DIAGN MICROBIOLINFECT DIS 1992;15:673-678
673
A Rapid Culture Alternative to the Shell-Vial Method for the Detection of Herpes Simplex Virus F. Brent Johnson and Elizabeth M. Visick
The rapid test for detection of herpes simplex virus (HSV) in clinical specimens based on infection of cells in suspension (SI test) was compared to the shell-vial culture (SVC) method and conventional culture. Mink lung cells were used throughout the study. Detection of HSV was not significantly different whether using SI or SVC. The sensitivity of SI in detecting HSV, when compared with conventional culture, was 93.0%
using 0.1 ml inocula and 98.3% using 0.5 ml inocula. The time to obtain a final result with both SI and SVC was 1 day compared with 1-7 days by conventional culture. The SI method detected both HSV type-1 and HSV type-2 clinical isolates. The SI technique is a simple method for the rapid detection of HSV and can yield diagnostic results with a minimum of technical manipulations.
INTRODUCTION
the detection of infected cells hours or days prior to the appearance of virus-induced cytopathic effects. Furthermore, centrifugation of the culture, which is contained in shell vials or tissue culture plates, enhances virus replication and shortens the time to a positive culture result (Gleaves et al., 1985b; Oefinger et al., 1988; Pruneda and Almanza, 1987; Tse et al., 1989). Such a procedure is often referred to as spin amplification or the shell-vial centrifugation culture (SVC). SVC has been used for the detection of cytomegalovirus (CMV) (Gleaves and Meyers, 1987; Gleaves et al., 1984 and 1985a; Paya et al., 1988; Swensen and Kaplan, 1987) and HSV (Gleaves et al., 1985b; MacDonald et al., 1988; Oefinger et al., 1988; Peterson et al., 1988; Salmon et al., 1986; Woods and Mills, 1988) as well as Chlamydia trachomatis, the original application for SVC. Disadvantages of SVC include the requirement for advance preparation of shell-vial monolayer cultures. This requires an advance estimation of the number of cultures likely to be submitted to the laboratory for diagnosis, or requires a 1-day delay after receipt of the specimens to prepare the monolayers, or results in the use of older monolayers that may not be as sensitive to virus replication as are freshly prepared monolayers. Another disadvantage with SVC is the time-consuming centrifugation step itself. Recently, a rapid culture alternative to SVC was described (Luker et al., 1991) that involves infecting the culture cells before attachment while they are still in suspension by mix-
Herpes simplex virus (HSV) is among the easiest of viruses to isolate in the diagnostic virology laboratory; however, in the past several years, laboratories have been challenged with the task of detecting HSV in clinical specimens as rapidly as possible. Conventional culture has been the standard method of isolation because of its sensitivity, but, even so, attempts have been made to optimize culture conditions to make it more rapid and sensitive by employing the most sensitive cells (Espy et al., 1991; Ginnow Johnston and Siegel, 1990; Ginnow Johnston et al., 1990; Peterson et al., 1988). Although advances have identified cell types that are more sensitive to HSV than others, conventional culture remains a relatively slow method. In recent years, the general approach to obtain more rapid results, to immunostain infected cells with fluorescent antibody or immunoperoxidase stains, has been refined. These techniques enable From the Department of Microbiology, Brigham Young University, Provo; and MicroVir Laboratories, Orem, Utah, USA. Address reprint requests to Dr. F.B. Johnson, 887 WIDB, Department of Microbiology,Brigham Young University, Provo, UT 84602, USA Received 12 December 1991; revised and accepted 14 February 1992. © 1992Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010 0732-8893/92/$5.00
674
ing the virus inoculum with the cells and then allowing the cells to settle through the cell culture medium without centrifugation. The cells then form a monolayer that is immunostained after 18-24 hr and is assessed for infected cells. This technique was referred to as the suspension-infection (SI) method. The present study compared the performance of the SI and SVC methods in a clinical virology laboratory for the detection of HSV and referenced the results obtained by these methods to conventional culture results.
MATERIALS A N D METHODS Specimens Clinical specimens were gathered from hospitals and clinics in northern and central Utah. The specimens were primarily from female genital tract sources collected from pubescent and gravid females, although specimens from genital lesions in males and nongenital sites in both males and females were submitted. Specimens were transported in Richards liquid transport medium or on swabs in transtubes such as Virocult and Culturette (Johnson, 1990). Specimens were received in the laboratory and processed within 1-2 days of collection.
Conventional Culture The method and conditions used for conventional culture previously described (Johnson et al., 1985) were modified in the present study as follows: Culture was performed in mink lung cells (MV 1 Lu, ATCC CCL 64; American Type Culture Collection, Rockville, MD) contained in flat-sided cell culture tubes (110 x 16 ram; Nunc, Roskilde, Denmark). All cell cultures were prepared in house and were used within 2 days of preparation. The cell culture medium was Dulbecco modified minimal essential medium (Sigma Chemical Company, St Louis, MO) with 2% fetal bovine serum (HyClone Laboratories, Logan, UT), 0.11% sodium bicarbonate, 2 mM HEPES (N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid) buffer, and 50 p~g of gentamicin per milliliter. Cultures were inoculated with 0.1-0.5 ml (depending on the experiment) of specimen clarified by centrifugation (60 g for 5 min), virus adsorption was allowed to proceed for 60 rain, and then the cultures were renewed with 2 ml of fresh medium. They were incubated at 36°C and observed daily for cytopathic effect. Upon the appearance of cytopathic effect or at the end of 7 days, all cultures were immunostained to identify infected cells. The method for immunostaining, an immunoperoxidase method, was previously described (Luker et al., 1991).
