Prevalence and clinical relevance of cholelithiasis in cats: A multicenter retrospective study of 98 cases

2.1 Case selection criteria

A retrospective review of medical records from 3 veterinary teaching hospitals (Veterinary Teaching Hospital of the National Veterinary School of Alfort, Maison-Alfort, France; Veterinary Teaching Hospital of VetAgro Sup, Campus Vétérinaire de Lyon, France and Veterinary Teaching Hospital of Oniris, Nantes, France) was performed to search for cats diagnosed with cholelithiasis by US between January 2010 and December 2021. Cases were identified by searching the electronic patient database for the terms “cholelith,” “cholelithiasis,” “gallbladder stone,” “biliary stone,” and “bile stone.” Cats were excluded if cholelithiasis was detected by an imaging modality other than US, or if results of abdominal US were equivocal. Abdominal US was performed by a board-certified radiologist, a senior resident, or an advanced practitioner in diagnostic imaging.

2.2 Medical records review

Data collected from the medical records included signalment (age, breed, sex, body weight, and body condition score [BCS]), clinical signs (anorexia, vomiting, diarrhea, lethargy, and weight loss), physical examination findings (pyrexia as defined by a rectal temperature >39.2°C, jaundice, and abdominal pain) and results of clinicopathologic tests. Given that the study was multicentric, involving various analyzers and reagents for blood testing, results were recorded as “normal,” “increased” or “decreased” relative to each reference interval. For the most relevant biochemical variables (alanine aminotransferase [ALT]), alkaline phosphatase [ALP]), and bilirubin), a subcategory of “>2-fold increase” relative to their upper reference limit was added to consider their amplitude of increase for the multivariate analysis. Abdominal US reports were reviewed for estimated liver size, echogenicity, and homogeneity; gallbladder volume, wall thickness, and echogenicity; intrahepatic biliary duct, cystic duct, or common bile duct (CBD) dilatation and wall thickness; estimated pancreas size and echogenicity; pancreatic duct dilatation and wall thickness; and ileus and wall thickness of the digestive tract. The location and number of choleliths were recorded. Choleliths were considered as obstructive when biliary tract obstruction (BTO) was present (BTO was defined as dilatation of intrahepatic or extrahepatic biliary ducts or both [CDB dilatation ≥5 mm] reported upstream of the cholelith on US). Results of aerobic and anaerobic bacterial cultures (on bile, liver, or gallbladder samples), cytology (on bile or liver samples), and histology (on liver, gallbladder, pancreas, and digestive tract samples) were recorded.

For each case, cholelithiasis was classified as a “symptomatic” (SC) or an “incidental” (IC) finding. Cats were considered to be symptomatic for their cholelithiasis based on a combination of clinical signs (vomiting, diarrhea, abdominal pain, lethargy, anorexia, jaundice, or some combination of these), biliary tract US abnormalities (increased thickness or echogenicity of the biliary tract walls, dilatation of the gallbladder, cystic duct, CBD, or some combination of these), biochemical evidence of biliary disease (increased blood bilirubin concentration, or liver enzyme activities, or both), and if no other disease was identified to explain the clinical presentation.⁶ Final diagnosis or diagnoses (confirmed or presumptive) and concurrent diseases were recorded for all cases. Incidental cholelithiasis was considered in cats that did not have any clinical or biochemical findings attributable to cholelithiasis or when those changes could be attributed to another disease process and for which the biliary tract did not show other US abnormalities. Classification was supervised by a board-certified internist at each institution (Ghita Benchekroun, Emilie Krafft, and Juan Hernandez in veterinary teaching hospital of the national veterinary school of Alfort, veterinary teaching hospital of VetAgro Sup and Veterinary Teaching Hospital of Oniris, respectively). When classification of a case was not straightforward, an internist from another institution also reviewed the cat's medical record, and only cases for which both clinicians agreed were included. Treatments related to cholelithiasis were reviewed, including surgical procedures, use of ursodeoxycholic acid (UDCA), type and duration of antimicrobials, and administration of corticosteroids. Outcome, either survival to discharge or death during hospitalization (euthanasia or natural death), was evaluated.

