Thrombosis Research 134 (2014) 1316–1322

Contents lists available at ScienceDirect

Thrombosis Research journal homepage: www.elsevier.com/locate/thromres

Regular Article

A two-step approach (Enzyme-linked immunosorbent assay and confirmation assay) to detect antibodies against von Willebrand factor in patients with Acquired von Willebrand Syndrome Franca Franchi b,⁎, Eugenia Biguzzi a, Francesca Stufano a, Simona M. Siboni a, Luciano Baronciani a, Flora Peyvandi a,b a b

Angelo Bianchi Bonomi Hemophilia and Thrombosis Center, Fondazione IRCCS Ca' Granda, Ospedale Maggiore Policlinico, Italy Angelo Bianchi Bonomi Hemophilia and Thrombosis Center, Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Milan, Italy

a r t i c l e

i n f o

Article history: Received 5 February 2014 Received in revised form 26 August 2014 Accepted 20 September 2014 Available online 28 September 2014 Keywords: AVWS Autoantibodies ELISA Confirmation assay Specificity

a b s t r a c t Background: Acquired von Willebrand Syndrome is a rare bleeding disorder, which arises in individuals with no personal or family history of bleeding, associated with lymphoproliferative and myeloproliferative disorders or other diseases. Aim: To develop a two-step approach assay to detect autoantibodies against VWF and to verify their prevalence in AVWS. Methods: AVWS definition: negative personal or family history of bleeding diathesis, VWF below normal range and recent history of bleeding manifestations. Twenty-three consecutive patients affected by AVWS were enrolled. An ELISA assay (first step) using recombinant VWF protein immobilized on plates and sheep/goat polyclonal anti-human IgG or IgM labelled with peroxidase was developed. A group of 40 healthy subjects was tested to calculate the floating cut point value. A confirmation assay (with addition of purified VWF vs buffer) was performed (second step). Results: Twenty–one patients (93%) had an associated disease, two patients had idiopathic AVWS. Anti-VWF autoantibodies were detected in 9 patients (39%). Of these, eight (89%) had VWF:RCo levels b 10%, but none of them resulted positive using Bethesda assay (neutralizing antibodies). The confirmation test confirmed the positive results obtained with ELISA and resulted negative in those patients with negative results and in the controls. Conclusion: With the present two-step approach assay nine out of 23 (39%) patients affected with AVWS resulted positive for anti-VWF autoantibodies. This ELISA assay might be used as an additional confirmation tool in the diagnostic procedure in patients affected by AVWS or in the follow-up of congenital and acquired patients exposed to replacement therapy. © 2014 Elsevier Ltd. All rights reserved.

Introduction Acquired von Willebrand Syndrome (AVWS) is a rare bleeding disorder, characterized by late onset in individuals with no family or personal history of bleeding. Clinical symptoms of AVWS are not specific and, like in congenital von Willebrand Disease (VWD), include Abbreviations: ELISA, Enzyme-linked immunosorbent assay; AVWS, acquired von Willebrand Syndrome; FVIII:C, factor VIII coagulant activity; VWF, von Willebrand Factor; VWF:RCo, von Willebrand Factor Ristocetin Cofactor; VWF:Ag, von Willebrand Factor antigen; VWF:CB, von Willebrand Factor collagen binding; VWFpp, von Willebrand Factor propeptide; PBS, phosphate buffer saline; BSA, bovine serum albumin; SD, Standard Deviation; OD, Optical Density; RT, room temperature; NF, normalization factor; MGUS, monoclonal gammopathy of undetermined significance; CLL, chronic lymphocytic leukemia; ET, essential thrombocytemia. ⁎ Corresponding author at: Angelo Bianchi Bonomi Hemophilia and Thrombosis Center, Fondazione IRCCS Ca' Granda, Ospedale Maggiore Policlinico, Via Pace 9, 20122 Milano, Italy. Tel.: +39 02 55035464. E-mail address: [email protected] (F. Franchi).

http://dx.doi.org/10.1016/j.thromres.2014.09.020 0049-3848/© 2014 Elsevier Ltd. All rights reserved.

