ORIGINAL ARTICLE Clinical haemophilia. Introduction

Transcription

1 Haemophilia (2014), 20, DOI: /hae ORIGINAL ARTICLE Clinical haemophilia Long-standing prophylactic therapy vs. episodic treatment in young people with severe haemophilia: a comparison of age-matched Danish and Russian patients J. INGERSLEV,* S. LETHAGEN, L. HVITFELDT POULSEN,* B. SØRENSEN,* E. LOPATINA, I. TENTSOVA, O. YASTRUBINETSKAYA and O. P. PLYUSHCH *National Haemophilia Center, Skejby University Hospital; National Haemophilia Center, University Hospital, Rigshospitalet, Copenhagen, Denmark; Center for Hematological Research, Moscow, Russia; and Center for Haemostasis and Thrombosis, St. Thomas Hospital, London, UK Summary. Two distinctly different substitution principles are commonly used in haemophilia: treatment at bleeding episodes only referred to as on-demand treatment, and prophylactic factor administration. The aim of the cross-sectional study which was undertaken in young patients suffering severe haemophilia A was to challenge our hypothesis that on-demand treatment is inferior to prophylactic substitution in prevention of chronic joint disease at young age. The method involved an investigation of 40 patients from Russia (n = 27) and Denmark (n = 13) born between 1975 and 1990 with no history of inhibitors; Russian patients had exclusively received factor VIII on demand, while Danish patients were managed with prophylactic treatment during a mean period of 16 years since median age of 5 years. The study endpoints were clinical joint scores, Quality of Life scores and functional independence scores. Matched by identical age (1 year) 13 Danish and 13 Russian patients were compared, while 14 age similar Russian patients served as controls. Demographic data among all groups were quite comparable. The results are that Russian patients presented with clinical joint scores at () while matched Danish counterparts scored (), differences being highly significant. The number of joint bleeds in recent 5 years were () vs (). Likewise, Quality of Life and functional independence scores were significantly higher in patients on prophylaxis as compared to on-demand treatment. In conclusion, the study outcomes confirmed our hypothesis. Longer term prophylactic factor administration during childhood and adolescence prevents joint destruction. Keywords: haemophilia, on demand, prophylaxis, substitution therapy, treatment Introduction In clinical management of persons suffering from severe haemophilia, substitution therapy with a factor VIII (FVIII) or factor IX concentrate is required to control bleeding episodes. For this purpose, safe substitution factor concentrates are abundantly available in many countries. However, careful management of bleeds alone may not protect the patient from development of progressive destruction of joints and Correspondence: Jørgen Ingerslev, MD, DMSc, Centre for Haemophilia and Thrombosis, Aarhus University Hospital Skejby, DK-8200 Aarhus N, Denmark. Tel.: ; Ingerslev@ki.au.dk Accepted after revision 5 July 2013 muscles leading to irreversible disability. With the aim to prevent the occurrence of spontaneous bleeds, treatment programmes have been implemented using frequent administration of factor concentrates that were introduced in some countries in northern Europe as early as the s. Initial data from the Haemophilia Centre in Malm o, Sweden, illustrated a highly protective effect of prophylactic treatment regimens consisting of IU kg 1 of FVIII administered three times weekly in patients with haemophilia A [1,2]. Importantly, in patients subjected to prophylactic factor replacement, joint destruction could largely be avoided. Follow-up several years later documented that an early start of prophylaxis from before the second year of life was more effective than if adopted later in childhood [3,4]. Consistent with the Swedish experience, likewise, Dutch haemophilia patients John Wiley & Sons Ltd

2 PROPHYLAXIS VS. ON-DEMAND TREATMENT 59 undergoing prophylactic treatment during childhood demonstrated quite minimal signs of arthropathy [4]. A direct comparison between the two programmes utilizing same frequency of factor supplementation, but applying different prophylactic doses, demonstrated that both groups of patients were very well protected while physical joint scores were more favourable using the Swedish model as compared to patients on the somewhat lower dosage regimen utilized in the Netherlands [5]. As most published reports on treatment outcomes of childhood prophylaxis were retrospective and uncontrolled, a Cochrane review published in 2005 [6] concluded that no convincing medical evidence had been detected in literature demonstrating a benefit of prophylactic factor treatment in haemophilia. However, a more recent prospective randomized controlled multicentre clinical trial compared prophylactic factor administration in early childhood in severe haemophilia [7] with episodic treatment at bleeding episodes