Sensitivity and specificity of ultrasonography for the diagnosis of malabsorption syndrome in the horse

Malabsorption syndrome in horses is frequently due to infiltrative bowel disease (IBD): the infiltration of the intestinal mucosa impairs the absorptive function with the onset of chronic wasting syndrome. Histology represents the gold standard diagnostic procedure and it can be performed by surgical biopsy of the small intestine, rectal biopsy or endoscopic duodenal biopsy. Biopsy techniques present many practical disadvantages in equine practice, therefore finding more practical complementary diagnostic tests to be used for screening purposes is advisable. Abdominal ultrasonography is both easy to perform and capable of detecting intestinal wall thickening that is a pathological feature of infiltrative disorders(1). The aim of the present study was to assess the sensitivity and specificity of ultrasonographic measurement of small intestinal wall thickness (UMSIT) to evaluate IBD, using the histology of rectal biopsies as the reference standard(2). 55 horses with a history of chronic wasting were retrospectively selected. In all patients, repeatability of UMSIT was statistically evaluated. All the UMSIT were analyzed and an “individual scoring system” defining the distribution of thickened intestinal loops (>3 mm) in each horse was elaborated to subdivide the population into two groups: general group (GG) including all horses and subgroup A (SA) including only those horses which had a “diffuse” thickening of both duodenum and jejunum. Eventually, sensitivity (Se), specificity (Sp) and ROC curve were statistically calculated both for GG and SA. Since no statistical differences were observed among measurements, UMSIT resulted repeatable (p < 0.05). For GG, the best UMSIT cut off was 5.7 mm with a Sp of 87.5% and a Se of 36.8%. For SA, the best UMSIT cut off was 5 mm with a Sp of 100% and a Se of 50%. The repeatability of UMSIT, previously demonstrated under experimental conditions(3), was confirmed in the present study in a clinical population. Specificity of UMSIT was high but linked to inclusion criteria. Sensitivity of UMSIT was low due to variable ultrasonographic explorability of the small intestine and to the different extension of the thickening: sensitivity was higher for SA because only horses with “diffuse” lesions were included. Study limitations include the use of rectal biopsy as the reference standard (whose selection was due to clinical practical reasons) and the absence of a standardized protocol for intestinal ultrasonographic evaluation (due to the retrospective nature of the study). However, this study demonstrates the UMSIT promising diagnostic potential, highlighting the necessity for more detailed perspective studies in the future