Dengue Literature Review

Berlin2007 - Autoantibodies in Nonautoimmune Individuals during Infections

6 min read

Berlin2007 - Autoantibodies in Nonautoimmune Individuals during Infections

Full citation: Berlin T, Zandman-Goddard G, Blank M, Matthias T, Pfeiffer S, Weis I, Toubi E, Singh S, Asherson R, Shoenfeld Y. Autoantibodies in nonautoimmune individuals during infections. Ann N Y Acad Sci. 2007;1108:584–593. doi:10.1196/annals.1422.061

Raw file: [[raw/berlin2007.pdf]]

Summary

This study measured a broad panel of autoantibodies in patients hospitalised with acute infections — bacterial, viral, parasitic, or rickettsial — and compared them against healthy blood donor controls. The goal was to characterise the degree to which acute non-autoimmune infections transiently induce autoantibodies, with the implicit mechanistic question of whether infections can trigger autoimmunity in otherwise healthy individuals.

Autoantibodies were detected at substantially higher rates in infected patients than in controls across nearly all tested specificities. ANA positivity was found in 21.7% of viral infection patients and 20% of bacterial infection patients, compared with only 3.8% of healthy controls. A novel finding was the high prevalence of anti-annexin-V and anti-prothrombin antibodies across infection types, suggesting antiphospholipid-type autoimmunity as a common infection-triggered phenomenon. Importantly, prior literature cited by the authors suggests that some of these infection-induced autoantibodies are transient — dropping from 20.5% to 6.4% ANA positivity between acute and convalescent viral hepatitis samples.

The study draws on a clinically diverse cohort spanning HAV, HBV, HCV, multiple bacterial species, parasites, and rickettsiae, providing a rare cross-infection comparison of autoantibody profiles.

Study Design

  • Type: Cross-sectional; prospective case-control
  • Sample size: 88 infected patients + 80 healthy blood donor controls (n=168 total)
    • Bacterial: n=41 (Salmonella typhi, Streptococcus, Staphylococcus, Pseudomonas, Klebsiella, Acinetobacter)
    • Viral: n=23 (HAV=10, HBV=5, HCV=8)
    • Parasitic: n=17 (Plasmodium falciparum, Toxoplasma, Leishmania, microfilaria)
    • Rickettsial: n=7
  • Setting: Hospital inpatient; Israel (single centre implied); time period not specified
  • Population: Adults with confirmed acute infections; no prior history of autoimmune disease

Key Findings

ANA Prevalence by Infection Type

  • Viral infections: 21.7% ANA positive (P<0.013 vs. controls)
  • Bacterial infections: 20% ANA positive (P<0.006 vs. controls)
  • Parasitic infections: 17.6% ANA positive (not significant vs. controls)
  • Rickettsial infections: 0% ANA positive
  • Healthy controls: 3.8% ANA positive
  • Method: ELISA (ANA 8 Pro panel, Euroimmun) at 1:100 dilution; panel covers anti-SSA, anti-SSB, anti-Sm, anti-U1RNP, anti-Sm/RNP, anti-centromere, anti-Jo-1, anti-Scl-70

ANA Specificity Details

  • HAV patients (n=10): positive for all 8 antigens tested in ANA 8 Pro panel
  • One HBV patient: positive for anti-Sm, anti-U1RNP, anti-Sm/RNP, and anti-centromere
  • These specific nuclear antigens (Sm, U1RNP, centromere) are typically considered disease-associated autoantibodies — their transient detection during acute viral infection is notable

Anti-Annexin-V and Anti-Prothrombin (Most Prevalent Autoantibodies)

  • Anti-annexin-V: viral 60.9%, parasitic 47.1%, rickettsial 57.1%, bacterial 39% — substantially elevated across all infection types
  • Anti-prothrombin: viral 69.6%, bacterial 22%, parasitic 23.5%
  • These antiphospholipid-associated autoantibodies were the most commonly detected overall

Other Autoantibodies Elevated in Infections

  • Anti-Saccharomyces cerevisiae antibodies (ASCA): viral 26.1%, bacterial 22%
  • Anti-β2-glycoprotein-I (anti-β2GPI): elevated in multiple infection groups
  • Multiple additional specificities (anti-cardiolipin, anti-phosphatidylserine, anti-aPT) were measured and elevated in infection groups vs. controls

