Prevalence of celiac disease and its endoscopic markers among patients having routine upper gastrointestinal endoscopy

THE AMERICAN JOURNAL OF GASTROENTEROLOGY © 1999 by Am. Coll. of Gastroenterology Published by Elsevier Science Inc.

Vol. 94, No. 8, 1999 ISSN 0002-9270/99/$20.00 PII S0002-9270(99)00404-9

Prevalence of Celiac Disease and Its Endoscopic Markers Among Patients Having Routine Upper Gastrointestinal Endoscopy William Dickey, M.D., F.A.C.G., and Dermot Hughes, F.R.C.Path. Departments of Gastroenterology and Histopathology, Altnagelvin Hospital, Londonderry, Northern Ireland, United Kingdom

OBJECTIVE: The aim of this study was to determine the prevalence of duodenal villous atrophy (VA) among patients undergoing routine upper gastrointestinal (GI) endoscopy and the value of endoscopic markers for VA in selecting patients for duodenal biopsy. METHODS: One hundred and fifty adult patients with upper GI symptoms or iron-deficiency anemia had inspection and biopsy of the second part of the duodenum during endoscopy. Endoscopic markers for VA sought were mosaic or nodular mucosa, scalloping of duodenal folds, and reduction in number or absence of duodenal folds. RESULTS: Endoscopic markers were seen in seven patients (5%): scalloped folds with mosaic pattern mucosa (three patients), scalloped folds, reduced in number with mosaic pattern mucosa (three patients), and nodular mucosa with reduction in fold numbers (one patient). All seven patients had partial, subtotal, or total VA. One of 143 patients with no endoscopic abnormality had patchy VA. The prevalence of VA was thus 1:19 (8 of 150). Endoscopic markers had a sensitivity of 87.5% (7 of 8), specificity of 100% (142 of 142), positive predictive value of 100% (7 of 7), and negative predictive value of 99% (142 of 143). Of the eight patients with VA, the indications for endoscopy were upper GI symptoms in seven patients (two with anemia) and anemia without GI symptoms in one. After 6 months of dietary gluten exclusion, improvement by at least one criterion was documented in all eight patients. CONCLUSIONS: Careful inspection of the duodenum during routine upper GI endoscopy allows accurate selection of patients for biopsy but may not detect patchy VA or milder enteropathy. Celiac disease should be considered as a cause of dyspeptic and reflux symptoms, as well as of iron-deficiency anemia. (Am J Gastroenterol 1999;94:2182–2186. © 1999 by Am. Coll. of Gastroenterology)

INTRODUCTION Celiac disease may present with nonspecific upper gastrointestinal (GI) symptoms of dyspepsia, abdominal pain, and nausea (1), and also with iron-deficiency anemia in the

absence of GI symptoms (2). Many patients will, therefore, have upper GI endoscopy as an initial investigation, which provides an opportunity to biopsy the second part of the duodenum. Although routine biopsies in all patients undergoing endoscopy would have significant resource implications, endoscopic abnormalities of the distal duodenum associated with villous atrophy (VA) have been described (3), and these may be used to select patients for biopsy. We prospectively sought these markers and obtained duodenal biopsies in a series of patients undergoing routine endoscopy for upper GI symptoms and for anemia. We wished to determine firstly how common VA was in this group of patients, and secondly whether endoscopic markers could be used reliably to select patients for duodenal biopsy.

METHODS We studied consecutive patients undergoing first-time endoscopy for upper GI symptoms or for iron-deficiency anemia by one gastroenterologist in a district general hospital in Northern Ireland. Patients undergoing endoscopy specifically for duodenal biopsy, i.e., with diarrhea, macrocytic anemia, positive family history of celiac disease, or known positive celiac serology (antigliadin or endomysial antibody), were excluded. All examinations were performed with Olympus GIF-XQ230 or XQ200 videoscopes (Olympus KeyMed, Southend-On-Sea, UK). The second part of the duodenum was examined after maximum insufflation for mosaic pattern or nodularity of the mucosa, reduction in number of duodenal folds, and scalloping of duodenal folds. Three biopsies were taken from the second part of the duodenum with standard forceps. They were oriented and mounted on filter paper before submission in formalin for examination by a consultant histopathologist who was unaware of clinical details or endoscopic findings. Histological abnormalities sought included partial, subtotal, and total villous atrophy (PVA, STVA, TVA), crypt hyperplasia, and increased intraepithelial lymphocytes (IELs). We reviewed patients with any of these abnormalities to inquire about symptoms associated with celiac disease, associated conditions, and whether there was a family history. Patient height (meters) and weight (kilograms) were measured for calcu-

