Continuing Education: Liver Enzyme Interpretation and Liver Function Tests - Today's Veterinary Practice

Continuing Education: Liver Enzyme Interpretation and Liver Function Tests

May/June 2017   •   (Volume 7, Number 3)

Brigitte B. McAtee, DVM
Jonathan A. Lidbury, BVMS, MRCVS, PhD, DACVIM, DECVIM-CA
Texas A&M University

Hepatobiliary disease is an important cause of morbidity and mortality in dogs and cats and can present a diagnostic challenge for two main reasons. First, patient signalment varies because liver disease and dysfunction can occur in cats and dogs of any age, sex, or breed (see Case Studies). Despite this, the patient’s signalment can sometimes give important clues because certain breeds have disease predispositions; for example, Labrador retrievers are predisposed to copper-associated chronic hepatitis. Second, elevations of serum liver enzyme activities are commonly encountered in small animal practice but are not specific for primary liver disease. However, early in the course of liver diseases, such as chronic hepatitis, patients may have no or only subtle, nonspecific clinical signs, such as intermittent anorexia or lethargy. In these patients, increased liver enzyme activities may be the first indicator of a problem. More liver-specific clinical signs, such as icterus, ascites, edema, polyuria/polydipsia, and hepatic encephalopathy, tend to occur late in the course of disease, when it is often too late to prevent its progression. Therefore, early diagnosis of liver disease often relies on serum biochemical testing, which may prompt further diagnostics, including liver function testing. This article reviews the interpretation and limitations of serum liver enzyme activity and liver function tests.

Case Studies

Case 1

Signalment and Presentation

A 3-month-old female intact Irish wolfhound presents for stunted growth and episodes of intermittent lethargy and disorientation.

Results of Diagnostic Testing

A serum biochemistry panel is performed, with the results in Table A. The fasted ammonia concentration is 175 mcg/dL (normal range, 0–50 mcg/dL). Preprandial and postprandial (2-hour) SBA are 40 mcmol/L (normal, 0–8 mcmol/L) and 102 mcmol/L (normal, 0–30 mcmol/L), respectively.

Interpretation

The combination of hypoalbuminemia, decreased BUN, and hypocholesterolemia suggests decreased hepatic synthetic capacity. The ALT and AST activities are within normal limits, making hepatocellular damage unlikely; the ALP activity is only mildly elevated, probably because the dog is growing.

The ammonia concentration and SBA results suggest portosystemic shunting and/or hepatic insufficiency.

Given the patient’s signalment, clinical findings, and laboratory abnormalities, a congenital portosystemic shunt is likely and imaging (ultrasonography and/or computed tomography) is warranted.

Case 2

Signalment and Presentation

An 8-year-old male neutered Labrador retriever presents for a 3-month history of decreased appetite and weight loss.

Results of Diagnostic Testing

A serum biochemistry panel is performed, with the results in Table B. The fasted ammonia concentration is <15 mcg/dL (normal range, 0–50 mcg/dL). Preprandial and postprandial (2-hour) bile acids are 2.9 mcmol/L (normal, 0–8 mcmol/L) and 14.5 mcmol/L (normal, 0–30 mcmol/L), respectively.

Interpretation

The ALT activity is 2.4 times the upper limit of the reference interval, while the ALP activity is only 1.3 times the upper limit of the reference interval. This, along with the increased serum AST activity, is consistent with a hepatocellular damage pattern.

The ammonia concentration and SBA results rule out portosystemic shunting and do not support the presence of severe liver dysfunction. However, hepatobiliary disease is not excluded and further testing is indicated.

Abdominal ultrasonography would be a logical next step. If the ALT is persistently increased and no evidence supports the presence of extrahepatic disease, liver biopsy would be indicated.

LIVER DISEASE AND DYSFUNCTION can occur in cats and dogs of any age, sex, or breed.

BACKGROUND

The liver has a wide variety of metabolic functions (Box 1). Because of these diverse metabolic roles, liver dysfunction is associated with a variety of sequelae and clinicopathologic abnormalities.

The liver is unique in that it receives much of its blood supply (75%) from the portal venous system, which drains abdominal organs, such as the gastrointestinal (GI) tract, spleen, and pancreas.1,2 This means that diseases of the pancreas and GI tract can secondarily affect the liver. The liver also metabolizes and/or excretes a variety of exogenous substances (ie, drugs and toxins) that may cause secondary liver injury.

