Hepatocellular carcinoma national burden across different geographical regions in the United States between 2001 and 2020

Abstract

BACKGROUND

While prior data showed an increasing incidence of hepatocellular carcinoma (HCC) in the United States, there are limited comprehensive and comparative data on the geographical variations of HCC trends in different demographic-specific populations.

AIM

To evaluate sex and age-specific incidence rates and time trends in different geographical regions in the United States.

METHODS

Age-adjusted HCC incidence rates were collected from the United States Cancer Statistics (USCS) database which covers approximately 98% of the population in the United States. HCC rates were stratified by sex, age, and geographical region. annual percentage change (APC) and average APC (AAPC) were estimated using Joinpoint Regression. A pairwise comparison was conducted between sex-specific trends.

RESULTS

There were 467344 patients diagnosed with HCC in the United States in the USCS database between 2001 and 2020. The rates and trends varied by geographical region. When looking at the West region (115336 patients), incidence rates of HCC were overall increasing and also increasing in older adults. However, when evaluating younger adults, HCC incidence rates decreased in men but not in women with a sex-specific absolute AAPC-difference of 2.15 (P = 0.005). When evaluating the Midwest region (84612 patients), similar results were seen. While incidence rates were increasing in the overall population and in older adults as well, they were decreasing in younger men but not in women with a sex-specific absolute AAPC-difference of 1.61 (P < 0.001). For the Northeast region (87259 patients), the analysis showed similar results with decreasing HCC incidence rates in younger men but not counterpart women (Sex-specific AAPC-difference = 3.26, P < 0.001). Lastly, when evaluating the south (180137 patients), the results were also decreasing in younger men but not in women (Sex-specific AAPC-difference = 2.55, P < 0.001).

CONCLUSION

Nationwide analysis covering around 98% of the United States population shows an increasing incidence of HCC across all geographical regions, most notably in the South. While younger men experienced decreasing HCC incidence, younger women had a stable trend and this was noted across all regions as well. Our study offers insight into the epidemiology of HCC in different demographic groups across various United States geographical regions. While the reasons contributing to our findings are unclear, they can be related to sex and regional disparities in healthcare access and utilization. Future research is warranted to characterize the temporal change in HCC risk factors across different United States regions.

Key Words: Hepatocellular carcinoma; Incidence; Epidemiology; Health disparity; Geography

Core Tip: In this retrospective study of the United States Cancer Statistics database (which covers approximately 98% of the United States population), we analyzed sex and age-specific hepatocellular carcinoma (HCC) incidence across different United States regions between 2001-2020. HCC incidence rates were significantly increasing in the West, Midwest, Northeast, and South of the United States, most notably in the South. While younger men experienced decreasing HCC incidence, younger women had non-decreasing incidence, and this was noted across all regions. While this can be due to regional disparities in healthcare access/utilization, future research is needed to investigate regional HCC risk factors, especially in younger adults.

INTRODUCTION

Hepatocellular carcinoma (HCC) accounts for around 80% of liver cancers and its incidence has tripled since 1980[1]. Although HCC can occur sporadically, almost 90% of cases can be attributed to underlying liver diseases such as hepatitis C virus (HCV), hepatitis B virus (HBV), alcohol-associated liver disease, and metabolic dysfunction associated steatotic liver disease (MASLD). With the routine vaccination against HBV and an increase in the proportion of HCV patients with sustained virologic response, coupled with increased prevalence of MASLD, epidemiological changes in HCC risk factors are expected in the coming years[2].

HCC burden is unequally distributed with disparities occurring at various steps of the cancer care continuum including implementation of screening programs, access to specialist care, timely diagnosis, and treatment. HCC is usually rare under the age of 40 years, and its incidence increases with age before plateauing around the age of 70 years[3]. The incidence of HCC is disproportionately greater in males, with associated mortality reported to be three times higher than in women[4]. These differences are attributed to a variation in the prevalence of risk factors such as alcohol use, smoking, viral hepatitis, and MASLD which disproportionally affect men and women. The evolving trends in sex- and age-specific HCC incidence rates call for further exploration and analysis of data at a national level to guide future interventions.

American Cancer Society projects cancer deaths in 2024 from liver and bile duct cancer to be the highest in California (3580), followed by Texas (2960) and Florida (2180)[1]. Limited data comparing regional variation in HCC incidence show rates are highest in Texas, followed by Hawaii, New Mexico, and California; argued to be in part due to their racial/ethnic diversity[5]. While recent data showed an increasing overall incidence of HCC with variations based on age and sex[4], literature is scarce on demographic-specific trends across geographic regions. Understanding the epidemiological differences at a national level is imperative to identify regional variabilities in epidemiology and study their impact on HCC-associated morbidity and mortality. Therefore, we aimed to perform a comprehensive analysis of HCC incidence rates and time trends stratified by sex and age in different geographical regions in the United States using the United States Cancer Statistics (USCS) database.

MATERIALS AND METHODS

We report a time trend comprehensive analysis of national incidence rates of HCC in the United States between 2001 and 2020 across various regions in the United States using publicly available and de-identified data from the USCS database which covers nearly 98% of the United States population[6]. HCC incidence rates were age-adjusted to the standard 2000 United States population using SEER Stat software [version 8.4.2, National Cancer Institute (NCI)]. The rates were categorized by sex and age into younger and older adults (defined with an age cutoff 55 years)[4]. HCC incidence rates were also stratified by geographical region in the United States into West, Midwest, Northeast, and South. Time-trends were reported as annual percentage change (APC) and average APC (AAPC) and were generated using Joinpoint Regression Software (version 5.0.2, NCI) via the weighted Bayesian Information Criteria (BIC) method (which is a data-driven analytical method used to estimate trends over time that is recommended to be used in large databases)[7-9]. A pairwise comparison was done between the sex-specific trends using the tests of parallelism and coincidence with a two-sided P-value cutoff at 0.05[10].

RESULTS

During the study period of 2001-2020, there were 467344 patients diagnosed with HCC in the United States. The majority of the patients were men (74.0%) and were diagnosed in the South (38.5%). HCC incidence rates and time trends varied between the cohorts across different geographical regions.

In the West regions (115336 patients; 26.9% women), HCC incidence rate per 100000 population significantly increased in the overall population from 5.57 to 6.48 between 2001 and 2020, and also in older adults (95152 patients; 28.2% women) from 19.66 to 26.31 between 2001 and 2020 (Figure 1). However, in younger adults (19160 patients; 19.8% women), while HCC incidence rates decreased in men from 3.77 to 1.97 between 2001 and 2020 (AAPC = -2.92, P < 0.001), the rates did not decrease in women (AAPC = -0.78, P = 0.10) with a significant difference between the sexes (Sex-specific absolute AAPC-difference = 2.15, P < 0.005) (Table 1 and Figure 2).

Figure 1 Sex-specific time-trends and age-adjusted incidence rates per 100000 population for hepatocellular carcinoma in older adults aged < 55 years categorized by United States geographical region. A: Western United States; B: Midwest United States; C: Northeast United States; D: Southern United States.

Figure 2 Sex-specific time-trends and age-adjusted incidence rates per 100000 population for hepatocellular carcinoma in younger adults aged < 55 years (15-54 years) categorized by United States geographical region. A: Western United States; B: Midwest United States; C: Northeast United States; D: Southern United States.

Table 1 Hepatocellular carcinoma incidence trends in different age and sex-specific groups across United States geographical regions.

Age group, year	Cases (n = 467344)1, (%)	Trends2			Sex-specific AAPC difference (95%CI)3	Pairwise comparison P values
		Time period	APC	AAPC		Sex-specific AAPC difference	Coincidence4	Parallelism5
West
All ages
Women	31008 (6.64)	2001-2014	2.66a	0.96a	-0.02 (-0.86 to 0.82)	0.95	< 0.001	0.008
		2014-2020	-2.61a
Men	84328 (18.04)	2001-2009	4.42a	0.94a
		2009-2014	1.66a
		2014-2018	-1.72
		2018-2020	-8.68a
Aged ≥ 55 years
Women	26804 (5.74)	2001-2014	3.10a	1.36a	0.61 (-0.42 to 1.65)	0.24	< 0.001	0.002
		2014-2020	-2.30a
Men	68348 (14.62)	2001-2013	4.95a	1.97a
		2013-2018	-0.44
		2018-2020	-8.91a
Aged < 55 years
Women	3786 (0.81)	2001-2009	3.26a	-0.78	-2.15a (-3.65 to -0.64)	0.005	< 0.001	< 0.001
		2009-2020	-3.61a
Men	15383 (3.29)	2001-2004	9.67a	-2.92a
		2004-2011	-1.91a
		2011-2020	-7.54a
All ages
Women	3786 (0.810)	2001-2009	3.26a	-0.78	-2.15a (-3.65 to -0.64)	0.005	< 0.001	< 0.001
		2009-2020	-3.61a
Men	15383 (3.291)	2001-2004	9.67a	-2.92a
		2004-2011	-1.92a
		2011-2020	-7.54a
Midwest
All ages
Women	22675 (4.851)	2001-2017	2.42a	1.12a	0.50 (-0.54 to 1.55)	0.34	< 0.001	0.02
		2017-2020	-5.56a
Men	61937 (13.252)	2001-2012	3.90a	1.62a
		2012-2018	0.99
		2018-2020	-8.38a
Aged ≥ 55 years
Women	19403 (4.151)	2001-2017	2.82a	1.53a	1.12a (0.04 to 2.19)	0.04	< 0.001	< 0.001
		2017-2020	-5.08a
Men	51228 (10.961)	2001-2013	4.89a	2.65a
		2013-2018	2.15a
		2018-2020	-8.72a
Aged < 55 years
Women	2942 (0.629)	2001-2010	3.02a	-0.35	-1.61 (-3.57 to 0.36)	0.10	< 0.001	< 0.001
		2010-2020	-3.28a
Men	10227 (2.188)	2001-2003	12.75a	-1.95a
		2003-2009	0.88
		2009-2020	-5.89a
Northeast
All ages
Women	21729 (4.645)	2001-2013	2.67a	0.74	0.14 (-0.90 to 1.17)	0.79	< 0.001	0.2
		2013-2020	-2.48a
Men	65530 (14.021)	2001-2007	4.77a	0.88a
		2007-2016	1.25a
		2016-2020	-5.49a
Aged ≥ 55 years
Women	18703 (4.001)	2001-2013	3.01a	1.13a	0.76 (-0.26 to 1.78)	0.14	< 0.001	< 0.001
		2013-2020	-2.00a
Men	53688 (11.487)	2001-2008	5.17a	1.89a
		2008-2016	2.62a
		2016-2020	-4.97a
Aged < 55 years
Women	2763 (0.591)	2001-2007	6.63a	-0.13	-3.26a (-5.08 to -1.43)	< 0.001	< 0.001	<0.001
		2007-2020	-3.10a
Men	11462 (2.452)	2001-2008	2.39a	-3.38a
		2008-2020	-6.60a
South
All ages
Women	45975 (9.837)	2001-2015	3.74a	2.08a	0.46 (-0.53 to 1.45)	0.36	< 0.001	0.01
		2015-2020	-2.42
Men	134162 (28.707)	2001-2007	6.14a	2.54a
		2007-2014	3.66a
		2014-2018	0.05
		2018-2020	-6.50a
Aged ≥ 55 years
Women	38809 (8.304)	2001-2016	3.99a	2.53a	1.12a (0.10 to 2.13)	0.03	< 0.001	< 0.001
		2016-2020	-2.77
Men	108315 (23.176)	2001-2014	5.89a	3.65a
		2014-2018	1.4
		2018-2020	-5.74a
Aged < 55 years
Women	6576 (1.407)	2001-2009	4.77a	0.5	-2.01a (-3.77 to -0.24)	0.02	< 0.001	< 0.001
		2009-2015	-1.2
		2015-2020	-6.00a
Men	25030 (5.355)	2001-2006	7.81a	-2.05a
		2006-2014	-2.16a
		2014-2020	-9.46a

¹Data are presented as count numbers followed by percentages of the count numbers from the total cases of hepatocellular carcinoma in the database.

²Time-trends were computed using Joinpoint Regression Program (version 5.0.2, National Cancer Institute) with 3 maximum joinpoints allowed (4-line segments).

³A negative value indicates a greater average annual percentage change in women compared to men.

⁴Tests whether incidence trends were identical. A significant P-value indicates that the trends were not identical (i.e., they had different incidence rates and coincidence was rejected).

⁵Tests whether incidence trends were parallel. A significant P-value indicates that the trends were not parallel (i.e., parallelism was rejected).

^aP < 0.05.

APC: Annual percentage change; AAPC: Average annual percentage change.

When evaluating the Midwest (84612 patients; 26.8% women), similar results were seen where HCC incidence rates per 100000 population were significantly increasing in the overall population from 3.84 to 5.16 between 2001 and 2020, and also in older adults (70631 patients; 27.5% women) from 13.65 to 20.83 between 2001 and 2020 (Figure 1). However, in younger adults (13169 patients; 22.3% women), while HCC incidence rates were decreasing in men from 2.40 to 1.61 between 2001 and 2020 (AAPC = -1.95, P = 0.002), the rates did not decrease in women (AAPC = -0.35, P = 0.65) with an absolute sex-specific AAPC -difference of 1.61, P = 0.1 (Table 1 and Figure 2).

In the Northeast (87259 patients; 24.9% women), similar results were also seen where HCC incidence rates per 100000 population were increasing in the overall population from 5.01 to 5.74, and in older adults (72391 patients; 225.8% women) from 17.36 to 23.41, between 2001 and 2020 (Figure 1). However, in younger adults (14225 patients; 19.4% women), HCC rates were decreasing in men from 3.70 to 1.73 between 2001 and 2020 (AAPC= -3.38, P < 0.001), and this was not seen in women who experienced a stable trend (AAPC= -0.13, P = 0.87) with an absolute sex-specific AAPC-difference of 3.26, P < 0.001 (Table 1 and Figure 2).

Lastly, in the south (180137 patients; 25.5% women), similar results were also seen with increasing HCC rates per 100000 population in the overall population from 4.31 to 6.81 between 2001 and 2020, and also in older adults from 14.69 to 27.68 between 2001 and 2020 (Figure 1). However, younger adults (467344 patients; 20.8% women) experienced decreasing HCC incidence rates in men from 3.28 in 2001 to 2.14 in 2020 (AAPC = -2.05, P < 0.001), but not in women who had a stable trend (AAPC = -0.05, P = 0.95) with an absolute sex-specific AAPC-difference of 2.55, P < 0.001 (Table 1 and Figure 2).

DISCUSSION

Our nationwide analysis of data covering nearly all HCC patients in the United States shows increasing overall incidence of HCC in older adults across all geographical United States regions, most notably in the South. However, in younger adults, HCC incidence was decreasing in men but not in counterpart women, and that sex-specific difference was seen across all geographical regions in the United States.

As of 2020, primary liver cancer is the fifth most common cancer in men worldwide, the ninth most common cancer in women worldwide, and the sixth most common cancer diagnosed overall[11]. According to the NCI, in 2023, there were about 41000 diagnoses and about 29000 deaths related to primary liver cancer in the United States. A recent analysis using the USCS database showed increasing overall HCC incidence, and decreasing rates in younger men but not women[4]. Recent data also suggests that HCC incidence is trending away from male predominance in younger adults, with a decreasing male-to-female incidence ratio[5]. We add to prior literature showing that the greatest increase in overall HCC was in the South and the greatest decrease in HCC incidence in younger men was in the Northeast. We also show that the stable trend in younger women was noted across all United States regions. These differences are multifactorial, in large part due to the racial and geographical disparities in access to and quality of care. The findings of non-decreasing HCC incidence rate in younger females may be attributable to the changing prevalence of risk factors over recent years across all regions of the United States[4].

