20423306, 2022, 5, Downloaded from https://beva.onlinelibrary.wiley.com/doi/10.1111/evj.13538 by Readcube (Labtiva Inc.), Wiley Online Library on [15/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Received: 7 May 2021 | Accepted: 18 October 2021 DOI: 10.1111/evj.13538 ORIGINAL ARTICLE Peritoneal bile acids concentration in adult horses with hepatic and gastrointestinal disorders Maria L. Rodríguez-­Pozo1 | Lara Armengou1 | Judit Viu1 | Jose Ríos3 | Eduard Jose-­Cunilleras1,2 1 Servei de Medicina Interna Equina, Fundació Hospital Clínic Veterinari, Summary Universitat Autònoma de Barcelona, Spain Background: Peritoneal bile acids concentration (PBAC) has not been previously re- 2 Departament de Medicina i Cirurgia ported in horses. A case of liver lobe torsion in which increased PBAC was detected Animals, Universitat Autònoma de Barcelona, Spain prompted us to study PBAC in horses. Objectives: (a) To determine a reference range of PBAC in horses; (b) to compare 3 Laboratory of Biostatistics and Epidemiology, Statistics and Methodology Support Unit, IDIBAPS, Hospital Clinic, PBAC from horses with either hepatic or gastrointestinal disease and healthy horses Barcelona, Spain and (c) to assess the prognostic and diagnostic values of PBAC. Correspondence Study design: Prospective case-­control. E. Jose-­Cunilleras, Departament de Methods: Prospective observational clinical study. Bile acids concentrations were Medicina i Cirurgia Animals, Universitat Autònoma de Barcelona, 08193 Bellaterra, measured in both plasma and peritoneal fluid in selected clinical patients with hepatic Barcelona, Spain. or gastrointestinal disease (n = 108) and healthy horses (n = 11). Sixty-­eight of 108 Email: eduard.jose.cunilleras@uab.cat patients survived to hospital discharge, and the remaining 40 were nonsurvivors. Present address Additionally, other haematological and biochemistry analyses were performed. Maria L. Rodríguez-­Pozo, Hospital Veterinario San Vicente, Alicante, Spain Results: Sick horses were classified according to diagnosis into hepatic (n = 13), gas- Jose Ríos, Hospital Veterinario Sierra de trointestinal (GI) obstructive (n = 48) and GI ischaemic-­inflammatory (n = 47) groups. Madrid, Madrid, Spain The hepatic group had significantly higher PBAC (6.8 [2.3-­9.4]; median [IQR]) than the control (1.0 [0.6-­1.5]) and GI obstructive groups (1.2 [0.8-­1.7] µmol/L; P < .001). Moreover, the GI ischaemic-­inflammatory group (3.3 [1.4-­5.5]) also had significantly higher values than the control and GI obstructive groups (P < .001). Regarding out- come, the nonsurvivor group (n = 40) had significantly higher median PBAC value than the survivor group (n = 68, 4.1 [1.6-­6.5] vs 1.3 [0.8-­3]; P < .001). Main limitations: A higher number of horses with abdominal disease is required to confirm the clinical significance of these findings. Conclusions: PBAC may have a role in the diagnosis of hepatic and gastrointestinal disease and as a prognostic tool in horses with abdominal pain. KEYWORDS colic, horse, liver, peritoneal fluid This is an open access article under the terms of the Creat​ive Commo​ns Attri​butio​n-­NonCo​mmerc​ial-­NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-­commercial and no modifications or adaptations are made. © 2021 The Authors. Equine Veterinary Journal published by John Wiley & Sons Ltd on behalf of EVJ Ltd. 914 |  wileyonlinelibrary.com/journal/evj Equine Vet J. 2022;54:914–921. | 20423306, 2022, 5, Downloaded from https://beva.onlinelibrary.wiley.com/doi/10.1111/evj.13538 by Readcube (Labtiva Inc.), Wiley Online Library on [15/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License RODRÍGUEZ-­POZO et al. 915 1 | I NTRO D U C TI O N the diagnostic workup were excluded. Additionally, a group of 11 healthy horses from a teaching herd was enrolled as a control group. Bile acids are amphophilic molecules produced by the liver that serve All procedures were conducted in compliance with the guidelines as detergents and make up to 90% of the organic portion of bile. These established by the institution animal experimentation ethics committee. acids are present in high concentration in the portal circulation, are extracted by the liver with a high level of efficiency and are normally transported away from the liver through the biliary circulation. The en- 2.2 | Classification terohepatic circulation normally removes more than 90% of bile acids.1 Concentrations of total plasma bile acids between 5 and 28 µmol/L Clinical cases were classified according to their final diagnosis into 2 have been reported in normal adult horses. Increased plasma bile 3 different groups: horses with liver disease (hepatic group), horses acids concentration are highly specific for the presence of liver insuf- with colic due to simple intestinal obstruction or gastric