Clin Chem Lab Med 2017; 55(11): 1652–1668

Review

Jake T.B. Collie*, Ronda F. Greaves, Oliver A.H. Jones, Que Lam, Glenn M. Eastwood and Rinaldo Bellomo Vitamin B1 in critically ill patients: needs and challenges

DOI 10.1515/cclm-2017-0054 Summary: A total of 777 published articles were identified; Received January 19, 2017; accepted March 21, 2017; previously 122 were included in this review. The most common pub- ­published online April 22, 2017 lished method is HPLC with florescence detection. Two of Abstract the six EQA organisations include a thiamine measure- ment programme, both measuring only whole-blood thia- Background: Thiamine has a crucial role in energy produc- mine pyrophosphate (TPP). No standard measurement tion, and consequently thiamine deficiency (TD) has been procedure for thiamine compound quantification was associated with cardiac failure, neurological disorders, identified. oxidative stress (lactic acidosis and sepsis) and refeeding Outlook: Overall, there is an absence of standardisation in syndrome (RFS). This review aims to explore analytical measurement methodologies for thiamine in clinical care. methodologies of thiamine compound quantification and Consequently, multiple variations in method practises are highlight similarities, variances and limitations of current prohibiting the comparison of study results as they are not techniques and how they may be relevant to patients. traceable to any higher order reference. Traceability of cer- Content: An electronic search of Medline, PubMed and tified reference materials and reference measurement pro- Embase databases for original articles published in peer- cedures is needed to provide an anchor to create the link reviewed journals was conducted. MethodsNow was between studies and help bring consensus on the clinical used to search for published analytical methods of thia- importance of thiamine. mine compounds. Keywords for all databases included Keywords: chromatography; critically ill; mass spectro- “thiamine and its phosphate esters”, “thiamine meth- metry; standardisation; thiamine; vitamin B1. odology” and terms related to critical illness. Enquiries were also made to six external quality assurance (EQA) programme organisations for the inclusion of thiamine measurement. Introduction

The role of thiamine and more specifically thiamine *Corresponding author: Jake T.B. Collie, School of Health and pyrophosphate (TPP), the active form of what is com- Biomedical Sciences, RMIT University, PO Box 71 Bundoora, monly known as vitamin B1, in carbohydrate metabolism Melbourne, Victoria 3083, Australia, Phone: +610399257080, E-mail: [email protected]; and Dorevitch Pathology, is well documented. However, a broader understanding Special Chemistry, 18 Banksia St, Heidelberg, Melbourne, Victoria of the biological role(s) of thiamine outside of carbohy- 3084, Australia drate metabolism is not yet available. Thiamine is a water-­ Ronda F. Greaves: School of Health and Biomedical Sciences, RMIT soluble vitamin which is passively absorbed in the small University, Bundoora, Australia; Murdoch Children’s Research intestine at high concentrations or actively absorbed Institute, Parkville, Victoria, Australia; and Australasian Association through thiamine transporter proteins at lower concentra- of Clinical Biochemists Vitamins Working Party Oliver A.H. Jones: Australian Centre for Research on Separation tions [1]. It exists in multiple forms through the addition Science, School of Science, RMIT University, Melbourne, Australia of one or more phosphate groups. These include the non- Que Lam: Department of Pathology, Austin Health, Heidelberg, phosphorylated or free thiamine, thiamine monophos- Australia phate (TMP), TPP or diphosphate (TPP) and thiamine Department of Intensive Care, Austin Health, Glenn M. Eastwood: triphosphate (TTP) (Figure 1). Collectively, these will be Heidelberg, Australia Rinaldo Bellomo: Department of Intensive Care, Austin Health, referred to here as thiamine compounds. Heidelberg, Australia; and School of Medicine, The University of The human adult can store around 30 mg of thiamine Melbourne, Parkville, Australia in muscle tissue, liver and kidneys [1]. These stores can

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM Collie et al.: Vitamin B1 in critically ill patients: needs and challenges 1653

Figure 1: The chemical structure and currently understood biochemical pathway of thiamine, thiamine monophosphate, thiamine pyrophos- phate and thiamine triphosphate. CSID:1099, http://www.chemspider.com/Chemical-Structure.1099.html (accessed 11:50, Jan 10, 2017). *Thiamine monophosphatase and thiamine triphosphosynthase enzymes are still yet to be identified or verified. however become depleted in as little as 18 days after the Patients presenting to an ICU are often immobile and in cessation of thiamine intake [2, 3]. This leads to a variety an unconscious state. These populations frequently require of clinical features that are often vague and sometimes nutritional support typically prescribed through parenteral non-specific [4–6]. In recent years, there has been increas- or enteral administration. Despite this, the bio-availability ing recognition of its importance in neurological function of thiamine may be reduced due to fluid loss, increased as well as the discovery of new thiamine adenine nucleo- metabolic demand, secondary to intensive care treatments tide species [6–10]. A greater understanding of human or the development of RFS upon commencing parenteral or thiamine biochemistry is highly desirable and clinically enteral nutrition. RFS itself is a multifaceted condition with relevant as thiamine deficiency (TD) is still reported numerous pathophysiological features and a similar cohort worldwide. Despite long-term thiamine fortification of a presentation to TD. The associated biochemistry of RFS variety of common household foods (meats, vegetables, is complex with the common biochemical profiles of low milk products, cereals and whole grains), TD continues to serum concentrations of magnesium, phosphate, potas- be present even in modern, western societies [11–14]. sium and thiamine all considered indicators of RFS [19, 20]. Classically, TD causes the disease known as beriberi, Identifying RFS and TD therefore requires keen observation but this disease is really one of several thiamine-­deficiency- and a sound knowledge of potential risk factors, comple- related conditions, which may occur concurrently. Key con- mented with appropriate biochemical markers. tributors to TD include hospitalisation, moderate to severe The significance of thiamine measurements in those alcohol consumption, and improper nutritional intake, at risk of RFS in the ICU may be more important than e.g. a high carbohydrate load with restricted vitamin and previously thought as states of subclinical deficiency or mineral ingestion. Age, as well as co-morbidities, such as relative deficiency may also exist. If this is true, then addi- cardiac and liver dysfunction, surgery (e.g. bariatric), lactic tional research questions might include: acidosis, sepsis, trauma and refeeding syndrome (RFS) 1. Should thiamine be routinely measured in hospital- may also be associated risk factors for TD [11, 15–18]. All ised patients to assess their status? of these issues are commonplace in critically ill patients 2. If this is so, should the measurand be total thiamine admitted to an intensive care unit (ICU). (where all thiamine compounds are reduced to form

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM 1654 Collie et al.: Vitamin B1 in critically ill patients: needs and challenges

free un-phosphorylated thiamine), or a profile of all Table 1: Search terms for publications in Embase, PubMed, Medline thiamine compounds to assess the biological avail- and MethodsNow. ability of each compound? Database Search terms 3. Finally, what current methodologies allow for the accurate measurement, and therefore clinical inter- Embase – Critically ill OR Critical illness OR Critically ill pretation, of thiamine? PubMed patients OR Critical Care OR Intensive care OR Medline Intensive care unit AND The aim of this review is to explore analytical methodolo- – Thiamine OR Thiamin OR Vitamin B1 OR Anurine gies of thiamine compound quantification and highlight OR Thiamine monophosphate OR Thiamine similarities, variances and limitations of current tech- pyrophosphate OR Thiamine diphosphate OR niques. A brief history of the vitamin and an examination Thiamine triphosphate OR Thiamine deficiency OR of the relationship between levels of thiamine compounds Thiamine ester OR Thiamine derivative AND in the blood of critically ill patients will also be presented. – Peer reviewed AND – Human Review methodology AND – English MethodsNow – Thiamine OR Thiamin OR Vitamin B1 OR Anurine Literature OR Thiamine monophosphate OR Thiamine pyrophosphate OR Thiamine diphosphate OR Thiamine triphosphate OR Thiamine deficiency OR The study objectives were achieved by electronically Thiamine ester OR Thiamine derivative searching the Medline, PubMed and Embase databases AND for original articles published in peer-reviewed journals – HPLC OR Florescence OR UV OR Ultra violet, OR with no limitations on year of publication. The Methods­ mass spectrometry Now website was also used as a tool to search for pub- AND lished methods of thiamine compound quantification in a – Urine OR Serum OR Plasma OR Blood OR Whole Blood OR Erythrocytes OR Washed erythrocytes variety of matrices [21]. Search keywords for all databases included “thiamine and its phosphate esters”, “thia- mine methodology” and terms related to critical illness The RCPAQAP and SKML schemes consist of a (Table 1). Manual searching of articles, including scan- sample set of six linearly related levels analysed twice ning the reference lists of cited publications, was also per cycle with two cycles each year [22, 23]. All par- performed to avoid omissions. Overall, 777 papers were ticipants are presented with a survey to classify each identified (723 using these databases, 54 through manual component of their method. This survey information is searching). A title scan of these papers reduced the total to divided into four categories: analytical principle, meas- 234 which was further reduced to 160 after the examina- urement system, reagent source and calibrator source. tion of abstracts and the removal of duplicate results. Of After each sample set, participants are provided with a these 160 papers, 122 were included in this review. survey report, indicating their performance in each of the method classifications. EQA programmes Both agencies provided de-identified result sets and median values for their latest thiamine programme To further examine the current status of the analytical cycles for the purpose of this review. Results from the methods, enquiries were made to six EQA organisations RCPAQAP Cycle 33 and SKML Cycles 20151 and 20152 pro- to ascertain the inclusion of a thiamine related compound grammes were used to assess the performance of current in their programme. Of these, two were found to include thiamine methodologies. As the results provided repre- thiamine in the form of whole-blood TPP: The Dutch sent an identical batch of lyophilised sample material, Foundation for Quality Assessment in Medical Laborato- the individual reported results were combined to gen- ries (SKML) and The Royal College of Pathologists Austral- erate an overall performance of the participating labo- asia Quality Assurance Programs (RCPAQAP). These EQA ratories. For note, prior to analysing these results each programmes co-operate by using a common lyophilised individual cycle was subject to a one-way ANOVA test to material, purchased in bulk and distributed to participat- ensure no population differences occurred among the ing laboratories in numerous countries. cycles (p > 0.05).

