Gallbladder Sludge in Dogs: Ultrasonographic and Clinical Findings in 200 Patients

Introduction

The biliary tract is routinely examined during ultrasonographic (US) scans in dogs, and the contents of the gallbladder (GB) are easily evaluated. These may be uniformly nonechogenic, with no evidence of solid debris. Alternatively, variable amounts of biliary sludge, defined as gravity-dependent hyperechoic variably particulate material without acoustic shadowing, may be noted.¹ Less commonly, GB contents may appear solid and non–gravity dependent. The presence of radial hyperechoic striations within this solid material and a hypoechoic rim suggests a GB mucocele (MC).^1,2 This is defined as an accumulation of mucin within the GB accompanied by hyperplasia of the mucus-secreting epithelium. Although choleliths appear to be relatively uncommon in dogs, these are characterized as markedly hyperechoic structures with acoustic shadowing. Depending on the density of the surrounding bile, these are generally mobile and gravity dependent.

In human patients, biliary sludge is a complex mixture of cholesterol crystals, bile pigments, and bile salts, embedded in mucin.³ This material often includes particles ≥1 mm, referred to as microliths, which are thought to precede gallstone formation.³ People with sludge and microlithiasis may report colicky pain and postprandial discomfort. Intervention is recommended to prevent cholelithiasis and often includes ursodeoxycholic acid (ursodiol) and cholecystectomy.⁴ The composition of spontaneous biliary sludge in dogs has not been reported, but low incidence of gallstones in this species compared to people suggests that canine sludge may have a different composition or etiology.⁵ This is supported by experimental studies indicating that canine biliary sludge does not contain substantial amounts of cholesterol.^6–8

The clinical implication of biliary sludge in dogs is unclear. An early study suggested a weak association between sludge and patient age but did not identify a connection to naturally occurring hepatobiliary disease.⁹ It was concluded that GB sludge should be considered an incidental finding in this species. This is somewhat contradicted by a later study, in which mobile sludge or precipitate was reported in 24/45 dogs with GB disease; nine of these 24 (37.5%) ruptured.¹ More recently, abnormal GB contents (both sludge and MC) were associated with impaired emptying following a meal challenge in dogs.¹⁰ However, it was not determined if decreased GB contractility was the cause or consequence of the abnormal contents.

In recent years, both genetic and metabolic factors have been implicated in MC formation in dogs. An insertion mutation in the ABCB4 gene was identified in 14/15 affected Shetland sheepdogs, but not in 20/21 unaffected individuals, and may explain the predisposition of this breed to MC formation.¹¹ This mutation impairs secretion of a protective phospholipid across the canalicular membrane, thereby making the GB epithelium more vulnerable to the irritant properties of bile. In a retrospective study of 78 dogs diagnosed with MC, age- and breed- matched controls were used to identify the impact of a concurrent endocrinopathy.¹² In this population, the odds of MC in dogs with hyperadrenocorticism (HAC) were 29 times that of dogs without HAC. A weaker association was found for hypothyroidism (HT), with a threefold increased risk. Although various sources suggest that sludge may precede MC formation, there is presently little direct evidence to support this hypothesis, and similar influences have not been reported for biliary sludge.^13,14

The goals of this study were to determine the prevalence of echogenic material (“sludge”) within the GB of dogs undergoing routine diagnostic abdominal US examination at a veterinary teaching hospital, to describe population characteristics for dogs with minimal versus substantial (>25%) GB sludge, to evaluate and compare selected clinicopathological data for dogs from these subsets, and to identify the impact of concurrent endocrine disorders on GB contents.

Materials and Methods

A grading scheme was created to allow crude quantification of the percentage of hyperechoic material within the GB, as determined during real-time US examination (Table 1). This grading tool was then used to score GB contents in 200 sequential examinations performed on canine patients undergoing routine abdominal studies at Texas A&M University Veterinary Medical Teaching Hospital from November 2009–April 2010. All US scans were performed by Diplomates of the American College of Veterinary Radiology or a radiology resident under direct supervision of a board-certified radiologist. Two different US systems^a,b were used, depending on availability and radiologist preference. Transducer type and frequency were selected by the attending radiologist based on patient size, configuration, and imaging characteristics.

