Background: The aim of the study was to evaluate the new hormonal entity oxyntomodulin-like immunoreactivity in malabsorption states, and to assess its potential in celiac disease management.
Methods: We measured basal and postprandial oxyntomodulin-like immunoreactivity values in 35 children divided into 3 groups: group 1 was composed of 13 children with celiac disease, either under a gluten-free diet (8 patients) or normal diet (5 patients); group 2 was composed of 8 children hospitalized for gastroenteritis or chronic diarrhea, without biological evidence of malabsorption nor abnormal jejunal mucosa; group 3 was composed of 22 control subjects.
Results: Fasting and meal-stimulated levels in the control group were 71 ± 10 and 130 ± 26 pmol/l, respectively. Mean concentrations were elevated in patients with celiac disease (basal = 349 ± 254 pmol/l, postprandial = 446 ± 332 pmol/l) and in the group 2 (basal = 139 ± 58 pmol/l, postprandial = 218 ± 85 pmol/l), but the difference with control subjects did not reach statistical significance. In children with celiac disease, basal and stimulated values correlated with the degree of malabsorption as assessed by hemoglobin (p = 0.006 and p = 0.01, respectively) and serum folate concentrations (p = 0.03 and p = 0.02, respectively).
Conclusions: Oxyntomodulin-like immunoreactivity is noticeably higher in healthy children than previously measured in healthy adult subjects. This hormonal parameter is not an adequate diagnostic tool in celiac disease. Nevertheless, in the context of celiac disease, its elevation reflects the degree of malabsorption and may provide a quantitative approach of the extent of mucosal damage.
*INSERM U 376, Hôpital Arnaud de Villeneuve; †Service de Médecine Interne A, Hôpital Saint-Eloi; ‡Service de Pédiatrie II, Hôpital Arnaud de Villeneuve, CHU de Montpellier; and §Département de l'Information Médicale, Hôpital Gaston Doumergue, CHU de Nîmes, France
Received October 1, 1997; revised May 7, 1998; accepted May 8, 1998.
Address correspondence and reprint requests to Pr. A. Le Quellec, Service de Médecine Interne A, Hôpital Saint-Eloi, 2 Avenue Bertin-Sans, 34 295 Montpellier Cedex 5, France.
We recently defined oxyntomodulin-like immunoreactivity (OLI) as a group of proglucagon-derived intestinal peptides that may participate in an enterogastrone function (1). This hormonal entity is secreted by ileal and colonic L cells in response to intraluminal nutrient stimuli, predominantly lipids and carbohydrates. In human adults, OLI plasma concentration increases twofold within the first hour after a meal, and then gradually declines to baseline (1). Meal-stimulated OLI values are at the same levels as concentrations needed to inhibit gastric acid secretion when experimentally injected (2-4).
To date, information about OLI has only been available in adults, and attention has been focused mostly on duodenal ulcer disease (5). Yet, other pathologic situations characterized by absorption deficiency or transit-time acceleration are known to be associated with important variations of gut-hormone concentrations, especially dumping syndrome (6), short-bowel syndrome (7), and tropical sprue (8). Celiac disease is responsible for complex abnormalities in intestinal peptide release (9). Because of proximal villous atrophy, increased amounts of unabsorbed nutrients may reach the distal gut mucosa and trigger transitory hyperactivation of the endocrine cells, resulting in an exaggerated release of several peptides.
To evaluate OLI variations in malabsorption conditions and to assess their potential interest in celiac disease management, we have measured basal and meal-stimulated OLI concentrations in two groups of children with celiac disease or with various digestive diseases without malabsorption. Children admitted for staturo-ponderal retardation or weight abnormalities in which normal peroral jejunal biopsies were obtained were used as a control group.
MATERIALS AND METHODS
This prospective study took place between 1993 and 1996 in our department of pediatric gastroenterology. All children whose pathologic state justified gastroduodenoscopy with intestinal biopsies were asked to enter the study. The subjects were classified into three groups according to final diagnosis:
* Group 1: Celiac disease assessed by classic criteria: total or subtotal villous atrophy associated with hyperplasia of the crypts and raised intraepithelial lymphocyte count and clinical remission and improvement of villous structure after institution of a gluten-free diet (10).
