Sex, racial, and ethnic disparities in United States liver transplantation clinical trials

Abstract

BACKGROUND

Regulatory agencies are increasingly recognizing that minority trial representation is inadequate, contributing to healthcare disparities. The scope of minority population disparities in clinical trial participation remains unclear, as previous studies have compiled enrollment data from published trials, which frequently do not report participant race and ethnicity.

AIM

To evaluate sex, racial and ethnic inequities in liver transplantation (LT) trials participation in the United States.

METHODS

We used data from completed United States liver transplant clinical trials registered and reported on the National Institute of Health (NIH) website (clincaltrials.gov). Demographic data, including race, ethnicity, sex, and age were collected. To make inferences to a larger population, 95%CIs were computed for estimates in each demographic group using the Wilson method for binomial proportions. We also computed the simultaneous 95%CIs by applying a Bonferroni correction to reflect the multinomial distribution of race proportions. The numbers and percentages of racial/ethnic minority and female individuals compared with United States census data from 2010 and 2018. Secondary outcome measures were inclusion by trial funding source and year of completion.

RESULTS

A total of 69 United States based clinical trials involving 6990 participants were included in the analysis. Of these, 35 trials (51%) were randomized, and 26 (38%) were conducted across multiple United States regions. All trials reported sex, while 42 (61%) reported race and 27 (39%) reported ethnicity. Compared to United States census data, Asian individuals were overrepresented (9.3%; 95%CI: 8.1%-10.5%), whereas African American (7.8%; 95%CI: 6.7%-8.9%) and American Indian or Alaska Native individuals (0.4%; 95%CI: 0.1%-0.6%) were underrepresented. The proportion of White participants (75.9%; 95%CI: 74.1%-77.7%) was consistent with census estimates. Hispanic participants were underrepresented (13.3%; 95%CI: 12.2%-14.5%) regardless of the census year referenced. In industry-sponsored trials, Asian representation was three times higher than in the general population (15%). NIH funded trials showed overrepresentation of White participants (83.8%) and underrepresentation of Black participants (4.1%) relative to census data. Women comprised 31.1% of all participants (95%CI: 30.0%-32.2%), indicating underrepresentation. Among trials that reported racial data, 62 (90%) did not include participants of American Indian or Alaska Native, Native Hawaiian, or Pacific Islander descent.

CONCLUSION

Our analysis indicates that women, African Americans, and Hispanic individuals are underrepresented in LT clinical trials compared to the general United States population. These results highlight the need for regulatory initiatives aimed at enhancing the inclusion of historically marginalized racial and ethnic groups in clinical research.

Key Words: Inequities; Liver transplant; Clinical trials; Gender; Race; Ethnicity

Core Tip: This cohort study demonstrates that women, African American, and Hispanic individuals were underrepresented in liver transplantation clinical trials compared to the general United States population. These results underscore the need for regulatory efforts to enhance the inclusion of historically underrepresented racial and ethnic groups in clinical research.

INTRODUCTION

Liver transplantation (LT) is a life-saving intervention for patients with acute or chronic end-stage liver disease (ESLD), a condition associated with high morbidity and mortality[1,2]. In 2021, more than 9000 patients underwent LT in the United States, while approximately 25000 were listed for transplantation[3]. The importance of LT lies in its capacity to significantly extend survival and improve the quality of life for patients with ESLD. According to the Centers for Disease Control and Prevention (CDC), over 27000 deaths in 2019 were directly attributed to chronic liver disease and cirrhosis[4-6]. Over the past five years, the number of LTs performed in the United States has increased by 18%[1], reflecting both a growing demand and expanding clinical capacity.

Despite these gains, a persistent and troubling disparity remains between the number of patients in need and the number who ultimately receive a transplant. Each year, approximately 12000 individuals are added to the LT waitlist, yet nearly 3000 die or are removed before receiving an organ[8]. Mounting evidence suggests that this gap is influenced not only by organ availability but also by systemic disparities related to race, ethnicity, gender, and geography[9].

Research shows that women may be less likely than men to be evaluated and listed for LT, even when controlling for disease severity and socioeconomic status[10]. A large-scale analysis from the Organ Procurement and Transplantation Network revealed that women were significantly less likely to receive a transplant than men[11]. Similarly, African American and Hispanic patients are underrepresented among LT recipients compared to their White counterparts, as demonstrated in national registry data from the United Network for Organ Sharing[12-14].