spontaneous skin and mucosal bleeding as ecchymosis, epistaxis, gengivorrhagia, menorrhagia, gastrointestinal haemorrhage and postsurgical bleeding. AVWS usually occurs in association with a variety of underlying disorders, in particular with clonal haematoproliferative diseases, solid tumours, immunological disorders, hypothyroidism and cardiovascular diseases [1,2]. Nevertheless, some cases are considered idiopathic, when an underlying disorder cannot be identified [3]. Except in patients with hypothyroidism with a decreased synthesis or release [4], data from the literature suggest that von Willebrand Factor (VWF) is synthesized normally in most AVWS patients and rapidly removed from plasma through different pathogenic mechanisms, leading to decreased circulating levels of VWF and, eventually, to bleeding symptoms. Up to now three main pathogenic mechanisms have been reported [5]: 1) specific circulating antibodies directed against functional or non-functional domains of VWF; 2) non specific antibodies forming complexes with VWF which are rapidly cleared from the circulation; 3) selective absorption of large and intermediate VWF

F. Franchi et al. / Thrombosis Research 134 (2014) 1316–1322

multimers to malignant cells or absorption to activated platelets (as in thrombocytemia or in valve stenosis due to high shear rates). However, none of the suggested pathogenic mechanisms are disease-specific and the same mechanism may be responsible for AVWS in different diseases. Moreover, different laboratory patterns can be found. Unlike acquired haemophilia, which is always characterized by neutralizing anti-FVIII autoantibodies, in AVWS neutralizing autoantibodies against VWF are rarely identified by mixing tests which detect the VWF functional activity. These tests are time-consuming, might fail to detect low-titre (but clinically important) neutralizing antibodies and cannot detect non-neutralizing antibodies. Therefore the mixing VWF:RCo studies are rarely useful in the clinical practice. In the nineties, a competitive ELISA was set up to reveal anti-VWF antibodies, but this method was not further used in our knowledge [6]. Recently, Siaka and Tiede described two ELISA assays, using purified VWF bound to the plates, to detect anti-VWF autoantibodies [7,8] independently of their neutralizing activity. The presence of anti-VWF autoantibodies, together with a late onset bleeding diathesis, negative bleeding family history and laboratory findings mimicking congenital VWD and the presence of a disease usually associated to AVWS, could enable the investigator to make a correct diagnosis of AVWS. Besides, the follow-up of levels of VWF and anti-VWF antibodies could help the clinicians to monitor the therapy efficacy. The aims of the present study were 1) to set up a new ELISA assay to detect anti-VWF autoantibodies, followed by a confirmation assay able to detect false positive samples and 2) to evaluate the prevalence of such autoantibodies in patients affected by AVWS. Patients and methods Study design Retrospective observational study including all patients affected by AVWS, consecutively diagnosed at the A. Bianchi Bonomi Haemophilia and Thrombosis Center of Milan between January 2006 and December 2012. Medical history and haemorrhagic manifestations were collected by the clinicians during the first and follow up visits. Inclusion criteria were: decrease of VWF:Ag and VWF:RCo associated with recent history of bleeding diathesis; previously negative personal haemorrhagic history and negative family haemorrhagic history, according to ISTH recommendation [5]. All patients were evaluated before any treatment. At the time of diagnosis blood samples were obtained from all patients and factor VIII coagulant activity (FVIII:C), von Willebrand factor antigen (VWF:Ag), von Willebrand Factor Ristocetin Cofactor (VWF:RCo), von Willebrand Factor Collagen Binding (VWF:CB) and von Willebrand Factor Propeptide (VWFpp). Anti-VWF neutralizing autoantibodies were measured using VWF:RCo assay in mixing experiment with normal plasma. Anti-VWF IgG and IgM autoantibodies were searched using a new two-step ELISA assay. All individuals gave their informed consent to the study, which was approved by the Institutional Review Board of the Ospedale Maggiore Policlinico of Milan. Routine laboratory tests FVIII:C, VWF:Ag, VWF:RCo and VWF:CB were performed as previously described [9,10], VWF propeptide was evaluated with a commercially available ELISA kit (Sanquin, Amsterdam, the Netherlands). A ratio between VWFpropeptide and VWF antigen (VWFpp/VWF:Ag) was calculated. Table 1 shows a flowchart of the methodological procedure that has been followed, in the pre-analytical and analytical phase. Anti-VWF antibodies: ELISA (first step) A direct-binding non-competitive direct ELISA was developed. The VWF multimeric profile may influence the efficiency of binding to anti-

