Comprehensive Analysis of Radiologic Cancer-Free Status through Various Treatment Approaches in Advanced-Stage Hepatocellular Carcinoma

Primary liver cancer, notably hepatocellular carcinoma (HCC), remains one of the most concerning malignancies with increasing incidence and mortality rates globally. The majority of HCC cases are linked to underlying liver diseases such as hepatitis B or C virus infections, alcohol abuse, or fatty liver due to metabolic syndrome [1, 2]. Current guidelines, both in the East and West, emphasize the importance of regular surveillance of high-risk populations [3‒6]. This approach has improved the early detection of HCC, making timely and curative treatment possible [7]. Nonetheless, the inherent metastatic potential of HCC and the high carcinogenic risk associated with chronic liver diseases pose challenges. Moreover, many patients experience recurrence even when early-stage HCC is effectively treated. Over time, this can progress to advanced-stage HCC, marked by macrovascular invasion (MVI), including the portal vein, hepatic vein, and bile duct, as well as potential extrahepatic spread (EHS). Furthermore, the absence or failure of surveillance often results in patients being initially diagnosed with advanced-stage HCC. The increasing incidence of nonviral HCC associated with fatty liver disease due to metabolic syndrome, coupled with the lack of a standardized monitoring protocol for the large number of patients with fatty liver disease globally, suggests that the diagnosis of advanced HCC may persist or even increase in the near future [8].

The Barcelona Clinic Liver Cancer (BCLC) staging system, a globally recognized guideline, considers the presence of MVI or EHS in HCC as challenging to address curatively. It advocates for systemic therapy as the primary treatment modality for advanced-stage HCC in the presence of EHS or MVI [9]. However, in primarily Asian regions, there is a longstanding belief that certain subsets of HCC, even those corresponding to BCLC advanced-stage, remain amenable to curative treatments. Indeed, numerous studies in these regions have demonstrated promising outcomes with curative resections, even when HCC presents with MVI or EHS [10‒13]. Some Asian guidelines even recommend curative surgery, particularly for HCC subsets with MVI [5‒7]. Recently, with the advent of highly effective systemic therapies such as atezolizumab combined with bevacizumab and lenvatinib, there have been reports of successful conversions of even advanced HCC with MVI or EHS to states amenable to curative treatment [14, 15].

In the realm of oncology, the pinnacle of therapeutic success is achieving a state where the patient is entirely free of cancer, i.e., “cancer free.” This benchmark should not only apply to early-stage malignancies but should also be the standard aim for advanced-stage diseases, including HCC. However, there is a glaring gap in the research landscape concerning the understanding of what cancer-free status truly means for patients with advanced-stage HCC. The complex nature of advanced-stage HCC, characterized by its varied progression patterns, resulting in diverse tumor statuses, may contribute to this oversight. Recognizing this knowledge gap, our study sheds light on the profound implications of patients with advanced-stage HCC reaching a cancer-free status using imaging examination results. Using a detailed patient database, we explored the varied clinical courses leading to advanced-stage HCC. We also examined the outcomes for patients who achieved a cancer-free state by radiologic assessments following advanced-stage HCC.

We established a comprehensive database at Chiba University Hospital to meticulously track the clinical course of all patients with HCC treated since April 2003. In this database, HCC diagnoses were confirmed either by histopathologic evidence or by characteristic imaging markers identified in dynamic computed tomography or magnetic resonance imaging scans. This database was integrated with the hospital’s electronic imaging system, allowing for a comprehensive review of all relevant radiology studies.

