Global, regional, and national burden and changing trends of hepatoblastoma and retinoblastoma from 1990 to 2021, and projections until 2050: findings from the Global Burden of Disease Study 2021

Introduction

Hepatoblastoma and retinoblastoma are two extracranial embryonal tumors, which mainly occur in children, especially in infants and young children (1) and pose serious threats to their life and health. Embryonal tumors originate from immature tissue and exhibit microscopic morphological features resembling developing embryonic and fetal tissues (2). Hepatoblastoma arises from immature liver tissue and is the most common primary liver cancer in children, accounting for nearly 80% of cases (3). According to previous studies, the incidence of hepatoblastoma has been rising gradually in recent decades, particularly in North America, Asia, and Europe (4). This trend may be attributed to improved diagnostic techniques and increasing survival rates of premature and low-birth-weight infants that had higher risk of developing hepatoblastoma (5). Due to advances in treatment, survival rates of localized hepatoblastoma have exceeded 70–80%. However, for advanced or metastatic cases, the prognosis remains unfavorable, with 5-year survival rates limited to 30–50% (4).

Retinoblastoma, the most common pediatric intraocular malignancy, originates from immature retinal cells (6). When detected early and managed properly, retinoblastoma is curable, with successful eye and vision preservation in most cases. However, advanced retinoblastoma requires more specialized treatments, often leading to eye enucleation (still a key strategy to cure unilateral retinoblastoma) and even significantly lower disease control and survival rates (7,8). Over the past century, survival rates in developed countries for retinoblastoma have increased dramatically, from under 5% to over 95% with agents such as topotecan and vincristine introduced for advanced intraocular retinoblastoma (9). However, in developing countries, which carry the highest tumor burden, survival remains poor, with some countries reporting rates as low as 0–5% (10).

An increasing number of studies have systematically evaluated pediatric oncology services across different countries, identifying profound systemic barriers and stark healthcare infrastructure disparities between high-income countries (HICs) and low- and middle-income countries (LMICs) (11). A systematic review found that inadequate healthcare infrastructure, workforce shortages, and financial constraints are the key barriers to healthcare access in LMICs, further exacerbated by socioeconomic disparities, geographic and racial inequities, and insufficient clinician training (12). In LMICs, which hold approximately 80% of the global retinoblastoma burden, late diagnosis due to cultural or socioeconomic barriers, low awareness at primary care level, deficient referral pathways and lack of multidisciplinary programs commonly lead to poor outcomes (13). Simultaneously, delay in cancer care for children in LMICs is also a critical issue (14). These disparities underscore the urgent need to understand the epidemiology and trends of these pediatric cancers for guiding prevention strategies, resource allocation, and healthcare planning. Previous epidemiological studies on hepatoblastoma and retinoblastoma have been constrained by fragmented regional registries, single-center designs, or lack of long-term tracking. For retinoblastoma, national registry gaps in Ethiopia, Brazil, and India lead to substantial under-reporting (15-17); for hepatoblastoma, epidemiological data similarly suffer from incomplete regional coverage and limited longitudinal follow-up (18). To overcome these limitations, the Global Burden of Disease (GBD) Study 2021—a comprehensive dataset covering a wide range of diseases across 204 countries and territories grouped into 21 GBD regions-harmonizes diverse global data sources into a unified and comparable format, enabling consistent cross-regional and long-term trend analysis from 1990 to 2021. It offers a valuable opportunity to examine the global, regional, and national burden of hepatoblastoma and retinoblastoma over the same period. Previous GBD analyses have explored the epidemiology of various childhood cancers, highlighting disparities in disease burden driven by socioeconomic status, healthcare infrastructure, and population demographics (19-21).

In the present study, we used data from GBD 2021 to explore the epidemiological patterns and trends (past and future) of hepatoblastoma and retinoblastoma, focusing on incidence, prevalence, mortality, and disability-adjusted life-years (DALYs). By analyzing these two cancers together, we sought to uncover shared patterns and disparities in their epidemiology and the cumulative disease burden. We aim to help develop targeted disease control strategies and improve outcomes of affected children globally. We present this article in accordance with the GATHER reporting checklist (available at https://tp.amegroups.com/article/view/10.21037/tp-2026-0286/rc).

Methods

Data source and processing

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. We accessed estimates of incidence, prevalence, deaths, and DALYs for hepatoblastoma and retinoblastoma by age, sex, region and country via the Global Health Data Exchange (GHDx) query tool (http://ghdx.healthdata.org/gbd-results-tool). Data were collected from various sources, including vital registration systems, verbal autopsies, censuses, household surveys, disease-specific registries, health service contact records, and other sources, covering 204 countries and territories and 811 subnational locations for each year from 1990 to 2021. The relevant data sources can be accessed via the GBD 2021 Sources Tool (22). To address the issue of sparse or poor-quality data in LMICs, the GBD framework implements two key methodological corrections. First, to address the widespread issue of ill-defined causes of death—so-called “garbage codes” (e.g., unspecified cancers or senility) that mask true underlying causes—the GBD study has developed a systematic redistribution algorithm that reclassifies such deaths into plausible underlying causes based on epidemiological patterns from well-defined data (23). Second, for location-years with insufficient empirical mortality data, the GBD framework employed model life tables that borrow information across age, time, and geography to synthesize heterogeneous data sources into internally consistent estimates (24). According to the International Classification of Diseases versions 9 and 10 (ICD-9 and ICD-10), hepatoblastoma was coded as C22.2 from ICD-10 and retinoblastoma was coded as 190.5 from ICD-9 as well as C69.2–C69.22 from ICD-10 in GBD 2021 (25). A brief overview of data processing and disease modelling strategies was detailed in GBD 2021 methods appendices (26) and supplementary appendix (Appendix 1).

