Journal of Pediarric Gos1roemerology and Nulrilion
`12:23~236 0 1991 Raven Press, Ltd., New York
`
`Tumor Necrosis Factor-a Is Not Elevated in Children with
`Inflammatory Bowel Disease
`
`JeffreyS. Hyams, William R. Treem, *Evangeline Eddy, *Nancy Wyzga, and
`*Robert E. Moore
`
`Departments of Pediatrics and *Pathology Hartford Hospital, Hartford, and
`University of Connecticut Health Center, Farmington, Connecticut, U.S.A.
`
`Summary: Chronic undernutrition and high-dose daily
`corticosteroid therapy are well-accepted causes of growth
`failure in children with inflammatory bowel disease. Oc(cid:173)
`casionally, children are seen with minimal gastrointesti(cid:173)
`nal symptoms but in whom severe anorexia and profound
`growth impairment are evident. Recent observations that
`elevated serum levels of tumor necrosis factor-a (TN F) in
`cachexia associated with a number of disease states have
`suggested a similar possible role in inflammatory bowel
`disease. Accordingly, we determined TNF levels in 45
`children and adolescents with inflammatory bowel dis(cid:173)
`ease (18 ulcerative colitis, 27 Crohn's disease) at varying
`times during their clinical course and compared them to
`
`values obtained from a group of 25 children with func(cid:173)
`tional bowel disease. No differences were noted in serum
`TNF levels between the children with inflammatory
`bowel disease and the control population. Values were
`generally within the range of the lower limit of detection
`of the assay. In the children with inflammatory bowel
`disease, there was no significant correlation between
`TNF levels and disease activity or growth parameters.
`Our observations suegest that elevated TNF levels are
`not associated with inflammatory bowel disease in chil(cid:173)
`dren. Key Words: Tumor necrosis factor-Inflammatory
`bowel disease-Children-Anorexia-Cachexia.
`
`Important causes of growth failure in children
`with inflammatory bowel disease (IBD) include
`chronic undernutrition and the daily use of cortico(cid:173)
`steroid therapy (1-3). Poor nutrition is usually due
`to inadequate protein and calorie intake as well as
`varying degrees of malabsorption. Frequently, chil(cid:173)
`dren are seen with minimal gastrointestinal symp(cid:173)
`toms but in whom severe anorexia and profound
`growth impairment are evident (4). While the etiol(cid:173)
`ogy of this anorexia has been unclear, recent re(cid:173)
`ports of raised serum levels of tumor necrosis
`factor-a (TNF) in cachexia associated with chronic
`infection (5) and acquired immunedeficiency syn(cid:173)
`drome (6) have suggested a similar possible role in
`IBD. In some patients with cancer-associated wast(cid:173)
`ing, increased TNF levels have been found (7),
`
`Address correspondence and reprint requests to Dr. J. S.
`Hyams at Division of Pediatric Gastroenterology & Nutrition,
`Department of Pediatrics, Hartford Hospital, 80 Seymour Street,
`Hartford, Cf 06115, U.S.A.
`Accepted for publication at the office of the Founding Editor.
`
`while in others, normal levels have been demon(cid:173)
`strated (8). These observations have prompted us to
`measure TNF in a large group of children with ul(cid:173)
`cerative colitis and Crohn's disease.
`
`METHODS
`
`Patients
`
`Forty-five children and adolescents with IBD (18
`ulcerative colitis, 27 Crohn's disease) were re(cid:173)
`cruited from the clinical practice of the Division of
`Pediatric Gastroenterology & Nutrition at Hart(cid:173)
`ford HospitaL The patients with ulcerative colitis (9
`males) ranged in age from 3 to 17 years (mean of
`ll.5 years) and those with Crohn's disease (16
`males) from 7.5 to 18 years (mean of l3 years).
`Thirty-seven serum samples were obtained from
`the patients with ulcerative colitis (range of 1 to 5
`per patient, mean of 1.8) and 57 samples from the
`patients with Crohn's disease (range of 1 to 5 per
`
`233
`
`

`

`234
`
`J. S. HYAMS ET AL.
