Vitamin D and the Parenteral Nutrition Patient

GASTROENTEROLOGY 2009;137:S79 S91 Vitamin D and the Parenteral Nutrition Patient HECTOR F. DELUCA Department of Biochemistry, University of Wisconsin Madison Vitamin D is a prohormone produced in the skin epidermis when irradiated with sunlight or ultraviolet light B. It is also absorbed from food or supplements. Vitamin D must be converted to 25-hydroxyvitamin D 3 (circulating form) and finally to the hormone, 1a,25- dihydroxyvitamin D 3, before it can function. This hormone acts through a single nuclear receptor. The details of these conversions and molecular biology of the action of vitamin D 3 are summarized. The physiologic functions of vitamin D have been expanded beyond the mineralization of the skeleton to include modulation of the immune system, terminal differentiation in several tissues, suppression of malignant cells, and anabolic activity in the skeleton and in the renal cardiovascular system. Epidemiologic studies have associated vitamin D deficiency with an increased risk of colorectal and breast cancers and an increased risk of autoimmune diseases, such as multiple sclerosis, type 1 diabetes, and cardiovascular events. Thus, vitamin D is essential not only for the skeleton but also many other organ systems. Recommendations for 25-hydroxyvitamin D 3 levels for PN patients are presented. Vitamin D and Health Over the past 5 years, there has been an enormous rebirth of interest in vitamin D from the public health point of view. The basis for the current excitement harkens back to the discovery of the vitamin D endocrine system. Before that discovery, the role of vitamin D in animal and human physiology was deemed to be the mineralization of the skeleton and hence prevention of the disease rickets in children and osteomalacia in the adult. 1 In fact, some textbooks indicated that vitamin D is not required after childhood because the growth of the skeleton had ceased. 2 This, of course, is false; vitamin D is required throughout life for skeletal health because it plays a major role in the remodeling system in addition to the requirement for mineralization that repairs the skeleton. 3 5 Vitamin D Metabolism The delineation at the physiologic level of the mechanism of action of vitamin D occurred with the discovery that vitamin D 3 itself is biologically inactive 3 and must be metabolized first in the liver to 25-hydroxyvitamin D 3 (25-OH-D 3 ) and subsequently in the kidney to 1,25-dihydroxyvitamin D 3 (1,25-(OH) 2 D 3 ) before it can carry out its well-known functions in calcium and phosphorus metabolism resulting in the healing of rickets and osteomalacia. 4,5 There is little doubt then that vitamin D plays an essential role in bone health, and in subsequent sections, the molecular mechanisms whereby it brings these actions about are discussed. The Vitamin D Receptor Continued pursuit of the vitamin D endocrine system based in the kidney revealed that the vitamin D hormone, that is, 1,25-(OH) 2 D 3 functions primarily, if not exclusively, through a nuclear receptor. 6,7 This vitamin D receptor (VDR) was first described in 1975 8 and 1976, 9 and several laboratories attempted isolation of the VDR for approximately a decade. Importantly, radiolabeled 1,25-(OH) 2 D 3 when administered to animals was found to localize in the nuclei in the target organs, suggesting a nuclear mechanism. 10 Although the receptor was extensively purified by 2 laboratories, it was never purified to homogeneity. These laboratories used the partially purified receptor to produce monoclonal antibodies directed to the VDR from 1982 to 1986. 11,12 This resulted in the cloning of the VDR in rats in the DeLuca group and in human cells by the O Malley, Haussler, and Pike group in 1988. 13,14 Of great interest was the finding, using either radiolabeled ligand or the antibodies to the receptor, that the receptor is found in many tissues not in any way associated with calcium and phosphorus homeostasis required for bone health and growth. 15 17 This led to the speculation that vitamin D may play other important roles in addition to skeletal mineralization and bone remodeling. Expanded View of Vitamin D Function The discovery that VDR in tissues not involved with calcium and bone ultimately resulted in discovery of new functions of vitamin D. Clearly, vitamin D plays a Abbreviations used in this paper: 1, 25-(OH) 2 D 3,1,25-dihydroxyvitamin D 3 ; 25-OH-D 3, 25-hydroxyvitamin D 3 ; CYP450, cytochrome P-450; PN, parenteral nutrition; PTH, parathyroid hormone; TPN, total parenteral nutrition; VDR, vitamin D receptor; VDRE, vitamin D-responsive element. 2009 by the AGA Institute 0016-5085/09/$36.00 doi:10.1053/j.gastro.2009.07.075

S80 HECTOR F. DELUCA GASTROENTEROLOGY Vol. 137, No. 5 significant role in the immune system, cellular differentiation, the development of the giant osteoclast, and some circumstances, as a regulator of cell cycle being useful in the treatment of proliferative diseases such as cancer and psoriasis. These findings have led epidemiologists to examine the relationship of vitamin D to a wide number of abnormalities giving rise to the idea that vitamin D may be very important in reducing the incidence of autoimmune disease, colorectal cancer, breast cancer, and heart attacks. 18 20 This, therefore, signals the importance of vitamin D for all subjects including those relying on parenteral nutrition (PN). This presentation is devoted to that idea. The Production and Metabolism of Vitamin D Although vitamin D was originally discovered as a vitamin by McCollum et al, 21 our current knowledge reveals very clearly that vitamin D is not a vitamin but a prohormone that is normally produced in skin under the influence of ultraviolet light. 22 As illustrated in Figure 1, the production of vitamin D in skin is from the cholesterol metabolite, 7-dehydrocholesterol, which absorbs ultraviolet light between 282 and 310 nm producing previtamin D, a compound that has little biological activity, but that slowly isomerizes to the vitamin D form. 23 Vitamin D thus produced in the skin is transported to the liver where it begins it metabolic alterations for function. In the liver, it undergoes 25-hydroxylation primarily due to a cytochrome P-450 (CYP450) enzyme yet to be positively identified but believed to be the CYP450 2R1. 24,25 However, that enzyme has not yet been knocked out and, therefore, we cannot yet conclude that it is the enzyme responsible for the initial activation step of vitamin D to produce 25-OH-D 3, which is the circulating form of vitamin D. Although another candidate has been suggested as the CYP450 that activates vitamin D to 25-OH-D 3, it is clear that that enzyme, that is, the CYP27A1, is not required for the activation of vitamin D as revealed by a null mutant mouse model. 26 25-OH-D 3 is the form of vitamin D that is measured to determine the vitamin D status of any patient. However, 25-OH-D 3 under normal circumstances is not biologically active, but must be metabolized further to its final hormonal form, 1,25-(OH) 2 D 3. A biological activity due exclusively to 25-OH-D has not been found; thus, it is currently believed that 1,25-(OH) 2 D 3 is the functional form of vitamin D. The process producing 1,25-(OH) 2 D 3 takes place primarily, if not exclusively, in the kidney, particularly in the proximal convoluted tubule cells. 