Abnormal von Willebrand factor multimer pattern without VWF variants in autosomal-recessive cutis laxa type IIA Open Access

TO THE EDITOR:

Cutis laxa comprises a heterogeneous group of connective tissue disorders characterized by inelastic, wrinkled skin.¹ Autosomal-recessive cutis laxa type IIA (ARCL2A; Online Mendelian Inheritance in Man number 219200) results from biallelic pathogenic variants in ATP6V0A2, which encodes the a2 subunit of the vacuolar H⁺-ATPase (V-ATPase) proton pump.^2-4 Loss of V-ATPase function impairs Golgi acidification, disrupting proper N- and O-glycosylation of secretory proteins and resulting in congenital disorders of glycosylation.^2,3,5 Although ARCL2A is primarily recognized for its characteristic dermatologic findings, developmental delay, and skeletal abnormalities, a few reports suggest a relationship with abnormal bleeding, including easy bruising, gingival bleeding, menorrhagia, and delayed wound healing, as well as with abnormal von Willebrand factor (VWF) parameters.^6-8 

An association between ARCL2A and VWF dysfunction is mechanistically plausible given that VWF biosynthesis, multimerization, and storage are critically dependent on acidic pH within the Golgi apparatus and Weibel-Palade bodies.^9-11 VWF undergoes extensive posttranslational processing, with monomers initially forming dimers in the endoplasmic reticulum before further maturation and multimerization in the acidic environment of the Golgi and post-Golgi compartments.^9,12,13 Detailed VWF multimer analyses have been limited in patients with ARCL2A, and standard VWF antigen and activity levels may appear normal or even elevated, particularly during physiologic stress. We present results of a VWF evaluation including multimer analysis for a patient with ARCL2A caused by a novel ATP6V0A2 frameshift variant. The results revealed a distinctive multimer pattern and have important clinical implications for bleeding risk assessment and perioperative management in patients with this disorder.

A female infant born at 33 weeks to consanguineous Egyptian parents was diagnosed with ARCL2A at age 3 months based on characteristic craniofacial dysmorphism, microcephaly, and developmental delays. Next-generation sequencing using the Intellectual Disability, Epilepsy, and Autism panel with reflex to exome sequencing identified a homozygous frameshift variant in ATP6V0A2, c.721del (p.Ile241Tyrfs∗26; NM_012463.3). The variant, which was predicted to cause nonsense-mediated decay, was absent from gnomAD and ClinVar databases and had not been previously reported in the literature. At age 13 months, the patient was referred to pediatric hematology due to the potential association of ARCL2A with bleeding.^6,8 The patient reportedly experienced prolonged bleeding (∼3 hours) after routine finger-stick blood sampling but had no other symptoms of abnormal hemostasis.

Initial hematologic evaluation revealed a normal hemoglobin (12.2 g/dL; reference, 10.5-13.5 g/dL), platelet count (396 × 10³/μL; reference, 150 × 10³/μL to 400 × 10³/μL), prothrombin time (12.1 seconds; reference, 11.9-14.5 seconds), and activated partial thromboplastin time (27.9 seconds; reference, 23.5-33.5 seconds). Factor VIII (FVIII) activity (305%; reference, 50%-150%) and VWF antigen (183%; reference, 45%-160%) were elevated, whereas ristocetin cofactor activity (90%; reference, 45%-150%) and fibrinogen (213 mg/dL; reference, 188-450 mg/dL) were normal. The ratio of VWF ristocetin cofactor activity to VWF antigen was reduced (0.5; normal >0.7), whereas the ratio of FVIII activity to VWF antigen (1.7) was normal. Multimer analysis performed at a reference laboratory reportedly showed a full spectrum of multimers but with increased staining intensity of the low-molecular-weight species (high-molecular-weight multimers, 8%; intermediate-molecular-weight multimers, 36%; low-molecular-weight multimers, 56%). The overall multimer pattern was described as unusual.

At age 31 months, during hospitalization for fever and obstructing renal calculus, repeat testing confirmed the abnormal VWF pattern. FVIII activity (335%), VWF antigen (222%), and fibrinogen (471 mg/dL) were elevated, whereas VWF activity by glycoprotein Ib (GP1b) binding assay using recombinant mutated GP1b (84%; reference, 50%-150%) was normal. The ratio of VWF activity to VWF antigen (0.38) was again significantly reduced, whereas the ratio of FVIII activity to VWF antigen (1.5) was normal. In-house multimer analysis revealed a striking distribution pattern characterized by marked predominance of the lowest-molecular-weight band, which likely represents the VWF dimer (Figure 1, left panel). A full spectrum of multimers was present, but the concentration of high- and intermediate-molecular-weight multimers was reduced relative to the lowest band (Figure 1, right panel). This pattern differs markedly from the pattern associated with classic type 2A von Willebrand disease, which typically shows selective loss of the largest multimers (Figure 1, middle panel).^14-16 Targeted next-generation sequencing spanning the entire VWF and GP1BA genes identified no pathogenic or likely pathogenic variants. The limited amount of plasma available precluded additional testing, and we did not perform platelet lumiaggregometry to assess the effects of ARCL2A on platelet function.