only. Boys randomized to prophylaxis suffered significantly fewer bleeds and presented with a significantly better musculoskeletal outcome at the age of 6 years. Until now, no studies have attempted a direct comparison of the musculoskeletal outcomes in younger adult patients who had been offered prophylactic factor administration during upbringing with age-comparable patients who had exclusively been managed with episodic factor infusion at bleeding episodes. The purpose of the present investigation was to challenge a hypothesis that on-demand treatment conserves patient s physical capabilities less well than prophylactic factor administration during childhood and adolescence in patients with severe haemophilia A. The study was designed to compare physical health in Danish (DA) patients who had been subjected to prophylactic factor administration for greater than 5 years during growth with age-matched patients from a Russian (RU) haemophilia centre where prophylactic factor administration had so far not been utilized in children and young people. Study subjects were to be recruited among patients born between 1975 and Patients Study subjects aged years suffering from severe haemophilia A were enrolled at the Haemophilia Centre of Moscow, Russia, as well as in the Danish National Haemophilia Centres in Aarhus and Copenhagen. The study was granted approval by the Regional Biomedical Ethics Committee of Aarhus, Denmark (Project No ). Following written and spoken information, informed consent was obtained in each of 27 Russian (RU) and 13 Danish (DA) patients eligible for study. All patients suffered from severe haemophilia A (residual factor VIII:C < 0.01 IU ml 1 ) and no patient had a history of inhibitors. Among RU patients, 13 were found to be age-matched with 13 DA patients according to our preset matching criterion of identical year of birth (1 year). During their entire lifetime, RU patients had received treatment only to resolve acute bleeding episodes (on-demand treatment), but had never been subjected to a prophylactic factor infusion programme. In contrast, all DA patients had received regular prophylaxis for a substantial number of years during growth (Table 3). Methods Demographic characteristics such as year of birth, type and severity of haem5ophilia, height and body weight, as well as history of inhibitors were collected. In this study, the primary efficacy parameters selected were the following: musculoskeletal status as assessed by the Gilbert joint score [8] assessing six joints, and the Orthopaedic Issues score, which was derived from the Gilbert score in each individual by eliminating scores related to pain and bleeding; mobility status as assessed using the Functional Independence Score in Haemophilia (FISH) [9]; and Quality of Life was assessed utilizing the Hemo- QoL-A scoring system [10]. Historical demographic bleeding and treatment variables were collected such as age at first bleeding in life, age at first joint bleed and age at first treatment. In addition, the age at start of home treatment was recorded. The recent bleeding history was collected as annual number of bleeds during previous 5 years together with FVIII consumption per year during the same time period. In prophylactic treated persons, information on individual treatment doses and total annual amounts of concentrate consumed was collected. Observations were also made concerning possible development of target joints, here defined as 3 bleeds in same joint during any 6 month s period. Furthermore, information was recorded on surgical interventions aiming at improving patient s physical condition throughout his lifetime. Days of absence from work or school during most recent 6 months were also collected. All data gathered were retrieved from patient chart records in respective haemophilia centres, through musculoskeletal investigation, patient interviews and questionnaires. Patient weight and height were measured. Factor consumption during five previous years was collected either from patient s diaries or hospital pharmacy registries (Denmark). Physical joint scores were assessed by physicians in respective study sites. Language-specific protocols, questionnaires and Case Record Forms were used throughout. Statistics As only few essential observations were missing, an Intent-To-Treat analysis approach was adopted in the calculation of data. Pairs of matched patient scores on 2013 John Wiley & Sons Ltd Haemophilia (2014), 20,