TABLE 1 Grading Scale for Ultrasonographic Assessment of Gallbladder Sludge in Dogs

View Fullscreen

Patients were positioned in dorsal or lateral recumbency during the examination, depending on radiologist preference. Sedation was administered as necessary at the direction of the attending clinician. Although fasting is recommended at this institution prior to US examination, patients were variably fed within the preceding 8 hr. Contemporary standard of care was provided to each animal.

For each dog, the contents of the GB were assessed during the US scan, and the findings were recorded immediately following the examination. The amount of sludge (if present) was estimated as a percentage of total GB contents during several seconds of scanning time, using views from various angles, including longitudinal and sagittal views. The GB contents were classified as consistent with a MC if the overall appearance matched that described in the current veterinary literature, i.e., essentially solid, non–gravity dependent, hyperechoic bile with hyperechoic striations, creating a starfish/kiwifruit appearance.^1,2 Results were excluded from the final analysis if the patient had previously been scored. Still images were collected by the radiologist during the study and stored in a digital image archive. A detailed written description of the US findings was generated for each case within 3 d of the examination and included in the computerized medical record.

GB volume was estimated from stored digital images using a modification of the method proposed by Dodds et al., in which width was substituted for height.¹⁵ Measurement of length and width were determined from long axis images in which the neck of the GB was evident. If more than one long axis image was available, the one with the greater longitudinal dimension was used. GB volume was calculated using the following equation:

GB volumes were then indexed for body weight in kilograms.

Information regarding patient age, breed, gender, and weight was obtained from computerized medical records. A body condition score (BCS) was derived from documents within the medical record and converted (if necessary) to a 9-point scale. Clinicopathological data reported within 24 hr of US scanning were reviewed; serum cholesterol and bilirubin concentrations and alkaline phosphatase (ALP) and gamma-glutamyl transferase (GGT) activities were recorded. Serum biochemical analyses were all performed at the in-house clinical pathology laboratory using an automated analyzer^c. Medications administered within the previous 8 wk were noted, including sedatives administered to facilitate ultrasonography.

One of the authors (A.K.C. a board-certified internist) reviewed each medical record and classified the patient into one of eight disease groups based on the reason(s) for performing ultrasonography, clinical complaints, and final diagnoses. These disease groups were hepatobiliary (icterus, increased enzyme activity, hyperammonemia), endocrine (diabetes mellitus [DM], HAC), urinary (renal, bladder, prostatic), gastrointestinal (including pancreatic), neurologic, neoplastic, thoracic, and other (immune mediated/ nonspecific/systemic/undefined).

Patients were classified as having spontaneous HAC if the attending clinician listed this diagnosis in the visit summary. Patients diagnosed by the attending clinician with inadequate thyroid function and those receiving thyroid supplementation at the time of examination were classified as HT. The appropriateness of a prior diagnosis of HT was not investigated, and the efficacy of therapy for dogs on thyroid supplementation was not evaluated. A diagnosis of DM was listed for any patient with blood glucose >200 mg/dl (reference range: 65–130 mg/dl) or currently receiving insulin therapy.

Statistical analysis of the data was performed using a commercial software program^d; findings were considered significant if P < 0.05. Where appropriate, data sets were tested for normality using the D'Agostino and Pearson omnibus normality test. The median values of selected data sets were compared using the Mann-Whitney test for unpaired data. Fischer's exact test was used to compare population characteristics. Odds ratios are reported with a 95% confidence interval (CI).

Results

Abnormal GB contents were noted in 133 dogs (66.5%) (Table 2). Almost half (47%) of the dogs examined had a small amount of echogenic material within the GB lumen (<25% of total GB contents). Four dogs (2%) had >75% hyperechoic bile (grade 5 sludge), and four (2%) were categorized as having a MC. For statistical purposes, dogs with normal to minimally abnormal bile (i.e., grades 1 and 2; n = 161: 80.5%) were compared to those with >25% sludge (grades 3–5; n = 35: 17.5%). Dogs with a MC (n = 4) were considered separately.

