Variation in dengue virus plaque reduction neutralization testing: systematic review and pooled analysis.

BACKGROUND: The plaque reduction neutralization test (PRNT) remains the gold standard for the detection of serologic immune responses to dengue virus (DENV). While the basic concept of the PRNT remains constant, this test has evolved in multiple laboratories, introducing variation in materials and methods. Despite the importance of laboratory-to-laboratory comparability in DENV vaccine development, the effects of differing PRNT techniques on assay results, particularly the use of different dengue strains within a serotype, have not been fully characterized.

METHODS: We conducted a systematic review and pooled analysis of published literature reporting individual-level PRNT titers to identify factors associated with heterogeneity in PRNT results and compared variation between strains within DENV serotypes and between articles using hierarchical models.

RESULTS: The literature search and selection criteria identified 8 vaccine trials and 25 natural exposure studies reporting 4,411 titers from 605 individuals using 4 different neutralization percentages, 3 cell lines, 12 virus concentrations and 51 strains. Of 1,057 titers from primary DENV exposure, titers to the exposure serotype were consistently higher than titers to non-exposure serotypes. In contrast, titers from secondary DENV exposures (n = 628) demonstrated high titers to exposure and non-exposure serotypes. Additionally, PRNT titers from different strains within a serotype varied substantially. A pooled analysis of 1,689 titers demonstrated strain choice accounted for 8.04% (90% credible interval [CrI]: 3.05%, 15.7%) of between-titer variation after adjusting for secondary exposure, time since DENV exposure, vaccination and neutralization percentage. Differences between articles (a proxy for inter-laboratory differences) accounted for 50.7% (90% CrI: 30.8%, 71.6%) of between-titer variance.

CONCLUSIONS: As promising vaccine candidates arise, the lack of standardized assays among diagnostic and research laboratories make unbiased inferences about vaccine-induced protection difficult. Clearly defined, widely accessible reference reagents, proficiency testing or algorithms to adjust for protocol differences would be a useful first step in improving dengue PRNT comparability and quality assurance.