INHIBITORS

IN HEMOPHILIA A AND B

An Introductory Discussion

2016

Carol K. Kasper, M.D.

2

What are inhibitors? Inhibitors are antibodies that neutralize the activity of a clotting factor. Inhibitors to fac-tor VIII or IX may arise as allo-antibodies in patients with hemophilia A or B, respectively, who have been treated with exogenous factor VIII or IX. Inhibitor antibodies to factor VIII may arise as auto-antibodies in non-hemophilic persons. Non-neutralizing antibodies also may arise and may be detected with suitable tests. Incidence and Prevalence The number of new inhibitors in a population in a given period of time (incidence) reflects transient inhibitors as well as the more common enduring ones. If inhibitor testing is frequent, a few more transient inhibitors are detected. The number of inhibitors present in a population at any given time (prevalence) pri-marily reflects long-standing inhibitors. The incidence of inhibitors to factor VIII (FVIII) in patients with severe hemophilia A within the first few years of life is reported as 20-30%. The prevalence is lower, and nowadays it may reflect the success of specific treatment to eradicate inhibitors. The incidence of inhibitors in severe hemophilia B is about 4%. Clinically-apparent FVIII auto-antibodies are uncommon; the incidence is difficult to assess because patients are not centralized. Some observers estimate one clinically-apparent case per mil-lion population per year. Factor IX auto-antibodies are rare. Predisposing influences The level of risk for inhibitor formation is influenced by the following: (1) the patient’s hemophilia mutation, (2) his race, (3) his im-mune system and (4) the nature of his exoge-nous factor replacement product. Patterns of factor replacement (i.e. intensity of treatment, age of prophylaxis onset) may also have some influence. The first three issues are genetic, so it

should not be surprising that strong inhibitors often arise after only a few exposures to ex-ogenous clotting factor. In prospective stud-ies of previously-untreated patients (“PUPs”, mostly babies), those developing inhibitors did so after a median of 8-9 exposure-days. In one registry, half of patients with severe hemophilia who developed inhibitors did so under age 5 years. The mutation causing hemophilia determines whether some of the clotting fac-tor is produced, and thus is accessible to the thymus during fetal life, to be recognized as “self”. Large deletions and nonsense muta-tions causing premature stop codons are associated with the highest incidence of in-hibitors. In the common intron-22 inversion mutation, most of the protein is produced intracellularly but does not reach the plasma; presence in cells suffices for thymic recogni-tion. Missense mutations allow a full-length protein to be made and circulated although it has a mistake in it. The incidence of inhibi-tors in patients with missense mutations is low, but a few specific missense mutations are associated with a high frequency of in-hibitors. (One such mutation prevents bind-ing of T cells to FVIII, thus interfering with tolerance development in fetal life.) The patient’s race/ethnicity also influ-ences the propensity for inhibitor develop-ment, but the reasons are not yet known. Patients who are of black African descent, or are Latin-American, or native-American have a higher incidence of inhibitors than do Cau-casians. As yet, no study has compared per-sons of Asian descent to Caucasians. The immune system of the patient is doubtless involved in his propensity to inhibi-tor formation. Some variants of immune competent genes and HLA phenotypes are more common than others in patients with inhibitors. The tendency for inhibitors to run in families may be related to inheritance of immune system characteristics. The role of the specific concentrate

3

has been questioned. Inhibitors are somewhat less likely to develop with the use of plasma-derived concentrate, compared to recombinant products. (Recombinant clotting factors are not identical to naturally-occurring ones.) Two plasma-derived FVIII concentrates made in Europe in the early 1990’s, however, did provoke inhibitors in some previously-heavily-treated patients (PTPs) thought to be at low-risk. One such product, made in by the controlled-pore-silica fractionation method had not elicited new inhibitors in PTPs when viral-inactivated with dry heat but did so when pas-teurized. The other product became more anti-genic when pasteurization was added to a sol-vent-detergent viral inactivation process. The FVIII in these products was denatured slightly in production. Non-hemophilic persons Inhibitors occasionally arise as auto-antibodies in post-partum women, in patients with auto-inmmune disorders or other ill-nesses, or in older adults without obvious un-derlying disorders. Immune response to FVIII The initiation of an immune response and formation of antibodies requires endocyto-sis of the “foreign” FVIII molecule by antigen presenting cells and presentation of antigen-derived peptides via the HLA class II mole-cules on the cell surface to the CD4 T cells. In addition, for the CD4 T cells to become acti-vated, and acquire the ability to stimulate anti-gen specific B cells, additional triggers, “danger signals”, are required, such as are released by inflammation, stress or tissue damage. T-regulatory cells have a modulating effect. Inhibitors are mainly IgG antibodies of the IgG1 and IgG4 subtypes. Anti-FVIII allo-antibodies bind primarily with epitopes in the FVIII A2 or C2 domains and auto-antibodies bind with the C2 domain.

The union of FVIII with its inhibitor is not associated with allergic reactions. The union of factor IX with its inhibitor, however, may cause severe allergic reactions includ-ing anaphylaxis. Such reactions can occur with the first infusion of exogenous factor IX given after the inhibitor develops, that is, be-fore it has been diagnosed. Factor IX-inhibitor complexes can cause nephrotic syndrome. Children with hemophilia B are treated cautiously at a clinic, not at home, during the first 50 or so exposure-days. The reaction of FVIII with its inhibitor is time-dependent both in vitro and in vivo, a fact relevant to measurement and to clinical treatment. The higher the level of the inhibi-tor, the more rapidly it inactivates FVIII. If a patient’s inhibitor level is low to moderate, a high therapeutic dose of FVIII may have a chance to take part in coagulation before being neutralized. Two patterns of reaction kinetics are seen (Figure 1). Inhibitors with “Type 1” or “simple” kinetics completely neutralize FVIII and are neutralized themselves in the reac-tion. Most inhibitors in hemophilia patients are Type 1. Inhibitors with “Type 2” or “complex” kinetics do not completely neutral-ize FVIII and, after reaction, retain some ability to neutralize additional FVIII. Although some FVIII may remain measurable in the plasma in the presence of type 2 inhibitors, the patient may bleed as profusely as if he had no FVIII at all. Type 2 reaction kinetics are more common in auto– than in allo-antibodies. Clinical Presentation In hemophilia, an inhibitor is sus-pected when a patient’s hemorrhage is not controlled with his usual dose of clotting fac-tor. The presence of an inhibitor does not change the typical site, frequency or severity of bleeding. The inhibitor makes control of hemorrhages more difficult. Patients with auto-antibodies to FVIII

4

Figure 1. Simple and complex reaction kinetics.

Simple reaction kinetics:

FVIII Ag + inhibitor Ab = AgAb; (no factor VIII activity, no inhibitor activity)

Complex reaction kinetics:

FVIII Ag + inhibitor Ab = Ag (low factor VIII activity) + Ab (diminished inhibitor activity)

-10

0

10

20

30

40

50

60

1 3 5 7 9 11 13 15

Inh

ibit

or

Level, B

U

Time, Weeks10

Dose of Factor VIII

Dose of Factor VIII

Dose of Factor VIII

High Responder

Low Responder

0

Figure 2. “Low-responders” and “high-responders”. When repeated doses of FVIII are given

to a low responder, as depicted by the solid line, anamnesis is not seen. The first dose of FVIII, given to

a high responder (dashed line) when his inhibitor level is low, does provoke anamnesis but subsequent

doses, given while the inhibitor level is still high, at the patient’s current maximum production, may not

provoke further rises in the inhibitor level.

5

present much differently. They may have se-vere bleeding in multiple sites, especially in muscles and soft-tissue, with extensive bruis-ing. “Elder abuse” may be suspected. The startling pattern of bleeding in these patients is unexplained. Course in hemophilia The immune response to exogenous FVIII (or IX) may be weak or robust. Most in-hibitors are “high responders”, that is , within a few days of repeated exposure to the exoge-nous clotting factor, the inhibitor level rises briskly (“anamnesis”), peaking within a month. If there is no further exposure to the exogenous factor, the inhibitor level gradually falls (in most patients) and, after years, may be undetectable.

A few low-level inhibitors in hemo-philia patients are “low responders”, that is, their levels do not rise notably after exposure to the exogenous factor. Such patients can continue treatment with the factor, in a higher-than-usual dose (Figure 2.) Auto-antibodies usually do not rise after exposure to exogenous clotting factor. Laboratory Diagnosis A common laboratory screening test for inhibitors is an activated partial throm-boplastin time (APTT) on an equal mixture of patient and normal plasma incubated to-gether for two hours at 37̊ C. After incuba-tion, in the presence of an inhibitor, the APTT is prolonged compared to controls without inhibitor, as in the actual examples in Figure 3.

INCUBATION MIXTURE: APTT at outset of incu-

bation. seconds:

APTT after two

hours, seconds:

Normal plasma alone 32 40

Factor VIII deficient plasma alone 90 95

Normal plasma plus

factor VIII deficient plasma with

No inhibitor

1 Bethesda unit

5 Bethesda units

20 Bethesda units

37

37

43

54

45

53

64

92

Figure 3 : APTTs with plasmas containing various FVIII inhibitor levels.

6

In a person with hemophilia, a pro-longed APTT of a mixture of patient and nor-mal plasma strongly suggests the presence of an inhibitor to the deficient clotting factor. In a non-hemophilic person, however, such a result suggests the presence of some inhibitor, but not the identity of the antigen. Similar results might be seen with an anti-phospholipid (“lupus”) inhibitor. Specific tests for such inhibi-tors can be performed. Assays of the antigenic clotting factor are low whether the patient’s plasma is tested at a small or a great dilution. However, assay systems for other clotting factors may be af-fected by the inhibitor. If a patient plasma is low in FVIII and also contains antibody to FVIII, the effects on the FVIII and other clotting factor assays might be as shown in Figure 4. These results show the apparent levels of four clotting factors as-sayed in an actual plasma sample containing an inhibitor to FVIII. Each clotting factor was measured in a one-stage APTT-based assay. The apparent level of FVIII is similarly low whether the patient plasma is slightly or greatly diluted in buffer. The apparent levels of factors IX, XI and XII increase as patient plasma is diluted. As the inhibitor in the patient plasma is “diluted out”, it has a decreasing det-rimental effect on the FVIII present in the re-agent plasmas deficient in factor IX, in factor XI or factor XII. Each reagent plasma is defi-cient in one factor and has a normal amount of all other clotting factors, including FVIII.

If plasma from the patient in Figure 4 is assayed at only two dilutions, for example, 1:5 and 1:10, the laboratory might report that the patient has a severe deficiency of FVIII and a mild deficiency of other clotting fac-tors. . Inhibitors are quantified by the “Bethesda” test, introduced in 1975. Normal pooled plasma (as a source of FVIII) is incu-bated with an equal amount of undiluted pa-tient plasma for two hours at 37 degrees C and then assayed for residual FVIII. A con-trol consists of normal plasma incubated with buffer. One inhibitor unit (Bethesda Unit, BU) is defined as the amount that destroys half the FVIII in the incubation mixture, corrected for the deterioration of FVIII in the control mixture. The assay can be modified to measure inhibitors to factor IX. Inhibitors with simple reaction kinetics are quantified easily by the Bethesda test but those with complex kinetics can be quantified only roughly. In the “Nijmegen” modification of the Bethesda test, introduced in 1995 and now widely used, the control consists of normal plasma incubated with immune-depleted FVIII deficient plasma instead of with buffer, and the normal plasma is buffered with imi-dazole to pH 7.4.

Patient plasma, diluted: Factor VIII Factor IX Factor XI Factor XII

1:5 <1 28 20 38

1:10 <1 36 29 47

1:20 <1 42 38 54

1:50 <1 60 51 72

1:100 <1 74 63 84

Figure 4. Apparent clotting factor concentration in four clotting factor assays at vari-ous dilutions of a plasma sample containing an inhibitor to FVIII.

7

10

100

0 0.5 1 1.5 2

Bethesda Units per mL Plasma

Co

rre

cte

d P

erc

en

t R

es

idu

al

Fa

cto

r V

III

75

50

25

Figure 5. The reference graph for the Bethesda test.

If the residual FVIII, after incubation, is 100% of the level in the control incubation mixture, then the

inhibitor level is zero. If the residual FVIII is 50% of that of the control, then the inhibitor level is one

Bethesda unit. The graph is used between 25 and 75% (solid line). Results of more than 75% are

within the error of the assay.

If the result is less than 25%, then the patient plasma is tested at various dilutions until the result can

be read off the graph. The result then is multiplied by the dilution to assign Bethesda units. For exam-

ple, if a plasma sample is diluted 1:5 before incubation and the residual FVIII then is 50%, or one

unit, 1 X 5 = 5 Bethesda units.

8

MANAGEMENT OF HEMORRHAGES Hemorrhages in the presence of inhibi-tors are managed by lowering inhibitor levels (available only in a few countries), by raising plasma levels of the deficient factor or by “bypassing” the need for the deficient factor. Lowering inhibitor levels It is possible to lower inhibitor levels, but not practical in the USA due to the com-plexities of health insurance. If equipment is available and the situa-tion is not urgent, the level of an inhibitor may be lowered in order to make it feasible to neu-tralize the remaining circulating inhibitor and raise the plasma clotting factor level. Ex-change plasmapheresis can be performed with continuous-flow cell-separation centrifuges which can replace a liter of plasma in 1/2 to one hour in a typical adult, if venous access is good. Replacement of 3 to 4 liters of plasma in typical adults temporarily reduces plasma inhibitor levels by 40% or more. If time per-mits, e.g. before an urgent but not emergency

surgical operation, exchanges may be per-formed on 2 or 3 consecutive days, if needed. Anamnesis is likely to be provoked within a few days of exposure to normal plasma, thus, FVIII (or IX) concentrate should be given promptly after the final plas-mapheresis to achieve a maximal peak plasma level. Any surgery should be timed to co-incide with that peak. In well-equipped centers, the efficacy of plasmapheresis can be improved by pour-ing the plasma, while outside the body, through columns containing protein A sepha-rose, which binds most IgG. Over a period of 6 or more hours, some 5 to 9 liters of patient plasma can be processed with reduction of inhibitor levels by two-thirds or more. This technology, pioneered in Sweden, is primar-ily available in northern Europe. Intravenous infusion of normal hu-man gamma globulin in high doses, by itself or after plasmapheresis, has been reported to lower inhibitor levels immediately, to some extent perhaps by the action of anti-idiotypic antibodies. Some degree of long-term inhibi-tor suppression also has been reported.

STRATEGIES for managing bleeding in the presence of an inhibitor (see also Figure 6):

Assay inhibitor levels periodically when the patient is not bleeding, so that when a hemorrhage

occurs, a decision about appropriate therapy can be made quickly.

For minor or routine hemorrhages, avoid stimulating anamnesis, so that the level of the in-

hibitor is as low as possible if and when a serious hemorrhage occurs. For routine hemor-

rhages, use a bypassing agent instead of FVIII in hemophilia A. Use recombinant activated

factor VII, which contains no factor IX, instead of FEIBA® which does contain factor IX, in

hemophilia B. If red cell replacement is needed, washed red cells are preferred over whole

blood or packed red cells which introduce plasma which contains the antigenic clotting factors.