1317

VWF antibodies. Three different VWF preparations (two recombinant and one plasma-derived preparations) were tested for coating during the setting of our procedure. Recombinant VWF (kindly provided by dr Scheiflinger, Baxter, Wien, Austria) was chosen since it gave the best dose-response curve using a plasma which was previously found strongly positive for neutralizing anti-VWF antibodies with the Bethesda assay. Briefly, 10 μg/ml rVWF in phosphate buffered saline (PBS pH 7.4) were immobilized to Maxisorp Immuno plates (Nunc, Roskilde, Denmark) at 4 °C overnight. These plates were chosen based on the results previously obtained by Siaka et al. [7]. After washing with PBS 0.1% Tween 20% (Sigma Aldrich, St Louis, USA), the wells were blocked using 2.5% bovine serum albumine (BSA fraction V, Sigma) in PBS. After washing, patients’ plasma samples, positive and negative controls diluted 1/20 using PBS 1% BSA were added to the plate and incubated for 2 hours at RT. IgG or IgM autoantibodies anti-VWF were then detected using sheep antihuman IgG (GE Healthcare, Little Chalfont, UK) or goat anti-human IgM (Sigma Aldrich,) horse-radish peroxidase labelled, diluted 1/3000 in PBS (BSA 1%) and o-phenylenediamine (Sigma Aldrich) substrate. The efficiency of rVWF coating on Maxisorp plates was evaluated using a plasma strongly positive for neutralizing anti-VWF antibodies. Peroxidase-labelled anti-human IgG/IgM antibodies were used for tagging. In the described ELISA, the reagents’ concentrations and the 1/20 plasma sample dilution were selected in the pre-analytical phase, using normal plasma samples, as the best compromise between optical density readings and minimal plasma sample dilution (in order not to reduce the sensitivity of the assay). Supplementary Fig. 1 shows the dose-response curves of an anti-VWF antibody positive plasma sample with increasing concentrations of coating rVWF. According to the procedure of the screening cut point [11] for discriminating positive from negative samples, the substraction of the unspecific binding of the samples (uncoated wells) was not performed. The screening cut point specifies the signal (OD) of the ELISA, above which a test sample is defined “potentially positive” for the presence of antibodies to VWF. Due to uncontrolled variability in colour developing between plates caused by multiple factors in the ELISA assay, a floating (or plate-specific) cut point was preferred in the present study to a fixed cut point [11]. In the pre-analytical phase plasmas from 40 normal individuals (20 males and 20 females) were tested to detect the anti-VWF IgG or IgM autoantibodies in a single ELISA plate. Their absorbance (OD) mean and standard deviation (SD) were determined. In the same plate a positive control (plasma of a VWD patient with a high titre of anti-VWF neutralizing antibodies detected with Bethesda assay) and a negative control were included. After exclusion of two far outliers (OD more than 95° centile), normal plasma samples were pooled, aliquoted and frozen. This pool was defined Negative Pool. All normal plasmas with OD less than 95° centile were tested again in duplicate, together with the Negative Pool repeated 8 times in the same plate for IgG or IgM anti-VWF autoantibodies. The cut point was calculated as the mean OD of these normal samples + 1.645 x SD; a normalization factor (NF) was calculated as cut point/mean OD of the Negative Pool. In the analytical phase, the Negative Pool sample (repeated 8 times) was always included when patients’ samples were screened. The floating cut point (that could be slightly different in the different plates) was calculated as mean OD of the Negative Pool samples (relative to that plate) multiplied for the NF (determined in the pre-analytical phase). A negative and a positive control for IgG and IgM autoantibodies were run in each plate. Patients’ samples and negative control were diluted 1:20 and tested in triplicate. The positive control was diluted 1:100 for IgG and 1:500 for IgM. Samples with absorbance greater than the floating cut point were considered “potentially positive” for anti-VWF antibodies. Subsequently, confirmed-positive samples were tested at serial dilutions to assess the antibody titre. Analytical variability was determined with intra and inter-assay variance in the pre-analytical phase. The intra-assay coefficient of variation was 2.9% and 3.3% for IgG and IgM respectively. The inter-assay coefficient of variation was 8% and 6% (negative control), and 18% and 14% (positive control) for IgG and IgM,

1318

F. Franchi et al. / Thrombosis Research 134 (2014) 1316–1322 Table 1 Flow chart of the methodological procedure.

respectively. The inter-assay coefficient of variation was calculated (ratios between the negative control OD and the Negative Pool OD mean obtained in the different plates) on 10 working days. The mean OD for 7 independent plates for IgG for and IgM was 0.219 (range 0.1740.255) and 0.087 (range 0.069-0.136), respectively. Confirmation assay (second step)

Table 2 Clinical characteristics of AVWS patients.