HEPATIC ENZYMOLOGY

Serum liver enzymes are sensitive but not necessarily specific markers of primary hepatobiliary disease. They are not direct markers of liver function. Alanine aminotransferase (ALT) and aspartate aminotransferase (AST) are markers of hepatocellular damage, whereas alkaline phosphatase (ALP) and gamma-glutamyltransferase (GGT) are markers of cholestasis.3 Each individual enzyme can provide information on whether liver disease is present and may provide clues as to the most likely differential diagnosis.

Alanine Aminotransferase

ALT is a cytoplasmic enzyme found mainly in hepatocytes. However, it is also found in other cells, such as skeletal muscle, renal, and red blood cells, in smaller amounts. ALT is released into the circulation when there is hepatocyte necrosis or increased cell membrane permeability and therefore is a sensitive marker of hepatocellular injury. ALT is the most liver specific of the liver enzymes, but occasionally severe muscle damage or ex vivo hemolysis may increase ALT activity.4 Concurrent evaluation of creatine kinase activity may help discriminate between muscle disease and liver disease because creatine kinase activity is expected to increase with muscle damage. ALT activity can also be increased in patients with extrahepatic diseases that secondarily affect the liver (eg, feline hyperthyroidism). The reported half-life of ALT has been reported to be about 60 hours in dogs and 3.5 hours in cats.3 These relatively short half-lives are useful when monitoring recovery after acute liver injury. Conditions that can cause an increase in ALT activity include those listed in Table 1.

Aspartate Aminotransferase

AST is a cytoplasmic and mitochondrial enzyme found in hepatocytes and other cells. Reversible or irreversible damage to the liver causes release of the cytoplasmic AST; however, only irreversible damage to the cell will cause release of mitochondrial AST. These two sources of AST are not distinguishable by measuring serum AST activity on a routine biochemistry panel.

Increases in AST activity generally parallel those of ALT. However, AST is less specific for liver injury than ALT because increases in activity of AST may also be due to cardiac or skeletal muscle injury4 or ex vivo hemolysis. The half-life of AST is about 22 hours in dogs and 80 minutes in cats.3 The shorter half-life compared with ALT means that AST activity decreases and returns to normal before that of ALT in patients with acute liver injury. Conditions that can cause an increase in AST activity include those listed in Table 1.

Alkaline Phosphatase

ALP is an enzyme found in hepatocytes that line the bile canaliculi. It is released into the circulation during intra- or extrahepatic cholestasis. This enzyme is sensitive for hepatobiliary disease in dogs (80%), but because of the possible contributions of bone and glucocorticoid-induced isoenzymes to serum ALP activity, its specificity is low (51%).5 In young, growing animals, ALP activity is normally increased because of the bone isoenzyme, with 71% of dogs younger than 1 year having ALP activity >150 U/L.6 Bone ALP may also be elevated in patients with osteomyelitis or osteosarcoma. Dogs with hyperadrenocorticism and those receiving glucocorticoids can be expected to have increased ALP activity due to the glucocorticoid-induced isoenzyme. Conditions that can cause an increase in ALP activity include those listed in Table 1.

The highest activities of ALP have been reported with conditions such as cholestasis, steroid hepatopathy, chronic hepatitis, and hepatic necrosis.7 This lack of tissue specificity can make increases in activity of ALP hard to interpret. The half-life of ALP is approximately 70 hours in dogs and 6 hours in cats.3 In cats, which lack the glucocorticoid-induced isoenzyme with a shorter half-life, increases of serum ALP activity are more specific for hepatobiliary disease than in dogs and are generally clinically relevant.