Socio-demographic index (SDI) is a comprehensive metric reflecting social, economic, and demographic conditions of a country or region, with an impact on health outcomes. SDI is calculated as the geometric mean of three key components: the total fertility rate among individuals under 25 years, the average years of education for those aged 15 years and above, and the lag-distributed income per capita. It ranges from 0 to 1, with higher values indicating greater socio-economic and demographic advancement. Based on the SDI in 2021, the 204 countries and territories in GBD 2021 were categorized into five quintiles: low-SDI, low-middle-SDI, middle-SDI, high-middle-SDI and high-SDI groups (27).

Statistical analysis

Given that the GBD 2021 estimates are model-based and carry inherent uncertainty, particularly in regions with historically poor cause-of-death reporting systems, we explicitly accounted for margins of error by using 95% uncertainty intervals (UIs) for all burden estimates and 95% confidence intervals (CIs) for trend analyses. The number of cases and age-standardised rates (ASRs) were utilized to measure the incidence, prevalence, deaths, and DALYs of hepatoblastoma and retinoblastoma. We calculated ASRs per 100,000 persons according to the formula:

ASRper100,000persons=∑i=1NαiWi∑i=1NWi[1]

In the equation, α_idenotes the age-specific rate in the ith age group, while W_isignifies the count of individuals within the same age group as per the GBD 2021 standard population. N is the total number of age groups (28).

The estimated annual percentage changes (EAPCs) of ASRs were calculated to assess the average trend of changes over a defined time period. The EAPCs were calculated using the formula: EAPC =100 × (𝑒^β − 1), with 95% CIs derived from the estimation of β. A positive trend was identified when both the EAPC and the lower bound of its 95% CI exceeded zero. Conversely, a negative trend was determined if the EAPC and the upper bound of its 95% CI were less than zero. Otherwise, the change was considered statistically insignificant (28).

Local regression smoothing models (loess) using the “geom_smooth” function of the package “ggplot2” were applied to examine the relationship between disease burden (including ASRs in 2021 and their EAPCs from 1990 to 2021) and the SDI in 2021 across 21 regions and 204 countries or territories. The expected values were calculated by considering both the SDI and disease burden of all locations. Spearman correlation analysis was performed to obtain r indices and P values for the association between burden and SDI, with P<0.05 considered statistically significant.

The Bayesian age-period-cohort (BAPC) model was applied to project future burden of hepatoblastoma and retinoblastoma from 2022 to 2050, along with the calculation of 95% CIs. This model considers age, period, and cohort factors, which follows a log-linear Poisson framework to examine the multiplicative effects of age, period and cohort:

log(λij)=μ+αi+βi+γk[2]

Here, 𝑖 (1 ≤ 𝑖 ≤ 𝐼) represents time points, 𝑗 (1 ≤ 𝑗 ≤ 𝐽) denotes age groups, 𝜇 is the intercept, α_𝑖, β_𝑗 and γ_𝑘 indicate the age, period and cohort effects, respectively. Integrated nested Laplace approximation (INLA) was employed for posterior marginal distribution approximation, avoiding issues of mixing and convergence with lower error rates. Projections were performed using the R-BAPC and R-INLA packages (29). All statistical analyses and graph drawings were carried out using R software (version 4.2.1).

Results

Global burden of hepatoblastoma and retinoblastoma

For hepatoblastoma, the global ASR of incidence per 100,000 population decreased from 0.11 (95% UI: 0.09–0.13) in 1990 to 0.06 (95% UI: 0.05–0.08) in 2021, with an EAPC of −2.03 (95% CI: −2.17 to −1.89). The ASR of deaths per 100,000 population declined from 0.08 (95% UI: 0.06–0.09) to 0.04 (95% UI: 0.03–0.05) over the same period (EAPC: −2.54, 95% CI: −2.68 to −2.40) (Table 1). The ASRs of prevalence and DALYs in 2021 were 0.52 (95% UI: 0.28–0.68) and 3.27 (95% UI: 2.61–4.10) per 100,000 population, with similar decreased trends [EAPCs: −1.84 (95% CI: −1.98 to −1.70) and −2.53 (95% CI: −2.67 to −2.39)] (Table S1). The corresponding case counts also followed downward trends (Table 1, Table S1).

For retinoblastoma, the ASRs of incidence and prevalence per 100,000 population increased from 0.08 (95% UI: 0.05–0.10) and 0.69 (95% UI: 0.45–0.88) in 1990 to 0.11 (95% UI: 0.07–0.15) and 1.03 (95% UI: 0.65–1.34) in 2019, followed by notable declines to 0.09 (95% UI: 0.06–0.13) and 0.86 (95% UI: 0.53–1.15) in 2021. The overall trends remained upward, with EAPCs of 1.35 (95% CI: 1.13–1.56) for incidence and 1.36 (95% CI: 1.15–1.57) for prevalence. The case numbers followed similar trends during the study period. However, the ASRs of deaths and DALYs per 100,000 population decreased from 0.05 (95% UI: 0.03–0.07) and 4.54 (95% UI: 2.60–5.94) in 1990 to 0.04 (95% UI: 0.03–0.06) and 3.65 (95% UI: 2.21–4.96) in 2021, with EAPCs of −0.44 (95% CI: −0.55 to −0.33) and −0.42 (95% CI: −0.53 to −0.31), respectively (Table 1, Table S1).

