Primary Care Screening for Von Willebrand Disease in African American Adolescents with Heavy Menstrual Bleeding

Introduction: Von Willebrand Disease (VWD) is the most common inherited bleeding disorder; however, establishing a laboratory diagnosis can be challenging. While the VWF Ag test is highly reproducible, VWF:RCo activity is prone to both intra- and inter-laboratory variation. A landmark study by Flood et al.demonstrated a common polymorphism in African American controls, D1472H, affecting approximately 50% of healthy controls as well as subjects previously labeled as type 1 VWD. This sequence variation results in decreased VWF:RCo levels and, ultimately, lower RCo:Ag ratio in healthy controls without an increase in their bleeding score. A newer GPIbM assay assesses VWF activity which does not depend on ristocetin binding. This assay is hence unaffected by D1472H polymorphism and provides more accurate assessment of VWF activity. However, GPIbM assay is not widely available, and most labs still rely on VWF:RCo assays to measure activity.

In this analysis of an ongoing study examining the population prevalence of VWD in adolescents with heavy menstrual bleeding (HMB), we aim to compare VWF activity measured by RCo and GPIbM assay in African American participants with HMB and low RCo activity. Our hypothesis is that many of these subjects may have lower VWF:RCo activity (presumably due to D1472H polymorphism) but normal GPIbM activity resulting in false diagnosis of type 1 VWD.

Methods: In this large, ongoing population study, 9 to 21-year-old menstruating adolescents presenting for routine care were recruited from five primary care clinics across Nationwide Children’s Hospital (NCH) (Columbus, OH, USA). Patients with a prior diagnosis of a bleeding disorder were excluded.

Participants completed the ISTH Bleeding Assessment Tool (BAT). Those with a score of ≥1 on the menses question of the ISTH BAT were considered to have HMB and underwent testing for VWD. Whole blood samples were collected and VWF Ag and RCo tests were performed locally by the NCH clinical lab. Additionally, samples were sent to Versiti research lab (Milwaukee, WI) for VWF Ag and GPIbM activity testing. VWF Ag and activity results obtained from both sites were then compared. Participants with VWF Ag or GPIbM levels of ≤50% were considered to have VWD per our study protocol. Data were summarized with standard descriptive statistics and nonparametric analyses were used to compare labs and clinical characteristics in patients with low VWF:RCo who did and did not have VWD.

Results: To date, 695 adolescents agreed to participate in this study and, among these, 433 self-identified as African American (62%). In the African American group, 204 subjects (47%) met the study definition of HMB and were considered eligible for further VWD testing. While some were lost to follow-up, 160 adolescents agreed to and underwent VWD testing. Of these, 32 adolescents (20%) had VWF:RCo ≤50%. Upon further testing at Versiti, 9 (28%) had low GPIbM activity as well and met the criteria for VWD. The remaining 23 participants (72%) had normal GPIbM activity, and hence, no VWD.

On comparing the two groups (9 with VWD and 23 without VWD), age at menarche and years since menarche were similar in both groups with a median of 16 years and 5 years, respectively. Bleeding scores (ISTH BAT scores) were also similar in both groups. Median VWF Ag at our local NCH lab was significantly higher at 83% (IQR: 71-99%) for those who did not have VWD compared to 54% (IQR: 52-57%) for those who truly had VWD (p <0.001). This observation was also seen with the Versiti labs where median VWF Ag was higher at 80% (IQR: 64-103%) for those without VWD compared to 47% (44-49%) for those with VWD (p <0.001). All the participants without VWD (n=23, 100%) had VWF Ag above 60% on original testing at our local NCH lab.

Conclusion: In this small primary care population, we found that VWD is not uncommon in African American adolescents with HMB, occurring in 1 out of 20 participants who underwent testing. However, our data again demonstrate that African American individuals may have lower RCo activity, likely due to the D1472H polymorphism which could result in false labeling as VWD. Hematology consultation and further testing with GPIbM assay should be pursued for accurate diagnosis of VWD. VWF Ag may give some clues and in individuals with Ag 60% or higher but low RCo, a falsely low RCo should be considered, and confirmatory testing should be pursued.