Sage Journals: Discover world-class research

Abstract

Beta–gamma bridging (β-γ bridging) on serum protein electrophoresis is touted as being virtually pathognomonic for hepatic disease. However, the criteria for β-γ bridging are not defined, and few publications support a relationship between β-γ bridging and liver disease. The goal of this retrospective study was to evaluate the prevalence of hepatic pathology in animals with β-γ bridging. All serum protein electrophoretograms from clinical patients generated at the University of Georgia between 1994 and 2008 were evaluated for the presence of β-γ bridging, defined as (1) an albumin:globulin ratio below the reference interval; (2) indistinct separation between all β and γ globulin fractions or between the β₂ and γ fractions, with a negative shoulder slope of < 5%; and (3) predominance of γ proteins versus β proteins. Of the 237 electrophoretograms examined, 25 (11 dogs, 11 cats, 3 horses) met the inclusion criteria for β-γ bridging. Patients were classified into disease categories on the basis of biochemical, cytologic, and/or histologic findings. Positive predictive values of β-γ bridging for hepatic and infectious diseases were determined with a one-sided exact binomial test. Of 25 animals, 8 had evidence for hepatic disease, whereas 9 had infectious diseases. As such, the positive predictive value of β-γ bridging for hepatic disease was 32.0%, with a 95% confidence interval of 15.0% to 53.5% (P < .001), whereas for infectious disease, the positive predictive value was 36.0%, with a similar confidence interval. Beta–gamma bridging is not pathognomonic for liver diseases and is as frequently found with infectious diseases.

Keywords

serum protein electrophoresis beta–gamma bridging hepatic pathology liver disease

Beta–gamma (β-γ) bridging, also known as β-γ linking⁴ and β-γ fusion,¹¹ is an uncommon pattern observed on electrophoretograms following serum protein electrophoresis. Despite its frequent usage, the term β-γ bridging is actually poorly defined. The use of this terminology is commonly based on the lack of demarcation between the peaks of the β and γ globulins on an electrophoretogram, along with a decreased albumin:globulin ratio (A:G).²⁹ According to many human and veterinary medicine and diagnostic pathology textbooks published over the past 60 years, the pattern of β-γ bridging is considered “virtually pathognomonic” for chronic active hepatitis or hepatic cirrhosis.^5,14,15,18 Reports of this correlation date back to the early 1950s and are based on studies by Demeulenaere and Wieme, utilizing paper serum protein electrophoresis,⁴ and Sunderman and Sunderman, utilizing sucrose gradient electrophoresis of sera.²⁸ The relationship between liver disease and this electrophoretic pattern is presumed to result from increased levels of various protein fractions, predominantly immunoglobulins A and M, resulting in poor separation between the β and γ fractions.^18,19 Despite the reported occurrence of β-γ bridging in other nonhepatic diseases—including lymphoma, mycobacteriosis, heartworm disease, leishmaniasis, and γ heavy chain disease—human medical sources suggest that the presence of a β-γ bridge on an electrophoretogram warrants a liver biopsy as the next appropriate diagnostic step.^2,8,12,18 The original studies of β-γ bridging in people have been widely accepted as being applicable to domestic species, but to our knowledge, it has not been evaluated in animals.

The positive predictive value (PPV) is defined as the probability that a positive test result indicates that the animal has the condition being evaluated.²⁶ Unlike test sensitivity, PPV varies with disease prevalence: When disease prevalence is high, a patient is more likely to have the disease, regardless of the test result; thus, false-negative results are more likely. For a test to be considered pathognomonic for a disease, the PPV should be 100%; all animals with the disease test positive, whereas all animals without the disease test negative. If β-γ bridging is virtually pathognomonic for hepatic pathology, the PPV should be high. The aim of this retrospective study was to determine the prevalence of hepatic pathology in animals with β-γ bridging, based on biochemical, cytologic, and histologic findings consistent with liver disease.

