An international collaborative study to compare different von Willebrand factor glycoprotein Ib binding activity assays: the COMPASS-VWF study.

Essentials New VWF activity assays are increasingly used but information on their comparability is limited. This is an ISTH SSC-organized study (expert labs, 5 countries) to compare all available assays. VWF activity by six assays correlated well with each other. The new assays show improved characteristics - minor differences are noted. SUMMARY: Background Several new assays have become available to measure von Willebrand factor (VWF) activity. The new assays appear to have improved performance characteristics compared with the old reference standard, ristocetin cofactor activity (VWF:RCo), but information is limited about how they compare with VWF:RCo and each other. Methods The von Willebrand factor Subcommittee of the International Society for Thrombosis and Haemostasis (ISTH) Scientific and Standardization Committee (SSC) designed a collaborative study involving expert laboratories from several countries to compare available tests with each other and with VWF:RCo. Eight laboratories from five countries were provided with blinded samples from normal healthy individuals and well-characterized clinical cases. Laboratories measured VWF activity using all tests available to them; data from six laboratories, not affected by thawing during transportation, are included in this study. Results All tests correlated well with VWF:RCo activity (r-values ranged from 0.963 to 0.989). Slightly steeper regression lines for VWF:Ab and VWF:GPIbM were clinically insignificant. The new assays showed improved performance characteristics. Of the commercially available assays, the VWF:GPIbR using the AcuStar system was the most sensitive and could reliably detect VWF activity below 1 IU dL-1 . The lower limit of the measuring interval for the VWF:GPIbM and the VWF:GPIbR assays was in the 3-4 and 3-6 IU dL-1 range, respectively. Inter-laboratory variation was also improved for most new assays. Conclusion All VWF activity assays correlated well with each other and the VWF:RCo assay. The slight differences in characteristics found in the COMPASS-VWF study will assist the VWF community in interpreting and comparing activity results.

A comparative evaluation of a new fully automated assay for von Willebrand factor collagen binding activity to an established method.

INTRODUCTION: Laboratory diagnosis of von Willebrand disease (VWD) is made by the measurement of von Willebrand factor (VWF) protein level and its activities. Current VWF activity tests include ristocetin cofactor and collagen binding (VWF:CB) assays. AIM: We have undertaken an evaluation of a new fully automated VWF:CB assay relative to an established enzyme-linked immunosorbent assay (ELISA) method. METHODS: The two analytical systems operate with different detection principles: a chemiluminescent method performed on ACL AcuStar Analyzer (the former) and a colorimetric ELISA by Asserachrom Stago (the latter) (type III collagen from human placenta). The HemosIL AcuStar VWF:CB assay is a chemiluminescent 2-step immunoassay that uses magnetic particles coated with a type III collagen triple-helical peptide. VWF:CB levels were determined in 50 healthy subjects and 100 VWD patients (22 type 1, 73 type 2 and 5 type 3). RESULTS: Eleven VWD samples reported VWF:CB values below the lower detection limit of one or both methods. The new method showed a good correlation with the ELISA method (r > .9, mean bias 3.85 IU/dL) in both healthy and VWD samples. One of 150 samples gave inconsistent results using the two assays, leading to an uncertain diagnosis of VWD type 1 (ELISA method) or type 2 MCB (fully automated method). CONCLUSION: The new assay is rapid and simple to use, with its ready-to-use reagent cartridges. This VWF:CB assay, in addition to the measurement of VWF:Ag and VWF:RCo made on the same platform, gives additional information for the diagnosis of VWD in both nonspecialized and reference laboratories.

Molecular diagnosis of von Willebrand disease.

The role of molecular characterization in the diagnosis of von Willebrand disease (VWD) is not essential if the patients have been extensively investigated using phenotypic analysis. On the other hand, if some of these phenotype assays are not available, the identification of the mutation causing the disease could be crucial for an accurate diagnosis. Nevertheless, there are several reasons for performing molecular analysis in patients phenotypically well characterized, e.g. to identify the mutation causing VWD can be useful for patients and their family members when prenatal diagnosis is required (type 3 or severe type 2). In this manuscript, we report the techniques used for the molecular characterization of suspected VWD patients. We describe the use of online von Willebrand factor database and online single nucleotide variation databases, the former to verify whether a candidate mutation has been previously identified in other VWD patients and the latter to ascertain whether a putative mutation has been reported earlier in healthy individuals. We listed the available in silico analysis tools, to determine the predicted pathogenicity of a sequence variant and to establish its possible negative effect on the normal splicing process. We also report the strategy that can be used to identify VWD type 2 patients' mutations in subjects who have been fully characterized using the phenotype assays.

von Willebrand disease type 1 mutation p.Arg1379Cys and the variant p.Ala1377Val synergistically determine a 2M phenotype in four Italian patients.

