Duodenitis in children: clinical, endoscopic, and pathological aspects

0016-5107/87/3305-0366$02.00

GASTROINTESTINAL ENDOSCOPY Copyright © 1987 by the American Society for Gastrointestinal Endoscopy

Duodenitis in children: clinical, endoscopic, and pathological aspects Giuseppina Oderda, MD, Marco Forni, MD Laura Farina, MD, Domenico Dell'Olio, MD Nicoletta Ansaldi, MD Torino, Italy

In a retrospective study of 320 gastroduodenoscopies performed on children in the last 2 years, 32 cases of duodenal damage, either isolated or associated with esophagitis, gastritis, or both, were selected. Twenty matched children with dyspeptic symptoms but an endoscopically normal duodenum were chosen as controls. Histological sections of fiberoptic biopsy specimens were submitted to double-blind examination by two observers: only four fulfilled accepted criteria for histological duodenitis. The concordance between endoscopic and histological resuhs in our patients was 13.8%. No changes of duodenal endoscopic appearance predictive of duodenitis were identified. In 8 more cases (6 patients and 2 controls) the histological examination showed an isolated lymphocyte and plasma cell infiltration of the lamina propria without any additional damage. We called this picture "minimal change duodenitis" and considered it as a variation of normal. In 17 patients and 11 controls, basal and maximal acid outputs were calculated and no significant differences were found. We concluded that duodenitis in children may be present, although rare, and its diagnosis requires histological examination of duodenal mucosa. (Gastrointest Endosc 1987;33:366-

369)

Fiberoptic endoscopy has made possible direct visualization of the duodenal bulb and accurate target biopsy in adults and children.l,2 Duodenal inflammatory changes without ulceration have been described in adults and variously named duodenitis, nonspecific duodenitis, or active duodenitis. 3,4 In adults the frequency of duodenitis at duodenoscopy varies between 1.9%5 and 30%6; it is associated with normal or increased gastric acid secretion and with variable dyspeptic symptoms which mimic peptic ulcer disease. 7 Endoscopically, the mucosa shows patchy or diffuse erythema, friability, superficial erosions, and nodularity.8,9 The histological picture has been carefully defined by Whitehead et al. lO and Hasan et alY Clinical significance, diagnostic criteria, natural history, and treatment of this entity are still controversial. 12 Only Received May 6, 1986. For revision July 2, 1986. Accepted August 4, 1986.

From the Cattedra di Puericultura, Servizio di Gastroenterologia, Universita di Torino, and the Servizio di Anotomia Patologica, Ospedale Infantile Regina Margherita di Torino, Italy. Reprint requests: Dr. Giuseppina Oderda, Cattedra di Puericultura, Servizio di Gastroenterologia, Universita di Torino, Piazza Polonia 94, 10126 Torino, Italy.

366

a few cases of duodenitis have been reported in children, 13,14 and clinical and pathological descriptions are lacking. To define frequency, clinical picture, and histological and endoscopic patterns of nonspecific duodenitis in children, we have retrospectively reviewed gastroduodenoscopies performed at our pediatric gastrointestinal unit in the last 2 years. PATIENTS AND METHODS

Three hundred twenty consecutive gastroduodenoscopies all performed by the same endoscopist (G,O.) in the last 2 years were reviewed and 32 cases with endoscopic duodenal changes were found. Twenty-one were male and 11 were female, aged 2 to 15 years (mean, 8 years 9 months). Indications for gastroscopy were severe abdominal pain, pernicious vomiting, anorexia with weight loss or failure to thrive, diarrhea, and occult blood in the stools. Endoscopically, 9 showed isolated duodenal damage, while the others also had esophagitis (5 cases), superficial gastritis (8 cases), or both (10 cases). Twenty children matched for age and sex with chronic esophagitis (7 cases), superficial gastritis (4 cases), or both (5 cases) or a normal gastroduodenoscopy (4 cases) were chosen as controls; all had an endoscopically normal duodenum. Thirteen had various dysGASTROINTESTINAL ENDOSCOPY

