“Bleeding in the jungle”

A 55-year-old male, with Hispanic background, presented to one of our facilities with recurrent episodes of epistaxis and hematuria that could not be explained by any anatomical abnormalities. Initial history of bleeding included persistent bleeding following tooth extractions, orthopedic surgery (which did not require reintervention), and several episodes of gastrointestinal bleeding, while living in a rural zone of his country of origin, with limited access to health services. The initial approach to a patient who presents with a history of bleeding starts with a comprehensive review of the bleeding history. A life-long history of bleeding suggests a congenital cause. The type of bleeding provides an important clue as to what the underlying cause may be: deep tissue bleeding such us muscular hematomas and hemarthrosis suggests deficiencies of the plasma phase of hemostasis. On the other hand, mucocutaneous bleeding points to defects of primary hemostasis, including platelet disorders. Unexpected bleeding in previous surgical procedures is of particular importance. For females, additional challenges such as heavy/prolonged menstrual bleeding and post-partum bleeding should also be assessed, although this is not relevant to our particular case. Medications with antithrombotic or antiplatelet activity should be considered. The age of the patient, ethnicity, and family history are also important. Further assessment of the patient's life-long bleeding history using the bleeding assessment tool recommended by the International Society on Thrombosis and Haemostasis (ISTH-BAT) 1 revealed a significant lifelong history of mucocutaneous bleeding including bleeding from tooth extractions, surgical procedures, and the gastrointestinal tract. The total bleeding score was 14. This patient was not on any medications. Although there was a lack of family history of bleeding disorders, this was largely because both parents were deceased and there were few other close relatives. Platelet count and platelet morphology are important to exclude a primary hemostasis disorder. In order to help exclude abnormalities of coagulation, basic screening tests such as the prothrombin time (PT) and the activated partial thromboplastin time (APTT) should be performed initially 2. Initial coagulation studies including PT, APTT, fibrinogen, and platelet count were all within normal limits. Platelet morphology was unremarkable (data not shown). Disorders of platelet function and von Willebrand disease (VWD) cannot be excluded based on these test results. Because of the significant bleeding history, additional tests to exclude VWD or platelet dysfunction are required. Platelet function screening analysis (PFA-100®, Siemens) showed persistently prolonged closure times (>300 sec) with both Col/Epi and Col/ADP cartridges. Initial tests showed absence of VWF activity by ristocetin cofactor (VWF:RCo), but high normal levels of VWF antigen (VWF:Ag) and factor VIII (FVIII:C) (both >150 U/dL). Of additional interest, another functional test for VWF, the collagen binding assay (VWF:CB), also yielded high normal values (>150 U/dL) (Table 1). The maximal closure times for both PFA Col/Epi and Col/ADP are consistent with any of the following: low platelet count, low hematocrit, antiplatelet medication, VWD, or platelet dysfunction. Previous normal blood counts excluded a low hematocrit and platelet count and the patient denied taking any antiplatelet medication. Current VWF test result pattern is unusual, in that there usually is not such a large disparity between the antigen and the activity results. Even with type 2 VWD, which characteristically does have a disparity, the difference between antigen and activity is not usually this striking. Nonetheless, the current test pattern would exclude types 1 and 3 VWD, but might reflect: a rare case of type 2 VWD, an acquired von Willebrand syndrome (AVWS), or might alternatively have arisen from an analytical/pre-analytical problem. Extensive platelet aggregation testing is warranted, as is repeat and extended VWF testing. Other analyses including platelet glycoprotein analysis might be useful. Repeat and extended testing was performed, including use of alternate methodologies at another laboratory (see Table 1 notes for details). VWF:Ag, VWF:RCo, VWF:CB, and FVIII:C, using different methodologies and several different time-point collected samples, consistently showed similar test results to original findings (Fig. 1A). An alternative VWF activity test, based on a gain of function mutation of recombinant glycoprotein Ib (rGPIb) but not involving ristocetin (INNOVANCE® VWF Ac, Siemens) 3, also consistently showed absence of activity. VWF multimer analysis showed an essentially normal pattern (no absence of high molecular weight multimers; data not shown). Platelet aggregation studies showed normal responses to all agonists tested except for ristocetin, where there was a complete lack of response (Fig. 1B), using an extended concentration range from 0.25 up to 2.5 mg/mL. Platelet glycoprotein (GP) expression analysis (using antibodies to GPIb, GPIb/IX, GPIIb/IIIa and GPIIIa) by flow cytometry was normal (data not shown). Panel A: Summary of all phenotypic test results for VWF:Ag, VWF:CB, VWF:RCo, VWF Ac (left axis in U/dL), and assay ratios (CB/Ag, RCo/Ag, and Ac/Ag) (right axis), as performed at two different sites, using a range of methods (see Table 1 footer for summary of methodologies employed). Dashed horizontal lines indicate “representative” cut-offs for distinguishing normal vs. low VWF and FVIII (∼50 U/dL; left axis) or for distinguishing normal vs. low activity/Ag ratios (∼0.7; right axis). Panel B: Aggregation traces from the initial platelet aggregation study showing normal aggregation to arachidonic acid (trace 1), adenosine diphosphate (traces 2 and 3), epinephrine (trace 4), collagen (trace 5), but absent aggregation to ristocetin (traces 6-8). Repeat aggregation testing confirmed this pattern, and overall included a range of ristocetin concentrations from 0.25 to 2.5 mg/mL, all of which failed to elicite an aggregation response. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] These results, although suggestive of VWD, cannot completely exclude AVWS. The absence of VWF GPIb binding, as expressed by absent VWF:RCo and VWF Ac, was unusual given the normal to high levels of VWF:Ag, VWF:CB, and FVIII:C. Exclusion of antibodies to VWF affecting GPIb binding and otherwise indicating AVWS would be useful 4. The platelet aggregation test results might alternatively be consistent with Bernard–Soulier syndrome (i.e., platelets lacking GPIb), but this is unlikely given the normal platelet count and morphology, plus abnormal VWF test pattern. Furthermore, normal flow cytometry for GPIb/IX essentially excludes Bernard–Soulier syndrome, although flow cytometry only detects the presence of the glycoproteins, and generally does not assess whether these components are dysfunctional. Assessment of potential inhibitory antibodies by modified Bethesda assays for various VWF activities 4 failed to identify any antibodies (data not shown). One the other hand, the mechanism for the decrease in VWF in AVWS from VWF antibodies is sometimes increased clearance of VWF rather than inhibition of VWF activity, and these would not be detected in a Bethesda assay, and this could not be assessed. Overall results are consistent with an unusual case of (type 2) VWD. VWD is the most common inherited bleeding disorder, with an estimated prevalence ranging between 0.05 and 1.0% of the population, and is caused by defects and/or deficiency in the plasma protein von Willebrand Factor (VWF) 5, 6. In turn, VWF is a large and complex multimeric protein, the main functions of which are to promote adhesion of platelets to the vasculature at sites of injury and to stabilize factor VIII (FVIII). In order to achieve these basic functions, VWF contains multiple binding sites, most notably for platelet GPIb, collagen, and FVIII, but additionally, heparin and GPIIb/IIIa 7. The large multimeric complexes are subject to intricate processes of storage, secretion, proteolysis, and clearance that are only partially elucidated 7, 8. There are six different types of VWD according to the most recent classification scheme 8: the quantitative disorders are defined as type 1 (partial deficiency) and type 3 (“total” deficiency), whereas qualitative disorders are defined as type 2, comprising 2A (selective deficiency of high-molecular-weight [HMW] multimers