Racial Diversity in the Decline in Hepatocellular Carcinoma and Increasing Age at Diagnosis in a Primarily African American Medical Center Population

Simple Summary

Hepatocellular Carcinoma (HCC) occurs primarily in individuals with cirrhosis due to known risk factors. The most predominant in Western areas are Hepatitis C virus infection, excessive alcohol consumption, and Metabolic Dysfunction-Associated Steatohepatitis. This study focuses on a predominately African American Health Center and demonstrates that AA and non-AA individuals have racial disparity in the three major risk factors for development of HCC. The study also demonstrates that the decline in HCC incidence is most prominent in AA due to the elimination of HCV by anti-viral treatment. This contrasts with MASH and Alcohol, which remained constant as risk factors primarily in non-AA and requires addressing with lifestyle modifications as opposed to the use of therapeutic agents.

Abstract

Background: Hepatocellular carcinoma (HCC) remains a significant global health burden, particularly among vulnerable populations. This retrospective study investigates trends in HCC incidence and age at diagnosis within an urban medical center population, focusing on the impact of hepatitis C virus (HCV) treatment and racial disparities. Methods: The study includes 484 patients diagnosed with HCC between 2000 and 2023. Results: A significant decline in HCC incidence was observed with a peak in incidences between 2015 and 2017 (p < 0.02). The increase and subsequent decline were driven by a decline in HCV-related cases, particularly among the African American (AA) population. This trend was not seen for patients with other risk factors for HCC. An increase in age at diagnosis in HCV patients but not other risk patients was observed in AA (62 vs. 69 years p = 0.001) but not non-AA patients (66 vs. 67 p = 0.16). This increase in age for AA HCV patients could be due to an aging population, changing risk factor profiles, and/or limitations in surveillance and early detection of HCC. Conclusions: This study highlights the critical role of HCV treatment in reducing HCC incidence, particularly within the AA population. These findings emphasize the need for sustained efforts in surveillance, early detection, and targeted prevention strategies to address the evolving epidemiology of HCC and improve outcomes across all populations.

1. Introduction

Hepatocellular carcinoma (HCC) is a significant global health burden and the prognosis with respect to survival is poor [1,2,3,4,5]. Literature with respect to risk and the development of HCC provides a few consistent observations [6,7,8,9]. With respect to geographic differences in HCC incidence and risk factors, in the US, they are present due to racial/ethnic diversity of the populations in various states as related to risk factor variability. While alcohol abuse is a consistent risk factor in the population, risk factor variability is also predicted because of the shifting of risk from viral disease to metabolic disease. The male gender is dominant with respect to the development of HCC regardless of the etiology of risk although that may be shifting as metabolic syndromes become more dominant as risk factors. The presence of cirrhosis is also a common precursor to the development of HCC and is considered of significant enough risk that screening of patients with cirrhosis is recommended. There are also several, predominantly virus-related components of the development of HCC which offer the possibility of intervention in development of HCC. These include the successful eradication of hepatitis C virus (HCV) by short-term treatment with highly effective direct acting antivirals and the vaccination against hepatitis B. These interventions have resulted in a decline in HBV as a cause of HCC, especially in Asian patients as compared to other populations. Multiple studies have also identified that an important component of the poor prognosis is a lack of surveillance in at-risk individuals with cirrhosis, diagnosis of tumor sizes greater than 3 cm, and epidemiological variations in risk factors for various populations. An underappreciated aspect of survival and surveillance is the role of racial diversity in the risk factors for African Americans (AA) as compared to non-AA individuals. This is especially important with respect to hepatitis C virus (HCV) infection, where AA are significantly more likely to be infected as compared to non-AA [10,11,12,13,14,15,16,17,18,19,20,21,22,23]. This study investigates trends in HCC incidence and age at diagnosis within an urban medical center population with a focus on racial diversity and its impact on incidence over time. Given the enhanced risk of HCV in AA population, the impact of the possibility that successful treatment of hepatitis C virus (HCV) treatment will reduce the incidence of HCC in the medical center due to racial differences in HCV infection is a major focus of the project. A second issue to be evaluated in this study is the underlying risk factor of Metabolic Dysfunction-Associated Steatosis Liver Disease (MASLD) with respect to the development of HCC [24,25,26,27]. While both AA and non-AA are at risk for the disease with respect to the prevalence of obesity and diabetes, multiple studies have demonstrated that non-AA in comparison to AA are more at risk for advanced fibrosis (Metabolic Dysfunction-Associated Steatohepatitis (MASH)). The role that this observation plays in HCC incidence is incompletely understood. The third area of potential racial diversity that can be evaluated in our dataset is the role of excess alcohol consumption in both populations [28,29,30,31].