For both cancers combined, the ASRs of incidence, prevalence, deaths, and DALYs per 100,000 population in 2021 were 0.16 (95% UI: 0.11–0.20), 1.38 (95% UI: 0.81–1.83), 0.08 (95% UI: 0.05–0.10), and 6.93 (95% UI: 4.82–9.06). From 1990 to 2021, these ASRs declined significantly, with EAPCs of −0.33 (95% CI: −0.50 to −0.17), −0.14 (95% CI: −0.31 to −0.02), −1.58 (95% CI: −1.67 to −1.49), and −1.56 (95% CI: −1.65 to −1.47), respectively. In comparison, the differences between hepatoblastoma and retinoblastoma on these ASRs were not pronounced in either 1990 or 2021 (Table 1, Table S1). Regarding sex-specific trends, the ASRs of incidence, prevalence, deaths, and DALYs from 1990 to 2021 for both hepatoblastoma and retinoblastoma showed similar global trends between males and females (Figure S1). All cases reported in this study were under 9 years old. Among different age groups, the highest numbers of incidence, prevalence, deaths, and DALYs of hepatoblastoma and retinoblastoma in 2021 were observed in children aged 2–4 years. The highest rates of hepatoblastoma occurred in children aged 0–6 days and then declined sharply. The rates of retinoblastoma were elevated both in children aged 0–6 days and in those aged 2–4 years (Figure S2). In infants under 6 days, hepatoblastoma rates were much higher than retinoblastoma rates. In other age groups, the rates of them were similar. The combined numbers of incidence, prevalence, deaths, and DALYs of both cancers were highest in age group of 2–4 years, while the combined rates were highest in children under 6 days (Figure S3A-S3D).

Regional burden of hepatoblastoma and retinoblastoma

In 2021, the highest ASRs of incidence, prevalence, deaths, and DALYs of hepatoblastoma were all in Western Sub-Saharan Africa, as 0.10 (95% UI: 0.07–1.14), 0.83 (95% UI: 0.25–1.21), 0.07 (95% UI: 0.05–0.09) and 6.26 (95% UI: 4.53–8.29) per 100,000 population respectively. Among the 5 SDI regions and 21 GBD regions, the ASRs of incidence, prevalence, deaths, and DALYs for hepatoblastoma went upward from 1990 to 2021 in 6, 6, 4 and 4 regions, respectively. The largest increases of these ASRs were all observed in High-income North America, with EAPCs of 2.10 (95% CI: 1.83–2.36) for incidence, 2.13 (95% CI: 1.86–2.40) for prevalence, 1.04 (95% CI: 0.91–1.17) for deaths, and 1.07 (95% CI: 0.94–1.20) for DALYs (Table 1, Figure 1, Table S1, Figure S4).

Figure 1 ASRs of incidence and deaths in 2021, and their EAPCs from 1990 to 2021 for hepatoblastoma and retinoblastoma, globally and by 26 regions. ASRs of incidence (A) and deaths (B), and EAPCs of ASRs of incidence (C) and deaths (D). ASRs, age-standardised rates; EAPCs, estimated annual percentage changes; SDI, socio-demographic index.

For retinoblastoma, the highest ASRs of incidence, prevalence, deaths, and DALYs in 2021 were observed in Eastern Sub-Saharan Africa, as 0.20 (95% UI: 0.12–0.31), 1.82 (95% UI: 1.13–2.85), 0.17 (95% UI: 0.11–0.26) and 14.67 (95% UI: 9.70–22.54) per 100,000 population, respectively. Among the 5 SDI regions and 21 GBD regions, the ASRs of incidence, prevalence, deaths, and DALYs of retinoblastoma increased between 1990 and 2021 in 16, 16, 2, and 2 regions, respectively. The most notable increases of ASRs of incidence and prevalence were observed in East Asia, with EAPCs of 5.55 (95% CI: 4.81–6.29) and 5.57 (95% CI: 4.83–6.31), respectively. For ASRs of deaths and DALYs, the most significant rises occurred in Southern Sub-Saharan Africa, with EAPCs of 1.94 (95% CI: 1.52–2.35) and 1.95 (95% CI: 1.53–2.37) (Table 1, Figure 1, Table S1, Figure S4).

For both cancers combined, the ASRs of incidence, prevalence, deaths, and DALYs in 2021 were highest in Eastern Sub-Saharan Africa, as 0.28 (95% UI: 0.17–0.45), 2.47 (95% UI: 1.29–4.08), 0.22 (95% UI: 0.14–0.36) and 19.61 (95% UI: 12.59–31.20) per 100,000 population respectively. In low-SDI regions, the ASRs of deaths and DALYs of retinoblastoma were both higher than those of hepatoblastoma. Conversely, in high-SDI regions, hepatoblastoma had higher ASRs of deaths and DALYs. Among the 26 regions, the combined ASRs of incidence, prevalence, deaths, and DALYs of two cancers increased from 1990 to 2021 in 9, 10, 3, and 3 regions, with Southern Sub-Saharan Africa showing the largest increases (Figure 1, Figure S4).

National burden of hepatoblastoma and retinoblastoma

At national level, the highest ASRs of incidence, prevalence, deaths, and DALYs of hepatoblastoma in 2021 were all found in Mali, as 0.36 (95% UI: 0.21–0.54), 2.94 (95% UI: 0.74–4.94), 0.25 (95% UI: 0.15–0.38) and 22.27 (95% UI: 13.04–33.42) per 100,000 population respectively. The largest increases in the ASRs of incidence, prevalence, deaths, and DALYs of hepatoblastoma occurred in Belarus [EAPC: 5.38 (95% CI: 4.56–6.19)], Belarus [EAPC: 5.45 (95% CI: 4.65–6.26)], Guam [EAPC: 4.85 (95% CI: 4.29–5.42)] and Guam [EAPC: 4.88 (95% CI: 4.31–5.46)], respectively (Figure 2A,2B, Tables S2,S3). However, most countries demonstrated decreased hepatoblastoma burden from 1990 to 2021.

Figure 2 ASRs of incidence in 2021, and their EAPCs from 1990 to 2021 for hepatoblastoma and retinoblastoma, by country. ASRs of incidence (A) and their EAPCs (B) of hepatoblastoma; ASRs of incidence (C) and their EAPCs (D) of retinoblastoma. ASRs, age-standardised rates; EAPCs, estimated annual percentage changes.

