Dengue Literature Review

Dengue Vaccine Candidates

8 min read

Dengue Vaccine Candidates

Overview

Dengue vaccine development is complicated by the need to simultaneously protect against all four serotypes and to do so without inducing partial immunity that could predispose vaccinees to antibody-dependent enhancement (ADE) upon breakthrough infection. As of 2016, one vaccine (CYD-TDV/Dengvaxia) had been approved in multiple countries; others were in phase II/III trials. The key challenge is that the correlates of dengue protection are incompletely understood — high-titre neutralising antibodies do not always predict protection, and the role of T cell immunity is increasingly recognised as important.

Key Points from Literature

The fundamental challenge

  • A dengue vaccine must provide balanced tetravalent immunity — unbalanced responses leave individuals partially immune and potentially at ADE risk for the under-protected serotype
  • Waning immunity is itself a risk: the longer the interval between sensitisation (whether by natural infection or vaccine) and a subsequent heterotypic exposure, the greater the DHF risk (DENV2 secondary infection after 20-year vs 4-year interval → 8× higher DHF rate)
  • No validated animal model for human dengue disease; no single established correlate of protection (see Guzman2016 - Dengue Infection)

CYD-TDV (Dengvaxia; Sanofi Pasteur) — approved

  • Live attenuated chimeric vaccine: prM and E genes of each DENV serotype inserted into 17D yellow fever vaccine backbone (ChimeriVax platform)
  • Phase III in >30,000 children across 10 countries; efficacy 56.5% (Asia) / 60.8% (Americas); >80% against DHF
  • Critical safety signal: seronegative children ≤5 years at vaccination had 5× the hospitalised breakthrough disease rate of controls; mechanism is ADE — vaccine primes naïve individuals with cross-reactive non-neutralising antibodies, enhancing viral entry during subsequent natural infection
  • Approved for ages 9–45 years in five countries; recommended only where >70% of target population is seropositive
  • See dedicated page: CYD-TDV (see Guzman2016 - Dengue Infection)

DENVax / TAK-003 (Takeda) — phase II at publication

  • Live attenuated chimeric vaccine: DENV-2 PDK-53 attenuated backbone with prM/E genes of wild-type DENV-1, -3, -4 substituted in (three chimeric serotypes + attenuated DENV-2)
  • Well-tolerated in children and adults 1.5–45 years; neutralising antibody seroconversion to all four serotypes; T cell-mediated cross-reactive responses
  • Phase III trials initiated in Asian countries (note: TAK-003 subsequently completed phase III and received approval in multiple countries after the Guzman2016 publication date) (see Guzman2016 - Dengue Infection)

TV003 / TV005 (US NIAID / Instituto Butantan) — phase II at publication

  • Live attenuated vaccine using 3′ UTR deletion (Δ30 mutation) to attenuate DENV-1 and DENV-4; recombinant chimeric DENV-2 and DENV-3
  • Human challenge study: single dose of TV003 fully protected all 21 vaccinated volunteers against virologically confirmed dengue infection; all 20 unvaccinated controls became infected
  • TV005 is a higher-titre formulation
  • Does not use yellow fever backbone — different immune priming profile than CYD-TDV (see Guzman2016 - Dengue Infection)

Subunit vaccines

  • DEN-80E (Hawaii Biotech / Merck): truncated recombinant E protein subunit (80% of E, lacking transmembrane domain); induces neutralising antibodies in mice and non-human primates; phase I of DENV-1-80E completed; tetravalent phase I began 2012
  • DIII-C (IPK Cuba / CIGB): domain III of E protein fused to capsid protein; induces serotype-specific antibody and cellular immunity in Balb/c mice and Vervet monkeys; advanced preclinical stage; developed by the same IPK group as the Garcia2009/2010 cohort studies
  • Advantages: no ADE risk from viral replication; balanced tetravalent responses feasible; accelerated immunisation schedule
  • Disadvantages: require adjuvants and multiple doses; may not induce as durable immunity as live attenuated vaccines (see Guzman2016 - Dengue Infection)

Inactivated vaccines

  • DPIV (purified formalin-inactivated; WRAIR/GSK): immunogenic in rhesus macaques; phase I completed 2011; no replication risk; multiple doses required; adjuvants needed (see Guzman2016 - Dengue Infection)

Implications for protective immunity (relevant to Sungnak2025)

  • Sungnak2025 suggests that CD8 T cell effector responses (non-exhausted, functional, directed against NS protein epitopes) are the hallmark of protective asymptomatic dengue, while IGHG1+ plasmablast expansion is a marker of pathogenic DHF
  • This aligns with the expectation that vaccines including NS protein antigens (TV003’s live attenuated design; some DNA vaccine approaches) may induce more protective T cell responses than vaccines relying solely on E protein (subunit 80E, inactivated virus)
  • CYD-TDV, which uses only prM/E in the yellow fever backbone without NS proteins, may therefore be structurally less capable of inducing the NS-epitope CD8 T cell responses that appear protective in natural infection (see Sungnak2025 - Distinct Immune Responses Asymptomatic Symptomatic Dengue)

