GASTROENTEROLOGY 2001;120:636 – 651

Current Approaches to Diagnosis and Treatment of Celiac

Disease: An Evolving Spectrum

ALESSIO FASANO* and CARLO CATASSI‡

*Center for Celiac Research and Division of Pediatric Gastroenterology and Nutrition, University of Maryland, Hospital for Children, Baltimore,
Maryland; and ‡Department of Pediatrics, University of Ancona, Ancona, Italy

Celiac disease (CD) is a syndrome characterized by dam- junctions serves as the main barrier to the passage of
age of the small intestinal mucosa caused by the gliadin macromolecules such as gluten. During this healthy
fraction of wheat gluten and similar alcohol-soluble pro- state, quantitatively small but immunologically signifi-
teins (prolamines) of barley and rye in genetically sus- cant fractions of antigens cross the defense barrier. These
ceptible subjects. The presence of gluten in these sub- antigens are absorbed across the mucosa along 2 func-
jects leads to self-perpetuating mucosal damage,
tional pathways. The vast majority of absorbed proteins
whereas elimination of gluten results in full mucosal
(up to 90%) cross the intestinal barrier through the
recovery. The clinical manifestations of CD are protean
in nature and vary markedly with the age of the patient,
transcellular pathway, followed by lysosomal degrada-
the duration and extent of disease, and the presence of tion, which converts proteins into smaller, nonimmuno-
extraintestinal pathologic conditions. In addition to the genic peptides. The remaining portion of peptides is
classical gastrointestinal form, a variety of other clinical transported as intact proteins, resulting in antigen-spe-
manifestations of the disease have been described, in- cific immune responses. This latter phenomenon uses the
cluding atypical and asymptomatic forms. Therefore, paracellular pathway that involves a subtle but sophisti-
diagnosis of CD is extremely challenging and relies on a cated regulation of intercellular tight junctions that leads
sensitive and specific algorithm that allows the identifi- to antigen tolerance. When the integrity of the tight
cation of different manifestations of the disease. Sero- junction system is compromised, such as in CD,3,4 an
logic tests developed in the last decade provide a non- immune response to environmental antigens (i.e., gluten)
invasive tool to screen both individuals at risk for the may develop. The up-regulation of zonulin, a recently
disease and the general population. However, the cur-
described intestinal peptide involved in tight junction
rent gold standard for the diagnosis of CD remains
regulation,5 seems to be responsible, at least in part, for
histologic confirmation of the intestinal damage in se-
rologically positive individuals. The keystone treatment the increased gut permeability characteristic of the early
of CD patients is a lifelong elimination diet in which food phase of CD.6 This zonulin-dependent increased perme-
products containing gluten are avoided. ability may also be responsible for the increased inci-
dence of autoimmune disorders reported in untreated CD
patients.7
eliac disease (CD) is an autoimmune enteropathy
C triggered by the ingestion of gluten-containing
grains in susceptible individuals. The gliadin fraction of
Another important factor for the intestinal immuno-
logic responsiveness is the major histocompatibility com-
plex (MHC). Human leukocyte antigen (HLA) class I
wheat gluten and similar alcohol-soluble proteins in and class II genes are located in the MHC on chromo-
other grains are the environmental factors responsible for some 6. These genes code for glycoproteins, which bind
the development of the intestinal damage. The disease is peptides, and this HLA–peptide complex is recognized
associated with HLA alleles DQA1*0501/DQB1*0201, by certain T-cell receptors in the intestinal mucosa.8,9
and in the continued presence of gluten the disease is Susceptibility to at least 50 diseases, including CD, has
self-perpetuating.1 The typical intestinal damage charac- been associated with specific HLA class I or class II
terized by loss of absorptive villi and hyperplasia of the
crypts completely resolves upon elimination of gluten-
Abbreviations used in this paper: AEA, antiendomysium antibody;
containing grains from the patient’s diet. AGA, antigliadin antibody; CD, celiac disease; ESPGHAN, European
It is now evident that CD is the result of an inappro- Society of Pediatric Gastroenterology, Hepatology, and Nutrition; GFD,
priate T cell–mediated immune response against in- gluten-free diet; Ig, immunoglobulin; tTG, tissue transglutaminase.
© 2001 by the American Gastroenterological Association
gested gluten.2 Under physiologic circumstances, the 0016-5085/01/$35.00
intestinal epithelium with its intact intercellular tight doi:10.1053/gast.2001.22123
February 2001 DIAGNOSIS AND TREATMENT OF CELIAC DISEASE 637

Table 1. Possible Clinical Manifestations of CD

Typical symptoms Atypical symptoms Associated conditions
Chronic diarrhea Secondary to malabsorption Possibly gluten dependent
Failure to thrive Sideropenic anemia IDDM
Abdominal distention Short stature Autoimmune thyroiditis
Osteopenia Autoimmune hepatitis
Recurrent abortions Sjögren syndrome
Hepatic steatosis Addison disease
Recurrent abdominal pain Autoimmune atrophic gastritis
Gaseousness Autoimmune emocytopenic diseases
Independent of malabsorption Gluten independent
Dermatitis herpetiformis Down syndrome
Dental enamel hypoplasia Turner syndrome
Ataxia Williams syndrome
Alopecia Congenital heart defects
Primary biliary cirrhosis IgA deficiency
Isolated hypertransaminasemia
Recurrent aphthous stomatitis
Myasthenia gravis
Recurrent pericarditis
Psoriasis
Polyneuropathy
Epilepsy (with or without intracranial calcifications)
Vasculitis
Dilatative cardiomyopathy
Hypo/hyperthyroidism

alleles. The primary HLA association in CD is to the first months of life. Symptoms begin within weeks to
HLA-DQA1*0501, DQB1*0201 genes encoding a few months after the introduction of weaning foods
DQ2 molecules.1 Non-HLA genes together appear to containing prolamines, and soon there is a progressive
contribute more to genetic susceptibility than do the decrease in weight gain with a decline in the child’s
HLA genes, but the contribution from each single, percentile for weight and weight for height. On ex-
predisposing non-HLA gene appears to be modest.10 amination, the children are often pale and noticeably
Dieterich et al.11 recently demonstrated that one of the thin with a protuberant abdomen, decreased subcuta-
targets of the autoimmune response in CD is the tissue neous fat, and reduction in muscle mass. The stools are
transglutaminase (tTG).11 The deamidating activity of characteristically pale, loose, bulky, and highly offen-
this enzyme seems to generate gliadin peptides that sive because of fat malabsorption. In the very young
bind to DQ2 to be recognized by disease-specific infant with early onset of symptoms there may be
intestinal T cells.10 frank watery diarrhea with dehydration and electrolyte
imbalance. A small number of these infants also have
Clinical Presentations severe hypoproteinemia and edema and may present in
The clinical manifestations of CD vary markedly a shocklike state that has been termed “celiac crisis.”
with the age of the patient, the duration and extent of Laboratory signs of the malabsorption include iron
disease, and the presence of extraintestinal pathology deficiency anemia, hypoalbuminemia, hypocalcemia,
(Table 1). Depending on the features at the time of and vitamin deficiencies. The pathologic changes are
presentation, together with the histologic and immuno- most marked in the duodenum and upper jejunum, but
logic abnormalities at the time of diagnosis, CD can be the extent of mucosal damage is highly variable, and in
subdivided into the following clinical forms. some cases the entire small intestine may be involved.
The histologic changes in CD range from minor villous
Classical (Typical) Form blunting to subtotal or total villous atrophy (see below,
The onset of symptoms in the classical form Figure 5), decreased villous height-to-crypt depth ratio,
generally occurs between 6 and 18 months of age. This crypt hyperplasia with increased mitosis, significantly
form is typically characterized by chronic diarrhea, increased plasma cell and lymphocyte infiltration in the
failure to thrive, anorexia, abdominal distention, and lamina propria, and a pronounced increase in the number
muscle wasting. Growth is usually normal during the of intraepithelial lymphocytes.
638 FASANO AND CATASSI GASTROENTEROLOGY Vol. 120, No. 3

