Clinical phenotype and pathophysiological mechanisms underlying qualitative low VWF Open Access

Patients with mild-to-moderate reductions in plasma von Willebrand factor (VWF) antigen (VWF:Ag) and/or activity (VWF:Act) continue to pose significant clinical challenges.^1,2 The current evidence suggests that most individuals with plasma VWF levels in the 30 to 50 IU/dL range do not exhibit bleeding complications.^1,3-5 Conversely, however, cohort studies have clearly demonstrated that a subset of individuals with VWF levels in the 30 to 50 IU/dL range do have significantly increased bleeding phenotypes, notably with reference to heavy menstrual bleeding and postpartum hemorrhage.^6-9 These patients with mild-to-moderate reductions in plasma VWF levels and increased bleeding have been classified previously as having low VWF.^10,11 

The American Society of Hematology (ASH)/International Society on Thrombosis and Haemostasis (ISTH)/National Hemophilia Foundation (NHF)/World Federation of Hemophilia (WFH) guidelines published in 2021 were based on a systematic review of the available evidence and addressed several specific questions that were prioritized after an initial international survey.^12,13 Consistent with previous guidelines, the panel recommended that individuals with plasma VWF:Ag levels <30 IU/dL should be diagnosed with type 1 von Willebrand disease (VWD).¹² However, in contrast with previous guidelines,^10,11 the ASH/ISTH/NHF/WFH panel removed the low VWF category and recommended that patients with plasma VWF:Ag and/or VWF:Act in the 30 to 50 IU/dL range and abnormal bleeding should be diagnosed as type 1 or type 2 VWD rather than as low VWF.¹² Notwithstanding this recommendation, the ASH/ISTH/NHF/WFH guidelines panel also noted the low certainty provided by the available published evidence and critically highlighted specific areas in this field that require additional research.

Low VWF has classically been regarded as a partial quantitative reduction in plasma VWF:Ag levels.^3,5,14 Importantly, however, previous reports have shown that some patients with low VWF and significant bleeding were actually diagnosed based on persistent reductions in plasma VWF:Act levels in the 30 to 50 IU/dL range.⁷ This subgroup all had plasma VWF:Ag levels of >50 IU/dL and thus had qualitative, rather than quantitative, low VWF. Although functional VWF defects are well recognized in patients with historic VWF:Act levels of <30 IU/dL as distinct subtypes of type 2 VWD, no previous studies have explored the clinical implications of mild functional VWF defects in patients with qualitative low VWF (low VWF–QL).

To address this clinically important question, we combined data sets from the low VWF in Ireland cohort (LoVIC), the low VWF in Erasmus MC (LVEMC), and the Willebrand in the Netherlands (WiN) studies.^6,7,14 Using these data, we specifically investigated the bleeding phenotype and pathogenic mechanisms in patients with low VWF–QL in comparison with patients with either (1) quantitative low VWF (low VWF–QT; ie, with plasma VWF:Ag levels in the 30 to 50 IU/dL range) or (2) type 2 VWD (type 2A, 2B, or 2M). Importantly, our findings demonstrate that many patients with reduced plasma VWF:Act levels in the 30 to 50 IU/dL range have significant bleeding phenotypes, although their plasma VWF:Ag levels are within the normal range. In addition, we further showed that low VWF–QL is a distinct entity compared to type 2 VWD. Finally, our studies highlight that low VWF–QL is predominantly caused by abnormalities in VWF biosynthesis within endothelial cells that occur largely independent of identifiable pathological VWF sequence variants.

The methodologies of the LoVIC, LVEMC, and WiN studies have been described previously.^6,7,15 Briefly, the LoVIC and WiN studies were national cross-sectional studies that were conducted in Ireland and The Netherlands, respectively.^7,16 The LVEMC study was a single-center cohort study that was performed at the Erasmus University Medical Center in Rotterdam, The Netherlands.⁶ For the LoVIC and LVEMC study, the inclusion criteria required historically lowest plasma VWF:Ag and/or VWF:Act in the 30 to 50 IU/dL range, accompanied by an abnormal ISTH-Bleeding Assessment Tool (BAT) (defined as a score of ≥3 in children, ≥4 in adult males, and ≥6 in adult females). The WIN study required patients to have either a personal bleeding history or a positive family history, along with documented lowest VWF:Ag, VWF:Act, and/or VWF collagen binding (VWF:CB) levels of ≤30 IU/dL. All studies were approved by the respective medical ethics committees, and written informed consent was obtained.^6,7,15 

