Keywords esophagitis - gastritis - histologic changes
Introduction
The field of pediatric gastroenterology and hepatology has evolved substantially over
the last two decades. As a result, more gastrointestinal (GI) procedures, such as
esophagogastroduodenoscopy (EGD), considering the standard diagnostic test to evaluate
children and adolescents with GI complaints, are being performed. The increase in
GI procedures performed in children and adolescents has resulted in reporting many
histologic findings with unclear clinical significance, such as chronic gastritis
and lymphocytic duodenosis (LD).[1 ]
[2 ]
[3 ] LD is defined as an increased number of (>25) intraepithelial lymphocytes with intact
villous architecture per 100 epithelial cells.[4 ]
[5 ]
[6 ]
[7 ] It is not clear if the increase in the number of EGDs performed has any significant
effect on the rate of reporting these histologic findings, including LD, in children
and adolescents.
Previous studies have shown that LD is associated with chronic gastritis due to Helicobacter pylori , celiac disease (CD), inflammatory bowel disease, and other autoimmune diseases.[8 ]
[9 ] However, it is not clear if LD is associated with concurrent abnormal gastric and
esophageal pathology, such as gastritis and/or esophagitis. While in adults the presenting
symptoms will likely impact a decision to biopsy the esophagus, stomach, and small
bowel (duodenum), it has been a long-standing practice in children and adolescents
to biopsy the esophagus, stomach, and small bowel in all diagnostic EGDs. This different
approach in children and adolescents allows assessment of 1) the correlation between
the number of EGDs performed and the rate of LD reporting and 2) the association of
LD with abnormal gastric and esophageal pathology.
Methods
We performed a single-center retrospective study using the Mayo Clinic electronic
health record and pathology database, reviewing all EGD procedures performed in children
and adolescents (<18 years) from January 1, 2000, through December 31, 2018. Pathology
reports were reviewed by a study member (J.D.) to identify individuals with LD on
their first diagnostic EGD (LD group), before reviewing their clinical notes and laboratory
results. We selected age- and sex-matched children and adolescents who underwent EGD
with normal duodenal biopsies during the same period (control group).
All slides for both groups were initially reviewed and reported by experienced GI
pathologists who were trained to review adult and pediatric GI biopsies. These pathologists
were not blinded to the patient's clinical information. Clinical diagnoses and documentation
of proton pump inhibitor (PPI) exposure within 3 months from EGD were reviewed and
compared between the two groups.
We excluded individuals with known GI diagnoses, such as CD, Crohn's disease, and
H pylori . The number of EGDs performed in the LD and control groups were tabulated for each
year from 2000 through 2018. A linear graph was generated to graphically explore a
change in counts over the time period ([Fig. 1 ]). The correlation between the number of EGDs per year and the percentage of LD-positive
biopsies was assessed using Pearson's correlation coefficient. Logistic regression
analysis assessed the association of gastric or esophageal histologic abnormalities
with the presence of LD. A type I error rate of 0.05 was used to assess statistical
significance. Analyses were performed using SAS v9.4 M6 (SAS Institute Inc).
Fig. 1 Number of esophagogastroduodenoscopies performed and percentage of lymphocytic duodenosis
between 2000 and 2018. (A ) Number of EGDs performed. (B ) Percentage of lymphocytic duodenosis. EGD, esophagogastroduodenoscopy; LD, lymphocytic
duodenosis.
Results
During the study period, 15,979 EGDs were performed in 11,870 children and adolescents.
Of those, 6,008 (51%) were female and 5,862 (49%) male. Three hundred thirty-eight
(3%) individuals had LD identified on their first diagnostic EGD and were included
in the study, comprising the LD group. Three hundred ninety (3%) individuals were
randomly selected as the control group. The mean (SD) age was 9.6 (5.3) years at the
time of small bowel biopsy in the LD group, and 198 (59%) were female. In the control
group, the mean (SD) age at the time of small bowel biopsy was 11.7 (5.0) years, and
237 (61%) were female ([Table 1 ]).
