Circulating cytomegalovirus (CMV) neutralizing activity in bone marrow transplant recipients: comparison of passive immunity in a randomized study of four intravenous IgG products administered to CMV-seronegative patients

Abstract

Abstract Forty-two cytomegalovirus (CMV)-seronegative bone marrow transplant (BMT) recipients were randomized in a double-blind fashion to receive one of four commercially available intravenous Ig (IVIgG) products (Gamimmune N, Immune Globulin Intravenous, Gammagard, or Sandoglobulin) at a dose of 500 mg/kg every other week. The four treatment groups were similar in distribution of patient ages, weights, autologous versus allogeneic donor type, and underlying diseases. Every other week administration of IVIgG provided total serum IgG levels within the physiologic range for age. CMV titers by latex agglutination were stable (average geometric mean titer of 18.4 after the second IVIgG dose), with no statistically significant differences among the four product groups. CMV neutralizing activity (CMVNA) and CMV enzyme-linked immunosorbent assay (ELISA) titers were determined on a subset of sera from 27 study patients representing the four product groups. Patient serum samples obtained before IVIgG infusions and 2 weeks after the second IVIgG dose (ie, 3 weeks post-BMT) were assayed for CMVNA and CMV ELISA titers. Geometric mean titers of CMVNA and CMV ELISA varied among the product groups. The highest mean 50% CMVNA was 1:43 for product B, whereas the lowest mean 50% CMVNA was 1:14 for product A; two of the IVIgG product groups showed intermediate 50% mean titers of 1:27 (product C) and 1:26 (product D) for an overall P = .02. CMV ELISA titers (expressed as Paul Ehrlich International units [PEI U]) also showed the highest mean of 2.95 PEI U/mL for product B and the lowest mean of 1.34 PEI U/mL for product A. Intermediate mean values of 2.27 PEI U/mL and 2.03 PEI U/mL were obtained with products C and D, respectively (overall P = .003). The CMV ELISA titers show a minimal correlation (r = .566) to the observed CMVNA titers. We conclude that commercially available IVIgG products provide passive CMVNA, and that the level of circulating CMVNA is affected by the IVIgG product used.