Veterinary hnmunology and hnmunopathology, 28 ( 199 ! ) 37-43

37

Elsevier Science Publishers B.V., Amsterdam

Use of multiple antigen substrates to detect ai tinuclear antibody in canine sera D. Scott McVey and Wilma Shuman Department of Laboratory Medicine, College of Veterinary Medicine, Kansas State | "eterinao, Medical Center and Kansas Agricultural Experiment Station, Kansas State University Manhattan, KS 66506, USA (Accepted 30 May 1990)

ABSTRACT McVey, D.S. and Shuman, W., 1991. Use of multiple antigen substrates to detect antinuclear antibody in canine sera. Vet, Immunol. ImmunopatLol., 28: 37-43. The presence of antinuclear antibody in serum has been used to serologically support the clinical diagnosis of systemic lupus erythematosus. Most often, diagnostic support is confered if the titer is above a "cut-off" value determined by the particular laboratory. The precision estimates of these commonly used serological assays are high. The arithmetic and geometric precisions estimates of fluorescent antinuclear antibody (FANA) assays were determined utilizing sera from dogs with polysystemic signalment suggestive of SLE. A simple score assay was found to have improved precision over FANA endpoint titrations on Hep-2 or cryostat, rat liver substrates. In one case, the precision improved from an arithmetic coefficient of variation of 25.6% (HEp-2, endpoint dilution, FANA titers) or 19.7% (rat liver cryostat substrate FANA titers) to ! 5°5%. Geometric coefficients of variation followed similar trends. Further, the score values for dogs with SLE were considerably higher (mean = 3.28) than for clinically normal dogs (mean = 0.20). The score system was more precise than FANA endpoint titers and therefore may be useful for diagnostic purposes, especially when FANA titers are negative or low.

INTRODUCTION

Numerous serologic assays have been used to detect antinuclear antibody (ANA) in sera from dogs, primarily to aid in the diagnosis of systemic lupu~ erythematosus (SLE; ShuU et a!., 1983; Costa et al., 1984). These techniques were recently reviewed by Toth and Rebar (1987). Even for fluorescent antinuclear antibody (FANA) assays, numerous types of substrates have been employed. Each of these substrates has many different antigens. There have been resultant problems with specificity and sensitivity. In one review (Quimby et al., 1980), only 56% of dogs with a clinical diagnosis of SLE had significant ANA titers. Shull et al. (1983) reported that no correlation between ANA titer, % DNA 0165-2427/91/$03.50

© 1991 m Elsevier Science Publishers B.V.

38

D.S. MCVEY AND W. SHUMAN

binding, or anti-dAdT titer (a synthetic double-stranded DNA analogue) could be documented. These investigators concluded that canine ANA are heterogeneous and are not as specific for SLE as ANA in human sera. FANA and specific ANA are only two indicators of connective tissue disease (Toth and Rebar, 1987). Positive ANA titers are observed in numerous inflammatory and infectious diseases (Toth and Rebar, 1987). However, the levels (titers) of antibodies to histones, Sm antigens, soluble nuclear antigens, and antigens of intact, fixed cells are consistently higher in dogs with classic, multisystemic symptoms of SLE (Quimby et al., 1980; Costa et al., 1984). For approximately 20 years, ANA testing has been used in clinical immunology laboratories, but the results have been difficult to interpret. One possible source of error is the precision of ANA assays. Proficiency testing by various FANA procedures in multiple laboratories in the U.S.A. has revealed differences up to 64-fold in titers of replicate human sera (Bonifacio, 1986). Clinical interpretation is difficult in veterinary medicine because there is little method standardization and no well-established reference ranges from healthy or ill dogs (Toth and Rebar, 1987). Quantitation by serial dilution may not be necessary for diagnosis and may contribute to poor precision of FANA assays. The purpose of this study was ( 1 ) to determine the precision of FANA techniques often employed at the Kansas State University Veterinary Hospital and (2) to describe a simple ANA scoring system modeled after that devised by Bonifacio (1986). MATERIALS AND METHODS

