Choice of Reference Assay for the Detection of Rotavirus in Fecal ...

6 downloads 0 Views 769KB Size Report
Dec 5, 1989 - a drop of 2% aqueous phosphotungstic acid (pH 6.2 to 6.5) for 30 s ... Results of testing stool specimens by ElAs or EM for detection of rotavirus.
Vol. 28, No. 6

JOURNAL OF CLINICAL MICROBIOLOGY, June 1990, p. 1280-1283 0095-1137/90/061280-04$02.00/0 Copyright © 1990, American Society for Microbiology

Choice of Reference Assay for the Detection of Rotavirus in Fecal Specimens: Electron Microscopy versus Enzyme Immunoassay PENELOPE H. DENNEHY,l.2* DOROTHY R. GAUNTLETT,' AND SARA E. SPANGENBERGER3 Division of Pediatric Infectious Diseases, Department of Pediatrics,' and Central Research Laboratories,3 Rhode Island Hospital, 593 Eddy Street, Providence, Rhode Island 02903, and Department of Pediatrics, Brown University, Providence, Rhode Island 029122 Received 5 December 1989/Accepted 9 March 1990

Two previously demonstrated sensitive and specific enzyme immunoassays (EIAs) for rotavirus, one using polyclonal and monoclonal antisera (TestPack Rotavirus [TPK]; Abbott Laboratories) and the other using only monoclonal anti-rotavirus antibodies (Rotaclone [RTC]; Cambridge BioScience Corporation), were evaluated as potential reference assays for rotavirus testing in comparison with direct negative-staining electron microscopy (EM), the current laboratory standard. Two hundred and seven stool samples collected consecutively during the winter of 1989 from children with acute diarrhea admitted to a ward for infants from 0 to 2 years of age were tested by the EIAs and by EM. TPK specimens were read visually; RTC results were read spectrophotometrically. Specimens with discordant EIA and EM results were further evaluated by a fluorescent focus assay. Specimens positive by EM and those negative by EM but positive by fluorescent focus assay were considered to be positive for rotavirus. Of the 207 stools tested, 35 (17%) were positive for rotavirus by these criteria. EM had a sensitivity of only 80%. Specificities were 100% for RTC and EM and 89% for TPK. These findings indicate that EM, although very specific, is relatively insensitive compared with a highly sensitive monoclonal antibody-based EIA. An EIA with high sensitivity and specificity, such as RTC, is a more appropriate reference standard for rotavirus testing. Rotavirus is a major cause of gastroenteritis in children throughout the world (1, 13) and is a frequently reported cause of nosocomial outbreaks of diarrhea (9, 25). In recent years, immunoassays have become the standard method for the detection of rotavirus in stool specimens. Commercial kits for detecting rotavirus are now available and include enzyme immunoassays (EIAs) and latex agglutination assays (8, 12, 16, 23, 27). The sensitivities and specificities of newly introduced immunoassays are usually evaluated against direct electron microscopy (EM), the current laboratory standard. However, the newer assays may be more sensitive than EM and require the use of a supplementary assay, such as a blocking test (3, 5-8, 15, 18, 21) or spin-amplified culture with fluorescent focus assay (4, 14, 16, 20, 28), to confirm this increased sensitivity. Two previously demonstrated sensitive and specific EIAs for rotavirus (6-8, 16, 18) were evaluated as potential reference assays for rotavirus testing in comparison with direct negative-staining EM.

was mixed with diluent and filtered, the latex particles, conjugate, and stool specimen were briefly incubated in a sample cup to permit binding of rotavirus to the antibody coated on the surface of the latex and to the conjugate. This mixture was poured from the sample cup through a plastic focuser onto the disposable reaction disk cartridge, which is designed to retain the latex particles and to allow rapid efficient washing of unreacted stool components and conjugate. A chromogenic substrate was then added. If the specimen contains rotavirus, a purple plus sign appears in the center of the reaction disk. A negative specimen is indicated by a purple minus sign. Specimens are read visually and graded on a scale of - to 4+. Specimens with readings of 1+ to 4+ were considered positive. Specimens giving a +/- reaction were considered to be inconclusive and were repeated. Rotaclone (Cambridge BioScience Corporation) uses murine monoclonal anti-rotavirus antibody directed against VP6, the group-specific antigen for all group A human rotaviruses, coated on plastic microdilution wells as the solid phase with the same monoclonal antibody conjugated to horseradish peroxidase as the detector antibody. The assay was run by following the manufacturer's instructions. Results were read spectrophotometrically at 450 nm on a Bio-Tek EL-308 plate reader. Specimens with an A450 greater than 0.150 were considered positive, as directed in the package insert. EM was performed as follows. A 10% stool suspension in phosphate-buffered saline was sonicated for 30 s and then g briefly to remove debris. A drop of spun at 8,800 supernatant was placed on Parafilm, and a 200-mesh Formvar-carbon-coated copper grid was floated on the surface for 2 min. The grid was air dried for 20 min and then floated on a drop of 2% aqueous phosphotungstic acid (pH 6.2 to 6.5) for 30 s, blotted, and then air dried for 10 min. Grids were examined for virus particles by using a Philips 300 electron