A critical barrier to addressing these inequities is the underreporting and inconsistent inclusion of race, ethnicity, and gender data in clinical trials related to liver disease and transplantation. This lack of demographic representation contributes to a limited understanding of how diverse populations respond to treatment, raising concerns about the generalizability and safety of therapies for minority groups[15]. CDC surveillance data from 2018 further underscores these disparities, showing that liver disease disproportionately affects men, individuals aged 65-74, Hispanic/Latino populations, and those with lower educational attainment[16].

These findings highlight the urgent need for equitable access to LT and more representative clinical research. As liver disease affects a diverse cross-section of the United States population, achieving equity in transplant outcomes requires a comprehensive understanding of how gender and racial/ethnic disparities impact access to care. Therefore, this study aims to assess current disparities in LT within the United States, focusing specifically on the representation of women and minority groups. Through this analysis, we seek to contribute to the ongoing effort to ensure that LT is accessible and equitable for all who may benefit from it.

MATERIALS AND METHODS

In this cross-sectional study, data were used from completed United States liver transplant clinical trials through December 2020, which were registered and reported on clinicaltrials.gov. On October 10, 2023, the National Institute of Health (NIH) website (clinicaltrials.gov) was searched for United States clinical and surgical trials that were done in the past two decades, January 2000 to December 2020. The terms “liver transplant” and “liver transplantation” were used to identify human trials on the website. To be included, trials had to be categorized as “Completed” or “Active, Not recruiting” status, indicating an established, finalized patient cohort. Only trials with available results have been included. Exclusion criteria included: (1) Trials that were completed but without results reported; (2) Trials with unknown status, and those that were not yet recruiting, still recruiting, enrolling by invitation, active but not recruiting, suspended, terminated, or withdrawn; and (3) Trials based primarily outside the United States. In compliance with the 45 CFR §46, this study was exempt from institutional review board approval as the data were deidentified and publicly available. Informed consent was not required, given that the data used did not contain any identifiable information. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology reporting guideline.

Data collection

Demographic data—including race, ethnicity, sex, and age—were independently collected by two reviewers (Alsakarneh S and Khalid Z) to ensure accuracy. Sex was classified as male or female based on the designation reported on ClinicalTrials.gov, aligning with biological classification. While sex reflects biological differences, we recognize that gender is a socially constructed identity, which may influence healthcare access and outcomes, but was not consistently reported across trials. For race and ethnicity, we adhered to the Office of Management and Budget Standards for the Classification of Federal Data on Race and Ethnicity, as used on ClinicalTrials.gov. Ethnicity was categorized as Hispanic/Latino or non-Hispanic/Latino. Race was categorized as White, African American, Asian, American Indian or Alaska Native, Native Hawaiian or Other Pacific Islander, more than one race, or unknown/missing/not reported. Trials were further classified by funding source into four categories: NIH, other United States federal agencies, industry, and other (individuals, universities, or organizations). We also classified trials by phase (1, 2, 3, or 4) and recorded both the calendar month and year for trial initiation and conclusion to assess changes over time. For population comparison, demographic distributions were referenced using United States population data from the 2010 and 2018 American Community Surveys.

Statistical analysis

For statistical purposes, we used SPSS, version 27.0 (IBM) and R, version 4.3.1 (R Project for Statistical Computing). In cases where the cumulative counts of distinct ethnic groups did not add to the overall number of participants in the trial, we adjusted that by reallocating them within the "other/unknown/missing" category. This was done to ensure that the total number of participants across all ethnic groups match the total number of participants under analysis in the trial. To make inference from our study to a larger population, using the Wilson method for binomial proportions, we computed 95%CIs for estimates of the percentage in each demographic group. To reflect multinomial distribution of race proportions, simultaneous 95%CIs were computed by applying a Bonferroni correction using a 2-sided α level of .05 divided by 7 for racial group estimates.

RESULTS

As detailed in Figure 1, the initial search identified 87 trials, of which 18 studies were excluded. 69 US based clinical trials with a total of 6990 participants met the inclusion criteria and were included in the final analysis. Of those, 35 (51%) were randomized and 26 (38%) included multiple regions in the United States. Only 12 studies included international testing sites (Figure 2). All included trials reported the funding source, with NIH being the least funding source (9, 13%) and other (individuals, universities, organizations) being the highest (39, 57%). The percentage of studies reported age (63, 91%), sex (69, 100%), race (42, 61%), and ethnicity (27, 39%) varied.