disease (GI ficiency in horses but is not specific for the type of liver disease.1,3,4 obstructive group) and horses with colic due to ischaemic/inflamma- Certain metabolites and enzymes present in the equine perito- tory disorders (GI ischaemic-­inflammatory group). neal fluid have been evaluated as indicators of ischaemia, inflam- The hepatic group included horses with a diagnosis of liver disease mation or other abdominal disorders.5-­9 Some of these parameters based on at least one of the following criteria: (a) plasma hepatic pro- include lactate,7 pancreatic enzymes (amilase, lipase)10 and hepatic file suggestive of hepatopathy or hepatic insufficiency, or (b) evidence enzymes (alkaline phosphatase).11 In addition, peritoneal:serum of significant hepatopathy on histopathologic examination of samples creatinine ratio is very useful to confirm uroabdomen in neonatal obtained by ultrasound guided percutaneous biopsy or (c) liver sam- foals.12 However, peritoneal bile acids concentration (PBAC) has not ples obtained at necropsy. Horses were not included in the hepatic been reported previously in horses. group solely based on post-­mortem findings. Hepatic profile included Bile peritonitis has been described in human medicine after trau- plasma bile acids, total and direct bilirubin concentrations, gamma-­ matic gallbladder or bile duct rupture by either a penetrating wound glutamyltransferase (GGT) and glutamate dehydrogenase (GLDH) ac- or a closed injury, by leakage following surgical biliary tract proce- tivities. Horses in this group had at least two of these five parameters dures, by transudation of bile through the walls of a gangrenous but above the reference range. Normal plasma bile acids were considered nonperforated gallbladder, after rupture of a subcapsular cholangitic <20 µmol/L.19 Post-­mortem examinations and liver histopathology 13,14 abscess of the liver or as an idiopathic disorder. Bile peritonitis were performed by an independent board-­certified pathologist with has been described in dogs and cats mainly due to disruption of the knowledge of the clinical findings and diagnostic workup, but unaware biliary tract secondary to trauma or necrotising cholecystitis.15,16 of the PBAC results. More recently, liver lobe torsion has been reported in horses as a Horses in the GI obstructive and GI ischaemic-­inflammatory 17,18 cause of colic and bile peritonitis. Peritoneal bile acids concentra- groups were classified based on clinical history, findings of a complete tion was found to be elevated in an individual case of liver lobe tor- physical examination, rectal palpation and nasogastric intubation, and sion in our hospital while the remainder of the hepatic enzymes and results of ancillary diagnostic tests (CBC, biochemistry and blood gas plasma bile acids concentration were not significantly increased. In analysis, gastroscopy, abdominal ultrasonography and peritoneal fluid that case, PBAC was comparable to those found in plasma of horses analysis). In those horses that underwent abdominal surgery, the di- with severe hepatic insufficiency. However, no previous data of nor- agnosis was confirmed. Liver disease was ruled out in both groups mal PBAC was available for the correct interpretation of that finding. based on either normal plasma hepatic profile and liver ultrasound in Therefore, the main objectives of the present study were: (a) surviving horses, or absence of hepatic abnormalities on post-­mortem to determine a reference range of PBAC in horses; (b) to compare examination of nonsurviving cases. The diagnosis of ischaemic, inflam- PBAC from horses with either hepatic or gastrointestinal disease and matory, or obstructive gastrointestinal and hepatic disorders was done healthy horses and (c) to assess the prognostic and diagnostic valid- by the attending clinician based on the aforementioned diagnostic ity of PBAC in horses with hepatic or gastrointestinal disease. tests, abdominal surgery and post-­mortem examination where appro- priate, and blinded to the results of the PBAC. The GI obstructive group included horses with nonstrangulating 2 | M ATE R I A L S A N D M E TH O DS and noninflammatory disorders of the intestinal tract without signs of intestinal devitalisation and horses with gastric ulcers or impac- 2.1 | Study population tion. The GI ischaemic-­inflammatory group included horses with strangulating GI lesions and horses with inflammation of the small Horses admitted to the Unitat Equina, Fundació Hospital Clínic intestine, large intestine or peritonitis. Veterinari, Universitat Autònoma de Barcelona, Spain for colic Horses included in the control group were healthy teaching horses or investigation of suspected hepatic disease between June 2010 whose soundness was assessed by a complete physical examination, and June 2014 