Prognostic impact of neoadjuvant therapy vs upfront surgery in patients with huge hepatocellular carcinoma undergoing curative resection

Abstract

BACKGROUND

Huge hepatocellular carcinoma (HCC) is associated with a high postoperative recurrence rate and poor long-term survival. The role of neoadjuvant therapy in patients with huge HCC eligible for curative resection remains uncertain.

AIM

To compare long-term outcomes after neoadjuvant therapy vs upfront surgery in huge HCC.

METHODS

This single-center retrospective surgical cohort study included patients with huge HCC who underwent curative liver resection from January 2019 to November 2024. Huge HCC was defined as a maximum tumor diameter ≥ 10 cm on pre-treatment enhanced computed tomography or magnetic resonance imaging. Patients were divided into neoadjuvant therapy and upfront surgery groups. The primary endpoint was recurrence-free survival (RFS), with overall survival (OS) and surgery-related outcomes as secondary endpoints.

RESULTS

A total of 103 patients were included, with 56 in the neoadjuvant therapy group and 47 in the upfront surgery group. After a median follow-up of 30 months, the RFS and OS rates in the neoadjuvant therapy group were superior to those in the upfront surgery group. The 1-year and 3-year RFS rates were 76.8% vs 55.3% and 59.2% vs 41.5%, respectively. The corresponding 1-year and 3-year OS rates were 98.2% vs 87.2% and 88.9% vs 71.7%, respectively. Multivariable analysis identified upfront surgery and lack of postoperative adjuvant therapy as independent risk factors for poorer RFS and OS. Surgery-related outcomes were comparable between groups.

CONCLUSION

In this retrospective surgical cohort study, neoadjuvant therapy was associated with improved survival without significantly increasing perioperative burden in patients with huge HCC.

Key Words: Hepatocellular carcinoma; Huge hepatocellular carcinoma; Neoadjuvant therapy; Liver resection; Recurrence-free survival

Core Tip: Huge hepatocellular carcinoma (HCC) is a particularly aggressive form of liver cancer, and surgery alone often fails to prevent early recurrence. Whether treatment before surgery can improve outcomes in these patients remains unclear. In this retrospective study, we compared patients with huge HCC who underwent surgery after neoadjuvant therapy with those who received upfront surgery. Neoadjuvant therapy was associated with better recurrence-free and overall survival, without significantly increasing operative time, blood loss, transfusion, or postoperative hospital stay. The benefit appeared to be greater in patients with aggressive disease features such as vascular invasion, portal vein tumor thrombus, and Barcelona Clinic Liver Cancer stage C disease. These findings suggest that multimodal neoadjuvant therapy may be a useful perioperative strategy for selected patients with huge HCC and warrant further prospective validation.

INTRODUCTION

Hepatocellular carcinoma (HCC) accounts for approximately 80% of all primary liver cancers and is the third leading cause of cancer-related mortality worldwide[1]. For early HCC patients with adequate hepatic functional reserve, liver resection (LR) remains the standard treatment; however, only about 15% of HCC cases are diagnosed at a stage suitable for curative intervention, including LR or liver transplantation[2,3]. Even among patients undergoing surgical resection, the risk of recurrence within 5 years can be as high as 80%[4], with a five-year overall survival (OS) rate ranging from 33.6% to 76.4%, highlighting the limited long-term efficacy of surgery alone[5,6]. These observations underscore the clinical importance of exploring comprehensive perioperative treatment strategies to improve outcomes beyond curative resection.

Huge HCC is defined as a tumor with a maximum diameter exceeding 10 cm, representing a particularly challenging subgroup. These patients often exhibit high-risk pathological features such as vascular invasion, satellite nodules, and poor differentiation, resulting in higher postoperative recurrence rates and poorer long-term prognosis. Consequently, even with successful curative resection, the long-term survival benefit for patients with huge HCC remains lower than that for patients with smaller tumors[7]. Previous studies comparing giant and non-giant HCC after resection have shown that, although perioperative safety is acceptable, huge HCC is associated with poorer long-term oncological outcomes, supporting the need for effective perioperative strategies in this high-risk population[8]. For this high-risk group, reliance on surgery alone is often insufficient to offset the recurrence risk associated with aggressive tumor biology, highlighting the necessity for effective perioperative interventions.

Neoadjuvant therapy has been widely adopted in other malignancies, including breast cancer, colorectal cancer, and esophageal cancer, to reduce tumor burden and improve prognosis in patients at high risk of recurrence[9-11]. Its core advantages include: Reducing tumor burden and inducing necrosis of the primary tumor with good blood supply, eliminating existing micro-metastases, increasing the likelihood of achieving R0 resection, and ultimately improving long-term prognosis. Furthermore, assessing the tumor’s response to neoadjuvant therapy provides valuable in vivo information regarding treatment sensitivity, which can guide the choice of postoperative adjuvant therapy[12]. These theoretical advantages have prompted the gradual adoption of neoadjuvant therapy in HCC, with the aim of offering survival benefits superior to upfront surgery for high-risk patients.