INTRODUCTION: We characterized five patients affected with von Willebrand disease (VWD) carrying the p.Arg1379Cys mutation. One was diagnosed as VWD type 1 and four as type 2M. The 2M patients also have the variant p.Ala1377Val in cis with p.Arg1379Cys. AIM: To evaluate the role of p.Ala1377Val and p.Arg1379Cys von Willebrand factor (VWF) variants to explain patients' phenotype. METHODS: Conventional phenotype tests were used to evaluate patients' plasma and platelets. Direct sequence analysis of exon 28 was carried out. The allele frequency of p.Ala1377Val was evaluated using online database. pcDNA3.1-VWF-WT and mutant (A1377V, R1379C and A1377V-R1379C) expression vectors were transiently transfected in HEK293 cells. The capacity of WT and mutant recombinant (r)VWF (along with patients' plasma VWF) to bind glycoprotein Ibα (GpIbα) were evaluated, using two ELISA assays. One with a wild-type (WT) recombinant (r)GpIbα at increasing ristocetin concentrations (from 0 to 1.50 mg mL-1 ) and the other with a gain-of-function mutant rGpIbα (VWF:GPIbM). RESULTS: The substitution c.4130C>T (p.Ala1377Val) was reported as rare variant in online databases. At 0.25 mg mL-1 of ristocetin, WT, A1377V and R1379C showed 6, 7.5 and 12-fold increased binding to rGpIbα, respectively. A1377V-R1379C rVWF showed no increased binding to rGpIbα at the same ristocetin concentration and reached the highest binding, of only 3-fold increased, at 1.50 mg mL-1 of ristocetin. The VWF:GPIbM showed strongly reduced values for the A1377V-R1379C rVWF and the 2M patients' plasma. CONCLUSION: Our study showed that the presence of both p.Ala1377Val and p.Arg1379Cys mutations (synergistic effect) abolishes the binding of rVWF to rGpIbα, explaining patients' 2M phenotype.

Evaluation of an heterogeneous group of patients with von Willebrand disease using an assay alternative to ristocetin induced platelet agglutination.

BACKGROUND: Diagnosis of von Willebrand disease (VWD) type 2 usually relies on the discrepancy between the von Willebrand factor (VWF) ristocetin cofactor activity (VWF:RCo) and VWF antigen (VWF:Ag). Type 2B patients can be discriminated from other qualitative VWD variants by using ristocetin-induced platelet agglutination (RIPA) test. The major limitation of RIPA is the requirement of fresh blood sample. OBJECTIVES: In this study, we evaluated the VWF gain-of-function mutant GPIb binding (VWF:GPIbM) and VWF:RCo assays to investigate whether the VWF:GPIbM/VWF:RCo ratio was able to identify the type 2B variant among an heterogeneous VWD population, previously characterized following the ISTH-SSC guidelines. PATIENTS/METHODS: Seventy-six VWD patients and 31 healthy subjects were evaluated by using VWF:Ag, VWF:RCo, and VWF:GPIbM assays. RESULTS: The mean (minimum-maximum values) VWF:GPIbM/VWF:RCo ratio was higher in type 2B patients (2.53, 0.84-6.11) than in healthy controls (1.05, 0.87-1.34), type 1 (0.85, 0.51-1.15), 2A (1.20, 0.36-2.82), and 2M (1.07, 0.91-1.38) (P < 0.0001). Type 2B variants were divided into four groups (A, B, C, and D) according to their different multimeric patterns. The mean value of the VWF:GPIbM/VWF:RCo ratio in the four groups showed an increasing trend from group A (1.08) to D (3.69), proportional to the loss of high molecular weight multimers. Among 32 type 2B patients, previously diagnosed with RIPA, 8 (mainly with a type I New York/Malmö phenotype) were not confirmed using the VWF:GPIbM/VWF:RCo ratio. CONCLUSIONS: Whenever the RIPA test is not feasible, the VWF:GPIbM/VWF:RCo ratio might help to identify severe type 2B VWD patients.