peptic symptoms, while the other 7 had no complaints. Four of the latter were investigated after treatment with H 2receptor antagonists for peptic ulcer disease, 2 to rule out celiac disease and 1 for follow-up after caustic injury of the esophagus. Clinical scores of the frequency and severity of the major symptoms are given in Table 1. Basal and maximal acid output (BAO and MAO) after pentagastrin stimulation was studied in 17 patients and 11 controls. 15 Ulcerogenic drug ingestion and food allergy were excluded by clinical history. Bacterial, fungal, or parasitic diseases were ruled out by means of appropriate laboratory tests. Gastroscopy was performed with an Olympus GIF-P3. Endoscopic scores of the duodenal mucosa considering color, friability, edema, and bulb deformity are given in Table 2. Multiple mucosal specimens were taken with a fenestrated biopsy forceps from the duodenal bulb. Two mucosal specimens were obtained in 23 cases and three specimens in the remaining 29. Sections, routinely fixed and processed, were stained with hematoxylin and eosin.1 6 In 6 cases (3 patients and 3 controls) the specimens were not adequate for evaluation because of mechanical artifacts or insufficient sampling. All the biopsies included the mucosal layer and portions of the muscularis mucosae. Only occasional fragments of submucosal glands were also present. The histological parameters considered and the diagnostic criteria for duodenitis are shown in Table 3. The significance of the differences between patients and controls was assessed by means of the Mann and Whitney test1 7 for clinical and endoscopic scores, Fisher's exact test for the frequency of histological diagnosis of duodenitis,18 and Student's t test for BAO and MAO values. 19 Spearman's rho and Kendall's tau were calculated to evaluate the correlation between clinical and endoscopic scores in the patient groUp.20 RESULTS

The endoscopic appearance of the duodenum was the basis for allotting each case to the patient or

control group. The difference in the endoscopic score between the groups was highly significant (mean endoscopic score: controls = 4, patients = 8.6 p < 0.001). Duodenal endoscopy in patients showed mucosa friability (25 cases, 81 %), erosions with or without bleeding (19 cases, 59%), edema (17 cases, 53%), hyperemia (16 cases, 50%), bulb deformity (4 cases, 12%), and nodularity (2 cases, 6%) variously associated. The difference in clinical score between the groups was statistically significant (mean clinical score: controls = 5.4, patients = 6.7, p = 0.025). Abdominal pain was the most frequent symptom in patients (25 cases, 81 %). It was more commonly periumbilical, less frequently epigastric, usually present when arising in the morning, not related to food ingestion, and showed variable duration, severity, and frequency. In all subjects the pain lasted for longer than 6 weeks. Less frequent symptoms were vomiting Table 2. Endoscopic score adopted to assess duodenal damage Score Color Normal Pale gray Patchy hyperemia Marked diffuse hyperemia Friability Absent Mild Marked Erosions Edema Absent Mild Marked Bulb deformity Absent Mild Marked

1 2

3 4 1

2 3 4

1 2

3 1 2 3

Table 1. Clinical score adopted to evaluate the severity of symptoms in patients and controls Symptoms

Frequency

Severity

Abdominal pain

Frequent Frequent Occasional Occasional Absent Frequent Occasional Uncommon Absent

Severe Mild Severe Mild

Vomiting

Anorexia

Diarrhea Nausea Headache Halitosis VOLUME 33, NO. 5, 1987

4

3 3 2 1 4

3 2 1

Severe Present Mild Absent Present Present Present Present

Score

4

3 2 1 1 1 1 1

367

Table 3. Histological criteria for the diagnosis of duodenitis Histological parameters Villi Surface epithelium

Lamina propria

Neutrophil polymorph infiltration

Duodenitis· Short, stubby, or flattened Flattened or eroded epithelium, intraepitheliallymphocytes Severe increase in inflammatory cells: lymphocytes, plasma cells, and polymorphs Marked

Minimal change duodenitis Normal villi Normal surface epithelium

Increased cellularity with lymphocyte and plasma cell infiltration Absent

• From Whitehead et al lO and Hasan et al.l\ (modified).