For life-threatening hemorrhages in hemophilia A, especially if the inhibitor level is not very

high, try to raise the FVIII level with infusions of FVIII concentrate. Consider use of porcine

FVIII. Consider giving a large bolus of FVIII even if the inhibitor level is high.

9

- 2

0

2

4

6

8

1 0

1 2

1 4

1 6

1 8

2 0

2 2

2 4

2 6

2 8

3 0

0 1 2 3 4 5 6 7 8 9 1 0 1 2 3 4 5 6 0 1 2 3 4 5 6 7 8 9 1 0 1 2 3 4 5

Hemorrhage:

Given Factor

VIII

Hemorrhage:

Given Factor

VIII

Hemorrhage:

Given APCC

0

10

20

30

0

20

40

60

80

Days DaysYears Months

TIME

Pla

sma

Fa

cto

r V

III,

U/d

LIn

hib

ito

r, B

thes

da

U

0 1 2 3 4 5 6 7 8 9 1 0 1 2 3 4 5 6 0 1 2 3 4 5 6 7 8 9 1 0 1 2 3 4 5

Figure 6. Avoiding inhibitor stimulation for all but critical hemorrhages.

The hemophilia A patient with a high-responding inhibitor whose course is illustrated above received

bypassing agents in the first few years after inhibitor diagnosis for “routine” hemorrhages to help

control bleeding and to avoid stimulating further production of inhibitor antibodies. On two occa-

sions, years apart, when his inhibitor level was low, life-threatening hemorrhages were controlled

with infusions of FVIII concentrate that raised his plasma FVIII level to 30%. He gained many years

of life, but the third life-threatening hemorrhage occurred not long after the second, while his inhibi-

tor level still was high. A bypassing agent was ineffective and he bled to death.

10

Raising plasma factor levels DDAVP (desmopressin) DDAVP mediates the release of FVIII and von Willebrand factor from cellular storage sites into the plasma. Its use may elevate plasma FVIII levels temporarily in those inhibi-tor patients who have the fundamental capac-ity to make FVIII, that is, non-hemophilic pa-tients or those with mild hemophilia A, in the presence of a low-level inhibitor (less than five Bethesda units). In these special and unusual circumstances, DDAVP may release enough FVIII to neutralize the circulating inhibitor and raise the plasma level of FVIII slightly, suffi-cient to stop minor bleeding or allow minor sur-gical procedures. Exogenous human factors A large bolus of human FVIII (for he-mophilia A) or human factor IX (for hemophilia B) may be infused directly, if the inhibitor level is not too high (under 5 BU, sometimes higher) to try to achieve a plasma level of 30 U/dl or more. (Factor IX concentrate should not be used in those patients whose factor IX inhibi-tors provoke allergic reactions.) The amount of concentrate needed is only roughly corre-lated with the inhibitor level. A suggested initial FVIII dose is 20-40 U/kg plus 20 U/kg per BU. The factor level should be measured immedi-ately after infusion so that further concentrate can be given if needed. Once an inhibitor is neutralized with boluses of concentrate, the dose needed for subsequent infusions the same day and over the next few days will be lower, until anamnesis supervenes. Some clinicians infuse FVIII in the presence of an inhibitor without attempting to attain measurable plasma FVIII levels, with some success. If the inhibitor level is not very high, FVIII may have time to act in coagulation before it is neutralized. In one such protocol, patients with FVIII inhibitors of less than 30 BU receive a bolus of 70-140 FVIII U/kg followed by a continuous infusion of 4-14 U/kg/hr. He-

mostasis often is achieved although meas-ureable FVIII levels are not always reached. Porcine factor VIII Porcine FVIII interacts well in the hu-man coagulation sequence but attaches poorly to anti-human-FVIII antibodies. Inhibi-tors arising in persons with hemophilia A neutralize, on average, only 20-25% as much porcine FVIII as human FVIII; the range is wide (see Figure 7.) Cross-reactivity of auto-antibody inhibitors tends to be even lower that that of hemophilic allo-antibodies. For many years, porcine FVIII made from the plasma of pigs was available as “Hyate-C®”. It was discontinued in 2004. Recently, a re-combinant porcine B-domain-deleted FVIII, “Obizur®”, was licensed for use in patients with auto-antibodies to FVIII. Porcine FVIII is given to patients with low anti-porcine-FVIII inhibitor levels, if such assays are available, in doses sufficient to neutralize that inhibitor activity and achieve a therapeutic circulating FVIII level. If such in-hibitor levels are not available, an initial dose of 200 U/kg is recommended, with further dosing depending on plasma FVIII levels or clinical response. Some patients develop antibodies to porcine FVIII after several days of treatment or after several episodes of treatment, and are no longer responsive (see Figure 8). Bypassing agents Activated clotting factors such as Xa or VIIa may trigger some degree of coagula-tion in the absence of factor VIII or IX or in the presence of an inhibitor to one of them. Thus, such activated factors may “bypass” the need for factor VIII or IX (see Figure 9.) Prothrombin complex concentrate (PCC) contains factors II (prothrombin), VII, IX and X, which travel together in plasma fractionation. PCC was developed to treat

11

2.0

4.3 4.4

5.6

6.5

7.0 7

.5 8.0 8

.4

9.0

10

.0

10

.0

12

.0

12

.0 12

.5

0.5 0.7

0.0

0.7

0.7 1

.0

1.9

1.8

3.0

1.6

1.3 1

.8

0.8

3.0

1.7

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

11.0

12.0

13.0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Inh

ibit

or,

Beth

esd

a U

nit

s

Human

Porcine2

5.0

27

.0

27

.0 33

.0

34

.5

36

.0 40

.0

53

.0

69

.0

96

.0

10

5.0 1

16

.0

13

8.0

14

0.0

2.3

0.5 2.2

2.7

1.6 4.0 5.0 6.5

0.0

45

.0

17

.0

40

.0

24

.0

16

.0

0

10

20

30

40

50

60

70

80

90

100

110

120

130

140

150

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Inh

ibit

or,

Be

the

sd

a U

nit

s

Figure 7. Comparison of anti-human and anti-porcine inhibitor levels in 29 patients.

The above graph depicts anti-human-FVIII-inhibition (solid bars) and anti-porcine-FVIII-inhibition

(grey bars) in my laboratory in the first 29 inhibitor patients studied. Patients with lower levels of an-

tibody to human FVIII are shown in the upper graph and those with higher levels on the lower graph.

Inhibition of porcine FVIII was lower than that of human FVIII in all instances. The difference was

dramatic in some instances.

12

Figure 8. Course of inhibitor patient treated with porcine FVIII for emergency surgery.

A man with a high-responding inhibitor needed emergency surgery. At the time, his inhibitor level was 8

BU . The first doses of porcine FVIII neutralized the inhibitor and his plasma FVIII level was adequate

for surgery. By post-op day 5, his level of anti-human FVIII began to rise. By post-op day 7, his level of

anti-porcine FVIII was rising. By post-op day 6, porcine FVIII was no longer effective. His later course

included some oozing from the incision which was controlled with infusions of prothrombin complex

concentrate.

13

hemophilia B. It has been superceded, for that purpose, by concentrates of factor IX alone. A portion of the factors in PCC activate sponta-neously during processing. Activation can be enhanced deliberately to create “activated” prothrombin complex (APCC), also known as anti-inhibitor coagulant complex. PCC and APCC were used to treat bleeding in patients with inhibitors starting in the early 1970’s. Nowadays, PCC is no longer widely available. Of the two APCCs once manufactured, Autoplex® was discontinued in 2005 but FEIBA® is still widely used. A single factor can be separated and activated. A concentrate of plasma-derived activated factor VII was developed in the early 1980’s and used successfully in patients with inhibitors in Scandinavia. A recombinant factor VIIa (rVIIa) concentrate was then prepared in

Denmark (NovoSeven®). FVIIa may need tissue thromboplastin to initiate the activation of factor X. Tissue thromboplastin is present at sites of injury but is not abundant else-where, thus, the coagulant effect of exoge-nous FVIIa may be focused on sites of in-jury, a probable safety advantage. Activated factor X also has been used for hemostasis but only experimentally. Whereas the potency of PCC was described in factor IX units, the potency of APCCs is described in factor VIII correc-tional or bypassing units, unique to each brand. The dose of FEIBA® recommended in the package insert is 50-100 units/kg at intervals of 6 to 12 hours. The potency of rVIIa is expressed in weight in micrograms. A dose of 90 mcg/kg, which can be repeated at intervals of 2.5 to 3 hours, is recom-

Figure 9. A simplified coagulation scheme. Factor X may be converted to activated factor X either by the contact-activation (“intrinsic”) se-

quence on the left, or the factor VII-tissue thromboplastin (“extrinsic”) sequence on the right. (These

sequences are depicted as separate for clarity; in reality, they interact.) Activated factor X is one of

the essential catalysts of the conversion of prothrombin to thrombin. Activated factor VII, in the pres-

ence of thromboplastin, may trigger some formation of factor Xa in the absence of factor VIII or IX.

14

mended in the package insert. Many patients use double or triple that amount as an initial dose. The European Union approved single doses of 270 mcg/kg as an alternative to re-peated smaller doses. Patients with inhibitors who bleed fre-quently may benefit from prophylaxis, that is, a dose of APCC or rVIIa given daily or every other day. Sometimes FEIBA® and No-voSeven® are given alternatively, for treat-ment of bleeding or for prophylaxis. A patient often feels that one or the other agent is more effective in his case, or that alternating doses is helpful. In some institutions, measurement of “global” coagulation, as by thromboelasto-graphy, may be used to follow a patient, but there is no specific laboratory test to confirm that hemostasis is being promoted. One must observe the patient clinically. Other brands of rVIIa concentrate, and modifications of rVIIa to prolong its half-life, are under development. Efficacy The efficacy of bypassing agents has been difficult to establish. How do we know just when internal bleeding (e.g., in a joint) stops? Diminution of pain is the first, albeit subjective, indicator. Range of motion im-proves more slowly. By the late 1970’s, doubts had arisen that PCC was effective at all, so comparison to albumin placebo was justified. The first controlled trial of PCC for joint hemorrhages in persons with hemophilia and inhibitors (see Figure 10) confirmed that a single dose of PCC was associated with clinical improvement within a few hours in about half of instances. Trials of Autoplex® (an APCC) versus Proplex® in the USA, and of FEIBA® versus Prothromblex® (a PCC) in Europe gave similar results. Reports of the efficacy of repeated doses of APCC, in un-controlled studies, with evaluation performed after two or three days, showed a higher suc-cess rate.

Clinical trials of NovoSeven® all fea-tured repeated doses at intervals of 2.5 to 3 hours. In rough summary, two doses of rVIIa successfully controlled about half of joint hemorrhages. In a trial of a single stan-dard dose of FEIBA® versus two standard doses of NovoSeven®, outcomes were simi-lar. Some clinicians give double or triple doses of Novo-Seven® at the outset, rather than giving it in repeated doses. Summaries of the most comparable completed trials may be found in Figure 10. Side-effects An occasional side-effect of PCC or APCC, seen only in a few sensitive patients with hemophilia A, is an anamnestic rise in the FVIII inhibitor. These plasma-derived products may contain residual FVIII. Such patients are better served by the use of rVIIa. Patients with inhibitors to factor IX may sometimes have allergic reactions when treated with plasma-derived products which contain factor IX, so they, too, are better served by the use of rVIIa. Other adverse side-effects of bypass-ing agents, seen predominantly after inten-sive use, may include disseminated intravas-cular coagulation (DIC), deep vein thrombo-sis (DVT) and, in a few rare patients, includ-ing youths, receiving PCC or APCC, myo-cardial infarction. Autopsies revealed hemor-rhagic myocardial infarction without coronary thrombosis. Because of this mysterious but catastrophic complication, closely-spaced doses of PCC or APCC are avoided. Fewer instances of DIC and DVT have been re-ported with rVIIa than with PCC and APCC. The frequency of adverse events with these products is the subject of ongoing contro-versy, and vigilance is essential.

15

Figure 10. Key trials of bypassing agents for joint hemorrhages with a FVIII inhibitor

Lusher et alia, 1980, observed at six hours after a single infusion

“effective” (subjective)

Range of motion(ROM) improved 10º or more

Hemophilia A, no inhibitor, open-label, usual dose of factor VIII

100% 65%

Hemophilia A, inhibitor, double-blind

Konyne (PCC) 75 U/kg 48% 34%

Proplex (PCC) 75 U/kg 53% 33%

Albumin placebo 29% 18%

Lusher et alia, 1983, observed at six hours after a single infusion

“effective” (subjective)

ROM improved 10º or more

Hemophilia A, inhibitor, double-blind

Proplex (PCC) 75 U/kg 50% 43%

Autoplex (APCC) 50 U/kg 56% 48%

Autoplex (APCC) 75 U/kg 52% 52%

Sjamsoedin et alia,1981, observed at 24 hours after a single infusion

“effective” (subjective)

ROM improved 30º or more

Hemophilia A, inhibitor, double-blind

Prothromblex (PCC) 48 U/kg 46% 7%

FEIBA (APCC) 88 U/kg 66% 25%

Santagostino et alia, 1999, Italy Subjective efficacy

Hemophilia A, inhibitor, open label, home treatment

NovoSeven, 90 ug/kg, every 3 hours until effective

Median 2 doses needed

Key et alia, 1998, USA Subjective efficacy

Hemophilia A, inhibitor, open label, home treatment

NovoSeven, 90 ug/kg, every 3 hours until effective Median 2.2 doses needed

Astermark et alia, 2006, international Effective at 2 hrs

(before 2nd dose rVIIa)

Effective at 6 hrs

Hemophilia A, inhibitor, open-label, home treatment

NovoSeven, target 105 mcg/kg, 2 doses, second at 2 hrs 60% 79%

FEIBA 85 U/kg, one dose 75% 81%

16

Surgical operations Both types of bypassing agents have been used to promote hemostasis for emer-gency surgical operations. NovoSeven® has been used for elective surgical operations, with very frequent dosing. Reports and guide-lines mention occasional “oozing”, that is, he-mostasis may not be quite normal. Some sur-geons who are willing to undertake operations on soft tissue under cover of rVIIa are not will-ing to implant a “total” joint because imperfect hemostasis with a resulting small collection of blood at the site of the implant may set the stage for late infection. Such infections are more common in persons with hemophilia than in non-hemophilic patients, perhaps due to the provision of good bacterial breeding grounds in small hematomas. IMMUNE TOLERANCE INDUCTION Bonn (and similar) protocols Tolerance to the deficient clotting factor can be achieved in a large majority of patients with allo-antibodies to FVIII by giving them fre-quent, usually daily, large doses of FVIII over a prolonged period of time, a protocol pio-neered in Bonn, Germany. Anamnesis usually occurs in the first month of treatment but in-hibitor levels typically fall dramatically in the second month. Thereafter, they may fall more slowly. Complete tolerance consists of a sus-tained zero inhibitor level and a normal half-life of infused FVIII.