Atypical Forms chemical abnormalities associated with hepatic damage

In recent years there has been a noticeable change has been reported in a high percentage of pediatric
in the age of onset of symptoms and the clinical presen- patients with CD who adhered to a strict GFD.34
Osteoporosis. Patients with CD are at high risk
tation of CD. Mäki et al.12 first reported an up-shift of
for developing low bone mineral density and bone turn-
age at diagnosis in Finland to 5– 6 years, with fewer than
over impairment. Persistent villous atrophy is associated
50% of new cases presenting with typical gastrointesti-
with low bone mineral density. In adult patients respon-
nal symptoms. Reports from Scotland,13 England,14 Can-
sive to diet, the bone density seems comparable to that of
ada,15 and the United States16 have also shown that
healthy individuals.35 Children who followed a GFD for
almost 50% of patients with newly diagnosed CD do not
at least 5 years had normal bone mineralization and bone
present with gastrointestinal symptoms.
turnover.36 Of 86 consecutive patients with newly diag-
Dermatitis herpetiformis. Dermatitis herpetifor-
nosed, biopsy-confirmed CD, 40% had osteopenia and
mis is currently regarded as a variant of CD (“skin CD”).
26% osteoporosis.36 No differences between male and
It is a blistering skin disease characterized by pathogno-
female patients or between fertile and postmenopausal
monic granular immunoglobulin (Ig) A deposits in un-
women were observed. Even in postmenopausal women,
involved skin.17 The most typical sites of the rash are the
GFD led to significant improvement in bone mineral
elbows, knees, and buttocks. Intestinal symptoms are not
density.37 In these cases, supplement treatment with
common, but a varying degree of enteropathy, ranging
vitamin D and Ca2⫹ is indicated.
from the infiltrative-type lesion to flat mucosa, can be Neurologic problems. Gluten sensitivity is com-
found on small intestinal biopsy in almost 100% of cases. mon in patients with neurological diseases of unknown
Both the enteropathy and the rash slowly clear with a cause and may have etiologic significance.38 Pellecchia et
gluten-free diet (GFD) and relapse when patients return al.39 recently reported that 3 of 24 patients with idio-
to a regular diet.18 pathic cerebellar ataxia had CD.
Iron-deficiency anemia. Iron deficiency with or Other extragastrointestinal symptoms. A delay
without anemia, typically refractory to oral iron supple- in onset of puberty secondary to CD has been described
mentation, can be the only presenting sign of CD.19 in a number of adolescent patients.12,22,40,41 Recurrent
Short stature. Short stature is well described as abortions42,43 and reduced fertility40,42 caused by CD
the only symptom of CD in some older children and have also been reported in this age group. Recently,
adolescents, and it is believed that as many as 9%–10% Ciacci et al.44 have reported that the relative risk of
of those with “idiopathic” short stature have CD.20 –22 In abortion in women affected by CD is 8.9 times higher
these patients, both the bone age and growth velocity are than in healthy subjects, and a GFD reduced the relative
significantly impaired.20,22,23 Some patients have also risk of abortion.44
demonstrated impaired growth hormone production af-
ter provocative stimulation testing.23 This value returns Asymptomatic (Silent) Form
to normal after introduction of a GFD.24 This form is characterized by the presence of
Dental enamel hypoplasia. Dental enamel hypo- histologic changes, probably limited to the proximal
plasia has been found in up to 30% of untreated patients intestine, that occur in individuals who are apparently
with CD.25,26 asymptomatic.45– 47 Most cases in this category have been
Arthritis and arthralgia. CD has been described identified through screening programs involving appar-
in 1.5%–7.5% of patients with rheumatoid arthri- ently healthy subjects. However, a more careful clinical
tis.27–29 These symptoms were reported by Mäki et al.30 anamnesis typically reveals that many of these “silent”
as the only presentation of CD in 7 adolescent patients. cases are indeed affected by low-intensity illness often
In each case, the symptoms resolved on introduction of a associated with decreased psychophysical well-being.
GFD and all other anti-inflammatory medications could Common findings include (1) iron deficiency with or
be discontinued. without anemia; (2) behavioral disturbances, such as
Chronic hepatitis and hypertransaminasemia. tendency to depression, irritability, or impaired school
Idiopathic chronic hepatitis as the initial presentation of performance in children; (3) impaired physical fitness,
CD has been reported occasionally.31,32 Vajro et al.33 “feeling always tired,” and easy fatigue during exercise;
describe 3 children with cryptogenetic chronic hepatitis and (4) reduced bone mineral density.48,49 A 24-month
secondary to CD. In all cases, GFD induced complete follow-up study showed that adolescents with screening-
remission with normalization of the biochemical and detected CD who were apparently symptomless at diag-
histologic changes of hepatitis. Resolution of the bio- nosis often reported improved physical and psychologic
February 2001 DIAGNOSIS AND TREATMENT OF CELIAC DISEASE 639

conditions once they began following a GFD.50 The most diabetes mellitus, whose clinical manifestations appear
common changes included increased weight and height after the patient has produced an autoimmune response
velocity, increased appetite, mood amelioration, and im- to various autoantigens (i.e., anti-insulin, anti–␤ cell),
proved physical and school performance.50 Finally, cur- and might also be present in CD. This could explain the
rent evidence suggests that subjects with “silent” CD are high incidence of autoimmune diseases (Table 1) and the
at risk to develop the same long-term complications presence of a large number of organ-specific autoantibod-
experienced by individuals with typical symptoms. ies in a certain number of celiac subjects on a gluten-
containing diet.
Based on this evidence, it is tempting to hypothesize
Associated Diseases
that the range of gluten-dependent autoimmune disor-
A number of medical conditions are significantly ders present in genetically predisposed individuals goes
associated with CD (Table 1). For some of these condi- well beyond the classic enteropathy of CD (Table 1).
tions, sensitivity to gliadin has been conclusively proven Furthermore, recent data suggest that the prevalence of
or may be implicated (Table 1). autoimmune diseases among patients with CD is propor-
tional to the time of exposure to gluten.7
Complications Associated With
Unrecognized CD
The Epidemiology of CD
Malignancies. The persistence of mucosal injury
with or without typical symptoms can lead to serious Epidemiology of CD in Europe
complications, and gastrointestinal malignancies (partic- In the past 3 decades, a substantial number of
ularly lymphoma) have been reported in 10%–15% of epidemiologic studies have been conducted in Europe to
adult patients with known CD who do not strictly establish the frequency of CD, and interesting contro-
comply with a GFD.51 However, the increased risk for versies have arisen. Earlier investigations measured the
malignancy in the gastrointestinal tract in patients with incidence of CD, namely the number of “new” diagnoses
CD has been questioned recently; therefore, the precise in the study population during a certain period. One of
magnitude of this complication remains uncertain (see the oldest epidemiologic studies on CD conducted in
diagnosis section below). Nevertheless, it has been re- 1950 established that the cumulative incidence of the
ported that the mortality rate in CD patients is almost disease in England and Wales was 1/8000, whereas an
double (1.9⫻) the rate calculated for the general popu- incidence of 1/4000 was detected in Scotland.53 The
lation, mainly because of the occurrence of neoplasms.52 diagnosis at that time was entirely based on the detection
Data from Logan et al.52 have shown that when appro- of typical symptoms and confirmed by complicated and
priate treatment for CD was instituted in childhood and sometimes nonspecific tests. The awareness of the disease
strictly followed, the mortality rate of these subjects was greatly increased in the 1960s when more specific tests
no different from that expected in the general popula- for malabsorption and the pediatric peroral biopsy tech-
tion, and no deaths from intestinal lymphoma were nique became available.54 Consequently, an elevated in-
recorded. cidence of the disease (which in the middle 1970s
Autoimmune diseases. CD seems to meet the reached peaks of 1/450 –500) was reported in studies
criteria of a true autoimmune disease for which the from Ireland,55 Scotland,56 and Switzerland.57 This in-
genetic predisposition (HLA), exogenous trigger (glu- creased incidence of CD prompted changes in the dietary
ten), and autoantigen (tTG) are known. It seems that habit, based on the hypothesis that delayed exposure to
tTG is only one of the autoantigens involved in gluten- gluten could prevent the onset of the disease. For the first
dependent autoimmune reactions. Other autoantigens time in 25 years, a decrease in the incidence of CD was
that are normally “cryptic” can be unmasked and cause a reported in the United Kingdom and Ireland58 – 60 after a
self-aggressive immunologic response following the gli- late introduction of gluten in infant diet. Unfortunately,
adin-initiated inflammatory process. In fact, persistent this decrease was deceptive because subsequent screening
stimulation by some proinflammatory cytokines such as studies showed that the reduction in typical cases in
interferon ␥ and tumor necrosis factor ␣ can cause further infants was counterbalanced by the increase of atypical
processing of autoantigens and their presentation to T forms of CD with the onset of the symptoms occurring in
lymphocytes by macrophage-type immunocompetent older children or in adults.61 Because of the development
cells (the so-called antigen-presenting cells). The phe- of sensitive serologic tests, it has recently become possi-
nomenon of antigen spreading has been described in ble to evaluate the prevalence of CD (number of affected
well-defined natural models such as insulin-dependent persons, including subclinical cases, in a defined popu-
640 FASANO AND CATASSI GASTROENTEROLOGY Vol. 120, No. 3

Table 2. Prevalence of CD Based on Clinical Diagnosis or Epidemiology of CD in the United States