(9 cases, 28%), anorexia (8 cases, 18%), headache (3 cases, 9%), and halitosis in 1 case. The same symptoms were reported by controls, but the severity and the frequency were different. Abdominal pain was present in 9 cases (45%), headache in 6 cases (30%), vomiting in 5 cases (25%), halitosis in 3 cases (15%), nausea in 2 cases (10%), and anorexia in 1 case. Clinical symptoms were more severe in the 10 children with esophagogastroduodenitis (mean clinical score = 7) than in the 9 children with isolated duodenal damage (mean score = 6.4). Within the patient groups the clinical and endoscopic scores were not significantly correlated (rho = 0.025, tau = 0.066). In 17 patients mean BAO was 0.107 mEqjkg/hour and mean MAO was 0.373 mEq/kg/hour and in 11 controls mean MAO was 0.346 mEqjkg/hour; the differences were not statistically significant. In particular, MAO was elevated in 10 patients (59%) and 4 controls (36%); the difference was not significant. Histological examination of duodenal biopsy of 29 patients showed normal mucosa in 19 and a mild chronic inflammation with increased infiltration of plasma cells in the lamina propria in 6 which was labeled "minimal change duodenitis." The Whitehead et a1. 10 and Hasan et a1. 11 criteria for the histological diagnosis of duodenitis were satisfied in only 4 patients. Their biopsies showed severely increased cellularity of the lamina propria, abnormalities of the surface epithelium with flattened cells and an increased number of intraepitheliallymphocytes, blunted villi, and infiltration of the epithelium and lamina propria by neutrophils. Only occasionally rare eosinophils were noted in the lamina propria. Brunner glands, when included in the sample, showed no alterations in both groups. In 17 controls histological examination showed normal mucosa in 15 and "minimal change duodenitis" in 2. Gastric superficial epithelial metaplasia was never observed. Table 4 shows clinical symptoms, BAO and MAO values, and endoscopic appearance of the duodenum in the 4 patients with histological duodenitis. 368

Concordance between endoscopic and histological results in our patients was 13.8%. There were no endoscopic changes predictive of the histological lesion. DISCUSSION

Our retrospective study lacks a true control group since, for ethical reasons, we did not examine asymptomatic children. Thus, the controls were subjects with gastrointestinal symptoms sufficiently severe to require endoscopy and whose duodenum had a normal endoscopic appearance. Our data show that duodenitis in childhood does not show any typical clinical symptom complex, that more severe symptoms are associated with a concomitant esophagitis or gastritis or both, and that there is no peculiar symptom in the 9 cases of isolated endoscopic duodenal damage or in the 4 cases of histological duodenitis. The observed frequency of endoscopic duodenal damage is 10% of total upper gastrointestinal endoscopies performed in the last 2 years at our pediatric gastrointestinal unit; the prevalence of duodenitis in the population studied is 4/320. The endoscopic picture of a damaged duodenum is not uniform since hyperemia, edema, congestion, friability, and erosions, variably associated, may be seen. In our cases mucosal friability was the most common finding followed by erosions. We did not make the distinction between chronic duodenitis (with plasma cell and lymphocyte infiltration) and acute duodenitis (with neutrophil polymorph infiltration) as some authors have. 3 • 21 We observed 8 cases (6 patients and 2 controls) with isolated lymphocyte and plasma cell infiltration of the lamina propria without any additional change. We called it "minimal change duodenitis" and temporarily considered it as a variation of norma1. 22 However, it is conceivable that in children this may represent the initial phase of duodenitis and that additional inflammation may appear later to give a typical histological picture. According to the strict histological criteria adopted, GASTROINTESTINAL ENDOSCOPY

Table 4. Clinical, endoscopic data and gastric secretion of four patients with histological duodenitis Patient no.

Age, sex

Endoscopy

Symptoms

4.5, M

Gastritis, duodenitis

Abdominal pain, vomiting Anorexia, diarrhea Abdominal pain, vomiting, irregular bowel movements Abdominal pain, vomiting

2

13.0, M

Duodenitis

3

8.0, F

Gastritis, duodenitis

4

1l.5, M

Gastritis, duodenitis

a concordance between endoscopic and histological findings was observed in 13.8% of the cases. This concordance has been reported in adults as varying from 82%6 to 70%22 to 57%3 when broader histological criteria were used, while it was found to be 27% by Cheli and Giacosa 21 who followed our same criteria. We observed histologically normal duodenal mucosa in the presence of an abnormal endoscopic appearance, but we never observed the opposite, while in adults 25% of endoscopically normal duodenums may show histological damage. 3 According to McCallum et a1.,3 the low level of correlation observed may be due to two different causes: (1) The endoscopic examination can be traumatic. The initial inspection of the bulb should be performed from the pyloric canal and excessive suction or insufflation avoided. (2) The reliability of endoscopic duodenal biopsy specimens is often questionable. This is particularly true using a pediatric instrument with its small forceps. We also investigated the relationship between endoscopic duodenal damage and gastric acid secretion in 50% of the cases, but the differences between patients and controls were not significant. Some authors suggest that duodenitis is a peptic lesion due to the frequent association between gastric acid hypersecretion, peptic ulcer, and duodenitis7.24-26 and to the similarity of the histological changes in ulcer-associated duodenitis and nonspecific duodenitis. lI However, other groups have found a high incidence of gastric acid hyposecretion in their patients and suggest that duodenitis is not a peptic lesion. 27,28