Gamma-Glutamyltransferase

GGT is associated with the cell membranes of hepatocytes that form the bile canaliculi and bile ducts, as well as periportal hepatocytes. It is a marker of intrahepatic (eg, feline hepatic lipidosis) or extrahepatic (eg, bile duct obstruction) cholestasis. In dogs, it has a higher specificity (87%) and lower sensitivity (50%) for hepatobiliary disease compared with ALP.7 In general, GGT is a more sensitive marker of feline hepatobiliary disease than ALP. However, in cats with feline hepatic lipidosis, GGT is generally only mildly elevated.8 No definitive studies determining the half-life of GGT have been performed in cats or dogs. However, serum GGT and ALP activities decrease after liver injury at a similar rate in dogs, suggesting that they have a similar half-life.9

Interpreting Liver Enzyme Elevations

The degree of the increase in hepatocellular-damage enzyme activities may help stratify disease severity as follows5:

  • Mild: 2- to 3-fold elevation in activity
  • Moderate: 5- to 10-fold elevation in activity
  • Marked: >10-fold elevation

However, such increases do not always correlate with severity of disease. This is true in dogs and cats with portosystemic shunting and dogs with end-stage chronic hepatitis, in which hepatocytes are replaced by fibrous tissue. Therefore, the degree of liver enzyme increase should be interpreted with caution.

Because the liver has a large regenerative capacity, the degree of liver enzyme elevation should also not be used to indicate prognosis. For example, a dog with acute liver injury may have severely increased serum ALT activity but can still make a full recovery. Longitudinal monitoring trends in liver enzyme activities can help in determining chronicity and monitoring disease progression and/or response to treatment.

In evaluating liver enzymes, it is important to determine what type of elevation pattern is present (ie, hepatocellular damage versus cholestasis). A relatively greater increase in ALT and AST activity indicates hepatocellular damage, while a greater increase in ALP and GGT activity indicates cholestasis, which could be intrahepatic or extrahepatic. Establishing the pattern may help narrow the differential diagnosis. However, some liver diseases can display a mixed pattern (eg, cholangitis, phenobarbital hepatopathy).

LIVER FUNCTION TESTING

Routine biochemical testing can give clinicians an insight into many liver functions. Box 2 presents common abnormal results of biochemical tests that can have liver-related causes as well as important differential diagnoses to consider for these test results. However, because of the liver’s functional reserve capacity, these tests are not sensitive for liver insufficiency. Abnormal results can also be caused by other conditions and thus also lack specificity.

It is important for clinicians to not only look for analytes that are flagged as being outside their respective reference intervals but also look at their actual values. For example, serum albumin, cholesterol, and blood urea nitrogen (BUN) concentrations toward the lower limit of the reference interval suggest hepatic insufficiency or portosystemic shunting. Monitoring trends in these values over time can also be informative.

Because of the limited sensitivity and specificity of biochemical tests, patients with confirmed or suspected liver disease sometimes require additional liver function testing to better characterize their disease.

Serum Bile Acids

Measurement of the total concentrations of serum bile acids (SBA) aids in the diagnosis of patients with portosystemic shunts and in the assessment of hepatic function. Potential indications for SBA measurement include:

  • Suspicion for portosystemic shunting (eg, seizures, other signs of encephalopathy)
  • Persistently increased liver enzyme activities, especially ALT
  • Severe hypoalbuminemia (<2.0 g/dL) in dogs
  • Unexplained ammonium urate urolithiasis
  • Hyperbilirubinemia when hemolysis cannot be definitely diagnosed/excluded (uncommon)

In a healthy patient, SBA are synthesized from cholesterol. In dogs, bile acids are conjugated to glycine or taurine and then stored in the gallbladder, whereas in cats they are conjugated almost exclusively with taurine.10 After a meal, the gallbladder contracts because of secretion of cholecystokinin, emptying bile into the duodenum. Bile acids are absorbed in the ileum. They are transported via the portal circulation to the liver, where they are subsequently reabsorbed. Normally this process is about 95% to 98% efficient.

The enterohepatic recirculation of bile acids is impeded in dogs without gallbladders and patients with ileal disease or that have had ileal resection, causing a decrease in SBA concentration. Other conditions that can cause decreased SBA concentrations include GI malabsorption and decreased gastric motility.2 Causes of increased total SBA concentrations are listed in Box 2. Diseases that cause intrahepatic cholestasis (lipidosis, diabetes mellitus, lymphoma, histoplasmosis, cirrhosis) or extrahepatic cholestasis (cholangitis, bile duct carcinoma, liver flukes, cholelithiasis, pancreatitis) can cause decreased bile acid excretion, despite no decrease in functional hepatic mass. In patients with hyperbilirubinemia, once hemolysis has been ruled out, measuring SBA is not indicated because their concentration will be predictably increased.

