[ Original Research Critical Care ]

Vitamin C Pharmacokinetics in
Critically Ill Patients
A Randomized Trial of Four IV Regimens
Harm-Jan de Grooth, MD; Wai-Ping Manubulu-Choo, PharmD; Anthe S. Zandvliet, PharmD, PhD;
Angélique M. E. Spoelstra - de Man, MD, PhD; Armand R. Girbes, MD, PhD; Eleonora L. Swart, PharmD, PhD;
and Heleen M. Oudemans-van Straaten, MD, PhD

BACKGROUND: Early high-dose IV vitamin C is being investigated as adjuvant therapy in

patients who are critically ill, but the optimal dose and infusion method are unclear. The
primary aim of this study was to describe the dose-plasma concentration relationship and
safety of four different dosing regimens.
METHODS: This was a four-group randomized pharmacokinetic trial. Patients who were
critically ill with multiple organ dysfunction were randomized to receive 2 or 10 g/d vitamin
C as a twice daily bolus infusion or continuous infusion for 48 h. End points were plasma
vitamin C concentrations during 96 h, 12-h urine excretion of vitamin C, and oxalate
excretion and base excess. A population pharmacokinetic model was developed using
NONMEM.
RESULTS: Twenty patients were included. A two-compartment pharmacokinetic model with
creatinine clearance and weight as independent covariates described all four regimens best.
With 2 g/d bolus, plasma vitamin C concentrations at 1 h were 29 to 50 mg/L and trough
concentrations were 5.6 to 16 mg/L. With 2 g/d continuous, steady-state concentrations were
7 to 37 mg/L at 48 h. With 10 g/d bolus, 1-h concentrations were 186 to 244 mg/L and trough
concentrations were 14 to 55 mg/L. With 10 g/d continuous, steady-state concentrations were
40 to 295 mg/L at 48 h. Oxalate excretion and base excess were increased in the 10 g/d dose.
Forty-eight hours after discontinuation, plasma concentrations declined to hypovitaminosis
levels in 15% of patients.
CONCLUSIONS: The 2 g/d dose was associated with normal plasma concentrations, and the 10
g/d dose was associated with supranormal plasma concentrations, increased oxalate excre-
tion, and metabolic alkalosis. Sustained therapy is needed to prevent hypovitaminosis.
TRIAL REGISTRY: ClinicalTrials.gov; No.: NCT02455180; URL: www.clinicaltrials.gov
CHEST 2018; 153(6):1368-1377

KEY WORDS: ascorbate; ascorbic acid; clinical trial; intensive care; pharmacology

ABBREVIATIONS: DBE = change of the arterial blood base excess from
baseline; IQR = interquartile range
AFFILIATIONS: From the Departments of Intensive Care (Drs de
Grooth, Spoelstra - de Man, Girbes, and Oudemans-van Straaten) and
Anesthesiology (Dr de Grooth), VU University Medical Center,
Amsterdam; the Department of Clinical Pharmacology and Pharmacy
(Drs Manubulu-Choo, Zandvliet, and Swart), VU University Medical
Center, Amsterdam; and the Department of Pharmacy (Dr Manubulu-
Choo), Westfriesgasthuis, Hoorn, The Netherlands.
Drs de Grooth and Manubulu-Choo contributed equally to this
manuscript.
Part of the results from this article has been presented in abstract form
(de Grooth HJ, Choo WP, Spoelstra - de Man AM, Swart EL,
Oudemans-van Straaten HM. Intensive Care Med Exp. 2016;4(suppl
1):A52).
FUNDING/SUPPORT: This study was performed on departmental
funding.