were considered for enrolment in the study. Foals peritoneal fluid (PF) analysis (cytology and total protein concentration) younger than 12 weeks of age and horses admitted for colic or he- and evaluation of plasma hepatic profile (bile acids, total and direct patic disease in which abdominocentesis was not performed during bilirubin concentrations, GGT and GLDH activities). | 20423306, 2022, 5, Downloaded from https://beva.onlinelibrary.wiley.com/doi/10.1111/evj.13538 by Readcube (Labtiva Inc.), Wiley Online Library on [15/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 916 RODRÍGUEZ-­POZO et al. Horses were also classified according to clinical outcome. The groups were analysed by means of one-­way ANOVA with a nonpara- survivor group included all clinical cases that were discharged from metric approach using rank transformation and pair-­wise comparison. the hospital. The nonsurvivor group included those cases that died A Mann-­Whitney U test was used to explore the association of PBAC during hospitalisation or were euthanised due to poor prognosis. and outcome. Correlation between peritoneal and plasma bile acid Horses from the control group were not included in this classification. concentrations was determined by Spearman correlation coefficient. Statistical analysis was performed using SPSS version 20.0 (SPSS Inc) and a two-­sided Type I error of 5% was used in all analyses (P < .05). 2.3 | Collection and processing of samples Additionally, sensitivity, specificity and predictive values of PBAC for the diagnosis of liver or gastrointestinal disease and out- Blood samples were collected from all horses by routine jugular come were calculated. venipuncture into 1mL lithium heparin tubes. Once centrifuged for 15 minutes at 1,000 g, plasma samples were analysed immediately or otherwise frozen at –­20ºC and analysed within 48 hours. Plasma col- 3 | R E S U LT S lected from horses of the hepatic group was used for biochemical anal- ysis (bile acids, total and direct bilirubin concentrations, GGT and GLDH A total of 119 animals were included in the study, 108 sick horses activities) with an automated chemistry analyser (AU400, Olympus). and 11 healthy horses. Twenty-­seven out of 119 were stallions Peritoneal fluid was obtained aseptically in all horses by abdomino- (22.7%), 51/119 were geldings (42.9%) and 41/119 were mares centesis with a sterile blunt teat cannula, ultrasound guided, approx- (34.5%). Horses were between 3 months and 30 years old (9.6 ± 6.0, imately 2 cm to the right of the midline at the most dependent area average ±SD). The breed distribution reflected the hospital's referral of the ventral abdomen. Each PF sample was collected into both 1mL population, with 37 Andalusian (31.1%), 29 cross-­bred (24.4%), 14 lithium heparin and 1mL K3-­EDTA tubes. Lithium heparin samples were Warmblood (11.8%), 13 Arabian (10.9%), 9 ponies (7.6%), 7 Draught-­ analysed for bile acids concentration using the same enzymatic colori- bred horses (5.9%) and 10 horses of other breeds (8.4%). Sick horses metric method (Randox) as for plasma determination immediately after were classified according to diagnosis into hepatic (n = 13), GI ob- centesis or otherwise frozen at –­20ºC and analysed within 48 hours. structive (n = 48) and GI ischaemic-­inflammatory (n = 47) groups. Samples in K3-­EDTA tubes were used for peritoneal total protein The hepatic group included 13 horses with the following diagnosis: concentration, white cell count and cytology. White cell counts were liver lobe torsion (2), cholangiohepatitis (4), hepatic lipidosis (2), lym- obtained using an automatic cell counter (Advia 120 analyser, Bayer Lab) phoma (1), metastatic hepatic melanoma (1), hepatic abscess due to S and total protein concentration was measured using a refractometer. equi equi (1), and uncharacterised hepatopathy and hepatic insufficiency (2). The diagnosis of liver disease was confirmed by histopathology in 9 out of 13 cases (ultrasound guided biopsy n = 4, necropsy n = 5). The 2.4 | Validation of determination of bile acids GI obstructive group included 48 horses with diagnosis of large colon concentration in peritoneal fluid impaction (22), large colon displacement (18), gastric ulcers (4), small colon obstruction (3) and stomach impaction (1). The “GI ischaemic-­ Biochemical measurement of bile acids concentration in horse peri- inflammatory” group included 47 horses with diagnosis of intestinal vol- toneal fluid was validated as part of this study following the meth- vulus or torsion (21), epiploic foramen entrapment (2), inguinal hernia odology elsewhere described. 