• canine;
  • fine needle aspiration;
  • liver cytology;
  • ultrasonography


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion

While abdominal ultrasound and ultrasound-guided fine-needle aspiration cytology are often combined to help determine the type of liver disease in dogs, little is known about the relationship that may exist between the results of these tests. We hypothesized that specific sonographic findings, or combinations of findings, may predict results of liver ultrasound-guided fine-needle aspiration cytology. Hepatic and extrahepatic sonographic findings were recorded prospectively using a standardized form in 70 dogs with clinically suspected liver disease and in which liver ultrasound-guided fine-needle aspiration cytology was performed. The predictive value of sonographic findings in regard to the category of cytology results was assessed with stepwise logistic regression analysis. Sonographic detection of a hepatic mass (≥3 cm; risk ratio [RR] 3.83, 95% Wald confidence intervals [95% CI] 2.42–3.93, P=0.0036), ascites (RR 3.82, 95% CI 1.94–4.28, P=0.0044), abnormal hepatic lymph node(s) (RR 3.01, 95% CI 1.22–4.88, P=0.0262), and abnormal spleen (RR 3.26, 95% CI 1.20–3.85, P=0.0274) were the most predictive of liver neoplasia on cytology. Conversely, sonographic detection of hepatic nodules (<3 cm; RR 1.97, 95% CI 0.95–2.96, P=0.0666) was most predictive of vacuolar hepatopathy on cytology. In dogs with suspected liver disease, several sonographic findings, alone or combined, are thus predictive of liver ultrasound-guided fine-needle aspiration cytology results. In the light of the fact that ultrasound-guided fine-needle aspiration cytology of the liver has limitations, these predictabilities could influence the selection of diagnostic tests to reach a reliable diagnosis.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion

Abdominal ultrasound has a significant role in the diagnosis of canine liver disease.1–4 While specific sonographic signs have been described for certain hepatic diseases,5–15 little is known about the overall value of abdominal ultrasound in that context. In 22 dogs with liver disease in which a definitive diagnosis was obtained, that diagnosis was predicted with ultrasound in only 11 (50%).3 Also, a significant association between sonographic signs assessed on still images and types of diffuse liver disease was not identified in 229 dogs.16

Ultrasound-guided fine-needle aspiration cytology is widely used for evaluation of canine hepatic disease, even if its accuracy, when compared with histopathology, is controversial.17–21 The diagnostic value of cytologic samples is limited by the inability to assess tissue architecture and vascularity, as well as the poor ability to detect inflammation.17,20,22 Furthermore, samples can be contaminated with blood, particularly when suction is used.20

Although liver fine-needle aspiration cytology is routinely used in combination with sonographic findings in the diagnosis of hepatic diseases in dogs, the relationship that may exist between the results of these tests remains undetermined. It is possible that sonographic signs, or combinations of signs, may predict the results of liver fine-needle aspiration cytology. To test this hypothesis, we examined prospectively the relationship between specific abdominal sonographic signs and results of fine-needle aspiration cytology in dogs with suspected liver disease.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion

Inclusion criteria for this study were dogs with suspected hepatic disease, based on clinicopathologic data that underwent abdominal ultrasonography and fine-needle aspiration cytology of the liver between October 2005 and April 2006. Exclusion criteria were dogs with suspected congenital portosystemic shunt, as this abnormality cannot be assessed with cytology;2 or those receiving glucocorticoids, as these induce hepatic vacuolar changes that can interfere with sonographic evaluation of focal lesion(s).23

Seventy dogs of several different breeds and conformations were studied. There were 26 females and 44 males. Age ranged from 8 months to 15 years, with a mean and standard deviation (SD) of 9.1±3.2 years. Twelve (17.1%) had no serum chemistry abnormality. These were dogs undergoing staging of tumors that commonly spread to the liver, dogs with clinical signs such as icterus that were eventually attributed to extrahepatic disease, and dogs for which laboratory data were available only after the ultrasound exam and fine-needle aspiration procedures. The other 58 (82.9%) had at least one of the following: hypoalbuminemia, increased alkaline phosphatase, alanine aminotransferase, gammaglutamyl-transferase, total bilirubin, and/or cholesterolemia.