TABLE 2 Results of Ultrasonographic Gallbladder Scoring in 200 Canine Patients

View Fullscreen

Suitable images for estimating GB volume were available for all but six dogs (all graded 1 or 2). The median volume for dogs graded 1 and 2 was 0.73 ml³/kg (range 0.08–3.65 ml³/kg); this was significantly less than for those graded 3–5 (1.04 ml³/kg; range 0.31–7.86 ml³/kg; P < 0.001; Figure 1).

View Figure

• Download as Powerpoint
• Download Figure

Patients ranged in age from 0.25–17 yr, with a median of 9 yr. The median age for dogs graded 1 and 2 was 8 yr; this was significantly lower than for the dogs with grade 3–5 sludge (11 yr; P < 0.0002; Figure 2). A total of 41 breeds were listed in the study population, with Labrador retriever (n = 20), Yorkshire terrier (n = 16), and miniature schnauzer (n = 15) most frequently reported. Thirty-one dogs were of mixed breeding. Breed lists for dogs with grade 1 and 2 sludge were similar to those with grade 3–5, and no breed predispositions were noted. Overall gender distribution was evenly divided, with 54% female and 46% male, and was similar for dogs with grade 1 and 2 sludge and those with grade 3–5. Median body weight for both subsets was 12.1 kg, with a range of 1–63.8 kg. BCS was available for 50 dogs graded 1 and 2 (median: 5; range 2–9) and 11 dogs graded 3–5 (median 6; range 5–9); these results were not significantly different.

View Figure

• Download as Powerpoint
• Download Figure

Contemporaneous laboratory findings were available for 142 dogs graded 1 and 2 (88.2%) and 34 dogs graded 3–5 (97.1%). Median serum cholesterol concentration (reference range: 120–147 mg/dl) for dogs graded 1 and 2 was 219 mg/dl (range: 53-861 mg/dl); this was not significantly different from that for dogs graded 3–5 (242 mg/dl; range 88–643 mg/dl). Serum bilirubin concentrations (reference range: 0–0.8 mg/dl) were similar for the two patient subsets, with medians of 0.2 (range: <0.1–14.6 mg/dl) and 0.3 mg/dl (range: <0.1–12.9 mg/dl), respectively. ALP activities (reference range: 24–147 U/L) were also similar: the median for dogs graded 1 and 2 was 161 U/L (range 20–5259 U/L); the median for dogs graded 3–5 was 183 U/L (range: 20–5202 U/L). Median GGT activity (reference range: 0–25 U/L) was not significantly different for dogs graded 1–2 (11 U/L; range 8–284 U/L) compared to the grade 3–5 group (12 U/L; range 8–122 U/L).

The majority of patients (124; 62%) were sedated prior to ultrasonography, primarily with opioids such as butrophanol (n = 74), buprenorphine (n = 27) and fentanyl (n = 9). Sedation did not appear to impact GB score, with a median score of 2 for both sedated and nonsedated dogs. GB volumes were also similar for the sedated and nonsedated dogs, with median values of 0.89 ml/kg and 0.75 ml/kg, respectively.

Only 26 dogs (13%) underwent ultrasonography for suspected or established hepatobiliary disease. None of these patients had an MC, but the proportion of dogs with grade 3–5 sludge was higher than for dogs without hepatobiliary disease (30.7% versus 15.9%; odds ratio: 2.65; CI 1.03–6.80; P = 0.048). The most common reason for ultrasonography was gastrointestinal disease, with 49 (24.5%) of dogs, followed by 41 (20.5%) dogs with neoplastic disease and 34 (17%) with urinary tract disorders.

Eleven dogs underwent ultrasonography for evaluation of a suspected or known endocrinopathy, although concurrent HAC, HT, or DM was reported in 26 dogs (13%). Three patients were diagnosed with more than one endocrinopathy. HT was the most common endocrinopathy, affecting 16 (8%) dogs, all of which were diagnosed prior to this visit; DM was reported in eight dogs (4%) and HAC in five dogs (2.5%).

