The present systematic review was conceived to collect the available data from the four major diagnostic tests, namely 75SeHCAT, C4, fecal BA assay and FGF19, proved to have clinical validity in identifying BAM in patients with chronic diarrhea due to various causes, mainly including functional (i.e., IBS-D, functional diarrhea) and inflammatory (Crohn’s disease) disorders. It should be noted, however, that data on BAM detection are still scarce and based mainly on small-size observational studies. Relevant calculation of the diagnostic accuracy of tests for BAM with true positive, true negative, false positive and false negative cases, which are essential parameters to assess sensitivity and specificity, have been reported in very few studies. Sensitivity and specificity were directly pooled from the included studies that reported diagnostic accuracy, and were shown as mean values in our systematic review. Nonetheless, because of the importance of this topic in daily practice and the possibility to improve the management of patients with chronic diarrhea, this systematic review aimed to demonstrate the actual diagnostic yield of few, though well established diagnostic tests for BAM, in patients with chronic diarrhea. Overall, data on 75SeHCAT, C4, fecal BA assay and FGF19 are encouraging although it should be stressed that any available diagnostic approach does not provide information as to whether BAs accumulate in the intestinal lumen as a result of true malabsorption or an increased secretion. Apart from this pathophysiological aspect, current available tests for BAM detection can provide objective information in patients with chronic diarrhea.
The final objective of studies involving tests for BAM should be the identification of high diagnostic accuracy, i.e. ‘biomarker(s)’, supporting the diagnosis of this underlying condition. The diagnostic test with the highest prevalence of positive results and highest diagnostic accuracy was 75SeHCAT retention, usually <10% as a best cut-off. This is an external scintigraphic approach with two phases of detection, usually at 3 h and 7 days after oral administration of the 370 Kbq 75Se gamma-emitting synthetic BA (homocholic acid taurine). 75SeHCAT resulted to have an average sensitivity and specificity of 87.32% of 93.2% [21][22][23]. Our data confirm previous analysis. In fact, another systematic review with meta-analysis (36 studies with a total of 5028 patients with functional bowel disorder and diarrhea) estimating biomarkers for BAM, Valentin et al. [24] showed that <10% 75SeHCAT (twenty-four studies) had the highest diagnostic yield (0.308 [0.247 to 0.377 CI]), followed by C4 (six studies) (0.171 [0.134 to 0.217 CI], fecal BAs at 48 hours (two studies) (0.255 [0.0.071 to 0.606 CI] and, finally, FGF19 (three studies) (0.248 [0.147 to 0.385 CI].
75SeHCAT did not show good accuracy in detecting BAM in patients with ileal resection. In the study of Borghede et. al. [25], the authors showed that 75SeHCAT yielded positive retention results indicative of BAM regardless the length of ileal resection. In fact, 39 out of 43 Crohn’s disease patients operated on ileal resection had a positive 75SeHCAT in a comparable percentage to that of Crohn’s disease but without ileal surgery. Hence, one can conclude that BAM-related chronic diarrhea in patients with Crohn’s ileitis occurs regardless ileal resection.
The most commonly used technique to determine BAM, 75SeHCAT, has some important limitations. First, this diagnostic approach shows several different thresholds of retention, which affect the results of the test. In a previous systematic review of 43 studies for a total of 1223 IBS patients, 75SeHCAT retention thresholds were as follows: 122 (10%) with <5%; 339 (27%) with <10% and 163 (13%) with <15% retention [26]. In our evaluation of the diagnostic accuracy we selected a 75SeHCAT retention rate cut-off <10%, which is most widely accepted for the diagnosis of BAM. Notably, a 75SeHCAT retention <10% correlates with a faster colonic transit time [27]. This retention threshold is believed to perform best in order to identify BAM in patients with chronic diarrhea. Secondly, it is known that 75SeHCAT is not available in many countries apart from few tertiary referral centers. The limited availability of 75SeHCAT prompted research to other tests for BAM, which may be of more practical use. Thirdly, 75SeHCAT requires a nuclear medicine department, highly expensive equipment, trained personnel, it is time consuming for the patient (as the test consists of two phases of scintigraphic recording at day 0 and after 7 days) and, last but not least, it has an unavoidable radiation risk. Finally, about 50% of the published studies do not compare 75SeHCAT to the new diagnostic tests.
