Safety and adverse events
Regarding safety and tolerability of the high dose ascorbic acid, median of VC levels of ASTRALI group were significantly higher than control at T48 (p< 0.0001), T96 (p< 0.0001) and T120 (p< 0.0001). At T144, VC levels returned to normal with was no statistically significant difference between the two groups (p= 0.739). (Figure 3) The two only reported adverse effects were hypernatremia and AKI, with similar distributions across groups (p >0.05).
In a recent pharmacokinetic trial, IV 10 gm/d ascorbic acid bolus or continuous infusion showed varying decrease in VC levels in all groups 48 hrs after the end of infusion, and no significant difference in VC level. This varying reduction may be due to differences in all pharmacokinetic parameters. In patients without a history of nephrolithiasis, the clinical relevance of hyperoxaluria and stone formation is negligible. (53)
In phase I safety trial, 24 septic ICU patients were randomized to receive IV ascorbic acid infusions every 6 hours for 4 days as 50 mg/kg/d (n = 8), or 200 mg/kg/d (n = 8), or placebo (n = 8). Patients were monitored for hypernatremia, hypotension, tachycardia, and GI adverse effects. Mean baseline plasma VC levels for all patients were 17.9 ± 2.4 μM (normal 50-70 μM). Ascorbic acid infusion rapidly and significantly raised plasma VC levels with no adverse events. (51) In Zhang et al. trial, daily infusion of 24 gm ascorbic acid for 7 days was not associated with significant incidence of AKI, liver injury, cardiac injury, septic shock, or coagulation disorders. (35) In two other trials, no ascorbic acid-related adverse events were identified. (43, 49)
Reports showed that maintaining plasma VC levels in critically ill patients requires increased frequency of dosing (4 times or continuous infusion) due to impaired distribution volume and increased clearance. There is currently no upper limit of IV dose, although nephrolithiasis at doses > 100 gm was reported. (54)
This study had several limitations. Although most of the primary outcomes were significantly different, the small sample size used may affect the significance of secondary outcomes, which may be more clinically important. Some baseline characteristics were not equally distributed across groups such as diabetes. Because one of our enrollment sites was a hepatology center, 25% of patients were hepatic. These deviations may affect the generalizability of our results to general population of critically ill patients with TRALI. It was not easy to enroll patients with TRALI. Although all possible means were used to confirm the diagnosis of TRALI, we cannot guarantee that there were no other forms of lung injury such as sepsis-induced ALI, ARDS or TACO. TRALI is a very difficult diagnosis. About 21.3% of patients were initially admitted with sepsis. Another 13.8% were trauma patients. Although no patients were enrolled with lung contusions and no radiological evidence of lung injury at the time of enrollment, both conditions are risk factors for ARDS. In this study, we focused on levels of pro-inflammatory and anti-inflammatory markers as primary outcomes. We did not add more secondary outcomes to avoid the problems of multiple testing and finding an outcome by chance. Organ failure assessment was not followed-up, it was not the scope of this trial.