Tolerance is then maintained with low-level prophylaxis. Without prophylaxis, inhibitors may recur, but typically at lower levels, easily suppressed by resuming the tolerance regimen. Some patients achieve and maintain partial tolerance, that is, their inhibitor does not disappear completely but persists at a low level, behaving as a low-responding inhibitor. The ideal dosage is highly debated. Doses used range from 300 FVIII U/kg/day in Bonn to 25 FVIII U/kg every other day in The Netherlands. In the USA, the first dose tried was 50 U/kg/day but nowadays a dose of 100 U/kg/day is commonly used. Some centers add short courses of low-dose corti-costeroids if the fall in the inhibitor level is sluggish. Immunosuppressive drugs alone, without FVIII, are not helpful. Reports agree that the lower the in-hibitor level at the outset of tolerance at-tempts, and the lower the inhibitor level in the patient’s past history, the more likely he is to achieve tolerance, and the faster the process will be. A trial comparing the low Dutch dose with the high German dose showed that success could be attained with the lower dose but a longer period of treat-ment was needed. A registry maintained by Prof. Guglielmo Mariani of Italy (Figure 11) showed that higher doses are correlated with a greater chance of success. (He also re-ported that higher doses were associated with shorter periods of treatment.)

Figure 11. Prof. Guglielmo’s Registry as of 1993:

Proportion of patients achieving tolerance on daily factor VIII dose of:

Historic highest inhibitor

level, Bethesda units:

Under 50 units / kg 50 – 200 units / kg Over 200 units/ kg

Under 50 38.7 % 77.8% 88 %

50-500 25% 51.9% 83%

Over 500 Nil 6.3% 56%

17

Figure 13. Induction of immune tolerance with daily factor VIII (50 U/kg).

The course of three actual patients on induction of immune tolerance, 50 FVIII units/kg/day, is de-

picted. The lower, solid line depicts a patient whose inhibitor level rose sharply during the first

month, fell sharply in the second and then gradually fell to zero, a response seen in most patients with

a low inhibitor level at the outset. The middle, dashed line depicts a patient whose inhibitor also rose

in the first month and fell in the second but thereafter fell slowly, with some inhibitor still detected

after two years. The upper, dotted line depicts a patient whose inhibitor never fell notably and who

later proved to have a large factor VIII gene deletion. A relative with hemophilia and an inhibitor

also failed tolerance induction at another institution.

-10

0

10

20

30

40

50

60

70

80

90

100

0 1 2 3 4 5 6 7 8 9 10 11 12

Time, Months

Inh

ibit

or,

Be

the

sd

a U

nit

s

18

There is some evidence that plasma-derived FVIII containing von Willebrand factor is more effective in inducing tolerance than is recombinant FVIII. Nowadays, tolerance pro-grams usually are started as soon as an inhibi-tor is diagnosed. Typically such programs start with whatever FVIII concentrate the patient has been receiving. If tolerance is difficult to achieve, the patient may be switched to plasma-derived concentrates. In Bonn, only plasma-derived products containing von Wille-brand factor are used for tolerance. The probability of success in inducing tolerance is related to the underlying hemo-philia-causing mutation. Patients with muta-tions, such as large deletions, associated with a very high likelihood of inhibitor development also are the patients least likely to respond to tolerance protocols. They may require high doses, such as are used in Germany, and a prolonged period of treatment. Malmö protocol In Malmö, Sweden, an intensive two to four week inpatient tolerance protocol was used for many years. The inhibitor level was lowered with plasmapheresis, then FVIII was given in large doses to maintain a low-normal plasma level, and cyclophosphamide and in-travenous gamma globulin were administered. Tolerance was achieved in more than half of patients with the first round of treatment and in 70% with one or more additional rounds. The

protocol was appropriate for patients with inhibitors to factor IX, because plasmaphere-sis removed the inhibitor antibody that tends to cause allergic reactions; tolerance was achieved in 86%. Once all existing inhibitor patients in Sweden had been treated, new inhibitor pa-tients tended to be babies, for whom the Bonn protocol is more suitable. Tolerance for factor IX inhibitors Attempts to induce tolerance with the Bonn-style protocol have sometimes re-sulted in a nephrotic symdrome. Variants of the Malmö protocol, with immunosuppres-sive agents suitable to the age of the patient, are sometimes successful and are safer. Autoimmune inhibitors Autoimmune inhibitors arising in postpartum women respond quickly to corti-costeroids. Most autoimmune inhibitors arise in elderly persons, idiopathic in more than half of instances, or associated with autoim-mune or malignant disorders. Suppression of autoimmune inhibitors may be achieved in about three-quarters of patients with corti-costeroids alone or a combination of such steroids with cyclophosphamide. Other im-munosuppressive agents such as rituximab have sometimes been used successfully.

19

Recent reviews Green D. Factor VIII inhibitors: a 50-year perspective. Haemo-philia 2011 ; 17:831-7.

Lollar, P. Pathogenic antibodies to coagulation factors. Part one: Factor VIII and Factor IX. J Thromb Haemost 2004; 2:1082-5 Waters B, Lillicrap D. The molecular mechanisms of immuno-modulation and tolerance induction to factor VIII. J Thromb Haemost 2009; 7:1446-1456 Kruse-Jarres, R. Current controversies in the formation and treatment of alloantibodies to factor VIII in congenital hemo-philia A. Am Soc Hematol Educ Program 2011; 407-412 Witmer C, Young G. Factor VIII inhibitors in hemophilia A: rationale and latest evidence. Ther Adv Hematol 2013; 4:59-72. Sahud MA. Laboratory diagnosis of inhibitors. Semin Thromb Hemost 2000; 28:195-203. The diagnosis and management of factor VIII and IX inhibitors: a guideline from the UK haemophilia centre directors’ organi-zation. Br J Haematol 2000; 111:78-90.

Incidence and Prevalence

Briet E, Rosendaal FR, Kreuz W et alia. High titer inhibitors in severe haemophilia A : A meta-analysis based on eight long-term follow-up studies concerning inhibitors associated with crude or intermediate purity factor VIII products. Thromb Haemost 1994; 71:162-3. Data from 8 studies on 451 patients with severe hemophilia A followed from birth with continuous supervision are combined. The cu-mulative incidence of high-responding inhibi-tors was 10% by age 3 years and 20% by age 18years. This report covered an era when plasma-derived concentrates were used. Ehrenforth S, Kreuz W, Sharrer I, et alia. Incidence of develop-ment of factor VIII and factor IX inhibitors in haemophiliacs. Lancet 1992; 339:594-597. In Frankfurt, Germany, from 1976 to 1992, (an era when plasma-derived FVIII concen-trates were used), inhibitors developed in 14/27

(52%) of boys with severe hemophilia A fol-lowed prospectively. (This incidence is the highest ever reported and is often quoted, note that the number of subjects is small.) Kreuz W, Ettingshausen CE, Zyschka A et alia. Inhibitor development in previously untreated patients with hemo-philia A: a prospective long-term follow-up comparing plasma-derived and recombinant products. Semin Thromb Haemost 2002; 285-290. German patients from several centers were evaluated over a 23-year period. Inhibi-tors developed in 22/72 (32%) of all hemo-philia A patients after a median of 15 expo-sure days. Among patients with severe hemo-philia, 43% developed inhibitors. No differ-ence in incidence was seen in this small sam-ple between patients on plasma-derived vs. recombinant concentrate. UK Haemophilia Centre Doctors’ Organization. The inci-dence of factor VIII and factor IX inhibitors in the hemophilia population of the UK and their effect on subsequent mortal-ity, 1977-99. J Thromb Haemost 2: 1047-1054, 2004. In the UK, all hemophilia patients are tracked centrally. The cumulative risk of inhibitor devel-opment in an era of treatment with plasma-derived concentrates was 16% at age 5 years and 36% at age 75 years in severe hemophilia A. The cumulative risk of inhibitor develop-ment was 6% at age 5 years and 8% at age 75 years in severe hemophilia B. In patients with severe hemophilia, without HIV infection, in-hibitor development doubled mortality in the years 1977-92 but did not affect mortality rates in 1993-99 perhaps due to more treatment options. Hay CR, Palmer B, Chalmers E et alia. The incidence of factor VIII inhibitors throughout life in severe hemophilia A in the United Kingdom. Blood 2011; 117:6367-70. Between 1990 and 2009, 315 new in-hibitors were reported in 2528 patients with severe hemophilia A in the UK. The age of highest incidence was under 5 years, with 51% of the inhibitors arising at that time. In-hibitors could arise throughout life, and were less likely to arise in HIV-infected patients.

REFERENCES (CKK’s comments are in italics.)

20

Vézina C, Carcao M… Rivard GE et alia. Incidence and risk factors for inhibitor development in previously untreated severe haemophilia A patients born between 2005 and 2010. Haemo-philia 2014; 20:771-6. In Canada, 34 of 99 severe hemophilia A PUPs developed inhibitors (24 of high-titer) at median age 13.6 months. Six of seven aboriginal(native American) patients developed inhibitors. Miller CH, Hooper WC…. Soucie JM et alia. F8 and F9 muta-tions in US haemophilia patients: correlation with history of in-hibitor and race/ethnicity. Haemophilia 2012; 18:375-382. In subjects with severe hemophilia A, inhibitors were reported in 56/307 white non-Hispanic subjects, in 11/33 black non-Hispanic subjects and 11 of 28 Hispanic subjects. The differences between whites and non-whites were significant. Only 5% of subjects with severe he-mophilia B had inhibitors. Carpenter SL, Soucie MJ, Sterner S, Presley R. Increased prevalence of inhibitors in Hispanic patients with severe haemo-philia A enrolled in the Universal Data Collection database. Haemophilia 2012; 18:e260-5. Among 5651 males with severe hemo-philia A, the prevalence of high-titer inhibitors among Hispanic participants was 24.5% com-pared to 16.4% for white non-Hispanic patients. This report and that above come from a USA regis-try and the lower overall prevalence of inhibitors re-flects, in part the exclusion of low-titer inhibitors, and, perhaps, less intensive surveillance. Kempton CL. Inhibitors in previously treated patients: a review of the literature. Haemophilia 2010:16:61-65. Review New inhibitor formation in persons with hemophilia A and more than 150 lifetime expo-sures to FVIII concentrate is rare, occurring be-tween 1.55 and 3.8 per 1000 person-years. Previ-ously-treated patients (PTPs) are the first to re-ceive new FVIII concentrates in clinical studies under the rationale that a product with increased antigenicity is more likely to be detected be-cause of a low baseline inhibitor risk compared to PUPs. Xi M, Makris M… Mannucci PM, Iorio A. Inhibitor development in previously treated hemophilia A patients : a systemic review,

meta-analysis, and meta-regression. J Thromb Haemost 2013; 11:1655-1662. The authors searched for studies of at least 25 previously-treated patients who were treated with any type of FVIII concentrate. They found 33 cohorts with 4323 subjects who had 43 de novo inhibitors, for an inci-dence of three such inhibitors per 1000 per-son-years. Katz J. Prevalence of factor IX inhibitors among patients with haemophilia B: results of a large-scale North American study. Haemophilia 1996;2:28-31. Inhibitors were present in 28/735 pa-tients with severe hemophilia B (3.8%), 1/644 with moderate hemophilia B and none in mild hemophilia B DiMichele D. Inhibitor development in haemophilia B: an orphan disease in need of attention. Br J Haematol 2007; 138:305-315. (Review) Chitlur MN, Warrier I, Rajpurkar M, Lusher JM. Inhibitors in factor IX deficiency: A report of the ISTH-SSC international FIX inhibitor registry (1997-2006). Haemophilia 2009;15: 2017-1031. In an international registry of 94 per-sons with hemophilia B and inhibitors, the median age at inhibitor detection was 19.5 months with a median of 11 exposure-days. (Anaphylaxis was reported in 56 patients. Large gene deletions and null mutations con-veyed the greatest risk of anaphylaxis. Of 39 patients who underwent some form of induc-tion of immune tolerance, only five were com-pletely successful. Fischer K, Iorio A, Lassila R, et alia. Inhibitor development in non-severe haemophilia across Europe. Thromb Haemost 2015; 114:670-5. Over a five-year period (2008-2012), participating centers reported 7,969 patients with mild-moderate hemophilia A of whom 37 developed inhibitors (0.43/100 treatment-years) and 1,863 patients with mild-moderate hemophilia B, of whom one developed an in-hibitor (0.05/100 treatment years.) Inhibitors occurred at a median age of 35 years after a

21

median of 38 exposure-days. Eckhardt CL, van Velzen AS, Peters M et alia. INSIGHT Study group. Factor VIII gene (F8) mutation and risk of inhibi-tor development in nonsevere hemophilia A. Blood 2013: 122: 1954-1962. In an international survey of 1112 non-severe hemophilia A patients, 59 were found to have an inhibitor (5.3%) after a median of 24 exposure-days. Among 214 different F8 mis-sense mutations, 19 were associated with in-hibitor development.

F8—F9 genotypes and inhibitors Gill JC. The role of genetics in inhibitor formation. Thromb Haemost 1999; 82:500-504. Review. Oldenburg J, Schroder J, Brackmann HH et alia. Environ-mental and genetic factors influencing inhi-bitor development. Semin Hematol 2004; 41:82-8. The inhibitor risk in hemophilia A is greatest (88%) with large deletions covering multiple domains. The most common hemo-philia mutation, intron 22 inversion, was associ-ated with 21% inhibitors. Missense mutations in the C1/C2 domain were associated with a 10% inhibitor incidence whereas missense muta-tions elsewhere had only a 3% risk. Gouw SC, van den Berg HM, Oldenburg J et alia. F8 gene mutation type and inhibitor development in patients with se-vere hemophilia: systematic review and meta-analysis. Blood 2012; 119:2922-2934. Data from 30 studies on 5383 patients, of whom 1029 had inhibitors, showed that the risk was highest with large deletions and non-sense mutations and lowest with small dele-tions/insertions and missense mutations. Schwaab R, Pavlova A...Oldenburg J. Significance of F8 mis-sense mutations with respect to inhibitor formation. Thromb Haemost 2013; 109:464-70. Missense mutations (374 different ones) were identified in 1135 hemophilia A patients; of whom 46% had mild hemophilia. Of the 720

whose inhibitor status was known, only 5% had inhibitors. Inhibitor prevalence was four-fold higher for severe hemophilia than for milder types (perhaps due to less exposure to exogenous clotting factors in milder hemo-philia). Mutations associated with inhibitor formation were especially clustered within the C1/C2 domain (8.7% inhibitors vs. 3.6% for non-C1/C2 domain.) Inhibitor risk is higher is the amino acid substitution belongs to a dif-ferent physioco-chemical class. Yada K, Nogami K...Shima M. Mild hemophilia A patient with novel Pro1809Leu mutation develops an anti-C2 anti-body inhibiting allogeneic but not autologous factor VIII ac-tivity. J Thromb Haemost 2015; 13: 1843-53. A patient with mild hemophilia A and a missense mutation in the A3 domain devel-oped an antibody that recognized a C2 epi-tope. It neutralized allogenous but not auto-genous FVIII. This is not the only example of an inhibitor in mild hemophilia A that solely inhibits exogenous FVIII. Viel KR, Ameri A,...Howard TE. Inhibitors of factor VIII in black patients with hemophilia A. New Engl J Med 2009; 360: 1618-27. Six wild-type FVIII proteins were found and designated haplotypes 1-6 (H1-6). H1 and H2 were found in all racial groups and were the only types found in white people. H1 is the basis of Kogenate and H2 of Recombi-nate/Advate recombinant FVIII concentrates. H3, H4 and H5 (which have substitutions in the immunogenic regions of the protein) were found only in black people. (H6 was found in some Chinese persons.) Among 78 black he-mophilia A patients, 24% had a background H3 or H4 haplotype. The odds ratio for preva-lence of inhibitors in H3-4 patients was 3.6 vs. those with an H1-2 background. Perhaps standard FVIII concentrate is mismatched to the H3-4 patients. The preponderance of the haplotypes H3-4 among black inhibitor patients has not been substantiated in later studies, but all studies suffer from small numbers of subjects. Warrier I. Management of haemophilia B patients with in-hibitors and anaphylaxis. Haemophilia 1998;4:574-6.