`
`patient, mean of 1.9) at varying times during their
`clinical course. If more than one specimen was ob(cid:173)
`tained from a particular patient, the second one was
`at least 6 months after the initial sample. At the time
`of serum sampling, data on height, weight, growth
`velocity, Tanner stage, medications, and disease
`activity were recorded. Height measurements were
`performed by taking three separate readings on a
`wall-mounted stadiometer at the time of each visit.
`The activity of disease was classified as mild, mod(cid:173)
`erate, or severe by previously established criteria
`(9,10). A complete blood count, erythrocyte sedi(cid:173)
`mentation rate, serum albumin level, and serum C(cid:173)
`terminal type I procoUagen concentration were gen(cid:173)
`erally obtained at the time of each venipuncture.
`Six serum samples were obtained from individuals
`who had reached fuU skeletal maturity and these
`were not used in the analysis of the relationship of
`growth parameters to TNF.
`Twenty-five children (mean age of 10.5 years)
`with a history, physical examination, and labora(cid:173)
`tory evaluation consistent with a diagnosis of func(cid:173)
`tional bowel disease (e.g., irritable bowel syndrome
`and constipation) and normal growth served as our
`control population. A single serum sample was ob(cid:173)
`tained from each of these patients at their initial visit.
`Informed consent was obtained in all cases, and
`the study was approved by the Investigational Re(cid:173)
`view Committee at Hartford Hospital.
`
`Tumor Necrosis Factor-o: Assay
`
`Serum samples were frozen at - 20°C within l-3
`h of collection and then stored at - 70°C until the
`performance of further studies. The assay for TNF
`was performed using a commercially available kit,
`CENTOCOR TNF RIA (CENTOCOR, Malvern,
`PA, U.S.A.), which employs a solid phase radioim(cid:173)
`munoassay in a "forward sandwich" configuration.
`A standard curve is constructed plotting counts per
`minute on the Y axis vs. TNF concentration in pg/
`ml on the X axis. The concentration of controls and
`unknowns is read by direct interpolation from the
`graph. The curve covers a dynamic range from 0 to
`4,500 pg/ml ofTNF. Samples exceeding the highest
`standard are diluted using the zero standard as dil(cid:173)
`uent. The analytical sensitivity, as defined by the
`zero standard plus 2 SD, is equivalent to approxi(cid:173)
`mately 36 pg/ml. TNF levels of less than 100 pg/ml
`are considered unremarkable using this assay sys(cid:173)
`tem. The coefficient of variation between assays is
`less than 10%. Cross-reactivity data supplied by the
`
`J Pediatr Gastroentero/ Nutr, Vol. /2 , No. 2, 1991
`
`manufacturer indicate that there is detectable cross(cid:173)
`reactivity with TNF of Rhesus monkey origin, but
`that there is no detectable cross-reactivity with tu(cid:173)
`mor necrosis factor-~.
`
`Type I Procollagen Assay
`
`The radioimmunoassay for the C-terminal frag(cid:173)
`ment of type I procollagen, which reflects growth
`velocity in children with IBD, was performed as
`previously reported (11,12).
`
`Statistical Analysis
`
`Statistical significance of differences between
`mean values was determined using t tests, analysis
`of variance, and the Scheffe procedure for multiple
`comparisons. Pearson correlation coefficients were
`used to examine the relationship of TNF to disease
`activity, corticosteroid therapy, growth velocity,
`erythrocyte sedimentation rate, serum albumin, and
`C-terminal type I procollagen.
`
`RESULTS
`
`As can be seen in Fig. 1, TNF values for patients
`with either mild or moderate/severe disease were
`similar to those found in the control population.
`Values were generally within the range of the lower
`limit of detection of the assay. There was no signif(cid:173)
`icant correlation between TNF and growth velocity
`(p > 0.1), use or nonuse of corticosteroids (p > 0.1),
`C-terminal type I procollagen (p > 0.3), erythrocyte
`sedimentation rate (p > 0.4), or serum albumin (p >
`0.1). TNF values at the time of diagnosis of eight
`patient's with Crohn's disease and growth failure
`were no different than those in the control group.
