Foscarnet Versus Ganciclovir for Severe Congenital Cytomegalovirus Infection: Short- and Long-Term Follow-Up

Cytomegalovirus (CMV) infection is the most common and serious congenital infection, with universal screening in pregnancy, standardized therapy, and a vaccine still lacking.

In the 1990s, we noted that intravenous ganciclovir did not cure some children with severe sequelae due to congenital cytomegalovirus (CMV) infection. Therefore, we performed an open randomized trial using intravenous foscarnet as an alternative to intravenous ganciclovir in 24 infants (12 in each therapy group), all with severe neurological manifestations due to congenital CMV infection. Nine and five infants, belonging to the foscarnet or ganciclovir group, respectively, had abnormal hearing. One infant in each group also had chorioretinitis. Concomitantly, 12 CMV-infected infants with similar manifestations, who did not receive any therapy, were used as controls. The results of short-term (2 years) and long-term (7-29 years, mean 22.2) follow-up are reported herein. Short-term results: Neurological outcomes were normal in five of the twelve children who were treated with foscarnet, compared to nine of the twelve children given ganciclovir. None of the untreated children were healthy. There was a statistically significant difference (p = 0.023) between the treated and untreated children. Hearing was normal in four of the twelve children treated with foscarnet, seven of the twelve children treated with ganciclovir, and two untreated children. Long-term-results: Two children in both therapy groups died before the age of 17 years, and six untreated children died between 7 and 26 years of age. Neurological outcomes were normal in three of the ten children treated with foscarnet, in two of the ten treated with ganciclovir, and in none of the untreated children. Hearing was normal in two children treated with foscarnet, in six children treated with ganciclovir, and in one untreated child.

Intravenous ganciclovir and foscarnet were found to be safe at long-term follow-up and appeared to be capable of mitigating the neurological and auditory consequences of congenital CMV disease at the short-term follow-up. However, there was progressive worsening of the symptomatology in all three groups, with a statistically significant increase in the number of deaths (p = 0.035) among 4 of the 24 children in the therapy groups and 6 of the 12 untreated children.