Can you grow out of von Willebrand disease?

The problem of diagnosing von Willebrand disease (VWD) is well known and has spawned a proliferation of guidelines.1-3 The problem relates mainly to the diagnosis of milder forms of type 1 VWD and their distinction from "low VWF" levels which may contribute to bleeding risk but not cause VWD.4 The question "can you grow out of VWD" arises from the well-known effect of increasing age on plasma levels of von Willebrand factor (VWF). Is this another diagnostic problem for VWD? One of the few benefits of growing older? Or is it a natural experiment giving insight into the relationship between VWF and haemostasis? Inasmuch as VWD can be defined as a bleeding disorder arising from a deficiency of VWF haemostatic activity, this problem can be broken down into the following questions: Twin studies indicate that 60% of the population variation in VWF is genetic, one-third of which is attributable to ABO blood group, leaving a considerable amount attributable to acquired influences.5 A study by Gill of 1117 blood donors confirmed the effect of ABO group but noted that within blood groups the only significant correlation was a positive association with age.6 Several studies have confirmed that levels of VWF increase by approximately 10% per decade.7 The largest cross-sectional study, by Davies et al, measured VWF in 5052 blood donors in South Wales. This demonstrated a rise in VWF with age that was most marked after the age of 40.8 The rise was only 0.03 iu/mL between the third and fourth decade, increasing to 0.15 iu/mL between the sixth and seventh decade. However, Albanez et al have examined this in more detail and found that the effect of age was dependent on blood group. Although VWF in non-O individuals rose progressively with age, reaching a 1.71-fold rise in old age (mean 71±7, 55-87 years), the corresponding figure for group O was only 1.25. Interestingly, the well-documented ABO difference was barely evident in younger individuals but became more marked with age.9 Importantly, from a haemostatic point of view, the levels of Factor VIII remained closely correlated with VWF and showed largely parallel rises with age. The mechanism underlying this rise is not fully known but appears to be multifactorial. Studies of VWF propeptide (VWFpp) have shown that VWF secretion increases with age but again this is significant only in the oldest age group and only in non-O individuals.9 The reason for the increase in production with age is again not known but may reflect vascular damage or inflammation.10 There is an age-related rise in inflammatory mediators which may be responsible and blood group A has been linked to the development of atherosclerosis.11 Despite the rise in secretion, the VWFpp:Ag ratio falls with age indicating an increase in half-life as well.9, 12 This is consistent with the reported increase in FVIII half-life with age because FVIII survival is largely determined by VWF.13 The reduction in clearance is unexplained but once more is significantly more marked in the non-O group. The net effect reported by Albanez was that non-O individuals can expect their VWF:Ag to rise from a mean of 0.98 to a mean of 1.68 iu/mL between childhood and old age, while group O rise from mean 0.85 to 1.11 iu/mL.9 These well-documented changes describe the effect of age on VWF in the normal population but are unlikely to be reproduced in many patients with VWD. Clearly patients with type 3 VWD do not have the capacity to exhibit a rise under any circumstances and patients with significant defects in VWF synthesis (type 1 or type 2) are also unlikely to be able to produce rises of similar degree. In many patients with type 1 VWD, an increase in VWD clearance has been shown to be a contributory factor to the reduction in levels: this would tend to blunt any rise due to increased synthesis but may allow them to benefit from an age-related reduction in clearance, provided the normal physiological mechanisms are involved.14 Patients with type 2 variants may increase production, but any rise in activity is likely to be proportionately less. Historically, many patients with VWF levels in the range 0.3-0.5 iu/mL have been diagnosed with VWD. Although this is now often regarded as "low VWF" (indicating that it is a risk factor that may contribute to, but not fully explain, a bleeding tendency)2 these patients may have a positive bleeding history and might be in the best position to benefit from the age-associated VWF rise. Of note, approximately half of these patients do not have a variant in VWF and so are more likely to experience a rise similar to normals.15 On the other hand, they are also more frequently group O and so will experience a less marked rise than non-O individuals. The WiN study comprised 664 patients with a diagnosis of VWD and historical level <0.3 iu/mL (activity or antigen).16 A subset of 66 patients had historical results for comparison over time. As expected, no significant rise was seen in patients with type 2 VWD. In type 1 patients, the rise in VWF:Ag was 0.025 iu/mL per decade which as expected, is slightly less than seen in the normal population. The rise in VWF:RCo of 0.095 iu/mL per decade was slightly higher and although this is likely to be less accurate, it is interesting that ADAMTS13 levels fall with age; so some separation of antigen and function may be real.17 Indeed a study by Favaloro suggested a trend for the Ag:function ratio to fall with age, in keeping with this hypothesis.18 A Canadian study of patients with VWD documented a rise in mean VWF:Ag from 0.44 iu/mL to 0.71 iu/mL over a mean of 11 years follow-up. However, although in the group with baseline VWF:Ag <0.3 iu/mL the mean VWF:Ag rose