High prevalence of hereditary thrombotic thrombocytopenic purpura in central Norway: from clinical observation to evidence

Transcription

1 Journal of Thrombosis and Haemostasis, 14: DOI: /jth ORIGINAL ARTICLE High prevalence of hereditary thrombotic thrombocytopenic purpura in central Norway: from clinical observation to evidence A. S. VON KROGH,* P. QUIST-PAULSEN,* A. WAAGE,* Ø. O. LANGSETH,* K. THORSTENSEN, R. BRUDEVOLD, G. E. TJØNNFJORD, ** C. R. LARGIAD ER, B. L AM M L E and J. A. K R E M E R HOVINGA *Department of Cancer Research and Molecular Medicine, Norwegian University of Science and Technology; Department of Haematology, St Olavs Hospital Trondheim University Hospital; Department of Clinical Chemistry, St Olavs Hospital Trondheim University Hospital, Trondheim, Norway; Department of Haematology, Møre and Romsdal Hospital Trust, Alesund; Department of Haematology, Oslo University Hospital; **Institute of Clinical Medicine, University of Oslo, Oslo, Norway; Department of Clinical Chemistry, Bern University Hospital and University of Bern, Inselspital; Department of Haematology and Central Haematology Laboratory, Bern University Hospital and University of Bern, Inselspital, Bern, Switzerland; Center for Thrombosis and Hemostasis (CTH), University Medical Center, Mainz, Germany; and Department of Clinical Research, University of Bern, Bern, Switzerland To cite this article: von Krogh AS, Quist-Paulsen P, Waage A, Langseth ØO, Thorstensen K, Brudevold R, Tjønnfjord GE, Largiader CR, L ammle B, Kremer Hovinga JA. High prevalence of hereditary thrombotic thrombocytopenic purpura in central Norway: from clinical observation to evidence. J Thromb Haemost 2016; 14: Essentials The population prevalence of hereditary thrombotic thrombocytopenic purpura (TTP) is unknown. We studied the prevalence of hereditary TTP and population frequencies of two ADAMTS-13 mutations. A high frequency of hereditary TTP related to ADAMTS-13 mutation c.4143_4144dupa was found. Vicinity of ABO blood group and ADAMTS-13 loci may facilitate screening of ADAMTS-13 mutations. Summary. Background: Hereditary thrombotic thrombocytopenic purpura (TTP) caused by ADAMTS-13 mutations is a rare, but serious condition. The prevalence is unknown, but it seems to be high in Norway. Objectives: To identify all patients with hereditary TTP in central Norway and to investigate the prevalence of hereditary TTP and the population frequencies of two common ADAMTS-13 mutations. Patients/Methods: Correspondence: Anne-Sophie von Krogh, Department of Haematology, St Olavs Hospital Trondheim University Hospital, PO Box 3250 Sluppen, N-7006 Trondheim, Norway. Tel.: ; fax: anne-sophie.v.krogh@ntnu.no Received 3 July 2015 Manuscript handled by: L. Zheng Final decision: P. H. Reitsma, 22 October 2015 Patients were identified in a cross-sectional study within the Central Norway Health Region by means of three different search strategies. Frequencies of ADAMTS-13 mutations, c.4143_4144dupa and c.3178 C>T (p.r1060w), were investigated in a populationbased cohort (500 alleles) and in healthy blood donors (2104 alleles) by taking advantage of the close neighborhood of the ADAMTS-13 and ABO blood group gene loci. The observed prevalence of hereditary TTP was compared with the rates of ADAMTS-13 mutation carriers in different geographical regions. Results: We identified 11 families with hereditary TTP in central Norway during the 10-year study period. The prevalence of hereditary TTP in central Norway was persons. The most prevalent mutation was c.4143_4144dupa, accounting for two-thirds of disease causing alleles among patients and having an allelic frequency of 0.33% in the central, 0.10% in the western, and 0.04% in the southeastern Norwegian population. The allelic frequency of c.3178 C>T (p.r1060w) in the population was even higher (0.3 1%), but this mutation was infrequent among patients, with no homozygous cases. Conclusions: We found a high prevalence of hereditary TTP in central Norway and an apparently different penetrance of ADAMTS-13 mutations. Keywords: ADAMTS-13 protein, human; congenital thrombotic thrombocytopenic purpura; mutation; prevalence study; Upshaw Schulman syndrome.

