Diagnosis and treatment of acquired haemophilia


Francesco Baudo, Thrombosis Hemostasis Unit, Niguarda Hospital, Milano, Italy.
Tel.: +39 335 225 606; fax: +39 02 6444 3948;
e-mail francesco.baudo@ospedaleniguarda.it


Summary.  Acquired haemophilia (AH) is an autoimmune syndrome characterized by acute bleeding in patients with negative family and personal history, and factor VIII depletion. Its incidence is 1.6 × 106 population per year. AH is associated with autoimmune diseases, solid tumours, lymphoprolipherative diseases, pregnancy; 50% of the cases idiopathic. Spontaneous or after minor trauma severe bleeding associated with a prolonged activated partial thromboplastin time, not corrected by incubation with normal plasma, with a normal prothrombin time are the diagnostic hallmarks. The goals of management are the control of bleeding and the suppression of inhibitor. First-line haemostatic treatment includes recombinant factor VIIa and activated prothrombin complex concentrate. Prednisone ± cyclophosphamide and other immunosuppressive agents are the standard intervention for inhibitor eradication.


Acquired haemophilia is a rare bleeding disorder caused by autoantibodies to factor VIII (FVIII) in a majority of cases. Antibodies to other clotting factors (factor V and IX) are exceedingly rare. Bleeding is acute, may be mild, but when severe, carries a high risk of death.

The incidence according to the UK registry (2007) is 1.6 × 106 population per year [1]. The median age varies in the reported series between 55 and 78 years with no difference between genders, except in the younger age because of the cases related to pregnancy. Fifty per cent of cases are idiopathic. Frequent associations are autoimmune diseases, solid tumours and lymphoprolipherative diseases (Table 1) [2].

Table 1.   Acquired haemophilia: underlying conditions (% of patient-number).
  1. *Diabetes, hepatitis, glomerulonephritis, MGUS, blood transfusion, cytomegalovirus infection, surgery, Gram-negative sepsis, nephrotic syndrome, thrombotic thrombocytopenic purpura, nephrotic syndrome, chronic obstructive pulmonary diseases, Herpes Zoster, ischaemic heart disease.

Autoimmune disorders16.6
Dermatological disorders2.2

Clinical presentation

The diagnosis of acquired haemophilia should be suspected in patients with negative family and personal history who experience either sudden spontaneous bleeding or after trivial trauma (intramuscular injection, positioning of a venous catheter) or surgery. Any site can be involved. Bleeding is defined minor or major, according to the specific site, extension and intensity. Minor bleeding, usually spontaneous, involves mainly the skin (ecchymoses); mucoses and muscles can also be affected (melaena, haematuria, methrorrhagia, epistaxis, gengivorrhagia). Major bleeding occurs in a majority of patients (65.5%) and is either spontaneous or secondary to trauma or surgery [2]. When the skin is involved, ecchymoses may diffuse to an entire limb or a vast area of the chest or abdomen. Retroperitoneal and compartmental bleeding (retropharingeal, extensive musle haematomata with compression of nervous and vascular structure) are challenging and often misdiagnosed. Haemarthroses are less frequent than in congenital haemophilia: in the Italian series, 7.3% of 96 patients at presentation and 16.1% of 31 patients at relapse [3]. Intracranial bleeding is rare (2% in the Italian survey), but catastophic. Severe bleeding can ensue surgical interventions because of an overlooked prolonged activated partial thromboplastin time (APTT). In the Italian survey, the preoperative APTT, when carried out, was always prolonged [3]. Anaemia is frequent; its severity is related to the severity of bleeding. The diagnosis of acquired haemophilia is easily suspected when the APTT is prolonged in the proper clinical setting.

Laboratory diagnosis

A prolonged APTT with normal prothrombin time should always arouse suspicion. The APTT is reported as the ratio of patient/normal plasma. The subsequent step is to carry out the mixing test. The APTT is carried out in a mixure (1:1) of normal and patient’s plasma, after incubation for 2 h at 37°C (the interaction inhibitor-FVIII:C is temperature- and time-dependent) with no correction. Further testing of FVIII:C and the inhibitor titre is available only in specialized laboratories. A prolonged APTT is also found in the lupus anticoagulant (LA), the anti phospholipid (APA) syndromes and during unfractionated heparin therapy. The LA and APA syndromes are acquired thrombophilic factors and heparin treatment is easily ruled out.

Management of acquired haemophilia

The objectives of the management are control of bleeding and suppression of inhibitor. The sudden onset of clinically significant bleeding with a prolonged APTT, suggesting acquired haemophilia, requires urgent consultation and transfer of the patient to the reference centre.

Control of the haemorrhages

Control of the acute bleeding is the priority because of the high early mortality. The inhibitor titre and the FVIII:C levels are not correlated with the clinical picture; therefore, being at variance with patients with congenital haemophilia, they are not valuable in guiding therapy [1]. The most significant criteria are clinical: the site and the intensity of bleeding.

First-line haemostatic treatment includes recombinant activated FVII recombinant factor VIIa (rFVIIa) and activated prothrombin complex concentrate (APCC). The published studies are retrospective and include a limited number of patients with different primary clinical conditions. No prospective randomized trial comparing their efficacy has been carried out to date as such a study is not practicable. Neither one is effective in every patient. The dosage of these agents is derived largely from the experience of the treatment of congenital haemophilia with inhibitor. No information on duration of treatment is available. Therapy should be continued after bleeding is controlled, depending on the site and severity of the bleeding (range of practice 24–72 h if the response is good). No laboratory tests are available to monitor rFVIIa and APCC therapy; the efficacy of therapy is judged clinically (Table 2).

Table 2.   Acquired haemophilia: first-line anti haemorrhagic therapy.
 Initial doseSubsequent dosesResponse rateTherapy durationLaboratory monitoringThrombotic side effects*
  1. *Myocardial infarction, stroke, pulmonary embolism, deep vein thrombosis, disseminated intravascular coagulation.

  2. rFVIIa, recombinant factor VIIa; APCC, activated prothrombin complex concentrate.

rFVIIa90 mcg kg−1Every 2–6 h75–100%Clinical evaluationNo<1% in haemophiliacs [10] <1/10 000 in non-haemophiliacs
APCC50–100 IU kg−1Every 8–12 h85–100%Clinical evaluationNo4–8 × 105 infusions [11]

The efficacy of rFVIIa was demonstrated in haemophilic patients with inhibitor, but its use in acquired haemophilia is not well defined. Treatment of acquired haemophilia with rFVIIa is reported in an overview of compassionate-use programs, the Hemophilia and Thrombosis Research Society Registry, and few publications [4,5]. Treatment was given for spontaneous bleeding in majority of cases. Regimens included bolus injection (46–150 mcg kg−1 every 2–24 h) or continuous infusions (8–50 mcg kg h−1). Response was categorized by the supervising physician as effective, partial response, or ineffective, by clinical examination, monitoring of vital signs, full blood count and ultrasonography or CT scanning, when appropriate. The overall response to the treatment was rated as effective or partially effective in 90% of non-surgical cases and 86% of surgical cases. Two earlier studies reported its efficacy as first-line or as salvage therapy. An effective response was observed in 100% of the episodes in which rFVIIa was used as first-line therapy; effective or a partial response was observed in 75% and 17% of the episodes when used as salvage therapy respectively [6]. A prospective study was carried out in Italy in 2001. Fourteen patients (20 bleeding episodes), selected according to the severity of bleeding, received rFVIIa as first-line therapy and one patient after failure of desmopressin (DDAVP or 1-deamino-8-D-arginine) and porcine FVIII. Treatment was very effective or effective in 13/15 patients (86.6%) and in 18/20 bleeding episodes (90.0%), in the majority of cases within 24 h without difference between patients treated by intermittent or continuous infusion [7].

The APCC is used in the treatment of FVIII inhibitors in congenital haemophilia, but again, experience in acquired haemophilia is quite limited. The standard dose ranges between 50 and 100 IU kg−1 every 8–12 h. In retrospective studies, as first line therapy, the response rate in severe and moderate bleeding ranges between 86% and 100% [8].

According to post-marketing surveillance and retrospective analysis, rFVIIa and APCC are well tolerated with few adverse events. At present, there is no conclusive evidence about the comparative thrombogenicity of APCC and rFVIIa; however, the risk is low with either agent [4,5,9–11]. Sporadic cases of myocardial infarction, stroke, pulmonary embolism, deep vein thrombosis and disseminated intravascular coagulation occurred in haemophiliacs or in non-haemophilic-patients with predisposing risk factors treated for off-label indications [10,12,13]. These indications were elective surgery (prostatectomy, hepatectomy, liver transplantation), severe bleeding secondary to oesophageal varices, complications of oral anticoagulant therapy, trauma, spontaneous cerebral haemorrhage, post partum critical bleeding. In elderly patients and in patients with underlying cardiovascular disease and other risk factors for thrombosis, these agents should be used, when strictly indicated. Anamnestic response with the increase in FVIII inhibitor after APCC treatment has been reported only in the haemophilic patients [14].

Data on the use of FVIII replacement therapy in acquired haemophilia are scanty. Its use should be attempted only in case of low inhibitor titre (<5 BU mL), minor bleeding and no bypassing agents availability. According to the experience in congenital haemophilia with alloantibodies, a loading dose should be given as bolus to neutralize the inhibitor and to achieve the haemostatic level, followed by subsequent doses given by bolus or by continuous infusion for maintenance [15]. The recovery and half-life of the infused FVIII:C cannot be predicted because of the variable kinetics of FVIII:C. In case of no satisfactory response within 24–48 h, one should resort to a by-passing agent.

Desmopressin, a synthetic vasopressin analogue, releases FVIII/von Willebrand factor from the vascular endothelium. Its use in acquired haemophilia is anecdotal; the indications are the same as for FVIII concentrates [16]. When infused intravenously or administered subcutaneously or intranasally, FVIII:C increases three- to five-fold above the baseline and reach a value sufficient to treat minor bleeding. The tachyphylaxis phenomenon limits it use to 3 or 4 consecutive days. The antidiuretic and vasomotor side-effects require caution in older patients.

The response to high-dose immunoglobulins has been attributed to the presence of anti-idiotype antibodies in the pooled immunoglobulins, but at present, there is no evidence for its use as a single agent in acquired haemophilia [17]. A possible application is as an integral component of immune tolerance induction protocol [18–20].

Immune suppressive therapy

The aim of the immunosuppressive therapy is the eradication of the inhibitor. Spontaneous complete remission (e.g. children, post-partum, drug-associated cases) were reported up to 36% of the patients [21], but are unpredictable and the patients remain at great risk of severe bleeding if the inhibitor persists [1,22,23]. Therefore, immunosuppressive therapy should be initiated as soon as the diagnosis is established. No prospective, controlled studies evaluating the efficacy of the different therapeutic agents have been published. Prednisone as monotherapy or in combination with cyclophosphamide and azothioprin is the standard intervention [1,24] (Table 3). The therapy should be carried out with adequate doses and duration: previous experience in haemophiliacs points to the importance of carrying out the treatment according to haematological tolerance [25]. Complete remission rate is higher and overall mortality is lower in the treated patients. Response rate with prednisone alone is high, but a sustained remission after prednisone discontinuation is rare. Data on the follow-up duration and relapse rate are limited. A low initial inhibitor titre and a short interval between the appearance of the inhibitor and the start of therapy seem to be positive predictive factors. The problems of infectious complications and therapy-related mortality were addressed, but data are scanty.

Table 3.   Acquired haemophilia: current strategy for inhibitor eradication.
TherapyType of therapyDosageDuration weeksComplete response %
  1. p.o., per oral.

1st linePrednisone1 mg kg−1 p.o.4–658.2–76 [1,24]
2nd linePrednisone + cyclophosphamide1.5–2 mg kg−1 p.o.669.2–75 [1,24]
3rd lineRituximab375 mg sqm week−1480.9 [27]

In a randomized prospective multicentre trial [26], 31 patients with newly diagnosed acquired haemophilia were treated with prednisone 1 mg kg day−1 for 3 weeks; 20 non-responders were randomized: four patients with prednisone (1 mg kg day−1); six patients with cyclophosphamide 2 mg kg day−1; 10 patients with prednisone + cyclophosphamide for additional 6 weeks. The inhibitor disappeared in three patients (75%) treated with prednisone and in eight patients (50%) treated with cyclophosphamide or cyclophosphamide + prednisone. No information on the follow-up was given.

