Gender Disparity in Enrollment in Clinical Trials for Hairy Cell Leukemia Treatments in the Last 40 Years

ABSTRACT

Background

Hairy Cell Leukemia (HCL) is a B‐cell lymphoproliferative disorder that predominantly affects males, yet recent evidence suggests a notable gender participation gap in HCL clinical trials. This study aims to characterize that disparity and explore potential factors contributing to the under‐enrollment of females.

Methods

In this descriptive, retrospective study, we searched EMBASE, PUBMED, Cochrane Central, and ClinicalTrials.gov from January 1983 to December 2023 for publications on clinical trials (CT) in HCL, descriptive statistical analysis of all the sociodemographic variables was performed.

Results

We analyzed 57 clinical trials totaling 4595 HCL patients, with 79.1% male and 20.9% female participants. The male‐to‐female ratio declined from 5.91 (1983–1993) to 4.19 (2014–2023). Although the gender gap narrowed over time, female participation slightly decreased to 19.2% in the most recent period (2014–2023).

Conclusions

Female enrollment in HCL clinical trials remains disproportionately low compared to incidence rates, underscoring the need to address underlying barriers to improve equity in clinical research and treatment outcomes.

Trial Registration

The authors have confirmed clinical trial registration is not needed for this submission.

1. Introduction

Hairy cell leukemia (HCL) is a B‐cell lymphoproliferative disorder characterized by infiltration of the bone marrow, spleen, lymph nodes, and extranodal tissues by malignant B cells with hairy cell cytoplasmic projections [1]. This condition manifests in two primary forms: hairy cell leukemia (HCLc), characterized by the BRAF V600E mutation, and the variant hairy cell leukemia (HCLv), devoid of the previously mentioned mutation and recently subclassified by the WHO into a separate unrelated B‐cell disorder, splenic lymphoma with prominent nucleoli (SLPN). This entity constituted only 10% of all HCL cases and was associated with a less favorable prognosis. However, because of its prior clinical association with HCL, patients with SLPN were included in many of the clinical trials (CT) evaluating treatment and outcomes in HCL [2].

When originally described by Bouroncle et al. [3] in 1958, patients with HCL had a median survival of only four years due to the lack of effective treatments. Since then, new treatment approaches have greatly improved progression‐free survival; however, these regimens remain non‐curative, and relapse continues to pose a significant challenge. With fewer than 2000 new cases annually in the United States, HCL is considered a rare disease [4].

A notable feature of HCL is the striking gender disparity, often reported as approximately four times more cases in males than in females [5]. Yet this discrepancy cannot be attributed to incidence alone. Furthermore, over the past four decades of HCL research, a pronounced gap in clinical trial enrollment has emerged, with male participation consistently outpacing that of females. Although the precise reasons behind this prevalence gap remain elusive, possible contributing factors include genetic predisposition, hormonal influences, and a range of environmental exposures, either singly or in combination [6, 7, 8]. For instance, studies from the early 2000s have suggested an increased risk of cattle farmers or individuals exposed to pesticides, petroleum products, diesel, and ionizing radiation [9]. Interestingly in a descriptive study in Mexico, HCL was particularly more common in farming areas, despite being a country with lower HCL incidence compared to other geographical areas [10]. Some studies also suggest males are more susceptible to proto‐oncogenic mutations and potentially more prone to a heightened incidence of hematologic malignancies [11, 12].

Despite multiple studies failing to establish a causative link between these occupational risk factors, the difference in gender‐related occupational exposures could offer a plausible explanation for the gender disparity in HCL prevalence. The 2023 analysis of the HCL Patient Data Registry (HCL‐PDR) reveals gender‐related differences in response to therapy, finding more favorable outcomes for females [5]. The indolent course of the disease in certain patient populations may also contribute to female under‐representation in CT as patients with less aggressive forms are less likely to necessitate experimental treatments.

The root factors behind persistent gender disparities in both HCL incidence and clinical trial enrollment remain unclear and may be multifactorial. To investigate this significant knowledge gap we conducted a descriptive, retrospective study analyzing HCL CT between January 1983 and December 2023. This analysis explores key physio‐pathological, sociodemographic, and accessibility factors, as well as awareness limitations, that may collectively contribute to the observed under‐representation of women in HCL studies. By identifying and examining these factors, we aim to inform strategies for more equitable research designs and ultimately improve outcomes for all patients with HCL.