Our review encompassed radiology data for patients with HCC from April 2003 to June 2022, with a focus on identifying the initial occurrence of MVI or EHS. The detection of either MVI or EHS was considered the time of diagnosis of advanced-stage HCC, serving as the starting date for our analysis. In cases where MVI or EHS was present at the initial diagnosis, that date was assigned as the starting point for our study. The analysis concluded with a data cutoff date of 31st December 2022. This study was approved by the Research Ethics Committee of Chiba University (Approval No. 2247). Given the nature of this study and in accordance with Japan’s Ethical Guidelines for Medical and Biological Research Involving Human Subjects, formal consent via written signature was not required. Instead, participating sites provided detailed study information to the subjects, giving them the option to opt out if they did not wish to participate (Statement of Ethics is also described below).

Treatment decisions adhered to the most current Japanese clinical practice guidelines and were the result of a collaborative effort of a multidisciplinary team, including hepatologists, surgeons, and radiologists [7]. For advanced-stage HCC, we initially assessed the appropriateness of surgical resection, which is the most curative option. If surgery was not feasible, systemic therapy, which became the standard of care in 2009 [16, 17], was the primary option. In cases where systemic therapy was deemed unsuitable due to poor liver function or where transcatheter interventions, such as transarterial chemoembolization (TACE) or hepatic arterial infusion chemotherapy (HAIC), were considered effective for vascular invasion, alternative treatments were selected.

We divided the patients with advanced-stage HCC into three groups based on the identification date of the initial instances of MVI or EHS (period 1: 2003–2008, period 2: 2009–2016, and period 3: 2017–2022). Period 1 reflects the era when systemic therapy has not become the standard of care; period 2 reflects the era when sorafenib was the sole systemic therapy; and period 3 reflects the era when multiple VEGF-TKIs and combination immunotherapy were available. Overall survival (OS) was defined as the time from the date of diagnosis of advanced-stage HCC to the date of death. The censoring date was defined as the date of the last follow-up. The frequency of achieving cancer-free status from imaging examinations was investigated, categorized by patient background and stage progression pattern to advanced stage.

We assessed all images taken after the identification of advanced HCC using the modified Response Evaluation Criteria in Solid Tumors (mRECIST) for HCC [18]. For this study, radiologic cancer-free (rCF) was defined radiologically by the following criteria: in surgical cases, the absence of tumors and no recurrence in imaging evaluations spaced a minimum of 2 months apart from resection; and in nonsurgical cases, two consecutive imaging evaluations (a minimum of 2 months apart) showed no tumors, with both evaluations indicating a complete response (CR) as per mRECIST. To evaluate the disappearance of cancer in radiologic imaging, rCF was defined based on two different types. Type A, assessed by RECIST v1.1 criteria, involved the complete disappearance of all lesions. Type A typically occurs following curative treatment or systemic therapy. Type B, evaluated by mRECIST criteria, was characterized by persistent nodules but complete disappearance of arterial enhancement in all lesions. Type B was commonly observed after systemic therapy or local treatments such as TACE.

The Pearson χ² test and Fisher’s exact probability test were used to compare demographic and clinical characteristics. OS was defined as the time from diagnosis with advanced-stage HCC until death, with the censoring date defined as the date of the last follow-up. Estimates such as cumulative survival curves, median OS, and annual survival proportions are obtained using the Kaplan-Meier method. p values were estimated using a log-rank test. Decision tree analysis was performed using data mining methods based on previous reports [19‒21]. Clinical factors that exhibited significant differences in univariate analysis (Fisher’s exact probability test) were used as explanatory variables, while the achievement of rCF status was used as the objective variable. The analysis was conducted to predict patients achieving rCF status. The software algorithm identified the most significant predictor from the analytical database, which was then used to stratify patients into two distinct groups. This predictor revealed that a developed rCF prediction model visually represents patient subgroups with varying rCF rates in a flowchart structure. The level of significance was considered two-sided at 5%. All statistical analyses were performed using JMP software (JMP Pro version 17, SAS Institute Inc., Cary, NC, USA) and R software (version 4.2.3; R Foundation for Statistical Computing, Vienna, Austria).