In all 3 studies, plasma VWF:Ag, VWF:Act, VWF:CB, and factor VIII (FVIII) activity (FVIII:C) were centrally measured.^6,7,15,17,18 Samples were not freeze-thawed more than once, and the VWF:Ag and VWF:Act measurements were performed on the same sample. In the LoVIC study, VWF ristocetin cofactor (VWF:RCo) was measured at diagnosis and at inclusion in the study. In addition, in a random selection of patients from the LoVIC study, VWF:GPIbM and/or VWF:GPIbR were measured at inclusion in the study. The LVEMC study recorded all measured VWF:Act assays ever performed for their patients in a single central specialized hemostasis laboratory. Based on these records, VWF:Act was measured using the VWF:RCo assay in samples before 2005, the VWF antibody assay (HemosIL VWF:Act; Werfen IL) was used from 2005 until 2012, and VWF:GPIbM (INNOVANCE VWF Ac) was used from 2012 onward.⁶ VWF propeptide (VWFpp) was measured in the WiN study and LoVIC study as previously described.^7,16 

In addition, the Desmopressin trials were performed in a subset of patients across the 3 studies, as outlined previously.^6,19-21 Platelet light transmission aggregometry and platelet nucleotide analysis were performed as detailed in the supplemental Materials and Methods, available on the Blood website. The VWF gene was sequenced in the LoVIC and WiN studies but not in the LVEMC study, as previously described.^7,22 

Continuous data were reported as mean ± standard deviation, whereas categorical data were presented as number (percentage). Statistical analyses were performed using SPSS Statistics version 25 (IBM Corp, Armonk, NY) with significance defines as a P value <.05. Detailed statistical methodologies are provided in the figure legends.

A total of 214 patients with low VWF, defined as a historically lowest VWF:Ag and/or VWF:Act level in the 30 to 50 IU/dL range, were studied, including 118 patients from the LoVIC study and 96 from the LVEMC study. Most patients were female (182 of 214; 85%), and the mean age at enrollment was 37.1 years (range, 4-93 years). In addition, 295 patients with type 2A, 2B, and 2M VWD from the WiN study were used as a control group. The baseline characteristics for both groups are presented in supplemental Table 1. Consistent with previous studies, blood group O (84.8%) and female sex (85.0%) were both significantly (P < .001) more common in the low VWF cohort than in the type 2 VWD subgroup (47.6% and 52.9%, respectively).

In our total cohort of 214 patients with low VWF, 103 subjects were diagnosed with low VWF based on isolated but persistent reductions in plasma VWF:Act in the 30 to 50 IU/dL range and plasma VWF:Ag levels of > 50 IU/dL (Figure 1A, red box). Based on these observations, we subdivided our total low VWF cohort into 2 subgroups for the subsequent analyses, namely (1) low VWF–QT (reduced plasma VWF:Ag level in the 30 to 50 IU/dL range) and (2) low VWF–QL (reduced VWF:Act in the 30 to 50 IU/dL range but plasma VWF:Ag levels >50 IU/dL). Overall, patients with a low VWF–QL accounted for 48.1% of our total low VWF cohort (Figure 1B). VWF:Act was assessed using several different activity assays in our low VWF cohort. Importantly, in the low VWF–QL subgroup, we observed that 63 patients (61.2%) had normal VWF:Ag levels but reduced plasma VWF:Act detected by at least 2 different VWF functional assays (Figure 1C). Only 16 patients were diagnosed with low VWF–QL based on a reduced VWF:RCo alone (Figure 1C). VWF sequencing in 10 of these patients revealed that none carried the p.D1472H polymorphism shown to inhibit ristocetin binding. Furthermore, the difference in the VWF:Act/VWF:Ag ratio observed between the low VWF–QL and low VWF–QT subgroups was consistent over time (Figure 1D). Notably, among the 214 patients with low VWF in our study, 202 had ≥3 VWF measurements performed, whereas 12 patients had 2 VWF measurements performed. The 12 patients with only 2 measurements were primarily young children with a family history of VWD of whom 5 had low VWF–QT and 7 had low VWF–QL (P = .313). Finally, no differences were observed in age (P = .739), ABO blood group (P = .605), or sex (P = .819) between patients with low VWF–QL and those with low VWF–QT (supplemental Figure 1A-C). Collectively, these data demonstrate that many patients with low VWF have plasma VWF:Ag levels >50 IU/dL and are diagnosed based on mild isolated functional VWF defects only. Moreover, in most of these patients, the qualitative VWF dysfunction is seen in >1 functional assay.