Compared with plasma ammonia, SBA are easy to measure and do not not require special sample handling. Paired preprandial and 2-hour postprandial SBA measurements are usually performed to increase the sensitivity of this test (Box 3). While SBA measurement is arguably the best test of liver function and portosystemic shunting in dogs and cats, increased concentrations are not specific for any single hepatobiliary disease. Therefore, this test can be helpful for evaluating the likelihood of hepatobiliary disease; however, it cannot definitively determine the underlying liver disease. Additionally, this test does not provide a truly quantitative assessment of hepatic function. Because of the hepatic reserve capacity, it is possible for dogs with normal SBA concentrations to have hepatobiliary disease; therefore, this test should not be used to screen patients for hepatobiliary disease. However, the sensitivity of SBA measurement for portosystemic shunts (congenital and acquired) is high and in one study was reported to be 93% and 100% for dogs and cats with congenital portosystemic shunt, respectively.11

Ammonia

Blood ammonia concentration can be increased because of portosystemic shunting, severe hepatic insufficiency, or urea cycle enzyme deficiencies (Box 4).12 Potential indications for plasma ammonia measurement include:

  • Suspicion for portosystemic shunting
    (eg, seizures, other signs of encephalopathy)
  • Suspicion for urea cycle enzyme deficiency
    (eg, cat with feline hepatic lipidosis)
  • Unexplained ammonium urate urolithiasis

Ammonia is mainly produced by catabolism of glutamine by enterocytes and bacterial degradation of urea and proteins in the large bowel. Therefore, blood coming from the splanchnic circulation is rich in ammonia.13 The liver detoxifies ammonia through two pathways: (1) the urea cycle, which converts ammonia into urea and (2) consumption of ammonia during glutamine synthesis by hepatocytes. In animals with portosystemic shunting or severe hepatic dysfunction, the liver is unable to synthesize sufficient glutamine or urea, leading to hyperammonemia. Because ammonia freely passes across membranes, including the blood–brain barrier, hyperammonemia contributes to the development of clinical signs of hepatic encephalopathy.

Fasting Ammonia Measurement

Ammonia testing requires heparinized tubes, transfer of the sample on ice, and urgent separation of plasma and is ideally performed within 30 minutes of sample collection. These requirements can make this diagnostic assay difficult to perform in private practice. Increased serum ammonia is a sensitive marker for congenital and acquired portosystemic shunts, with a reported sensitivity of 83% to 98%.11,14 However, in the absence of portosystemic shunting, ammonia is not a sensitive test of liver disease.

Ammonia Tolerance Test

When ammonia is administered orally or rectally to a normal dog, it should be efficiently extracted from the portal circulation by the liver. However, dogs with a portosystemic shunt or decreased hepatic functional mass cannot extract the additional ammonia, leading to an excessive increase in plasma ammonia concentration.

The main indication for this test is concern for hepatic insufficiency that is not supported by routine laboratory testing. This test is unnecessary for dogs with increased baseline ammonia, in addition to posing a risk for hepatic encephalopathy in these patients. Disadvantages of oral ammonia administration include15:

  • Absorption depends on gastric emptying.
  • Vomiting can occur.
  • It is stressful to the patient.
  • The taste of the ammonium chloride is unpleasant.

The rectal ammonia tolerance test avoids these problems (Box 5).16 However, we do not routinely perform either test in dogs or cats.

Postprandial Venous Ammonia Tolerance Test

The postprandial ammonia tolerance test involves a procedure similar to that of the oral or rectal ammonia tolerance test except that digested food provides the ammonia challenge and the disadvantages of oral administration are avoided. The patient is fed a commercial diet containing about 30% protein to provide 33 kcal/kg, and a blood sample is collected 6 hours after feeding. This test was reported to have 91% sensitivity for the detection of portosystemic shunting, but in the absence of portosystemic shunts it is not as sensitive for detecting hepatic insufficiency.17

Protein C

Protein C is an anticoagulant protein produced by the liver. Measurement of protein C provides information regarding liver function and perfusion. In one study,18 dogs with congenital and acquired portosystemic shunts, hepatic failure, and chronic hepatitis had decreased levels of protein C, which distinguished them from dogs with microvascular dysplasia (portal vein hypoplasia) and those without hepatobiliary disease. With use of a cutoff value of 70% activity, protein C could distinguish dogs with congenital portosystemic shunt from those with microvascular dysplasia with a sensitivity of 93% and a specificity of 88%.