AJG – August, 1999

lation of body mass index (weight/height2), and ranges for underweight, normal, and overweight, as defined by Garrow (4), were applied. If not already done, blood was taken for hemoglobin concentration and serum ferritin, as well as for IgA class endomysial antibody (EmA) testing by indirect immunofluorescence. An explanation of the celiac condition was given and referral made to hospital dietitians for advice regarding dietary gluten exclusion. The response to gluten-free diet was assessed at outpatient follow-up after 6 months.

RESULTS Patients One hundred and fifty patients were studied over 16 wk. Their mean age was 57 yr (SD 15, range 12–90). Eightynine (59%) were women. One hundred and twenty-six patients had outpatient endoscopy; these were referred by primary care practitioners for open access endoscopy (OAE) (64 patients), from gastroenterology clinics (51 patients), or from other hospital physicians (11 patients). Twenty-four patients had inpatient endoscopy. Indications were one or more of reflux symptoms (62 patients), ulcer-type symptoms— epigastric pain, dyspepsia (48 patients), iron-deficiency anemia (43 patients), dysphagia (19 patients), nausea and/or vomiting (17 patients), weight loss (eight patients), and hematemesis and/or melaena (three patients). Thirtyone patients had iron-deficiency anemia with no GI symptoms. Endoscopic Findings Apart from gastritis (12 patients, 8%) and hiatus hernia (37 patients, 25%) with no other findings, endoscopic abnormalities proximal to the distal duodenum were present in 46 patients (31%) and included esophagitis (17 patients, 11%), peptic esophageal stricture (eight patients, 5%), Barrett’s esophagus (three patients, 2%), esophageal candida (two patients, 1%), gastric erosions or benign ulcer (four patients, 3%), gastric cancer (two patients, 1%), congestive gastropathy (one patient, 1%), and duodenal erosions or ulcer (nine patients, 6%). Seven patients (5%) had suspicious mucosal abnormalities in the second part of the duodenum (Fig. 1). Six had mosaic pattern mucosa and scalloping of duodenal folds, of whom three also had reduction in fold numbers. The seventh had loss of folds and nodular mucosa. None of these seven patients had any other endoscopic abnormality except hiatus hernia in two (one who presented with reflux, and one with anemia and no GI symptoms). Histology The seven patients with endoscopic markers had PVA (one patient), STVA (three patients), or TVA (three patients), with crypt hyperplasia and increased IELs. Of the 143 patients without endoscopic abnormality of the second part of the duodenum, one had patchy VA, with villous pattern ranging from normal to STVA, although all three biopsy

Endoscopic Markers of Celiac Disease

2183

samples had increased IELs (Fig. 2). This patient had no endoscopic abnormality of any type. None of the remaining 142 patients without endoscopic markers had VA, crypt hyperplasia, or increased IELs. Thus, the prevalence of enteropathy was 1:19 (8 of 150), manifest as VA. Endoscopic markers had sensitivity for VA of 87.5% (7 of 8), positive predictive value of 100% (7 of 7), specificity of 100% (142 of 142), and negative predictive value of 99% (142 of 143). Clinical Details Features of presentation in the eight patients with VA are listed in Table 1. No patient reported diarrhea or a family history of celiac disease. Two had had osteoporosis confirmed by bone densitometry. Body mass index was in the underweight range (⬍20) for one patient, normal (20 –24.9) for three, and in the overweight range (ⱖ25) for four patients. Hemoglobin concentration was ⬍12 g/dl in five patients, although it had been identified as low before endoscopy in only three. Three patients, with TVA or STVA, were EmA negative. During the study period, we confirmed VA in a further seven patients, who underwent duodenal biopsy because of positive EmA, a history of diarrhea, or a family history of celiac disease: thus, 47% (7 of 15) of patients were diagnosed celiac because of duodenal abnormalities seen during endoscopy. Response to Treatment All seven patients with GI symptoms reported improvement at 6-month review, irrespective of symptom type. Weight gain was recorded in five patients (range 2– 6 kg), and EmA had disappeared in four of the five initially EmA-positive patients.