Patients and sera Sera submitted from dogs with polysystemic symptoms and histories suggestive of SLE were selected for these experiments. Criteria for inclusion in this study were fever (unresponsive to antibiotics) and any combination of two or more of the following: a symmetric, nonerosive polyarthritis, peripheral lymphadenopathy, dermatologic disease, hemolytic anemia, or thrombocytopenic purpura. These observations are among the most common with canine SLE as reported by Lewis and Pieut (1989). Twelve dogs with positive ANA tests (titers of 10 or greater on HEp-2 cells and/or positive LE cell tests; Coles, 1986) and six dogs with negative ANA tests were chosen. Also, sera from 50 dogs (various ages, 29 male, 21 female) that were clinically normal were tested, and score values were calculated. Fluorescent antinuclear antibody (FANA) tests Eight-well slides with composite, microtome-cryostat sections of liver, stomach (smooth muscle), and kidney from rats were obtained commercially (MeDiCa, Carlsbad, CA). Similar slides with the hemoflagellate Crithidia

DETECTION OF ANTINUCLEAR ANTIBODY IN CANINE SERA

39

luciliae were also obtained commercially (MeDiCa, Carlsbad, CA). Also, eight-well slides with monolayer HEp-2 cells (approximately 24-h cultures) were prepared. These slides were used for indirect immunofluorescent ANA (FANA) tests as described by Shull et al. (1983) at appropriate dilutions of sera. Passive hemagglutination assays Sera were diluted to 1:20 in 0.15 M NaCI with 0.01 M Na:HsPO4 buffer (pH 7.0) containing 0.5% bovine serum albumin. Serial twofold dilutions of the sera were then performed in the same dilutent. H u m a n erythrocytes (0negative) sensitized with nDNA or ssDNA antigens (Hemagen Diagnostics, Inc., Waltham, MA) were suspended in saline, and a volume of 50/zl was added to 50/d of the appropriate serum dilution. After mixing in a 96-well, V-bottomed microplate, the suspension was incubated for 90 min at 22 °C on a vibration-free surface. At that time, the plate was observed for agglutination patterns. A positive reaction was recorded only if agglutination of the erythrocytes was complete. Precision and FANA score assays Replicate tests (30) were performed to determine titers (by endpoint titration) and score values of a strong FANA positive serum, weak FANA positive serum, and a negative FANA serum. All the sera were from dogs with symptoms suggestive of SLE, as reported above. The replicate tests were performed on HEp-2 and rat liver antigen substrates as described above. For the ANA score system, sera were tested at only 1 : 20 dilutions for indirect immunofluorescence on HEp-2, liver, stomach (smooth muscle), kidney, and C. luciliae kinetoplasts. All sera were also tested for agglutinating antibody activity at a 1 : 20 dilution for nDNA and ssDNA by passive hemagglutination. The results were tabulated and compared. The score values were also calculated for 30 replicate tests at 1 : 40 dilution The tests were performed by three laboratory personnel over a 12-month interval. Statistical methods The results of replicate tests were averaged, and the coefficient of variation was expressed as the standard deviation as a percentage of the arithmetic mean (Tizard, 1982). Precision was also determined by calculation of geometric standard deviation (Taylor, 1986). Comparison of the mean score values of the different categories of dogs and HEp-2 or liver FANA titers was accomplished using analysis of variance, and least significant differences were calculated.

D.S. MCVEYAND W. SHUMAN

40 RESULTS

Precision o f FANA assays The estimates of precision of the respective assays are presented in Table 1. The CoV for endpoint titration of FANA assays using HEp-2 substrates were surprisingly high, 25.6% (175-207) and 71.6% (181-325). The CoV's for similar endpoint titration of FANA assays using liver cryostat-sections were lower, but still unacceptable. The CoV's of score values were much lower and more acceptable. The precision of the score assays was also calculated at a 1:40 dilution of the respective sera, and the CoV's were improved to 13.4% ( 175-207, mean = 4.0 ) and 13.2% ( 181-325, mean = 2.1 ). There were no significant differences (by analysis of variance) among personnel conducting the assays. Geometric CoV's demonstrated the same trend. The CoV's for serum 175207 were acceptable in all tests. A twofold difference in titer could be considered significant at P

Use of multiple antigen substrates to detect antinuclear antibody in canine sera.

The presence of antinuclear antibody in serum has been used to serologically support the clinical diagnosis of systemic lupus erythematosus. Most ofte...
455KB Sizes 0 Downloads 0 Views