* Group 2: Various digestive diseases including gastroenteritis, lactose intolerance, or colonic motor dysfunction; normal findings in analysis of jejunal biopsy specimens.
* Group 3: Children admitted for staturoponderal retardation or weight loss; no evidence of digestive disease, normal findings in endoscopy, and no mucosal damage.
Children with history of abdominal surgery, diabetes mellitus or renal insufficiency were not included in the study. Any drug that affects intestinal function was discontinued 3 days before the hormonal evaluation.
Small Bowel Biopsies
Biopsies were performed with an endoscope. The mucosal findings were histologically divided into four groups: normal, partial villous atrophy, subtotal villous atrophy, and total villous atrophy.
Characterization of Malabsorption in Group 1
Routine laboratory tests included: red cell count, partial prothrombin time, blood concentrations of hemoglobin, folic acid, vitamin B12, calcium, iron, albumin, and cholesterol. IgA and IgG antigliadin antibodies were quantified in each subject by enzyme-linked immunosorbent technique.
Measurement of Oxyntomodulin-like Immunoreactivity
Three days or more after the gastroduodenoscopy OLI was measured. An indwelling antecubital venous catheter was placed at 11:30 A.M. Blood samples were drawn just before feeding and 60 minutes after the beginning of a standardized meal (55% carbohydrates, 30% fat, and 15% protein) at 12 Noon. The energy value of the meal represented 15% of the total theoretical daily energy intake according to age, body weight, and body height (11). Blood samples were drawn into chilled evacuated glass tubes containing 15% ethylenediamine tetraacetic acid (EDTA) and 250,000 IU aprotinin (Vacutainer A3206SV186; Becton Dickinson, Meylan, France). Samples were centrifuged at 800g for 10 minutes at 4°C within 15 minutes of venipuncture and stored at -30°C until assay.
Radioimmunoassay of Oxyntomodulin-like Peptides
The level of OLI in plasma was evaluated in triplicate using the radioimmunoassay (RIA) that we developed in the laboratory, as previously described (1). Oxyntomodulin consists of the entire glucagon sequence, extended by a C-terminal basic octapeptide that is also present in other proglucagon-derived fragments (Fig. 1) and supports the biologic specificity of the molecule (12). Briefly, the assay used a polyclonal rabbit antiserum directed against the synthetic octapeptide (13). The C-terminal nonadecapeptide fragment of oxyntomodulin (oxyntomodulin 19-37) and the mono-(125I)-Tyr-oxyntomodulin (19-37) were used as standard and tracer, respectively. Sensitivity of the RIA for OLI was 1 pmol/l, the mid range of the assay was 13.4 pmol/l, and the mean slope of the dose-response curve was 1.03 (n = 6). The within-assay precision was 8% in plasma with an OLI concentration of 24 ± 2 pmol/l (n = 6), and the between-assay precision of the same plasma was 13.6% with a mean OLI concentration of 22 ± 3 pmol/l (n = 7). Cross-reactivity of our antiserum with oxyntomodulin and glicentin was 105% and 97%, respectively (14). No cross-reactivity existed with glucagon.
All values are represented as mean ± SEM. Statistical analyses were performed using nonparametric tests: Mann-Whitney test for the comparison of basal and meal-stimulated OLI concentrations among the three groups, and the Spearman rank order method for evaluating correlations. A value of p < 0.05 was considered statistically significant.
The study was approved by the local Ethics Committee (Comité Consultatif de Protection des Personnes). Written parental informed consent was obtained for each subject, according to French legislation on biomedical research (Loi Huriet).
Group 1 was composed of 13 subjects: 7 boys, 6 girls; mean age, 6 years; range, 10 months to 13 years. Group 2 was composed of 8 children: 5 boys, 3 girls; mean age, 2 years; range, 2 months to 6 years. Group 3 was composed of 22 subjects: 10 boys, 12 girls; mean age, 2 years; range, 2 months to 7 years. In the three groups taken as a whole, meal-stimulated plasma OLI concentrations were closely correlated with basal values (p = 0; r = 0.85; n = 38). Meal sensitivity was maintained even in the case of high OLI basal level. In group 1, basal and postprandial plasma OLI concentrations were negatively correlated with age (p = 0.02 and p = 0.03, respectively).