2. Methods

We developed a search strategy using Medical Subject Headings (MeSH) related to HCL and CT. We searched four databases (EMBASE, PubMed, ClinicalTrials.gov, and Cochrane Central) using the following search string: (Hairy Cell Leukemia) AND (Clinical Trials). The search results were downloaded into Rayyan software to remove duplicates. The search includes all CT published between the January 1, 1983 and December 31, 2023.

2.1. Outcomes

The primary outcome was the proportion of female participants in HCL CT over four decades. Secondary outcomes included regional differences, study design characteristics, and changes in male‐to‐female ratios over time.

2.2. Study Selection

Two reviewers (O.B.M. and A.B.) independently searched the title and abstract for potentially eligible CT, on treatments for HCL. Studies indexed in the databases were eligible in this review if they met the following criteria: (1) Describing therapeutic efficacy and/or safety of HCL treatments, and (2) gender not a part of exclusion criteria. Single‑arm trials were not excluded provided they met the above criteria. Studies only among pregnant women were excluded. Studies that did not report the number of male and female participants were also excluded. The review was not limited to language.

2.3. Data Extraction

Reviewers (O.B.M., A.B., and T.E.G.) independently extracted data into a Microsoft Excel spreadsheet. Data extracted included the year of publication of the abstract or full manuscript, country, study design, number of participants, and the proportion of male and female participants. To ensure consistency, extracted data was compared between reviewers, and disagreements were discussed until a consensus was reached.

2.4. Quality Appraisal

Reviewers (O.B.M. and A.B.) independently assessed each included study for the risk of bias. A third reviewer (T.E.G.) arbitrated possible differences. Randomized controlled trials (RCT) were evaluated using the Cochrane Collaboration's Risk‐of‐Bias Tool 2.

2.5. Risk of Bias Assessment

The risk of bias in studies included was assessed by adapting the Cochrane Risk of Bias tool for randomized controlled trials. Since patient‐centered outcome variables were not analyzed in this review, we evaluated the risk of bias in randomized studies based on three domains related to participant enrolment: random sequence generation, allocation concealment, and blinding of participants and personnel. Two reviewers (O.B.M. and A.B.) independently assessed the risk of bias in each article (See Figure S1).

2.6. Descriptive Statistical Analysis

Descriptive statistics were performed to obtain proportions and ratios. For the male‐to‐female ratio, we used a weighted average, taking into account the number of participants in every clinical trial within the time range. The enrollment incidence disparity index was calculated for both male and females.

3. Results

The search rendered 629 studies, 42 duplicates were removed and 587 were deemed ineligible after screening titles and abstracts. The remaining 66 full texts were screened, of which 57 studies were selected for data extraction and analysis (see Figure 1), comprising 4595 individuals with different subtypes of HCL (see Table S1). Publication dates were 1986–2023. Studies were conducted in more than 15 countries with 33.3% of the studies in the United States, 31.57% in Western Europe, 5.26% in Eastern Europe, 3.50% in Japan, 1.75% in Iran, and the remaining 24.56% being multicenter, multi‐country studies. Of the total 57 CT, 79.10% (n = 3635) of the patients were male and 20.89% (n = 960) were female. The median number of patients per study was 44 ranging from 10 to 861 with a standard deviation of: 127.16. Two CT included only male participants [13, 14].

FIGURE 1.

PRISMA flow chart to illustrate the flow of studies through the review and the selection process.

The male‐to‐female ratio weighted average was 5.91 for clinical trials (CT) performed between 1983 and 1993, 4.02 for CT performed between 1994 and 2003, 3.84 for CT performed between 2004 and 2013, and 4.19 for CT performed between 2014 and 2023. A total of 23% (n = 13/57) of the CT were randomized. Male participation consistently remained higher throughout the four decades (See Figure 2 and Table S2).

FIGURE 2.

This figure illustrates the percentage of male and female participants in hairy cell leukemia (HCL) clinical trials across four time periods. Male participation consistently remained higher throughout the four decades. In the earliest period (1983–1993), males comprised 84.3% of trial participants, while females accounted for only 15.6%. Over the following decades, there was a slight increase in female participation, peaking at 24.6% between 2004 and 2013. However, this trend reversed slightly in the most recent period (2014–2023), with female representation decreasing to 19.2%, while male participation rose to 80.7%.