Predictors of von Willebrand disease diagnosis in individuals with borderline von Willebrand factor plasma levels.

BACKGROUND: In individuals with borderline von Willebrand factor (VWF) plasma levels, second-level tests are required to confirm or exclude von Willebrand disease (VWD). These tests are time-consuming and expensive. OBJECTIVE: To assess which parameters can predict VWD diagnosis in individuals with borderline VWF levels (30-60 IU dL(-1) ). METHODS: Nine hundred and fifty individuals with bleeding episodes or abnormal coagulation test results were investigated with first-level tests (blood count, prothrombin time, activated partial thromboplastin time, blood clotting factor VIII, VWF ristocetin cofactor activity [VWF:RCo], and VWF antigen), and 93 (62 females and 31 males; median age, 28 years; interquartile range 15-44) had borderline VWF:RCo levels. All underwent second-level investigations to confirm or exclude VWD. A multivariable logistic regression model was fitted with sex, age, bleeding score, family history, VWF:RCo and ABO blood group as predictors, and used to predict VWD diagnosis. RESULTS: Forty-five of the 93 individuals (48%) had VWD (84% type 1). A negative linear relationship between VWF:RCo levels and risk of VWD diagnosis was present, and was particularly evident with blood group non-O [adjusted odds ratio 7.00 (95% confidence interval [CI] 1.48-33.11) for every 5 IU dL(-1) decrease in VWF:RCo]. The other variable clearly associated with VWD diagnosis was female sex (adjusted odds ratio 5.76 [95% CI 1.47-22.53]). The area under the receiver operating characteristic curve of the full logistic model was 0.89 (95% CI 0.82-0.95). CONCLUSIONS: In individuals with borderline VWF, the two strongest predictors of VWD diagnosis are low VWF:RCo levels (particularly in those with blood group non-O) and female sex. This predictive model has a promising discriminative ability to identify patients with borderline VWF levels who are likely to have VWD.

A two-centre comparative evaluation of new automated assays for von Willebrand factor ristocetin cofactor activity and antigen.

von Willebrand disease (VWD) is caused by a quantitative and/or qualitative deficiency of the von Willebrand factor (VWF). The laboratory diagnosis of VWD is dependent on the measurement of VWF antigen (VWF:Ag) and ristocetin cofactor activity (VWF:RCo). The aim of this study was to undertake a two-centre evaluation of two new automated VWF:Ag and VWF:RCo assays systems from Instrumentation Laboratory (Bedford, USA). Using the two new analytical systems that operated with different detection principles: immunoturbidimetric (TOP500 analyser) and chemiluminescent (AcuStar analyser), VWF:Ag and VWF:RCo levels were determined in samples from 171 healthy normal subjects, 80 VWD patients (16 type 1, 58 type 2 and 6 type 3) and 7 acquired von Willebrand syndrome patients. With commercial lyophilized normal and pathological plasmas VWF: Ag and VWF:RCo assays performed on both analysers exhibited low levels of inter-assay imprecision (AcuStar: CV% range 3.3-6.9; TOP500: CV% range 2.6-6.3). Samples from normal healthy subjects (range: VWF:Ag 44.6-173.9 IU dL(-1) ; VWF:RCo 43.1-191.5 IU dL(-1)) and patients (range: VWF:Ag <0.3-115.1 IU dL(-1) ; VWF:RCo <0.5-57.2 IU dL(-1)) showed a good correlation between the two VWF:Ag and VWF:RCo methods (rs = 0.92 and 0.82 respectively), with only a few inconsistent cases among the patients' samples evaluated. The chemiluminescent assays had a lower limit of detection for both VWF:Ag and VWF:RCo compared to immunoturbidimetric tests (0.3 IU dL(-1) vs. 2.2 IU dL(-1) and 0.5 IU dL(-1) vs. 4.4 IU dL(-1) respectively). The TOP500 and AcuStar VWF:Ag and VWF:RCo assays were precise and compare well between centres, making these systems suitable for the diagnosis of VWD in non-specialized and reference laboratories.

The type 2B p.R1306W natural mutation of von Willebrand factor dramatically enhances the multimer sensitivity to shear stress.