2 74 A. S. von Krogh et al Introduction Thrombotic thrombocytopenic purpura (TTP) caused by severe ADAMTS-13 deficiency is characterized by microangiopathic hemolytic anemia, platelet consumption, and varying degrees of organ dysfunction as a result of microvascular thrombosis [1]. Severe ADAMTS-13 deficiency (< 10% of the normal) is the consequence of circulating autoantibodies inhibiting ADAMTS-13 activity or increasing ADAMTS-13 clearance in acquired TTP and the result of homozygous or compound heterozygous ADAMTS-13 mutations in hereditary TTP (Upshaw Schulman syndrome; OMIM #274150). The incidence of idiopathic TTP with a severe acquired ADAMTS-13 deficiency is 2.17 cases per 1 million per year [2]. Although prevalence data for hereditary TTP are lacking, it is assumed that this form accounts for < 5% of all ADAMTS-13-deficient TTP cases [3]. Plasma exchange (PEX) with replacement with fresh frozen plasma has become the mainstay of TTP treatment [4]. The presence of severe thrombocytopenia together with microangiopathic hemolytic anemia without another apparent cause is nowadays considered sufficient to initiate PEX to prevent death or permanent morbidity, and many patients require daily PEX for at least 1 2 weeks or longer to achieve remission. This is in clear contrast to the clinical experience with TTP in central Norway, where many patients, though often experiencing recurrent disease episodes, respond quickly to plasma infusions alone, making more-intensive plasma treatment by PEX very often dispensable [5], which is reminiscent of the hereditary form of TTP. We thus hypothesized that in (central) Norway, hereditary TTP might be more common than elsewhere and decided to perform the first systematic study on the prevalence of hereditary TTP and to assess the allelic frequency of two common ADAMTS-13 mutations c.3178 C>T (p.r1060w) and c.4143_4144dupa, by screening healthy blood donors from different Norwegian regions and in a population-based cohort (The Nord-Trøndelag Health Study). Patients and methods First, the electronic coding system for discharge diagnoses of the eight Central Norway Health Region hospitals were file searched for patients discharged between January 1, 1998, and December 31, 2007, using the International Classification of Diseases, Ninth and Tenth Revisions codes ICD Thrombotic microangiopathy and ICD-10 M31.1 Thrombotic thrombocytopenic purpura. In parallel, the laboratory database of the Hemostasis Research Laboratory at the Department of Hematology and Central Hematology Laboratory, Inselspital, Bern University Hospital, was scanned for patients from Norway referred for ADAMTS-13 activity determination. ADAMTS-13 activity for the first Norwegian patient (Table 1, patient of family 1) was determined in December 2000 [7,8]. As of 2008, ADAMTS-13 activity is assessed in all Norwegian patients with a suspected diagnosis of TTP or unexplained thrombocytopenia, according to national guidelines. Patients with hereditary TTP were identified after review of the medical records and laboratory results based on the following criteria: (i) one or more acute TTP-like episodes with thrombocytopenia and microangiopathic hemolytic anemia, (ii) an ADAMTS-13 activity of < 10% of the normal in the absence of a functional ADAMTS-13 inhibitor, (iii) homozygous or compound heterozygous ADAMTS-13 mutation(s), or (iv) a plasma C HUNT2 Study area and population The Central Norway Health Region is one of four health administration bodies in Norway (Fig. 1) and encompasses an area of km 2 and a population of on January 1, 2008 (data from Statistics Norway [6]). W CENTRAL NORWAY HEALTH REGION Hereditary TTP cases and families To identify subjects and families with confirmed or suspected hereditary TTP, we used three different search strategies, two of which were interlocked (Fig. 2). SE Fig. 1. Norway with study area Central Norway Health Region and study area for the second Nord-Trøndelag Health Study (HUNT2): C, central Norway; SE, southeastern Norway; W, western Norway.

3 Prevalence of hereditary TTP in Norway 75 Population in Central Norway Search strategy I Computer search ICD-10 M31.1 ICD Search strategy II Laboratory database Search strategy III Extended computer search Hits 16 Hereditary TTP cases from Norway 16 Hits 1300 Aquired TTP 5 Already identified by search strategy I 10 Living outside Central Norway 6 Not fulfilling selection criteria 1270 Dead 1 Hereditary TTP 11 Additional cases from Central Norway 0 Patients eligible for further testing 29 Dead 1 Patients tested 17 Eligible patients 10 Confirmed hereditary TTP 1 Patients with informed consent 10 Fig. 2. Flow chart of the case finding strategy for patients with hereditary thrombotic thrombocytopenic purpura (TTP) in central Norway. Identification of patients from central Norway with hereditary TTP was performed using the following flow chart. Norwegian hospitals use the International Classification of Diseases (ICD) system for discharge diagnoses. For the 10-year study period , the Ninth Revision and 10th Revision of the ICD were used. A computer-based search was performed, followed by review of the medical records of patients fulfilling inclusion criteria of search strategy I and III. infusion trial with full-recovery and a plasma half-life of infused ADAMTS-13 of 2 4 days. (v) Given the possibility of residual ADAMTS-13 activity, especially in adultonset hereditary TTP, a patient was also considered to have Upshaw Schulman syndrome when a suitable patient history was associated with homozygous or compound heterozygous ADAMTS-13 mutation(s) previously reported in hereditary TTP cases. Finally, we performed an extended computer-based diagnosis search with predefined stepwise selection to identify atypical or misdiagnosed cases (Fig. 2, search strategy III). The search strategy was founded on associated diagnoses, observed among patients who later were confirmed to have hereditary TTP but whose diagnosis had been delayed. The search included: Thrombocytopenia (ICD-10: D69.3 6, P61.0 and ICD 9: ), Hemolytic uremic syndrome (ICD-10: D59.3 and ICD-9: ), Disseminated intravascular coagulation (ICD-10: D65 and ICD-9: 286.6), Hemolytic anemia (ICD-10: D59.0 1, D59.4 6, D59.8 9, and ICD-9: 282.9, ), Acute stroke (ICD-10: I and ICD-9: , , 436), Transient ischemic attack (ICD-10: G and ICD-9: 435), Acute myocardial infarction (ICD-10: I21.0 9, I22.0 9, and ICD-9: 410, 411), and