A total of 24.56% (n = 14/57) of CT included exclusively patients with relapsed disease. Only 10.52% (n = 9/57) of the studies excluded pregnant women in their exclusion criteria. The enrollment incidence disparity and the enrollment incidence ratio for females were: −9.10 and 0.69 respectively (See Table 1 and Figure 3).

TABLE 1.

Enrollment incidence disparity for males and females in HCL clinical trials.

	Males	Females
Enrollment proportion in clinical trials	0.791	0.208
Estimated incidence in the general population	0.7	0.3
EID a	+9.10	−9.10
EIR b	1.13	0.69

^a EID: Enrollment incidence disparity, calculated as the difference between the enrollment proportion and the expected proportion based on incidence.[15]

^b EIR: Enrollment incidence ratio, which reflects the ratio of enrollment to the expected proportion, is 1.13 for males and 0.69 for females, further demonstrating the gender imbalance in clinical trial participation.[15]

FIGURE 3.

*EID: Enrollment Incidence Disparity, calculated as the difference between the enrollment proportion and the expected proportion based on incidence. °EIR: Enrollment Incidence Ratio, which reflects the ratio of enrollment to the expected proportion, is 1.13 for males and 0.69 for females, further demonstrating the gender imbalance in clinical trial participation.

4. Discussion

This systematic review synthesizes four decades of CT in HCL, revealing a persistent underrepresentation of females relative to incidence rates, despite some improvement in more recent trials. We identified 57 clinical treatment trials of HCL published from 1983–2023, enrolling 4595 patients, and found that more than two‐thirds of the patients were males. The study also reveals that the average weighted male‐to‐female ratio has changed over the decades, with a more significant disparity seen in the initial studies performed in the 20th century followed by an increase in female participation in some contemporary trials. HCL is known to be significantly more prevalent in biological males and previous observational studies have suggested a male‐to‐female ratio of 4:1 [4, 5, 16]. Our findings suggest that while the biological underpinnings of HCL may partially explain the overall higher incidence in males, multifactorial societal and methodological factors likely drive the observed under‐enrollment of females in CT. Addressing these barriers is crucial for the equitable advancement of HCL therapies.

Several theories have been proposed to account for the gender difference in prevalence, but the evidence behind this disparity remains unclear. Potential determinants such as biological, epigenetic, and occupational factors have been suggested [8, 16]. In our study, the male‐to‐female ratio of patients participating in HCL CT between 1983 and 1993 was 5.91. Considering the treatment options available during that timeframe and the probable demographics of the studied population it is reasonable to infer that the observed gender gap was anticipated. For instance, observational studies have linked pesticide exposure to HCL risk [9]. Historical data indicates a predominance of males in farming in the United States during the early 20th century [17]. This, coupled with the delayed implementation of farming mechanization leads us to hypothesize that the initial clinical trial participants may have encountered distinct occupational exposures compared to later generations [18]. Furthermore, an epidemiological study conducted in Norway from 1927 to 2007 revealed significantly higher tobacco consumption in males, particularly before 1970, estimating that over 70% of cigarettes were smoked by men in Norway since 1927 [19]. This underscores potential explanations for gender disparity in HCL prevalence.

Despite the well‐known gender differences in prevalence, our investigation demonstrated that there is an underrepresentation of females in HCL CT enrollment when compared to the current incidence [20]. The calculated enrollment incidence disparity index was −9.10 in females, which implies underrepresentation. Previous studies have examined the gender disparity in the enrollment of CT directed to cancer‐related therapies [21, 22]. Women were less likely to be enrolled in FDA‐approved studies for colorectal cancer, lung cancer, and surgical oncology trials [22, 23]. Multiple factors could account for this underrepresentation like barriers to enrollment, such as inconvenience, distrust of researchers, the need to maintain daily logs or specific treatment regimens, the amount of travel, the number and length of follow‐up visits, and disruption of daily life activities as well as family responsibilities [24].

