Objective: The aim of this study was to assess whether power Doppler ultrasound (PDU) can serve as a reliable replacement for endoscopy in follow-up assessment of disease activity in children with Crohn disease.
Methods: Nineteen children (13 boys), median age 14.8 (5.4–15.8) years, with macroscopically diagnosed Crohn disease were included in the study. Clinical parameters, histological evaluation, and graded PDU assessments were undertaken in all patients at diagnosis and following their initial treatment. Discriminant analysis was used to build predictive models from the PDU data for the histological evaluation.
Results: The median Pediatric Crohn's Disease Activity Index (PCDAI) was 31.5 (15.5–42.0) at diagnosis. All clinical and ultrasonographic parameters and the histological evaluation showed an improvement between pre- versus posttreatment results; 1-way analysis of variance showed a significant difference because of treatment (P < 0.005) for all variables apart from the superior mesenteric artery flow (SMA); paired sample t tests indicated that these differences were statistically significant (P < 0.001), with the exception of SMA (P = 0.178). There was a statistically significant correlation (P < 0.001) between the platelet count and the bowel wall stratification (STRAT). Significant correlation was also observed between the histology findings and the mean mucosal, transmural, and segmental flow (MFL) and STRAT and between platelets and both MFL and small bowel thickening (SBT) (P < 0.01 in all cases). There was a statistically significant correlation (P < 0.05) between C-reactive protein and MFL and between histology and SBT, MFL, and STRAT. Discriminant analysis using discriminating factors SBT, STRAT, and MFL could assign 84.6% of cases to the correct classification of “no/mild inflammation” or “medium/severe inflammation.”
Conclusions: Results obtained using power Doppler ultrasonography and endoscopy showed statistically significant correlations. Power Doppler sonography, in the hands of an experienced examiner, can be used for follow-up assessment of disease activity in children with Crohn disease.
*Centre for Paediatric Gastroenterology, Sheffield Children's NHS Trust, Sheffield
†Centre for Paediatric Gastroenterology, Royal Free Hospital, London
‡Section of Paediatrics, Division of Medicine, Imperial College Faculty of Medicine, Chelsea & Westminster Campus, London, UK.
Address correspondence and reprint requests to Dr Mike Thomson, Centre for Paediatric Gastroenterology, Sheffield Children's NHS Trust, Western Bank, Sheffield S10 2TH, UK (e-mail: firstname.lastname@example.org).
Received 6 September, 2009
Accepted 23 August, 2010
The authors report no conflicts of interest.
Crohn disease is a chronic inflammatory disease of the bowel that follows a pattern of relapse and remission characterised by asymmetric, focal, transmural, and sometimes granulomatous inflammation. Assessment of inflammatory activity in children with Crohn disease is based mainly on clinical parameters, usually requiring the measurement of many clinical parameters to define the Pediatric Crohn's Disease Activity Index (PCDAI); however, some of the criteria used to evaluate the index in the PCDAI are subjective, affecting the reliability of the PCDAI and thus limiting its clinical usefulness (1). Other laboratory indices that are used to estimate disease activity, such as the estimated sedimentation rate (ESR), C-reactive protein (CRP), and platelet count, demonstrate correlations with Crohn disease activity, and are nonspecific and therefore more useful in combination rather than as individual indices (2,3).
Incomplete and/or suboptimal treatment could have important clinical consequences for the paediatric patient. Morbidity, growth impairment, and numerous potential complications of structuring and fistulae need to be avoided. Overzealous treatment with immunosuppressants is clearly undesirable, particularly in children. Repeated endoscopy is considered too invasive for regular monitoring of the progress of treatment or diagnosing each disease flare-up. Because of the lack of a single noninvasive and absolute reference method to assess disease activity, multiple parameters are often used.
Radiological imaging techniques that have proven useful in assessing activity in Crohn disease include barium meal and follow-through, scintigraphy, magnetic resonance imaging, and computerised tomography scans (4). Ultrasonography, in the last 15 years, has been shown to be a useful adjunct in the investigation of Crohn disease. Indeed, Doppler ultrasound has advanced considerably, making it possible to reveal neovascularization and inflammatory hyperaemia. This has proven to be valuable for the assessment of disease activity in inflammatory bowel disease. Avoiding exposure of patients to ionising radiation, especially children, is an obvious advantage over other radiological techniques. Bowel wall thickness, stratification, reduced compressibility, loss of peristalsis, and haustrations assessed by ultrasonography have shown correlations with the Crohn's Disease Activity Index (CDAI), other clinical indices, and histological findings, particularly in moderate to severe mucosal inflammation (5–7).
