Prophylaxis in the haemophilia population



    1. Division of Hematology/Oncology, Hospital for Sick Children, Toronto
    2. Deparmtent of Pediatrics, University of Toronto, Toronto, ON, Canada
    Search for more papers by this author

Dr Victor S. Blanchette, Division of Hematology/Oncology, Hospital for Sick Children, 555 University Avenue, Toronto, ON, Canada M5G 1X8.
Tel.: +1 416 813 5852; fax: +1 416 813 5327;


Summary.  Prophylaxis is recommended as preventive therapy for young boys with severe haemophilia in countries where safe factor concentrates are available. This recommendation is supported by results from a randomized, controlled study that compared on-demand therapy with full-dose prophylaxis (Manco-Johnson MJ, Abshire TC, Shapiro AD et al. N Engl J Med 2007;357:535). It is important to distinguish primary vs. secondary prophylaxis. Primary prophylaxis refers to preventive treatment started before the onset of joint damage, whereas secondary prophylaxis refers to treatment started after joint damage has occurred. Whereas the benefits of primary prophylaxis are well documented, data relating to secondary prophylaxis are limited, especially in the adolescent/adult haemophilia population. Failure of prophylaxis may relate to several variables, including: (i) underlying status of the joints; (ii) poor compliance; (iii) participation in high-risk activities and (iv) unfavourable pharmacokinetics (PK), i.e., too rapid elimination of infused coagulation factors. There is evidence that the risk of joint bleeding in individuals with severe haemophilia A relates to time spent with factor levels < 1% (Collins PW, Blanchette VS, Fischer K et al. J Thromb Haemost 2009;7:413); this variable is strongly influenced by frequency of factor infusions and the individual’s PK profile. Key ongoing questions relating to prophylaxis include: (1) what is the optimal regimen for initiating primary prophylaxis; (2) role of prophylaxis in the adolescent/young adult haemophilia population and (3) role of prophylaxis in individuals with severe von Willebrand’s disease and other rare inherited coagulation disorders. The role of novel long-acting factor concentrates for prophylaxis will also need to be evaluated.


Prophylaxis, derived from the Greek work prophulaktikos, relates to the prescription of medicine or a course of action tending to prevent disease or other misfortune [1]. This literary definition is apt in the context of the disorder haemophilia. This review will update previous reviews of prophylaxis published following World Federation of Haemophilia Congresses in 2004, 2006 and 2008 [2–4].

Definitions of prophylaxis

Prophylaxis is defined as ‘treatment by intravenous injection of factor concentrates in anticipation of and in order to prevent bleeding’ [5]. In this context, the administration of factor concentrates prior to surgery constitutes prophylaxis; however, the most common use of factor prophylaxis in the haemophilia population, and the one discussed in this review, is the use of long-term prophylaxis to prevent arthropathy. An important and still contentious matter is the definition of primary vs. secondary prophylaxis. Definitions were proposed at a Consensus Conference on prophylaxis held in London, UK in 2002 [5] and have since been updated by the European Paediatric Network for Haemophilia Management (PEDNET) (Table 1). These definitions, although useful, merit reconsideration. As joint damage can occur after only a very few bleeds, and because it is recognized that some joint bleeding is subclinical [7], it may be appropriate to define primary prophylaxis as the regular infusion of factor concentrates started before the occurrence of joint damage and with the intent of administering prophylaxis continuously, defined as >45 weeks year−1 [8]. This definition incorporates the elements of both the underlying joint status and duration of prophylaxis and distinguishes primary prophylaxis from on-demand treatment and short-term prophylaxis that may be used in individuals with haemophilia and target joint bleeding. If this definition of primary prophylaxis is accepted, secondary prophylaxis would refer to prophylaxis started after the onset of objectively determined joint damage and with the intent of administering prophylaxis continuously defined as >45 weeks year−1 [8].