4 76 A. S. von Krogh et al Table 1 Characteristics of Norwegian patients with hereditary thrombotic thrombocytopenic purpura (TTP) (Upshaw Schulman syndrome) Area Fam Pat Sex Age (y) at Diagnosis Present NExT Clinical course Plasma prophylaxis ABO ADAMTS-13 act (%) ADAMTS-13 genotype Ref C 1 F 1 15 No Intermediate Yes BB < 1 p.c804r p.c804r 7, 8 C 2 a M 51 D at No Severe Yes A 1 B 1.8 p.c804r p.c1213y 59 y C b M D at Yes n.a. n.d. n.d. 0y C 3 M Yes Severe Yes BB 1.8 p.c804r c.4143insa* C 4 F No Mild No BB < 1 c.4143insa c.4143insa 32 C 5 M Yes Intermediate No BB < 1 c.4143insa c.4143insa C 6 F Yes Mild No BB < 1 c.4143insa c.4143insa 32 C 7 M No Intermediate Yes BB < 1 c.4143insa c.4143insa C 8 M Yes Mild No BB < 1 c.4143insa c.4143insa C 9 a F 23 D at No Intermediate No BB < 1 c.4143insa c.4143insa 5, y C b M 2 D at Yes Severe Yes B n.d n.d y C 10 M Yes Mild No BB < 1 c.4143insa c.4143insa C 11 M No Intermediate Yes A 1 A p.l232q p.r1060w C 12 F No Mild No A 2 B < 1 p.r1060w c.4143insa W 13 M No Severe Yes BB < 1 c.4143insa c.4143insa SE 14 a F Yes Severe Yes O 1 B < 1 p.r507q c.4143insa 17, 32 SE b F D at n.a. n.a. n.d. n.d. 0y SE 15 F Yes Intermediate Yes BB 1.0 c.4143insa c.4143insa 17, 32 SE 16 F 4 9 No Intermediate Yes BB < 1 c.4143insa c.4143insa SE 17 M No Mild Yes BB < 1 c.4143insa c.4143insa SE 18 M No Mild No A 1 A p.v88m p.v88m Fam, family; Pat, patient; NExT, neonatal exchange transfusion; ABO, genotyped ABO blood group; Ref, reference where patient has been described previously; C, central Norway; W, western Norway; SE, southeastern Norway; D at, died at age; n.a., not available; n.d., not done. *Old nomenclature used to save space. Patient diagnosed with hereditary TTP after January 1, Phenotypic blood group. Acute renal insufficiency (ICD-10: N and ICD-9: , ). Thrombocytopenia ICD-10 and ICD-9 diagnoses included all age groups, whereas for the other diagnoses only patients aged < 50 years at the time of diagnosis were considered. Validity of diagnoses and eligibility for study participation was evaluated by reviewing the medical records. Patients were found eligible for study participation if they had (i) one of the listed diagnoses together with unexplained thrombocytopenia, hemolytic anemia, or episodes of abdominal pain or (ii) two or more of the listed diagnoses. Identified cases were compared with the death register, and all living patients were invited to participate in the study. Clinical data were obtained from the patients medical records and updated as of January We created a semiquantitative severity score based on the following clinical end points: history of neonatal blood exchange (1 point); age at first TTP episode requiring plasma therapy (2 points for age < 5 years, 1 point for age 5 18 years, and 0 points for age > 18 years); receiving regular prophylactic plasma infusions (2 points); evidence of TTP episode (s) after initiation of prophylactic plasma therapy (2 points); evidence of permanent organ damage (2 points); and death of disease (3 points). Patients with 3 points were considered to have a mild, patients with 4 6 points an intermediate, and patients with > 6 points a severe clinical course. Healthy blood donor cohorts and HUNT2 population-based cohort EDTA whole blood samples from healthy blood donors were obtained from University Hospital blood banks in the health administration regions of central, western, and southeastern Norway (Fig. 1) between March 2008 and November The ABO blood group gene locus lies in the close neighborhood of the ADAMTS-13 gene on chromosome 9q34 (Fig. S1). All patients homozygous for the c.4143_4144dupa mutation were of phenotypic blood group B, which was confirmed by genotyping to be BB (Table 1). As in the two-first studies describing the c.4143_4144dupa mutation[9,10], neither of the authors observed the mutation in 100 screened random control alleles; to increase the number of b-alleles in our donor population, we screened healthy blood donors of phenotypic blood groups B and AB. Samples for a resident population cohort were randomly picked from the second Nord Trøndelag Health Study (HUNT2), a large health survey conducted in

5 Prevalence of hereditary TTP in Norway 77 Nord Trøndelag County (Fig. 1), central Norway, between 1995 and The study, described in detail elsewhere [11], includes questionnaire-based information and biological material of > participants, constituting 71% of the Nord Trøndelag County population. For the HUNT2 study, no information on the subjects blood group is available. The study was approved by the Regional Committee for Medical and Health Research Ethics, which also granted the two electronic database searches limited to the time window of January 1, 1998, to December 31, Written informed consent was obtained before inclusion from all eligible subjects. In case patients were dead or younger than 18 years, consent was obtained from next of kin. Blood donors signed either a project specific or a general consent, according to local guidelines, and their blood samples were handled anonymously. ADAMTS-13 parameters and ABO blood groups ADAMTS-13 activity was determined in citrated plasma by the quantitative immunoblotting assay and/or the slightly modified FRETS-VWF73 assay as previously described [12 15]. FRETS-VWF73 is a fluorescence resonance energy transfer assay using a synthetic 73-aminoacid VWF peptide. The detection limit of this assay is 1%. All samples with an ADAMTS-13 activity of 20% were subjected to a screening test for the presence of an ADAMTS-13 inhibitor [15]. DNA extraction from leukocytes, ADAMTS-13 gene amplification, and sequencing were performed using standard methods. Blood donors and the HUNT2 cohort were genotyped for the ADAMTS-13 mutations c.3178 C>T (p.r1060w) and c.4143_4144dupa using custom-designed TaqMan Ò assays (Applied Biosystems, now Thermo Fisher Scientific, Waltham, MA, USA). Genotyping for the latter mutation was done in addition with an in-house, FRETbased, melting curve analysis assay on the Light Cycler 2.0 system (Roche Diagnostics, Indianapolis, IN, USA). Primer sequences and detailed assay conditions are available upon request. The ABO blood group locus was genotyped using a commercially available assay (ABO-SSP; Innotrain Diagnostik, Kronberg, Germany) in hereditary TTP patients as well as in ADAMTS-13 mutation carriers of the HUNT2 cohort. Statistical methods Allele frequencies of ADAMTS-13 mutations c.3178 C>T (p.r1060w) and c.4143_4144dupa were obtained by gene counting in the different cohorts and assessed for deviation from the Hardy Weinberg equilibrium. The genotype frequencies in the cohorts were compared by means of v 2 analysis or Fisher exact test where