1368 Original Research [ 153#6 CHEST JUNE 2018 ]

Early high-dose IV vitamin C is under growing interest
as potential adjuvant therapy in critically ill patients
with sepsis, trauma, burns, or ischemia/reperfusion
injury, but pharmacokinetic data in these patients are
surprisingly scarce.
Vitamin C (ascorbic acid) is a water-soluble molecule
with pleiotropic functions. It is an essential antioxidant
and has anti-inflammatory and immune-supporting
properties.1 Vitamin C is a cofactor or cosubstrate in the
biosynthesis of collagen, catecholamines, vasopressin,
and other peptide hormones2 and is essential in the
stress response by promoting the synthesis of cortisol
and norepinephrine and restoring receptor sensitivity.3,4
Critically ill patients have low plasma vitamin C because
of decreased intake and absorption, acutely increased
metabolism, and redistribution.5-7 Patients with sepsis
may have scurvy-level vitamin C depletion.8,9
Some clinical trials have shown beneficial effects of
vitamin C in patients with sepsis, after major surgery or
trauma, but not all studies were positive.10 Variations in
timing, dose, and administration route may play a role
because direct radical scavenging depends on plasma
concentrations > 175 mg/L (1,000 mmol/L).11 Oral
regimens cannot increase plasma concentrations to
Patients, Methods, and Materials
Patients
This randomized factorial pharmacokinetic trial was performed between
March 2015 and November 2016 in the mixed medical/surgical ICU of
VU University Medical Center in Amsterdam, The Netherlands. The
analysts determining the vitamin C and oxalate concentrations were
blinded to treatment allocation. The trial protocol was registered at
ClinicalTrials.gov before patient inclusion (identifier NCT02455180).
The study was approved by the medical ethics committee of VU
University Medical Center (registration NL50578.029.15, decision
2014.539). All participants or their legal representative provided
written informed consent before inclusion.
Eligible for inclusion were patients admitted with severe sepsis or after
major surgery or trauma with a Sequential Organ Failure Assessment
score > 6 (excluding neurologic score to avoid sedation bias) and an
expected ICU stay > 96 h.21
Exclusion criteria were use of vitamin C supplements in the week
before inclusion, major active bleeding, preexistent renal insufficiency
(estimated glomerular filtration rate < 30 mL/min/1.73 m2),
expected need for renal replacement therapy within 48 h, known
glucose 6-phosphate dehydrogenase deficiency, history of urolithiasis
or oxalate nephropathy, and hemochromatosis.
CORRESPONDENCE TO: Harm-Jan de Grooth, MD, VU University
Medical Center, Department of Intensive Care, De Boelelaan 1117,
1081 HV Amsterdam, The Netherlands; e-mail: h.degrooth@vumc.nl
Copyright Ó 2018 American College of Chest Physicians. Published by
Elsevier Inc. All rights reserved.
DOI: https://doi.org/10.1016/j.chest.2018.02.025
chestjournal.org
normal (let alone supranormal) levels because
transported-mediated enteral uptake is rate-limited and
possibly impaired in critical illness.6,11-13
Recently, three small studies using pharmacologic
doses have rekindled interest in vitamin C during
critical illness. In these studies, patients with sepsis
receiving IV vitamin C (50-200 mg/kg/d) showed a
dose-dependent reduction in organ failure14,15 and
28-day mortality.15,16 Along with a body of preclinical
research, these studies mandate further investigation
into the clinical benefit and safety of pharmacologic
doses of IV vitamin C.17 Oxidized vitamin C can be
converted into oxalic acid in concentration-dependent
enzymatic reactions,18,19 but up to now, data on the
effect of short-term IV regimens on oxalate excretion
have not been available.
Despite growing interest, pharmacokinetic data on IV
vitamin C in patients who are critically ill are scarce and
unsystematic.20 Therefore, the primary objective of this
study was to determine the plasma concentration profile
over time of four different dosing regimens of IV
vitamin C in critically ill patients. Secondary objectives
were to study associated oxalate excretion and acid-base
balance as safety parameters.
Study Procedures
Patients were randomized using a computer-generated randomization
table available to a blinded agent. The investigational treatment
consisted of IV vitamin C for 48 h at a dose of 2 g/d in twice daily
bolus infusions of 1 g (2 g/d bolus group) or as continuous infusion
(2 g/d continuous group), or a dose of 10 g/d in twice daily bolus
infusions of 5 g (10 g/d bolus group) or as continuous infusion (10
g/d continuous group). Bolus infusions were administered over 15 min.
Vitamin C (ascorbic acid 100 mg/mL; Centrafarm BV) was diluted
with NaCl 0.9% to 50 mL in opaque light-protected syringes.
Plasma vitamin C samples were taken at baseline and predefined
intervals up to 96 h after the start of the study infusion. Twelve-
hour urine samples for vitamin C, oxalate, and creatinine
concentrations were collected during the first 12 h and during the
last 12 h of therapy. Arterial blood gas samples and hemodynamic
and organ function parameters were collected at baseline, after 4 h,
and at 12-h intervals. The sample handling and laboratory
procedures are described in e-Appendix 1.
All patients received standard concomitant treatment including enteral
nutrition initiated after hemodynamic stabilization (e-Appendix 1).
Reference Values and Statistical Analyses
Descriptive statistics are reported as mean with SD or median and
interquartile range (IQR) where appropriate. Means and SEs are
reported in the figures to facilitate between-group comparisons.
Vitamin C concentrations are reported in milligrams per liter for
comparison with the administered dose (multiply by 5.7 to convert
mg/L to mmol/L).
1369
We defined hypovitaminosis C as < 3.7 mg/L (21 mmol/L) and the upper
limit of normal as 17.6 mg/L (100 mmol/L).6,22 Increased oxalate excretion
was defined as > 90 mg/d (45 mg every 12 h), the value associated
with increased risk of stone formation in primary hyperoxaluria.23,24
The primary outcomes were the attained plasma concentrations over a
96-h period and the resulting pharmacokinetic parameter estimates.
For population pharmacokinetic analysis, total vitamin C
concentrations (ascorbic acid þ dehydroascorbic acid) were used.
Pharmacokinetic modeling was based on the baseline corrected
concentration, ie, supplemental vitamin C above the baseline plasma
concentrations. NONMEM version 7.2.0 (Icon plc) was used for
population pharmacokinetic analysis.
Details of the pharmacokinetic modeling and other statistical methods
are provided in e-Appendix 1.