The potential benefits of neoadjuvant therapy are particularly pronounced in huge HCC. Appropriate preoperative treatment can not only improve long-term prognosis but also facilitate surgical resection. Chiang et al[13] reported that a novel hepatic arterial infusion chemotherapy (HAIC) regimen used as neoadjuvant therapy significantly reduced early postoperative recurrence in patients with huge HCC, extending the median disease-free survival from 10 months to 41 months (P = 0.016) and OS from 46 months to 96 months, with an objective response rate (ORR) of 66.7%. Conversely, Na et al[14] found that transarterial chemoembolization (TACE) decreased the incidence of microvascular invasion but did not improve recurrence-free survival (RFS) or OS, suggesting that upfront surgical resection may be preferable for some patients. Taken together, the evidence regarding the role of neoadjuvant therapy in huge HCC remains limited and controversial.

Most previous studies on neoadjuvant therapy for HCC included heterogeneous populations with varying tumor sizes and stages[15], and few specifically analyzed patients with huge HCC. Moreover, many of these studies were single-arm studies lacking direct comparisons with cohorts undergoing upfront surgery at the same time[16]. As a result, it remains unclear whether patients with huge HCC can derive significant benefits from neoadjuvant therapy, the magnitude of such benefits, and which patients are ideal candidates. To address this knowledge gap, the present study retrospectively compared the long-term outcomes of patients with huge HCC who underwent curative resection after neoadjuvant therapy with those who underwent upfront curative resection, aiming to assess the potential clinical value of neoadjuvant therapy in this high-risk population.

MATERIALS AND METHODS

Patient selection

This single-center retrospective surgical cohort study included patients with huge HCC who underwent curative LR at the Cancer Hospital of the Chinese Academy of Medical Sciences from January 2019 to November 2024. The aim of this study was to compare the postoperative long-term outcomes between patients who underwent R0 resection after neoadjuvant therapy and those who underwent upfront R0 resection. Therefore, only patients who eventually completed curative LR were included in the final analysis.

Inclusion criteria: (1) Age ≥ 18 years; (2) Histopathologically confirmed HCC in the resected specimen; (3) Maximum tumor diameter ≥ 10 cm on enhanced computed tomography (CT) or magnetic resonance imaging (MRI) before neoadjuvant therapy or upfront surgery; (4) Child-Pugh class A liver function; and (5) Curative LR (R0) was defined as complete macroscopic removal of the tumor with histologically negative resection margins.

Exclusion criteria: (1) Maximum tumor diameter < 10 cm; (2) Concomitant malignancies in other organs; and (3) Incomplete follow-up data.

Patients who received neoadjuvant therapy but were unable to undergo surgery due to tumor progression, deterioration of liver function, treatment-related toxicity, or other contraindications were not included in this surgical cohort. Accordingly, the present study should not be interpreted as an intention-to-treat analysis for all patients initially considered for neoadjuvant therapy.

This study was approved by the Ethics Committee of the National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College (Approval No. NCC2025C-401). As this is a retrospective analysis and all identifiable patient information was anonymized, the Ethics Committee waived the requirement for written informed consent. The study was conducted in accordance with the ethical standards of the Declaration of Helsinki.

Baseline characteristics

Baseline data collected included age, sex, hypertension, diabetes, American Society of Anesthesiologists classification, hepatitis B virus (HBV) and hepatitis C virus status, liver cirrhosis, number of tumors, vascular invasion [radiologically assessed macrovascular invasion, including portal vein tumor thrombosis (PVTT) and hepatic vein tumor thrombus (HVTT)], PVTT and HVTT grading, Barcelona Clinic Liver Cancer (BCLC) staging, hepatitis B e antigen status, preoperative alpha-fetoprotein levels, anatomical LR, extensive LR, surgical approach, and postoperative adjuvant therapy. Liver cirrhosis was confirmed histologically. The severity of PVTT was classified according to the Japanese “VP” system: VP0, no tumor thrombus in the portal vein; VP1, thrombus limited to distal secondary branches; VP2, thrombus involving secondary branches; VP3, thrombus involving primary branches (right or left main trunk)[17]. Anatomical resection ranged from standard partial LR to extended hemihepatectomy[18], while extensive LR was defined as resection of three or more hepatic segments[19].

Procedures

Patients receiving neoadjuvant therapy underwent targeted therapy, immunotherapy, interventional therapy, radiotherapy, or multimodal combination therapy. Targeted agents included lenvatinib, donafenib, sorafenib, and bevacizumab; immunotherapy included sintilimab, toripalimab, tislelizumab, and camrelizumab; interventional therapy included TACE, HAIC, and transcatheter arterial embolization. Radiotherapy, when used, employed volumetric modulated arc therapy, which was determined by a multidisciplinary team (MDT) based on tumor location, liver function, and risks to adjacent organs (Supplementary Table 1). Adverse events were recorded during neoadjuvant therapy and graded according to the Common Terminology Criteria for Adverse Events[20]. Special attention was given to grade 3-4 adverse events, immune-related adverse events, liver function injury, and treatment interruptions or delays. Treatment regimens were individualized by the MDT based on tumor burden, vascular invasion, liver function, and overall patient condition. Surgical arrangements were made at least three weeks after the last administration of PD-1 inhibitors, one week after small-molecule targeted therapy, 6 weeks after large-molecule targeted therapy, four weeks after radiotherapy, or two weeks after interventional therapy.