Because no previous studies compared patients with low VWF–QL and those with low VWF–QT, we subsequently investigated the bleeding scores in the 2 subgroups. We found that most patients with low VWF–QL had significant bleeding histories, as determined using the ISTH-BAT (Figure 2A). For example, among 75 female patients with low VWF–QL, 32 subjects (42.7%) had ISTH-BAT scores of ≥10. Importantly, there was no significant difference in the overall ISTH-BAT bleeding scores at diagnosis between patients with low VWF–QL and those with low VWF–QT (P = .645; Figure 2A). Moreover, there were no significant differences in bleeding scores within any of the individual ISTH-BAT domains (Figure 2B). In contrast, bleeding scores in patients with low VWF–QL were significantly (P < .001) lower than those with type 2A/2B/2M VWD (supplemental Figure 1D).

In the LVEMC study, patients in the low VWF cohort were followed up after an average duration of 6.6 ± 3.4 years after diagnosis.⁶ All bleeding events and the related treatments documented in the electronic patient files were collected and analyzed. Interestingly, despite the fact that the ISTH-BAT scores were similar between the 2 groups, a significantly (P = .038) higher proportion of patients with low VWF–QL experienced bleeding episodes during follow-up than those with low VWF–QT (Figure 2C-D). In addition, bleeding episodes that required treatment during follow-up were also significantly (P = .045) increased in the low VWF–QL subgroup when compared with the low VWF–QT subgroup (Figure 2E). To further investigate the bleeding phenotype, platelet aggregometry was performed in 119 patients with low VWF. Overall, abnormalities (predominantly subtle and of dubious clinical significance) were identified in only 12 patients (10.1%) (supplemental Figure 2A). Importantly, there was no significant difference (P = .830) in the prevalence of platelet function abnormalities between patients with low VWF–QT (n = 6; 10.7%) and those with low VWF–QL (n = 6; 9.5%). In addition, no abnormalities were found in 89 patients who underwent platelet nucleotide testing. Mild coagulation factor deficiencies were identified in only 9 patients (4.2%) in the total low VWF cohort with no significant difference (P = .821) observed between individuals with low VWF–QT (n = 5; 4.5%) and those with low VWF–QL (n = 4; 3.9%; supplemental Figure 2B). Together, these findings highlight that some patients with low VWF–QL in the 30 to 50 IU/dL range demonstrate a significant bleeding phenotype despite having plasma VWF:Ag levels of >50 IU/dL.

The 2021 ASH/ISTH/NHF/WFH guidelines suggest that patients with plasma VWF:Ag and/or VWF:Act levels in the 30 to 50 IU/dL range and bleeding phenotypes should be subdivided into 2 groups based on the VWF:Act/VWF:Ag ratios (ie, ≥0.7 or <0.7).¹² Of our total cohort of 103 patients with low VWF–QL (defined as having reduced VWF:Act in the 30 to 50 IU/dL range but plasma VWF:Ag levels >50 IU/dL), we observed that 51 (49.5%) had VWF:Act/VWF:Ag ratios of <0.7 (shown in purple; Figure 3A). In contrast, 52 (50.5%) patients with low VWF–QL had plasma VWF:Act/VWF:Ag ratios of ≥0.7. Of note, most of the patients in this latter subgroup had plasma VWF:Ag in the 50 to 60 IU/dL range (shown in orange; Figure 3A). Importantly, there was no significant (P = .372) difference in the ISTH-BAT scores between the patients with low VWF–QL and VWF:Act/VWF:Ag ratios of <0.7 and those with low VWF–QL and VWF:Act/VWF:Ag ratios of ≥0.7 (Figure 3B). Cumulatively, these data suggest that application of a VWF:Act/VWF:Ag ratio threshold (ie, ≥0.7 or <0.7) for patients with low VWF–QL may be of limited clinical utility.

