical resection (Table 2). Although the difference in OS was somewhat diminished, a consistent finding was observed when comparing the OS of patients who underwent resection or ablation with those who achieved a complete response following transarterial therapy (aHR, 1.66; 95% CI, 1.09 to 2.55) (Supplementary Table 1, Supplementary Fig. 1B).

Figure 3. Overall survival (OS) of patients recommended for liver transplantation (LT). (A) The Kaplan-Meier curve compares OS between patients treated with LT and other treatments. (B) The Kaplan-Meier curve compares OS between patients treated with LT, resection, ablation, and transarterial therapy.

3. Subgroup analysis of LT-recommended group

In patients with a single lesion and impaired liver function, the 5-year survival rates for resection (n=126), ablation (n=68), and TACE (n=75) were significantly different (76.5%, 69.1%, and 57.0%, respectively) (Fig. 4A). In a Cox regression analysis, the TACE group showed a significantly higher risk of mortality compared to the resection group after adjusting for potential covariates (aHR, 3.24; 95% CI, 1.69 to 6.19; p=0.001). The TACE group also showed a higher mortality risk than the ablation group in the unadjusted model; however, this association was not statistically significant in the adjusted model (aHR, 1.62; 95% CI, 0.78 to 3.37; p=0.51). There was no significant difference in mortality risk between the resection and ablation group (aHR, 1.47; 95% CI, 0.69 to 3.13; p=0.96).

Figure 4. Overall survival of patients recommended for liver transplantation by treatment modality (resection vs ablation vs transarterial therapy), stratified by the number of mass (A) single lesion without preserved liver function (B) multiple lesions.

The 5-year survival rates for patients with multiple HCC were 84.3% for resection (n=62), 80.2% for ablation (n=113), and 57.4% for TACE (n=181) (Fig. 4B). In the unadjusted model, the TACE group appeared to have a higher risk of mortality compared to both resection (HR, 2.33; 95% CI, 1.04 to 5.22; p=0.10) and ablation groups (HR, 2.28; 95% CI, 1.22 to 4.28; p=0.03). However, statistical significance was not achieved across the three modalities in the adjusted model (p=0.42 for resection vs TACE and p=0.09 for ablation vs TACE) (Table 4).

Table 4 . Risk of Mortality by Treatment Modality in Patients Recommended for Liver Transplantation, Stratified by the Number of Tumors.

	Unadjusted HR (95% CI)	p-value*	Adjusted HR (95% CI)†	p-value*
Single tumor without preserved liver function
Resection (n=126) vs ablation (n=68)	1.72 (1.04–2.84)	0.111	1.61 (0.86–3.01)	0.329
Resection (n=126) vs transarterial therapy (n=75)	3.92 (1.62–4.23)	<0.001	2.13 (1.22–3.70)	0.027
Ablation (n=68) vs transarterial therapy (n=75)	1.67 (1.01–2.75)	0.134	1.34 (0.76–2.36)	>0.999
Multiple tumors
Resection (n=62) vs ablation (n=113)	1.29 (0.67–2.50)	>0.999	0.75 (0.33–1.71)	>0.999
Resection (n=62) vs transarterial therapy (n=181)	2.97 (1.65–5.35)	0.001	2.13 (1.14–3.97)	0.050
Ablation (n=113) vs transarterial therapy (n=181)	2.30 (1.49–3.54)	0.001	2.16 (1.38–3.39)	0.002

HR, hazard ratio; CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; ALBI, albumin-bilirubin; AFP, α-fetoprotein; PIVKA-II, protein induced by vitamin K antagonist-II..

*Corrected using Bonferroni’s method due to subgroup analysis; ^†Adjusted for age (≤65 years vs >65 years), sex (male vs female), ECOG performance status (0 vs ≥1), diabetes (yes vs no), etiology (viral vs non-viral), ALBI grade (1 vs ≥2), tumor size (<3 cm vs ≥ 3 cm), AFP (≤200 ng/mL vs >200 ng/mL), and PIVKA-II (≤100 mAU/mL vs >100 mAU/mL)..