Table 1
Characteristics of patients with abnormal gastric and esophageal biopsies
Patient characteristics
LD group (n = 338)
Control group (n = 390)
p -Value
Age at EGD, mean (SD), y
9.6 (5.3)
11.7 (5.0)
<0.001[a ]
Female sex
198 (59)
237 (61)
0.548[b ]
PPI exposure
150 (44)
113 (29)
<0.001[b ]
Abnormal gastric histology
138 (41)
76 (19)
<0.001[b ]
Chronic gastritis
95 (28)
45 (12)
Reactive gastritis
43 (13)
29 (7)
Lymphocytic gastritis[c ]
0 (0)
1 (<1)
Focal intestinal metaplasia
0 (0)
1 (<1)
Abnormal esophageal histology
54 (16)
45 (12)
0.081[b ]
Reflux esophagitis
33 (10)
26 (7)
EOE
20 (6)
18 (5)
Candida esophagitis
0 (0)
1 (<1)
Pancreatic acinar metaplasia
1 (<1)
0 (0)
Abbreviations: EGD, esophagogastroduodenoscopy; EOE, eosinophilic esophagitis; LD,
lymphocytic duodenosis; PPI, proton pump inhibitor.
a Equal variance t -est.
b χ2 test.
c A form of chronic gastritis with dense infiltration of foveolar epithelium by T lymphocytes.
The frequency of EGD procedures and LD findings were similar between males and females.
Counts of EGD procedures and percentage of LD from 2000 through 2018 are shown in
[Fig. 1A ] and [B ], respectively. The number of LD-positive biopsies per year was highly correlated
with the number of EGD procedures performed yearly (ρ = 0.931; 95% CI, 0.826–0.974;
P < 0.001). Beginning in 2010, the percentage of LD found on EGD began to increase,
and this coincided with an increase in the number of EGDs performed each year.
When comparing previous exposure to PPIs between the two groups, we found that PPI
use (44% vs. 29%, P < 0.001) was more common in the LD group ([Table 1 ]). In the LD group with abnormal gastric histology, 119 of 138 (86%) were taking
PPIs at the time of EGD. In the LD group, the median (range) of follow-up was 2.7
(0.0–190.9) months. The most common diagnoses in both groups were functional GI disorders,
such as chronic abdominal pain, irritable bowel syndrome, and constipation. A comparison
of clinical diagnoses between the two groups is summarized in [Table 2 ]. During the follow-up period, none were diagnosed with CD, Crohn's disease, or H . pylori .
Table 2
Final outcomes[a ]
Diagnosis
LD group (n = 338)
Control group (n = 390)
p -Value[b ]
EOE
20 (6)
19 (5)
0.613
Reflux
57 (17)
45 (12)
0.664
Fructose malabsorption
12 (4)
23 (6)
0.082
Small bowel bacterial overgrowth
3 (1)
17 (4)
0.001
IBS
12 (4)
23 (6)
0.008
Rumination syndrome
0 (0)
8 (2)
0.004
Functional GI disorders
140 (41)
152 (39)
0.444
Lactose intolerance
20 (6)
39 (10)
0.014
Abbreviations: EOE, eosinophilic esophagitis; GI, gastrointestinal; IBS, irritable
bowel syndrome; LD, lymphocytic duodenosis.
a Data are presented as No. (%) unless otherwise indicated.
b χ2 test.
Of the 338 patients in the LD group, abnormal pathology was reported in 138 (41%)
and 54 (16%) of gastric and esophageal biopsies, respectively. In addition, abnormal
gastric histologic findings included chronic gastritis in 95 (28%) and reactive gastropathy
in 43 (13%). Meanwhile, abnormal esophageal histologic findings included reflux esophagitis
in 33 (10%) and eosinophilic esophagitis (EOE) in 20 (6%) ([Table 1 ]).