Most (61.2%) of our low VWF–QL cohort had reduced VWF:Act that was detected in 2 separate VWF functional assays (Figure 1C). Based on a lower coefficient of variation and higher reproducibility, the 2021 ASH/ISTH/NHF/WFH guidelines recommend using newer assays to measure the platelet-binding activity of VWF (eg, VWF:GPIbM or VWF:GPIbR) instead of the VWF:RCo assay for patients suspected of having VWD.¹² Consequently, we next specifically investigated plasma VWF:GPIbM levels and VWF:GPIbM/VWF:Ag ratios in our total low VWF cohort. Importantly, we observed that 50 of the 150 (33.3%) in our low VWF cohort had reduced VWF:GPIbM levels but normal VWF:Ag levels (Figure 4A). Thus, even if the analysis was restricted to a VWF:GPIbM assessment of VWF:Act, many patients with low VWF have evidence of a mild functional defect. Furthermore, a range of interpatient variability was seen in the distribution of the VWF:GPIbM/VWF:Ag ratio results in our low VWF cohort (Figure 4B). If we subdivided our total low VWF cohort based on the VWF:GPIbM/VWF:Ag ratio as per the ASH/ISH/NHF/WFH guideline recommendations, 16.7% (n = 25) of our low VWF cohort had VWF:GPIbM/VWF:Ag ratios <0.7 (Figure 4A,C). Similarly, 19.0% (n = 22) of our total low VWF cohort had VWF:RCo/VWF:Ag ratios of <0.7 (supplemental Figure 3). Notably, no significant difference (P = .782) was seen in the ISTH-BAT scores between patients with low VWF with VWF:GPIbM/VWF:Ag ratios of <0.7 and those with low VWF with VWF:GPIbM/VWF:Ag ratios of ≥0.7. (Figure 4D). Altogether, these data further support the hypothesis that there is a significant subgroup of patients with low VWF with abnormal bleeding who have normal VWF:Ag levels but isolated reduced VWF:GP1bM activity levels in the 30 to 50IU/dL range.

The ASH/ISTH/NHF/WFH guidelines suggest that patients with plasma VWF levels in the 30 to 50 IU/dL range with VWF:GPIbM/VWF:Ag ratios of <0.7 should progress to type 2 VWD subtyping.¹² In our total low VWF cohort, 59 (27.6%) had plasma VWF:Act/VWF:Ag ratios of <0.7. Importantly, however, the relationship between VWF:Act/VWF:Ag ratios of <0.7 in patients with low VWF and type 2 VWD has not been defined previously. Consequently, we next investigated the pathogenic mechanisms that underlie low VWF with VWF:GPIbM/VWF:Ag ratios of <0.7. Potentially pathogenic VWF sequence variants were identified in only 50% of our low VWF cohort with VWF:GPIbM/VWF:Ag ratios of <0.7, which was not significantly different from our patients with low VWF with ratios of ≥ 0.7 (41.7%; P = .628) (Figure 5A). This contrasts markedly with what is observed in patients with type 2 VWD (all with VWF:Act <30 IU/dL) among whom pathogenic VWF variants were identified in 99.1% (Figure 5A). In addition, there was no significant difference in the VWF:GPIbM/VWF:Ag ratio between patients with low VWF with and those without VWF sequence variants (0.89 ± 0.36 vs 0.82 ± 0.10, respectively; P = .392). No classical type 2 VWD sequence variants were seen in our low VWF cohort with VWF:GP1bM/VWF:Ag ratios of <0.7.²³ Moreover, none of the 8 VWF variants identified in patients with low VWF with VWF:GPIbM/VWF:Ag ratios of <0.7 were seen in patients with type 2 VWD (Figure 5B). Conversely, 5 of the 8 VWF variants identified in this subgroup were also found in patients with low VWF with VWF:GP1bM/VWF:Ag ratios of ≥0.7 (Figure 5C).²² Collectively, these findings demonstrate that the genetic basis that underlies cases of low VWF with VWF:GP1bM/VWF:Ag ratios of <0.7 is distinct from that of type 2 VWD.

Previous studies have used the VWFpp/VWF:Ag and FVIII:C/VWF:Ag ratios to gain insights into the roles of reduced VWF synthesis and/or enhanced VWF clearance in VWD pathogenesis.^16,22,24 We observed no significant differences in VWF synthesis/secretion (as assessed by the FVIII:C/VWF:Ag ratio) between patients with low VWF with VWF:GPIbM/VWF:Ag ratios of <0.7 and those with low VWF with VWF:GPIbM/VWF:Ag ratios of ≥0.7 (Figure 6A). Conversely, reduced synthesis/secretion of VWF was significantly (P < .001) more marked in patients with type 2A/2B/2M VWD. Similarly, there was no difference in VWF clearance (assessed by the VWFpp/VWF:Ag ratio) between the 2 low VWF subgroups (Figure 6B). However, pathological enhanced VWF clearance was also significantly (P < .001) more marked in patients with type 2A/2B/2M VWD (Figure 6B). Altogether, these findings demonstrate that low VWF with VWF:GPIbM/VWF:Ag ratios of <0.7 is a distinct entity that is different from type 2 VWD.

