Global scientific trends on hepatocellular carcinoma research from 2004 to 2023: A bibliometric and visualized analysis

Abstract

BACKGROUND

Hepatocellular carcinoma (HCC) is a major cause of cancer-related mortality worldwide, and the research landscape has rapidly evolved over the past two decades. Despite significant progress, an in-depth analysis of global research trends, collaborative networks, and emerging themes in HCC remains limited. This study aimed to fill this gap by conducting a bibliometric analysis to map the research output, identify key contributors, and highlight future directions in HCC research. We hypothesized that the analysis would reveal a growing focus on molecular mechanisms and immunotherapy, with increasing contributions from specific countries and institutions.

AIM

To investigate global research trends, collaborative networks, and emerging themes in HCC from 2004 to 2023.

METHODS

A bibliometric analysis was performed using 93987 publications from the Science Citation Index Expanded Database of the Web of Science Core Collection. Data were analyzed using the VOSviewer software to identify publication trends, leading contributors, and research themes. Key metrics included annual publication output, country and institutional contributions, journal impact, and thematic clusters. Statistical analysis was carried out to quantify trends and collaborations.

RESULTS

The number of annual publications increased from 2341 in 2004 to 8756 in 2023, with 65583 papers (69.78%) published between 2014 and 2023. China, the United States, and Japan were the top contributors, constituting 58.3% of total publications. PLOS One published the most studies (n = 2145), while Gastroenterology had the highest average number of citations (78.4 citations per paper). Fudan University was the most prolific institution (n = 1872). Thematic analysis identified five main clusters, namely molecular mechanisms, therapeutic strategies, prognosis and immunology, risk factors, and diagnostic approaches.

CONCLUSION

This study highlights the growing focus on HCC research, particularly in immunotherapy and molecular mechanisms, underscoring the significance of international collaboration to advance diagnosis and treatment strategies.

Key Words: Hepatocellular carcinoma; Bibliometric analysis; VOSviewer; Research trends; Oncology

Core Tip: This study offers a comprehensive bibliometric and visual analysis of hepatocellular carcinoma research from 2004 to 2023, covering 93987 publications. It identified key trends, influential journals, leading countries and institutions, and international collaboration patterns. The analysis highlighted evolving research hot spots, including molecular mechanisms, diagnosis, therapeutic strategies, and immunology. Notably, immune-related studies have increasingly integrated with prognostic research in recent years, reflecting a shift toward personalized treatment. These findings provide valuable insights into the development of hepatocellular carcinoma research and its future clinical applications.

INTRODUCTION

Liver cancer mainly comprises hepatocellular carcinoma (HCC) and intrahepatic cholangiocellular carcinoma, with HCC accounting for most liver cancers. According to the World Health Organization estimates, HCC ranks as the fifth most prevalent cancer and is the fourth leading cause of cancer-related deaths globally[1]. The prognosis for HCC patients is generally poor, with a mortality rate nearly matching its worldwide incidence[2]. Due to the insidious onset of the disease, the majority of patients are in the middle to late stages of disease at the time of diagnosis. Cirrhosis and viral hepatitis are the most important risk factors for HCC, and there has been some success in eliminating the risks associated with the major risk factors. Nevertheless, the increased prevalence of other risk factors rises the incidence of HCC worldwide. Recently, nonalcoholic fatty liver disease (NAFLD) and nonalcoholic steatohepatitis (NASH) have emerged as the primary causes of HCC[3]. With the approval of new first- and second-line drugs and the refinement of the standard of care for immune checkpoint inhibitor (ICI)-based therapies, the outlook for the treatment of HCC has somewhat improved. However, more effective systemic therapies and predictive biomarkers are still needed for effective detection and prevention strategies[4]. Therefore, given the widespread interest in HCC, it is crucial to find more impactful publications to deepen our understanding of the molecular mechanisms, risk factors, and diagnosis and treatment of HCC.

Bibliometrics, a method that can combine mathematical and statistical methods with data visualization, has widely been used in the medical field[5,6]. However, none of the previous studies have applied bibliometric analysis to the entirety of HCC research, and most studies have only focused on a specific direction of HCC research. Therefore, in the current study, we employed comprehensive bibliometric analysis of HCC research using VOSviewer to analyze HCC-related literature from 2004 to 2023. The objectives were to reveal the publication trends in HCC research over the past two decades; identify influential journals, countries, and institutions; and explore international research collaborations, hot spots, and emerging themes.