For retinoblastoma, by country, the highest ASRs of incidence, prevalence, deaths, and DALYs in 2021 were respectively recorded in Tokelau [1.19 (95% UI: 0.25–3.86) per 100,000 population], Tokelau [10.93 (95% UI: 2.32–35.38) per 100,000 population], Malawi [0.40 (95% UI: 0.16–0.88) per 100,000 population] and Malawi [34.90 (95% UI: 14.34–77.09) per 100,000 population]. The most significant increases in ASRs of all four metrics were observed in Armenia, with EAPCs of 18.07 (95% CI: 14.74–21.49), 18.07 (95% CI: 14.74–21.50), 14.17 (95% CI: 11.04–17.39), and 14.29 (95% CI: 11.15–17.52), respectively (Figure 2C,2D, Tables S2,S3). In most countries, the ASRs of incidence and prevalence of retinoblastoma went upward but those of deaths and DALYs went downward from 1990 to 2021.

For both cancers combined, Tokelau reported the highest ASRs of incidence and prevalence in 2021, as 1.20 (95% UI: 0.26–3.90) and 11.04 (95% UI: 2.36–35.73) per 100,000 population, respectively. Malawi had the highest ASRs of deaths and DALYs in 2021, as 0.50 (95% UI: 0.21–1.11) and 43.71 (95% UI: 18.16–97.42) per 100,000 population, respectively. In addition, Mauritius saw the largest rises in combined ASRs of incidence (EAPC: 4.83, 95% CI: 1.76–7.99) and prevalence (EAPC: 4.92, 95% CI: 1.84–8.09) of both cancers. Guam reported the most significant increases in combined ASRs of deaths (EAPC: 3.66, 95% CI: 3.21–4.11) and DALYs (EAPC: 3.68, 95% CI: 3.23–4.14) of both cancers (Tables S2,S3). The combined burden of both cancers showed descending trends in most countries.

The association between ASRs, EAPCs and SDI

At regional level, V-shaped associations were seen between SDI and both ASRs of incidence and prevalence of hepatoblastoma from 1990 to 2021, which had significant negative correlations overall (incidence: r=−0.241, P<0.001; prevalence: r=−0.159, P<0.001). The ASRs of deaths (r=−0.632, P<0.001) and DALYs (r=−0.629, P<0.001) of hepatoblastoma decreased with rising SDI across regions. Similarly, there were also V-shaped associations between SDI and ASRs of incidence and prevalence of retinoblastoma but the overall correlations were positive (incidence: r=0.221, P<0.001; prevalence: r=0.224, P<0.001). The ASRs of deaths (r=−0.817, P<0.001) and DALYs (r=−0.809, P<0.001) of retinoblastoma declined when SDI increased (Figure 3).

Figure 3 ASRs of incidence, prevalence, deaths, and DALYs of hepatoblastoma and retinoblastoma, globally and for 21 GBD regions, by SDI [2021], from 1990 to 2021. ASRs of incidence (A), prevalence (B), deaths (C), and DALYs (D) of hepatoblastoma, by SDI; ASRs of incidence (E), prevalence (F), deaths (G), and DALYs (H) of retinoblastoma, by SDI. Expected values with 95% CI, based on SDI and rates in all locations, are shown as a solid line and shaded area; 32 points are plotted for each location and show the observed ASRs of incidence, prevalence, deaths, or DALYs for each year from 1990 to 2021. Points above the solid line show a higher-than-expected burden, and those below the line represent a lower-than-expected burden. ASRs, age-standardised rates; CI, confidence interval; DALYs, disability-adjusted life-years; GBD, Global Burden of Diseases; SDI, socio-demographic index.

At national level, slight V-shaped associations were found between SDI and ASRs of both incidence and prevalence of hepatoblastoma in 2021, with significant negative correlations (r=−0.289, P<0.001 and r=−0.241, P<0.001, respectively). Additionally, significant negative correlations were observed between SDI and ASRs of deaths of both hepatoblastoma (r=−0.624, P<0.001) and retinoblastoma (r=−0.729, P<0.001). Similarly, the correlations between SDI and ASRs of DALYs of both hepatoblastoma (r=−0.620, P<0.001) and retinoblastoma (r=−0.713, P<0.001) were also significantly negative. However, the correlations between SDI and ASRs of incidence or prevalence of retinoblastoma were not significant (incidence: r=−0.107, P=0.13, prevalence: r=−0.105, P=0.14) (Figure S5).

From 1990 to 2021, significant positive correlations were observed between the EAPCs of ASRs of incidence (r=0.416, P<0.001), prevalence (r=0.410, P<0.001), deaths (r=0.152, P=0.03), and DALYs (r=0.157, P=0.03) of hepatoblastoma and SDI in 2021, by country. Similar positive correlations were also shown between the EAPCs of ASRs of incidence (r=0.223, P=0.001) and prevalence (r=0.221, P=0.001) of retinoblastoma and SDI. Conversely, the EAPCs of ASRs of deaths (r=−0.282, P<0.001) and DALYs (r=−0.192, P=0.006) of retinoblastoma were both negatively correlated with SDI (Figure S6).

Global burden projection for hepatoblastoma and retinoblastoma to 2050

Using BAPC model, the ASRs of incidence, prevalence, and deaths of hepatoblastoma were predicted to follow continuous downward trends, reaching 0.04 (95% UI: 0.00–0.10), 0.31 (95% UI: 0.00–0.76), and 0.02 (95% UI: 0.00–0.04) per 100,000 population respectively by 2050 (Figure 4A-4C). Meanwhile, for retinoblastoma, the ASRs of incidence and prevalence were predicted to rise gradually, reaching 0.11 (95% UI: 0.00–0.25) and 1.02 (95% UI: 0.00–2.27) by 2050, approximately equal to their historical maximum values. In contrast, the ASRs of deaths of retinoblastoma were projected to decline steadily, reaching 0.04 (95% UI: 0.00–0.07) by 2050 (Figure 4D-4F).