DISCUSSION All seven patients with GI symptoms and VA reported improvement after a gluten-free diet. This suggests that their symptoms were caused by celiac disease, especially as no other endoscopic abnormality was seen except hiatus hernia in one patient. Presentation of celiac disease (i.e., VA resulting from gluten sensitivity) with mild or nonspecific symptoms is well recognized. None of seven patients identified from Swedish blood donor screening had symptoms apart from slight bloating in a study by Grodzinsky et al. (5). Fewer than half of new patients in our own clinical practice report diarrhea (6, 7) and upper GI symptoms are common. A minority of patients is underweight, and many, particularly men, are overweight (8). Over 20% of men and 40% of women had dyspepsia in one series studied by Ciacci et al. (1). Of 36 adult patients surveyed by Usai et al. (9), 76% reported abdominal pain, 50% dysphagia, 50% nausea, 30% vomiting, and 14% noncardiac chest pain. Although some of these symptoms are common in the general population, there is objective evidence of gut dysmotility in celiac disease. Duodenal and jejunal manometry showed an in-

2184

Dickey and Hughes

AJG – Vol. 94, No. 8, 1999

Figure 1. Endoscopic abnormalities of the second part of duodenum: (A) Mosaic pattern mucosa; (B) Nodular mucosa; (C) Scalloping of duodenal fold; (D) Loss of duodenal folds.

creased frequency of migrating motor complexes in adult patients compared with controls, and abnormal discrete clustered contractions, giant jejunal contractions, and bursts of nonpropagated contractions were common (10). Esophageal manometry in 18 adult patients showed motor abnormalities, including nutcracker esophagus, repetitive contractions, and a hypotonic lower esophageal sphincter, in twothirds (9). Iovino et al. (11) reported esophageal symptoms in 45% of 22 adult patients, which after dietary gluten

exclusion fell to 9% with a significant rise in lower esophageal sphincter pressures. Endoscopic forceps biopsy of the second part of the duodenum has largely replaced x-ray guided capsule biopsy and allows demonstration of endoscopic abnormalities associated with VA. Brocchi et al. (12) looked for fold loss (reduction in number or absent folds) in 65 patients undergoing endoscopy specifically for duodenal biopsy. Although their analysis categorized PVA (in most cases resulting from

AJG – August, 1999

Endoscopic Markers of Celiac Disease

2185

Figure 2. Villous pattern ranging from normal (A) to STVA (B) in a patient with no endoscopic markers. All biopsies had increased intraepithelial lymphocytes. Magnification ⫻25 (A); ⫻40 (B).

treated celiac disease) along with normal duodenal histology, fold loss was in fact present in 88% of STVA and 54% of PVA cases, with 97% of 35 patients with normal histology having no fold loss. McIntyre et al. (13) studied 75 patients suspected of having celiac disease and found fold loss to have a sensitivity of 73%, specificity of 97%, and positive predictive value of 85% for TVA, which was present in 15 patients. Maurino et al. (3) studied 100 patients undergoing duodenal biopsy and found that of 36 with STVA or TVA, 27 had loss of folds, 12 had scalloped folds, 14 had mosaic

pattern, and five had visible underlying blood vessels: the presence of any marker had a sensitivity of 94%, specificity of 92%, positive predictive value of 87%, and negative predictive value of 97%. No patient in the study had PVA. However, these figures are based on patient groups with a high prevalence of VA, where the predictive value of any screening test is likely to be high and may not be applicable to less selected populations. Assessment of endoscopic markers in unselected dyspeptic populations is required. Brocchi et al. (12) found four patients with fold loss among 873 undergoing upper GI endoscopy who had STVA, giving

Table 1. Clinical Characteristics of Patients With Histological Confirmation of Villous Atrophy Age/Sex Source

