RESEARCH IN CONTEXT

Vitamin-based therapies, in particular, ascorbic acid (vitamin C) and thiamin, have undergone extensive trials in septic adults. The role of vitamin C and thiamin deficiency in children with sepsis is largely unknown. Given that vitamin-based therapies are being used off-label in critically ill children, we aimed to assess the association of vitamin C and thiamin levels with multiple organ dysfunction syndrome (MODS) in children with infection.

WHAT THIS STUDY MEANS

In this observational study of 221 children with suspected sepsis, vitamin C and thiamin levels were comparable to reference values in the absence of organ dysfunction. Vitamin C levels were inversely correlated with severity as measured by pediatric Sequential Organ Failure Assessment, and low vitamin C levels on presentation were associated with MODS in multivariable analyses. We failed to identify an association between thiamin levels and severity.

Mortality in children with septic shock admitted to PICUs in Australia and New Zealand was around 17% in population-based studies covering 2003–2015 (1–3). In adults with sepsis, treatment with intravenous vitamin C, thiamin, and hydrocortisone has been proposed as rescue therapy. The rationale for this intervention stems from the idea that vitamin C has antioxidant and immunomodulatory effects that lead to improved microvascular function, and enhanced effectiveness of glucocorticoid and catecholamine pathways (4–6). In addition, thiamin has a key role in cellular energy metabolism, and low levels are associated with lactatemia in critically ill patients (7,8). However, results from randomized controlled trials (RCTs) have been predominantly negative (9), but observational data in septic children suggest potential benefit of vitamin C and thiamin (10,11). Given the differences between children and adults in host susceptibility, immunology, epidemiology, and nutritional status, it is essential to generate pediatric data on vitamin levels in critically ill children.

Therefore, we aimed to measure vitamin C and thiamin levels in children evaluated for suspected sepsis and investigate the association between low vitamin C or thiamin levels with multiple organ dysfunction syndrome (MODS). We hypothesized that vitamin C and thiamin levels correlate with severity.

MATERIALS AND METHODS

Single-center prospective observational study nested within the “Rapid Pediatric Infection Diagnosis in Sepsis (RAPIDS)” study (approved June 06, 2019, by Children’s Health Queensland Human Research Ethics Committee Brisbane, Australia; HREC/17/QRCH/85). The study was performed in accordance with the ethical standards of the committee and with the Helsinki Declaration of 1975. Reporting follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines (12). Prospective consent or consent-to-continue was obtained from the parents of study participants. Children greater than 1 month old and less than 17 years old were eligible if they underwent blood culture sampling for suspected sepsis upon presentation to the emergency department (ED) or the PICU at Queensland Children’s Hospital, between September 25, 2017, and February 05, 2020. Two milliliters of serum were collected concomitantly with blood cultures. Immunosuppressed patients, neonates, patients with hospital-acquired infection, and children where parents were unavailable to provide consent were excluded. Patient data were collected into the Research Electronic Data Capture database (Vanderbilt University, Nashville, TN), including demographics, history and comorbidities, infection data, and severity. Severity was assessed using the pediatric Sequential Organ Failure Assessment (pSOFA) at the time of blood sampling, and 24 hours after sampling (13,14). Adjudication of the type and severity of infection was performed by two assessors using available clinical and microbiological information at discharge (15). Serum was centrifuged immediately at 1,000g for 15 minutes before aliquoting into cryotubes and transferred into a minus 80°C freezer until analysis.

Serum ascorbate concentrations were quantified with an in-house validated method using ultra-performance liquid chromatography (UPLC) coupled with a mass spectrometer. Sample preparation consisted of protein precipitation with 10% meta-phosphoric acid containing internal standard ascorbic acid-13C6 in a 96-well plate. Chromatographic separation was performed on an Acquity UPLC Atlantis T3 (Waters Limited, Wilmslow, United Kingdom) 3.0 µm 2.1 × 100 mm column using a water/formic acid/acetonitrile gradient over a run time of 2 minutes and the mass detected with a QDa spectrometer (Waters Corporation, Milford, MA). The measurement range starts at 0 mg/L and is linear up to 1,000 mg/L. The limit of the blank was 0.04 mg/L and the inter-run precision (coefficient of variation) was less than 7% across four quality control levels. Whole blood thiamin concentrations were quantified using Chromsystems vitamin B1 Whole Blood Kit (Munich, Germany) as per the manufacturer’s instructions.

