Is the incidence and prevalence of inhibitors greater with recombinant products? No


Jeanne M. Lusher, Division of Hematology/Oncology, Children's Hospital of Michigan, 3901 Beaubien Boulevard, Detroit, MI 48201, USA.
Tel.: +1 313 745 5515; e-mail:

Over the past 20 years, as each new factor (F)VIII concentrate or new viral inactivation process has been developed, there has been concern that it might be more immunogenic, resulting in a higher percentage of FVIII inhibitors. Fortunately, this has not proven to be the case. Many of us first became aware of this concern when some physicians treating hemophiliacs expressed worry about dry heat-treated concentrates, introduced in the USA in 1984. Then, when a short-term prelicensure hepatitis, and HIV safety study with a monoclonal antibody-purified, plasma-derived FVIII concentrate (Monoclate®) showed a cumulative incidence of inhibitor formation of 18% in previously untreated patients (PUPs) [1], more worry ensued. No doubt resulting from such concerns, the multicenter, multinational prelicensure clinical trials with the first two recombinant (r)FVIII products (Baxer's Recombinate and Bayer's Kogenate) incorporated more frequent inhibitor assays and much longer periods of observation (5 years or 100 exposure days). These studies with the first two full-length rFVIII products revealed that 28–30.5% of severely affected PUPs developed FVIII inhibitors during the course of the study. However, over half of the inhibitors were low titer and many were transient [2,3]. Thus, it was apparent that many of these inhibitors could have been missed if they had not been looked for at 3-month intervals during the entire course of the trials. These PUP studies (which began in 1990 and 1989, respectively) then served as a model for the design of subsequent trials to assess safety and inhibitor development with new FVIII products in PUPs [4].

As the data from these and other multicenter clinical studies in PUPs with hemophilia A were analyzed, certain important facts became apparent: most inhibitors develop in persons with severe hemophilia A (baseline FVIII < 0.01 U mL−1); most develop after relatively few exposure days (ED) to FVIII (median 9–11 ED); most (but not all) develop in early childhood; inhibitors occur more frequently in persons of African descent; they are also more likely to occur in hemophilic individuals whose hemophilic brother(s) has an inhibitor. It was also apparent that some inhibitor patients are ‘high responders’, having a brisk anamnestic response to FVIII, while others are ‘low responders’, with inhibitor concentrations never exceeding a few Bethesda Units (BU). Furthermore, some inhibitors (usually low-titer inhibitors) disappeared over time, despite continued episodic treatment with FVIII.

We now know that the single most important determinant of inhibitor development is the individual's FVIII gene defect. Large gene deletions, stop codons and nonsense mutations are associated with a higher incidence of inhibitor formation. Such mutations result in failure to synthesize FVIII [5]. Thus, the individual's immune system recognizes infused FVIII as a foreign protein. Other ‘patient factors’ include race. In each of several studies (and with various types of FVIII products) it has been shown that persons of African descent are more likely to develop an inhibitor than are Caucasians or persons of other racial groups [4]. While the reason for this remains unclear, it may reflect differences in HLA type, or the influence of modifying genes.

Additionally, environmental factors, such as infections or immunizations at or around the time of FVIII infusion, may play a role in inhibitor development by activation of the immune system. Some have postulated that an inflammatory response to a hemorrhage, or frequent large doses of FVIII given for a severe bleeding episode or major surgery, may predispose an individual to develop an inhibitor.

It is perhaps surprising that the incidence of inhibitor development is so similar in prospective, multicenter PUP studies with inhibitor assays being done at specified intervals (usually every 3 months or less), regardless of the product being tested. In most of these studies, 28–30.5% of severely affected PUPs have developed inhibitors; however, the prevalence of inhibitors at the end of these trials has been considerably less, being 9–19%[2–4,6]. This prevalence range is quite comparable to prevalence figures of 10–15% cited in publications from the 1960s and 1970 s [4,7,8] in patients with severe hemophilia A being treated with cryoprecipitates, crude or intermediate purity plasma-derived FVIII concentrates.

While several other studies in PUPs receiving predominantly intermediate purity plasma-derived FVIII products have been published in the 1990s, two are particularly noteworthy. One is the relatively small but well-designed prospective study by Ehrenforth et al. [9], in which 52% of patients with severe hemophilia A developed inhibitors over the course of the study.