When available, follow-up information was collected. Recurrence of clinical signs and development of complications attributed to cholelithiasis were recorded. Long-term outcome was recorded until last available follow-up or death.

2.3 Statistical analysis

Statistical analyses were performed using R 4.1.2 software (R Foundation for Statistical Computing, Vienna, Austria). Non-parametric statistics were used to describe the population. Data were examined for normal distribution using the Shapiro-Wilk test. The means of continuous normally-distributed variables were compared using a t test. Non-normally distributed continuous data were reported as medians and were compared using the Mann-Whitney U test. Biological findings were summarized as the number and percentage of cases with a measured variable above, within, or below the reference interval and were compared using Fischer's exact test. Categorical variables were expressed as number and percentage and compared using Pearson's χ² test or Fischer's exact test.

The overall prevalence of cholelithiasis was calculated by dividing the number of cats with evidence of cholelithiasis by the total number of cats that underwent abdominal US over the same time period.

Multivariate analysis was performed to determine potential risk factors for clinical expression of cholelithiasis and short-term survival. Data from all included cats were used to evaluate factors associated with SC. Short-term survival was defined as survival to discharge and was evaluated only in the population with SC. Because of the retrospective nature of the data leading to a considerable proportion of missing information in medical records, separate multivariate logistic regression models were fit for each of the 5 types of information collected (epidemiology, clinical signs, CBC, biochemistry, and US findings) to maximize the number of cats with no missing data in each multivariate model. All variables collected with <50% missing information were included in the corresponding complete multivariate logistic model and variable selections were performed using the MuMin (Multi-Model Inference) package in R with the automated model selection option (dredge) to calculate the corrected second-order Akaike information criterion (AICc) for all possible sub-models. The final model was chosen as the most biologically and medically relevant among all sub-models with a difference of <2 AICc compared with the 1 with the lowest AICc.²⁹

In the presence of separation (when the number of cats with certain risk factors was equal to 0, respectively, in SC or non-survivors), inference of the final multivariate model was performed using the logistf package³⁰; otherwise inference was performed using the glm function of R. Both linear models were checked for homoscedasticity and random distribution of residuals, and correct fit of the logistic regression was assessed using Pearson residual analysis. All variables were null-checked using a Wald test and considered significant at a P-value <.05.

For treatment effect, a final model then was constructed as mentioned above, including all significant risk factors for non-survival previously identified and the implementation of a medical or surgical treatment or both. To assess the impact of the inclusion of non-treated cats in the survival analysis, the corresponding models were run again after excluding them.

3.1 Population description and demographic data

Ninety-eight cats were included in the study, with 51 cats from Veterinary Teaching Hospital of the National Veterinary School of Alfort, 29 from Veterinary Teaching Hospital of VetAgro Sup, and 18 from Veterinary Teaching Hospital of Oniris. Over the study period, 9805 cats underwent abdominal US, which led to an overall prevalence of cholelithiasis of 0.99% (95% CI, 0.79-1.19). Cholelithiasis was classified as incidental (IC) in 40/98 (41%) and as symptomatic (SC) in 58/98 (59%) cats.

Demographic data are presented in Table 1 and breeds are given in Supplementary Materials S1. Age (P = .35), sex (P = 1), body weight (P = .47), and BCS (P = .92) were not statistically different between IC and SC on univariate analysis.

Final diagnosis in cases with SC is presented in Table 2. Forty-eight cats were diagnosed with concurrent hepatobiliary disease (either confirmed [n = 26] or presumptive [n = 22]) including cholangitis/cholangiohepatitis (CCH; n = 27; 56%), cholecystitis (n = 11; 23%), both CCH and cholecystitis (n = 9; 19%) and hepatic lipidosis (n = 1; 2%). Based on cytologic or histologic results or both, it was possible to further classify concurrent hepatobiliary disease in 16 cases: CCH was classified as neutrophilic in 11 cases, lymphocytic in 2, and mixed in 3 cases. Other concurrent diseases included pancreatitis in 6 cats (presumptive in 4 and confirmed in 2), or chronic enteropathy (CE) in 8 cats (presumptive in 6 and confirmed in 2). In 10 cats with SC, cholelithiasis was the sole diagnosis, of which it was obstructive (BTO) in 8 cases and non-obstructive in the other 2. Biliary tract obstruction was reported in 24 cases (43%) with SC. In cats with IC, diagnosis was achieved in 39 of 40 cases (98%) and is presented in Table 3.