affecting platelet adhesion), 2B (VWF with an increased affinity for platelets), 2N (defective binding of VWF with FVIII), and 2M (VWF with decreased activity not associated with loss of HMW multimers). Type 2M VWD is classically considered a rare form of VWD, but increasing evidence suggests that 2M VWD is probably as common as 2A VWD, but is most often misdiagnosed as either type 1 or 2A VWD 5, 9, 10. In this particular case, the abnormal PFA-100 closure times appeared to be due to an absence of VWF—platelet GPIb binding, as also highlighted by undetectable VWF:RCo, VWF Ac, and ristocetin induced binding in platelet aggregation studies; this being despite high normal test values for FVIII:C, VWF:Ag, and VWF:CB (all >150 U/dL). The lack of increase in ristocetin aggregation is not characteristic of type 2B VWD, where enhanced aggregation to ristocetin would be expected. Normal multimer analysis is not characteristic of type 2A or 2B VWD, where a loss of HMW VWF would be expected. A diagnosis of type 2M VWD was made (and 2A and 2B VWD excluded) based on this differential test pattern. Most cases of type 2M VWD identified in the literature display decreased VWF-dependent platelet adhesion due to some loss of GPIb binding, although a handful of cases have been described that show a selective loss of collagen binding with normal GPIb binding 5. The VWF gene is located in the short arm of the chromosome 12 (12p13.3). It spans approximately 178 kb of DNA and contains 52 exons that account for 8,439 base pairs. The VWF protein sequence contains 2,813 amino acids. The first 22 constitute the signal peptide, which is cleaved before the formation of dimers. VWF propeptide (VWFpp) includes amino acids 23–763 and is cleaved from VWF during the process of multimerization. The mature molecule of VWF includes amino acids 764–2,813 7, 8. Domains A1 and A2 are encoded by exon 28. Domain A1 and part of A2 are responsible for interaction with GPIb, collagen, and heparin. Type 2B and 2M VWD are associated with mutations in the region that codes for interaction with GPIb. The A2 domain also contains the ADAMTS13 sensitive bond tyrosine 1605-methionine 1606, and mutations in this region are associated with type 2A VWD, due to increased cleavage of VWF by ADAMTS13. Type 2M mutations cause loss of function, e.g. reduced binding of VWF to GPIb, whereas type 2B mutations cause increased binding of VWF to GPIb on platelets. The increased binding in type 2B VWD causes VWF to spontaneously bind to platelets, and these platelets are cleared from the circulation, leading also to loss of VWF. Genetic studies are useful to confirm the presence of mutations associated with type 2 VWD, including type 2M. Genetic analysis of the VWF gene by sequencing of exon 28, in both forward and reverse direction, in this patient revealed a homozygous missense variant (c.3974C>T), causing the amino acid substitution p.Ser1325Phe (Ser562Phe using an older nomenclature). The online VWF database mutation registry 11 identifies that 33 of 42 mutations registered for type 2M VWD reside in exon 28. None of these mutations involve the amino acid residue 1325. A further search of the literature also failed to identify any previous reports of a mutation at this residue. Two mutations have been reported involving the amino acid residue 1324. One causes the substitution Gly for Ser and the other Gly for Ala. The p.Gly1324Ala has been further investigated: an elegant study of gene expression by Hilbert et al. 12 in France proved that this specific amino acid substitution was the culprit for the phenotype of a family with VWD. The phenotypic expression of that family also identified absent ristocetin based GPIb binding, although fewer VWF activity tests were performed, and VWF:Ag results were low (21–61 U/dL) in all family members. Genetic studies in our case confirmed a missense mutation in exon 28 of the VWF gene that explains the abnormal interaction of the protein with GPIb and the clinical phenotype of the patient. To estimate to what extent amino acid substitution p.Ser1325Phe would affect the VWFA1 domain and the interaction with GPIb, we examined the crystal structure of VWFA1 13 and of the complex between these two proteins. Figure 2A shows that p.Ser1325 lies at