Figure 4 Global burden projections for hepatoblastoma and retinoblastoma to 2050. Projections of ASRs of incidence (A), prevalence (B), and deaths (C) of hepatoblastoma; projections of ASRs of incidence (D), prevalence (E), and deaths (F) of retinoblastoma. ASRs, age-standardised rates.

Moreover, we specifically focused on projections for the USA and China, large developed and developing countries, respectively. The model forecasted continued increases of ASRs of incidence, prevalence, and deaths of hepatoblastoma in the USA from 2022 to 2050, extending the upward trend in the previous 30 years. In China, these ASRs of hepatoblastoma were expected to follow a declining pattern similar to the global trend (Figure S7). For retinoblastoma, the model predicted continued decreases in the ASRs of incidence, prevalence, and deaths in both the USA and China, diverging from the global projections for incidence and prevalence (Figure S8).

Discussion

This study assessed the global burden of two cancers mainly seen in children with similar embryonal origins, hepatoblastoma and retinoblastoma. We explored the temporal trends of their disease burden over the past three decades and also in the future. The primary findings can be summarized as follows: first, the global burden of the two cancers mostly decreased from 1990 to 2021, except for the incidence and prevalence of retinoblastoma. Second, substantial regional disparities were observed in the burden of the two cancers, driven largely by socioeconomic and demographic factors. Third, the disease burden also varied across different age groups. Fourth, predictive modeling projected a decrease in the global burden of hepatoblastoma and retinoblastoma by 2050, with the exception of the incidence and prevalence of retinoblastoma. Different future trends were predicted in specific countries such as the USA and China. These findings highlight the need for tailored and region-specific control measures for the two cancers.

From 1990 to 2021, the combined global burden of these two cancers in this population significantly decreased primarily due to decreased hepatoblastoma burden. Moreover, in most regions and countries, hepatoblastoma burden went downward, which may be attributed to improvements of environmental sanitation, maternal nutrition (30), screening, and treatment. Standard chemotherapy regimens including cisplatin, surgical resection and liver transplantation have contributed to improved cure rates (31). To date, only a limited number of studies have focused on the epidemiologic trends of hepatoblastoma over recent decades. In a recent study using data from the Surveillance, Epidemiology, and End Results (SEER) registry, the incidence of hepatoblastoma has increased from 1975 to 2019 with an average percentage change (APC) of 2.31 in American children (32). In our study, the hepatoblastoma incidence in America also followed an increased trend from 1990 to 2021, consistent with the findings from SEER data. This may be a consequence of improved diagnostic techniques and better survival of premature infants with low birth weight or birth defects (33,34).

For retinoblastoma, the global ASRs of incidence and prevalence initially increased from 1990 to 2019, followed by declines from 2019 to 2021, which exhibited significant upward trends overall. The uptrend until 2019 may be due to enhanced health awareness, comprehensive child examinations and improved diagnostic capabilities. The declines after 2019 may partly result from the coronavirus disease 2019 (COVID-19) pandemic, which disrupted healthcare systems globally and likely affected the diagnosis and reporting of retinoblastoma. Substantial evidence indicates that the COVID-19 pandemic disrupted routine pediatric healthcare services globally (35). It likely reflects a statistical artifact rather than a genuine epidemiological shift. In England, childhood cancer diagnoses dropped significantly during the first pandemic wave (36). Such disruptions were not observed in hepatoblastoma possibly because hepatoblastoma is an extremely rare cancer (compared to retinoblastoma), making it less susceptible to short-term reporting artifacts (37). The global ASRs of deaths and DALYs of retinoblastoma decreased gradually from 1990 to 2021 mainly attributed to medical progress especially in care and treatment. In most regions and countries, the ASRs of retinoblastoma incidence and prevalence showed an upward trend, while those of deaths and DALYs exhibited downward trends between 1990 and 2021. Among the regions, East Asia had the greatest increases of incidence and prevalence and Southern Sub-Saharan Africa showed the largest increases of deaths and DALYs. According to previous studies, most patients with retinoblastoma were reported in Africa, Asia, and Latin America but North America, Europe, and Oceania represented less than 10% of cases. The survival rates of retinoblastoma also varied dramatically acrosscountries, exceeding 95% in developed countries like the USA, the UK, and Japan, but remaining much lower in developing countries like Uganda, Senegal, and Nepal (38). The large disparity may depend on the lifestyle, environmental risk factors (39), and medical level. In low-SDI regions, the retinoblastoma deaths and DALYs were higher than those of hepatoblastoma, suggesting that the retinoblastoma burden was more affected by socioeconomic factors. To address this problem, international collaborative organizations and global strategies are needed to promote resource and knowledge sharing. For example, the Rb Network (Rb-NET) was established to improve retinoblastoma outcomes in 6 Sub-Saharan African countries, linking them with experts from the UK, India, and Israel. The platform hosts regular virtual multidisciplinary team meetings that support local teams in diagnosing and managing complex cases (https://cehc.lshtm.ac.uk/dr-links/rbnet/). Besides, the World Health Organization launched the Global Initiative for Childhood Cancer (GICC) in 2018 under the CureAll framework, aiming to achieve at least 60% survival for children with cancer by 2030. Retinoblastoma was selected as one of six tracer cancers. The CureAll framework guides countries to strengthen centers of excellence and care networks, expand universal health coverage, develop management regimens and roadmaps, and establish evaluation and monitoring mechanisms (https://www.who.int/initiatives/the-global-initiative-for-childhood-cancer).