27,28 The enzyme responsible for this conversion is the CYP27B1. It has been cloned, and has been rendered null by mutation in mice. Most important is that vitamin D-dependency rickets type I, which was discovered in 1953 by Prader represents a defect in this enzyme 29 and, Figure 1. The production of vitamin D in skin and its conversion through the liver to the major circulating form, 25-OH-D 3, and its conversion to the final vitamin D hormone, 1,25-(OH) 2 D 3, in the proximal convoluted tubule cells of the kidney. in fact, many kindreds have now been identified illustrating the mutations resulting in this form of disease. 30,31 This disease can be cured by large amounts of vitamin D or 25-OH-D 3 but only physiologically with 1,25-(OH) 2 D 3. These mutations themselves illustrate the vitamin D endocrine system and the essential role played by the CYP27B1. Vitamin D-dependency rickets type I and the 3 different reports of the null mutant mice 28,32,33 prove conclusively that the vitamin D endocrine system is true and is certainly functional in man. There is considerable interest in the possibility that small amounts of the CYP27B1 might exist in other cells and tissues, giving rise to an autocrine/paracrine function in the vitamin D system. 34 Although this may still be true, conclusive in vivo evidence is lacking. Certainly cells and tissues explanted in vitro can produce 1,25-(OH) 2 D 3. Furthermore, in certain disease states, extrarenal production of 1,25-(OH) 2 D 3 is clearly evident, as, for example, in

November Supplement 2009 VITAMIN D AND THE PN PATIENT S81 sarcoidosis and certain malignancies. 34,35 However, in normal animals, proof is lacking that there is an autocrine/paracrine function of the CYP27B1. In nephrectomized rats given high specific activity radiolabeled 25- OH-D 3, no 1,25-(OH) 2 D 3 has been detected in blood, whereas if the same degree of uremia is produced by ligating the ureters, 1,25-(OH) 2 D 3 is clearly produced. 36,37 More recently, the CYP27B1 has been replaced with a -galactosidase gene under the control of the CYP27B1 promoter. 28 In normal mice, the only place that this enzyme is expressed is in the proximal convoluted tubule cells or in the placenta, sites that clearly have been shown to possess the 1 -hydroxylase by biochemical means. The jury is not in regarding the autocrine/paracrine function of vitamin D; so far, conclusive evidence in vivo is lacking. Fortunately, we know from the work of Sir Edward Mellanby in 1919 that vitamin D can be efficiently absorbed, as confirmed many times because orally administered vitamin D is quite effective. 4 6,38 It is for this reason that McCollum, who inactivated vitamin A by aeration and heating cod liver oil, concluded that another vitamin existed in cod liver oil, called vitamin D, that is necessary for skeletal mineralization. 21 It is clearly evident that vitamin D is absent from our natural food supply except in fish liver oils. Some is found in egg yolk, very little is found in milk, little is found in meats, and almost none is found in plant-based foods. We, fortunately, are able to fortify foods and we are able to absorb the vitamin D so that we must consider that vitamin D is available not only through photolysis of 7-dehydrocholesterol in skin, but also from dietary means, usually supplements. Figure 2. Regulation of serum calcium by the calciotropic hormones, parathyroid hormone (PTH), and 1,25-(OH) 2 D 3. Plasma calcium is kept constant at 10 mg% in normal animals and man. This level is required to prevent hypocalcemic tetany on the one hand and to ensure mineralization of newly formed collagen fibrils of bone. 1,25-(OH) 2 D 3 accomplishes this by activating the enterocyte to transport calcium against electrochemical potential gradient in the small intestine to the plasma compartment. If calcium is not available from that source, continued stimulation of the parathyroid results in high levels of PTH and 1,25- (OH) 2 D 3 causes mobilization of calcium. Both hormones are required in vivo for this result. Furthermore, the reabsorption of the final 1% of the filtered load of calcium in the kidney is under the influence of 1,25- (OH) 2 D 3 and the PTH. Both are also required for this function. These elevated plasma calcium levels result in the mineralization of the skeleton and prevention of hypocalcemic tetany. The Function of Vitamin D in Mineralizing the Skeleton As illustrated in Figure 2, we now understand at the physiologic level how the active form of vitamin D functions in the healing of bone and hence in bone health. In rickets, the osteoblasts of the skeleton are able to produce approximately normal collagen fibers and associated proteins responsible for the organic matrix of bone. However, in vitamin D deficiency, there is a failure of the collagen matrix to acquire calcium and phosphorus in the form of hydroxyapatite. After many years of investigation, it became abundantly clear that the mineralization process is not vitamin D dependent, but results from plasma levels of calcium and phosphorus that are supersaturating with regard to hydroxyapatite and that the mineralization process is one not involving vitamin D. 39,40 The function of vitamin D is to elevate plasma calcium and phosphorus to supersaturating levels. It is also of great importance to recognize that a deficiency of vitamin D results not only in failure of bone mineralization, but also in a failure in neuromuscular function, giving rise to hypocalcemic tetany, a devastating disorder that results in death unless immediately corrected. 41 Vitamin D, therefore, is responsible for elevating plasma calcium and, not shown in this figure, plasma phosphorus to levels that are supersaturating. These levels of calcium prevent hypocalcemic tetany on the one hand and prevent rickets or osteomalacia on the other. The manner in which vitamin D functions to raise serum calcium is by increasing the active transport of calcium from the lumen of intestine to the plasma compartment and independently the transport of phosphate from the lumen of intestine to the plasma compartment. To prevent hypocalcemic tetany in the absence of dietary calcium, we must be able to mobilize a source of calcium that is always available. The skeleton serves in this capacity. Apart from its structural role, bone is a warehouse for calcium. When calcium is absent from the intestine, the body mobilizes calcium from the skeleton by a process requiring both the parathyroid hormone (PTH) and 1,25- (OH) 2 D 3. Both are required, resulting in bone resorption and pumping calcium from the skeleton back into the plasma. 42,43 Similarly, the conservation of the last 1% of the filtered load of calcium in the kidney is under control of 1,25-(OH) 2 D 3 and the PTH. 44 By these processes, calcium and phosphorus are raised in the plasma to levels that support mineralization of the skeleton.