3 60 J. INGERSLEV et al. efficacy variables: Gilbert score, Hemo-Qol-A and FISH scores were analysed using a t-test for paired differences among treatment groups by two-sided approach at a = In addition to the parametric t- test, Wilcoxon s non-parametric test was adopted to ensure the statistical significance in most of the group comparisons and in particular, to assess homogeneity among Russian patients matched with Danish patients vs. those not matched. All other study data were analysed using descriptive statistics. Results In total, 40 patients were enrolled, among whom 27 were RU patients and 13 were DA, all patients fulfilled the criteria of being born between 1975 and 1990, and suffering from severe haemophilia A with no history of inhibitors. Among RU patients, 13 were found to be age-matched with 13 DA patients according to our preset matching criterion of identical year of birth (1 year). All 27 RU patients were investigated according to the protocol and data utilized in statistical evaluation of homogeneity among all Russian patients enrolled to ensure that no selection bias was present. Demographic characteristics of study subjects are provided in Table 1, showing good comparability in terms of mean age and severity of haemophilia. Danish patients had slightly higher numerical body mass indexes compared to Russian patients, but the difference was not significant. Bleeding characteristics The age at earliest bleeding manifestations in life of patients are provided in Table 2. While bleeds appeared to occur almost at the same age in all patients, treatment was instituted somewhat earlier in the DA cohort compared with all RU cases, whether matched or not. Importantly, the access to home treatment was provided much earlier in DA cases (mean age 4 years) compared to RU patients (mean age 19 years). The mean annual number of joint bleeds in DA patients was significantly lower (1.6 bleeds per year) compared with Russian patients (40 bleeds per year). Treatment practices As shown in Table 3, all DA patients had received prophylactic factor concentrate for a mean time period of 16.4 years at the time of study from a mean age of 5 years. In two cases, prophylaxis was primary, i.e.: instituted before bleeding manifestations occurred, while 11 patients had secondary prophylaxis after a mean number of four joint bleeds. The median prophylactic infusion dose was 24 IU kg 1 b.w. and was administered three times per week or more often in 11 patients while two patients had two weekly injections only. Five patients had decided to stop their usual prophylaxis at a median age of 18 years. The mean annual consumption of FVIII was at IU in RU patients and IU in DA cases. Table 1. Demographic presentation of study participants. Patients studied Danish Age-matched Russian All Russian Numbers Residual FVIII:C <0.01 IU ml 1 <0.01 IU ml 1 <0.01 IU ml 1 FVIII inhibitors None None None detected Age, years, BMI, kg m 2, Primary outcome joint status The numerical outcome data on joint function (Fig. 1) illustrate that the mean values of joint scores were significantly lower in patients subjected to prophylaxis as compared to those who had received episodic treatment only, confirming our hypothesis with a very high significance, as assessed by the mean Gilbert scores (P<0.0001). Homogeneity of joint scores was found between matched RU and the entire population of RU patients participating in the study, pointing out that selection bias was most likely not present. In addition, Table 2. Early and recent bleeding characteristics. Danish Age-matched Russian All Russian N Early bleeding and treatment history Age at first bleed, months, Age at first joint bleed, months Age at first treatment, months, Home treatment since age, years No. target joints in the past/no. patients 2/2 28/10 76/24 Recent 5 years bleeding history All bleeds during last 5 years (9) (217) (217) Joint bleeds during last 5 years (5) (186) (175) Haemophilia (2014), 20, John Wiley & Sons Ltd

4 PROPHYLAXIS VS. ON-DEMAND TREATMENT 61 Table 3. Treatment practices utilized in Danish and Russian patients. Danish patients Russian Patients All Russian Patients N Schedule Prophylaxis On demand On demand FVIII use/year last 5 years Prophylactic treatment data Prophylaxis (duration), years (15.0) Prophylactic dose, IU kg (24.0) Prophylaxis discontinued (n = 5) at age, years (18.0) Fig 1. Gilbert scores (left panel) and orthopaedic issue scores (Gilbert scores corrected for bleeding and pain scores) (right panel) recorded in 13 Danish patients and 13 age-matched Russian patients. For comparison, the scores obtained in a total of 27 Russian patients are provided. orthopaedic issue scores were also significantly higher in on-demand treated cases compared with those using prophylaxis. The presence of a target joint was reported in 10 of 13 RU matched patients with 28 historical target joints in this group, and 76 target joints in the entire cohort of RU patients, while only two among 13 DA patients had suffered one target joint episode each. Surgical interventions were conducted on nine occasions in 13 matched RU patients and 23 surgical episodes were reported in the entire group of 27 RU participants with the aim to improve their physical condition. No orthopaedic surgeries had been required in any DA patient. Psycho-social characteristics As determined by Hemo-QoL A using Russian and Danish language translations, respectively, quality of life responses were recorded in all patients