We identified 795 patients with advanced-stage HCC in our database. The baseline characteristics of these patients, all of whom received some form of treatment, are detailed in Table 1. The median age of the patients was 70 years, with a male predominance of 80.4%. A significant 73.3% of patients were classified as Child-Pugh Class A. MVI and EHS were observed in 57.8% and 49.4% of patients, respectively, with 12.3% presenting with both conditions. The primary sites of EHS were, in descending order of frequency, the lung, bone, adrenal gland, and peritoneum. HCV was the most common etiology accounting for 42.8% of cases. The distribution of patients among the three periods was as follows: 107 in period 1, 380 in period 2, and 308 in period 3. Online supplementary Table S1 (for all online suppl. material, see https://doi.org/10.1159/000542577) shows the detailed demographics for each period, highlighting a progressive decrease in the proportion of HCV-positive patients.

Our research encompassed a comprehensive analysis of medical records, focusing on treatments administered at initial diagnosis and during subsequent management of advanced-stage HCC. Figure 1 illustrates the treatment strategies immediately after diagnosing advanced-stage disease. Although some patients received more than one type of treatment, this figure represents the initial treatment. A notable trend observed was the increasing use of systemic therapy over time, coupled with a decreasing preference for resection as the initial treatment. Additionally, a significant decrease in the percentage of patients with advanced-stage HCC treated with transarterial therapies, such as TACE or HAIC, was noted. Online supplementary Figure S1 shows the Kaplan-Meier survival curves for OS over the three time periods. The OS for patients with advanced-stage HCC in periods 1, 2, and 3 was 9.5 months (95% CI: 7.2–13.8), 12.3 months (95% CI: 9.2–14.8), and 17.0 months (95% CI: 15.6–23.1), respectively, with significant differences observed between these time periods (p = 0.001).

In our study cohort, 8.7% (69 patients) were determined to have achieved the rCF status based on our defined criteria. The clinical outcomes of the two rCF types defined in this study were examined after achieving type A (disappearance of all lesions) and type B (disappearance of arterial enhancement of all lesions). No clear difference in cumulative recurrence incidence and cumulative survival was observed between the two types (online suppl. Fig. S2). The results revealed that the clinical outcomes of the two types were largely consistent.

A total of 81 patients achieved curative treatment or radiologic CR (online suppl. Fig. S3). Among them, 12 patients experienced early recurrence (within 2 months). Comparing the prognosis of these patients with those with early recurrence (12 cases) and those without recurrence for >2 months after two imaging assessments (69 cases), a clear difference can be observed (online suppl. Fig. S4). Notably, the OS of patients in our cohort who achieved rCF status was significantly longer, with the median OS not reached (95% CI: 67.4–NE) compared to those who did not achieve this status, who had a median OS of 11.7 months (95% CI: 9.8–13.8), indicating a significant survival benefit (p < 0.001) (Fig. 2). The OS was significantly prolonged in the subgroup of patients with MVI and EHS with confirmed rCF (online suppl. Fig. S5). Online supplementary Figure S6 shows a Kaplan-Meier curve comparing patients who have recurrence-free period >6 or 12 months with those who did not, with results similar to those shown in Figure 2, defined as >2 months. For these reasons, patients with early recurrence were excluded from rCF. Based on our analysis results, our definition of the rCF criteria was considered appropriate for conducting further analyses.

Of the 69 patients who achieved rCF, 58 had all three markers (AFP, AFP-L3, and PIVKA-II) measured at the time of confirmation. A total of 34 patients exhibited all marker levels within the normal range. A comparison of recurrence rates between patients with all normal markers and those with at least one elevated marker revealed no statistically significant difference in time to recurrence. The median time to recurrence for the marker-negative and nonnegative groups was 16.2 months (95% CI: 12.7–23.0) and 22.2 months (95% CI: 9.3–34.0), respectively, without statistically significant difference in cumulative recurrence rates observed (p = 0.624), although this subgroup analysis is limited by the small sample size. Remarkably, among 24 patients with abnormal markers, 11 exhibited elevated AFP levels.