tector antibodies. After the stool specimen

MATERIALS AND METHODS Two hundred seven stool samples collected consecutively during the winter of 1989 from children (aged 1 week to 24 months) with acute diarrhea admitted to a ward for infants 0 to 2 years of age were tested by the EIAs by following the manufacturer's instructions and by direct negative-staining EM. Swab specimens were excluded. Stools were stored at 4°C and tested within 48 h. For the evaluation, samples were taken for each assay and the remaining stool was stored undiluted at -70°C. TestPack Rotavirus (Abbott Laboratories) is an EIA which uses a guinea pig polyclonal anti-rotavirus antibodycoated latex particles as a solid-phase immunosorbent and both murine monoclonal and bovine polyclonal anti-rotavirus antibodies conjugated to alkaline phosphatase as de*

X

Corresponding author. 1280

VOL. 28, 1990

REFERENCE ASSAY FOR ROTAVIRUS DETECTION: EM VERSUS EIA

1281

TABLE 1. Results of testing stool specimens by ElAs or EM for detection of rotavirus Test method

Rotaclone TestPack Rotavirus EM

No. of:

%

FalseTrueFalsepositives positives negatives negatives negatives positives negatives positivesa

Positive

True35 35 28

172 135 172

0 17 0

0 0 7

Sensitivity

Specificity

Negative

predictive predictive value

value

100

100

100

100

100 80

89 100

67 100

100 96

Diagnostic

No. of inconclusive

accuracy

results

82 97

0 20 0

acrc 100

a True-positive, Positive by EM or by one or both EIAs and fluorescent focus assay. b True-negative, Negative by EM.

microscope at a magnification of x52,000. A minimum of 15 grid squares were examined on each grid over a minimum period of 10 min. Discordant specimens were further evaluated by a fluorescent focus assay. Stools were diluted 1:10 in phosphatebuffered saline and vortexed. After centrifugation, 0.2 ml of supernatant was inoculated onto MA 104 cells seeded on cover slips in shell vials and spun at 3,000 x g for 60 min; minimal essential medium containing 0.5 ,ug of trypsin per ml was added, and the vials were incubated at 37°C for 24 h. The cover slip was then fixed with cold acetone, stained with anti-rotavirus monoclonal antibody followed by an anti-mouse fluorescein isothiocyanate-conjugated antibody (Chemicon International, Inc.), and read by fluorescence microscopy. A cover slip with 21 cell with typical cytoplasmic fluorescence was positive for rotavirus. To determine the sensitivities, specificities, and predictive values of the assays, each stool specimen was classified as either true-positive or true-negative. Specimens positive by EM and those negative by EM but positive by one or both EIAs and by the confirmatory fluorescent focus assay were considered to be positive for rotavirus. RESULTS By direct EM, 28 of the 207 stool samples from children with diarrhea were positive for rotavirus, while 179 were negative. A total of 35 stool samples were positive by Rotaclone, and 52 were positive by TestPack Rotavirus. A fluorescent focus assay was done on the 24 specimens in which the results of EM and at least one of the immunoassays were discordant. The results of the fluorescent focus assay were in complete agreement with those of the Rotaclone EIA but not with those of EM or TestPack Rotavirus. Seven specimens positive by Rotaclone and TestPack Rotavirus but negative by EM contained rotavirus, while none of the 17 TestPack Rotavirus-positive, Rotaclone- and EMnegative specimens had detectable virus by fluorescent focus assay. Five of the seven specimens positive by this assay but negative by EM had low absorbance readings (c0.500 absorbance units) in the Rotaclone EIA, while only 1 of 28 specimens positive by EM had a reading of less than 0.500. Twenty additional specimens which were inconclusive by TestPack Rotavirus but negative by EM and Rotaclone were also tested by fluorescent focus assay and found to be negative for rotavirus. The results of the EIAs and EM are summarized in Table 1. Both Rotaclone and TestPack Rotavirus detected all 35 true-positives for rotavirus (defined as either positive by EM or positive by one or both EIAs and by fluorescent focus assay). Seven specimens positive by the immunoassays but negative by EM had virus by fluorescent focus assay. TestPack Rotavirus had 17 false-positives, which reduced