Reason for Endoscopy

Endoscopic Abnormality

55/F OAE

Reflux, abdominal pain

None seen

60/F OAE

Reflux

57/F OAE

Reflux

Mosaic, reduced fold nos., scalloped folds Mosaic, scalloped folds

60/M OAE

Abdominal pain

Mosaic, scalloped folds

63/M OAE

Nausea, vomiting

Reduced fold nos., nodular

58/F GE

Anemia

34/M GE

Reflux, anemia

Mosaic, reduced fold nos., scalloped folds Mosaic, scalloped folds

52/F OAE

Abdominal pain, anemia

Mosaic, reduced fold nos., scalloped folds

Histology, EmA Status

Hb Status, BMI

Patchy VA EmA ⫹ve TVA EMA ⫹ve

Hb 14.3, ferritin 10.1 BMI 25 Hb 10.8, ferritin 10.2 BMI 23 Hb 11.3, ferritin 14.3 BMI 21 Hb 15.3, ferritin 20.6 BMI 30 Hb 14.8, ferritin 65 BMI 26 Hb 9.1, ferritin 6.2 BMI 24 Hb 11.9, ferritin 4.7 BMI 26 Hb 10.3, ferritin 7.9 BMI 19

TVA EmA ⫹ve STVA EmA ⫹ve PVA EmA ⫹ve TVA EmA ⫺ve STVA EmA ⫺ve STVA EmA ⫺ve

Response to Diet Symptoms improved EmA ⫹ve Symptoms improved, weight gain 6 kg EmA ⫺ve Symptoms improved EmA ⫺ve Symptoms improved EmA ⫺ve Symptoms improved, weight gain 3 kg EmA ⫺ve Weight gain 2 kg

Symptoms improved, weight gain 4 kg Symptoms improved, weight gain 5 kg

M ⫽ male; F ⫽ female; GE ⫽ gastroenterology clinic; OAE ⫽ open access endoscopy; BMI ⫽ body mass index; STVA ⫽ subtotal villous atrophy; PVA ⫽ partial villous atrophy; TVA ⫽ total villous atrophy; EmA ⫽ endomysial antibody; Hb ⫽ hemoglobin concentration (g/dl). Ferritin normal range 10 –130 ng/ml (women), 27–399 ng/ml (men).

2186

Dickey and Hughes

a prevalence of 1:218. Of 146 patients undergoing endoscopy for dyspepsia (3), only one had a marker (mosaic mucosa), which was falsely positive. We obtained biopsies if markers were present among 500 OAE patients and found a prevalence of marker positive VA of 1:63 (14). These studies did not obtain biopsies from patients without markers, so they may have underestimated the true prevalence of VA. To our knowledge, the present study is the first to assess the performance of endoscopic markers by obtaining biopsies from all patients having routine endoscopy. The prevalence of VA with endoscopic abnormality (7 of 150, 1:21) was much higher than in our earlier OAE study. This may reflect a higher proportion of patients with anemia, although most patients with VA had normal hemoglobin concentrations or had not had hemoglobin concentration checked before endoscopy. It is also possible that the endoscopist’s ability to identify endoscopic markers has improved. Whereas earlier studies focused on endoscopic markers as predictors of STVA or TVA, two of our patients had PVA. Both patients also had EmA. We routinely treat such patients as celiac as they have other histological indicators of gluten-sensitive enteropathy, most notably IELs as described by Marsh (15), and respond symptomatically to dietary gluten exclusion. Three of the five patients with TVA or STVA were EmA negative and would have been unlikely to have been diagnosed without endoscopic detection. Although the markers had satisfactory sensitivity and specificity for PVA as well as STVA, they are unlikely to be of value for milder degrees of enteropathy (IELs without villous abnormality: Marsh type 1) and were not seen in one patient where the villous abnormality was patchy.

CONCLUSIONS Our study was performed in a population where celiac disease is very common, with a prevalence of 1:150 on population screening (16), so the prevalence of 1:19 among our patients may not be reproducible elsewhere. However, the prevalence of celiac disease in Northern Ireland estimated by screening is less than twice that of other western European countries [1:300 Italian school children (17); 1:250 Swedish blood donors (5)], and a recent study of serum EmA in 2000 American blood donors suggests that the true prevalence of celiac disease in the United States is also of the order of 1:250 (18). It is, therefore, likely that a significant number of patients elsewhere undergoing endoscopy for dyspeptic symptoms, as well as for anemia, will have celiac disease, which is responsible for their symptoms. Endoscopic markers appear to perform sufficiently well in this situation to be able to select patients for duodenal biopsy, although they will not identify patients with non-VA enteropathy and may miss VA if it is patchy. As well as improving symptoms prompting referral, early diagnosis and treatment will prevent the long-term complica-