appropriate. To deduce the frequency of different ABO blood group alleles, we used phenotypic ABO blood group data from Kornstad [16] and information provided by the respective blood banks on their donor population for a 2-year period, ABO allele frequencies were calculated by means of the Hardy Weinberg equation for three alleles and a maximum likelihood estimator (S2 ABO estimator; ABOestimator/). We estimated the prevalence of hereditary TTP in central Norway by using all living Upshaw-Schulman syndrome cases identified between January 1, 1998, and December 31, 2007, as the numerator and the January 1, 2008, central Norway population as the denominator [6]. Results Hereditary TTP patients and their families During a period of 10 years ( ), 12 cases of hereditary TTP in 11 families were identified in central Norway through our different search strategies (Fig. 2). Ten patients had been diagnosed after recurrent TTP episodes leading to ADAMTS-13 investigation. One patient, the brother of a confirmed patient homozygous for the c.4143_4144dupa mutation, had died after frequently recurring TTP episodes before ADAMTS-13 activity was measured (Table 1, patient 9b). His TTP episodes usually had responded rapidly to plasma infusions; he had been of phenotype blood group B, and a diagnosis of hereditary TTP was assumed to be very likely. The 12th patient (Table 1, patient of family 10) was identified through the extended search strategy. He had undergone an exchange transfusion for neonatal jaundice and had had recurrent, self-limiting episodes of microangiopathic hemolytic anemia and thrombocytopenia during childhood. His ICDcoded diagnosis was Recurrent hemolytic uremic syndrome. During the same 10-year period, five patients were diagnosed with acquired TTP with severe ADAMTS-13 deficiency due to an inhibitory autoantibody After 2008, one additional hereditary TTP case in central Norway was discovered, presenting with adult-onset disease (Table 1, patient of family 12). Outside of central Norway, six hereditary TTP cases were identified through the laboratory database search (Table 1, families 13 18). Finally, two patients (patients 2a and 14a) reported death of a sibling within the first year of life in the 1950s. We were unable to retrieve detailed health information on these two infants, and they are not further considered. Of note, however, is the fact that in one of them an exchange transfusion had been performed because of neonatal jaundice. To the best of our knowledge, neither the 18 Norwegian Upshaw Schulman syndrome families nor the parents of index cases are related.

6 78 A. S. von Krogh et al The most prevalent ADAMTS-13 mutation in these 18 families was c.4143_4144dupa in exon 29, which accounted for 25 (69%) of 36 disease-causing alleles and was present in homozygous form in 11 families, of whom seven were from central Norway. The homozygous patients shared the same ancestral ADAMTS-13 haplotype as the patients described by Schneppenheim et al. [17], and the allele structure in three compound heterozygotes did not exclude identity with this common haplotype. The remaining alleles displayed six different missense mutations (Table 1): c.262 G>A (p.v88m; two alleles), c.695 T>A (p.l232q), c.1520 G>A (p.r507q), c.2410 T>C (p.c804r; four alleles), c.3178 C>T (p.r1060w; two alleles), and c.3638 G>A (p.c1213y). In view of the recent discussion on a genotype phenotype correlation in hereditary TTP [18], it is of interest that the clinical phenotype in the 11 c.4143_4144dupa homozygotes is variable: Five of 11 patients (45.5%) had had an exchange transfusion in the neonatal period; the clinical course was mild in five and intermediate or severe in six patients. Five patients are on regular prophylactic plasma infusions. Although the genotype of patient 9b was not formally proven, he was likely homozygous for c.4143_4144dupa and displayed the most severe course, eventually resulting in premature death at age 40. Prevalence of ADAMTS-13 mutations c.4143_4144dupa and c.3178 C>T (p.r1060w) in central Norway Genotype distribution for the two ADAMTS-13 mutations in the HUNT2 and the different blood donor cohorts were in Hardy Weinberg equilibrium in all cohorts; data are summarized in Table 2. In central Norway, we identified 15 heterozygous c.4143_4144dupa mutation carriers in the first and 17 in the replication cohort of blood group AB and B blood donors, while this mutation was not observed in the HUNT2 cohort or in the cohort of blood group A and O donors. In the two AB and B blood group donor cohorts, the prevalence of the c.4143_4144dupa mutation was 2.87% and 3.62%, respectively. Based on the estimated prevalence of the blood group b-allele in the central Norwegian population (inferred from the present-day blood bank donor population and Kornstad [16], for details see Table S1), this amounts to an allelic frequency among blood group b- alleles of 5.15% and 7.02% in the investigated cohorts, and of 0.33% and 0.41% in the general central Norwegian population, respectively. The prevalence of the c.4143_4144dupa mutation among AB and B blood donors in western Norway, where one hereditary TTP family (index patient homozygous for c.4143_4144dupa) has been identified so far, was Table 2 ADAMTS-13 mutations c.3178 C>T (p.r1060w) and c.4143_4144dupa in different cohorts in central Norway and comparison with other regions in Norway ADAMTS-13 c.3178 C>T (p.r1060w) ADAMTS-13 c.4143_4144dupa Region Cohort Alleles (n) Alleles (n) Allele frequency in Allele frequency in/among Cohort (95% CI) Population * Alleles (n) Cohort (95% CI) b-alleles Population P Central Western Southeastern HUNT2; % ( %) 1% 0 random cases; 50% males Aor $ 0.36% ( %) 0.34% 0 O blood donors AB or B blood donors First cohort 522 n.d % ( %) 5.15% 0.33% Replication cohort % ( %) 7.02% 0.41% AB or 736 n.d % ( %) 1.87% 0.10% B blood donors AB or 660 n.d % ( %) 0.59% 0.04% < B blood donors Genotype success rates were % in all cohorts; n.d., not done. *Calculated using the observed c.3178 C>T (p.r1060w) allele frequency in the cohort and that of the phenotypic blood groups A and O in the population. Calculated using the observed c.4143_4144dupa allele frequency, the estimate on the blood group b-alleles in the cohort and in the population. Comparison of allelic frequency of ADAMTS-13 c.4143_4144dupa between central Norway and western and southeastern Norway (v2 test). Genotyped blood groups of the nine heterozygous c.3178 C>T (p.r1060w) carriers: four times O 1 A 2, twice A 2 A 2, and once O 1 O 1,O 1 A 1, and A 1 A 2, each.