100

% of patients with spontaneous bleeding

80 60

% of patients with provoked bleeding

40

Clinical characteristics of AVWS patients Number of patients Age at diagnosis (years), median (IQ range) Sex (male/female) Monoclonal gammopathy − of undetermined significance (MGUS) − Waldestrom’s disease − Chronic lymphocytic leukemia Other diseases Idiopathic

120

%

Positive samples in the ELISA assay (OD above the floating cut-point) were also tested with the confirmation assay [12], based on the binding of exogenous VWF to anti-VWF autoantibodies. This second step was mandatory to exclude the 5% false positive results obtained with the ELISA assay in the first step. In the pre-analytical phase, recombinant and plasma purified VWF preparations were tested with similar results (Supplementary Table 1). Plasma purified VWF (Wilfactin, LFB, Les Ulis, France) was chosen for feasibility reasons. The concentration of exogenous VWF and the time of incubation were determined as follows: exogenous VWF was added to a positive sample (for IgG and IgM antibodies) to obtain a final concentration (f.c.) of 5, 10, 20, 50 U/ml. In a parallel tube an equal volume of 1% BSA in PBS pH 7.4 was added to the same positive sample. All tubes were incubated for 2 hours at 37 °C. The residual

anti-VWF autoantibodies (corresponding to the residual OD), were measured by ELISA assay and expressed as percentage of the sample to which PBS was added. The results were 59%, 52%, 40%, 31% for IgG and 66%, 56%, 47%, 42% for IgM, respectively. The final concentration of 20 U/ml was chosen for the confirmation assay of potentially positive samples. To choose the incubation time, a positive sample to which 20 U/ml of exogenous purified VWF (or 1% BSA in PBS) were incubated at 37 °C for 30 minutes, 1 hour and 2 hours. Similar percentages of residual anti-VWF OD (around 40%) were obtained. Therefore, one hour incubation was chosen arbitrary for feasibility reason (data not shown). Similarly, in the analytical phase, two aliquots of plasma samples identified as positive in the ELISA assay, were diluted according to their titre, so that the OD fell in the linear part of the dose-response curve. Then one aliquot was incubated with 20U/ml of plasma derived VWF (Wilfactin), the other with buffer. An ELISA assay was performed and the residual OD (of the aliquot to which VWF was added) was

20 23 55 (45-70) 12/11 13 (57%) 11 1 1 8 (35%) 2 (8%)

0 all patients (n=23)

autoAb yes (n=9)

autoAb no (n=14)

Fig. 1. Percentage of patients with spontaneous or provoked bleeding manifestation at diagnosis: A) all patients, B) patients with anti-VWF autoantibodies, C) patients without anti -VWF autoantibodies.

Table 3 Demographic data, clinical characteristics and laboratory data of AVWS patients (n = 23). sex

age at diagnosis

Blood group

FVIII: C%

VWF: Ag %

VWF:RCo % (value at diagnosis)

VWF: CB %

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

M M F F M M F F M F M M M M F F F F F M F M M

55 74 55 45 54 69 66 38 53 56 71 67 74 70 46 87 50 22 38 70 71 70 41

A+ B+ AB+ A+

10 33 13 29 36 22 60 74 32 25 59 43 38 19 51 17 52 53 54 63 48 69 40

3 35 9 22 31 27 42 70 23 6 7 45 20 12 32 12 56 51 53 52 29 74 30

b6 7 b6 b6 8 b6 19 50 (b6) 6 b6 b6 35 b6 b6 17 b6 52 (36) 40 (26) 47 (36) 31 b6 46 21

8 23 b1

4.7 1.4 9.5 3.5

7 19

4.0 2.9 1.1 4.0 27 4.8 1.9 4.5 17 1.8 16 1.3 1.5 1.1 1.6 3.0 2.6 6.2

A+ A+ AB+ A+ A+ 0+ 0+ 0+ AB+ 0+ 0+ 0+

0+ 0-

1 3 55 b1 4

48 32 52 33 6 16 25

VWFpp/Ag ratio

Spontaneous bleeding

Provoked bleeding

Underlying disease

Monoclonal component

Monoclonal component mg/dL

Anti-VWF autoAb IgG (titre)

Anti-VWF autoAb IgM (titre)

neutralizing antibodies

no yes no yes yes yes yes yes yes yes yes yes yes yes yes yes no yes yes yes yes yes no

yes yes yes no yes no no no no no yes no yes no yes no yes no no no no yes yes

MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS Waldestrom CLL Crohn CV CV ET ET + Ehler-Danlos ET ET neoplasia idiopathic idiopathic