Screening Data
In the American scientific community it is gen-
Prevalence on Prevalence on
Geographic area clinical diagnosisa screening data erally believed that CD is a rare disorder in the United
States, which is reflected by the limited number of
Brazil ? 1:400
Denmark 1:10,000 1:500 scientific papers published from the new continent in the
Finland 1:1000 1:130 30-year period from 1965 to 1995.73 Only 2 epidemio-
Germany 1:2300 1:500 logic studies of CD were published during this period,
Italy 1:1000 1:184
Netherlands 1:4500 1:198 both between 1993 and 1994. The first study was con-
Norway 1:675 1:250 ducted by Rossi et al.74 in 1993 on a pediatric population
Sahara ? 1:70 from the western New York area with symptoms possi-
Slovenia ? 1:550
Sweden 1:330 1:190 bly related to CD, such as chronic diarrhea, failure to
United Kingdom 1:300 1:112 thrive, short stature, and diabetes.74 Although the prev-
United States 1:10,000 1:111 alence of CD among patients with symptoms possibly
Worldwide (average) 1:3345 1:266
associated to the disease was lower than reported in
aClassical gastrointestinal symptoms.
Europe, the concurrence of CD and insulin-dependent
Data from references 81– 85, 125–131.
diabetes mellitus was comparable to that previously re-
ported from the old continent. These data suggest that
other atypical presentations of CD and eventually late
lation at a certain point). Screening studies show a high onset of the disease after an asymptomatic phase during
prevalence of CD among both healthy children62– 64 and childhood may account for the low occurrence of CD
adults.65 The prevalence of CD throughout the old con- reported in this study. The second American epidemio-
tinent seems to be more homogeneous than previously logic study published in 1994 was based on a retrospec-
thought (Table 2). Furthermore, these screenings showed tive evaluation (1960 –1990) of the incidence of CD
that CD is one of the most frequent genetically based among the population of Olmsted County, Minnesota,
diseases,62,66 occurring in 1 of 130 –300 in the European using the medical record of the Rochester Epidemiolog-
population67,68 (Table 2). In a serologic screening study ical Project.75 Case definition was limited to those indi-
involving more than 17,000 Italian schoolchildren, the viduals presenting typical gastrointestinal symptoms
prevalence of CD was 1 in 184,48 and the ratio of known (i.e., chronic diarrhea and weight loss) or dermatitis
to undiagnosed CD cases was 1 to 7. The European herpetiformis whose intestinal biopsies showed flat mu-
experience taught that, despite common genetic and cosa.75 Using these restrictive parameters, the authors
environmental factors, the clinical presentation of CD in identified only 3 cases among the pediatric population
neighboring countries may greatly diverge. A typical (calculated incidence rate, 0.4 per 100,000 person-years),
example of this phenomenon is the Danish epidemiologic whereas the overall age- and gender-adjusted incidence
case. Until a few years ago, CD was regarded as rare in was 1.2 per 100,000 person-years. Based on these results,
Denmark, with an estimated incidence based on clinical the authors concluded that CD is relatively rare in the
evidence (i.e., presence of classical symptoms) of United States (prevalence ⬃1:10,000). Unfortunately,
1/10,00069 (Table 2). At the same time, the incidence of both studies failed to consider the protean clinical man-
the disease in neighboring countries (including Sweden ifestations of CD. By focusing on specific symptoms, the
and Finland) that share similar genetic backgrounds authors may have missed what is currently defined as the
increased after a decrease in breast feeding practice and submerged part of the so-called celiac iceberg (Figure 1).
increased consumption of gluten during infancy.70,71 Recently, a series of epidemiologic studies conducted
Subsequent serologic screening studies suggested that using more appropriate experimental designs and pow-
CD is as frequent in Denmark as in Sweden, with a erful screening tools showed that CD is as frequent in the
reported prevalence of 1/50072 (Table 2). These results United States as in Europe in both risk groups76 –78 and
suggest that in Denmark most cases of CD were previ- the general population79,80 (Table 2). Our center for
ously undiagnosed, presumably because of lack of typical celiac research is currently conducting a large, multi-
gastrointestinal symptoms. Factors such as type of cow’s center study on the prevalence of CD in both risk groups
milk formulas, breast feeding, age at gluten introduc- (i.e., subjects with either symptoms or complications
tion, quantity of gluten and quality of cereals, and quan- associated with CD, first- and second-degree relatives of
tity of wheat gluten may all influence the clinical pre- patients with biopsy-proven CD, etc.) and the general
sentation of the disease.71 population. The results generated on a large number of
February 2001 DIAGNOSIS AND TREATMENT OF CELIAC DISEASE 641

The Iceberg Model

The epidemiological changes of CD are efficiently
conceptualized by the iceberg model, originally intro-
duced by Richard Logan in 1991.86 The prevalence of
CD can be conceived as the overall size of the iceberg,
which is primarily influenced by the frequency of the
predisposing genotypes in the population. Indeed, CD
seems to be more common wherever the frequency of the
HLA-DR3 (and DQ2) is high, such as in Europe, the
Figure 1. The CD iceberg model. United States, and North Africa. The dimension of this
iceberg also depends, to a lesser extent, on disease defi-
individuals screened so far suggest that the prevalence of nition, i.e., whether subjects with so-called latent or
CD in the United States is similar to that reported in potential CD47 or those with gluten sensitivity and mild
Europe if not even higher, both among risk groups and enteropathy87 are “counted” as affected individuals. In
in the general population81 (Table 2). countries where a substantial part of the population is of
Epidemiology of CD in the Rest of European origin, the prevalence of CD is likely to be
the World more stable than previously thought, roughly in the
Because CD is the result of the interaction be- range of 0.5%–1% of the general population. A sizable
tween genetic (both HLA and non–HLA-associated number of these cases are properly diagnosed because of
genes) and environmental factors (gluten-containing suggestive complaints (e.g., chronic diarrhea, unex-
grains), it would be reasonable to evaluate the world plained iron deficiency) or other reasons (e.g., family
distribution of these 2 components to identify areas “at history of CD). These cases make up the visible part of
risk” for CD. The coincidence of the CD HLA aplotypes the celiac iceberg, in quantitative terms expressed by the
(Figure 2A) and the level of wheat consumption (Figure incidence of the disease. However, as previously reported,
2B) clearly confirm Europe as a region at risk for CD. screening studies show that in Western countries, for
However, the coexistence of the 2 key components in- each diagnosed case of CD, an average of 5–10 cases
volved in CD pathogenesis (Figure 2A and B) is also remain undiagnosed (the submerged part of the iceberg).
notable in regions where CD has been historically con- The “water line,” namely the ratio of diagnosed to un-
sidered rare. This apparent paradox can be explained by diagnosed cases, depends on several factors: (1) awareness
the limited number of scientific studies performed in of CD: “think of CD and you will find it” is an aphorism
some of those countries in which CD is perceived as a rare worth remembering88; differing awareness, and conse-
disorder (Figure 2C). Recent epidemiologic studies con- quently variable thresholds for serologic CD testing, is
ducted in areas at risk (Figure 2A and B), such as South likely to explain a substantial part of the wide differences
America,82 North Africa,83 and Asia,84,85 suggest that in incidence between countries; (2) availability of diag-
CD was indeed underdiagnosed. nostic facilities: lack of both laboratory equipment and
Combined together, these studies suggest that CD is personnel trained in CD diagnosis is a major problem in
still underestimated in areas where large epidemiologic large areas of the world, e.g., North Africa, the Middle
studies are lacking. The European experience taught us East, and India, where the frequency of CD is currently
that despite common genetic and environmental factors, underestimated; (3) variations in clinical intensity: at
the clinical presentation of CD in neighboring countries both individual and population levels, the higher the
may greatly diverge and could explain the different amount of ingested gluten, the higher the intensity of
disease prevalence previously reported. A comparison the clinical picture, thereby increasing the chances that
between the estimated prevalence (based on the occur- CD can be diagnosed on clinical grounds. This has been
rence of typical symptoms) and the serologic screening clearly shown by the “epidemic” of CD observed in
data (where available) shows that CD is a common dis- Sweden during the 1980s and early 1990s, in relation-
ease but its gastrointestinal presentation is relatively ship with the gluten load that infants received with
rare, particularly in countries in which CD was consid- follow-up formulas.89 Because of the variable relevance of
ered a negligible pathology (Table 2). Worldwide, CD these factors, the water line is much more unstable than
“out of the intestine” is 15 times more frequent than CD the overall size of the iceberg, thereby explaining the
“in the intestine” (Table 2), making the diagnosis ex- reported wide fluctuations in space and time of CD
tremely challenging. incidence. What remains to be evaluated is the effect of
Figure 2. (A ) CD-associated
HLA-DR3. Percentage of genic
frequency of HLA-DR3 in the
world. (Data provided by Dr.
Francesco Cucca, Department
of Pediatrics, University of Ca-
gliari, Cagliari, Italy.) (B) World
distribution of grain consump-
tion. The intensity of color is
directly related to the amount
of wheat products consumed
(expressed as percent of daily
energy supply). Numbers in pa-
rentheses represent the num-
ber of countries that have the
wheat consumption shown on
the left. (Data from The Sixth
World Food Survey; Rome, It-
aly: Food and Agriculture Orga-
nization of the United Nations,
1996.) (C ) Scientific produc-
tion on CD worldwide during the
period from 1966 to the
present. The intensity of color
is directly related to the num-
ber of articles found in a MED-
LINE search for CD and the
name of the country. Numbers
in parentheses represent the
number of countries that have
published within the range of
manuscripts shown on the left.
February 2001 DIAGNOSIS AND TREATMENT OF CELIAC DISEASE 643

other factors (e.g., intestinal infections and nutrient in- GFD, the clinical symptoms had to resolve, results of
takes) on the clinical presentation and, even more in- screening tests had to return to within normal limits,
triguingly, whether environmental variables can influ- and a second intestinal biopsy showing complete healing
ence the prevalence of CD, therefore assessing the of the histologic damage was recommended (phase 2)
fascinating possibility of primary prevention of this dis- (Figure 3). Phase 3 was then started by a gluten chal-
order. lenge with subsequent return of symptoms, pathologic
screening test results, and intestinal damage (Figure 3).
How to Diagnose CD? The diagnosis was confirmed only if all the criteria listed
The diagnosis of CD is based on 3 key parameters: in the 3 phases were completely satisfied.
(1) case identification, (2) screening tests, and (3) defin- The Present
itive tests. These parameters have substantially changed
Development of serologic tests. In the past
during the past 50 years, thanks to better understanding
of the clinical presentation of the disease and the advent 10 –15 years we have learned that the clinical expression
of more sensitive and specific diagnostic tools and con- of CD is more heterogeneous than previously thought.86
firmative tests. Beside the classical gastrointestinal form, a series of other
clinical manifestations of the disease have been described
The Past thanks to the advent of innovative serologic screening
Until a few decades ago, there was the general tests, such as assays for antigliadin antibody (AGA) and
perception that the clinical presentation of the disease antiendomysium antibody (AEA). The combined use of
was quite uniform. Case identification was based entirely serum AGA IgG (good sensitivity) and IgA (good spec-
on the search for symptoms such as chronic diarrhea, ificity) resulted in a reliable screening test for diagnosis
abdominal distention, and weight loss (or poor weight of CD.91 Based on the use of this new tool, we have
gain) occurring in young children a few months after the learned that the clinical presentation of CD is more
introduction of solid food to their diet. To confirm protean than previously thought, including previously
clinically suspected CD, unspecific screening tests aimed unrecognized atypical and asymptomatic forms (see
at establishing the digestive/absorptive functions of the above). Moreover, these studies show that CD is not
proximal small intestine (i.e., glucose tolerance test, limited to the pediatric population; the onset of disease
D-xylose test, fecal fat) were used. Given the lifelong
may occur during adulthood, after years of silent disease.
nature of the disease, in 1970 the European Society of Because it has been demonstrated recently that tTG is
Pediatric Gastroenterology, Hepatology, and Nutrition the target of a specific autoimmune response (see be-
(ESPGHAN) dictated specific guidelines by identifying low),11 this enzyme has also been used to develop inno-
3 CD diagnostic phases (Figure 3).90 To meet the criteria vative diagnostic tools. The routine use of the AEA assay
of the first phase, the presence of gastrointestinal symp- is limited by elevated costs, time-consuming protocols
toms compatible with CD, positive results of pathologic unsuitable for testing large numbers of samples, poor
screening tests, and confirmation of the diagnosis by sensitivity in young children (⬍2 years of age) and in
intestinal biopsy showing histologic evidence of flat mu- IgA-deficient individuals (the AEA assays routinely per-
cosa were required (Figure 3). Upon establishment on a formed are of the IgA class), and use of the esophagus of
an endangered species (such as the monkey) as the sub-
strate for the immunofluorescent analysis. Even if this
last issue has been resolved by using the human umbil-
ical cord as a valid alternative to the monkey esopha-
gus,80 it has been reported that the subjective interpre-
tation of the AEA assay may lead to unacceptable
variability among laboratories that perform this test.92
Therefore, major effort has been concentrated on devel-
oping a tTG-based ELISA, using either the commercially
available guinea pig tTG93,94 or human recombinant
tTG.95,96
The currently available serologic tests for the diagnosis
of CD remain within the province of the specialized
diagnostic laboratory. Given the projected high preva-
Figure 3. CD diagnostic protocol proposed by the ESPGHAN in 1970. lence of the disease and its protean nature, a simple
644 FASANO AND CATASSI GASTROENTEROLOGY Vol. 120, No. 3