Details about prior medications are provided in Supplementary Materials S2.

3.2 Clinical and biological data

Frequencies of clinical signs at presentation and physical examination findings are presented in Table 4. Lethargy, vomiting, anorexia, jaundice, and pyrexia were significantly more frequent in cats with SC than in cats with IC on univariate analysis (P < .05). Clinical pathology results are presented in Table 5. As anticipated based on the inclusion criteria, increased serum bilirubin concentration and liver enzyme activity, including ALT, ALP, and aspartate aminotransferase (AST), were more common in cats with SC than in cats with IC on univariate analysis (P < .05).

3.3 Abdominal ultrasound findings

Number and location of choleliths are reported in Table 6. According to imaging findings, choleliths most frequently were located within the gallbladder (69 cats; 70%), CBD (42 cats; 43%), cystic duct (24 cats; 24%), and intrahepatic biliary tract (22 cats; 22%). Cats with SC more often had multiple choleliths (P < .001) in multiple locations (P < .001) than did cats with IC on univariate analysis. Choleliths located within the CBD were more commonly symptomatic, compared with other locations (P < .001). Biliary tract obstruction was reported in 24 cats (41%) with SC, either alone (n = 8) or in combination with another hepatobiliary disease (n = 16), and in no cats with IC (P < .001).

Abdominal US findings are presented in Table 7. In addition to choleliths, biliary tract, gallbladder, liver, digestive tract, and pancreatic abnormalities were identified in 53 (54%), 64 (65%), 64 (65%), 46 (47%), and 43 (44%) cats, respectively. Ultrasonographic abnormalities of the biliary tract (SC, 44/58; IC, 9/40; P < .001), gallbladder (SC, 47/58; IC, 17/40; P < .001), liver (SC, 24/58; IC, 19/58; P < .01), and pancreas (SC, 31/58; IC, 12/40; P = .02) were more common in cats with SC compared with cats with IC on univariate analysis.

3.4 Microbiological and histopathological data

Twenty-nine cats (30%) underwent cholecystocentesis, of which 28 cases had SC and 1 had IC. Bile cytology was performed in 18 cats (18%) and bile cultures (both aerobic and anaerobic) in 28 (29%) cats. Bile cytology identified neutrophilic inflammation in 6/18 samples and bacteria in 8/18. Bacterial cultures were positive in 17/28 samples and identified 1, 2, or 3 bacterial species in 12, 4 and 1 samples respectively. Cultures were positive in all 8 samples for which bacteria were identified on cytology. Results of bacterial culture are presented in Table 8. For the single cat with IC that underwent cholecystocentesis and for 2/10 cats where the sole diagnosis was cholelithiasis, bile cytology was unremarkable, and culture was negative.

Liver, gallbladder, pancreatic, and intestinal biopsies were performed in 13, 4, 2, and 2 cats, respectively. Liver biopsy was performed in 2 cats with IC and no cats with a final diagnosis of only SC. Results are presented in Supplementary Materials S2.

3.5 Treatment

Among the 58 cats with SC, 33 (57%) received medical treatment alone and 13 (22%) required surgical intervention, combined (9 cats) or without (4 cats) medical management. Twelve cats (21%) did not receive any treatment, of which 7 were euthanized or died shortly after arrival and the others were discharged from the hospital. Among the cats with BTO, only 7 (29%) underwent surgical intervention, 10 (42%) received medical management alone, and 7 (29%) were not treated, including, respectively, 2, 2, and 4 cats without suspicion or confirmation of concurrent hepatobiliary disease.