To further investigate the pathobiology that underpins cases of low VWF with VWF:GPIbM/VWF:Ag ratios of <0.7, we subsequently investigated the VWF:Act/VWF:Ag ratios before and after desmopressin. One hour after desmopressin administration, there was significantly increased plasma VWF:Ag and VWF:Act levels among patients with VWF:GPIbM/VWF:Ag ratios of ≥0.7 and <0.7. Consistent with increased secretion of stored high molecular weight VWF multimers from Weibel Palade bodies, the VWF:Act/VWF:Ag ratio also increased after desmopressin in both groups (Figure 6C). Importantly, however, we observed that the VWF:Act/VWF:Ag ratio after desmopressin remained significantly (P < .001) reduced over time in the low VWF group with baseline VWF:GPIbM/VWF:Ag ratios of <0.7 when compared with the group with baseline ratios of ≥0.7 (Figure 6C). In addition, the desmopressin response observed in the low VWF subgroup with VWF:GPIbM/VWF:Ag ratios of <0.7 was again markedly different from that observed in patients with type 2 VWD (Figure 6C). Taken together, these data suggest that, in patients with low VWF with VWF:GPIbM/VWF:Ag ratios of <0.7, desmopressin triggers secretion of VWF with reduced functional activity from endothelial cells. Furthermore, this reduced VWF function occurs in many cases independent of identifiable pathological VWF sequence variants.

No loss of high molecular weight multimers was observed in any of our patients in the low VWF cohort with VWF:Act/VWF:Ag ratios of <0.7. Consequently, based on the ASH/ISTH/NHF/WFH guidelines, these 59 patients would all be diagnosed as type 2M VWD.¹² Consequently, we subsequently compared the subgroup with low VWF and VWF:GPIbM/VWF:Ag ratios of <0.7 (n = 25) with patients with type 2M (n = 35). Blood group O phenotype and female sex were both significantly more common in the low VWF subgroup with VWF:GPIbM/VWF:Ag ratios of <0.7 than in the type 2M cohort (Figure 7A). Conversely, VWF sequence variants were significantly (P = .003) more frequent among patients with type 2M VWD (Figure 7B). Moreover, none of the VWF sequence variants identified in the low VWF group with VWF:GPIbM/VWF:Ag ratios of <0.7 were seen in patients with type 2M VWD (Figure 7C). Finally, the FVIII:C/VWF:Ag ratio (Figure 7D), VWFpp/VWF:Ag ratio (Figure 7E), and desmopressin responses (Figure 7F) were all significantly (P < .001) different between the low VWF subgroup with VWF:GPIbM/VWF:Ag ratios of <0.7 and the type 2M cohort. Cumulatively, these findings demonstrate that the low VWF subgroup with VWF:GPIbM/VWF:Ag ratios of <0.7 is a distinct clinicopathological entity that is different from type 2M VWD.

Desmopressin response analysis demonstrated that, at baseline, plasma VWF:Act levels were lowest in patients with type 2A/B/M VWD (15 ± 7 IU/dL), intermediate in individuals with low VWF and VWF:GPIbM/VWF:Ag ratios of <0.7 (41 ± 8 IU/dL), and highest in those with low VWF and VWF:GPIbM/VWF:Ag ratios of ≥0.7 (52 ± 15 IU/dL; P < .001; supplemental Figure 4). Importantly, although the postdesmopressin VWF:Act levels remained significantly different (P < .001) among these 3 groups, patients with low VWF–QL demonstrated a strong initial response (158 ± 40 IU/dL) and a sustained increase (115 ± 33 IU/dL) at 4 hours after administration (supplemental Figure 4). Collectively, these data suggest that patients with low VWF–QL can be effectively treated with desmopressin.