S82 HECTOR F. DELUCA GASTROENTEROLOGY Vol. 137, No. 5 Vitamin D Endocrine Systems and Calcium Homeostasis Figure 3 illustrates the regulators of this process and essentially describes the vitamin D endocrine system. The calcium-sensing organ of the body is the parathyroid gland. A calcium-sensitive protein detects even the slightest decrease in ionized calcium of the plasma and by a G-protein coupled mechanism signals the parathyroids to secrete PTH into the plasma, a process requiring only seconds to achieve. 45 The PTH has many functions in the renal tubule, but one of the most important is to activate the CYP27B1 to produce 1,25-(OH) 2 D 3, the calcium heavyweight. This compound then goes to the intestine, bone, and kidney where it facilitates the functions shown in Figure 2. By this process, serum calcium is raised and, not shown in this figure, serum phosphorus is raised, which then results in normalization of bone formation and neuromuscular function. If calcium rises above the required level of 10 mg/dl, the parafollicular cells of the thyroid react to secrete calcitonin, a 32-amino acid peptide hormone that blocks bone resorption as its major function to bring calcium back into the normal range. Calcium rising above this range is dangerous, resulting in hypercalcemia and mineralization of tissues such as heart, kidney, aorta, skin, and other sites. 46 One of the most important discoveries has been the finding of the VDR in the parathyroid glands. 47 It is now known that 1,25-(OH) 2 D 3, through the VDR, acts to negatively control the parathyroid gene and parathyroid proliferation. 48,49 This represents another important feedback mechanism where the vitamin D hormone suppresses production of PTH and also suppresses proliferation of the parathyroid cells, keeping the vitamin D endocrine and calcium homeostatic system in balance. Molecular Mechanism of Action of Vitamin D and the Role of the VDR The VDR is a member of the nuclear receptor family and is the smallest member of that family. It is most closely related to the retinoid receptor and the thyroid receptor. 50 A great deal is known about the biochemistry of the receptor and interested readers are referred to other reviews that present more detail on the VDR itself. 50,51 The essentiality of the VDR in the function of vitamin D is clearly illustrated by the disease, vitamin D-dependency rickets type II, which was discovered in the early 1980s when patients with severe rickets, alopecia, and stunted growth presented with abnormally high levels of 1,25-(OH) 2 D 3 in their blood. 52,53 In the most brittle cases, depending on the site of the mutation of the VDR, they are completely unresponsive to the vitamin D hormone and the only choice available to physicians is to infuse calcium and phosphorus into their plasma to mineralize their skeletons. 54 Other mutations in the VDR occur elsewhere that reduces its effectiveness Figure 3. Diagrammatic representation of the regulation of serum calcium by 1,25-(OH) 2 D 3. Plasma level of calcium is shown as a thermometer. Very slight movements of calcium in the low range activate a calcium-sensing protein in the parathyroids that cause immediate secretion of the PTH. PTH activates the 1 -hydroxylase in the proximal convoluted tubule cells that causes a synthesis 1,25-(OH) 2 D 3. 1,25- (OH) 2 D 3 then acts on the intestine, bone, and kidney as described in Figure 2 to raise plasma calcium. When calcium rises above the normal range, secretion of PTH is shut off. Higher than normal calcium levels also activate a calcium-sensing protein in the parafollicular cells of the thyroid, which secretes calcitonin. Calcitonin then blocks bone resorption, bringing calcium into the normal range. Calcitonin also activates small amounts of the 1 -hydroxylase to produce a sustaining amount of 1,25-(OH) 2 D 3 for its functions other than in calcium homeostasis. 1,25- (OH) 2 D 3 feedback regulates the biosynthesis of PTH and parathyroid proliferation. and can be treated in part by increasing levels of 1,25- (OH) 2 D 3. 54 Alopecia is a failure of the hair follicle system, not because of absence of the ligand but absence of the VDR itself that plays a role in determining the hair cycle. 55 At least 3 different null mutant mice models in regard to the VDR have been produced and demonstrate that vitamin D-dependency rickets type II in man is identical to the null mutant mouse models. It is, therefore, quite clear that the major function if not exclusive function of vitamin D is by virtue of 1,25-(OH) 2 D 3 interacting with its nuclear receptor to induce transcription of target genes that then carry out the functions of vitamin D. Furthermore, there are genes whose transcription is actually suppressed by the vitamin D hormone as, for example, the preproparathyroid gene. Thus, vitamin D through its hormonal form interacting with the receptor brings about induction of certain target genes and suppression of others that brings about the phenotype of vitamin D treatment. The molecular mechanism of how the VDR is believed to act in the transcription system is illustrated in Figure 4. Thus, the VDR that is found either in the cytoplasm or nucleus interacts with the ligand, thereby undergoing a conformational change that allows it to bind to vitamin