The second is the publication from Sultan and The French Hemophilia Study Group, in which they reported inhibitor prevalence data (obtained by questionnaire) for 3435 hemophilia patients in France. Of 1565 with severe hemophilia A, 198 (12.8%) had an inhibitor. The authors noted that this survey was done before the introduction of ‘ultra-purified’ plasma-derived or recombinant products in France. It is also interesting to note that 31% of the inhibitors were reported to have been detected after 20 years of age [10].

In the multicenter, prospective previously treated patient (PTP) studies with rFVIII products, it is noteworthy that 0–1 new inhibitors were reported in each trial. In the Kogenate PTP study, among 103 PTPs (who had a median of 4.6 years of follow up), one patient developed an inhibitor [11]. In the Recombinate study, none of 69 PTPs developed a de novo inhibitor; however, one subject with a prior history of an inhibitor developed a transient, low-titer inhibitor with decreased recovery but normal half-life [12]. In the ReFacto PTP study, 113 patients were followed for a median duration of 1711 days, and had a median of 385 ED. A FVIII inhibitor was noted in one individual after 113 ED to B-domain deleted (BDD)rFVIII and 35 months on study. It was subsequently noted that he had a low-titer inhibitor (0.6 BU) before entering the trial, and also had monoclonal gammopathy of unknown significance (MGUS) [6]. The two PTPs (one in the Recombinate study and one in the ReFacto study) who were later noted to have had low-titer inhibitors prior to study entry point out the difficulties in obtaining an accurate history in this regard, even when the individual is carefully screened for a clinical trial.

Among 152 PTPs evaluated in several prospective studies with Bayer's Kogenate FS [13], no patient developed a new inhibitor. Of 108 PTPs in Baxter's pivotal study with ADVATE, only one subject developed a low-titer inhibitor [14]. These data compare favorably with other publications which describe a range of inhibitor formation of 1–3% in PTPs treated with cryoprecipitates and/or crude and intermediate plasma-derived products over varying periods of time [15,16].

Only rarely has a particular FVIII product resulted in an increased incidence of inhibitors. This occurred with each of two intermediate purity plasma-derived concentrates (used in PTPs in Belgium, the Netherlands and Germany in the early 1990s), which had undergone double virus inactivation with solvent-detergent treatment plus pasteurization [17–19]. The combination of these two viral inactivation processes was later shown to modify the FVIII molecule, resulting in heightened antigenicity of the C2 domain [19].

With the development of so-called second and third generation rFVIII products such as BDDrFVIII (Wyeth's ReFacto), products formulated without albumin (Wyeth's ReFacto®AF), or albumin-free (Bayer's Kogenate FS and Baxter's ADVATE) there was again some concern that these might be more immunogenic. To date, this does not appear to be the case, with incidence and prevalence figures from prospective multicenter clinical trials being no higher than with full-length rFVIII products or plasma-derived products in PTPs or PUPs.

In view of (i) the substantial influence of patient factors (severity of hemophilia, gene defect, race, etc.) in determining which PUPs will develop inhibitors, (ii) the observation that 28–32% of severely affected PUPs will develop an inhibitor after a median of 9–11 ED if inhibitors are being looked for at frequent intervals, and (iii) the much lower incidence of new inhibitors in heavily treated PTPs, it would seem that heavily treated PTPs would be the best group in which to look for increased immunogenicity of a new product. This is now the published recommendation of the International Society on Thrombosis and Haemostasis' FVIII and FIX Scientific Subcommittee [20], and governmental regulatory bodies such as the US Food and Drug Administration's Center for Biologics Evaluation and Research, and the European Agency for the Evaluation of Medicinal Products, have taken note of this [21].

While variables can be kept to a minimum in well-designed, prospective clinical trials, in the postmarketing setting many variables contribute to the complexity of interpretation of reports. Many assumptions are made, often leading to inaccurate estimates of new inhibitor formation with various products. It will take a concentrated effort by many to make such postmarketing reports truly meaningful.

There is no doubt that the development of an inhibitor (particularly a high-titer inhibitor) is viewed as the most serious complication of hemophilia treatment today. Fortunately, the body of information from carefully designed, long-term prospective trials has shown that the risk of inhibitor formation is no greater with any of the recombinant FVIII products than with plasma-derived FVIII products.