Mild haemophilia: a disease with many faces and many unexpected pitfalls

Epidemiology

The incidence of haemophilia is around 1/5000 male births. The reported prevalence of haemophilia, however, shows major differences across the world depending largely on economic status and reporting procedures [1]. The reported proportion of mild haemophilia is highly variable. In a review on 16 115 haemophilia A patients from 147 haemophilia treatment centres worldwide, 32% had mild haemophilia [2]. In some registries, the percentages of mild haemophilia is much higher: the Canadian Hemophilia Registry reports mild haemophilia in 51% of 1594 registered haemophilia A patients and in Spain 51.5% of 2905 patients from the national registry are registered as mild [3]. In contrast, the reported incidence of mild haemophilia is much lower in less wealthy parts of the world with only 16% mild haemophilia of 5043 reported patients in The National Hemophilia Registry in China [4].

The life expectancy of patients with mild haemophilia without hepatitis or HIV equals the life expectancy of the non-haemophilic population at least in developed countries [5].

Diagnosis

Mild haemophilia is usually diagnosed during family investigation or following a bleeding episode. Diagnosis of mild haemophilia is generally made at an older age than in severe haemophilia. In a French cohort study in a paediatric population, the median age of diagnosis was 28.6 months vs. 5.8 months in severe haemophilia [6]. It is not unusual that patients are diagnosed with mild haemophilia at an adult age when suffering major haemorrhagic complications after surgery or trauma.

In haemophilia, generally, a prolonged activated partial thromboplastin time (APTT) is found. Prolongation of APTT occurs when an individual has <0.30 IU mL⁻¹ factor VIII, whereas in individuals with 0.30 IU mL⁻¹ factor IX, APTT often is within normal ranges. The sensitivity for low levels of factor VIII or factor IX of the different thromboplastins used for APTT measurement is very variable. Whenever there is clinical suspicion of mild haemophilia, factor VIII and factor IX levels should be measured despite an APTT within the normal range.

As factor VIII is an acute phase protein, levels are increased during episodes of bleeding, trauma and inflammation and this may interfere with a diagnosis of mild haemophilia in such circumstances. There is no simple relationship between the measured level of factor VIII and bleeding tendency in mild haemophilia. In haemophilia carriers, increasing bleeding tendency is found for levels of factor VIII between 0.41 and 0.60 IU mL⁻¹ [7]. In mild haemophilia, an age-dependent increase in FVIII activity is described, however, without apparent effect on the number of bleedings [8].

In approximately 40% of patients with mild haemophilia A, discrepant factor VIII levels are measured when a one-stage clotting assay is used as compared with a two-stage (chromogenic) assay [9–11]. In most of these discrepant cases, the factor VIII levels are reduced by 50% or more when measured by a two-stage assay as compared with a one-stage assay and this can lead to missing the diagnosis of mild haemophilia A when a one-stage assay is used as a screening method. The reverse situation, higher factor VIII levels with a two-stage method than with a one-stage method, is less frequent. Mutations and molecular mechanisms of many of these discrepant cases have been resolved [12–14]. However, it remains unclear which assay is the best reflection of the bleeding phenotype. Using thrombin generation assays in patients with the more common discrepancy pattern (FVIII lower by two-stage assay), the most significant correlation was found between the one-stage FVIII assay and thrombin generation [12]. In two families with the ‘reversed discrepancy’ (FVIII higher by two-stage assay) and contrasting clinical histories (one family bleeding and one non-bleeding), impaired thrombin generation reflected the bleeding phenotype [15].

Molecular mechanisms in mild haemophilia A

Characterization of the molecular mechanisms resulting in low FVIII levels have helped to identify regions of the factor VIII gene critical for proper factor VIII biosynthesis, thrombin activation, intramolecular stability as well as binding regions for important partners such as von Willebrand factor, factor IXa and the phospholipid surface [16]. In patients with the common presentation of mild haemophilia A with reduced FVIII activity in a two-stage assay as compared with a one-stage assay, a number of missense mutations mainly clustered within the A domains have been described that lead to defective stability of FVIIIa. Conversely, mutations impairing FVIII activation by thrombin result in higher FVIII activity in a two-stage than in a one-stage assay [14].

