Methods

Top Methods Results Discussion Author & Article Info References

We included all nonhemophiliac patients known to have acquired factor VIII inhibitors who were admitted or referred to the teaching hospitals at the University of Texas in Houston between December 1992 and July 1995. The criteria for diagnosis were a life- or limb-threatening bleeding disorder with no history of bleeding, a prolonged activated partial thromboplastin time that was not correctable with admixture of normal human plasma, normal Russell viper venom time, severely decreased factor VIII coagulant activity, and a result of a Bethesda assay that was positive for anti-factor VIII antibodies.

Moderate acute hemorrhages were treated with activated prothrombinase complex concentrates, 30 to 50 U/kg of body weight; major episodes were treated with 75 to 100 U/kg. Doses were reduced or omitted if fibrinogen levels or platelet counts decreased rapidly. Additional doses of activated prothrombinase complex concentrates, human factor VIII, porcine factor VIII, or recombinant factor VIIa were given as needed.

Therapy with oral cyclophosphamide (100 to 200 mg/d) and prednisone (50 to 80 mg/d) was started during the initial hospitalization or on referral. As successful experience accumulated, the lower ends of these ranges became the standard therapy to minimize side effects. Cyclophosphamide and prednisone were continued at full doses until the factor VIII inhibitor titer decreased to zero; after this, the doses of prednisone and cyclophosphamide were slowly decreased. Patients were hospitalized only to control hemorrhage.

Patients were followed weekly, then monthly as bleeding symptoms and inhibitor titers decreased. At each visit, a complete blood count was done and factor VIII coagulant activity and inhibitor titer were measured. The blood glucose level was measured after therapy with prednisone was started and at each follow-up visit if the level was abnormal.

Results

Top Methods Results Discussion Author & Article Info References

Patients were 50 to 79 years of age; four were men and five were women. All required hospitalization for severe spontaneous hemorrhage (Table 1). One patient had rheumatoid arthritis, but no other patients had illnesses known to be associated with formation of inhibitor. Therapies used before referral included intravenous immunoglobulin (four patients), corticosteroids (five patients), and cytotoxic chemotherapy (two patients). Clinically worsening hemorrhage or an increasing titer of factor VIII inhibitor showed that these therapies were not effective. All other therapies were stopped before cyclophosphamide and prednisone therapies were started. At presentation, titers of human factor VIII inhibitor ranged from 2.5 to 1040 Bethesda units (mean, 174 Bethesda units; median, 38 Bethesda units). Inhibitors to porcine factor VIII developed in all patients; only three patients had previous exposure to porcine factor. All patients had extensive soft tissue bleeding, including massive retroperitoneal bleeding (two patients), hemarthrosis (two patients), massive hematuria (one patient), hemopericardium with tamponade (one patient), and compartment syndrome of the forearm (one patient). All patients required therapy with coagulation factors to control severe hemorrhage.

All patients achieved complete remission (that is, no measurable inhibitor titer could be detected and plasma factor VIII activity returned to normal) (Figure 1). Median time to resolution of all bleeding symptoms was 3 weeks (range, 2 to 10 weeks); median time to complete loss of measurable factor VIII inhibitor levels was 12 weeks (range, 3 to 37 weeks). Two patients had relapse as the cyclophosphamide dose was tapered, but remission quickly returned with reinstitution of cyclophosphamide and prednisone. Resolution of high titers of inhibitors seemed to require a longer duration of treatment.

View larger version (16K): [in this window] [in a new window]   Figure 1. Titers of factor VIII inhibitor in all patients after initiation of therapy. Therapy with cyclophosphamide and prednisone was started on day 0. Numbers in the key refer to patient numbers.

Seven patients are no longer receiving therapy (median time without therapy, 91 weeks [range, 61 to 164 weeks]). One patient requires low-dose prednisone for chronic obstructive pulmonary disease. Another patient died of chronic obstructive pulmonary disease but had had no measurable inhibitor level for 37 weeks.

Few side effects were observed. Two of three patients whose fasting glucose levels were less than 9 mmol/L were not treated; a third patient required an increased dose of glyburide. A 77-year-old patient developed herpes zoster and had worsening cataracts that required surgery after 45 weeks of therapy with prednisone. One patient who received cyclophosphamide had hematuria that was related to the inhibitor, not to cyclophosphamide cystitis. Another patient receiving cyclophosphamide developed reversible dysplastic changes in bone marrow after 104 weeks. One patient had an anaphylactoid reaction to porcine factor VIII, and one patient had an anaphylactoid reaction to activated prothrombinase complex concentrate.

Discussion

Top Methods Results Discussion Author & Article Info References

Acquired autoantibodies to coagulation factors are uncommon. The incidence of acquired factor VIII inhibitors is estimated to be one case per 1 million persons per year [1]. This Figure 1 may be an underestimate because symptoms can vary and spontaneous remissions can occur in persons with low titers of inhibitors.