AJG – Vol. 94, No. 8, 1999

tions of celiac disease such as malignancy (19) and osteoporosis (20). Reprint requests and correspondence: William Dickey, M.D., F.A.C.G., Altnagelvin Hospital, Londonderry BT47 6SB, Northern Ireland, UK. Received Dec. 8, 1998; accepted Mar. 30, 1999.

REFERENCES 1. Ciacci C, Cirillo M, Sollazzo R, et al. Gender and clinical presentation in adult celiac disease. Scand J Gastroenterol 1995;30:1077– 81. 2. Corazza GR, Valentini RA, Andreani ML, et al. Subclinical coeliac disease is a frequent cause of iron-deficiency anaemia. Scand J Gastroenterol 1995;30:153– 6. 3. Maurino E, Capizzano H, Niveloni S, et al. Value of endoscopic markers in celiac disease. Dig Dis Sci 1993;38:2028 – 33. 4. Garrow JS. Treatment of obesity. Lancet 1992;340:409 –13. 5. Grodzinsky E, Franzen L, Hed J, et al. High prevalence of celiac disease in healthy adults revealed by antigliadin antibodies. Ann Allergy 1992;69:66 –70. 6. Dickey W, McConnell JB. How many hospital visits does it take before celiac sprue is diagnosed? J Clin Gastroenterol 1996;23:21–3. 7. Dickey W, McMillan SA, McCrum EE, et al. Association between serum levels of total IgA and IgA class endomysial and antigliadin antibodies: Implications for coeliac disease screening. Eur J Gastroenterol Hepatol 1997;9:559 – 62. 8. Dickey W, Bodkin S. Prospective study of body mass index in patients with coeliac disease. Br Med J 1998;317:1290. 9. Usai P, Bassotti G, Usai Satta P, et al. Oesophageal motility in adult coeliac disease. Neurogastroenterol Mot 1995;7:239 – 44. 10. Bassotti G, Castellucci G, Betti C, et al. Abnormal gastrointestinal motility in patients with celiac sprue. Dig Dis Sci 1994;39:1947–54. 11. Iovino P, Ciacci C, Sabbatini F, et al. Esophageal impairment in adult celiac disease with steatorrhea. Am J Gastroenterol 1998;93:1243–9. 12. Brocchi E, Corrazza GR, Caletti G, et al. Endoscopic demonstration of loss of duodenal folds in the diagnosis of celiac disease. N Engl J Med 1988;319:741– 4. 13. McIntyre AS, Ng DPK, Smith JA, et al. The endoscopic appearance of duodenal folds is predictive of untreated adult celiac disease. Gastrointest Endosc 1992;38:148 –51. 14. Dickey W. Diagnosis of coeliac disease at open-access endoscopy. Scand J Gastroenterol 1998;33:612–5. 15. Marsh MN. Gluten, major histocompatibility complex, and the small intestine. A molecular and immunobiologic approach to the spectrum of gluten sensitivity (‘celiac sprue’). Gastroenterology 1992;102:330 –54. 16. Johnston SD, Watson RGP, McMillan SA, et al. Preliminary results from follow-up of a large-scale population survey of antibodies to gliadin, reticulin and endomysium. Acta Paediatr Suppl 1996;412:61– 4. 17. Catassi C, Ratsch IM, Fabiani E, et al. Coeliac disease in the year 2000: Exploring the iceberg. Lancet 1994;343:200 –3. 18. Not T, Horvath K, Hill ID, et al. Celiac disease risk in the United States: High prevalence of antiendomysium antibodies in healthy blood donors. Scand J Gastroenterol 1998;33:494 – 8. 19. Holmes GKT, Prior P, Lane MR, et al. Malignancy in coeliac disease—Effect of a gluten-free diet. Gut 1989;30:333– 8. 20. Valdimarsson T, Lofman O, Toss G, et al. Reversal of osteopenia with diet in adult coeliac disease. Gut 1996;38:322–7.