Although the highest numbers of incidence, prevalence, deaths, and DALYs in 2021 for both cancers were observed in children aged 2 to 4 years, distributions of rates were different due to the demographic differences in different age groups. For hepatoblastoma, the highest rates were seen in infants aged 0–6 days, with a sharp decline after that. For retinoblastoma, children aged 0–6 days and 2–4 years both had elevated rates. In infants under 6 days, the combined burden of both cancers was highest and the burden of hepatoblastoma was more severe than those of retinoblastoma. Therefore, more advanced screening methods for hepatoblastoma and retinoblastoma in newborns under 6 days are crucial, especially in high-risk populations (premature infants, families with familial adenomatous polyposis, retinoblastoma or RB1 mutation in the family and so on) (33,40,41). More potentially effective treatments including targeted therapy, immunotherapy, chemotherapy and focal therapy are needed to save the life and sight of newborn babies (42). For retinoblastoma, children aged 2–4 years also require special attention. This may be due to the non-heritable type of retinoblastoma, which comprises 60–70% and often presents later (43).

Although significant negative correlations were observed between SDI and hepatoblastoma burden overall, V-shaped associations were seen in detail between SDI and hepatoblastoma incidence and prevalence, with higher rates in both low- and high-SDI regions than in middle-SDI regions. These patterns may be related to healthcare access and environmental conditions. In low-SDI areas, certain environmental risk factors (such as parental exposure to harmful chemicals or infectious diseases) and poor maternal nutrition could be more common, causing increased incidence and prevalence (30). The rapid urbanization in many low-SDI countries has also aggravated air pollution exposure. In high-SDI areas with advanced healthcare systems, higher survival rates of premature and low-birth-weight neonates, early detection, and better treatment contributed to higher reported incidence and improved prevalence (5). In addition, delayed childbearing, a common trend in high-SDI areas, is associated with increased risk of childhood solid tumor. A meta-analysis supports this link, which may partially explain the higher hepatoblastoma incidence and prevalence in high-SDI regions (44). Middle-SDI regions lie in the midst of epidemiological transition. Their health systems, although undergoing expansion, still lacked the diagnostic capacity and resource allocation to fully capture these rare pediatric cancers due to competing health priorities such as infectious diseases and maternal health services (45). Similar V-shaped associations were seen in the relationships of SDI and retinoblastoma incidence and prevalence at regional level, which could be attributable to the poor environment and nutrition habits in low-SDI regions (43,46) and better screening, diagnostic, and therapeutic practices in high-SDI regions. The penetration of global health initiatives, such as the GICC, in low-SDI regions enhances local diagnostic capacity and thus increases disease detection. Additionally, retinoblastoma is often triggered by a biallelic mutation in the RB1 gene or MYCN oncogene amplification, which can be heritable or non-heritable (42). Around 40% of affected children carry a germline RB1 mutation and have a 50% chance of passing the mutation to their offspring. In high-SDI regions, a high survival rate of more than 95% means that patients who carry pathogenic genes may survive to childbearing age due to better healthcare access, leading to more familial cases and higher reported incidence (6). High-SDI regions had lower mortality and DALYs of both hepatoblastoma and retinoblastoma than low-SDI regions possibly due to better healthcare systems. This finding aligned with the report of Global Retinoblastoma Study Group which demonstrated that the survival of retinoblastoma children in low-income countries was far lower than that in HICs (38). About the correlations between SDI and the average changing trends of disease burden from 1990 to 2021, low-SDI countries experienced more decreases in incidence and prevalence of both cancers, as well as in hepatoblastoma deaths and DALYs. Although low-SDI countries have poorer healthcare infrastructure and environmental conditions compared to high-SDI countries, the former have undergone rapid development and significant progress over the past decades, leading to more improvements. However, the reduction of retinoblastoma deaths and DALYs was less pronounced in low-SDI countries, indicating that the intervention strategies for retinoblastoma remained underdeveloped in these countries and more medical resources should be allocated to improving the treatment of retinoblastoma.

Predictive modeling indicated continued decreases of hepatoblastoma burden and retinoblastoma deaths, but rising retinoblastoma incidence and prevalence by 2050. These findings indicated that retinoblastoma will remain a substantial challenge to child health. More efforts are needed to promote disease prevention and improve quality of life of affected children. Moreover, regional and national disparities still exist in the predicted future burden of hepatoblastoma and retinoblastoma, necessitating more equitable medical resources allocation, region-specific health policies, and better global collaboration for sharing knowledge and technologies about the diseases. Feasible strategies (especially in low-SDI regions) include: strengthening antenatal examination and early screening for premature and low-birth-weight infants; providing preconception counseling or prenatal diagnosis for families with a history of retinoblastoma and for older mothers to prevent affected births or enable early detection; implementing environmental and nutritional improvements; and enhancing primary healthcare training, referral networks, and multidisciplinary consultation services through international collaborative organizations.

There are several limitations in this study. First, data availability and quality vary across regions. In many LMICs, vital statistics, disease registries, and healthcare infrastructure are often limited or unreliable, leading to potential underreporting or misclassification of health data, especially in the early 1990s (47). In such regions, the GBD study relies on modeling assumptions and extrapolations, which can introduce uncertainties. Another limitation of this study is the inherent time lag of GBD data, as GBD 2021 estimates do not reflect the most current epidemiological status beyond 2021. Third, our study only focused on the burden of two extracranial embryonal tumors. Fourth, GBD 2021 data on hepatoblastoma and retinoblastoma only include individuals aged under nine but these cancers can also be diagnosed in older children and adults in practice, particularly when the tumor grows slowly or presents without early symptoms. This may result in an underestimation of their global burden. Finally, this study did not involve tumor type or staging details due to the lack of relevant data in GBD 2021, restricting our deeper exploration.

Conclusions

This study reveals the global trends of hepatoblastoma and retinoblastoma burden over the past three decades and projections for the future. Although the global burden has generally decreased, exceptions such as the rising incidence and prevalence of retinoblastoma call for more effective management. Notably, regional and age-related disparities in disease burden, driven by socioeconomic and demographic factors, emphasize the need for targeted, region- and age-specific interventions. Feasible strategies including antenatal examination, early screening, genetic counseling for high-risk families, environmental and nutritional improvements, and international collaboration to strengthen healthcare training, referral networks, and multidisciplinary consultation should be implemented especially in low-SDI regions to reduce inequalities and promote children health.