IgG κ IgG λ + Ig oligoclonal IgG λ IgM κ + IgG λ IgM κ IgM κ IgG κ IgG κ IgM κ IgG λ IgG κ IgM λ IgM κ + IgM λ

0.41 0.36 0.2 0.84 2.1 0.78 1.1

absent absent absent 1:8000 1:10000 1:200 absent absent 1:1600 absent absent absent 1:12000 1:200 absent 1:80 absent absent absent absent 1:400 1:20 absent

absent absent absent 1:80000 1:25000 1:1600 absent absent 1:12800 absent absent absent 1:50000 absent absent 1:100 absent absent absent absent 1:1600 1:100 absent

no

0.27 0.66 4.4 0.5 + 0.5

no NA no

no no no no no no

F. Franchi et al. / Thrombosis Research 134 (2014) 1316–1322



no

Normal range (blood group = 0*; blood group = non-0**): Factor VIII:C: 60%-150%; von Willebrand Factor Antigen (VWF:Ag): 40%-169%* and 55%-165%**; von Willebrand Factor Ristocetin-Cofactor activity (VWF:RCo):41%-160%* and 53%-188%**; von Willebrand Factor Collagen-Binding activity (VWF:CB): 45%-170%* and 53%-188%**; VWF:propeptide/VWF:Ag: 0.5-1.5.

1319

1320

F. Franchi et al. / Thrombosis Research 134 (2014) 1316–1322

A (IgG)

B (IgM)

1,2

1,2

1

1 0,8 OD

OD

0,8 0,6

0,6

0,4

0,4

0,2

0,2 0

0 4

5

6

13

14

21

16

22

9

neg pts

NC

21

4

6

13

5

9

22

16

neg NC pts

Fig. 2. Confirmation assay: anti-VWF autoantibodies positive patients (numbers presented in Table 3), negative patients (mean value of 14 patients) and normal controls (mean value of 20 individuals) after incubation of plasma with 20U/ml f.c of exogenous VWF (black bars), or buffer (grey bars). A) IgG positive patients (diluted in the order 1:400, 1:800, 1:100, 1:3000, 1:40, 1:200, 1:20, 1:20, 1:400), IgG negative patients and normal controls (diluted 1:20) B) IgM positive patients (diluted in the order 1:200, 1:500, 1:100, 1:10000, 1:640, 1:3000, 1:20, 1:20), IgM negative patients and normal controls (diluted 1:20). The horizontal line indicates the mean value of the floating cut point (7 independent runs).

measured and expressed as percentage relative to the aliquot to which buffer was added. The same experiment was performed in the plasma samples identified as negative in the ELISA assay (OD below the floating cut-point) and in 20 healthy controls to verify specificity of our assay for anti-VWF antibodies.

Table 3 reports the characteristics of the patients and laboratory findings. Nine out of 23 patients with AVWS (39%) were positive for anti-VWF antibodies. Of these, eight patients (89%) had both antiVWF IgG and IgM autoantibodies, one patient was positive only for IgG autoantibodies. Of the patients with autoantibodies, eight (89%) had VWF:RCo levels b10% (compared to 4/14, 29% in patients without autoantibodies). Positivity to anti-VWF autoantibodies was higher in patients with IgM monoclonal gammopathy (5/6, 83%), compared to patients with IgG monoclonal gammopathy (1/8, 13%). Table 3 shows details of anti-VWF autoantibodies in relation to underlying disease, monoclonal component and VWF levels. Fig. 2 shows the results of the confirmation assay. The OD values of IgG and IgM positive patients appeared reduced following the addition of purified plasma derived VWF, compared to the same sample with only buffer added, whereas no significant variance was shown in plasma samples of normal subjects or patients with AVWS negative with the ELISA assay (OD values below the floating cut-point). The residual OD detected in patients with anti-VWF autoantibodies, evaluated with the confirmation assay, ranged from 40% to 64% and and from 18% to 52% for IgG and IgM autoantibodies, respectively (Fig. 3). Fig. 4 shows the distribution of FVIII:C, VWF:Ag, VWF:RCo, VWF:CB in all patients with AVWS, in relation to the presence of anti-VWF autoantibodies. Positive patients had lower median values for all tests than negative patients: FVIII:C (median 32% vs 51%), VWF:Ag (median 23% vs 38%), VWF:RCo (median 3% vs 20%), VWF:CB (median 5% vs 25%). There was a significant difference between the two groups of patients with and without autoantibodies regarding VWF:RCo (P b 0.05) and VWF:CB (P b 0.01) values. FVIII:C and VWF:Ag levels were lower in the positive patients compared to the negative ones, although the differences were not statistically significant. VWFpropeptide/VWF:Ag ratio levels were increased in 8/9 (89%) patients with anti-VWF autoantibodies and in 8/14 (57%) patients without antibodies (Fig. 5). The VWFpropeptide/VWF:Ag median value ratio

