Abstract
Background
Hemophilia carriers (HCs) experience varied bleeding tendencies affecting their quality of life, yet their bleeding phenotype and management remain poorly understood. Since majority of the studies in carriers have focused on women of reproductive age, an unmet need exists to characterize their clinical, laboratory phenotype and treatment patterns across the lifespan.
Objectives
The objective of our study was to investigate the age-dependent clinical and laboratory phenotype of hemophilia carriers across the lifespan. We also aimed to study the longitudinal treatment trends in HCs through analysis of the American Thrombosis and Hemostasis Network dataset.
Methods
We investigated the age-dependent variation of bleeding phenotype, coagulation factor levels, and trends of utilization of factor concentrates and nonfactor hemostatic therapies (antifibrinolytics, 1-desamino-8-D-arginine vasopressin) in HCs using the American Thrombosis and Hemostasis Network dataset from 2010 to 2020. The study included 3663 HCs (2728 hemophilia A, 935 hemophilia B) divided into 3 age groups: 0 to 12, 13 to 49, and >50 years.
Results
Joint bleeding was prevalent across all ages of HCs. While bleeding events were more frequent among HCs within the reproductive-age group, hemophilia B carriers received factor therapy more frequently than hemophilia A carriers (P = .03). Factor (F)VIII activity levels in hemophilia A carriers increased significantly with age (P < .001). Analysis of treatment trends from 2010 to 2020 showed increased utilization of factor concentrates and nonfactor hemostatic therapies among HCs (P < .001 and P = .02, respectively).
Conclusion
Our study confirms prior reports of increased bleeding in reproductive-age HCs and identifies lifelong joint bleeding risk. While the rising trend in hemostatic therapy use suggests growing awareness of bleeding tendencies, age-dependent increase in FVIII levels in hemophilia A carriers emphasizes the need for developing tailored management strategies across the lifespan for this population.
Keywords: hemarthrosis, hemophilia, hemostatic, female, lifespan
Essentials
-
•
HCs experience various types of bleeding.
-
•
This study focuses on how the bleeding tendencies and factor levels of carriers change with age.
-
•
Carriers bleed the most during reproductive age, and factor levels seem to increase with age.
-
•
We found increased utilization of factor concentrates and hemostatic therapies in carriers.
1. Introduction
Hemophilia, due to its X-linked pattern of inheritance, has historically been viewed as a male disorder where the bleeding phenotype of female carriers has been overlooked. This bias has led to under recognition of bleeding tendencies in hemophilia carriers (HCs) and limited research focusing on them, particularly across their lifespan. Recent studies have demonstrated that HCs experience increased bleeding tendencies, affecting their quality of life [[1], [2], [3]]. A few studies have highlighted that factor levels do not reliably predict bleeding severity in HCs [4,5]. In response to the growing awareness about bleeding risk and variable bleeding phenotype in HCs, the International Society of Thrombosis and Hemostasis (ISTH) proposed a new classification system in 2021 where HCs are classified based on their bleeding symptoms and baseline plasma coagulation factor levels (normal levels are 0.50-1.50 IU/mL) [6]. The resulting nomenclature distinguishes 5 clinically relevant HC categories: mild, moderate, or severe hemophilia (factor [F]VIII/FIX < 0.40 IU/mL, 0.01-0.05 IU/ml, and < 0.01 IU/mL, respectively), symptomatic and asymptomatic HC (FVIII/FIX ≥ 0.40 IU/mL with and without a bleeding phenotype, respectively). It is noteworthy that the National Bleeding Disorders Foundation uses a cutoff of <50% or 0.50 IU/mL for mild hemophilia, which is followed by some Hemophilia Treatment Centers (HTCs).
Despite this progress with the new ISTH classification, there remains an unmet need to identify HCs and characterize the bleeding and factor level variation of HCs across their lifespan. Furthermore, data on utilization trends of factor concentrates and hemostatic therapies in HCs are limited. Our study aimed to address this knowledge gap by analyzing the national American Thrombosis and Hemostasis Network (ATHN) dataset. Our objective was to examine the age-dependent variation of bleeding and coagulation factor levels and to assess the longitudinal trend of hemostatic therapy utilization in HCs in the United States from 2010 to 2020. The goal is to provide a comprehensive overview of HC phenotype across the lifespan. Our project also aimed to identify data gaps and enrich the ATHNdataset to improve the quality of care of HCs.
