Abstract
Introduction
Surgical resection following systemic therapy is feasible in patients with initially unresectable hepatocellular carcinoma (HCC). However, postoperative tumor recurrence is common after surgery, and the factors affecting this recurrence remain unclear. This study aimed to assess factors influencing postoperative outcomes in patients with initially unresectable HCC undergoing hepatectomy after systemic therapy.
Methods
This study retrospectively enrolled patients with initially unresectable HCC who underwent hepatectomy after targeted therapy plus immunotherapy (with or without locoregional therapy). Multivariate Cox regression analyses were used to identify the independent prognostic factors for recurrence-free survival (RFS) and overall survival (OS). Machine learning was used to determine the RFS rates at different intervals for different radiographic responses.
Results
Eighty-one patients who underwent R0 hepatectomy after systemic therapy were included. With a median follow-up of 17.4 (interquartile range: 7.2–22.3) months, median RFS and OS were not reached. Preoperative tumor downstaging and achieving pathological complete response were associated with improved RFS and OS. Multivariate Cox analyses identified radiographic response as an independent prognostic factor for RFS and OS. Furthermore, a radiographic response >40% (assessed using the Response Evaluation Criteria in Solid Tumors, version 1.1) or >50% (assessed using the modified Response Evaluation Criteria in Solid Tumors) was associated with a longer RFS (p = 0.006 and 0.003, respectively).
Conclusion
Radiographic response depth was an independent prognostic factor in patients with initially unresectable HCC who underwent hepatectomy following targeted therapy plus immunotherapy, and the response to systemic therapy may be the determining factor for patient prognosis after surgery.
Keywords: Hepatectomy, Hepatocellular carcinoma, Radiographic response, Recurrence, Systemic therapy
Introduction
Hepatocellular carcinoma (HCC) is a common malignant tumor and a leading cause of cancer mortality worldwide and in China [1, 2]. Although surgery is the main curative treatment option for patients with HCC, most patients are diagnosed at an advanced stage, rendering them ineligible for surgery and limiting treatment options [3].
Recently, systemic therapies, such as targeted therapy plus immunotherapy, have significantly improved the treatment efficacy for advanced or unresectable HCC. With an objective response rate of approximately 30% [4–6], many patients with advanced HCC, that is otherwise unresectable, have the opportunity to undergo surgery [7–9].
In our previous study, we concluded that selected patients with initially unresectable HCC can undergo hepatectomy after systemic therapy with a combination of tyrosine kinase inhibitor plus anti-PD-1 antibodies, and this treatment regimen is associated with a favorable prognosis [10]. Another clinical trial verifying the efficacy and safety of lenvatinib plus anti-PD-1 antibodies suggested that among the patients with a successful conversion, the R0 resection rate was 85.7%, and pathological complete response (pCR) was achieved in 38.1% of the 21 patients who underwent surgery after systemic therapy [11].
Although preoperative systemic therapy could improve resectability, some patients who received systemic treatment and underwent surgical resection still experienced tumor recurrence after surgery. However, perioperative factors affecting postoperative tumor recurrence remain unclear. For patients with HCC who underwent upfront hepatectomy, postoperative high-risk recurrence factors include largest tumor >5 cm, more than three tumors, microvascular invasion, macrovascular invasion, and grade 3/4 pathology [12]. However, whether these factors determine the risk of postoperative recurrence in patients with HCC who undergo hepatectomy after systemic therapy remains unclear. Therefore, understanding the potential perioperative factors that affect postoperative tumor recurrence after hepatectomy following systemic therapy is of clinical significance, which may contribute to optimizing the preoperative evaluation, timing of surgery, and treatment regimen.
This study aimed to identify and understand perioperative factors that influence postoperative tumor recurrence in patients with initially unresectable HCC who received targeted therapy plus immunotherapy and underwent R0 resection. The relationship between perioperative factors, including common high-risk factors for postoperative recurrence, changes between baseline and preoperative variables, and patient postoperative prognosis were analyzed to identify independent prognostic factors of postoperative recurrence in patients with initially unresectable HCC who underwent hepatectomy following targeted therapy plus immunotherapy.
Methods
Patient Enrollment
This was a retrospective single-center study. Consecutive patients with initially unresectable HCC from Zhongshan Hospital, Fudan University (Shanghai, China) between December 2018 and August 2023 were eligible if they received first-line targeted therapy plus immunotherapy, either with or without locoregional therapy, followed by hepatectomy, were unresectable owing to surgical or oncological reasons, and achieved R0 resection after hepatectomy. Briefly, surgical reasons for initial unresectability include insufficient resection margin or insufficient future liver remnant [13–15], and the oncological reason for initial unresectability is that resection is not the preferred treatment recommendation according to the Barcelona Clinic Liver Cancer (BCLC) guideline (i.e., BCLC stage B and C) [16]. Patients with extrahepatic disease before the initiation of systemic therapy, incomplete clinicopathological data, or a history of cancers other than HCC were excluded.
The study was conducted in accordance with the Declaration of Helsinki, and written informed consent was obtained from all patients before inclusion in the study. The study protocol was approved by the Zhongshan Hospital Research Ethics Committee (Approval No. B2020-177R).
Systemic and Locoregional Therapy
All patients received combination therapy with targeted therapy plus immunotherapy, with or without locoregional therapy, prior to liver resection. After the discussion with multidisciplinary liver tumor boards, final regimens were chosen by patients themselves according to their own willing and financial ability. One of the following antiangiogenic drugs was administered: lenvatinib (8 mg/day, orally) [17], bevacizumab (15 mg/kg, intravenously) [18], apatinib (250 mg/day, orally) [19], and donafenib (400 mg/day, orally) [20]. One of the following anti-PD-(L)1 antibodies was intravenously administered: nivolumab 3 mg/kg [21], camrelizumab 200 mg [22] every 2 weeks, pembrolizumab 200 mg [17], sintilimab 200 mg [18], toripalimab 240 mg [23], tislelizumab 200 mg [24], or atezolizumab 1,200 mg [4] every 3 weeks. Locoregional therapy includes transarterial chemoembolization and hepatic artery infusion chemotherapy. All patients with an active HBV infection received concomitant antiviral therapy with entecavir or tenofovir.
All patients were treated and monitored regularly as previously described [8]. Briefly, complete blood count, liver, renal, thyroid, adrenal, and cardiac functions, and tumor markers were monitored every 2–3 weeks before each anti-PD-(L)1 antibody treatment cycle. Tumor response was evaluated via contrast-enhanced magnetic resonance imaging/computed tomography and chest computed tomography every 2 months (±2 weeks). Tumor response was assessed according to the Response Evaluation Criteria in Solid Tumors, version 1.1 (RECIST v1.1) [25] and modified Response Evaluation Criteria in Solid Tumors (mRECIST) [26] as the best overall response of intrahepatic lesion(s). An objective response was defined as a complete response or partial response.