Surgical resection was performed by experienced hepatobiliary surgeons. Resectability was determined by the MDT based on future liver remnant, extent of vascular involvement, technical feasibility of achieving R0 resection, and overall patient condition. Some patients underwent anatomical resection using standard techniques to ensure negative margins, while others underwent non-anatomical resection due to tumor involvement in multiple liver segments. For patients with PVTT or HVTT, tumor thrombus resection was performed, and affected vascular walls were resected and reconstructed as necessary to achieve R0 resection. R0 resection is defined as having histologically negative margins. Intermittent hepatic blood flow occlusion was performed using the Pringle maneuver (15 minutes of clamping, 5 minutes of release). Postoperatively, patients underwent routine blood tests, liver and kidney function assessments, coagulation function analysis, and pathological examinations. Postoperative adjuvant therapy mainly included targeted therapy, immunotherapy, interventional therapy, or their combinations, determined by the MDT according to postoperative pathological risk factors, liver function, and recurrence risk. Pathologic complete response (pCR) is defined as the complete absence of viable tumor cells, while major pathologic response (MPR) is defined as ≤ 10% viable tumor cells[21]. Pathological response was assessed on hematoxylin-eosin-stained sections of the resected tumor specimens. The percentage of residual viable tumor was estimated as the proportion of the area of viable tumor to the total area of the treated tumor bed, including viable tumor, necrosis, fibrosis, hemorrhage, and inflammatory changes. The tumor necrosis/regression ratio was calculated as the proportion of the area of non-viable tumor to the total area of the treated tumor bed. Pathological assessment was performed independently by two experienced pathologists who were blinded to survival outcomes. Any discrepancies were resolved through joint review and consensus.

Follow-up

The primary endpoint was RFS, while secondary endpoints include OS and surgical outcomes (intraoperative bleeding, duration of surgery, intraoperative transfusion, and postoperative length of stay). For the upfront surgery group, the start of treatment was defined as the date of surgery; for the neoadjuvant therapy group, the start of treatment is defined as the date of initiation of neoadjuvant therapy. RFS was defined as the interval from the start of treatment to the first radiological or pathological evidence of tumor recurrence; OS was defined as the interval from the start of treatment to death from any cause or the last follow-up. RFS and OS were calculated from the start of treatment to reflect the entire perioperative treatment strategy.

Patients in the neoadjuvant therapy group were followed up monthly after completing neoadjuvant therapy. All patients, including those in the neoadjuvant therapy group and the upfront surgery group, underwent enhanced CT or MRI one month postoperatively, along with laboratory tests including complete blood count, liver and kidney function, coagulation function, and tumor markers. Subsequent follow-ups were conducted every three months for the first two years, and then every six months until death or the end date of the study (November 30, 2025).

The tumor response was assessed using the modified RECIST criteria[22], with responses classified as complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD). ORR was defined as the proportion of patients achieving CR or PR, while the disease control rate (DCR) was defined as the proportion of patients achieving CR, PR, or SD. Imaging assessments were independently reviewed by two experienced readers, including a radiologist and a surgeon.

Statistical analysis

All statistical analyses were performed using SPSS version 29.0 (IBM Corp., Chicago, IL, United States) and R version 4.5.0 (Institute for Statistics and Mathematics, Vienna, Austria). Two-sided tests were used, and P < 0.05 was considered statistically significant. Categorical variables were expressed as n (%) and compared using Pearson’s χ² test or Fisher’s exact test. Continuous variables were presented as mean ± SD or median (range), and intergroup comparisons were performed using independent-samples t test or Wilcoxon rank-sum test. Kaplan-Meier analysis was used to estimate RFS and OS, with intergroup differences assessed using the log-rank test. Univariable and multivariable Cox regression analyses were performed to identify prognostic factors associated with RFS and OS, with variables showing P < 0.05 in univariable analysis or deemed clinically relevant included in the multivariable model.

