Corticosteroids in cardiac surgery: a continuing controversy

Abstract

Cardiac surgery leads to significant improvements in symptoms of cardiac disease and quality of life, but is still associated with a substantial risk of adverse events and postoperative disability. The perioperative systemic inflammatory response syndrome (SIRS) likely plays a role in the development of several of these adverse outcomes. In order to prevent a SIRS, high-dose corticosteroids (‘steroids’) have been routinely used during cardiac surgery for several decades. However, as demonstrated in the systematic review in this thesis, good clinical studies on the effect of steroids on important clinical outcomes were lacking. We therefore performed a clinical trial, the DExamethasone for Cardiac Surgery (DECS) study, in which 4494 cardiac surgical patients were randomised to receive either intraoperative high-dose dexamethasone or placebo. In this study, no statistically significant effect of dexamethasone could be demonstrated on major adverse events at 30 days, although a trend towards benefit was present (relative risk (RR) 0.83; 0.67-1.01). Moreover, multiple effects were found on several important secondary endpoints, of which the overall pulmonary benefit of dexamethasone was probably the most marked effect. This effect was demonstrable as a reduced incidence of respiratory failure, a shorter duration of mechanical ventilation, a reduced ICU and hospital length of stay, and a reduced risk of pneumonia. However, dexamethasone also increased the risk of other adverse outcomes, such as surgical re-interventions for (late) tamponade (RR 1.84; 1.30-2.61). Also, steroid treatment induced marked hyperglycaemia. Furthermore, the often-claimed benefit of dexamethasone of increased postoperative haemodynamic stability could not be demonstrated in a sub study of blood pressure variability in the early post-cardiopulmonary bypass period. In the analysis of the outcomes at 1 year, no effects of dexamethasone past its effects in the immediate postoperative period were demonstrable. However, the early effects of dexamethasone on pulmonary outcomes and length of stay contributed to a € 1,084 per patient cost benefit. In an analysis of age subgroups, younger patients (up to 75 years) had most benefit from dexamethasone treatment, with a reduced incidence of both the primary endpoint and 30 day mortality. However, in patients aged >80 years there was a possible harmful effect. Similar age effects were observed in a recent large study of methylprednisolone in cardiac surgery. We hypothesized that the phenotype of the perioperative inflammatory response may be different between these age groups, which could have resulted in differential effects of steroid treatment. This hypothesis was confirmed in two additional studies into the association between age and the postoperative inflammatory response. In one study in 985 DECS study patients, increasing patient age was associated with a reduced early CRP response, and also with a lower risk of developing a high early CRP response. In a second study in a separate cohort of 25,095 cardiac surgical patients, a strong association was demonstrated between increasing age and a lower incidence of postoperative clinical SIRS. Given the apparent phenotype variability, prediction of an individual’s risk of a deregulated inflammatory response is a crucially important next step towards developing more precise SIRS prophylaxis strategies.