Hepatectomy
During combination therapy, patients were classified as resectable, as previously described [8], if (1) R0 resection could be achieved with sufficient remnant liver volume and function, (2) intrahepatic lesions were evaluated as having a partial response or stable disease for at least 2 months, (3) no severe or persistent adverse effects occurred from systemic therapy, and (4) no contraindications for hepatectomy existed. The surgical resection procedure has been described in our previous study [27]. Briefly, intraoperative ultrasonography was used to visualize the tumor location and its relationship with major vascular structures and to detect satellite nodules. Parenchyma transection was conducted by alternating the ultrasonic dissector and clamp-crushing techniques. Complete hemostasis was achieved via ligation or electrocoagulation. The Pringle maneuver or hepatic vein occlusion was used to control bleeding from the inflow or outflow vessels, if necessary. pCR was defined as the absence of residual viable tumor cells on hematoxylin and eosin-stained slide sections among all resected primary tumor(s), tumor thrombosis, and metastatic lesions [28].
Definitions of Some Analyzed Factors
Tumor Stage and Migration
The tumor stage was assessed using the BCLC and China Liver Cancer (CNLC) stage system [29]. Tumor stage per RECIST v1.1, represents the entire tumor area considered, and tumor stage per mRECIST represents only the enhanced tumor area considered. The tumor stage was assessed at baseline and before surgery. Tumor downstaging represents a reduction in the cancer stage from a more to a less threatening stage. Tumor upstaging represents an increase in the cancer stage from a less to a more threatening stage.
Radiographic Response
Radiographic response was defined as the percentage reduction in tumor size on preoperative imaging compared with baseline imaging. The formula for calculating the radiographic response was as follows:
A positive radiographic response represents tumor regression before surgery compared to baseline and vice versa. The radiographic response per RECIST v1.1 indicated the percentage reduction in the tumor size of the entire tumor area. The radiographic response per mRECIST indicated the percentage reduction in tumor size in the enhanced tumor area.
Hepatic lesions (≥1 cm) were selected as target lesions if they could be accurately measured in at least one dimension, with the final selection of target lesions based on size (the two with the longest diameter) and suitability for accurate repeated measurements. The tumor size was calculated as the sum of the diameters of the hepatic target lesions.
AFP and PIVKA-II Responses
The α-fetoprotein (AFP) and PIVKA-II responses were modified from our previous study [30] and calculated using the following formulas:
Demonstrating Recurrence-Free Survival Rate Using Machine Learning
To determine the recurrence-free survival (RFS) rate at different years after surgery for different radiographic responses, a random forest was employed to build a prediction model using all patients in this study (training set). Multi-fidelity tuning via hyperband with 10-fold cross-validation was used for hyperparameter optimization in the training process, and the proportion of the training set was used as a model-agnostic fidelity parameter. After model training and hyperparameter optimization, nested resampling was used to estimate model performance in an unbiased manner. The above machine learning processes were accomplished using the mlr3 ecosystem based on R.
Follow-Up
Patients were followed up every 60 days (±7 days) after initiation of combination therapy. Overall survival (OS) was calculated from the date of surgery to death owing to any cause or censored at the last follow-up. RFS was calculated from the date of surgery to the first documented disease recurrence, death from any cause, or censorship at the final follow-up.
Statistical Analysis
Continuous variables are expressed as mean (standard deviation) or median (interquartile range [IQR]) and compared using Student’s t test or the Mann-Whitney U test, as appropriate. Categorical variables are expressed as counts and percentages and were compared using Pearson’s χ2 analysis, Fisher’s exact test, or Mann-Whitney U test, as appropriate. Paired variables were compared using McNemar’s χ2 test, Mann-Whitney U test, or Student’s t test as appropriate. Survival curves were calculated using the Kaplan-Meier method and compared using the log-rank test. Variables with a p value <0.1 in univariate Cox regression analyses and that may have prognostic significance were evaluated by multivariate Cox regression analyses to identify independent prognostic factors of RFS and OS. No evidence of multicollinearity was found in the multivariate Cox regression analyses after checking the variance inflation factor [31]. The concordance index (C-index) was used to determine the discrimination ability of the prediction model for RFS. Maximally selected rank statistics were used to determine the optimal cutoff value of continuous variables for RFS. A p value of <0.05 was considered statistically significant. All statistical analyses were performed using R software (version 4.3.1).
Results
Patient Characteristics and Survival
Of the 108 patients who received targeted therapy plus immunotherapy, with or without locoregional therapy, followed by hepatectomy, 81 were eligible for this study (Fig. 1). The patients’ baseline and perioperative characteristics are presented in Table 1.
Fig. 1.
Flowchart of patient enrollment.
Table 1.
Patient baseline and perioperative characteristics
| Variables | At baseline (n = 81) | Before surgery (n = 81) | After surgery (n = 81) | p valuea |
|---|---|---|---|---|
| Age, mean±standard deviation, years | 57.5±9.8 | |||
| Sex | ||||
| Female | 3 (3.7) | |||
| Male | 78 (96.3) | |||
| ECOG PS | ||||