Materials and Methods

All mammalian serum protein electrophoresis profiles were evaluated as generated on hospitalized patients at the University of Georgia’s clinical pathology laboratory between September 1994 and July 2008. In all cases, protein electrophoresis was performed with cellulose acetate gels (Titan III Zip Zone Plates, Helena Laboratories, Beaumont, TX) stained with Ponceau S and sulfosalicylic acid (Helena Laboratories), and electrophoretograms were produced by a scanning densitometer utilizing ultraviolet light detection (Cliniscan 2, Helena Laboratories). All electrophoretograms were evaluated by two board-certified clinical pathologists for β-γ bridging, without prior knowledge of any medical information or previous interpretation. Electrophoretogram analyses were performed under the premise that all proteins to the cathode side of midline were considered to be β and γ globulins.²⁶ Based on the literature and a collective discussion among the pathologists, the following criteria were used to define a β-γ bridge (Fig. 1 ):

The A:G ratio was ≤ 0.6 (the low end of the reference interval established at the University of Georgia’s College of Veterinary Medicine for all included species).

There was a lack of obvious separation between all β and γ globulin fractions (Fig. 2; type 1 β-γ bridge) or between the β2 and γ fractions (Fig. 3 ; type 2 β-γ bridge). If any visible trough on the β-γ bridge was present, it was less than 5% of the height of the β peak, based on the sensitivity of our densitometer. In these cases, a distinct separation between the β and γ regions could not be definitively identified (Fig. 4 ; type 3 β-γ bridge).¹⁰

Figure 1.

Decision tree illustrating classification process for beta–gamma (β-γ) bridging. Solid arrows indicate yes decisions. Open arrows indicate no decisions. A:G, albumin:globulin.

Figure 2.

Type 1 beta–gamma bridge: lack of distinction between all beta and gamma components. A, midpoint; B, shoulder. Albumin, 2.2 g/dl; alpha, 2.1 g/dl; beta–gamma, 11.0 g/dl.

Figure 3.

Type 2 beta–gamma bridge: lack of distinction between beta 2 and gamma fractions. A, midpoint; B, shoulder. Albumin, 2.43 g/dl; alpha, 1.13 g/dl; beta–gamma, 5.44 g/dl.

Figure 4.

Type 3 beta–gamma bridge: trough less than 5% of the height of the beta fraction. A, midpoint; C, trough. Albumin, 1.84 g/dl; alpha, 1.42 g/dl; beta–gamma, 4.14 g/dl.

Based on the area under the curve, the highest concentration of proteins is found in the quartile most closely associated with the cathode, indicating a γ predominance rather than a β predominance.

For the cases that met the initial inclusion criteria, medical records were examined to determine biochemical abnormalities and/or cytologic or histologic evidence for liver disease, along with a final diagnosis. The animals that met the inclusion criteria and had adequate follow-up data were then categorized as having hepatic or nonhepatic disease. For inclusion as a case of liver disease, animals had to possess either cytologic or histologic evidence for hepatic pathology, or they had to exhibit at least one of the major, or two of the minor, biochemical indicators of hepatic damage or dysfunction, as outlined in Table 1 . Cutoff values for biochemical parameters were generally based on twofold changes in magnitude above or below the upper and lower ends of the reference interval, respectively. Patients not demonstrating microscopic or biochemical evidence for liver involvement were classified as having nonliver disease and were further subclassified as having infectious, inflammatory, or other etiologies, based on the final diagnosis. If a final diagnosis was not reached, patients were classified as simply having hepatic or nonhepatic disease, with no further subclassification. Potential associations between all protein and biochemical parameters were visually assessed on scatter plots using Cartesian coordinates, as generated with spreadsheet software (Microsoft Office Excel 2007). Then, linear regression analysis was performed to evaluate any potentially significant associations; PPV and a 95% confidence interval were determined with a one-sided exact binomial test (SAS 9.1). For this analysis, we defined virtually pathognomonic as an expected value of 95%.