22

Total gene deletions and major gene derangements were found in the majority. In 30 patients, anaphylaxis developed at a median of 11 exposure-days. Hemorrhages are best man-aged with recombinant activated factor VII. Attempts at immune tolerance in such patients were rarely successful and sometimes compli-cated by the nephrotic syndrome. Saini S, Hamasaki-Katagiri H...Sauna ZE et alia. Genetic de-terminants of immunogenicity to factor IX during the treatment of haemophilia B. Haemophilia 2015; 21:210-18. Mutations in patients with hemophilia B with known inhibitor status from an interna-tional database were classified. Inhibitors were associated with large deletions of more than 50 basepairs (39%) but not with smaller deletions. Inhibitors were associated with 60% of non-sense mutations if synthesis was terminated before the first 100 amino acids but only 23% if synthesis was terminated later. Inhibitors were associated with 27% of frameshift mutations. Inhibitors were associated with only 1.2% of missense mutations, the type of mutation which accounts for 56% of hemophilia B.

Immunology Pratt KP, Thompson AR. B-cell and T-cell epitopes in anti-factor VIII immune response. Clinic Rev Allerg Immunol 2009; 37:80-95. Review Wroblewska A, Reipert BM, Pratt KP, Voorberg J. Dangerous liaisons: how the immune system dweals with factor VIII. J Thromb Haemost 2012; 11:47-55. Review Reipert BM. Risky business of inhibitors: HLA haplotypes, gene polymorphisms, and immune response. Am Soc Hematol Educ Program 2014; 372-378. Review Astermark J. FVIII inhibitors: pathogenesis and avoidance. Blood 2015; 125:2045-2050. Review Tinlin S, Webster S, Giles AR. The development of homolo-gous (canine/anti-canine) antibodies in dogs with haemophilia A (factor VIII deficiency): A ten year longitudinal study. Thromb Haemost 1993; 69: 21-24. A purebred miniature Schnauzer with severe hemophilia A was bred to produce he-

mophilic descendants. He was mated to a normal Brittany spaniel; six of her nine hemo-philic descendants developed significant in-hibitors after treatment with canine cryopreci-pitate. The dog also was mated to his sister and only one of their 16 treated hemophilic descendants had an inhibitor, at a trace level, present only transiently. These results suggest that the two batches of the founder dog’s grand-sons inherited their grandmother’s immune sys-tems, one system conveying susceptibility to in-hibitors, and one not.

Pavlova A, ... Astermark J, Oldenburg J et alia. Impact of

polymorphisms of the major histocompatibility complex class

II, interleukin-10, tumor necrosis factor-alpha and cytotoxic

T-lymphocyte antigen-4 genes on inhibitor development in

severe hemophilia A. J Thromb Haemost 2009; 7:2006-15.

In 260 patients with severe hemo-philia, patients with inhibitors were matched with patients with the same mutation without inhibitors. Significantly higher frequencies of the DRB1*15 and DQB1*0602 alleles were found in inhibitor patients. The A allele of the 308G>A polymorphism of TNF-alpha was al-most twice as frequent in the inhibitor group as in case controls. High production of TNF-alpha/IL-10 was found in 12% of inhibitor pa-

tients and only 3% of non-inhibitor patients.

Bafunno V, …. Peyvandi F, Margaglione M. Polymorphisms

in genes involved in autoimmune disease and the risk of

FVIII inhibitor development in Italian patients with haemo-

philia A. Haemophilia 2010, 16:469-473

Distributions of single-nucleotide polymorphisms (SNPs) in the CRLA4, PTPN22, IL10, TNFα, FOXP3 amd IRF5 genes, all previously reported to be associated with autoimmune disorders, were not found to dif-fer between 115 hemophilia A patients with and 328 patients without inhibitors in Italy. Lozier JN, Rosenberg PS, Goedert JJ, Menashe I. A case-control study reveals immunoregulatory gene haplotypes that influence inhibitor risk in severe haemophilia A. Haemo-philia 2011; 17:641-9. In a study of severe hemophilia A in

23

Caucasians in the USA, interleukin haplotypes in 302 patients with inhibitors of more than one BU were compared to those in 633 patients with no inhibitor. Haplotypes associated with an in-creased inhibitor risk were found in IL10, IL12, and IL1α. Protective haplotypes were found in IL2 and IL1β. One rare F8 haplotype was seen in 3.1% of patients with inhibitors and only 0.25% of non-inhibitor subjects. Astermark J, Donfield S, Gomperts ED et alia. The polygenic nature of inhibitors in hemophilia A: results from the Hemo-philia Inhibitor Genetics Study (HIGS) combined cohort. Blood 2013:1446-1454. In a study of 833 hemophilia A patients, of whom 457 had a current inhibitor or a history thereof, 53 of 13,331 single nucleotide polymor-phisms (SNPs) in primary immune response and immune modifier genes predicted inhibitor status. Of the 13 most highly-significantly-related SNPs, 8 showed protective effects and 5 indicated increased risk; the latter were located on genes CD44, CSF1R, DOCK2, MAPK9 and IQGAP2. Whelan SFJ, Hofbauer CJ...Reipert BM et alia. Distinct char-acteristics of antibody responses against factor VIII in healthy individuals and in different cohorts of hemophilia A patients. Blood 2013; 121:1039-1048 Non-neutralizing FVIII antibodies were found in 116 of 600 normal blood donors (19%) (not the first such report, but the largest numeri-cally). Most appeared to be directed against the B-domain. Such non-neutralizing antibodies also were also found in 34% of persons with hemophilia who did not have neutralizing anti-bodies. Only patients with neutralizing antibod-ies had IgG4 antibodies.

Plasma-derived concentrates with high

antigenicity

Rosendaal FR, Nieuwenhuis HK... Briet E et alia. A sudden increase in factor VIII inhibitor development in multitransfused hemophilia A patients in The Netherlands. Blood 1993; 81:2180-2186. The Dutch Red Cross produced FVIII concentrate by controlled pore silica fractiona-tion. In 1990, viral inactivation was changed

from dry heat to pasteurization. Dutch patient surveillance showed that in the previous 27 months on dry-heated concentrate, only 4 inhibitors developed, all in newly-exposed patients, that is, the most vulnerable. During 18 months of use of pasteurized concentrate, inhibitors were reported in 11 patients, all but two of whom were previously-heavily-exposed patients, those believed to have low vulnerability.

Peerlinck K, Arnout J, Gilles JG, et alia. A higher than ex-pected incidence of factor VIII inhibitors in multitransfused haemophilia A patients treated with an intermediate purity pasteurized factor VIII concentrate. Thromb Haemost 1993;

69:115-118. The above pasteurized Dutch concen-trate also was used in Belgium starting in 1990. Five of 50 patients with severe hemo-philia A, all heavily transfused in the past, developed new high level inhibitors. Belgium then switched totally to solvent-detergent-treated (SD) concentrate.

Peerlinck K...Saint-Remy JMR, Vermylen J et alia. Factor VIII inhibitors in previously treated haemophilia A patients with a double virus-inactivated plasma derived factor VIII concentrate. Thromb Haemost 1997; 77:80 -86. In 1995, Belgium replaced the SD con-centrate mentioned above with a concentrate doubly viral inactivated with both SD and pas-teurization. Eight of 140 previously-heavily-transfused patients with severe hemophilia A developed inhibitors shortly after changing products. Raut S, DiGiambattista M, Bevan SA et alia. Modification of factor VIII in therapeutic concentrates after virus inactivation by solvent-detergent and pasteurization. Thromb Haemost 1998; 80:624-31. Investigations of the above event showed that the combination of SD plus pas-teurization induced changes in FVIII not found after other single or double viral inacti-vation processes. Josic D, Buchacher A, Kannicht C, et alia. Degradation products of factor VIII which can lead to increased immuno-genicity. Vox Sang 1999; 77 (suppl 1): 90-99

24

On investigation, plasma obtained from certain blood banks was found to already con-tain FVIII breakdown products. (Plasma may not have been separated from red cells promptly.) Concentrate produced from these sources was associated with inhibitor formation. The reason for the increased antigenicity of this product was debated; this report is from the manufacturer.

Recombinant concentrates and inhibitors Below are a few of the many papers comparing inhibitor rates in PUPs treated with plasma-derived vs. recombinant concentrates. Results showed ei-ther equal risk, or greater risk with recombinants. Iorio A, Halimeh S Holzhauer S et alia. Rate of inhibitor devel-opment in previously-untreated hemophilia A patients treated with plasma-derived or recombinant factor VIII concentrates. A systematic review. J Thromb Haemost 2010, 8:1256-65. In this meta-analysis of 24 studies cov-ering 1965 PUPs treated with plasma-derived FVIII and 420 with recombinant FVIII, inhibitor incidence was 14.3% with plasma-derived FVIII

and 27.4% with recombinant FVIII. Gouw SG, van der Bom JG, Ljung R et alia. Factor VIII prod-ucts and inhibitor development in severe hemophilia A. N Engl J Med 2013; 368:231-239. In 574 consecutive PUPs born between 2000 and 2010, inhibitors developed in 177 (32.4%). Plasma-derived and recombinant FVIII concentrates were associated with similar risks, but second-generation full-length recom-binants were associated with a higher risk of inhibitors. Collins PW, Palmer BP...Hay CRM et alia. Factor VIII brand and the incidence of factor VIII inhibitors in previously un-treated UK children with severe hemophilia A, 2000-2011. Blood 2014; 124:3389– 3408. Inhibitors developed in118 of 407 (29%) PUPs with severe hemophilia A born in the UK between 2000-9011 after a median of 16 expo-sure-days. All patients were on recombinant products. (The UK had had a recent “mad cow disease” epidemic and feared transmission of ab-normal prions in plasma products.) A second-

generation recombinant FVIII concentrate, Kogenate/Helixate, was associated with a higher incidence (35%) of inhibitors than a third-generation product, Advate (24%). This report and the previous one cast suspicion on second generation products, but another report disputed the association. Peyvandi F, Mannucci PM...Rosendaal RF et alia. Source of factor VIII replacement (PLASMATIC OR RECOMBINANT) and incidence of inhibitory alloantibodies in previously un-treated patients with severe hemophilia A: The multicenter randomized Sippet study. American Society of Hematology, 57th annual meeting, December 2015, Abstract # 5. Between 2010 and 2014, 251 PUPs from five continents were randomized to re-ceive a VWF-containing plasma-derived FVIII concentrate vs. recombinant FVIII. Inhibitors developed in 76 patients, of whom 50 had high titers. Recognized risk factors were equally divided between the two product groups. Recombinant FVIII was associated with 87% higher inhibitor incidence compared to plasma-derived FVIII. The difference per-sisted even when subjects on second genera-tion full-length recombinant concentrates were excluded from analysis. Oldenburg J, Lacroix-Desmazes S, Lillicrap D. Alloantibod-ies to therapeutic factor VIII in hemophilia A: the role of von Willebrand factor in regulating factor VIII immunogenicity. Haematologica 2015; 100:149-156. Binding of VWF to the C2 domain of FVIII may reduce immunogenicity by steric hindrance. Normally, some 94% of FVIII in the circulation is bound to VWF. In recombinant products, some 20% of FVIII may remain un-bound, which may facilitate an immune re-sponse. The authors suggest a few other mechanisms by which VWF may exert a pro-tective effect. See also: Shi Q, Kuether EL...Montgomery RR. Factor VIII inhibitors: von Willebrand factor makes a difference in vitro and in vivo. J Thromb Haemost 2012; 10:2328-2337. Ragni MV. Von Willebrand factor: factor VIII protector and friend. J Thromb Haemost 2012; 10: 2324-7.

25

Other treatment-related issues Batarova A, Holme P, Gringeri A, Richards M, Hermans C, Altisent C, Lopez-Fernandez M, Fijnvandraat K. Continuous infusion in haemophilia: current practice in Europe. Haemo-philia 2012; 18:753-9. Thirteen hemophilia centers reported a total of 1079 episodes of treatment with con-tinuous infusion of concentrate for surgery or major hemorrhages. Inhibitors developed in 0.45% of 659 patients with severe hemophilia A and in 7.2% of patients with mild hemophilia. (One presumes that those with mild hemophilia had less prior exposure than those with severe hemo-philia; previous to this report, there had been suspi-cion that continuous infusion was a risk factor.) Gouw SC, van den Berg HM, Fischer K et alia. Intensity of factor VIII treatment and inhibitor development in children with severe hemophilia A: The RODIN study. Blood 2013; 121:4046-55. In a multicenter study of 606 consecu-tive PUPs born between 2000 and 2010, the in-cidence of inhibitor development was 32% (high titer 22%). High-dose intensive FVIII treat-ment for bleeding or surgery was associated with an increased risk of inhibitor formation. During the first 20 exposure-days, patients us-ing on-demand treatment vs. prophylaxis had an equal risk of inhibitor development but thereafter prophylaxis was associated with a decreased risk. The authors suggest that some patients, based on their genetics, are highly unlikely to ever develop inhibitors whereas others are at very high risk, but the group in the middle is most susceptible to such influences as the mode of treatment. Aznar JA, Moret A, Ibáñez F et alia. Inhibitor development after switching of FVIII concentrate in multitransfused patients with severe hemophilia A. Haemophilia 2014; 20:624-9 In Spain, adequate medical records were available on 97 hemophilia A patients who had more than 150 exposure days at the time their concentrate brand was switched some time between 1970 and 2007. Multiple plasma-derived and recombinant products were used in this era and switches often were multiple. Nine patients developed inhibitors (9%), all low-level. Neither the number of different FVIII products

nor switching seemed to influence the risk. This paper and the next are included to refute the widespread “old wives’ tale” that switching con-centrate brands promotes inhibitor formation. Hay CRM, Palmer BP...Collins PW et alia. The incidence of factor VIII inhibitors in severe haemophilia A following a major switch from full-length to B-domain-deleted factor VIII: a prospective comparison. Haemophilia 2015; 21:219-26 Half of UK patients switched FVIII con-centrate brands in 2010 because of national contracting. Patients were tested for inhibitor prior to switching and at 6 month intervals thereafter. In patients with more than 50 prior exposure-days, five new low-titer inhibitors were reported, four in switchers and one in a non-switcher. The incidence rates did not dif-fer from each other or from the historical rate of 6/1000 treatment years.