Fasting and postprandial OLI levels in the control group were 71 ± 10 and 130 ± 26 pmol/l, respectively. Mean plasma OLI concentrations were higher than normal in group 1 and group 2, but there was some scattering of the results in those two groups, and the differences did not reach statistical significance (Table 1). Because of this overlapping, sensitivity and specificity of any OLI threshold were too low for the diagnosis of celiac disease. One 1 year-old child with celiac disease had highly elevated OLI basal and stimulated levels (3382 pmol/l and 3755 pmol/l, respectively). He had untreated celiac disease and complete jejunal villous atrophy. This child had no additional disease or particular complication during the 2 years after hormonal evaluation.
Patients with celiac disease (group 1) were included in the study when the diagnosis was established, just before beginning a gluten-free diet (three patients); when disease relapsed, always as a consequence of noncompliance with diet (two patients); and after a period of consuming a gluten-free diet, when admitted for assessing regeneration of jejunal mucosa (eight patients). So, the spectrum of clinical and biologic malabsorption syndromes was very wide (Table 2). We found no correlation between OLI levels and the intensity of jejunal villous atrophy or of quantitative measurement of antigliadin antibodies. Levels of OLI were higher in the subgroup of children with untreated or relapsing celiac disease (n = 5) than in the group prescribed a gluten-free diet (n = 8), but statistical significance was not reached in results of nonparametric tests. Nevertheless, basal and stimulated OLI values correlated with the degree of malabsorption assessed by hemoglobin (p = 0.006 and p = 0.01, respectively) and serum folate concentrations (p = 0.03 and p = 0.02, respectively), as illustrated in Figure 2. A correlation was also found between stimulated OLI values and vitamin B12 (p = 0.03; r = 0.76; n = 8).
We report here for the first time data on the activity of oxyntomodulin-like peptides in children. In adults, plasma OLI concentrations range from 15 pmol/l (basal) to 40 pmol/l (postprandial) (1). The values we observed in children are higher (70-130 pmol/l in the control group). Similar observations have been made in infants of activities of several intestinal peptides including motilin, gastrin, neurotensin, and enteroglucagon (15). Furthermore, digestive endocrine cells are more sensitive to intraluminal stimuli in newborns than in adults (16). These molecules, just like OLI, exert precise regulatory functions on the gastrointestinal tract, and some of them are implicated in mucosal trophism and development. Accordingly, a hypertrophic or hyperreactive endocrine cell mass is thought to be necessary during childhood, when developmental processes take place (15).
Variations in the gut-hormone profiles can influence the pathophysiology and the clinical course of digestive diseases. For example, the postprandial responses of motilin, enteroglucagon, neurotensin, and peptide YY are significantly increased in children with cystic fibrosis. Altered gut hormone secretion may play a role in the pathophysiology of intestinal dysmotility in such patients (17). In celiac disease, secretin (9,18), vasoactive intestinal peptide (19), somatostatin (18), and gastric inhibitory peptide (9,18) plasma levels are diminished, whereas neurotensin (18) and some proglucagon-derived peptides (9) are markedly elevated.
Oxyntomodulin-like immunoreactivity is produced by L cells, which are predominantly located in the distal small intestine (20). Because mucosal atrophy is more pronounced in the proximal part of the small intestine in celiac disease, proximal malabsorption causes a reinforced distal nutrient stimulus that is combined with an intact OLI production capacity. In our study, a significant difference between children with and without (control subjects) celiac disease may have been masked because patients with celiac disease were somewhat older than control subjects and because OLI concentrations decrease with age. Yet, we found that basal and stimulated OLI concentrations were increased in celiac disease only when biologic signs of malabsorption were present, and that OLI hypersecretion correlated quantitatively with the intensity of the malabsorption. Neither the degree of villous atrophy (total, subtotal, or partial) at the duodenojejunal level nor adherence to the gluten-free diet seemed to be the primary factors responsible for plasma OLI increase. Furthermore, OLI concentrations were independent of IgA and IgG antigliadin antibodies levels, although the latter is a sensitive indicator of gluten reactivity. We observed the same discrepancy in adult celiac disease, with OLI plasma values markedly elevated in three patients whose antigliadin antibodies were within the normal range (unpublished data). In fact, follow-up studies of postpubertal patients with celiac disease have shown that clinical and biologic signs of malabsorption frequently do not correlate with dietary compliance and flattening of mucosa (21-23). Oxyntomodulin-like immunoreactivity must then be considered a marker of intestinal malabsorption and does not reflect the immune reactions to gliadin.