RESULTS

Baseline characteristics

Between January 2019 and November 2024, a total of 129 patients with a preoperative imaging-documented maximum tumor diameter of ≥ 10 cm were screened and subsequently underwent curative LR at the Department of Hepatobiliary Surgery, Cancer Hospital, Chinese Academy of Medical Sciences. Among them, 26 patients were excluded for not meeting the inclusion criteria (5 patients had concomitant malignancies in other organs, and 21 patients had incomplete data). Ultimately, 103 patients were included in the study, 47 patients in the upfront surgery group and 56 patients in the neoadjuvant therapy group (Figure 1). In the neoadjuvant therapy group, the treatment regimens were as follows: 12 patients received only interventional therapy, 7 patients received combined targeted therapy and immunotherapy, 14 patients received a combination of targeted therapy, immunotherapy, and interventional therapy, 3 patients received a combination of targeted therapy, immunotherapy and radiotherapy, 9 patients received a combination of targeted therapy and radiotherapy, 8 patients received a combination of targeted therapy and interventional therapy, and 3 patients received a combination of immunotherapy and radiotherapy. The baseline characteristics of all patients are summarized in Table 1. In the overall cohort, the proportions of patients undergoing anatomical LR (57.4% vs 62.5%, P = 0.602), extensive LR (59.6% vs 57.1%, P = 0.803), and laparoscopic LR (29.8% vs 26.8%, P = 0.736) were comparable between the upfront surgery group and the neoadjuvant therapy group.

Figure 1 Selection of the study population. HCC: Hepatocellular carcinoma.

Table 1 Patient clinical characteristics, n (%).

Variable	Neoadjuvant therapy (n = 56)	Upfront surgery (n = 47)	P value
Age (year)			0.243
< 60	35 (62.5)	24 (51.1)
≥ 60	21 (37.5)	23 (48.9)
Sex
Male	47 (83.9)	41 (87.2)	0.636
Female	9 (16.1)	6 (12.8)
Hypertension			0.588
Absent	39 (69.6)	35 (74.5)
Present	17 (30.4)	12 (25.5)
Diabetes
Absent	50 (89.3)	44 (93.6)
Present	6 (10.7)	3 (6.4)
ASA			0.195
I	35 (62.5)	35 (75.5)
II	21 (37.5)	12 (25.5)
HBV			0.232
Absent	7 (12.5)	10 (21.3)
Present	49 (87.5)	37 (78.7)
HCV			0.238
Absent	52 (92.9)	46 (97.9)
Present	4 (7.1)	1 (2.1)
Hepatocirrhosis			0.662
Absent	10 (17.9)	10 (21.3)
Present	46 (82.1)	37 (78.7)
Tumor number			0.232
Multiple	11 (19.6)	14 (29.8)
Single	45 (80.4)	33 (70.2)
Vascular invasion			0.192
Absent	39 (69.6)	38 (80.9)
Present	17 (30.4)	9 (19.1)
PVTT			0.056
VP0	40 (71.4)	41 (87.1)
VP1	1 (1.8)	0 (0)
VP2	3 (5.4)	4 (8.5)
VP3	12 (21.4)	2 (4.3)
HVTT			0.823
Absent	48 (85.7)	41 (87.2)
Present	8 (14.3)	6 (12.8)
BCLC stage			0.232
A	31 (55.4)	28 (59.6)
B	7 (12.5)	10 (21.3)
C	18 (32.1)	9 (19.1)
HBeAg			0.703
Negative	48 (85.7)	39 (83)
Positive	8 (14.3)	8 (17)
AFP (ng/mL)			0.064
≤ 400	28 (50)	32 (68.1)
> 400	28 (50)	15 (31.9)
Anatomical resection			0.602
Absent	21 (37.5)	20 (42.6)
Present	35 (62.5)	27 (57.4)
Major liver resection			0.803
Absent	24 (42.9)	19 (40.4)
Present	32 (57.1)	28 (59.6)
Surgical method			0.736
Laparoscopy	15 (26.8)	14 (29.8)
Laparotomy	41 (73.2)	33 (70.2)
Adjuvant therapy			0.361
Absent	24 (42.9)	16 (34)
Present	32 (57.1)	31 (66)

ASA: American society of Aneshesiologists; HBV: Hepatitis B virus; HCV: Hepatitis C virus; PVTT: Portal vein tumor thrombosis; HVTT: Hepatic vein tumor thrombus; BCLC: Barcelona Clinic Liver Cancer; HBeAg: Hepatitis B e antigen; AFP: Alpha-fetoprotein.

RFS and OS

The median follow-up time for the entire cohort was 30 months (range: 1-84 months). Compared to the upfront surgery group, patients in the neoadjuvant therapy group had significantly prolonged RFS and OS (Figure 2A and B; RFS: P = 0.012, OS: P = 0.045). The median RFS for the upfront surgery group was 20 months (95%CI: 4.703-35.297), whereas the median RFS for the neoadjuvant therapy group had not yet been reached. The median OS for both groups had not been reached, but the survival curves were clearly separated. In the neoadjuvant therapy group, the 1-year and 3-year RFS rates were 76.8% and 59.2%, respectively, while the upfront surgery group had rates of 55.3% and 41.5%. The corresponding 1-year and 3-year OS rates in the neoadjuvant therapy group were 98.2% and 88.9%, while the upfront surgery group had rates of 87.2% and 71.7%. Among patients in the neoadjuvant therapy group, those achieving MPR/pCR exhibited superior RFS compared to those who did not achieve MPR/pCR (median not reached vs 24 months, P = 0.046), whereas no significant difference in OS was observed (P = 0.344), likely due to the limited sample size and insufficient death events (Figure 2C and D).