Inhibitors in healthy persons Algiman M...Sultan Y, Kazatchkine MD et alia. Natural anti-bodies to factor VIII (anti-hemophilic factor) in healthy indi-viduals. Proc Natl Acad Sci USA 1992; 89: 3795-9.

Inhibitor activity against FVIII, found in 85/500 (17%) of plasma samples from healthy blood donors, ranged from 0.4 to 2 BU. Mean levels of FVIII activity were not dif-ferent in persons with and without these anti-bodies. Dietrich G, Algiman M...Kazatchkine MD et alia. Origin of anti-idiotypic activity against anti-factor VIII autoantibodies in pools of normal human immunoglobulin G )IVIg). Blood 1992; 79:2946-2951. A higher frequency of anti-idiotypic antibodies to FVIII inhibitors was found in do-nors in older age groups and in multiparous women (reflecting the population who tend to acquire auto-antibodies to FVIII). These donor groups may be good sources of such anti-bodies for therapeutic IV gamma globulin. Krudysz-Amblo J, Parhami-Seren B…Mann KG et alia. Quantitation of anti-factor VIII antibodies in human plasma. Blood 2009; 113:2587-94.

26

An immunoassay for the quantitation of anti-FVIII antibodies was developed. This test captures neutralizing and non-neutralizing antibod-ies. Detectable non-neutralizing antibodies were found in 4/150 healthy donors and in 13/39 in-hibitor-free hemophilia A subjects. Hofbauer CJ, Whelan SFJ...Reipert BM. Affinity of FVIII-specific antibodies reveals major differences between neutral-izing and nonneutralizing antibodies in humans. Blood 2015; 125:1180-1187 Antibody affinity was studied using a competition-based ELISA test. FVIII-specific antibodies found in inhibitor patients (with con-genital and acquired hemophilia) had an up to 100-fold higher affinity than those found in non-inhibitor hemophilia patients and healthy per-sons.

Characteristics of inhibitor antibodies

Gawryl MS, Hoyer LW. Inactivation of factor VIII coagulant activity by two different types of human antibodies. Blood 1982; 60:1103-1109. Type 1 antibodies completely destroy FVIII when antibody is present in high concen-tration, and may react with antigenic determi-nants near sites for procoagulant activity. Type 2 antibodies do not completely inactivate factor VIII; they react with more distant sites. Allain J-P, Frommel D. Antibodies to factor VIII. V. Patterns of immune response to factor VIII in hemophilia A. Blood 1976; 47:973-981.

In this classic article, “high-responding” patients were defined as those in whom “the antibody titers increased after each antigenic stimulation or persisted for years in the absence of transfusion” whereas “low-responding” patients were those in whom anti-body titers remained low, and in whom there was no significant difference in individual titers before and 8-10 days following transfusion.

Laboratory Diagnosis

Kasper CK, Aledort LM, Counts RB et alia. A more uniform measurement of factor VIII inhibitors. Thrombos Diathes Haemorrh (Thromb Hemost) 1975; 34:869-872. (letter)

The Bethesda test and “Bethesda” units are described and defined. The commit-tee of U.S. hematologists agreed to use it henceforth. Prior to this time, each lab tested inhibitors by their own method and assigned their own unitage.

Lossing TS, Kasper CK, Feinstein DI. Detection of factor VIII inhibitors with the partial thromboplastin time. Blood 1977; 49:793-797. The time-dependence of the reaction of factor VIII inhibitors with factor VIII, and its effect on APTT screening tests, is described.

Kasper CK. Laboratory tests for factor VIII inhibitors, their variation, significance and interpretation. Blood Coagulation and Fibrinolysis, 1991; 2:7-10. Review.

Verbruggen B, Novadova I...Mauser-Bunschoten E et alia. The Nijmegen modification of the Bethesda assay for factor VIII:C inhibitors: improved specificity and reliability. Thromb Haemost 1995; 73: 247-251. False-positive Bethesda tests were encountered in The Netherlands and attrib-uted to rising pH and to FVIII activation dur-ing incubation. Two test modifications were introduced (and have since become popular) including buffering of the normal plasma in incubation mixes with imidazole buffer and using, as a control, immuno-depleted FVIII-deficient plasma instead of buffer. Kasper CK. In vivo recovery and early half-life of infused factor VIII in haemophilia A. Haemophilia 1995; 1:14-16. In studies of 95 adults with severe he-mophilia A, who had not been transfused in the previous few days, the early phase T 1/2 of infused FVIII, as concentrate, was meas-ured at 3.25 hours. The early T ½ was used as a sensitive, routine pre-operative screen for inhibitor.

27

Sahud MA. Laboratory diagnosis of inhibitors. Semin Thromb Hemost 2000;26;195-203. Technical review.

Plasmapheresis

Cobcroft R,Tamagnini G, Dormandy KM. Serial plasmaphere-sis in a hemophiliac with antibodies to FVIII. J Clin Path 1977; 30:763-765.

Inhibitor levels were lowered with plas-mapheresis alone using older equipment. Freiburghaus C, Berntorp E, Ekman M, Gunnarsson M, Kjell-berg BM, Nilsson IM. Immunoadsorption for removal of inhibi-tors: Update on treatments in Malmo-Lund between 1980 and 1995. Haemophilia 1998; 4:16-20.

Inhibitors levels were lowered in 10 pa-tients on 19 occasions with plasmapheresis coupled with extracorporeal adsorption of anti-bodies given over a treatment period of 1-6 days. In all instances, it was possible either to eliminate the inhibitor totally or to reduce it to low levels at which remaining inhibitor was neutralized easily with concentrate infusions.

Jansen M, Schmaldienst S, Banyal S, Quehenberger P, Pabin-ger I, Derfler K, Hoerl WH, Knoebl P. Treatment of coagulation inhibitors with extracorporeal immunoadsorption (Ig-Therasorb). Br J Haematol 2001; 112:91-97.

An excellent description is provided of the use of plasmapheresis with extracorporeal adsorption of antibodies in 13 patients with FVIII inhibitors measuring 18 to 540 Bethesda units, treated on a total of 89 occasions. At each session, lasting about 3.8 hours, about seven liters of plasma were processed with a mean reduction in inhibitor level of about 72% per session.

DDAVP De la Fuente B, Panek S, Hoyer LW. The effect of 1-deamino 8 D-arginine vasopressin (DDAVP) in a nonhaemophilic pa-tient with an acquired type II factor VIII inhibitor. Brit J Haema-tol 1985; 59:127-131.

A patient with a FVIII auto-antibody of 1.8 Bethesda units was given DDAVP on two occasions and had a seven- to nine-fold rise in

plasma FVIII (to over 50 FVIII U/dL). Plasma FVIII levels remained in the hemostatic range for three hours after each infusion, allowing dental procedures to be performed.

Mudad R, Kane WH. DDAVP in acquired hemophilia A: Case report and review of the literature. Am J Hematol. 1993; 43: 295-299 The authors present an excellent sum-mary of the use of DDAVP in 22 patients with auto-antibodies to FVIII. Successful hemosta-sis was correlated with low inhibitor levels.

Human factor VIII

Berntorp E, Ekman M, Gunnarsson M, Nilsson IM. Variation in factor VIII inhibitory reactivity with different commercial factor VIII preparations. Haemophilia 1996; 2:95-99 Stimulated by in vivo observations, the authors incubated seven brands of FVIII concentrate in vitro with seven plasmas with different inhibitors to FVIII. Several inhibitors caused notably less neutralization of FVIII when it was present together with VWF (in concentrates “Humate-P” and “Koate HP”) than when it was presented as FVIII alone, in plasma-derived or recombinant concentrates. The authors suggest that some inhibitor pa-tients may respond better clinically to con-centrates containing FVIII with VWF.

Porcine factor VIII Giangrande P. Porcine factor VIII. Haemophilia 2012; 18:305-309. Review. Mannucci PM. Recombinant porcine factor VIII: a new in-stallment of a long story. Haemophilia 2015; 21:149-51. Review.

Kernoff PBA, Lilley TPA, Matthews KB, Goldman E, Tud-denham EGD. Clinical experience with polyelectrolyte-fractionated porcine factor VIII concentrate in the treatment of hemophiliacs with antibodies to factor VIII. Blood 1984; 63:31-41. A polyethylene-glycol fractionated porcine FVIII concentrate was first used in London in 1980. The chance of hemostasis after infusion was related to the ability to

28

raise plasma FVIII to a measurable level, and that response was related to a low anti-porcine-FVIII inhibitor level. Some patients had no an-amnesis after infusions and could use the con-centrate repeatedy.

Gatti L, Mannucci PM. Use of porcine factor VIII in the man-agement of seventeen patients with factor VIII antibodies. Thromb Haemost 1984; 51:379-384. Median cross-reactivity was 32%. An-amnesis occurred after 9/22 (41%) of treat-ments. Severe thrombocytopenia was seen in two instances.

Morrison AE, Ludlam CA, Kessler C. Use of porcine factor VIII in the treatment of patients with acquired hemophilia. Blood 1993; 81:1513-1520. In 65 patients with auto-antibodies to FVIII, treated with porcine FVIII for a total of 74 acute hemorrhages, there was a good to excel-lent response in 78%. These authors did not find a relationship between the probability of hemostasis and the initial antibody level or

plasma FVIII level attained after infusion. Hay CRM, Lozier JN… Kasper CK et alia. Safety profile of porcine factor VIII and its use as hospital and home-therapy for patients with haemophilia-A and inhibitors: The results of an international survey. Thromb Haemost 1996; 75:25-9. A retrospective international survey showed that the median cross-reactivity in 137 patients was 15%. No rise in antibodies to por-cine factor VIII after infusion was seen in 29% of recipients, an intermediate response was seen in 40% and a brisk response in 31%. Seven patients who did not have anamanesis were treated on-demand at home for 1.5 to 13 years (median 6.2 years) and another 23 pa-tients were treated regularly with porcine FVIII in the hospital for 2-7 years (median 3). These non-anamestic patients used porcine FVIII for 2000 bleeding episodes. The risk of transfusion reaction was dose-related. Such reactions were rare in the home setting, where dosage was low. A post-infusion fall in platelet count was common but usually transient and clinically insignificant. Occasional marked falls in plate-let counts were seen, usually with intensive re-placement therapy.

Kruse-Jarres R, St-Louis J...Ewenstein B et alia. Efficacy and safety of OBI-1, an antihaemophilic factor VIII (recombinant), porcine sequence, in subjects with acquired haemophilia A. Haemophilia 2015; 21:1-9. A recombinant, porcine B-domain de-leted factor VIII concentrate (subsequently trade-named “Obizur”) was successful in con-trolling hemorrhages in 24/28 subjects with acquired FVIII inhibitors. Ten patients had baseline neutralizing antibodies against por-cine FVIII; all but one responded well to treat-ment. No thrombocytopenia occurred. A stan-dard initial dose of 200 units/kg was used, after which dosage was titrated to maintain desired plasma FVIII levels. Five patients de-veloped de novo anti-porcine antibodies. The product currently is licensed for use in acquired hemophilia only.

Bypassing agents

Lusher JM, Shapiro SS, Palascak JE et alia. Efficacy of

prothrombin-complex concentrates in hemophiliacs with

antibodies to factor VIII: A multicenter therapeutic trial. New

Engl J Med 1980; 303:421-425.

In this first controlled study of by-passing agents, patients with severe hemo-philia A and inhibitor in the USA were treated for acute joint hemorrhages. As a preliminary, 20 non-blind control patients who did not have inhibitors were treated with a single dose of factor VIII and observed the same way as inhibitor patients; at six hours, all had symptomatic improvement and 65% had im-proved joint range of motion (ROM) of at least 10 degrees. None had any change in joint cir-cumference. In the double-blind controlled group, 51 patients with inhibitors were treated for 157 joint hemorrhages with a single dose of either Konyne® (a PCC) 75 factor IX U/kg, Proplex® (a PCC) 75 factor IX U/kg, or an in-travenous albumin placebo. Six hours after the infusion, patients judged the treatment efficacious in 29% of hemorrhages treated with albumin, 48% with Konyne®, and 53% with Proplex®. Range of motion was im-proved in 18% on placebo, 34% on Konyne® and 33% on Proplex®. Results with the two PCCs were significantly better than with albu-min.

29

Lusher JM, Blatt PM, Penner JA et alia. Autoplex versus

Proplex: A controlled, double-blind study of effectiveness in

acute hemarthroses in hemophiliacs with inhibitors to factor

VIII. Blood 1983; 62: 1153-1138

In a trial of similar design to that above, acute joint hemorrhages were treated with sin-gle doses either of Proplex® (75 factor IX units/kg) or Autoplex® (either 50 or 75 “factor eight correctional units” per kg.) Six hours after the infusion, patients judged the treatment effica-cious in 50% of hemorrhages treated with Proplex®, 56% with lower-dose Autoplex®, 52% with higher-dose Autoplex®. Differences were not significant.

Sjamsoedin LJM, Heijnen L, Mauser-Bunschoten EP et alia. The effect of activated prothrombin-complex concentrate (FEIBA) on joint and muscle bleeding in patients with hemo-philia A and antibodies to factor VIII: A double blind clinical trial. New Engl J Med 1981; 305:717-721. In this European trial, a large majority of hemorrhages treated were into joints. Hemor-rhages were treated with single doses either of Prothromblex® (a PCC, 48 factor IX units/kg) or FEIBA® (88 factor eight inhibitor bypassing units/kg). At 24 hours, Prothromblex® was ef-fective or partly effective in 46% and FEIBA® in 66% of instances. A 30-80% improvement in range of motion was seen in 7% of hemor-rhages treated with Prothromblex® and in 25% of those treated with FEIBA®. Giles AR, Mann KG, Nesheim ME. A combination of factor Xa and phophtidylcholine-phophatidylserine vesicles bypassed factor VIII in vivo. Br J Haematol 1988; 69:491-497. Various doses of infused activated fac-tor X and phospholipids were given to hemo-philic and to normal dogs to determine the therapeutic dose (shortening of the cuticle bleeding time) and toxic dose (causing DIC). (Other experimental bypassing agents containing Xa have been proposed, but none has been li-censed.)

Hedner U, Bjoern S, Bernvil SS, Tengborn L, Stigendahl L. Clinical experience with human plasma-derived factor VIIa in patients with hemophilia A and high titer inhibitors. Haemosta-sis 1989; 19:335-343.

Purified factor VIIa from human plasma was given in doses of 9-20 ug/kg (equivalent to 700-1000 U/kg). Plasma levels of factor VII were 330% to 610% 15 minutes after infusion. With the lowest dose level, a second dose clearly was needed for hemosta-sis. As dose levels went up, a second dose was not always clearly needed.