Overall, the prevalence of an endocrinopathy (DM, HT, or HAC) in dogs with grade 3–5 sludge was similar to that in dogs with grade 1 and 2 sludge (17.1 % and 10.6%, respectively). However, dogs with HAC and/or HT were substantially more likely to have grade 3–5 sludge (n = 9) than grade 1 and 2 (n = 11) when compared to the rest of the study population (odds ratio 5.04; CI: 1.90–13.33; P < 0.002%).

Fifteen dogs had received exogenous glucocorticoids for >4 wk prior to examination. Agents administered included prednisone (n = 13; median dose 1.1 mg/kg/d; range 0.2–2.2 mg/kg/d), dexamethasone (n = 1; 0.16 mg/kg), and budesonide (n = 1; 0.15 mg/kg). The administration of exogenous steroids for >4 wk was not associated with increased sludge.

Discussion

Echogenic luminal contents were noted in two-thirds of canine patients undergoing routine diagnostic US examination of the abdomen at this institution. This is substantially higher than the prevalence reported in a recent study from Brazil, in which sludge was reported in 35% of 1021 dogs undergoing abdominal US evaluation.¹⁶ This may reflect different patient demographics or other characteristics. Certainly, animals examined at our institution may not reliably reflect the canine population in general, and conclusions regarding the true prevalence of sludge in this species should be made with caution. It is interesting to note that four dogs (2% of those examined) had US evidence of a MC. This number is higher than expected and does suggest that findings from a tertiary referral center are subject to substantial bias.

In this report, dogs with >75% hyperechoic bile and those with US findings indicating a MC were classified separately. In fact, the diagnosis of a MC requires documentation of mucinous hyperplasia of the GB epithelium and cannot be definitively determined on the basis of US or gross findings alone.¹⁷ However, the majority of canine GB MCs have such a distinctive US appearance that this imaging modality is considered to be highly specific for this condition.^1,2 Although it is possible that some of these eight patients were incorrectly categorized, it was important to attempt to separate dogs with severe sludge (>75%) from those with a MC. Failure to separate these patients would have implied some commonality between these two conditions and may have biased our results.

Overall, dogs with grade 3–5 sludge had larger GB volumes (indexed for body weight) compared to those with grade 1 and 2 sludge. Since compromised GB emptying has been associated with microlithiasis in people, this is the most likely explanation for this finding in this patient population.⁵ GB emptying is triggered by the arrival of ingesta in the proximal small intestine and is mediated by cholecystokinin. Impaired emptying may reflect a derangement in the normal signaling process or an inability to effectively move abnormally dense contents. Prospective, dynamic studies using timed meals or cholecystokinin analogues are needed to investigate potential associations between GB sludge and emptying in dogs.

In common with previous veterinary studies, our findings indicate that biliary sludge is more likely to be noted in older dogs.^9,16 However, the development of sludge cannot be predicted simply on the basis of advancing age, as there was no evidence of a linear relationship between these parameters, and several dogs with uniformly hypoechoic bile were >10 yr of age. The development of biliary sludge in people is thought to be multifactorial, but increasing age is consistently identified as a strong risk factor. In fact, the incidence of gallstones increases 4–10-fold after age 40.⁵

Obesity and insulin resistance are commonly implicated in human microlithiasis. In our patient population, the median BCS for dogs with grade 1 and 2 sludge was not significantly different from those graded 3–5. Unfortunately, BCS was not available for >2/3 of dogs, which limited statistical power. It is interesting to note that none of the dogs graded 3–5 were judged to be underweight (i.e., BCS ≤4).

All of the serum biochemical parameters evaluated (i.e., cholesterol, bilirubin, ALP, and GGT) were not apparently influenced by the extent of biliary sludging. As both ALP and GGT activity indicate cholestatic processes within the liver and biliary tree, it seems unlikely that sludge formation is necessarily related to compromised bile flow.¹⁸ Hypercholesterolemia is strongly associated with biliary sludging and cholelithiasis in people, and cholesterol is a primary constituent of human biliary sludge.⁵ The fact that serum cholesterol concentrations were not related to sludge severity in our patient population concurs with the findings of a similar study and supports the hypothesis that cholesterol is unlikely to be a major component of canine biliary sludge.¹⁶