The rCF rates for periods 1, 2, and 3 were 9.4% (10 of 107 patients), 7.4% (28 of 380 patients), and 10.1% (31 of 308 patients), respectively. The treatment procedure leading to an rCF status is depicted in Figure 3 and online supplementary Table S2. Among the patients who achieved rCF status, 79.7% (55 of 69 patients) were treated with a single therapeutic approach. The predominant treatment strategy was resection alone, followed by systemic therapy. A smaller subset of patients achieved rCF status with either TACE or HAIC alone, although there were cases in which these treatments did not result in rCF status. Online supplementary Table S3 presents the treatment choices made immediately after reaching the advanced stage and the rCF achievement rates resulting from these treatments. The proportion of the first treatment before rCF was as follows: resection had the highest rate (42.9%; 36/84), followed by TACE (5.9%; 16/254). Systemic therapy had a rate of 3.7% (13/351). Regarding specific systemic therapy regimens, the rCF rates were 2.4% for sorafenib, 1.6% for lenvatinib, and 7.0% for the combination of atezolizumab and bevacizumab. The rCF was achieved with systemic therapy alone in 11/69 (15.9%) patients (2 achieved by systemic therapy plus other treatment). In these 11 patients, the median response duration and PFS were 26.5 months and 28.0 months, respectively. Moreover, 7/11 (63.6%) patients were managed as drug-free status after achieving rCF.

In addition, the median time to cumulative recurrence for patients who achieved rCF status according to our criteria was 19.1 months (95% CI: 14.1–22.6). When comparing cumulative recurrence rates, there was no statistically significant difference between patients who underwent resection to achieve rCF status and those who did not. Specifically, the median cumulative recurrence time was 14.1 months (95% CI: 9.5–22.2) in patients without resection and 19.3 months (95% CI: 15.0–30.6) in patients who underwent resection (p = 0.349) (online suppl. Fig. S7).

The demographic and clinical characteristics of patients who achieved rCF status and those who did not are detailed in Table 2. In patients with advanced-stage disease, the achievement of an rCF status is contingent upon the tumor characteristics. To further understand this, we conducted a comparative analysis of patient backgrounds in relation to rCF status, stratifying data based on the presence or absence of EHS, as shown in online supplementary Table S4. The clinical characteristics associated with the achievement of rCF status from these analyses are summarized in Figure 4. We observed that the rCF rate was 10.1% in patients without EHS compared to 7.3% in patients with EHS. Interestingly, patients with EHS but no intrahepatic lesions had an rCF rate of 25.9%. In addition, patients with a single EHS nodule had a higher rCF rate of 36.7%.

To highlight the prediction of rCF achievement rate by the patient background, decision tree analysis was employed to explore the conditions necessary to achieve rCF status in patients with advanced-stage HCC (Fig. 5). The decision tree demonstrated that the number of intrahepatic tumors was a critical determinant of achieving rCF status. Interestingly, the presence of bone metastases significantly influenced the likelihood of achieving an rCF status in patients with intrahepatic tumor number of 7 or less. Patients without intrahepatic lesions but with a single metastatic lesion other than bone exhibited a notably high rCF rate of 50%. However, only 2.1% of patients with 8 or more tumors achieved rCF status. The analysis revealed that factors including the tumor diameter, such as maximum size of intrahepatic lesions ≥50 mm or up-to-seven criteria, did not emerge as robust predictors of the rCF status.

For further analysis, a logistic regression analysis was performed on factors associated with achieving the rCF status (online suppl. Table S5). Consistent with the results of the decision tree analysis, the number of intrahepatic tumors of ≥8 was a more important factor than the tumor diameter or the up-to-seven criteria. Liver function, bone metastasis, and AFP were also extracted as significant factors associated with achieving rCF in multivariate analysis, supporting the results of the decision tree analysis.