For many years, expert consensus guidelines for VWD diagnosis have recommended that VWF:Ag and VWF:Act assays be performed.^10,11,25 Based on these tests, patients with plasma VWF levels of <30 IU/dL and reduced VWF:Act/VWF:Ag ratios (typically <0.7) have been diagnosed with type 2 VWD.^10,11,25 The rationale for classifying patients with qualitative type 2 VWD as distinct from those with quantitative type 1 VWD was strong and had important implications in terms of inheritance, VWF sequence variants, pathogenic mechanisms, and clinical management.^10,11,25-28 In contrast, however, no previous studies have explored the clinical significance of mild qualitative vs quantitative reductions in plasma VWF:Act levels that fall within the range of 30 to 50 IU/dL. Consequently, the novel data from this combined analysis of the Dutch and Irish low VWF cohort studies have direct translational implications. In particular, we demonstrated that approximately 50% of all patients with low VWF in both studies had plasma VWF:Ag levels of >50 IU/dL. These patients were previously diagnosed with low VWF because of persistent reductions in plasma VWF:Act in the 30 to 50 IU/dL range and thus had qualitative rather than quantitative low VWF.

Laboratory variability in VWF functional assays is well recognized and may form a potential limitation of the current study, particularly with respect to the VWF:RCo assay.^12,17,29 Importantly, however, we observed that more than 60% of our low VWF–QL cohort had reduced plasma VWF:Act that was detected in several different functional assays. Furthermore, 16 of the 103 patients with low VWF–QL were diagnosed based on a persistently reduced VWF:RCo assay alone and none of these patients who underwent VWF sequencing carried the p.D1472H polymorphism.^30,31 Collectively, these findings support the hypothesis that, regardless of which VWF:Act assays are used, there is a significant cohort of patients who have mild functional VWF defects despite concurrent plasma VWF:Ag levels that are within the normal range. Importantly, our findings highlight the clinical utility of measuring VWF:CB in patients with mild VWF defects, because 9 patients were diagnosed with low VWF–QL solely based on a reduced VWF:CB measurement. Furthermore, the difference in the VWF:Act/VWF:Ag ratio observed in these patients is consistent over time.

A key finding of this study is that many patients with mild to moderate VWF functional defects in the 30 to 50 IU/dL range have a significant bleeding phenotype. In fact, we observed that the ISTH-BAT scores were not significantly different between patients with low VWF–QL and those with low VWF–QT. In both groups, the highest scores were observed in the menorrhagia domain. Additional studies will be necessary to determine the biological mechanisms that underlie this significant gynecological bleeding in women with low VWF–QL.⁸ Intriguingly, follow-up data from the LVEMC study suggest that bleeding in patients with low VWF–QL may even be more marked than in patients with low VWF–QT. Although previous studies have shown that elevated bleeding scores can predict future bleeding episodes, we observed differences in the ISTH-BAT scores and the prospective follow-up bleeding data in our cohort.^32,33 We hypothesize that this may be related, in part, to saturation of the ISTH-BAT domains with aging. Thus, future studies are needed to investigate the bleeding phenotypes of patients with low VWF–QL and those with low VWF–QT in greater detail.³⁴ 

Our data highlighted that some patients with low VWF–QL have a bleeding phenotype that seems to be discrepantly severe when compared with the mild to moderate reductions observed in their plasma VWF:Act levels.^2,35 These clinical observations suggest that modest reductions in plasma VWF:Act levels may not fully explain the bleeding phenotype in these subjects.³⁶ Their bleeding phenotype could also not be attributed to mild platelet function defects or other coagulation factor deficiencies. Further studies, including more specialized platelet function testing, are required to elucidate the pathological mechanisms that underlie the increased bleeding phenotype observed in the low VWF–QL cohort. Nonetheless, our data suggest that patients with low VWF–QL with an increased bleeding phenotype should be registered and followed up in comprehensive care centers, although their plasma VWF:Ag levels are consistently within the normal range.³⁵ 

Overall, our findings support the current ASH/ISTH/NHF/WFH guideline recommendation that patients with a personal bleeding history and either VWF:Ag or VWF:Act levels in the 30 to 50 IU/dL range should be diagnosed with VWD.¹² The ASH/ISTH/NHF/WFH guidelines further recommend that patients with plasma VWF:Act levels in the 30 to 50 IU/dL range with VWF:Act/VWF:Ag ratios of <0.7 should be diagnosed with type 2 VWD.¹² Our findings raise important questions regarding the clinical utility of applying a VWF:Act/VWF:Ag threshold of <0.7 to subclassify patients with mild to moderate reductions in plasma VWF:Act. In particular, the data presented herein highlight that low VWF–QL represents a distinct entity that is different from type 2 VWD, likely attributable, in part, to differences in the severity of their functional VWF defects (ie, mild vs severe). It should be noted that although previous VWD consensus guidelines have proposed various VWF:Act/VWF:Ag ratio thresholds for distinguishing between type 1 and type 2 VWD, our analysis adheres to the 0.7 cutoff recommended in the current ASH/ISTH/NHF/WFH guidelines.¹² Notable differences between patients with low VWF–QL and those with type 2 VWD include (1) less severe bleeding phenotype in low VWF–QL; (2) less prevalent and different VWF sequence variants in low VWF–QL; (3) significantly attenuated VWF biosynthetic and clearance defects in low VWF–QL, and (4) improved desmopressin responses in low VWF–QL. Conversely, our findings suggest that the low VWF–QT and low VWF–QL subgroups are broadly similar in these respects. Thus, in our opinion, the clinical rationale to subclassify patients with low VWF according to their VWF:Act/VWF:Ag ratios is questionable.

