Antibody sugar patterns separate sepsis from COVID-19 and may flag fatal outcomes in ICU patients

Predicting which critically ill patients will deteriorate or die remains difficult, whether the cause is sepsis or severe COVID-19. Immunoglobulin G (IgG), the most abundant serum antibody, carries sugar chains called N-glycans that shape how strongly it activates immune cells such as natural killer cells, through the FcγRIIIa receptor. Changes in galactose, sialic acid, fucose, bisecting GlcNAc and high-mannose content can push IgG toward pro- or anti-inflammatory behavior, making these patterns candidate prognostic biomarkers.

The researchers analyzed serum from 86 people: 15 healthy volunteers, 19 ICU patients with sepsis, 25 ICU patients who recovered from COVID-19, and 27 who died of it. COVID-19 samples were taken on ICU admission day; sepsis samples were taken between ICU days 1 and 4. IgG was purified, N-glycans were released with PNGase F, labeled with procainamide, and measured by online HILIC solid-phase extraction with HILIC separation, fluorescence detection and mass spectrometry. Groups were compared with Kruskal–Wallis tests and Dunn's post hoc comparisons, with data adjusted for age by linear regression.

Galactosylation was lowest in sepsis and somewhat reduced in fatal COVID-19 after age adjustment, while healthy and recovered COVID-19 groups were higher. It was the only total trait that distinguished fatal from recovered COVID-19. Both COVID-19 groups had lower core fucosylation than healthy and sepsis groups, with sepsis showing a slight rise in fucosylation. Sialylation increased in COVID-19, driven by disialylated glycans, but fell in sepsis. High-mannose glycans (M5 and M6) were slightly higher in COVID-19 and nearly absent in sepsis. Total bisecting glycans did not differ between groups after age correction.

The opposing fucosylation patterns suggest that sepsis and COVID-19 have distinct IgG glyco-signatures despite shared inflammation. Reduced galactosylation and fucosylation could favor FcγRIIIa binding and antibody-dependent cellular cytotoxicity, which may contribute to hyperinflammation. Even so, the authors judge IgG glycosylation a contributor rather than a dominant driver of inflammation. The rise in sialylation in COVID-19 was counterintuitive given its usual anti-inflammatory role.

The main caveats are small group sizes and large age differences: the healthy group had a median age of 30 versus 61–70 in the patient groups. Age affects IgG glycosylation and age correction can itself distort results; some findings changed with or without correction. Sample timing differed for sepsis, and the study is cross-sectional without validation. Larger, age-matched cohorts are needed before any clinical use.