Abstract
Immune globulin intravenous (IGIV) has grown steadily in the volume of use and numbers of clinical indications since the first modern commercial preparation was approved by FDA in 1981. All US-marketed IGIV products are approved by FDA to treat primary immune deficiency; seven are also approved for immune thrombocytopenia purpura. Currently, most IGIV infusions are for treating autoimmune diseases.
IGIV has historically been isolated from pools of human plasma (at least 1,000 donors, but typically ranging from 10,000 donations in the US up to 60,000 donations in some EU nations) using Cohn-Oncley cold ethanol fractionation. Most firms now supplement this process with chromatographic steps to remove Factor XIa, other promoters of complement activation, and other Ig isotypes.
Despite high purity, a range of adverse symptoms, varying widely by product, is associated with IGIV infusions. To date, a possible correlation between adverse event profiles and contaminant profiles for IGIV products has not been studied and remains poorly understood. FDA requires donor blood tests for absence of hepatitis, HIV and other viral diseases. Final container immunoglobulin products must show potency against measles, diphtheria and polio virus and contain no prekallikrein activator. However, no further characterization of contaminant proteins is currently required.
We therefore undertook comparative component analyses of 5% and 10% solution IGIV products approved or in development for the US market. Component analysis was performed using high resolution mass spectrometry (HRMS) and very sensitive multiplexed ELISA (mELISA) surveys to identify and quantitate trace contaminants in individual products. Factor XIa antigen and activity tests as well as functional assays were performed to assess the thrombogenic potential of the products. Subgroup and Ig isotype distribution were assessed by ELISA, and size exclusion chromatography was used to determine the extent of molecular aggregation.
HRMS and mELISA are complementary but fundamentally different survey methods. HRMS is less "biased" in that it will detect any molecular entity more abundant than the limit of detection, whereas mELISA detected primarily biomarkers included in lists of assays preset by the vendor. Among these were several tumor markers, a wide range of inflammatory modulators, and a number of metabolic hormones.
For a complex biologic such as polyclonal IGIV, HRMS typically detects entities present at levels of 10 - 20 molar ppm, while the lower limit of quantitation (LLOQ) of the mELISA is typically 3 - 5 orders of magnitude lower. Therefore, the mELISA survey of each IGIV product typically detects a substantially greater number of contaminant entities than the HRMS assay. If mELISA found a protein at ≥ 10 ppm for a product, it was generally also among the contaminants detected by HRMS. Thus, HRMS and mELISA results correlated satisfactorily. With both methods, detected levels of any given contaminant varied up to several orders of magnitude among the IGIV products tested.
The most abundant contaminants in most preparations were IgA, IgD, β 2-glycoprotein-1, α 2-macroglobulin and albumin. Residual albumin and IgM were found in some but not all products. Inflammatory modulators were largely absent from the IGIV products surveyed. The interleukins IL6, IL-8 and IL-10 were detected in a few products, but in no case more than 4-fold above the mELISA LLOQ. The tumor markers carcinoembryonic antigen (CEA), α-fetoprotein (αFP), CA-125, CA 19-9 and tumor necrosis factor-α were not found in any product.
The results of these assays are discussed with respect to their potential relationship to the observed adverse event and safety profiles of the individual products.
Abramson: GC Mogam, Incorporated: Consultancy, Research Funding. Van Cott: GC Mogam, Incorporated: Research Funding.