| 0 | 74 (91.4) | |||
| 1 | 7 (8.6) | |||
| Child-Pugh class | ||||
| A | 80 (98.8) | |||
| B | 1 (1.2) | |||
| HBsAg, IU/mL | 1,851.0 (538.0, 2,407.0) | 1,623.0 (143.0, 2,251.0) | 0.001 | |
| HBsAg | 0.250 | |||
| Negative | 8 (9.9) | 13 (16.0) | ||
| Positive | 69 (85.2) | 68 (84.0) | ||
| NA | 4 (4.9) | 0 (0.0) | ||
| HBV DNA | <0.001 | |||
| ≤1,000/mL | 41 (50.6) | 72 (88.9) | ||
| >1,000/mL | 37 (45.7) | 5 (6.2) | ||
| NA | 3 (3.7) | 4 (4.9) | ||
| BCLC stage per RECIST v1.1 | 0.740 | |||
| A | 12 (14.8) | 16 (19.8) | ||
| B | 33 (40.7) | 27 (33.3) | ||
| C | 36 (44.4) | 38 (46.9) | ||
| CNLC stage per RECIST v1.1 | 0.215 | |||
| Ia | 0 (0.0) | 2 (2.5) | ||
| Ib | 12 (14.8) | 14 (17.3) | ||
| IIa | 15 (18.5) | 19 (23.5) | ||
| IIb | 18 (22.2) | 8 (9.9) | ||
| IIIa | 36 (44.4) | 36 (44.4) | ||
| IIIb | 0 (0.0) | 2 (2.5) | ||
| BCLC stage per mRECIST | <0.001 | |||
| Tumor-freeb | 0 (0.0) | 10 (12.3) | ||
| 0 | 0 (0.0) | 4 (4.9) | ||
| A | 12 (14.8) | 24 (29.6) | ||
| B | 33 (40.7) | 21 (25.9) | ||
| C | 36 (44.4) | 22 (27.2) | ||
| CNLC stage per mRECIST | <0.001 | |||
| Tumor-freeb | 0 (0.0) | 10 (12.3) | ||
| Ia | 0 (0.0) | 9 (11.1) | ||
| Ib | 12 (14.8) | 19 (23.5) | ||
| IIa | 15 (18.5) | 15 (18.5) | ||
| IIb | 18 (22.2) | 6 (7.4) | ||
| IIIa | 36 (44.4) | 20 (24.7) | ||
| IIIb | 0 (0.0) | 2 (2.5) | ||
| Macrovascular invasion per RECIST v1.1 | 0.999 | |||
| No | 46 (56.8) | 44 (54.3) | ||
| Yes | 35 (43.2) | 37 (45.7) | ||
| Macrovascular invasion per mRECIST | 0.004 | |||
| No | 46 (56.8) | 60 (74.1) | ||
| Yes | 35 (43.2) | 21 (25.9) | ||
| Extrahepatic disease | 0.999 | |||
| No | 81 (100.0) | 79 (97.5) | ||
| Yes | 0 (0.0) | 2 (2.5) | ||
| AFP, ng/mL | 180.5 (9.32, 1,451.5) | 8.0 (3.1, 161) | <0.001 | |
| AFP | <0.001 | |||
| ≤400 ng/mL | 49 (60.5) | 67 (82.7) | ||
| >400 ng/mL | 31 (38.3) | 14 (17.3) | ||
| NA | 1 (1.2) | 0 (0.0) | ||
| PIVKA-II, mAU/mL | 2,908.0 (321.3, 23,724.8) | 196.0 (33.0, 1,899.0) | <0.001 | |
| PIVKA-II | <0.001 | |||
| ≤1,000 mAU/mL | 28 (34.6) | 53 (65.4) | ||
| >1,000 mAU/mL | 52 (64.2) | 28 (34.6) | ||
| NA | 1 (1.2) | 0 (0) | ||
| Tumor number per RECIST v1.1 | 0.039 | |||
| ≤3 | 61 (75.3) | 69 (85.2) | ||
| >3 | 20 (24.7) | 12 (14.8) | ||
| Tumor number per mRECIST | 0.007 | |||
| ≤3 | 61 (75.3) | 72 (88.9) | ||
| >3 | 20 (24.7) | 9 (11.1) | ||
| Tumor size per RECIST v1.1, cm | 9 (6.8, 14.3) | 6.57 (4.18, 10.0) | <0.001 | |
| Tumor size per RECIST v1.1 | <0.001 | |||
| ≤5 cm | 11 (13.6) | 25 (30.9) | ||
| >5 cm | 70 (86.4) | 56 (69.1) | ||
| Tumor size per mRECIST, cm | 9 (6.8, 14.0) | 4.7 (2.0, 8.2) | <0.001 | |
| Tumor size per mRECIST | <0.001 | |||
| ≤5 cm | 11 (13.6) | 42 (51.9) | ||
| >5 cm | 70 (86.4) | 39 (48.1) | ||
| Antiangiogenic drug used | ||||
| Apatinib | 3 (3.7) | |||
| Bevacizumab | 35 (43.2) | |||
| Donafenib | 1 (1.2) | |||
| Lenvatinib | 42 (51.9) | |||
| Anti-PD-(L)1 antibody used | ||||
| Atezolizumab | 18 (22.2) | |||
| Camrelizumab | 10 (12.3) | |||
| Nivolumab | 1 (1.2) | |||
| Pembrolizumab | 19 (23.5) | |||
| Sintilimab | 28 (34.6) | |||
| Tislelizumab | 3 (3.7) | |||
| Toripalimab | 2 (2.5) | |||
| Locoregional therapy | ||||
| No | 71 (87.7) | |||
| Yes | 10 (12.3) | |||
| Time to surgery in months | 3.1 (2.5, 4.4) | |||
| Tumor responsec per RECIST v1.1 | ||||
| CR | 2 (2.5) | |||
| PR | 34 (42) | |||
| SD | 45 (55.6) | |||
| Objective responsec per RECIST v1.1 | ||||
| No | 45 (55.6) | |||
| Yes | 36 (44.4) | |||
| Tumor responsec per mRECIST | ||||
| CR | 11 (13.6) | |||
| PR | 40 (49.4) | |||
| SD | 30 (37.0) | |||
| Objective responsec per mRECIST | ||||
| No | 30 (37.0) | |||
| Yes | 51 (63.0) | |||
| AFP response | 0.7 (0.4, 0.9) | |||
| PIVKA-II response | 0.8 (0.6, 1.0) | |||
| Radiographic responsed per RECIST v1.1, % | 21.1 (9.1, 39.0) | |||
| Radiographic responsed per mRECIST, % | 43.3 (13.5, 81.2) | |||
| Microvascular invasion | ||||
| No | 52 (64.2) | |||
| Yes | 29 (35.8) | |||
| Poor tumor differentiation (grade 3 or 4) | ||||
| No | 63 (77.8) | |||
| Yes | 18 (22.2) | |||
| Pathological complete response | ||||
| No | 64 (79.0) | |||
| Yes | 17 (21.0) | |||
Data are presented as n (%) or median (IQR) unless indicated otherwise.
AFP, α-fetoprotein; BCLC, Barcelona Clinic Liver Cancer; CNLC, China Liver Cancer; CR, complete response; ECOG PS Eastern Cooperative Oncology Group performance status; HBsAg, hepatitis B surface antigen; HBV DNA, hepatitis B virus DNA copy number; mRECIST, modified Response Evaluation Criteria in Solid Tumors; NA, not available; PD-(L)1, programmed death-(ligand) 1; PIVKA-II, protein induced by vitamin K absence or antagonist-II; PR, partial response; RECIST v1.1, Response Evaluation Criteria in Solid Tumors, version 1.1; SD, stable disease.
a p value represented the comparison between variables at baseline and before surgery.
bTumor-free indicates no enhanced area of the tumor(s) on imaging.
cTumor response was assessed as the best overall response before surgery.
dPositive radiographic response represents tumor regression before surgery compared to baseline and vice versa.
Twelve (14.8%) BCLC stage A patients were initially unresectable due to surgical reasons (i.e., insufficient resection margin or insufficient future liver remnant), and the remaining 69 (85.2%) patients were initially unresectable owing to oncological reasons. Ten (12.3%) patients received concurrent locoregional therapy at baseline. The objective response rates before surgery per RECIST v1.1 and mRECIST were 44.4% (36/81) and 63.0% (51/81), respectively, with a pCR rate of 21.0% (17/81). The median time to surgery was 3.1 (IQR: 2.5–4.4) months. The hepatitis B surface antigen titer, hepatitis B virus (HBV) DNA copy number (HBV DNA), tumor stage per mRECIST, macrovascular invasion per mRECIST, AFP, prothrombin induced by vitamin K absence or antagonists-II (PIVKA-II), tumor number, and tumor size decreased after combination therapy (Table 1).