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM Collie et al.: Vitamin B1 in critically ill patients: needs and challenges 1655

Early work on thiamine many others, was not content with that nomenclature and in 1935 proposed the name aneurin (from anti-polyneu- The discovery of thiamine and indeed vitamins in general ritisvitamin) [32, 33]. As Jansen was half of the team that began with Eijkman’s early studies of the causes of beri- first isolated the crystalline structure, the rights to name beri in Java [24]. Eijkman originally sought a bacterial the vitamin seemed appropriate. However, a year later, cause of the disease, a logical conclusion at the time, as Williams succeeded in identifying the chemical structure diseases arising from nutrient deficiency had not yet been and synthesising it [34, 35]. In 1938, Williams formally recognised [24]. The key discovery in Eijkman’s research introduced the term thiamin due to the fact it was a sul- occurred when the similarities between the symptoms phur-containing vitamin [36]. Whilst some older tests use of polyneuritis in his laboratory chickens and the poly- the term anurine, today it is more commonly presented as neuritis occurring in beriberi were noticed. The chicken thiamine (including the addition of an ‘e’ from the original polyneuritis was traced to an accidental change in diet (of name). polished rather than unpolished rice) and when this was The work of Eijkman, Hopkins, Grijns and others reversed, the birds recovered in a few days. Eijkman incor- brought vitamins to the forefront of health research and rectly concluded that the polished rice contained a toxic with it evidence of the existence of nutrient deficiency as starch which was the cause of the leg weakness whilst a cause for disease. On a broader scale, this (much later) the bran and germ of the whole rice neutralised this toxin resulted in the addition of vitamins in foodstuffs, the [25, 26]. Nowadays, it is understood that the polished existence of vitamin supplements as preventative meas- white rice feed was deficient in thiamine, whilst the bran ures to beriberi (vitamin B1 deficiency), rickets (vitamin and germ in the unmilled brown rice was thiamine rich. D deficiency), scurvy (vitamin C deficiency) and pellagra Although Eijkman’s original hypothesis was incorrect,­ he (vitamin B3 deficiency), as well as vitamin recommended was jointly awarded the Nobel Prize in Physiology and dietary intakes (RDI) [1, 2, 30, 37–39]. Medicine in 1929 with Sir Frederick Hopkins for the dis- The distribution of thiamine compounds in the body covery of essential nutrient factors (vitamins) needed in varies considerably. After ingestion, free thiamine is con- animal diets to maintain health [27]. verted to TPP by thiamine pyrophosphokinase and is then It was Gerrit Grijns (Eijkman’s successor) who correctly utilised by numerous metabolic pathways as a cofactor. identified the disease as a lack of vital nutrients (and not TPP can also be further phosphorylated or dephosphoryl- the work of a toxin) in 1901 and attempted to isolate the ated to form any of the thiamine compounds (Figure 1). missing nutritional element by introducing minerals and Gangolf et al. [40] and Tallaksen et al. [41] conducted fats to the diets of the diseased chickens. Neither proved extensive studies on human cells and body fluids which effective [28]. Grijns’ later success in curing the condition determined that the most abundant intracellular thiamine using mung beans led to his conclusion that the underly- species was TPP. TMP and free thiamine were the only ing cause was a deficiency in a compound or compounds, detectable species in the extracellular fluid. The majority as yet unidentified, present in some foods and absent of TPP in the body is found in erythrocytes and accounts from others [29]. Despite Grijns’ pioneering research, his for approximately 80% of the total bodies stores [14, contributions were largely overlooked. His work, however, 42–44]. There is no consensus on reference intervals for allowed for the emerging scientific movement of nutri- thiamine compounds in the human body. As an example tional deficiencies to progress. Notable attributions that published reference intervals for whole-blood TPP range stemmed from both Eijkman and Grijns’ works include the from a lower limit of 63–105 nmol/L to an upper limit of discovery of other B vitamins and the coining of the term 171–229 nmol/L [40, 41, 45–48]. Reference intervals for the ‘vitamin’ (from vital amine) by Funk in 1912 (although less abundant compounds in whole blood and serum have many vitamins were later shown to not be amines) and also been established each with similar variations in their the eventual isolation of crystallised thiamine in 1926 by lower and upper limits [40–42, 45, 47–49]. Jansen and Donath [30, 31]. Clinical practise publications related to nutritional supplementation state that patients with suspected or at risk of developing TD should be provided with thia- Physiology of thiamine mine [50–52]. Malnutrition (starvation or reduced intake, increased loss or impaired absorption), alcoholism, bari- In the 1920s–1930s, thiamine was still known as either atric surgery, refeeding, congestive heart failure, lactic anti-neuritic vitamin or vitamin B1 as it was the first of the acidosis, neuropathy, renal failure with dialysis and the B complex vitamins to be discovered. Jansen, as well as critically ill are all risk factors for the development of TD.

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM 1656 Collie et al.: Vitamin B1 in critically ill patients: needs and challenges

However, there is no consensus on the dose, frequency or covered elsewhere and will not be reviewed here. For the duration of supplementation. For mild symptoms, 20 mg interested reader, papers by Betterndorf [56], Lonsdale [9] a day up to 100 mg 2 or 3 times a day has been suggested. and Manzetti [8] provide excellent overviews of thiamine More advanced symptoms include intravenous doses of in both cofactor and non-cofactor roles. Recent discoveries 50 mg a day to >500 mg twice a day. These recommenda- have also identified thiamine containing adenine nucleo- tions are dependent on the risk factor, with alcoholism tides, but as there is little known in regard to their func- and bariatric surgery calling for higher frequency and tion and mechanisms they will not be discussed further in larger doses, as there is the potential of impaired absorp- this review [10]. tion, compared to cardiac and renal failure. Thiamine’s function as a cofactor, in the form of its diphosphate ester TPP, is the most broadly recognised physiological role. TPP is a cofactor for numerous enzy- Pathophysiology related to TD matic reactions including pyruvate dehydrogenase (PDH), transketolase, α-ketogluterate dehydrogenase (KGDH) The diphosphate esters of thiamine are involved in numer- and branched-chain α-keto acid dehydrogenase (BCKDH) ous enzymatic reactions. In the mitochondria PDH, KGDH [9, 53–55]. Its involvement in these metabolic pathways and BCKDH are heavily involved in the production of and its intracellular abundance has made TPP a popular energy in the tricarboxylic acid cycle (TCA), or the Krebs analyte for assessing thiamine status. However, there are cycle, feeding reduced hydrogen carrier nicotinamide also proposed non-cofactor roles of thiamine compounds adenine dinucleotide (NAD) into the electron trans- within the immune system, gene regulation, oxidative port chain for adenosine triphosphate (ATP) production stress response, cholinergic activity, chloride channels (Figure 2). Transketolase is also involved in energy pro- and neurotransmission [8, 56–61]. duction oxidising two reactions in the pentose phosphate The physiology of thiamine compounds regarding pathway transporting carbohydrates for glycolysis. A synthesis, transport and energy metabolism has been deficiency in thiamine can potentially lead to a decline

Figure 2: TPP biochemical pathways. Four enzymes (transketolase, pyruvate dehydrogenase, α-ketogluterate dehydrogenase and branched-chain keto-acid dehydrogenase) require cofactor roles of thiamine pyrophosphate for pathways in both the mitochondria and cytosol. These pathways include glycolysis, pentose phosphate pathway, TCA (Krebs) cycle and branched-chain amino acid metabolic pathways.