Results at baseline. The 2 g/d dose resulted in concentrations in

Twenty patients (five per dosage regimen) were included
and randomized (Fig 1). Baseline characteristics are
shown in Table 1. All patients received the allocated
study medications. No patients died during the 96-h
study period.
Vitamin C Pharmacokinetics
Measured vitamin C plasma concentrations are shown
in Figure 2. Ten patients (50%) were vitamin C deficient
the normal range, with 1-h concentrations of 29 to
50 mg/L and trough concentrations of 5.6 to 16 mg/L in
the bolus group and steady-state concentrations of 7 to
37 mg/L at 48 h in the continuous group. Infusions of 10
g/d resulted in supranormal concentrations, with 1-h
concentrations of 186 to 244 mg/L and trough
concentrations of 14 to 55 mg/L in the bolus group and
steady-state concentrations of 40 to 295 mg/L at 48 h in
the continuous group.

Enrollment Assessed for eligibility (n = 50)

Excluded (n = 30)
♦ Did not meet inclusion criteria (n = 20)
♦ Declined to participate (n = 10)

Allocation Randomized (n = 20)

Allocated to 2 g/d bolus Allocated to 2 g/d cont. Allocated to 10 g/d bolus Allocated to 10 g/d cont.
(n = 5) (n = 5) (n = 5) (n = 5)
♦ Received allocated ♦ Received allocated ♦ Received allocated ♦ Received allocated
intervention (n = 5) intervention (n = 5) intervention (n = 5) intervention (n = 5)

Follow-Up

Lost to follow-up (n = 0) Lost to follow-up (n = 0) Lost to follow-up (n = 0) Lost to follow-up (n = 0)

Discontinued intervention Discontinued intervention Discontinued intervention Discontinued intervention

(n = 0) (n = 0) (n = 0) (n = 0)

Analysis

Analyzed (n = 5) a Analyzed (n = 5) Analyzed (n = 5) Analyzed (n = 5) a

Figure 1 – Flow diagram of recruitment, allocation, and follow-up. aOne patient with severe rhabdomyolysis because of limb ischemia (randomized to 2
g/d bolus) rapidly progressed to anuric renal failure, and another patient (randomized to 10 g/d cont.) developed anuric renal failure after 36 h. These
patients were included in the primary outcome (plasma pharmacokinetic) analyses but were excluded from the urine vitamin C and oxalate analyses.
cont. ¼ continuous.

1370 Original Research [ 153#6 CHEST JUNE 2018 ]

TABLE 1 ] Baseline Characteristics of Included Patients and Baseline Vitamin C Plasma Concentrations
2 g/d Bolus 2 g/d Continuous 10 g/d Bolus 10 g/d Continuous
Characteristic (n ¼ 5) (n ¼ 5) (n ¼ 5) (n ¼ 5)
Male/female 4/1 2/3 3/2 4/1
Age, y 64  17 61  19 69  9 60  16
Weight, kg 90  22 86  28 70  15 79  15
Chronic comorbidities
Diabetes mellitus 0 0 0 0
Chronic kidney insufficiencya 1 0 2 0
Liver cirrhosis 0 0 0 0
COPD 0 0 0 0
Cardiovascular disease 1 0 2 1
Cancer 0 1 0 1
Hematologic malignancy 0 2 1 1
Immunocompromised 0 2 1 0
Admission diagnoses
Pneumonia 1 4 4 2
Acute abdomen 0 0 1 1
Soft tissue infection 0 0 1 0
Systemic infection 1 3 2 2
Major surgery 4 1 3 2
b
Acute kidney injury 1 0 0 0
Cerebrovascular incident 1 1 0 0
Clinical characteristics
SOFA score (excluding neurological component)c 8.8  2.2 8.0  1.0 7.0  1.2 10.0  3.0
APACHE II score 26.0  2.9 23  3.4 21.8  3.6 25.2  4.2
Heart rate, bpm 98  15 96  17 94  24 93  18
Mean arterial blood pressure, mm Hg 84  7.5 76  12 94  19 90  18
Serum creatinine, mg/dL 1.10  0.47 0.81  0.43 0.91  0.49 1.10  0.46
Measured creatinine clearance 0-12 h, mL/min 85  62 114  56 96  38 94  45
Arterial pH 7.434  0.032 7.426  0.08 7.404  0.036 7.374  0.070
Arterial base excess, mmol/L 3.1  1.2 1.2  5.0 3.0  3.7 1.1  3.9
Arterial lactate, mmol/L 1.9  0.6 1.5  1.2 1.4  0.30 1.5  0.34
Patients on norepinephrine 3 4 3 2
Norepinephrine dose, mg/kg/min 0.17  0.25 0.17  0.16 0.03  0.04 0.05  0.10
Patients on dopamine 1 0 0 0
Patients on enoximone 1 0 0 0
Baseline plasma vitamin C
Baseline plasma vitamin C, mg/Ld 8.5  4.2 4.3  1.6 4.9  3.0 3.4  1.3
Vitamin C hypovitaminosis C (< 3.7 mg/L) 1 3 2 4
Vitamin C deficiency (< 1.9 mg/L) 0 1 1 0

Values are expressed as counts or mean  SD. APACHE ¼ Acute Physiology and Chronic Health Evaluation; bpm ¼ beats per minute; SOFA, Sequential
Organ Failure Assessment.
a
Chronic kidney insufficiency was defined as a preexisting estimated glomerular filtration rate between 30 and 60 mL/min/1.73 m2 (patients with a
preexisting estimated glomerular filtration rate < 30 mL/min/1.73 m2 were ineligible for inclusion).
b
Two patients developed acute kidney injury during the trial, but only one patient had acute kidney injury criteria at baseline.
c
SOFA score was calculated without the neurologic component to exclude sedation effects.
d
For conversion to micromole per liter, multiply by 5.7.