The ASH/ISTH/NHF/WFH guidelines removed the low VWF category used in previous consensus guidelines and instead recommended that patients with plasma VWF:Ag levels in the 30 to 50 IU/dL range and abnormal bleeding should be diagnosed as type 1 or type 2 VWD based on their VWF:Act/VWF:Ag ratio.¹² Based on our results, it seems appropriate that patients with mild to moderate reductions in VWF function in the 30 to 50 IU/dL range should all be classified as type 1 VWD, regardless of their VWF:Act/VWF:Ag ratios. However, in certain cases, such as a family history of type 2 VWD or a discrepantly severe bleeding phenotype, additional analysis may be required to further characterize the patient’s phenotype. It is also important to acknowledge that this approach is somewhat at odds with the traditional perception of type 1 VWD as being caused by a partial quantitative reduction in plasma VWF levels.^3,5,25,37-41 An alternate approach is that patients with a bleeding phenotype and plasma VWF:Ag or VWF:Act levels in the 30 to 50 IU/dL range could be diagnosed as low VWF. Based on our data, there is little clinical rationale for further classifying patients with low VWF into low VWF–QT and low VWF–QL subgroups.

Previous studies have highlighted that plasma VWF:Ag levels increase in many patients with type 1 VWD and low VWF with increasing age.^14,42 Conversely, however, although plasma VWF:Ag levels increase with aging in patients with type 2 VWD, there is typically a limited increment in VWF:Act.⁴³ Consequently, bleeding has been reported to increase with aging in patients with type 2 VWD.⁴³ Given that low VWF–QL is a functional VWD disorder, further research will be required to define the effects of aging on VWF:Ag, VWF:Act, and bleeding in this cohort.

In terms of the underlying pathogenesis, our findings suggest that the etiologies of low VWF–QL and low VWF–QT are similar. Based on the FVIII:C/VWF:Ag and VWFpp/VWF:Ag ratios, it seems likely that reduced VWF biosynthesis is more common than pathological enhanced clearance in patients with low VWF–QL. Nonetheless, from a clinical treatment perspective, it is important to note that patients with low VWF–QL still responded very well to desmopressin treatment. Interestingly, however, although plasma VWF:Ag levels increased after desmopressin, the VWF:Act/VWF:Ag ratios remained significantly reduced in the low VWF–QL subgroup when compared with the low VWF–QT cohort. This observation is again consistent with the concept that low VWF–QL may be caused, at least in part, by alterations in VWF biosynthesis and/or secretion from endothelial cells. Because most patients with low VWF–QL do not have pathological VWF sequence variants, further studies will be required to determine the intracellular mechanisms within endothelial cells (particularly with respect to posttranslational modification and glycosylation) that contribute to the secretion of dysfunctional VWF in these patients.^44,45 It is important to note that the pathogenicity of VWF variants identified in our low VWF cohort remains unclear because of the absence of experimental validation. In addition, in keeping with recent reports, variability in the penetrance of some VWF variants that we observed (eg, Y1584C) may impact interpretation.^46,47 

In conclusion, to our knowledge, this is the first study to specifically investigate the clinical implications of mild functional VWF defects. Critically, our findings demonstrate that many patients with reduced plasma VWF:Act levels in the 30 to 50 IU/dL range have significant bleeding phenotypes, although their plasma VWF:Ag levels are within the normal range. Our data further demonstrate that low VWF–QL is distinct from type 2 VWD. Cumulatively, these novel observations have important clinical implications for the diagnosis and management of patients with mild functional VWF defects.

The visual abstract was created with BioRender.com.