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM Collie et al.: Vitamin B1 in critically ill patients: needs and challenges 1657 in the activity of these TPP-dependent enzymes, causing Metabolic acidosis a decreased energy yield or forcing alternate reactions to ensue. No study has been able correlate low thiamine levels with Beriberi is the clinical disease which manifests oxidative stress or lactic acidosis in the general ICU pop- from TD and presents in three forms. Genetic beriberi is ulation. However, a study by Donnino et al. [16] reported a rare condition that prevents the body from absorbing a significant negative correlation between lactic acido- thiamine. The two more commonly acquired forms of the sis and thiamine levels within a sub population without disease are wet and dry beriberi. Wet beriberi affects the liver dysfunction. Moskowitz et al. [67] also showed a cardiovascular system and can cause cardiac failure. Dry negative correlation on lactate and thiamine levels in beriberi damages the peripheral nervous system and can patients with diabetic ketoacidosis. Furthermore, there lead to muscle paralysis. Dry beriberi may co-present with have been several reported cases where a metabolic aci- neuropathies such as Wernicke’s encephalopathy and dosis has been corrected with thiamine supplementation Korsakoff’s syndrome [62]. Wernicke-Korsakoff syndrome [68, 74, 77, 78]. may occur if both neurologic and psychiatric symptoms are present. The clinical symptoms related to beriberi are associ- Sepsis and septic shock ated with a reduction in the mitochondrial energy pro- duction leading to cardiac and cerebral susceptibility. There is little evidence to support or reject thiamine sup- Peripheral vasodilation, biventricular myocardial failure, plementation in the treatment and prevention of sepsis volume overload, tachycardia, depression of left ventric- and septic shock. Several studies have been described but ular function with low ejection fraction, sudden onset the therapeutic role of thiamine is still inconclusive. cardiac failure and acute renal failure are just some of 1. A retrospective study by Marik et al. [79] on the use of the signs of wet beriberi [6, 63, 64]. In contrast, signs of vitamin C with thiamine and hydrocortisone for the dry beriberi include ophthalmoplegia, ataxia, confusion, treatment of sepsis and septic shock showed a sig- memory loss and confabulation in chronic settings [6, nificant difference in mortality between the treated 65, 66]. These symptoms seem to be variable depending (n = 47) and controlled groups (n = 47). This suggests on age and the spectrum of the deficiency and may only that the intervention may reduce mortality of patients precipitate during periods of increased energy demand with severe sepsis and septic shock. or metabolism in individuals with only a subclinical defi- 2. Donnino et al. [72] observed no differences in lactate ciency. Although cases of beriberi are somewhat of a rarity levels, severity of illness or mortality between the thi- in modern medicine, there are still populations that are at amine treatment (n = 43) and control groups (n = 45). risk of developing TD. These include patients post-gastric However, a sub population for patients with baseline surgery, the elderly, alcoholics, malnourished, diabet- TD was also examined and a significant difference in ics, prolonged hospitalisation, severe fluid loss or fluid lactate levels at 24 h and time to death was observed replacement therapy [3, 15, 67–71]. in this group. Considerations of thiamine status in critical illness is 3. A post-hoc analysis was performed by Moskowitz et al. not a new approach. Studies of low thiamine compound [80] on this study to identify whether thiamine sup- levels include association with oxidative stress, sepsis, plementation reduced the severity of kidney injury in metabolic acidosis, cardiac failure, and neurodegen- septic shock. A significant difference was seen for the eration [16, 62, 72–74]. The prevalence of TD in the ICU, requirement of renal replacement therapy between hospitals and the elderly (>76 years of age) has also been the control and intervention groups (p = 0.04) indicat- examined [15, 75, 76]. A retrospective study by Cruickshank ing that thiamine supplementation may reduce the et al. [63] observed a 20% prevalence of TD in intensive risk of sepsis associated kidney injury. However, no care patients requiring nutritional support. A mortality difference was observed between the groups for in- rate of 72% was seen in patients with TD compared to 50% hospital mortality (p = 0.45). in the non-TD group from 158 patients. In the intensive care and general hospital setting, TD The prevalence of TD in sepsis and septic shock has has been linked to specific presentations including meta- also been determined. A study on patients with sepsis bolic acidosis, sepsis and septic shock, cardiac failure and in the ICU reported a TD prevalence of 10% upon surgery, neurodegeneration, and RFS; each of these will ­admission which increased to 20% within the first be discussed in turn below. 72 h post-admission­ [16]. This study improved on a

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM 1658 Collie et al.: Vitamin B1 in critically ill patients: needs and challenges retrospective study by aforementioned Cruickshank Neurodegeneration paper by ­incorporating direct measurement techniques using high-pressure liquid chromatography (HPLC). A The neurodegeneration patterns of TD in Wernicke’s more recent study by Costa et al. [81] on septic shock encephalopathy and Korsakoff’s syndrome (and the also concluded that the prevalence of TD in the ICU was joint symptoms of Wenicke-Korsakoff’s syndrome) 20% upon admission but increased to 71.3% during the share some similarities with other neurodegenerative course of the study. diseases such as Alzheimer’s disease [62, 90]. All three present with reduced brain TPP, KGDH, PDH and tran- sketolase concentrations, reduced glucose metabolism Cardiac failure and surgery as well as the presence of twisted tau protein fibres, commonly referred to as tangles [62, 66, 91, 92]. Studies There is an association but no clear relationship between have hypothesised that a decrease in brain thiamine TD and cardiac failure. Ahmed et al. [3] reviewed the compounds can lead to a cascade of reactions ultimately prevalence of TD in cardiac failure patients reporting an leading to neuronal cell death through excitotoxicity, overall prevalence of 33%–91% in those that are hospi- inflammation and oxidative damage [91, 93–95]. Recent talised and 21%–27% in a consortium of inpatients and comprehensive publications have reviewed the role of outpatients with cardiac failure. Multiple studies have TD in impaired oxidative metabolism, excitotoxicity and shown improvement with supplementation whilst others blood-brain barrier alterations causing neurodegenera- show no difference between the disease and non-disease tion [62, 90, 96]. groups [64, 73, 82–85]. Interestingly, only selective areas of neurodegen- 1. A perspective observational study by Donnino et al. eration in the brain due to TD have been reported [66, [86] showed a statistically significant difference in 91, 93–95]. This has led to theories that the neurode- thiamine status before and 24 h post-cardiac surgery. generation from TD is due to the localisation of either The enrolled participants were specific to a relatively thiamine-dependent enzymes, with particular atten- healthy population prior to surgery, as they were non- tion to KDGH, or alterations in the phosphokinase or emergency operations with a rapid recovery progno- ­phosphotransferase enzymes required to maintain sis. For this reason the study could not concluded thiamine brain homeostasis [97]. One study has shown that supplementation is warranted in critically ill and that thiamine supplementation does not significantly post-surgery populations. increase brain thiamine compound levels [66]. Yet there However, they did highlight that physical stress, have been reports that thiamine administration reduces even in healthy populations, can cause a significant and even reverses motor-sensory polyneuropathy in change in the thiamine status of a patient. Wenicke-Korsakoff’s syndrome and post-gastrectomy 2. A perspective, double blind, randomised, placebo [98, 99]. This highlights that brain thiamine homeosta- controlled, single centre trial by Berger et al. [87] inves- sis by thiamine specific transporters, enzymatic phos- tigated the outcomes of early antioxidant supplemen- phorylation and de-phosphorylation is still not fully tation in the critically ill. The supplementation regime understood. Verification and identification of specific encompassed several micronutrients (100 mg of thia- enzymes and mechanisms, such as thiamine monophos- mine) for 5 days. A cardiac surgery cohort was also phatase and thiamine triphosphosynthase, is yet to be included. The study concluded that the intervention achieved [100]. did not reduce early organ dysfunction but there was Besides studies into the levels of thiamine compounds a significant reduction in the inflammatory response in Wernicke’s encephalopathy, Korsakoff’s syndrome in cardiac and trauma patients. and in the brains of Alzheimer’s patients, there is little 3. Two double-blind, randomised, placebo-controlled research in the area of neurodegeneration associated with studies (200- and 300-mg doses) were performed TD in the critically ill. The neurological complications that with a main endpoint of post-cardiac surgery blood arise from TD may be subclinical in the critically ill due lactate levels [88, 89]. Both studies concluded no to the latent nature of the deficiency. These symptoms significant differences in blood lactate levels or may only become apparent once periods of increased clinical outcome between intervention and placebo energy requirement or decreased nutritional intake are groups. However, significantly higher blood thia- encountered. Consequently these may not be routinely mine levels were seen in the intervention group in checked for upon admission and may be detrimental to both studies. the patient’s outcome.

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM Collie et al.: Vitamin B1 in critically ill patients: needs and challenges 1659