Our results are consistent with those in previous studies in which proglucagon-derived peptides were evaluated in celiac disease. Investigators used either an N-terminal glucagon-specific antibody, or the enteroglucagon concept. With the former method, a correlation was found between glucagon-like immunoreactivity and the biologic parameters of malabsorption after a gluten challenge in children (24). Enteroglucagon activity was measured as the difference between total plasma glucagon immunoreactivity and pancreatic glucagon immunoreactivity using two separate radioimmunoassays. Its basal and postprandial concentrations were elevated in children with untreated celiac disease (25). In another study in children, investigators found that meal-stimulated plasma enteroglucagon levels correlated with intestinal mucosal morphology, but there was a substantial overlap in values from treated and challenged patients (26). In adults, postprandial enteroglucagon correlated with fecal fat excretion, however basal enteroglucagon concentration did not vary before and after a gluten-free diet (27). It is important to note, however, that the OLI and the enteroglucagon concepts are not identical notwithstanding the recognition of glicentin and oxyntomodulin by N-terminal-directed but not by C-terminal-directed antiglucagon antibodies. Our antiserum directed against the C-terminal oxyntomodulin octapeptide bind all the proglucagon fragments exerting an inhibitory action on acid secretion, including oxyntomodulin (19-37) and shorter peptides that are devoid of glucagon-like activity (12). In contrast, the C-terminal-truncated form of glucagon, glucagon (1-21), which does not possess OLI biologic and immunologic specificity, is an enteroglucagon molecule (28). Oxyntomodulin-like immunoreactivity differs from that of enteroglucagon in that it is an immunologic concept based on common biologic features, with the epitope superimposed over the minimally active structure.
Because proglucagon-derived peptides were initially thought to have a role in mucosal growth regulation, the increased mucosal turnover in celiac patients was considered to be a consequence of their elevation (9). Nevertheless, the trophic effects of the enteroglucagon entity have subsequently been questioned (29). A recent report (30) suggests that GLP-2, one of the proglucagon-derived peptides produced by intestinal L cells, may be responsible for the observed mucosal growth when enteroglucagon peptides are released-that is, when the L cells are stimulated. The OLI peptides (oxyntomodulin, glicentin, and their C-terminal fragments) are characterized by the presence of the C-terminal octapeptide that bears both their biologic activity (12) and the epitope recognized by our radioimmunoassay (1,5). The specificity of the OLI peptides is directed toward the regulation of the digestive physiology at the level of the intestinal mucosa, where they modulate ionic fluxes (31) of the smooth muscles of the gut wall (32) and of the gastric oxyntic gland, where they inhibit gastric acid secretion (2,12). Accordingly, elevated plasma OLI concentrations observed in malabsorptive states could cause hypochlorhydria. However, regulation of gastric acid secretion depends on many different peptides, and gastric secretory capacity is either unchanged (33) or enhanced (34,35) in patients with celiac disease because of impaired secretin secretion (34) caused by the destruction of proximal secretin cells by the mucosal inflammatory process (36,37). The achlorhydric atrophic gastritis that frequently occurs in dermatitis herpetiformis associated with celiaclike enteropathy is not correlated with the quantity of gluten ingested or with the degree of villous atrophy and probably has a different immunopathogenesis (38).