4. Impact of CSPH on all-cause mortality in patients who underwent surgical resection for early-stage HCC

In patients who underwent surgical resection for early-stage HCC, either CSPH or ≥ALBI grade 2 did not show significantly lower OS compared to those with no CSPH and ALBI grade 1 (OS 86.5% vs 82.7% vs 73.0% at 5-year) (Supplementary Fig. 4). However, the OS at 5 years was significantly lower at 47.4% among the 18 patients who had both CSPH and a grade of 2 or higher on the ALBI scale. In a multivariable Cox analysis, both CSPH and ALBI grade ≥ 2 were independent risk factors for all-cause mortality, with an aHR of 3.24 (95% CI, 1.68 to 6.24), along with age >65 years and diabetes (Supplementary Table 2).

This study evaluated the survival outcome of patients with early-stage HCC based on the recently proposed algorithm for surgical treatment by the AASLD. The majority (85.8%) of patients with single HCC and preserved liver function underwent surgical resection, as recommended by the AASLD algorithm, and had a better OS rate than those who received alternative treatments. However, only 3.4% of patients diagnosed with early HCC underwent LT as their initial treatment, despite being recommend by the AASLD algorithm. Resection, ablation, and transarterial therapy were performed as alternative treatment options for patients who did not receive LT. When comparing survival outcomes of these three modalities, resection and ablation had comparable OS rates, while transarterial therapy had a worse OS. In the subgroup analysis of LT-recommended patients, resection was more effective than transarterial therapy for treating a single lesion, while ablation was more effective than transarterial therapy for treating multiple lesions.

Studies have shown that preserved liver function is related to excellent outcomes after surgical resection,¹² and liver resection has been considered a gold standard for solitary HCC without CSPH, according to both the AASLD algorithm and the Barcelona Clinic Liver Cancer strategy.⁵ We found that about 85% of patients with single HCC sized 2 to 5 cm, and preserved liver function, underwent surgical resection in real-life practice. When patients had an inadequate future liver remnant, an indeterminate intrahepatic nodule, or a high surgical risk due to comorbidities or advanced age (≥80 years old), alternative treatments were performed other than surgical resection, such as ablation, transarterial therapy, radiation therapy, or LT. Surgical resection offered an independent survival benefit compared to other alternative therapies after adjusting for potential confounders. Of note, the 5-year OS for patients who underwent resection was outstanding, up to 90%, which is significantly superior to the previously reported, 60%, for 1,012 patients with Barcelona Clinic Liver Cancer stages 0-A.¹³ Given that the previous study included patients who had CSPH, including esophageal varices, or tumor sizes greater than 5 cm, which were excluded as surgical candidates in the AASLD guideline, this study validated the prognostic significance of surgical resection criteria by applying the new AASLD algorithm to a single tumor with a size range of 2 to 5 cm.

In principle, LT provides an optimal treatment strategy for patients with early HCC and who are ineligible for surgical resection. This is due to its ability to eliminate both the tumor tissue and risk of de novo carcinogenesis arising from a cirrhotic liver. In real-life situation, however, LT was performed in a minority of patients (3.4%) diagnosed with early HCC. One reason for the extremely low likelihood of LT for these patients is the favorable clinical outcomes of early HCC, with an OS rate of more than 70% after 5 years.¹⁴ This study also found a 5-year survival rate of approximately 70% in patients who underwent alternative treatment other than LT in the LT-recommended group. Another factor is the challenges involved in allocating liver grafts from deceased donors to HCC patients, unless they exhibit a Model for End-Stage Liver Disease score exceeding 30, due to the low rate of deceased organ donation in Korea.³

Notably, surgical resection was mainly performed for single HCC and CTP class A with CSPH. Although comparative analysis was limited due to a small sample size of LT patients, the 5-year OS for patients with surgical resection was 79.0%, which was comparable to the 5-year OS rate of 81.8% for LT patients. Notably, surgical resection showed a remarkable clinical outcome for patients with ALBI grade 1 irrespective of CSPH; the 5-year OS was 86.5% and 82.7%, respectively without or with CSPH. This finding suggested that surgical resection could be effective for selected patients recommended for LT, as determined by the AASLD algorithm. As minimally invasive surgical approaches, including laparoscopy and robotic assisted hepatectomy decreased the risk of postoperative complications by minimizing physiologic disruption,¹⁵ the resectability criteria can be extended, particularly for patients with impaired liver function.^16,17 In this study, approximately 80% of patients who had surgical resection as an alternative to LT underwent laparoscopic or minor liver resection. However, current clinical guidelines do not establish a specific cutoff for mild CSPH. Further studies that identify surgical criteria in patients with early HCC and mild CSPH, to guide safe surgical resection with a minimally invasive surgical approach.