Although some baseline characteristics had significant differences among different regimens (i.e., bevacizumab plus atezolizumab/sintilimab, lenvatinib plus an anti-PD-1 antibody, and systemic therapy plus locoregional therapy), the preoperative tumor response, radiographic response, and pathological response were not significantly different among different regimens in this cohort, with systemic therapy plus locoregional therapy having a trend of higher radiographic response than systemic therapy alone (online suppl. Table 1; for all online suppl. material, see https://doi.org/10.1159/000541300).
Radiographic responses between surgically (n = 12) and oncologically (n = 69) unresectable HCC patients were not significantly different (median [IQR]: 18.9% [11.6%–21.9%] vs. 23.0% [9.1%–45.3%] per RECIST v1.1, p = 0.239; 19.5% [8.3%–87.6%] vs. 45.3% [15.9%–80.6%] per mRECIST, p = 0.576), while oncologically unresectable patients having a trend of higher radiographic response than surgically unresectable patients, indicating that oncologically unresectable patients may require a higher radiographic response than those surgically unresectable to meet the criteria for hepatectomy.
The median follow-up as of September 12, 2023, was 17.4 (IQR: 7.2–22.3) months; 12 (14.8%) patients died, and 22 (27.2%) patients had recurrence. The median RFS and OS were not reached (Fig. 2a, b). The 1-, 2-, and 3-year RFS rates were 70.5% (95% confidence interval [CI]: 60.6–82.1%), 57.3% (95% CI: 44.7–73.4%), and 57.3% (95% CI: 44.7–73.4%), respectively; the 1-, 2-, and 3-year OS rates were 91.0% (95% CI: 84.4–98.2%), 84.7% (95% CI: 75.7–94.7%), and 74.4% (95% CI: 60.2–91.9%), respectively.
Fig. 2.
Recurrence-free and overall survival plots. RFS (a) and OS (b) in all patients. RFS (c) and OS (d) grouped by whether achieving pathological complete response. OS, overall survival; RFS, recurrence-free survival.
Tumor Downstaging Is Associated with Improved RFS and OS
Sankey diagrams of tumor stage migration are shown in Figure 3a, b. Two (2.5%) patients without extrahepatic disease at baseline developed extrahepatic disease preoperatively. The intrahepatic lesions of these 2 patients shrank by ∼30% preoperatively. The extrahepatic lesions of these 2 patients were completely resected during surgery, and R0 resection was achieved. The relationship between tumor stage migration and patient survival is shown in Figure 3c–f. In the case of RECIST v1.1, patients with tumor downstaging had a trend of better RFS and OS, while patients with tumor upstaging had a trend of worse RFS and OS than patients with unchanged tumor stages (Fig. 3c, d). In the case of mRECIST, patients with tumor upstaging had a trend of the worst OS, and the trend of OS was similar between patients with tumor downstaging and unchanged tumor stages. Patients with tumor downstaging and upstaging had trends of better and worse RFS, respectively, than patients with unchanged tumor stages (Fig. 3e, f).
Fig. 3.
Relationship between tumor stage migration and patient survival. Sankey diagrams of tumor stage migration per RECIST v1.1 (a) and mRECIST (b), respectively. Relationship between tumor stage migration and RFS (c)/OS (d) per RECIST v1.1. Relationship between tumor stage migration and RFS (e)/OS (f) per mRECIST. BCLC, Barcelona Clinic Liver Cancer; mRECIST, modified Response Evaluation Criteria in Solid Tumors; OS, overall survival; RECIST v1.1, Response Evaluation Criteria in Solid Tumors version 1.1; RFS, recurrence-free survival.
Achieving pCR Is Associated with Improved RFS and OS
Achieving pCR was significantly associated with improved RFS (hazard ratio [HR]: 0.115, 95% CI: 0.016–0.852, p = 0.011; Figure 2c), which was consistent with our previous report [28] but was marginally related to improved OS (HR: 0.218, 95% CI: 0.027–1.746, p = 0.120; Fig. 2d), which may be due to relatively short follow-up time and small sample size. However, achieving pCR was neither an independent prognostic factor for RFS nor OS in the case of either RECIST v1.1 (Table 2) or mRECIST (Table 3).
Table 2.
Univariate and multivariate analysis of perioperative factors for RFS and OS according to RECIST v1.1
| Variables | RFS | OS | ||||||
|---|---|---|---|---|---|---|---|---|
| univariate analysis | multivariate analysis | univariate analysis | multivariate analysis | |||||
| p value | HR | 95% CI | p value | p value | HR | 95% CI | p value | |
| Before surgery | ||||||||
| HBsAg, positive versus negative | 0.155 | 0.465 | ||||||
| HBV DNA, >1,000/mL versus ≤1,000/mL | <0.001 | 7.288 | 1.478–35.941 | 0.015 | 0.407 | |||
| BCLC stage | ||||||||
| B versus A | 0.140 | 0.861 | ||||||
| C versus A | 0.194 | 0.375 | ||||||
| Macrovascular invasion, yes versus no | 0.985 | 0.436 | ||||||
| AFP, >400 ng/mL versus ≤400 ng/mL | 0.007 | 1.990 | 0.598–6.626 | 0.262 | 0.773 | 0.728 | 0.123–4.304 | 0.726 |
| PIVKA-II, >1,000 mAU/mL versus ≤1,000 mAU/mL | 0.003 | 0.953 | 0.232–3.919 | 0.946 | 0.525 | 0.482 | 0.099–2.341 | 0.365 |
| Tumor number, >3 versus ≤3 | 0.879 | 1.140 | 0.287–4.523 | 0.853 | 0.759 | 1.890 | 0.349–10.232 | 0.460 |
| Tumor size, >5 cm versus ≤5 cm | 0.009 | 10.837 | 1.367–85.907 | 0.024 | 0.278 | 5.984 | 0.67–53.476 | 0.109 |
| Time to surgery, months | 0.309 | 0.697 | ||||||
| Objective response, yes versus no | 0.009 | 3.412 | 0.653–17.827 | 0.146 | 0.519 | 3.211 | 0.498–20.691 | 0.220 |
| AFP response | 0.002 | 1.262 | 0.727–2.191 | 0.409 | 0.900 | |||
| PIVKA-II response (multiplied by 10) | <0.001 | 1.212 | 0.885–1.659 | 0.232 | 0.151 | |||
| Radiographic response (divided by 10) | <0.001 | 0.913 | 0.842–0.991 | 0.030 | 0.005 | 0.900 | 0.825–0.981 | 0.017 |
| After surgery | ||||||||
| MVI, yes versus no | <0.001 | 1.271 | 0.341–4.739 | 0.721 | 0.109 | 3.404 | 0.618–18.747 | 0.159 |
| Poor tumor differentiation (grade 3 or 4), yes versus no | 0.009 | 2.539 | 0.873–7.384 | 0.087 | 0.062 | 1.649 | 0.424–6.412 | 0.471 |
| Pathologic complete response, yes versus no | 0.034 | 0.470 | 0.042–5.299 | 0.541 | 0.151 | 0.288 | 0.024–3.423 | 0.324 |
AFP, α-fetoprotein; BCLC, Barcelona Clinic Liver Cancer; CI, confidence interval; HR, hazard ratio; HbsAg, hepatitis B surface antigen; HBV DNA, hepatitis B virus DNA copy number; MVI, microvascular invasion, PIVKA-II, protein induced by vitamin K absence or antagonist-II; RECIST v1.1, Response Evaluation Criteria in Solid Tumors, version 1.1.