20 For the calculation of intra-­assay (4) and caecocolic intussusceptions (3). Horses with enteritis (2), colitis CV a low concentration sample (7.2 µmol/L) and high concentration (9) and septic peritonitis (6) were also included in this GI ischaemic-­ sample (187.5 µmol/L) were analysed 10 times. For the calculation inflammatory group. Considering outcome, a total of 68/108 sick of inter-­assay CV the same samples were analysed during 5 con- horses survived to discharge and were included in the survivor group. secutive days. Mean values for intra-­assay CV and inter-­assay CV Seventeen out of 68 survivors were in the GI ischaemic-­inflammatory were 0.5% and 0.7% respectively. An individual sample with an ini- group, 44 in the GI obstructive group and 7 in the hepatic group. Forty tial concentration of 187.5 µmol/L was used to perform serial dilu- horses died or were euthanised due to poor prognosis. Thirty of the tions up to 1:128 (1.5 µmol/L) of the analyte in order to determine nonsurviving horses were from the GI ischaemic-­inflammatory group, assay linearity. Using linear regression analysis an r value of 0.998 four from the GI obstructive group and six from the hepatic group. was obtained. Finally, the enzymatic colorimetric assay yielded a Finally, 11 healthy horses were enrolled in the control group. recovery ranging from 77% to 98% from spiked samples. 3.1 | Plasma bile acids concentration 2.5 | Data analysis The median value of plasma bile acids concentration in the hepatic Results for qualitative variables are described by absolute frequencies group (9.66 µmol/L) was significantly higher than the control (4.16; and percentages and as median and interquartile range [Percentiles P = .04), GI obstructive (2.64; P = .003) and GI ischaemic-­inflammatory 25th, 75th] for quantitative variables. Differences between diagnostic (2.63; P = .016) groups (Figure 1; Table 1). No other statistically | 20423306, 2022, 5, Downloaded from https://beva.onlinelibrary.wiley.com/doi/10.1111/evj.13538 by Readcube (Labtiva Inc.), Wiley Online Library on [15/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License RODRÍGUEZ-­POZO et al. 917 Plasma bile acid concentration TA B L E 1 Median and IQR of the plasma and peritoneal bile acids concentration obtained in control horses, gastrointestinal disease and hepatic disease groups. Horses are also classified as Reference 100 range 0-­20 80 4.14 [1.58-­6.53]c 3.72 [1.31-­9.28] 60 [Bile acid] Nonsurvivor µmol/L 40 group 20 0 2.81 [1.59-­5.27] Survivor group 1.34 [0.82-­3] l ive ic tic ro em pa nt ct co tru ha he isc s ob Group F I G U R E 1 Plasma bile acid concentration (µmol/L) of control and sick horses classified into four different groups. The median 1.00 [0.59-­1.49]a,b 4.16 [1.00-­5.80]a bile acid concentration is significantly higher in horses with hepatic Control group disease relative to control and gastrointestinal groups significant differences were observed among diagnostic groups. No differences were observed between survivor and nonsurvivor groups. inflammatory group 2.63 [1.50-­6.40]a 3.2 | Peritoneal bile acids concentration 3.29 [1.36-­5.5] GI ischaemic-­ Determination of PBAC was possible in all the samples obtained without technical limitations related to the type of fluid (PF) ana- lysed (ie transudate, modified transudate and exudate). The median PBAC of the hepatic group was significantly higher than 1.20 [0.75-­1.66]a,b 2.64 [1.37-­4.39]a in the control and GI obstructive groups (6.8 vs. 1.0 and 1.2 µmol/L re- GI obstructive spectively; P < .001). Moreover, the GI ischaemic-­inflammatory group Median significantly different from the GI ischaemic-­inflammatory group. also had significantly higher median PBAC than the control and GI ob- group structive groups (3.3 vs. 1.0 and 1.2 µmol/L respectively; P < .001). No differences were observed between the GI ischaemic-­inflammatory and the hepatic group and between control and GI obstructive groups (Figure 2). Regarding outcome, the nonsurvivor group had significantly Abbreviation: PBAC, peritoneal bile acids concentration. Median significantly different from the survivor group. 9.66 [3.37-­31.22] Median significantly different from the hepatic group. 6.77 [2.27-­9.44] higher median PBAC than the survivor group (4.1 vs 1.3 µmol/L respec- Hepatic group tively; P < .001). Following conventional statistical criteria, we considered as a possible PBAC cut-­off value 2.28 µmol/L, that corresponded to 2 standard deviation over the mean of control group (1.04 µmol/L; survivor and nonsurvivor groups SD 0.62). Most of the hepatic (9 of 13; 69%) and the GI ischaemic-­ Plasma bile acid concentration inflammatory (30 of 47; 64%) horses had a PBAC ≥2.28 µmol/L, whereas only 7 out of 48 horses (15%) with GI obstruction had PBAC ≥2.28 µmol/L. Related to outcome, 68% of nonsurvi- vor horses had PBAC≥2.28 µmol/L, in contrast to 28% of the PBAC (µmol/L) survivors. (µmol/L) Correlation of plasma and peritoneal bile acid concentrations was low when all horses were included (rho = 0.44, P < .001), and sub- jectively most of the obstructive colics had normal PBAC and plasma b a c | 20423306, 2022, 5, Downloaded from https://beva.onlinelibrary.wiley.com/doi/10.1111/evj.13538 by Readcube (Labtiva Inc.), Wiley Online Library on [15/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 918 RODRÍGUEZ-­POZO et al. Peritoneal bile acid concentration bile acids, whereas higher increases in PBAC relative to plasma were 200 observed in the ischaemic-­inflammatory group (Figure 3). 