Table 1.

Biochemical Indicators of Hepatic Damage or Dysfunction Used to Support a Diagnosis of Liver Disease in Patients With Beta–Gamma Bridging^a

Major Criteria, U/liter	Canine	Feline	Equine
Alanine aminotransferase	> 218	> 194	SDH^b > 16
Alkaline phosphatase	> 228	> 90	GGT^c > 64
Total bilirubin	> 0.6	> 0.2	> 6.4

^a Minor criteria for canine, feline, and equine (mg/dl): glucose < 50, blood urea nitrogen < 6, cholesterol < 120.

^b SDH, sorbitol dehydrogenase.

^c GGT, gamma glutamyltransferase.

Results

In summary, 237 electrophoretograms were performed between September 1994 and July 2008. Based on the defined criteria, 36 cases were initially identified as having a β-γ bridge. Of these initial 36 cases, only 25 met all the inclusion criteria. Table 2 presents the results for these 25 cases. Eleven dogs (ages, 1–11 years), 11 cats (ages, 7 months–14 years), and 3 horses (ages, 2 months–16 years) were included. Eight of these animals (32%) had microscopic or biochemical evidence for liver disease: Case Nos. 1 and 2 had intrahepatic neoplasia; case No. 3 had hepatic lipidosis; case Nos. 4 and 5 had infectious diseases directly involving the liver (leishmaniasis and bacterial septicemia); and case Nos. 6–8 had an undetermined disease process involving the liver. Case No. 6 was a dog with dramatically increased liver enzyme activities and a provisional clinical diagnosis of cholangiohepatitis. Unfortunately, this patient was discharged without further diagnostics or a final diagnosis. The second animal with unclassified liver disease (case No. 7) was a cat with increased alanine aminotransferase and a palpably enlarged liver. The cat had a provisional diagnosis of feline infectious peritonitis (FIP). This animal was euthanized without a postmortem examination. The third animal with unspecified liver disease (case No. 8) was a horse with extensive nodular hepatic lesions. A liver biopsy revealed multifocal inflammatory foci and fibrosis, which were interpreted as “potentially consistent with chronic active hepatitis.” Hairy vetch toxicosis was a primary consideration but never confirmed.

Table 2.

Case Information for 25 Animals With Beta–Gamma Bridging: Type, Disease Category, Basis for Classification, and Final Diagnosis

Case No. / Species	Sex^a	Age, Years	Type of Bridge^b	Disease Category^c	Basis for Classifying^d	Final Diagnosis
1. Canine	FS	1	2	Liver	M, B	B-cell lymphoma with liver involvement
2. Feline	MN	3	1	Liver	M, B	Round cell neoplasia with liver involvement
3. Feline	MN	13	1	Liver	M, B	Suspect feline infectious peritonitis with hepatic lipidosis
4. Canine	M	9	1	Liver	M, B	Leishmaniasis with liver involvement
5. Equine	M	0.2	3	Liver	M, B	Bacterial sepsis with liver involvement
6. Canine	MN	11	2	Liver	B	Chronic liver disease, suspect cholangiohepatitis
7. Feline	M	0.6	1	Liver	B	Suspect feline infectious peritonitis
8. Equine	M	9	2	Liver	M, B	Diffuse granulomatous disease involving viscera
9. Canine	FS	Adult	1	Infectious	B	Ehrlichiosis, whipworms
10. Canine	FS	9	1	Infectious	M, B	Ehrlichiosis
11. Canine	FS	6	1	Infectious	B	Ehrlichiosis
12. Canine	FS	6	1	Infectious	B	Ehrlichiosis, mass in brain, hydrocephalus
13. Canine	MN	2	1	Infectious	B	Leishmaniasis
14. Feline	FS	1	2	Infectious	B	Ileocolic feline infectious peritonitis
15. Feline	F	5	1	Infectious	B	Feline infectious peritonitis
16. Feline	MN	Adult	2	Infectious	B	Chronic bacterial infections, bone marrow necrosis
17. Feline	M	14	1	Infectious	M, B	Atypical mycobacteriosis, chronic renal failure
18. Feline	MN	4	2	Inflammatory	B	Periodontal disease
19. Canine	M	5	1	Inflammatory	B	Pulmonary infiltrates with eosinophilia
20. Equine	MN	16	1	Other	B	Carcinoma in lung
21. Canine	M	4	1	Other	B	Osteosarcoma
22. Feline	FS	10	2	Other	B	Early chronic renal failure, possible nasal allergy
23. Feline	MN	6	2	Nonliver: Open	B	Sublingual fibroma, possible bartonellosis
24. Feline	MN	2	2	Nonliver: Open	B	Suspect feline infectious peritonitis
25. Canine	MN	9	2	Nonliver: Open	M, B	Incompletely resected cutaneous mast cell tumor