Table 1.   Definitions of primary and secondary prophylaxis [6].
Primary prophylaxis A
Regular continuous treatment started after the first joint bleed and before the age of 2 years
Primary prophylaxis B
Regular continuous treatment started before the age of 2 years without previous joint bleed
Secondary prophylaxis A
Regular continuous (long-term) treatment started after two or more joint bleeds or at an age >2 years
Secondary prophylaxis B
Intermittent regular (short-term) treatment, because of frequent bleeds

Pathogenesis of hemophilic arthropathy

The pathogenesis of haemophilic arthropathy is increasingly better understood. Older studies, involving careful clinical and pathological observations in individuals with haemophilia, established that recurrent bleeding into joints results in a destructive arthropathy that is often painful and disabling [9,10]. Recent studies, including in vitro studies and studies in animals, have provided insights into the complexity of haemophilic arthritis [11–14]. Even though the exact mechanisms that result in haemophilic arthropathy remain to be fully elucidated, it appears that haemoglobin-derived iron released from erythrocytes following recurrent bleeding into joints results in an inflammatory synovitis that leads to cartilage damage and bone destruction. A number of inflammatory mediators are involved in this process, and angiogenesis, induced by growth factors such as vascular endothelial growth factor is key to the development of synovitis and resultant joint damage. Of interest, there is evidence from in-vitro studies to suggest that immature articular cartilage may be more susceptible to blood induced damage than mature articular cartilage [15].

An understanding of the consequences of acute bleeding into joints may be very important in the design of optimal prophylaxis regimens. Based on the results of experimental studies of blood induced joint damage [11,12,14], it is possible that enhanced episodic therapy for breakthrough bleeding in young boys with severe haemophilia started on primary prophylaxis regimens, as given in the US Joint Outcome Study and the Canadian dose-escalation primary prophylaxis study may be important with respect to preventing subclinical or overt joint bleeding (i.e., rebleeding) following an acute joint bleed [7,16]. This possibility is supported by studies that demonstrate that wound healing is abnormal in mice with haemophilia B and suggests that ongoing coagulation function needs to be maintained to limit bleeding into granulation tissue during tissue remodelling [17]. It is possible that the ‘inflammatory storm’ and stimulation of new blood vessel formation (angiogenesis) that occurs as a result of acute bleeding into a joint may act as a risk factor for subclinical bleeding and rebleeding into the affected joint. Adequate clotting factor cover during this immediate ‘at-risk’ period following an acute joint haemorrhage may therefore be important in ensuring an optimal long-term musculoskeletal outcome.

Prophylaxis studies with long-term follow-up

The field of prophylaxis owes a great debt to the pioneering studies of Professor Inga Marie Nilsson and her colleagues from Malmö, Sweden and Professor van Creveld and his co-workers in Utrecht, the Netherlands. These two groups began programmes of prophylaxis in boys with severe haemophilia in the late 1950’s/1960’s, the results of which have been reported after more than two decades of careful follow-up [18–21]. In both haemophilia treatment centres, prophylaxis was started in boys with a history of some joint bleeding (i.e., secondary prophylaxis), but evolved to programmes where factor infusions were given before, or after a very few, clinically reported joint bleeds. The two prophylaxis programmes differed significantly with respect to age at introduction of prophylaxis and intensity of regimen, as described below.

Swedish high-dose (‘Malmö’) prophylaxis protocol

In Sweden, prophylaxis was given as high-doses of factor VIII (FVIII) (25–40 IU kg−1) on alternate days, minimum three times per week for haemophilia A patients and 25–40 IU kg−1 of factor IX (FIX) twice weekly for haemophilia B cases. Prophylaxis was initiated at a very young age (generally ≤2 years) or at the time of the first index joint bleed if this occurred before age 2 years. The ‘high-dose’ Malmö prophylaxis protocol was aimed to maintain trough FVIII and FIX levels above 1%. Breakthrough joint bleeds were treated with one or more infusions of FVIII or FIX (25–40 IU kg−1) according to severity and until bleeding had stopped. It was recommended that prophylaxis continue life-long.

Dutch intermediate-dose prophylaxis protocol

In the Netherlands, prophylaxis was started at an early age according to the individual’s bleeding pattern, generally after the occurrence of at least one or two joint bleeds. The Dutch regimen involved the administration of 15–25 IU kg−1 of FVIII two or three times a week for haemophilia A cases, and 30–50 IU kg−1 of FIX once or twice a week for haemophilia B cases. The intensity of prophylaxis was adjusted based on spontaneous breakthrough bleeding into joints and not increased according to the subject’s body weight alone. Trough levels of FVIII or FIX were not taken into consideration when adjusting prophylactic treatment. It was recommended that prophylaxis continue throughout adulthood.

