|Home | About | Journals | Submit | Contact Us | Français|
The small bowel follow-through (SBFT) is a non-invasive imaging modality for evaluating terminal ileum (TI) inflammation. The accuracy of this modality in pediatric patients is not well established.
We retrospectively determined the sensitivity and specificity of SBFT for detecting TI inflammation diagnosed on histology in 93 pediatric patients followed in a single institution.
The mean age at the first study was 12.9 years [range: 1.1-20.9 years]. 45% were female. Twenty-five patients (27%) had abnormal TI's on SBFT. Seventeen patients (18%) had TI inflammation diagnosed by biopsy. The sensitivity of SBFT was 59% and the specificity was 80% for detecting TI inflammation diagnosed on histology. Sensitivity and specificity did not change by demographic factors, final diagnoses, presenting symptoms, or laboratory parameters reflecting the presence of intestinal inflammation.
The sensitivity and specificity of SBFT in pediatric patients were poor and did not vary with demographic factors, final diagnoses, presenting symptoms, or laboratory parameters. Prospective longitudinal studies comparing various imaging modalities (SBFT, magnetic resonance (MR) enterography, and capsule endoscopy) are required to determine which is the most effective tool for evaluating pediatric patients for TI inflammation.
The terminal ileum (TI) is the most commonly affected area in Crohn's disease1,2. Intubation and biopsy of the terminal ileum during colonoscopy is considered the gold standard for diagnosing TI inflammation3-5. Less invasive imaging modalities, such as the small bowel follow-through (SBFT), are also utilized to evaluate for TI inflammation. However, the sensitivity of the SBFT for detecting inflammation of the TI in pediatric patients is not well established. We estimated the sensitivity and specificity of SBFT in diagnosing TI inflammation documented by histology of ileal biopsies in children and adolescents with no history of medical treatment for intestinal inflammation.
Patients less than age 21 years undergoing both initial SBFT and colonoscopy with intubation and biopsy of the TI between November, 1992 and January, 2008 at UCSF Children's Hospital at the University of California, San Francisco (UCSF) were retrospectively identified and included in the study. Patients were excluded if only one study was performed, if the SBFT and ileal biopsy were performed greater than 10 months apart, or if the patients had prior medical therapy for intestinal inflammation. Ninety three patients met study criteria. This study was approved by the UCSF Institutional Review Board.
SBFT and histology of the TI biopsy reports were reviewed and dichotomized into normal (absence of inflammatory changes) or abnormal (presence of inflammatory changes). Abnormal SBFT findings included TI thickening, ulcerations, strictures, or fistulas. Abnormal histology included features of acute or chronic inflammatory infiltrates.
Laboratory parameters (obtained ± 14 days of the date of SBFT or ileal biopsy, whichever came first) included white blood cell count, hematocrit, albumin, erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP) levels. Serum albumin levels, hematocrit, ESR, and CRP were dichotomized into normal or abnormal. White blood cell (WBC) count was dichotomized into elevated or not elevated.
Documented patient- and/or parent-reported presenting symptoms of abdominal pain, perianal fissure, anorexia, bloody stool, constipation, diarrhea, fatigue, fever, joint pain, nausea, poor growth, tenesmus, vomiting, and weight loss were dichotomized into present or absent.
Final diagnoses are reported as inflammatory bowel disease (IBD), Crohn's disease (CD), ulcerative colitis (UC), IBD-unspecified (IBD-U), or non-IBD.
Sensitivity and specificity of SBFT for diagnosing TI inflammation documented on ileal biopsies in pediatric patients were estimated, with exact 95% binomial confidence intervals, in the total population and by diagnoses, demographics, year of evaluation, presenting symptoms, and laboratory parameters. Fisher's exact test was used to compare sensitivity and specificity across complementary subgroups. We considered subgroups defined by final diagnosis (CD, UC, IBD-U, non-IBD), procedure date, patient age (<13 years vs ≥13 years), gender, race (white vs non-white), presence or absence of presenting symptoms (abdominal pain, diarrhea, bloody stool, weight loss, anorexia, vomiting, fatigue, nausea, fever, constipation, joint pain, tenesmus, poor growth, perianal fissure), and abnormal laboratory parameters (WBC, albumin, ESR, HCT, CRP).
