Saito2004 - PAIgG and PAIgM in Secondary Dengue

Full citation: Saito M, Oishi K, Inoue S, Dimaano EM, Alera MTP, Robles AMP, Estrella BD Jr, Kumatori A, Moji K, Alonzo MT, Buerano CC, Matias RR, Morita K, Natividad FF, Nagatake T. Association of increased platelet-associated immunoglobulins with thrombocytopenia and the severity of disease in secondary dengue virus infections. Clinical and Experimental Immunology, 2004; 138:299–303.

Raw file: [[raw/saito2004.pdf]]

Summary

This prospective hospital-based study from San Lazaro Hospital and St Luke’s Medical Center, Manila, Philippines (September 2002–November 2003) examined the roles of platelet-associated IgG (PAIgG) and IgM (PAIgM) in thrombocytopenia and disease severity in secondary dengue infection. The study builds directly on Oishi2003 - PAIgG and Thrombocytopenia in Secondary Dengue — corresponding author K. Oishi is shared, and the 2004 study explicitly extends the earlier PAIgG finding to include PAIgM. Of 135 enrolled patients, 78 with confirmed acute-phase secondary dengue (HI titer ≥1:2,560) completed paired acute and convalescent platelet-associated immunoglobulin measurements.

Both PAIgG and PAIgM were significantly elevated in secondary dengue patients versus healthy controls and inversely correlated with platelet count. Both isotypes fell significantly from acute to convalescent phase, mirroring platelet count recovery. Anti-dengue virus IgG and IgM activity was confirmed in platelet eluates from 10 secondary-infection patients, establishing that neither PAIgG nor PAIgM represents an anti-self (autoantibody) response — both carry anti-dengue virus specificity and form on platelets via immune complex deposition.

Crucially, PAIgM was the only immunoglobulin independently associated with DHF development by multivariate logistic regression. A cut-off of >20 ng/10⁷ platelets predicts DHF with 92.1% specificity (48.6% sensitivity). The authors propose that PAIgM may contribute to the vascular permeability characteristic of DHF via activated platelets adhering to dengue-infected endothelial cells, acting through complement rather than Fc receptor pathways.

Study Design

• Type: Prospective hospital-based cohort
• Sample size: 78 secondary dengue patients (40 DF, 38 DHF) + 43 age-matched healthy volunteers
• Setting: San Lazaro Hospital (Blood Borne Diseases Ward, Manila) and St Luke’s Medical Center (Quezon City, Research and Biotechnology Division), Philippines; September 2002 – November 2003
• Population: Paediatric and young adult (mean age 18.2 ± 5.9 years); secondary infection confirmed by HI titer ≥1:2,560; acute phase 3–7 days from symptom onset; all DHF cases were grades I–II (no shock)

Key Findings

• PAIgG and PAIgM both significantly elevated in secondary dengue vs. healthy volunteers: PAIgG 30.9 ± 23.1 vs. 6.4 ± 4.2 ng/10⁷ platelets (P < 0.001); PAIgM 17.5 ± 20.4 vs. 4.2 ± 3.8 ng/10⁷ (P < 0.001)
• Inverse correlation with platelet count: PAIgG r = −0.256, P = 0.023; PAIgM r = −0.231, P = 0.046 (n = 78 and n = 75 respectively)
• Parallel acute-to-convalescent recovery: platelet count 47.9 → 259.5 × 10³/μl (P < 0.001); PAIgG 30.9 → 13.3 ng/10⁷ (P < 0.001); PAIgM 17.5 → 9.7 ng/10⁷ (P < 0.001)
• Anti-dengue virus activity confirmed in platelet eluates (OD 1.54 ± 0.35 anti-dengue IgG; 0.35 ± 0.20 anti-dengue IgM) vs. healthy volunteer eluates (OD 0.20 ± 0.10 and 0.09 ± 0.05) — both PAIgG and PAIgM are immune complexes carrying anti-dengue specificity, not anti-self antibodies
• Higher in DHF than DF: PAIgG 37.1 ± 26.0 vs. 25.0 ± 18.4 ng/10⁷ (P < 0.01); PAIgM 24.8 ± 24.5 vs. 10.3 ± 11.9 ng/10⁷ (P < 0.001)
• PAIgM independently predicts DHF by logistic regression (P < 0.01); cut-off >20 ng/10⁷ platelets: sensitivity 48.6% (18/37), specificity 92.1% (35/38)
• PAIgM mechanism bypasses Fc receptors: IgM pentamer structure precludes Fc receptor engagement; platelet clearance is proposed to proceed via complement receptor- and/or complement-mediated lysis — the same pathways as PAIgG but without any FcγR involvement
• PAIgG is the dominant contributor to thrombocytopenia overall: only 19.2% of patients had normal PAIgG levels, vs. 50.7% with normal PAIgM — confirming that while both contribute, PAIgG is the primary pathway
• Dengue virus RNA detected by RT-PCR in 42.8% of purified platelet samples (21 secondary-infection patients), supporting direct dengue-platelet localisation as the substrate for immune complex formation