diagnostic test that could be used at the general practi-

tioner’s site would represent a great advance. The recent
report of a human tTG dot blot test based on the
detection of anti-tTG antibodies in serum or in 1 drop of
whole blood97 opened new horizons for the diagnosis of
CD. These preliminary results show that the test seems
to be extremely sensitive (100%) and reasonably specific
(96%). If these data are confirmed, this test holds great
potential because it is quick (30 minutes) and inexpen-
sive, requires minimal handling, and in view of its high
sensitivity and specificity could easily be introduced into Figure 4. Revised criteria for the diagnosis of CD proposed by the
the general practitioner’s armory for ambulatory screen- ESPGHAN.
ing of CD.
Current guidelines for serologic diagnosis and
follow-up of CD. Because the guinea pig– based tTG criteria, if the symptoms (either typical or atypical) and
ELISA has been only recently commercialized and the screening results are suggestive, a single intestinal biopsy
human-based tTG ELISA is still experimental, serologic followed by a favorable response to the GFD is sufficient
diagnosis of CD still relies on the combined use of AGA to definitely confirm the diagnosis (Figure 4). However,
and AEA assays. Interpretation of these assays should total villous atrophy, once considered the only histologic
take into account the fact that AEA can have false- finding compatible with a diagnosis of CD, is now
negative results in both IgA-deficient subjects and chil- considered only the extreme of a continuous spectrum of
dren younger than 2 years of age, whereas AGA (partic- tissue damage that can be detected during the acute
ularly the IgG subclass) can yield false-positive results in phase of the disease (Figure 5). Furthermore, the possible
gastrointestinal conditions other than CD, including patchy characteristics of intestinal damage99 and the
cow’s milk protein intolerance and parasite infections. importance of correct orientation of the biopsy for ap-
Once a definitive diagnosis is established (see below), use propriate evaluation of the intestinal damage both add
of these serologic tests is recommended to verify com- further challenge to a conclusive histologic diagnosis
pliance with the GFD, which should be evaluated on a of CD.
yearly basis or every time patients experience symptoms
Who Should Be Tested?
possibly related to gluten exposure. If the preliminary
data so far reported on the sensitivity and specificity of At-risk groups. Serologic testing is indicated for
the tTG ELISA (Table 3) are confirmed on a large scale, subjects with symptoms suggestive of CD, as well as for
it is likely that this test will make the AGA and possibly those with CD-associated diseases (Table 1). However,
the AEA assays obsolete. small intestinal biopsies should always be performed if
Algorithm for the definitive diagnosis of CD. the clinical suspicion is strong, regardless of the serology
Given the high sensitivity and specificity reported for results. Some at-risk groups showing a particularly high
some of the screening tools currently available (Table 3), prevalence of associated CD (Figure 6) deserve a special
the ESPGHAN has recently proposed a revised CD di- mention: (1) first- and second-degree relatives of patients
agnostic protocol98 (Figure 4). Based on these revised with CD: younger siblings can be checked at age 2 years
or earlier if CD is clinically suspected; (2) patients and
Table 3. Sensitivity, Specificity, and Positive and Negative relatives of patients with type I diabetes100 and patients
Predictive Values of Serologic Screening Tests with immune thyroid or liver disorders; (3) patients with
Reported in the Literature for the Diagnosis of CD Sjögren syndrome and other connective tissue diseases:
Test Sensitivity Specificity PPV NPD in a recent Finnish series, 5 (15%) of 34 patients with
AGA IgG 57–100 42–98 20–95 41–88 Sjögren syndrome were found to have CD,101 although
AGA IgA 53–100 65–100 28–100 65–100 ongoing inflammation was often present in the small
AEA IgAa 75–98 96–100 98–100 80–95 intestinal mucosa of patients without CD101; (4) subjects
Guinea pig tTGb 90.2 95
Human tTGb 98.5 98
with either Down or Turner syndrome; and (5) subjects
with selective IgA deficiency, who show a 10-fold in-
PPV, positive predictive value; NPD, negative predictive value. creased risk of associated CD.102 In these cases, the
aPatientsolder than 2 years.
bIgG ⫹ IgA antibodies. screening test should be an IgG class antibody, e.g.,
Data from references 132–138. AGA IgG or anti-tTG IgG.86
February 2001 DIAGNOSIS AND TREATMENT OF CELIAC DISEASE 645

Figure 5. Histologic grades of intestinal mucosa damage in patients with CD (courtesy of Dr. Karoly Horvath).

A single negative result of the serologic markers can- penic anemia secondary to the combination of periodic
not always rule out the possibility of CD on a lifelong blood drawings and the malabsorption condition typical
basis. This has been elegantly shown by a recent fol- of the disease.
low-up study of 275 patients with type I diabetes, in Case finding or mass screening? How to deal
which only 2 of 9 patients found to have CD during a with the submerged part of the celiac iceberg is currently
6-year period had an AEA-positive test result at the time a matter of debate in the scientific community. An
of diabetes onset.103 increasing number of experts is in favor of early, mass
Finally, we suggest that serologic testing for CD screening of CD because this condition apparently fulfills
should be performed routinely in people joining blood the requirements for a worthwhile screening program:
donor groups. Because the celiac enteropathy often im- (1) it is a common disorder causing significant morbidity
pairs iron absorption, CD should be identified as soon as in the general population; (2) early detection is often
possible in these subjects to avoid the onset of a sidero-
difficult on a clinical basis; (3) if not recognized, the
disease can manifest itself with severe complications that
are difficult to manage (e.g., infertility, osteoporosis,
lymphoma); (4) there is an effective treatment, the GFD;
and (5) sensitive and simple screening tests are available,
e.g., the anti-tTG test.
However, several issues need further clarification to
correctly establish the cost/benefit ratio for CD screen-
ing. Although it is well established that patients with
untreated CD may develop complications, the natural
history of undiagnosed CD is currently unclear. Avail-
able studies have necessarily been limited to patients
Figure 6. Prevalence of CD in some at-risk groups. Dots represent the
prevalence found in different studies, and lines show the mean val-
with clinically diagnosed CD (i.e., the tip of the iceberg),
ues. eventually leading to biased estimate of the risks. For
646 FASANO AND CATASSI GASTROENTEROLOGY Vol. 120, No. 3