chestjournal.org 1371
A B
2 g/d bolus group 2 g/d continuous group
Plasma vitamin C (mg/L)

Plasma vitamin C (mg/L)

60 Trough concentrations 60 Infusion
300 300

40 40

µmol/L

µmol/L
200 200

20 100 20 100

0 0 0 0
0 1 2 4 12 24 36 48 72 96 0 1 2 4 12 24 36 48 72 96
Time (h) Time (h)

C D
10 g/d bolus group 10 g/d continuous group

Plasma vitamin C (mg/L)

Plasma vitamin C (mg/L)

300 300
Trough concentrations Infusion
1,500 1,500

200 200

µmol/L
µmol/L
1,000 1,000

100 500 100 500

0 0 0 0
0 1 2 4 12 24 36 48 72 96 0 1 2 4 12 24 36 48 72 96
Time (h) Time (h)

Figure 2 – A-D, Vitamin C plasma concentrations by dosage regimen (five patients per regimen): (A) 2 g/d bolus group, (B) 2 g/d continuous group, (C)
10 g/d bolus group, and (D) 10 g/d continuous group. Arrows represent infusions (four times bolus or 48-h continuous). The dashed lines repre-
sent the lower and upper limit of normal vitamin C plasma concentrations at 3.7 mg/L (21 mmol/L) to 17.6 mg/L (100 mmol/L). Plasma vitamin C was
not sampled at 36 h in the continuous groups.

The dose-concentration relationship was linear. A different between the patients treated with 10 g/d
two-compartment pharmacokinetic model best vs those treated with 2 g/d: median of 10.3 mg/L (IQR,
described all four dose regimens. Creatinine clearance 8.2-15.0 mg/L) vs 8.7 mg/L (IQR, 3.8-10.7 mg/L),
(based on 12-h urine collection) and body weight were respectively (Mann-Whitney P ¼ .423).
determinants of total vitamin C clearance, and body
weight was a determinant of creatinine clearance, based Three patients had plasma concentrations that returned
on 12-h urine collection and the volume of distribution to below the hypovitaminosis threshold at 96 h (two
(V1 and V2). Total vitamin C clearance was constant patients from the 2 g/d dose group and one patient from
over time and independent of dose and infusion method, the 10 g/d dose group). The plasma concentrations were
indicating first-order linear kinetics. The estimates of the 3.1, 2.4, and 0.68 mg/L, and these patients had 36- to 48-h
pharmacokinetic model parameters are shown in creatinine clearances of 140, 197, and 128 mL/min,
Table 2. Figure 3 shows the predicted and actual respectively, indicating that all three patients had
supplemental vitamin C plasma concentrations (change moderate to severe glomerular hyperfiltration.
from baseline) based on the pharmacokinetic model.
Measured urine vitamin C excretion and renal vitamin C
The best-fit model underestimated the 1-h
clearance are described in e-Figures 1 and 2. In short,
concentrations after short-term infusion, which suggests
mean urinary vitamin C excretion was higher in the 10
that initial fast distribution to peripheral compartments
g/d group (bolus: mean, 3,385  1,287 mg at 0-12 h and
is not optimally described by a two-compartmental
6,295  1,286 mg at 36-48 h; continuous: mean, 1,179 
model. A three-compartment model was however not
498 mg at 0-12 h and 4,268  2,688 mg at 36-48 h) than
supported by the data because of the limited sample size
in the 2 g/d group (bolus: mean, 386  288 mg at 0-12 h
and because 1-h concentrations were only measured
and 751  439 mg at 36-48 h; continuous: mean, 73 
after the first dose.
110 mg at 0-12 h and 497  283 mg at 36-48 h), with
At 96 h, 48 h after stopping the study medication, bolus compared with continuous infusion and at 36 to
plasma vitamin C concentrations were not significantly 48 h compared with 0 to 12 h. Renal vitamin C clearance