Refeeding syndrome (RFS) Methods for the measurement

RFS is thought to be caused by the sudden shifts of mole- of thiamine compounds cules from the extracellular to intracellular space leading The therapeutic benefit of thiamine supplementation is to serum nutrient displacement. During periods of starva- still being debated with no consensus on patient groups, tion the body decreases the release of insulin due to the dosage or the duration of supplementation. This diver- cessation of carbohydrate ingestion [17]. Fats and proteins gence is fundamentally due to discordant results from become the sole energy source where reduced intracellu- inconsistencies in methodologies. Coordinating an agree- lar phosphate levels ensue [101, 102]. Upon recommenc- ment on the role of thiamine compounds in the critically ing carbohydrate nutrition insulin is secreted, stimulating ill can only be attained with evidence resulting from cellular uptake of serum electrolytes as well as the reacti- methods deemed fit for the purpose for thiamine quan- vation of numerous pathways, including glycolysis, pro- tification. This is achievable by the standardisation of moting uptake of serum thiamine and magnesium. Severe thiamine quantification practises including the imple- hypophosphataemia, hypokalaemia, hypomagnesaemia mentation of a certified reference material and reference and TD are the predominant pathophysiological outcomes method to which all current methods can be made trace- of these fluid shifts. However, these biochemical abnor- able to. malities are also present in sepsis, post-gastric surgery, The analytical techniques for thiamine compounds post-trauma as well as liver disease to name but a few pre- are discussed in this section in relation to the total testing senting conditions. process. This encompasses the pre-analytical, analyti- Clinical features of RFS are also non-specific but cal and post-analytical phases based on specific consid- can include respiratory failure, cardiac failure, arrhyth- erations in order to provide reliable thiamine results to mias, seizures, coma or sudden death [103]. A systematic the clinician. Important considerations across the total review of 27 RFS case studies highlighted the inconsisten- testing process for direct measurement of thiamine com- cies of reported signs and symptoms [104]. These reports pounds can be found in Table 2. emphasise the clinical spectrum of this disorder indicat- ing that not all cases will present with RFS from classi- cal malnutrition as some might be asymptomatic. Other clinical features that can be associated but not limited Pre-analytical to RFS, which are also common place in the ICU, include fluid retention and hypotension as well as heart, brain The pre-analytical phase consists of specimen handling and liver injury. Overall, its prevalence in the ICU is esti- such as sample collection and storage. Special considera- mated at 34% [105]. tions for thiamine analysis in this phase include the type As the human body does not store an excessive of sample collected, the collection tubes and the stability amount of thiamine, and RFS is known to be caused by of the sample. a state of nutritional starvation, patients in intensive care are already likely to have some degree of TD. As glycolysis is re-engaged when initiating refeeding the requirement Sample type of thiamine-dependent enzymes intensifies, exhausting any thiamine stores available, potentially leading to the The choice of matrix depends on the thiamine compound associated metabolic and neurological complications of of interest. TMP, TTP and thiamine are mainly concen- TD. Although there is no direct evidence that links thia- trated in serum, plasma and urine [110, 112, 113]. Thus, mine supplementation to improving the status of RFS, rec- these matrices should be used when quantifying one or ommendations from case reports, reviews and guidelines several of those analytes. Whole blood or washed eryth- all support thiamine administration prior to commencing rocytes can be used to measure TPP as it is predominantly and during the course of refeeding [2, 50, 51, 71, 106–108]. found in red blood cells. However, whole blood is con- In addition, parenteral or enteral refeeding is one of the sidered a more suitable sample type as it permits the main risk factors associated with RFS, along with alco- quantification of all thiamine compounds and washing holism, post-bariatric surgery, cancer and anorexia [101]. erythrocytes in saline can reduce the levels of the thia- These clinical situations make the diagnosis of RFS chal- mine compounds in these cells [114]. A study by Ihara lenging as individuals in intensive care do not always et al. [114] showed no difference in the values obtained present with typical biochemical profiles. from sodium and potassium ethylenediaminetetraacetic

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM 1660 Collie et al.: Vitamin B1 in critically ill patients: needs and challenges

Table 2: Important considerations for pre-analytical, analytical and post-analytical stages of the total testing process for the direct ­measurement of thiamine compounds.

Testing procedure Considerations

Pre-analytical – Sample stability can be an issue if not stored correctly prior to analysis as thiamine is light sensitive. Samples should be stored at a minimum of –70 °C and must be analysed within 48 h of sample preparation [45, 109] – The choice of sample type must be appropriate for the analyte being measured – Lack of standardisation introduces various storage conditions and stability parameters Analytical – No certified reference material or reference method available – No reliability on calibration accuracy – The lack of standardisation introduces various methodologies without traceability – Very few internal standard options are available – Thiamine is unstable in alkaline conditions, however the intensity of the fluorescence increases with alkalinity, plateauing at pH 9.0 [49] – Column performance and longevity can be reduced under highly alkaline conditions – Derivitisation requires hazardous oxidising chemicals such as cyanogen bromide and mercuric chloride [110]. Although there are alternate chemicals which are less dangerous to replace it, such as potassium ferrocyanide (or hexacyanoferrate), this consequently reduces the fluorescence intensity of the trichromes [14, 41] – Post-column derivatisation methods prolong column life and performance but requires a second chromatographic pump to mix the effluent with the derivitising agent to carry it to the detector – Pre-column derivitisation provides sharper peaks, better resolution but requires equipment for a gradient elution [49, 111]. This usually comes standard with modern HPLC equipment – Imprecision must fit within at least the minimum range of biological variation – No data on the commutability of calibrators and standards Post-analytical – No standardised common reference intervals including those for gender, ethnicity and age – No decision limit to determine therapeutic levels

acid (EDTA) or sodium heparin anticoagulant tubes traceability of results should all be considered. Thiamine when measuring from whole blood, plasma or washed compounds are present in nmol/L, thus require highly erythrocytes. sensitive methodologies to detect and quantify them reli- ably. Currently there are several analytical techniques for the measurement of thiamine status in the human body. Stability The transketolase activation test, or transketolase activity assay, and the pyrophosphate effect test are both indirect There is limited data on the stability of thiamine com- methods which monitor the status of TPP. Liquid chro- pounds in biological fluids or chemical standards. Mul- matographic techniques, with a range of separation and tiple publications describe the immediate storage of detection methods, are direct methods of quantification samples at either −20 or −70°C for the lysis of erythrocytes and are the most commonly described [45, 52, 111]. prior to analysis [45, 49, 109, 110, 115]. Standard thiamine compound solutions have also been reported to be stable from 1 to 5 months when stored in either a whole blood or Indirect measurement plasma matrix or in a 0.1-M hydrochloric acid (HCl) solu- tion. The storage conditions of these standards ranged Indirect measurement procedures for TPP only estimate from 4 to −70°C [45, 49, 109, 116, 117]. Overall, there is little the activity of the enzyme and are not quantitative. Tran- evidence to suggest that these storage conditions improve sketolase testing was first introduced in the mid-1950s as analyte stability. an in vitro method for evaluating TD [118]. It estimates the functional activity of the transketolase enzyme in erythrocytes by measuring the rate that hexoses are Analytical produced from the TPP-dependent pentose phosphate pathway; with a decreased result indicating TD [119]. For The analytical phase consists of all laboratory diagnostic the pyrophosphate effect test, a baseline measurement processing. Methodologies, internal standard use, cali- of transketolase activity is first recorded, then a second bration, biological variation, quality assurance, and the reading is taken after the addition of exogenous TPP. The

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM Collie et al.: Vitamin B1 in critically ill patients: needs and challenges 1661 difference is transcribed as a percentage increase in tran- of the participants utilised pre- or post-column deriviti- sketolase activity; with an increase of around 23% or more sation with florescence detection. Of the remaining EQA indicating TD. However, the upper limit of normal has participants, 2% stated the use of liquid chromatography been greatly debated with published values ranging from tandem mass spectrometry (LC-MS/MS), 3% were ambigu- 15.5% to 40% of increased activity [3, 115]. ous in their report and 6% did not state the methodology Further considerations are required when interpreting used [120, 121]. results from these tests as they are known to be affected by There are three recent published methods for the patients with diabetes or liver dysfunction as well as any measurement of thiamine, TMP and TPP using LC-MS/MS other condition impeding the synthesis of transketolase [109, 117, 131]. Tashirova et al. [117] measured thiamine in [14, 115]. Recent end-of-cycle reports from the RCPAQAP plasma using an acetonitrile-based mobile phase and a and SKML EQA vitamin programmes indicate that no C18 column. Puts et al. [109] measured TPP in whole blood participating laboratory is currently reporting TPP values using a methanol-based mobile phase and a C18 column. using indirect methodology [120, 121]. Lindhurst et al. [131] measured TMP and TPP from a post- mitochondrial supernatant using an acetonitrile based mobile phase and a Hypercarb porous graphite carbon Direct measurement column. These methods offer low limits of quantification and fast acquisition times (from 1.3 to 2.8 min). As men- The first HPLC method for the direct measurement of tioned earlier, this methodology is already present in some thiamine compounds was reported in the late 1970s [122]. clinical laboratories with two laboratories in the latest Unlike indirect methods, serum, plasma and urine can SKML cycle listing their instrumentation as LC-MS/MS for be used to measure other thiamine compounds beside TPP analysis. Although these methods only measure one TPP [123, 124]. This permits for the total thiamine status or two analytes in the thiamine family from a single matrix to be calculated or the biological availability of all thia- type, the introduction of tandem MS provides a much- mine compounds to be assessed. The limit of quantifica- needed platform for a thiamine quantification reference tion for all thiamine compounds has been reported to be method to be built upon. <3.0 nmol/L with linearity surpassing physiologically pos- The remainder of the analytical section will focus sible biological concentrations [45, 49, 111, 125]. on the quality of the direct measurement of thiamine The direct measurement of thiamine compounds compounds. by HPLC can be coupled with either fluorescence, ultra violet (UV), photo-diode array (PDA) detection and more recently, mass spectrometry (MS) [109, 116, 126]. Separa- Traceability tion can be accomplished using gradient or isocratic tech- niques, with or without ion-pairing chemicals and on a A reference point is required to ensure certainty that all range of columns including octadecyl carbon chain silica aspects of the method, including commercially available based (C18), polymer, or amino. Sample preparation pro- kits, are traceable through a documented unbroken chain cedures include cloud point extraction (using the cloud of calibrations from an established hierarchy [132]. The phenomena of surfactants), solid phase extraction and Joint Committee for Traceability in Laboratory Medicine derivitisation (which can occur pre- or post-column) [112, (JCTLM) is an established database of higher order refer- 127–129]. There are also an array of mobile phase solutions ence materials, measurement procedures and services. and chemicals for sample preparation with variations in Currently, there is no higher-order reference method, or pH levels from 2 to 13, even when identical chemicals are materials, listed in their database for thiamine and its utilised. An extensive flow chart of the differences and esters on their website [133]. This is a major limitation similarities for a wide variety of these published methods with all current thiamine quantitative methods. Without can be seen in Figure 3. certified reference materials there is no anchor for com- The use of HPLC with pre- or post-column deriviti- mercially available secondary calibrators to be traceable sation coupled with florescence detection is the most to. Calibration accuracy is a necessity for the comparisons popular method for the detection and quantification of of results from multiple sources. thiamine compounds. The principle of this method is the A top-down approach for standardisation of all thia- oxidation of the thiamine compounds in alkaline condi- mine compound methods in several matrix types (whole tions into highly florescent trichromes [32, 113, 130]. In blood, serum and urine) will allow for analytical quality the EQA programmes offered by SKML and RCPAQAP 89% goals to be established. This will support target setting for

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM 1662 Collie et al.: Vitamin B1 in critically ill patients: needs and challenges

Figure 3: Flow chart of published HPLC methods for the quantification of thiamine compounds. Further differences in methods not included in this flow chart extend to chemicals for washing erythrocytes, protein precipitation and ­additional steps in sample preparation. thiamine EQA programmes and assess bias. It can then accurate reference intervals or recommended levels for provide empirical evidence towards the effectiveness of healthy populations as well as expected values in critical thiamine supplementation, the establishment of more illness.