Ultrasound examinations were performed by a radiologist (M.A.D., K.A., or G.B.). Dogs were in dorsal recumbency and restrained manually, with tranquilization when necessary. Hair was clipped from the xyphoid process to the pubis, and included the most caudal intercostal spaces when the liver could not be fully assessed using a subcostal approach. All examinations were performed using high-resolution ultrasound equipment and a 5–8 MHz curvilinear transducer.* Image presets, overall gain, time gain compensation, and number and location of focal zones were adjusted for each patient and during each examination to optimize image quality. A separate standardized form was used for each dog to record and grade every sonographic sign systematically. The hepatic volume was described qualitatively as normal, increased, or decreased, considering its extent in regard to the costal arch, the position of the organs in the cranial abdomen as well as taking into account body conformation.24,25 Liver contour was recorded as normal, rounded, or irregular. The hepatic parenchyma was also characterized by its uniformity (homogeneous or heterogeneous), its echogenicity (normal, increased, decreased, or mixed), and its attenuation (qualitatively normal or increased). The qualitative assessment of liver echogenicity was based on its relative echogenicity with the spleen, renal cortices, and falciform fat; and on the visibility of intrahepatic portal markings. Focal hepatic lesions were classified as nodule (diameter <3 cm), mass (diameter equal to or exceeding 3 cm), or cyst (well-defined, anechoic, nearly spherical lesion with acoustic enhancement). The echogenicity of solid focal lesions was noted as hypoechoic, isoechoic, or hyperechoic relative to the surrounding parenchyma. The biliary tract was classified as abnormal if one or more abnormalities were detected, such as intrahepatic bile duct dilation, common bile duct exceeding 3 mm,26 gallbladder mucocele, cholelithiasis, or gallbladder wall thickness exceeding 3 mm.27 Hepatic lymph nodes and spleen were classified subjectively as normal or abnormal based on their size, shape, contour, echogenicity, uniformity, and echotexture, including the presence of focal abnormalities. Finally, the presence or absence of ascites was recorded.

Fine-needle aspiration cytology was performed with a 21 G needle attached or not to an air-filled 10 ml syringe, with or without an extension tube.28,29 The aspiration technique was performed without suction unless the initial sample seemed poorly cellular, in which instance suction was also performed. The use of suction was avoided whenever possible to minimize dilution of the sample with blood. Blood coagulation was assessed before fine-needle aspiration cytology in several dogs and sedation use was variable, but these data were not collected specifically. Fine-needle aspiration cytology samples were obtained from the liver parenchyma and if present, separately, from focal lesions. At least two or three aspirates were collected for each site to obtain at least three cytosmears of moderate to high cellularity, based on visual assessment. Air-dried slides were stained using a modified Wright–Giemsa stain and examined by a clinical pathologist (C.B. or M.D.). Slides with excessive blood contamination were not interpreted. Cytology results were classified into five categories: normal, vacuolar hepatopathy (mild, moderate, marked, or severe), lipidosis (alone or together with vacuolar hepatopathy), inflammation (any type of inflammation alone or together with vacuolar hepatopathy), and neoplasia (primary or metastatic neoplasm). When results included both neoplastic and nonneoplastic changes, only neoplastic changes were considered for the classification. Similarly, inflammation was recorded as the only finding in dogs with additional nonneoplastic changes, e.g. vacuolar hepatopathy.

