Dengue Literature Review

Hung2008 - Anti-Platelet Anti-Endothelial Autoantibodies Vietnam

7 min read

Hung2008 - Anti-Platelet Anti-Endothelial Autoantibodies Vietnam

Full citation: Hung NT, Lan NT, Lin YS, Lin CF, Lien LB, Huang KJ, Yeh TM, Ha DQ, Huong VTQ, Chen LC, My LT, Huang JH, Liu CC, Lei HY. Anti-Platelet and Anti-Endothelial Cell Autoantibodies in Vietnamese Infants and Children with Dengue Hemorrhagic Fever. American Journal of Infectious Diseases. 2008;4(1):41–49. doi:10.3844/ajidsp.2008.41.49

Raw file: [[raw/Hung2008.pdf]]

Summary

This study verifies the presence of anti-platelet and anti-endothelial cell (anti-EC) autoantibodies in two distinct age and infection-order groups of DHF/DSS patients from an endemic dengue region (Ho Chi Minh City, Vietnam): infants (<12 months) with predominantly primary dengue infections, and children (4–15 years) with predominantly secondary dengue infections. The paper extends earlier NCKU Taiwan findings (Lin2001, Lin2006) by confirming that: (1) anti-platelet IgM autoantibodies are not confined to secondary infection or to Taiwan’s DENV-3 outbreak context but arise in primary infection in infants in a hyperendemic setting, across multiple serotypes; and (2) anti-EC autoantibodies show an infection-order-dependent isotype pattern — IgM only in infants/primary, IgM + IgG in children/secondary.

The study also provides direct in vivo evidence of endothelial structural damage via elevated serum thrombomodulin (TM) levels in both groups. Critically, neither autoantibody levels nor TM levels correlated with DHF severity grade (non-shock DHF vs. DSS), platelet counts, or haematocrit increase, systematically dissociating autoantibody quantity from acute clinical outcomes.

Study Design

  • Type: Cross-sectional observational cohort (two prospective cohorts)
  • Sample size: 50 infants (<12 months; May 1998 – March 2002) + 37 children (4–15 years; September–November 2001); anti-EC autoantibody subset n=20 infants + 33 children; TM subset n=24 infants + 34 children; controls n=10 infants + 7 children
  • Setting: Department of Dengue Haemorrhagic Fever, Children’s Hospital No.1, Ho Chi Minh City, Vietnam; autoantibody assays performed at National Cheng Kung University (NCKU), Tainan, Taiwan
  • Population: Dengue-confirmed hospitalised paediatric patients (WHOcriteria DHF/DSS); infants predominantly primary, children 89.2% secondary

Key Findings

  • Anti-platelet IgM elevated in all patients: Infants 16.5 ± 8.6% vs controls 1.5 ± 1.2% (p<0.001); children 11.0 ± 8.2% vs controls 3.2 ± 1.0% (p<0.001). The dominant isotype is IgM in both groups; IgG anti-platelet Abs are either absent (children) or present at very low levels (infants)
  • First demonstration of anti-platelet IgM in infants with primary dengue in an endemic region (n=49/50 primary; DENV-3, DENV-4 infections); extends the Lin2001 Taiwan DENV-3 primary-infection finding to Vietnam and multiple serotypes
  • Anti-EC isotype shift by infection order:
    • Infants (primary): IgM 17.6 ± 10.5% vs controls 4.0 ± 1.9% (p=0.003); IgG not significantly elevated
    • Children (secondary): IgM 46.0 ± 15.3% vs controls 4.7 ± 0.5% (p<0.001); IgG 25.1 ± 8.5% vs controls 2.6 ± 0.5% (p<0.001) — substantially higher overall than infants
    • The addition of anti-EC IgG in secondary infection probably reflects immune memory producing affinity-matured IgG in a previously primed host
  • Infection-order independence for anti-platelet IgM: In children, levels were not significantly different between primary (n=3) and secondary (n=33) infections (p=1.0), consistent with Cheng2015 - NS1 P311-330 Anti-PDI Autoantibodies in DHF (caveat: only 3 primary-infected children)
  • Infection-order independence for anti-EC IgM and IgG in children: p=1.0 and p=0.5 respectively between primary (n=4) and secondary (n=29) children
  • Serotype independence: Anti-platelet IgM and anti-EC IgM/IgG were not significantly different between DEN-3 vs DEN-4 (infants) or DEN-2 vs DEN-4 (children)
  • No severity correlation: Anti-platelet IgM and anti-EC IgM/IgG were not significantly different between non-shock DHF and DSS in either age group
  • No platelet count correlation: Linear regression showed no association between anti-platelet IgM levels and lowest platelet count in either group
  • No plasma leakage correlation: Anti-EC levels not correlated with haematocrit increase (evidence of plasma leakage)
  • Thrombomodulin elevated — in vivo endothelial structural damage confirmed: Infants 6.1 ± 1.7 vs controls 2.7 ± 0.4 pg/mL (p<0.001); children 8.8 ± 2.1 vs controls 2.4 ± 0.06 pg/mL (p=0.01). TM did not correlate with severity grade or with anti-EC levels