Some particular FVIII missense mutations, mainly located within the region encoding for the light chain of factor VIII, contribute to an unexpectedly high incidence of inhibitors in mild haemophilia A [16,17]. Genetic testing might thus become an important key feature in the management of mild haemophilia A patients.

Treatment of mild haemophilia

Most patients with mild haemophilia A respond well to the administration of desmopressin which typically results in a 2–6-fold increase of FVIII levels over baseline [18]. The peak postdesmopressin levels of FVIII depend on the patient’s basal FVIII level [19] and postdesmopressin FVIII half-life, typically around 5–8 h, is positively related to basal and peak von Willebrand Factor Antigen levels and patient age [20]. Young children often have a markedly lower response to desmopressin than adults [21].

Postdesmopressin FVIII levels >0.30 IU mL⁻¹ are considered clinically adequate at least for the treatment of spontaneous or posttraumatic bleeding, whereas a postdesmopressin FVIII level of at least 0.50 IU mL⁻¹ is required for the treatment of major surgery.

Predictors of the magnitude of response to desmopressin in the individual patient are not well studied. The underlying molecular defect seems to be important as the response tends to be similar within families and as a reproducible pattern of response was shown in patients with the same mutation even when they were unrelated [20]. In the individual adult patient, the magnitude of FVIII response to desmopressin is stable over time [22]. Repeated doses administered at 8–12-hourly intervals, however, will lead to decreased responsiveness (tachyphylaxis). The factor increase after the second dose is approximately 30% less than the response after the first dose without further decrease upon following repeated doses [18]. Prior to therapeutic use, the magnitude and the half-life of the FVIII response to desmopressin should be obtained in the individual patient.

Formulations of desmopressin are available for intravenous (i.v.), subcutaneous (s.c.) and intranasal routes of administration. The s.c. and especially the intranasal formulations are important tools for patient coping, but are unfortunately not available in all countries.

Desmopressin has few side effects such as facial flushing, headache, limited decrease in blood pressure and increase in heart rate. Desmopressin has an antidiuretic effect, which lasts 24 h after a single dose and may induce hyponatremia, especially in young children. To avoid hyponatremia, some fluid restriction following desmopressin administration is mandatory and the concomitant administration of non-steroidal anti-inflammatory drugs should be avoided [19].

Desmopressin should be used whenever possible in the treatment of mild haemophilia A, not only to avoid the high cost of FVIII concentrates but also to minimize the exposure to exogenous factor VIII and thus the risk for inhibitor development.

The association of antifibrinolytic agents is effective especially in mucosal bleedings, with the exception of haematuria where it may provoke obstruction of the urinary tract [23].

In patients unresponsive to desmopressin or if long-term correction of FVIII levels is mandatory in major surgery or after major trauma, the administration of factor VIII concentrates is the treatment of choice. Evidence-based data on target levels and frequency of administration are non-existent. Guidelines generally recommend similar target levels as in severe haemophilia, the frequency of administration should be guided by the measured plasma concentration of FVIII. Especially in the milder forms of haemophilia A, the postsurgery levels of FVIII increase significantly as a result of the acute phase response of factor VIII. Whether the age-dependent increase in FVIII levels in patients with mild haemophilia should influence the transfusion practices remains to be studied [8]. Although some data are suggestive of an increased risk for inhibitor formation in mild haemophilia A if FVIII concentrates are given by continuous infusion [24], analysis of larger databases is needed to confirm these findings.

In patients with mild haemophilia B, bleeding episodes are treated with factor IX concentrates. Antifibrinolytics can be used as a concomitant treatment, especially for mucosal bleeding. Patients with mild haemophilia B do not respond to desmopressin.

