A Bleed With No History: Reasoning a Path to Type 2M von Willebrand Disease

A Bleed With No History: Reasoning a Path to Type 2M von Willebrand Disease

When von Willebrand factor is present in a normal amount and normally assembled, yet still cannot grip the platelet.

By William C. Aird, MD

Case Presentation

A man in his late thirties, with obstructive sleep apnea, undergoes uvulopalatopharyngoplasty. He does well at first. On the tenth day after the operation he bleeds from the surgical bed and returns to the operating room for cautery. Days later he bleeds again and goes back a second time for suture ligation.

He has never bled like this before. His wisdom teeth came out uneventfully. He recalls only a few nosebleeds after minor trauma in his youth. No one in his family bleeds. Screening studies are unremarkable: normal PT, normal aPTT, normal platelet count, and a normal blood count and film.

One point is worth holding in mind. Delayed bleeding after this particular operation is not by itself rare, because the healing eschar tends to separate around this time. The question is not only whether the patient has a bleeding disorder, but whether something in him turned an expected surgical risk into one that required the operating room twice.

Given a significant surgical bleed, a normal PT and aPTT, and no personal or family bleeding history, which statement is most accurate?

Explanation

PT and aPTT report on the coagulation factors, not on the interaction between von Willebrand factor and the platelet. A qualitative VWF defect can leave both screening tests entirely normal. A negative family history offers little reassurance either: penetrance is variable, and a carrier may simply never meet a hemostatic challenge large enough to expose the defect. A real bleed in a real surgical field is reason enough to look specifically at VWF.

Reveal: the von Willebrand panel

Targeted testing follows. Across several draws over the next months, a pattern holds.

The signal is in the gap. Activity sits well below antigen, while factor VIII is comparatively preserved.

A VWF antigen that is only mildly reduced, sitting well above a much lower VWF activity, indicates which of the following?

Explanation

When the ratio of activity to antigen falls, the problem is quality, not quantity. Type 1 lowers antigen and activity together and preserves their ratio. Type 3 leaves almost no protein to measure. Here there is plenty of protein doing too little work, which places the patient in type 2. That is not the end of the reasoning, though. Type 2 is not one disease but four, and each of the four fails at a different step in the life of the molecule.

Reveal: the discriminating data

More returns.

A qualitative (type 2) VWF defect with normal multimers, a persistently normal platelet count, and preserved factor VIII points to which subtype?

Explanation

This is the center of the case, and it is best reasoned by subtraction, one preserved function at a time.

Type 2A and type 2B both lose the largest multimers. This patient’s multimers are intact on repeated testing, so both are excluded. Type 2B tends, in addition, to clear platelets and lower the count; his count is normal, which removes the last support for 2B.

Type 2N spares the ratio of activity to antigen and instead drops factor VIII by impairing the site where VWF carries factor VIII. His factor VIII is preserved and his activity to antigen ratio is clearly reduced, the opposite pattern. Type 2N is out.

What remains is type 2M, and within it a distinction matters. Type 2M gathers the qualitative variants that lose a specific binding function while keeping their multimers, and the lost function can be binding to platelet glycoprotein Ib or binding to collagen. This patient’s low ristocetin cofactor activity beside a normal ratio of antigen to collagen binding places his lesion on the platelet side. VWF is made in adequate amount and assembled into a full range of multimers, yet it binds glycoprotein Ib poorly. Sequencing of exon 28, which encodes the A1 domain that contacts glycoprotein Ib, confirms a pathogenic variant and settles the diagnosis.

It is worth naming what makes 2M distinctive. It localizes the defect to a single binding surface and leaves the rest of the molecule intact. In this patient the quantity is normal, multimer assembly is normal, factor VIII carriage is normal, and collagen binding is normal, so only the platelet contact fails. That also explains the laboratory picture that first looked contradictory.

Reveal: the desmopressin challenge

A desmopressin trial is performed. Antigen and factor VIII climb steeply. Activity rises only a little.

Why does desmopressin raise this patient’s VWF antigen substantially while his VWF activity rises only modestly?