entering the study. Results are outlined in Table 5, showing a statistically significant higher overall Hemo-Qol score in patients on prophylaxis in comparison with those using on-demand treatments. Results of the functional independence interview (Table 6) also points to more requirements for assistance among RU patients compared to DA patients. Among all 40 persons investigated three had no occupation i.e. employed or receiving an education, in society: three RU patients who were severely disabled and one DA patient because of HIV-related health problems. Days off from school or work were reported in 31 of 40 study subjects only. Among matched RU patients, 4 of 13 (44%) had >2 days of absence from work/school, altogether 190 days during previous 6 months. In comparison, among DA subjects one had 2 days off work/school and one patient had 1 day off, while seven reported that they had no absence at all from school or work during the past 6 months. In addition, among DA patients 5 of 13 (38%) had 1 visit and two had >3 visits to the hospital in the last 6 months, while among RU patients 6 of 13 (46%) had 1 visit and five had >4 visits during previous 6 months. Discussion The working hypothesis of this study was that the outcome of on-demand treatment with factor replacement was thought to be inferior to prophylactic factor administration. We investigated a total of 40 patients born between 1975 and 1990 suffering severe haemophilia A having no history of inhibitors. Among these, 27 lived in Russia and 13 in Denmark. Important differences exist between the treatment technologies adopted in the respective countries, as RU patients had received treatment with FVIII only in management of actual bleeds while DA patients had received prophylactic FVIII administration for several years 2013 John Wiley & Sons Ltd Haemophilia (2014), 20,

5 62 J. INGERSLEV et al. ( years). Among all patients, 13 RU patients were found to match with 13 DA patients by age identical criterion (1 year), and this group was explored to challenge the hypothesis. The Gilbert joint score constituted the primary endpoint of the investigation and the outcome finding was that patients in the age-matched RU group (n = 13) presented with a mean score at while the corresponding figure in DA patients was at The mean Gilbert score in all RU patients (n = 27) was found at Based on these data and their corresponding significance evaluation (Fig. 1, Table 4) the hypothesis was definitively confirmed. Comparing DA patients with all RU patients, significance was equally strong, indicating that selection bias could be ruled out. Likewise, omitting the pain and bleeding questions from the Gilbert Score here reported as orthopaedic issues did not change the strength of significance between the group on prophylaxis and the on-demand treatment groups (Table 4). It should be underlined, however, that an important difference exists between treatment practices in DA and RU patients that may have influenced the outcome of the study; home treatment was instituted at a median age of 4 years in DA patients while home therapy was set forth from a median age of 19 years in RU patients. Early access to quick management of a new bleed in the home setting may represent an additional benefit in prevention of chronic joint disease. However, at present time we have no controlled studies demonstrating the magnitude of importance of home treatment, and a recent review article dealing with various aspects of home treatment offered no clear recommendation [13]. Only a few studies exist that address the relationship between prophylactic treatment and on-demand management. A retrospective cross-sectional comparison of Norwegian patients managed exclusively by an on-demand principle (median age 35) was compared to a Swedish group of patients who had been utilizing prophylactic treatment for most of their lives (median Table 4. Joint scores and orthopaedic issues recorded in Danish and Russian patients. Patients studied Danish Age-matched Russian All Russian Numbers Gilbert score, (2) (29) (30) P-value < < P -value < < < Orthopaedic (1) (18) (18) issues*, P-value < < P-value < < < Median values provided in brackets. *Orthopaedic issues calculated by subtracting scores on bleeding and pain from the total Gilbert score. age 23 years) [11]. Although no assessment of joint disease using a formal scoring system was utilized, indirect evidence for a higher incidence of joint disease in the Norwegian patients was witnessed by the number of orthopaedic procedures required; 80 surgeries had been performed in the 61 Norwegian patients as compared to 9 in the 95 Swedish patients. Another retrospective study compared Dutch patients [12] among whom the majority had been using prophylactic factor administration with a large group of French patients who had predominantly been using ondemand treatment. Patients were of quite comparable age (mean 22.3 years) and age distribution. In this study, a clinical joint score of 2.0 (mean) was noted in Dutch patients while French patients had a mean score of 8.0. The Dutch-French study was not designed as a pure prophylaxis vs.