ABSTRACT: Annually, ∼0.6% of newborns in the United States and European Union are congenitally infected with human cytomegalovirus (CMV); ∼20% of these infected newborns are symptomatic at birth or have developmental sequelae. The lack of an effective therapy for maternal infection hinders routine serologic screening of pregnant women. A previous study found that CMV-specific hyperimmune globulin administered to pregnant women with primary infection decreased the rate of mother-to-fetus transmission and risk for congenital disease. This randomized, controlled clinical trial was designed to verify the efficacy of CMV-specific hyperimmune globulin in preventing fetal infection.
The Congenital HCMV Infection Prevention (CHIP) trial was conducted at 11 centers in Italy. Pregnant women with a primary infection at 5 to 26 weeks of gestation, with onset in the prior 6 weeks, received hyperimmune globulin or placebo. The hyperimmune globulin preparation contained 50 U of anti-CMV immunoglobulin G (IgG) antibody per milliliter (≥95% IgG); the intravenous dose was 100 U (2.0 mL) per kilogram. The placebo was a 0.9% saline solution, given intravenously at the same dose. Parturients received globulin or placebo every 4 weeks until 36 weeks’ gestation, detection of CMV in the amniotic fluid, or spontaneous termination of pregnancy. Levels of virus-specific IgG and IgM antibodies and IgG avidity were determined. Diagnosis of primary CMV infection was based on seroconversion or the concomitant presence of CMV-specific IgM antibodies and low IgG avidity. The primary end point was the number of infected fetuses/newborns in each group. Secondary outcomes included CMV DNA level, levels of specific antibodies and cell subpopulations, clinical outcomes, and drug safety. The intention-to-treat principle was used for efficacy end points, and the on-treatment principle was used for safety end points.
Of 338 parturients with primary CMV infection, 124 enrolled and 123 completed the study. The 2 groups did not differ significantly in baseline characteristics. In both groups, the diagnosis of primary infection based on testing for seroconversion was made later during pregnancy than was diagnosis based on CMV-specific IgM antibody level and IgG avidity (14 weeks [range, 7–26] vs 8 weeks [range, 5–18] in the study group and 17 weeks [range, 6–26] vs 8 weeks [range, 6–17] in the placebo group). Seventy-two women with primary infection (59%) had symptoms or abnormal laboratory findings indicating primary CMV infection. Intrauterine transmission occurred in 45 cases (37%). Congenital infection was diagnosed before birth in 18 fetuses and at birth in 27 newborns. Three (8%) of 40 newborns negative for CMV on amniocentesis were infected at birth. Eighteen congenital infections (30%) occurred in fetuses/newborns of 61 women in the study group and 27 congenital infections (44%) occurred in fetuses/newborns of 62 women in the placebo group, a difference that was not statistically significant (P = 0.13). The viral DNA load in the amniotic fluid of infected fetuses or levels of viral DNA in urine or blood or virus-specific IgM in infected newborns did not differ between the 2 groups, nor were differences seen in CMV-specific IgG and IgM, IgG avidity, or levels of neutralizing antibodies. The groups did not differ in gestational week at infection onset, number of weeks between diagnosis and treatment, or any of the other immunologic or virologic variables investigated. Twenty adverse events were reported in 16 women, including 11 serious adverse events in 10 women. Serious adverse events occurred in 7 (13%) of 53 women in the study group and in 1 (2%) of 51 women in the placebo group (P = 0.06); in the study group, these included preterm delivery, intrauterine growth restriction, induction of preterm delivery, intrahepatic cholestasis of pregnancy, and postpartum eclampsia. Those in the placebo group were spontaneous abortion after amniocentesis, arthralgia of the upper and lower limbs, and pregnancy-induced hypertension. The newborns did not differ significantly in sex, weight, gestational age, or Apgar scores. Seven (15%) of 48 babies in the study group and 1 (2%) of 47 babies in the placebo group were premature (P = 0.06). Six premature babies in the study group were not infected with CMV. Three (30%) of 10 CMV-infected newborns in the study group and 4 (24%) of 17 CMV-infected newborns in the placebo group had symptoms at birth.
This controlled trial of virus-specific hyperimmune globulin to prevent congenital CMV infection found no significant between-group differences in either the primary outcome or the secondary clinical and biologic outcomes examined.