Because of only two dogs with lipidosis, this category was excluded from statistical analysis. Logistic regression analysis30 was performed to examine the sonographic determinants of the presence of the remaining cytologic anomalies: vacuolar hepatopathy, inflammation, and neoplasia. These outcomes were examined separately by modeling the log-odds of a positive finding. The following sonographic signs were each used as predictor variables: altered hepatic volume, shape, parenchyma, echogenicity, and attenuation, presence of focal lesion(s), ascites, abnormal biliary tract, hepatic lymph nodes, and spleen; as well as all possible dual interactions of these 10 variables. The logistic regression model was built by testing the statistical relevance of the above predictors with a stepwise-forward algorithm, using a threshold of P=0.15 for including these predictors in the model, and, following this step, a threshold of P=0.20 for their subsequent removal.30 An interaction of sonographic predictors was considered in the model only when both of them were already individually included in the model. Predictors were ineligible for removal from the model as long as they were involved in a significant interaction already included in the model. We additionally performed residual analysis to verify the validity of significant relationships.

Afterwards, the odds ratio (OR) and associated 95% Wald confidence interval (95% CI) of statistically significant predictors of each cytologic anomaly were calculated (not shown),27 taking into account the presence of a significant interaction, i.e. the global OR is the sum of the marginal and interaction effects. Then, because the prevalence of each disease category was relatively high (>10%) in the study population, risk ratio (RR) were calculated using a standard approach.31


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion

The distribution of dog cytologic categories and sonographic features are compiled in Table 1. The presence of a hepatic mass(es), ascites, abnormal hepatic lymph nodes, abnormal spleen, interaction of mass with abnormal biliary tract, interaction of mass with abnormal hepatic lymph nodes, and abnormal biliary tract were predictive of neoplasia (Table 2), and certain sonographic determinants: ascites, hepatic nodule(s), abnormal hepatic lymph nodes, and interaction of abnormal hepatic lymph nodes with ascites were predictive of vacuolar hepatopathy (Table 3).

Table 1.   Sonographic Findings in 70 Dogs Distributed in Function of Categories of Results of Ultrasound-Guided Fine-Needle Cytology
Sonographic FindingsCytological Classification (Number of Dogs)
Normal (7)Vacuolar Hepatopathy (23)Inflammatory (15)Neoplasia (23)Lipidosis (2)
  • *

    An individual dog could have more than one ultrasonographic finding in these categories.

  • †A splenectomy had been performed in one of the dogs.

Hepatic volume
Hepatic shape*
Hepatic parenchyma
Hepatic echogenicity
Liver attenuation
Hepatic focal lesions*
Hypoechoic nodules011662
Hyperechoic nodules03032
Hypoechoic masses00100
Hyperechoic masses01150
Mixed masses01040
Not detected79760
Hepatic lymph node
Biliary tract
Not detected52213122
Table 2.   Logistic Regression Estimates, Associated Risk Ratios, and Statistical Predictors of a Hepatic Neoplasia as Found with Ultrasound-Guided Fine-Needle Aspiration Cytology
Sonographic PredictorsMaximum Likelihood Estimate (Standard Error)Risk Ratio (95% Wald Confidence Limits)P-Value
Mass (solid focal lesion ≥3 cm)4.7333.830.0036
(1.625)(2.42, 3.93) 
(1.117)(1.94, 4.28) 
Abnormal hepatic LN2.1183.010.0262
(0.953)(1.22, 4.88) 
Abnormal spleen1.9273.260.0274
(0.873)(1.20, 3,85) 
Mass × abnormal biliary tract−4.1560.530.0382
(2.007)(0.03, 2.10) 
Mass × abnormal hepatic LN−3.4183.650.0986
(2.070)(1.001, 3.98) 
Abnormal biliary tract−1.3980.360.1800
(1.043)(0.05, 1.50) 
Table 3.   Logistic Regression Estimates, Associated Risk Ratios, and Statistical Predictors of Vacuolar Hepatopathy as Found with Ultrasound-Guided Fine-Needle Aspiration Cytology
Sonographic PredictorsMaximum Likelihood Estimate (Standard Error)Risk Ratio (95% Wald Confidence Limits)P-Value
(1.122)(0.01, 0.79) 
(0.602)(0.95, 2.96) 
Abnormal hepatic LN−1.6290.290.0689
(0.896)(0.06, 1.08) 
Abnormal hepatic LN × Ascites2.5430.280.1559
(1.792)(0.11, 0.64) 