Lusher JM, Roberts HR, Davignon G et alia. A randomized, double-blind comparison of two dosage levels of recombi-nant factor VIIa in the treatment of joint, muscle and muco-cutaneous haemorrhages in persons with haemophilia A or B, with and without inhibitors. Haemophilia 1998; 4:790-798. Doses of 35 or 70 ug of rVIIa/kg were given at intervals of 2.5 hours, up to a maxi-mum of 6 doses, for treatment of 119 joint hemorrhages in 33 patients with hemophilia A (of whom 27 had an inhibitor) or B (of whom half had an inhibitor). Response was judged by assessing subjective pain and measuring decrease in size of hemorrhage at 8-14 hours after initiating treatment. In 144 hemarthroses (20 of which were in non-inhibitor patients), treatment was judged ef-fective or excellent in 71% of those given 35 ug/kg and 71% of those given 70 ug/kg. The median number of doses given for effective or excellent responses was two.

Key NS, Aledort LM, Beardsley D et alia. Home treatment of mild to moderate bleeding episodes using recombinant fac-tor VIIa (Novoseven) in haemophiliacs with inhibitors. Thromb. Haemost. 1998; 80:912-918. NovoSeven® (rVIIa), 90 ug/kg, was given every three hours for a maximum of four doses. Once the patient judged it effec-tive, one more dose was given for “maintenance” and then treatment was stopped. In about 47-48% of the courses judged effective, a median of 2.2 doses had been given prior to the maintenance dose.

Santagostino E, Gingeri A, Mannucci PM. Home treatment with recombinant activated factor VII in patients with factor VIII inhibitors: the advantages of early intervention. Brit J Haematol 1999; 104:22-26. NovoSeven®, 90 ug/kg, was given every three hours for a maximum of four

30

doses for hemorrhages treated within an hour of onset. The authors state: “two infusions were sufficient to achieve a successful out-come in more than half of the bleeding epi-sodes.”

Astermark J, Donfield SM, DiMichele DM et alia. A randomized comparison of bypassing agents in hemophilia complicated by an inhibitor: the FEIBA NovoSeven Comparative (FENOC) Study. Blood 2007; 109: 546-551. In an international, non-blinded, ran-domized crossover study of 96 joint hemor-rhages in 48 patients, treatment was either of a single dose of FEIBA® at about 85 U/kg, or No-voSeven® at about 105 U/kg given twice, the

second dose two hours after the first. At 6 hours, FEIBA® was effective in 81% and NovoSeven® in 79%. Kenet G, Lubetsky A, Luboshitz J, Martinowitz U. A new ap-proach to treatment of bleeding episodes in young hemophilia patients: a single bolus megadose of recombinant activated factor VII (NovoSeven®). J Thromb Haemost 2003; 1:450-455. In three young hemophilia patients with inhibitors, 95/114 hemorrhages responded to a single dose of NovoSeven® at 300 U/kg. A sec-ond dose was required in 19 episodes. This protocol gave faster relief than two doses of 90 U/kg.

Negrier C, Goudemand J, Sultan Y et alia. Multicenter retro-spective study on the utilization of FEIBA in France in patients with factor VIII and factor IX inhibitors. Thromb Haemost. 1997; 77:1113-1119. In France, FEIBA® was used between 1978 and 1993 in 433 bleeding episodes in 60 patients. In joint hemorrhages, bleeding was controlled with one dose in 50.7% of instances, 2 doses in 31.2 %, 3 doses in 7.4 %, and more than three in 10.7% of instances. In five of six major surgeries, FEIBA® provided effective he-mostasis; the sixth patient had excessive bleeding but a favorable outcome. Adverse events included anamnesis (more than a 50% rise in inhibitor level) after 31.5% of evaluable treatments, three instances of DIC and one myocardial infarct. The authors do not encour-age the use of FEIBA® in elective orthopedic surgery.

Schneiderman J, Rubin E, Nugent DJ, Young G. Sequential therapy with activated prothrombin complex concentrates and recombinant FVIIa in patients with severe haemophilia and inhibitors: update of our previous experience. Haemo-philia 2007; 13:244-248. In four patients, during a total of 35 admissions, activated prothrombin complex concentrate and rVIIA were given within six hours of each other for hemorrhages that had not responded to one of those agents given at home. After failing to respond to a median of three days of therapy with a single agent, hemorrhages resolved in a median of three days on combined therapy. No thrombotic events were seen but D-dimer elevations over 5 ug/ml were noted in 42% of courses lasting more than three days.

Ingerslev J, Sorensen B. Parallel use of by-passing agents in haemophilia with inhibitors: a critical review. Br J Haema-tol 2011; 155:256-62. In patients refractory to either No-voSeven® or FEIBA® used alone, these agents have been used together, either simul-taneously or in sequence. The authors review 17 reports in the literature of such use (in mixed doses) in 49 patients. Among nine pa-tients with acquired hemophilia, one patient developed deep vein thrombosis (DVT) and pulmonary embolism (PE), one developed fatal cerebral thrombosis and three devel-oped DIC. Among 40 patients with congenital hemophilia, one had a fatal myocardial infarc-tion, another had a fatal PE, another recov-ered from PE, one recovered from DVT and one recovered from DIC.

Martinowitz U, Lvnat T, Zivelin A, Kenet G. Concomitant infusion of low doses of rFVII and FEIBA in haemophilia patients with inhibitors. Haemophilia 2009; 15:904-10. Five hemophilia A patients with high inhibitor titers were treated with a combina-tion of 30 to 70 ug/kg FVIIa and 20-30 U FEIBA. This combination was effective in most hemorrhages and no adverse events occurred.

Holmstrom M, Tran HTT, Holme PA. Combined treated with

APCC (FEIBA) and tranexamic acid in patients with haemo-

31

philia A with inhibitors and in patients with acquired haemo-

philia A — a two-centre experience. Haemophilia 2012; 18:544

-549.

FEIBA® was given in combination with tranexamic acid in 6 hemophilia A patients with inhibitors and one patient with an acquired in-hibitor. Hemostasis was excellent in 10/11 bleeding episodes. No thrombosis developed and thromboelastography and thrombin gen-eration assays showed no signs of hyperco-agulability.

Valentino LA, Holme PA. Should anti-inhibitor coagulant com-

plex and tranexamic acid be used concomitantly? Haemophilia

2015; 21:709-14.

Reports of small series of patients re-ceiving bypassing agents plus tranexamic acid, in an attempt to gain a synergistic effect, are reviewed. There has been no increased inci-dence of thrombotic events. The authors con-clude that it might be tried in patients who fail with one bypassing agent.

Prophylaxis with Bypassing Agents

Leissinger C, Gringeri A, Antmen B et alia. Anti-inhibitor co-agulant complex prophylaxis in hemophilia with inhibitors. New Engl J Med 2011; 365:1684-1692.

In a prospective, randomized cross-over study, patients with high-titer inhibitors were either treated for hemorrhages (on demand) with about 85 U of FEIBA® /kg or given prophy-laxis three days a week with the same dose. On prophylaxis, patients had 62% fewer hemor-rhages than with on demand therapy. Target joint bleeding had been described in 70% of patients and such bleeding was reduced by 72%. No thrombotic complications were re-ported. Ewing E, Escuriola-Ettingshausen C, Kreuz W. Prophylaxis with FEIBA in paediatric patients with haemophilia A and in-hibitors. Haemophilia 2015; 21:1-7. Sixteen children under age 13 years with hemophilia A and inhibitors were given FEIBA prophylactically at doses of 75U/kg three

times a week (in USA) or 70-100 U/kg every day or every other day (Germany), adjusted as needed. Median number of joint hemor-rhages per year decreased from four to one. No thrombotic complications were seen. Antunes SV...Ewenstein B, Wong W-Y et alia. Randomized comparison of prophylaxis and on-demand regimens with FEIBA NF in the treatment of haemophilia A and B with inhibitors. Haemophilia 2014; 20:65-72. In Brazil, Ukraine, Russia and New Zealand, a total of 36 child and adult patients with hemophilia A or B and inhibitors who had a history of 12 or more hemorrhages in the past year were randomized to on-demand vs. prophylactic FEIBA (the latter dose was 85 U/kg every other day.) The median annual bleed rate in subjects on demand was 28.7 and in those on prophylaxis was 7.9. No thrombotic complications were seen. Konkle BA, Ebbesen LS, Erhardtsen E et alia. Randomized, prospective clinical trial of recombinant factor VIIa for sec-ondary prophylaxis in hemophilia patients with inhibitors. J Thromb Haemost 2007; 5:1904-13. Patients with hemophilia A or B and inhibitors of 2 BU or more, from 11 countries, who had at least two hemorrhages per month during a 3-month preliminary observation pe-riod, were randomized to receive either 90 or 270 ug/kg rVIIa, double-blind, once daily for three months. Each arm contained eleven pa-tients. Prophylaxis with 90 ug/kg rVIIa re-duced bleeds from 5.6 to 3.0 per month, and prophylaxis with 270 ug/kg reduced bleeds from 5.3 to 2.2 per month. The reduction in bleeds was a significant difference, but the effects of two dosages were not significantly different. No thrombotic complications were encountered. A very careful study. Young G, Auerswald G, Jimenez-Yuste V et alia. PRO-PACT: Retrospective observational study on the prophylac-tic use of recombinant factor VIIa in hemophilia patients with inhibitors. Thromb Res 2012; 130:864-70. In a retrospective study, data was re-viewed from 86 patients with hemophilia and inhibitors in 14 countries who had used pro-phylaxis with rVIIa. The median dose fre-

32

quency in children was 7 times/week, in adoles-cents 5.5 times/week and in adults 3 times/week. Individual doses varied. Overall, the fre-quency of bleeding was reduced from 1.37 to 0.74 bleeds/month. In patients who had had es-pecially frequent hemorrhages, the rate was reduced from 2.1 to 1.01 bleeds/month.

Surgery with Bypassing Agents Shapiro AD, Gilchrist GS, Hoots WK, Cooper HA, Gastineau DA. Prospective, randomized trial of two doses of r VIIa (NovoSeven) in haemophilia patients with inhibitors undergo-ing surgery. Thromb Haemost 1998; 80:773-778. NovoSeven® was given in doses of 35 ug/kg or 70 ug/kg (double-blind) pre-operatively, intra-operatively and for 48 hours post-operatively, every 2 hours as needed. If hemostasis was inadequate, an open-label “escape” dose of up to 180 ug/kg could be given. After the first 48 post operative hours, NovoSeven® was given at the same dose at intervals of two to six hours, as determined by the investigator, for another three days. After post-operative day 5, a dose of 90 ug/kg, open-label, was given as often as the investigator determined. The 70ug/kg initial dose proved adequate for 8 minor and 6 major surgeries through post-operative day 4. The 35 ug/kg dose was less satisfactory. Six patients left the study, five as treatment failures who were sub-sequently managed on FVIII or FEIBA®, and the sixth because of thrombosis of the internal jugular during central venous catheter place-ment.

Giangrande PLF,...Tuddenham EGD. Goddard NJ et alia.

Consensus protocol for the use of recombinant activated factor

VII [eptacog alfa (activated); NovoSeven® ] in elective ortho-

paedic surgery in haemophilic patients with inhibitors. Haemo-

philia 2009, 15:501-508

A protocol for use of rVIIa for surgical operations was described. Concomitant use of fibrinolytic agents is recommended to improve hemostasis. An initial bolus of 120-180 ug/kg immediately pre-op is followed by 90 ug/kg at two hour intervals during the surgical operation and for 48 hours thereafter. The dosing then is extended to every three hours. At about 5 days post-op, the interval is extended to every 4

hours. At about 8 days post-op, the interval is lengthened to every 6 hours. The patient usu-ally is discharged at day 10-12. If there is per-sistent oozing after surgery, physiotherapy is stopped, and the next dose brought forward. Increased doses can be considered, or, a product change to FEIBA®. This protocol was used in 10 operations with satisfactory intra-operative hemostasis in all. There were two instances of post-operative bleeding, one oc-curring in one of the three patients who had not received anti-fibrinolytic medications. In-creasing the dose resolved the bleeding. In another patient, bleeding at 8 hours post-op was treated with FEIBA® but persisted for 5 days. Valentino LA, Cooper DL, Goldstein B. Surgical experience with rFVIIa (NovoSeven) in congenital haemophilia A and B patients with inhibitors to factors VIII or IX. Haemophilia 2011; 17:579-589. Data from clinical trials and published reports was combined. It covered 395 proce-dures, of which 261 were surgical, including 100 orthopedic procedures. In the latter, he-mostasis was judged effective in 72, partially effective in 9, ineffective in 13 and not re-ported in 6. The incidence of thrombotic events was low (0.4% of patients.) Caviglia H, Mandela M... Bianco RP et alia. Elective ortho-paedic surgery for haemophilia patients with inhibitors: sin-gle centre experience of 40 procedures and review of the literature. . Haemophilia 2011, 17:910-919 Of 40 surgical procedures in Argen-tina, 15 were major (but none was a “total” joint replacement); rVIIa was the hemostatic cover in all major cases. Bleeding was greater than usually encountered in nine of the major procedures. With minor surgical procedures, bleeding was no greater than expected. Van Veen JJ, Maclean RM… Makris M et alia. Major sur-gery in severe haemophilia A with inhibitors using a recom-binant factor VIIa and activated prothrombin complex con-centrate hybrid regimen. Haemophilia 2014; 20:587-92. In the UK, eight surgical procedures (including six total knee replacements) were performed on four inhibitor patients under

33

cover of rVIIa together with tranexamic acid for the first 2-6 postoperative days and FEIBA for the remaining period. FEIBA was used to permit less frequent dosing at this period of recovery and to reduce costs. Five episodes of post-operative bleeding occurred, four in the same patient in connection with four different opera-tions. Tadedani H, Shima M, Horikoshi Y et alia. Ten-year experi-ence of recombinant activated factor VII use in surgical pa-tients with congenital haemophilia with inhibitors or acquired haemophilia in Japan. Haemophilia 2015; 21:374-9. Recombinant factor VIIa was judged effective for surgical hemostasis in 34/38 pro-cedures in patients with hemophilia A, 10/13 procedures in hemophilia B and 4/5 procedures in acquired hemophilia. A majority of patients were treated with repeated bolus doses in the range of 100-140 ug/kg but nine, including two joint replacements, were treated with continu-ous infusion of 29-36 U/kg/hr, by way of a saline line. Tranexamic acid was used concomitantly in 36/56 procedures. Those not responding ade-quately were treated with an activated prothrombin complex concentrate (APCC) or a combination of rVIIa and APCC. One patient developed mild thrombophlebitis at the site of infusion.

Complications of bypassing agents Green, D. Complications associated with the treatment of hae-mophiliacs with inhibitors. Haemophilia 1999: 5(suppl 3), 11-17. Review Dr. Green reviews reports of anamnesis, DIC and thrombosis after bypassing agents and of thrombocytopenia and severe allergic reac-tions after porcine factor VIII. Kasper CK. Effect of prothrombin complex concentrate on factor VIII inhibitor levels. Blood 1979; 54: 1358-1368. In the USA, FVIII inhibitor levels were measured before and after prothrombin com-plex concentrate treatment of 261 bleeding epi-sodes in 75 patients with hemophilia and inhibi-tor. A rise in inhibitor level to twice baseline or more was seen in 13.5% of episodes, in 27 pa-tients. (The plasma-derived concentrate contained traces of factor VIII.)

