SAN FRANCISCO – Perioperative infusions of sodium bicarbonate failed to reduce the risk of kidney injury in patients undergoing cardiac surgery in a multicenter randomized, double-blind, placebo-controlled trial in 427 patients.
The bicarbonate infusion increased the pH of both blood and urine in the 215-patient treatment group compared with 213 patients in a control group who got sodium chloride (saline) infusions, but 45% of the bicarbonate group and 44% of the placebo group developed kidney injury, a nonsignificant difference, Dr. Shay McGuinness and his associates reported at an international conference of the American Thoracic Society.
Dr. Shay McGuinness
The study excluded patients with end-stage renal disease; patients having emergency cardiac surgery or planned off-pump cardiac surgery; and patients with known blood-borne infectious disease, chronic inflammatory disease, immunosuppression, or chronic moderate- to high-dose corticosteroid use.
"We cannot recommend the use of perioperative infusions of sodium bicarbonate to reduce cardiac surgery–associated kidney injury in these patients, and we do not believe further investigation of this therapy is justified," said Dr. McGuinness, an intensive care specialist at Auckland City Hospital, New Zealand.
The study defined kidney injury as an increase in creatinine of at least 25% from baseline or at least 0.5 mg/dL within the first 5 postoperative days.
The bicarbonate group and placebo group did not differ significantly in mean time on ventilation (21 and 25 hours, respectively), length of stay in the ICU (2 days each), length of stay in the hospital (13 days each), mortality in the ICU (3% and 2%, respectively), or 90-day mortality (4% and 2%).
The infusion strategy had started to catch on in New Zealand and Australia, but it’s unclear if anyone in the United States has pursued it, he said.
The study identified a high-risk group, got a plausible physiological treatment effect, and had good compliance and follow-up rates, but the clinical results were "absolutely negative," Dr. McGuinness said.
The investigators still are analyzing subgroups in the study, but "my gut feeling is that this is a completely negative study. There’s not even a hint of benefit. Walk away from it and find something else to study," he said.
To be in the study, patients having cardiac surgery at three centers in New Zealand and Australia had to have one or more risk factors for associated kidney injury. The rates of risk factors were similar between groups, including age over 70 (a mean of 58% of patients), preexisting renal impairment (14%), left ventricular ejection fraction below 35% (6%), valvular surgery with or without coronary artery surgery (72%), previous cardiac surgery involving sternotomy (16%), or insulin-dependent diabetes mellitus (6%). Measures of baseline renal function were similar between groups.
"What you see is standard cardiosurgical higher-risk patients" in the cohort, Dr. McGuinness said.
Infusions commenced at the start of anesthesia with a 1-mmol/mL solution, followed by 0.5 mmol/kg for 1 hour and 0.2 mmol/kg per hour for 23 hours.
Acid-base status and plasma levels of bicarbonate were similar between groups at baseline, but statistically and clinically significant differences emerged between groups at several time points after the infusion started.
Mean plasma bicarbonate levels in the bicarbonate and placebo groups were 25.72 mmol/L and 25.91 mmol/L at baseline, respectively, 27.03 and 24.35 mmol/L at 6 hours, 29.74 and 23.7 mmol/L at 24 hours, and 29.14 and 25.35 mmol/L at 48 hours.
Mean plasma pH levels in the bicarbonate and placebo groups were 7.40 and 7.41 at baseline, 7.40 and 7.37 at 6 hours, and 7.44 and 7.36 at 24 hours, respectively.
Mean urinary pH measures were 5.8 and 5.5 at baseline, 6.5 and 5.8 at 6 hours, and 7.3 and 5.2 at 24 hours, respectively.
Up to half of the 1 million patients who undergo open heart surgery each year will develop associated kidney injury with increased risk for further morbidity or death. The overall incidence of cardiac surgery–related kidney injury is approximately 5%-10% and probably is increasing, he said.
A previous pilot study by Dr. McGuinness and his associates of 100 patients at a single site had suggested that prophylactic perioperative infusions of sodium bicarbonate might reduce the risk of kidney injury. The investigators conducted the phase II trial before deciding whether or not to pursue a large phase III trial, which will not be happening based on these results.
The study was funded by Fisher & Paykel Healthcare and New Zealand medical organizations. Dr. McGuinness reported having no disclosures.