F.A. is supported by a Rubicon grant (452022310) from the Netherlands Organization for Health Research and Development (ZonMw). J.S.O. is supported by a Science Foundation Ireland Frontiers for the Future (FFP) Award (20/FFP-A/8952) and the National Institutes of Health, National Heart, Lung, and Blood Institute for the Zimmerman Program (HL081588). The WiN study was supported, in part, by research funding from the Dutch Hemophilia Foundation (Stichting Haemophilia), Shire (Takeda), and CSL Behring (unrestricted grant).

Contribution: F.A., F.W.G.L., and J.S.O. designed the research and wrote the article; F.A. performed the statistical analysis; F.A., R.B., C.B.v.K., A.-M.H., D.D., M.L., J.G.v.d.B., N.M.O., J.d.M., K.R., S.E.M.S., W.L.v.H., M.D., M.B., F.C.J.I.H.-M., K.P.M.v.G., M.H.C., K.F., K.M., J.E., F.W.G.L., and J.S.O. contributed to patient enrollment, literature review, data interpretation, final draft writing, and critical revision; R.J.S.P., R.I.B., P.J., and J.D.P. contributed to literature review, data interpretation, final draft writing, and critical revision; and all the authors have participated sufficiently in this work, take public responsibility for the content, and gave consent to the final version of the article.

Conflict-of-interest disclosure: F.W.G.L. reports receiving unrestricted grants/research funding from CSL Behring, UniQure, Sobi, and Takeda; receiving consultancy fees from BioMarin, CSL Behring, Takeda, and UniQure (all fees to the institution); and serving as data safety and monitoring board member for a study sponsored by Roche. J.S.O. reports serving on the speaker’s bureau for Baxter, Bayer, Novo Nordisk, Sobi, Boehringer Ingelheim, Leo Pharma, Takeda, and Octapharma; serving on the advisory boards of Baxter, Sobi, Bayer, Octapharma, CSL Behring, Daiichi Sankyo, Boehringer Ingelheim, Takeda, and Pfizer; and receiving research grant funding awards from 3M, Baxter, Bayer, Pfizer, Shire, Takeda, and Novo Nordisk. D.D. reports receiving honoraria from Takeda and receiving educational support sponsorship from Novo Nordisk and Amgen. M.L. reports receiving consultancy fees from Sobi, Band Therapeutics, and CSL Behring; receiving honoraria from CSL Behring and Pfizer; and receiving indirect funding for the development of educational content from Takeda. J.G.v.d.B. reports receiving research funding from Novo Nordisk. N.M.O. reports serving on the advisory boards of Sobi, F. Hoffmann-La Roche Ltd, UniQure, and Freeline and serving on the speaker’s bureau for Novo Nordisk. F.C.J.I.H.-M. reports receiving an unrestricted grant from Octapharma. S.E.M.S. reports receiving research funding from Bayer. R.I.B. reports that the institution received research support/clinical trial funding from Bayer, Takeda, Pfizer, Daiichi Sankyo, CSL Behring, Roche, Amgen, AstraZeneca, AbbVie, Sanofi, Acerta Pharma, Janssen-Cileg, Bristol Myers Squibb, Boehringer Ingelheim, Werfen, and Technoclone unrelated to the current study. M.B. reports receiving consultancy fees from Sobi. K.M. reports receiving speaker fees from Alexion, Bayer, and CSL Behring; participating in trial steering committees for Bayer and AstraZeneca; receiving consulting fees from UniQure and Therini Bio; and participating in data monitoring and end point adjudication committee for Octapharma (all fees paid to the institution). J.E. reports receiving unrestricted research funding from CSL Behring. P.J. reports receiving research funding from Bayer and receiving consultancy fees from Band/Guardian Therapeutics, Star/Vega Therapeutics, and Roche. K.F. reports receiving unrestricted grants/research funding from CSL Behring, Sobi, and Takeda for research unrelated to the current study and receiving consultancy fees from Sobi, Sanofi, Takeda, Novo Nordisk, and Roche (all fees to the institution). K.P.M.v.G. reports receiving an unrestricted research grant from Octapharma. The remaining authors declare no competing financial interests.

Correspondence: Ferdows Atiq, Irish Centre for Vascular Biology, Royal College of Surgeons in Ireland, 123 St Stephen’s Green, Dublin, Ireland; and Department of Haematology, Erasmus University Medical Center-Erasmus CM, Dr Molewaterplein 40, 3015 GD Rotterdam, The Netherlands; email: f.atiq@erasmusmc.nl/ ferdowsatiq@rcsi.ie.