Systemic Inflammatory Mediators and Bone Homeostasis in Intestinal Failure  
From the * University of Pennsylvania School of
Nursing, Philadelphia, Pennsylvania; Correspondence: Charlene Compher, PhD, University of Pennsylvania School of Nursing, 420 Guardian Drive, Philadelphia, PA 19104-6096. Electronic mail may be sent to compherc{at}nursing.upenn.edu.
Background: A proinflammatory state has been described in patients
with intestinal failure. The prevalence of metabolic bone disease in this
group is considerable. It is not known whether this proinflammatory state is
related to bone parameters, though bone disease is recognized as a
proinflammatory process in postmenopausal women. The purpose of this study was
to examine whether inflammation was related to bone disease. Methods:
Eight patients with parenteral nutrition (PN)-dependent intestinal failure but
no recent infections or immunosuppressive medications had serum assayed for
interleukin-6 (IL-6), tumor necrosis factor (TNF)- Patients with intestinal failure (IF) due to loss of intestinal mass as a result of thrombosis or disease or to loss of absorptive function as a result of severe malabsorption syndromes are dependent on parenteral nutrition (PN). A proinflammatory state has been identified in small groups of long-term PN patients.1–3 In addition, bone diseases such as osteoporosis, osteomalacia, and hyperparathyroidism may complicate their course. Osteopenia prevalence has been reported as high as 84% and osteoporosis as high as 67% in home PN (HPN) patients.4–6 Whether the proinflammatory state is associated with bone disease in patients with IF is not known.
Bone disease is a proinflammatory process that limits vitamin D
availability. Inadequate activated vitamin D results in elevated parathyroid
hormone (PTH), negative calcium balance, and reduced bone
formation.7 It has
been hypothesized that during chronic limited vitamin D nutrition states, PTH
secretion results in stimulation of an acute-phase response mediated by
interleukin-6 (IL-6), followed by resorption of mineral from
bone.8 When vitamin
D is low, IL-6 has paracrine effects on osteoclasts, resulting in bone
resorption, along with further production of IL-6 in hepatocytes, hepatic
endothelial cells, and Kupffer cells. IL-6 also stimulates release of
C-reactive protein (CRP) and
fibrinogen.8 Tumor
necrosis factor (TNF)- The purpose of this study was to examine whether the proinflammatory state in HPN patients is associated with bone mineral homeostasis. We hypothesized that proinflammatory cytokine concentrations would be associated with bone disease.
Patients who were managed by the HPN program at the Hospital of the University of Pennsylvania, Philadelphia, PA, were eligible for study if they had received PN for  2 years, had no clinical signs or culture-confirmed
catheter infection within 3 months, were >18 years old, and did not have
active cancer or inflammatory bowel disease or immunosuppressive medications.
The institutional review board approved the protocol, and informed consent was
obtained from all participants.
After a 12-hour fast, 20 mL of blood was collected into heparinized
vacutainers for determination of serum cytokine levels and isolation of
lymphocytes. Serum IL-6, soluble TNF- For the lymphocyte proliferation and flow cytometry assays, 20 mL blood was diluted with a 50% volume of sterile phosphate-buffered saline and centrifuged at 670 x g for 30 minutes over Ficoll-Hypaque (Pharmacia, Piscataway, NJ). Cells from the interface were harvested. Total T cells (CD3+), T helper cells (CD4+), and cytotoxic T cells (CD8+) were enumerated with the fluorescently conjugated CD antigen-specific monoclonal antibody using a FACScan (Becton-Dickinson, San Jose, CA) and data analyzed with Cell quest (Becton-Dickinson) and WinMDI software. To assess cell proliferation, lymphocytes were prepared at a dilution of 2 x 106 cells/mL and labeled with 2.5 µmol/L carboxyfluorescein diacetate succinimidyl ester (CFSE) for 10 minutes at 37°C in serum-free RPMI. After washing twice in 10% fetal bovine serum, the cells were plated and stimulated in the presence of 1.0 µg phytohemagglutinin. Total T-cell proliferation was determined 5 days later by assessing the CFSE dilution pattern by flow cytometry on the gated CD3+ population. Propidium iodide was included to discriminate dead from live cells. In addition, calcium, phosphorus, magnesium, and alkaline phosphatase concentrations were assayed by a clinical laboratory by spectrophotometric assay in serum drawn within 3 months of the inflammation studies. Intact PTH was assayed by immunochemical luminometric assay, 25-hydroxy vitamin D by radioimmunoassay, and 1,25-dihydroxy vitamin D by radioreceptor assay. When patients had a dual-energy x-ray absorptiometry (DXA) scan of bone density within 6 months of the inflammation studies, these bone scan results were included in the data. Because the DXA scans were not made by the same scanner, however, the T-scores were not evaluated statistically as they were not optimally comparable. Results are reported as mean ± SD. Pearson's correlation was used to compare biochemical bone profiles to the inflammatory cytokines and T-cell function. p Values <.05 were considered statistically significant. SPSS-12.0 (SPSS, Chicago, IL) was used for statistical analysis.