In the last few years, power Doppler ultrasonography has been shown to be a highly sensitive method for evaluating the flow in vessels that is poorly imaged by conventional colour Doppler techniques (8). Additional information has also been obtained with power Doppler ultrasound (PDU), for example, assessment of superior mesenteric artery flow or mucosal and transmural perfusion, both of which have been shown to be significantly increased in actively inflamed bowel segments, including inflammatory bowel disease (9). This has been illustrated by cross-sectional studies comparing Doppler ultrasound findings in disease states with clinical or endoscopic indices. Recently, the use of PDU has been extended to the identification of complications of Crohn disease such as internal fistulas and mucosal stricturing (10).
At the Centre for Paediatric Gastroenterology at the Royal Free Hospital (London, UK), the practice involves routine repeat of ileocolonoscopy to ensure mucosal healing following the first course of treatment. This is advantageous in correlating findings in individual patients before and after treatment. Change in disease activity is assessed clinically, histologically, and radiologically. The purpose of this study was to use PDU to evaluate disease activity in Crohn disease and to assess whether this noninvasive tool can reliably replace endoscopy for the follow-up assessment of disease activity.
A prospective study of 20 children with a new diagnosis of Crohn disease was undertaken at the Centre for Paediatric Gastroenterology at the Royal Free Hospital; 19 of the 20 children were recruited to take part. A clinical suspicion of Crohn disease was confirmed histologically. After diagnosis, an 8- to 10-week treatment period ensued followed by a clinical reassessment that included a repeat ileocolonoscopy to confirm mucosal and histological remission. Children were enrolled in the study during a 24-month period. All children with newly diagnosed Crohn disease were included for whom parental consent was obtained. Ethical approval was obtained from the joint trust and medical school ethics committee.
At diagnosis, the PCDAI was calculated and, where possible, a barium meal and follow-through were performed. Other parameters of disease activity tabulated included serological inflammatory markers such as CRP, ESR, and platelets. A blinded segmental histological assessment of disease severity was made retrospectively by 1 histopathologist. All of the patients underwent graded compression and PDU within 2 to 3 weeks of diagnosis. Following the course of treatment, all of the children were reassessed for disease activity using serological inflammatory markers, ileocolonoscopy for segmental histological assessment, and repeat graded compression and power Doppler ultrasonography; the PCDAI was not recalculated at this time. Patients who did not undergo both pre- and posttreatment sonographic examinations were not included in the study.
All of the patients were treated with exclusive Modulen IBD (casein based) enteral nutrition (Nestlé Nutrition, Vevey, Switzerland). Three of the 19 patients required steroids because of poor/partial response to the exclusive elemental diet.
The ultrasonography was performed by an experienced paediatric radiologist (L.B.) who was blinded to clinical and histological evaluation of disease. To confirm validity, the graded compression images were stored on hard copy and reviewed retrospectively by another blinded radiologist. The following measurements were made: small bowel compressibility, bowel wall thickness, indicating the measurements and site of thickening, bowel wall stratification, peristaltic activity, and the presence of loop dilatation. The presence of free intraperitoneal fluid, lymph nodes, echogenic fat, and bladder contour were also measured. Using the power Doppler, superior mesenteric artery peak systolic flow was measured. Measurements were taken 2 cm distal to its origin from the aorta. Segmental mucosal, transmural, and serosal flow were also measured, and a collective overall assessment for flow was used in the analysis. An ATL 5000 scanner (National Ultrasound, Duluth, GA) was used. Graded compression scans were carried out using a 7.5-MHz linear array probe to assess bowel wall thickness and stratification. To measure Superior mesenteric artery (SMA) flow, a 3.5-MHz curvilinear probe was used. The peak flow in the vessel was measured 2 cm from the origin of the SMA from the aorta for the maximum sample volume. The angle of visualisation was oriented along the line of the vessel. This was multiplied by the cross-sectional diameter of the SMA to give the flow in millilitres per second. SPSS for Windows version 14.0 (SPSS Inc, Chicago, IL) was used for statistical analysis of the data.