Data were collected and stored in Microsoft Excel 2002 (Microsoft Corporation, Redmond, Washington). Statistical analysis was performed in STATA, Version 9 (STATA Corporation, College Station, TX).
Demographic information is shown in Table 1. The mean age at the first study was 12.9 years [range: 1.1-20.9 years].
43 patients were diagnosed with IBD. 50 patients carried a non-IBD final diagnosis. The median days between procedures was 9 days (25th percentile= 2 days; 75 percentile = 37 days). The days between procedures did not vary by the order of procedures (p=.09). Seventeen patients (18.3%) had inflammation diagnosed by TI biopsies. Of the twenty-five patients (26.9%) with abnormal TI on SBFT, ten (40.0%) had inflammation on TI biopsies. Sensitivity of SBFT was 59% (10/17, 95% CI= 33%-85%), while specificity was 80% (61/76, 95% CI= 70%-89%). Sensitivity and specificity were generally similar across subgroups defined by final diagnoses (Table 2), procedure date, gender, age at first study, race/ethnicity, presenting symptoms, and laboratory parameters. We found weak evidence (Fisher's exact p=.03) that specificity was higher in biopsy-negative patients with constipation (15/15, 100%, 95% CI =78%-100%) compared with biopsy-negative patients without constipation (46/61, 75%, 95% CI =63%-86%).
SBFT was performed in 41 patients prior to TI biopsy and in 51 patients after TI biopsy. One patient had a TI biopsy and SBFT on the same day. Sensitivity of the SBFT was 50.0% when it was done prior to TI biopsy and 66.7% when it was done after TI biopsy (p=.64). Specificity of SBFT was 66.7% when it was done prior to TI biopsy and 90.5% when the SBFT followed TI biopsy (p=.02).
The results presented in Table 2 were similar when patients who underwent the studies greater than one month apart were excluded, decreasing the total number of patients from 93 to 65.
Our study of 93 pediatric patients suggests that the sensitivity and specificity of SBFT for the detection of inflammation in the terminal ileum diagnosed by histology are poor in pediatric patients. Furthermore, sensitivity and specificity did not vary with gender, age at first study, race/ethnicity, final diagnoses, or laboratory parameters reflecting the presence of intestinal inflammation. Specificity may be somewhat higher in patients presenting with constipation, which may reflect that constipation is a less common presenting symptom of inflammatory bowel disease. Sensitivity or specificity did not vary with other presenting symptoms.
Sensitivity of SBFT for the detection of TI inflammation diagnosed by histology did not vary by order of procedures. Specificity improved when SBFT followed TI biopsy. This finding most likely represents a selection bias for proceeding with a TI biopsy when SBFT is done first. If the SBFT is positive, a clinician may be more inclined to proceed with TI biopsy; this would select for more patients with a positive SBFT to be included in this study. A patient with a negative SBFT would less likely be captured in this study if it was done prior to the TI biopsy because the clinician may be less likely to proceed with a TI biopsy if the SBFT is negative. This selection can potentially decrease specificity of the SBFT if it is done prior to TI biopsy.
Previous studies have shown widely varying results regarding the accuracy of barium radiography in detecting terminal ileal inflammation3, 5-15. Byrne et al. retrospectively studied 46 patients (age 16-78 years) who underwent SBFT and ileocolonoscopy within two weeks of one another and found that SBFT failed to detect 41% of cases of TI inflammation diagnosed by ileoscopy with biopsy5. From a cohort of 121 adults with known Crohn's disease, Halligan et al. retrospectively studied 23 patients who underwent both SBFT and ileoscopy and reported that SBFT had a 100% sensitivity and 90% specificity for TI disease diagnosed by ileoscopy15.