`Values were similar in children with growth arrest
`( < 0.1 em/month) and those with adequate growth
`(>0.5 em/month).
`After analysis of the initial 94 serum samples and
`failure to detect elevated levels of TNF, we ob(cid:173)
`tained 13 additional samples and immediately mixed
`them with a protease inhibitor to prevent possible
`TNF degradation. Upon assay, none of the 13 sera
`revealed elevated TNF levels.
`
`DISCUSSION
`
`Tumor necrosis factor-a, a cytokine released by
`activated macrophages, has many important bio(cid:173)
`logic effects in mediating inflammation (13). It also
`
`

`

`TUMOR NECROSIS FACTOR
`
`235
`
`..... .
`. . . . . . .. .
`~nt~l +---------+---------+---------+---------+---------+---------+
`.. ....
`. . . . . . . . . . . . ...
`. . . . . . . . . . . . . . . . .
`. . . . . . . . . . . . .. . . . . . . .
`Mild +---------+---------+---------+---------+---------+---------+
`Severa +---------+---------+-------·--+---------+---------+---------+·
`30
`40
`50
`60
`70
`80
`90
`4S.1 (7. 0)
`
`x(SD)
`
`46.S (3. 5)
`
`47.9 (8.0)
`
`Moderate/
`
`•
`
`• •
`
`• ' • '
`
`•
`
`•
`
`•
`
`•
`
`FIG. 1. Serum levels of tumor necrosis facto r-Ot in 25 control children with functional bowel disease (25 samples) and 45 children
`with Inflammatory bowel disease (94 samples: 79 mild disease, 15 moderate/severe disease). No significant differences were
`noted.
`
`,., (pg/111)
`
`has potent catabolic effects on fat and muscle cells
`resulting in lipolysis and glycogenolysis, respec(cid:173)
`tively (14). Chronic administration of TNF to ex(cid:173)
`perimental animals results in anorexia, decreased
`nitrogen balance, and wasting (15). TNF also has
`significant effects on collagen metabolism (16-18)
`that might be expected to affect growth negatively.
`In the present study, no differences were found in
`TNF levels between children with IBD and an age(cid:173)
`matched control population of children with func(cid:173)
`tional bowel disease. In addition, there was no cor(cid:173)
`relation between clinical and biochemical growth
`parameters in children with IBD and TNF levels.
`No elevation ofTNF was seen in eight children with
`Crohn's disease and growth failure at the lime of
`initial presentation and before any therapy was in(cid:173)
`stituted.
`While our findings suggest that TNF may not be
`associated with disease activity or growth impair(cid:173)
`ment in children with IBD, additional studies may
`be necessary. It is possible that IBD patients might
`be exquisitely sensitive to low levels of TNF, or
`that there may be some diurnal variation in TNF
`elevation that was missed during our sampling pe(cid:173)
`riods. Additionally, many of our patients were re(cid:173)
`ceiving prednisone therapy at the time of serum
`sampling and glucocorticoids are potent inhibitors
`of TNF production (19}. While it is possible that
`TNF may be labile during storage, recent reports
`indicate that TNF is unlikely to degrade during fro(cid:173)
`zen storage (8,20). The serum samples stored with a
`protease inhibitor also did not reveal elevated TNF
`levels.
`In conclusion, raised serum levels of TNF were
`not detected in a large group of children with IBD
`
`with varying degrees of clinical severity and growth
`impairment. Alternative explanations for the an(cid:173)
`orexia commonly seen in IBD patients must await
`further study.
`
`Acknowledgment: The authors thank Dr. Burton Gold·
`berg for performance of the C-terminal type I procollagen
`assays and Ms. Debbie Clough and Ms. Judy Therrien for
`preparation of the manuscript.
`
`REFERENCES
`
`I. Kelts DG, Grand RJ, Shen G, Watkins JB, Werlin SL,
`Boehme C. Nutritional basis of growth failure in. children
`and adolescents with Crohn's disease. Gastroenterology
`1979;76:720-7.