Pre- and posttreatment values for platelet count, CRP, histology, small bowel thickening (SBT), mean flow level (MFL), SMA, and stratification (STRAT) were compared using Student paired samples t test, to identify any significant changes after treatment. The SMA values were also tested for differences using 1-way analysis of variance (ANOVA) statistics. The strength of the relations between the various parameters was measured using the Pearson correlation for paired samples. Discriminant analysis was used to build a predictive model of the level of disease severity assessed histologically, using the independent sonographic parameters SBT, MFL, STRAT, and SMA and calculating the observed characteristics for each case. The procedure generated a discriminant function (or, for more than 2 groups, a set of discriminant functions) based on linear combinations of the predictor variables that provided the best discrimination between the groups. The functions were generated from a randomly selected sample of cases for which group membership was known and then tested on the remaining cases. The percentage of correctly assigned cases indicated how effectively the predictor variables obtained by PDU assigned cases to the correct histologically assessed grouping.
Nineteen patients were recruited for analysis in the study. Clinical parameters, histological evaluation, and PDU data from the 19 children (6 girls) who took part in this study are shown in Table 1. The initial diagnosis of Crohn disease was confirmed macroscopically and histologically before the patient entered the study. The mean PCDAI at diagnosis was 31.5 (range 5.4–15.8); these values were coded into 4 categories (0 = not measured: 1 = 20–30: 2 = >30–40: 3 = >40), and all of the comparisons were based on these categories. The distribution of macroscopically inflamed mucosa consisted of 10 ileal, 6 colonic, and 4 ileocolonic bowel wall segments. The histology confirmed disease activity in all macroscopically identified diseased segments. Barium studies were undertaken in only 9 of 20 patients at diagnosis; proximal small bowel disease was found in 1 patient who also had terminal ileal disease.
Paired-sample t tests for pre- and posttreatment clinical and sonographic parameters are tabulated in Table 1. All of the parameters showed a significant difference in pre- versus posttreatment results, P < 0.001, except for the SMA flow, which was not significant. Pre- and postmedian values for PCDAI, histology, SBT, STRAT, and MFL are shown in Figure 1.
When the SMA values were tested for differences between the means using 1-way ANOVA, there was no statistically significant difference between the pre- and posttreatment results, (P > 0.2). This may be because the posttreatment parameters were skewed, because not all children responded to the same degree, and the nonresponders were still included in the analysis. This measurement proved technically difficult to accurately determine in 2 patients. Eighteen of the 19 patients who were recruited in the study showed clinical improvement and had achieved clinical remission following the course of initial treatment. One patient who required surgical intervention, having failed medical therapy, did not provide sufficient data for analysis and was therefore excluded from the study.
There were highly significant correlations (P < 0.001), between the histology findings and MFL and between platelets and both SBT and MFL. There were significant correlations (P < 0.01) between PCDAI and both SBT and MFL, and between histology and SBT.
The number of cases was insufficient to use all of the classifications of the histology results, so cases were divided into 2 groups as follows: “no/mild inflammation” or “moderate/severe inflammation.” Discriminant analysis indicated that a combination of SBT, STRAT, MFL, and SMA would correctly classify 71.4% of cases to the correct degree of inflammation, but if the variable SMA was excluded, the number correctly classified rose to 84.6% of the cases.
Historically, vascular lesions and microvascular changes, such as granulomatous vasculitis, neovascularisation, and dilatation of arteries and veins, have been recognised as well known features in the pathogenesis of Crohn disease. In 1970, the vascularity in the gut wall of subjects with severe Crohn disease was found to be increased, whereas there were less marked vascular changes in subjects with moderate or mild disease (11). Histological examinations have established a changed vascular architecture in chronically inflamed bowel segments in both Crohn disease and ulcerative colitis (12–14). The extent and pattern of bowel wall changes visualised with sonography in Crohn disease correspond with these histologically described changes (15–17).
