Blog

The HYPOLYTE trial, released in JAMA in March, raises more questions than it answers. The trial, conducted in 7 ICUs over 3 years, randomised 150 patients to "stress dose" steroids or placebo. Patients who were defined as "non-responders" based on baseline or 250mcg synacthen stimulation tests continued the regime for 7 days, while those who had an appropriate response ceased the intervention. The primary endpoint was hospital acquired pneumonia, which was significantly reduced at 28 days (35% vrs 51%, p=0.007). This effect was particularly noted in the patients with severe traumatic brain injury. Mortality was unchanged, but given the low number of patients, and the surprisingly low mortality rate of 8% in the intervention group, the study appears underpowered to detect a difference. A number of interesting questions are raised. Should treatment be based on Synacthen testing, given the results of the Corticus study (the findings of which were released after HYPOLYTE commenced)? Why is pneumonia reduced so significantly in this group when it wasn't in the sepsis trials? Is this simply an "etomidate-protective" effect? Why are 70% of patients in this trial "non-responders"? Why was the death rate so low when compared to similar cohorts? The trial again piques the interest in steroids for conditions associated with Systemic Inflammatory Response Syndrome, but for the time being, seems to raise more questions than it answers.