`2. Kirschner BS, Klich JR, Kalman S, DeFavaro MY, Rosen(cid:173)
`berg I H. Reversal of growth retardation in Crohn's disease
`witb therapy emphasizing oral nutritional restitution. Gas(cid:173)
`troenterology 1981;80: 10-5.
`3. Hyams JS, Moore RE, Leichtner AM, Carey DE, Goldberg
`BD. Relationship of type I procollagen to corticosteroid
`therapy in children with inflammatory bowel disease. J Pe(cid:173)
`diatr 1988;112:893-8.
`4. KanofME, Lake AM, Bayless TM. Decreased height veloc(cid:173)
`ity in children and adolescents before the diagnosis of
`Crohn's disease. Gastroenterology 1988;95: 1523-7.
`5. Scuderi P, Sterling K E, Lam KS, et al. Raised serum levels
`of tumour necrosis factor in parasitic infections. Lancet
`1986;2: 1364-5.
`6. Lahdevirta J, Maury CPJ, Teppo AM, Repo H. Elevated
`levels of circulating cachectin tumor necrosis factor in pa(cid:173)
`tients with acquired immunodeficiency syndrome. Am J
`Med 1988;85:289-91.
`7. Balk will F, Osborne R, Burke F, et al. Evidence of tumour
`necrosis factor/cachectin production in cancer. Lancet
`1987;2: 1229-32.
`8. Socher SH, Martinez D, Craig JB, Kuhn JG, Oliff A. Tumor
`necrosis factor not detectable in patients with clinical cancer
`cachexia. JNCI 1988;80:595-8.
`9. Leonard-Jones JE, Ritchie JK, Hilder W, Spicer CC. As(cid:173)
`sessment of severity in colitis: a preliminary study. Gut
`1975;16:579-84.
`
`J Pediatr Ga11roentero/ Nutr. Vol. 12, No. 2, 1991
`
`

`

`236
`
`J. S. HYAMS ET AL.
`
`10. Harvey RF, Bradshaw JM. A simple index of Crohn's dis(cid:173)
`ease activity. Lancet 1980;1:514.
`II. Taubman MB, Goldberg B, Sherr CJ. Radioimmunoassay
`for human procollagen. Science 1974;186:1115-7.
`12. Taubman MB, Kammerman S, Goldberg B. Radioimmuno(cid:173)
`assay of procollagen in serum of patients with Paget's dis·
`ease of bone. Proc Soc Exp Bioi Med 1976; 152:284-7.
`13. Tracey KJ, Vlassara H, Cerami A. Cachectin/tumour necro(cid:173)
`sis factor. Lancet 1989;1:1122-5.
`14. Fong Y, Lowry Sl', Cerami A. Cachectin/TNF: a macro(cid:173)
`phage protein that induces cachexia and shock. J Parenter
`Enter Nutr 1988;12:728-7.
`15. Tracey KJ, Wei H, Manogue KR. Cachectin/tumor necrosis
`factor induces cachexia, anemia, and inflammation. J Exp
`Med 1988;167:1211-27.
`
`16. Bertolini DR, Ned win GE, Bringman TS, Smith DD, Mundy
`GR. Stimulation of bone resorption and inhibition of bone
`formation in vitro by human tumour necrosis factors. Nature
`(Lond) 1986;319:516-8.
`17. Canalis E. Effects of tumor necrosis factor on bone forma·
`lion in vitro . Endocrinology 1987;11:1596-604.
`18. Buck M, Rippe R, Chojkier M, Brenner DA. Inhibition of
`collagen gene expression by tumor necrosis factor: molecu(cid:173)
`lar mechanisms. Hepatology 1989;10:617.
`19. Beutler B, Krochin N, Milsork IW, Luedke C, Cerami A.
`Control of cachectin (tumor necrosis factor) synthesis:
`mechanisms of endotoxin resistance. Science 1986;232:977-
`80.
`20. Teppo AM, Maury CPJ. Radioimmunoassay of tumor necro(cid:173)
`sis factor in serum. Clin Chem 1987;33:2024-7.
`
`J Pediatr Gastroenterol Nutr, Vol. 12, No.2, /99/
`
`