MATERIALS AND METHODS

Data source and search strategy

The first step in this study was to define the field of science and the data to be used from the number of scientific networks databases. As a database of the latest data sources, it covers approximately 250 discipline categories across science, social sciences, art, and humanities, including many medical documents. In the present study, all of the obtained publications were retrieved from the Science Citation Index Expanded Database in the Web of Science Core Collection. The following search string was used in this study: Title = (“Hepatocellular Carcinoma$” or “Hepatoma$” or “Liver Cell Carcinoma$” or “Hepatocarcinoma$” or “Hepatic Carcinoma$” or “Hepatocellular neoplasm$” or “Hepatocellular Cancer$”) or Author Keyword = (“Hepatocellular Carcinoma$” or “Hepatoma$” or “Liver Cell Carcinoma$” or “Hepatocarcinoma$” or “Hepatic Carcinoma$” or “Hepatocellular neoplasm$” or “Hepatocellular Cancer$”) or Keyword Plus = (“Hepatocellular Carcinoma$” or “Hepatoma$” or “Liver Cell Carcinoma$” or “Hepatocarcinoma$” or “Hepatic Carcinoma$” or “Hepatocellular neoplasm$” or “Hepatocellular Cancer$”). We set the scope of document publishing from January 1, 2004, to December 31, 2023. The initial search yielded a total of 188575 records (86270 + 46033 + 56272). After removing duplicates, 144025 unique records remained. Afterward, 1349 non-English publications were excluded, leaving 142676 records. Further exclusions were made based on document type, removing 48779 records, including meeting abstracts, review articles, and other nonrelevant materials. This resulted in a final dataset of 93987 publications included in the bibliometric analysis. We downloaded the entire literature measurement data in the “Full Records and City References” format on August 9, 2024, so as to avoid the impact of database updates. Figure 1 displays the overall search strategy.

Figure 1 Flow diagram of search strategy. A total of 93987 articles retrieved from the Web of Science Core Collection were subjected to bibliometric analysis.

Data analysis and visualization

As classic literature measurement analysis software, VOSviewer is extensively used in visual literature analysis and research[7]. Using VOSviewer 1.6.20, in this study, we conducted citation analysis of journals, coauthorship analysis of institutions, coauthorship analysis of references, and co-occurrence analysis of keywords. We also conducted a descriptive analysis on publishing years, journals, countries, institutions, references, and funding agencies. All of the original data are available upon reasonable request.

RESULTS

Annual publication

The search yielded 93987 pertinent publications, with Figure 2 depicting the trend in annual publications from 2004 to 2023. The publication trend is shown in red. There were 1785 publications on HCC research in 2004, compared with 6680 publications in 2023. In 2021, the number of publications was 7724 - 4.32 times higher than the number of publications in 2004 - which reached the annual peak in publication volume. The yearly count of publications on HCC research showed an overall fluctuating upward trend over the 20-year period. We also compared the number of publications between 2004-2013 and 2014-2023, with 28404 and 65583 publications identified, respectively, accounting for 69.78% of all publications from 2004 to 2023. The number of publications in the latter decade was 2.30 times higher than that in the former. Although the number of publications has declined from 2022 to 2023, this area still needs to be given broader attention when considering the overall number of publications.

Figure 2 Annual global publications on hepatocellular carcinoma from 2004 to 2023. The annual number of publications exhibited a fluctuating upward trend overall, with the red line representing the trend in publication volume.

Contributions of countries and regions

From 2004 to 2023, authors from 160 countries/regions were involved in HCC research, and their global distribution is shown in Figure 3. Authors from Central Asia, North America, and Europe published more articles, whereas those from Africa published fewer. Furthermore, 45 countries/regions had more than 100 publications, and three countries had more than 10000 publications, among which China (n = 42902) ranked the first, followed by the United States (n = 16848) and Japan (n = 10862). Despite occupying the second place, the United States was still far from China (more than twice).

Figure 3 The geographical distribution and visualization of productive countries/regions. Global distribution of publications between 2004-2023. Color gradient from blue to red indicates increasing publication volume from low to high.

Table 1 presents the top 20 countries/regions with the highest publication counts from 2004 to 2013 and from 2014 to 2023. The top 20 countries did not change much overall, with China, the United States, and Japan consistently ranking the first, second, and third, respectively. Two new countries/regions entered the top 20 in the last decade, namely Saudi Arabia and Brazil. China had the largest increase in the number of publications, with 27782 articles, representing an increase of 3.67 times, far exceeding the overall growth rate. Between 2014 and 2023, China’s ratio of publications and citations to those of the United States, Japan, and South Korea was much higher than that between 2004 and 2013. Additionally, Egypt moved from the 17^th to the 10^th place. The number of publications increased by 988 articles in the last decade, representing a growth rate of 3.55 times, which was also much higher than the overall growth rate. The United States economy grew by 0.71 times and that of Japan grew by 0.10 times - much slower than the overall economy growth rate. Although China had a high number of citations (n = 771552), its citation/publication ratio (n = 21.8) was still low. From 2004 to 2013, Spain had the highest citation/publication ratio (n = 64.9) among the 20 countries/regions, although Spain published a relatively small number of papers, which is indicative of the high quality of its publications. China led HCC research with over 42000 publications, doubling the United States output. Its growth (2014-2023) far outpaced that of the United States and Japan. Egypt and Saudi Arabia showed significant progress, while Spain achieved the highest citation ratio, indicating high-quality research. The top 20 contributors remained stable, with Brazil and Saudi Arabia joining recently.