REFERENCES 1. Tedesco FJ, Goldstein PO, Gleason WA, et a1. Upper gastrointestinal endoscopy in the pediatric patient. Gastroenterology 1976;70:492-4. 2. Ament ME, Christie DL. Upper gastrointestinal fibreoptic endoscopy in pediatric patients. Gastroenterology 1977;72:12448. 3. CottonPB, Price AB, Tighe JR, Beales JSM. Preliminary evaluation of "duodenitis" by endoscopy and biopsy. Br Med J 1973;3:430-3. 4. Gelzayd EA, Gelfand OW, Rinaldo JA Jr. Non·specific-duodenitis. A distinct clinical entity. Gastrointest Endosc 1973;19:1318. 5. Kasugai T, Kuno N, Aoki A, Kizu M, Kobayashi S. Fibroduo-

VOLUME 33, NO.5, 1987

6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.

BAO

MAO

Normal

Mildly elevated

Elevated

Elevated

Normal

Normal

Mildly elevated

Severely elevated

denoscopy: analysis of 353 examinations. Gastrointest Endosc 1971;18:9-11. Gregg JA, Garabedian N. Duodenitis. Am J Gastroenterol 1974;61:177-84. Thomson WO, Joffe SN, Robertson AG, Lee FD, Imrie CW, Blumgart LH. Is duodenitis a dyspeptic myth? Lancet 1977; 1:1197-8. Gelzayd EA, Biederman MA, Gelfand OW. Changing concepts of duodenitis. Am J Gastroenterol 1975;64:213-6. Forrester AW, Joffe SN, Lee FD. The endoscopic and histological features of peptic duodenitis. Scand J Gastroenterol 1979;14(suppI54):18-22. Whitehead R, Roca M, Meikle DO, Skinner J, Truelove SC. The histological classification of duodenitis in fibreoptic biopsy specimens. Digestion 1975;13:129-36. Hasan M, Hay F, Sircus W, Ferguson A. Nature of the inflammatory cell infiltrate in duodenitis. J Clin Pathol 1983;36:2808. Palmer ED. Common duodenitis of any clinical import? JAMA 1974;230:599-602. Hargrove CB, Ulsen MH, Shub MD. Upper gastrointestinal endoscopy in infants: diagnostic usefulness and safety. Pediatrics 1984;74:828-31. Miller V, Doig CM. Upper gastrointestinal endoscopy. Personal practice. Arch Dis Child 1984;59:1100-2. Ansaldi N, Santini B, Barbera C, et a1. Iperacidita gastrica nell'infanzia e Ie sue conseguenze sulla funzionalita intestinale. Italian J Pediatr 1975;2:11-9. Luna LG. Manual of histologic staining methods of the Armed Forces Institute of Pathology. New York: McGraw-Hill, 1968. Mann HB, Whitney DR. On a test of whether one of two random variables is stochastically larger than the other. Ann Math Statist 1947;18:50-60. Barnard GA. Significance test for 2 x 2 tables. Biometrika 1947;34:123-38. Armitage P. Statistical methods in medical research. Oxford: Blackwell, 1980:116-26. Marascuilo LA, McSweeney ME. Nonparametric and distribution-free methods for the social sciences. Monterey: Brooks/ Cole, 1977:431-46. Cheli R, Giacosa A. Inflammatory cell count and identification in chronic non-specific duodenitis. Endoscopy 1977;9:91-5. Jenkins 0, Goodall A, Gillet FR, Scott BB. Defining duodenitis: quantitative histological study of mucosal responses and their correlations. J Clin Pathol 1985;38:1119-26. McCallum WR, Singh 0, Wollman J. Endoscopic and histologic correlations of the duodenal bulb. The spectrum of duodenitis. Arch Pathol Lab Moo 1979;103:169-72. Bayeli PF, Bargagli 0, Benelli R, Abate L. Studio clinico su 33 quadri istologici di duodenite cronica non specifica. Giorn Ital End Dig 1983;6:133-40. Kreuning J. Chronic non-specific-duodenitis. Delf: Meinema, 1978. Duodenitis-any progress? Lancet 1985;2:1222-3. Cheli R, Aste H, Ciancamerla G. Value of endoscopy in the diagnosis of chronic non-specific duodenitis. Rend GastroenteroI1975;7:23-6. Cheli R, Aste H. Duodenitis. Stuttgart: G Thieme, 1976.

369