A previous systematic review addressed the cost-effectiveness analysis of the relationship between 75SeHCAT and response to cholestyramine treatment in patients with BAM. Only three studies, based on a limited number of patients, covered this topic and the results showed that, in the long-term, there were no consistent cost-effective differences among the main options including trial of treatment with cholestyramine; no use of 75SeHCAT; use of 75SeHCAT with a 15% cut-off [28]. Therefore, the ultimate choice is upon physicians and depending on resources available. Despite significant heterogeneity, in a recent meta-analysis, aimed to determine the proportion of patients with BAM amongst 361 cases with watery diarrhea and previous cholecystectomy, the authors showed a pooled BA diarrhea rate of 70% (95% CI 56% - 82%) regardless a 10% or 15% 75SeHCAT cut-off. Five out of eight studies (166 patients) demonstrated that cholestyramine treatment achieved a pooled response rate of 79% (95% CI 63% to 91%), thus confirming the usefulness of this BA sequestrant in patients with post-cholecystectomy associated BAM [29].
The C4 and FGF19 are the two serological tests, which can be assessed by HPLC and ELISA, respectively, to evaluate patients with BAM related chronic diarrhea. C4 yielded very positive results showing a mean reported sensitivity and specificity of 85.2% and 71.1%, respectively. Although with a lower diagnostic accuracy compared to 75SeHCAT, C4 assay is increasingly tested in patients with BAM because it is relatively simple, is not invasive, and exhibits an appreciable diagnostic accuracy [21].
Another serum marker of BAM is given by FGF19. The pathophysiological concept supporting this assay is based on the evidence that elevated levels of FGF19 inversely correlate with C4 expression, a mechanisms leading to inhibition of BA synthesis. Thus, low serum levels of FGF19 causes a reduced inhibition of BAs thereby promoting BAM and related diarrhea [20] [27] [30]. Lenicek et al. [20] demonstrated an indirect correlation between C4 (increased) and FGF19 (decreased) levels, suggesting that when combined these two techniques enhance the diagnostic sensitivity of BAM. The overall FGF19 sensitivity and specificity were 63.75% and 72.25%, with the ROC curve analysis showing that a cut-off of FGF19 <145 pg/mL is a specific and sensitive marker for BAM [20] [27] [30]. FGF19 can be assessed by commercially available ELISA kits; it is an easy, non-invasive assay, and relatively not expensive diagnostic technique. However it should be emphasized that FGF19 levels show significant variations due to meal consumption. That is the reason why FGF19 should be assessed during fasting [31].
The evaluation fecal BA excretion is based on a high-fat (usually 100g / day for four days) diet followed by the collection of the whole amount of feces in 48 hours. HPLC is needed to evaluate BAs in fecal samples. Based on three studies, this test showed an overall sensitivity and specificity of 66.6% 79.3%, thus lower compared to 75SeHCAT and C4, but higher than FGF19. Routine use of this diagnostic approach is limited by its laborious organization (i.e. patients should adhere to a high-fat diet, stools need to be collected in 48 hours and, finally, BAs measured with HPLC, which is not readily available in any gastroenterological center) [19][32]. Moreover, the amount of 100 g of dietary fat has been shown not to be able to evoke an increased entero-hepatic circulation in any patient / subject, thus undermining their actual effectiveness as challenge dose. Finally, some authors suggested a single fecal sample can be measured for BA assay to unravel patients with BAM, however this approach is still far from clinical application and further study is needed to test its diagnostic accuracy [19][32][24].
Our systematic review has two major drawbacks, which we would like to acknowledge. First, the lack of real gold standard tests for BAM represents the main reason hampering comprehensive meta-analysis studies comparing the diagnostic accuracy of available tests for BAM. Heterogeneity was high for all diagnostic tests, indicating that included patients with different etiologies of diarrhea caused a high degree of variability across the majority of studies, thus lowering the specificity of the BAM tests. Large specificity variation is also due to lack of gold standard tests. Secondly, very few studies (only 25%) compared head-to-head the available diagnostic methods for BAM diagnosis. In this line, it is important to point out that methods to assess patients with an underlying BAM are currently used in few tertiary referral centers. Despite these limitations, herein we provided clinically useful data about the average diagnostic accuracy reported from each included study and the expected prevalence of the positive test that can help physicians to the appropriate diagnostic work-up for patients with BAM related watery diarrhea.