7 Prevalence of hereditary TTP in Norway %, amounting to 1.87% of blood group b-alleles and a prevalence in the resident population in western Norway of 0.10%. The corresponding numbers for southeastern Norway, with three hereditary TTP families with c.4143_4144dupa homozygotes, are 0.30%; 0.59%, and 0.04%, respectively. The difference in c.4143_4144dupa prevalence between central Norway and western or southeastern Norway was statistically significant (P = and P < , respectively). In central Norway, the index patients of two families were identified to be compound heterozygous for c.3178 C>T (p.r1060w) and a second mutation. The HUNT2 cohort sample, the blood group A and O donor cohort, as well as the replication cohort of blood group AB and B donors, were genotyped for this ADAMTS-13 mutation. Nine heterozygous carriers were identified: five in the HUNT2 cohort (allelic frequency in cohort and population 1%) and four in the blood group A and O donor cohort (allelic frequency in cohort and population 0.36% and 0.34%, respectively). The blood groups of all nine c.3178 C>T (p.r1060w) carriers were genotyped and found to be O1A2 in four, A2A2 in two, and A1A2, O1A1 or O1O1 in one carrier each. Although the ADAMTS-13 sequence variation haplotype of c.3178 C>T (p.r1060w) mutation carriers would be compatible with a founder mutation, the linkage with the blood group gene locus is less stringent than in c.4143_4144dupa carriers hinting at an older age of the c.3178 C>T (p.r1060w) mutation. Estimate on the prevalence of hereditary TTP in Norway During our 10-year search period, 12 hereditary TTP cases had been identified in central Norway, of which 11 were alive as of January 1, 2008, providing a point estimate of prevalence of hereditary TTP in this region of 16.7 cases per 1 million. In the same period, five different patients were diagnosed with acquired TTP in this region, amounting to an annual incidence of a first acute TTP bout of 0.77 case per 1 million. At present, there are 16 known patients with hereditary TTP alive in Norway (population as of January 2015 [6]). Although a systematic case-finding strategy has not been performed for the whole country, this number would account to a point estimate of prevalence of hereditary TTP in Norway of 3.1 cases per 1 million. Using the observed allele frequencies and the blood group distribution, we estimated the number of c.4143_4144dupa homozygous carriers to be 7 11 in central Norway, in line with the actually observed number. The contrary was true for the c.3178 C>T (p.r1060w) mutation, where the estimated number of homozygotes was 7 66, but not one such patient was observed. This would, however, be in line with a reduced phenotypic penetrance. Discussion This is the first population-based study on hereditary TTP combining estimates based on genotype frequencies in the population and a structured case-finding procedure. We found a prevalence of 16.7 hereditary TTP cases per 10 6 inhabitants in central Norway, which is at least times higher than previous estimates [3,19 22]. These previous estimates were mainly based on the ratio of acquired and hereditary TTP observed in regional or national registries, enrolling TTP patients referred for PEX or ADAMTS-13 activity determination. In these registries, hereditary TTP cases represent only 5 10% of all TTP cases with a severe ADAMTS-13 deficiency. In central Norway, we encountered more than twice as many cases of hereditary (n = 12) compared with acquired (n = 5) TTP cases in the 10-year study period. The incidence of acquired, idiopathic TTP of 0.77 case per 1 million per year in central Norway is thus slightly lower than the Oklahoma incidence rate for white patients of 1.24 cases per 1 million per year [2]. Hereditary TTP has been observed on all continents, and to date, > 120 different causative ADAMTS-13 mutations are known, of which the majority are confined to single families. Consequently, homozygous patients were primarily encountered in areas with high prevalence of consanguinity. Two mutations observed in hereditary TTP patients of European ancestry represent exceptions: (i) the single nucleotide insertion in ADAMTS-13 exon 29, c.4143_4144dupa resulting in a frameshift and premature termination at amino acid 1386, and (ii) the missense mutation c.3178 C>T (p.r1060w) in exon 24. Both mutations have been described in a number of unrelated families of European ancestry [9,10,17,23 31]. Despite the high prevalence of heterozygous c.3178 C>T (p.r1060w) carriers in the HUNT2 cohort (1%), to date no Upshaw Schulman syndrome patient homozygous for this mutation has been identified in Norway. This missense mutation, though leading to a severely impaired ADAMTS-13 synthesis and secretion, is usually associated with some residual ADAMTS-13 activity in plasma. The majority of reported patients with this mutation have remained asymptomatic into adulthood, when strong triggers such as pregnancy or pneumonia provoked a first acute episode. While one of the Norwegian Upshaw Schulman syndrome patients carrying one c.3178 C>T (p.r1060w) allele had a mild clinical course, the other (Table 1, patient of family 11) was diagnosed after several severe TTP episodes and strokes, despite having residual ADAMTS-13 activity of ~5% on several occasions during remission. Among the 18 Norwegian Upshaw Schulman syndrome families, 11 index patients were homozygous for the c.4143_4144dupa mutation (Table 1). Most of them had, in retrospect, a disease onset in the neonatal period or in early childhood, although a formal diagnosis was