Of the 390 patients in the control group, abnormal pathology was reported in 76 (19%)
and 45 (12%) gastric and esophageal biopsies, respectively. In addition, abnormal
gastric histology included chronic gastritis in 45 (12%) and reactive gastropathy
in 29 (7%). Abnormal esophageal histologic findings included reflux esophagitis in
26 (7%) and EOE in 18 (5%) ([Table 1 ]).
Seventy-one patients in the LD group (21%) underwent a second EGD within a median
(range) of 11.0 (0.0–146.0) months after their first EGD. Indication for a second
EGD was persistent symptoms in 54 individuals and follow-up EGD in 17 (8 for EOE and
9 for gastroesophageal reflux disease). Of those, 58 of 71 (82%) had normal duodenal
biopsies, and 13 (18%) had persistent LD on subsequent EGD. In the persistent symptom
group (54 individuals) follow-up EGD showed LD resolution in 45 (83%) with a median
(range) of 10.6 (1.4–146.0) months between the two EGDs. None were on gluten-free
diets at the time of either their first or second EGD. Those with persistent symptoms
were treated with standard PPIs or acetaminophen, and patients were diagnosed with
functional GI disorders, such as chronic abdominal pain or functional dyspepsia, after
follow-up EGD.
Based on logistic regression analysis, abnormal gastric histology was associated with
the presence of LD (odds ratio, 2.85; 95% CI, 2.05–3.97; P < 0.001), while the presence of abnormal esophageal histology was not statistically
significantly associated with LD (odds ratio, 1.46; 95% CI, 0.92–2.23; P = 0.081).
Discussion
Intraepithelial lymphocytes are a normal part of the small bowel's defense and surveillance
system and respond to injury. LD is a common pathologic finding in children and adolescents
undergoing EGD and is more common in those with CD, Crohn's disease, and H . pylori .[10 ] Reporting of abnormal pathologic changes, such as reactive gastritis, chronic gastritis,
and LD, is more commonly encountered. It is not clear if that is due to an increase
in GI procedures, a change in pathology reporting practice, or a true increase in
the prevalence of those histologic findings.
In our cohort, the rate of LD reporting increased over the study period. When we correlated
the rate of LD reporting with the number of yearly EGDs performed, we found that the
more EGDs performed, the more pathology reports with LD were encountered. We are not
aware of any substantial change in our endoscopy practice or pathology reporting that
could have affected these results. A previous study with a smaller cohort suggested
that LD might be higher in the presence of esophagitis and/or gastritis, but no statistical
significance was found when compared with controls with normal biopsy.[5 ] In our larger cohort, the LD rate was higher in both children and adolescents with
esophagitis and gastritis but was only statistically significant in those with abnormal
gastric biopsies, suggesting that LD can be a manifestation of a dispersed, nonspecific,
immunoinflammatory response, as well as a normal host response to different triggers
such as medications and infections. We believe that the presence of intraepithelial
lymphocytes in the duodenum are a normal part of the small bowel defense mechanism
and LD is more likely a transient reaction to pharmacologic, infectious, or environmental
triggers. Many of the children with LD were diagnosed with functional GI disorders.
It is not clear if the finding of LD should change the approach or the outcome of
those cases when compared with children with functional diagnoses and normal SBB histology.
In our cohort, PPI use was more common in the LD group. Many children and adolescents
with GI complaints are treated with PPI empirically before EGD.[11 ] It is not clear if the association between the rate of LD reporting and PPI use
is due to the presence of gastritis or due to the direct effects of PPIs on the small
bowel.[12 ] In both groups, functional GI diagnoses such as chronic abdominal pain and irritable
bowel syndrome were more common.
The major limitation of our study was the retrospective nature of the available data.
Because this was a single-center study and there was turnover among pathologists reviewing
the data, we believe larger prospective studies with follow-up endoscopic assessment
of LD cases are needed to clarify the association with PPI use and address long-term
implications. Ideally, the pathology findings would be reviewed by a consistent team
of pathologists to limit interobserver variation.
In conclusion, LD reporting is correlated with the number of EGDs performed and is
more likely seen in children and adolescents with abnormal gastric histology.