To evaluate the prognostic impact of achieving the rCF status, univariate and multivariate Cox proportional hazards analyses for OS were performed. In addition to the tumor condition, two rCF types were added: rCF type A (disappearance of all lesions) and rCF type B (disappearance of arterial enhancement of all lesions). In multivariate analysis, both types A and B were significant independent factors that reduced the risk of OS (online suppl. Table S6). The impact of our definition of rCF on survival was demonstrated.

To highlight the importance of achieving the rCF status, a good control arm resembling the rCF cohort is needed. Propensity score matching was adjusted for patient background using critical factors extracted from the decision tree analysis. As a result, cohorts of two groups were created with well-adjusted backgrounds: the presence of rCF (n = 64) and the absence of rCF (n = 64) (online suppl. Table S7). OS was compared in the matched cohort; in the matched cohort, OS could also be stratified by the rCF status (online suppl. Fig. S8). In the matched cohort, achieving the rCF status was still associated with improved survival.

In this cohort, 37.6% (299 out of 795) of patients were initially diagnosed with advanced-stage HCC (Fig. 6, Pattern I). The remaining patients were initially diagnosed with HCC at an early or intermediate-stage but later progressed to advanced-stage HCC following different clinical courses. We categorized these progression patterns as follows: pattern II comprises cases initially diagnosed at an early stage that progressed directly to an advanced stage without an intermediate-stage; pattern III includes cases initially diagnosed at an intermediate stage that later progressed to an advanced stage; and pattern IV comprises cases initially diagnosed at an early stage that later progressed to the advanced stage after passing through an intermediate stage. Table 3 shows the patient backgrounds at the time of advanced stage for each treatment course leading to this stage. Pattern II patients were more likely to have no intrahepatic lesions. In contrast, pattern I was characterized by a significantly higher number of patients with ≥50 mm intrahepatic tumor and MVI. In contrast, patterns III and IV showed an increased number of patients with intrahepatic tumor counts of ≥8. Online supplementary Table S8 shows the initial treatment strategy by stage progression patterns. The rCF rates for these patterns were 11.0%, 12.6%, 2.7%, and 0.8%, respectively (Fig. 6).

This study represents a unique approach using a large cohort of patients to identify the point at which HCC was diagnosed at an advanced stage and to analyze the transition from that point to rCF status. Our analysis differs from mainstream clinical research on systemic therapy for advanced HCC, which commonly includes a mix of advanced- and intermediate-stage patients and typically excludes surgical candidates. The results indicated that although the overall rate of patients with advanced-stage HCC achieving rCF status was low, the prognostic impact was significant. We found that patients with certain tumor characteristics had a higher success rate in achieving the rCF status. Although several studies have concentrated on particular interventions for advanced-stage HCC, comprehensive clinical data encompassing a range of treatment modalities, as in our study, appear to be limited. It is important to recognize that not every HCC diagnosis occurs at an advanced stage; many patients progress to this stage after a variable clinical course from earlier stages. The distribution of tumor conditions at diagnosis varies based on the pattern of progression, which affects the likelihood of achieving the rCF status. The distribution of tumor conditions during the diagnosis varies based on the progression pattern, affecting the likelihood of achieving the rCF status. To the best of our knowledge, no reports have comprehensively analyzed the different liver function and tumor conditions of advanced-stage HCC and the achievement of rCF status. We believe that these findings underscore the importance of a multidisciplinary treatment approach for advanced-stage HCC and represent a significant advancement in its management.