Lipson, et al. retrospectively studied 46 pediatric patients with suspected chronic IBD, in which patients underwent ileocolonoscopy and SBFT within a 14 day period3. There was agreement between SBFT and ileal histology in 80% of cases. Authors reported that the sensitivity and specificity of SBFT were 90% and 96%, respectively, for diagnosing Crohn's disease of the terminal ileum. In contrast, our study of pediatric patients demonstrates that the sensitivity and specificity of SBFT for the detection of inflammation diagnosed on TI histology in pediatric patients are poor. Furthermore, sensitivity and specificity did not vary with demographic factors, final diagnosis, or laboratory parameters reflecting the presence of intestinal inflammation. Although specificity improved in patients presenting with constipation, sensitivity and specificity did not change with other presenting symptoms. Furthermore, we examined sensitivity and specificity by combinations of symptoms (data not shown), e.g. abdominal pain, weight loss, and diarrhea, and did not identify any improvement in sensitivity or specificity.
SBFT is only one of a growing number of radiographic studies that are used to evaluate the TI. The capsule endoscopy has shown encouraging results in adults10-14, with studies suggesting higher sensitivity of capsule endoscopy compared with SBFT11,14. One study reported a sensitivity of 89.6% and a specificity of 100% of capsule endoscopy for detecting small bowel inflammation suggestive of Crohn's disease10.
Studies conducted in pediatric patients also suggest that capsule endoscopy may be superior to SBFT in evaluating small bowel inflammation16-18. Furthermore, Cohen et al. reported capsule endoscopy may lead to reclassification of IBD from UC or indeterminate colitis to definitive Crohn's disease in a cohort of patients diagnosed with IBD between ages 2 and 18 19. Considering the risks of early onset inflammatory bowel disease 20, using techniques involving less radiation (e.g., capsule endoscopy and MR enterography) is important to monitor evolution of disease type and activity in young children.
Although CT evaluation of the small bowel is being studied in adult patients12,21-24 the utility of CT enterography in pediatric patients is not well established. The amount of radiation exposure is not insignificant25, which raises concern for using this imaging modality in the pediatric population. Magnetic resonance imaging is another imaging modality that may be used to evaluate small bowel disease24, 26-28. Laghi et al. performed a prospective study of 75 pediatric patients (age 8-17 years) with known Crohn's disease, in which patients underwent contrast enhanced MRI of the bowels, and found that MRI had a sensitivity of 84% and specificity of 100% for detecting erosive ileitis, as documented by ileoscopy27. This modality is appealing because of the avoidance of ionizing radiation.
The major limitation of our study is its retrospective study design. The timing and order of SBFT and ileal biopsy were not standardized, but obtained at the discretion of the pediatric gastroenterologist based on clinical indication. The result of one test would naturally influence the provider's decision to obtain the second test. Furthermore, the severity of a patient's symptoms or progression of symptoms may influence the timing and order of the tests. A further limitation of our study includes its relatively small size, which limits the precision of our sensitivity and specificity estimates, as well as power to detect subgroup effects29.
Our retrospective study shows poor sensitivity and specificity of SBFT to detect inflammation of the terminal ileum in pediatric patients. Large prospective longitudinal studies comparing the utility of SBFT, MR enterography, and capsule endoscopy, with TI biopsy are required in the pediatric population to determine which of these modalities is the most effective tool for diagnosing TI inflammation in children and adolescents.
This project was supported in part by NIH grant DK077734 (ng), NIH grant DK060617 (mh), Children's Digestive Health and Nutrition Foundation/Crohn's and Colitis Foundation of America (CCFA) Award for New Investigators (ng), CCFA Career Development Award (ng), and NIH/NCRR UCSF-CTSI Grant Number UL1 RR024131.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.