ORLANDO – Thiamine supplementation improved in-hospital survival for thiamine-deficient patients with septic shock, and reduced progression to dialysis regardless of baseline thiamine status, in a pilot study.
The prospective, randomized, double-blind placebo-controlled trial compared daily 200-mg intravenous doses of thiamine to placebo for 88 patients hospitalized with septic shock. Thiamine was given for 7 consecutive days, or until discharge. On entry into the clinical trial, median lactate was the same in both the thiamine (n = 43) and the placebo arm (n = 45) at 4.1 mmol/L, Dr. Michael Donnino said during a late-breaking session at the Society for Critical Care Medicine’s Critical Care Congress.
At 24 hours, median lactate had dropped to 2.5 mmol/L in the thiamine group and 2.6 mmol/L in the placebo group, a nonsignificant difference. There was no significant difference in shock reversal between groups, nor was there a difference in time to shock or cumulative vasopressor dose.
However, the thiamine group had fewer patients with markedly elevated lactate at 24 hours: 17 (38%) of the placebo group had lactate levels greater than 4 mmol/L, compared with 9 (21%) of the thiamine group. “There were many more ‘lactate outliers’ in the placebo arm, and those virtually disappeared in the thiamine treatment arm,” said Dr. Donnino, director of the center for resuscitation science at Beth Israel Deaconess Medical Center, Boston.
For patients who were thiamine deficient and received thiamine repletion, a preplanned subgroup analysis of these 28 patients (35%) showed significantly lower lactate at 24 hours (median, 2.1 mmol/L vs, 3.1; P = .03). “In this group we saw a lot different physiology,” said Dr. Donnino. “Change was different; median was different; any way you looked at these data, lactate was lower in the thiamine group.”
For the full group of participants, there was no significant between-group difference in overall survival. Of the 15 thiamine-deficient patients who received thiamine, 2 (13%) patients died, while 6 (46%) of the 13 thiamine-deficient patients in the placebo arm died, a nonsignificant difference (P = .10), but there was a significant difference between the survival curves of the two groups (P = .047).
In a post hoc analysis not contained in the concurrently published article , only one of the patients in the thiamine arm (3%) progressed to renal injury requiring dialysis, while eight patients in the placebo arm (23%) needed dialysis (P = .04) (Crit Care Med. 2016 Feb;44[2]:360-7. doi: 10.1097/CCM.0000000000001572).
“I would argue that an often forgotten – but key – component of pyruvate dehydrogenase is thiamine. … Without thiamine, you get refractory hypotension, lactic acidosis, and death,” Dr. Donnino said.
Adult patients were included in the study if they had a suspected or confirmed source of infection, and remained hypotensive (with a systolic blood pressure of less than 90 mm Hg) after a 2-L fluid bolus, with subsequent vasopressor dependence. Vasopressor dependence in the study was defined as continuous infusion of norepinephrine or epinephrine, or of dopamine at 5 mcg/kg per minute or greater.
The final criterion for inclusion was lactate greater than 3 mmol/dL after the initial fluid bolus. “Interestingly, if you change that 3 to 2, this is very similar to the current definition of septic shock,” said Dr. Donnino.
The primary endpoint of the study was median lactate level at 24 hours, with secondary endpoints of lactate change at 24 hours, shock reversal, and mortality. Adverse events and mortality were the safety endpoints.
Baseline characteristics were similar for both study arms, except that the thiamine group had a higher proportion of patients with diabetes than the placebo group (17/43 [40%] vs. 6/45 [13%]).
The study excluded cases in which lactic acidosis might be due to reasons other than septic shock, such as recent seizures, use of medications that can cause lactic acidosis, ischemia, and carbon monoxide poisoning. Patients who had another indication for thiamine administration, such as those in acute alcohol withdrawal, were excluded as well.
Some data analysis was hampered by the high in-hospital mortality rate of 43%. Also, the study involved a target population of critically ill patients, drawn mostly from a single center.
Bringing the discussion to the real world, moderator Dr. Timothy Buchman, director of the critical care center at Emory University, Atlanta, asked, “Today, when you see a patient in sepsis, do you routinely give thiamine?”
Dr. Donnino responded, “When I see someone in septic shock with a persistent lactic acidosis who I consider to be in a risk category, my clinical practice is to give them thiamine now. I’m not necessarily saying that I advocate that to the wider critical care community based on our data, but that’s just based on my experience and what I’ve seen over the years.”