Biliary sludge has been induced in dogs by cystic duct ligation or through dietary manipulation.^6,7 In both models, the sludge consisted primarily of bilirubinate and mucin, variably complexed with calcium carbonate. As this material coalesced into pigmentary gallstones, it may have limited similarities to spontaneous sludge. GB contents were analyzed in one Shetland sheepdog with a MC; the material was shown to have a similar composition to the experimentally induced sludge.¹⁹ However, as this breed has a strong genetic predisposition for defective phoshatidylcholine secretion and subsequent MC formation, this finding may not reliably reflect events in other breeds.¹¹

Both HAC and HT have been identified as a risk factor for MC formation in dogs.¹² In our study, the presence of HAC or HT was significantly associated with substantial GB sludge (odds ratio 5.04). This finding suggests a link between biliary sludge and the development of a MC. Various sources have proposed that the two are interconnected and that substantial sludge precedes the generation of a MC. This has not been proven, although sludge has been reported around or embedded within MCs.¹³ In experimental studies, glucocorticoids have been shown to alter bile acid profiles in dogs, with a shift towards the more irritant hydrophobic forms.²⁰ This change may impact GB epithelial function and mucus production. Subnormal thyroid levels have been associated with delayed emptying of bile into the intestine in people, and it has been suggested that chronic bile stasis may lead to the formation of sludge and excessive mucus secretion.²¹ Although 16 dogs in this study population were classified as HT, there was inadequate information regarding the initial diagnosis and the duration and efficacy of supplementation. One clinical report suggests that effective thyroid supplementation may result in a resolution of GB MC, but its impact on sludge has not been described in dogs.¹³ Prospective studies looking at GB contents at the time of diagnosis and following treatment are needed.

In this patient population, the administration of exogenous glucocorticoids for at least 4 wk did not appear to impact the formation of sludge. However, this finding may reflect a relatively short duration of administration in some dogs and substantial variation in glucocorticoid dosing. A recent study of dogs receiving high doses of exogenous glucocorticoids for 3 mo reported that all treated dogs (n = 6) developed GB sludge.⁸ However, this was not significantly different to the control group (n = 6), half of which also developed sludge over the same time period. Further work is needed to assess the impact of prolonged administration of exogenous glucocorticoids on GB status in dogs.

Although this study provides some useful insight into canine biliary sludge, it has several limitations. As this was an observational study, and not a prospective one, variables such as duration of fasting and administration of sedation prior to US evaluation could not be controlled. Sedation did not appear to impact US findings (% sludge, GB volume) in this patient population, but recent feeding may certainly influence these parameters. In addition, several radiologists participated in the grading process and may not have been consistent with their assessments. Conclusions regarding GB volume must be made cautiously, as volumes were estimated used a modification of a reported formula, the reliability of which has not been established.

For statistical purposes, dogs with >25% sludge (grades 3, 4, and 5) were considered together. This was necessary because of small group sizes but may have limited our ability to detect differences between groups regarding the biochemical parameters assessed. Larger studies may be needed to identify associations between routine clinicopathologic data and GB sludge in dogs.

One of the major limitations to our understanding of canine GB sludge is a lack of information regarding its composition. It is possible that different biochemical constituents are present in various dogs and may reflect varying etiologies. Additional prospective longitudinal studies with larger patient numbers are needed to establish the clinical impact of biliary sludge in dogs. In particular, the relationship (if any) between sludge and the development of the GB MC needs further investigation.

Conclusion

This study suggests that echogenic liminal material is routinely noted within the GB of dogs during transabdominal US examination at a veterinary teaching hospital, with echogenic contents reported in two-thirds of patients. Sludge formation was associated with increasing patient age but was not apparently influenced by patient weight or obesity. Dogs with clinical evidence of hepatobiliary disease were more likely to have hyperechoic GB contents than the general population, but the presence of substantial sludge was not associated with routine biochemical markers of biliary status (i.e., serum cholesterol and bilirubin concentrations; serum ALP or GGT activities). Similar to reports of dogs with MCs, biliary sludge was strongly associated with spontaneous HAC. Abnormal emptying was suggested by increased median GB volume in dogs with >25% sludge. Further studies are needed to determine the indications for intervention in dogs with biliary sludge and its possible connection to MC formation or other biliary tract disorders.