180 3.3 | Sensitivity, specificity, positive and negative predictive values of PBAC [Bile acid] µmol/L 160 Using the population of sick horses (hepatic and nonhepatic groups), 40 the sensitivity, specificity and predictive values of PBAC for the di- agnosis of hepatic disease were calculated. Using the cut-­off value 20 of 2.28 µmol/L, PBAC showed modest sensitivity (69%, [39%-­91%]; [95%CI]) and specificity (61%, [51%-­71%]). Although the positive pre- 0 dictive value (PPV) was poor (19.6%, [14%-­27%]), the negative pre- dictive value (NPV) was fairly good (94%, [86%-­97%]). l ive ic tic ro em Considering that the two groups including horses with more severe pa nt ct co tru ha he isc abdominal disease (ie hepatic and GI ischaemic-­inflammatory groups) s ob Group had the highest percentage of horses with increased PBAC, we inves- control tigated the usefulness of this parameter to diagnose severe abdominal obstructive disease. The sensitivity and NPV were moderate (63% [52%-­77%] and ischaemic 72% [63%-­78%] respectively) but fair to good values of specificity and hepatic PPV (89% [77%-­95%] and 84% [73%-­92%] respectively) were observed. F I G U R E 2 Peritoneal bile acid concentration (µmol/L) of control In addition, the sensitivity, specificity and predictive values and sick horses classified into four different groups. The median of PBAC for the prognosis of sick horses (hepatic and nonhepatic PBAC is significantly higher in horses with hepatic disease relative groups) were calculated. Using the cut-­off value of 2.28 µmol/L to to control and gastrointestinal obstructive groups. In addition, predict outcome, PBAC showed modest sensitivity (67% [51%-­81%]) median PBAC is significantly higher in horses with ischaemic and PPV (58% [48%-­69%]). Similarly, the specificity (73% [60%-­ gastrointestinal disease relative to control and obstructive gastrointestinal groups 82%]) and NPV (79% [70%-­86%]) were moderate. [18.9,185.5] 100 15 Hepatic Obstructive 90 Ischaemic 80 Control 10 70 Peritoneal bile acids (µmol/L) 60 5 50 40 0 30 0 5 10 15 20 F I G U R E 3 Scatterplot of plasma and peritoneal bile acid concentrations. Note 10 that the inset graph is a zoom in to display most of the values in the lower range. An 0 outlier value [X,Y] is depicted outside of 0 20 40 60 80 100 the main graph. The green box represents Plasma bile acids (µmol/L) the reference range of peritoneal and plasma bile acid concentration | 20423306, 2022, 5, Downloaded from https://beva.onlinelibrary.wiley.com/doi/10.1111/evj.13538 by Readcube (Labtiva Inc.), Wiley Online Library on [15/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License RODRÍGUEZ-­POZO et al. 919 4 | D I S CU S S I O N lactate concentration in PF are less likely to survive. 26-­28 Horses with peritoneal lactate concentration over 4 mmol/L after admission to a The main findings of this study were: (a) healthy horses have neg- referral hospital were reported to be three times more likely to suf- ligible peritoneal bile acids concentrations (three times lower than fer a strangulating intestinal lesion. 29 Similarly, as ischaemia and hy- plasma bile acids), (b) about 70% of horses with ischaemic or inflam- poxia progress, intestinal wall devitalisation could result in increased matory gastrointestinal disorders or hepatic disease have altered intestinal wall permeability and diffusion of intraluminal substances PBAC, and (c) increased PBAC is associated with nonsurvival in (ie bile acids) into the peritoneal fluid. Given their biochemical prop- horses with colic or hepatic disease. erties, bile acids should not easily diffuse across cell membranes, Peritoneal bile acids concentration has not been established be- and increased bile acids concentration in peritoneal fluid could be fore in horses. One of the main objectives of the study was to de- a marker of bile duct disruption or loss of intestinal wall integrity (ie termine PBAC in horses and to compare concentrations from horses devitalisation). Therefore, increased bile acids concentration could with either hepatic or gastrointestinal disease and healthy horses. perhaps be suggestive of more advanced intestinal necrosis and not Abdominocentesis and laboratorial analysis of PF are common diag- simply early stages of ischaemia and hypoxia. nostic procedures in horses admitted to a