Inhibitors to factor VIII are of two types: alloantibodies [2], which occur in patients with congenital factor VIII deficiencies who receive exogenous factor VIII, and autoantibodies, which cause the autoimmune disease observed in our patients. Autoantibodies usually occur with no identifiable cause. Their incidence increases with age; most patients with this condition are older than 60 years of age. Anti-factor VIII antibodies can occur in conjunction with other autoimmune diseases, especially rheumatoid arthritis, lupus erythematosus, and inflammatory bowel disease. A transient and usually mild form of factor VIII inhibitors is seen in postpartum women. Factor VIII inhibitors are also associated with lymphoid malignant conditions and such medications as penicillin and phenytoin [3, 4].

Successful treatment of factor VIII inhibitors requires immediate control of hemorrhage and long-term suppression of the autoantibody. Hemostasis may be achieved in some patients with inhibitor titers of 5 Bethesda units or less by administering high doses of human factor VIII. If patients have inhibitor levels greater than 5 Bethesda units or do not respond adequately to human factor VIII, then porcine factor VIII, activated prothrombinase complex concentrates, or recombinant factor VIIa are necessary. Activated coagulation factors should be used initially if the titer of antiporcine antibodies is high or unknown, but repeated doses should be given judiciously because of the risk for thrombosis. Porcine factor VIII is advantageous because it predictably shortens the activated partial thromboplastin time and produces a measurable level of factor VIII in plasma. However, alloantibodies and cross-reacting anti-human antibodies can render it ineffective [5]. Recombinant human factor VIIa is investigational in the United States.

Other immunosuppressive regimens have been used in this disorder with reasonable success. Lian and colleagues [6] administered monthly intravenous cyclophosphamide, vincristine, prednisone, and factor VIII until remission was achieved. After 1 to 3 courses of therapy, 11 of 12 patients (92%) responded. Unlike the patients in our study, the patients studied by Lian and colleagues received factor VIII and had lower titers of factor VIII inhibitor. A randomized trial [7] compared the effects of oral prednisone, oral cyclophosphamide, or both. In that study, patients who had high titers of antibodies also received infusions of factor VIII. Thirty patients initially received oral prednisone, 1 mg/kg, for 3 weeks. In 10 of 30 patients, the titer of factor VIII inhibitor was reduced to 1 Bethesda unit or less. The remaining 20 patients were randomly assigned to 6 additional weeks of therapy with prednisone (4 patients); oral cyclophosphamide, 2 mg/kg (6 patients); or both (10 patients). Titers of factor VIII inhibitor less than 1 Bethesda unit were seen in 2 patients who received prednisone, 3 patients who received cyclophosphamide, and 5 patients who received both. The median titer of factor VIII inhibitor was 3 Bethesda units for patients who responded to any of the three therapies and 50 Bethesda units for nonresponders.

Our study differs from previous studies in several ways. First, all patients in our study had life- or limb-threatening hemorrhage that required short-term and long-term therapy. Other series have included patients whose inhibitor titers were low enough to be of uncertain clinical significance [1, 4]. Second, we continued to administer immunosuppressive agents at full doses until complete remission (no detectable inhibitor and normal factor VIII antibody titer) was achieved. Third, we did not use empirical factor VIII infusions to stimulate immune response because endogenous factor VIII production was ongoing.

The importance of the duration of treatment must be emphasized. Results of an early study [8] suggested that response to immunosuppressive therapy occurs within 6 weeks of treatment. We and others [9] have shown that high titers of factor VIII inhibitor may require prolonged continuous therapy. Immunosuppression should be continued until the inhibitor cannot be detected and should then be discontinued gradually while patients are observed for return of the antibody. In two patients in our series, antibodies rapidly returned after discontinuation of cyclophosphamide therapy but quickly resolved with retreatment.

Our series is limited by the lack of a control group. Randomization is logistically difficult in studies of rare disorders, and no studies with completely comparable controls have been published. In previous studies, factor VIII was used or therapy was discontinued before factor VIII activity completely returned to normal [6, 7]. In addition, because our series is small it is potentially not representative of all patients with this condition.

Low titers of factor VIII inhibitor may resolve spontaneously [1, 4]. Asymptomatic patients with low inhibitor titers may therefore be closely monitored without treatment. However, patients with high titers of factor VIII inhibitor, low factor VIII activity, or hemorrhage usually require immediate intervention with aggressive immunosuppression to prevent severe illness or death. Activated factor concentrates or porcine factor VIII given during the early phases of treatment can control acute hemorrhage. Prolonged therapy may be necessary, but even patients who have high titers of factor VIII inhibitor and severe hemorrhage can achieve sustained remission.

We show that immunosuppression with continuous low-dose oral cyclophosphamide and prednisone is successful in treating patients with autoantibodies to factor VIII. We attribute the high rate of success to continuation of therapy until complete normalization. This conclusion is supported by the observed progressive decline in titers of factor VIII inhibitor seen with continued therapy and the resolution of recurrences of inhibitors with retreatment.

From the University of Texas-M.D. Anderson Cancer Research Center and University of Texas-Houston Medical School, Houston, Texas.

Dr. Phillips: University of Texas-Houston Medical School, Division of Hematology, 6431 Fannin Street, Room 5.278 MSB, Houston, TX 77030.