The primary outcome was defined as MODS at the time of sampling, with MODS being defined as a pSOFA score greater than 1 in at least two organ systems. Low levels of vitamin C were defined as less than 23 µmol/L and deficiency was defined as less than 11 µmol/L (16). Thiamin deficiency was defined as levels less than 70 nmol/L (8). Categorical data are reported as percentages and numbers, with continuous data presented as medians with interquartile ranges (IQRs). Spearman’s correlation coefficient was used to investigate unadjusted associations between vitamin C and thiamin levels with pSOFA. Group comparisons were undertaken using the Kruskal-Wallis test. We investigated associations of vitamin levels with MODS at the time of sampling, and at 24 hours after sampling, using bivariable and multivariable logistic regression. Backward stepwise regression was used to develop the final model; independent variables were chosen based on existing literature and those with clinical importance. In the instance where independent variables with strong clinical or statistical collinearity were available for inclusion in the multivariable model, the most clinically relevant variable was selected. All resulting models met the required assumptions. No sample size calculation was performed, and analyses were not adjusted for multiple comparisons. As such, results are considered exploratory, rather than confirmatory. All analyses were conducted using StataSE, version 17.0 (Stata Corp, College Station, TX). p values less than 0.05 were considered statistically significant.

RESULTS

Two hundred twenty-one children were enrolled with a median age of 3.5 (range 0.08, 16.7) years; and 101 of 221 (45.7%) were female. Comorbidities were present in 27 of 221 (12.2%) patients. Samples were collected in the ED in 144 of 221 (65%) patients, and the other 77 of 221 upon admission to PICU. The median (IQR) time from ED/PICU admission to blood sampling was 160 (IQR 88, 264) minutes. The median length of stay in PICU was 4.5 (IQR 2.5, 9.8) days. Overall, 4 of 221 patients (1.8%) died; and 82 of 221 patients (37.1%) had organ dysfunction at the time of sampling, with a median pSOFA score of 6 (IQR 3, 10).

Vitamin C levels ranged from 0 µmol/L to 151.3 µmol/L with a median (IQR) of 29.9 (13.5, 52.0) µmol/L. Vitamin C levels were inversely correlated with pSOFA scores (Fig. 1, Spearman’s rho = –0.16, p = 0.018). Median (IQR) vitamin C levels were 35.7 (17.9, 54.1) µmol/L, 36.1 (21.4, 53.7) µmol/L, and 17.9 (6.6, 43.0) µmol/L in children without organ dysfunction, single organ dysfunction, and MODS at admission, respectively (p = 0.017, Fig. 1A). Low vitamin C levels were associated with MODS at time of sampling, and MODS at 24 hours, as was vitamin C deficiency in bivariate analyses (Supplementary Table 1, https://links.lww.com/PCC/C420). In multivariable analyses, low vitamin C levels were associated with greater odds of MODS at the time of sampling (aOR 3.04; 95% CI, 1.51–6.12), and vitamin C deficiency was associated with greater odds of MODS at 24 hours (aOR 3.38; 95% CI, 1.53–7.47). C-reactive protein was inversely correlated with vitamin C levels (Spearman`s rho = –0.538, p < 0.001), but we failed to identify such a correlation for thiamin levels (Spearman`s rho = –0.136, p = 0.062).

Figure 1.:

Serum levels of vitamin C (A, C) and thiamin (B, D) are shown for n = 221 children evaluated for sepsis in relation to the degree of organ dysfunction at the time of blood sampling (A, B: no organ dysfunction, n = 139; single organ dysfunction, n = 19, multiple organ dysfunction, n = 63), and in relation to the pediatric Sequential Organ Failure Assessment score (pSOFA, C, D). Box-whisker plots indicate the median and interquartile range. The dashed horizontal lines indicate the threshold to define vitamin C deficiency (< 11 µmol/L), and thiamin deficiency (< 70 nmol/L), respectively. pSOFA = pediatric Sequential Organ Failure Assessment.