Figure 1 Flow diagram of study selection. A total of 87 trials were identified through ClinicalTrials.gov (July 2002-July 2018). After screening, 18 trials were excluded due to irrelevant populations or interventions. Sixty-nine trials met inclusion criteria and were included in the final analysis.

Figure 2 Distribution of clinical trials across regions of the United States.

Race and ethnicity

The number and percentage of participants in each demographic group for the 69 trials compared with US census data from 2010 and 2018 are presented in Table 1. Among trials reporting race, Asians were over-represented at 9.3% (95%CI: 8.1%-10.5%). While African American participants (7.8%; 95%CI: 6.7%-8.9%) and American Indian and Alaska Native participants (0.4%; 95%CI, 0.1%-0.6%) were underrepresented, White participants appeared to be equally represented (75.9%; 95%CI: 74.1%-77.7%), independent of year of census. Similarly, Native Hawaiian and Pacific Islander participants (0.2%; 95%CI, 0.0%-0.3%) and those reporting multiple races (0.4%; 95%CI: 0.1%-0.7%) were underrepresented as well. For ethnicity, Hispanic or Latino participants (13.3%; 95%CI: 12.2%-14.5%) were underrepresented independent of year of census.

Table 1 The number and percentage of participants by demographic group, n (%)¹.

Characteristics	Participants, (95%CI) (n = 4048)	United States population in 2010 (%)	United States population in 2018 (%)
Race
White	3098 (75.9) (74.1-77.7)	75.5	76.3
Black or African American	319 (7.8) (6.7-8.9)	13.6	13.9
Asian	379 (9.3) (8.1-10.5)	6.3	5.9
American Indian or Alaska Native	15 (0.4) (0.1-0.6)	1.3	1.3
Hawaiian or Pacific Islanders	8 (0.2) (0.0-0.3)	0.3	0.2
Multiracial (> 1 race)	17 (0.4) (0.1-0.7)	3.0	2.8
Other, unknown, missing	248 (6.1) (5.1-7.1)
Ethnicity
Hispanic or Latino	2823 (13.3) (12.2-14.5)	19.1	18.5
Sex
Female	2175 (31.1) (30.0-32.2)	50.4	50.8

¹Wilson 95%CIs are given. For race, simultaneous 95%CIs were calculated using a Bonferroni correction for the 7 racial categories. From United States census data in the respective year. Sex was reported for all participants (100%).

Among the 41 trials reporting data regarding race, only 7 (10%) trials included American Indian and Alaska Native participants, and 7 (10%) included Native Hawaiian and Pacific Islander participants. Conversely, 100% of trials included White participants. During the study period, there was significant variability in percentage of trials reporting race and ethnicity (Figure 3). For trials conducted in 2002-2005, 0% trials reported ethnicity. However, 50% of trials done 2015-2018 reported ethnicity. In most years less than half of trials report ethnicity.

Figure 3 Reporting of race/ethnicity of participants in liver transplantation trials.

When stratified by funding source, racial and ethnic minorities enrollment varied widely. Hispanic or Latino participants were underrepresented regardless of the funding source (Figure 4). In industry funded trials, Asian participants (15%) were overrepresented as triple as their representation in the United States population. In NIH funded trials, White (83.8%) participants were overrepresented and Black (4.1%) participants were underrepresented compared to the general United States population.

Figure 4 Participant demographics by funding source. Industry-funded trials showed higher Asian participant representation, while National Institute of Health-funded trials demonstrated underrepresentation of African American participants and overrepresentation of White participants.

Sex

Overall, females were underrepresented in liver transplant clinical trials (31.1%; 95%CI: 30.0%-32.2%) compared to the general United States population, independent of year of census. When analyzed by funding source of clinical trial, females remained underrepresented in all trials regardless of funding source (Figure 4). Similarly, across years, women were underrepresented independent of year of trial (Figure 5).

Figure 5 Participants’ representation in liver transplantation trials based on demographics and year. Women accounted for 31.1% of all trial participants, significantly below their proportion in the general United States population.