Table 1 Top 20 productive countries/regions on hepatocellular carcinoma from 2004 to 2013 and 2014 to 2023.

Rank	Country/region, 2004-2013	Counts, 2004-2013	Citations, 2004-2013	Average Citation, 2004-2013	Country/region, 2014-2023	Counts, 2014-2023	Citations, 2014-2023	Average citation, 2014-2023
1	China	7560	355391	47.0	China	35342	771552	21.8
2	United States	6209	455449	73.4	United States	10639	429837	40.4
3	Japan	5159	220507	42.7	Japan	5703	154556	27.1
4	South Korea	2099	100453	47.9	South Korea	3821	109114	28.6
5	Taiwan	1961	104682	53.4	Taiwan	3008	93549	31.1
6	Germany	1713	112577	65.7	Germany	2794	105542	37.8
7	Italy	1671	117403	70.3	Italy	2712	112447	41.5
8	France	1130	90151	79.8	France	1890	114798	60.7
9	United Kingdom	800	66208	82.8	United Kingdom	1647	86530	52.5
10	Spain	623	60447	97.0	Egypt	1266	21957	17.3
11	Canada	605	47512	78.5	Spain	1156	75027	64.9
12	India	369	16565	44.9	Canada	1105	48454	43.8
13	Australia	365	29447	80.7	India	1028	25435	24.7
14	Singapore	322	23992	74.5	Australia	796	32681	41.1
15	Switzerland	316	32718	103.5	Switzerland	749	44638	59.6
16	Netherlands	285	21782	76.4	Singapore	687	30524	44.4
17	Egypt	278	7396	26.6	Turkey	662	15840	23.9
18	Turkey	248	5655	22.8	Netherlands	650	25842	39.8
19	Belgium	229	20888	91.2	Saudi Arabia	569	9031	15.9
20	Sweden	207	13899	67.1	Brazil	552	12688	23.0

Contributions of journals

During the 20-year research period, a cumulative total of 3275 journals published research articles on HCC. In compliance with Bradford’s law, we set the threshold of journal publications at 370 to obtain 44 high-yield journals as core-area journals[8]. The 44 high-yield journals published 31185 articles, accounting for 33.18% of all of the publications. An overlay visualization map (Figure 4) was constructed based on a citation analysis of the 44 highly productive journals. The diameter of each circle in Figure 4 indicates the publication count, while the color spectrum, ranging from blue to red, indicates the average citation count, with blue signifying lower and red denoting higher citation rates. PLOS One (n = 1961) was the most productive journal, followed by World Journal of Gastroenterology (n = 1763), Oncotarget (n = 1520), and Hepatology (n = 1500). Thirteen journals (shown in red, Figure 4) had on average more than 40 citations, with Gastroenterology (n = 189) having the highest average citation count, followed by Hepatology (n = 113) and Journal of Hepatology (n = 109).

Figure 4 Network visualization map of the 44 most productive journals on hepatocellular carcinoma. Visualization map of publications across 44 journals. Each node represents an individual journal, where the circle size corresponds to the publication volume (larger circles indicate higher publication counts). Node color represents citation frequency, with a gradient from blue to red indicating increasing average citation counts from low to high. The thickness of the links denotes the collaboration strength, with thicker lines indicating a higher degree of cooperative activity between the two journals.

Table 2 presents the top 20 journals, ranked according to their publication count. The journals ranking in the top 20 in the previous decade had 7768 articles, constituting 27.34% of the total number of publications in that period, and the top 20 journals in the last decade published 14801 articles, accounting for 22.57% of the total number of publications in the last decade, highlighting the core influence of these journals. Twelve journals entered the top 20 in the last decade. The number of HCC publications published by Oncotarget, Frontiers in Oncology, and Scientific Reports increased rapidly over the last decade. These three journals were founded after 2010 and quickly became the top 10 in terms of publications in the last decade, reflecting their greater interest and focus on HCC. PLOS One was a particularly prolific journal in the field. During the past two decades, despite the relatively low number of publications in Hepatology and Journal of Hepatology, their average citations have been consistently high, emphasizing their influence and quality of publications in the field. Among the 3275 journals publishing HCC research, the 44 high-yield journals contributed 33.18% of the publications. PLOS One led in productivity, while Gastroenterology, Hepatology, and Journal of Hepatology had the highest citation number, reflecting their influence. In the last decade, 12 new journals entered the top 20, with Oncotarget and Frontiers in Oncology showing rapid growth. Despite fewer publications, Hepatology and Journal of Hepatology maintained high citation rates, highlighting their impact.

Table 2 Top 20 productive Journals on hepatocellular carcinoma from 2004 to 2013 and 2014 to 2023.