Thiamin levels on admission ranged from 53 nmol/L to 438 nmol/L with a median (IQR) of 161 (129, 190) nmol/L. Median (IQR) thiamin levels were 162 (138, 192) nmol/L, 185 (143, 200) nmol/L, 136 (110, 179) nmol/L in children without organ dysfunction, single organ dysfunction, and MODS, respectively (p = 0.06 [Fig. 1]). We failed to identify any association between thiamin deficiency and greater odds of MODS at time of sampling (OR 2.52; 95% CI, 0.15–40.86), or greater odds of MODS at 24 hours (OR 2.96; 95% CI, 0.18–48.18; Supplementary Table 1, https://links.lww.com/PCC/C420).

DISCUSSION

In this prospective observational study of children evaluated for sepsis, we observed that a substantial proportion of children had vitamin C levels below the normal range. Vitamin C levels correlated inversely with pSOFA scores and low levels were associated with greater odds of MODS. For thiamin serum levels, we failed to identify any consistent association with severity.

Vitamins have gained interest as potential adjunctive therapies in critically ill patients (4). A network meta-analysis in adults with sepsis indicated that high-dose vitamin C treatment was associated with decreased mortality, albeit with low certainty (17). In an animal model, vitamin C used in much higher doses than in previous RCTs was effective in mitigating sepsis-related organ dysfunction (18). In children, systematic reviews indicate a high safety profile (7,10,19,20). Our study was designed to characterize vitamin C and thiamin levels in a prospective cohort of children evaluated for sepsis, with different microbiological etiologies, age groups, comorbidities, and variable degrees of illness severity. Vitamin measurements were determined at admission to the ED or PICU to minimize the effect of critical illness-associated malnutrition on vitamin levels. Children with organ dysfunction in our cohort had even lower vitamin C levels compared to previous studies of general PICU patients which revealed overall decreases in micronutrients (11,21,22), and compared to adults with sepsis (16). Endotoxinemia during sepsis leads to widespread activation of immune and endothelial cells, platelets, and the coagulation system, all of which cause a dramatic increase in multiple oxidative mechanisms and contribute to vitamin C depletion (18).

Our study has several limitations. First, vitamin measurements were only obtained upon admission, and we were thus unable to longitudinally assess serum vitamin levels. Second, we did not obtain measures of skin-fold thickness, body mass indices, blood transfusions, or a detailed nutritional history of patients. The cohort represents a metropolitan pediatric population in a high-income country, and we did not collect information on eligible children who were not enrolled. Hence, the vitamin levels may not be generalizable to children in less-resourced settings who may be more prone to malnutrition. Third, although we observed associations between vitamin C levels and pSOFA, a causal link between organ dysfunction and consumption of vitamin C cannot be elucidated with our data. Notably, an association of low vitamin levels with greater odds of MODS does not necessarily justify vitamin supplementation, as levels may be directly affected by factors relating to critical illness and therapies, and may reflect adaptive rather than pathological mechanisms.

In conclusion, in this cohort of children with suspected sepsis, serum vitamin C levels were frequently low. Whether vitamin C-based therapies in children with sepsis improve outcomes needs to be assessed in future RCTs.

ACKNOWLEDGMENTS

The authors thank the parents and children participating in this trial and the medical and nursing and research teams in the participating sites for their help in study setup, recruitment, data collection, and monitoring of study data.

