Reports

This report is based on medical evidence presented at sanctioned medical congress, from peer reviewed literature or opinion provided by a qualified healthcare practitioner. The consumption of the information contained within this report is intended for qualified Canadian healthcare practitioners only.

PRIORITY PRESS - 51st Annual Meeting of the American Society of Hematology

New Orleans, Louisiana / December 5-8, 2009

Formation of factor VIII (FVIII) antibodies is regarded as the most serious complication of treatment in patients with hemophilia A. New tests appear to be more sensitive and specific, particularly for their ability to identify non-neutralizing antibodies (NNAs) undetectable with the Bethesda assay, the standard method for detecting antibodies, and have the potential to better characterize how these antibodies adversely affect recombinant FVIII (rFVIII) therapies.

Results from Bethesda, ELISA Assays

Canadian studies with the ELISA test suggest NNA formation differs for full-length rFVIII and B domain-deleted rFVIII (BDDrFVIII) agents. “The presence of anti-FVIII antibodies has been evaluated by the functional Bethesda clotting assay for the past 30 years, [but] there is a growing appreciation that the anti-FVIII immune response may also include the appearance of NNAs in some patients,” stated Dr. David Lillicrap, Richardson Laboratory, Queen’s University, Kingston, Ontario, during the ASH meeting.

As primary author of a study with an ELISA-based assay supported by the Association of Hemophilia Clinic Directors of Canada, Dr. Lillicrap reported progress in providing a more complete characterization of the anti-FVIII antibody response. Of plasma collected from 392 patients with hemophilia A initially tested with the Bethesda assay, 24 (6%) were positive for the antibody. The 368 negative samples were then tested with a FVIII ELISA-binding assay. All samples were tested in duplicate and positive samples underwent serial dilution testing. About 25% of the samples were independently tested using the same ELISA assay in a second laboratory.

On ELISA, 48 (13%) of the Bethesda-negative samples were positive for NNAs, but there was a higher proportion of antibodies positive to the rFVIII formulations using full-length proteins than to the one based on a BDD protein.

Based on this study, the explanation for this difference is that “some of the antibodies bind only to the FVIII B domain,” Dr. Lillicrap reported. Specifically, the ELISA-based assay found anti-FVIII NNAs to rFVIII plasma/albumin-free method (rAHF-PFm) in 11.5% of the patients, to the sucrose formulation (FS) rFVIII in 8.5% of the patients, but only 4.6% of the patients to the BDDrFVIII moroctocog alfa. Only 2.8% of the patients had antibodies to all three rFVIII formulations.

Overall, “this study of a large Canadian hemophilia A population has demonstrated the presence of anti-FVIII NNAs in 11% of the population,” Dr. Lillicrap told delegates. Although there was a good correlation for the presence of inhibitory antibodies with the Bethesda and ELISA assays, the study demonstrates that only the ELISA assay detects both high- and low-titre neutralizing antibodies (NAs) as well as high- and low-titre NNAs. The clinical significance of the NNAs, which may include an adverse effect on the pharmacokinetics of rFVIII concentrates, “must await further study with clotting factor recovery and half-life analysis,” but this study confirms that ELISA provides better discrimination of antibodies to FVIII than the Bethesda assay. The ELISA test can be expected to be useful for better characterizing the antibody response if both NAs and NNAs are found to be relevant to rFVIII therapy.

Supportive Data

Highly compatible results were generated in a study by many of the same authors that focused on patients with hemophilia A in the province of Quebec. Plasma specimens from 127 patients were evaluated with the Bethesda and ELISA assays using the two full-length rFVIII concentrates, rAHF-PFm and rFVIII FS, and BDDrFVIII as the coating antigens. Plasma from six individuals without hemophilia was used as controls. Unlike the Canada-wide study which accepted a positive Bethesda assay by a value of 0.6 Bethesda units (BU)/mL, a positive test in the Quebec study was predefined as <u>></u>0.4 BU/mL. An ELISA test was considered positive if absorbance was at least three standard deviations (SDs) above the mean of the normal plasma.