Another important consideration is the sex‐related outcomes of HCL. In a recent observational study, women were shown to have a significantly longer progression‐free survival than male patients [5]. An Israeli study also found statistically significant better overall survival in females [25]. From our review, in the last 40 years, many of the studied treatments were focused on HCL patients with relapsed disease. The percentage of CT that enrolled exclusively patients with refractory disease was remarkable (24.56%). Consequently, it is possible to infer that males would be more likely to be enrolled as they tend to have a higher incidence of refractory disease and unfavorable outcomes, requiring second and third‐line treatments and also experimental therapies [5].

The geographical areas where the CT were performed are also important to note. Most CT on HCL were performed in the United States and Western Europe at 33.3% and 31.57%, respectively. These are areas with a higher percentage of Caucasians, known to be affected by the disease at higher rates than other ethnicities [3, 4, 16]. The second largest percentage of studies were multi‐center and multi‐country trials, accounting for 24.56% of the reviewed studies. Cultural differences in certain countries can impact the representation of a particular gender in the design of the trial. For instance, a recent study examined gender disparity in Middle Eastern countries to evaluate the underrepresentation of Arab women in cardiovascular disease‐related CT. They found that only 38.02% of patients in 71 trials were female [26]. However, in our study, there was only one Iranian study with no significant differences in gender parity when compared with the trials in other geographical regions.

The reason for the underrepresentation of women in HCL trials can be attributed to multiple different causes. Women in different societies are more likely to be partially or totally immersed in childcare and support responsibilities, which explains additional barriers to an accurate gender representation in cancer‐related CT [27]. Our systematic review provides a window into how gender disparities in HCL CT have evolved—yet persist—over the past four decades. These findings hold broad implications for clinical practice, health policy, and patient advocacy. The imbalance likely reflects a confluence of social and logistical barriers, as well as limited awareness and referral pathways for female patients. Moving forward, comprehensive strategies are necessary to achieve equitable representation in HCL studies, including active efforts to engage women during trial recruitment, closer monitoring of gender‐specific trial metrics by regulatory agencies, and broader patient advocacy initiatives. By bridging these gaps, we can facilitate more robust data generation, ultimately leading to improved, tailored outcomes for all individuals affected by HCL.

4.1. Strengths and Limitations

This study has several strengths, including its extensive timeframe covering four decades of HCL CT, rigorous literature search across multiple databases, and detailed demographic analysis highlighting gender disparities. However, there are notable limitations to consider. Potential publication bias exists due to exclusion of trials that did not report gender distribution, possibly influencing representation. Geographical bias is evident, as most studies originated from the United States and Western Europe, reducing generalizability to other regions. In addition, lack of adjustment for disease severity may partially account for female underrepresentation, as women often experience less aggressive disease courses. Addressing these limitations through expanded geographical representation, inclusion of previously unreported trials, severity‐adjusted analyses, and focused efforts to reduce selection and referral biases will enhance future research. We utilized enrollment incidence disparity index (EID) and the enrollment incidence ratio (EIR) to quantify gender disparities due to their simplicity and suitability for directly comparing trial enrollment proportions against known incidence rates. These indices align with our strictly descriptive objective: to determine whether the proportion of women and men enrolled in HCL studies mirrors epidemiologic reality. Given that our dataset comprises aggregate counts and that many trials include small samples, conventional inferential tests such as chi‑square or trend analyses would add limited value and could produce unstable estimates. We therefore focused on the magnitude and direction of disparity rather than formal hypothesis testing, a choice we acknowledge as a limitation while emphasizing its practical relevance for rare‑disease research.

5. Conclusions

After screening 587 studies from various databases and extracting data from 57 CT on HCL treatments we have concluded that there has been female underrepresentation in HCL CT in the last 40 years with a percentage of 20.89% of female participants. The average weighted male‐to‐female ratio has changed over the decades noticing more gender disparity in the initial CT done in the 1980s decade. The reasons behind this gender disparity are wide and difficult to predict. Having a considerable number of CT focusing on treatments for relapsed disease has also played a role in female underrepresentation, as there is a higher percentage of female patients having better outcomes and likely not requiring second‐line treatments. Efforts directed towards reducing barriers for female representation should be continued and encouraged when performing CT.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supplemental Figure 1. Assessment of the risk of bias in clinical trials. Quality tool used: Cochrane risk‐of‐bias tool for randomized trials Version 2.

Supporting Information

Data Availability Statement

As no new primary data were generated, there are no additional data underlying this article. Interested researchers may request further details regarding data collection and analysis methods from the corresponding author.