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM Collie et al.: Vitamin B1 in critically ill patients: needs and challenges 1663

In all, a reference anchor will allow for traceability, the performance of these kits can still be considered as a leading to standardisation and harmonisation between major factor in the variability in the results obtained. laboratory methods, and will support consistent clinical The inclusion of an internal standard in HPLC and interpretations and permit the bridging of results from LC-MS/MS methods is broadly recognised to be good multiple methods. laboratory practise as it improves assay imprecision as corrective action is taken for variations during sample preparation and instrument response. A survey of partici- Quality goals pants enrolled in the RCPAQAP whole-blood programme revealed that only one of nine respondents included an The biological variability (BV) of an analyte can be applied internal standard [22]. Furthermore, from our literature to set quality standards for a method, such as imprecision, search, only four publications reported the use of an inter- bias and total error. Fraser described this fine tuning of nal standard [109, 125, 128, 129]. goals from minimal, desirable and optimal [134]. The BV Non-isotopic internal standards, acetylaneurine, for TPP has been assessed with an individual CV of 4.8% pyrithiamine and theobromine, were used in the three and a between individual CV of 12.0% [135]. These equate HPLC-florescent published methods [125, 128, 129]. Only to desirable analytical goals of imprecision of 2.4%, bias of one of the three LC-MS/MS publications cited the use of an 3.2% and total error of 7.2%. Given the current variability isotopic labelled (d3) internal standard [109]. This is also between methods and lack of traceability, bias (and also the most recent LC-MS/MS publication. The absence of total error) comparison goals are difficult to interpret for internal standards in thiamine compound methodologies whole-blood TPP. is an unfortunate circumstance that may be caused by the Based on the BV data, the analytical imprecision difficulty in sourcing appropriate, commercially accessi- goals for TPP are <1.2% (optimal), 2.4% (desirable) and ble, chemicals. This lack of availability is highlighted by 3.6% (minimum). To compare actual performance to these each of these publications employing a different internal quality goals for imprecision, cycles of both EQA pro- standard compound. grammes were reviewed, these being the RCPAQAP Cycle It must be noted that the commercial kits discussed 33 (2016) and the SKML programme Cycle 20151 and 20152 above have very recently (2016) been updated to include (2015). As a whole, the programmes achieved a median an internal standard for TPP and have switched from a coefficient of variance (CV) of 7.5% (n = 925 reported post-column to a pre-column derivitisation step. Whether results) [120, 121]. Interestingly, 69% of the participants this is attributed to the company’s own decision from pub- indicated that they were using the same commercial lished information or from consumer demand to improve brand reagent and calibrator kits. Overall, these com- accuracy and precision is not known [22]. What can be mercial kits achieved a median CV of 6.7%. A summary acknowledged is that this is a necessity to improve the of the best, median and worst CV’s for each programme performance of these quantitative methods. cycle and for the commercial kit results only is displayed in Table 3. With a large percentage of the population using identical reagent and calibrator kits, the major vari- Commutability ables appear to be associated with instrumentation and sample preparation techniques. However, these commer- A further potential limitation not discussed in the litera- cial kits do not include an internal standard for TPP, thus ture is the commutability of thiamine compound reference

Table 3: Overview of the performance of the RCPAQAP and SKML TPP programmes for all participants and for participants only using com- mercially available kits.

CV% RCPAQAP Cycle 33, 2016 SKML Cycle 20151/2, 2015 Combined programmes

Whole programme Commercial kits Whole programme Commercial kits Whole programme Commercial kits

Best 1.1 1.5 3.1 3.1 1.1 1.5 Median 8.5 7.9 6.4 6.5 7.5 6.7 Worst 24.2 21.8 22.2 22.2 24.2 22.2 Participants 20 16 (80%) 63 41 (65%) 83 57 (69%)

Data presented have had outliers removed. These data have been reproduced with permission of the RCPAQAP chemical pathology pro- gramme and SKML.

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM 1664 Collie et al.: Vitamin B1 in critically ill patients: needs and challenges materials. This applies to any non-patient material i.e. cal- difference between the young and old populations ibration standards, quality controls and EQA materials. (p < 0.001; Student’s unpaired t-test) and suggested that A material is commutable when it behaves with the same lower TPP levels in the elderly were related more to age numerical relationship as is seen with patient samples. than a co-existing illness. This is an additional concern This commutability needs to be demonstrated experimen- when determining whether the lower thiamine compound tally. Our review of the literature and product information levels are due to age or the disease state. related to thiamine did not locate any data that certified It should be noted that the elderly participants in the the commutability of calibration standards or quality con- study described above were recruited from a general prac- trols used for thiamine compound quantification. tise, not a hospital setting. Furthermore, the blood TPP The RCPAQAP and SKML vitamin material is based on results from the elderly sub-group were 117–142 nmol/L the relevant matrix (e.g. whole blood) but the material has which is well within all published reference intervals of undergone a range of manipulations including multiple 63–229 nmol/L [41, 45, 47]. Although this study provides analyte additions, mixing procedures and lyophilisation, clear evidence that there is an age-related decrease in any of which may affect commutability of the analyte. As TPP, the elderly population did not appear to be clinically this EQA material has not been formally verified for com- thiamine deficient. A likely explanation would be that in mutability the organisers specify that the material cannot a community setting for the elderly the co-morbidities are be assumed to be commutable. functional illnesses and thus less severe than in a nursing, The commutability of calibration and also EQA mate- full-time care or hospital setting. This is supported by an rial is an important consideration not to be overlooked as observational study be Lee et al. [75] who found a 14% we move forward with improving the harmonisation and prevalence of TD in elderly patients admitted to hospital eventual standardisation of thiamine compound methods. from a nursing facility. This highlights that the predisposi- tion to reduced TPP levels in the elderly increases the risk of TD in critical care scenarios. Post-analytical Although the lack of agreement between analyti- cal methods continues, a higher level common reference The post-analytical phase consists of the generation of interval or decision limit for interpretation cannot be reports and interpretation. This involves specifically how introduced. Therefore, reference intervals and expected we interpret our thiamine compound result. To turn the values of disorders associated with TD cannot be gener- numerical value into a clinically meaningful result we ated nor broadly adopted. Further to such work, a critical need an appropriate decision point(s) such as a reference result limit for immediate clinical notification and action interval. A survey of RCPAQAP participants measuring could be identified. vitamins B1, conducted in 2012, found a general consen- sus towards reported reference intervals for vitamin B1 [22]. The survey concluded that the apparent harmonisa- tion of reference intervals was based upon their adoption Conclusions from the reagent kit inserts, as 66% of the responding participants utilised a common commercially available A better understanding of TD and RFS in the critically ill reagent. This is considered as the lowest level of harmo- is warranted as these appear to be disorders with a wide nisation by the Milan consensus statement for analytical spectrum and the likelihood of sub-clinical deficiencies. performance specifications [136]. There is currently no clear evidence as to which ICU popu- In the current literature, there are limited data on ref- lations are at risk of developing further complications from erence intervals based on gender or ethnicity [41, 45, 47]. TD and RFS or whether any are asymptomatic. Investiga- However, there is evidence of age-related differences [15]. tions into thiamine compound levels in ICU populations The elderly (>76 years of age) are known to have a reduced is essential to understand the role of thiamine in these total thiamine compound concentration that can be 21%– situations. A lack of consensus on the benefits of thiamine 26% lower than that of the middle aged [57, 137]. A longi- supplementation to prevent RFS and TD in the critically ill tudinal study of over 300 participants spanning 3 years by stems from absence of empirical evidence. A strong link Wilkinson et al. [15] examined a younger, healthy popula- between thiamine compound levels and TD and RFS is tion (mean age 41.5 years) of 100 volunteers and compared possible and may be likely but remains to be established. it to an elderly population (mean age 76 years) of 221 par- This evidence is paramount to provide consensus on thia- ticipants. The study concluded that there was a significant mine supplementation leading to ICU-relevant protocols