CONCLUSIONS

Increased liver enzyme activities are common results in small animal practice and can suggest patterns of liver disease, including hepatocellular damage, cholestasis, or both. Liver enzymes, especially ALP, are not specific for primary liver disease. To evaluate their clinical significance, a combination of history, clinical signs, physical examination, diagnostic imaging, and other liver function test results must be considered. Changes such as hypocholesterolemia or hypoalbuminemia can suggest hepatic dysfunction. Measuring SBA or ammonia concentrations provides a more accurate assessment of liver function, but it is important to be aware that patients with normal liver function test results can still have liver disease. Although these laboratory tests play an important role in diagnosing canine and feline liver disease, definitive diagnosis usually requires a combination of diagnostic imaging and cytologic or histologic assessment of liver tissue.

References

  1. Hall JE, Guyton AC. Guyton and Hall Textbook of Medical Physiology, 12th ed. Philadelphia: Saunders/Elsevier; 2011.
  2. Allison RW. Laboratory evaluation of the liver. In: Thrall M, Weiser, G, Allison RW, Campbell TW (eds): Veterinary Hematology and Clinical Chemistry, 2nd ed. Oxford: John Wiley & Sons; 2012:401-424.
  3. Lidbury JA, Steiner JM. Diagnostic evaluation of the liver. In: Washabau RJ, Day MJ, eds: Canine & Feline Gastroenterology. St. Louis, MO: Elsevier Saunders; 2013:863-875.
  4. Valentine BA, Blue JT, Shelley SM, et al. Increased serum alanine aminotransferase activity associated with muscle necrosis in the dog. J Vet Intern Med 1990; 4(3):140-143.
  5. Webster CRL, Cooper JC. Diagnostic approach to hepatobiliary disease. In: Bonagura J, Twedt D, eds. Kirk’s Current Veterinary Therapy, 15th ed. St. Louis, MO: Elsevier; 2014:569-575.
  6. Comazzi S, Pieralisi C, Bertazzolo W. Haematological and biochemical abnormalities in canine blood: frequency and associations in 1022 samples. J Small Anim Pract 2004; 45(7):343-349.
  7. Center SA, Slater MR, Manwarren T, et al. Diagnostic efficacy of serum alkaline phosphatase and gamma-glutamyltransferase in dogs with histologically confirmed hepatobiliary disease: 270 cases (1980-1990). JAVMA 1992; 201(8):1258-1264.
  8. Center SA, Baldwin BH, Dillingham S, et al. Diagnostic value of serum gamma-glutamyl transferase and alkaline phosphatase activities in hepatobiliary disease in the cat. JAVMA 1986; 188(5):507-510.
  9. Kaneko JJ, Harvey J, Bruss ML. Diagnostic enzymology of domestic animals. In: Clinical Biochemistry of Domestic Animals, 6th ed. St. Louis, MO: Elsevier; 2008:358-361.
  10. Rabin B, Nicolosi RJ, Hayes KC. Dietary influence on bile acid conjugation in the cat. J Nutr 1976; 106(6):1241-1246.
  11. Ruland K, Fischer A, Hartmann K. Sensitivity and specificity of fasting ammonia and serum bile acids in the diagnosis of portosystemic shunts in dogs and cats. Vet Clin Pathol 2010; 39(1):57-64.
  12. Center SA, ManWarren T, Slater MR, et al. Evaluation of twelve-hour preprandial and two-hour postprandial serum bile acids concentrations for diagnosis of hepatobiliary disease in dogs. JAVMA 1991; 199(2):217-226.
  13. Lidbury JA, Cook AK, Steiner JM. Hepatic encephalopathy in dogs and cats. J Vet Emerg Crit Care (San Antonio) 2016; 26(4):471-487.
  14. Gerritzen-Bruning MJ, van den Ingh TS, Rothuizen J. Diagnostic value of fasting plasma ammonia and bile acid concentrations in the identification of portosystemic shunting in dogs. J Vet Intern Med 2006; 20(1):13-19.
  15. Meyer DJ, Strombeck DR, Stone EA, et al. Ammonia tolerance test in clinically normal dogs and in dogs with portosystemic shunts. JAVMA 1978; 173(4):377-379.
  16. Rothuizen J, van den Ingh TS. Rectal ammonia tolerance test in the evaluation of portal circulation in dogs with liver disease. Res Vet Sci 1982; 33(1):22-25.
  17. Walker MC, Hill RC, Guilford WG, et al. Postprandial venous ammonia concentrations in the diagnosis of hepatobiliary disease in dogs. J Vet Intern Med 2001; 15(5):463-466.
  18. Toulza O, Center SA, Brooks MB, et al. Evaluation of plasma protein C activity for detection of hepatobiliary disease and portosystemic shunting in dogs. JAVMA 2006; 229(11):1761-1771.