-on-demand study as deviations in both directions were mentioned. A rather favourable joint score found in French patients may have been caused by extended episodic treatment, shorter periods of prophylactic treatment, or early instalment of home therapy. Discussing these two studies in the context of this study, the Swedish-Norwegian study identified 14 times more orthopaedic surgery in the on-demand group, which should be compared with our finding of nine orthopaedic surgery episodes among 13 RU patients while none had been required in the 13 DA patients. The controlled, randomized study of Manco-Johnson et al. [7] clearly demonstrated an improved physical outcome of haemophilia up to the age of 6 years following early onset prophylaxis. From older as well as more recent studies [e.g. 14] we may anticipate that haemophilic joint deterioration in children is accelerating from around the age of 10 and onwards. A median joint score of 2 in DA patients appears to be a quite favourable result and concordant with those observed in younger haemophilia patients in the Netherlands and Sweden. However, one could criticize that is was not found to be at zero level. A reason may well be, that prophylaxis was started relatively late at around 5 years following a median number of four joint bleeds in the DA group of patients enrolled. The mean annual number of joint bleeds in RU patients was 40 bleeds while DA patients had 1.6 joint bleeds only per year. In consequence of joint-related disability, quality of life measures in patients receiving episodic replacement therapy gave distinctly lower results as compared to those obtained in patients who had received prophylactic administration (Table 5). Likewise, functional independence scores were significantly higher in prophylactic treated patients as opposed to scores obtained in patients receiving on-demand treatment (Table 6). A weakness of this study was the low number of patients recruited in the Danish Centres. Enrolment Haemophilia (2014), 20, John Wiley & Sons Ltd

6 PROPHYLAXIS VS. ON-DEMAND TREATMENT 63 Table 5. Quality of Life estimates (Hemo-QoL% of maximal score) in Danish and Russian patients. Treatment N Mean SD Median P-value (t-test) On demand (RU) * Prophylaxis (DK) Difference On demand (RU), not matched patients *P-value for unpaired two sample t-test for equality of means (on demand vs. prophylaxis). P-value for paired t-test for equality of means (on demand vs. prophylaxis). P-value for unpaired two sample t-test for equality of means (matched vs. unmatched RU patients). Table 6. Functional independence (FISH) scores obtained in Danish and Russian patients. FISH score N Mean SD Median P-value (t-test) On demand (RU) <0.0001* Prophylaxis (DK) Difference On demand (RU), not matched patients *P-value for unpaired two sample t-test for equality of means (on demand vs. prophylaxis). P-value for paired t-test for equality of means (on demand vs. prophylaxis). P-value for unpaired two sample t-test for equality of means (matched vs. unmatched RU patients). was offered to all patients meeting the age criterion but most unfortunately, recruitment was met with difficulties because of lacking interest among patients. However, as data so clearly demonstrate that the primary outcome measure of joint damage was statistically different between the groups with no overlapping values, the low number of DA patients is less critical. Furthermore, the joint status of Danish patients detected in this study was quite identical to those identified in a joint Dutch-Danish study that enrolled 22 Danish patients who had been subjected to prophylactic treatment during upbringing [15]. Conclusions This collaborative study was established to characterize age-matched severe haemophilia A patients (age group years) in Russia (n = 13) receiving exclusively on-demand treatment for their bleeds and Danish haemophilia A patients (n = 13) who had been managed with prophylactic FVIII administration for close to 16 years since their fifth year of life. The primary outcome characteristic of preserved joint function was assessed by clinical joint scoring that revealed a highly significant difference (P ) between the two groups in favour of prophylactic treatment, having a median Gilbert score of 2 vs. a median score at 29 in the age-matched group receiving on-demand treatment only. Corrected for pain and bleeding, the net median joint scores were 1 and 18, respectively, with an equally strong significance. The median annual number of joint bleeds in patients on prophylaxis was 1 while patients managed with ondemand treatment suffered a median of 37 joint bleeds per year. The quality of life estimate further showed a statistical significant advantage of prophylaxis (P < 0.001) and patient s functional independence significantly better than