BACKGROUND: Shear stress triggers conformational stretching of von Willebrand factor (VWF), which is responsible for its self-association and binding to the platelet receptor glycoprotein (GP)Ibα. This phenomenon supports primary hemostasis under flow. Type 2B VWF natural mutants are considered to have increased affinity for platelet GPIbα. OBJECTIVES: To assess the mechanism responsible for the enhanced interaction of the p.R1306W VWF mutant with the platelet receptor. METHODS: The interaction of GPIbα with wild-type (WT) and p.R1306W VWF multimers and A1-A2-A3 constructs was investigated with surface plasmon resonance spectroscopy. Analysis of the static VWF conformation in solution was performed with dynamic light scattering spectroscopy. The shear stress-induced self-association of VWF multimers was investigated with atomic force microscopy (AFM) over a 0-60 dyn cm(-2) range. RESULTS: WT VWF did not interact with GPIbα under static conditions, whereas the mutant at ~ 2 µg mL(-1) already bound to the receptor. By contrast, the WT and p.R1306W-A1-A2-A3 constructs showed comparable affinities for GPIbα (Kd  ~ 20 nm). The hydrodynamic diameter of resting R1306W VWF multimers was significantly greater than that of the wild type (210 ± 60 nm vs. 87 ± 22 nm). At shear forces of < 14 dyn cm(-2) , the p.R1306W multimers rapidly changed conformation, entering a regime of self-aggregation, which, in contrast, was induced for WT VWF by shear forces of > 30 dyn cm(-2) . Mechanical stretching AFM experiments showed that p.R1306W multimers needed less energy per length unit (~ 10 pN) to be stretched than the WT protein. CONCLUSIONS: The increased affinity of p.R1306W VWF for GPIbα arises mostly from higher sensitivity to shear stress, which facilitates exposure of GPIbα binding sites.

A synonymous (c.3390C>T) or a splice-site (c.3380-2A>G) mutation causes exon 26 skipping in four patients with von Willebrand disease (2A/IIE).

INTRODUCTION: We characterized four unrelated patients with von Willebrand disease type 2A/IIE, sharing the same von Willebrand factor (VWF) in-frame deletion (p.[P1127_G1180delinsR];[=]) resulting from exon 26 skipping (Δ26). OBJECTIVES: To identify the VWF mutations and how they caused the mRNA splicing alteration, to evaluate the deletion by in vitro expression studies, and to assess whether or not the heterogeneity of the patients' phenotype might be related to a different degree of expression of the deleted subunit in patient plasma VWF. METHODS: Sequence analysis was performed with patient genomic DNA and platelet mRNA. Semiquantitative RT-PCR was also carried out to compare the expression of the wild-type (WT) and Δ26 alleles in the four patients. In silico analysis was performed with prediction splicing programs. Expression studies were performed to evaluate mutant recombinant VWF (rVWF) (Δ26 and Δ26/WT) as compared with WT rVWF. RESULTS: Three patients shared the synonymous single-nucleotide substitution (SSS) c.[3390C>T];[=], whereas the novel mutation c.[3380-2A>G];[=] was present in the fourth patient. Semiquantitative RT-PCR of platelet mRNA revealed a different ratio of the WT and Δ26 alleles in the patients, consistent with the different VWF:FVIIIB values present in patient plasma. Expression studies confirmed reduced VWF-FVIII binding of rVWF-Δ26/WT. CONCLUSIONS: SSS can induce alternative splicing, and those like c.3390C>T, which impact on the poorly conserved splicing regulatory elements, are difficult to predict, so that their role can be evaluated only by mRNA analysis. Moreover, these mutations seem to have different effects on the efficiency of alternative splicing, producing heterogeneous VWF variants among the four patients.

Type 2A (IIH) von Willebrand disease is due to mutations that affect von Willebrand factor multimerization.