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM Collie et al.: Vitamin B1 in critically ill patients: needs and challenges 1665 ensuring that these potentially important deficiencies are 9. Lonsdale D. A review of the biochemistry, metabolism and addressed. clinical benefits of thiamin(e) and its derivatives. Evid Based Complement Alternat Med 2006;3:49–59. To successfully investigate and predict clinical out- 10. Bettendorff L, Wirtzfeld B, Makarchikov AF, Mazzucchelli G, Fré- comes associated with TD, several key foundations are dérich M, Gigliobianco T, et al. Discovery of a natural thiamine necessary; the capacity to quantify total thiamine status in adenine nucleotide. Nature Cell Biol 2007;3:211–2. biological fluids, universally accepted reference intervals 11. Hanninen SA, Darling PC, Sole MJ, Barr A, Keith ME. The preva- for healthy populations and expected values for disor- lence of thiamin deficiency in hospitalized patients with conges- ders associated with TD. For these to occur, standardisa- tive heart failure. J Am Coll Cardiol 2006;47:354–61. 12. Lee HS, Lee SA, Shin HS, Choi HM, Kim SJ, Kim HK, et al. A case tion of thiamine compound methodologies is required to of cardiac beriberi: a forgotten but memorable disease. Korean allow for traceable results. This is currently unachievable Circ J 2013;43:569–72. without an accredited reference method and material. 13. Faigle R, Mohme M, Levy M. Dry beriberi mimicking Guillain- Barre syndrome as the first presenting sign of thiamine defi- Acknowledgments: We thank Mrs Sabrina Koetsier from ciency. Eur J Neuro 2012;19:e14–5. 14. Lynch PL, Young IS. Determination of thiamine by high-performance­ the RCPAQAP and Dr Cas Weykamp from the SKML for liquid chromatography. J Chromatogr A 2000;881:267–84. their constant support, for providing de-identified cycle 15. Wilkinson TJ, Hanger HC, George PM, Sainsbury R. Is thiamine reports and for reviewing parts of this manuscript prior deficiency in elderly people related to age or co-morbidity? Age to submission. We would also like to thank the Australian Ageing 2000;29:111–6. government research training programme scholarship for 16. Donnino MW, Camey E, Cocchi MN, Barbash I, Chase M, Joyce N, awarding an Australian Postgraduate Award to J.T.B.C. et al. Thiamine deficiency in critically ill patients with sepsis. J Crit Care 2010;25:576–81. Author contributions: All the authors have accepted 17. Mallet M. Refeeding syndrome. Age Aging 2002;31:65–6. responsibility for the entire content of this submitted 18. Raziel A. Thiamine deficiency after bariatric surgery may lead to manuscript and approved submission. wernicke encephalopathy. Isr Med Assoc J 2012;14:692–4. Research funding: None declared. 19. McCray S, Walker S, Parrish CR. Much ado about refeeding. Pract Employment or leadership: None declared. Gastroenterol 2005;XXVIII:26–44. 20. Crook MA. Refeeding syndrome: problems with definition and Honorarium: None declared. management. Nutrition 2014;30:1448–88. Competing interests: The funding organization(s) played 21. SCIFINDER. MethodsNOW 2017. Available from: www.cas.org/ no role in the study design; in the collection, analysis, and methodsnow. interpretation of data; in the writing of the report; or in the 22. Hoad KE, Johnson LA, Woollard GA, Walmsley TA, Briscoe S, Jolly decision to submit the report for publication. LM, et al. Vitamin B1 and B6 method harmonization: comparison of performance between laboratories enrolled in the RCPA Qual- ity Assurance Program. Clin Biochem 2013;46:772–6. 23. SKML. Personal communication. Dutch Foundation for Quality References Assessment in Medical Laboratories, 2016. 24. Carpenter KJ, Sutherland B. Eijkman’s contribution to the discov- 1. Kohlmeire M. Water-soluble vitamins and nonnutrients. Nutrient ery of vitamins. J Nutr 1995;125:155–63. metabolism: structures, functions and genes, 2nd ed. Oxford, UK: 25. Eijkman C. Report of the investigations carried out in the labora- Academic Press, 2015:580–9. tory of pathology and bacteriology, Weltevreden, during the year 2. Sriram K, Manzanares W, Joseph K. Thiamine in nutrition therapy. 1895. VI. Polyneuritis in chickens. New contributions to the etiol- Nutr Clin Pract 2012;27:41–50. ogy of the disease. 1896. Nutr Rev 1990;48:243–6. 3. Ahmed M, Azizzi-Namini P, Yan AT, Keith M. Thiamin deficiency 26. Eijkman C. Report of the investigations carried out in the labora- and heart failure: the current knowledge and gaps in literature. tory of pathology and bacteriology, Weltevreden, during the year Heart Fail Rev 2015;20:1–11. 1889. VI. Polyneuritis in chickens. 1890. Nutr Rev 1990;48:242–3. 4. Sechi GP, Serra A. Wernicke’s encephalopathy: new clinical set- 27. Nobelprize.org. The Nobel Prize in Physiology or Medicine 1929 tings and recent advances in diagnosis and management. Lancet 2014. Available from: http://www.nobelprize.org/nobel_prizes/ Neurol 2007;6:442–54. medicine/laureates/1929/. 5. Schattner A. An unlikely culprit – the many guises of thiamine 28. Sutherland B. A micronutrient deficiency in chickens (Grijns, deficiency. Am J Emergency Med 2013;31:635.e5–6. 1896–1901). J Nutr 1997;127:1023S–5. 6. Singleton CK, Martin PR. Molecular mechanisms of thiamin utili- 29. Carpenter KJ. The discovery of thiamin. Ann Nutr Metab zation. Curr Mol Med 2001;1:197–207. 2012;61:219–23. 7. Lonsdale D. Thiamin(e): the spark of life. In: Stranger O, editor. 30. Semba RD. The discovery of the vitamins. Int J Vitam Nutr Res Water soluble vitamins: clinical research and future application. 2012;82:310–5. 56: Subcell Biochem, New York, USA: Springer, 2012:199–227. 31. Funk C. The etiology of the deficiency diseases. Beri-Beri, poly- 8. Manzetti S, Zhang J, Van Der Spoel D. Thiamin function, neuritis in birds, epidemic dropsy, scurvy, experimental scurvy ­metabolism, uptake and transport. Biochemistry 2014;53: in animals, infantile scurvy, ship beri-beri, pellagra. J State Med 821–35. 1912;20:341–68.

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM 1666 Collie et al.: Vitamin B1 in critically ill patients: needs and challenges