the specificity of the test to 89%, the positive predictive value to 67%, and the diagnostic accuracy to 82%. DISCUSSION The choice of a reference standard for the evaluation of rotavirus assays is important in determining the clinical utility of assays and in comparisons of evaluations of assays done in a variety of laboratory settings. Reference standards used in prior evaluations of rotavirus assays have included direct EM (5, 6, 15, 22, 24, 27), immune EM (2, 3, 20), solid-phase immune EM (11), polyacrylamide gel electrophoresis of rotaviral RNA (21), a spin-amplified culture with fluorescent focus assay (16), or a reference EIA (8, 10, 12, 19, 29). The ideal reference method for detection of rotavirus should have high degrees of sensitivity and specificity and should be a method which can be consistently performed in the laboratory setting. High degrees of sensitivity are desirable in rotavirus assays, as specimens from patients late in the course of the illness (29) or from adults may have low levels of viral shedding. Specificity is equally important in a reference assay if false-positives and reruns due to inconclusive results are to be avoided. Each of the methods used as reference standards in previous evaluations of rotavirus testing have both advantages and disadvantages. Of the EM methods used as reference assays, direct EM has been most widely used. Direct EM is easily standardized and can be performed consistently from laboratory to laboratory; sensitivity may be a problem, however. Direct EM detects approximately 108 virus particles per ml (24) and has been found to be less sensitive than immune electron microscopy (IEM), EIA, or solid-phase IEM (SPIEM) in previous evaluations (2, 3, 11, 20, 26). IEM detects approximately 106 particles per ml (24), an increase over direct EM of approximately 100-fold (3) and similar to the levels of sensitivity of EIA (12, 24). IEM, however, has the disadvantage of difficulties in standardization, as the antibodies used differ from laboratory to laboratory and may vary in their abilities to recognize all strains and subgroups of rotavirus. In addition, if the concentration of viral particles in the fecal suspension is low, the few aggregates formed in the reaction may be missed on the grid. Finally, optimum antibodyantigen concentrations must be achieved for aggregation to occur, and this requires careful evaluation of antibody dilutions to be used in IEM. SPIEM has similar advantages and disadvantages compared with IEM. Both Svensson et al. (26) and Gerna et al. (11) found SPIEM to be 10 to 30 times more sensitive than direct EM, but Svensson et al. note that the procedure works well only when careful titration of protein A and serum is done to achieve the highest virustrapping efficiency. Therefore, although sensitive, SPIEM would be difficult to standardize from laboratory to labora-

1282

DENNEHY ET AL.

tory unless single lots of antibody and protein A were used universally. Culture methods are the preferred reference standard in evaluations of diagnostic methods in other viral diseases but have not been used in rotavirus infections until recently because of the difficulties in growing rotavirus in culture. Spin-amplified culture methods with detection of infected cells by an anti-rotavirus antibody have high sensitivity and specificity. A fluorescent focus assay described by Keswick and colleagues was able to detect 10 infectious virus particles per ml of inoculant (14), while that described by Ward et al. detected 5 x 102 infectious particles per ml (28). Assays using immunoperoxidase and 1251I-conjugated antibodies in place of fluorescein-conjugated anti-rotavirus antibody have been found to be equally capable of detecting rotavirus in stool specimens (4, 17). Ward et al. have estimated that an average of 1 of every 46,000 rotavirus particles is infectious (28); therefore, the fluorescent focus assay can detect specimens with approximately 105 to 107 virus particles per ml, a figure confirmed in a recent study by Lipson and ZelinskyPapez (16). This method has the disadvantage of being very labor intensive, and it is difficult to use in the evaluation of a large number of specimens. The EIAs evaluated in this study have previously demonstrated 100% sensitivity when confirmed with blocking assays (7, 8) and are superior to direct EM for detection of rotavirus when confirmed by fluorescent focus assay in this evaluation. The fluorescent focus assay has been used to confirm specimens positive by EIA but negative by EM. Fluorescent focus assay results suggest that the EIAs were accurate in detecting rotavirus in seven discordant specimens. In five of the seven specimens, absorbance data would suggest that a small number of viral particles were present. In previous studies, monoclonal antibody-based EIAs have been found to be capable of detecting rotavirus in specimens containing 10-fold-fewer visions than those positive by direct EM or by a polyclonal antibody-based EIA and to be comparable with SPIEM in detecting rotavirus in clinical specimens (10). The completely monoclonal antibody-based EIA (Rotaclone) evaluated in this study appears to be superior in specificity to the monoclonal combined with polyclonal antibody-based test (TestPack Rotavirus) for detection of rotavirus. The superiority of the monoclonal antibody-based assay is likely due to the differences in antigen recognition by monoclonal and polyclonal antisera. Polyclonal antisera contain antibodies directed against multiple antigenic determinants and with a wide range of affinities, while monoclonal antibodies are directed against a single epitope and have a single affinity. As a result, an EIA using monoclonal antibodies is usually more specific and has fewer inconclusive results than an EIA with polyclonal detecting antibodies which may have cross-reactivity with other components of stool (30). TestPack Rotavirus had an unexpectedly high false-positivity rate and a large number of stools with inconclusive test results, both likely the result of cross-reactivity with other substances in the stool. These data indicate that inconclusive TestPack Rotavirus results should be considered negative. The findings of this study indicate that direct EM, although very specific, is relatively insensitive compared with a highly sensitive monoclonal antibody-based EIA. An EIA with high sensitivity and specificity, such as Rotaclone, is a more appropriate reference standard for rotavirus assay evaluation.

