Gas6 evaluation in patients with acute dyspnea due to suspected pulmonary embolism.

BACKGROUND: Gas6 protein is involved in pulmonary embolism (PE) and acute inflammation in animal models.

METHODS: We enrolled 82 consecutive patients with acute dyspnea and suspected PE (Geneva score with high (HCP) or low/intermediate clinical probability (LICP)+D-dimer >or=0.5microg/mL) and 29 age-matched healthy volunteers. According to clinical and instrumental evaluations the following diagnoses were obtained: heart failure (HF), pulmonary or systemic infection (I), PE, or no illness (N). Twenty-two patients were excluded due to oral anticoagulation (9), lack of CT angiography or pulmonary scintigraphy (6), plasma creatinine >or=3mg/dL (3), and pulmonary cancer (4). Plasma Gas6 was measured with a validated enzyme-linked immunoassay. Non-parametric tests and accuracy measures were calculated.

RESULTS: Out of 60 patients included, 8 were N, 12 HF, 11 I and 29 PE. Gas6 median value in the N group (20.4ng/mL, interquartile range 17.6-21.6) matched that of healthy volunteers, 19.1 (17.2-21.4). Median Gas6 values in HF, 26.4 (21.6-33.3) and I groups, 34.1 (30.0-38.7), were significantly higher than those in PE 18.2 (16.3-23.3) or N (Kruskal-Wallis test p<or=0.05) groups. Gas6 test improved PE diagnosis with an area under the curve of 0.80 and 0.91 (in all and LICP patients). A 24ng/mL threshold excluded PE in 33% of LICP patients without loosing any diagnosis.

CONCLUSIONS: The data link Gas6 protein to infection/inflammation, but not to PE, in humans. Gas6 assay was useful in PE diagnosis, improving D-dimer accuracy particularly in LICP patients, and limiting false positives.