In our large cohort, the prognosis of patients with advanced-stage HCC has improved over time. Before the approval of sorafenib, when systemic therapy was not the standard of care for advanced HCC, treatment consisted mainly of hepatic resection or transarterial therapy. However, with the introduction of sorafenib as the cornerstone systemic therapy for advanced HCC, systemic therapy has become the primary standard of care, and now combined immunotherapy is recognized as a first-line treatment approach [22]. Research by Kobayashi et al. [23] suggests that strategic sequencing of systemic therapies within a multidrug regimen can prolong treatment duration and significantly improve survival rates, particularly in cases of advanced-stage HCC. This is underscored by recent prospective clinical trials in which first-line treatment with a combination of immunotherapy and lenvatinib was shown to result in a median survival of approximately 20 months [24, 25]. These results demonstrate that advances in systemic therapy over the past decade have significantly improved the outlook for patients with advanced-stage HCC.

The present study defined rCF status based on the complete absence of detectable cancer in imaging examinations rather than on the presence or absence of biological cancer cells. This definition comprised two scenarios: patients without cancer were visualized in radiologic assessments (type A) and instances where residual cancer traces were present; however, arterial blood flow had ceased entirely (type B). The adopted definition, designated rCF, has not been reported in HCC. Approaches using imaging-based CR criteria have been used in several studies of other malignancies to assess treatment efficacy and predict outcomes. CR by imaging has been reported as radiologic CR in other cancer types [26‒28]. By adopting these concepts, our study aimed to adapt to the broader oncologic literature while acknowledging the limitations of imaging-based assessments. The validity of our definition was substantiated by two key analyses. First, the cumulative recurrence-free rate was almost equivalent between patients classified as cancer-free based on imaging and those exhibiting complete cessation of the arterial blood flow to the tumor site. Second, the significance of repeated radiologic assessments performed at predetermined intervals further corroborates the robustness of our definition. These results collectively support the clinical utility and prognostic value of our proposed definition of the rCF status in the context of this study.

In our study, the rate of achieving rCF status in advanced-stage HCC was low at 8.7%. Despite the low rate, the prognosis for patients who achieved this status was significantly favorable, providing important insights for treatment strategies aimed at achieving rCF status in patients eligible for systemic therapy. Combining atezolizumab and bevacizumab demonstrates a tendency toward elevated rCF rates in comparison to VEGF tyrosine kinase inhibitors, such as sorafenib and lenvatinib. This trend suggests the potential for improvement in rCF rates with advances in combination immunotherapy therapy. Moreover, immunotherapies are known to exhibit a “long-tail” effect, whereby a subset of patients experience durable responses. Recent studies have reported that the advent of combination immunotherapy facilitated 10–30% of patients who respond to treatment in clinical practice and can transition to curative therapies [21, 29]. It should be noted that the aforementioned cohorts differ from our own in that they either consisted solely of intermediate-stage patients or included approximately half intermediate-stage patients. Our data demonstrated very few patients transitioned from systemic therapy to curative treatment. This discrepancy might be largely attributed to two key characteristics of our study cohort. First, our analysis exclusively focused on patients with advanced-stage disease, without including earlier stages. Second, the majority of data presented in this study were derived from patients treated before the widespread clinical implementation of combination immunotherapy. Furthermore, our study defined rCF based on at least two imaging evaluations conducted at intervals of >2 months, which may have resulted in a lower incidence of patients meeting this rigorous definition. In parallel with these developments, as our results show, surgical resection alone or in combination with other therapies remains the most effective treatment for achieving the rCF status. The efficacy of resection has been previously reported [10‒13], even in cases with MVI or EHS in which systemic therapy is recommended in Asian guidelines [5‒7]. Our analysis further supports the critical role of resection in the management of advanced HCC and underscores its importance, even in the presence of MVI or EHS. As new immunotherapy regimens are developed, those that increase the proportion of patients achieving this long-tail effect may further improve rCF rates.