example, the relative risk of developing a lymphoma completely absent, diagnosis of CD becomes more diffi-
complication was reported to be as high as 30 –100,53,104 cult and the diet treatment is significantly delayed. In
whereas an ongoing case-control multicenter Italian these subjects, exposure to gluten will continue for a
study investigating the prevalence of CD in patients with prolonged period, with a subsequent increase in the risk
lymphoma seems to indicate only a slight increase in the of complications.
risk of this malignancy (odds ratio ⬃2) in comparison
with the general population.105 Despite the high sensi-
tivity of the serologic CD markers, the positive predic- The Treatment
tive value of these investigations decreases when they are Total lifelong avoidance of gluten ingestion re-
used in the general population rather than in at-risk mains the cornerstone treatment for the disease. The diet
groups.106 The appropriate age to screen for CD also requires ongoing education of patients and their families
remains to be established, as well as whether periodic by both doctors and dieticians. Regional CD support
repetition of the screening would be required to rule out groups are instrumental sources of information and sup-
a “late onset” gluten sensitization.107 Because of the port. One of the major controversies in the treatment of
ethical implications of mass screening, the difficulties of CD relates to the amount of gluten allowed in the diet of
treating patients with apparently silent CD should not CD patients. The National Food Authority has recently
be overlooked. A recent 5-year follow-up study revealed redefined the term “gluten-free.” Previously, ⬍0.02%
a 30% decrease in adherence to the GFD in patients with gluten was considered gluten-free, but gluten-free now
screening-detected CD compared with age-matched, means no gluten, and ⬍0.02% is currently labeled “low
hospital-detected CD cases.108 Wherever products con- gluten.” However, the stringency of gluten restriction
taining wheat flour represent the staple food, treatment (zero tolerance versus low gluten ingestion) is an issue
with a GFD is likely to interfere with quality of life, that is far from being resolved because opinions differ
especially in adults, and it has been shown that adults among scientists and CD support groups worldwide.
with CD undergoing long-term treatment fail to attain These controversies are attributable to a lack of solid
the same degree of subjective health as the general pop- scientific evidence for a threshold of gluten consumption
ulation.109 below which no harm occurs. The gliadin fraction of
Finally, mass screening for CD will depend on the wheat gluten and similar alcohol-soluble proteins (pro-
results of comprehensive, well-performed cost-effective-
lamins) in other grains are the environmental factors
ness analyses. Currently, the “best buy” approach to the
responsible for the development of intestinal damage.
submerged portion of the iceberg of undiagnosed CD
Prolamins are found in a variety of widely used grains
seems to be a systemic process of case finding, as sug-
(Table 4). Therefore, products labeled “wheat-free” are
gested by a recent study developed in a primary care
not necessarily gluten-free. They may contain gluten as
setting in central England.110 By simply investigating
well as other grains that are not allowed. Wheat, rye, and
at-risk subjects, e.g., those with anemia, fatigue, thyroid
barley are the predominant grains containing toxic pep-
disease, diabetes, or a family history of CD, Hin et al.
observed a 4-fold increase in the number of CD diagnoses tides. Both in vivo challenges and in vitro immunologic
during a 1-year period.110 Increased awareness of the studies support the possibility that oats (once considered
extraintestinal manifestations of CD, coupled with a low toxic for CD patients) can be ingested safely.111 How-
threshold for serologic testing, uncovers a large portion ever, because of uncontrolled harvesting and milling
of the submerged iceberg.110 procedures, cross-contamination of oats with gluten is a
concern. Triticale (a combination of wheat and rye),
Why Early Diagnosis Is Important kamut, and spelt112 (sometimes called farro) are also
Our better understanding on the pathogenesis of toxic. Other forms of wheat are semolina (durum wheat),
CD2 and the observation that CD patients’ risk of devel- farina, einkorn, bulgur, couscous, and any form that
oping autoimmune diseases11 and intestinal lympho- includes wheat in the name, such as wheat germ, wheat
mas51,52 is proportional to the time of exposure to gluten bran, whole wheat, and cracked wheat. Foods made from
suggest that prompt diagnosis is crucial to minimize if rye and barley are toxic. Malt is also toxic because it is a
not prevent serious complications. Based on epidemio- partial hydrolysate of barley prolamins. It may contain
logic data, it might be hypothesized that if CD develops 100 –200 mg of barley prolamins /100 g of malt.113 In
early with typical gastrointestinal symptoms, prompt general, an ingredient with malt in its name (barley
diagnosis and thus timely prescription of a GFD are more malt, malt syrup, malt extract, malt flavorings) is made
likely. If, on the other hand, symptoms are atypical or from barley.
February 2001 DIAGNOSIS AND TREATMENT OF CELIAC DISEASE 647

Table 4. General Guidelines for the CD Diet

Not allowed Allowed
Wheats (Triticum family) Rice, wild rice
All forms, including Einkorn wheat (Triticum monococcum) Corn (maize)
Wheat flour
Wheat germ
Wheat bran
Cracked wheat
Emmer wheat (Triticum dicoccon) Sorghum
Couscous (endosperm of durum wheat) Millet
Kamut (Triticum polonicum) Buckwheat (kasha)
Spelt (farro, drinkle) Beans, peas, and bean flours
Semolina (durum wheat) Quinoa
Rye (Secale cereale) Potato
Triticale (wheat-rye hybrid) Soybean
Barley (Hordeum vulgare) and malt Tapioca
Amaranth
Teff
Nuts
Fruits
Milk (cheesesa)
Plain meat
Fish
Egg
Oat (Avena sativa)b
a The coat of some cheeses may contain gluten.
bAwaiting definitive scientific confirmation and regulation to avoid cross-contamination.

Gluten in Medications consumers and health care professionals to registered

Medications and vitamin and mineral supple- dietitians who have expertise in special diseases. The
ments may also contain gluten as an inactive ingredient. Consumer Nutrition Hotline can also provide phone
The inactive ingredients of these products can be numbers and addresses of companies within the food
changed by the manufacturers without warning because industry to help clarify the ingredients of a given food
there are no regulations on the formulation of inactive product and how it has been processed.
drug components. Nebulous (questionable) ingredients, Problems in Practical Dietary Management
such as vegetable gum and modified food starch, can
contain gluten. All medications should be checked for Possible gluten contamination of products that
nebulous ingredients, especially if they must be taken for are presumed to be gluten-free is a recurrent problem.
a long period. It is imperative to know the lot number of This cross-contamination can happen in farms where the
nonprescription medications when contacting the man- grains are grown and harvested, in mills where grains are
ufacturer for clarification of the inactive ingredients. processed into flours, or on food processing lines where
Prescription medications purchased through a phar- one line produces a food that includes gluten and the line
macy come with an ingredient list on the package insert. next to it produces a gluten-free product. Contamination
However, different batches of medications may contain might also occur in stores where grains are available from
different ingredients. open bins, in restaurants, at salad bars, or any place
The limited expertise of health care professionals re- where a variety of different meals are produced or differ-
garding celiac diet and the absence of federal regulations ent ingredients come together.114
for accurate food and drug labeling both represent sig-
nificant challenges for patients with newly diagnosed Refractory Sprue
CD. Despite the efforts of celiac support groups, there are In a minority of adult patients, CD does not
still no laws regulating gluten-free labeling in the respond to treatment with a gluten-free diet. The most
United States. The American Dietetic Association’s Na- likely cause of nonresponsiveness is continued gluten
tional Center for Nutrition and Dietetics Consumer Nu- ingestion, which can be voluntary or inadvertent. Other
trition Hotline at 1-800-366-1655 is a valuable source of causes of nonresponsiveness that must be considered
updated information on the treatment of CD. One of the include other food intolerance diseases (e.g., milk, soya),
functions of the Consumer Nutrition Hotline is to refer pancreatic insufficiency, enteropathy-associated T-cell
648 FASANO AND CATASSI GASTROENTEROLOGY Vol. 120, No. 3

Table 5. Research Priorities Identified at the 9th cussed both at the 9th International Symposium on CD
International Symposium on CD that was held on August 10 –13 in Baltimore123 and at
Area of research the first World Congress of Pediatric Gastroenterology,
Searching for the CD genes
Hepatology, and Nutrition in Boston.124 Although some
Developing a vaccine against CD
Who, when, and how to screen for CD of these goals are in an advanced state of development
Engineering gluten-free grains (i.e., engineering gluten-free grains), others (i.e., the
Gaining more insight on CD pathogenesis search for the CD genes) are extremely challenging and
Developing noninvasive, fast, and reliable tests for the
diagnosis and follow-up of CD
will require an international task force to generate mean-
Web information ingful data. Nevertheless, the appreciation that CD is not
http://www.celiaccenter.org a disease confined in Europe but a global problem affect-
http://www.nowheat.com/grfx/nowheat/index.htm
http://www.niddk.nih.gov/health/digest/pubs/celiac/index.htm
ing continents such as North and South America, Africa,
http://www.fastlane.net/homepages/thodge/archive.shtml and Asia, where it was historically considered an ex-
tremely rare condition, is catalyzing the scientific atten-
tion of new generations of investigators who will surely
help achieve these challenging targets.
lymphoma, refractory sprue, and ulcerative jejunitis. Pa-
tients with CD in whom the lack of compliance to a GFD
References
has been ruled out belong to the refractory sprue cate-
1. Sollid L, Thorsby E. HLA susceptibility genes in celiac disease:
gory. An aberrant clonal intraepithelial T-cell population genetic mapping and role in pathogenesis. Gastroenterology
can be found in up to 75% of patients with refractory 1993;105:910 –922.
sprue, a condition that is currently classified as cryptic 2. Shuppan D. Current concepts of celiac disease pathogenesis.
Gastroenterology 2000;119:234 –242.
enteropathy-associated T-cell lymphoma.115 These pa- 3. Madara JL, Trier JS. Structural abnormalities of jejunal epithelial
tients typically undergo pharmacologic therapies, in- cell membranes in celiac sprue. Lab Invest 1980;43:254 –261.
cluding treatment with steroids116,117 or immunosup- 4. Schulzke JD, Bentzel CJ, Schulzke I, Riecken EO, Fromm M.
pressants, such as azathioprine118 and cyclosporine.119 If Epithelial tight junction structure in the jejunum of children with
acute and treated celiac sprue. Pediatr Res 1998;43:435– 441.
patients do not respond to these treatments, the ultimate 5. Wang W, Uzzau S, Goldblum SE, Fasano A. Human zonulin, a
treatment is total parenteral nutrition. None of these potential modulator of intestinal tight junctions. J Cell Sci 2000;
therapies have been subjected to rigorous controlled 113:4435– 4440.
6. Fasano A, Not T, Wang W, Uzzau S, Berti I, Tommasini A,
studies.120 Goldblum SE. Zonulin, a newly discovered modulator of intesti-
In young children with villus atrophy whose symp- nal permeability, and its expression in coeliac disease. Lancet
toms do not respond to a gluten-free diet, diseases that 2000;358:1518 –1519.
7. Ventura A, Magazzu G, Greco L. Duration of exposure to gluten
must be considered include tufting enteropathy and and risk for autoimmune disorders in celiac patients. Gastroen-
other congenital ultrastructural abnormalities of the en- terology 1999;117:303–310.
terocyte,121 unrecognized chronic giardiasis, and autoim- 8. Bjorkman PJ, Saper MA, Samraoui B, Bennet WS, Strominger JL,
mune enteropathy. Wiley DC. The foreign antigen binding site and T-cell recognition
regions of class I immunohistocompatibility antigens. Nature
For a more comprehensive overview on refractory 1987;329:512–518.
sprue, the reader is referred to a recently published 9. Cuvelier C, Barbatis C, Mielants H, De Vos H, Roels H, Veys E.
review.122 Histopathology of intestinal inflammation related to reactive
arthritis. Gut 1987;28:394 – 401.
10. Molberg O, McAdam SN, Sollid LM. Role of tissue transglutami-
Future Directions nase in celiac disease. J Pediatr Gastroenterol Nutr 2000;30:
232–240.
A multidisciplinary research effort to understand 11. Dieterich W, Ehnis T, Bauer M, Donner P, Volta U, Riecken E,
the pathogenesis of CD is currently taking place world- Schuppan D. Identification of tissue transglutaminase as the
wide. This effort is fueled by the appreciation that CD autoantigen of celiac disease. Nat Med 1997;3:797– 801.
12. Mäki M, Kallonen K, Lahdehao ML, Visakorpi JK. Changing
represents a unique example of an autoimmune disease in pattern of childhood coeliac disease in Finland. Acta Paediatr
which the environmental factor(s) that induce the im- Scand 1988;77:408.
mune response has been identified. Therefore, scientists 13. Logan RF, Tucker G, Rifkind E, Heading RC, Ferguson A.
Changes in clinical features of coeliac disease in adults in
view CD as a model to tackle key questions on the Edinburgh and the Lothians 1960-79. BMJ 1983;286:95–97.
pathogenic mechanisms involved in other autoimmune 14. Swinson CM, Levi AJ. Is coeliac disease underdiagnosed? BMJ
diseases (i.e., multiple sceloris, diabetes mellitus, rheu- 1980;281:1258 –1260.
matoid arthritis, etc.) whose environmental triggers are 15. Pare P, Douville P, Caron D, Lagace R. Adult coeliac sprue:
changes in the pattern of clinical recognition. Gastroenterology
still unknown. Future directions in CD research (Table 1988;10:395.
5) have been clearly identified and were recently dis- 16. Berti I, Horvath K, Green PHR, Sblattero D, Not T, Fasano A.
February 2001 DIAGNOSIS AND TREATMENT OF CELIAC DISEASE 649