November Supplement 2009 VITAMIN D AND THE PN PATIENT S83 Figure 4. The molecular mechanism whereby 1,25-(OH) 2 D 3 is believed to regulate gene transcription. The VDR changes configuration on binding of 1,25-(OH) 2 D 3, which then causes rejection of the co-repressor. The liganded VDR then binds to the vitamin D- responsive element (VDRE) usually in the promoter region of a target gene. There it forms a heterodimer with retinoid X receptor. This resulting complex then binds several proteins that in turn cause transacetylation of the histones of chromatin loosening the structure. This is followed by the leaving of those proteins and attachment of a new group of transcription factors that then either stimulate transcription or, in the case of parathyroid, inhibit transcription. D-responsive elements (VDRE) that are found primarily in the promoter region of target genes, but that may also be found elsewhere as enhancers in the introns or distal regions of the promoter. The liganded receptor binds to the proximal arm of the VDRE, whereas the retinoid X receptor binds to the distal element of the VDRE. The VDRE is described by consensus sequence in which there are 6 nucleotides separated by 3 unspecified nucleotides from another hexamer that is similar to the first hexamer. When the retinoid X receptor VDR binds to the responsive elements, there is immediate binding of several important transcriptional factors. The first group of factors is responsible for histone acetylation, making available the DNA sequence for transcriptive regulation. That group of proteins leaves and another group of proteins binds that then initiates transcription or suppresses transcription. The RNAs produced or not produced then are responsible for the pattern of proteins that bring about the function of vitamin D. There have been reports of nongenomic actions of vitamin D. 56 Initially, it seemed that high concentrations of the vitamin D hormone could act on the membranes of cells to permit calcium penetration and thereby activating a number of kinases. This has been reported to cause very rapid actions of vitamin D, such as an increase in intestinal calcium absorption a few minutes after vitamin D administration. 57 In our laboratory experiments, no such rapid actions of vitamin D have been found. 58,59 Many of the nongenomic actions are carried out with high concentrations of vitamin D hormone, which are physiologically not possible. Although they represent interesting experimental exercises, whether they actually occur in patients or in vivo remains to be determined. Currently, there is no function of vitamin D delineated that has been clearly related to the nongenomic action of vitamin D. Functions of Vitamin D and Bone Resorption The first demonstration that vitamin D might play a role in terminal cell differentiation was found by Suda et al, in which promyelocytes could be caused to differentiate to the monocyte under the influence of the vitamin D hormone. 60 At the same time that the hormone caused differentiation, there was a suppression of cellular growth that led to excitement that vitamin D might play an important role in prevention of malignant proliferation. Furthermore, continued study of the promyelocyte differentiation led to the discovery that the vitamin D hormone plays a major role in the development of the giant osteoclast (Figure 5). 61 It is now believed that the vitamin D hormone interacts with the osteoblast and stromal cells of the bone marrow to produce RANKL, a protein that regulates osteoclastogenesis and the activation of resting osteoclasts. This mechanism goes a long way to describe the molecular events that occur in the vitamin D-induced bone calcium mobilization required to support calcium homeostasis. It also illustrates the importance of vitamin D in regulation of bone remodeling. In the intestine, vitamin D regulates the induction of proteins that cause the enterocyte of the small intestine to transport calcium against an electrochemical potential gradient to the plasma compartment. This was believed to be catalyzed by a calcium channel protein called TYPV6 that is induced by 1,25-(OH) 2 D 3, a calcium-binding protein (calbindin D 9k ), and a basal lateral membrane calcium adenosine triphosphatase. 62 However, the elimination of calbindin D 9k and the TRPV6 protein has failed to reduce or prevent the role of 1,25-(OH) 2 D 3 in inducing intestinal calcium transport. 63 65 Therefore, the molecular mechanism of calcium transport in the intestine remains as yet undefined.

