Does prior dengue virus exposure worsen clinical outcomes of Zika virus infection? A systematic review, pooled analysis and lessons learned

Abstract

Zika virus (ZIKV) recently caused a pandemic complicated by Guillain-Barre syndrome (GBS) and birth defects. ZIKV is structurally similar to the dengue viruses (DENV) and in vitro studies suggest antibody dependent enhancement occurs in ZIKV infections preceded by DENV; however, the clinical significance of this remains unclear. We undertook a PRISMA-adherent systematic review of all current human and non-human primate (NHP) data to determine if prior infection with DENV, compared to DENV-naïve hosts, is associated with a greater risk of ZIKV clinical complications or greater ZIKV peak viremia in vivo. We identified 1146 studies in MEDLINE, EMBASE and the grey literature, of which five studies were eligible. One human study indicated no increase in the risk of GBS in ZIKV infections with prior DENV exposure. Two additional human studies showed a small increase in ZIKV viremia in those with prior DENV exposure; however, this was not statistically significant nor was it associated with an increase in clinical severity or adverse pregnancy outcomes. While no meta-analysis was possible using human data, a pooled analysis of the two NHP studies leveraging extended data provided only weak evidence of a 0.39 log10 GE/mL rise in ZIKV viremia in DENV experienced rhesus macaques compared to those with no DENV exposure (p = 0.22). Using a customized quality grading criteria, we further show that no existing published human studies have offered high quality measurement of both acute ZIKV and antecedent DENV infections. In conclusion, limited human and NHP studies indicate a small and non-statistically significant increase in ZIKV viremia in DENV-experienced versus DENV-naïve hosts; however, there is no evidence that even a possible small increase in ZIKV viremia would correlate with a change in ZIKV clinical phenotype. More data derived from larger sample sizes and improved sero-assays are needed to resolve this question, which has major relevance for clinical prognosis and vaccine design.