Intravenous immune globulin and risk factors for adverse events in the pediatric population: Considering ABO and secretor status

In 1890, von Behring and Kitasato immunized rabbits with Clostridium tetani, collected their blood, and injected it into the abdominal cavity of mice, which protected them against lethal doses of C. tetani. This landmark finding set the stage for the eventual development of intravenous immune globulin (IVIG). Over half a century later, Cohn et al. developed a process by which to fractionate human plasma into classes of proteins, including the gamma globulin fraction, commonly known as Cohn fraction II. Following the discovery that Cohn fraction II containing immune serum globulin (ISG) could prevent or reduce the severity of measles in exposed children, optimization of the manufacturing process culminated in the ability to deliver gamma globulin intravenously. The revelation that IVIG could effectively improve the platelet count in patients with immune thrombocytopenia in the 1980s led to its investigation as a therapy for a myriad of autoimmune conditions.

Today, most IVIG products are produced via Cohn cold ethanol fractionation and undergo variable purification steps to reduce certain proteins such as albumin and IgA. Despite these processes, differences may be seen amongst manufacturers and amongst lots of IVIG from the same manufacturer. For example, the distribution of IgG subclasses may vary among preparations, and the variable concentrations of substances (eg, IgE, IgM, solvents, detergents, buffers) may influence tolerability among patients and across different products.

As such, contemporary IVIG products are still associated with potential side effects, some of which may be serious. Thus, we were intrigued to read the recent study in which Côté and colleagues retrospectively evaluated adverse events following IVIG administration in children.¹ In particular, we noted that the authors described hemolysis as a potential adverse effect, but did not include recipients’ ABO blood type as one of the variables in their study. This variable is crucial to elucidate a possible hemolytic association, as prior studies have shown that there is an increased risk of IVIG-mediated hemolysis in patients who are non-O blood group, and particularly group AB.^2–4 This occurs because IVIG contains anti-A and anti-B antibodies, which can cause hemolysis. IVIG is produced from the fractionation of human plasma, and is sourced from large numbers of human donors (either whole blood donors or plasma donors) to create pooled a pooled product. While the pool size is generally proprietary information from the manufacturers, the minimum pool is 1000 donors and may include tens of thousands of donors.² Given that ABO blood type varies amongst donors, so too will the level of ABO isohemagglutinins.

The incidence of hemolysis following IVIG administration varies widely among studies,⁵ which have focused primarily on the adult population. For example, a recent prospective study that included patients from 1 pediatric and 5 adult centers identified that 32% of IVIG infusions to non-blood group O patients resulted in hemolysis.³ The incidence of IVIG-induced hemolysis in the cohort examined by Côté et al. was 1% (5/478) of all infusions, notably lower than what is described in the literature; however, these hemolytic cases comprised 33% (5/15) of all serious adverse events. It is possible that the true incidence of hemolysis observed by Côté et al. was higher than what was observed. Sensitivity in the current study may be limited by its retrospective nature, as is noted by the authors, and/or an overrepresentation of patients with O blood group. Similarly, enrichment for patients who are secretors (ie, those who express the FUT2 gene which codes for a fucosyltransferase that elaborates secreted A and B substance) could, at least hypothetically, explain the lower prevalence of hemolysis; though it is important to note that while this neutralization is observable in vitro, it has been suggested that in vivo hemolysis may still occur as the anti-A and anti-B IgG may not be as readily neutralized as their IgM counterparts.⁶ Moreover, understanding the immunohematologic composition of the entire cohort could help inform why the number of infusions associated with hemolysis was lower in this pediatric cohort.

Additionally, it is unclear if, and to what extent, the risk of IVIG-induced hemolysis differs between the pediatric population—or those with lower total blood volumes—and the adult population.^7–9 IVIG is dosed on a weight-adjusted basis, and blood volume generally decreases in proportion to body weight with age. However, as with any discussion of pediatric physiology, additional complexity exists regarding if, and how, other relevant variables vary with age (e.g., intravascular/extravascular IVIG distribution kinetics, IVIG clearance mechanisms, ABO antigen density, rheological properties of circulating red blood cells, immune effector function, etc.), making risk estimation challenging. Furthermore, as noted by Côté et al., previously reported incidences of IVIG adverse events in pediatric cohorts have been variable in magnitude. As such, additional in-depth study of IVIG-induced hemolysis in this population is warranted.

As IVIG inherently includes isohemagglutinins, limits on the acceptable level of anti-A and anti-B titers have been implemented; further some manufacturers may use different purification methods to reduce titers in the IVIG product.⁶ However, there are a variety of challenges associated with defining a regulatory upper limit of isohemagglutinin titers, including intra- and interlaboratory variation in testing and the need for a standardized reference agent.⁶ As both the amount of ABO antigens present on red blood cells and the presence of soluble ABO antigens that may neutralize isohemagglutinins will differ amongst recipients, evaluating the ABO type of pediatric individuals in IVIG studies could be informative. These studies could also impart additional understanding as to whether additional immunologic effects (either beneficial or adverse) are associated with ABO type.

Given these considerations, future studies, especially in the pediatric population, should consider reporting ABO blood type, particularly if they are prospective in nature and/or examine specific IVIG products not previously investigated. Additionally, while secretor status has not shown to be a predictor for IVIG-mediated hemolysis in adults,⁴ the effect of secretor status in patients who are low blood volume (LBV) (eg, pediatrics) is less certain. As such, future analyses are necessary to delineate the risk stratification of IVIG based on both ABO and FUT2 genetic testing of the LBV recipient. Likewise, prospective studies will ideally include direct antiglobulin test results, and where positive, follow with an elution study, which will include testing against A1 and B cells to confirm the presence of passively acquired isohemagglutinins.