Differences of celiac disease’s clinical presentation among pe- Zaccaria T, Di Stefano M, Isaia CC. The effects of 1-year gluten
diatric and adults relatives of CD patients in U.S.A. J Invest Med withdrawal on bone mass, bone metabolism and nutritional
2000;48:215A. status in newly-diagnosed adult coeliac disease patients. Ali-
17. Fry L. Dermatitis herpetiformis. Baillières Clin Gastroenterol ment Pharmacol Ther 2000;14:35– 43.
1995;9:371–393. 38. Hadjivassiliou M, Gibson A, Davies-Jones GA, Lobo AJ, Stephen-
18. Reunala T, Collin P. Diseases associated with dermatitis herpet- son TJ, Milford-Ward A. Does cryptic gluten sensitivity play a part
iformis. Br J Dermatol 1997;136:315–318. in neurological illness? Lancet 1996;347:369 –371.
19. Carroccio A, Iannitto E, Cavataio F, Montalto G, Tumminello M, 39. Pellecchia MT, Scala R, Filla A, De Michele G, Ciacci C, Barone
Campagna P, Lipari MG, Notarbartolo A, Iacono G. Sideropenic P. Idiopathic cerebellar ataxia associated with celiac disease:
anemia and celiac disease: one study, two points of view. Dig lack of distinctive neurological features. J Neurol Neurosurg
Dis Sci 1998;43:673– 678. Psychiatry 1999;66:32–35.
20. Groll A, Candy DC, Preece MA, Tanner JM, Harries JT. Short 40. Farthing MJ, Rees LH, Edwards CR, Dawson AM. Male gonadal
stature as the primary manifestation of coeliac disease. Lancet function in coeliac disease. 2. Sex hormones. Gut 1983;24:
1980;2:1097. 127–136.
21. Cacciari E, Salardi S, Lazzari R, Cicognani A, Collina A, Pirazzoli 41. Auricchio S, Greco L, Troncone R. Gluten-sensitive enteropathy
P, Tassoni P, Biasco G, Corazza GR, Cassio A. Short stature and in childhood. Pediatr Clin North Am 1988;35:157–187.
coeliac disease: a relationship to consider even in patients with 42. Infertilities and CD (editorial). Lancet 1983;1:453.
no gastrointestinal symptoms. J Pediatr 1983;103:708. 43. Gasbarrini A, Sanz Torre E, Trivellini C, De Carolis S, Caruso A,
22. Stenhammar C, Fallstrom SP, Jansson G, Lindberg T. Celiac Gasbarrini G. Recurrent spontaneous abortion and intrauterine
disease in children of short stature without gastrointestinal fetal growth retardation as symptoms of coeliac disease. Lancet
symptoms. Eur J Pediatr 1986;145:185. 2000;256:399 – 400.
23. Verkasalo M, Kuitenen P, Leisti S, Perheentupa J. Growth failure 44. Ciacci C, Cirillo M, Auriemma G, Di Dato G, Sabbatini F, Mazza-
from symptomless coeliac disease. Helv Paediatr Acta 1978; cca G. Celiac disease and pregnancy outcome. Am J Gastroen-
47:489 – 495. terol 1996;91:718 –722.
24. Prader A, Tanner JM, Von Harnack GA. Catch-up growth in coe- 45. Auricchio S, Mazzacca G, Tosi R, Deritis G. Coeliac disease as
liac disease. Acta Paediatr Scand 1969;58:311–316. a familial condition: identification of asymptomatic coeliac pa-
25. Smith DMH, Miller J. Gastroenterology, coeliac disease and tients within family groups. Gastroenterol Int 1988;1:25.
enamel hypoplasia. Br Dent J 1979;147:91–95. 46. Hed J, Leiden G, Ottosson E, Strom M, Walan A, Groth O,
26. Aine L. Dental enamel defects and dental maturity in children Sjogren F, Franzen L. IgA anti-gliadin antibodies and jejunal
and adolescents with coeliac disease. Proc Finn Dent Soc mucosal lesions in healthy blood donors (letter). Lancet 1986;
1986;82:222–229. 2:215.
27. George EK, Hertzberger-ten Cate R, Van Suijlekom-Smit LW, Von 47. Ferguson A, Arranz E, O’Mahony S. Clinical and pathological
Blomberg BM, Stapel SO, Van Elburg R M, Mearin ML. Juvenile spectrum of coeliac disease—active, silent, latent, potential.
chronic authritis and coeliac disease in the Netherlands. Clin Gut 1993;34:150 –151.
Exp Rheumatol 1996;14:571–575. 48. Maki M, Kallonen K, Lahdeaho ML, Visakorpi JK. Changing
28. Lepore L, Martelossi S, Pennesi M, Falcini F, Ermini ML, Ferrari pattern of childhood coeliac disease in Finland. Acta Paediatr
R, Perticarari S, Presani G, Lucchesi A, Lapini M, Ventura A. Scand 1988;77:408 – 412.
Prevalence of celiac disease in patients with juvenile chronic 49. Mustalahti K, Collin P, Sievanen H, Salmi J, Mäki M. Osteopenia
arthritis. J Pediatr 1996;129:311–313. in patients with clinically silent coeliac disease warrants screen-
29. O’Farrelly C, Marten D, Melcher D, McDougall B, Price R, Gold- ing. Lancet 1999;354:744 –745.
stein AJ, Sherwood R, Fernandes L. Association between villous 50. Fabiani E, Catassi C, Villari A, Gismondi P, Pierdomenico R,
atrophy in rheumatoid arthritis and a rheumatoid factor and Rätsch IM, Coppa GV, Giorgi PL. Dietary compliance in screen-
gliadin-specific IgG. Lancet 1988;2:819 – 822. ing-detected coeliac disease adolescents. Acta Paediatr Suppl
30. Mäki M, Hallstrom O, Verronen P, et al. Reticulin antibody 1996;412:65– 67.
arthritis and coeliac disease in children. Lancet 1988;1:479 – 51. Swinson CM, Slavin G, Coles EC, Booth CC. Coeliac disease and
480. malignancy. Lancet 1983;1:111–115.
31. Maggiore G, De Giacomo C, Scotta MS, Sessa F. Coeliac dis- 52. Logan RFA, Rifkind EA, Turner ID, Ferguson A. Mortality in coe-
ease presenting as chronic hepatitis in a girl. J Pediatr Gastro- liac disease. Gastroenterology 1989;97:265–271.
enterol Nutr 1986;5:501–503. 53. Davidson LSP, Fountain JR. Incidence of sprue syndrome with
32. Leonardi S, Bottaro G, Patane’ R, Musumeci S. Hypertransami- some observation on the natural history. BMJ 1950;1:1157–
nasemia as the first symptom in infant coeliac disease. J Pedi- 1161.
atr Gastroenterol Nutr 1990;11:404 – 406. 54. Meeuwisse GW. Diagnostic criteria in coeliac disease. Acta
33. Vajro P, Fontanella A, Greco L. Hepatite chronique cryptogene- Paediatr Scand 1970;59:461– 463.
tique revelant une maladie coeliaque. Proceedings of Groupe 55. Mylotte M, Egan-Mitchell B, McCarthy CF, McNicholl B. Inci-
Francophone de Gastroenterologie et Nutrition Pediatrique. dence of coeliac disease in the west of Ireland. BMJ 1973;1:
Lyon, France, July 2–3, 1990. 703–705.
34. Fontanella A, Vajro P, Ardia E, Greco L. Danno epatico in corso 56. Logan RFA, Rifking EA, Busuttil A, Gilmous HM, Ferguson A.
di malattia celiaca. Studio retrospettivo in 123 bambini. Riv Ital Prevalence and ‘incidence’ of celiac disease in Edinburgh and the
Pediatr 1987;5:80 – 85. Lothian region of Scotland. Gastroenterology 1986;90:334 –342.
35. Valdimarsson T, Lofman O, Toss G, Strom M. Reversal of os- 57. Van Stirum J, Baerlocher K, Fanconi A, Gugler E, Shmerling DH.
teopenia with diet in adult coeliac disease. Gut 1996;38:322– The incidence of coeliac disease in children in Switzerland. Helv
327. Paediatr Acta 1982;37:421– 430.
36. Mora S, Barera G, Beccio S, Proverbio MC, Weber G, Bianchi C, 58. Littlewood JM, Crollick AJ, Richards IDG. Childhood coeliac dis-
Chiumello G. Bone density and bone metabolism are normal ease is disappearing. Lancet 1980;2:1359.
after long-term gluten-free diet in young celiac patients. Am J 59. Dossetor JFB, Gibson AAM, McNeish AS. Childhood coeliac
Gastroenterol 1999;94:398 – 403. disease is disappearing. Lancet 1981;1:322–323.
37. Sategna-Guidetti C, Grosso SB, Grosso S, Mengozzi G, Aimo G, 60. Stevens FM, Egar-Mitchell B, Cryan E, McCarthy CF, McNicholl B.
650 FASANO AND CATASSI GASTROENTEROLOGY Vol. 120, No. 3