Inhibitor development in patients with mild haemophilia A

Until the late 1990s, inhibitors in mild haemophilia A were considered very rare. However, since the publication of Hay et al. [25] in 1998 on behalf of the UK Haemophilia Centre Directors Organisation, it has been appreciated that inhibitors in mild/moderate haemophilia are more frequent than previously thought. Clinical problems associated with inhibitors in mild haemophilia are often considerable, as in the majority of cases, adult patients are confronted with a change in phenotype from mild-to-severe and they suddenly experience spontaneous severe bleeding.

Patients with mild haemophilia are at lower risk of inhibitor development than are severely affected patients. The prevalence of these inhibitors has been estimated to be between 3% and 13% [26–28]. In a prospective study of inhibitor incidence among 1306 haemophilia A patients, only 6% of the inhibitors were found in patients with factor VIII >0.03 IU mL⁻¹ [29]. Sixteen (28%) of 57 new inhibitors reported between January 1990 and January 1997 in the UK Haemophilia Centre Doctors’ Organisation (UKHCDO) inhibitor register arose in patients with mild or moderate haemophilia. The annual incidence of inhibitors in the UK was 3.5 per 1000 registered with severe haemophilia and 0.84 per 1000 patients registered with mild/moderate haemophilia [30].

Usually, the presence of an inhibitor in patients with mild haemophilia is suggested by a change to a severe bleeding pattern with spontaneous bleedings or uncontrollable postsurgery bleeding. This change in bleeding pattern is explained by cross-reactivity of the inhibitor with the mutated factor VIII of the patient resulting in a residual factor VIII level of <0.01 IU mL⁻¹ [31–33]. The bleedings occur often in muscles and joints as in severe congenital haemophilia, but sometimes, the bleeding pattern is more reminiscent of acquired haemophilia with the occurrence of large cutaneous bruising, gastrointestinal and urogenital bleeding [25]. Occasionally, there is no change in residual factor VIII level, but an inhibitor is detected in the Bethesda assay and/or there is lack of efficiency of factor VIII transfusions [33–35]. In some cases, the specificity of the immune response reverts over time from neutralization of both mutated self and transfused normal factor VIII to tolerance to self, resulting in a recovery of the original basal factor VIII level and response to desmopressin, despite the persistence of antibodies to exogenous FVIII [25,31,33,34].

Inhibitors in mild haemophilia occur more commonly later in life and an episode of intensive treatment with factor VIII concentrate (for bleeding, trauma or surgery) seems to precede detection of the inhibitor in most reported cases. In the series reported by Hay et al. [25], 16 of 26 inhibitors were detected after such intensive replacement therapy and in this series, no particular concentrate was implicated. Intensive exposure to factor VIII as a risk factor for inhibitor development in mild haemophilia A was confirmed in a publication from Canada [36]. The overall incidence of inhibitors in their population of boys (age between 0 and 18 years) with mild haemophilia A (n = 54) was 7.4%. When the analysis was restricted to patients exposed to factor VIII, the incidence was 14% (4/29) and patients who received factor VIII as a continuous infusion developed inhibitors in four of five (57%) cases. In a retrospective cohort study of 138 patients with mild haemophilia A, intensive use of factor VIII was associated with an increased risk for inhibitor development, especially in the perioperative setting and when used as a continuous infusion [24].

In patients with mild haemophilia, certain missense mutations seem to predispose to inhibitor formation. In the series of Hay et al. [25], seven of nine mutations were clustered in a region at the junction between the C1 and C2 domain. The two remaining mutations affected the A2 domain. Clustering of the mutations in these regions has been confirmed in most other reported cases of mild haemophilia with inhibitor and some particular mutations such as Arg2150His and Arg593Cys seem to be overrepresented [25,31,33,34,37–39]. Arg593Cys was a risk factor together with intensive perioperative factor VIII administration, in the retrospective cohort study from Amsterdam [24].

To understand why some mutations predispose to inhibitor formation, B cell and T cell responses to FVIII were studied in patients with some of these mutations who developed inhibitors. Analysis of FVIII produced by patients with mild haemophilia A demonstrated that mutations at residues Arg2150, Arg2159 or Ala2201 eliminates FVIII epitopes (antigenic determinants) recognized by monoclonal inhibitor antibodies [40–42] and patients’ polyclonal antibodies [31,34,43].