Of the 127 patients, 12 (9.4%) were positive for antibodies on the Bethesda assay; all were positive on ELISA. Of the 115 who were negative for antibodies on the Bethesda assay, 12 (10.4%) were positive on ELISA. However, the positive antibodies were limited entirely to the full-length rFVIII concentrates. None of the negative patients had antibodies to BDDrFVIII. Although lead author Dr. Anne-Marie Vincent, Department of Hematology, CHUM-Hôpital Ste-Justine, Montreal, Quebec, noted that “the clinical significance of this observation is presently unknown,” pharmacokinetic studies to evaluate an influence on clinical effect are underway.

“The development of anti-FVIII antibodies is the most serious complication of the treatment of hemophilia A,” Dr. Vincent confirmed. Based on studies so far, she suggested, “The ELISA assay may be more sensitive,” but she also cautioned that ELISA might be “detecting antibodies against both functional and non-functional epitopes of FVIII. We are now looking into this.”

Overcoming Limitations

One of the limitations of the Bethesda assay addressed by ELISA is that non-inhibitory antibodies “could be responsible for low in vivo recovery of FVIII or a pharmacokinetic pattern suggestive of rapid clearance of FVIII,” Dr. Vincent explained. In this situation, ELISA may be more useful for assessing relative response to available treatment, particularly in identifying those who might be better candidates for BDDrFVIII, which was shown in both studies to be less susceptible to NNAs.

If ELISA assays are found more clinically useful than the Bethesda assay, one potential limitation is the shipping of plasma samples when hospitals are not equipped with ELISA. Although these specimens can be readily transported on dry ice, this was characterized by Dr. Vincent as “both costly and cumbersome.” To test an alternative, she and other investigators, including Dr. Lillicrap, undertook a study of the stability of serum specimens adsorbed and dried onto filter paper at room temperature. In the study, 213 plasma samples from 97 patients with congenital hemophilia and five patients with acquired hemophilia were tested with ELISA after two methods of transport (plasma from six normal individuals was used as controls). The first was the commonly used methodology involving freezing at -80C; the second involved drying samples on filter paper for one month at room temperature. The coating antigens were from a full-length rFVIII FS preparation and BDDrFVIII.

“The correlation between the numbers of SDs obtained with frozen specimens and specimens dried on filter paper was excellent,” Dr. Vincent reported. She indicated that if ELISA emerges as a new standard for testing of FVIII antibodies in patients with hemophilia A, this method of transporting samples to the laboratory may increase convenience and reduce costs.

Investigators from France used two assays to identify NNAs undetectable with the Bethesda assay. One assay was designed to test NNA reactivity to the heavy chain and the light chain of FVIII. The second assay was designed to identify NNA reactivity to the B domain of FVIII. In samples from 187 hemophilia A patients (133 had severe disease and 31 had moderate disease), about two thirds had been treated with rFVIII and one third with plasma-derived FVIII. All were negative for antibodies on the Bethesda assay as defined by <0.6 BU/mL.

“The prevalence of NNAs in this series was 24%,” reported Dr. Aurélien Lebreton, Laboratoire Sysdiag, Montpellier, France. Although the majority of the NNAs were found to be directed at the heavy chain of the FVIII, 9% of the NNAs were directed at the B domain. Like the Canadian investigators, Dr. Lebreton speculated there is a potential for these NNAs to influence the pharmacokinetics of rFVIII agents but prospective studies are needed.

“NNAs in hemophiliacs are still being debated, but these antibodies could form immune complexes with FVIII and be responsible for an increased FVIII clearance, inducing a shortened half-life,” Dr. Lebreton suggested. “Prospective studies could highlight a difference according to the type of treatment.”

Summary

Antibodies directed against FVIII are one of the most significant treatment-related complications in the management of hemophilia A. While inhibitory antibodies neutralize the pro-coagulant activity of FVIII, defeating the efficacy of rFVIII agents, NNAs, which are not detected by the Bethesda assay, may also have clinical significance. Several research groups are now pursuing new, more sensitive approaches to the detection of antibodies, which may provide greater insight into the relative vulnerabilities of current therapies.