Canadian dose-escalation primary prophylaxis protocol

The Canadian dose-escalation primary prophylaxis study was started in 1997. In this single arm, prospective study, boys’ ages 1 year to 30 months with severe haemophilia A, no evidence of a circulating inhibitor to FVIII and absence of any overt joint disease were started on once weekly infusions of FVIII (50 IU kg−1). If clinically significant bleeding into muscles and/or joints occurred, the frequency of FVIII infusions was increased to twice weekly (dose 30 IU kg−1); continuation of bleeding resulted in escalation of the prophylaxis regimen to 25 IU kg−1 on alternate days. Criteria for escalation included: ≥ 3 clinically determined bleeds into any one joint over a consecutive 3-month period; ≥ 4 significant soft tissue/joint bleeds over a consecutive 3-month period and ≥ 5 bleeds into any one joint while on the same dosage (step) of factor therapy over any period of time. The interim results of this study have been reported [16], and 10-year follow-up results were presented at the 2009 International Society on Thrombosis and Hemostasis Congress [22]. The Canadian primary prophylaxis study is now closed to patient accrual, but follow-up of enrolled cases is ongoing.

Key results from these three long-term prophylaxis studies are as follows:

  • 1Compared to on-demand therapy, intermediate-dose prophylaxis (the Dutch protocol) started at an early age in boys with severe haemophilia results in significantly fewer joint bleeds, a better joint status and a more favourable health-related quality of life [20].
  • 2Compared to the intermediate-dose Dutch prophylaxis regimen, the Swedish high-dose prophylaxis regimen is associated with a significantly lower rate of joint bleeding [21]. However, the FVIII consumption, and therefore cost, is approximately twofold higher for the high-dose prophylaxis regimen and, at least after a follow-up period of nearly 20 years, the extent of haemophilic arthropathy measured by a radiologic scale is clinically similar for the two prophylaxis regimens (Table 2).
  • 3The percentage of boys with severe haemophilia A on once weekly, twice weekly and alternate day treatment in the Canadian dose-escalation cohort at 10 years from the start of the study is 37% (21/56 cases), 34% (19/56 cases) and 29% (16/56 cases), respectively. The median follow-up for the cohort at the time of analysis was 4.6 years (range 0.25–10 years).
Table 2.   Comparison of an intermediate-dose to a high-dose prophylaxis regimen: long-term follow-up results.
 Dutch regimen (‘intermediate-dose’)Swedish regimen (‘high-dose’)P-value
  1. Values are medians; values in parentheses are interquartile ranges [21].

Number of cases4218 
Age at start of prophylaxis (years)4.6 (3.1–6.2)1.2 (0.8–1.7)<0.001
Number of joint bleeds per year3.7 (1.7–5)0.2 (0–0.3)<0.001
Pettersson score0 (0–5)0<0.001
Annual clotting factor use (IU kg−1 year−1)2126 (1743–2755)4616 (4105–5571)<0.001

The results of the retrospective Swedish and Dutch cohort studies continue to be debated. At the centre of this debate is the issue of when should primary prophylaxis be started in boys with haemophilia A? Data from the Swedish and Dutch studies suggest that primary prophylaxis should be started at an early age but can be individualized based on the bleeding pattern in the individual child [23,24]. In the Dutch cohort studies, an early start to prophylaxis resulted in complete prevention of joint damage for 70% of boys compared with 31% for boys who started prophylaxis after 3 or more bleeds [24]. After two decades of follow-up, the radiological (Pettersson) joint score was 8% higher for every year prophylaxis was postponed after the first joint bleed [24]. Based on the results of long-term prophylaxis studies reported to date, it seems reasonable to commence primary prophylaxis in boys with severe haemophilia A after 1–2 joint bleeds using an infusion frequency of at least once weekly. An advantage of a once weekly infusion protocol to initiate a programme of primary prophylaxis is the opportunity to avoid the need for a central venous access line in a majority of cases [25].

Tailored dosing in prophylaxis

The rationale for the Swedish high-dose prophylaxis (‘Malmö’) regimen was the observation, reported by Ahlberg in 1965, that patients with moderate forms of haemophilia A or B, i.e., with a FVIII or FIX level of 1–5%, experienced few spontaneous joint bleeds and rarely developed clinically significant arthropathy [10]. This led to the hypothesis that if the plasma level of FVIII or FIX could be artificially kept at, or above, 1% in severe haemophilia A or B cases, it should be possible to convert the severe to a moderate bleeding phenotype with a significant reduction in spontaneous joint bleeding and bleed-associated arthropathy.