the corresponding value for the on-demand treated group. It should be underlined, however, that patients receiving prophylactic substitution therapy commenced home treatment very early in life (mean age 4 years), while those receiving on-demand therapy adopted home treatment much later in life (mean age 19 years). The favourable long-term outcome identified is most probably influenced by the concerted effect of home therapy and prophylaxis. In conclusion, this relatively pure comparison of on-demand and prophylactic treatment demonstrates that prophylactic treatment offers a considerable guarantee for a well-conserved physical joint health up to a median age of 23 years with much less bleeds, pain and physical discomfort. Longer term studies should aim at re-investigating adult patients who had prophylaxis during childhood and adolescence as they reach a more advanced age. Acknowledgements We wish to acknowledge the efforts of our haemophilia centre colleagues and nurses for their outstanding help during the study. This project was supported by an unrestricted grant offered by Bayer Health Care who also provided logistic help. Disclosures Dr. Ingerslev has received honoraria as a speaker on behalf of Bayer, Novo Nordisk and Baxter in the past and been a member of Advisory Committees for Bayer and Novo Nordisk. The current study received initial financial support from Bayer to cover for staff hours. L. Hvitfeldt received fees for speaking at Baxter symposium. S. Lethagen received fees for speaking or acting as chairman at meetings from Bayer, Baxter, ZLB Behring and Novo Nordisk. He received a research grant from CSL Behring. He is a fulltime employee and holds shares at Novo Nordisk. B. Sorensen has acted as a paid consultant for Baxter, Bayer, CSL and Novo Nordisk. He was an employee of Baxter from May 2012 to October E. Lopatina, I. Tentsova, O. Yastrubinetskaya, and O. Plyushch stated that they had no interests which might be perceived as posing a conflict or bias. References 1 Nilsson IM, Hedner U, Ahlberg A. Hemophilia prophylaxis in Sweden. Acta Paediatrica Scand 1976; 65: Nilsson IM, Berntorp E, L ofqvist T, Pettersson H. Twenty-five year s experience of prophylactic treatment in severe haemophilia A and B. J Int Med 1992; 232: Astermark J, Petrini P, Tengborn L, Schulman S, Ljung R, Berntorp E. Primary prophylaxis in severe haemophilia should be started at an early age but can be individualised. Br J Haematol 1999; 105: John Wiley & Sons Ltd Haemophilia (2014), 20,

7 64 J. INGERSLEV et al. 4 Fischer K, van der Bom JG, Mauser-Bunschoten EP et al. The effects of postponing prophylactic treatment on long-term outcome in patients with severe haemophilia. Blood 2002; 99: Fischer K, Astermark J, van der Bom JG et al. Prophylactic treatment for severe haemophilia: comparison of an intermediatedose to a high-dose regimen. Haemophilia 2002; 8: Stobart K, IorioA, Wu JK. Clotting factor concentrates given to prevent bleeding and bleeding -related complications in people with haemophilia A or B. Cochrane Database Syst Rev 2005; 2: CD Manco-Johnson MJ, Abshire TC, Shapiro AD, Riske B, Hacker MR, Kilcoyne R, Ingram DJ, Manco-Johnson ML, Funk S, Jacobson L, Valentino LA, Hoots WK, Buchanan GR, DiMichele D, Recht M, Brown D, Leissinger C, Bleak S, Cohen A, Mathew P, Matsunaga A, Medeiros D, Nugent D, Thomas GA, Thompson AA, McRedmond K, Soucie JM, Austin H, Evatt BL. Prophylaxis versus episodic treatment to prevent joint disease in boys with severe hemophilia. NEJM 2007; 357: Gilbert MS. Musculoskeletal complications of haemophilia: the joint. Haemophilia 2000; 6(suppl. 1): Poonnoose PM, Manigandan C, Thomas R et al. Functional Independence Score in Haemophilia: a new performance-based instrument to measure disability. Haemophilia 2005; 11: Rentz A, Flood E, Altisent C et al.; Members of the HAEMO-QoL-A Steering Committee. Cross-cultural development and psychometric evaluation of a patient-reported healthrelated quality of life questionnaire for adults with haemophilia. Haemophilia 2008; 5: Steen Carlsson K, H ojgaard S, Glomstein A et al. On-deman vs prophylactic treatment for severe haemophilia in Norway and Sweden: differences in treatment characteristics and outcome. Haemophilia 2003; 9: Fischer K, van der Bom J, Molho P et al. Prophylactic versus on-deman treatment strategies for severe haemophilia: a comparison of costs and long-term outcome. Haemophilia 2002; 8: Teitel J, Barnard D, Israels S, Lillicrap D, Poon M-C, Sek J. Home management of haemophilia. Haemophilia 2004; 10: SaulyteTrakymiene S, Ingerslev J, Ragaliene L. Utility of the Haemophilia Joint Health Score in study of episodically treated boys with severe haemophilia. Haemophilia 2010; 16: van Dijk K, Fischer K, van der Bom JG, Scheibel E, Ingerslev J, van den Berg HM. Can long-term prophylaxis for severe haemophilia be stopped in adulthood? Results from Denmark and the Netherlands. Br J Haematol 2005; 130: Haemophilia (2014), 20, John Wiley & Sons Ltd