Localized Inflammation in Dengue Vaccine-Induced Skin Rash Is Not Associated with Continuous Presence of Dengue Virus Genome

Abstract

Vaccination with the tetravalent live-attenuated dengue virus (DENV) vaccines TV003 and TV005 causes a mild, relatively localized erythematous maculopapular skin rash in most dengue-naïve vaccinees. Human challenge model DENV strains, DENV2Δ30 and DENV3Δ30, trigger a confluent skin rash over most of the body in most unvaccinated participants. To determine the etiology of these rashes, we performed in situ hybridization for DENV genome and assessed cellular infiltration by H&E staining in skin biopsies from humans infected with live-attenuated dengue vaccine DENV2Δ30 or DENV3Δ30 challenge strains. Sixty-three biopsies from 40 participants were included in the study, of which 43 biopsies from 32 patients contained intact RNA. Of these, 1 sample taken from a nonerythematous site from a DENV2Δ30-infected participant experiencing a rash was positive for DENV2 genome. Incidence and severity of lymphocytic infiltration were highest in rash biopsy samples than in those from noninvolved areas in participants experiencing a rash or from those taken from participants not experiencing a rash. These results indicate that the rash associated with infection with live-attenuated dengue vaccines or challenge strains is predominantly lymphocyte-driven perivascular dermal inflammation without local concomitant active viral replication.

Keywords: Challenge virus; Dengue tetravalent attenuated vaccine; Neutralizing antibodies; Viremia; immunogenicity.

Figure 1 |. Overview of dengue vaccine rash study procedures.

(a) Participants without a rash had one biopsy taken. Those with a rash had biopsies taken from a rash site and a non-rash site. (b) Anterior upper trunk shoulders area of a TV003 vaccinee with a maculopapular dengue vaccine rash and hematoxylin and eosin staining of a biopsy taken from a rash site (solid circle) or a non-erythematous non-rash site (dashed line circle). Scale bars: 500 μm for larger image (4× magnification), 100 μm for inset (20× magnification). Written consent stated that photographs may be taken and that research using data and materials collected from volunteers may be presented in publications in a non-identifiable way. (c) Overview of DENV genome RNA in situ hybridization of formalin-fixed paraffin-embedded (FFPE) sections from skin biopsies.

Figure 2 |. Detection of DENV infection of human cells and tissue by RNAScope.

(a) K562 cells stably expressing DC-SIGN were infected at multiplicity of infection (MOI) = 1 with the indicated virus for three days. DENV infection was detected by flow cytometry with an Alexa488-labelled 2H2 anti-DENV mAb in cells that were fixed with paraformaldehyde and permeabilized with 0.5% saponin. (b) Cell buttons from K562 DC-SIGN cells were processed for RNAScope using PPIB (RNA integrity control), dapB (bacterial RNA negative control), DENV1–4 pooled probe or DEN2-NS2A probe. Scale bar = 50 μm (c) PPIB and DENV pooled probe staining in human skin explants were exposed to inoculation with vial culture medium (no virus), live DENV1–4 tetravalent admixture, or heat-inactivated DENV1–4. Scale bar = 60 μm

Figure 3 |. Clinical presentation and skin biopsy sampling participants with DENV vaccine-induced rash.

The time course of clinical presentation and skin biopsy sampling and results of DENV1–4 pooled probe, DENV2-NS2A probe, and PPIB probe staining is presented for sixteen subjects with rash following exposure to either TV003, TV005, DEN2Δ30 or DENVΔ30. Box with rash graphics denotes period after exposure in which rash was observed and filled box with “X” denotes time of “rash” and “non-rash” biopsies (see Figure 1). nd, not determined; +, staining present, −, staining not present.

Figure 4 |. Viremia and serum neutralizing antibodies following vaccination with TV003 or TV005, or challenge with DENV2Δ30 or DENV3Δ30.

(a) Serum viremia (all serotypes) measured by viral focus assay (presented in focus-forming units, FFU/mL) and (b) neutralizing antibodies (all serotypes) measured by focus reduction neutralization 50% assay (FRNT)₅₀.

Figure 5 |. Evidence of DEN2 genome in skin of a participant with DENV2Δ30-induced rash.

DENV2-NS2A and DENV1–4 pooled probes were used to stain non-rash and rash sites from biopsies taken at day 12 following DENV2Δ30 infection (participant 287.D2-D). Hematoxylin and eosin (H & E) staining also shown. Upper left panel arrows, DENV2 RNA. Scale bars for RNAscope images is 50 μm and is 600 μm for H&E panels.

Figure 6 |. Histological inflammation in DENV vaccine associated rash.

(a), Grade 0 (no rash sample), (b) Grade 1 rash, (c) Grade 2 rash, (d), Grade 3 rash. Key features are perivascular and interstitial lymphomononuclear infiltrate within the dermis. Scale bar: 500 μm (4× magnification). (e) Summary of histology results for all samples (n = 63).