The following hypothetic clinical situations illustrate our results and the use of RR. When compared with dogs with none of the significant predictors listed in Table 2, dogs with a hepatic mass (focal solid lesion ≥3 cm in diameter) are 3.83 (CIRR=[2.42–3.93]) times at risk of having cytologic findings consistent with neoplasia. This risk of finding neoplasia on cytology is similar for dogs in which a hepatic mass and ascites are both identified (RR=3.44; CIRR=[3.23–3.45]), but are lower for dogs with a combination of a hepatic mass, ascites, and abnormal hepatic lymph nodes (RR=3.13; CIRR=[2.66–3.14]). The risk for dogs with ascites, and abnormal hepatic lymph nodes and spleen (RR=3.31; CIRR=[3.07–3.32]) is similar to dogs with mass and ascites.

When compared with dogs having none of the significant predictors listed in Table 3, dogs with nodule(s) (focal solid lesion <3 cm in diameter) are 1.97 times (CIRR=[0.95–2.96]) more likely to have cytologic findings consistent with vacuolar hepatopathy. However, dogs with hepatic nodule(s) and abnormal hepatic lymph nodes are 1.45 times (CIRR=[0.62–29.84]) less at risk of having cytologic findings consistent with vacuolar hepatopathy than dogs without sonographic predictors, and this risk is similar for dogs with a combination of nodule(s), abnormal hepatic lymph nodes, and ascites (RR=1.51; CIRR=[0.43–14.62]).

No sonographic sign was found to be predictive of inflammation on cytology and no target lesions were encountered in our study.

Of the nine neoplastic masses (carcinoma, n=5; hemangiosarcoma, n=2; undifferentiated sarcoma, n=1; hepatocellular adenoma, n=1), concurrent splenic abnormalities were found in six, ascites in five, a combination of splenic anomaly and ascites in four, and hepatic lymphadenopathy in two. Neoplasia was diagnosed with cytology in six other dogs in the absence of a focal lesion (lymphoma, n=4; histiocytic sarcoma, n=1; and metastatic carcinoma of unknown origin, n=1). Of these six dogs, five had a normal liver volume while one had hepatomegaly (lymphoma). Four were hypoechoic (lymphoma,2 histiocytic sarcoma, and metastatic carcinoma), one was hyperechoic (lymphoma), and one was normal (lymphoma). Of these same dogs with diffuse liver neoplasia, other findings included abnormal lymph nodes (5), abnormal spleen (5), or both of these findings (4).

Twelve dogs with cytologic findings consistent with vacuolar hepatopathy had hepatic nodules. Nine of these dogs (75%) had hypoechoic nodule(s), while one dog (11%) had hyperechoic nodule(s), and two (22%) had a combination of hypoechoic and hyperechoic nodules.

Seven dogs had a normal liver on cytology, of which four (57%) had no sonographic liver abnormality, two (28%) had a small liver, and one (14%) had a small, irregular liver with a hypoechoic and heterogeneous parenchyma, as well as biliary tract abnormalities. On the same dogs, there was no sign of focal lesion or abnormal hepatic lymph nodes.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion

In this prospective study, no sonographic sign was pathognomonic for a given cytologic diagnosis. However, there was a significant relationship between several sonographic findings or combinations of findings and cytologic features of the presence, or absence, of hepatic neoplasia. In general, there was not a significant association between sonographic findings and cytologic findings consistent with diffuse liver disease.