Of 10 adults who were originally studied and whose immune system–associated inflammatory marker results were reported as significantly higher than healthy age- and sex-matched controls,2,9 8 were patients from the University of Pennsylvania HPN program and had laboratory panels within 3 months of the inflammation tests. The 8 patients included 3 men and 5 women (4 of whom were postmenopausal), with a mean age of 55.0 ± 14.23 years. Diagnoses included short bowel syndrome in 6 (radiation enteritis in 1 of these) and dysmotility in 2 patients (Table I). The duration of IF was 9.65 ± 9.16 years. Risks for osteoporosis included the primary risk of severe malabsorption in all, postmenopausal female status in 4, and a 10-year-distant history of steroid use in 1.
The mean concentrations of calcium = 9.0 ± 0.66 and phosphorus = 3.41 ± 0.66 mg/dL were within the reference range (Table II). The circulating albumin concentration was 3.41 ± 0.66 g/dL and low in 3 patients. The mean 25-hydroxy vitamin D concentration was low, at 10.05 ± 3.65 ng/mL, although the 1,25-dihydroxy vitamin D was normal, at 27.57 ± 18.10 pg/mL. The mean intact PTH concentration was 8.77 ± 5.19, greater than the laboratory reference range of 1.3–6.8 pmol/L and elevated in 4 patients. Mean alkaline phosphatase was 155.25 ± 85.45 U/L and above the laboratory reference range of 35–125 U/L. The T scores at the lumbar spine ranged from –1.6 to 1.0 SD, whereas T scores at the femur were –1.6 to 0.95 (data not shown). Patient 3 had a history of multiple fractures after falls, and patient 4 had no fractures but loss of height.
The immune and cytokine data are presented in
Table III and correlations in
Table IV. Calcium was inversely
correlated with sTNFR-II (R =–0.817; p = .013) and TNF-
In this group of HPN patients, the inflammatory state was linked to vitamin D deficiency, hyperparathyroidism, and low calcium. Although these findings must be considered preliminary because they are based on statistical associations only, they are nonetheless consistent with the hypothesis elaborated by McCarty8 that vitamin D deficiency, resulting in elevated PTH levels, stimulates an inflammatory response and the loss of bone mineral. Unfortunately, the cross-sectional nature of this study did not permit evaluation of the timing of inflammation relative to vitamin D deficiency.
Vitamin D has anti-inflammatory effects in immune cells. Vitamin D
receptors (VDRs) are found on all types of immune cells, where the VDR
mediates vitamin D
action.10
Specifically, antigen-induced proliferation and cytokine (IL-2, IL-12,
interferon- In addition, vitamin D status affects susceptibility to infection. The innate immune response, through activation of toll-like receptors, triggers the expression of the VDR and conversion of 25-hydroxy vitamin D to its active metabolite to enable intracellular bacterial killing.12 Furthermore, sera from vitamin D–deficient individuals had limited ability to induce mRNA for the cathelicidin gene that codes for the protein which kills intracellular bacteria. Patients with PN-dependent IF have a mean of 0.8 catheter-related bloodstream infections per 1000 catheter-days,13 one of their most common causes of hospitalization.14,15 However, there is great variation in infection rates among individual patients trained in similar techniques of catheter and infusion management, ranging from no infections over 20 years to several each year. We have found no reports in the literature linking catheter-related bloodstream infections to vitamin D status in this population. In this small group of patients, an exclusion criterion for the study was any catheter infection within the prior 3 months, suggesting that recent infection was not likely to have affected vitamin D status. The mean 25-hydroxy vitamin D concentration observed among patients in the current investigation (10.61 ± 3.65 ng/mL) is not very different from the Haderslev et al5 report in a group of patients with intestinal insufficiency (not PN dependent). Serum vitamin D concentration was deficient at 13.4 ± 9.7 ng/mL,5 approximately half that of the healthy control group. Furthermore, in 38% of these patients, 25-hydroxy vitamin D was <8 ng/mL and associated with elevated PTH, markers of bone turnover and reduced bone density. Also, in the Haderslev et al5 data, the mean 1,25-dihydroxy vitamin D concentration did not vary, and means across all patient groups were 21–23 pg/mL by comparison to our mean value of 27.57 ± 18.10 pg/mL. This finding of normal 1,25-dihydroxy vitamin D in the context of 25-hydroxy D deficiency suggests that the bone mineralization defect seen with osteomalacia may not have occurred. We have extended the Haderslev et al5 observation, however, by including patients with more severe IF (PN dependence) and by correlating proinflammatory cytokines with low calcium. The mean value for the circulating 1,25-dihydroxy vitamin D concentration was normal in this study, although 2 individual values were low. This normal 1,25-dihydroxy vitamin D concentration suggests that these patients do not have difficulty activating vitamin D due to renal impairment but that their intake of vitamin D may be too limited. Alternatively, the vitamin D deficiency observed among patients in the current study might have occurred as a secondary response to the inflammatory condition previously observed for HPN patients. Although the actual cause of inflammation in patients with IF is not understood, it is possible that the inflammatory condition identified among these individuals stimulates an increased compensatory conversion of vitamin D to activated vitamin D for immune modulation. If this scenario occurs, the vitamin D provided intravenously and the small amount absorbed by oral diet or obtained from sunlight exposure might fail to provide an adequate intake to replace the physiologic demand for vitamin D's considerable functions in maintenance of bone mass and modulation of immunity and inflammation.