Yoshioka A, Kamisue S, Tanaka I, Kato M, Kohmura I, Shima M, Fukui H. Anamnestic response following infusion of prothrombin complex concentrates (PCC) and activated prothrombin complex concentrates (APCC) in haemophilia A patients with inhibitors. Blood Coagulation and Fibrinolysis 1991; 2 (suppl 1): 51-58. Six of 20 patients with high-responder inhibitors to FVIII occasionally had anamnes-tic inhibitor responses after treatment with PCC or APCC. FVIII antigen, measured in vi-als of one PCC and the two APCCs, was 11.7, 14.1 and 117 units/vial, respectively. FVIII was present predominantly as the light chain. Only small amounts of von Willebrand factor were found, and only in one APCC. Thus, the factor VIII present in these concentrates is modified, degraded and dissociated from vWF.

Chavin SI, Siegel DM, Rocco TA, Olson JP. Acute myocar-dial infarction during treatment with an activated prothrom-bin complex concentrate in a patient with factor VIII defi-ciency and a factor VIII inhibitor. Amer J Med 1988; 85: 245-249. Report and review. In the first nine myocardial infarct pa-tients reported, six were under age 30 years and four under age 20. Pathology findings were available for five patients. Myocardial hemorrhages were seen, but no evidence of atherosclerosis or thrombosis. Some patients also had old infarcts. Myocardial infarction tended to be associated with use of high doses of concentrate over several days.

Hough RE, Hampton KK...Makris M et alia. Recombinant VIIa concentrate in the management of bleeding following prothrombin complex concentrate-related myocardial infarc-tion in patients with haemophilia and inhibitors. Brit J Haematol 2000; 111: 974-979. Sixteen published instances of myo-cardial infarct after use of PCC or APCC are reviewed. Peerlinck K, Vermylen J. Acute myocardial infarction follow-ing administration of recombinant activated factor VII (NovoSeven) in a patient with haemophilia A and inhibitor. Thromb Haemost 1999; 82:1775-6. A 72 year old patient was given a bo-

34

lus of rVIIa at 102 ug/kg and then a total of 1 mg by continuous infusion, as well as tranexamic acid, prior to a dental extraction. Immediately after the extraction, he had a myocardial infarct. The patient survived so no pathologic report is available. The authors wonder whether tissue factor, exposed on atherosclerotic plaques, at-tracts activated factor VII and thus leads to myocardial infarct.

Diness V, Bregengaard C, Erhardtsen E, Hedner U. Recombi-nant human factor VIIa (rFVIIa) in a rabbit stasis model. Thromb Res 1992; 67:233-241. Typical doses of rVIIa or of FEIBA® were infused into rabbits with isolated vein seg-ments. Both drugs were associated with minor thrombus formation after 10 minutes and defi-nite thrombus formation at 30 minutes. At three hours, FEIBA® was associated with a signifi-cant decrease of platelet counts and fibrinogen levels whereas rVIIa was not.

Ehrlich HJ, Henzl MJ, Gomperts ED. Safety of factor VIII in-hibitor bypassing activity (FEIBA): 10-year compilation of thrombotic adverse events. Haemophilia 2002; 8:83-90. For the most recent 10 year period of post-marketing surveillance of FEIBA®, 16 thrombotic adverse events were documented, an incidence of 4.05 per 105 infusions. Included were 7 instances of DIC and 5 of myocardial infarction. Known risk factors were present in 13/16 instances, including overdose in half of instances.

Abshire T, Kenet G. Safety update on the use of recombinant

factor VIIa and treatment of congenital and acquired deficiency

of factor VIII or IX with inhibitors. Haemophilia 2008; 14:898-

902.

Safety data from licensure of No-voSeven® to April 2003 revealed 25 throm-boembolic events from over 700,000 doses given. The current report covers May 2003 through 2006. A total of 30 thromboembolic events, six of them fatal, occurred from about 800,000 doses. The drug may be more danger-ous when “used in off-label indications and when activated coagulation is present or when coagulation parameters are normalized during

the process of restoring hemostasis.”

O’Connell KA, Wood JJ, Wise RP, Lozier JN, Braun MM. Thromboembolic adverse events after use of recombinant human coagulation factor VIIa. JAMA 2006; 295:293-298. A total of 431 adverse events reports about recombinant VIIa were received by the FDA, of which 168 described 185 thromboem-bolic events. Seventeen events occurred in patients with hemophilia, 59 occurred in non-hemophilic patients enrolled in post-licensure trials, and, in the remainder, the drug had been used for off-label indications. In such instances, with many pre-existing medical conditions, arterial and venous thrombosis often resulted in serious morbidity and mor-tality. A close temporal association was noted between the dose of rVIIa and the throm-boembolic event in many instances where it could be assessed.

Aledort LM. Comparative thrombotic event incidence after infusion recombinant factor VIIa versus factor VIII inhibitor bypass activity. J Thromb Haemost 2004; 2:1700-1708. Data from the FDA, supplemented by case reports, were used in conjunction with estimates of the total number of infusions to assess comparative incidence of thrombotic adverse events in patients using NovoSeven® vs those using FEIBA® in the period April 1999 through June 2002. Incidence rates were 24.6 per 100,000 infusions for NovoSeven® vs 8.24 per 100,000 infusions for FEIBA®. (This report does NOT segregate recipients with hemo-philia from others who may have serious pre-existing medical conditions predisposing to thrombosis; NovoSeven® has been used more often in such “off-label” situations than has FEIBA®.)

Induction of Immune Tolerance : Bonn

Protocol and Similar Programs

Brackmann HH, Gormsen J. Massive factor VIII infusion in haemophilia with factor VIII inhibitor, high responder. Lancet 1977; 2:933.

35

The first deliberate induction of toler-ance at Bonn, Germany, is described. An excel-lent graph shows that the inhibitor rose in the first few days of intensive treatment and then fell, reaching one unit by three months.

Sultan Y, White GC, Aronstam A et alia. Hemophilic patients with an inhibitor to factor VIII treated with high dose factor VIII concentrate: Results of a collaborative study for the evaluation of factor VIII inhibitor titre, recovery and half life of infused factor VIII. Nouv Rev Fr Hematol 1986; 28:85-59. In this excellent study, performed be-cause of doubts about the success of immune tolerance, 18 patients were evaluated, most from the Bonn center, who had completed in-duction of tolerance. Twelve no longer had de-tectable inhibitor and six had questionable or very low inhibitors (maximum 2.4 Bethesda units) behaving as low-responders. All now could be treated with factor VIII.

Oldenburg J, Schwaab R, Brackmann HH. Induction of im-mune tolerance in haemophilia A inhibitor patients by the ‘Bonn protocol’: Predictive parameters for therapy duration and outcome. Vox Sang 1999; 77 (suppl 1): 49-54. In early years, the Bonn protocol con-sisted of two phases. In phase one, 100 FVIII U/kg and 50 FEIBA® U/kg were given twice daily until the inhibitor level fell to less than one BU and FVIII could be measured in the plasma after infusion. Thereafter, in phase two, 150 FVIII U/kg was given daily without FEIBA® until the inhibitor disappeared completely and the T ½ of infused FVIII was normal. In recent years, Bonn patients are put on the second phase dosage immediately and do not receive FEIBA®. In 60 patients (36 high responders and 24 low responders) induction of tolerance was either achieved (# 52, 86.7%) or stopped.(# 8). In 41 patients whose treatment was continuous, the median time until the inhibitor was less than one BU was 3 months and the median time to a normal T ½ was 11.6 months. In 11 patients whose tolerance programs were interrupted, the median time until the inhibitor was less than one BU was 18.8 months and the median time to a normal T ½ was 39.9 months. They did achieve complete tolerance in five patients whose inhibitor at the outset was over 100 BU. Central lines were used in five patients and all lines became infected.

Ewing NP, Sanders NL, Dietrich SL, Kasper CK. Induction of immune tolerance to factor VIII in hemophiliacs with in-hibitors. JAMA 1988; 259: 65-68

In this first Western Hemisphere se-ries, nine of 12 patients achieved complete tolerance in less than 11 months on a dose of 50 FVIII units/kg/day. The three unsuccessful patients included one with a large gene dele-tion who did not respond at all, and two who had had intensive FVIII treatment ending two months before the tolerance regimen was started. Those two patients did respond, but slowly, over years. Van Leeuwen EF, Mauser-Bunschoten EP...Sixma JJ et alia. Disappearance of factor VIII:C antibodies in patients with haemophilia A upon frequent administration of factor VIII in intermediate or low dose. Br J Haematol 1986; 64:291-297. In The Netherlands, 18 patients with hemophilia A and inhibitors under 9 BU were given 25 FVIII units/kg every other day. Treat-ment was stopped if the inhibitor level peaked at more than 80 BU. At the time of publica-tion, 12 patients had achieved tolerance.

Ter Avest PC, Fischer K, Gouw SC, van Dijk K, Mauser-Bunschoten EP. Successful low dose immune tolerance induction in severe haemphilia A with inhibitors below 40 Bethesda units. Haemophilia 2010; 16:71-79. The Dutch study experience is up-dated. Over 26 years, 21 patients were treated with 25-50 units FVIII/Kg every other day or three times a week. Tolerance was achieved in 18 patients. A low inhibitor level at the out-set and a low (under 40 BU) maximum rise during tolerance both were associated with faster time to tolerance. Mariani G, Scheibel E...Brackmann HH et alia. Immunetol-erance as treatment of alloantibodies to factor VIII in hemo-

philia. Seminars in Hematology 1994; 31 (suppl 4): 62-64. This international registry included 204 patients who underwent immune toler-ance. Of these, 82.3% were high responders with historic inhibitor peaks of more than 10 BU. Higher rates of success were seen in pa-tients with lower inhibitor levels at the outset

36

and in patients in whom higher doses of FVIII, over 100 U/kg/day, were used. DiMichele DM, Kroner BL. The North American immune toler-ance registry: practices, outcomes, outcome predictors. Thromb Haemost 2002;87:52-7 Data on 188 courses of tolerance induc-tion were collected from 1993-99. The success rate was 70% for hemophilia A and 31% for he-mophilia B. Lower historical peak inhibitor titers, pre-induction inhibitor titers, and peak titers on induction all were predictive of suc-cess in hemophilia A. An inverse correlation between daily FVIII factor dose and success was found, surprisingly, with less success us-ing higher doses, however, the range of American doses tends to be narrow, and lower than German ones. In hemophilia B, allergic events were common and were the major reason for toler-ance failure. Three/17 subjects developed nephrotic syndrome.

Hay Cr, DiMichele DM. The principal results of the Interna-tional Immune Tolerance Study: a randomized dose compari-son. Blood 2012; 119:1355-44. The “German” dose style of 200 FVIII U/kg/day was compared prospectively with the “Dutch” style of 50 U/kg three times a week; 66 inhibitor patients completed the protocol. Suc-cess did not differ between these treatment arms but the time to achieve a negative titer and normal in vivo recovery was shorter with the high dose arm. Kreuz W, Ettinghausen CE, Auerswald G et alia. Epidemiology of inhibitors and current treatment strategies. Haematologica June 2003; 8:EREPO4 “A longitudinal study of ITI at our center showed a significantly decreased success rate since the introduction of high purity plasma derived and recombinant FVIII products using the Bonn protocol. In inhibitor patients who showed an unsatisfactory response to treat-ment with FVIII concentrates with very little or no VWF the change to concentrates containing high amounts of VWF increased success rates up to 90%. These observations raise the ques-tion of whether VWF plays an important role in the induction of immune tolerance.”

Franchini M, Mengoli C, Lippi G et alia. Immune tolerance with rituximab in congenital haemophilia with inhibitors: a systematic literature review based on individual patients’ analysis. Haemophilia 2008; 14:903-12. In 49 patients with inhibitors, refrac-tory to standard immune tolerance induction, reported in 29 studies, rituximab was used to aid induction of tolerance, and was success-ful in 53%. No severe adverse reactions to rituximab were reported. Coppola A, MargaglioneM, Santagostino E et alia. Factor VIII gene (F8) mutations as predictors of outcome in im-mune tolerance induction of hemophilia A patients with high-responding inhibitors. J Thromb Haemost 2009; 7:1809-15 In 86 patients, the success of induc-tion of tolerance was related to the underlying mutation: 17% success with large deletions (n=6), 33% with splice site mutations (n=3), 48% with inversions (n=50) , 64% with non-sense mutations (n=11), 71% with small dele-tions (n=7), 87% with small insertions (n=87), 100% with missense mutations (n=1). Oldenburg J, Austin SK, Kessler CM. ITI choicefor the opti-mal management of inhibitor patients– from a clinical and pharmacoeconomic perspective. Haemophilia 2014; 20 (suppl 6):17-26. In Europe, using plasma-derived VWF-containing FVIII concentrate, primary immune tolerance induction was completely or partly successful in 87% of children and 89% of adults with hemophilia A and inhibitors. Use of that concentrate as “rescue” immune toler-ance, when programs with other products had failed, resulted in 70% success in chil-dren and success in one adult who tried it. Nearly all subjects received concentrate daily. There was no clear relationship between the daily dose and the rate of success. The me-dian time to complete success was about 18 months for primary protocols and 26 months in the rescue protocol. Favorable results with plasma-derived VWF-containing FVIII concen-trate (as opposed to recombinant FVIII which contains no VWF) continue to be reported, both for primary and for rescue immune toler-ance. Induction of immune tolerance is ex-pensive, but, over a lifetime, projected to be about 40% less costly than on-demand ther-

37

apy with bypassing agents, and about half as costly as prophylaxis with bypassing agents. Nakar C, Manco-Johnson MJ…Shapiro A et alia. Prompt im-mune tolerance induction at inhibitor diagnosis regardless of titre may increase overall success in haemophilia A compli-cated by inhibitors: experience of two US centres. Haemophilia 2015;21:365-73. Induction of immune tolerance over a 30-year period was analyzed. Among those with a high-responding inhibitor who started immune tolerance within one month of detection (n=23), 96% succeeded although 13/23 had an inhibitor titer of 10 or more BU at the outset. Among those with a high-responding inhibitor who started immune tolerance more than 6 months after diagnosis of an inhibitor, 64% succeeded. Thus, an inhibitor titer of 10 BU should not be an impediment to starting tolerance (as had been previously suggested.)