Results from fine-needle aspiration cytology of the liver are usually classified into eight diagnostic categories: normal, inflammatory, neoplastic, lipidosis, vacuolar hepatopathy, hyperplastic, hematopoietic tissue, and pigment abnormalities.2,20,32 In our study, five types of abnormalities at the organ level (normal, inflammatory, neoplastic, lipidosis, and vacuolar hepatopathy) were considered; hyperplasia, hematopoiesis, and pigment accumulation are most often combined with the other processes and were therefore not considered as separate entities. In addition, mixed reactions were classified according to the most clinically important finding, i.e. neoplasia vs. inflammation.2

The sonographic detection of a liver mass (focal solid lesion ≥3 cm) was most predictive of neoplasia. In our population of nine dogs with a hepatic mass, carcinoma was diagnosed with cytology in five (56%). This prevalence is lower than expected based on data from 15 dogs with hepatic neoplasia where 14 (93%) had carcinoma.12 Also those 14 carcinomas were hyperechoic or mixed in echogenicity, which suggested those sonographic patterns were specific for carcinoma.12 Similarly, in our dogs, all carcinomas were hyperechoeic or mixed focal masses, although these changes were also found with hemangiosarcoma (mixed), hepatocellular adenoma (hyperechoic), and undifferentiated sarcoma (hyperechoic), which is consistent with other studies.3,8,15 Hence, although hepatic masses tend to be hyperechoic or mixed in echogenicity and carcinoma appears to be the most common liver mass, these signs are not specific to this group of malignancies.

Ascites was another sign that predicted cytologic evidence of hepatic neoplasia, similar to what has been found in cats.33 Ascites associated with hepatic neoplasia is mainly due to vasculitis, capillary obstruction or bleeding, resulting in modified transudate, chyle, or hemoperitoneum.34,35 Ascites is also a frequent sign of cirrhosis and hypoprotenemia.4,34 As fine-needle aspiration of the liver is not typically performed when cirrhosis is suspected, this might have biased our results toward an overestimation of ascites as a predictor of neoplasia. For this reason, our results apply only to dogs in which fine-needle aspiration cytology is indicated and not to the canine population with liver disease as a whole.

The presence of abnormal hepatic lymph nodes or an abnormal spleen predicted cytologic evidence of hepatic neoplasia. In dogs, hepatic metastases are almost as frequent as primary hepatic tumors, and primary hepatic tumors can metastasize to regional lymph nodes, lung and other abdominal organs.35 In dogs with various types of hepatic neoplasia, changes to the hepatic lymph nodes and spleen were observed commonly.8,9,13 Similarly, in cats with suspected liver disease, the presence of splenic or lymph node abnormalities suggested hepatic neoplasia.35 Despite this, the additional sonographic identification of abnormal hepatic lymph nodes in dogs with a hepatic mass did not increase the risk ratio of finding hepatic neoplasia. This likely reflects the lack of specificity associated with the sonographic detection of abnormal hepatic lymph nodes.

There was a statistically significant negative association between ascites, and/or abnormal hepatic lymph nodes and cytologic evidence of vacuolar hepatopathy. In dogs not treated with glucocorticoids, or affected with hyperadrenocorticism, vacuolar changes are often associated with a primary liver condition such as neoplasia or other acquired hepatobiliary disease.23 The physiologic stress associated with extrahepatic diseases can also induce vacuolar hepatopathy.23 Ascites and lymphadenopathy are not expected with hyperadrenocorticism,36 consistent with our findings in dogs with hepatic vacuolar hepatopathy.

Among all sonographic signs investigated in our dogs, none was significantly predictive of hepatic inflammation. Our inability to detect any significant predictor of inflammation at the fine-needle aspiration cytology may reflect the difficulty of this specific cytological diagnosis.17,20

In one dog, the liver was cytologically normal yet sonographically abnormal. Although, it is possible that sonographic signs were erroneously recognized in this dog, limitations in the technique of fine-needle aspiration cytology must be emphasized. Indeed, focal lesions can be challenging to target with ultrasound, because of their size and/or location. Furthermore, normal or abnormal liver cells located along the path of needle insertion can be included in the sample. Cells corresponding to a focal lesion, regardless of its depth, can therefore be diluted in the smear, or even absent (i.e., nonexfoliating lesion), resulting in a cytologic diagnosis that may inaccurately represent the targeted lesion.