2. Methods
The ATHNdataset is a deidentified, secure, national database compliant with the Health Insurance Portability and Accountability Act that includes data of patients with bleeding and clotting disorders from ATHN-affiliated HTCs. Patients at ATHN-affiliated HTCs are enrolled in the ATHNdataset after a process of informed consent. Data is entered by an HTC staff member through review of the electronic medical records. This study was funded by the Dataset Research Engagement and ATHN Mentorship award, a collaborative initiative between the Hemostasis and Thrombosis Research Society and the ATHN, which supports research using the ATHNdataset to advance care for patients with bleeding and clotting disorders [7]. We conducted a retrospective multicenter cohort study and analyzed the data of HCs from 3 HTCs in the United States: University of Iowa, University of Michigan, and Michigan State University. Institutional review board approval was obtained at each participating HTC. The study was conducted from April 2021 through December 2022.
We performed an internal audit comparing patient data from electronic medical records with corresponding ATHNdataset entries for HCs across all 3 participating HTCs. This audit aimed to identify discrepancies between electronic medical records and ATHNdataset entries and enrich the ATHNdataset by entering missing variables, the latter being one of the goals of the Dataset Research Engagement and ATHN Mentorship award. Data from the audit was not included in the analysis.
HCs were stratified into 3 age groups: 0 to 12, 13 to 49, and >50 years. The rationale for choosing the 3 age cohorts was to include carriers across the entire lifespan and not only the reproductive-age group. The cutoff points were arbitrary, based on the average ages of menarche and menopause, which are around 12 years and 50 years, respectively [8,9]. There were no exclusion criteria. The following cutoffs were used for mild, moderate, and severe hemophilia: FVIII/FIX < 0.40 IU/mL, 0.01-0.05 IU/mL, and < 0.01 IU/mL, respectively. We collected demographic data including race, ethnicity, and use of a medical alert device in HCs. Additionally, clinical and laboratory data was collected using ATHNdataset’s core data elements: primary diagnosis, bleed data (frequency of bleed events, anatomical location of bleed events, factors associated with bleeding), infusion data (factor and nonfactor treatment used for bleed events), and laboratory tests (baseline FVIII levels in hemophilia A carriers [HA-Cs], FIX levels in hemophilia B carriers [HB-Cs], and repeat factor levels when available). Statistical analyses included descriptive statistics for categorical variables, 3-sample proportion tests to test the equality of proportions across 3 age groups, a pairwise 2-proportion Z-test to compare 2 observed proportions among 3 age groups, and logistic regressions to analyze the association between age and (i) bleed events, (ii) FVIII/FIX treatment, (iii) nonfactor hemostatic therapies (1-desamino-8-d-arginine vasopressin [DDAVP], aminocaproic acid, and tranexamic acid).
We compared the cross-sectional differences in factor activity levels over time from baseline, which was defined as initial reported FVIII/FIX activity for each participant. Time intervals from baseline were categorized as “within one year,” “one to three years,” and “over three years.” These cutoffs were selected arbitrarily based on distribution of data available to perform temporal analyses. The Wilcoxon test was performed between the baseline mean value and the mean values at each time frame, with statistical significance set at P < .05. We further used a linear mixed model to test the association between time and factor activity, which was repeatedly measured over time in all carriers. In this model, the fixed effects were time differences and carrier status, while the random effect was unique participant ID to adjust for repeated measurements. The time differences were normalized to Z-score using the mean and standard deviation.
We examined the longitudinal treatment trends in HCs. Treatment data was extracted from the ATHNdataset’s core elements: bleeds and infusion data, surgeries and procedures, and medications. Factor therapy was defined as any treatment involving FVIII or FIX products. Nonfactor hemostatic therapy was defined as any treatment using nonfactor products, such as DDAVP, aminocaproic acid, and tranexamic acid. The significance of changes in the proportions of factor treatment or hemostatic therapies from 2010 to 2020 was tested using a linear regression model, where the outcomes were the changes in the proportions of factor treatment or hemostatic therapies, and the predictor was the year of treatment.
Missing values were not imputed and were treated as “unknowns” in the descriptive analysis. These values were excluded from the linear mixed model and regression analysis. The robustness of the linear regression and linear mixed models were checked by residual analysis.
3. Results
As of October 13, 2023, a total of 3663 HCs were identified from the ATHNdataset, including 2728 HA-Cs and 935 HB-Cs (Table). Details about carrier diagnosis, whether it was based on family history, factor levels, and/or genetic testing, were not available. Majority of carriers (>80%) belonged to White race and non–Hispanic ethnicity. A small percentage of carriers (∼9% of HA-Cs and ∼12% of HB-Cs) reported using a medical alert device. The rate of medical device usage in HA-Cs was 23 (7.3%) for the 0 to 12 years age group, 156 (8.5%) for the 13 to 49 years group, and 60 (10%) for those aged >50 years. No statistically significant difference in usage rates across age groups was observed (chi-squared test, P = .21).