referral hospital for gas- The results obtained from healthy horses suggest that in normal trointestinal (GI) disorders and other intraabdominal conditions. It physiological conditions, bile acids might be present in small concen- is well recognised that colicky horses with a simple obstructive GI tration in the peritoneal cavity. On average the ratio of peritoneal lesion (eg pelvic flexure impaction) will usually have normal PF cy- to plasma bile acid concentrations in normal horses was 1:3. In con- tology (transudate) and low lactate concentration; in contrast, those trast, the median PBAC of the hepatic group was significantly higher with an ischaemic GI lesion (eg small intestinal volvulus) could have than the control and GI obstructive groups, and the GI ischaemic-­ abnormal WBC count and TP concentration and/or high lactate con- inflammatory group had significantly higher values than the control centration relative to plasma.7 Similarly, determination of creatinine and GI obstructive groups (Table 1). A large proportion of horses concentration in PF, relative to that in plasma, is of diagnostic value in with GI ischaemic-­inflammatory disorders had high ratio of peri- cases with suspected uroabdomen.12 The observation of unremark- toneal to plasma bile acids concentration (1:1 and higher). Higher able plasma liver enzymes but increased peritoneal and plasma bile median PBAC in the hepatic and GI ischaemic-­inflammatory groups acids concentration in a horse with severe abdominal pain caused by compared to other groups could be related with the compromised liver torsion spurred us to investigate the diagnostic and prognostic liver structure or diffusion of bile acids from the intestinal lumen value of PBAC in horses. At necropsy, complete rupture of the liver across a compromised gut wall into the peritoneal cavity. lobe capsule secondary to liver lobe torsion was observed and, in the The NPV of PBAC for the diagnosis of hepatic disease was fairly authors’ opinion, it could had been the definitive cause of the release good (94.5%) meaning that in this population, a horse with a PBAC of bile acids into the abdomen. Our hypothesis was that determina- below 2.28µmol/L has a 94.5% chance of not having hepatic disease. tion of PBAC could be useful to diagnose specific gastrointestinal Although the positive predictive value (PPV) was low (19.6%) it was disorders (eg ischaemic lesions) and specific liver diseases (eg liver higher than the prevalence of the hepatic disease in this population torsion or cholelithiasis) or could be associated with outcome. of hospitalised horses (ie 12%) meaning that this parameter could be Bile acids, amphophilic derivatives of cholesterol, are synthesised valuable as a screening test for the diagnosis of hepatic disease in within hepatocytes, conjugated with either glycine or taurine, and this type of population. are excreted into the bile.1,21 Given their “detergent-­like” properties, The specificity and PPV of PBAC to diagnose severe abdominal bile acids have an important role in digestion and absorption of lipids disease were fair to good (88.7% and 84% respectively). A practical within the small intestine. 22 Nondegraded bile acids are efficiently interpretation would be that a horse with a PBAC >2.28µmol/L has absorbed by specific transporters of the ileal mucosa and trans- 84% possibility of having an inflammatory/ischaemic intestinal dis- ported back to the liver via the portal circulation, and about 90% of ease or hepatic disorder. the bile acid pool is restricted to the enterohepatic circulation. 21,22 Related to prognosis, 67.5% of nonsurvivor horses had PBAC Since hepatic metabolism is the only source of bile acids synthesis, higher than 2.28 µmol/L, in contrast to 27.9% of the survivors. and the vast majority of bile acids remain in the enterohepatic circu- Lactate has been studied as prognostic indicator in colic patients.7 lation, increased plasma bile acids concentration has been described Considering the results herein described, PBAC could be used in as useful biomarker of liver disease and function in horses.3,21,23,24 combination with peritoneal lactate concentration in order to pre- More recently, plasma bile acids and other hepatic biomarkers have dict the chances of survival in colic patients. been investigated in horses undergoing colic surgery. 25 In this pilot Plasma bile acids concentration have been previously determined study, horses with markedly increased admission plasma bile acid in horses by the colorimetric method, with values of 5-­28 μmol/L concentrations and SDH activities did not survive, however, the small considered normal. 