REFERENCES 1. Schlapbach LJ, Straney L, Alexander J, et al.; ANZICS Paediatric Study Group: Mortality related to invasive infections, sepsis, and septic shock in critically ill children in Australia and New Zealand, 2002-13: A multicentre retrospective cohort study. Lancet Infect Dis. 2015; 15:46–54 2. Weiss SL, Peters MJ, Alhazzani W, et al.: Surviving sepsis campaign international guidelines for the management of septic shock and sepsis-associated organ dysfunction in children. Intensive Care Med. 2020; 46:10–67 3. Schlapbach LJ, MacLaren G, Festa M, et al.; Australian & New Zealand Intensive Care Society (ANZICS) Centre for Outcomes & Resource Evaluation (CORE) and Australian & New Zealand Intensive Care Society (ANZICS) Paediatric Study Group: New Zealand Intensive Care Society Paediatric Study: Prediction of pediatric sepsis mortality within 1 h of intensive care admission. Intensive Care Med. 2017; 43:1085–1096 4. Mehta NM: Resuscitation with vitamins C and B1 in pediatric sepsis—hold on to your “HAT.”. Pediatr Crit Care Med. 2022; 23:385–389 5. Wilson JX: Mechanism of action of vitamin C in sepsis: Ascorbate modulates redox signaling in endothelium. Biofactors. 2009; 35:5–13 6. May CN, Bellomo R, Lankadeva YR: Therapeutic potential of megadose vitamin C to reverse organ dysfunction in sepsis and COVID‐19. Br J Pharmacol. 2021; 178:3864–3868 7. Weiss SL, Blowey B, Keele L, et al.: Matched retrospective cohort study of thiamine to treat persistent hyperlactatemia in pediatric septic shock. Pediatr Crit Care Med. 2019; 20:e452–e456 8. Donnino MW, Andersen LW, Chase M, et al.; Center for Resuscitation Science Research Group: Randomized, double-blind, placebo-controlled trial of thiamine as a metabolic resuscitator in septic shock: A pilot study. Crit Care Med. 2016; 44:360–367 9. Schlapbach LJ, de Oliveira CF, Raman S, et al.: Metabolic resuscitation in pediatric sepsis: A narrative review. Transl Pediatr. 2021; 10:2678–2688 10. Wald EL, Sanchez-Pinto LN, Smith CM, et al.: Hydrocortisone-ascorbic acid-thiamine use associated with lower mortality in pediatric septic shock. Am J Respir Crit Care Med. 2020; 201:863–867 11. Fathi A, Downey C, Rabiee Gohar A: Vitamin C deficiency in critically ill children: Prospective observational cohort study. Pediatr Crit Care Med. 2022; 23:395–398 12. Vandenbroucke JP, von Elm E, Altman DG, et al.; STROBE Initiative: Strengthening the Reporting of Observational Studies in Epidemiology (STROBE): Explanation and elaboration. Epidemiology. 2007; 18:805–835 13. Schlapbach LJ, Straney L, Bellomo R, et al.: Prognostic accuracy of age-adapted SOFA, SIRS, PELOD-2, and qSOFA for in-hospital mortality among children with suspected infection admitted to the intensive care unit. Intensive Care Med. 2018; 44:179–188 14. Matics TJ, Sanchez-Pinto LN: Adaptation and validation of a pediatric Sequential Organ Failure Assessment score and evaluation of the sepsis-3 definitions in critically ill children. JAMA Pediatr. 2017; 171:e172352 15. Herberg JA, Kaforou M, Wright VJ, et al.; IRIS Consortium: Diagnostic test accuracy of a 2-transcript host RNA signature for discriminating bacterial vs viral infection in febrile children. JAMA. 2016; 316:835–845 16. Carr AC, Rosengrave PC, Bayer S, et al.: Hypovitaminosis C and vitamin C deficiency in critically ill patients despite recommended enteral and parenteral intakes. Crit Care. 2017; 21:300 17. Fujii T, Salanti G, Belletti A, et al.: Effect of adjunctive vitamin C, glucocorticoids, and vitamin B1 on longer-term mortality in adults with sepsis or septic shock: A systematic review and a component network meta-analysis. Intensive Care Med. 2022; 48:16–24 18. Lankadeva YR, Peiris RM, Okazaki N, et al.: Reversal of the pathophysiological responses to Gram-negative sepsis by megadose vitamin C. Crit Care Med. 2021; 49:e179–e190 19. Schlapbach LJ, Gibbons K, Ridolfi R, et al.: Resuscitation in paediatric sepsis using metabolic resuscitation—A Randomized Controlled Pilot Study in the Paediatric Intensive Care Unit (RESPOND PICU): Study protocol and analysis plan. Front Pediatr. 2021; 9:663435 20. Yanase F, Raman S, Naorungroj T, et al.: Efficacy and safety of parenteral high-dose vitamin C therapy in pediatric patients: A scoping review. Pediatr Crit Care Med. 2021; 22:561–571 21. Valla FV, Bost M, Roche S, et al.: Multiple micronutrient plasma level changes are related to oxidative stress intensity in critically ill children. Pediatr Crit Care Med. 2018; 19:e455–e463 22. Equey L, Agyeman PKA, Veraguth R, et al.; Swiss Pediatric Sepsis Study Group: Serum ascorbic acid and thiamine concentrations in sepsis: Secondary analysis of the Swiss Pediatric Sepsis Study. Pediatr Crit Care Med. 2022; 23:390–394