Patients
Twenty-two IBD patients (11 CD, 10 UC, and 1 indeterminate colitis) were identified
through a gastroenterology departmental IBD database as subjects who had an initial
inconclusive endoscopy performed from January 2002 to December 2008 prior to a later
confirmatory diagnosis of IBD. Patients who had a definitive diagnosis of IBD on the
initial endoscopy were excluded from the study. Approximately 650 patients were diagnosed
with IBD during this time period. The comparison group was comprised of 20 patients
with a diagnosis of functional abdominal pain identified through a gastroenterology
departmental database matched for age, gender, and study type (EGD, colonoscopy, or
both). Patients in this comparison group underwent endoscopy between January 2003
and December 2004 and were followed for a time period of at least five years, having
not developed IBD during this time. This study was approved by Children’s Mercy Pediatric
Institutional Review Board who waived the need for informed consent due to the retrospective
nature of the study and analysis of de-identified data and tissue samples.
Tissue specimens
Biopsy specimens obtained from 37 esophagogastroduodenoscopies (EGDs) and 31 colonoscopies
were studied. All biopsies analyzed were obtained during the initial non-diagnostic
endoscopy. The EGD specimens included multiple grasp biopsies of the distal esophagus,
gastric antrum, and the duodenum. Colonoscopy specimens were obtained from various
areas including terminal ileum, cecum, ascending colon, transverse colon, descending
colon, and rectosigmoid. Release of tissue samples was approved by the Institutional
Review Board and the Chairman of the Department of Pathology at Children’s Mercy Hospital.
Histopathological evaluation
Sections from the biopsy specimens which had been formalin-fixed and paraffin-embedded
in the usual fashion including staining with hematoxylin and eosin were reviewed by
a pediatric pathologist, blinded to diagnostic group. Specimens were evaluated for
the presence of gastritis, duodenitis, lymphoid hyperplasia, basal plasmacytosis,
eosinophilia, cryptitis, crypt abscess, and crypt distortion.
Mucosal (lamina propria) eosinophils of the stomach, duodenum, and rectosigmoid areas
were further quantified by the primary investigator (J.A.B.). Eosinophils were identified
by the characteristic prominent eosinophilic cytoplasmic granules and a typical bi-lobed
nucleus. Densities were determined by counting eosinophils in what appeared to be
the most involved area after scanning the entire specimen. Three consecutive high
power fields (each high power field approximately 0.15 square millimeters at x400
magnification) were evaluated with the peak count defined as the highest count of
the three fields and the mean count as the average of the three fields.
Immunohistochemistry
Additional slides were prepared from the gastric antrum and rectosigmoid biopsy specimens
for IHC staining for tumor necrosis factor-alpha (TNF-?) and matrix metalloproteinase-9
(MMP-9). Serial 4-?m thick sections were cut from paraffin embedded tissue blocks
for IHC staining.
IHC staining for TNF-? was performed on the Bond-MAX automated stainer (Leica Corporation,
Melborne, Australia). The sections were deparaffinized followed by heat induced antigen
retrieval using citrate buffer for 20Â min. Mouse monoclonal anti-human TNF-alpha (clone
P/T2:AbCam, Cambridge, MA, USA) was used as the primary antibody and applied at a
1:400 dilution. The Bond Polymer Refine Detection kit (Cat. No DS9800 Vision BioSystems
BondTM, Newcastle-upon-Tyne, UK) was used as the detection system and included peroxide
block, post primary enhancer, poly-HRP anti-mouse/rabbit IgG, DAB chromogen, and hematoxylin
counterstain. The stained sections were evaluated for the presence of TNF-? immunoreactive
cells and were graded as negative, focal, or diffuse by a pediatric pathologist. The
entire specimen was scanned for reactivity. Focal activity was defined as areas of
immunoreactive positive cells confined to 1–2 high power fields. Diffuse activity
was defined as greater than 2 high power fields of involvement by the immunoreactive
cells.
IHC was performed manually for MMP-9. The sections were deparraffinized and then rehydrated
in alcohol to tris-buffered saline. Endogenous peroxidase activity was blocked using
3Â % hydrogen peroxide followed by a protein block with 5Â % goat serum. Residual biotin
and avidin activity were quenched using avidin and biotin block, respectively. Affinity
purified polyclonal, mono specific rabbit anti-human MMP-9 (HPA001238; Sigma-Aldrich,
St. Louis, MO, USA) was used as the primary antibody and applied at a 1:200 dilution
overnight at ~4 °C. Labeled streptavidin-biotin (LSAB) was used for the detection
system with diaminobenzidine tetrahydrochloride (DAB) as the chromogen. Sections were
counterstained with hematoxylin. The stained sections were evaluated for the presence
of MMP-9 immunoreactive cells and graded as negative, focal or diffuse by the primary
investigator. The grading scheme was similar to that followed for TNF-?.
Statistical analysis
Comparisons between groups were made by chi-square and Fisher’s exact test as appropriate
for categorical variables. Mann–Whitney U test was used to compare continuous variables. A significance level of p??0.05 was established for all statistical comparisons. All calculations were performed
using the SPSS software package version 17 (SPSS Inc, Chicago, IL).