Brigitte B. McAtee received her DVM from Auburn University. She is currently a second-year internal medicine resident at Texas A&M University in College Station, Texas. Her clinical and research interests include infectious and immune-mediated diseases.

 

 

Jonathan A. Lidbury graduated from the University of Glasgow and completed his internal medicine residency at Texas A&M University. Jonathan is an assistant professor of small animal internal medicine and the associate director for clinical services of the Gastrointestinal Laboratory at Texas A&M University. His clinical interests are small animal hepatology and gastroenterology, and he is involved in a wide range of research in these fields.

 


Liver Enzyme Interpretation and Liver Function Tests

The article you have read has been submitted for RACE approval for 1 hour of continuing education credit and will be opened for enrollment when approval has been received. To receive credit, take the approved test online at vetmedteam.com/tvp.aspx (CE fee of $5/article).

TOPIC Overview

Increased serum liver enzyme activities are sensitive markers of hepatocellular injury or cholestasis. However, these elevations may also be present in a wide variety of extrahepatic disorders and are therefore nonspecific. Hypoalbuminemia, hypocholesterolemia, hypoglycemia, hyperbilirubinemia, and decreased blood urea nitrogen suggest hepatic dysfunction. However, these changes are not always present in dogs and cats with liver disease. Blood ammonia and bile acids are more sensitive tests of liver function. This article reviews the interpretation of serum liver enzyme activities and commonly used liver function tests.

Learning Objectives

After reading this article, participants should be able to:

  1. Interpret serum liver enzyme activities in dogs and cats and recognize the limitations of these tests
  2. Recognize changes consistent with hepatic dysfunction on a serum chemistry panel
  3. Recognize the indications for further liver function testing
  4. Interpret serum bile acid measurements and recognize the limitations of these tests
  5. Interpret plasma ammonia measurements and recognize the limitations of these tests

 

  1. At least _____ of hepatic tissue must be nonfunctional for hypoglycemia to be seen.

A. 20%

B. 50%

C. 75%

D. 99%

  1. In cats, bile acids are conjugated almost exclusively with which amino acid?

A. Taurine

B. Asparagine

C. Glycine

D. Glutamine

  1. True or false: Protein C testing distinguishes between hepatocellular carcinoma and congenital portosystemic shunts.

 

  1. Which of the following is not a cause of decreased serum bile acids?

A. Jejunal resection

B. Lack of gallbladder

C. Decreased gastric motility

D. Ileal disease

  1. Which of the following liver enzymes is most specific for detecting hepatocellular damage?

A. ALP

B. GGT

C. ALT

D. AST

  1. Which of the following conditions is not a reason for elevated serum ALT activity?

A. Neoplasia

B. Hepatitis

C. Increased osteoblastic activity

D. Diabetes mellitus

  1. True or false: AST has a shorter half-life than ALT.

 

  1. Which of the following conditions has not been associated with hyperammonemia?

A. Urea cycle enzyme deficiency

B. Hepatic dysfunction causing decreased glutamine synthesis

C. Acquired portosystemic shunting

D. Extrahepatic bile duct obstruction

  1. Which of the following findings suggests hepatobiliary dysfunction?

A. Decreased bilirubin

B. Hyperglycemia

C. Hypercholesterolemia

D. Increased bilirubin

 

  1. What would be the most accurate interpretation of the following test results: ALT, 160 U/L (normal range, 10–130 U/L); ALP, 1800 U/L (24–147 U/L); GGT, 34 U/L (0–25 U/L); AST, 18 U/L (10–34 U/L)?

A. Hepatocellular damage

B. Cholestasis

C. Cholangitis

D. Extrahepatic bile duct obstruction

 

SHARE ARTICLE
  • Like us on facebook

  • Latest tweets