Methods Used

• IgM-IgG Serology ELISA (IgM-capture ELISA for dengue diagnosis; Bundo & Igarashi 1985 protocol)
• RT-PCR (Morita et al. 1991 protocol; dengue virus serotyping and platelet-associated viral RNA detection)
• Hemagglutination Inhibition Test (Clarke & Casals 1958; DENV-1, -2, -3, -4 acetone-extracted antigens; HI ≥1:2,560 = secondary infection classification)
• Platelet-Associated Immunoglobulin ELISA (competitive ELISA for PAIgG and PAIgM; wells coated with standard human IgG or IgM; platelet sample competes with coated standard for HRP-conjugated anti-human IgG or IgM Fc antibody; read at OD 492 nm; reported as ng/10⁷ platelets)

Entities Mentioned

• FcγRIIa Receptor (PAIgM bypasses Fc receptors via IgM pentamer structure; contrast with PAIgG which operates through both Fc receptor- and complement-mediated macrophage clearance and complement-mediated platelet lysis)

Concepts Addressed

• Secondary Dengue Infection (central context; all 78 patients secondary-confirmed; extends Oishi2003 PAIgG mechanism to include PAIgM isotype)
• Dengue Pathophysiology (dual PAIgG/PAIgM thrombocytopenia mechanism; proposed link of PAIgM-coated activated platelets to vascular permeability via endothelial cell adhesion)
• Autoimmunity in Dengue (platelet-associated IgM in secondary infection is anti-dengue virus specificity, NOT anti-platelet autoantibody — mechanistically distinct from primary-infection IgM; no autoimmune component in secondary PAIgM)
• Dengue Clinical Classification (WHO 1997 DHF criteria: platelet nadir <100,000/μl, haemorrhagic manifestations, haematocrit increase ≥20% or pleural effusion/ascites; DF defined by haematocrit increase <20% and no pleural effusion)

Relevance & Notes

Saito2004 is the direct follow-on to Oishi2003 - PAIgG and Thrombocytopenia in Secondary Dengue from the same Nagasaki University/Manila collaboration. Together the two papers establish a complete two-isotype picture: in secondary dengue, both PAIgG and PAIgM are elevated, both carry anti-dengue virus specificity (confirmed by platelet eluate), and both inversely correlate with platelet count. The PAIgG mechanism is established as quantitatively dominant (50.7% of patients have normal PAIgM vs. 19.2% normal PAIgG).

The most important contribution relative to existing wiki content is the clarification that the IgM elevated on platelets in secondary infection is anti-dengue virus IgM immune complex — not anti-platelet autoantibody IgM (the Lin2001 primary-infection mechanism). This requires refining the thrombocytopenia bifurcation model:

• Primary infection: IgM anti-platelet autoantibody (NS1 molecular mimicry; Lin2001)
• Secondary infection: PAIgG (anti-dengue IgG immune complex; Oishi2003) + PAIgM (anti-dengue IgM immune complex; Saito2004); both FcγRII-independent at platelet docking step; PAIgM additionally FcγR-independent at clearance step

The PAIgM-DHF specificity (92.1%) is notable as a potential bedside predictor but sensitivity is too low (48.6%) for standalone screening. In a secondary-infection patient with virological confirmation, a PAIgM >20 ng/10⁷ would strongly suggest impending DHF.

Limitations: All DHF cases grades I–II (no shock); extrapolation to DSS uncertain. The PAIgM–haematocrit correlation (r = 0.23, P = 0.045 overall) disappears within DF and DHF subgroups, raising doubt about whether PAIgM directly drives vascular permeability or is merely a severity marker.

Questions Raised

• Does FcγRIIa genotype differentially influence PAIgG vs. PAIgM-mediated platelet clearance? (FcγRIIA polymorphism modulates IgG clearance; IgM clearance would rely on complement receptor 1 [CR1] polymorphisms instead)
• Can PAIgM >20 ng/10⁷ be validated as a DHF predictor in an independent cohort, particularly including DSS grades III–IV?
• Are both PAIgG and PAIgM simultaneously present in most secondary-infection patients, or do they represent distinct immunological subsets?
• If dengue virus RNA is in 42.8% of platelet samples, does direct viral platelet infection contribute to platelet activation and adhesion to endothelium — independent of immune complex deposition?