S84 HECTOR F. DELUCA GASTROENTEROLOGY Vol. 137, No. 5 Figure 5. The mechanism whereby 1,25-(OH) 2 D 3 and PTH cause bone resorption. These 2 hormones stimulate the secretion of RANKL which binds to a receptor encoded by nuclear factor RANK. This then facilitates the differentiation of the myelocytes to form precursor osteoclasts. These early osteoclasts are stimulated further to become active osteoclasts. Furthermore, RANKL also stimulates inactive osteoclasts to become active. This results in the resorption of bone and the return of bone calcium into the plasma compartment. (Courtesy of Dr Jackie A. Fretz [J.W. Pike Lab]). A role for the vitamin D hormone in the immune system is now clear. 1,25-Dihydroxyvitamin D 3 was first shown to markedly reduce delayed hypersensitivity in the mouse induced by nitrobenzene. 66 This led to a demonstration that 1,25-(OH) 2 D 3 can suppress experimental immune encephalomyelitis, a model of multiple sclerosis. 67 Similarly, prevention of type 1 diabetes can be blocked by the vitamin D hormone. 68 Unfortunately, both treatments are accompanied by hypercalcemia. Thus, the primary function of the vitamin D hormone is to manage calcium homeostasis by the elevation of plasma calcium concentrations, and this property makes therapy of autoimmune disease difficult. Other autoimmune diseases that are ameliorated or suppressed by the vitamin D hormone include rheumatoid arthritis and inflammatory bowel disease such as Crohn s disease. 69,70 The VDR has been found in virtually all cells of the immune system, but especially T cells, being highest in concentration in the cytotoxic cells. 71 Furthermore, the presence of the VDR in skin illustrates a role of vitamin D in skin. 72 Thus, topical application of vitamin D derivatives has been shown to markedly ameliorate psoriasis. 73 Multiple functions of vitamin D are clearly present; the possible role of vitamin D in the prevention of malignancy or suppression of malignancy may exist. 74 These possibilities have been explored extensively in epidemiologic studies. These studies report an association between plasma levels of 25-OH-D 3 of the order of 40 60 ng/ml with a reduced risk of colorectal cancer. 19 Furthermore, high sunlight areas near the equator have the lowest incidence of certain autoimmune diseases such as multiple sclerosis and inflammatory bowel disease, as well as type 1 diabetes. 18 20,75 A role for vitamin D in the prevention of cardiac dysfunction has recently been signaled by an epidemiologic study showing that the risk of heart attack is reduced in patients with higher blood levels of 25-OH-D 3. 20 Studies of this type have strongly suggested that these blood levels of 25-OH-D 3 should be achieved for the benefit of public health. This currently represents a controversial area, but certainly we must reexamine our recommended intakes of vitamin D. Unfortunately, there is not sufficient evidence of where the safety line is for vitamin D. Vitamin D intoxication is well known. 76 In an animal study, it was clearly shown that plasma levels of 25-OH-D 3 must be between 250 and 400 ng/ml for toxicity with a normal range being approximately 30 ng/ml. Current epidemiologic evidence suggests that we should strive for blood levels of between 30 and 100 ng/ml, but not above without a clear indication of where the safety limit is in man. It is evident that measurements of plasma 25-OH-D 3 are essential in the determination of when the danger for vitamin D intoxication can be expected. In the article by Shepard et al, 77 it is clear that 1,25-(OH) 2 D 3 is very low in animals given toxic doses of 25-OH-D 3, suggesting that 1,25-(OH) 2 D 3 is not the toxicant in vitamin D intoxication but rather it is 25-OH-D 3 that, at high concentrations, interacts with the VDR giving rise to hypercalcemia. 77 Assessment of Vitamin D Status: What Should Be Measured, and What Are the Levels From the section on regulation, it is evident that measurement of 1,25-(OH) 2 D 3, which is found in picogram range in plasma, is not very useful, either for diagnosis or for assessment of vitamin D status. For example, in hypervitaminosis D, 1,25-(OH) 2 D 3 is probably not even measureable. 77 Furthermore, under conditions of low calcium intake and high parathyroid secretion, plasma levels of 1,25-(OH) 2 D 3 are very high, possibly in

November Supplement 2009 VITAMIN D AND THE PN PATIENT S85 the range of 150 200 pg/ml. 77 Another factor is the short half-life of 1,25-(OH) 2 D 3 (2 4 hours) as compared with the half-life of 25-OH-D 3 (15 days). 17 In short, the measurement of 1,25-(OH) 2 D 3 is only useful when examined in regard to plasma levels of calcium, PTH, phosphorus, and vitamin D intake or exposure. However, 25-OH-D 3, which is the circulating form of vitamin D 3, is by far the best assessment of vitamin D status available. 78 There are a large number of articles reporting on the levels of 25-OH-D 3 in various countries, under various conditions, and in various disease states. Plasma levels 10 ng/ml can be associated with vitamin D deficiency. 79 Plasma levels of approximately 20 40 ng/ml can be associated with reduced vitamin D function, but adequate in preventing rickets and osteomalacia. 79,80 However, plasma levels between 30 and 100 ng/ml can be associated with the maximum protection against malignancy, autoimmune diseases, and diseases of inflammation. 81,82 Plasma levels 200 ng/ml signal possible vitamin D intoxication. 77 It is important to realize that vitamin D 3 and vitamin D 2 are stored in the adipose tissue. 83 Measurement of 25-OH-D 3 does not provide an assessment of the amount of vitamin D stored in the body. In fact, there is no adequate way to assess this pool of vitamin D. However, continued monitoring of 25-OH-D 3 would give an indication of whether intoxication is going to become a problem. Unfortunately, we are faced with the problem of giving 1 form of vitamin D, namely, vitamin D 3, and measuring another 25-OH-D 3 and an undetermined amount of vitamin D 3 goes into the storage compartment. Exactly what happens as one increases the doses of vitamin D 3 in regard to the distribution between storage and plasma levels remains unknown. Impact of Various Conditions on Vitamin D Measurements in the Plasma All types of malabsorption syndrome, of course, interfere with vitamin D absorption in the gastrointestinal (GI) tract and, unless ultraviolet irradiation is being used, the source of vitamin D is lost to the patient and results in deficiency. In the case of PN, very likely these patients do not receive ultraviolet light and do not absorb vitamin D from the GI tract. Therefore, the danger of vitamin D deficiency is high. Intramuscular injections of vitamin D are possible, but not currently available. It is also of great importance that orally administered vitamin D in normal subjects does not elevate plasma 25- OH-D 3 sharply. 