Dengue Literature Review

Secondary Dengue Infection

11 min read

Secondary Dengue Infection

Overview

Secondary dengue infection refers to infection with a DENV serotype different from a prior primary infection. It is the single most important risk factor for severe dengue (DHF/DSS). During secondary infection, pre-existing cross-reactive but non-neutralising antibodies from the primary infection can enhance viral uptake into Fc receptor-bearing cells (antibody-dependent enhancement, ADE), leading to higher viraemia, greater immune activation, and increased risk of vascular permeability and haemorrhagic complications.

Key Points from Literature

Epidemiological risk

  • Severe dengue (DHF/DSS) occurs predominantly in secondary heterotypic infections — this is the core observation underlying the ADE hypothesis (see Antibody-Dependent Enhancement)
  • Third and fourth infections are typically mild or asymptomatic, suggesting that broad immunity develops after exposure to 2+ serotypes (see Guzman2016 - Dengue Infection)
  • A longer interval between primary and secondary infection increases DHF risk: in Cuba, a 20-year interval (DENV-2, 1981→2001) produced ~8× higher DHF rates than a 4-year interval; waning neutralising antibody levels may shift the balance toward enhancement (see Guzman2016 - Dengue Infection)

Immunological features

  • Anamnestic IgG response: In secondary infection, IgG rises within 1–2 days of illness onset at high titres (≥1,280 by HI), unlike primary infection where IgG appears around day 7–10 (see IgM-IgG Serology ELISA)
  • IgG:IgM ratio > 1.2 is used in research settings to classify secondary infection
  • Complement activation proceeds via the classical pathway in secondary infection (immune complexes), vs. alternative pathway in primary (see Dengue Pathophysiology)
  • sNS1 blood levels correlate with disease severity specifically in secondary infection, not in primary (see Viraemia)

Serotype-specific patterns

Asymptomatic secondary infection

  • Not all secondary infections are symptomatic: in the 2006 Cuban DENV-4 epidemic, among individuals with prior DENV-1 exposure (secondary dengue), the overt:subclinical ratio was approximately 1:1 (see Garcia2010 - Asymptomatic Dengue FcγRIIa Polymorphism)
  • FcγRIIa-RR131 genotype was significantly associated with asymptomatic outcome in secondary infection, suggesting host genetics modulate whether ADE leads to clinical disease (see FcγRIIa Receptor)

Anti-prM as dominant cross-reactive ADE-mediating antibody in secondary infection context

Dejnirattisai2010 - Anti-prM Antibodies Enhance Dengue ADE provides the mechanistic explanation for why cross-reactive antibodies from a prior primary infection are so effective at mediating ADE during secondary infection. Approximately 60% of anti-dengue structural antibodies in DENV-infected humans are directed against the prM protein — fully cross-reactive across all four serotypes — and cannot neutralise above a 10–60% ceiling regardless of titre. These antibodies mediate up to 10^5-fold enhancement in primary monocytes and DCs.

In the secondary infection context, the implication is that the dominant non-neutralising cross-reactive antibody pool that encounters a new heterotypic serotype is not primarily sub-threshold anti-E antibody (the classical model) but anti-prM antibody that is structurally incapable of neutralisation. See Antibody-Dependent Enhancement and prM Protein for the full mechanism.

Antibody kinetics: primary vs secondary divergence (Bos2025)

Bos2025 - Longitudinal Antibody Dynamics After Dengue (PREPRINT) provides the most detailed comparison of antibody trajectories across primary and secondary infection currently in this wiki, in a Nicaraguan pediatric cohort (n=79, DENV-1/DENV-3, followed to 18M):

PREPRINT — not peer reviewed.

  • Secondary infection kinetics are more compressed and attenuated at 18M relative to primary — consistent with rapid memory recall and faster contraction in a primed host, rather than the sustained expansion seen after primary exposure
  • Cross-reactive E-IgG (XR E-IgG) rises 6–18M post-primary (t½ = −2.13y), suggesting active accumulation of the cross-reactive pool in the inter-infection window — the pool most likely to mediate ADE upon secondary challenge (see Cross-Reactive Antibodies)
  • NS1-IgG wanes post-primary (t½ ≈ 2.1y), following the classical type-specific decay model; secondary kinetics show similar waning pattern
  • The divergence between primary (rising XR E-IgG, waning NS1-IgG) and secondary (attenuated 18M responses overall) suggests that the immunological state entering a potential tertiary infection is qualitatively different from the state entering secondary infection, in ways the classical ADE model does not fully capture

Thrombocytopenia mechanism in secondary infection — PAIgG and PAIgM

Two consecutive prospective studies from Manila (same research group, same institutions) establish a two-isotype immune complex model for thrombocytopenia in secondary dengue.

Oishi2003 - PAIgG and Thrombocytopenia in Secondary Dengue (n=53, Manila 2001):

  • Platelet-associated IgG (PAIgG) is the proximate driver: a significant inverse correlation exists between PAIgG levels and platelet count (r = −0.570, P < 0.0001); absent in healthy volunteers.
  • Anti-dengue virus IgG — not anti-platelet autoantibody — is responsible: anti-platelet IgG autoantibody was present in plasma in only 1/53 patients (1.9%). IgG eluted from platelets of all 8 tested patients was confirmed anti-dengue virus IgG by indirect ELISA.
  • Mechanism: immune complexes of dengue antigen + anti-dengue IgG bind to platelets via direct dengue-platelet interaction — FcγRII is explicitly NOT required for this step (citing Wang 1995). PAIgG formation subsequently triggers platelet clearance by macrophages and/or complement-mediated platelet lysis.
  • Longitudinal confirmation: platelet counts recovered (43 → 307 × 10³/μl, P = 0.0022) while PAIgG fell (1.49 → 0.31 ng/10⁷ platelets, P = 0.0037) together from acute to convalescent phase (4 days).
  • PAIgG level does not correlate with HI titre (r = −0.130, P = 0.361), implying circulating antibody concentration alone is insufficient — viral antigen availability on platelets appears rate-limiting.