Figure 2 Kaplan-Meier curves of survival outcomes in patients with huge hepatocellular carcinoma. A and B: Recurrence-free survival (RFS) and overall survival (OS) in the neoadjuvant treatment group vs the upfront surgery group; C and D: RFS and OS in patients with and without major pathologic response/pathologic complete response after neoadjuvant therapy. RFS: Recurrence-free survival; OS: Overall survival; MPR: Major pathologic response; pCR: Pathologic complete response after neoadjuvant therapy.

Surgery-related and tumor response

No significant increase in perioperative burden was observed after neoadjuvant therapy. There were no significant differences between the neoadjuvant therapy group and the upfront surgery group in terms of intraoperative bleeding, surgical time, intraoperative transfusion rate, and postoperative hospital stay (Table 2). Among the 56 patients in the neoadjuvant therapy group, the tumor responses were as follows: 7 achieved CR, 37 achieved PR, 11 had SD, and 1 had PD. The ORR was 78.6%, and the DCR was 98.2% (Table 3). Adverse events related to treatment were recorded during neoadjuvant therapy and summarized in Supplementary Table 2.

Table 2 Intraoperative bleeding and postoperative hospital stay of patients, median (interquartile range).

Variable	Upfront surgery (n = 47)	Neoadjuvant therapy (n = 56)	P value
Blood loss (mL)	521.00 (200.00, 600.00)	421.00 (100.00, 600.00)	0.894
Operation time (minute)	258.00 (198.00, 311.00)	258.00 (191.50, 300.00)	0.625
Postoperative hospitalization time (day)	12.00 (7.00, 15.00)	9.00 (6.00, 11.00)	0.052
Intraoperative blood transfusion, n (%)	17 (36.2)	23 (41.1)	0.611

Table 3 Tumor response.

mRECIST (n = 56)	n (%)
CR	7 (12.5)
PR	37 (66.1)
SD	11 (19.6)
PD	1 (1.8)
ORR	44 (78.6)
DCR	55 (98.2)

CR: Complete response; DCR: Disease control rate; ORR: Objective response rate; PD: Progressive disease; PR: Partial response; SD: Stable disease; mRECIST: Modified RECIST.

Univariable and multivariable analyses of OS and RFS

In the univariable Cox regression analysis, the lack of postoperative adjuvant therapy was significantly associated with poorer OS (HR = 2.51; 95%CI: 1.081-5.926; P = 0.032). For RFS, the absence of hypertension (HR = 2.248; 95%CI: 1.092-4.627; P = 0.028), multiple tumors (HR = 2.157; 95%CI: 1.205-3.860; P = 0.010), and lack of postoperative adjuvant therapy (HR = 1.891; 95%CI: 1.084-3.299; P = 0.025) were significantly associated with poorer RFS. In the multivariable Cox regression analysis, upfront surgery (HR = 2.860; 95%CI: 1.178-6.943; P = 0.020) and absence of postoperative adjuvant therapy (HR = 3.406; 95%CI: 1.400-8.284; P = 0.007) were independent risk factors for OS. Upfront surgery (HR = 2.167; 95%CI: 1.208-3.888; P = 0.010), absence of hypertension (HR = 2.163; 95%CI: 1.047-4.471; P = 0.037), multiple tumors (HR = 2.838; 95%CI: 1.549-5.200; P = 0.001), and lack of postoperative adjuvant therapy (HR = 2.540; 95%CI: 1.407-4.583; P = 0.002) were independent risk factors for RFS (Tables 4 and 5).

Table 4 Univariate and multivariate analyses of risk factors for recurrence-free survival.

Variable	Univariate		Multivariate
	HR (95%CI)	P value	HR (95%CI)	P value
Treatment group (surgery vs NT)	1.577 (0.904-2.752)	0.109	2.167 (1.208-3.888)	0.010
Age (< 60 years vs ≥ 60 years)	1.072 (0.611-1.880)	0.808
Sex (male vs female)	1.335 (0.569-3.314)	0.506
Hypertension (absence vs presence)	2.248 (1.092-4.627)	0.028	2.163 (1.047-4.471)	0.037
Diabetes (absence vs presence)	1.007 (0.362-2.799)	0.989
ASA (I vs II)	1.720 (0.898-3.292)	0.102
Hepatitis B (absence vs presence)	0.558 (0.238-1.311)	0.181
Hepatitis C (absence vs presence)	3.289 (0.454-23.827)	0.239
Hepatocirrhosis (absence vs presence)	0.442 (0.188-1.039)	0.061
Tumor number (multiple vs single)	2.157 (1.205-3.860)	0.010	2.838 (1.549-5.200)	0.001
Vascular invasion (absence vs presence)	1.434 (0.717-2.867)	0.308
PVTT		0.114
0 vs 3	2.257 (0.810-6.289)
1 vs 3	11.709 (1.258-109.010)
2 vs 3	1.148 (0.210-6.269)
HVTT (absence vs presence)	1.152 (0.491-2.703)	0.746
BCLC stage		0.581
A vs C	1.228 (0.592-2.547)
B vs C	2.844 (1.256-6.439)
HBeAg (absence vs presence)	0.648 (0.324-1.298)	0.221
AFP group (< 400 ng/mL vs ≥ 400 ng/mL)	0.633 (0.381-1.156)	0.147
Anatomical resection (no vs yes)	0.791 (0.444-1.409)	0.426
Extensive liver resection (no vs yes)	0.600 (0.334-1.078)	0.088
Surgical method (scope vs octomy)	1.382 (0.760-2.512)	0.289
Adjuvant therapy (absence vs presence)	1.891(1.084-3.299)	0.025	2.540 (1.407-4.583)	0.002