J. CLIN. MICROBIOL.

ACKNOWLEDGMENTS This study was supported in part by a grant from Cambridge BioScience Corporation. We thank Georges Peter for review of the manuscript. LITERATURE CITED 1. Brandt, C. D., H. W. Kim, W. J. Rodriguez, J. O. Arrobio, B. C. Jeffries, E. P. Stallings, C. Lewis, A. J. Miles, R. M. Chanock, A. Z. Kapikian, and R. H. Parrott. 1983. Pediatric viral gastroenteritis during eight years of study. J. Clin. Micro-

biol. 18:71-78. 2. Brandt, C. D., H. W. Kim, W. J. Rodriguez, L. Thomas, R. H. Yolken, J. O. Arrobio, A. Z. Kapikian, R. H. Parrott, and R. M. Chanock. 1981. Comparison of direct electron microscopy, immune electron microscopy, and rotavirus enzyme-linked immunosorbent assay for detection of gastroenteritis viruses in children. J. Clin. Microbiol. 13:976-981. 3. Brooks, R. G., L. Brown, and R. B. Franklin. 1989. Comparison of a new rapid test (TestPack Rotavirus) with standard enzyme immunoassay and electron microscopy for the detection of rotavirus in symptomatic hospitalized children. J. Clin. Microbiol. 27:775-777. 4. Cevenini, R., F. Rumpianesi, R. Mazzaracchio, M. Donati, E. Falceri, and I. Sarov. 1984. A simple immunoperoxidase method for detecting enteric adenovirus and rotavirus in cell culture. J. Infect. 8:22-27. 5. Chernesky, M., S. Casticiano, J. Mahony, and D. DeLong. 1985. Examination of the Rotazyme Il enzyme immunoassay for the diagnosis of rotavirus gastroenteritis. J. Clin. Microbiol. 22: 462-464. 6. Chernesky, M., S. Casticiano, J. Mahony, M. Spiewak, and L. Schaefer. 1988. Ability of TESTPACK ROTAVIRUS enzyme immunoassay to diagnose rotavirus gastroenteritis. J. Clin. Microbiol. 26:2459-2461. 7. Dennehy, P. H., and D. R. Gauntlett. 1988. Evaluation of a new enzyme immunoassay (TESTPACK Rotavirus) for the detection of rotavirus in fecal specimens. Diagn. Microbiol. Infect. Dis. 11:201-203. 8. Dennehy, P. H., D. R. Gauntlett, and W. E. Tente. 1988. Comparison of nine commercial assays for detection of rotavirus in fecal specimens. J. Clin. Microbiol. 26:1630-1634. 9. Dennehy, P. H., and G. Peter. 1985. Risk factors associated with nosocomial rotavirus infection. Am. J. Dis. Child. 139:935-939. 10. Doern, G. V., J. E. Herrmann, P. Henderson, D. Stobbs-Walro, D. M. Perron, and N. R. Blacklow. 1986. Detection of rotavirus with a new polyclonal antibody enzyme immunoassay (Rotazyme II) and a commercial latex agglutination test (Rotalex): comparison with a monoclonal antibody enzyme immunoassay. J. Clin. Microbiol. 23:226-229. 11. Gerna, G., A. Sarasini, N. Passarani, M. Torsellini, M. Parea, and M. Battaglia. 1987. Comparative evaluation of a commercial enzyme-linked immunoassay and solid-phase immune electron microscopy for rotavirus detection in stool specimens. J. Clin. Microbiol. 25:1137-1139. 12. Gilchrist, M. J. R., T. S. Bretl, K. Moultney, D. R. Knowlton, and R. L. Ward. 1987. Comparison of seven kits for detection of rotavirus in fecal specimens with a sensitive, specific enzyme immunoassay. Diagn. Microbiol. Infect. Dis. 8:221-228. 13. Kapikian, A. Z., H. W. Kim, R. G. Wyatt, W. L. Cline, J. O. Arrobio, C. D. Brandt, W. J. Rodriquez, D. A. Sack, R. M. Chanock, and R. H. Parrott. 1976. Human reovirus-like agent as the major pathogen associated with "winter" gastroenteritis in hospitalized infants and young children. N. Engl. J. Med. 294:965-972. 14. Keswick, B. H., L. K. Pickering, H. L. DuPont, and W. E. Woodward. 1983. Survival and detection of rotaviruses on environmental surfaces in day care centers. Appl. Environ. Microbiol. 46:813-816. 15. Knisley, C. V., J. Bednarz-Prashad, and L. K. Pickering. 1986. Detection of rotavirus in stool specimens with monoclonal and polyclonal antibody-based assay systems. J. Clin. Microbiol. 23:897-900. 16. Lipson, S. M., and K. A. Zelinsky-Papez. 1989. Comparison of