Multiplicity of Autoantibodies

  • 34/88 (38.6%) infected patients had ≥2 autoantibodies simultaneously:
    • 16 patients: exactly 2 autoantibodies
    • 7 patients: exactly 3 autoantibodies
    • 8 patients: exactly 4 autoantibodies
    • 2 patients: exactly 5 autoantibodies
    • 1 patient: exactly 6 autoantibodies

Transience of Infection-Triggered Autoantibodies

  • Authors cite a prior study (not fully detailed) showing ANA positivity of 20.5% during acute viral hepatitis, dropping to 6.4% in convalescence — consistent with transient, infection-triggered autoantibody production rather than persistent autoimmunity
  • This temporal pattern is central to the interpretation: infection-triggered autoantibodies may resolve with clearance of infection

Methods Used

  • ELISA (ANA 8 Pro panel, Euroimmun; detects anti-SSA, anti-SSB, anti-Sm, anti-U1RNP, anti-Sm/RNP, anti-centromere, anti-Jo-1, anti-Scl-70; 1:100 dilution) (note: this is a specific-antigen ELISA, not the same as Indirect Immunofluorescence ANA Test used in population-reference studies; no equivalent method page currently exists in the wiki)
  • Additional ELISA panels for antiphospholipid antibodies (anti-annexin-V, anti-prothrombin, anti-β2GPI, anti-cardiolipin)
  • ASCA ELISA

Entities Mentioned

(None specific to the topics of this wiki — paper covers bacterial, viral, parasitic, and rickettsial infections; no dengue cases included)

Concepts Addressed

Relevance & Notes

This paper is the most directly relevant comparator to the 23.1% ANA positivity reported in Garcia2009 - Long-term Clinical Symptoms Post-Dengue. Garcia2009 measured ANA 2 years post-acute infection in symptomatic patients; Berlin2007 measures ANA during acute infection in hospitalised patients. The 21.7% ANA rate during acute viral infections closely matches Garcia2009’s post-dengue finding, raising the question: is the Garcia2009 finding best understood as a persisting effect of acute-infection-triggered autoimmunity, or a specifically dengue-driven phenomenon?

Key interpretive points:

  1. Berlin2007’s 3.8% control rate is lower than all major healthy-population estimates from NHANES-based studies (13.8% Satoh2012; 16.1% Dinse2022) — likely because the ELISA ANA 8 Pro method targets only 8 specific nuclear antigens, whereas IIF on HEp-2 cells detects all antinuclear reactivities. This methodological difference makes direct comparison of absolute rates unreliable; the 21.7% vs. 3.8% ratio (fold-change) is more informative than the absolute 21.7% figure.
  2. Transience evidence from cited prior data suggests that acute-infection ANA may not persist. Garcia2009 measures ANA at 2 years post-infection — if still elevated, this is more likely to reflect persistent post-dengue immune dysregulation than the transient acute-infection phenomenon described in Berlin2007.
  3. The specificity of ANA (detecting disease-associated antigens like anti-Sm, anti-U1RNP, anti-centromere) during HAV infection is unexpected and supports the hypothesis that infections can induce genuine autoantibodies, not just non-specific nuclear staining.

Limitation: All infection groups are small (n=7–41). The viral group combines HAV, HBV, and HCV — three diseases with very different immunopathology. No dengue cases are included. The ELISA method (8 antigens) is less sensitive than IIF for total ANA but more specific for disease-associated nuclear antigens.

Questions Raised

  • Do dengue-specific post-acute ANA findings (Garcia2009) persist longer than the transient patterns seen during other viral infections? What is the temporal course of dengue-induced ANA?
  • Are the specific ANA antigens elevated in dengue patients the same as those seen in HAV (where all 8 ANA 8 Pro antigens were detected) or a different subset?
  • Why are rickettsial infections (n=7) ANA-negative when other infection types are not? Does this relate to the intracellular nature of rickettsiae, or to the small sample?
  • The 3.8% ANA positivity in controls by ELISA (8-antigen panel) vs. 13.8–16.1% by IIF at 1:80 in population studies — what proportion of healthy-population IIF-positive ANA is directed against specificities outside the 8-antigen panel?

Graph View

Backlinks