Our decision tree and multivariate analyses of patients with advanced-stage HCC revealed that achieving the rCF status was challenging in numerous (≥8) intrahepatic tumors. However, patients with EHS but without intrahepatic tumors, excluding bone metastases, demonstrated high rCF rates. These findings, in conjunction with those of Kim et al. [30] on the benefits of lung metastasis resection, indicate the possibility of employing more aggressive treatments in patients with a single EHS lesion (the so-called oligometastases). For MVI, aggressive resection might be indicated when intrahepatic tumors are limited, without bone metastases, and preserved liver function. Notably, rCF status achieved even when resection was initially contraindicated. In patients with bone metastases, prioritizing tumor control over radical resection might be more appropriate. These observations highlight the remarkable heterogeneity of advanced-stage HCC. In clinical settings, treatment decisions should be based on a precise assessment of the disease state, considering tumor distribution, extrahepatic metastases, and liver function. This comprehensive approach is essential for determining the optimal treatment strategy for each patient’s unique presentation of advanced HCC.

Conversely, the number of intrahepatic tumors is emerging as a critical determinant of treatment efficacy. Indeed, we previously demonstrated that patients with intermediate-stage HCC who have more than seven liver tumors had worse treatment outcomes [31, 32]. Extending this analysis to advanced-stage HCC, we observed a clear impact of tumor burden on treatment success. This trend was particularly evident in the analysis of patients’ disease progression pathways to advanced stage. Specifically, those who progressed to an advanced stage after passing through an intermediate-stage exhibited a higher number of tumors and a lower frequency of achieving rCF. This observation indicates that for patients diagnosed at early or intermediate stages who face a prolonged course of HCC treatment, one potential avenue for improving long-term prognosis may be the implementation of early intervention strategies to prevent an increased tumor burden. Currently, the TACE combined with immunotherapy are currently combined for early and intermediate stages of HCC. Should these therapeutic developments prove successful, they may contribute to an improved prognosis by preventing an increased tumor burden. Although not a primary factor in our decision tree analysis, univariate analysis underscored a significantly higher rCF rate in patients with MVI presenting with Vp2 or lesser degrees of invasion. Our results, combined with those of previous studies, underscore the importance of considering both the number of liver tumors and liver function when selecting aggressive treatments for patients with MVI, aiming for rCF status. This approach suggests a nuanced strategy for curative treatment of advanced HCC, particularly in patients with MVI, to optimize outcomes.

In this study, the determination of rCF status was based exclusively on imaging diagnostics, without any reliance on tumor markers. The analysis of rCF patients revealed no statistically significant difference in cumulative recurrence rates between patients with all three tumor markers within the normal range and those without. Notably, a considerable number of patients without all markers in the normal range exhibited AFP levels exceeding the upper normal limit. This phenomenon can be attributed to the intrinsic characteristics of AFP as a biomarker. In some patients with chronic hepatitis, AFP levels may elevate prior to the development of cancerous cells, and these values may not fully normalize even in the absence of recurrence. As AFP, AFP-L3, and PIVKA-II remain the most effective tumor markers currently available, further analysis based exclusively on these markers presents significant challenges. Nevertheless, an increasing focus on the potential of circulating tumor cells and cell-free DNA as diagnostic tools can determine the absence of cancer across a range of oncologic contexts, including HCC [33, 34]. Future developments may facilitate more accurate identification of true cancer-free status by integrating these novel techniques with imaging findings.