The use of Doppler sonography has become widespread in recent years as a useful noninvasive means of evaluating splanchnic haemodynamics, particularly inflammatory bowel disease. Various measurement parameters have been used by a number of investigators to assess the activity of Crohn disease; however, when comparing active versus quiescent disease, conflicting results have been reported (18–21). The use of greyscale sonography in the primary investigation of adult patients with suspected Crohn disease or recurrent Crohn disease was reported to have up to 87% sensitivity and 91% specificity for detection of disease (22). Using colour Doppler sonography in a small group of children, it was first suggested that mucosal or transmural hypervascularity could be seen in several inflammatory bowel processes (23). Most authors have concentrated on recurrent disease with sonography in children (24,25). It was reported that greyscale and colour Doppler sonography aided in the differentiation of causes of bowel wall thickening in children, but it was most useful when considered with the clinical situation, particularly the age of the patient and the location of the disease (6). With advances in technology and increasing experience with Doppler sonography, analysis has evolved to a more reliable technique for the detection of inflamed bowel wall using PDU (Fig. 2) (25,26). Greyscale and colour sonography or PDU can show changes in disease activity in children and young adults undergoing treatment for inflammatory bowel disease (27).
Most paediatric power Doppler studies in inflammatory bowel disease thus far have been limited by the lack of an independent gold standard to confirm bowel inflammation. In the present study, for the first time in children, we have used histological changes in the architecture of the bowel wall to correlate with the sonographic vascular changes occurring during active and quiescent Crohn disease. Our results show that in Crohn disease, PDU can be a useful and reliable tool for the evaluation of bowel wall thickness and stratification as well as hyperaemia of the inflamed gut wall in the assessment and follow-up of posttreatment patients. We have demonstrated statistically significant changes in bowel wall thickness and stratification, hyperaemia in quiescent disease, which is comparable to histological changes, and other clinical parameters of disease activity.
It must be acknowledged that the present study has certain limitations. In the series of children in whom serial PDU was undertaken during ongoing treatment for inflammatory bowel disease, a decrease was seen in both bowel wall thickness and Doppler grade as early as 6 days after initial sonography (27). The timing of the ultrasound after the endoscopy was variable in the present study and therefore could have feasibly affected the ultrasound findings, particularly because the comparative endoscopic and ultrasound data were used in the analysis of the results. Most patients would have had the ultrasound performed within 1 week of undergoing colonoscopy; however, some children would have already completed 2 to 3 weeks of treatment at the time of the initial ultrasound, which could therefore affect the sensitivity of the ultrasound study.
All of the ultrasound studies were performed by the same radiologist who was blinded to clinical and histological evaluation of disease. For practical reasons, the study was not independently performed by the second blinded radiologist, but instead images were stored on hard copy and reviewed retrospectively. This could have introduced further bias.
Another problem encountered was that in most patients, evaluation of the peak flow of the SMA proved to be technically difficult. In this instance, it was found that the overall thickness of the diseased loops of bowel decreased with clinical improvement of the patient. The examiner (L.B.) commented that the difficulty in evaluation of SMA flow in children resulted partly from the small vessel size and also from the presence of increased movement artefact caused by the problems some children had in holding their breath during the study.
There was also the concern regarding possibility that the entire extent of the bowel disease may not be identified; however, it may be feasible to infer that if a response to treatment is seen in 1 segment, then a similar response could be anticipated in other areas of involvement.
Contrast-enhanced PDU has a high sensitivity and specificity in the detection and evaluation of inflammatory abdominal masses. The improved spatial resolution and sensitivity of this technique may allow for more accurate assessment of disease progression and response to treatment in children with inflammatory bowel disease. The absence of radiation dose to the patient allows multiple and serial examinations to be undertaken, thereby facilitating assessment of treatment response. Our findings suggest that PDU in the hands of an experienced examiner represents a new tool for the assessment and follow-up of Crohn disease, which can facilitate the often-difficult management decisions in paediatric patients. The information yielded from PDU grading of the bowel wall correlates with the most common clinical parameters of disease activity and histology in children with Crohn disease. This study further illustrates that sonography shows therapy-induced changes in the bowel disease activity in children with inflammatory bowel disease. Further development with contrast enhancers and sonographic techniques, will increase the diagnostic strength of transabdominal intestinal wall sonography, will permit quantitative evaluation of bowel wall vascularisation, and may have a future role in directing therapy.