In our study, the large overlap in OLI concentrations from the three groups and the high OLI values observed in some subjects with other digestive disorders make this hormonal entity an ineffective diagnostic tool for celiac disease, with low sensitivity and poor specificity. It must be recorded that absorption tests such as the xylose and the steatocrit tests are known to be of no use in screening for celiac disease (39). In fact, basal and meal-stimulated OLI provide information complementary to that obtained by immunologic or histologic means. The high level of OLI is related to the extent of jejunoileal lesions and thus allows a quantitative evaluation of the malabsorption syndrome. It could facilitate the follow-up of children after gluten withdrawal and improve the sensitivity of standard malabsorption tests (40). Because many patients do not adhere to a life-long gluten-free diet, OLI elevation could be an indication to advise a strict gluten diet when slight alterations of jejunal mucosa are found in an asymptomatic child. Indeed, atrophic mucosa accumulates carcinogenic oxidized metabolites of xenobiotics in celiac disease (41), and the risk of cancer could be linked to the expense of flattened mucosa.
Acknowledgment: This work was supported in part by a grant from the Association pour le Développement de l'Evaluation Thérapeutique (ADET).
1. Le Quellec A, Kervran A, Blache P, Ciurana AJ, Bataille D. Oxyntomodulin-like immunoreactivity: Diurnal profile of a new potential enterogastrone. J Clin Endoc Metab
2. Jarrousse C, Carles-Bonnet C, Niel H, et al. Inhibition of gastric acid secretion by oxyntomodulin and its [19-37] fragment in the conscious rat. Am J Physiol
3. Schjoldager BTG, Baldissera FGA, Mortensen PE, Holst JJ, Christiansen J. Oxyntomodulin: A potential hormone from the distal gut. Pharmacokinetics and effects on gastric acid and insulin secretion in man. Eur J Clin Invest
4. Veyrac M, Ribard D, Daures JP, et al. Inhibitory effect of the C-terminal octapeptide of oxyntomodulin on pentagastrin-stimulated gastric acid secretion in man. Scand J Gastroenterol
5. Le Quellec A, Kervran A, Blache P, Ciurana AJ, Bataille D. Diurnal profile of oxyntomodulin-like immunoreactivity in duodenal ulcer patients. Scand J Gastroenterol
6. Lawaetz O, Blackburn AM, Bloom SR, Aritas Y, Ralphs DNL. Gut hormone profile and gastric emptying in the dumping syndrome. A hypothesis concerning the pathogenesis. Scand J Gastroenterol
7. Andrews NJ, Irving MH. Human gut hormone profiles in patients with short bowel syndrome. Dig Dis Sci
8. Besterman HS, Cook GC, Sarson DL, et al. Gut hormones in tropical malabsorption. BMJ
9. Besterman HS, Bloom SR, Sarson DL, et al. Gut-hormone profile on coeliac disease. Lancet
10. Working group of European Society of Paediatric Gastroenterology and Nutrition. Revised criteria for diagnosis of coeliac disease. Arch Dis Child
11. Recommended dietary allowances
. 9th ed. Washington DC: National Academy of Sciences. 1980:1-185.
12. Bataille D. Oxyntomodulin and its related peptides. In: Lefebre PJ, ed. Glucagon III: Handbook of experimental pharmacology
. Heidelberg: Springer 1996;123:327-40.