During ureteral stent placement, prophylactic antihemophilic factor/VWF complex (Humate-P; 60 units/kg) was administered once before the procedure, based on the abnormal VWF studies per the treating provider. No postoperative doses were given, and the surgical procedure was completed without bleeding complications.

This is, to our knowledge, the first report of the ATP6V0A2 frameshift variant c.721del (p.Ile241Tyrfs∗26) causing ARCL2A. In addition to typical phenotypic features, the disorder in this patient is associated with an unusual VWF multimer distribution that indicates inefficient VWF multimerization. The dimer-predominant pattern may represent a pathophysiologic mechanism distinct from those associated with inherited VWF disorders caused by mutations in the VWF gene, particularly those responsible for type 2A von Willebrand disease.^15-17 The multimer pattern is consistent with an incomplete defect in early steps of VWF multimerization. Loss of ATP6V0A2 function abolishes acidification of the Golgi and distal secretory pathway, creating an alkaline shift that disrupts pH-dependent glycosyltransferase activity and pH-regulated cargo-receptor interactions.^5,7,10 This may impair the efficiency of dimer-to-multimer conversion while preserving the capacity to form the full spectrum of multimer sizes, albeit with reduced efficiency. It is important to note that the multimer electrophoresis technique shown here used agarose gels under nonreducing conditions. Such a system may not be able to resolve the VWF dimer from low-molecular-weight proteolytic fragments due to ADAMTS13 cleavage or protein instability. These possibilities are not excluded by our analysis. However, Wagner et al demonstrated that raising the pH of the trans-Golgi apparatus in cultured vascular endothelial cells disrupted VWF multimerization but not dimerization.⁹ In a culture system, VWF cleavage by ADAMTS13 is likely to be minimal, supporting the interpretation that the VWF abnormality in ARCL2A is caused by a defect in VWF multimerization of the VWF dimer.

The normal to elevated VWF antigen and activity levels measured for our patient using standard assays may provide false reassurance that the hemostatic capacity of VWF is intact in patients with ARCL2A. The reduced activity-to-antigen ratio (using both the ristocetin cofactor activity and the GP1b binding assay using recombinant mutated GP1b assay) and the abnormal multimer distribution are consistent with a qualitative VWF defect that could compromise hemostasis, particularly during surgical procedures or trauma. The absence of disease-associated variants in the VWF and GP1BA genes further underscores the importance of a complete VWF assessment, including multimer studies. Based on our findings and a review of the literature (Table 1), we recommend that patients with ARCL2A undergo comprehensive VWF assessment including multimer analysis, particularly before surgical procedures. The inefficiency of endogenous VWF multimerization suggests that factor concentrate replacement may be more effective than desmopressin (DDAVP) for hemostatic support, because DDAVP may increase the release of endogenous structurally abnormal VWF.

This report provides detailed characterization of VWF dysfunction in a single patient with ARCL2A but cannot establish the prevalence or clinical significance of these findings across the broader population of patients with ARCL2A. Future multicenter studies correlating ATP6V0A2 genotype with VWF phenotype severity may better inform personalized bleeding risk assessments and management strategies. Additionally, investigation of other proteins requiring Golgi acidification for proper processing may reveal additional abnormalities in ARCL2A that could contribute to bleeding risk. Finally, we did not have sufficient clinical material to assess the possibility that processes distinct from multimerization, such as proteolytic cleavage or protein instability, may have contributed to the abnormal multimer pattern.

This case highlights the importance of a comprehensive coagulation evaluation, particularly, a thorough evaluation of VWF, in patients with ARCL2A. The distinctive VWF multimer pattern provides a potential biomarker for bleeding risk that may guide perioperative management decisions. Therefore, we recommend VWF studies including multimer analysis for all patients with ARCL2A, with particular attention to the activity-to-antigen ratio and multimer distribution pattern. For patients requiring surgical procedures, prophylactic factor concentrate should be considered based on the degree of VWF dysfunction demonstrated.

Acknowledgments: This work was supported by award HL140025 from the National Heart, Lung, and Blood Institute of the National Institutes of Health and by the Ernest W. Goodpasture Chair in Experimental Pathology for Translational Research from Vanderbilt University.

Contribution: J.W.J. and D.G. analyzed the literature, drafted the manuscript, and provided supervision; L.A.B., N.S., R.D., and G.S.B. revised the manuscript; and all authors approved the final version.

Conflict-of-interest disclosure: D.G. is a consultant for several pharmaceutical companies with interests in developing drugs that target factor XI and prekallikrein (unrelated to this article). The remaining authors declare no competing financial interests.

Correspondence: Jeremy W. Jacobs, Department of Pathology, Microbiology, and Immunology, Vanderbilt University Medical Center, 1211 Medical Center Dr, Nashville, TN 37232; email: jeremy.w.jacobs@vumc.org.