^a M, male; F, female; MN, male neutered; FS, female spayed.

^b Type 1: Bridging is present between all β and γ fractions. Type 2: Bridging is present between β₂ and γ fractions. Type 3: Visible trough of < 5% of the height of the β peak.

^c Other is nonliver disease that is neither infectious nor inflammatory. Nonliver: Open has no final diagnosis and no evidence for hepatic disease or involvement.

^d M, microscopic evaluation of the liver (cytology, histopathology, or both); B, biochemical. Patients with both M and B designations were classified on the basis of microscopic and biochemical findings.

The remaining 17 animals (68%) with a β-γ bridge on electrophoresis did not have microscopic or biochemical evidence of liver disease. Of these animals without liver involvement, 9 (53%) had infectious diseases that did not result in clinical hepatic disease, based on biochemical and microscopic findings. Of the 9 animals with infectious diseases but not liver disease, 4 dogs (case Nos. 9–12) had ehrlichiosis; 1 dog (case No. 13) had leishmaniasis; 2 cats (case Nos. 14 and 15) had FIP; and 2 cats (case Nos. 16 and 17) had bacterial septicemia. Of the 17 animals without cytologic, histologic, or biochemical evidence for liver disease, 2 (12%) had inflammatory diseases not associated with a proven infectious etiology: Case No. 18 was a cat with severe periodontal disease; case No. 19 was a dog diagnosed with pulmonary infiltrates with eosinophilia. Of the 17 animals, 3 (18%) had diseases classified as other (neither hepatic nor inflammatory/infectious disease): 2 (12%; case Nos. 20 and 21) were diagnosed with nonhepatic neoplasia, and 1 (6%; case No. 22) was a cat diagnosed with chronic renal failure. The remaining 3 patients did not have definitive diagnoses but lacked cytologic, histologic, or biochemical evidence for hepatic disease: 2 (12%; case Nos. 23 and 24) were cats suspected of having infectious diseases (bartonellosis and FIP, respectively), but neither was confirmed; the third animal (case No. 25) was a dog with an incompletely resected cutaneous mast cell tumor.

Based on visual inspection of scatter plots, a positive correlation appeared between total protein concentration and increases in the γ and β-γ fractions of liver and nonliver patients. This relationship was confirmed with linear regression analysis (P < .001). Linear regression analysis did not reveal correlations between any other biochemical parameter and the total serum protein concentration or protein fraction concentration. The PPV of β-γ bridging for hepatic disease was calculated to be 32% (P < .001), with a 95% confidence interval of 15.0% to 53.5%. The PPV of β-γ bridging for infectious disease not specifically associated with the liver was determined to be 36.0%, with a 95% confidence interval of 18.0% to 57.5% (P < .001).

Discussion

Based on the results of this retrospective study and despite the propagation of this information in the veterinary literature, β-γ bridging can occur in the absence of overt liver disease. In the present study, only 8 of 25 animals (32%) had hepatic disease, and only 1 of these was suspected to have chronic active hepatitis. None of the animals had clinical, biochemical, or microscopic evidence for hepatic cirrhosis.