Table 3.
Univariate and multivariate analysis of perioperative factors for RFS and OS according to mRECIST
| Variables | RFS | OS | ||||||
|---|---|---|---|---|---|---|---|---|
| univariate analysis | multivariate analysis | univariate analysis | multivariate analysis | |||||
| p value | HR | 95% CI | p value | p value | HR | 95% CI | p value | |
| Before surgery | ||||||||
| HBsAg, positive versus negative | 0.155 | 0.465 | ||||||
| HBV DNA, >1,000/mL versus ≤1,000/mL | <0.001 | 8.656 | 1.727–43.395 | 0.009 | 0.407 | |||
| BCLC stagea | ||||||||
| 0 versus tumor-free | 0.351 | NE | NE | NE | 0.270 | |||
| A versus tumor-free | 0.428 | 0.770 | 0.05–11.773 | 0.851 | 0.942 | |||
| B versus tumor-free | 0.091 | 0.958 | 0.052–17.499 | 0.977 | 0.716 | |||
| C versus tumor-free | 0.124 | 0.267 | 0.014–4.974 | 0.377 | 0.287 | |||
| Macrovascular invasion, yes versus no | 0.551 | 0.461 | ||||||
| AFP, >400 ng/mL versus ≤400 ng/mL | 0.007 | 3.255 | 0.82–12.92 | 0.093 | 0.773 | 1.023 | 0.167–6.273 | 0.980 |
| PIVKA-II, >1,000 mAU/mL versus ≤1,000 mAU/mL | 0.003 | 1.180 | 0.321–4.34 | 0.804 | 0.525 | 0.642 | 0.151–2.721 | 0.547 |
| Tumor number, >3 versus ≤3 | 0.266 | 1.912 | 0.464–7.877 | 0.369 | 0.352 | 2.274 | 0.426–12.14 | 0.337 |
| Tumor size, >5 cm versus ≤5 cm | 0.018 | 2.765 | 0.78–9.796 | 0.115 | 0.822 | 0.719 | 0.176–2.928 | 0.645 |
| Time to surgery, months | 0.309 | 0.697 | ||||||
| Objective response, yes versus no | 0.012 | 2.105 | 0.549–8.066 | 0.277 | 0.585 | 1.868 | 0.343–10.167 | 0.470 |
| AFP response | 0.002 | 1.125 | 0.604–2.095 | 0.711 | 0.900 | |||
| PIVKA-II response (multiplied by 10) | <0.001 | 1.142 | 0.781–1.668 | 0.494 | 0.151 | |||
| Radiographic response (divided by 10) | <0.001 | 0.943 | 0.898–0.991 | 0.019 | 0.006 | 0.952 | 0.909–0.998 | 0.042 |
| After surgery | ||||||||
| MVI, yes versus no | <0.001 | 1.323 | 0.347–5.043 | 0.682 | 0.109 | 3.493 | 0.595–20.517 | 0.166 |
| Poor tumor differentiation (grade 3 or 4), yes versus no | 0.009 | 4.109 | 1.056–15.984 | 0.041 | 0.062 | 1.346 | 0.324–5.596 | 0.683 |
| Pathologic complete response, yes versus no | 0.034 | 0.688 | 0.056–8.472 | 0.770 | 0.151 | 0.406 | 0.038–4.367 | 0.457 |
AFP, α-fetoprotein; BCLC, Barcelona Clinic Liver Cancer; CI, confidence interval; HR, hazard ratio; HbsAg, hepatitis B surface antigen; HBV DNA, hepatitis B virus DNA copy number; mRECIST, modified Response Evaluation Criteria in Solid Tumors; MVI, microvascular invasion; NE, not evaluable; PIVKA-II, protein induced by vitamin K absence or antagonist-II.
aTumor-free indicates there is no enhanced area of the tumor(s) on imaging.
Radiographic Response as an Independent Prognostic Factor of Both RFS and OS
None of the baseline variables were identified as an independent prognostic factor of RFS or OS in the case of either RECIST v1.1 or mRECIST (online suppl. Table S2, S3). In the multivariate analyses of perioperative variables, HBV DNA >1,000/mL (HR: 7.288, 95% CI: 1.478–35.941, p = 0.015) was independently associated with an unfavorable RFS, and the radiographic response was independently associated with favorable RFS (HR: 0.913, 95% CI: 0.842–0.991, p = 0.030) and OS (HR: 0.900, 95% CI: 0.825–0.981, p = 0.017) in the case of RECIST v1.1 (Table 2). HBV DNA >1,000/mL (HR: 8.656, 95% CI: 1.727–43.395, p = 0.009) was independently associated with an unfavorable RFS, and the radiographic response was independently associated with favorable RFS (HR: 0.943, 95% CI: 0.898–0.991, p = 0.019) and OS (HR: 0.952, 95% CI: 0.909–0.998, p = 0.042) in the case of mRECIST (Table 3).
Notably, the results indicated that a preoperative HBV DNA level of >1,000/mL was an independent prognostic factor for RFS. All patients with active HBV infection received concomitant antiviral therapy (entecavir or tenofovir). Five (6.2%) patients had an HBV DNA level of >1,000/mL before surgery. Of these 5 patients, four (80.0%) died or had tumor recurrence, and one (20.0%) remained tumor-free and alive as of the date of the data cutoff. The tumor response of these 5 patients was stable disease per RECIST v1.1 or mRECIST (online suppl. Table S4).
RFS Rate at Different Years after Surgery for Different Radiographic Responses
Although the radiographic response was an independent prognostic factor for RFS, it was difficult to determine the RFS rate at different time points after surgery for a specific radiographic response. To further explore the relationship between radiographic response and RFS rate, a random forest model using radiographic response to predict RFS rate was constructed after hyperparameter optimization (n = 81). The C-indexes for radiographic response per RECIST v1.1 and per mRECIST were 0.779 and 0.800, respectively.
The prediction models are presented in Table 4. At the same postoperative time point, the greater the radiographic response, the greater the RFS rate. The RFS rate improved significantly and remained almost unchanged when the radiographic response was higher than 40% (HR: 0.167, 95% CI: 0.039–0.710, p = 0.006; Fig. 4a) or 50% (HR: 0.247, 95% CI: 0.093–0.657, p = 0.003; Fig. 4c) per RECIST v1.1 or mRECIST, respectively, which was almost in accordance with the optimal cutoff value of radiographic response for RFS determined using the maximally selected rank statistics (online suppl. Fig. S1a, b), that is, radiographic response >34.8% or >42.5% per RECIST v1.1 or mRECIST, respectively.