The value of the 1% FVIII threshold as a risk factor for spontaneous bleeding into joints has generated lively debate. In a study of 51 patients with haemophilia A and 13 with haemophilia B, Ahnström and colleagues found only a weak correlation between trough FVIII and FIX levels and the incidence of joint bleeding, even after stratification of the patients according to joint score [27]. Some patients did not bleed in spite of a trough level of <1%, and others did in spite of trough levels >3%. The investigators concluded that a standard prophylaxis regimen should be implemented only after careful clinical consideration and with a high readiness for re-assessment and individualized dose tailoring. Collins and co-workers have provided statistical evidence that the risk of hemarthrosis in both children and adults is associated with the time per week an individual spends with a FVIII below 1% [28]. The investigators cautioned, however, that even though the data implied that the risk of breakthrough bleeding on prophylaxis is increased as the time spent with a low FVIII level increases, it should not be interpreted as a confirmation that a factor level of 1% is a critical threshold above which bleeds are prevented and below which bleeds occur [28]. Several factors likely influence the bleeding patterns seen in individuals with severe haemophilia, including the individual PK profile of the patient, the musculoskeletal status of the underlying joint and the patients’ activity profile. As an example, it is well known that target joints, defined as joints with active synovitis as a result of frequent bleeding, bleed more than normal joints and may require higher threshold levels of FVIII or FIX to suppress recurrent bleeding [29]. Subsequently pharmacokinetic data on FVIII from 147 individuals with haemophilia A (48 children ages 1–6 years of age and 99 individuals ages 10–65 years of age) were used for simulations of commonly used prophylactic regimens to calculate their effect on FVIII levels during prophylaxis [30]. The results of the simulations demonstrated that individual half-life of infused FVIII and frequency of dosing have a much larger effect on FVIII trough levels and time per week with FVIII levels <1% than recovery and infused dose.

Given the significant variation of individual patients’ FVIII and FIX pharmacokinetic profiles, attention to frequency of infusions should allow a more cost-effective use of FVIII and FIX in prophylaxis regimens. The concept that pharmacokinetically tailored dosing of FVIII and FIX could result in considerable savings of factor concentrates compared to standard (‘fixed’) prophylaxis protocols is supported by publications of Carlsson, Björkman, Berntorp and co-workers [30–33]. A challenge to PK directed therapy that would allow easy alteration in prophylaxis regimens to achieve, for example, higher threshold (‘trough’) FVIII or FIX levels, is the perceived need to perform very demanding conventional PK studies on individual patients. This problem can be overcome by using Bayesian PK analysis, utilizing a population pharmacokinetic model that allows a sparse blood sampling protocol [34].

Prophylaxis in the adolescent/adult haemophilia population

Use of prophylaxis in the late adolescent/adult haemophilia population is increasing particularly in countries with unrestricted access to safe FVIII and FIX concentrates [35,36]. As an example in a recent survey of 2663 persons with haemophilia A or B followed in Canadian Comprehensive Care Hemophilia Treatment Centres, 53% of individuals with severe haemophilia A and 20% with severe haemophilia B >18 years of age were identified to be receiving prophylaxis defined as the infusion of FVIII or FIX at least once weekly for >45 weeks during the year 2006 [37].

An important question, in the context of prophylaxis use in the adult haemophilia population, is whether prophylaxis can be safely discontinued in individuals who have been receiving intermediate or full-dose prophylaxis from an early age of life. Data reported from Denmark and the Netherlands are instructive in this regard. In Denmark, patients with severe haemophilia are treated using the high-dose Swedish prophylaxis protocol, whereas the Dutch patients, as described earlier in this review, receive an intermediate-dose prophylaxis regimen. Of a total 49 Dutch patients who received intermediate-dose prophylaxis from an early age in life, 11 (22%) were able to permanently discontinue prophylaxis [38]. The median age of the cohort was 23.4 years at the time of analysis, and the median follow-up off prophylaxis was 3.2 years. The median number of joint bleeds in the last year of follow-up was 3.7. Of interest, the patients who were able to permanently discontinue prophylaxis appeared to have a milder bleeding phenotype as evidenced by having a later start of prophylaxis, requiring a lower weekly dose of factor replacement and experiencing a lower joint bleed frequency on prophylaxis [38]. In Denmark, 10 of 22 cases (45%) studied at a median age of 26.2 years were able to permanently discontinue prophylaxis [39]. Long-term studies are now required to determine the musculoskeletal consequences of discontinuing long-term factor prophylaxis in early adulthood.