8 80 A. S. von Krogh et al often only made in adulthood [32]. However, despite the shared genotype and a documented severe ADAMTS-13 deficiency of < 1% of the normal, the clinical courses in these patients were variable, including premature death, hinting at thus far unknown disease-modifying factors. To our knowledge, there are no asymptomatic homozygous c.4143_4144dupa cases known, suggesting that the two common ADAMTS-13 mutations are associated with different clinical penetrance. This may help explain the fact that we observed as many homozygous c.4143_4144dupa cases in central Norway as expected from the prevalence of the heterozygous mutation carriers in the blood donor cohorts, whereas c.3178 C>T (p.r1060w) homozygotes were not encountered despite a mutation carrier frequency of up to 1% in the population (Table 2). Both ADAMTS-13 mutations have a widespread geographical distribution. ADAMTS-13 c.3178 C>T (p.r1060w) has been observed in patients from Turkey, central Europe, Scandinavia, Great Britain, France, and Italy and in North Americans of European descent [22,30,31]. Moreover, it was found in the National Heart, Lung, and Blood Institute GO Exome Sequencing Project population of North America as well as in the CEPH sample in the 1000 Genomes Browser, where it is listed as a single nucleotide polymorphism (rs ). In a recent study from the Netherlands, it was included in an exome array of rare variants in the ADAMTS-13 gene with a reported frequency of 0.06% in that cohort [33]. The c.4143_4144dupa mutation has been described in case series from central and northern Europe as well as in an Australian patient with German forebears [9,10,17,23,24,27,28]. Haplotype analyses of ADAMTS-13 c.4143_4144dupa alleles suggested a founder mutation in a common ancestor, possibly from central Europe [17]. The high prevalence of patients with this mutation in central Norway, where consanguineous marriages are rare [34], matched by a high frequency of heterozygous carriers in the blood donor cohorts is remarkable and challenges this assumption. Given the significantly lower c.4143_4144dupa rate in healthy blood donors in the western and southeastern parts of the country, it is possible that the mutation arose in central Norway. A spreading pattern of mutation following the migratory activities of the Vikings, similar to what has been proposed for the HFE gene p.c282y mutation, is possible [35]. Another possible explanation is a population bottleneck phenomenon with a substantial population size reduction, such as by the plague that rummaged in central Norway in the 14th century, and relative isolation followed by reproductive compensation, thereby increasing the number of already existing c.4143_4144dupa alleles. All other ADAMTS-13 mutations found in our hereditary TTP patients have been reported as causative for severe congenital ADAMTS-13 deficiency, but their appearances were sporadic [36,37]. Therefore, the ADAMTS-13 c.4143_4144dupa mutation is likely the main cause of the clustering of hereditary TTP in central Norway. The widespread distribution of both ADAMTS-13 mutations among persons of European ancestry despite a clear selective disadvantage (e.g., severe pregnancy complications in most homozygous or compound heterozygous females reaching adulthood [32]) raises the question how these mutations escaped negative selection pressure. From a clinical perspective and current knowledge, it seems unlikely that the ADAMTS-13 mutations themselves conferred a selective advantage. Some hitchhiking effect due to close linkage with positively selected alleles at the neighboring ABO blood group locus may provide a more likely explanation. Indeed, the adaptive importance of this locus, especially regarding conferring (relative) resistance toward different infectious diseases, such as cholera, plague, or malaria, has long been recognized [38 40]. This study has some limitations. Research on rare diseases is inevitably associated with small patient numbers and, therefore, assumptions are subject to uncertainty. The cross-sectional study design of the case-finding procedure represents a weakness. Even though we made every effort to facilitate participation in the extended case-finding procedure, only 17 (59%) of 29 eligible patients took part. Bias was attempted to be minimized by using three different search strategies. Asymptomatic cases were not identified by our search strategies, nor were asymptomatic siblings, as genetic testing of family members is restricted by the Norwegian biotechnology act. Finally, as blood donors represent a highly selected, healthy population of young and middle-aged adults, screening blood donors instead of a general population sample likely slightly underestimated the mutation frequencies. Taken together, the mentioned limitations rather tend to underestimate than overestimate the number of hereditary TTP cases in Norway. The longstanding clinical observation that the response to FFP infusion in most Norwegian patients with acute TTP is good, obviating the need for PEX, contrasted the general experience in adult TTP patients elsewhere. This discrepancy set off this study, which confirmed that hereditary TTP is more common than the acquired form in central Norway, with an estimated prevalence of 16.7 hereditary TTP cases per 1 million, which is the highest reported so far. This is matched by the high frequency of heterozygous ADAMTS-13 mutation carriers [c.4143_4144dupa or c.3178 C>T (p.r1060w)] in the population. The latter observation and the apparently differing clinical penetrance of the investigated ADAMTS-13 mutations open up the possibility of further, thus-farunrecognized cases, not only in Norway but also in other areas, particularly so where c.3178 C>T (p.r1060w) has been observed. The association of certain blood groups with specific ADAMTS-13 mutations may be used to facilitate the screening procedure and estimates of their