2. Materials and Methods

A retrospective cohort analysis was conducted using a dataset of 484 patients diagnosed with HCC between 2000 and 2023 at the Detroit Medical Center and recorded in the EMR. Patients for evaluation were identified by ICD-10 billing codes (ICD-9 155.0 and ICD-10 CC22.0). Confirmation of primary hepatocarcinoma by biopsy or imaging was required for inclusion in the study. Cholangiocarcinoma and metastatic liver carcinoma were excluded from the dataset. Since the data was collected over a period of time by multiple individuals, the data from the EMR was extracted using a case report form to assure that there was consistency of data extraction. The focus of data extraction for this project was demographic data including gender, age at diagnosis, self-reported race, and date of diagnosis. Risk factors for HCC were gathered from the EMR with the final ranking in the case of multiple risk factors being HCV as the major risk factor. In cases where no risk factor was readily available, patients were designated as cryptogenic primarily due to a lack of medical history in the EMR. HCV treatment prior to diagnosis, surveillance defined as imaging within one year of diagnosis, and tumor size at diagnosis were also collected on the case report forms. SAS-based data analysis software (JMP Pro 18.0.2; JMP Statistical Discovery LLC, Cary, NC, USA) was used for data analysis. The assessment of difference between risk for AA and non-AA was evaluated using Contingency Analysis and the significance of the variation was defined using the Pearson ChiSquare Probability evaluation. The evaluation of the age at diagnosis was plotted and evaluated using Oneway Analysis and ANOVA fit. Significance was defined using Probability for analysis of variance. This analysis was used for the relationship between age at year of diagnosis and the two factors: Race (AA bs non-AA) and HCC vs. Other risk factors.

3. Results

3.1. Demographic Overview

The 484 HCC patient population in the DMC were primarily African American and Male (

Table 1. Racial Diversity in Hepatocellular Carcinoma Diagnosis between 2000 and 2023.

Race	Number	% of Total	Age (Years)	Gender (M)	Etiology (%HCV)	HCV (%Not Treated)	Surveillance	Tumor Size (>3 cm)
AA	282	49	65	72%	81%	85%	11%	85%
Non-AA	86	18	64	71%	52%	84%	22%	78%
Unknown	16	3	67	72%	56%	78%	13%	67%

). The median age at diagnosis was 66 years and was similar between AA and non-AA patients. The primary etiology was HCV, although AA patients had a higher incidence than non-AA patients. Regardless of race, the majority of the patients with HCV-related HCC were not treated for HCV prior to HCC diagnosis. The rate of surveillance was only 12% overall but there was a lower rate for AA as compared to non-AA. The poor surveillance rate also contributed to the observation that regardless of etiology and race, tumor size at diagnosis for the majority of patients was >3 cm. The racial data presented in the subsequent graphs utilized only patients of known race and represent overall 96% of the patients seen at the DMC.

3.2. Risk Factor Distribution

The dominant risk factor for the development of HCC for both races was HCV (Table 1 and Figure 1). Metabolic Dysfunction-Associated Steatohepatitis (MASH) and Cryptogenic were the more likely to be risk factors in non-AA individuals as compared to AA. In this population, cryptogenic were likely MASH due to a history of metabolic risk factors (obesity, diabetes, hypertension, and lipidemia) which occur in the presence of steatoses, but a definitive diagnosis could not be determined due to a lack of early imaging data and early medical history for patients who were first seen as inpatients but were not previous patients in the medical center clinics. Alcohol risk was similar in both AA and non-AA individuals but the possibility that alcohol was a contributing factor in patients with other risk factors could not be excluded.

3.3. Temporal Trends

An increase in age at diagnosis was observed for AA patients (p = 0.001) who had HCV as their risk factor. (Figure 2). Age did not change in non-HCV AA patients, suggesting the possibility of a different time course (i.e., slower) for development related to HCV as compared to the other risk factors. In contrast, non-AA patients regardless of whether the risk was HCV or other did not show an increase in age over time. The increase is attributed to an aging population who acquired HCV significantly earlier, and despite the efficacy of anti-viral treatment, they were not treated.

3.4. Shifts in HCC Incidence

Figure 3 presents the shift in the incidence of HCC diagnosis in the medical center population over time. A significant shift in HCC incidence overall was observed in the medical center population with a peak in numbers between 2015 and 2017 (p < 0.02). The increase and subsequent decline were primarily driven by a shift in HCV-related cases. This trend was not as prominent for patients with other risk factors for HCC.

With respect to potential racial disparity, the African American (AA) population had a dominant role in the decline, due primarily to the higher incidence of HCV in the population (Figure 4). The continued high numbers of HCC patients in the AA population is attributed to a failure to actively identify and treat AA patients who continue to be dominant in the Medical Center population.

4. Discussion

The dominant observation from our study in the diagnosis of Hepatocellular Carcinoma (HCC) in our predominantly African American (AA) patient population is the decline in HCC over time. Since the medical center includes a National Cancer Institute-designated cancer center, there is no obvious reason that patients would go to a different medical center for diagnosis and treatment. The Medical Center also has seen an increase in overall patients over time and continues to see similar populations. It is unlikely that this decrease in HCC diagnosis represents a decline in the number of total patients being seen in the medical center and reflects a change in the number of HCC diagnosed patients. Thus, our study focused on the hypothesis that the decline was a result in a shift in the risk factors for disease development that resulted in fewer patients developing HCC due to HCV. The shifting epidemiology of hepatocellular carcinoma in our predominately African American population reflects both progress in disease prevention and emerging challenges. Nationally, HCC incidence increased over the past three decades but started to plateau or decline in younger and middle-aged adults, with the most notable increase occurring in adults over 60 years [1,2,3,4,5,6]. In our population, there was an increase in age at diagnosis but only for patients with HCV as their risk factors, which were more dominant in AA as compared to non-AA. Regardless of race, there was a steady age at diagnosis for patients with the other risk factors. This suggests that moving forward, we may see the diagnosis of HCC in younger populations as the slow-development cirrhosis and HCC incidence of HCV declines. With respect to the literature, this data is consistent with the national trend suggesting that earlier age at diagnosis may be seen as there is a reduction in age which is currently driven primarily by AA patients with untreated HCV. This may indicate an aging at-risk population driven by HCV infection, evolving risk factor profiles towards an age-related MASLD population, and/or gaps in early detection efforts within this demographic. Our data also suggests that additional efforts to target HCV treatment in our population will further decrease the incidence of HCC in the medical center. Given the dramatic impact of the decline in HCV as a risk factor in AA patients, increased effort in that area should pay dividends.

HCC incidence and mortality remain disproportionally high among racial and ethnic minorities [1,2,3,4,5,6,7,8,9,10]. Notably, the highest increase in HCV attributable HCC has been observed in Black patients (+5.3% annually), followed by White patients (+3.7%), while incidence has declined in Asian populations (−4.3%) [6]. Our cohort, which was predominantly AA (79%) and male (72%), aligns with the broader national data demonstrating higher HCV-driven HCC burden in minority populations, particularly Black individuals [1,2,3,4,5,6]. Our study, however, does not address the issue of whether AA patients have a higher incidence of HCC since the medical center population is predominantly AA (approximately 80%) and our HCC incidence is 76% AA. It does, however, allow us to compare potential differences in the % of risk factors in each population.

Consistent with most studies, our study also demonstrated a peak in HCC incidence between 2015 and 2017, followed by a decline. This observation is driven primarily by HCV-related cases within the medical center population. An issue that our data does not address is related to patients seen prior to 2010. Thus, even though the increase and decrease are driven by HCV, it is unclear whether the still significant number of HCV patients in 2010 and presumably prior to the data available in our study still reflects increasing incidence of HCV prior to 2010. Our data does reflect the broader national trend in which HCV-related HCC incidence initially rose then began to decline following the widespread adoption of direct-acting antivirals [5,6,7,8,9,10,11,12,13,14,15,16,17,18,19]. DAAs achieve sustained virologic response (SVR) in >90% of patients and reduce HCC risk by approximately 70%, though the risk persists in patients with cirrhosis who were treated with anti-virals [9,10,11,12].

Our data also demonstrates that despite these advances with respect to rapid and highly efficient cure rates for HCV, a substantial number of HCV-infected individuals remain untreated and continue to contribute to the development of HCC even in the recent era. Given that the DMC Medical Center provides services particularly among minority and socioeconomically disadvantaged populations, our study is consistent with the lower rates of HCV treatment among those populations. Our data also reflects that even though there is a decline in HCV-driven HCC, there continues to be a significant number of aging AA individuals yet to be treated. This is confirmed by the observation that HCV patients who develop HCC continue to increase in age with time. Given that HCV-related HCC continues to rise in certain subgroups, expanding access to HCV screening and treatment remains a critical priority. Our study reinforces the importance of sustained efforts to identify and treat at-risk individuals, particularly within the AA population where HCV has been the dominant HCC risk factor and has not continued to decline over the past two years.

As the incidence of HCV declines, non-viral risk factors, such as alcohol consumption and MASH, have been shown to increase in respect to the % of individuals with those risk factors in our population [24,25,26,27,28,29,30,31,32]. Our data is also consistent with the literature suggesting these are major drivers of HCC as the incidence of HCV-mediated HCC declines [1,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32]. In the U.S., a report from 2007 to 2017 showed that HBV/HCV-unrelated HCC, linked to metabolic and alcohol-related liver disease, is increasing annually, with notable increases among patients with diabetes or obesity [6,7,8,9,10,25,26,27,28,31]. Our findings, when combining MASH and cryptogenic cases, align with this national data, demonstrating that MASLD-related HCC is rising worldwide. MASH has become the second-leading cause of HCC deaths [1,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32]. With respect to the issue of cryptogenic cases, while they do not have a specific risk factor listed in the EMR, most are not viral- or alcohol-driven. The lack of an accurate medical history due to patients being diagnosed as inpatients without clinic visits in our system means we cannot accurately confirm the presence of metabolic syndrome, and the cirrhosis precludes the use of imaging to define the presence of liver steatosis. This issue of metabolic impact on the development of fatty liver and potential for fibrosis and cirrhosis raises the importance of more careful evaluation of the cryptogenic patients for co-morbidities such as diabetes and obesity. More recently, in 2022, MASH surpassed HBV as the third-leading cause of HCC-related death. Current trends predict that MASH will overtake HCV by 2032 as the second-leading cause, while alcohol-related liver disease, which was the second-leading cause of HCC-related mortality in 2022, is predicted to become the leading cause by 2026 [4,5,6,7,8,9]. In contrast to many studies, our patient population has similar rates of alcohol risk in both AA and non-AA patients. In non-AA patients, alcohol is lower as a risk factor than MASLD-related patients. This data, however, requires a caveat in that patients with HCV and a mention of alcohol were scored with HCV as their primary risk factor. Thus, alcohol as a co-risk factor needs to be further evaluated in both HCV and MASLD patients.

MASH-related HCC poses unique challenges because it can develop in the absence of cirrhosis, making traditional cirrhosis-based screening protocols insufficient to detect such cases [24,25,26,27]. There is also the possibility that alcohol at levels lower than that defined for designating alcohol as the dominant risk factor could influence the development of HCC. In fact, due to this possibility, recent studies have started to identify such patients as having MetALD and considering them as different from pure MASLD patients and alcohol abuse risk patients. Studies to identify and follow these patients are important since they could represent another area for risk stratification, intervention, and surveillance. Given that MASLD is becoming a more common condition, with an estimated prevalence exceeding 34% in the United States and 30% globally, expanding screening and surveillance strategies to include high-risk metabolic patients will be essential in reducing HCC-related mortality. Such an approach might only target certain populations such as Hispanic individuals, who face the highest burden of MASLD, while Black individuals face the lowest burden [5,24,25,26,27,28,29,30,31,32]. Also to be further evaluated is the relationship between MASLD and MASH and the value of using fibrosis evaluation to define patients who would benefit from HCC surveillance despite not having cirrhosis. It is also important to note that several newer therapies related to MASLD/MASH have been approved. Theu include a variety of GLP-1 agonists which can result in significant weight loss and control of diabetes and Resmetirom which is approved for reducing fibrosis in MASLD/MASH patients.

Our data is also somewhat surprising in that alcohol-related HCC is lower as compared to both HCV and MASLD/MASH patients. While significant public health efforts aimed at reducing alcohol-related liver disease (ALD) have been attempted, their success rates have been somewhat mixed. Programs like policy changes to decrease exposure to alcohol advertisements, taxes, and limiting alcohol sales have been associated with a lower ALD mortality [5,28,29,30,31]. Other options such as improving awareness of ALD as an HCC risk factor and expanding treatment options for alcohol-related liver disease are also crucial in addressing this potential burden of HCC in the future. An area of concern but underappreciated and poorly evaluated is the possibility that even moderate alcohol can impact both HCV and MASLD/MASH as drivers of HCC development. Further evaluation of the role of alcohol as an important co-factor in the development of HCC remains an important area for further research.

5. Conclusions

Our study provides insight into the evolving landscape of HCC risk factors and racial disparities by utilizing a predominately AA clinic population. The decline in HCV-related HCC highlights the success of antiviral therapies, but the continued burden of an aging untreated HCV population underscores the need for ongoing screening and treatment efforts. As MASH appears to be increasingly driving HCC incidence, adapting surveillance programs to identify high-risk metabolic patients and promote lifestyle interventions will be essential. Implementing better screening strategies for metabolic-associated HCC continues to be challenging in most patient populations. Our study also demonstrated the need for addressing racial and socioeconomic disparities with respect to improving early detection and treatment access.