Rank	Journal, 2004-2013	Counts, 2004-2013	Citations, 2004-2013	Average citation, 2004-2013	Journal, 2014-2023	Counts, 2014-2023	Citations, 2014-2023	Average citation, 2014-2023
1	World Journal of Gastroenterology	1021	28851	28.3	Oncotarget	1503	48089	32.0
2	Hepatology	744	96533	129.7	PLOS One	1258	29416	23.4
3	PLOS One	703	33594	47.8	Scientific Reports	1188	27195	22.9
4	Hepato-Gastroenterology	569	6518	11.5	Frontiers in Oncology	1109	10136	9.1
5	Journal of Hepatology	508	47909	94.3	Oncology Letters	1002	13871	13.8
6	Oncology Reports	400	11305	28.3	Cancers	792	8688	11.0
7	Hepatology Research	363	9757	26.9	Hepatology	756	73071	96.7
8	Journal of Gastroenterology and Hepatology	346	12074	34.9	World Journal of Gastroenterology	742	20879	28.1
9	Liver International	335	13490	40.3	BMC Cancer	680	14145	20.8
10	International Journal of Cancer	282	18689	66.3	Medicine	665	8256	12.4
11	International Journal of Oncology	278	9393	33.8	Oncology Reports	605	13853	22.9
12	Biochemical and Biophysical Research Communications	277	10296	37.2	Hepatology Research	546	10326	18.9
13	Cancer Letters	273	12845	47.1	Tumor Biology	544	12421	22.8
14	Journal of Vascular and Interventional Radiology	264	12117	45.9	International Journal of Molecular Sciences	533	8025	15.1
15	American Journal of Roentgenology	259	12377	47.8	Molecular Medicine Reports	518	8555	16.5
16	Clinical Cancer Research	251	24477	97.5	Journal of Hepatology	510	63932	125.4
17	Cancer Research	242	33590	138.8	Oncotargets and Therapy	497	8634	17.4
18	Annals of Surgical Oncology	223	13574	60.9	Liver International	467	12675	27.1
19	Gastroenterology	218	44010	201.9	Cell Death & Disease	444	17814	40.1
20	Anticancer Research	212	5163	24.4	Biochemical and Biophysical Research Communications	442	11284	25.5

Contributions of institutions

We set the threshold of an institution’s publication count at 200 and identified 173 highly productive institutions among the total of 33870 institutions. We used VOSviewer to perform a coauthorship analysis on the 173 highly productive institutions and constructed an overlay visualization map (Figure 5). The diameter of each circle shown in Figure 5 indicates the publication count, while the color transition from blue to red indicates the average citation count, ranging from lower to higher values. Fudan University (n = 3188) was the most productive institution, followed by Sun Yat-sen University (n = 3169), Zhejiang University (n = 1963), and Shanghai Jiao Tong University (n = 1865). Seventeen institutions, highlighted in red in Figure 5, had an average citation count over 70, with Kindai University (n = 117) being the most frequently cited on average, followed by the University of Barcelona (n = 112), Paris Descartes University (n = 94), and the University of California, Los Angeles (n = 92). Fudan University had the highest number of collaborations with the highly productive institutions at 150. It was followed by the University of Hong Kong with 148 collaborations with the high-yielding institutions. Harvard University had 142 collaborations with the high-yielding institutions, and the Chinese University of Hong Kong had 138 collaborations with the high-yielding institutions.

Figure 5 Network visualization map of the 173 most productive institutions on hepatocellular carcinoma. Visualization map of publications from 173 high-yield institutions. Each node represents an individual journal, with circle size proportional to publication volume (larger circles denote higher output). Node coloration reflects citation frequency, transitioning from blue to red to indicate increasing average citation rates from low to high. The thickness of the links represents the strength of collaboration, with thicker lines indicating more intensive cooperation between two institutions.

The 20 institutions with the highest number of publications, sorted by the number of publications from 2004 to 2013 and from 2014 to 2023, are listed in Table 3. Fudan University and Sun Yat-sen University have always been prolific in this field. Sun Yat-sen University has been the fastest-growing institution in the last decade, with 3.95 times more publications in the latter decade than in the former decade. Comparing the top 20 institutions between the two time periods, considerable changes were noted. Twelve institutions newly entered the global top 20 for HCC research in the last decade, all of which were Chinese institutions. Between 2014 and 2023, a total of 19 Chinese institutions were included in the top 20. Despite the high volume of publications from the Chinese institutions, their average citation value was low, suggesting that the quality of publications from these institutions still needs to be improved. The institution with the highest average citation from 2004 to 2013 was the National Cancer Institute (n = 119.1), and that with the highest average citation from 2014 to 2023 was Seoul National University (n = 35.8). Fudan University and Sun Yat-sen University led the way in HCC research, with Sun Yat-sen showing the fastest growth in the last decade. Seventeen institutions, including Kindai University and the University of Barcelona, achieved high average citations, reflecting the quality of research. Although Chinese institutions dominated the top 20 in 2014-2023, their low citation rates indicate that quality needs to be improved. The National Cancer Institute and Seoul National University had the highest average citations in their respective decades.

Table 3 Top 20 productive institutions on hepatocellular carcinoma from 2004 to 2013 and 2014 to 2023.