Acknowledgments

We would like to thank the GBD research team and the Institute for Health Metrics and Evaluation (IHME) for their valuable efforts in providing comprehensive global health data.

Footnote

Reporting Checklist: The authors have completed the GATHER reporting checklist. Available at https://tp.amegroups.com/article/view/10.21037/tp-2026-0286/rc

Peer Review File: Available at https://tp.amegroups.com/article/view/10.21037/tp-2026-0286/prf

Funding: This work was supported by grants from Guangzhou Women and Children’s Medical Center Clinical Doctor Initiation Scientific Research Fund (No. 2024BS016, to D.L.) and Guangdong Medical Science and Technology Research Fund (No. C2025064, to X.Z.).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tp.amegroups.com/article/view/10.21037/tp-2026-0286/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Awounou D, Mancini M, Lacour B, et al. Residential proximity to vines and risk of childhood embryonal tumours in France - GEOCAP case-control study, 2006-2013. Environ Res 2024;240:117417. [Crossref] [PubMed]
2. Scotting PJ, Walker DA, Perilongo G. Childhood solid tumours: a developmental disorder. Nat Rev Cancer 2005;5:481-8. [Crossref] [PubMed]
3. Jia M, He E, Sun W, et al. Hepatoblastoma regional trends: dynamic SDI & joinpoint regression analysis. BMC Cancer 2025;25:1148. [Crossref] [PubMed]
4. Fu Y, Francés R, Monge C, et al. Metabolic and Epigenetic Mechanisms in Hepatoblastoma: Insights into Tumor Biology and Therapeutic Targets. Genes (Basel) 2024;15:1358. [Crossref] [PubMed]
5. Berkemeyer S. Primary Liver Cancers: Connecting the Dots of Cellular Studies and Epidemiology with Metabolomics. Int J Mol Sci 2023;24:2409. [Crossref] [PubMed]
6. Global Retinoblastoma Study Group. Global Retinoblastoma Presentation and Analysis by National Income Level. JAMA Oncol 2020;6:685-95. [Crossref] [PubMed]
7. Sullivan EM, Wilson MW, Billups CA, et al. Pathologic risk-based adjuvant chemotherapy for unilateral retinoblastoma following enucleation. J Pediatr Hematol Oncol 2014;36:e335-40. [Crossref] [PubMed]
8. Naseripour M, Aghajani A, Kasraei H, et al. Retinoblastoma survival trend: a 30-year analysis from a referral single center in Iran. Int J Retina Vitreous 2025;11:113. [Crossref] [PubMed]
9. Qaddoumi I, Billups CA, Tagen M, et al. Topotecan and vincristine combination is effective against advanced bilateral intraocular retinoblastoma and has manageable toxicity. Cancer 2012;118:5663-70. [Crossref] [PubMed]
10. Chantada GL, Villanueva G, Abramson DH. Challenges for Early Diagnosis in Retinoblastoma in Low- and Middle-Income Countries. Pediatr Blood Cancer 2025;72:e31859. [Crossref] [PubMed]
11. Rodriguez-Galindo C, Friedrich P, Alcasabas P, et al. Toward the Cure of All Children With Cancer Through Collaborative Efforts: Pediatric Oncology As a Global Challenge. J Clin Oncol 2015;33:3065-73. [Crossref] [PubMed]
12. Kucukkaya A, Goktas P, Semerci Şahin R. Exploring Barriers and Inequalities in Access to Comprehensive Care for Pediatric Oncology Patients: A Systematic Review. Semin Oncol Nurs 2025;41:151852. [Crossref] [PubMed]
13. Ramanathan S, Vijayasekharan K. Retinoblastoma: Poster-Child Tumor of the Low-Middle Income Countries (LMICs)!!! South Asian J Cancer 2022;11:279-80. [Crossref] [PubMed]
14. Cotache-Condor C, Rice HE, Schroeder K, et al. Delays in cancer care for children in low-income and middle-income countries: development of a composite vulnerability index. Lancet Glob Health 2023;11:e505-15. [Crossref] [PubMed]
15. Sherief ST, Asfaw G, Abateneh A, et al. Incidence and geographic distribution of retinoblastoma in Ethiopia. BMC Ophthalmol 2023;23:231. [Crossref] [PubMed]
16. Barbosa AC, de Magalhães-Barbosa MC, Moreira JPL, et al. Incidence of retinoblastoma in children and adolescents in Brazil: A population-based study. Front Pediatr 2022;10:1048792. [Crossref] [PubMed]
17. Khetan V, Sengupta S. Twenty-year trends of retinoblastoma under five hospital-based cancer registries: few observations. Indian J Cancer 2011;48:371-2. [Crossref] [PubMed]
18. Dasgupta P, Henshaw C, Youlden DR, et al. Global trends in incidence rates of childhood liver cancers: A systematic review and meta-analysis. Paediatr Perinat Epidemiol 2020;34:609-17. [Crossref] [PubMed]
19. Wu Y, Deng Y, Wei B, et al. Global, regional, and national childhood cancer burden, 1990-2019: An analysis based on the Global Burden of Disease Study 2019. J Adv Res 2022;40:233-47. [Crossref] [PubMed]
20. Shukla A, Zeidan RK, Saddik B. Pediatric and adolescent cancer disparities in the Middle East and North Africa (MENA) region: incidence, mortality, and survival across socioeconomic strata. BMC Public Health 2024;24:3602. [Crossref] [PubMed]
21. GBD 2023 Childhood Cancer Collaborators. Global burden of cancer in children and adolescents aged 0-19 years, 1990-2023: a systematic analysis for the Global Burden of Disease Study 2023. Lancet 2026;407:1360-73. [Crossref] [PubMed]
22. Global Burden of Disease Study 2021 (GBD 2021) Sources Tool. [Accessed December 25 2024]. Available online: https://ghdx.healthdata.org/gbd-2021/sources