84 Large amounts are needed to raise plasma levels even 10 ng/ml. Thus, in PN, it is likely that because 25-OH-D 3 is not available for clinical use, vitamin D 3 of the order of 3,000 4,000 IU/d would be required to result in levels of 30 100 ng/ml of plasma 25-OH-D 3, although this has not been directly tried in parenterally fed patients. Although the recommended daily oral allowances are 200 to not 600 IU/d, these have been set for the purpose of preventing the diseases rickets and osteomalacia. 85 They have never been set for the assessment of protection against the diseases discussed. From the literature, it is quite clear that we should strive for blood levels of 25-OH-D 3 between 30 and 100 ng/ml (75 250 nmol/l). This may be achieved either by oral administration or in the case of patients unable to absorb vitamin D by either ultraviolet exposure of skin, which brings about a risk of skin malignancy, or intramuscular injections of large amounts of vitamin D. The time has come for us to consider for total PN (TPN) that the vitamin D levels must be increased substantially to provide 25-OH-D 3 levels of 30 100 ng/ml (75 250 nmol/l). It is at this level that currently available evidence suggests that we may receive protection against a number of diseases including malignancy and autoimmune disease. Vitamin D and the Patient on PN Although the mineral levels in PN patients are well managed, vitamin D is needed for functions excluding intestinal absorption of calcium. Although not determined in PN nutrition patients directly, it is reasonable to expect that vitamin D functions in PN patients in the same way as for ordinary subjects who have full GI function. As discussed, vitamin D is required for bone remodeling, bone synthesis, and immunomodulation, and may be required for cardiovascular function. It may also reduce the risk of several malignancies and the risk of autoimmune diseases such as multiple sclerosis, type I diabetes, and lupus. 7 These functions are independent of serum calcium levels. Current epidemiologic evidence suggests that in normal subjects plasma or serum levels of 30 100 ng/ml of 25-OH-D may offer protection against disease. This cannot be achieved by 400 IU/d vitamin D, the level currently in intravenous vitamin preparations. 84 Unfortunately, no parenteral form of vitamin D (above the intravenous multivitamin) is available in the United States. The only current alternative is ultraviolet light exposure to gather the public health benefits of vitamin D. Thus, PN patients could achieve 30-ng/ml of 25-OH-D 3 by UV exposure as recommended. 84 There is some debate over the equivalence of vitamin D 2 versus D 3. Certainly in several species they are not equivalent, 85 but in man they are equal in potency for curing rickets. 85 They may not be equivalent in raising blood levels of 25-OH-D. 86 However, this view is controversial. 87 Because vitamin D 3 is the naturally produced form in skin, it would be the form of choice. Calcium levels in serum regulate PTH secretion that in turn regulate the CYP27B1 or the 25-OH-D 1 -hydroxylase that produces the vitamin D hormone, 1,25- (OH) 2 D 3. 7 A constant and high level of serum calcium might be expected to suppress PTH and the 1-hydroxylase. However, high calcium levels also stimulate calci-

S86 HECTOR F. DELUCA GASTROENTEROLOGY Vol. 137, No. 5 tonin secretion that can also activate the CYP27B1, 88 providing maintenance levels of 1,25-(OH) 2 D 3. It is, therefore, unlikely that serum calcium levels that may change slightly in relation to meal consumption would provide more 1,25-(OH) 2 D 3 than occurs with a constant infusion of calcium. However, too much calcium could suppress PTH levels and consequently the conversion of 25-OH-D 3 to 1,25-(OH) 2 D 3. Metabolic bone disease has been reported in long-term PN. 89 These patients presented with osteomalacia, hypercalcemia, hypercalciuria, and negative calcium balance. Withdrawal of vitamin D from the PN solutions resulted in a disappearance of hypercalcemia and hypercalciuria. 90 The authors concluded that vitamin D might actually be responsible for osteomalacia and fractures. A later report suggested the disease to be a defect in matrix formation. 91 Clearly, the basis of the disease is unknown and it is unlikely that the underlying cause is vitamin D. Removal of vitamin D could have resulted in failure of bone remodeling and a failure of PTH to act on bone because vitamin D is required for that action. 42,43 In this manner, removal of vitamin D could result in the correction of hypercalcemia and hypercalciuria, but would be detrimental in the long term because bone repair by remodeling would be reduced without vitamin D. The removal of vitamin D would further deny the patient the more recently discovered benefits of vitamin D. The cause of the metabolic bone disease seen in some long-term TPN patients needs to be determined because it may be caused by other factors, including the calcium and phosphorus load presented to the patients. The conclusion that the bone disease of long-term TPN patients is due to vitamin D is premature and could be misleading. Recommendations Vitamin D is required by PN patients as well as other human subjects. It must be supplied in the PN solutions or by limited ultraviolet irradiation. Serum 25-OH-D levels should be monitored and kept in the range of 30 100 ng/ml serum. Measurements of 1,25-(OH)2D3 is not recommended for assessment of vitamin D status. Beside bone health, vitamin D may reduce the risk of colorectal cancer, breast cancer, autoimmune diseases, and heart attacks. Serum levels of 25-OH-D3 of 30 100 ng/ml may offer significant protection from these diseases. The availability of parenteral vitamin D preparation should be strongly pursued to provide doctors a means of achieving levels of 25-OH-D3 that provide these benefits. The real cause of the bone disease of long-term TPN needs to be determined. Question and Answer Session DR SHIKE: Thank you for your comments. I would just like to make a couple of points and I also have a question. The point about the bone disease is that it is really a very complex phenomena. As you may have noted in our research, the PTH was very suppressed in these patients. You rightly pointed out that it takes 2 keys to open the safe, both the 1,25-dihydroxyvitamin D and the PTH. Furthermore, because of the constant infusion of calcium during long hours, unlike the postabsorptive state, we have reordered