Rank	Institution, 2004-2013	Counts, 2004-2013	Citations, 2004-2013	Average citation, 2004-2013	Institution, 2014-2023	Counts, 2014-2023	Citations, 2014-2023	Average citation, 2014-2023
1	Fudan University	768	51212	66.7	Sun Yat-sen University	2637	65851	25.0
2	National Taiwan University	573	45852	80.0	Fudan University	2420	68376	28.3
3	Sun Yat-sen University	532	38484	72.3	Zhejiang University	1667	38404	23.0
4	University of Hong Kong	488	41159	84.3	Shanghai Jiao Tong University	1518	44882	29.6
5	Seoul National University	487	22224	45.6	Huazhong University of Science and Technology	1220	25413	20.8
6	The Second Military Medical University	482	26288	54.5	Nanjing Medical University	1203	31945	26.6
7	The University of Tokyo	431	38423	89.1	The Second Military Medical University	1129	38166	33.8
8	The Chinese University of Hong Kong	410	28461	69.4	Southern Medical University	1103	24952	22.6
9	National Yang Ming Chiao Tung University	358	17093	47.7	Guangxi Medical University	1081	17509	16.2
10	Yonsei University	355	20531	57.8	China Medical University	1036	21983	21.2
11	Shanghai Jiao Tong University	347	18708	53.9	Zhengzhou University	1015	20239	19.9
12	Chang Gung University	335	17672	52.8	Sichuan University	1008	19610	19.5
13	National Cancer Center	309	18556	60.1	Chinese Academy of Sciences	1007	32901	32.7
14	Chinese Academy of Sciences	301	16517	54.9	Shandong University	982	19957	20.3
15	National Cancer Institute	299	35623	119.1	Central South University	980	23574	24.1
16	Peking University	299	12927	43.2	Xi’an Jiaotong University	895	22981	25.7
17	Osaka University	296	16674	56.3	Capital Medical University	894	17993	20.1
18	Zhejiang University	296	16308	55.1	Seoul National University	862	30835	35.8
19	Sungkyunkwan University	287	15636	54.5	Peking University	821	20200	24.6
20	Johns Hopkins University	276	20132	72.9	Fujian Medical University	759	12370	16.3

Analysis of influential references

We set the threshold of the number of references being cited at 700 and identified 70 highly cited references out of the 1234282 references. We used VOSviewer to perform co-citation analysis on the 70 highly cited references and constructed a density visualization map, as shown in Figure 6. The colors from green to red represent the references cited from low to high. The most cited reference titled “Global cancer statistics” was published by Jemal et al[9] in 2011 with 6013 citations, followed by Llovet et al[10] in 2008, with 5521 citations. The use of sorafenib in HCC has recently attracted considerable attention. This topic was also mentioned in the third most cited article, Bruix et al[11]. It is noteworthy that articles by Josep M Llovet, Jordi Bruix, and Hashem B El-Serag have received many citations, demonstrating their significant contributions to the field of HCC.

Figure 6 Density visualization map of the top 70 co-cited references on hepatocellular carcinoma. Density visualization of the reference co-citation network. The collaborative network comprises 70 publications, with color gradients from green to red representing increasing citation frequencies from low to high.

Keyword analysis

The 93987 publications included 76288 author keywords. We set a threshold of 80 for high-frequency author keywords and obtained 391 high-frequency author keywords. Co-occurrence analysis of these 391 high-frequency keywords was performed, and the resulting co-occurrence network map of the high-frequency author keywords was constructed (Figure 7). This analysis revealed five distinct clusters, each of which was denoted by a unique color. Cluster 1 (red), cluster 2 (yellow), cluster 3 (purple), cluster 4 (green), and cluster 5 (blue) focused on the research topics, such as molecular mechanism, risk factors, diagnosis, therapeutic strategies, prognosis, and immunology, respectively. The searched keyword “hepatocellular carcinoma” was removed, and the following were identified as the 10 most frequent author keywords: Prognosis, apoptosis, hepatitis B virus (HBV), hepatectomy, transarterial chemoembolization, fibrosis, neoplasm metastasis, proliferation, sorafenib, and biomarkers.

Figure 7 Network visualization map of keywords co-occurrence analysis on hepatocellular carcinoma. Co-occurrence network map comprising 5 clusters (nodes with identical colors) formed by 391 high-frequency author keywords. Node labels display keywords, with node size proportional to keyword frequency, where larger nodes indicate higher occurrence rates.

We also constructed an overlay graph of the 391 high-frequency author keywords (Figure 8). In Figure 8, “Avg. pub. year.” refers to the average publication year of the documents in which a keyword occurs. The color of each node reflects the “Avg. pub. year.” of each keyword. Atezolizumab, immune infiltration, immune microenvironment, ICIs, and ferroptosis are all colored red in Figure 8, signifying that these research topics have garnered significant attention in recent years and have the potential to emerge as novel hot spots in the coming years. In contrast, the blue nodes in the graph represent keywords that have lost popularity over time, such as hepatitis and ribavirin. Analysis of the 391 high-frequency keywords identified five research clusters, namely molecular mechanisms, risk factors, diagnosis, therapy, and immunology. Top keywords included prognosis, apoptosis, and sorafenib. Emerging hot spots such as atezolizumab and ferroptosis (red) indicate recent trends, while older topics such as hepatitis (blue) have declined.