1372 Original Research [ 153#6 CHEST JUNE 2018 ]

TABLE 2 ] Estimated Population Pharmacokinetic The change of the arterial blood base excess from
Parameters baseline (DBE) over the 48-h treatment period is shown
Total in Figure 4C. The slope of the DBE from 0 to 48 h was
Population RSE
associated with the 10 g/d dose compared with the 2 g/d
Parameter (N ¼ 20) (%)
dose (b ¼ 0.039 mmol/h; 95% CI, 0.014-0.064 mmol/h;
Estimated parameters (theta)
P ¼ .0025), but not with bolus compared with
CLvitC, L/h 4.27 9.8
continuous infusion (b ¼ 0.016; 95% CI, 0.009 to
V1, L 31.6 13.3
0.041; P ¼ .210). At 48 h, there were no significant
V2, L 39.6 21.1
differences in DBE between the four treatment groups
Q, L/h 5.21 18.9
(P ¼ .554), or between the dosages (P ¼ .229), or
Interindividual variability (omega) between the infusion methods (P ¼ .625).
CLvitC, % 27.8 40.3
V1 and V2, % 53.6 62.7
Plots of (changes in) heart rate, norepinephrine support,
serum creatinine, and urine output are available as
Covariate effects
e-Figures 4-7.
CLvitC w weighta 1 (fixed) .
V1 and V2 w weightb,c 1 (fixed) .
a
Discussion
CLvitC w CLcr 0.446 55.6
Residual error Dose-Plasma Concentration Relationship
Proportional residual 25.5 20.1 In this four-group randomized pharmacokinetic trial, 20
error (SD), %
patients with multiple organ failure were randomized to
Additional residual error (SD) 0.496 33.3
receive either 2 or 10 g/d vitamin C administered as a
Pharmacokinetic parameters are shown as estimated population phar- twice daily 15-min bolus infusion or continuous
macokinetic parameters. CLcr ¼ creatinine clearance, based on 12-h urine infusion.
collection; CLvitC ¼ total vitamin C clearance; Q ¼ intercompartmental
clearance; RSE ¼ residual squared error; V1 ¼ central volume of distri- Two grams per day resulted in normal plasma
bution; V2 ¼ peripheral volume of distribution.
a
CLvitC ¼ 4.27  (CLcr/100)0.446  (weight/70)1. concentrations, and 10 g/d resulted in supranormal
b
V1 ¼ 31.6  (weight/70)1. concentrations. Bolus infusion caused higher 1-h
c
V2 ¼ 39.6  (weight/70)1. concentrations but also caused trough concentrations
that approached the hypovitaminosis cutoff in the 2 g/d
was significantly associated with renal creatinine dose group. Plasma concentrations in the 10 g/d dose
clearance in the 10 g/d group, but not in the 2 g/d group. group were in the supranormal range with bolus 1-h
concentrations > 175 mg/L (1,000 mmol/L), providing
Safety optimal concentrations for fast cellular uptake12,25 and
Figure 4A shows total oxalate excretion and Figure 4B increased radical scavenging.11
shows oxalate concentration in 12-h urine samples Our study confirms previous findings that IV doses of 2
collected between 0 and 12 h and between 36 and 48 h. to 3 g/d are required to normalize vitamin C plasma
In a multivariate linear model (R2 ¼ 0.36), 12-h oxalate concentrations in patients who are critically ill.6
excretion was independently higher in the 10 g/d dose However, optimal plasma concentrations during
compared with 2 g/d (b ¼ 18.2 mg; 95% CI, 2.65- overwhelming oxidative stress are not known, nor is
33.7 mg; P ¼ .023), and in the last (36-48 h) sampling whether peaks are more effective than lower but stable
period compared with the first (0-12 h) (b ¼ 30.2 mg; plasma concentrations. Two recent studies using
95% CI, 14.6-45.7 mg; P ¼ .0004), but not in bolus 100 mg/kg/d and 6 g/d as intermittent boluses in
compared with continuous infusion (b ¼ 9.45 mg; patients with sepsis reported a reduction in mortality,
95% CI, 6.1 to 25.0 mg; P ¼ .224). Using a separate but these findings and the dose-effect relationship need
model (R2 ¼ 0.18), urine oxalate concentration was confirmation in larger randomized controlled trials.15,16
found to be a linear function of the natural logarithm of
urine vitamin C concentration (b ¼ 48.0 mmol; 95% CI, High Plasma Concentrations Require Sustained
15.8-80.2 mmol; P ¼ .004) (details in e-Fig 3 in Therapy
e-Appendix 1). Mean oxalate excretion relative to We found a varying decline in plasma concentrations
vitamin C dose (mole/mole) varied between 0.3% and across all groups 48 h after the end of therapy (96 h),
2.3% (e-Appendix 1). with no significant difference in plasma vitamin C

chestjournal.org 1373
A B
2 g/d bolus group 2 g/d continuous group
change from baseline (mg/L)

change from baseline (mg/L)

60 60
300 300
Plasma vitamin C,

Plasma vitamin C,
40 40
200 200

µmol/L

µmol/L
20 100 20 100

0 0 0 0

0 12 24 36 48 72 96 0 12 24 36 48 72 96
Time (h) Time (h)

C 10 g/d bolus group

D 10 g/d continuous group
change from baseline (mg/L)

change from baseline (mg/L)

300 300
1,500 1,500
Plasma vitamin C,

Plasma vitamin C,
200 200

µmol/L

µmol/L
1,000 1,000

100 500 100 500

0 0 0 0

0 12 24 36 48 72 96 0 12 24 36 48 72 96
Time (h) Time (h)

Figure 3 – A-D, Predicted changes in vitamin C plasma concentrations from baseline by dosage regimen. The solid lines represent the predicted
changes in plasma concentrations from baseline for a hypothetical 70-kg patient with a creatinine clearance of 100 mL/min when administered 2 g/d
(A and B) or 10 g/d (C and D) vitamin C as twice daily bolus infusions (A and C) or 48-h continuous infusion (B and D). The points represent
the observed changes from baseline plasma concentration in the 20 included patients (five patients per dosage regimen). The measurements at 12, 24,
36, and 48 h in the bolus groups are trough concentrations.