9 Prevalence of hereditary TTP in Norway 81 population frequencies. Other consequences of a high population frequency of ADAMTS-13 mutation carriers, in terms of cardiovascular and cerebrovascular morbidity and mortality, are unclear and should be substance for future research. Addendum A. S. von Krogh, P. Quist-Paulsen, A. Waage, B. L ammle, and J. A. Kremer Hovinga designed the research. Ø. O. Langseth, R. Brudevold, G. E. Tjønnfjord, and A. S. von Krogh collected and analyzed patient data. K. Thorstensen and J. A. Kremer Hovinga performed experiments. A. S. von Krogh, C. R. Largiader, and J. A. Kremer Hovinga analyzed data. A. S. von Krogh and J. A. Kremer Hovinga wrote the manuscript. All authors read, commented on, and approved the final version of the manuscript. Acknowledgements We thank Irmela Sulzer, Gabriela M ader-heinemann, Hildegunn E. Pettersen, and Mona Kvitland for expert technical assistance. We thank the HUNT board and biobank for samples and technical help, as well as Rune Logan-Halvorsrud, Tor Hervig, and Richard Olaussen of the blood banks of St Olavs Hospital, Trondheim University Hospital; Haukeland University Hospital; and Oslo University Hospital for providing blood donor samples. The Regional Biobank of Central Norway is acknowledged for the collaboration regarding DNA extraction and storage. We thank Elisabeth Siebke, Møre and Romsdal Hospital Trust; Adriani Kanellopoulos, Einar Stensvold, Marit Rinde and Nina H. Schulz, Akershus University Hospital; Astrid Dale, Førde Hospital Trust; Einar Svarstad, Haukeland University Hospital; and Jan Rocke, Levanger Hospital Trust for providing clinical data. This work was supported by grants from Unimed Innovations (grant 10/ to A. S. von Krogh), the ISTH 2007 Presidential Fund (to J. A. Kremer Hovinga), and the Swiss National Science Foundation (grant 32003B to J. A. Kremer Hovinga and B. L ammle). Disclosure of Conflict of Interests A. Waage, J. A. Kremer Hovinga, and B. L ammle have served Baxter as consultants for radamts-13 development as treatment of hereditary TTP. The hereditary TTP registry ( ClinicalTrials.gov identifier NCT ) is supported by an unrestricted grant from Baxter to J. A. Kremer Hovinga. B. L ammle reports personal fees from Alecion and Siemens outside the submitted work and has an ADAMTS-13 patent issued. All other authors state that they have no conflict of interest. Supporting Information Additional Supporting Information may be found in the online version of this article: Fig. S1. Schematic presentation of the chromosome 9q34. Table S1. Distribution of phenotypic ABO blood groups and estimate of the corresponding a, b, o alleles in different regions in Norway. Allelic frequencies are estimated from the ABO phenotypes by the Hardy Weinberg equation and a maximum likelihood estimator. References 1 Moake JL. Thrombotic microangiopathies. N Engl J Med 2002; 347: Reese JA, Muthurajah DS, Kremer Hovinga JA, Vesely SK, Terrell DR, George JN. Children and adults with thrombotic thrombocytopenic purpura associated with severe, acquired Adamts13 deficiency: comparison of incidence, demographic and clinical features. Pediatr Blood Cancer 2013; 60: Galbusera M, Noris M, Remuzzi G. Inherited thrombotic thrombocytopenic purpura. Haematologica 2009; 94: Rock GA, Shumak KH, Buskard NA, Blanchette VS, Kelton JG, Nair RC, Spasoff RA. Comparison of plasma exchange with plasma infusion in the treatment of thrombotic thrombocytopenic purpura. Canadian Apheresis Study Group. N Engl J Med 1991; 325: Waage A, Siegel J, Thorstensen K, Lamvik J. Thrombotic thrombocytopenic purpura in 2 siblings: defective platelet function and plasma factor deficiency occurring simultaneously. Scand J Haematol. 1986; 36: Statistics Norway. Official statistics about Norwegian society since Accessed 14 April Studt JD, Kremer Hovinga JA, Alberio L, Bianchi V, L ammle B. von Willebrand factor-cleaving protease (ADAMTS-13) activity in thrombotic microangiopathies: diagnostic experience 2001/2002 of a single research laboratory. Swiss Med Wkly 2003; 133: Gulati S, Siebke E, Gamlem AM. A child with relapsing haemolytic anemia and thrombocytopenia. Tidsskr Nor Laegeforen 2007; 127: Assink K, Schiphorst R, Allford S, Karpman D, Etzioni A, Brichard B, van de Kar N, Monnens L, van den Heuvel L. Mutation analysis and clinical implications of von Willebrand factorcleaving protease deficiency. Kidney Int 2003; 63: Schneppenheim R, Budde U, Oyen F, Angerhaus D, Aumann V, Drewke E, Hassenpflug W, Haberle J, Kentouche K, Kohne E, Kurnik K, Mueller-Wiefel D, Obser T, Santer R, Sykora KW. von Willebrand factor cleaving protease and ADAMTS13 mutations in childhood TTP. Blood 2003; 101: Holmen J, Midthjell K, Kr uger O, Langhammer A, Holmen TL, Bratberg GH, Vatten L, Lund-Larsen PG. The Nord-Trøndelag Health Study (HUNT 2): objectives, contents, methods and participation. Norwegian Journal of Epidemiology (Norsk Epidemiologi). 2003; 13: Kokame K, Nobe Y, Kokubo Y, Okayama A, Miyata T. FRETS-VWF73, a first fluorogenic substrate for ADAMTS13 assay. Br J Haematol 2005; 129: Froehlich-Zahnd R, George JN, Vesely SK, Terrell DR, Aboulfatova K, Dong JF, Luken BM, Voorberg J, Budde U, Sulzer I, L ammle B, Kremer Hovinga JA. Evidence for a role of anti-adamts13 autoantibodies despite normal ADAMTS13 activity in recurrent thrombotic thrombocytopenic purpura. Haematologica 2012; 97:

10 82 A. S. von Krogh et al 14 Kremer Hovinga JA, Mottini M, L ammle B. Measurement of ADAMTS-13 activity in plasma by the FRETS-VWF73 assay: comparison with other assay methods. J Thromb Haemost 2006; 4: Kremer Hovinga JA, Vesely SK, Terrell DR, L ammle B, George JN. Survival and relapse in patients with thrombotic thrombocytopenic purpura. Blood 2010; 115: Kornstad L. Frequency of the blood group antigen K and the A1A2BO groups in the Norwegian counties. Gene Geography: A Computerized Bulletin on Human Gene Frequencies. 1997; 11: Schneppenheim R, Kremer Hovinga JA, Becker T, Budde U, Karpman D, Brockhaus W, Hrachovinova I, Korczowski B, Oyen F, Rittich S, von Rosen J, Tjønnfjord GE, Pimanda JE, Wienker TF, L ammle B. A common origin of the 4143insA ADAMTS13 mutation. Thromb Haemost 2006; 96: Lotta LA, Wu HM, Mackie IJ, Noris M, Veyradier A, Scully MA, Remuzzi G, Coppo P, Liesner R, Donadelli R, Loirat C, Gibbs RA, Horne A, Yang S, Garagiola I, Musallam KM, Peyvandi F. Residual plasmatic activity of ADAMTS13 is correlated with phenotype severity in congenital thrombotic thrombocytopenic purpura. Blood 2012; 120: Fujimura Y, Matsumoto M. Registry of 919 patients with thrombotic microangiopathies across Japan: database of Nara Medical University during Intern Med 2010; 49: Kokame K, Kokubo Y, Miyata T. Polymorphisms and mutations of ADAMTS13 in Japanese population and estimation of the number of patients with Upshaw-Schulman syndrome. J Thromb Haemost 2011; 9: Scully M, Yarranton H, Liesner R, Cavenagh J, Hunt B, Benjamin S, Bevan D, Mackie I, Machin S. Regional UK TTP registry: correlation with laboratory ADAMTS 13 analysis and clinical features. Br J Haematol 2008; 142: Lotta LA, Garagiola I, Palla R, Cairo A, Peyvandi F. ADAMTS13 mutations and polymorphisms in congenital thrombotic thrombocytopenic purpura. Hum Mutat 2010; 31: Pimanda JE, Maekawa A, Wind T, Paxton J, Chesterman CN, Hogg PJ. Congenital thrombotic thrombocytopenic purpura in association with a mutation in the second CUB domain of ADAMTS13. Blood 2004; 103: Peyvandi F, Ferrari S, Lavoretano S, Canciani MT, Mannucci PM. von Willebrand factor cleaving protease (ADAMTS-13) and ADAMTS-13 neutralizing autoantibodies in 100 patients with thrombotic thrombocytopenic purpura. Br J Haematol 2004; 127: Manea M, Kristoffersson A, Tsai HM, Zhou W, Winqvist I, Oldaeus G, Billstrom R, Bj ork P, Holmberg L, Karpman D. ADAMTS13 phenotype in plasma from normal individuals and patients with thrombotic thrombocytopenic purpura. Eur J Pediatr 2007; 166: Cermakova Z, Hrdlikova R, Sulakova T, Koristka M, Kovarova P, Hrachovinova I. Thrombotic thrombocytopenic purpura: incidence of congenital form of disease in north Moravia (region Moravia-Silesia). Prague Med Rep. 2009; 110: Lindblom A, Thorsen S, Hillarp A, Bj ork P. Minor stroke as singular manifestation of hereditary thrombotic thrombocytopenic purpura in a young man. International Angiology: A Journal of the International Union of Angiology. 2009; 28: Klukowska A, Niewiadomska E, Budde U, Oyen F, Schneppenheim R. Difficulties in diagnosing congenital thrombotic thrombocytopenic purpura. J Pediatr Hematol Oncol 2010; 32: Donadelli R, Banterla F, Galbusera M, Capoferri C, Bucchioni S, Gastoldi S, Nosari S, Monteferrante G, Ruggeri ZM, Bresin E, Scheiflinger F, Rossi E, Martinez C, Coppo R, Remuzzi G, Noris M. In-vitro and in-vivo consequences of mutations in the von Willebrand factor cleaving protease ADAMTS13 in thrombotic thrombocytopenic purpura. Thromb Haemost 2006; 96: Camilleri RS, Cohen H, Mackie IJ, Scully M, Starke RD, Crawley JT, Lane DA, Machin SJ. Prevalence of the ADAMTS-13 missense mutation R1060W in late onset adult thrombotic thrombocytopenic purpura. J Thromb Haemost 2008; 6: Tao Z, Anthony K, Peng Y, Choi H, Nolasco L, Rice L, Moake JL, Dong JF. Novel ADAMTS-13 mutations in an adult with delayed onset thrombotic thrombocytopenic purpura. J Thromb Haemost 2006; 4: von Krogh AS, Kremer Hovinga JA, Tjønnfjord GE, Ringen IM, L ammle B, Waage A, Quist-Paulsen P. The impact of congenital thrombotic thrombocytopenic purpura on pregnancy complications. Thromb Haemost 2014; 111: de Vries PS, Boender J, Sonneveld MA, Rivadeneira F, Ikram MA, Rottensteiner H, Hofman A, Uitterlinden AG, Leebeek FW, Franco OH, Dehghan A, de Maat MP. Genetic variants in the ADAMTS13 and SUPT3H genes are associated with ADAMTS13 activity. Blood 2015; 125: Bongard T. Life History Strategies, Mate Choice, and Parental Investment Among Norwegians Over a 300-Year Period. Dissertation Doctor Scientiarum. Department of Biology. Trondheim: Norwegian University of Science and Technology, 2005, Milman N, Pedersen P. Evidence that the Cys282Tyr mutation of the HFE gene originated from a population in Southern Scandinavia and spread with the Vikings. Clin Genet 2003; 64: Veyradier A, Lavergne JM, Ribba AS, Obert B, Loirat C, Meyer D, Girma JP. Ten candidate ADAMTS13 mutations in six French families with congenital thrombotic thrombocytopenic purpura (Upshaw-Schulman syndrome). J Thromb Haemost 2004; 2: Levy GG, Nichols WC, Lian EC, Foroud T, McClintick JN, McGee BM, Yang AY, Siemieniak DR, Stark KR, Gruppo R, Sarode R, Shurin SB, Chandrasekaran V, Stabler SP, Sabio H, Bouhassira EE, Upshaw JD Jr, Ginsburg D, Tsai HM. Mutations in a member of the ADAMTS gene family cause thrombotic thrombocytopenic purpura. Nature 2001; 413: Anstee DJ. The relationship between blood groups and disease. Blood 2010; 115: Holmner A, Mackenzie A, Krengel U. Molecular basis of cholera blood-group dependence and implications for a world characterized by climate change. FEBS Lett 2010; 584: Yamamoto F, Cid E, Yamamoto M, Blancher A. ABO research in the modern era of genomics. Transfus Med Rev 2012; 26: