Diagnosis

Incidence and co-morbidity

AHA is rare and the incidence is about 1.5/million per year [1]. It is primarily a disease affecting the elderly (median age 64–78 years) but can also be associated with pregnancy. Most cases are idiopathic (43.6%-51.9%) but AHA is also associated with malignancy (6.4%-18.4%), autoimmune disorders (9.4%-27.0%), infections and dermatologic conditions [1–6]. Rarely, AHA arises as idiosyncratic reactions to medication including antibiotics, psychiatric and immunomodulatory drugs [7]. The relatively high number of co-existing autoimmune disorders in the patients or among his/her relatives may be striking. Therefore, findings suggesting polymorphisms in immune regulatory genes to be associated with the incidence of acquired hemophilia may be of importance [8,9]. It has also been hypothesized that encounter of antigenically different, allogeneic FVIII after blood transfusion may challenge inhibitor formation after presentation on MHC class II and support for this assumption was found in a study by Tiede et al [10].

Symptoms, signs and clinical course

Easy bruising with extensive enlargement, muscle hematomas and profuse bleeds after trauma and surgery are the most common symptoms in a patient who has never had any signs of bleeding tendency earlier. Muscle bleeds and gastrointestinal (GI) bleeds are also common, but in contrary to congenital hemophilia, joint bleeds are rare [1,2,4,11]. Hematuria with intermittent clotting and occlusion of ureter(s) may decrease renal function. Massive bleeds may occur after intravenous venipuncture if special care is not taken. Severe bleeds may be life-threatening. Because of the second-order kinetics of the anti-FVIII antibodies, FVIII levels are not predictive of bleeding risk and patients can have clinically significant bleeding despite only modestly reduced FVIII activity levels.

In the Nordic study, mortality related to AHA was 21% [11]. Immunosuppressive therapy (IST) associated mortality was high, accounting for 16% of all deaths (4.2% of patients) in EACH2 [2] and 15% of all deaths (8.8% of patients) in the Nordic study [11]. Other fatal and nonfatal complications include thrombotic events such as myocardial infarction and stroke [3,6,11].

Laboratory screening methods

Activated partial thromboplastin time, APTT is usually prolonged, while Prothrombin time and PT (INR) are within the normal range. Platelet and leukocyte counts are generally normal. The erythrocyte count, hemoglobin and hematocrit values may be low due to bleeds.

APTT-mixing test: Patient plasma is mixed with an equal volume of normal plasma (NPP; 1+1) and APTT analyzed immediately and/or incubated for 1-2 h at 37°C. In normal conditions, the APTT will lengthen slightly if incubated 1-2 h at 37°C, because of a spontaneous decay of the labile FVIII. A prolonged APTT in a factor deficient patient plasma will be corrected when the plasma is mixed 1:1 with normal plasma, whereas presence of an inhibitory antibody against a coagulation factor or lupus anticoagulant, does not give a correction after mixing (anticoagulant in the plasma must be excluded). In acquired haemophilia, weak affinity inhibitors can sometimes only be detected if a longer incubation time (1-2h) is used prior to analysis (see below Characteristics of the autoantibodies) [12].

Specialized laboratory methods

The diagnosis should be confirmed at a specialized coagulation laboratory. For plasma-based coagulation assays the recommendations for sample collection and transport are the same as in the general recommendation for hemophilia (See NHC hemophilia Guideline Chapter 3).

FVIII assays

The functional activity of FVIII (FVIII:C) or, occasionally, other coagulation factors such as FIX, can be measured with either clot-based (OSA) or chromogenic assays (CSA), the latter being less prone to interference from i.e. lupus anticoagulant that can be a diagnostic problem. Autoantibodies to other factors are extremely rare. In the diagnostic situation the patient can have a significant amount of factor activity, which can interfere with the measurement of the titer of inhibitory antibodies.

FVIII inhibitor assay

The recommended procedure for diagnosing inhibitory antibodies to coagulation factors is the Bethesda-Nijmegen method [13]. The antibody titer is expressed in Bethesda units (BU). One Bethesda Unit is defined as the quantity of antibody, which reduces the FVIII activity by half in normal plasma. A test sample is prepared by mixing equal volumes of patient plasma with NPP and then measure the residual factor activity in the plasma mixture after 2h incubation. A control sample is prepared in parallel with NPP mixed with an equal volume of FVIII-deficient plasma. Both test and control samples are incubated for 2 h at 37ºC and then the factor activities in both samples are measured. The factor activity is measured using OSA or CSA method. The Bethesda assay is most suitable for patients that do not have measurable factor activity. If the patient has an activity > 0.05 IU/dL (5%), heat-inactivation of the plasma sample endogenous activity at 56˚C for 1h before analysis is recommended.

Interference

The specific treatments used in acquired haemophilia can affect measurement of both factor activity and inhibitor titre. This is relevant for many FVIII products as well as for bypassing agents. Of significant importance is that the clinicians discuss possible local measurement options before the initiation of a new treatment.

Emicizumab interferes in both APTT and APTT-based factor VIII activity assays and in CSA FVIII assays with reagents of human origin. A substantial shortening of APTT and overestimation of FVIII activity is observed.

For determination of FVIII activity and inhibitor level in patients receiving emicizumab, the use of a CSA containing bovine factor X is recommended. Confirmation of emicizumab levels requires a specific assay with emicizumab calibrators. Furthermore, porcine FVIII treatment can also give discrepant results in OSA and CSA and is not consistent between methods.

Characteristics of the autoantibodies

Autoantibodies against FVIII are composed predominantly of IgG, most often of the IgG4 subclass, and have a preponderance of kappa light chains. The main antigenic epitopes of the FVIII molecule are the A2 and C2 domains. Bound inhibitors block the binding of FVIII to phospholipids, von Willebrand factor and cofactors of the FVIII molecule. The autoantibodies follow a type II inactivation pattern, which means that there is incomplete neutralization of FVIII activity. Therefore, variable levels of residual FVIII activity may be detectable despite the concomitant presence of high titers of the FVIII inhibitor. The complex type II kinetics makes it difficult to clinically evaluate the inhibitor titer and factor levels. This contrasts with alloantibodies seen in patients with congenital hemophilia, which follow type I kinetics characterized by complete inactivation in a linear relationship between antibody titer and residual FVIII activity.

Recommendations for diagnosis

  • Acquired hemophilia should be suspected in patients with prolonged APTT, but normal INR/PT and signs of bleeding.

  • If coagulation factor analyses are not readily available, a mixing test could be performed to strengthen the suspicion. The presence of an inhibitory antibody against a coagulation factor does not give a correction of APTT after mixing with normal plasma.

  • The diagnosis should be confirmed at a specialized coagulation laboratory using specific methods for functional factor activity measurement (FVIII:C) and the Bethesda-Nijmegen method.

  • Type II kinetics of the antibodies results in poor correlation between antibody titer and FVIII:C and caution is warranted when interpreting the results.

  • The specific treatments used can affect measurement of both factor activity and inhibitor titre. The local measurement options need to be discussed before treatment initiation.