Plasma chondroitin sulfate predicts the effectiveness of fluid resuscitation strategies in patients with sepsis.

BACKGROUNDPlasma heparan sulfate, a glycosaminoglycan released during endothelial glycocalyx degradation, predicts sepsis mortality. Chondroitin sulfate is a circulating glycosaminoglycan not specific to glycocalyx degradation; its relevance to sepsis is unknown.METHODSWe studied the associations of plasma chondroitin sulfate with (a) mortality in patients with sepsis-associated hypotension and (b) the relative effectiveness of a randomly assigned liberal versus restrictive intravenous fluid resuscitation strategy. We selected 574 patients enrolled in the Crystalloid Liberal or Vasopressors Early Resuscitation in Sepsis trial using an outcome-enriched sampling strategy. We used liquid chromatography-mass spectrometry to quantify plasma chondroitin sulfate. In comparison, we measured hyaluronic acid as a glycocalyx degradation marker and IL-6 as an inflammatory biomarker. We conducted Cox proportional hazards regression analyses to examine associations of baseline biomarker concentrations with mortality and resuscitation strategy effectiveness. We used inverse probability of selection weights and generalized raking to account for the nonrepresentative sampling design.RESULTSPlasma chondroitin sulfate, hyaluronic acid, and IL-6 were associated with mortality within 90 days. As baseline chondroitin sulfate increased, subsequent randomization to a restrictive strategy was increasingly beneficial (P = 0.022): treatment effect hazard ratio (restrictive versus liberal) for mortality was estimated as 1.49 (95% CI, 0.98-2.27), 1.30 (95% CI, 1.00-1.69), 1.09 (95% CI, 0.82-1.44), 0.88 (95% CI, 0.66-1.16), and 0.71 (95% CI, 0.52-0.97) for 10th, 25th, 50th, 75th, and 90th percentiles of baseline chondroitin sulfate.CONCLUSIONPlasma chondroitin sulfate predicts sepsis mortality and may modify the response to a subsequent liberal versus restrictive intravenous fluid resuscitation strategy.TRIAL REGISTRATIONClinicalTrials.gov NCT03434028.FUNDINGNIH grants R01HL149422 and R01HL094786.

Duke Scholars

Author Christopher Lindsell Biostatistics & Bioinformatics, Division of Biostatistics