32. Jansen R. Identity of vitamin B2 and flavine and the nomencla- 53. Peters RA. The biochemical lesion in vitmain B1 deficiency appli- ture of vitamins. Nature 1935;135:267. cations of modern biochemical analysis in its diagnosis. Lancet 33. Drummond JC. The nomenclature of the so-called accessory food 1936;227:1161–5. factors (vitamin). Biochem J 1920;14:660. 54. Racker E, De La Haba G, Leder IG. Thiamine pyrophosphate, a 34. Williams RR. Structure of vitamin B1. Am J Clin Soc coenzyme of transketolase. J Am Chem Soc 1953;75:1010–1. 1936;58:1063–4. 55. McCormick DB. The co-enzyme function of thiamin (Peters et al., 35. Williams RR. Synthesis of vitamin B1. Am J Clin Soc 1929–1937). J Nutr 1997;127:1038S–9. 1936;58:1504–5. 56. Bettendorff L. Thiamine in excitable tissues: reflections on a 36. Williams RR. The chemistry and biological significance of thia- non-cofactor role. Met Brain Dis 1994;9:183–209. min. Science 1938;87:559–63. 57. Pan X, Fei G, Lu J, Jin L, Pan S, Chen Z, et al. Measurement of 37. Harris LJ, Leong PC. Vitamins in human nutrition the excretion blood thiamine metabolites for Alzheimer’s disease diagnosis. of vitamin B1 in human urine and its dependence on the dietary EBioMedicine 2016;3:155–62. intake. Lancet 1936;227:886–94. 58. Bunik VI, Tylicki A, Lukashev NV. Thiamin diphosphate-depend- 38. Food and Nutrition Board, Institute of Medicine. Dietary refer- ent enzymes: from enzymology to metabolic regulation, drug ence intakes for thiamin, riboflavin, niacin, vitamin B6, folate, design and disease models. FEBS J 2013;280:6412–42. vitamin B12, Pantothenic acid, biotin, and choline. Washington, 59. Pekovich SR, Martin PR, Singleton CK. Thiamine defi- DC: National Academy Press, 2000. ciency decreases steady-state transketolase and pyruvate 39. Yates AA, Schlicker SA, Suitor CW. Dietary Reference ­dehydrogenase but not alpha-ketoglutarate dehydrogenase intakes: the new basis for recommendations for calcium and mRNA levels in three human cell types. J Nutr 1998;128:683–7. related nutrients, B vitamins, and choline. J Am Diet Assoc 60. Mkrtchyan G, Aleshin V, Parkhomenko Y, Kaehne T, Di Salvo ML, 1998;98:699–706. Parroni A, et al. Molecular mechanisms of the non-coenzyme 40. Gangolf M, Czerniecki J, Radermecker M, Detry O, Nisolle M, action of thiamin in brain: biochemical, structural and pathway Jouan C, et al. Thiamine status in humans and content of phos- analysis. Sci Rep 2015;5:12583. phorylated thiamine derivatives in biopsies and cultured cells. 61. Bettendorff L, Wins P. Biological functions of thiamine deriva- PLoS One 2010;5:e13616. tives: focus on non-coenzyme roles. OA Biochem 2013;1:1–10. 41. Tallaksen CM, Bohmer T, Bell H, Karlsen J. Concomitant determi- 62. Gibson GE, Zhang H. Interactions of oxidative stress with thia- nation of thiamin and its phosphate esters in human blood and mine homeostasis promote neurodegeneration. Neurochem Int serum by high-performance liquid chromatography. J Chroma- 2002;40:493–504. togr 1991;564:127–36. 63. Cruickshank AM, Telfer AB, Shenkin A. Thiamine deficiency in 42. Bettendorff L. Thiamine triphosphate: a ubiquitous mole- the critically ill. J Intensive Care Med 1988;14:384–7. cule in search of a physiological role. Metab Brain Dis 64. Shimon I, Almog S, Vered Z, Seligmann H, Shefi M, Peleg E, et al. 2014;29:1069–82. Improved left ventricular function after thiamine supplementa- 43. Bâ A. Metabolic and structural role of thiamine in nervous tis- tion in patients with congestive heart failure receiving, long- sues. Cell Mol Neurobiol 2008;28:923–31. term furosemide therapy. Am J Med 1995;98:485–90. 44. Subramanian VS, Nabokina SM, Lin-Moshier Y, Marchant JS, 65. Hazell AS, Butterworth RF. Update of cell damage mechanisms Said HM. Mitochondrial uptake of thiamin pyrophosphate: phys- in thiamine deficiency: focus on oxidative stress, excitotoxicity iological and cell biological aspects. PLoS One 2013;8:e73503. and inflammation. Alcohol Alcohol 2009;44:141–7. 45. Lu J, Frank EL. Rapid HPLC measurement of thiamine and its 66. Gibson GE, Hirsch JA, Fonzetti P, Jordan BD, Cirio RT, Elder phosphate esters in whole blood. Clim Chem 2008;54:901–6. J. Vitamin B1 (thiamine) and dementia. Ann NY Acd Sci 46. Chromsystems. Instruction manual for the HPLC analysis of 2016;1367:21–30. vitamin B1 in whole blood. 2008:1–18. 67. Moskowitz A, Graver A, Giberson T, Berg K, Liu X, Uber A, et al. 47. Ihara H, Hirano A, Wang L, Okada M, Hashizume N. Reference The relationship between lactate and thiamine levels in patients values for whole blood thiamine and thiamine phosphate esters with diabetic ketoacidosis. J Crit Care 2014;29:182e5–8. in Japanese adults. J Anal Bio-Sci 2005;28:241–6. 68. Amrein K, Ribitsch W, Otto R, Worm HC, Stauber RE. Severe 48. Brunnekreet JW, Eidhof H, Gerrits J. Optimized determination lactic acidosis reversed by thiamine within 24 hours. Crit Care of trichrome derivatives of thiamine and thiamine phosphates 2011;15:457. in whole blood by reverse-phase liquid chromatorgaphy with 69. Sum SS, Brody R. The link between thiamine supplementation precolumn derivitisation. J Chromatogr 1989;491:89–96. and treatment of wernicke-korsakoff syndrome in patients with 49. Van Landeghem BA, Puts J, Claessens HA. The analysis of thiamin alcoholism. Top Clin Nutr 2012;27:334–8. and its derivaties in whole blood samples under high pH condi- 70. Mancinelli R, Ceccanti M. Biomarkers in alcohol misuse: their tions of the mobile phase. J Chromatogr B 2005;822:316–21. role in the prevention and detection of thiamine deficiency. 50. Singer P, Berger MM, Van den Berghe G, Biolo G, Calder P, Alcohol Alcohol 2009;44:177–82. Forbes A, et al. ESPEN guidelines on parenteral nutrition: inten- 71. Ziegler TR. Parenteral nutrition in the critically Ill patient. N Engl sive care. Clin Nutr 2009;28:387–400. J Med 2009;361:1088–97. 51. National Collaborating Centre for Acute Care. Nutrition support 72. Donnino MW, Andersen LW, Chase M, Berg KM, Tidswell M, for adults: oral nutrition support, enteral tube feeding and par- Giberson T, et al. Randomized, double-blind, placebo-controlled enteral nutrition. National Collaborating Centre for Acute Care trial of thiamine as a metabolic resuscitator in septic shock: a (UK), 2006. pilot study. Crit Care Med 2016;44:360–7. 52. Frank LL. Thiamin in clinical practice. J Parenter Enteral Nutr 73. Seligmann H, Halkin H, Rauchfleisch S, Kaufmann N, Tal R, 2015;39:503–20. Motro M, et al. Thiamine deficiency in patients with congestive

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM Collie et al.: Vitamin B1 in critically ill patients: needs and challenges 1667

heart failure receiving long-term furosemide therapy: a pilot a double-blinded, randomised controlled pilot study. Eur J study. Am J Med 1991;91:151–5. Anaesthesiol 2015;32:543–8. 74. Romanski SA, McMahon MM. Metabolic acidosis and thiamine 90. Jhala SS, Hazell AS. Modeling neurodegenerative disease patho- deficiency. Mayo Clin Proc 1999;74:259–63. physiology in thiamine deficiency: consequences of impaired 75. Lee DC, Chu J, Satz W, Silbergleit R. Low plasma thiamine levels oxidative metabolism. Neurochem Int 2011;58:248–60. in elder patients admitted through the emergency department. 91. Heroux M, Rao R, Lavoie J, Richardson S, Butterworth RF. Acad Emerg Med 2000;7:1156–9. Alterations of thiamine phosphorylation and of thiamine- 76. Leite HP, De Lima LF. Thiamine (vitamin b1) deficiency in inten- dependent enzymes in Alzheimer’s disease. Metab Brain Dis sive care: physiology, risk factors, diagnosis, and treatment. 1996;11:81–8. Diet and Nutrition in Critical Care. New York, USA: Springer 92. Mastrogiacomo F, Bettendorff L, Grisar T, Kish SJ. Brain thia- Sciences, 2015:959–72. mine, its phosphate esters, and its metabolizing enzymes in 77. Mukunda BN. Lactic acidosis caused by thiamine deficiency in a Alzheimer’s disease. Ann Neurol 1996;39:585–91. pregnant alcoholic patient. Am J Med Sci 1999;317:261–2. 93. Todd KG, Butterworth RF. Mechanisms of selective neuronal 78. Madl C, Kranz A, Liebisch B, Traindl O, Lenz K, Druml W. Lactoc cell death due to thiamine deficiency. Ann NY Acad Sci acidosis in thiamine deficiency. Clin Nutr 1993;12:108–11. 1999;893:404–11. 79. Marik PE, Khangoora V, Rivera R, Hooper MH, Catravas J. 94. Gibson GE, Calingasan NY, Baker H, Gandy S, Sheu KF. Impor- ­Hydrocortisone, vitamin C and thiamine for the treatment of tance of vascular changes in selective neurodegeneration with severe sepsis and septic shock: a retrospective befroe-after thiamine deficiency. Ann NY Acd Sci 1997;826:516–9. study. Chest 2016, pii: S0012-3692(16)62564-3. 95. Calingasan NY, William WJ, Park LC, Uchida L, Gibson GE. doi: 10.1016/j.chest.2016.11.036. [Epub ahead of print]. Oxidative stress is associated with region-specific neuronal 80. Moskowitz A, Andersen LW, Cocchi MN, karlsson M, Patel PV, death during thaimine deficiency. J Neuropath Exp Neur Donnino MW. Thiamine as a renal protective agent in septic 1999;58:946–58. shock. A secondary analysis of a randomized, double-blind, 96. Abdou E, Hazell AS. Thiamine deficiency: an update of patho- placebo-controlled trial. Ann Am Thorac Soc 2017. physiologic mechanisms and future therapeutic considera- doi: 10.1513/AnnalsATS.201608-656BC. [Epub ahead of print]. tions. Neurochem Res 2014;40:353–61. 81. Costa NA, Gut AL, de Souza Dorna M, Pimentel JA, Cozzolino SM, 97. Gibson GE, Park LC, Sheu KF, Blass JP, Calingasan NY. The Azevedo PS, et al. Serum thiamine concentration and oxidative α-ketoglutarate dehydrogenase complex in neurodegenera- stress as predictors of mortality in patients with septic shock. J tion. Neurochem Int 2000;36:97–112. Crit Care 2014;29:249–52. 98. Koike H, Misu K, Hattori N, Ito S, Ichimura M, Ito H, et al. Post- 82. Schoenenberger AW, Schoenenberger-Berzins R, Auf der Maur gastrectomy polyneuropathy with thiamine deficiency. J Neurol C, Suter PM, Vergopoulos A, Erne P. Thiamine supplementation Neurosurg Psychiatry 2001;71:357–62. in symptomatic chronic heart failure: a randomized, double- 99. Ishibashi S, Yokota T, Shiojiri T, Matunaga T, Tanaka H, Nishina blind, placebo-controlled, cross-over pilot study. Clin Res K, et al. Reversible acute axonal polyneuropathy associated Cardiol 2012;101:159–64. with Wernicke-Korsakoff syndrome: impaired physiological 83. Berg K, Gautam S, Salciccioli J, Giberson T, Saindon B, Donnino nerve conduction due to thiamine deficiency? J Neurol Neuro- MW. Intravenous Thiamine is associated with increased oxygen surg Psychiatry 2003;74:674–6. consumption in critically Ill patients with preserved cardiac 100. Makarchikov AF. Vitamin B1: metabolism and functions. index. Annals ATS 2014;11:1597–601. Biomed Khim 2009;3:116–28. 84. Härdig L, Daae C, Dellborg M, Kontny F, Bohmer T. Reduced thia- 101. Byrnes MC, Stangenes J. Refeeding in the ICU: an adult mine phosphate, but not thiamine diphosphate in erythrocytes and pediatric problem. Curr Opin Clin Nutr Metab Care in elderly patients with congestive heart failure treated with 2011;14:186–92. furosemide. J Intern Med 2000;247:597–600. 102. Mehanna HM, Moledina J, Travis J. Refeeding syndrome: 85. Yue QY, Beermann B, Lindstrom B, Nyquist O. No difference in what it is, and how to prevent and treat it. Br Med J blood thiamine diphosphate levels between Swedish Caucasian 2008;336:1495–8. patients with congestive heart failure treated with furosemide 103. Hearing SD. Refeeding syndrome is underdiagnosed and and patients without heart failure. J Intern Med 1997;242:491–5. undertreated, but treatable. Br Med J 2004;328:908–9. 86. Donnino MW, Cocchi MN, Smithline H, Carney E, Chou PP, Salcic- 104. Skipper A. Refeeding syndrome or refeeding hypophos- coli J. Coronary artery bypass graft surgery depletes plasma phatemia: a systematic review of cases. Nutr Clin Pract thiamine levels. Nutrition 2010;26:133–6. 2012;27:34–40. 87. Berger MM, Soguel L, Shenkin A, Revelly J, Pinget C, Baines M, et al. 105. Boateng AA, Sriram K, Meguid MM, Crook MA. Refeeding syn- Influence of early antioxidant supplements on clinical evolution drome: treatment considerations based on collective analysis and organ function in critically ill cardiac surgery, major trauma, of literature case reports. Nutrition 2010;26:156–67. and subarachnoid hemorrhage patients. Crit Care 2008;12:R101. 106. Hershkowitz E, Reshef A, Munich O, Yosefi B, Markel A. 88. Andersen LW, Holmberg MJ, Berg KM, Chase M, Cocchi MN, Thiamine deficiency in self-induced refeeding syndrome, an Sulmonte C, et al. Thiamine as an adjunctive therapy in cardiac undetected and potentially lethal condition. Cas Rep Med surgery: a randomized, double-bline, placebo-controlled, phase 2014;605707:6. II trial. Crit Care 2016;20:1–11. 107. Crook MA, Hally V, Panteli JV. The importance of the refeeding 89. Luger M, hiesmayr M, Köpple P, Sima B, Ranz I, Wiess C, syndrome. Nutrition 2001;17:632–7. et al. Influence of intravenous thiamine supplementation 108. Stanga Z, Brunner A, Leuenberger M, Grimble RF, Shenkin A, on blood lactate concentration prior to cardiac surgery: Allison SP, et al. Nutrition in clinical practice – the refeeding