23. Scohy A, Lesnik T, Devleesschauwer B, et al. The importance of redistribution process of ill-defined deaths in national mortality database. Arch Public Health 2025;83:185. [Crossref] [PubMed]
24. Global age-sex-specific mortality, life expectancy, and population estimates in 204 countries and territories and 811 subnational locations, 1950-2021, and the impact of the COVID-19 pandemic: a comprehensive demographic analysis for the Global Burden of Disease Study 2021. Lancet 2024;403:1989-2056. [Crossref] [PubMed]
25. Global Burden of Disease Study 2021 (GBD 2021) Causes of Death and Nonfatal Causes Mapped to ICD Codes. [Accessed January 14 2025]. Available online: https://ghdx.healthdata.org/record/ihme-data/gbd-2021-cause-icd-code-mappings
26. GBD 2021 Methods Appendices. [Accessed January 9 2025]. Available online: https://www.healthdata.org/gbd/methods-appendices-2021/cancers
27. Global incidence, prevalence, years lived with disability (YLDs), disability-adjusted life-years (DALYs), and healthy life expectancy (HALE) for 371 diseases and injuries in 204 countries and territories and 811 subnational locations, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021. Lancet 2024;403:2133-61. [Crossref] [PubMed]
28. Sun P, Yu C, Yin L, et al. Global, regional, and national burden of female cancers in women of child-bearing age, 1990-2021: analysis of data from the global burden of disease study 2021. EClinicalMedicine 2024;74:102713. [Crossref] [PubMed]
29. Zhang T, Sun L, Yin X, et al. Burden of drug use disorders in the United States from 1990 to 2021 and its projection until 2035: results from the GBD study. BMC Public Health 2024;24:1639. [Crossref] [PubMed]
30. Heck JE, Wu CK, Huang X, et al. Cohort study of familial viral hepatitis and risks of paediatric cancers. Int J Epidemiol 2022;51:448-57. [Crossref] [PubMed]
31. Tang R, Qi SQ, Zhang T, et al. Progress in the study of therapeutic strategies for hepatoblastoma in children. World J Gastrointest Oncol 2025;17:107700. [Crossref] [PubMed]
32. Sultan I, Alfaar AS, Sultan Y, et al. Trends in childhood cancer: Incidence and survival analysis over 45 years of SEER data. PLoS One 2025;20:e0314592. [Crossref] [PubMed]
33. Heck JE, Meyers TJ, Lombardi C, et al. Case-control study of birth characteristics and the risk of hepatoblastoma. Cancer Epidemiol 2013;37:390-5. [Crossref] [PubMed]
34. Lupo PJ, Schraw JM, Desrosiers TA, et al. Association Between Birth Defects and Cancer Risk Among Children and Adolescents in a Population-Based Assessment of 10 Million Live Births. JAMA Oncol 2019;5:1150-8. [Crossref] [PubMed]
35. Fabian ID, Stacey AW, Bowman R, et al. Retinoblastoma management during the COVID-19 pandemic: A report by the Global Retinoblastoma Study Group including 194 centers from 94 countries. Pediatr Blood Cancer 2021;68:e28584. [Crossref] [PubMed]
36. Saatci D, Oke J, Harnden A, et al. Childhood, teenage and young adult cancer diagnosis during the first wave of the COVID-19 pandemic: a population-based observational cohort study in England. Arch Dis Child 2022;107:740-6. [Crossref] [PubMed]
37. Gatta G, Ferrari A, Stiller CA, et al. Embryonal cancers in Europe. Eur J Cancer 2012;48:1425-33. [Crossref] [PubMed]
38. The Global Retinoblastoma Outcome Study: a prospective, cluster-based analysis of 4064 patients from 149 countries. Lancet Glob Health 2022;10:e1128-40. [Crossref] [PubMed]
39. Koochakzadeh L, Yekta A, Hashemi H, et al. Epidemiological aspect of retinoblastoma in the world: a review of recent advance studies. Int J Ophthalmol 2023;16:962-8. [Crossref] [PubMed]
40. Trobaugh-Lotrario AD, López-Terrada D, Li P, et al. Hepatoblastoma in patients with molecularly proven familial adenomatous polyposis: Clinical characteristics and rationale for surveillance screening. Pediatr Blood Cancer 2018;65:e27103. [Crossref] [PubMed]
41. Skalet AH, Gombos DS, Gallie BL, et al. Screening Children at Risk for Retinoblastoma: Consensus Report from the American Association of Ophthalmic Oncologists and Pathologists. Ophthalmology 2018;125:453-8. [Crossref] [PubMed]
42. Nag A, Khetan V. Retinoblastoma - A comprehensive review, update and recent advances. Indian J Ophthalmol 2024;72:778-88. [Crossref] [PubMed]
43. Kumari A, Singh SP, Kumar P, et al. A Comprehensive Review of the Epidemiology, Pathophysiology, Risk Factors, and Treatment Strategies for Retinoblastoma. Diseases 2025;13:307. [Crossref] [PubMed]
44. Domingues A, Moore KJ, Sample J, et al. Parental Age and Childhood Lymphoma and Solid Tumor Risk: A Literature Review and Meta-Analysis. JNCI Cancer Spectr 2022;6:pkac040. [Crossref] [PubMed]
45. Omran AR. The epidemiologic transition: a theory of the epidemiology of population change. 1971. Milbank Q 2005;83:731-57. [Crossref] [PubMed]
46. Jung EM, Bunin GR, Ganguly A, et al. The association between maternal nutrient intake during pregnancy and the risk of sporadic unilateral retinoblastoma among offspring. Cancer Epidemiol 2023;85:102376. [Crossref] [PubMed]
47. Lozano R, Naghavi M, Foreman K, et al. Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012;380:2095-128. [Crossref] [PubMed]