13. Blache P, Kervran A, Martinez J, Bataille D. Development of an oxyntomodulin/glicentin C-terminal radioimmunoassay using a "thiol-maleoyl" coupling method for preparing the immunogen. Anal Biochem
14. Blache P, Kervran A, Le-Nguyen D, et al. The glucagon-containing peptides and their fragments in the rat gastro-enteropancreatic and central nervous system. Biomed Res
15. Lucas A, Bloom SR, Aynsley-Green A. Postnatal surges in plasma gut hormones in term and preterm infants. Biol Neonate
16. Lucas A, Bloom SR, Aynsley-Green A. Gut hormones and "minimal enteral feeding." Acta Paediatr Scand
17. Murphy MS, Brunetto AL, Pearson AD, et al. Gut hormones and gastrointestinal motility in children with cystic fibrosis. Dig Dis Sci
18. Hernanz A, Polanco I, Codoceo R, Lama R, Vazquez C. Gastrointestinal peptide profile in children with coeliac disease. J Pediatr Gastroenterol Nutr
19. Beck B, Villaume C, Debry G. Clinical aspects of VIP secretion. Acta Diabetol Lat
20. Green DW, Gomez G, Greeley GH. Gastrointestinal peptides. Gastroenterol Clin North Am
21. Kumar PJ, Walker-Smith J, Milla P, Harris G, Colyer J, Halliday R. The teenage coeliac: Follow-up study of 102 patients. Arch Dis Child
22. Mäki M, Lähdeaho ML, Hällstöm O, Viander M, Visakorpi JK. Postpubertal gluten challenge in coeliac disease. Arch Dis Child
23. Guandalini S, Ventura A, Ansaldi N, et al. Diagnosis of coeliac disease: Time for a change? Arch Dis Child
24. Carson DJ, Glasgow JFT, Buchanan KD, Sloan JM. Changes in N-terminal glucagon-like immunoreactivity and insulin during short-term gluten challenge in childhood coeliac disease. Gut
25. Kilander AF, Stenhammar L, Lindstedt G, Lundberg PA. Determination of enteroglucagon in plasma for detection of coeliac disease in children. Clin Chem
26. Stenhammar L, Strömberg L, Kilander AF, Lindstedt G, Lundberg PA. Plasma enteroglucagon in the control of childhood coeliac disease. J Pediatr Gastroenterol Nutr
27. Kilander AF, Dotevall G, Lindstedt G, Lundberg PA. Plasma enteroglucagon related to malabsorption in coeliac disease. Gut
28. Matsuyama T, Itoh H, Watanabe N, et al. Glucagon-(1-21)-peptide as an active enteroglucagon. Biomed Res
29. Gornacz GE, Ghatei MA, Al-Mukhtar MYT, et al. Plasma enteroglucagon and CCK levels and cell proliferation in defunctioned small bowel in the rat. Dig Dis Sci
30. Tsai CH, Hill M, Asa SL, Brubaker PL, Drucker DJ. Intestinal growth-promoting properties of glucagon-like peptide-2 in mice. Am J Physiol
31. Beauclair F, Pansu D, Rodier G, et al. Effects of oxyntomodulin and related peptides on the hydromineral transport through rat small intestine (abstract). Regul Pept
32. Rodier G, Magous R, Mochizuki T, et al. Effect of glicentin, oxyntomodulin and related peptides on isolated smooth muscle cells from rabbit antrum. Pflugers Arch
33. Kokkonen J, Similä S. Gastric function and absorption of vitamin B12 in children with coeliac disease. Eur J Pediatr
34. Rominger JM, Chey WY, Chang TM. Plasma secretin concentrations and gastric pH in healthy subjects and patients with digestive disease. Dig Dis Sci
35. Marcello U, Deganello A, Consolaro G, Zoppi G. Gastric secretory function in coeliac disease. Eur J Pediatr
36. Buchan AM, Grant S, Brown JC, Freeman HJ. A quantitative study of enteric endocrine cells in coeliac sprue. J Pediatr Gastroenterol Nutr
37. Sjölund K, Alumets J, Berg NO, Hakanson R, Sundler F. Duodenal endocrine cells in adult coeliac disease. Gut
38. Andersson H, Björkman AC, Gillberg R, Kastrup W, Mobacken H, Stockbrügger R. Influence of the amount of dietary gluten on gastrointestinal morphology and function in dermatitis herpetiformis. Hum Nutr Clin Nutr
39. Carroccio A, Iacono G, Montalto G, et al. Immunologic and absorptive tests in coeliac disease: Can they replace intestinal biopsies? Scand J Gastroenterol
40. Berg NO, Borulf S, Jacobson I, Lindberg T. How to approach the child suspected of malabsorption. Acta Paediatr Scand
41. Stahlberg MR, Hietanen E, Mäki M. Mucosal biotransformation rates in the small intestine of children. Gut
Keywords:© 1998 Lippincott Williams & Wilkins, Inc.
Celiac disease; Gastrointestinal hormones; Oxyntomodulin