Table.
Summary of characteristics of hemophilia A carriers and hemophilia B carriers
| Total number of carriers N = 3663 |
Hemophilia A carriers (HA-C) N = 2728 |
Hemophilia B carriers (HB-C) N = 935 |
|---|---|---|
| Demographic characteristics | ||
| Race | ||
| Black | 6.5% (180/2728) | 5% (45/935) |
| White | 81.5% (2226/2728) | 90% (842/935) |
| Other | 6.3% (172/2728) | 2.2% (21/935) |
| Unknown | 5.5% (150/2728) | 2.8% (27/935) |
| Ethnicity | ||
| Hispanic | 18% (491/2728) | 10.4% (98/935) |
| Not Hispanic | 79.3% (2165/2728) | 87.1% (815/935) |
| Unknown | 2.6% (72/2728) | 2.3% (22/935) |
| Medical alert device use | 8.7% (239/2728) | 11.6% (109/935) |
| Clinical characteristics | ||
| Frequency of bleeding | 82% (2439/2958) | 18% (519/2958) |
| Number of carriers with bleed events | 363 | 139 |
| Factor treatment for bleeds | 22.7% (556/2439) | 33% (172/519) |
| Nonfactor treatment for bleeds | 3.7% (91/2439) | 2.8% (15/519) |
| Laboratory characteristics | ||
| Baseline factor activity level | ||
| Severe (<1%) | 2.1% (40/1868) | 0.8% (6/673) |
| Moderate (1%-5%) | 1.5% (29/1868) | 1.4% (10/673) |
| Mild (5%-40%) | 38.7% (724/1868) | 49.1% (331/673) |
| Normal (>40%) | 57.1% (1067/1868) | 48.2% (325/673) |
| Unknown | 0.4% (8/1868) | 0.1% (1/673) |
HA-C, hemophilia A carrier; HB-C, hemophilia B carrier.
Similarly, in HB-Cs, the usage rates were 17 (12%) for the 0 to 12 years group, 75 (12%) for the 13 to 49 years group, and 17 (9.0%) for those >50 years. As with HA-Cs, no significant association between age group and device usage was found (chi-squared test, P = .99).We did not have data to correlate medical alert device use and bleeding symptoms.
A total of 2958 bleed events have been documented in those HCs. Below, we summarize the clinical characteristics and laboratory characteristics of HCs, followed by the trends of utilization of factor and hemostatic therapies in HCs.
3.1. Clinical characteristics of bleeding diatheses in HA-Cs
3.1.1. Frequency of bleed events
A total of approximately 82% (2439/2958) of reported bleed events were in HA-Cs. Bleed events were most frequent among the 13 to 49 years age group of HA-Cs (68%, 1663/2439) and least frequent among the 0 to 12 years group (4.3%, 105/2439). The >50 year group had 28% (671/2439) bleed events. Among 363 HA-Cs with bleed events, there were 61% (222/363) in the 13 to 49 years group, 27% (97/363) in the >50 years group, and 12% (44/363) in the 0 to 12 years group. The P value from 3-sample proportion test showed significant differences between the age groups, with the 13 to 49 years group having the highest proportion of bleed events (P < .001) followed by the >50 years group. The frequency of bleed events in each age group is represented in Figure 1.
Figure 1.
Frequency of bleed events among different age groups of hemophilia A carriers (HA-C).
3.1.2. Anatomical locations of bleed events
Joint bleeding was most prevalent across all ages with a notable increase in prevalence in older carriers: 20% (21/105) in 0 to 12 years, 29% (478/1663) in 13 to 49 years, and 38% (254/671) in the >50 years group of HA-Cs. Oral and nasal bleeding was most prevalent in the 0 to 12 years group (25%, 26/105). Genitourinary bleeding, which likely included reproductive tract bleeding (RTB), was the second most common location in the 13 to 49 years age group. Details of the various anatomical locations of bleed events in each age group of HA-Cs are shown in Figure 2.
Figure 2.
Anatomical locations of bleeding diatheses among hemophilia A carriers (HA-C).
3.1.3. Factors associated with bleed events
Trauma was the most common factor associated with bleeding in the 0 to 12 years group (55%, 58/105). The most common factor in the other 2 age groups was “unknown.”