from 0.23 to 0.57 iu/mL, only 2 of 8 patients had levels that rose from <0.3 to >0.5 iu/mL.19 This was a small study of 31 patients and it is not clear why the rises were higher than other studies. In summary, the level of VWF rises significantly with age, but the effects are not surprisingly less marked in those with definite VWD (<0.3), the majority of whom will have a variant in VWF. Those with "low VWF" are better able to increase VWF with age but are most commonly blood group O whose age-related rise is less. Given that the incidence of many morbidities increases with age, it would not be surprising if there was a parallel increase in bleeding symptoms. In the European MCDMC study, Tosetto et al20 found that bleeding score did indeed go up with age in patients with VWD, but not in the control group or unaffected family members. However, because the bleeding score is cumulative, this does not mean that bleeding became more common with age, only that it continued as the patients grew older. Nonetheless, some specific problems such as bleeding from angiodysplasia21 and haematuria22 are known to increase with age and age is an important component of a bleeding risk assessment tool for patients anticoagulated for atrial fibrillation.23 For example, in the Leiden cohort, age >80 years compared to <60 years was associated with an adjusted hazard ratio for bleeding of 2.9 (95% CI: 1.7-4.8) despite similar quality of anticoagulant control.24 Notably, the effect of age below 80 was not significant and this was also seen in the National Consortium of Anticoagulation Clinics study (RR for >80: 4.6 95% CI 1.2-18.1).25 In the Inception Cohort study, the effect of age was a non-significant increasing trend but with an excess of fatal haemorrhage in the elderly.26 In a web-based survey of younger, healthy individuals there was no increase in reported bleeding symptoms with age.27 In the WiN study, elderly (>65 years) patients with VWD reported more bleeding episodes in the year preceding the study than younger patients: 42% vs 31% (P=.048). Interestingly, this difference was entirely within the patients with type 2 disease and there was no difference for patients with type 1 VWD.16 Equally of note, the rise in VWF did not lead to a reduction in bleeding episodes in patients with type 1VWD. Elderly patients tended to bleed more frequently after surgery, had more GI bleeding and reported a higher number of comorbidities. If the threshold VWF level for increased bleeding risk rose with age, then different criteria for diagnosis and treatment could be applied in this group. In summary, there is little evidence for a general increase in bleeding associated with age independent of age-related comorbidities, but in patients with a bleeding diathesis such as VWD or anticoagulation, there may be an increase with advanced age. The question of whether a level of activity is low enough to cause or to explain bleeding is difficult. In the past many patients with levels below the lower limit of normal have been diagnosed with VWD on the basis of a reduced level alone or with coincidental symptoms but there are many individuals with levels <0.5 iu/mL who have no excessive bleeding tendency.4 The relative risk of bleeding for VWF:Ag below (vs above) 0.5 iu/mL is modest and of borderline significance: RR= 2.0 (P=.51).4 A similar result was obtained by Castaman et al who also noted that FVIII had a greater influence on the risk than VWF:Ag.28 In the RENAWI study, the bleeding score began to rise only when the VWF:RCo fell below 0.3 iu/mL.29 These data suggest that for patients with a significant bleeding history, a rise into the range 0.3-0.5 will likely reduce their bleeding risk and a rise to >0.5 will reduce it further but the possibility that other contributing risk factors are present means that it may not return entirely to normal. Most patients with "low VWF" can be satisfactorily treated with desmopressin which will correct the primary haemostatic defect by elevating VWF and possibly sensitising platelets. In older patients, this approach may be contraindicated and the physician must decide whether a bleeding risk still exists and if so whether it is primarily due to an inadequate level of VWF or to another defect such as platelet dysfunction. It might be observed, for example, that PFA100 closure times remain prolonged despite a rise of VWF activity into the normal range. They may then be able to recommend either VWF or platelet replacement as prophylaxis against bleeding. Most patients with definite VWD will not experience a rise in VWF activity that normalizes their bleeding risk. However, those with less deleterious gene variants or low levels not linked to VWF, will experience a beneficial rise even if they do not achieve fully haemostatic levels. For some, whose VWF was in the borderline range, VWF may enter the normal range which will reduce their bleeding risk and may be sufficient to compensate for other contributory factors such as impaired platelet function; however, it is difficult to say they are growing out of VWD, since we are not certain of the initial diagnosis. Sadly, those with the most severe VWD will benefit the least. Explaining these changes to patients requires time and tact and the question of growing out of VWD proves to be just as difficult to answer as making the diagnosis of VWD in general. In both cases, we are hampered by a lack of adequate data which will come only from large and long-term prospective studies of individuals both with and without VWD. Prof Laffan has received speaker fees from Pfizer and Bayer, advisory board fees from Shire and Sobi and travel support from LFB and Bayer.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.