A B C
0.8
Base excess change from baseline
12-h urine oxalate excretion (mg)

12-h urine oxalate concentration

80 4
0.6
60
(mmol/L)

(mmol/L)

2
0.4
40

0.2 0
20

0 0 –2
0-12-h 36-48-h 0-12-h 36-48-h 0 4 12 24 48
sample sample sample sample Time (h)

Infusion Dose
bolus 10 g/d
continuous 2 g/d

Figure 4 – A-C, Safety end points by intervention group; data points represent mean and SE. A, Urine 12-h oxalate excretion. Twelve-hour oxalate
excretion was independently higher in the 10 g/d dose compared with the 2 g/d dose (b ¼ 18.2 mg; 95% CI, 2.65-33.7 mg; P ¼ .023) and in the last
(36-48 h) sampling period compared with the first (0-12 h) sampling period (b ¼ 30.2 mg; 95% CI, 14.6-45.7 mg; P ¼ .0004), but not in bolus
compared with continuous infusion (b ¼ 9.45 mg; 95% CI, 6.1 to 25.0 mg; P ¼ .224). B, 12-h urine oxalate concentration. C, Change in arterial
base excess from baseline. In a linear model, the slope of the change in arterial base excess from 0 to 48 h was associated with the 10 g/d dose
(0.047 mmol/h compared with 2 g/d dose, P ¼ .0002), but not with the infusion method.

1374 Original Research [ 153#6 CHEST JUNE 2018 ]

between patients who were treated with 10 vs 2 g/d.
This varying decline may be because of differences in
metabolism, recycling, redistribution, or excretion.26,27
Remarkably, in three patients, vitamin C concentrations
decreased to hypovitaminosis levels as soon as 48 h after
discontinuation of the infusion. These patients had
glomerular hyperfiltration, which is a known cause of
increased drug clearance and may contribute to critical
illness-related hypovitaminosis C.28 Conversely, one
patient with anuric acute kidney injury at 36 h had very
high plasma vitamin C at 48 h (outlier in the 10 g/d
continuous group), confirming earlier reports.8
We also found that urinary excretion depended on dose
and on creatinine clearance and that renal vitamin C
clearance depended on glomerular filtration in the 10 g/d
group but not in the 2 g/d group. A likely explanation is
that vitamin C is freely filtered in the glomerulus, but its
reabsorption in the proximal convoluted tubule via the
sodium-dependent vitamin C transporter 1 is a satiable
process.12 Vitamin C excretion increases proportionally
more at 36 to 48 h in the higher-dose group because
higher steady-state concentrations increase vitamin C
concentrations in the primary urine and because the
transporter-mediated reabsorption cannot increase the
reabsorption rate (e-Fig 1).
In all, our findings indicate that supplementation > 48 h
is needed to maintain plasma concentrations in the
normal range, possibly as long as patients remain
critically ill.
Safety
Oxalate excretion was significantly higher in the 10 g/d
dose compared with the 2 g/d dose. However, the
combined influence of the molar product of calcium and
oxalate, urine volume, and urine pH has not yet been
investigated in patients receiving short-term IV vitamin
C therapy.29 We used ethylenediaminetetraacetic acid to
prevent in vitro conversion of vitamin C to oxalate, but
we cannot fully exclude that vitamin C was artefactually
converted to oxalate (section 3 of e-Appendix 1). The
proportion of variance in urine oxalate explained by
urine vitamin C was low (18%), suggesting that other
metabolic factors contribute to oxalate excretion. Based
on physiological reasoning, Marik et al16 added thiamine
to reduce oxalate excretion, but the effectiveness of this
therapy has not yet been empirically validated.
Above all, the clinical relevance of short-term
hyperoxaluria and microscopic calcium-oxalate
crystallization remains unclear because the risk of stone
chestjournal.org
formation appears negligible in patients without a
history of urolithiasis.30-34
The metabolic component of acid-base balance,
expressed as base excess, increased faster in patients
randomized to 10 g/d, indicating that the higher vitamin
C dose was associated with slight metabolic alkalosis
rather than acidosis. A possible explanation is the
sodium content of the vitamin C preparation used
(approximately 7 mmol Naþ per gram of vitamin C),
which contributes positively to the strong ion difference
and may thereby cause alkalosis.35
Strengths and Limitations
We included a heterogeneous group of critically ill
patients to make the study results generalizable to a
broad intensive care population. Multiple organ
dysfunction was the main inclusion criterion because we
previously found a relation between hypovitaminosis
and organ failure.36 In the absence of a valid real-time
vitamin C measurement method, we could not stratify
randomization according to baseline vitamin C status.
Therefore, we cannot show whether baseline differences
lead to different plasma concentrations. The small
sample size and the heterogeneous population also
preclude the detection of meaningful pharmacokinetic
profiles for specific subpopulations. Because vitamin C
excretion is related to glomerular filtration, excretion is
likely decreased in acute kidney injury.
A limitation of the study is that researchers, clinicians,
and participants were unblinded to treatment allocation.
However, measurements of the primary and safety end
points were performed by an independent laboratory
unaware of treatment allocations.
Although recent studies used a 6-h bolus regimen for up
to 96 h,14,16 we used a 12-h bolus regimen for 48 h
because our primary aim was the development of a
pharmacokinetic model. Because our estimated model
described all dosage regimens, it can be readily
translated to 6-h regimens.
Although the sample size per group was small (five
patients per dosage regimen), the factorial design of the
trial allowed us to cluster effects from the dose (10
patients per group) and infusion method (10 patients
per group), thereby increasing power. The relatively
small sample size precluded the evaluation of clinical
end points such as duration of vasopressor support or
recovery from organ failure and necessitated the use of
surrogate safety outcomes instead of patient-oriented
1375
outcomes (eg, oxalate excretion rather than the actual
risk of urolithiasis).
Conclusions
The present four-group randomized trial in patients
with multiple organ dysfunction found that 2 g/d of IV
vitamin C resulted in normal plasma concentrations,
whereas supraphysiologic concentrations were achieved
with 10 g/d. Although renal losses were substantial, high
concentrations can be achieved with higher IV doses in a
linear dose-concentration relationship. Sustained
vitamin C administration is needed to prevent a decline
to hypovitaminosis. Urine oxalate excretion was related
to the administered dose, but the risk of calcium-oxalate
crystallization with short-term therapy remains to be
investigated. The present pharmacokinetic data can be
used to determine the appropriate dose to achieve
desired plasma concentrations.