3.1.4. Treatment with factor and nonfactor hemostatic therapies for bleed events
The frequency of factor concentrate treatment for bleeding ranged between 22% and 25%, and nonfactor hemostatic therapy usage was between 2% and 4%. Logistic regression showed no significant difference in treatment frequency across the age groups of HA-C.
3.2. Clinical characteristics of bleeding diatheses in HB-Cs
3.2.1. Frequency of bleed events
A total of approximately 18% (519/2958) of bleed events occurred in HB-Cs. Bleed events were most frequent in the 13 to 49 years age group (59%, 307/519) and least frequent in the 0-12 years group (7.5%, 39/519). The >50 years group accounted for 33% (173/519) of bleed events. Among 139 HB-C with bleed events, there were 58% (81/139) in the 13 to 49 years group, 27% (37/139) in the >50 years group, and 15% (21/139) in the 0 to 12 years group. The P value from 3-sample proportion tests indicated significant differences (P < .001), with the 13 to 49 years group having the highest proportion of bleed events. The frequency of bleed events in each age group is shown in Figure 3.
Figure 3.
Frequency of bleed events among different age groups of hemophilia B carriers (HB-C).
3.2.2. Anatomical locations of bleed events
Joint bleeding was the most prevalent location of bleeding in the >50 year group (40%, 70/173). Mucosal bleeding from the oral cavity and nose were most prevalent in the 0 to 12 years group (51%, 20/39) and the 13 to 49 years group (20%, 62/307). Details of the various anatomical locations of bleed events in each age group of HB-Cs are shown in Figure 4.
Figure 4.
Anatomical locations of bleeding diatheses among hemophilia B carriers (HB-C).
3.2.3. Factors associated with bleed events
Trauma was the most associated factor with bleeding in all 3 age groups.
3.2.4. Treatment with factor and nonfactor hemostatic therapies for bleed events
The frequency of factor concentrate treatment for bleeding was 38% (118/307) in the 13 to 49 years group, and approximately 21% to 27% in the other 2 groups. The frequency of nonfactor hemostatic therapy use ranged between 2% and 5% in HB-Cs. Logistic regression showed a significant association between factor treatment and age, with HB-Cs in the 13 to 49 years group receiving the highest number of factor infusions (P = .03; odds ratio 2.42; 95% CI, 1.13-5.81).
3.3. Laboratory characteristics of HCs
3.3.1. Baseline coagulation factor levels in HA-Cs
There were 1868 HA-Cs with reported baseline FVIII levels. The median age of the first FVIII activity level test was 22 years (minimum age 0 and maximum age 89 years) for HA-Cs. Fifty-six percent (1067/1868) had FVIII levels within the normal range (>40%). It was unclear if the factor assays were single stage or chromogenic. The distribution of baseline FVIII levels in each age group of HA-Cs is demonstrated in Figure 5.
Figure 5.
Age-related distribution of factor VIII levels in hemophilia A carriers (HA-C).
3.3.2. Baseline coagulation factor levels in HB-Cs
There were 673 HB-Cs with reported baseline FIX levels. The median age of the first FIX level test was 17 years (minimum age 0 and maximum age 82 years) for HB-Cs. Majority in the 0 to 12 years group (58%, 64/110) and >50 years group (52%, 66/126) had FIX levels consistent with mild hemophilia (5%-40%), while majority in the 13 to 49 years group (51%, 222/437) had FIX levels in the normal levels (>40%). As with HA-C, data on the type of assays (single stage/chromogenic) was not available. The distribution of baseline FIX levels in each age group is shown in Figure 6.
Figure 6.
Age-related distribution of factor IX levels in hemophilia B carriers (HB-C).
3.3.3. Change in factor activity levels in HCs with time
A significant increase in the mean factor activity level was observed over time in all carriers, with the mean levels increasing in the 1 to 3 years and >3-year time frames (P < .001; estimate 5.59; SE 1.56), as shown in Figure 7. Further analysis among HA-Cs showed a significant association between time and FVIII activity in HA-C using a linear mixed model (P < .001; estimate 6.06; SE 1.74). Analysis among HB-Cs did not show a significant association between time and FIX activity levels (P = .47; estimate 2.49; SE 3.42).
Figure 7.
Change in factor VIII/IX activity levels with time in all carriers.