NT: Neoadjuvant therapy; ASA: American society of Aneshesiologists; PVTT: Portal vein tumor thrombosis; HVTT: Hepatic vein tumor thrombus; BCLC: Barcelona Clinic Liver Cancer; HBeAg: Hepatitis B e antigen; AFP: Alpha-fetoprotein.

Table 5 Univariate and multivariate analyses of risk factors for overall survival.

Variable	Univariate		Multivariate
	HR (95%CI)	P value	HR (95%CI)	P value
Treatment group (surgery vs NT)	2.334 (0.994-5.480)	0.052	2.860 (1.178-6.943)	0.020
Age (< 60 years vs ≥ 60 years)	0.794 (0.344-1.834)	0.589
Sex (male vs female)	0.946 (0.278-3.214)	0.929
Hypertension (absence vs presence)	1.334 (0.492-3.617)	0.571
Diabetes (absence vs presence)	2.466 (0.331-18.355)	0.378
ASA (I vs II)	1.002 (0.408-2.459)	0.997
Hepatitis B (absence vs presence)	0.258 (0.035-1.921)	0.186
Hepatitis C (absence vs presence)	21.789 (0.005-99916.054)	0.474
Hepatocirrhosis (absence vs presence)	0.392 (0.092-1.680)	0.207
Tumor number (multiple vs single)	1.823 (0.739-4.500)	0.193	1.866 (0.735-4.736)	0.190
Vascular invasion (absence vs presence)	0.910 (0.356-2.326)	0.843	0.828 (0.320-2.143)	0.697
PVTT		0.945
0 vs 3	1.002 (0.295-3.410)
1 vs 3	0.000 (0.000-)
2 vs 3	0.535 (0.055-5.159)
HVTT (absence vs presence)	0.713 (0.241-2.110)	0.541
BCLC stage		0.437
A vs C	0.836 (0.309-2.263)
B vs C	1.681 (0.509-5.558)
HBeAg (absence vs presence)	1.205 (0.356-4.074)	0.765
AFP group (< 400 ng/mL vs ≥ 400 ng/mL)	1.015 (0.434-2.376)	0.973
Anatomical resection (no vs yes)	0.991 (0.415-2.370)	0.984
Extensive liver resection (no vs yes)	0.600 (0.233-1.543)	0.289
Surgical method (scope vs octomy)	1.162 (0.451-2.992)	0.756
Adjuvant therapy (absence vs presence)	2.531 (1.081-5.926)	0.032	3.406 (1.400-8.284)	0.007

NT: Neoadjuvant therapy; ASA: American society of Aneshesiologists; PVTT: Portal vein tumor thrombosis; HVTT: Hepatic vein tumor thrombus; BCLC: Barcelona Clinic Liver Cancer; HBeAg: Hepatitis B e antigen; AFP: Alpha-fetoprotein.

Subgroup analyses

Exploratory subgroup analysis suggested that the association between neoadjuvant therapy and improved survival was more pronounced in selected high-risk subgroups, including patients with vascular invasion, PVTT-positive disease, HVTT-positive disease, and BCLC-C stage. Treatment-by-subgroup interaction P values were calculated for both RFS and OS and are presented in Figure 3.

Figure 3 Forest plot of subgroup analysis by baseline characteristics for recurrence-free survival and overall survival. PVTT: Portal vein tumor thrombosis; HVTT: Hepatic vein tumor thrombus; BCLC: Barcelona Clinic Liver Cancer; HBeAg: Hepatitis B e antigen; AFP: Alpha-fetoprotein.

Postoperative complications

Further assessment of postoperative safety outcomes included Clavien-Dindo grade III or higher complications, postoperative liver failure, bile leakage, fever, and 90-day mortality. No significant differences were observed between the neoadjuvant therapy group and the upfront surgery group for these postoperative safety endpoints (Table 6).

Table 6 Postoperative morbidity and mortality after liver resection, n (%).