Table 4.
RFS rate at different years after surgery for different radiographic responses
| Radiographic responsea | RFS rate, % | |||||||
|---|---|---|---|---|---|---|---|---|
| under radiographic response per RECIST v1.1 | under radiographic response per mRECIST | |||||||
| time after surgery | time after surgery | |||||||
| 12 months | 24 months | 36 months | 48 months | 12 months | 24 months | 36 months | 48 months | |
| 0% | 57.9 | 44.2 | 44.2 | 44.2 | 54.2 | 45.3 | 45.3 | 45.3 |
| 10% | 49.2 | 42.0 | 42.0 | 42.0 | 45.4 | 40.6 | 40.6 | 40.6 |
| 20% | 59.7 | 57.1 | 57.1 | 57.1 | 49.4 | 44.4 | 44.4 | 44.4 |
| 30% | 58.2 | 51.9 | 51.9 | 51.9 | 46.6 | 40.3 | 40.3 | 40.3 |
| 40% | 93.0 | 81.2 | 81.2 | 81.2 | 61.7 | 53.1 | 53.1 | 53.1 |
| 50% | 96.3 | 81.6 | 81.6 | 81.6 | 83.4 | 72.8 | 72.8 | 72.8 |
| 60% | 97.1 | 83.4 | 83.4 | 83.4 | 89.6 | 72.0 | 72.0 | 72.0 |
| 70% | 97.2 | 84.2 | 84.2 | 84.2 | 90.7 | 77.4 | 77.4 | 77.4 |
| 80% | 97.2 | 84.2 | 84.2 | 84.2 | 89.0 | 78.3 | 78.3 | 78.3 |
| 90% | 97.2 | 84.2 | 84.2 | 84.2 | 91.0 | 80.5 | 80.5 | 80.5 |
| 100% | 97.2 | 84.2 | 84.2 | 84.2 | 92.5 | 80.4 | 80.4 | 80.4 |
For example, patients with a radiographic response of 40% had RFS rates of 93.0% and 61.7% at 12 months after surgery, according to RECIST v1.1 and mRECIST, respectively. The numbers of patients in each radiographic response interval are shown in online supplementary Table S5.
mRECIST, modified Response Evaluation Criteria in Solid Tumors; RECIST v1.1, Response Evaluation Criteria in Solid Tumors, version 1.1.
aPositive radiographic response represents tumor regression before surgery compared to baseline and vice versa.
Fig. 4.
Relationship between radiographic response and patient survival. RFS (a) and OS (b) grouped by radiographic response per RECIST v1.1. RFS (c) and OS (d) grouped by radiographic response per mRECIST. mRECIST, modified Response Evaluation Criteria in Solid Tumors; OS, overall survival; RECIST v1.1, Response Evaluation Criteria in Solid Tumors version 1.1; RFS, recurrence-free survival.
Discussion
In this study, we identified radiographic response as an independent prognostic factor for RFS and OS in patients with initially unresectable HCC who underwent R0 hepatectomy following targeted therapy plus immunotherapy and exhibited RFS rates at different postoperative time points for specific radiographic responses. To the best of our knowledge, this is the first study to report independent prognostic perioperative factors in patients with initially unresectable HCC who received targeted therapy plus immunotherapy followed by surgical resection.
In this study, unresectability was classified as surgically or oncologically unresectable. Surgically unresectable HCC is mainly due to patient’s general condition or liver function intolerance, insufficient remaining liver volume, or insufficient resection margins [13–15]. The definition of surgically unresectable HCC is widely agreed, but for oncologically unresectable HCC, the definition is controversial. Generally, the oncological reason for unresectability was predicted efficacy of upfront hepatectomy may not surpass other nonsurgical treatment options [15]. However, there are no specific criteria for oncological unresectability. In the era of novel systemic therapies, one oncological resectability criteria [32] for HCC has been published recently. According to the criteria, oncological resectability for intrahepatic lesions includes single HCC regardless of the tumor size and up to 3 modules with each ≤3 cm (i.e., BCLC stage 0–A). Intrahepatic lesions exceeding these criteria are considered unresectable, or borderline resectable according to the original text, which is consistent with the definition of oncologically unresectability in this study. Furthermore, the criteria indicate that some patients with vascular invasion or bile duct invasion are also considered as oncologically resectable. However, these patients were considered as oncologically unresectable in this study.
In this study, neither the median RFS nor OS was achieved for all patients. The median OS reported with atezolizumab plus bevacizumab in the IMbrave150 trial [33] and with lenvatinib plus pembrolizumab in the LEAP-002 trial [34] were 19.2 and 21.2 months, respectively. All patients in this study had initially unresectable HCC but underwent surgery after receiving combination therapy, indicating that the efficacy of the studied patients was better than that of the overall population. Furthermore, our previous study showed that conversion surgery is associated with a favorable prognosis [10]. Therefore, the prognosis of the patients in this study may be superior to that reported in phase 3 clinical trials.
In this study, the independent prognostic factors of postoperative recurrence for patients with initially unresectable HCC who underwent hepatectomy following systemic therapy were radiographic response and HBV DNA level >1,000/mL before surgery, which were different from those for patients with HCC who underwent upfront hepatectomy [12]. Considering that at least two independent studies concluded that a deeper pathological response, which can be partially reflected by the radiographic response, after conversion surgery was associated with decreased postoperative recurrence in patients with HCC who received targeted therapy plus immunotherapy [28, 35], this result may imply that tumor shrinkage is more important than other tumor-related factors in patients with initially unresectable HCC who received hepatectomy following systemic therapy. Therefore, response to systemic therapy may be the determining factor of patient prognosis after surgery.
The radiographic response is an indicator of the efficacy of preoperative systemic therapy. Topalian et al. [36] reported that radiographic tumor regression could predict improved RFS in patients with Merkel cell carcinoma who received neoadjuvant therapy before surgery. Takigawa et al. [37] reported that early radiographic response during bevacizumab-containing chemoradiation could be a predictive indicator of patient survival in unresectable glioblastoma. A higher radiographic response indicates more significant tumor shrinkage before surgery, suggesting that patients with a higher radiographic response are likely to have fewer minimal residual lesions before surgery than patients with a lower radiographic response. Furthermore, although the duration of postoperative systemic therapy may vary among patients, all patients continued their preoperative systemic regimen after surgery. Therefore, in patients with a higher radiographic response, continued postoperative systemic therapy will also help eliminate minimal residual lesions. These findings support the idea that the radiographic response can serve as an independent prognostic factor for postoperative recurrence, indicating that the efficacy of systemic therapy is an important factor affecting the survival of these patients. Therefore, if the tumor reaches the desired radiographic regression (i.e., radiographic response), surgical resection should be performed. For patients in whom the desired radiographic response was not achieved, treatment should be continued under close observation to achieve a deeper tumor response, followed by surgical resection.