Figure 8 Overlay visualization map of keywords co-occurrence analysis on hepatocellular carcinoma. Co-occurrence overlay map constructed from 391 high-frequency author keywords, with node labels displaying corresponding keywords. Node coloration represents the average publication year, transitioning from blue to red to indicate progressively more recent publication dates.

DISCUSSION

Global research trends

Our analysis of the annual publications on HCC from 2004 to 2023 showed an overall fluctuating upward trend in terms of the number of publications per year, peaking in 2021. The number of publications from 2014 to 2023 was 2.3 times higher than the number of publications from 2004 to 2013, and the number of publications on HCC has increased considerably in recent years. The rising trend in HCC research is driven by several key factors as follows: The growing global burden of HCC due to an increasing prevalence of risk factors such as hepatitis B/C, NAFLD, and lifestyle changes; significant advancements in diagnostic tools and therapies; and expanded funding and international collaborations, particularly in high-prevalence regions. Additionally, the shift toward precision medicine and heightened awareness of HCC as a major public health issue have further accelerated research efforts.

Authors from China, the United States, and Japan have made significant contributions to HCC research through their numerous publications. Although China leads in the number of publications, its relatively lower citation rate warrants further consideration. This phenomenon may be attributed to factors such as disparities in research quality, language barriers, and citation practices. Moreover, the global impact of research originating from China needs to be enhanced, which could be achieved by increasing international visibility and fostering collaborative efforts. The low number of publications in Africa compared with other continents may be related to the lack of necessary infrastructure, funding, and researcher experience. Addressing the relevant factors is essential to advancing academic research in this region. We compared publication counts across countries for the time periods 2004-2013 and 2014-2023. In addition to the finding that China, the United States, and Japan held the top three positions, we found that the number of publications in China and Egypt was growing very rapidly, both at more than three times the overall rate of growth. Countries such as China, Japan, and Egypt have a high prevalence of HCC due to endemic hepatitis B/C infections and other risk factors such as aflatoxin exposure and NAFLD. This has prompted significant research efforts in these regions. The quality of Spain’s publications in this area has been consistently high. In addition to an increasing interest in research, a robust research infrastructure and extensive financial support have played an important role here.

High co-citation among journals indicates strong thematic connections and intellectual influence within the research field. Journals with high co-citation frequencies often represent core sources of knowledge that shape the research landscape, reflecting their crucial role in disseminating key findings and fostering interdisciplinary dialogue. This visualization helps identify central journals that drive innovation and collaboration in the field. The most frequently cited and influential journals in the field of HCC are Hepatology and Journal of Hepatology, both of which are known for their high citation rates and influence in the field, and are widely recognized by researchers in the field. PLOS One stands out as the journal with the highest number of publications, and most of the articles related to this field are considered for publication in this journal. Oncotarget, Frontiers in Oncology, and Scientific Reports have grown rapidly in terms of publications in the field of HCC in the last decade. Journals such as PLOS One and Scientific Reports have gained prominence due to their open-access models, which increases visibility and citation rates. The emergence of newer journals, such as Frontiers in Oncology, reflects the growing interest in HCC research and the expansion of specialized oncology publications.

International cooperation

From 2004 to 2023, Fudan University was successful in the field of HCC, with the highest number of publications and the closest collaboration and communication with other institutions. In addition to Fudan University, the University of Hong Kong, Harvard University, and the Chinese University of Hong Kong are centers of global institutional collaboration. Functioning as hubs for international collaboration, these institutions tend to achieve high citation rates and are likely to produce research of superior quality. By examining the top 20 high-productivity institutions in the past 20 years, we found that almost all of the high-productivity institutions in the past decade were from China. This also corresponds to the results of our analysis of the number of publications in countries and the rapid development of HCC in China in the past decade. Although institutions from China have more publications, Seoul National University has the highest average citation (n = 35.8), indicating that this institution has published higher-quality papers in this area of HCC and has been instrumental in the advancement of this field. In contrast, some academic institutions in China still need to strengthen cooperation and exchanges with other institutions.

Research hot spots and frontiers

Through cluster analysis of the 391 high-frequency author keywords that appeared over 80 times, we categorized these keywords into five primary clusters as follows: Cluster 1 (red cluster), molecular mechanisms; cluster 2 (yellow cluster), risk factors; cluster 3 (purple cluster), diagnosis; cluster 4 (green cluster), therapeutic strategies; and cluster 5 (blue cluster), prognosis and immunology. These keywords in co-occurrence analysis can indicate the primary research focus and popular topics being explored by researchers in this field. Through an examination of the overlay plot depicting the keywords of high-frequency authors, we observed that the keywords pertaining to immunity and prognosis appeared more frequently among the keywords with newer average publication years (red, Figure 8), indicating that immune-related research has increasingly gained prominence in recent years, frequently merging seamlessly with prognostic assessments in such studies. Next, we analyzed the aforementioned five clusters and discussed the prevalent research hot spots and emerging directions within the field of HCC.