2 In another study of horses undergoing colic sur- number of horses included and the bias towards only surgical cases, gery, normal plasma bile acids concentration was considered <8.6 does not allow for direct comparison with the present study. and mildly high 8.6-­20 μmol/L. 25 In the present study, only horses Peritoneal lactate concentration is considered a sensible bio- included within the “hepatic” group presented increased plasma bile marker of intestinal ischaemia in colicky horses and those with high acids concentration. The reason of admission in five out of 11 horses | 20423306, 2022, 5, Downloaded from https://beva.onlinelibrary.wiley.com/doi/10.1111/evj.13538 by Readcube (Labtiva Inc.), Wiley Online Library on [15/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 920 RODRÍGUEZ-­POZO et al. of the “hepatic group” was colic signs. Blood analysis and comple- PEER REVIEW mentary tests allowed us to classify as hepatic disease. In a previ- The peer review history for this article is available at https://publo​ ous study, 24 out of 26 horses with diagnosis of GI disease in which ns.com/publo​n/10.1111/evj.13538. plasma bile acids were determined resulted within the reference range.3 This finding is similar to our results in that horses in both DATA AVA I L A B I L I T Y S TAT E M E N T colic groups had unaltered plasma bile acids concentration. The data that support the findings of this study are available from The findings of the present study should be interpreted consid- the corresponding author upon reasonable request. ering several limitations. Firstly, the possibility of misclassification of cases to diagnostic categories, given that a full biochemical hepatic ORCID profile was not performed in all horses admitted for gastrointestinal Lara Armengou https://orcid.org/0000-0003-4715-7323 disorders, or histopathologic examination was not performed in 4 Judit Viu https://orcid.org/0000-0001-5166-4139 out of 13 hepatic cases. Secondly, the preliminary reference range Jose Ríos https://orcid.org/0000-0002-0716-8784 proposed in this study should be validated with an external popu- Eduard Jose-­Cunilleras https://orcid.org/0000-0002-4536-7717 lation, and the 2.28 μmol/L cutoff for PBAC should be interpreted with caution. And finally, these findings in a referral hospital case- REFERENCES load may not be representative of the general population of horses 1. Engelking LR. Evaluation of equine bilirubin and bile acid metabo- with gastrointestinal or hepatic disorders. lism. Compend Contin Educ Pract Vet. 1989;11:328–­36. 2. Kaneko JJ, Rudolph WG, Wilson DV, Carlson GP. Bile acid fraction- In conclusion, PBAC is not an accurate biomarker of hepatic dis- ations by high performance liquid chromatography in equine liver ease but it could be useful to diagnose horses with ischaemic or inflam- disease. Vet Res Commun. 1992;16:161. matory abdominal disease. In the present study horses with ischaemic 3. West HJ. Evaluation of total plasma bile acid concentration for the diag- or inflammatory gastrointestinal disease had increased PBAC and al- nosis of hepatobiliary disease in horses. Res Vet Sci. 1989;46:264–­70. 4. McGorum BC, Murphy D, Love S, Milne EM. Clinicopathological tered PBAC is associated with nonsurvival. A preliminary reference features of equine primary hepatic disease: a review of 50 cases. range of peritoneal bile acids concentration is 0-­2.28 μmol/L, which is Vet Rec. 1999;145:134–­9. three times lower than plasma bile acids concentration. A higher num- 5. Van Hoogmoed L, Rodger LD, Spier SJ, Gardner IA, Yarbrough TB, ber of horses with abdominal disease is required to better understand Snyder JR. Evaluation of peritoneal fluid pH, glucose concentration, and lactate dehydrogenase activity for detection of septic peritoni- the clinical significance of these findings in horses with gastrointesti- tis in horses. J Am Vet Med Assoc. 1999;214:1032–­6. nal disease without hepatic involvement. 6. Barton MH, Collatos C. Tumor necrosis factor and interleukin-­6 activ- ity and endotoxin concentration in peritoneal fluid and blood of horses ETHICAL ANIMAL RESEARCH with acute abdominal disease. J Vet Intern Med. 1999;13:457–­64. 7. Latson KM, Nieto JE, Beldomenico PM, Snyder JR. Evaluation of Procedures on healthy animals were approved by the Institution's peritoneal fluid lactate as marker of intestinal ischemia in equine animal experimentation ethics committee. colic. Equine Vet J. 2005;37:342–­6. 8. Delgado MA, Monreal L, Armengou L, Ríos J, Segura D. Peritoneal I N FO R M E D C O N S E N T D-­dimer concentration for assessing peritoneal fibrinolytic activity in horses with colic. J Vet Intern Med. 2009;23:882–­9. Explicit owner informed consent for inclusion of samples from ani- 9. Grulke S, Franck T, Gangl M, Péters F, Salciccia A, Deby-­Dupont mals in this study was not sought but owners were given the option G, et al. Myeloperoxidase assay in plasma and peritoneal fluid of to opt out of research. horses with gastrointestinal disease. Can J Vet Res. 2008;72:37–­42. 