NS1-based vaccine: protective AND pathogenic (Wan2012)

  • Anti-NS1 Abs fix complement and trigger complement-mediated lysis of DENV-infected cells — a protective mechanism distinct from neutralisation
  • Active immunisation with NS1 protein or passive transfer of anti-NS1 Abs protects mice against lethal DENV challenge
  • However, anti-NS1 Abs also cross-react with platelets, endothelial cells, and coagulation factors via molecular mimicry — pathogenic
  • Resolution requires epitope engineering: remove or modify the C-terminal aa 277–352 pathogenic epitopes while preserving the protective complement-fixing epitopes
  • A successful dengue vaccine must therefore address three constraints: tetravalent efficacy, no ADE, AND no autoimmune cross-reactivity (see Wan2012 - Autoimmunity in Dengue Pathogenesis)

Anti-prM ADE risk from native prM sequences (Dejnirattisai2010)

All three leading vaccine platforms — CYD-TDV, TAK-003, and TV003/TV005 — incorporate native prM sequences alongside E protein genes. Dejnirattisai2010 demonstrates that natural DENV infection primes anti-prM antibodies as the dominant structural antibody class (~60%), and that these antibodies are:

  • Fully cross-reactive across all four serotypes
  • Unable to neutralise above a 10–60% ceiling (structural limit from incomplete prM cleavage)
  • Capable of up to 10^5-fold ADE enhancement in primary monocytes and DCs

Any vaccine using native prM sequences will likely prime an anti-prM response with the same profile. The authors call for heterologous prM sequences or prM deletion in future vaccine constructs to avoid priming this ADE-prone antibody class. This is an unresolved challenge: as of the sources in this wiki, no approved dengue vaccine has adopted a modified or deleted prM approach.

See prM Protein and Antibody-Dependent Enhancement for full mechanistic context. This adds a specific structural-antibody-level constraint to the ADE concern documented for CYD-TDV seronegatives (see CYD-TDV). (see Dejnirattisai2010 - Anti-prM Antibodies Enhance Dengue ADE)

Quantitative ADE window — a rational safety benchmark for vaccine titre monitoring (Bhatt2020)

Bhatt2020 - Dengue Pathogenesis Review introduces the Katzelnick 2017 quantitative ADE window (1:21–1:80 as peak enhancement titer; n=6,684 Nicaraguan children) as a benchmark for evaluating vaccine safety over time. This converts the qualitative “waning immunity = ADE risk” concern into a testable quantitative criterion:

  • Vaccines that induce neutralising titres well above 1:1,280 initially may provide protection, but as titres wane (which occurs at different rates for different vaccines and recipients), any period spent in the 1:21–1:80 range represents a potential ADE vulnerability window
  • For seronegative CYD-TDV vaccinees specifically: the vaccine induced titres that, as they waned, passed through the enhancement range — consistent with the observed 5× hospitalisation risk in this group
  • For TAK-003 (Qdenga) and TV003: post-vaccination waning kinetics have not been benchmarked against the Katzelnick window in available wiki sources; whether their titres track through 1:21–1:80 as they wane is unknown
  • The window also contextualises the Bos2025 (PREPRINT) observation that XR E-IgG rises post-primary — if rising cross-reactive antibody titres are in the enhancement range, duration of risk may differ from the simple waning model

(see Antibody-Dependent Enhancement, CYD-TDV)

Contradictions & Debates

  • CYD-TDV seronegative risk: The ADE interpretation of the CYD-TDV safety signal is widely accepted but mechanistic proof in vaccinees (e.g., FcR pathway enrichment, IGHG1 plasmablast expansion as in DHF Sungnak2025 data) has not been directly demonstrated.
  • Neutralising antibody vs. T cell correlates: High neutralising antibody titres in vaccinees do not always predict protection (observed for DENV-2 in some CYD-TDV vaccinees); T cell immunity is increasingly recognised as important but is not yet a validated regulatory correlate.
  • Anti-prM vs. anti-E as dominant ADE mechanism: Standard vaccine immunogenicity readouts measure anti-E neutralising titres. If anti-prM is the more potent and numerically dominant ADE pathway (Dejnirattisai2010), current vaccine evaluation frameworks may be missing the most important enhancing antibody class. Whether anti-prM titres post-vaccination track with breakthrough DHF risk has not been tested.

Sources

Graph View

Backlinks