3.4. Trends of utilization of factor concentrates and nonfactor hemostatic therapies
A linear regression model showed a significant increase in utilization of both factor concentrates (P < .001; SE 0.51; estimate 3.55) and nonfactor hemostatic therapies (P = .02; SE 0.27; estimate 0.77) from 2010 to 2020. This increasing trend of utilization of factor and nonfactor hemostatic therapies among HCs is shown in Figure 8.
Figure 8.
Trends of utilization of factor VIII/IX concentrates and nonfactor hemostatic therapies (antifibrinolytics and 1-desamino-8-d-arginine vasopressin) in hemophilia carriers.
4. Discussion
HCs have been underrepresented in prior epidemiologic studies focusing on women with inherited bleeding disorders [10,11]. This study addressed this research gap by characterizing the phenotype of HCs across the entire lifespan.
4.1. Demographics and usage of medical alert devices
Our cohort was predominantly White and non–Hispanic, which was not unexpected. Interestingly, most carriers did not report using a medical alert device, reflecting the historical misconception that carriers do not experience serious bleeding. This finding underscores the importance of promoting the use of medical alert bracelets among HCs to ensure appropriate emergency care.
4.2. Clinical characteristics
Bleeding was most prevalent among HCs in the 13 to 49 years age group, likely due to hemostatic challenges associated with reproductive life. This finding is consistent with previous reports [4]. Our results demonstrate that joint bleeding was common across all age groups, with an increased incidence in carriers aged >50 years. It was not clear from our data whether clinician documentation of joint bleeding was based on clinical exam and/or imaging. It has previously been reported that HCs experience joint related comorbidities [12,13] and have imaging abnormalities suggestive of subclinical hemarthrosis [14]. Our findings suggest that older age and postmenopausal status may contribute to poor joint health and parallels the age-related increase in loss of motion in males with mild to moderate hemophilia not on prophylaxis [15]. It is possible that prior microbleeds could contribute to joint disease along with age-related comorbidities. Unlike prior studies [13], these findings are significant as this study characterizes joint bleeding by age in a large cohort of carriers. Although the risk of joint pathology was shown to correlate with factor activity levels in a previous registry study on Swedish HCs [12], our analysis was not able to confirm this correlation. We did not have data to correlate joint bleeding with the factor levels or severity of hemophilia, which is a limitation.
HCs from our cohort were noted to experience a wide spectrum of bleeding diatheses including mucosal, joint, muscle and soft tissue bleeding as has been previously described [2,16,17]. The frequency and severity of bleed events could not be correlated with factor levels, as data was not available. The ATHNdataset listed the following anatomical locations at the time of this study: genitourinary, head (extracranial), head (intracranial), joint, oral/nasal, other, spine, tissue-muscle, tissue-soft, tissue-unknown, and unknown. It was unclear how many patients with genitourinary tract bleeding experienced only RTB, as the latter was not specifically reported, highlighting the limitation of the current dataset. To enable clear documentation of bleed events, we recommend revising the anatomical locations in the ATHNdataset to include more specific sites. Some of the locations listed, including “other” and “unknown,” are not clear or specific. It is unclear if the term, “tissue-muscle” refers to the muscle or soft tissue. It would also be helpful to separate oral and nasal bleeding. We also recommend including sites within the female reproductive tract to enable better reporting and recognition of bleeding in women and girls in future ATHN studies. These revisions of data fields to capture bleeding sites will enhance phenotype granularity and inform clinical management.
Trauma was the primary bleeding trigger across all HB-C age groups and in HA-Cs aged 0 to 12 years, while triggers for other groups were less clear. About 1% to 2% of HCs in each age group had bleeding with surgery. The small numbers are likely due to underreporting, possibly due to challenges with classifying the factors associated with bleeding in the dataset. The ATHNdataset’s current categorization included the following factors associated with bleeding: dental procedure, pregnancy, spontaneous, surgery, trauma/injury, others, and unknown, but omits menstruation, which is a crucial trigger in females. RTB, including heavy menstrual, antepartum, and postpartum bleeding, affects approximately 10% to 57% of HCs [18]. However, our study did not adequately capture these symptoms, likely due to lack of a clear definition for RTB and challenges in entering RTB data in the ATHN Clinical Manager, a limitation reported previously [11].
To address this challenge, we recommend revising the dataset to broadly categorize bleeding causes as spontaneous or induced (trauma/surgery), with a detailed drop-down list of specific factors (such as type of dental procedure, type of surgery, type of trauma, menstruation, pregnancy, labor and delivery, postpartum hemorrhage) to improve the granularity of bleeding event characterization. An ongoing initiative supported by ATHN and Hemostasis and Thrombosis Research Society to incorporate a “women and girls specific form” is expected to improve tracking of these bleeding symptoms, potentially leading to enhanced surveillance of women and girls with bleeding disorders.