Methods Used

  • Flow Cytometry (anti-platelet and anti-endothelial autoantibody detection; FITC-conjugated anti-human IgM/IgG secondary Abs; FACScan Becton Dickinson; excitation 488 nm)
  • IgM-IgG Serology ELISA (E/M-specific capture IgM and IgG ELISA for dengue confirmation; IgM/IgG ratio ≥1.2 = primary; <1.2 = secondary)
  • NS1 Antigen Detection (NS1 serotype-specific IgG ELISA for serotyping; positivity = OD ratio highest:second-highest serotype ≥2.0; serotype specificity if ratio ≥1.2)
  • RT-PCR (virus isolation and/or RT-PCR for DEN-2, DEN-3, DEN-4 serotype confirmation in children, performed at Pasteur Institute Ho Chi Minh City and CDC Taipei)
  • ELISA (IMUBIND Thrombomodulin ELISA Kit for serum TM levels)

Entities Mentioned

Concepts Addressed

Relevance & Notes

This paper is the Vietnamese endemic-setting replication of the foundational NCKU Taiwan findings (Lin2001, Lin2006). Its two most important contributions to the wiki are:

  1. Infant/primary anti-platelet IgM in an endemic country: Lin2001 established this in a Taiwan outbreak (n=9 DENV-3, mixed severity). Hung2008 confirms it in 49 infants with primary DENV-3/4 infections in hyperendemic Vietnam. This strongly suggests the mechanism is not outbreak-specific or serotype-limited.

  2. Anti-EC isotype shift by infection order: Infants (primary) produce IgM anti-EC only; older children (predominantly secondary) produce both IgM and IgG with markedly higher overall levels. This pattern mirrors the well-known IgM→IgG isotype class switching that occurs upon recall responses and provides the first age/infection-order-stratified data for anti-EC autoantibodies. The NS1 absorption-confirmed mechanism from Lin2006 is cited as the explanatory framework.

The thrombomodulin finding is notable: TM is a marker of actual endothelial cell injury (not just permeability), and its elevation in both infant and child groups confirms that endothelial structural damage occurs in vivo in DHF/DSS — not merely functional permeability change. The absence of a TM-severity or TM-anti-EC correlation is consistent with a picture in which endothelial damage is present across the DHF/DSS severity spectrum rather than being the determining factor for shock progression.

The systematic lack of autoantibody-to-clinical-outcome correlations (no anti-platelet ~ platelet count, no anti-EC ~ hematocrit, no either ~ DHF grade) is important: it reinforces the multi-mechanism model of DHF. Autoantibodies are present and pathologically plausible, but their levels alone do not determine severity outcomes. Context: Lin2001 also found no platelet count correlation; the current paper extends this to the anti-EC/hematocrit dimension.

Limitation: The NS1-P1 (aa 1–15) epitope proposed in the discussion as the anti-platelet IgM driver (Huang et al. 1999) is a speculative mechanism citation, not tested in this paper. Lin2001 and Lin2006 established NS1 absorption confirmation in Taiwan; Hung2008 does not repeat this absorption assay in its Vietnamese cohort.

Questions Raised

  • Is the anti-EC IgG addition in secondary infection a direct consequence of immune memory class switching, or does a distinct secondary-infection-specific process generate anti-EC IgG? No experimental evidence distinguishes these in this paper.
  • Does the TM-anti-EC level dissociation mean that anti-EC autoantibodies are not the primary driver of endothelial damage, or merely that the assay is insufficiently sensitive? If another mechanism (e.g., cytokine storm, direct NS1 effects) is responsible for most TM release, autoantibodies may be a minor contributor.
  • Do anti-platelet IgM levels correlate with platelet counts in non-DHF dengue fever (DF) patients? The study enrolled only DHF/DSS patients — this question is unanswerable from this dataset.
  • Does the anti-EC IgG in secondary infection children correlate with the dengue E protein or NS1 protein specifically? No absorption assay was performed; the mechanistic basis follows Lin2006 by inference, not by direct NS1 absorption confirmation in this cohort.

Graph View

Backlinks