Medical management included antimicrobials in 39 (93%), UDCA in 21 (50%), and corticosteroids in 12 (29%) cases. Medical treatments are detailed in Supplementary Materials S3. Surgeries performed included cholecystoenterostomy in 8 cats, temporary bile duct stenting in 5 cats, and cholecystectomy, cholecystotomy, and choledochotomy in 2 cats each. Two or more surgical procedures were performed in 5 cases. None of the choleliths removed were submitted for composition analysis.

Among the 40 cats with IC, only 4 (10%) were prescribed UDCA for a median of 4 weeks (range, 4-8 weeks) and none underwent surgical removal of the choleliths.

3.6 Outcomes

Forty-three cats (74%) with SC and 30 (75%) with IC survived to discharge. When focusing on cases with SC that were treated (n = 46), 38 (83%) survived to discharge, including 9/13 with surgical treatment and 29/33 with medical management (P = .2).

Among the 73 cats that survived to discharge, 40 (55%) had available follow-up, including 26 cats (60%) with SC (median follow-up time, 122 days; range, 46-4262 days) and 14 (47%) with IC (median follow-up time, 173 days; range, 4-464 days). Detailed data for short and long-term outcomes are presented in Figures 1 and 2. Of the 14 cats with IC that had follow-up available for review, none developed clinicopathological signs related to cholelithiasis. Among them, 5 had US re-evaluation of the biliary tract; choleliths were still present in 2 cats. In contrast, for cats with SC, recurrent or persistent clinicopathological signs were reported in 5/26 cases (2 surgically and 3 medically treated), of which 3 had at least 2 relapses. One cat underwent cholecystectomy, 2 cats were medically managed (UDCA in 2, antibiotic in 1 and prednisolone in 1) and 2 were euthanized on days 46 and 122.

3.7 Factors associated with clinical expression of cholelithiasis

The final multivariate model included age, male sex, low and high BCS for epidemiologic variables; vomiting, lethargy, pyrexia, and jaundice for clinical variables; lymphopenia, hyperglobulinemia, and increased ALP and ALT activities for biological variables; and liver, gallbladder, and biliary tract abnormalities, multiple locations, location within gallbladder, and obstruction for US variables (Table 9). High BCS remained the only epidemiologic variable significantly associated with an increased risk of SC (OR, 3.86; 95% CI, 1.13-13.16; P = .03). For clinical signs, vomiting (OR, 9.89; 95% CI, 2.65-36.94; P = .001), pyrexia (OR, 10.52; 95% CI, 1.49-74.22; P = .02), and jaundice (OR, 5.79; 95% CI, 1.03-32.45; P = .05) were significantly associated with SC. Among biological variables, only increased ALT activity (OR, 15; 95% CI, 4.05-81.77; P < .001) was associated with higher risk of being symptomatic. Multiple location of the choleliths within the biliary tract (OR, 8.11; 95% CI, 2.32-34.15; P = .001) and BTO (OR, 18.47; 95% CI, 2.13-2413.34; P = .004) were the only US variables associated with SC.

3.8 Factors associated with survival in cats with symptomatic cholelithiasis

The first multivariate model included age, male sex, low and high BCS for epidemiologic variables; weight loss and lethargy for clinical variables; neutropenia, neutrophilia, and increased serum creatinine concentration and ALP activity for biological variables; and obstruction for US variables (Table 10). Among all factors, only neutropenia (OR, 48.33; 95% CI, 2.52-8173.54; P = .01), neutrophilia (OR, 6.61; 95% CI, 1.13-71.22; P = .04), and BTO (OR, 4.14; 95% CI, 1.19-14.44; P = .03) were associated with increased risk of death before discharge and were included in the final model. In this last model that included treatment effects, only BTO remained associated with increased risk of death (OR, 13.87; 95% CI, 1.54-124.76; P = .001) whereas medical management was associated with decreased risk (OR, 0.02; 95% CI, 0-0.23; P = .001; Table 11). Surgical management did not significantly decrease the risk of death before discharge (OR, 0.79; 95% CI, 0.14-4.44; P = .43). Main risk factors of death before discharge remained similar when excluding non-treated cats from the analyses (Tables 12 and 13).