This study had several limitations. First, it was a retrospective, single-center study with a limited number of patients, the results of which require validation in multicenter studies. Second, different systemic regimens were used across the study population, and preoperative systemic therapy may have been considered as neoadjuvant or conversion therapy for different patients. However, the key objective of this study was to assess the relationship between preoperative depth of radiographic response, rather than regimens themselves, and patient prognosis. Third, patients with extrahepatic disease at baseline were excluded. Radiographic response was defined based on intrahepatic lesions, which did not consider the impact of extrahepatic disease on the prognosis. Finally, more data are needed to validate whether high HBV DNA load before surgery is an independent prognostic factor for postoperative recurrence.
Conclusion
In patients with initially unresectable HCC who underwent R0 hepatectomy following targeted therapy plus immunotherapy, both tumor downstaging and pCR were associated with improved RFS and OS, and the radiographic response was an independent prognostic factor for RFS and OS. The response to systemic therapy, as reflected by the radiographic response, may be a determinant of patient recurrence and survival after surgical resection.
Acknowledgments
The authors wish to thank all the patients and their families.
Statement of Ethics
The study was conducted in accordance with the Declaration of Helsinki, and written informed consent was obtained from all patients before inclusion in the study. The study protocol was approved by the Zhongshan Hospital Research Ethics Committee (Approval No. B2020-177R). Consent for publication was obtained from all patients.
Conflict of Interest Statement
H.-C.S. received honorariums and lecture fees from Roche, Bayer, MSD, Eisai, Hengrui, Innovent, TopAlliance, Abbott, BeiGene, Gilead, and Zelgen during the last 5 years. J Fan is an Editorial Board Member of Liver Cancer. The other authors declare no conflicts of interest.
Funding Sources
This work was supported by the National Natural Science Foundation of China (81871929 and 82072667 to C.H. and 82372037 to H.-C.S.), the Special Research Fund for Liver Cancer Diagnosis and Treatment from the China Anti-Cancer Association (H2020-044 to C.H. and H2020-008 to H.-C.S.), the Clinical Research Special Fund of Zhongshan Hospital, Fudan University (2020ZSLC71 to H.-C.S.), and the 2022 Personalized Medical Incubator Project (the fund for Precision Medicine Research and Industry Development in SIMQ) (EH20 to H.-C.S.).
Author Contributions
B.X. and L.-N.W. conceived and designed the study, analyzed the data, and drafted and revised the manuscript. Z.-Y.W. and T.H. prepared and analyzed the data. X.-D.Z., Y.-H.S., J.Z., and J.F. revised the manuscript. H.-C.S. and C.H. designed the study, interpreted the results, and revised the manuscript. All authors provided critical comments regarding the manuscript and have read and approved this manuscript.
Funding Statement
This work was supported by the National Natural Science Foundation of China (81871929 and 82072667 to C.H. and 82372037 to H.-C.S.), the Special Research Fund for Liver Cancer Diagnosis and Treatment from the China Anti-Cancer Association (H2020-044 to C.H. and H2020-008 to H.-C.S.), the Clinical Research Special Fund of Zhongshan Hospital, Fudan University (2020ZSLC71 to H.-C.S.), and the 2022 Personalized Medical Incubator Project (the fund for Precision Medicine Research and Industry Development in SIMQ) (EH20 to H.-C.S.).
Data Availability Statement
The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the corresponding author (C.H. or H.-C.S.) upon reasonable request.
Supplementary Material.
Supplementary Material.
References
- 1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71(3):209–49. [DOI] [PubMed] [Google Scholar]
- 2. Zheng R, Zhang S, Zeng H, Wang S, Sun K, Chen R, et al. Cancer incidence and mortality in China, 2016. J Natl Cancer Cent. 2022;2(1):1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Vogel A, Meyer T, Sapisochin G, Salem R, Saborowski A. Hepatocellular carcinoma. Lancet. 2022;400(10360):1345–62. [DOI] [PubMed] [Google Scholar]
- 4. Finn RS, Qin S, Ikeda M, Galle PR, Ducreux M, Kim TY, et al. Atezolizumab plus bevacizumab in unresectable hepatocellular carcinoma. N Engl J Med. 2020;382(20):1894–905. [DOI] [PubMed] [Google Scholar]
- 5. Xu J, Shen J, Gu S, Zhang Y, Wu L, Wu J, et al. Camrelizumab in combination with apatinib in patients with advanced hepatocellular carcinoma (RESCUE): a nonrandomized, open-label, phase II trial. Clin Cancer Res. 2021;27(4):1003–11. [DOI] [PubMed] [Google Scholar]
- 6. Llovet JM, Kudo M, Merle P, Meyer T, Qin S, Ikeda M, et al. Lenvatinib plus pembrolizumab versus lenvatinib plus placebo for advanced hepatocellular carcinoma (LEAP-002): a randomised, double-blind, phase 3 trial. Lancet Oncol. 2023;24(12):1399–410. [DOI] [PubMed] [Google Scholar]
- 7. Zhang W, Hu B, Han J, Wang Z, Ma G, Ye H, et al. Surgery after conversion therapy with PD-1 inhibitors plus tyrosine kinase inhibitors are effective and safe for advanced hepatocellular carcinoma: a pilot study of ten patients. Front Oncol. 2021;11:747950. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Zhu XD, Huang C, Shen YH, Ji Y, Ge NL, Qu XD, et al. Downstaging and resection of initially unresectable hepatocellular carcinoma with tyrosine kinase inhibitor and anti-PD-1 antibody combinations. Liver Cancer. 2021;10(4):320–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Qu WF, Ding ZB, Qu XD, Tang Z, Zhu GQ, Fu XT, et al. Conversion therapy for initially unresectable hepatocellular carcinoma using a combination of toripalimab, lenvatinib plus TACE: real-world study. BJS Open. 2022;6(5):zrac114. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Zhu XD, Huang C, Shen YH, Xu B, Ge NL, Ji Y, et al. Hepatectomy after conversion therapy using tyrosine kinase inhibitors plus anti-PD-1 antibody therapy for patients with unresectable hepatocellular carcinoma. Ann Surg Oncol. 2023;30(5):2782–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Zhang W, Tong S, Hu B, Wan T, Tang H, Zhao F, et al. Lenvatinib plus anti-PD-1 antibodies as conversion therapy for patients with unresectable intermediate-advanced hepatocellular carcinoma: a single-arm, phase II trial. J Immunother Cancer. 2023;11(9):e007366. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Qin S, Chen M, Cheng AL, Kaseb AO, Kudo M, Lee HC, et al. Atezolizumab plus bevacizumab versus active surveillance in patients with resected or ablated high-risk hepatocellular carcinoma (IMbrave050): a randomised, open-label, multicentre, phase 3 trial. Lancet. 2023;402(10415):1835–47. [DOI] [PubMed] [Google Scholar]