10. Parry BW, Crisman MV. Serum and peritoneal fluid amylase and li- pase reference values in horses. Equine Vet J. 1991;23:390–­1. AC K N OW L E D G E M E N T S 11. Saulez MN, Cebra CK, Tornquist SJ. The diagnostic and prognostic The assistance of Dr Francisco Javier Mendoza with sample acquisi- value of alkaline phosphatase activity in serum and peritoneal fluid tion of four of the healthy teaching horses is gratefully acknowledged. from horses with acute colic. J Vet Intern Med. 2004;18:564–­7. 12. Kablack KA, Embertson RM, Bernard WV, Bramlage LR, Hance S, Reimer C O N FL I C T O F I N T E R E S T S JM, et al. Uroperitoneum in the hospitalised equine neonate: retrospec- tive study of 31 cases, 1988–­1997. Equine Vet J. 2000;32(6):505–­8. No competing interests have been declared. 13. Ellis H, Cronin K. Bile peritonitis. Br J Surg. 1960;48(208):166–­71. 14. Koch M, Garden OJ, Padbury R, Rahbari NN, Adam R, Capussotti AU T H O R C O N T R I B U T I O N S L, et al. Bile leakage after hepatobiliary and pancreatic surgery: a ML Rodríguez-­Pozo and E. Jose-­Cunilleras contributed to study de- definition and grading of severity by the International Study Group of Liver Surgery. Surgery. 2011;149(5):680–­8. sign, data collection and study execution, data integrity, data analy- 15. Kipnis RM. Cholelithiasis, gallbladder perforation, and bile peritoni- sis and interpretation. L. Armengou contributed to study design, and tis in a dog. Canine Pract. 1986;13:15–­27. data collection and study execution. J. Viu contributed to data col- 16. Ludwig LL, McLoughlin MA, Graves TK, Crisp MS. Surgical treat- lection and study execution, and data analysis and interpretation. J. ment of bile peritonitis in 24 dogs and 2 cats: a retrospective study (1987–­1994). Vet Surg. 1997;26:90–­8. Ríos contributed to data integrity, and data analysis and interpreta- 17. Tennent-­Brown BS, Mudge MC, Hardy J, Whelchel DD, Freeman tion. All authors contributed to the preparation of the manuscript DE, Fischer AT Jr. Liver lobe torsion in six horses. J Am Vet Med and gave their final approval. Assoc. 2012;241(5):615–­20. | 20423306, 2022, 5, Downloaded from https://beva.onlinelibrary.wiley.com/doi/10.1111/evj.13538 by Readcube (Labtiva Inc.), Wiley Online Library on [15/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License RODRÍGUEZ-­POZO et al. 921 18. Bentz KJ, Burgess BA, Lohmann KL, Shahriar F. Hepatic lobe tor- horses presenting with strangulating lesions of the small intestine. sion in a horse. Can Vet J. 2009;50(3):283–­6. Vet Surg. 2019;48(5):786–­94. 19. DeNotta SL. Divers TJ clinical pathology in the adult sick horse: the 27. Radcliffe RM, Divers TJ, Fletcher DJ, Mohammed H, Kraus MS. gastrointestinal system and liver. Vet Clin North Am Equine Pract. Evaluation of L-­lactate and cardiac troponin I in horses under- 2020;36(1):105–­20. going emergency abdominal surgery. J Vet Emerg Crit Care. 20. Boutet P, Heath F, Archer J, Villiers E. Comparison of quantitative 2012;22(3):313–­9. immunoturbidimetric and semiquantitative latex-­agglutination as- 28. Dunkel B, Mason CJ, Chang YM. Retrospective evaluation of the says for D-­dimer measurement in canine plasma. Vet Clin Pathol. association between admission blood glucose and l-­lactate concen- 2009;38:78–­82. trations in ponies and horses with gastrointestinal disease (2008–­ 21. Barton MH, LeRoy BE. Serum bile acids concentrations in healthy 2016): 545 cases. J Vet Emerg Crit Care. 2019;29(4):418–­23. and clinically ill neonatal foals. J Vet Intern Med. 2007;21(3):508–­13. 29. Peloso JG, Cohen ND. Use of serial measurements of peritoneal 22. Tennant BC, Center SA. Hepatic function. In: Kaneko JJ, Harvey fluid lactate concentration to identify strangulating intestinal le- JW, Bruss ML, editors. Clinical Biochemistry of Domestic Animals, sions in referred horses with signs of colic. J Am Vet Med Assoc. 6th edn. Amsterdam: Elsevier; 2008. pp. 379–­412. 2012;240(10):1208–­17. 23. Divers T. Biochemical diagnosis of hepatic disease and dysfunction in the horse. Equine Pract. 1993;15:15–­7. 24. Hoffman WE, Baker G, Rieser S, Dorner JL. Alterations in selected serum biochemical constituents in equids after induced hepatic dis- How to cite this article: Rodríguez-­Pozo ML, Armengou L, Viu ease. Am J Vet Res. 1987;48:1343–­7. J, Ríos J, Jose-­Cunilleras E. Peritoneal bile acids concentration 25. Underwood C, Southwood LL, Walton RM, Johnson AL. Hepatic in adult horses with hepatic and gastrointestinal disorders. and metabolic changes in surgical colic patients: a pilot study. J Vet Equine Vet J. 2022;54:914–­921. doi:10.1111/evj.13538 Emerg Crit Care. 2010;20(6):578–­86. 26. Pye J, Espinosa-­Mur P, Roca R, Kilcoyne I, Nieto J, Dechant J. Preoperative factors associated with resection and anastomosis in