Our analyses revealed that a significantly higher number of HB-Cs than HA-Cs were treated with factor concentrates for their bleed events. This difference could potentially be explained by the lack of nonfactor therapy such as DDAVP to treat FIX deficiency, leaving clinicians with only the choice of using FIX concentrates and antifibrinolytic therapy.
4.3. Characteristics of FVIII/FIX levels at baseline and during follow-up
Majority of HA-Cs from our cohort had normal factor activity levels in the 40% to 60% range, while most HB-Cs were in mild hemophilia range with factor activity levels ranging from 5% to 25%. Lower factor levels in HB-C could be due to inclusion of Amish cohorts in the Midwest, who have a high prevalence of moderate hemophilia B. Our results are consistent with previous studies showing that factor levels in HCs are ∼50% [19]. Our study is the first, to our knowledge, to investigate changes in factor levels over time in HCs and revealed a significant increase in factor activity levels beyond 3 years. Further subgroup analyses revealed a significant increase in FVIII levels in HA-Cs (P < .001) alone (and not in HB-Cs), which could potentially be related to age-associated increases in von Willebrand factor (VWF) and FVIII levels, which have previously been reported in patients with von Willebrand disease [20]. Our study lacked data on VWF levels in HCs; consequently, we could not clarify the association of VWF levels in HA-Cs over time. Nevertheless, previous research has established an association between FVIII and VWF levels in HA-Cs [21]. Furthermore, the clinical circumstances when the repeat factor levels were drawn were not known; as a result, we could not explore the impact of hemostatic stress. Perhaps longitudinal studies with larger cohorts exploring the hemostatic profile of HCs across their lifespan along with clinical circumstances for drawing FVIII/FIX levels would provide a deeper understanding of these associations.
4.4. Treatment trends
Emerging epidemiologic and clinical studies over the last 2 decades have led to increased awareness about the bleeding risk in HCs, especially with surgical intervention. In 2021, the Medical and Scientific Advisory Council (MASAC) of the National Bleeding Disorders Foundation recommended factor replacement prophylaxis prior to delivery for HCs if their factor levels are <50% [22]. Since our data was from prior to 2021, we could not evaluate the impact of MASAC recommendations on treatment trends in HCs. Our analysis reveals an increasing trend in the utilization of factor concentrates and other nonfactor hemostatic therapies, including DDAVP and antifibrinolytics among carriers. This upward trend likely reflects a growing awareness of bleeding risks in HCs and improved accessibility to hemophilia therapies, rather than a response to the 2021 MASAC recommendation. Perhaps our findings could be used as valuable baseline data for future studies evaluating the impact of these updated MASAC recommendations on treatment patterns and bleed events in HCs.
4.5. Strengths
Our study’s key strength lies in its analysis of a large, nationally representative cohort of HCs across all ages, addressing previously unanswered questions about bleeding phenotype and treatment trends across the lifespan.
4.6. Limitations
Our study has several limitations inherent to registry-driven observational studies. First, we used arbitrary cutoff points for menarche and menopause to classify the time points for factor assessments based on clinical relevance. Second, we faced a limitation of insufficient/missing data, potentially due to underreporting of bleeding events, longitudinal factor levels, and the ATHN requirement of basic data entry. Third, we did not have data to correlate bleeding symptoms with the factor levels as our study was aimed at providing a cross-sectional review rather than longitudinal analyses of patient data. Therefore, it was not clear from our data if the diagnosis of joint bleeding was based on clinical exam or imaging, so there is a possibility that our numbers are overstated. Our study was not robust enough to better characterize the group of symptomatic carriers with normal FVIII levels. Hopefully, combining multiple countrywide registries will lead to further recognition and better management of that subgroup. Finally, it is possible that patients included in the dataset were those who experienced bleeding events, which could have contributed to selection bias.