Study of the T cell response to FVIII in a mild haemophilia A patient carrying an Arg2150His substitution in the C1 domain and who presented with a high titre inhibitor towards normal but not self FVIII showed that Arg2150His FVIII and normal FVIII can be distinguished by the immune system not only at the B cell level but also at the T cell level [44]. Similar observations have been made with a patient carrying mutation A2201P [45].

These observations have demonstrated that both B cells and T cells can distinguish between self and wild-type FVIII molecules differing by a single point mutation, which provides a mechanism for the frequent occurrence of inhibitor in patients carrying some mutations.

Treatment of patients with mild haemophilia A with inhibitors

Bleeding episodes in patients with mild haemophilia who developed an inhibitor are often particularly severe and sometimes life-threatening. Bypass therapy with activated prothrombin complex concentrates or recombinant activated factor VII can be used to control bleeding and has the advantage of avoiding anamnesis. Some patients can be treated successfully with desmopressin, especially those patients whose basal factor VIII level did not significantly decrease and whose inhibitor does not seem to cross-react with their endogenous factor VIII [25,33,34] or once an adequate circulating factor VIII level has been restored. Desmopressin does not cause anamnesis in those patients despite the presence of high-responding inhibitors [25].

Published data on immune tolerance induction in patients with mild haemophilia and inhibitors are very scarce. In the series reported by Hay et al. [25], immune tolerance induction was attempted in eight patients using different regimens. The Malmo regime (high dose factor VIII combined with cyclophosphamide and i.v. IgG) was used successfully in two patients and with a partial response in further two patients. The Van Creveld regime (low dose factor VIII every other day) was used unsuccessfully in one patient and with partial success in a further patient and the Bonn regime was used unsuccessfully in one patient and with partial success in another patient. The overall success rate of immune tolerance of two of eight patients seems lower than the reported success rate in severe haemophilia.

Other reported treatments have included immunomodulatory drugs such as corticosteroids, cyclophosphamide, anti-CD20 monoclonal antibody rituximab [32,46–48] and avoidance of re-exposure to factor VIII using desmopressin and bypassing agents to treat bleeding episodes [49]. Currently available data are not sufficient to offer evidence-based advice on the optimal treatment of inhibitors in patients with mild haemophilia A and the management of these patients remains controversial at this point. Preliminary data from a retrospective and prospective data collection in France and Belgium [16,50] suggest that immune tolerance induction could be more effective than no specific treatment or immunomodulating drugs in preventing risk of anamnesis of the inhibitor after re-exposure to factor VIII. In a meta-analysis on the effectiveness of rituximab in patients with congenital haemophilia and inhibitors, complete responses were unexpectedly high in patients with mild haemophilia (12/16 patients) as compared with severe haemophilia (12/28) [51].

Maximal use of desmopressin for the treatment of patients with mild haemophilia A is certainly useful to prevent the development of inhibitors in these patients. Avoidance of intensive courses of treatment with factor VIII concentrates has to be considered especially in those patients known to harbour one of the high risk mutations or having a relative who developed an inhibitor.

Conclusion

Patients with mild haemophilia are facing a tricky itinerary full of unexpected pitfalls. Increasing the awareness for the peculiarities of this bleeding disorder can avoid major posttraumatic or postsurgery bleedings that occur in unrecognized and also in previously diagnosed mild haemophilia. A better understanding of the molecular mechanisms and the environmental risk factors contributing to the risk of inhibitor development will help in the design of an individual treatment course for each patient with mild haemophilia minimizing the inhibitor risk. The maximal use of desmopressin certainly is a cornerstone in this strategy. For inhibitor eradication, less invasive strategies than the standard ‘immune tolerance induction’ are urgently needed to decrease the morbidity in these often elderly patients.

The group of mild haemophilia patients requiring major surgery will further increase with increased life expectancy. Prevention of inhibitor formation in this vulnerable patient group is a challenge for the next decade.

Disclosures

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