Our study highlights the heterogeneity of tumor conditions in advanced-stage HCC. Although advanced-stage HCC typically correlates with a poor prognosis, our research methodology has identified subsets of patients with potential for the rCF outcome and those for whom this remains an unattainable goal. This differentiation is critical for understanding the pathogenesis of advanced-stage HCC and developing targeted treatment strategies. The constraints of our study should be recognized. As a retrospective, single-center cohort study conducted at an academic university hospital, our findings are subject to institutional biases. Remarkably, the shift in patient referral patterns, from a paucity of advanced-stage HCC patients before 2009 to an increase following the implementation of systemic therapy as a standard practice, has introduced a temporal bias in patient numbers and characteristics. Despite the consistent application of a multidisciplinary approach, decision-making processes may have varied over time. The extended study period encompasses eras of evolving systemic therapies, which may limit the applicability of our findings to current paradigms. A notable limitation is the potential for inherent bias in prior treatment exposure among the four identified progression patterns (Fig. 6), which may have influenced the observed differences in rCF rates. The presence of discrepancies in patient backgrounds, including tumor characteristics, further complicates the result interpretation. Considering these factors, we have chosen to refrain from making direct statistical comparisons of rCF rates among these patterns. Notwithstanding these limitations, our study significantly advanced the understanding of advanced-stage HCC, a previously underexplored area. As the first study to investigate cancer-free status by radiologic assessments in a large cohort, it offers valuable insights into disease progression and treatment outcomes. The heterogeneity in progression patterns and prior treatments complicates the interpretation but highlights the complex nature of HCC progression and the need for personalized treatment approaches. Our study lays the groundwork for a more comprehensive understanding of advanced-stage HCC disease. These findings were expected to stimulate further research, particularly prospective, multicenter studies with large cohorts. These studies will be crucial in validating our results, enhancing understanding, and refining treatment strategies for this challenging disease.

In conclusion, the prognosis of patients with advanced-stage HCC has steadily improved, owing to advances in systemic therapy. Although the advent of combination immunotherapy offers encouraging prospects for increasing the rate of rCF outcomes, notable improvements for patients with advanced-stage HCC are pending. Specifically, the rate of achieving the rCF outcome varies widely among patients with advanced HCC, a variance that is largely attributed to tumor characteristics. The knowledge gained from this research provides a solid foundation for developing multidisciplinary treatment strategies aimed at achieving the rCF status in patients with advanced-stage HCC.

We are grateful to Satomi Nakamura, Yuka Iwase, and Ryoko Arai for their contributions to data management. The authors would like to thank Enago (www.enago.jp) for the English language review.

All procedures were conducted in accordance with the ethical standards of the Institutional and/or National Research Committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The Chiba University Research Ethics Committee approved this study (clinical trial No. 2247). Formal consent by written signature was not required for this type of study based on the Ethical Guidelines for Medical and Biological Research Involving Human Subjects in Japan.

Sadahisa Ogasawara received honoraria from Bayer, Leverkusen, Germany; Eisai, Tokyo, Japan; Eli Lilly, Indianapolis, IN, USA; Chugai Pharma, Tokyo, Japan; AstraZeneca, Cambridge, UK; and Merck & Co., Inc., Kenilworth, NJ, USA; consulting or advisory fees from Bayer, Eisai, Merck & Co., Inc., Chugai Pharma, Eli Lilly, and AstraZeneca; and research grants from Bayer, AstraZeneca, and Eisai. Naoya Kato received honoraria from Bayer; Eisai; Sumitomo Dainippon Pharma, Tokyo, Japan; and Merck & Co., Inc.; consulting or advisory fees from Bayer and Eisai; and research grants from Bayer and Eisai. The other authors have no conflicts of interest to declare.

This study was not supported by any sponsor or funder.

Keisuke Koroki and Sadahisa Ogasawara contributed substantially to the study conceptualization. Masayuki Yokoyama provided guidance on statistical analysis. Ryo izai, Takuya Yonemoto, Teppei Akatsuka, Chihiro Miwa, Sae Yumita, Kisako Fujiwara, Masanori Inoue, Kazufumi Kobayashi, Soichiro Kiyono, Masato Nakamura, Naoya Kanogawa, Takayuki Kondo, Shingo Nakamoto, Shinichiro Nakada, Nozomu Sakai, Masayuki Ohtsuka, and Naoya Kato contributed substantially to data acquisition and interpretation. Keisuke Koroki and Sadahisa Ogasawara contributed to the manuscript drafting. Naoya Kato supervised the study. All the authors critically reviewed and revised the draft manuscript and approved the final version for submission.

All data generated or analyzed during this study are included in this article and its online supplementary material files. Further inquiries can be directed to the corresponding author.