In the Italian study [3], 65 of 90 patients were evaluable for the immunosuppressive therapy. Three patients died before starting treatment, one because of bleeding and two for reasons of the underlying disease. Eight patients with a low inhibitor titre (<10 BU) did not receive immunosuppressive therapy; three of them died because of bleeding complications. Information relevant to the response to the immunosuppressive therapy was missing in 14 patients. Results of the initial immunosuppressive therapy: complete remission 46 (70.7%), partial remission 13 (20%), failure 6 (9.3%). Four patients in partial remission achieved a complete remission after discontinuation of treatment. The other patients including the failures received alternative treatments (Table 4). Patients with low (<10 BU) or high (>10 BU) inhibitor titre did not differ in the rate of complete remission (30 and 22 patients respectively). Eleven patients (21.1%) relapsed; eight were rescued with additional therapy, one patient died because of bleeding and two achieved a spontaneous complete remission.

Table 4.   Acquired haemophilia. Results of the immunosuppressive therapy in the Italian study [3].
  1. Chemotherapy: cyclophosphamide, azathioprine, melphelan.

  2. CR, complete remission; PR, partial remission; F, failure; R, relapse.

Steroid ± immunoglobulins322723
Chemo ± combined therapy312344

Rituximab, an anti-CD 20 monoclonal antibody, has been used as salvage therapy. Sperr et al. compared Rituximab and prednisone + cyclophosphamide in 42 and 44 patients respectively reported in various studies in the literatures [27]. Results were similar: complete remission (CR) rate 78.6% and 84.1% without difference between patients who had (75%) or had not received previous treatment with other immunosuppressive drugs. The median treatment duration to CR was 8.3 and 6.3 weeks and the probability of CR at 2 years 66% and 94% with a plateau in the Kaplan–Mayer curve. The authors concluded that the use of Rituximab should be limited to failure of first/second line therapy. Few patients were treated with cyclosporine A or 2-chlorodeoxyadenosine.

Immune tolerance is an accepted and effective treatment of haemophilic patients with inhibitor, but has been rarely applied in acquired haemophilia. Evidence of its effectiveness and safety has been demonstrated in patients treated by the Budapest protocol (human FVIII combined with Cyclophosphamide and methylprednisolone) [18]. A complete and sustained remission has been obtained in 95% of the patients. Similar results have been obtained with a modified Malmö protocol (immunoadsorption, high doses of FVIII, high dose immunoglobulin, cyclophosphamide and corticosteroids) [20]. Bleeding was rapidly controlled with one or two aphaeresis sessions without recurrence. The inhibitor decreased to undetectable levels: median time to response 3 days; median duration of therapy, 14 days; complete response, 88%; median follow-up, 44 months (Table 5).

Table 5.   Immunotolerance in acquired hemophilia. Critical bleeding controlled rapidly during 1 or 2 aphaeresis sessions.
AuthorProtocolPatient numberBaseline inhibitor (BU mL−1)CR/PRTime to CRRelapseFinal CRFollow-up (months)Bleeding-death
  1. B, Budapest protocol (FVIII + Cyclophosphamide + methylprednisolone); BM, Bonn–Malmö protocol (immunoadsorption + high dose immunoglobulin + FVIII concentrate + cyclophosphamide + prednisolone); CR, complete remission defined as normal FVIII:C with undetectable inhibitor during a minimum F-up of 12 months; PR, partial remission defined as FVIII:C recovery up to 30%, inhibitor titre <5 BU mL−1 or both).

  2. *Severe side effects: infections (3), neutropenia (1), mucositis (1) successfully managed by antibiotics without discontinuation of MB.

  3. Two patients discontinued therapy for concomitant diseases.

  4. CR after aphaeresis and immunosuppression.

Nemes [18]B2025 (2.2–1128)193 (2–12) weeks21921.2 (2–93)0
Hute-Khune [19]BM819.5 (4.5–106)83 weeks0947 (8–52)0
Zeitler* [20]BM35146 (6–3600)31/43 (2–4) days23144 (5–86)0


The diagnosis of acquired haemophilia requires a high degree of suspicion. There is no standard therapy for either bleeding control or inhibitor eradication. The available data indicate the importance of the expert opinion in dealing with difficult problems and emphasize again the importance of early consultation with the reference centre.


F. Baudo has received reimbursement for attending symposia and fees for speaking from Bayer Healthcare and Novo Nordisk. The rest of the authors stated that they had no interests which might be perceived as posing a conflict or bias.