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM 1668 Collie et al.: Vitamin B1 in critically ill patients: needs and challenges

syndrome: illustrative cases and guidelines for prevention and 123. Bohrer D, do Nascimento PC, Ramirez AG, Medndonca JK, de treatment. Eur J Clin Nutr 2008;62:687–94. Carvahlo LM, Promblum SC. Determination of thiamine in blood 109. Puts J, de Groot M, Haex M, Jakobs B. Simultaneous serum and urine by high-performance liquid chromatography determination of underivatized vitamin B1 and B6 in whole with direct injection and post-column derivatization. Micro- blood by reversed phase ultra high performance liquid chem J 2004;78:71–6. chromatography tandem mass spectrometry. PLoS One 124. Ihara H, Matsumoto T, Kakinoki T, Shino Y, Hashimoto R, Hashi- 2015;10:e0132018. zume N. Estimation of vitamin B1 excretion in 24-hr urine by 110. Losa R, Sierra MI, Fernandez A, Blanco D, Buesa JM. Determi- assay of first-morning urine. J Clin Lab Anal 2008;22:291–4. nation of thiamine and its phosphorylated forms in human 125. Lynch PL, Trimble ER, Young IS. High-performance liquid plasma, erythrocytes and urine by HPLC and fluorescence chromatographic determination of thiamine diphosphate in detection: a preliminary study on cancer patients. J Pharm erythrocytes using internal standard methodology. J Chroma- Biomed Anal 2005;37:1025–9. togr B 1997;701:120–3. 111. Mancinelli R, Ceccanti M, Guiducci MS, Sasso GF, Sebastiani G, 126. Körner RW, Vierzig A, Roth B, Müller C. Determination of thia- Attilia ML, et al. Simultaneous liquid chromatographic assess- min diphosphate in whole blood samples by high-performance ment of thiamine, thiamine monophosphate and thiamine liquid chromatography-a method suitable for pediatric diag- diphosphate in human erythrocytes: a study on alcoholics. J nostics. J Chromatogr B 2009;877:1882–6. Chromatogr B 2003;789:355–63. 127. Samaddar P, Sen K. Cloud point extraction: a sustainable 112. Heydari R, Elyasi NS. Ion-pair cloud-point extraction: a new method of elemental preconcentration and speciation. J Ind method for the determination of water-soluble vitamins in Eng Chem 2014;20:1209–19. plasma and urine. J Sep Sci 2014;37:2724–31. 128. Chatzimichalakis PF, Samanidou VF, Verpoorte R, Papadoyannis 113. López-Flores J, Fernández-De Córdova ML, Molina-Diaz A. IN. Development of a validated HPLC method for the determina- Implementation of flow-through solid phase spectroscopic tion of B-complex vitamins in pharmaceuticals and biological transduction with photochemically induced fluorescence: fluids after solid phase extraction. J Sep Sci 2004;27:1181–8. determination of thiamine. Anal Chim Acta 2005;535:161–8. 129. Basiri B, Sutton JM, Hanberry BS, Zastre JA, Bartlett MG. 114. Ihara H, Matsumoto T, Shino Y, Hashizume N. Assay values Ion pair liquid chromatography method for the determina- for thiamine or thiamine phosphate esters in whole blood tion of thiamine (vitamin B1) homeostasis. Biomed Chrom do not depend on the anticoagulant used. J Clin Lab Anal 2016;30:35–41. 2005;19:205–8. 130. Bettendorff L, Peeters M, Jouan C, Wins P, Schoffeniels E. 115. Talwar D, Davidson H, Cooney J, St JO’Reilly D. Vitamin Determination of thiamin and its phosphate esters in cultured B1 ­status assessed by direct measurement of thiamin pyroph- neurons and astrocytes using an ion-pair reversed-phase high- osphate in erythrocytes or whole blood by HPLC: comparison performance liquid chromatographic method. Anal Biochem with erythrocyte transketolase activation assay. Clin Chem 1991;198:52–9. 2000;46:704–10. 131. Lindhurst MJ, Fiermonte G, Song S, Struys E, De Leonardis F, 116. Viñas P, López-Erroz C, Balsalobre N, Hernández-Córdoba M. Schwartzberg PL, et al. Knockout of Slc25a19 causes mito- Comparison of ion-pair and amide-based column reverse- chondrial thiamine pyrophosphate depletion, embryonic phase liquid chromatography for the separation of thiamine- lethality, CNS malformations, and anemia. Proc Natl Acad USA related compounds. J Chromatogr B 2001;757:301–8. 2006;103:15927–32. 117. Tashirova OA, Ramenskaya GV, Vlasov AM, Khaitov MR. Devel- 132. Greaves RF. A Guide to harmonisation and standardisation of opment and validation of an LC/MS methods for quantitative measurands determined by liquid chromatography – mass determination of thiamine in blood plasma. Pharm Chem J spectormetry in routine clinical biochemistry. Clin Biochem Rev 2013;46:742–4. 2012;33:123–32. 118. Smeets EH, Muller H. A NADH-dependent transketolase assay 133. Bureau International des Poids et Mesures. JCTLM database: in erythrocyte hemolysates. Clin Chem Acta 1971;33:379–86. Laboratory medicine and in vitro diagnostics. Available from: 119. Nordentoft M, Timm S, Hasselbalch E, Roesen A, Gammeltoft S, http://www.bipm.org/jctlm/. Hemmingsen R. Thiamine pyrophosphate effect and erythro- 134. Fraser CG. Biological variation: from principle to practice. cyte transketolase activity during severe alcohol withdrawal Washington: AACC Press, 2001. syndrome. Acta Psychiatr Scand 1993;88:80–4. 135. Talwar DK, Azharuddin MK, Williamson C, Teoh YP, McMillan 120. RCPA Quality Assurance Programs. Thiamine pyrophosphate DC, O’Reilly DS. Biological variation of vitamins in blood of end of cylce 33 report 2016. Available from: http://www. healthy individuals. Clin Chem 2005;51:2145–50. rcpaqap.com.au/chempath. 136. Sandberg S, Fraser CG, Horvath AR, Jansen R, Jones G, Ooster- 121. SKML. End of cycle vitamin B1 reports 20151 and 20152. Dutch huis W, et al. Defining analytical performance specifications: Foundation for Quality Assessment in Medical Laboratories, consensus statement from the 1st strategic conference of the 2016. European Federation of Clinical Chemistry and Laboratory 122. Roser RL, Andrist AH, Harrington WH, Naito HK, Lonsdale D. Medicine. Clin Chem Lab Med 2015;53:833–5. Determination of urinary thiamine by high-pressure liquid 137. Bettendorff L, Mastrogiacomo F, Kish SJ, Grisar T. Thiamine, chromatography utilizing the thiochrome fluorescent method. thiamine phosphates, and their metabolizing enzymes in J Chromatogr 1978;146:43–53. human brain. J Neurochem 1996;66:250–8.

Brought to you by | Monash University Library Authenticated Download Date | 2/19/20 5:25 AM