Caplacizumab Resistance in Immune Thrombotic Thrombocytopenic Purpura (TTP) Mediated By VWF A1 Domain Missense Mutation

Introduction: Caplacizumab has been shown to reduce the time to platelet normalisation, reduce TTP exacerbations and prevent refractory acute disease. Exacerbations have been rarely documented, but this is the first case associated with caplacizumab ‘resistance’.

A 65 year old woman with no history of a bleeding diathesis was diagnosed with iTTP and managed with urgent plasma exchange (PEX), steroids, caplacizumab & rituximab. Platelet count normalised by Day 5 but fell to 17 x10⁹/L on D14 despite good adherence to caplacizumab.

VWF activity (VWF:Ac) on D14 was normal at 0.8 IU/ml (NR 0.50 - 1.60) with VWF antigen (VWF:Ag) 2.29 IU/ml (NR 0.50-1.87) and VWF Ristocetin cofactor assay 39 IU/dl (NR 45-175). We retrospectively investigated serial VWF activity on frozen samples. VWF:Ag on D1 at initial presentation pre PEX was elevated at 2.13 IU/ml with VWF:Ac 1.91 IU/ml. After initiation of caplacizumab, VWF:Ag remained at 2.2 IU/ml with VWF:Ac reduced but largely in normal range (0.47-0.84). There was no evidence of more profound initial suppression of VWF activity with subsequent loss of caplacizumab effect, which might have suggested drug-neutralising antibody.

A microfluidic assay was used to analyse platelet binding to VWF under flow and high shear stress representing the microvasculature. In iTTP patients receiving caplacizumab, there is very low platelet coverage on both collagen-coated and anti VWF-coated surfaces, compared to TTP patients not on caplacizumab. Median surface coverage for collagen coated channels in iTTP patients on caplacizumab was 4.5% (range 0.3-6.9%) and 0% for anti VWF-coated channels. In our patient, platelet surface coverage was 19.6% on collagen and 24% on anti-VWF.

Platelet light transmission aggregometry was normal, including absent agglutination to low dose ristocetin. CD42b (GPIb alpha) expression by flow cytometry was 99.98%. Sanger sequencing of exon 28 encoding VWF A1 domain showed patient to be heterozygous for 2 missense variants: c.3797C>T, p.(Pro1266Leu) and c.3835G>A, p.(Val1279Ile).

The crystal structure of the VWF A1 domain in complex with caplacizumab was visualised in PyMOL; AlphaFold 3 used to predict the structure of the mutated A1 domain and alignments made in PyMOL using least-squares superposition. The crystal structure showed interactions between the A1 domain and Caplacizumab including P1266, which lies at the complex interface, with buried surface area of 60.78 Ų suggesting the mutation may directly affect A1 domain binding to Caplacizumab.

The patient recovered well, re-achieving clinical remission on D20 after 7 more PEX and continued immunosuppression, with ADAMTS13 remission on D30. P1266L missense variant has been reported in association with Type 2B VWD (although with mild phenotype and normal platelet levels) with total maximum allele frequency (all populations) of 0.066%. P1266L has been reported to affect the A1-GP1b binding site, although not in the same manner as a classical Type 2B VWD variant.

Conclusion: We report the first case of caplacizumab ‘resistance’ in a patient with iTTP with a failure of suppression of in vivo VWF activity and persistence of VWF-mediated platelet capture in a flow-based in vitro assay. Genetic analysis revealed a missense mutation P1266L in exon 28 of VWF, which is associated with normal platelet counts but affects the A1-GP1b binding site. This case presents a unique scenario relating to a VWF mutation affecting the A1 domain and site of platelet binding. VWF activity levels should be checked when TTP exacerbation occurs despite daily caplacizumab therapy.