For this analysis, we defined virtually pathognomonic as an expected PPV of 95% (ie, 95% of animals that have β-γ bridging truly have hepatic disease). In the study population of this report, the 95% confidence intervals for a PPV for hepatic disease ranged from 15.0% to 53.5%. Although some animals with liver disease certainly have β-γ bridging on electrophoretograms, these results indicate that most animals with bridging can be expected to have other diseases that do not affect the liver. Animals with confirmed liver disease and animals with infectious disease not involving the liver had similar PPVs and confidence intervals.

The lack of clearly defined universal criteria makes identification of β-γ bridging difficult, with marked variation among pathologists about how to define this phenomenon. At the time the analyses were performed, 23 of 237 (9.7%) were reported as having β-γ bridges. Only 10 of these met our narrow inclusion criteria. Others included in this report were not classified as β-γ bridges at the time of initial interpretation. For the purpose of this study, we utilized the most restrictive definition for β-γ bridging. During evaluation of electrophoretograms, two other potential manifestations of β-γ bridging appeared, which were not included in our study. Electrophoretograms in one excluded group met the first two inclusion criteria (decreased A:G ratio and indistinct separation between β and γ fractions) and were classified as a potential fourth type of β-γ bridge (Fig. 5 ). However, in this small group of cases (6 of 237), the predominance of β-γ proteins is within the third quartile, indicating β rather than γ protein predominance. These cases were excluded owing to uncertainty of the significance of this pattern. Electrophoretograms in a second excluded group (3 of 237) met the first and third inclusion criteria (decreased A:G ratio and predominance of γ proteins) but exhibited a trough > 5% of the height of the β peak. Because these may simply represent an overlap between distinct β and γ fractions, this group was not included. Unfortunately, the majority of these animals were lost to follow-up analysis.

Figure 5.

Increased proteins in the beta fraction as opposed to the gamma fraction. A, midpoint. Albumin, 2.5 g/dl; alpha 1, 0.2 g/dl; alpha 2, 1.3 g/dl; beta, 2.9 g/dl; gamma, 1.3 g/dl.

Few published studies have evaluated the strength of the correlation between β-γ bridging and hepatic disease. However, there have been many attempts to determine the nature of the substance that links the β and γ regions on the protein electrophoretograms of human patients with known hepatic disease—particularly, cirrhosis.^{6,21,22,27,30} Although often referred to as substance X or substance T, the substances migrating between these two fractions are, according to most investigators' hypotheses, simply high levels of immunoglobulins. Some have theorized that these immunoglobulins are synthesized because of an autoimmune reaction occurring upon hepatocyte damage, resulting in exposure of hepatocyte nuclei and mitochondria that were previously immunologically protected. Others have suggested that there are excess immunoglobulins migrating within this fraction because of increased antigenic stimulation from enteric-derived proteins that are not adequately removed by the Kupffer cells of the diseased liver or that bypass the liver via acquired hepatic shunts.^21,23 Still others have proposed that increased globulins are actually produced locally with liver disease by proliferative reticuloendothelial cells lining hepatic sinusoids.^3,7,20,27 Different investigators have suggested that β-γ bridging occurs secondary to increased production of β and γ globulins, in an attempt to compensate for the hypoalbuminemia caused by decreased hepatic synthesis and/or loss of albumin in ascitic fluid.⁹ Finally, some have theorized that liver disease—particularly, cholangiohepatitis—incites an acute inflammatory response, resulting in the increased production of fibrinogen, which is actually the substance responsible for bridging the β and γ regions.^3,7,20 It is possible that any or all of these proposed mechanisms may be correct in specific instances. Regardless, the nature of the elusive migrating substance (or substances) does not need to be clarified for the identification of a β-γ bridge and any associated disease. It is likely that proteomic analysis could help resolve the identity of this substance.