Neutralizing anti-VWF antibodies The neutralizing anti-VWF autoantibodies were searched by mixing experiment (Bethesda assay), as previously described (9), in all patients with VWF:RCo b6%. Statistical analysis was performed using SPSS version 20.0 (SPSS Inc., Chicago, IL, USA). Continuous variables were expressed as median and Interquartile value range. To compare continuous variables between groups, the P-value of t-test was used. When the VWF:RCo measured levels were less than 6%, the arbitrary value of 3% was used to perform statistical analyses, the value of 1% in the case of VWF:CB b 1%. Results Twenty-three patients out of 26, diagnosed as AVWS, were available. Demographic data and clinical characteristics are shown in Table 2. At the time of diagnosis, all patients presented with bleeding diathesis either spontaneous (19/23 patients, 83%) or provoked (subsequent a challenging maneuver, as dental extraction or surgery, 11/23 patients, 48%). Spontaneous bleeding manifestations were present in all patients positive for anti-VWF autoantibodies (9/9) and in 71% of the negative patients (10/14), whereas provoked bleeding was present in 3/9 (33%) and in 8/14 (57%), respectively (Fig. 1). In the past, 20/23 patients (87%) had underwent surgery and 9/23 (39%) had had tooth extraction without any bleeding.

A

B 120

120

100

100 80 % OD decrease

60

% OD residual

%

% OD

80

%OD decrease

60

40

40

20

20

% OD residual

0

0 4

5

6

13

14

21

16

22

9

21

4

6

13

5

9

22

16

Fig. 3. Confirmation assay: dark dotted histograms show the % residual OD in anti-VWF autoantibodies positive patients after addition of exogenous VWF (20 U/ml f.c.); light dashed histograms show the % of OD decrease. A) IgG positive patients; B) IgM positive patients.

F. Franchi et al. / Thrombosis Research 134 (2014) 1316–1322

FVIII:C

VWF:Ag

1321

VWF:RCo

VWF:CB

neg

neg

80 70 60

%

50 40 30 20 10 0

neg

pos

neg

pos

pos

pos

Fig. 4. FVIII:C, VWF:Ag, VWF:RCo, VWF:CB. For each test, the left column refers to patients without anti-VWF autoantibodies, the right one to patients with IgG and/or IgM anti-VWF autoantibodies. The median value is indicated by a solid line. Statistically significant differences were found for VWF:RCo (P b 0.05) and VWF:CB (P b 0.01).

in patients with autoantibodies was higher than in patients without, 4 and 1.85 respectively. All patients with VWF:RCo b6% were tested also for neutralizing anti-VWF antibodies using mixing experiments (Bethesda assay) and all resulted negative. Discussion AVWS is a rare, complex disorder with a multifactorial etiology, difficult to diagnose and challenging to treat successfully. The diagnosis of AVWS requires careful correlation of clinical information with the results of laboratory tests. At difference with acquired hemophilia where neutralizing antibodies are always detectable with the Bethesda neutralization assay, the presence of neutralizing antibodies against FVIII/VWF can rarely be detected in AVWS patients by means of mixing experiments using VWF: RCo, VWF:Ag or VWF:CB assays [13]. Indeed, in the present study none of the 10 patients with VWF:RCo b 6% resulted positive for neutralizing autoantibodies. This could be explained by lack of sensitivity of the mixing studies or by the presence of non-neutralizing autoantibodies, which could increase VWF clearance from the circulation without inhibiting its function. This is also confirmed by a higher median ratio of VWFpropeptide/VWF:Ag in patients with anti-VWF autoantibodies compared to patients without (4 vs 1.85). Up to now, two enzyme-linked immunosorbent assays for the detection of anti-VWF autoantibodies have been described [7,8]. In the present study, a new two-step ELISA assay was developed which differs propeptide/Antigen 30

ratio pp/Ag

25 20 15 10 5 0

neg

pos

Fig. 5. VWFpropeptide/VWF:Ag ratio median value: the left column refers to patients without anti-VWF autoantibodies, the right one to patients with IgG and/or IgM antiVWF autoantibodies. The median value is indicated by a solid line.