Use of secondary prophylaxis in adults with severe haemophilia is increasing in countries with access to safe FVIII and FIX concentrates. This practice is supported by results from prospective studies. In a longitudinal study of 477 patients under 25 years of age with severe haemophilia A, prophylaxis for >45 weeks year−1 significantly reduced the rate at which joints deteriorated both on physical and X-ray examinations [8]. Patients on long-term prophylaxis had significantly fewer days lost from work or school as well as fewer days spent in hospital. The investigators concluded that ‘a haemophiliac may well be better served with prophylaxis as the treatment regimen’ [8]. Recently, Collins and colleagues have reported the results of a cross-over study comparing on-demand treatment with full-dose prophylaxis in 20 adults ages 30–45 years of age with severe haemophilia A and an average of two bleeds per month [40]. Subjects received on-demand treatment for 6 months and were then switched to a high-dose prophylaxis regimen for 7 months. The first month of prophylaxis was considered a run-in period to allow stabilization on the prophylaxis regimen. Compared to on-demand treatment, prophylaxis was associated with a significant reduction in the frequency of joint bleeds (median 0 vs. a median of 15 for on-demand treatment). Of note, postinfusion FVIII trough levels were >5% at 48 h in 75% of cases and ≥2% at 72 h for 57% of patients. These studies provide a foundation in support of secondary prophylaxis in adults, and suggest that, in general, adults will require lower total doses of FVIII or FIX compared with children to maintain equivalent trough FVIII/FIX levels. The lower dose requirement in adults is predictable from the fact that FVIII half-life tends to be longer in adults than in small children 41, 42). Stabilization of clotting factor consumption in adulthood for subjects who receive early intensive prophylaxis has been reported by Dutch investigators [43].


The beneficial role of primary prophylaxis in young boys with severe haemophilia can no longer be questioned. The results of the USA Joint Outcome Study reported by Manco-Johnson and colleagues in 2007 [7] provide definitive evidence that primary full-dose prophylaxis is superior to on-demand therapy based on MRI detected osteochondral changes in index joints by age 6 years [7]. This randomized controlled trial has effectively silenced doubts about the benefits of prophylaxis raised by a 2006 Cochrane Collaboration review [44]. There is now global consensus that primary prophylaxis, started at a young age before the onset of overt joint disease, should be regarded as standard of care for boys with severe haemophilia A in countries where there is reliable access to safe FVIII concentrates. It is not possible to make a definitive statement for boys with severe haemophilia B as the majority of data regarding primary prophylaxis in the haemophilia population have been obtained from studies in patients with haemophilia A. This fact, together with the belief by some that the bleeding profile in patients with haemophilia B may be less severe than in comparable subjects with severe haemophilia A, may offer an explanation for the observation that fewer severe haemophilia B cases are placed on long-term primary prophylaxis, started at an early age of life, than equivalent patients with haemophilia A [37]. Well-designed long-term studies of prophylaxis in boys with haemophilia B are urgently needed.

The role of secondary prophylaxis remains to be defined. The benefits of secondary prophylaxis started in adolescent and adult haemophiliacs are very encouraging but, as with primary prophylaxis, prospective long-term studies are needed [45]. These studies should incorporate a battery of outcome measures such as objectively determined musculoskeletal disease and health-related quality of life measures [46].

A very important sub-group of patients are those with high-titre inhibitors to FVIII or FIX. Many of these cases are young boys with relatively good joint status. Approximately, two-thirds of subjects with high-titre inhibitors to FVIII can be rendered responsive to infused FVIII following a programme of immune tolerance induction (ITI) therapy. During the period of ITI, which in many cases may extend beyond one year, it may be very important to initiate a programme of prophylaxis with by-passing agents, either FEIBA or recombinant factor VIIa, in boys who manifest target joint bleeding.

Future directions

The greatest barriers to more widespread use of prophylaxis in young boys with severe haemophilia are the very high cost of this treatment approach and the challenge of venous access in very young boys started on full-dose prophylaxis. A possible solution may come from long-acting FVIII or FIX products, many of which are now in an advanced stage of development, and some of which have entered clinical trials. Given the anticipated degree of variability in PK profiles that is likely to be seen between individuals who are treated with these novel products, it will be important to consider PK directed therapy, perhaps using sparse blood sampling and Bayesian pharmacokinetic analysis. The impact of differences in half-life on time spent below a certain plasma factor level might be exacerbated with a longer half-life of infused clotting factors. Consequently, controlling target plasma levels may become more important as a means to reduce spontaneous bleeding into joints thus assuring a good long-term musculoskeletal outcome.


The author is grateful to Professors Sven Björkman, Peter Collins and Kathelijn Fischer for their helpful suggestions during preparation of this manuscript.


The author stated that he had no interests which might be perceived as posing a conflict or bias.