Saito2004 - PAIgG and PAIgM in Secondary Dengue (n=78, Manila 2002–2003) extends the Oishi2003 finding to include PAIgM:

  • PAIgM also inversely correlates with platelet count in secondary infection: r = −0.231, P = 0.046; elevated in acute phase (17.5 ± 20.4 ng/10⁷) vs. healthy controls (4.2 ± 3.8 ng/10⁷, P < 0.001); falls at convalescence (P < 0.001).
  • Anti-dengue IgM activity confirmed in platelet eluates from secondary-infection patients (OD 0.35 ± 0.20 vs. 0.09 ± 0.05 in controls) — PAIgM is also an immune complex carrying anti-dengue specificity, not anti-self.
  • PAIgM is quantitatively smaller: 50.7% of patients had normal PAIgM vs. only 19.2% with normal PAIgG — PAIgG is the dominant contributor to thrombocytopenia.
  • PAIgM mechanism bypasses Fc receptors entirely: IgM pentamer structure cannot engage FcγRs; clearance proceeds only via complement receptor- and complement-mediated lysis — completely FcγR-independent.
  • PAIgM independently predicts DHF (logistic regression, P < 0.01): cut-off >20 ng/10⁷ platelets gives 92.1% specificity and 48.6% sensitivity for DHF; both PAIgG and PAIgM are significantly higher in DHF than DF.
  • PAIgM >20 ng/10⁷ in virologically confirmed secondary infection is the highest-specificity clinical predictor of DHF severity in this data.

Contrast with primary infection (see Lin2001 - IgM Anti-Platelet Autoantibody in Dengue Patients): primary DENV-3 infection drives thrombocytopenia via IgM anti-platelet autoantibodies (NS1 molecular mimicry), not immune complexes. Saito2004 and Oishi2003 show that the IgM elevated on platelets in secondary infection is anti-dengue virus IgM (not anti-self), mechanistically distinct from Lin2001’s primary-infection IgM autoantibody. The bifurcation is now precisely defined:

  • Primary: IgM anti-platelet autoantibody (NS1 mimicry; Lin2001)
  • Secondary: PAIgG + PAIgM immune complexes, both carrying anti-dengue specificity (Oishi2003 + Saito2004)

IgG Autoantibody Paradox: Primary > Secondary

A protein microarray screen of 123 autoantigens in Cambodian children found that primary DENV infection was associated with significantly higher IgG autoantibody levels than secondary infection (p < 0.01 for total IgG NFI; 70 individual IgG autoantibodies elevated in primary vs. secondary) (see Vo2020 - Autoantibody Profiling in Dengue). This is counterintuitive: secondary infection is the context of greater disease severity, higher viral loads, and the ADE mechanism — all of which might be expected to drive greater immunological dysregulation including more autoantibody production.

The Vo2020 interpretation attributes the primary-infection IgG autoantibody excess to impaired B cell tolerance checkpoints during first exposure, where naïve autoreactive B cells may escape deletion more readily in the absence of memory competition. In secondary infection, rapid memory recall responses dominate the B cell expansion, possibly outcompeting autoreactive clones that lack this established memory advantage.

This finding sits in tension with the severity-autoantibody correlation documented by Wan2012 (anti-endothelial cell Ab levels higher in DHF/DSS than DF by flow cytometry) — but the two studies measure different autoantibody populations by different methods (protein array NFI vs. anti-endothelial cell flow cytometry), and the Vo2020 comparison is primary vs. secondary infection order, not DF vs. DHF severity. They are compatible if: (a) secondary infection produces less total IgG autoantibody but the specific anti-endothelial autoantibodies tracked by Wan2012 are higher in DHF; or (b) in secondary DHF, autoantibodies are rapidly consumed through immune complex formation (consistent with the Vo2020 platelet-count correlation), leaving the measured total pool depleted even as pathogenic activity is high.

Diagnostic classification

Contradictions & Debates

  • PAIgG vs. autoantibody thrombocytopenia: Oishi2003 shows near-absent anti-platelet IgG autoantibody in secondary infection, while Lin2001 documents IgM anti-platelet autoantibody in primary infection. The two mechanisms are not mutually exclusive — secondary infection patients may harbour residual IgM autoantibody pathway alongside the dominant PAIgG pathway, but no study has directly measured both simultaneously across infection order.
  • FcγRIIa genotype and platelets: Oishi2003/Wang 1995 state FcγRII is not required for dengue-platelet binding (and therefore for PAIgG formation). Yet Garcia2010 shows HH genotype dramatically increases DHF risk. FcγRIIa’s role in severity must therefore operate through a mechanism other than direct platelet binding — most likely ADE-mediated viral uptake in monocytes/macrophages (see FcγRIIa Receptor).
  • The relationship between secondary infection and severe disease is probabilistic, not deterministic — host genetics, infecting serotype sequence, inter-infection interval, and viral genotype all modulate outcome
  • Whether ADE is the primary mechanism or whether T-cell original antigenic sin and cytokine-mediated pathology contribute independently remains debated (see Original Antigenic Sin, T Cell Responses in Dengue)
  • CYD-TDV (Dengvaxia) vaccination of seronegatives created an ADE-like state mimicking secondary infection upon natural exposure — raising the question of whether vaccine-induced and naturally-acquired enhancing antibodies behave identically (see CYD-TDV)

Sources

Graph View

Backlinks