A prior analysis of the SEER database of 43868 patients between 2000-2012 evaluating the most at-risk group in the United States (Southern region) compared to other regions, showed that blacks comprised a larger proportion of HCC patients in the South compared to other areas (32.4% vs 10.1%) and were diagnosed at a younger age, with more advanced stage at diagnosis and more metastases[12]. Furthermore, black patients were 58% less likely to receive liver transplant and 36% less likely to receive surgical therapy for HCC compared to patients of white race[13]. Of interest, this study also saw similar radial disparities in HCC outcomes in geographical regions outside the Southern United States. Another analysis of the SEER database from 1975-2017 showed that HCC mortality rates were highest in the South, followed by the West, Northeast, and Midwest[14]. These differences are multifactorial, in large part due to the racial and geographical disparities in access to and quality of care[15]. Our study builds on these prior studies, using a significantly larger sample size over a more recent period and stratifying the analysis by sex and age, demonstrating that the greatest increase in HCC incidence was in the South.

The introduction of direct-acting-antiviral treatment for HBV and HCV along with the increasing obesity epidemic in the United States has shifted the leading causes of HCC from viral hepatitis toward MASLD[16,17]. However, access to treatment has been a subject of public health concern, as fewer Medicaid and Medicare recipients with HCV receive timely treatment compared to those privately insured[18]. Furthermore, black Medicaid recipients are less likely to receive direct-acting antiviral therapy than white counterparts[18,19]. Viral hepatitis also disproportionately affects the black population. A prior study using the National Health and Nutrition Examination Survey database found that African Americans had a two to three-fold higher prevalence of chronic HBV than the general population between 1999 and 2008[20]. According to the United States Census Bureau in 2021, nine of the ten poorest states are in the South. These factors could explain the greater increase in HCC incidence in the South. These findings demonstrate a need for targeted public health intervention and multidisciplinary care in this high-risk, underserved, largely uninsured population.

MASLD is a leading cause of HCC in Western countries[21]. Sex differences in metabolic risk factors for HCC may explain the trends we observed. In the more recent years of our study period, in all regions but the Midwest, the HCC incidence rate in women older than 55 years of age has decreased less than that of men over 55 years of age. This could be explained in part by decreased estrogen levels in postmenopausal women. Estrogen may have protective effects on hepatic fibrinogenesis, causing later onset of HCC in the female population[22,23]. A single-center study reported significantly higher rates of MASLD in older women compared to younger women and older men[24]. Patients without screenable etiologies to cirrhosis are disproportionately identified later in the disease course; postmenopausal women may be part of this population[25,26]. Having said that, it is hard to blame a single risk factor for the observed findings, especially with the variation in environmental exposures, biologics, and socioeconomic status between different cohorts in the United States in different regions.

Our study has several strengths which include the stratified analysis by age and sex across different geographical regions, in addition to the large sample size (467344 patients), recent period (2001-2020), and the use of joinpoint regression with the BIC method and comparative analysis. With that in mind, our study is limited in the lack of clinical variables to assess for HCC risk factors. However, our manuscript is observational, and its epidemiological retrospective design is hypothesis-generating and aims to guide future efforts toward further investigations of the contributions leading to increasing HCC incidence across different United States populations and geographical locations. Our study suffer from other limitations inherent in large databases such as loss of records and coding reliability[27]. However, all the data used in our analysis were obtained from the USCS database which is the official source of cancer incidence data in the United States and undergoes many processes to ensure high-quality standardization and coding per the standards of the North American Association of Central Cancer Registries.

CONCLUSION

Our study offers insight into the epidemiology of HCC in different demographic groups across various United States geographical regions. While overall HCC incidence was increasing across all geographical regions, Southern states experienced the steepest increase. The non-decreasing trend in younger women was noted across different regions, compared to counterpart younger men who experienced a decreasing trend. The reasons contributing to our findings are unclear and can be related to sex and regional disparities in healthcare access and utilization. Future research is warranted to characterize the temporal change in HCC risk factors across different United States regions.