the calcium concentrations in blood, which again as you point out, in normal physiology the calcium level is tightly regulated. My question is whether administration of vitamin D intravenously has a different effect than vitamin D from sunlight exposure or ingested from food? Do you think that the 25-hydroxyvitamin D is the active form when it comes to cellular proliferation and differentiation? DR DELUCA: First the paracrine question, a lot of that is still debatable, but remember the 1-hydroxylase in the paracrine is the same enzyme that is found in the kidney. There s only one 1-hydroxylase and so the K m for 25-hydroxy D should be the same for both of those. So far no one has shown a function for 25-hydroxyvitamin D. Having said that, there had to be an evolutionary reason for 25-hydroxyvitamin D. I think eventually a function for 25-hydroxyvitamin D will be discovered, but so far none has been identified. DR JEEJEEBHOY: You point out that the major source of D comes via the skin. In a number of these PN patients, we found withdrawal of IV vitamin D does not make the patients vitamin D deficient. The 25-hydroxyvitamin D concentration remains normal, but PTH increases, and probably as a result, bone mass increases as well. Therefore, I think we are over suppressing PTH by the intravenous administration of vitamin D. DR DELUCA: I believe that it is the intravenous calcium that is suppressing the PTH. Your patients are not producing 1,25-dihydroxyvitamin D, which is the suppressant of the parathyroid gland. 25-Hydroxyvitamin D really does not function in that capacity. I believe the issue is whether calcium can be delivered in a different manner or perhaps patients are provided with excessive calcium. DR JEEJEEBHOY: What happens is that when we provide a combination of calcium and vitamin D to these patients and they become relatively hypercalcemic. If we use low dose calcium 5 mmol/d, serum calcium concentrations fall below normal in many patients. However PTH remains suppressed, and this is the problem. DR BUCHMAN: Moshe and Khursh, don t you think that this is really a heterogeneous disease that we are talking about. If you look at Ed Lipkin s study from

November Supplement 2009 VITAMIN D AND THE PN PATIENT S87 Seattle, some of the patients have a low turnover osteoporosis, some have osteomalacia, and yet others have elevated PTH. I have not seen a depressed PTH in years. The low PTH was often attributed to aluminum contamination, which is no longer an issue. Do you think we are trying to treat different problems with the same approach? DR DELUCA: Well, there are a number of issues here. The point is that irrespective of whether it is osteoporosis, or osteomalacia, if you suppress PTH that has a series of actions that results in additional osteoporosis. As you know, osteoporosis today is being treated by the recombinant PTH. DR BUCHMAN: Do you think aluminum was a factor in those early experiments? DR SHIKE: Not in our study; we never used the casein hydrolysates with the high aluminum content. We saw a picture of high osteoid with little calcification. With regard to PTH, many patients with long-standing intestinal failure have secondary hyperparathyroidism, and when PN is started the amount of calcium they receive increases, and their PTH decreases. Having said that, I agree with you, there are many other factors to consider in the development of PN-associated metabolic bone disease. DR DELUCA: My question to those of you that prescribe PN, is there a way to infuse calcium whereby the serum ionized calcium concentration is not increased to the degree that PTH suppression occurs? I do not think leaving out the vitamin D is the answer. Do these longterm PN patients still develop metabolic bone disease in 2009? DR JEEJEEBHOY: PN patients still develop metabolic bone disease and there are longitudinal studies to support Dr Shike s contention that many factors still need to be considered but the question is this: Are we committed to intravenous vitamin D supplementation as long as the D levels are normal? In other words, isn t there another way for these patients to get vitamin D other than giving it IV? DR DELUCA: I think the best way to do it is UV radiation. ANONYMOUS: Dr DeLuca, you mentioned about many potential roles for vitamin D other than for bones and calcium metabolism. Do you things greater intakes of vitamin D are required than what is currently provided to reap these newly described benefits? The other question I have is that there is a lot of science now dealing with single gene polymorphism where small based paired deletions could theoretically make some individuals more predisposed to the effects of vitamin D deficiency. Do you see that as an area that we should be investigating? DR DELUCA: I do not think we have a full understanding of which of those polymorphisms are a problem or not. I think they are still under investigation, but you should be aware of them. I do think that in the next 10 years we will be recommending 25-hydroxy D levels of greater than the current 30 60 ng/ml, because of all the epidemiologic studies that show a correlation between higher levels and protection against serious diseases. DR SHULMAN: In the animal models, you showed there was a benefit from vitamin D supplementation in multiple sclerosis and other models, but the problem that occurred with vitamin D supplementation was hypercalcemia, so I presume that you were using very high doses of vitamin D there. Are there any animal data that show maintaining the kind of analogous levels as humans (ie, 80 100 ng/ml 25 OH vitamin D) has the same kind of benefit? DR DELUCA: First, those experiments were done with the active form of vitamin D, which is very, very potent; we gave 50 ng/d to mice and that produced a1mg% hypercalcemia, and prevented the multiple sclerosis-like disease. That is an unacceptable treatment for people. DR PIRONI: Does vitamin D interfere with bisphosphonate blockage of calcium resorption from bone? DR DELUCA: No. You can use the two together, and as far as I know it does not inhibit the CYP27 B1 or 1 -hydroxylase. DR BIESALSKI: You mentioned that food is a rather poor source for vitamin D for humans, so metabolism in the skin is the major source. To my knowledge, the ability of the skin to synthesize vitamin D decreases with age, so shouldn t it be recommended with special reference to home PN that with increasing age, additional vitamin D supplementation may be required? My second question regards acute intervention. There are some data that during trauma or critical illness vitamin D plasma concentrations decrease rapidly. Is that an indication for an acute intervention with parenteral vitamin D? DR DELUCA: Some acute diseases, especially those with proteinuria, have been associated with increased urinary vitamin D loss. However, it is unlikely that transient losses are associated with clinically significant sequelae. DR BUCHMAN: Many of the patients on long-term PN have Crohn s disease. Could you comment on the recent data on the potential role of vitamin D deficiency and increased intestinal permeability in the pathogenesis of Crohn s disease? DR DELUCA: There seems to be a benefit from vitamin D supplementation in animal models of inflammatory bowel disease, and the absence of vitamin D actually aggravates the disease symptoms. DR LIPMAN (Washington, DC): In what form are we delivering vitamin D in the current intravenous multivitamin preparations? DR HOWARD: Vitamin D 3 cholecalciferol. DR LIPMAN: So it is previtamin and it still has to be converted to its active form. What is the difference between vitamin D 2 and vitamin D 3, and does that have any