- 13. Shindoh J, Kawamura Y, Kobayashi Y, Kobayashi M, Akuta N, Okubo S, et al. Prognostic impact of surgical intervention after lenvatinib treatment for advanced hepatocellular carcinoma. Ann Surg Oncol. 2021;28(12):7663–72. [DOI] [PubMed] [Google Scholar]
- 14. Sun HC, Zhu XD. Downstaging conversion therapy in patients with initially unresectable advanced hepatocellular carcinoma: an overview. Front Oncol. 2021;11:772195. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Sun HC, Zhou J, Wang Z, Liu X, Xie Q, Jia W, et al. Chinese expert consensus on conversion therapy for hepatocellular carcinoma (2021 edition). Hepatobiliary Surg Nutr. 2022;11(2):227–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Reig M, Forner A, Rimola J, Ferrer-Fabrega J, Burrel M, Garcia-Criado A, et al. BCLC strategy for prognosis prediction and treatment recommendation: the 2022 update. J Hepatol. 2022;76(3):681–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Finn RS, Ikeda M, Zhu AX, Sung MW, Baron AD, Kudo M, et al. Phase Ib study of lenvatinib plus pembrolizumab in patients with unresectable hepatocellular carcinoma. J Clin Oncol. 2020;38(26):2960–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Ren Z, Xu J, Bai Y, Xu A, Cang S, Du C, et al. Sintilimab plus a bevacizumab biosimilar (IBI305) versus sorafenib in unresectable hepatocellular carcinoma (ORIENT-32): a randomised, open-label, phase 2-3 study. Lancet Oncol. 2021;22(7):977–90. [DOI] [PubMed] [Google Scholar]
- 19. Qin S, Li Q, Gu S, Chen X, Lin L, Wang Z, et al. Apatinib as second-line or later therapy in patients with advanced hepatocellular carcinoma (AHELP): a multicentre, double-blind, randomised, placebo-controlled, phase 3 trial. Lancet Gastroenterol Hepatol. 2021;6(7):559–68. [DOI] [PubMed] [Google Scholar]
- 20. Qin S, Bi F, Gu S, Bai Y, Chen Z, Wang Z, et al. Donafenib versus sorafenib in first-line treatment of unresectable or metastatic hepatocellular carcinoma: a randomized, open-label, parallel-controlled phase II-III trial. J Clin Oncol. 2021;39(27):3002–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. El-Khoueiry AB, Sangro B, Yau T, Crocenzi TS, Kudo M, Hsu C, et al. Nivolumab in patients with advanced hepatocellular carcinoma (CheckMate 040): an open-label, non-comparative, phase 1/2 dose escalation and expansion trial. Lancet. 2017;389(10088):2492–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Qin S, Ren Z, Meng Z, Chen Z, Chai X, Xiong J, et al. Camrelizumab in patients with previously treated advanced hepatocellular carcinoma: a multicentre, open-label, parallel-group, randomised, phase 2 trial. Lancet Oncol. 2020;21(4):571–80. [DOI] [PubMed] [Google Scholar]
- 23. Mai HQ, Chen QY, Chen D, Hu C, Yang K, Wen J, et al. Toripalimab or placebo plus chemotherapy as first-line treatment in advanced nasopharyngeal carcinoma: a multicenter randomized phase 3 trial. Nat Med. 2021;27(9):1536–43. [DOI] [PubMed] [Google Scholar]
- 24. Qin S, Finn RS, Kudo M, Meyer T, Vogel A, Ducreux M, et al. RATIONALE 301 study: tislelizumab versus sorafenib as first-line treatment for unresectable hepatocellular carcinoma. Future Oncol. 2019;15(16):1811–22. [DOI] [PubMed] [Google Scholar]
- 25. Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford R, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer. 2009;45(2):228–47. [DOI] [PubMed] [Google Scholar]
- 26. Lencioni R, Llovet JM. Modified RECIST (mRECIST) assessment for hepatocellular carcinoma. Semin Liver Dis. 2010;30(1):52–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Li XL, Xu B, Zhu XD, Huang C, Shi GM, Shen YH, et al. Simulation of portal/hepatic vein associated remnant liver ischemia/congestion by three-dimensional visualization technology based on preoperative CT scan. Ann Transl Med. 2021;9(9):756. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Huang C, Zhu XD, Shen YH, Xu B, Wu D, Ji Y, et al. Radiographic and α-fetoprotein response predict pathologic complete response to immunotherapy plus a TKI in hepatocellular carcinoma: a multicenter study. BMC Cancer. 2023;23(1):416. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Zhou J, Sun H, Wang Z, Cong W, Zeng M, Zhou W, et al. Guidelines for the Diagnosis and treatment of primary liver cancer (2022 edition). Liver Cancer. 2023;12(5):405–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Li XL, Zhu XD, Cai H, Li Y, Zhou J, Fan J, et al. Postoperative α-fetoprotein response predicts tumor recurrence and survival after hepatectomy for hepatocellular carcinoma: a propensity score matching analysis. Surgery. 2019;165(6):1161–7. [DOI] [PubMed] [Google Scholar]
- 31. Fox J, Monette G. Generalized collinearity diagnostics. J Am Stat Assoc. 1992;87(417):178–83. [Google Scholar]
- 32. Akahoshi K, Shindoh J, Tanabe M, Ariizumi S, Eguchi S, Okamura Y, et al. Oncological resectability criteria for hepatocellular carcinoma in the era of novel systemic therapies: the Japan liver cancer association and Japanese society of hepato-biliary-pancreatic surgery expert consensus statement 2023. Liver Cancer. 2024:1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Cheng AL, Qin S, Ikeda M, Galle PR, Ducreux M, Kim TY, et al. Updated efficacy and safety data from IMbrave150: atezolizumab plus bevacizumab vs. sorafenib for unresectable hepatocellular carcinoma. J Hepatol. 2022;76(4):862–73. [DOI] [PubMed] [Google Scholar]
- 34. Llovet JM, Kudo M, Merle P, Meyer T, Qin S, Ikeda M, et al. Lenvatinib plus pembrolizumab versus lenvatinib plus placebo for advanced hepatocellular carcinoma (LEAP-002): a randomised, double-blind, phase 3 trial. Lancet Oncol. 2023;24(12):1399–410. [DOI] [PubMed] [Google Scholar]
- 35. Ho WJ, Zhu Q, Durham J, Popovic A, Xavier S, Leatherman J, et al. Neoadjuvant cabozantinib and nivolumab converts locally advanced HCC into resectable disease with enhanced antitumor immunity. Nat Cancer. 2021;2(9):891–903. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Topalian SL, Bhatia S, Amin A, Kudchadkar RR, Sharfman WH, Lebbe C, et al. Neoadjuvant nivolumab for patients with resectable Merkel cell carcinoma in the CheckMate 358 trial. J Clin Oncol. 2020;38(22):2476–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Takigawa K, Hata N, Michiwaki Y, Hiwatashi A, Yonezawa H, Kuga D, et al. Volumetric study reveals the relationship between outcome and early radiographic response during bevacizumab-containing chemoradiotherapy for unresectable glioblastoma. J Neuro Oncol. 2021;154(2):187–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the corresponding author (C.H. or H.-C.S.) upon reasonable request.