There were some limitations of our study. First, as only one fine-needle aspiration cytology diagnosis was recorded for each liver, it is possible that the cytologic diagnosis did not relate to some of the sonographic changes recorded. For example, a targeted neoplastic mass might have been associated with overall liver parenchymal hyperechogenicity that was due to a vacuolar hepatopathy. In this instance, the fine-needle aspiration cytology result was neoplasia while the parenchymal changes caused by the concurrent vacuolar hepatopathy remained included in the list of detected signs for this particular dog.

Secondly, abdominal ultrasound is an examiner-dependent modality that can be associated with diagnostic variability.16 For instance, the presence of multiple ill-defined foci of abnormal echogenicity may be interpreted by one sonographer as nodules, but as parenchymal heterogeneity by another, a situation that yields different diagnostic results. In this study, to minimize the variation associated with examiner-dependent liver assessment, all sonographers used a standardized form with fixed criteria for each exam.

Another limitation is associated with the number of available dogs. Some types of sonographic features with low prevalence had to be withdrawn from the database, e.g. hepatic lipidosis, or be grouped into one statistical code, e.g. nodules, to achieve sufficient statistical power. Including more dogs with variable degrees of disease severity might have increased the level of significance and predictive value of some of these signs.

Finally, the use of cytology to categorize liver disease rather than histopathology may be questioned, as the agreement between these two tests ranges between approximately 30%19 and 60%.2,18 Part of this difference may be explained by the nature of the disease. Indeed, vacuolar hepatopathy, malignant neoplasia, and suppurative hepatitis, which were the diseases evaluated herein, can be more reliably diagnosed than nodular hyperplasia, adenoma, chronic inflammation, and fibrosis, which are difficult to identify on cytology.2,17,20,32 Despite its limitations, ultrasound-guided fine-needle aspiration cytology remains in frequent use in dogs because of its simplicity, rapidity, and safety,20,22 as opposed to hepatic biopsy that requires general anesthesia, is more expensive, and is associated with potential life-threatening complications.37,38 Ultrasound-guided fine-needle aspiration cytology techniques are also used routinely in humans,39–41 helping to differentiate benign from malignant lesions with high accuracy (96%), as well as accurately identify neoplasia type (93%).40 It must also be clearly pointed out that the objective of the present study was not to determine the accuracy of ultrasound-guided fine-needle aspiration cytology, but rather to identify its relationship with sonographic findings.

To conclude, in dogs with suspected liver disease, several sonographic findings are predictive of specific liver ultrasound-guided fine-needle aspiration cytology results. The predictability of these sonographic findings may guide clinicians to select appropriate tests that should be performed to reach a reliable diagnosis, including or excluding fine-needle aspiration cytology in light of its recognized limitations. Additional prospective studies using histology as gold-standard are warranted to clearly establish the role of specific sonographic signs in the detection and categorization of liver diseases in dogs.

  1. *Philips HDI 5000, Koninklijke Philips Electronics N.V., Amsterdam, the Netherlands.

  2. †SAS, system, version 9.1, SAS Institute Inc., Cary, NC.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion

The authors wish to thank Maxim Moreau (research assistant), Drs. Benoit Rannou and Caroline Piché (clinical pathology residents) for their contribution in the data collection and analysis.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  • 1
    Biller DS, Blackwelder T. Hepatic ultrasonography: a valuable tool in small animals. Vet Med 1998;93:646653.
  • 2
    Kristensen AT, Weiss DJ, Klausner JS, Hardy RM. Liver cytology in cases of canine and feline hepatic disease. Compend Contin Educ Pract Vet 1990;12:797808.
  • 3
    Vörös K, Vrabély T, Papp L, Horvàth L, Karsai F. Correlation of ultrasonographic and pathomorphological findings in canine hepatic diseases. J Small Anim Pract 1991;32:627634.
  • 4
    Webster CRL. History, clinical signs, and physical findings in hepatobiliary disease. In: EttingerSJ, FeldmanEC (eds): Textbook of veterinary internal medicine, Vol. 2. St. Louis, MO: Elsevier Saunders, 2005;14221482.
  • 5
    Schwartz SG, Mitchell SL, Keating JH, Chan DL. Liver lobe torsion in dogs: 13 cases (1995–2004). J Am Vet Med Assoc 2006;228:242247.
  • 6
    Besso JG, Wrigley RH, Gliatto JM, Webster CRL. Ultrasonographic appearance and clinical findings in 14 dogs with gallbladder mucocele. Vet Radiol Ultrasound 2000;41:261271.
  • 7
    Cuccovillo A, Lamb CR. Cellular feautures of sonographic target lesions of the liver and spleen in 21 dogs and a cat. Vet Radiol Ultrasound 2002;43:275278.
  • 8
    Lamb CR, Hartzband LE, Tidwell AS, Pearson SH. Ultrasonographic findings in hepatic and splenic lymphosarcoma in dogs and cats. Vet Radiol 1991;32:117120.
  • 9
    Sato AF, Solano M. Ultrasonographic findings in abdominal mast cell disease: a retrospective study of 19 patients. Vet Radiol Ultrasound 2004;45:5157.
  • 10
    Schwarz LA, Penninck DG, Leveille-Webster C. Hepatic abscesses in 13 dogs: a review of the ultrasonographic findings, clinical data and therapeutic options. Vet Radiol Ultrasound 1998;39:357365.
  • 11
    Stowater JL, Lamb CR, Schelling SH. Ultrasonographic features of canine hepatic nodular hyperplasia. Vet Radiol 1990;31:268272.
  • 12
    Whiteley MB, Feeney DA, Whiteley LO, Hardy RM. Ultrasonographic appearance of primary and metastatic canine hepatic tumors. A review of 48 cases. J Ultrasound Med 1989;8:621630.
  • 13
    Cruz-Arámbulo R, Wrigley R, Powers B. Sonographic features of histiocytic neoplasms in the canine abdomen. Vet Radiol Ultrasound 2004;45:554558.
  • 14
    Chapman BL, Hendrick MJ, Washabau J. Granulomatous hepatitis in dogs: nine cases (1987–1990). J Am Vet Med Assoc 1993;203:680684.
  • 15
    Waters CB, Morrison WB, DeNicola DB, Widmer WR, White MR. Giant cell variant of malignant fibrous histiocytoma in dogs: 10 cases (1986–1993). J Am Vet Med Assoc 1994;205:14201424.
  • 16
    Feeney DA, Anderson KL, Ziegler LE, Jessen CR, Daubs BM, Hardy RM. Statistical relevance of ultrasonographic criteria in the assessment of diffuse liver disease in dogs and cats. Am J Vet Res 2008;69:212221.
  • 17
    Weiss DJ, Blauvelt M, Aird B. Cytologic Evaluation of Inflammation in Canine Liver Aspirates. Vet Clin Pathol 2001;30:193196.
  • 18
    Roth L. Comparison of liver cytology and biopsy diagnoses in dogs and cats: 56 cases. Vet Clin Pathol 2001;30:3538.
  • 19
    Wang KY, Panciera DL, Al-Rukibat RK, Radi ZA. Accuracy of ultrasound-guided fine-needle aspiration of the liver and cytologic findings in dogs and cats: 97 cases (1990–2000). J Am Vet Med Assoc 2004;224:7578.
  • 20
    Weiss DJ, Moritz A. Liver cytology. Vet Clin North Am Small Anim Pract 2002;32:12671291.