the interface between VWFA1 and GPIb 14. The main chain carbonyl oxygen and nitrogen from p.Ser1325 hydrogen bonds to the main chain carbonyl oxygen and nitrogen of GpIb residue p.Met255 respectively and the p.Ser1325 side chain is buried at the interface in the complex. The substitution of p.Ser1325 by the more bulky Phe side chain likely prevents formation of the complex with GPIb via a steric mechanism and may also disrupt the local main chain conformation of p.Ser1325 precluding hydrogen bonding to p.Met255. This represents a selective abnormality of the A1 domain of VWF that is responsible for its interaction with platelet GPIb, and not an abnormality of collagen binding, which may otherwise arise from mutations in the A3 domain. The region is also highly conserved in mammals (Fig. 2B). Panel A: Cartoon diagram illustrating the crystal structure (pdb:1SQ0) of the complex between VWFA1 (orange) and GPIb (white). The amino acids p.Ser1325 (VWFA1) and p.Met255 (GPIb) are shown as spheres. The main chain nitrogen atom from p.Ser1325 is shown in blue which hydrogen bonds to the main chain oxygen (red) of p.Met255. The p.Ser1325Phe mutation will likely disrupt this VWFA1/GPIb interface. Panel B: High conservation of this region among mammals. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] We report an interesting new case of 2M VWD that shows an extreme phenotypic expression of undetectable GPIb binding, but high levels of VWF protein, collagen binding, and FVIII. This case is considered unique for the following reasons: (i) this is the first case of VWD identifying a mutation in the amino acid residue 1325 (p.Ser1325Phe), where the substitution of serine for phenylalanine has a deleterious effect in the function of the VWF on the binding to GPIb; (ii) there was undetectable binding of the patient's VWF to GPIb by several assays, VWF:RCo, VWF Ac (assay based on rGPIb binding), and in the platelet aggregation assay; (iii) the VWF:Ag level, VWF:CB activity, and FVIII were all in the high normal range (>150 U/dL). The reason for these high normal values are not clear, but the patient was identified to have A Rh positive blood type, which is known to be associated with relatively high VWF and FVIII levels 15. The clinical manifestations of this patient confirm the diagnosis of type 2M VWD. Despite the unique nature of this case, it provides several important lessons. First, the mutation in the A1 domain caused a defect in GPIb binding that did not affect collagen binding (A3 domain). Second, use of a limited VWF test panel comprising VWF:Ag, VWF:CB, and FVIII:C would have led to a missed VWD diagnosis, and performance of multimers would not have altered this missed diagnosis. Accordingly, laboratories are always encouraged to perform test panels that represent the full range of VWF activities, including both GPIb and collagen binding. Third, there is a role for genetic testing when investigating a patient with suspected type 2 VWD. Genetic testing in our patient helped identify a unique mutation, but would otherwise also help to differentiate between (carrier status for) hemophilia A and type 2N VWD, and platelet-type VWD from type 2B VWD 16, 17. In most cases, sequencing of exon 28 would be sufficient to characterize type 2B, 2M, and some cases of type 2A VWD. Exon 28 has the added advantage of being a relatively short segment of DNA, taking into account the length of the complete VWF gene. Genetic testing has a more limited role in other forms of VWD, although a role in the prenatal assessment or family investigations of type 3 VWD is acknowledged 16, 17. Alejandro Arbelaez, Johan Niemann, and Robert Freney were involved in the initial characterisation of the case, and the initial diagnosis of the case as a type 2 VWD, inclusive of clinical and laboratory evaluation at the Mater. Emmanuel Favaloro was involved in confirmation laboratory studies at Westmead, as well as co-ordinating additional investigations. Soma Mohammed was also involved in confirmation laboratory studies at Westmead. Maha Othman and Jonas Emsley were involved in analysing the structure-function implications of the VWF mutation reported. All authors contributed to writing of the manuscript.

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.