5. Conclusions
Our study is unique in its focus on investigating the phenotype of HCs through their lifespan. It reinforces prior observations that HCs bleed the most during their reproductive years. Joint bleeding was prevalent across all ages in our cohort with increasing incidence in older carriers, which highlights the need to closely monitor the musculoskeletal health of HCs. Our results suggest that bleeding was more frequent among HA-Cs than HB-Cs, although most carriers had factor levels close to the normal range in both groups. We found a low prevalence of medical alert device use among carriers, which highlights the need for increasing awareness that hemophilia carriership is a bleeding disorder. We observed that FVIII activity levels increased with age in carriers, which is interesting and novel and necessitates further study in determining if an age-related increase ameliorates the bleeding phenotype while decreasing the surgical risk of bleeding. The increase in the longitudinal trend of utilization of factor concentrates and nonfactor hemostatic therapies in HCs reflects increased awareness among clinicians about the bleeding risks of HCs. Overall, our findings underscore the importance of lifelong follow-up of HCs, ideally in HTCs, to facilitate timely management of bleeding with hemostatic agents and multidisciplinary care. Further research is essential to develop age-specific clinical guidelines for the management of HCs, addressing the needs of this long-neglected patient population across their life span.
Acknowledgments
We acknowledge the American Thrombosis and Hemostasis Network biostatistics team for help with data analyses. We acknowledge the work of research assistants from all three participating sites for their help with data entry.
Funding
N.S., MD, received the 2020 DREAM (Dataset Research Engagement and American Thrombosis and Hemostasis Network Mentorship) award from American Thrombosis and Hemostasis Network and HTRS (Hemostasis and Thrombosis Research Society), which was funded by Takeda.
Author contributions
N.S., P.K., and A.S. conceptualized and designed the study. N.S. and A.C. collected the data. J.H. performed the statistical analysis. R.K., S.P., P.K., and A.S. provided supervision and expert input through the course of the study. N.S. wrote the manuscript. All authors reviewed, edited and approved the final version of the manuscript prior to submission.
Relationship Disclosure
Authors N.S., A.C., J.H., P.K., and R.K. do not have any disclosures. S.P. has received consultancy fees from for Apcintex, ASC Therapeutics, Bayer, BioMarin, CSL Behring, HEMA Biologics, Freeline, Metagenomi, Novo Nordisk, Pfizer, Poseida Therapeutics, Regeneron/Intellia, Roche/Genentech, Sanofi, Takeda, Spark Therapeutics and uniQure, and research funding from Siemens and holds a membership on a Scientific advisory committee for GeneVentiv and Equilibra Bioscience. A.S. is site principal investigator and co-principal investigator for industry driven hemophilia studies (Pfizer, Amgen, NovoNordisk), and received honorarium for serving on advisory boards of CSL Behring, Pfizer, and Genentec.
Footnotes
Handling Editor: Dr Johnny Mahlangu
References
- 1.Mauser Bunschoten E.P., van Houwelingen J.C., Sjamsoedin Visser E.J., van Dijken P.J., Kok A.J., Sixma J.J. Bleeding symptoms in carriers of hemophilia A and B. Thromb Haemost. 1988;59:349–352. [PubMed] [Google Scholar]
- 2.Paroskie A., Gailani D., DeBaun M.R., Sidonio R.F., Jr. A cross-sectional study of bleeding phenotype in haemophilia A carriers. Br J Haematol. 2015;170:223–228. doi: 10.1111/bjh.13423. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Gilbert L., Paroskie A., Gailani D., Debaun M.R., Sidonio R.F. Haemophilia A carriers experience reduced health-related quality of life. Haemophilia. 2015;21:761–765. doi: 10.1111/hae.12690. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Plug I., Mauser-Bunschoten E.P., Bröcker-Vriends A.H., van Amstel H.K., van der Bom J.G., van Diemen-Homan J.E., et al. Bleeding in carriers of hemophilia. Blood. 2006;108:52–56. doi: 10.1182/blood-2005-09-3879. [DOI] [PubMed] [Google Scholar]
- 5.Olsson A., Hellgren M., Berntorp E., Ljung R., Baghaei F. Clotting factor level is not a good predictor of bleeding in carriers of haemophilia A and B. Blood Coagul Fibrinolysis. 2014;25:471–475. doi: 10.1097/MBC.0000000000000083. [DOI] [PubMed] [Google Scholar]