  • 21
    Menard M, Papageorges M. Ultrasound-guided liver fine needle biopsies in dogs; results of 600 cases. Vet Pathol 1996;33:570.
  • 22
    Stockhaus C, Ingh Tvd, Rothuizen J, Teske E. A multistep approach in the cytologic evaluation of liver biopsy samples of dogs with hepatic diseases. Vet Pathol 2004;41:461470.
  • 23
    Sepesy LM, Center SA, Randolph JF, Warner KL, Erb HN. Vacuolar hepatopathy in dogs 336 cases (1993–2005). J Am Vet Med Assoc 2006;229:246252.
  • 24
    D'Anjou MA. Liver. In: PenninckDG, D'AnjouMA (eds): Atlas of small animal ultrasonography. Ames, IA: Wiley-Blackwell, 2008;217262.
  • 25
    Nyland TG, Mattoon JS, Herrgesell EJ, Wisner ER. Liver. In: NylandTG, MattoonJS (eds): Small animal diagnostic ultrasound, 2nd edn. Philadelphia: Mosby-Saunders, 2002;93127.
  • 26
    Zeman RK, Taylor KJ, Rosenfield AT, Schwartz A, Gold JA. Acute experimental biliary obstruction in the dog: sonographic findings and clinical implications. Am J Roentgenol 1981;136:965967.
  • 27
    Spaulding KA. Ultrasound corner: gallbladder wall thickness. Vet Radiol 1993;34:270272.
  • 28
    Papageorges M, Gavin PR, Sande RD, et al. Ultrasound-guided fine needle aspiration: an inexpensive modification of the technique. Vet Radiol 1998;29:269271.
  • 29
    Menard M, Papageorges M. Ultrasound corner: technique for ultrasound-guided fine needle biopsies. Vet Radiol Ultrasound 1994;36:137138.
  • 30
    Hosmer DW, Lemeshow S. Applied logistic regression. New York: John Wiley & Sons Inc, 1989.
  • 31
    Zhang J, Yu KF. What is the relative risk? A method for correcting the odds ratio in cohort studies of common outcomes. J Am Med Assoc 1998;230:16901691.
  • 32
    Raskin RE. Liver cytology: interpreting needle biopsy samples. Vet Med 2000;95:244249.
  • 33
    Newell SM, Selcer BA, Girard H, Roberts GD, Thompson JP, Harrison JM. Correlations between ultrasonographic findings and specific hepatic diseases in cats: 72 cases (1985–1997). J Am Vet Med Assoc 1998;213:9498.
  • 34
    Kruth SA. Abdominal distension, ascites, and peritonitis. In: EttingerSJ, FeldmanEC (eds): Textbook of veterinary internal medicine, Vol. 1. St. Louis, MO: Elsevier Saunders, 2005;150153.
  • 35
    Scherk MA, Center SA. Toxic, metabolic, infectious, and neoplastic liver disease. In: EttingerSJ, FeldmanEC (eds): Textbook of veterinary internal medicine, Vol. 2. St. Louis, MO: Elsevier Inc, 2005;14641477.
  • 36
    Reush CE. Hyperadrenocorticism. In: EttingerSJ, FeldmanEC (eds): Textbook of veterinary internal medicine, Vol. 2. St. Louis, MO: Elsevier Saunders, 2005;15921611.
  • 37
    Bigge LA, Brown DJ, Penninck DG. Correlation between coagulation profile findings and bleeding complications after ultrasound-guided biopsies: 434 cases (1993–1996). J Am Anim Hosp Assoc 2001;37:228233.
  • 38
    Léveillé R, Partington BP, Biller DS, Miyabayashi T. Complications after ultrasound-guided biopsy of abdominal structures in dogs and cats: 246 cases (1984–1991). J Am Vet Med Assoc 1993;203:413415.
  • 39
    Bret PM, Sente JM, Bretagnolle M, Fond A, Labadie M, Paliard P. Ultrasonically guided fine-needle biopsy in focal intrahepatic lesions: six years' experience. Can Assoc Radiol J 1986;37:58.
  • 40
    Ding W, He X-J. Fine needle aspiration cytology in the diagnosis of liver lesions. Hepatobiliary Pancreat Dis Int 2004;3:9092.
  • 41
    Sbolli G, Fornari F, Civardi G, et al. Role of ultrasound guided fine needle aspiration biopsy in the diagnosis of hepatocellular carcinoma. Gut 1990;31:13031305.