BACKGROUND: Type IIH von Willebrand disease was reported 20 years ago as a novel variant characterized by the loss of the largest multimers in plasma and platelets and absence of the typical triplet structure. OBJECTIVES AND METHODS: The propositus and his daughter have been reinvestigated and characterized at the molecular level. The identified mutations were expressed in COS-7 cells to evaluate the mechanism of this variant. RESULTS AND DISCUSSION: The propositus had normal von Willebrand factor (VWF):ristocetin cofactor activity (RCo) and high VWF antigen (VWF:Ag) values, with a low VWF:RCo/VWF:Ag ratio (0.51). No abnormalities were found in his daughter, except for the reduced triplet structure in plasma VWF and diminished ultralarge VWF (ULVWF) multimers in platelets. Three mutations were identified in the propositus: 604C>T (R202W), 4748G>A (R1583Q), and 2546G>A (C849Y). The amounts of secreted recombinant VWF (rVWF) were apparently increased for R202W (130%), R202W-R1583Q (131%), and R202W-R1583Q/WT (121%), reduced for C849Y (72%) and C849Y/WT (83%), and normal for R1583Q (107%) and R202W-R1583Q/C849Y (102%). In cell lysates, higher values were found in association with the C849Y mutation. A normal multimeric pattern was found in R1583Q rVWF, mainly dimers in R202W rVWF, and intermediate molecular weight multimers in C849Y rVWF. Hybrid R202W-R1583Q/WT and C849Y/WT rVWFs had a nearly normal multimeric pattern, whereas in hybrid R202W-R1583Q/C849Y rVWF there was a loss of large/intermediate multimers. CONCLUSIONS: The propositus phenotype seems to be due to mutations R202W and C849Y, both affecting the VWF multimerization process and, for C849Y rVWF, intracellular survival. The absent triplet multimeric structure in the propositus and its reduction in his daughter appears to be related to the lack of ULVWF multimers, which mainly contribute to the formation of satellite bands.

Expression studies of missense mutations p.D141Y, p.C275S located in the propeptide of von Willebrand factor in patients with type 3 von Willebrand disease.

Missense mutations are not considered a common cause of type 3 von Willebrand's disease (VWD), the most severe defect of von Willebrand factor (VWF) characterized by undetectable levels of this protein in plasma and platelets. Nevertheless, several missense mutations have been identified in these patients. In this study, we report the cases of two Italian patients with type 3 VWD, both compound heterozygotes for different missense mutations and null alleles, p.D141Y/c.2016_2019del and p.C275S/p.W222X. We performed in vitro expression studies of the candidate missense mutations, both located in the D1 domain of VWF propeptide, to confirm their link with the disease and to understand the mechanisms of type 3 VWD responsible in these patients. Mutant and wild-type (WT) expression vectors were used for transient transfection and co-transfection studies in COS-7 cells. Single construct transfections of both missense mutations showed a strongly reduced but detectable secretion of recombinant (r)VWFs (approximately 15% of WT), with essentially only dimers being visualized on multimeric analysis. As expected, expression of a single construct of either mutation with the WT, showed mildly reduced secretion (approximately 40% of WT) and a full set of multimers. These expression studies indicate that the two amino acids D141 and C275 are key residues in the tertiary structure of the VWF propeptide. Their replacement with a tyrosine and a serine, respectively, might compromise propeptide folding, affecting both its intracellular survival and its capacity to mediate multimerization. Co-expression of hybrid rVWFs confirmed the recessive inheritance pattern of these missense mutations.

Biochemical characterization of a recombinant von Willebrand factor (VWF) with combined type 2B and type 1 defects in the VWF gene in two patients with a type 2A phenotype of von Willebrand disease.

BACKGROUND: In a patient previously diagnosed with type 2A von Willebrand disease (VWD) [absence of high and intermediate molecular weight von Willebrand factor (VWF) multimers and markedly reduced ristocetin-induced platelet aggregation (RIPA)], an infusion test of desmopressin was followed by mild thrombocytopenia. This led to further laboratory investigations of his affected brother and of family members, who showed different phenotypic patterns compatible with type 1, 2A, 2B and an uncertain classification of VWD. The two brothers were compound heterozygotes (C275R/P1337L), whereas the others members of the family were heterozygous for C275R (a novel mutation in the D1 domain) or P1337L (a type 2B mutation in the A1 domain). OBJECTIVE AND METHODS: To evaluate the role of the combined effect of the two mutations in the two brothers, C275R and P1337L recombinant (r) VWFs were transiently expressed in COS-7 cells. RESULTS: Recombinant VWF levels secreted in cell media were similar for wild-type (WT), P1337L and hybrid P1337L/WT rVWFs, reduced for hybrids C275R/P1337L and C275R/WT rVWFs, and strongly reduced for C275R rVWF. All rVWFs had a full set of multimers except C275R rVWF, which had only dimers. P1337L rVWF and C275R/P1337L rVWF showed the highest degree of binding to glycoprotein (GP) Ibalpha and the lowest to collagen, followed by P1337L/WT rVWF (with an intermediate level of binding to both ligands), and by WT rVWF with the lowest level of binding to GPIbalpha and the highest to collagen. CONCLUSION: These results suggest that the two compound heterozygous patients have a circulating VWF mainly mutated in the A1 domain (P1337L). This peculiar type 2B VWF variant showed a remarkably high affinity for the GPIbalpha platelet receptor, leading to the loss of high and intermediate molecular weight multimers and hence to decreased RIPA, as in type 2A VWD.