- 6.van Galen K.P.M., d’Oiron R., James P., Abdul-Kadir R., Kouides P.A., Kulkarni R., et al. A new hemophilia carrier nomenclature to define hemophilia in women and girls: communication from the SSC of the ISTH. J Thromb Haemost. 2021;19:1883–1887. doi: 10.1111/jth.15397. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.American Thrombosis and Hemostasis Network ATHNdataset. https://athn.org/what-we-do/national-projects/athndataset.html
- 8.Wang Z., Asokan G., Onnela J.P., Baird D.D., Jukic A.M.Z., Wilcox A.J., et al. Menarche and time to cycle regularity among individuals born between 1950 and 2005 in the US. JAMA Netw Open. 2024;7 doi: 10.1001/jamanetworkopen.2024.12854. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Appiah D., Nwabuo C.C., Ebong I.A., Wellons M.F., Winters S.J. Trends in age at natural menopause and reproductive life span among US women, 1959-2018. JAMA. 2021;325:1328–1330. doi: 10.1001/jama.2021.0278. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Byams V.R., Kouides P.A., Kulkarni R., Baker J.R., Brown D.L., Gill J.C., et al. Surveillance of female patients with inherited bleeding disorders in United States Haemophilia Treatment Centres. Haemophilia. 2011;17(Suppl 1):6–13. doi: 10.1111/j.1365-2516.2011.02558.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Haley K.M., Sidonio R.F., Jr., Abraham S., Cheng D., Recht M., Kulkarni R. A cross-sectional study of women and girls with congenital bleeding disorders: the American Thrombosis and Hemostasis Network cohort. J Womens Health (Larchmt) 2020;29:670–676. doi: 10.1089/jwh.2019.7930. [DOI] [PubMed] [Google Scholar]
- 12.Osooli M., Donfield S.M., Carlsson K.S., Baghaei F., Holmström M., Berntorp E., et al. Joint comorbidities among Swedish carriers of haemophilia: a register-based cohort study over 22 years. Haemophilia. 2019;25:845–850. doi: 10.1111/hae.13831. [DOI] [PubMed] [Google Scholar]
- 13.Sidonio R.F., Mili F.D., Li T., Miller C.H., Hooper W.C., DeBaun M.R., et al. Females with FVIII and FIX deficiency have reduced joint range of motion. Am J Hematol. 2014;89:831–836. doi: 10.1002/ajh.23754. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Gilbert L., Rollins L., Hilmes M., Luo Y., Gailani D., Debaun M.R., et al. Haemophilia A carriers demonstrate pathological and radiological evidence of structural joint changes. Haemophilia. 2014;20:e426–e429. doi: 10.1111/hae.12535. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wang M., Recht M., Iyer N.N., Cooper D.L., Soucie J.M. Hemophilia without prophylaxis: Assessment of joint range of motion and factor activity. Res Pract Thromb Haemost. 2020;4:1035–1045. doi: 10.1002/rth2.12347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Sharathkumar A., Hardesty B., Greist A., Salter J., Kerlin B., Heiman M., et al. Variability in bleeding phenotype in Amish carriers of haemophilia B with the 31008 C→T mutation. Haemophilia. 2009;15:91–100. doi: 10.1111/j.1365-2516.2008.01792.x. [DOI] [PubMed] [Google Scholar]
- 17.Staber J., Croteau S.E., Davis J., Grabowski E.F., Kouides P., Sidonio R.F., Jr. The spectrum of bleeding in women and girls with haemophilia B. Haemophilia. 2018;24:180–185. doi: 10.1111/hae.13376. [DOI] [PubMed] [Google Scholar]
- 18.James A.H. Women and bleeding disorders. Haemophilia. 2010;16(Suppl 5):160–167. doi: 10.1111/j.1365-2516.2010.02315.x. [DOI] [PubMed] [Google Scholar]
- 19.Rizza C.R., Rhymes I.L., Austen D.E., Kernoff P.B., Aroni S.A. Detection of carriers of haemophilia: a ‘blind’ study. Br J Haematol. 1975;30:447–456. doi: 10.1111/j.1365-2141.1975.tb01859.x. [DOI] [PubMed] [Google Scholar]
- 20.Biguzzi E., Siboni S.M., le Cessie S., Baronciani L., Rosendaal F.R., van Hylckama Vlieg A., et al. Increasing levels of von Willebrand factor and factor VIII with age in patients affected by von Willebrand disease. J Thromb Haemost. 2021;19:96–106. doi: 10.1111/jth.15116. [DOI] [PubMed] [Google Scholar]
- 21.Ay C., Thom K., Abu-Hamdeh F., Horvath B., Quehenberger P., Male C., et al. Determinants of factor VIII plasma levels in carriers of haemophilia A and in control women. Haemophilia. 2010;16:111–117. doi: 10.1111/j.1365-2516.2009.02108.x. [DOI] [PubMed] [Google Scholar]
- 22.Medical and Scientific Advisory Council of the National Hemophilia Foundation . National Bleeding Disorders Foundation; 2021. MASAC Guidelines for Pregnancy and Perinatal Management of Women with Inherited Bleeding Disorders and Carriers of Hemophilia A or B. MASAC Document #265. [Google Scholar]