Acknowledgments
Author contributions: H.-J. G. and H. M.
O.-S. take responsibility for the work as a
whole, from inception to published article.
H.-J. G., W.-P. M.-C., E. L. S., and H. M.
O.-S. contributed to study conception and
design. H.-J. G., W.-P. M.-C., A. M. E. S.-M.,
and H. M. O.-S. were responsible for the
inclusion of patients, study procedures, and
sample handling. W.-P. M.-C. and A. S. Z.
performed the pharmacokinetic modeling.
H.-J. G. performed the other analyses. A. R.
G. and E. L. S. provided administrative,
technical, and material support. All authors
contributed to the interpretation of the data,
contributed to the writing of the paper, and
approved the final version of the manuscript.
Financial/nonfinancial disclosures: The
authors have reported to CHEST the
following: H.-J. G., A. M. E. S.-M., A. R. G.,
and H. M. O.-S. received a research grant
from ZonMW, the Netherlands Organization
for Health Research and Development, to
perform a randomized controlled trial on
high dose vitamin C after cardiac arrest.
None declared (W.-P. M.-C., A. S. Z., E. L.
S.).
Role of sponsors: The sponsor had no role in
the design of the study, the collection and
analysis of the data, or the preparation of the
manuscript.
Other contributions: We thank Erna Alberts
and Ingrid van den Hul for their significant
contribution to the screening and inclusion of
patients and for their help with the study
procedures. We also thank Frans van der
Horst, PhD, and Frances van der Hoek from
the Clinical Chemistry Laboratory at Reinier
de Graaf Hospital for the vitamin C
measurements.
Additional information: The e-Appendix
and e-Figures can be found in the
Supplemental Materials section of the online
article.
References
1. Carr AC, Maggini S. Vitamin C and
immune function. Nutrients. 2017;9(11).
2. Carr AC, Shaw GM, Fowler AA,
Natarajan R. Ascorbate-dependent
vasopressor synthesis: a rationale for
vitamin C administration in severe sepsis
and septic shock? Crit Care. 2015;19:418.
3. Padayatty SJ, Doppman JL, Chang R, et al.
Human adrenal glands secrete vitamin C
in response to adrenocorticotrophic
hormone. Am J Clin Nutr. 2007;86(1):
145-149.
4. Jenkins JS. The effect of ascorbic acid on
adrenal steroid synthesis in vitro.
Endocrinology. 1962;70:267-271.
5. Levenson SM, Green RW. Ascorbic acid,
riboflavin, thiamin, and nicotinic acid in
relation to severe injury, hemorrhage, and
infection in the human. Ann Surg.
1946;124(5):840-856.
6. Long CL, Maull KI, Krishnan RS, et al.
Ascorbic acid dynamics in the seriously ill
and injured. J Surg Res. 2003;109(2):
144-148.
7. Schorah CJ, Downing C, Piripitsi A, et al.
Total vitamin C, ascorbic acid, and
dehydroascorbic acid concentrations in
plasma of critically ill patients. Am J Clin
Nutr. 1996;63(5):760-765.
8. Carr AC, Rosengrave PC, Bayer S,
Chambers S, Mehrtens J, Shaw GM.
Hypovitaminosis C and vitamin C
deficiency in critically ill patients despite
recommended enteral and parenteral
intakes. Crit Care. 2017;21(1):300.
9. Marik PE, Hooper MH. Doctor-your
septic patients have scurvy! Crit Care.
2018;22(1):23.
10. Oudemans-van Straaten HM, Man A, de
Waard MC. Vitamin C revisited. Crit
Care. 2014;18(4):460.
11. Jackson TS, Xu A, Vita JA, Keaney JF.
Ascorbate prevents the interaction of
superoxide and nitric oxide only at very
high physiological concentrations. Circ
Res. 1998;83(9):916-922.
12. Levine M, Padayatty SJ, Espey MG.
Vitamin C: a concentration-function
approach yields pharmacology and
therapeutic discoveries. Adv Nutr.
2011;2(2):78-88.
13. van Zanten AR, Sztark F, Kaisers UX,
et al. High-protein enteral nutrition
enriched with immune-modulating
nutrients vs standard high-protein enteral
nutrition and nosocomial infections in the
ICU: a randomized clinical trial. JAMA.
2014;312(5):514-524.
14. Fowler AA, Syed AA, Knowlson S, et al.
Phase I safety trial of intravenous ascorbic
acid in patients with severe sepsis. J Transl
Med. 2014;12(1):32.
15. Zabet MH, Mohammadi M, Ramezani M,
Khalili H. Effect of high-dose ascorbic acid
on vasopressor’s requirement in septic
shock. J Res Pharm Pract. 2016;5(2):
94-100.
16. Marik PE, Khangoora V, Rivera R,
Hooper MH, Catravas J. Hydrocortisone,
vitamin-c, and thiamine for the treatment
of severe sepsis and septic shock: a
retrospective before-after study. Chest.
2017;151(6):1229-1238.
17. Oudemans-van Straaten HM,
Elbers PWG, Spoelstra-de Man AM. How
to give vitamin C a cautious but fair
chance in severe sepsis. Chest.
2017;151(6):1199-1200.
18. Robitaille L, Mamer OA, Miller WH, et al.
Oxalic acid excretion after intravenous
ascorbic acid administration. Metabolism.
2009;58(2):263-269.
19. Lamarche J, Nair R, Peguero A,
Courville C. Vitamin C-induced oxalate
nephropathy. Int J Nephrol. 2011;2011:
146927.
20. Berger MM, Oudemans-van Straaten HM.
Vitamin C supplementation in the
critically ill patient. Curr Opin Clin Nutr
Metab Care. 2015;18(2):193-201.
21. Vincent JL, Moreno R, Takala J, et al. The
SOFA (Sepsis-related Organ Failure
Assessment) score to describe organ
dysfunction/failure. On behalf of the
Working Group on Sepsis-Related
Problems of the European Society of
Intensive Care Medicine. Intensive Care
Med. 1996;22(7):707-710.
22. Levine M, Rumsey SC, Daruwala R,
Park JB, Wang Y. Criteria and
recommendations for vitamin C intake.
JAMA. 1999;281(15):1415-1423.
23. Asplin JR. Hyperoxaluric calcium
nephrolithiasis. Endocrinol Metab Clin
North Am. 2002;31(4):927-949.
24. Hoppe B. An update on primary
hyperoxaluria. Nat Rev Nephrol.
2012;8(8):467-475.
25. Wilson JX. Mechanism of action of
vitamin C in sepsis: ascorbate modulates
redox signaling in endothelium.
Biofactors. 2009;35(1):5-13.
26. May JM, Qu ZC. Ascorbic acid prevents
increased endothelial permeability caused
by oxidized low density lipoprotein. Free
Radic Res. 2010;44(11):1359-1368.

