Specimens were frozen at ? 70C in 0.5 mL aliquots. no serologic correlate of protection has been defined for mumps. Measles Plaque Reduction Neutralization Assay Seroprotection against measles was assessed using a plaque reduction neutralization (PRN) assay, as this method has been clinically validated [15]. In brief, Vero cell monolayers were infected with a low-passage Edmonston measles virus strain and incubated with serially diluted serum specimens in duplicate. The 50% endpoint titers were interpolated using the Karber method [16]. Measles seroprotection was defined as a PRN titer of 120 milliCinternational units (mIU) of neutralizing antibody per milliliter of serum relative to Sodium orthovanadate World Health Organization II reference serum 66/202 (supplied by National Institute for Biological Standards and Control, South Mimms, United Kingdom) [16]. Statistical Analysis The prevalences of measles seroprotection, rubella seroprotection, and mumps seropositivity were compared between PHIV and HEU participants, with 95% exact binomial confidence intervals (CIs) and Fisher exact test. Demographic and clinical characteristics assessed at the times of first and last MMR dose and date of serologic specimen were compared between PHIV and HEU participants using Wilcoxon rank-sum and Fisher exact tests as appropriate. Among PHIV participants, HIV severity measures and ART use were also compared by MMR antibody status. The Acta2 prevalences of measles seroprotection, rubella seroprotection, and mumps seropositivity were compared by the number of vaccine doses received while on sustained cART with Fisher exact test. Further analyses assessed whether 1 or more (compared to zero) vaccine doses prior to participant exposure to sustained cART modified the relationship between seroimmunity and the number of vaccine doses received while on sustained cART. To identify key sets of covariates that would be most predictive of seroimmunity and to discriminate between children who have seroimmunity and those who may need to be reimmunized, multivariable models for immunity to each virus were built by initially including the number of doses while on sustained cART and subsequently adding covariates 1 at a time by descending order of their univariable c-statistic, which is analogous to the area under the curve in a receiver-operator curve (ie, a measure of discrimination). To obtain an efficient set of independent clinical predictors, covariates were retained if they were significant at = .05 and did not nullify the significance of any already included predictors. All analyses were conducted using SAS software version 9.2 (SAS Institute, Cary, North Carolina). RESULTS As of 10 October 2011, 428 PHIV and 221 HEU participants had serum specimens available for Sodium orthovanadate serologic testing. A serologic specimen from 1 PHIV participant did not contain a sufficient quantity of serum to be tested in the measles PRN assay; this participant’s data was included only in analyses Sodium orthovanadate of mumps seropositivity and rubella seroprotection. The 428 PHIV children, compared with the 221 HEU children, were more likely born before 1996 (when cART became available) and had a different racial/ethnic composition (Table ?(Table1).1). Sodium orthovanadate At the time the serologic specimen was obtained, the PHIV children were older (14.6 vs 12.2 years, .001) and had a lower BMI score (0.30 vs 0.85, .001). In both groups, 87% had received 2 doses of MMR, but the distribution of MMR doses was different between the groups ( .001), as PHIV children were more likely to have received.