from the previous ones for some characteristics: 1) use of human recombinant VWF (kindly gifted by Dr. Sheiflinger, Baxter, Wien, Austria), that contains high and ultra-large multimers [14] at difference with plasma-derived VWF concentrate [15]; in addition, recombinant VWF lacks ABO blood group antigens that may cause false positive results in the ELISA because of isoagglutinins; 2) determination of a cutpoint value assessed with a large number of healthy individuals (n = 40) and determination of a normalization factor that allowed the calculation of a floating cut-point; 3) the titration of the positive samples: 4) a second step confirmation assay developed to verify the specificity of the positive results obtained in the ELISA assay (first step). In this study, 23 patients with AVWS diagnosis and different underlying diseases were evaluated with the new two-step approach (ELISA assay and confirmation assay) and nine (39%) were found positive for anti-VWF autoantibodies; six out of nine patients (67%) with VWF:RCo b 6% were tested and found negative with Bethesda assay. Previous studies using ELISA assays by Siaka and Tiede [7,8] found anti-VWF autoantibodies in 8/10 (80%) and 4/27 (15%) of patients respectively. This discrepancy can be due to the limited number of studied patients due to the rarity of the disease or to the heterogeneity of patients affected by AVWS and finally to the different sensitivity of the ELISA assays. It is noteworthy that in the present study spontaneous bleeding at diagnosis was present in 100% of AVWS patients with autoantibodies compared to 57% of patients without autoantibodies. Provoked bleeding (after challenging maneuvers, such as surgery or dental extractions) was present in 33% and 50% of patients with and without autoantibodies. This is in accordance with the study of Tiede [8] who also found that the three patients with anti-VWF antibodies had major bleeding symptoms compared to those without, who did not. In this study, 5/6 (83%) patients with IgM monoclonal gammopathy were found positive for anti-VWF autoantibodies. On the contrary, only 1/8 (13%) of patient with IgG monoclonal gammopathy resulted positive. A lower prevalence of anti-VWF autoantibodies in patients with IgG gammopathy might be due to a lower sensitivity of our assay in the presence of IgG gammopathy or to an alternative mechanism of action causing low levels of VWF (i.e. presence of immunocomplexes). The specificity of the new ELISA assay, even if not formally evaluable for lack of a golden standard, was confirmed by the negative results obtained in 20 normal individuals and by the confirmation assay, able to separate false positive samples from truly positive samples. Interestingly, in the present group of AVWS patients, lower levels of FVIII:C, VWF:Ag, VWF:RCo and VWF:CB were found in patients with anti-VWF autoantibodies compared to patients without antibodies, although a statistically significant difference was reached only for VWF: RCo and VWF:CB levels. In accordance with the study of van Genderen

1322

F. Franchi et al. / Thrombosis Research 134 (2014) 1316–1322

et al. [16], increased median levels of VWFpropeptide/VWF:Ag ratio were found, more evident in patients with positive autoantibodies compared to negative patients. This could reflect a compensatory increase in VWF synthesis or increased perturbation of the endothelium, for instance as a consequence of circulating immune-complexes formation. The detection of anti-VWF autoantibodies (using ELISA associated to the confirmation assay) might improve the diagnosis of AVWS and allow us to perform a better follow-up study. In addition, this approach could help us to predict the bleeding severity of patients affected with AVWS. The drawbacks of the present study are the relatively small number of patients and their heterogeneity relatively to VWF plasma levels and underlying disease. In conclusion, this study describes a new combined approach (ELISA assay followed by the confirmation assay) to detect anti-VWF autoantibodies. In the present series of patients affected by AVWS, 39% resulted positive for anti-VWF autoantibodies. The new ELISA assay seems to detect IgG antibodies better than IgM antibodies. This assay might be used as an additional confirmation tool of the diagnosis of AVWS, which is in most cases difficult and always involves both clinical and laboratory parameters. Moreover, the presence of anti-VWF autoantibodies may be useful for the clinicians when choosing the best therapeutic approach and in the follow-up of the treatment. The present assay needs to be validated in future studies and could be also tested to search for alloantibodies in the framework of severe congenital VWD, where mixing studies cannot always reveal the presence of antibodies that can interfere with concentrate half-life [17]. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.thromres.2014.09.020. Conflict of interest statement Franca Franchi, Eugenia Biguzzi, Francesca Stufano, Simona M Siboni, Luciano Baronciani declare that they have no conflict of interest, Flora Peyvandi received speaker honoraria by Novo Nordisk, CSL Behring, Bayer and Baxter. Acknowledgements The author would like to thank Dr F. Scheiflinger from Baxter for kindly supplying the recombinant von Willebrand Factor and for his constructive comments, dr. M.T Canciani for helpful discussion and dr G.Cozzi for technical support. Authors contributions SMS and EB recruited patients; FF, EB, FS designed the study, analyzed the results and wrote the paper. FP and LB designed the study, took part in discussion and revised the paper.