1376 Original Research [ 153#6 CHEST JUNE 2018 ]

27. Padayatty SJ, Levine M. Vitamin C: the
known and the unknown and Goldilocks.
Oral Dis. 2016;22(6):463-493.
28. Udy A, Roberts JA, Boots RJ, Lipman J.
You only find what you look for: the
importance of high creatinine clearance in
the critically ill. Anaesth Intensive Care.
2009;37(1):11-13.
29. Daudon M, Letavernier E, Frochot V,
Haymann JP, Bazin D, Jungers P.
Respective influence of calcium and
oxalate urine concentration on the
formation of calcium oxalate
monohydrate or dihydrate crystals.
Comptes Rendus Chim. 2016;19(11-12):
1504-1513.
30. Stephenson CM, Levin RD, Spector T,
Lis CG. Phase I clinical trial to evaluate
chestjournal.org
the safety, tolerability, and
pharmacokinetics of high-dose
intravenous ascorbic acid in patients with
advanced cancer. Cancer Chemother
Pharmacol. 2013;72(1):139-146.
31. Hoffer LJ, Levine M, Assouline S, et al.
Phase I clinical trial of i.v. ascorbic acid in
advanced malignancy. Ann Oncol.
2008;19(11):1969-1974.
32. Riordan HD, Casciari JJ, González MJ,
et al. A pilot clinical study of continuous
intravenous ascorbate in terminal cancer
patients. P R Health Sci J. 2005;24(4):
269-276.
33. Tanaka H, Matsuda T, Miyagantani Y,
Yukioka T, Matsuda H, Shimazaki S.
Reduction of resuscitation fluid volumes
in severely burned patients using ascorbic
acid administration: a randomized,
prospective study. Arch Surg. 2000;135(3):
326-331.
34. Lamas GA, Goertz C, Boineau R, et al.
Effect of disodium EDTA chelation
regimen on cardiovascular events in
patients with previous myocardial
infarction: the TACT randomized trial.
JAMA. 2013;309(12):1241-1250.
35. Kellum JA. Determinants of blood pH in
health and disease. Crit Care. 2000;4(1):
6-14.
36. de Grooth HJ, Spoelstra-de Man AM, de
Waard MC, Oudemans-van
Straaten HM. Plasma vitamin C in
critically ill patients is related to organ
dysfunction and mortality [abstract].
Clin Nutr. 2014;33:S2.
1377