Explanation

Desmopressin empties the endothelial storage pool. In type 2M, that pool holds the same variant protein the patient makes all the time. So antigen and factor VIII surge while activity lags behind, and the ratio of activity to antigen does not normalize. It may even worsen at peak antigen, because you are releasing a larger quantity of a protein that still cannot do its one job. The response is genuine but blunted and unreliable. The practical consequence follows from the mechanism, not from a dosing table: desmopressin may serve as an emergency stopgap where concentrate is unavailable, but for a major hemostatic challenge, replacement with a VWF containing product is the sounder choice.

Reveal: genetics and inheritance

Exon 28 sequencing confirms a pathogenic A1 domain variant, and type 2M is inherited in an autosomal dominant pattern. Yet the family history is unremarkable, and both of the patient’s children later test negative.

A dominant, genetically confirmed disorder can still hide. Expressivity varies, and a mild qualitative defect may never surface across decades of ordinary life, including uneventful dental extractions, until an operation removes the usual margins of safety. His disease did not announce itself until the surgical challenge was large enough and the field vascular enough. The bleed did not reveal a new disease. It revealed the threshold at which an old one finally mattered.

Reveal: management over time

The mechanism, once named, steers every later decision.

For a major operation years afterward, a partial nephrectomy, therapy is a plasma product supplying both VWF and factor VIII, loaded toward a ristocetin cofactor goal near 100 percent, with tranexamic acid as a mucosal adjunct. Monitored trough levels run high and doses are held. Even in a qualitative disorder, replacement can overshoot, so levels guide therapy rather than a fixed schedule.

For endoscopic and colonic procedures later still, therapy shifts to recombinant VWF. Because his factor VIII is preserved, a pure VWF product restores the missing function without loading additional factor VIII the patient does not need. The choice matches the product to the specific deficit rather than treating “VWD” as a single generic entity.

Teaching Points

• Type 2M localizes the defect to one binding surface. The lost function can be platelet binding, through the A1 domain as in this patient, or collagen binding, through the A3 domain. Either way the multimers are preserved, which is what separates 2M from 2A. A collagen binding variant can slip through a standard panel. Ristocetin cofactor and glycoprotein Ib assays report platelet binding, so they read normal when the lesion sits in collagen binding, and a normal platelet based activity does not by itself exclude 2M. Catching that variant takes a collagen binding assay, which is why measuring it here localized the defect rather than merely confirming it.
• A collagen binding variant can slip through a standard panel. Ristocetin cofactor and glycoprotein Ib assays report platelet binding, so they read normal when the lesion sits in collagen binding, and a normal platelet based activity does not by itself exclude 2M. Catching that variant takes a collagen binding assay, which is why measuring it here localized the defect rather than merely confirming it.
• Multimers separate 2M from 2A. Both reduce the platelet dependent activity relative to antigen. Only 2A strips out the largest multimers, so the multimer study is the pivot on which the two diagnoses turn.
• Discordant assays are the diagnosis, not noise. A low ristocetin cofactor activity beside a normal multimer pattern is the signature of a binding defect, because the two tests report on different steps in the molecule’s function.
• Desmopressin raises the quantity of a qualitatively abnormal protein. Expect antigen and factor VIII to climb and activity to trail, and prefer replacement for major challenges.
• A dominant mutation can be phenotypically silent. Variable penetrance means a negative family history should not deter specific testing once a genuine bleed has occurred.

Why This Case Matters

Von Willebrand disease is usually taught as a deficiency, a missing quantity of a needed protein. Type 2M refuses that framing. The protein here is present, well made, and correctly sized, and the patient bleeds anyway, because a single molecular handshake fails. The case rewards a habit of mind the screening tests cannot supply. Read the antigen against the activity, then read the activity against the multimers, and let the pattern of what is preserved localize the defect as precisely as the pattern of what is lost. The diagnosis was not produced by any one test. It was reasoned, one intact function at a time, until only the platelet’s grip remained unaccounted for.