News
Soluble ST2 predicts ARDS ICU mortality
Major finding: Median soluble ST2 levels were significantly higher among ICU nonsurvivors than survivors at day 1 (4,934 pg/mL vs. 1,007 pg/mL)...
WASHINGTON – Plasma levels of an interleukin-33 receptor that’s involved in inflammation regulation appeared able to discriminate between acute respiratory distress syndrome and acute decompensated heart failure in an analysis with 72 patients.
In a second study, high plasma levels of the same interleukin-33 receptor, known as soluble suppressor of tumorgenicity 2 (sST2), identified acute respiratory distress syndrome (ARDS) patients who were sicker and more responsive to conservative fluid management, Sean D. Levy, MD, said at an international conference of the American Thoracic Society.
Mitchel L. Zoler/Frontline Medical News
Dr. Sean D. Levy
In order to assess the ability of sST2 to reliably distinguish patients with ARDS from those with acute decompensated heart failure, he and his associates selected 72 patients seen at the Massachusetts General Hospital in Boston with an initial diagnosis of acute decompensated heart failure accompanied by bilateral lung infiltrates and acute hypoxemia respiratory failure requiring endotracheal intubation and mechanical ventilation. The investigators measured the sST2 level in a plasma specimen from each patient. In addition, after each patient either left the hospital or died, their case underwent review by two critical care physicians who retrospectively rediagnosed the patients as either having ARDS or acute decompensated heart failure. This divided the cohort into 30 patients with ARDS and 42 with true acute heart failure. The two subgroups matched up fairly closely for most clinical measures and comorbidities, but APACHE III (Acute Physiology and Chronic Health Evaluation III) scores averaged significantly higher in the ARDS patients.
The plasma levels of sST2 showed a dramatic split between the two subgroups. The 30 patients retrospectively diagnosed with ARDS had an average level of 386 ng/mL with an interquartile range of 318-611 ng/mL. The 42 acute decompensated heart failure patients averaged a sST2 level of 148 ng/mL, with an interquartile range of 84-225 ng/mL. The area under the receiver operator curve for discriminating between ARDS and acute heart failure using a cutpoint of 271 mg/mL was 0.86, showing “good” discrimination, Dr. Levy said. This cutpoint had a sensitivity of 83% and specificity of 88% for correctly distinguishing between ARDS and acute heart failure.
In a second analysis, Dr. Levy and his associates looked at the ability of sST2 levels to separate out patients with acute lung injury who had a more robust response to either the conservative or liberal fluid-management strategies tested in the Fluid and Catheter Treatment Trial (FACTT), run by the National Heart, Lung, and Blood Institute’s ARDS Clinical Trials Network. The primary outcome of FACTT was death from any cause 60 days after entry, and this showed no significant difference between conservative (restricted fluids and increased urine output) and liberal (the reverse) fluid management strategies in acute lung injury patients (N Engl J Med. 2006 Jun 15;354[14]:2564-75). From among the 1,001 patients enrolled in FACTT, 826 had specimens available for measuring sST2 (Crit Care Med. 2013 Nov;41[11]:2521-31),
The researchers applied the sST2 cut point they derived in the first analysis to the FACTT cohort and identified 133 (16%) patients with a low sST2 level and 693 (84%) with a high level. The patients with high sST2 were sicker, with significantly higher APACHE III scores, worse acidemia, and worse renal function.
Patients with high sST2 levels had a significant increase in ventilator-free days on conservative fluid management, compared with liberal management, while the two management strategies produced virtually identical results in the patients with low levels of sST2. Patients with high sST2 also had a significantly quicker time to extubation on a conservative strategy compared with the liberal strategy, and again this correlation did not exist among patients with low sST2. However, as in the overall trial a conservative strategy had no discernible impact on 60-day mortality, compared with the liberal strategy, even in the subgroup with high sST2.
Dr. Levy had no disclosures.
On Twitter @mitchelzoler