The moment of transition of this proposed association between β-γ bridging and hepatic disease from the human into the veterinary literature is unknown. However, Johnson and Wicks¹³ compared the serum protein electrophoretic profiles of various mammalian species and humans, found marked species variation, and concluded that such a comparison (ie, between human and other mammalian protein profiles) is inappropriate for diagnostic purposes. Although arguments for the presence of increased levels of immunoglobulins seen with numerous diseases are reasonable explanations for β-γ bridging, the presence of other substances cannot be excluded. This is important to remember in interpreting the electrophoretogram of nonmammalian species in which plasma rather than serum is often utilized to evaluate fibrinogen levels. Fibrinogen migrates in the late β-globulin region and can bridge the area between the β and γ regions, particularly when increased levels of fibrinogen are present, making identification and interpretation difficult.^16,24 For this reason, all electrophoretograms from nonmammalian species were automatically excluded. Free hemoglobin–haptoglobin complexes resulting from hemolysis can occasionally migrate in the β-γ region and even within the α₂ region, often resulting in further diagnostic confusion.^1,17,25 To our knowledge, none of the samples in this study exhibited gross hemolysis. Additional electrophoretic studies using immunoelectrophoresis and capillary band electrophoresis may prove useful in further defining the migrating protein fractions to correctly determine the underlying pathologic process.

This study has shown, in contrast to information in the veterinary literature, that β-γ bridging does not have a strong predictive value for hepatic disease in dogs, cats, or horses. Based on the results of this study, the finding of β-γ bridging on an electrophoretogram should not be considered pathognomonic for hepatic disease in dogs, cats, or horses.

Footnotes

Acknowledgements

We thank Shay Bush, Jeff Duke, Kelly Gibson, and Laura Amlung for their assistance in obtaining medical information and Deborah Keys for her assistance with statistical analysis. We also thank Tom Drapela, Lari Cowgill, Jennifer Towe, and Drs Holly Brown and Julie Webb for assistance in preparing the article.

The authors declared that they had no conflicts of interest with respect to their authorship or the publication of this article.

The authors declared that they received no financial support for their research and/or authorship of this article.

References

Alanio-Brechot

Girard-Lamoulere

Abbed

Taoufik

Raphaël

Besson

: An atypical electrophoretic profile. Hematologie 12:424–428, 2006.

Bloch

Lee

Mills

Haber

: Gamma heavy chain disease-an expanding clinical and laboratory spectrum. Am J Med 55:61–70, 1973.

Cohen

Ohta

Singer

Popper

: Immunocytochemical study of gamma globulin in liver in hepatitis and postnecrotic cirrhosis. J Exp Med 111:285–293, 1960.

Demeulenaere

Wieme

: Serum electrophoretic anomalies in the serum of liver patients: a report of 1145 cases. Am J Dig Dis 6:661–675, 1961.

Evans

Duncan

: Proteins, lipids, and carbohydrates. In: Duncan & Prasse’s Veterinary Laboratory Medicine Clinical Pathology, ed. Latimer

Mahaffey

Prasse

, 4th ed., pp. 162–192. Blackwell Publishing, Ames, IA, 2003.

Feizi

: Immunoglobulins in chronic liver disease. Gut 9:193–198, 1968.

Frankilin

Bean

Paul

Routh

De La Huerga

Popper

: Electrophoretic studies in liver disease: II. Gamma₁ globulin in chronic liver disease. J Clin Invest 30:729–737, 1951.

Gicking

Foley

Harr

Raskin

Jacobson

: Plasma protein electrophoresis of the Atlantic Loggerhead sea turtle, Caretta caretta . J Herpetol Med Surg 14:13–19, 2004.

Gray

Barron

ESG

: The electrophoretic analyses of the serum proteins in diseases of the liver. J Clin Invest 22:191–200, 1943.

10.

Helena Laboratories. Cliniscan 2 Scanning Densitometer Operator’s Manual. Helena Laboratories, Beaumont, TX, 1993.

11.