DISCUSSION

A total of 6990 participants were included from 69 United States-based clinical trials. Significant disparities were noted in participant demographics compared to the general United States population. Women were consistently underrepresented, regardless of census year, trial year, or funding source. Racial and ethnic representation varied—Asians were overrepresented, Whites were equally represented, while African Americans, American Indian or Alaska Natives, and other minorities were underrepresented. Hispanic/Latino participants were also underrepresented across trials. Although all trials included White participants, only 10% included American Indian or Alaska Native participants. Recent trials showed improved reporting of ethnicity, suggesting growing attention to health equity.

Given the shifting epidemiology of liver disease, the gender gap in LT is alarming[17]. A national survey showed increased alcohol use among women before and after coronavirus disease-2019. Women are more vulnerable to alcohol-associated liver disease (ALD) at lower alcohol consumption[18], have faster fibrosis progression, and worse outcomes despite abstinence[19]. Between 2009-2015, alcohol-related cirrhosis rose by 50% in women vs 30% in men[20]. NHANES data showed ALD with stage ≥ 2 fibrosis in women doubled from 0.5% to 1.3% between 2015-2016[21]. Women also face an 83% higher risk of death from steatotic liver disease[22] and may be disadvantaged by the model for end-stage liver disease (MELD) score due to lower muscle mass and creatinine levels[23,24]. Despite this, sex-based MELD adjustments have not fully corrected the disparity—LT deficits in women have increased by 14% in the MELD era[24-26].

Ethnic minorities face similar challenges. Despite higher needs, they are underrepresented in LT trials and less frequently referred for evaluation[27,28]. A National Inpatient Sample (2016-2020) showed Black patients had lower LT rates for HCC and higher post-LT mortality (7% vs 1%) compared to Whites[29]. Hispanic candidates had higher waitlist mortality despite lower disease severity[28], and Hispanic, Asian, and American Indian/Alaska Native patients faced higher waitlist death risk than NH Whites[30]. The Stanford Integrated Psychosocial Assessment for Transplantation, a transplant psychosocial tool, may inadvertently disadvantage candidates based on race, education, and insurance status—penalizing non-White, uninsured, or less-educated individuals[31-33]. The NIH Revitalization Act (1993) aimed to boost minority trial participation[34]. The MELD system improved objectivity in organ allocation, but disparities persisted, prompting MELD 3.0 to incorporate sex as a variable—yet ethnicity remains unaddressed[35]. NIH strategies now include mandatory disaggregated data collection, improved equity in data reporting, and attention to social determinants[36-38].

Despite these efforts, our analysis reveals continued underrepresentation of women and minorities. Industry-funded trials notably overrepresented Asians, possibly due to geographic factors not captured in our data. NIH guidelines call for proportional inclusion, yet White participants comprised 83.8% of enrollees, while African Americans and Hispanics were underrepresented[39-41]. This highlights the need for stricter recruitment protocols across all funding types to ensure equitable trial representation[15].

These disparities undermine the generalizability of trial results and may perpetuate inequities in LT referral and outcomes. Poor female and minority representation can obscure sex- or ethnicity-specific risks and benefits, delaying needed clinical recognition. To improve equity, future trials must set inclusion targets, adopt culturally tailored recruitment, and enforce transparency and accountability from sponsors and regulators. Without reform, disparities in LT access and outcomes will persist, hindering equitable, patient-centered care.

Limitations

Our study should be interpreted in the context of some limitations. First, the socioeconomic status of the included patients could not be ascertained. Second, the epidemiology of liver diseases has changed over the 20 years of our analysis, which may have affected specific population inclusion. Third, different centers have varying inclusion criteria for LT, which may also change over time. Fourth, since most trials did not report ethnicity, our findings may not be generalizable to all patients. Finally, not all United States trials are registered in ClinicalTrials.gov, so our findings are biased towards those covered by FDAAA 801, although these trials are generally considered the most clinically relevant.

CONCLUSION

In conclusion, having a deeper understanding of the disparities in LT requires a detailed approach based, in part, on the specific disease indications for the transplant. This is essential to formulate effective research and clinical strategies for improvement. These findings highlight the necessity to re-evaluate present approaches to decrease the current disparities and promote more diversity. The majority of reported trials were performed in an era where the male population generally had more advanced liver diseases, which may explain their higher representation. This is specifically important to understand and develop effective, equitable, and generalizable LT strategies that represent the dynamic change in liver disease distribution in demographics. A collaborative effort from researchers, investigators, funding agencies, and national policy makers is required to implement more inclusive recruitment strategies with a standard reporting system to ensure equitable access for LT.