Malmö Protocol for Tolerance Induction Nilsson IM, Berntorp E, Zettervall O. Induction of immune tol-erance in patients with hemophilia and antibodies to factor VIII by combined treatment with intravenous IgG, cyclophos-phamide and factor VIII. New Engl J Med 1988; 318:947-950. In 9 of 11 patients with hemophilia A and inhibitor, the inhibitor disappeared after 2-3 weeks of an intensive inpatient regimen of treatment. Tolerance appeared to be stable on follow-up. Once such course of treatment suf-ficed in seven cases and two such courses were required in two patients. Two had inade-quate responses. The protocol was defined as follows: (1) inhibitors of more than 10 BU were reduced

by plasmapheresis with extracorporeal ad-sorption of antibodies;

(2) FVIII was given to neutralize remaining cir-culating inhibitor and raise the plasma FVIII level to 40-100 U/dL and maintain it on sub-sequent days at 30-80 U/dL;

(3) Cyclophosphamide was given just prior to the first dose of FVIII, at a dose of 12-15 mg/kg intravenously on each of two days, fol-lowed by 2-3 mg/kg orally for 8-10 days.

(4) Intravenous gamma globulin, 2.5 to 5 grams, was given immediately after the first dose of FVIII and then, starting day 4, at-doses of 0.4 grams/kg for 5 days.

After the intense course of treatment, FVIII was continued prophylactically to main-tain suppression. The total amount of FVIII used was equivalent to that used in a month or less on the Bonn protocol. Freiburghaus C, Berntorp E, Ekman M, Gunnarsson M, Kjellberg BM, Nilsson IM. Tolerance induction using the Malmo treatment model 1982-1995. Haemophilia 1999; 5:32-39. Results of induction of tolerance with the Malmö protocol are updated. Tolerance was achieved in 10 of 16 patients with hemo-philia A and inhibitors to FVIII and in 6 ot 7 patients with hemophilia B and inhibitors to factor IX. (One of the patients with hemophilia B relapsed at six months and another course of treatment did not re-induce tolerance.) Overall, 12 patients achieved tolerance after one course of the protocol and four more achieved tolerance after two or more courses. The chance of success was best in those with low inhibitor levels at outset, histories of low inhibitor levels and a long interval since pre-vious replacement therapy. Actual duration of treatment in successful courses ranged from 13 to 39 days. The average FVIII use over a mean of 20 days was 162,000 units. The aver-age factor IX use over a mean of 23 days was 219, 000 units. (Note that the rate of success in inducing tolerance in hemophilia B was good, in stark contrast to outpatient regimens of daily infusions.)

Nilsson IM, Berntorp E, Zettervall O. Induction of split toler-ance and clinical cure in high-responding hemophiliacs with factor IX antibodies. Proc Natl Acad Sci 1986; 83:9169-9173. At 6-19 days into tolerance induction in four patients with hemophilia B and inhibi-tors, a new non-neutralizing IgG4 antibody appeared, which complexed with infused fac-tor IX.

Nilsson IM, Berntorp E, Zettervall O, Dahlback B. Nonco-agulation inhibitory factor VIII antibodies after induction of tolerance to factor VIII in hemophilia A patients. Blood 1990; 75: 378-383.

A new non-neutralizing IgG4 antibody

38

was found in six patients who had undergone induction of tolerance. The antibody fused with infused FVIII but did not neutralize it or reduce its half-life.

Gilles JG, Desqueper B, Lenk H, Vermylen J, Saint-Remy JM. Neutralizing anti-idiotypic antibodies to factor VIII inhibitors after desensitization in patients with hemophilia A. J Clin In-

vest 1996; 97:1382-1388. Two patients with hemophilia A and in-hibitor were followed during induction of im-mune tolerance. Anti-idiotypic antibodies were elicited which neutralized the inhibiting capac-ity of anti-FVIII antibodies. The authors contend that anti-FVIII inhibitors have not disappeared after induction of tolerance, but are disarmed by anti-idiotypic antibodies.

Acquired factor VIII inhibitors Green D, Lechner K. A survey of 215 non-hemophilic patients with inhibitors to factor VIIIl. Thromb Haemost 1981; 45: 200-203. In this international survey, incidence was about equal in males and females. More than half of patients were over 50 years of age. Inhibitors were post-partum in 7.3%, associated with recognized auto-immune disease in 18%, and not associated with any underlying disor-der in 46.1%. Others had a variety of possibly-associated conditions. Major hemorrhages oc-curred in 87% of patients and 22% died.

Green D, Rademaker AW, Briet E. A prospective, randomized trial of prednisone and cyclophosphamide in the treatment of patients with factor VIII autoantibodies. Thromb Haemost 1993; 70: 753-757.

A prospective, randomized multicenter trial was conducted with 31 patients who had auto-antibodies to FVIII. All patients were treated with prednisone, 1 mg/kg for 3 weeks, and, if inhibitors still persisted, then were ran-domized either to continue prednisone, or to taper off it and begin cyclophosphamide 2 mg/kg, or to continue prednisone and add cyclo-phosphamide. Inhibitors disappeared in 10 pa-tients during initial prednisone therapy, in 3 of 4 randomized to continue on prednisone, in 3 of

6 continued on cyclophosphamide alone, and in 5 of 10 continued on both agents. Hauser I, Schneider B, Lechner K. Post-partum factor VIII inhibitors: A review of the literature with special reference of the value of steroid and immunosuppressive treatment. Thromb Haemost 1995; 73: 1-5. Post-partum inhibitors occur most common after the first pregnancy, within three months of delivery. They may disappear spontaneously, over months or years. Retro-spectively, the authors compared no treat-ment to treatment with corticosteroids and to use of other immunosuppressive drugs with or without corticosteroids. Use of immuno-suppressive drugs with or without corticos-teroids resulted in remission sooner than use of corticosteroids alone. No difference was seen in the rate of remission in patients treated with corticosteroids alone versus those not given medication. Shaffer LG, Phillips MD. Successful treatment of acquired hemophilia with oral immunosuppressive therapy. Ann In-tern Med 1997;127: 206-209. Nine consecutive patients, ages 50-79 years, with autoantibodies to factor VIII were given daily cyclophosphamide,100-200mg orally,and prednisone 50-60 mg, until inhibi-tors disappeared. Remission was achieved in all patients, with a median treatment time of 12 weeks, and a maximum time of 270 days. (Contains excellent graphs.)

Yee TT, Taher A, Pasi KJ, Lee CA. A survey of patients with acquired haemophilia in a haemophilia centre over a 28-year period. Clin Lab Haematol 2000; 22:275-278. In the major referral center in London, 24 patients with auto-antibodies to factor VIII were seen over a 28 year period. Two-thirds were female. The median age was 69 years. (This rate of acquired inhibitors, from a catch-ment area of millions of people, needing diagno-sis or therapy in the foremost referral center of the area, is not high.) Delgado J, Jimenez-Yuste V, Hernandez-Navarro F, Villar A. Acquired haemoophilia: Review and meta-analysis fo-

39

cused on therapy and prognostic factors. Br J Haematol 2003; 121:21-35. Review. Results of 20 surveys are summarized. The median age was 64 years, and 55% of pa-tients were female. Related conditions included pregnancy in 15%, malignancy in 18%, autoim-mune disorders in 11 % and unidentified or other in 58%. The over-all mortality rate was 20%; the inhibitor-related mortality rate was 11% (but only one of 32 patients with a post-partum inhibitor died.) The rate of complete re-missions was 74%. Wiestner A, Cho HJ, Asch AS, Michelis MA, Zeller JA, Peer-schke EIB, Weksler BB, Schechter GP. Rituximab in the treat-ment of acquired factor VIII inhibitors. Blood 2002; 100: 3426-3428. Auto-antibodies were suppressed suc-cessfully in three patients given rituximab to-gether with other immunosuppressive agents and also in patient with mild hemophilia who had both an auto-antibody to his own FVIII and an allo-antibody to exogenous FVIII. The allo-antibody persisted longer than the auto-antibody. The authors believe that addition of rituximab, a monoclonal antibody against CD 20 T cells, to their other immunosuppressive drugs either improved a sluggish response or resulted in a quicker response than would have been expected. Field JJ, Fenske TS, Blinder MA. Rituximab for the treatment of very high-titre acquired factor VIII inhibitors refractory to conventional chemotherapy. Haemophilia 2007; 13:46-50. Four non-hemophilic patients with peak inhibitor levels of 249 to 508 Bethesda units who had failed to respond to cyclophos-phamide, vincristine and prednisone were then treated with four weekly rituximab infusions. Each patient had an initial partial response but all but one relapsed thereafter. One patient sus-tained a partial response for 13 months.

Sultan Y, Kazatchkine MD, Maisonneuve P, Nydegger UE. Anti-diotypic suppression of autoantibodies to factor VIII (antihemophilic factor) by high-dose intravenous gammaglobu-lin. Lancet 1984; 2:765-768.

Two patients with auto-antibodies

treated with intravenous IgG had a rapid and prolonged but not total suppression of their inhibitors. Intravenous IgG was ineffective in

two hemophilia A patients with inhibitors. Schwartz RS, Gabriel DA, Aledort LM, Green D, Kessler CM. A prospective study of treatment of acquired (autoimmune) factor VIII inhibitors with high-dosage intrave-nous gammaglobulin. Blood 1995; 86: 797-804. A 25% or more reduction of the levels of autoimmune factor VIII inhibitors was achieved in 6 of 16 patients given high-dose intravenous IgG. Zeitler H, Ulrich-Merzenich G...Brackman H-H et alia. Treat-ment of acquired hemophilia by the Bonn-Malmo protocol: documentation of an in vivo immunomodulating concept. Blood 2005; 105:2287-2293. Thirty-five patients with high-titer ac-quired inhibitors were treated with immu-noadsorption for inhibitor elimination, FVIII substitution, intravenous immunoglobulin and immunosuppression with cyclophos-phamide and prednisolone. Bleeding was rap-idly controlled during one or two apheresis sessions and no subsequent bleeding epi-sodes occurred. Inhibitor levels decreased to undetectable levels within a median of three days, and factor substitution was stopped within a median of 12 days, and treatment was completed within a median of 14 days. On long-term follow-up, if cancer patients were excluded, the complete remission rate was 97%.

Collins P, Baudo F. Knoebl P et alia. Immunosuppression for acquired hemophilia A: results from the European Ac-quired Haemophilia Registry (EACH2). Blood 2012; 120:47-55. The European registry collected infor-mation on 501 patients from 117 centers be-tween 2003-2008. Patients’ median age was 73.9 years. Data on immunosuppression was available on 331 patients. Steroids plus cyclo-phosphamide resulted in a more stable com-plete remission (70%) than steroids alone (48%) or rituximab-based regimens (59%). The median time to complete remission was about five weeks for steroids with or without

40

cyclophosphamide but was about twice as long with rituximab. Immunoglobulin administration did not improve outcome. The likelihood of achieveing stable remission was not affected by underlying etiology but was influenced by the presenting inhibitor titer and FVIII level. Tay l, Duncan E...Lloyd J et alia. Twelve years of experience of acquired hemophilia A: trials and tribulations in South Aus-tralia. Semin Thromb Hemost 2009: 35:769-77. The incidence over 12 years in South Australia was 1.2 cases/million population/year, with a median age of 78 years and equal sex ratio. Of the total of 25 patients, 13 required he-mostatic agents. Of 18 patients treated with im-munosuppressive agents, 15 achieved remis-sion. (Note that not all patients required a hemo-static agent, the condition may be diagnosed and treated before a major hemorrhage occurs.) Knoebl P, Marco P, Baudo F et alia. Demographic and clinical data in acquired hemophilia A: results from the European Ac-quired Haemophilia Registry (EACH2). J Thromb Haemost 2012; 10:622-31. In Europe between 2003 and 2008, 501 patients with acquired hemophilia from 13 countries were entered in a registry. The me-dian age was 73.9 years. Among the non-pregnancy-related cases, 57% were male. The condition was idiopathic in 51.9%, related to malignancy in 11.8% and related to autoim-mune conditions in 11.6%. Of the 501 patients, 474 had presented with bleeding. Of the 477 patients who had immunosuppres-sive therapy, 72.6% achieved complete remis-sion. Tengborn L, Baudo F...Collins P et alia. Pregnancy-associated acquired haemophilia A: results from the European acquired haemophilia (EACH2) registry. BJOG 2012; 119:1529-37. In the European registry of acquired hemophilia cases, 42/501 were peri-partum. Ante-partum inhibitors were evident in eight women (and two babies had post-natal bleed-ing.) Mean time to diagnosis after delivery was 89 days (range 21-120). First-line immunosup-pressive therapy provided complete remission in 74% of cases. All women survived.

Collins P, Baudo F, Knoebl P et alia. Immunosuppression for acquired hemophilia A: results from the European Ac-quired Haemophilia Registry (EACH2). Blood 2012; 120:47-55. In the European registry, data on im-munosuppressive therapy in 331 patients could be analyzed. Steroids combined with cyclophosphamide resulted in more stable complete remission (70%) than steroids alone (48%) or rituximab-based regimens (59%). The median time to complete remission was about 5 weeks for steroids with or without cyclophosphamide; rituximab-based regi-mens took about twice as long. Immu-noglobulin administration did not improve outcome. Baudo F, Collins P, Huth-Kühne A et alia. Management of bleeding in acquired hemophilia A: results from the Euro-pean Acquired Hemophilia (EACH2) Registry. Blood 2012; 120:39-46. In the European registry, not all the patients had a bleeding episode, and not all such episodes required a hemostatic agent. Among 307 patients, 174 received rVIIa, 63 received activated prothrombin complex (APCC), 56 received factor VIII and 14 re-ceived DDAVP. Bleeding was controlled in 79.6% of patients with initial therapy. Bleed-ing control was similarly successful with rVIIa versus APCC. and more successful with those bypassing agents (93%) than with FVIII or DDAVP (68.3%). Thrombotic events oc-curred in 2.9% of patients on rVIIa and 4.8% on APCC. Tiede A, Klamroth R, Scharf RE et alia. Prognostic factors for remission and survival in acquired hemophilia A (AHA): results from the GTH-AH 01/2010 study. Blood 2015; 125:1091-7. In a prospective study in multiple cen-ters in Austria and Germany, data on 102 pa-tients with acquired hemophilia who were treated with an agreed standard protocol were analyzed. The primary end point was time to achieve a partial remission, i.e. FVIII activity restored to more than 50U/dl and no active bleeding. Prednisolone alone (75-150 mg/day according to body weight) was given for three weeks and if PR was not achieved,

41

then cyclophosphamide (100-200 mg/day ac-cording to weight) was added in weeks 4 to 6, and in week 7, prednisolone and rituzimab (4 weekly doses of 375 mg/sq meter) were given. As soon as PR was achieved, cyclophos-phamide or rituximab were stopped (if the pa-tient was on them) and prednisolone tapered. Remission was achieved with prednisolone alone in 47% of patients, of whom some two-fifths had a partial relapse during the steroid-tapering phase, and nearly all responded to in-creasing the steroid dose. Of the 44 patients

requiring second-line therapy, 35 received cyclophosphamide (and 21 achieved partial remission) and 9 received rituximab. Overall, partial remission was achieved less frequently and over a longer time in patients whose baseline FVIII level had been below 1 U/dL. There was a trend towards a lower partial remission rate with higher baseline inhibitors but it was not sta-tistically significant. Age, gender and underly-ing condition did not affect the probability of success.