Cluster 1 (red): Molecular mechanisms

Different modes of programmed cell death, such as pyroptosis, apoptosis, and necrosis, can contribute to the progression of liver disease through different mechanisms. Hepatocyte apoptosis can be mechanistically divided into an intrinsic pathway triggered through the b-cell lymphoma 2 family and an extrinsic pathway activated through the tumor necrosis factor family[12]. Contrary to common perception, hepatocyte apoptosis may not be very inert, and its progression may lead to HCC[13]. Preventive strategies in the early stages of HCC mainly interfere with and inhibit cell death. In treating HCC that has already occurred, cell death needs to be promoted in the cancerous tissue[14]. Hence, additional methods to induce specific HCC cell death are currently being investigated. Ketoconazole stimulates apoptosis in HCC cells by triggering mitochondrial phagocytosis in vitro and in vivo and has demonstrated significant antitumor effects both independently and along with sorafenib[15]. In a dose- and time-dependent fashion, ropivacaine induces apoptosis in HCC cells, and this effect also activates the caspase-3 signaling pathway by participating in the disruption of mitochondrial function in HCC cells[16]. Numerous studies have explored the molecular mechanism of HCC development through apoptosis, which also provides new research ideas for developing therapeutic strategies.

Network pharmacology is a discipline that studies drug-organism interactions using network analysis, systems biology, and pharmacology. The analysis reveals information about the mechanism of pharmacological action, the therapeutic target, and the therapeutic efficacy of drugs by constructing and analyzing multilevel biological networks, such as drug-target-disease[17]. This keyword has gained more attention in recent years, showing a red color in the overlay graph with an average publication year of 2,021.73. Network pharmacology has changed our dogmatic approach of “one disease-one target-one drug”[18]. This effective targeting of etiological modules will help achieve the goal of precision medicine.

Cluster 2 (yellow): Risk factors

The primary risk factors of HCC encompass cirrhosis, viral hepatitis, alcohol addiction, NAFLD, and exposure to food toxins. Viral hepatitis is the most important factor for HCC, with HBV and hepatitis C virus (HCV) infections causing 80% of HCC cases worldwide[19]. HBV infection causes approximately 55% of all HCC cases, and the incidence rate of HCC annually among patients with HBV- or HCV-induced cirrhosis ranges from 2% to 5%[20]. Patients with hepatitis are at a significant risk of developing HCC, with cirrhosis being a primary contributing factor. Hepatitis is the most common pathogenic cause of cirrhosis. Individuals infected with HCV may have a heightened propensity for developing cirrhosis compared with those with HBV, making the prevention of hepatitis particularly important[21]. The moderate preventability of these risk factors demonstrates the considerable potential of risk prevention in reducing the incidence of HCC and the need for effective tertiary preventive measures[22]. Although there is currently no effective primary prevention vaccine for HCV, researchers are still exploring new prophylactic agents and noninvasive markers to use for prediction[23]. Additionally, NAFLD and NASH have become one of the fastest-growing causes of HCC in recent years, with 20%-30% of NAFLD- and NASH-associated HCC cases occurring in the absence of cirrhosis[24]. Although the incidence of NAFLD-associated HCC is lower than that of HCC of other etiologies, such as viral hepatitis, NAFLD-associated HCC tends to be diagnosed at a later stage and has a poorer prognosis than HCC resulting from other liver diseases[25]. Therefore, early prevention strategies are necessary to address this growing problem.

Cluster 3 (purple): Diagnosis

Artificial intelligence is currently a widely used computer technology, and its applications in medicine mainly include two categories, namely imaging omics and deep learning[26,27]. Imaging histology is a machine-learning technology generated by the combination of medical data and artificial intelligence, which applies data representation algorithms to extract and process a large number of features from radiological images and model them to assist in disease diagnosis and prognosis[28]. Owing to rapid advancements in deep learning, the convolutional neural network technique in deep learning and medical images now fit better than before and outperform imaging histology in terms of accuracy[29]. Currently, deep learning is used in the field of HCC to construct models for predicting the risk of HCC occurrence, to predict the risk of postoperative recurrence and survival of HCC[30], and to predict the survival risk of radiofrequency ablation[31] and transcatheter hepatic artery chemoembolization[32]. In the era of continuous updating of information technology, the application of various models constructed by deep learning in the field of HCC research should be more in-depth and exert a more substantial influence on the diagnosis and treatment of HCC. Moreover, these tools leverage machine-learning algorithms and natural language processing techniques to accurately identify research hot spots, predict future trends, and generate visual maps that reveal key research focuses and developmental trajectories within the field. By further analyzing interdisciplinary connections and citation networks, as well as identifying high-impact research, they provide real-time monitoring and comprehensive analytical support for HCC researchers, thereby advancing scientific discovery and innovation.