An association of candidate gene haplotypes and bleeding severity in von Willebrand disease type 2A, 2B, and 2M pedigrees.

We analyzed the association of bleeding severity with candidate gene haplotypes within pedigrees of 11 index cases of von Willebrand disease (VWD) type 2 (two type 2A, three type 2B and six type 2M), using the QTL Association model (MENDEL 5.5). In addition to the 11 index cases, these pedigrees included 47 affected and 49 unaffected relatives, as defined by VWF mutations and/or phenotype. A bleeding severity score was derived from a detailed history and adjusted for age. Donors were genotyped using a primer extension method, and eight candidate genes were selected for analysis. VWF antigen (or ristocetin cofactor activity) levels had the strongest influence on bleeding severity score. After Bonferroni correction for multiple testing, only ITGA2 promoter haplotype -52T was associated with an increased bleeding severity score (P < 0.01). This association remained statistically significant when the three type 2B pedigrees were excluded (P = 0.012) or when gender-specific bleeding categories were excluded (P < 0.01). The major haplotypes of seven other candidate genes, GP1BA, ITGA2B, ITGB3, GP6, VWF, FGB, and IL6, were not associated with bleeding severity. These results establish that genetic differences in the expression of the integrin subunit alpha2 can influence the bleeding phenotype of VWD type 2 and complement our previous findings in VWD type 1. Genetically controlled attenuation of platelet collagen receptor expression can influence risk for morbidity in clinical settings where hemostasis is compromised.

Expression studies on a novel type 2B variant of the von Willebrand factor gene (R1308L) characterized by defective collagen binding.

A novel mutation, R1308L (3923G > T) was present in the heterozygous state in five members of a family with type 2B von Willebrand disease (VWD) characterized by a full set of von Willebrand factor (VWF) multimers in plasma and by the absence of thrombocytopenia before and after desmopressin (DDAVP). The defect (R1308L) was located at the same amino acid position of one of the most common mutations associated with type 2B VWD (R1308C), which is characterized by the loss of high molecular weight VWF multimers (HMWM) in plasma and the occurrence of thrombocytopenia. To understand the mechanisms of this defect, the novel (R1308L) and 'common' (R1308C) mutations were expressed in COS-7 cells, either alone or, to mimic the patients' heterozygous state, together with wild-type VWF. R1308L recombinant VWF (rVWF) had a higher affinity for the platelet glycoprotein Ibalpha (GPIbalpha) receptor than wild-type rVWF, R1308C rVWF showing an even higher affinity. A novel finding was that both mutant rVWFs showed a similarly reduced binding to collagen type I and type III in comparison with wild-type rVWF. The latter finding suggests a more important role than recognized so far for the VWF A1 domain in VWF binding to collagen, which may contribute to the in vivo hemostatic defect associated with type 2B VWD.

Molecular characterization of the PK-LR gene in sixteen pyruvate kinase-deficient patients.

We studied the PK-LR gene in 16 unrelated patients with congenital haemolytic anaemia associated with erythrocyte pyruvate kinase deficiency. Fifteen different mutations were detected among the 28 mutated alleles identified: two deletions (del 1010G, del 1042--1044); one four nucleotide duplication (nt 1515--1518, GGTC); one splice site [IVS6(-2)t]; nine missense (991A, 1003A, 1151T, 1160G, 1181T, 1181A, 1456T, 1483A, 1529A); and two nonsense (721T, 1675T) mutations. Eight of them [del 1010G, del 1042--1044, dupl 1515--1518, IVS6(-2)t, 1003A, 1160G, 1181T, 1181A] were novel. The deletion 1042-1044 causes the loss of Lys 348. Deletion 1010G and duplication 1515-1518 determine a frameshift and the creation of a stop codon at nucleotides 1019 and 1554 respectively. Mutation IVS6(-2)t leads to an alteration of the 5' and 3' splice site consensus sequence; the cDNA analysis shows a 67-bp deletion in the first part of exon 11 (del 1437--1503). All the four new missense mutations involve highly conserved amino acids. The most frequent mutation in Italy would appear to be 1456T. Correlation was made between mutations, biochemical characteristics of the enzyme and clinical course of the disease.