A recent meta-analysis of seven randomized controlled trials and 18 matched nonrandomized trials that compared resection and local ablation therapies for HCC within the Milan criteria, reported no significant difference in OS between the two treatments.¹⁸ These findings align with our observation of no difference in OS between resection and ablation in the LT recommendation group, akin to the Milan criteria group. According to a meta-analysis that included the majority of research on single lesions less than 3 cm, surgery was found to be superior to ablation, regarding 5-year OS rates.⁷ However there was no difference in OS or the 5-year OS rate between resection and ablation in two studies of 2–4 cm or 3–5 cm single lesions.^19,20 Ablation is associated with low trauma incidence, few postoperative complications, and a short hospital stay. Most importantly, it is more effective for patients with severe liver cirrhosis or impaired liver function.^7,21 In this study, there was no difference in OS between surgery and ablation in patients with single lesions of 2 to 5 cm. However, considering the compromised liver function and perioperative morbidity and mortality associated with surgery, ablation may be a preferred treatment option, alternative to LT, for patients with early HCC, particularly when the tumor size is small.

Clinically, transarterial therapy is an alternative treatment option for patients unsuitable for resection or ablation with small HCC tumors.²² It is the most common treatment strategy alternative to LT-recommended patients in real-life practice. Several retrospective studies have compared survival outcomes of TACE, resection,^23,24 and ablation.^25,26 However, these studies yielded inconclusive results. In this study, transarterial therapy exhibited notably inferior OS outcomes in the subset recommended for LT, compared to resection or ablation. The disparity of transarterial therapy compared to resection persisted even after adjusting for confounding variables and treatment response. Notably, this finding was particularly pronounced in cases involving single lesions, as revealed in subgroup analysis. In cases that involve multiple small lesions, it was observed that transarterial therapy led to a notable decrease in OS when compared to ablation.

Our study has several limitations that should be considered. Firstly, as this study was conducted retrospectively at a single tertiary center, potential selection and measurement bias may have influenced our study results. Furthermore, we integrated various treatment such as ablation, transarterial therapy, radiation therapy, and combination therapy other than resection into the alternative treatment group for patients with single lesions and intact liver function due to the limited sample size of each treatment. Additionally, the better outcomes of resection may be due to a careful selection of patients. Similarly, baseline characteristics, according to treatment modalities including resection, ablation, and transarterial therapy, were significantly different in LT-recommended patients. Although potential biases were adjusted, the better survival outcomes of resection compared to transarterial therapy could be due to patient selection rather than the modality itself. Randomized controlled trials will be required to address these limitations, but this will be challenging due to the current treatment recommendations and ethical concerns. Lastly, our findings should be cautiously generalized to different centers with different environments in terms of surgical techniques and postoperative care. Despite such limitations, this study demonstrated the clinical relevance and efficacy of a newly proposed AASLD algorithm for early-stage HCC treatment decisions. It also provides information for alternative options in patients who cannot adhere to the AASLD algorithm.

Surgical treatment in accordance with the recently proposed AASLD algorithm had better survival outcomes than alternative treatment approaches. Nevertheless, performing LT in real-world practice is extremely difficult, particularly in early-stage HCC. Resection or ablation could be effective for selected patients recommended for LT under careful consideration of patient features, tumor characteristics, and liver function.

No potential conflict of interest relevant to this article was reported.

Study concept and design: D.H.S., A.K., M.J.G. Data acquisition: A.K. Data analysis and interpretation: A.K., M.J.G., B.G.S. Drafting of the manuscript: A.K., D.H.S., M.J.G. Critical revision of the manuscript for important intellectual content: M.J.G., D.H.S., B.G.S., W.K., G.Y.G., Y.H.P., M.S.C., J.H.L. Statistical analysis: A.K., M.J.G. Administrative, technical, or material support; Y.H.P., M.S.C. study supervision: J.H.L., D.H.S. Approval of final manuscript: all authors.

Supplementary materials can be accessed at https://doi.org/10.5009/gnl240214.