S88 HECTOR F. DELUCA GASTROENTEROLOGY Vol. 137, No. 5 effect on the conversion to the activated form? Finally, what affect does chronic liver disease in PN patients have on vitamin D metabolism? DR DELUCA: Vitamin D 2 differs structurally from vitamin D 3 ; the side chain is different. It is made artificially from the plant sterol ergosterol. I believe that was used at one time in the parenteral multivitamin formulations, but I think currently most vitamin D being used is vitamin D 3, which is the form that is made in your skin. DR LIPMAN: Are there any clinical benefits to vitamin D 2 or should it not be used? DR DELUCA: In curing rickets, vitamin D 2 and vitamin D 3 are equally active, but there is an ongoing debate, with a similar dose of vitamin D 2 there is less 25-hydroxyvitamin D 2 in the blood than would be the case with an equal amount of vitamin D 3. The efficiency of conversion seems to be somewhat different in humans. Is that a functional problem? We are not sure. But I think from the practical point of view almost all vitamin D that is now being developed is vitamin D 3.D 2 is becoming less and less important. DR HOWARD: When serum 25-hydroxyvitamin D is measured, if there is vitamin D 2 present, do we get information about both D 2 and D 3? DR DELUCA: That depends on which laboratory runs the assays. Some groups can measure both 25-hydroxyvitamin D 2 and 25-hydroxyvitamin D 3 separately (this requires mass spectroscopy), but it is not always reported; one needs to request it specifically. DR SHIKE: Mild hepatic dysfunction does not affect the 25-hydroxylation of the provitamin D, and similarly, one would need a significant degree of renal dysfunction before 1,25-hydroxylation becomes impaired. Therefore, I do not think it is a major issue in these patients. Recently, there was a recall of 25-hydroxy D measurements done commercially. Could you comment whether we should be alarmed by this? DR DELUCA: I think we lack standardization of the 25-hydroxyvitamin D assays and they differ. Quest, which I think is the major company analyzing 25-hydroxyvitamin D in the United States, recalled thousands of assays because they had a 30% error on the high side; in other words, patients looked like they had 30% more vitamin D than they actually had. We in the vitamin D field are trying to get standardized 25-hydroxyvitamin D assays. We were contacted by a clinician who was puzzled by the same sample sent to 3 different groups who used different methods (chemical, high-performance liquid chromatography, and mass spectrometry), which resulted in 3 different values. DR BUCHMAN: I think that is actually a good point for all the assays of the trace metals and vitamins that we are discussing; anybody can give you a result, but the level that is provided is not always correct. DR CHRISTINA TENNYSON (Columbia University): Do you recommend 50,000 units of oral vitamin D compound once weekly in PN patients or those with malabsorption from diseases such as celiac sprue or a smaller supplement? DR BUCHMAN: In our practice, we provide patients with 50,000-unit dosing between once weekly and twice a day depending on the level of malabsorption. We have a couple of patients that are getting 200,000 U/d and still have 25-hydroxyvitamin D concentrations 2 ng/ml. That is a significant problem, especially in Chicago in the winter where nobody goes outside, and when they do go outside in the summer, they slather on sunscreen for obvious and otherwise appropriate reasons. We have started to recommend tanning salons for some patients. DR DELUCA: One thing I d like to suggest and that is if you are going to give these huge doses of vitamin D, be sure you monitor the patient because 25-hydroxy D levels will tell you if you re getting into the toxicity range. DR JEEJEEBHOY: In patients with malabsorption and low vitamin D concentrations, we start with 50,000 units once weekly, just as Dr Buchman does. I have never seen plasma 25-hydroxyvitamin D concentrations in excess of 50 60 ng/ml in these cases. In some of them, it is not enough and we increase the dose as needed every 3 4 months. BETH LYMAN, RN (Kansas City, KS): How much solar exposure is necessary? DR DELUCA: Mike Hollick has claimed in normal people that radiation of hands and face for 10 15 minutes midday in summer would produce your daily requirement for vitamin D, but that was when I think the daily requirement was thought to be somewhere around 400 600 U. I know when I recently went to the Caribbean for 10 days, when I came back, my serum 25- hydroxyvitamin D concentration jumped by 20 ng/ml. DR ZALOGA (Baxter, Deerfield, IL): You alluded to the ionized calcium measurement, which I think is very important. Clearly, calcium infusion rapidly and dramatically suppresses PTH, within a few minutes actually. The ionized calcium as a percentage of total calcium increases when the serum albumin concentration is decreased and in that situation even a minor change in total calcium say from 8 to 8.2 mg/dl could reflect a large change in ionized calcium, sufficient to suppress PTH. Ionized calcium concentration is often not measured because it has to be analyzed using a blood gas machine. It is possible that more patients on PN have elevated ionized calcium concentrations that we might appreciate. Is there a place for low vitamin D and calcium status in critical illness such as sepsis where an increased ionized calcium concentration has been associated with detrimental effects, possibly stimulating apoptotic and pyrolytic pathways? In animal models of acute sepsis, 1-hydroxylase concentrations decrease rapidly and significantly. It recovers as sepsis resolves. It has been thought