References [1] Franchini M, Lippi G. Acquired von Willebrand Syndrome: an update. Am J Hematol 2007;82:368–75. [2] Federici AB. Acquired von Willebrand syndrome: an underdiagnosed and misdiagnosed bleeding complication in patients with lymphoproliferative disorders. Semin Hematol 2006;43:S48–58. [3] Michiels JJ, Budde U, van der Planken M, van Vliet HH, Schroyens W, Berneman Z. Acquired von Willebrand syndromes: clinical features, aetiology, pathophysiology, classification and management. Best Pract Res 2001;14:401–36. [4] Manfredi E, van Zaane B, Gerdes VE, Brandjes DP. SquizzatoA Hypothyroidism and acquired von Willebrand's syndrome: a systematic review. Haemophilia 2008;14: 423–33. [5] Federici AB, Rand JH, Bucciarelli P, Budde U, van Genderen PJ, Mohri H, et al. Acquired von Willebrand Syndrome: data from an International Registry. Thromb Haemost 2000;84:345–9. [6] Stewart MW, Etches WS, Shaw AR, Gordon PA. vWf inhibitor detection by competitive ELISA. J Immunol Methods 1997;200:113–9. [7] Siaka C, Rugeri L, Caron C, Goudemand J. A new ELISA assay for diagnosis of acquired von Willebrand Syndrome. Haemophilia 2003;9:303–8. [8] Tiede A, Priesack J, Werwitzke S, Bohlmann K, Oortwijn B, Lenting P, et al. Diagnostic workup of patients with acquired von Willebrand syndrome: a retrospective singlecentre cohort study. J Thromb Haemost 2008;6:569–76. [9] Mannucci PM, Lombardi R, Bader R, Horellou MH, Finazzi G, Besana C, et al. Studies of the pathophysiology of acquired von Willebrand’s disease in seven patients with lymphoproliferative disorders or benign monoclonal gammopathies. Blood 1984;64: 614–21. [10] Federici AB, Canciani MT, Forza I, Mannucci PM, Marchese P, Ware J, et al. A sensitive ristocetin co-factor activity assay with recombinant glycoprotein Ibalpha for the diagnosis of patients with low von Willebrand factor levels. Haematologica 2004;89: 77–85. [11] Shankar G, Devanarayan V, Amaravadi L, Barrett YC, Bowsher R, Finco-Kent D, et al. Recommendations for the validation of immunoassays used for detection of host antibodies against biotechnology products. J Pharm Biomed Anal 2008;48:1267–81. [12] Jaki T, Lawo J-P, Wolfsegger MJ, Allacher P, Horling F. A comparison of methods for classifying samples as truly specific with confirmatory immunoassays. J Pharm Biomed Anal 2014;88:27–35. [13] Tiede A, Rand JH, Budde U, Ganser A, Federici AB. How I treat the acquired von Willebrand syndrome. Blood 2011;117:6777–85. [14] Turecek PL, Mitterer A, Matthiessen HP, Gritsch H, Varadi K, Siekmann J, et al. Development of a plasma- and albumin-free recombinant von Willebrand factor. Hamostaseologie 2009;29(Suppl. 1):S32–8. [15] Mazurier C, Poulle M, Samor B, Hilbert L, Chtourou S. In vitro study of a triplesecured von Willebrand factor concentrate. Vox Sang 2004;86(2):100–4. [16] Van Genderen PJ, Boertjes RC, van Mourik JA. Quantitative Analysis of von Willebrand Factor and Its propeptide in plasma in Acquired von Willebrand. Thromb Haemost 1998;80:495–8. [17] James PD, Lillicrap D, Mannucci PM. Alloantibodies in von Willebrand disease. Blood 2013;122:636–40.

A two-step approach (Enzyme-linked immunosorbent assay and confirmation assay) to detect antibodies against von Willebrand factor in patients with Acquired von Willebrand Syndrome.

Acquired von Willebrand Syndrome is a rare bleeding disorder, which arises in individuals with no personal or family history of bleeding, associated w...
2MB Sizes 0 Downloads 8 Views