Heremans

: Les globulines seriques du systeme gamma: leur nature et leur pathologie [Serum globulins of the gamma system, their nature and pathology]. Masson, Paris, France, 1960. (In French; chapter summaries in English).

12.

Hyperglobulinemia. Available at Cornell University College of Veterinary Medicine website: http://diaglab.vet.cornell.edu/clinpath/modules/chem/hypergl.htm. Accessed April 2, 2008.

13.

Johnson

Wicks

: Serum protein electrophoresis in mammals-taxonomic implications. Syst Zool 8:88–95, 1959.

14.

Johnston

Morris

: Alterations in blood proteins. In: Large Animal Internal Medicine, ed. Smith

, 2nd ed., pp. 489–497. Mosby, St Louis, MO, 1996.

15.

Kaneko

: Serum proteins and the dysproteinemias. In: Clinical Biochemistry of Domestic Animals, ed. Kaneko

Harvey

Bruss

, 5th ed., pp. 117–138. Academic Press, Boston, MA, 1997.

16.

Koay

ESC

Walmsley

: Handbook of Clinical Pathology, p. 163. PG Publishers, Singapore, 1989.

17.

Kyle

: Sequence of testing for monoclonal gammopathies. Arch Pathol Lab Med 123:114–118, 1999.

18.

Larson

: Serum proteins: diagnostic significance of electrophoretic patterns. Hum Pathol 5:629–640, 1974.

19.

Loo

SYT

Bhagavan

Scottolini

: Polyclonal gammopathy with β-globulin-γ-globulin bridging. Arch Pathol Lab Med 105:98–101, 1981.

20.

Paronetto

Rubin

Popper

: Local formation of γ-globulin in the diseased liver, and its relation to hepatic necrosis. Lab Invest 11:150–158, 1962.

21.

Parraga

Carlson

Thurmond

: Serum protein concentrations is horses with severe liver disease: a retrospective study and review of the literature. J Vet Intern Med 9:154–161, 1995.

22.

Powell

Mortimer

Harris

: Cirrhosis of the liver: a comparative study of the four major aetiological groups. Med J Aust 1:941–950, 1971.

23.

Sevelius

Andersson

: Serum protein electrophoresis as a prognostic marker of chronic liver disease in dogs. Vet Rec 137:663–667, 1995.

24.

Sibley

Hendrickson

: A comparative electrophoretic study of avian plasma proteins. Condor 72:43–49, 1970.

25.

Stern

Mais

Boggs

: “U” protein and migration ratios in paper electrophoresis. Clin Chem 3:599–608, 1957.

26.

Stockham

Scott

: Fundamentals of Veterinary Clinical Pathology, 2nd ed., p. 39. Blackwell Publishing, Ames, IA, 2008.

27.

Sunderman

Jr : Recent advances in clinical interpretation of electrophoretic fractionations of the serum proteins. In: Serum Proteins and Dysproteinemias, ed. Sunderman

Sunderman

Jr , 323–345. JB Lippincott Co, Philadelphia, PA, 1964.

28.

Sunderman

Jr Kniffen

: Electrophoretic fractionations of serum proteins in hepatic disease. In: Laboratory Diagnosis of Liver Disease, ed. Sunderman

Sunderman

Jr , pp. 15–27. Warren H. Green Inc, St Louis, MO, 1968.

29.

Teloh

: Serum proteins in hepatic disease. Ann Clin Lab Sci 8:127–129, 1978.

30.

Wieme

Demeulenaere

: Vier jaar ervaring met de electophorese in de hepatologie [Four year study with electrophoresis in the liver]. Beljisch tijdschrift voor geneeskunde 10:1032–1044, 1954. (In Dutch; abstract in English).

Evaluation of the Positive Predictive Value of Serum Protein Electrophoresis Beta–Gamma Bridging for Hepatic Disease in Three Domestic Animal Species

Abstract

Keywords

Materials and Methods

Results

Discussion

Footnotes

Acknowledgements

References