1. Otley A, Loonen H, Parekh N, et al. Assessing activity of pediatric Crohn's disease: which index to use? Gastroenterology
2. Hyams JS, Ferry GD, Mandel FS, et al. Development and validation of a Pediatric Crohn's Disease Activity Index. J Pediatr Gastroenterol Nutr
3. Beattie RM, Croft NM, Fell JM, et al. Inflammatory bowel disease. Arch Dis Child
4. Rioux M, Gagnon J. Imaging modalities in the puzzling world of inflammatory bowel disease. Abdom Imaging
5. Puylaert JBCM. Mesenteric adenitis and acute terminal ileitis: ultrasound evaluation using graded compression. Radiology
6. Siegel MJ, Friedland JA, Hildebolt CF. Bowel thickening in children: differentiation with ultrasonography. Radiology
7. Haber HP, Busch A, Ziebach R, et al. Bowel thickness measured by ultrasound as a marker of Crohn's disease activity in children. Lancet
8. Kratzer W, von Tirpitz C, Mason R, et al. Contrast-enhanced power Doppler sonography of the intestinal wall in the differentiation of hypervascularised and hypovascularised intestinal obstructions in patients with Crohn's disease. J Ultrasound Med
9. van Oosten JA, Wasser MN, van Hogezand RA, et al. Activity of Crohn disease assessed by measurement of superior mesenteric artery flow with Doppler ultrasound. Radiology
10. Byrne MF, Farrell MA, Abass S, et al. Assessment of Crohn's disease activity by Doppler sonography of the superior mesenteric artery, clinical evaluation and the Crohn's Disease Activity Index: a prospective study. Clin Radiol
11. Brahme F, Lindström C. A comparative radiographic and pathological study of intestinal vaso-architecture in Crohn's disease and in ulcerative colitis. Gut
12. Schafer K, Teidjen KU, Bohm E, et al. Angioarchetecture of the ileum and colon in Crohn's disease. Zentralbl Chir
13. Lim JH, Ko YT, Lee DH, et al. Sonography of inflammatory disease: findings and value in differential diagnosis. AJR Am J Roentgenol
14. Kruschewski M, Bushc C, Dorner A, et al. Angio-archetecture of the colon in Crohn's disease and ulcerative colitis. Light microscopy and scanning electron microscopy studies with reference to the morphology of the healthy large intestine. Langenbecks Arch Surg
15. Kimmey MB, Martin RW, Haggitt RC, et al. Histologic correlates of gastriointestinal ultrasound images. Gastroenterology
16. Kimmey MB, Wang KY, Haggitt RC, et al. Diagnosis of inflammatory bowel disease with ultrasound: an in vivo study. Invest Radiol
17. Hata J, Haruma K, Yamanaka H, et al. Ultrasonographic evaluation of inflammatory bowel disease: comparison of in vivo and in vitro studies. Abdom Imaging
18. Shankey EA, Dhillon AP, Anthony A, et al. Early mucosal changes in Crohn's disease. Gut
19. Faure C, Belarbi N, Mougenot JF, et al. Ultrasonographic assessment of inflammatory bowel disease in children: comparison with ileocolonoscopy. J Pediatr
20. Pera A, Cammarota T, Comino E, et al. Ultrasonography in the detection of Crohn's disease and in the diffrential diagnosis of inflammatory bowel disease. Digestion
21. Khaw KT, Yeoman LJ, Saverymuttu SH, et al. Ultrasonic patterns in inflammatory bowel disease. Clin Radiol
22. Sheridan MB, Nicholson DA, Martin DF. Transabdominal ultrasound as the primary investigation in patients with suspected Crohn's disease or recurrence: a prospective study. Clin Radiol
23. Quillan SP, Seigal MJ. Gastrointestinal inflammation in children: color Doppler ultrasonography. J Ultrasound Med
24. Dinkel E, Dittrich M, Peters H, et al. Realtime ultrasound in Crohn's disease: characteristic features and clinical implications. Pediatr Radiol
25. Babcock DS, Patriquin H, La Fortune M, et al. Power Doppler sonography: basic principles and clinical applications in children. Pediatr Radiol
26. Clautice-Engle T, Jeffrey RB Jr, Li KC, et al. Power Doppler imaging of focal lesions of the gastrointestinal tract: comparison with conventional color Doppler imaging. J Ultrasound Med
27. Ruess L, Nussbaum Blask AR, Kerzner B, et al. Inflammatory bowel disease in children and young adults. Correlation of sonographic and clinical parameters. AJR Am J Roentgenol