Variable	Neoadjuvant therapy (n = 56)	Upfront surgery (n = 47)	P value
Clavien-Dindo grade ≥ III	2 (3.6)	3 (6.4)	0.658
Post-hepatectomy liver failure	0 (0)	0 (0)	-1
Postoperative hemorrhage	5 (8.9)	6 (12.8)	0.543
Bile leakage	4 (7.1)	4 (8.5)	1.000
Fever	20 (35.7)	18 (38.3)	0.839
Pleural effusion	3 (5.4)	4 (8.5)	0.699
Ascites	8 (14.3)	6 (12.8)	1.000
Reoperation	0 (0)	0 (0)	-1
Hypoproteinemia	50 (89.3)	41 (87.2)	0.767
90-day mortality	0 (0)	2 (4.3)	0.206

¹Not calculated because no events occurred in either group.

DISCUSSION

This study indicates that among patients with huge HCC who ultimately underwent curative resection, those in the neoadjuvant therapy group had longer RFS and OS compared to patients in the upfront surgery group. At a median follow-up of 30 months, the RFS in the neoadjuvant therapy group was significantly better than that in the upfront surgery group (median not reached vs 20 months, P = 0.012) and OS (P = 0.045). Multivariable analysis confirmed that upfront surgery was an independent risk factor for poorer OS and RFS. These results indicate that preoperative therapy may help improve long-term prognosis in patients with huge HCC and suggest that this high-risk population may still benefit from preoperative intervention[23].

The survival benefits associated with neoadjuvant therapy in this study may be attributed to several mechanisms. First, preoperative treatment can reduce tumor activity and induce necrosis while the blood supply to the primary tumor remains intact, thereby decreasing the burden of residual tumor and potentially lowering the risk of dissemination during surgery and early postoperative recurrence[24]. Second, the treatment response itself may serve as a “biological selection” mechanism: Patients who achieve objective or MPR (MPR/pCR) preoperatively typically indicate that the tumor is more sensitive to systemic therapy, with lower residual malignant biological activity, thus reducing the risk of postoperative recurrence[25]. Finally, there may be synergistic effects among targeted therapy, immunotherapy, and local therapy, collectively enhancing tumor necrosis, improving the local immune microenvironment, and inhibiting the progression of micrometastases, thereby promoting postoperative tumor control[26]. Previous perioperative studies have demonstrated that preoperative immunotherapy is feasible and can induce clinically meaningful pathological responses[27]. In the broader framework of multimodal HCC management, the neoadjuvant approach evaluated in this study should be distinguished from conversion therapy or palliative systemic therapy. Our cohort includes patients who ultimately underwent curative-intent resection, with preoperative treatment used as part of a perioperative strategy to improve tumor control, rather than as definitive palliative therapy or conversion therapy. For high-risk HCC, systemic therapy, local therapy, antiviral optimization, and perioperative treatment may interact to improve surgical eligibility, reduce recurrence risk, and optimize long-term outcome[28].

In the neoadjuvant therapy group, patients who achieved MPR/pCR demonstrated superior RFS compared to those who did not, further supporting the correlation between pathological response and long-term benefits. This is consistent with prior studies; for instance, a large retrospective study by Zeng et al[29] showed that among patients receiving preoperative systemic therapy combined with local treatment, ≥ 90% MPR was closely associated with improved OS (P < 0.001). Pathological response not only reflects the morphological necrosis of the tumor but may also indicate sensitivity to systemic therapy and reduced risk of postoperative recurrence. In other words, a more complete pathological response corresponds to a lower risk of postoperative recurrence, suggesting that pathological response may play a dual role in the perioperative management of huge HCC: As a short-term efficacy indicator and a surrogate endpoint for long-term benefits. A prospective study also supports this concept, showing that ≥ 90% tumor regression after neoadjuvant immune checkpoint inhibitor therapy may serve as a surrogate indicator of RFS in HCC patients[25]. However, in this study, the sample size for this analysis was limited, and the OS difference did not reach statistical significance; thus, these findings need to be further validated in larger prospective studies.

Multivariable analysis indicated that upfront surgery and the absence of postoperative adjuvant therapy were associated with poorer OS and RFS, while multiple tumors further indicated a higher risk of recurrence. These findings suggest that the long-term outcomes for patients with huge HCC depend not only on the completion of curative resection but also on whether perioperative therapy is adequate and whether high-risk tumor biology is effectively controlled. Previous studies have also suggested that recurrence after curative HCC treatment is influenced by multiple factors, including residual micrometastases, tumor burden, and high-risk pathological features[30]. In this context, preoperative neoadjuvant therapy may further improve long-term outcomes by reducing preoperative tumor activity and selecting patients sensitive to systemic therapy[31]. Clinically, these findings indicate that the effects observed in multivariable analysis reflect the adequacy of the overall perioperative treatment strategy rather than any single treatment step: For huge HCC, merely completing curative resection is insufficient to offset the high recurrence tendency; integrating neoadjuvant therapy with postoperative adjuvant therapy into a continuous perioperative management pathway may be crucial for improving long-term prognosis. This finding also suggests that the survival benefits associated with neoadjuvant therapy should be interpreted within the broader context of perioperative treatment rather than as isolated preoperative effects. Considering the increasingly evident yet still evolving role of adjuvant therapy in HCC, neoadjuvant and postoperative adjuvant treatments can be viewed as complementary components for recurrence control in selected high-risk patients[32]. This also highlights the distinction between technical resectability and oncologic adequacy. Although R0 resection is technically feasible and perioperatively safe in selected patients with giant or even super-giant HCC, surgery alone may still be insufficient to overcome the high tumor burden and aggressive biological behavior of these tumors. Prior data from super-giant HCC cohorts further support this distinction, showing that despite resection being feasible, long-term outcomes remain poor, and tumor volume may have prognostic significance[33].