The negative associations among calcium and TNFR-II and TNF- The associations between PN duration and inflammation raise several concerns. First, it is possible that some component (lipid, amino acids, glucose, trace elements, multivitamins, electrolytes) or contaminant (possibly aluminum in the additives) in the PN infusion itself or the catheter causes inflammation. Even those patients who only rarely have bloodstream infections will experience them more often, with their associated inflammatory response, over a longer time period. In addition, chronic marginal nutrition intake of vitamin D or bone minerals might worsen over a longer time. As patients survive more years with PN, days of infusion may be reduced as intestinal adaptation progresses, leaving them with reduced IV vitamin D intake and the need for absorption of oral vitamin D by a diseased intestinal tract to avoid deficiency. This latter factor seems less likely in this patient group because the 3 patients with longest PN duration continued with daily IV vitamin D. The primary limitations of this study were the small sample size, nonconcurrent measurement of bone laboratory panels with inflammatory markers, and the fact that inflammatory markers were measured only once. Regarding the sample size, this investigation was a pilot study. In addition, patients with IF are estimated to number only around 40,000 in the United States.17 We limited our patient selection to stable, permanent IF, as determined by >2 years' duration of HPN. We excluded patients with recent bloodstream infections to avoid any impact of infectious inflammatory response on the data. We further eliminated patients who were using immunosuppressive medications. Thus, most recognized covariates of inflammation were controlled for the individuals who participated in the study, but the exclusion criteria limited the number of participants. The measurement of biochemical bone profiles was not concurrent with the inflammatory measures. It was necessary to measure inflammatory cytokines and T-cell function as a batch assay for standardized conditions; thus, all blood samples were drawn on the same day. Our clinical practice is to measure vitamin D studies only quarterly. Thus, the single-point measure of inflammation could not be exactly matched with the episodic measures of vitamin D, although all reported measurements occurred within a 3-month window. In spite of this limitation, the findings were consistent with a negative correlation of inflammation on bone biochemical parameters. The lack of clinically available inflammation assays up to the present time has limited the ability of investigators to do repeated measures. However, other investigators have also measured inflammation in patients with IF, using a single measurement,1,3 suggesting that these findings are not likely to be aberrations. A controlled trial with repeated measures of inflammatory markers in patients with IF is lacking. A final issue is whether any of these patients may have had blind loop syndrome or intestinal bacterial overgrowth. Although we have no proof that these conditions were not operative, there was essentially no difference in inflammatory markers in the patients who had dysmotility and those who had short bowel syndrome, so we feel that blind loop syndrome is an unlikely explanation of our findings. Many unanswered questions remain regarding inflammation and bone disease in patients with IF. Is there some component or contaminant in the PN solution itself that causes inflammation to occur? Does the presence of the catheter set up an inflammatory response? Is the proinflammatory state persistent over time? Does unappreciated vitamin D deficiency facilitate the inflammation? Alternatively, does the inflammatory process overwhelm the limited available vitamin D in these patients? We look forward to approaching definitive answers to some of these important questions. Not only will such determinations have the potential to improve the management of adult patients with IF, but they may also benefit the medical management of the sizable population of children with IF who are challenged to maintain height velocity in spite of these difficult metabolic conditions.
Weak associations between proinflammatory cytokine and serum calcium concentrations (in the context of low 25-hydroxy vitamin D) and stronger associations between duration of PN and proinflammatory cytokines were found. Longitudinal studies are needed to determine whether vitamin D deficiency causes inflammation or conversely, whether inflammation from another cause results in compromised vitamin D status. The work was supported in part by the Public Health Services Research Grant M01 RR00040 from the National Institutes of Health and by the University of Kansas Medical Center Research Fund (M. Hise, PI). Received for publication May 30, 2006. Accepted for publication October 5, 2006.
Journal of Parenteral and Enteral Nutrition, Vol. 31, No. 2,
142-147 (2007)
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
, and its receptors
(TNFR-I and TNFR-II), C-reactive protein, and whole blood for lymphocyte
proliferation. Routine clinical laboratory measures of vitamin D, parathyroid
hormone, serum calcium, and phosphorus within 3 months of the inflammatory
measures were compared by Pearson's correlation to the inflammatory measures.
Results: Mean values for calcium, phosphorus, and albumin were
normal, but 25-hydroxy vitamin D was reduced and parathyroid hormone and
alkaline phosphatase elevated. Serum total calcium was negatively related to
TNFR-II, TNF-
Top
MATERIALS AND METHODS
RESULTS
) production are inhibited and T-helper-1 maturation
reduced. In dendritic cells, CD-40, CD-80, and CD-86 are down-regulated and
IL-12 secretion is reduced, along with enhancement of the production of the
anti-inflammatory cytokine IL-10 in response to vitamin D. The nuclear VDR
inhibits IL-2 and granulocyte-macrophage stimulating factor gene expression in
response to vitamin
D.