Decreasing incidence of coeliac disease. Arch Dis Child 1987; Sblattero D, Not T, Pietzak M, Hill I, Murray J, Fine K, Fasano A.
62:465– 468. Celiac disease in U.S.A. among risk groups and the general
61. Greco L, Maki M, DiDonato F, Visakorpi JK. Epidemiology of population in USA (abstr). J Pediatr Gastroenterol Nutr 2000;
coeliac disease in Europe and the Mediterranean area. Dyn Nutr 31:S29.
Res 1992;2:25– 44. 82. Gandolfi L, Pratesi R, Cordoba JCM, Tavil PL, Gasparin M,
62. Catassi C, Ratsch IM, Fabiani E, Rossini M, Bordicchia F, Can- Catassi C. Prevalence of celiac disease among blood donors in
dela F, Coppa GU, Giorci PL. Coeliac disease in the year 2000: Brazil. Am J Gastroenterol 2000;95:689 – 692.
exploring the iceberg. Lancet 1994;343:200 –203. 83. Catassi C, Ratsh IM, Gandolfi L, Pratesi R, Fabiani E, El Asmar
63. Catassi C, Ratsch IM, Fabiani E, Ricci S, Bordicchia F, Pierdo- R, Frijia M, Bearzi I, Vizzoni L. Why is coeliac disease endemic in
menico R, Giorgi PL. High prevalence of undiagnosed coeliac the people of the Sahara? Lancet 1999;354:647– 648.
disease in 5280 Italian students screened by antigliadin anti- 84. Yachha SK, Mohindra S, Srivastava A, Krishnani N, Saxena A.
bodies. Acta Paediatr 1995;84:572–576. Effects of gluten-free diet on growth and small bowel histology in
64. Mazzetti di Pietralata M, Giorgetti GM, Gregori M, De Simone M, children with celiac disease in India. J Pediatr Gastroenterol
Leonardi C, Barcetta PA, Ricciardi MM, Sandri G. Sub-clinical Nutr 2000(suppl);31:S23.
coeliac disease. Ital J Gastroenterol 1992;24:352–354. 85. Shahbazkhani B, Maghari M, Nasseri Moghaddam S, Kamalian
65. Grodzinsky E, Franzen L, Hed J, Strom M. High prevalence of N, Sotoudeh M, Minapour M, Malekzadeh R. Prevalence of
celiac disease in healthy adults revealed by antigliadin antibod- celiac disease among Iranian patients with chronic diarrhea.
ies. Ann Allergy 1992;69:66 –70. J Pediatr Gastroenterol Nutr 2000(suppl);31:S4.
66. Ascher H, Krantz I, Kristiansson B. Increasing incidence of 86. Logan RFA. Problems and pitfalls in epidemiological studies of
celiac disease in Sweden. Arch Dis Child 1991;66:608 – 611. coeliac disease. Dyn Nutr Res 1992;2:14 –24.
67. Kolho KL, Farkkila MA, Savilahti E. Undiagnosed celiac disease 87. Picarelli A, Maiuri L, Mazzilli MC, Coletta S, Ferrante P, Di
is common in Finnish adults. Scand J Gastroenterol 1998;33: Giovanbattista F, Greco M, Torsoli A, Auricchio S. Gluten-sensi-
1280 –1283. tive disease with mild enteropathy. Gastroenterology 1996;
68. Catassi C, Fabiani E, Ratsch IM, Coppa GV, Giorgi PL, Pierdo- 111:608 – 616.
menico R, Alessandrini S, Iwanejko G, Domenici R, Mei E, Miano 88. Holmes GKT, Catassi C. Coeliac disease. Oxford, England:
A, Maran M, Bottaro G, Spina M, Dotti M, Montanelli A, Barbato Health Press—Fast Facts, 2000.
M, Viola F, Lazzari R, Vallini M, Guariso G, Plebani M, Cataldo F,
89. Ivarsson A, Persson LA, Nyström L, Ascher H, Cavell B, Daniels-
Traverso G, Ventura A, De Simone M, Leonardi C, Barcetta PA,
son L, Dannaeus A, Lindberg T, Lindquist B, Stenhammar L,
Ricciardi MM, Sandri G. The coeliac iceberg in Italy. A multi-
Hernell O. Epidemic of coeliac disease in Swedish children. Acta
centre antigliadin antibodies screening for coeliac disease in
Paediatr 2000;89:165–171.
school-age subsets. Acta Paediatr Suppl 1996;412:29 –35.
90. Meeuwisse GW. Diagnostic criteria in coeliac disease. Acta
69. Weile B, Krasilnifoff PA. Extremely low incidence rate of celiac
Paediatr Scand 1970;59:461– 463.
disease in the Danish population of children. J Clin Epidemiol
91. Burgin-Wolff A, Berger R, Gaze H, Huber H, Lentze MJ, Nussle D.
1993;46:661– 664.
IgA and IgG gliadin antibody determinations as screening test
70. Ascher H, Krantz I, Kristiansson B. Increasing incidence of
for untreated coeliac disease in children, a multicentre study.
coeliac disease in Sweden. Arch Dis Child 1991;66:608 – 611.
Eur J Pediatr 1989;148:496 –502.
71. Maki M, Holm K, Ascher H, Greco L. Factors affecting clinical
92. Murray JA, Herlein J, Goeken J. Multicenter comparison of se-
presentation of coeliac disease: role of type and amount of
rologic tests for celiac disease in the USA: results of phase 1
gluten-containing cereals in the diet. Dyn Nutr Res 1992;2:76 –
serological comparison (abstr). Gastroenterology 1997;112:A389.
82.
72. Ascher H, Kristiansson B. Childhood coeliac disease in Sweden. 93. Sulkanen S, Halttunen T, Laurila K, Kolho KL, Korponay-Szabo
Lancet 1994;44:340 –341. IR, Sarnesto A, Savilathi E, Colun P, Maki M. Tissue transglu-
73. Fasano A. Where have all American celiacs gone? Arch Dis Child taminase autoantibody enzyme-linked immunosorbent assay in
1996;412:20 –24. detecting celiac disease. Gastroenterology 1998;115:1322–
74. Rossi TM, Slbini CH, Kumar V. Incidence of celiac disease 1328.
identified by the presence of serum endomysial antibodies in 94. Troncone R, Maurano F, Rossi M, Micillo M, Greco L, Auricchio
children with chronic diarrhea, short stature, or insulin-depen- R, Salerno G, Salvatore F, Sacchetti L. IgA antibodies to tissue
dent diabetes mellitus. J Pediatr 1993;123:262–264. transglutaminase: an effective diagnostic test for celiac dis-
75. Talley NJ, Valdovinos M, Petterson TM, Carpenter HA, Melton L ease. J Pediatr 1999;134:166 –171.
Jr. Epidemiology of celiac sprue: a community-based study. Am J 95. Fasano A. Tissue transglutaminase: the holy grail for the diag-
Gastroenterol 1994;89:843– 846. nosis of celiac disease, at last? J Pediatr 1999;13:134 –135.
76. Hill ID, Horvath K, Fasano A. Epidemiology of celiac disease. 96. Sblattero D, Berti I, Trevisiol C, Marzari R, Tommasini A, Brad-
Am J Gastroenterol 1995;90:163–164. bury A, Fasano A, Ventura A, Not T. Human tissue transglutami-
77. Hill I, Fasano A, Schwartz R, Counts D, Glock M, Horvath K. nase ELISA: a powerful mass screening diagnostic assay for
Prevalence of celiac disease in at risk groups of children in celiac disease. Am J Gastroenterol 2000:98:1253–1257.
United States. Pediatr Res 2000;136:86 –90. 97. Baldas V, Tommasini A, Trevisiol C, Berti I, Fasano A, Sblattero
78. Berti I, Horvath K, Green PHR, Not T, Fasano A. Prevalence of D, Bradbury A, Marzari R, Ventura A, Not T. The development of
celiac disease among first and second degree relatives in the a novel rapid non-invasive screening test for celiac disease. Gut
U.S.A. (abstr). Gastroenterology 1999;116:A861. 2000;47:628 – 631.
79. Berti I, Horvath K, Green PHR, Sblattero D, Not T, Fasano A. 98. Walker-Smith JA, Guandalini S, Schmitz J, Shmerling DH, Visa-
Prevalence of celiac disease among risk groups and the general korpi JK. Revised criteria for diagnosis of coeliac disease. Arch
population in U.S.A. (abstr). J Invest Med 2000;48:220A. Dis Child 1990;65:909 –911.
80. Not T, Horvath K, Hill ID, Partanen J, Hammed A, Magazzu G, 99. Scott BB, Losowsky MS. Patchiness and duodenal-jejunal vari-
Fasano A. Celiac disease risk in the USA: high prevalence of ation of the mucosal abnormality in coeliac disease and derma-
antiendomysium antibodies in healthy blood donors. Scand J titis herpetiformis. Gut 1976;17:984 –992.
Gastroenterol 1998;33:494 – 498. 100. Hummel M, Bonifacio E, Stern M, Dittler J, Schimmel A, Ziegler
81. Gerarduzzi T, Berti I, De Gregorio E, Horvath K, Green PHR, AG. Development of celiac disease-associated antibodies in
February 2001 DIAGNOSIS AND TREATMENT OF CELIAC DISEASE 651