VOL. 28, 1990

17.

18. 19.

20. 21.

22. 23.

REFERENCE ASSAY FOR ROTAVIRUS DETECTION: EM VERSUS EIA

four latex agglutination and three enzyme-linked immunosorbent assays for the detection of rotavirus in fecal specimens. Am. J. Clin. Pathol. 92:637-643. Liu, S., C. Birch, A. Coulepis, and I. Gust. 1984. Radioimmunofocus assay for detection and quantification of human rotavirus. J. Clin. Microbiol. 20:347-350. Marchlewicz, B., M. Spiewak, and J. Lampinen. 1988. Evaluation of Abbott TESTPACK ROTAVIRUS with clinical specimens. J. Clin. Microbiol. 26:2456-2458. Miotti, P. G., J. Eiden, and R. H. Yolken. 1985. Comparative efficacy of commercial immunoassays for the diagnosis of rotavirus gastroenteritis during the course of infection. J. Clin. Microbiol. 22:693-698. Morinet, F., F. Ferchal, R. Colimon, and Y. Perol. 1984. Comparison of six methods for detecting human rotavirus in stools. Eur. J. Clin. Microbiol. 3:136-140. Pacini, D. L., M. T. Brady, C. T. Budde, M. J. Connell, V. V. Hamparian, and J. H. Hughes. 1988. Polyacrylamide gel electrophoresis of RNA compared with polyclonal- and monoclonalantibody-based enzyme immunoassays for rotavirus. J. Clin. Microbiol. 26:194-197. Pai, C. H., M. S. Shahrabadi, and B. Ince. 1985. Rapid diagnosis of rotavirus gastroenteritis by a commercial latex agglutination test. J. Clin. Microbiol. 22:846-850. Pickering, L. K., and B. H. Keswick. 1986. Rotaviruses and norwalk viruses, p. 301-327. In S. Specter and G. J. Lancz

1283

(ed.), Clinical virology manual. Elsevier, New York. 24. Rubenstein, A. S., and M. F. Miller. 1982. Comparison of an enzyme immunoassay with electron microscopic procedures for detecting rotavirus. J. Clin. Microbiol. 15:938-944. 25. Ryder, R. W., J. E. McGowan, Jr., M. H. Hatch, and E. L. Palmer. 1977. Reovirus-like agent as a cause of nosocomial diarrhea in infants. J. Pediatr. 90:698-702. 26. Svensson, L., M. Grandien, and C. Pettersson. 1983. Comparison of solid-phase immune electron microscopy by use of protein A with direct electron microscopy and enzyme-linked immunosorbent assay for detection of rotavirus in stool. J. Clin. Microbiol. 18:1244-1249. 27. Thomas, E. E., M. L. Puterman, E. Kawano, and M. Curran. 1988. Evaluation of seven immunoassays for detection of rotavirus in pediatric stool samples. J. Clin. Microbiol. 26:11891193. 28. Ward, R. L., D. R. Knowlton, and M. J. Pierce. 1984. Efficacy of human rotavirus propagation in cell culture. J. Clin. Microbiol. 19:748-753. 29. Yolken, R. H., and F. J. Leister. 1981. Evaluation of enzyme immunoassays for the detection of human rotavirus. J. Infect. Dis. 144:379. 30. Yolken, R. H., and P. J. Stopa. 1979. Analysis of nonspecific reactions in enzyme-linked immunosorbent assay testing for human rotaviruses. J. Clin. Microbiol. 10:703-707.