Cluster 4 (green): Therapeutic strategies

Hepatectomy, a strategy for HCC with curative potential, is generally used in patients with HCC without cirrhosis who can tolerate surgery. Considering the surgical indications for hepatic resection and the assessment of survival and risk, the application of other therapeutic strategies is particularly important[33]. Our keyword analysis revealed that atezolizumab had been the most prominent emerging hot spot in the field of HCC in recent years (Figure 8), with a mean publication year of 2022. IMbrave050, the first phase 3 study reporting positive results for adjuvant therapy in HCC, found that atezolizumab (targeting programmed death ligand 1) in combination with bevacizumab (targeting vascular endothelial growth factor), a T + A regimen, achieved improved recurrence-free survival compared with active surveillance among patients with an elevated risk of HCC recurrence[34]. The latest IMbrave150 study demonstrated the efficacy of this combination therapy in patients with unresectable HCC, exhibiting an acceptable safety profile and superior overall and progression-free survival outcomes compared with sorafenib treatment[35]. Dual blockade of vascular endothelial growth factor and programmed death ligand 1 may provide greater survival benefit to patients. The combination therapy of atezolizumab and bevacizumab has emerged as a significant advancement in clinical practice. Currently, the research focus has progressively transitioned from conventional therapies, such as sorafenib, to immunotherapy and precision medicine. The significance of ICIs and targeted therapies in HCC research has been increasingly emphasized[36].

Cluster 5 (blue): Immunology and prognosis

So far, numerous studies have shown that the pathology of HCC is influenced by interactions between hepatocytes and their surrounding microenvironment, leading to the transformation of hepatocytes to HCC, driving cancer progression, and ultimately lineage commitment[37]. The tumor immune microenvironment (TIME) of HCC includes various components, such as immune cell infiltrates, non-immune stromal constituents, exogenous cytokines, and exogenous chemokines. Cytokines attract regulatory T cells that infiltrate tumor tissue and construct the immunosuppressive microenvironment[38]. With a greater understanding of the role of TIME in the initiation and progression of HCC, the identification of prognostically relevant TIME subtypes by evaluating the phenotype, diversity, spatial arrangement, and functional activities of infiltrating cells has become pivotal in both basic and clinical research. The HCC immune landscape has also emerged as an important prognostic factor[39].

In 2012, Dixon first proposed a new cell death mode, ferroptosis, which is distinct from programmed cell death, such as apoptosis, necrosis, and autophagy. This novel process is driven by its reliance on iron-mediated lipid peroxidation. In ferroptosis, a large accumulation of reactive oxygen species disrupts the cellular membrane’s integrity and thereby induces cell death, which plays a key role in cancer, neurodegenerative diseases, and infectious and inflammatory diseases[40,41]. Iron death plays a role in Janus kinase-signal transducer and activator of transcription 1, a pathway typically occurring in HCC, which disrupts the oxidative/antioxidant balance of the body and enhances the efficacy of immunotherapy by downregulating solute carrier family 7 member 11 and solute carrier family 3 member 2[42]. During the exploration of sorafenib resistance mechanisms in recent years, we have found that identifying key targets in the mechanism of iron death onset inhibition by sorafenib helps to improve the anticancer efficacy of sorafenib in HCC[43,44]. The development and therapeutic regulation of iron death in HCC is complex, and a profound understanding of the regulatory network of iron death and its induction in HCC can provide a scientific basis for the precise treatment of HCC. Numerous studies have demonstrated that the induction of ferroptosis in combination with ICIs significantly enhances the response rate in HCC patients. For example, while the efficacy of atezolizumab monotherapy has been suboptimal, its combination with ferroptosis-inducing agents has been shown to promote tumor cell death, thus improving therapeutic outcomes[45]. Clinical investigations in this direction are currently underway, laying a solid foundation for future treatment strategies.

Study limitations

Similar to other bibliometric analyses, this study has some limitations. First, although the data from the Web of Science are very extensive, there are still publications that are not included, such as those indexed only in Scopus and PubMed, which may have caused bias in the results. Additionally, the names of the institutions were not written in a standard way, which may have affected citation statistics. Finally, in this study we selected the documents in English, which may have introduced language bias, and we acknowledge that this approach may have excluded valuable contributions from non-English-speaking regions.

CONCLUSION

The bibliometric analysis carried out in this research uncovered 93987 studies related to HCC, delineating the characteristics of publications on HCC from 2004 to 2023. We identified the countries, institutions, authors, references, and journals that have made significant contributions to this area, and we focused on five clusters of high-frequency and emerging keywords (molecular mechanisms, treatment strategies, immunity and prognosis, risk factors, and diagnosis). In future research, the diagnostic and therapeutic approaches to treating HCC should be updated and improved by further exploring the immune mechanisms of HCC.