offspring of parents with type I diabetes. Diabetologia 2000; agement of patients with non-responsive coeliac disease. Alim
43:1005–1011. Pharmacol Ther 1996;10:671– 680.
101. Iltanen S, Collin P, Korpela M, Holm K, Partanen J, Polvi A, Mäki 121. Patey N, Scoazec JY, Cuenod-Jabri B, Canioni D, Kedinger M,
M. Celiac disease and markers of celiac disease latency in Goulet O, Brousse N. Distribution of cell adhesion molecules in
patients with primary Sjögren’s syndrome. Am J Gastroenterol infants with intestinal epithelial dysplasia (Tufting enteropathy).
1999;94:1042–1046. Gastroenterology 1997;113:833– 843.
102. Cataldo F, Marino V, Ventura A, Bottaro GR, SIGEP, Club del Tenue 122. Ryan BM, Kelleher D. Refractory celiac disease. Gastroenterol-
Working Groups on Coeliac Disease. Prevalence and clinical fea- ogy 2000;119:243–251.
tures of selective immunoglobulin A deficiency in coeliac disease: 123. 9th International Symposium on Celiac Disease. J Pediatr Gas-
an Italian multicentre study. Gut 1998;42:362–365. troenterol Nutr 2000;31(suppl 3):S1–S35.
103. Viscardi M, Barera G, Bonfanti R, Menni L, Zappa L, Leone BE, 124. World Congress of Pediatric Gastroenterology, Hepatology and
Bianchi C, Chiumello G. Occurrence of coeliac disease after the Nutrition Proceedings: working group report on celiac disease.
onset of insulin-dependent diabetes mellitus: a six years fol- 125. Catassi C, Fabiani E, Ratsch IM, Coppa GV, Giorgi PL, Pierdo-
low-up (abstr). J Pediatr Gastroenterol Nutr 2000;31:S216. menico R. The coeliac iceberg in Italy. A multicentre antigliadin
104. Leonard JN, Tucker WFG, Fry JS, Coulter CA, Boylston AW, antibodies screening for coeliac disease in school-age subjects.
McMinn RM, Haffenden GP, Swain AF, Fry L. Increased inci- Acta Paediatr Suppl 1996;412:29 –35.
dence of malignancy in dermatitis herpetiformis. BMJ 1983; 126. Johnston SD, Watson RG, McMillan SA, Sloan J, Love AH.
286:16 –18. Prevalence of coeliac disease in Northern Ireland. Lancet 1997;
105. Catassi C, Corrao G, Barbato M, De Renzo A, Carroccio A, 350:1370.
D’Altilia MR, Guariso G, Baldassarre M, Caramaschi P, Berto- 127. Ascher H, Krantz I, Kristiansson B. Increasing incidence of
lani P, Bellentani S, Corazza GR. The risk of non-Hodgkin lym- coeliac disease in Sweden. Arch Dis Child 1991;66:608 – 611.
phoma (NHL) in patients with coeliac disease (CD): preliminary 128. Csizmadia CG, Mearin ML, von Blomberg BM, Brand R, Ver-
results of the Italian multicentre case-control study (abstr). loove-Vanhorick SP. An iceberg of childhood coeliac disease in
Gastroenterology 1999;116:G3770. The Netherlands. Lancet 1999;353:813– 814.
106. Colletti RB. The effect of prevalence on the value of serological 129. Hovdenak N, Hovlid E, Aksnes L, Fluge G, Erichsen MM, Eide J.
tests for celiac disease. J Pediatr Gastroenterol Nutr 2000; High prevalence of asymptomatic coeliac disease in Norway: a
31(suppl 3):S11. study of blood donors. Eur J Gastroenterol Hepatol 1999;11:
107. Gerarduzzi T, Horvath K, Rabsztyn A, Berti I, Kryszak D, Fasano 185–187.
A. Should risk group patients be retested for celiac disease? 130. Kolho KL, Farkkila MA, Savilahti E. Undiagnosed coeliac dis-
(abstr). J Pediatr Gastroenterol Nutr 2000;31:S29. ease is common in Finnish adults. Scand J Gastroenterol 1998;
108. Fabiani E, Taccari LM, Rätsch IM, Di Giuseppe S, Coppa GV, 33:1280 –1283.
Catassi C. Compliance with gluten-free diet in adolescents with 131. Rostami K, Mulder CJ, Werre JM, van Beukelen FR, Kerchhaert
screening-detected celiac disease: a 5-year follow-up study. J, Crusius JB, Pena AS, Willekens FL, Meijer JU. High prevalence
J Pediatr 2000;136:841– 843. of celiac disease in apparently healthy blood donors suggests a
109. Hallert C, Granno C, Grant C, Hulten S, Midhagen G, Strom M, high prevalence of undiagnosed Celiac disease in the Dutch
Svensson H, Valdimarsson T, Wickstrom T. Quality of life of population. Scand J Gastroenterol 1999;34:276 –279.
adult coeliac patients treated for 10 years. Scand J Gastroen- 132. Gandolfi L, Bocca AL, Pratesi R. Screening of celiac disease in
terol 1998;33:933–938. children attending the outpatient clinic of a university hospital
110. Hin H, Bird G, Fisher P, Mahy N, Jewell D. Coeliac disease in (abstr). J Pediatr Gastroenterol Nutr 2000;31:S212.
primary care: case finding study. BMJ 1999;318:164 –167. 133. Lindquist BL, Rogozinski T, Moi H, Danielsson D, Olcén P.
111. Hardman CM, Garioch JJ, Leonard JN, Thomas HJ, Walker MM, Endomysium and gliadin IgA antibodies in children with coeliac
Lortan JE, Lister A, Fry L. Absence of toxicity of oats in patients with disease. Scand J Gastroenterol 1994;29:452– 456.
dermatitis herpetiformis. N Engl J Med 1997;337:1884 –1887. 134. Grodzinsky E, Jansson G, Skogh T, Stenhammar L, Fälth-Mag-
112. Forssell F, Wieser H. Spelt wheat and celiac disease. Z Lebens- nusson K. Anti-endomysium and anti-gliadin antibodies as se-
mittel Untersuch Forsch 1995;201:35–39. rological markers for coeliac disease in childhood: a clinical
113. Ellis HJ, Doyle AP, Day P, Wieser H, Ciclitira PJ. Demonstration study to develop a practical routine. Acta Paediatr 1995;84:
of the presence of coeliac-activating gliadin-like epitopes in 294 –298.
malted barley. Int Arch Allergy Immunol 1994;104:308 –310. 135. Cataldo F, Ventura A, Lazzari R, Balli F, Nassimbeni G, Marino V.
114. Horvath K, Fasano A. Celiac disease in children. Curr Opt Gas- Antiendomysium antibodies and coeliac disease: solved and
troenterol (in press). unsolved questions. An Italian multicentre study. Acta Paediatr
115. Cellier C, Delabesse E, Helmer C, Patey N, Matuchansky C, 1995;84:1125–1131.
Jabri B, Macintyre E, Cerf-Bensussan N, Brousse N, French 136. Russo PA, Chartrand LJ, Seidman E. Comparative analysis of
Coeliac Disease Study Group. Refractory sprue, coeliac dis- serologic screening tests for the initial diagnosis of celiac dis-
ease, and enteropathy-associated T-cell lymphoma. Lancet ease. Pediatrics 1999:104:75–78.
2000;356:203–208. 137. Valdimarsson T, Franzen L, Grodzinsky E, Skogh T, Ström M. Is
116. Stuart BM, Gent AE. Atrophy of the coeliac mucosa. Eur J small bowel biopsy necessary in adults with suspected celiac
Gastroenterol Hepatol 1998;10:523–525. disease and IgA anti-endomysium antibodies? Dig Dis Sci
117. Mitchison HC, Al Mardini H, Gillespie S, Laker M, Zaitoun A, 1996;41:83– 87.
Record CO. A pilot study of fluticasone propionate in untreated 138. Troncone R, Ferguson A. Anti-gliadin antibodies. J Pediatr Gas-
coeliac disease. Gut 1991;32:260 –265. troenterol Nutr 1991;12:150 –158.
118. Vaidya A, Bolanos J, Berkelhammer C. Azathioprine in refractory
sprue. Am J Gastroenterol 1999;94:1967–1969.
119. Rolny P, Sigurjonsdottir HA, Remotti H, Nilsson LA, Ascher H, Received September 6, 2000. Accepted November 29, 2000.
Tlaskalova-Hogenova H, Tuckoua L. Role of immunosuppressive Address requests for reprints to: Alessio Fasano, M.D., Center for
therapy in refractory sprue-like disease. Am J Gastroenterol Celiac Research, University of Maryland School of Medicine, 685 West
1999;94:219 –225. Baltimore Street HSF Building, Room 465, Baltimore, Maryland
120. O’Mahony S, Howdle PD, Losowsky SM. Review article: man- 21201. e-mail: [email protected]; fax: (410) 328-1072.