In multivariable analysis, the lack of association between hypertension and poorer RFS is noted. One possible explanation is that patients with hypertension may be receiving antihypertensive treatment, such as beta-blockers or angiotensin receptor blockers, which have been reported in observational studies to be associated with improved cancer outcomes. Further research is needed, incorporating detailed medication data, to clarify the impact of antihypertensive therapy on tumor recurrence and survival[34,35].

In this study, neoadjuvant therapy showed more pronounced benefits in patients with vascular invasion, PVTT-positive status, and BCLC-C stage. This suggests that the advantages of neoadjuvant therapy are primarily observed in high-recurrence-risk, biologically aggressive huge HCC patients. This finding is biologically plausible, as vascular invasion and PVTT generally indicate a higher burden of micrometastasis and greater potential for recurrence[36]. Thus, our study does not support the mechanical application of neoadjuvant therapy to all huge HCC patients, but rather suggests selective application in high-risk subgroups, for whom surgery alone is unlikely to adequately control biological risks. Nevertheless, given the limited sample size and multiple comparisons in subgroup analyses, these findings should be considered exploratory results.

Notably, neoadjuvant therapy did not increase operative time, intraoperative blood loss, transfusion requirements, or postoperative hospital stay, suggesting that, with appropriate treatment cessation windows and multidisciplinary assessment, preoperative multimodal therapy does not significantly compromise perioperative safety. Recent perioperative studies similarly indicate that, when conducted by experienced teams, the integration of immunotherapy, targeted therapy, and locoregional therapy into the preoperative phase is generally feasible[37]. Consequently, the novelty of this study lies not in proving the superiority of any fixed regimen but in providing real-world evidence that, in the extremely high-risk and evidence-limited population of huge HCC, multimodal neoadjuvant strategies confer consistent directional benefits.

Several limitations should be acknowledged. First, this is a single-center retrospective study with a relatively small sample size, which may introduce selection bias. In particular, the neoadjuvant therapy group only included patients who successfully proceeded to curative resection; patients who lost surgical opportunities after neoadjuvant therapy were not included. Therefore, the results should be interpreted as postoperative outcomes in a selected surgical cohort, rather than definitive evidence of the overall efficacy of neoadjuvant therapy. Second, the regimens within the neoadjuvant therapy group were heterogeneous, making it impossible to further elucidate the independent contributions of each regimen or combination in huge HCC. Third, only patients who ultimately completed curative resection were included; patients who lost surgical opportunities due to disease progression, liver function deterioration, or other reasons were not represented. Thus, the results primarily reflect real-world outcomes of patients who successfully underwent surgery, rather than the overall benefits for all initially treated patients. Fourth, the sample size for subgroup and pathological response analyses was relatively small, serving more as hypothesis-generating evidence rather than definitive conclusions. Fifth, the cohort was primarily composed of HBV-related HCC, which is consistent with the epidemiology in China but limits the generalizability to populations with other etiologies. Additionally, the start time for the upfront surgery group was defined as the date of surgery, while the neoadjuvant therapy group was defined as the start of neoadjuvant therapy. This definition was intended to reflect the entire perioperative treatment strategy, including the preoperative treatment period for patients receiving neoadjuvant therapy, and is consistent with previous preoperative treatment studies. Although the time-zero points between the two groups are not the same, this methodological issue should be considered when interpreting survival outcomes. Future prospective multicenter studies are needed, incorporating pathological response, biomarkers, and immune microenvironment characteristics, to determine the optimal regimens, timing, and patient selection.

CONCLUSION

In this single-center retrospective surgical cohort of patients with huge HCC who underwent curative resection, neoadjuvant therapy was associated with improved RFS and OS without significantly increasing perioperative burden, with more pronounced associations observed in high-risk subgroups such as those with vascular invasion, PVTT-positive status, and BCLC-C stage. These results suggest that multimodal neoadjuvant strategies may provide valuable perioperative treatment options for selected patients with this high-recurrence-risk disease. However, given the retrospective design, potential selection bias, treatment heterogeneity, and the inclusion of only patients who ultimately underwent surgery, these findings should be interpreted with caution and require validation in prospective multicenter studies using an intention-to-treat design.