F.B. Johnson and E.M. Visick
Suspension-Infection Culture Mink lung cells for the SI cultures were prepared by standard trypsinization of confluent monolayer cultures. The loosened cell sheets were suspended in cell culture medium (10 ml per 75-cm 2 flask) and then stored at 2°-8°C in flint glass vials (19 x 65 mm; American Scientific Products, McGaw Park, IL), as previously described (Luker et al., 1991). As the cells were stored, they settled to the bottom of the vials. Upon use, they were resuspended in the original medium in the storage vial by gentle pipetting. The method for SI culture consisted of combining in flatsided Nunc culture tubes 2 ml of Dulbecco modified minimal essential medium that contained 10% fetal bovine serum, 0.11% sodium bicarbonate, 2 mM HEPES buffer, penicillin (100 U/ml), streptomycin (100 ~g/ml), nystatin (6 ~g/ml), and amphotericin B (0.5 p,g/ml), 0.2 ml of cell suspension, and 0.1-0.5 ml (depending on the experiment) of patient specimen clarified by low-speed centrifugation (60 g for 5 min). After the caps were securely tightened, the contents of the tubes were mixed by two gentle inversions, whereupon they were incubated at 36°C for 18-24 hr with no further agitation. Following incubation, the cultures were fixed and stained by the immunoperoxidase method previously described (Luker et al., 1991). The criterion used for defining a positive culture was either the presence of at least one immunostained focus consisting of three or more cells with intact cell morphology or the presence of at least two individually stained cells located in different areas of the monolayer. These latter cells were also required to exhibit cell morphology typical of the cultured cell line. In practice, most positive cultures exhibited far more foci than required by the minimum criterion.
Shell-Vial Culture Shell vials (screw-capped polypropylene vials, 57 x 16.5 mm; Sarstedt, Newton, SC) containing monolayers of mink lung cells on 12-mm coverslips were used for SVC. The vials were seeded with cells in our laboratory and used within 24 hr of seeding. The growth media were decanted prior to inoculation of the cultures with 0.1 ml of sample. The vials were then centrifuged for 60 min at 700 g at room temperature. After centrifugation, 1 ml of growth medium (described above in methods for suspensioninfection culture) was added to each vial, and the vials were incubated for 16-24 hr at 36°C. The culture media were then decanted, and the monolayers washed with 1 ml of phosphate-buffered saline (PBS) and then fixed in 1 ml of cold acetone. After airdrying, the cultures were immunostained with fluorescein-conjugated type-common HSV antibody
675
Detection of HSV by Suspension-Infection
(Dynatech Diagnostics, South W i n d h a m , ME) and evaluated by fluorescence microscopy.
Performance of S u s p e n s i o n - I n f e c t i o n and S h e l l - V i a l Culture
Herpes S i m p l e x Virus S e r o t y p i n g Specimens being cultured for routine evaluation by SI were occasionally a c c o m p a n i e d b y a request for HSV serotyping should the culture exhibit a positive result. In such cases, duplicate tubes were inoculated with specimen. At 16-24 hr, one tube was stained by the i m m u n o p e r o x i d a s e m e t h o d and evaluated for the presence of infected cells. If the culture was positive, then the second tube was incubated further until 3 + to 4 + cytopathic effects (CPEs) developed. The infected, d e t a c h e d cells were h a r v e s t e d by centrifugation of the culture tube (60 g for 5 min), washed with 2 ml of PBS, and recentrifuged. The infected cell pellets were s u s p e n d e d in 0.1 ml PBS, spotted in two wells each on teflon-coated glass slides, allowed to air-dry, fixed in cold acetone, and t h e n airdried again. Specimens prepared in this m a n n e r were stained with type-specific m o n o c l o n a l antibodies (Syva Corporation, Palo Alto, CA), one well with type-1 reagent and one well with type-2 reagent. The stained cells were e v a l u a t e d b y fluorescence microscopy.
Data A n a l y s i s The SI trial 1 data and the SVC data (Tables 1 and 2) were analyzed by the chi-squared test of independence.
TABLE I
RESULTS
Clinical specimens (n -- 530) in trial 1 w e r e tested in parallel for HSV b y conventional culture, SI, and SVC. Mink lung cells of identical passage levels were u s e d in all tests. Each culture was inoculated with 0.1 ml of specimen. Of the 115 positive specimens (overall isolation rate of 21.7%), 107 were detected in overnight SI cultures for a sensitivity value of 93.0% (Table 1). Of the 115 positive specimens, 109 were detected by SVC for sensitivity value of 94.8% (Table 2). The data (numbers of test-positive and negative specimens) obtained in trial 1 (Table 1) b y the SI m e t h o d were c o m p a r e d with the data obtained b y the SVC m e t h o d (Table 2) for statistical significance of the two sensitivity values. A chisquared value of 0.304 was obtained, indicating no significant difference b e t w e e n results obtained by SI and SVC m e t h o d s at a 95% confidence level (critical value = 3.841). The six true-positive specimens not detected by SVC were also not detected b y SI. In addition, two more positive specimens were not detected b y SI. All these samples contained low a m o u n t s of HSV and were referred to as low-level specimens. Lowlevel specimens were defined as true-positive specimens detected by either conventional culture after 3 or more days of incubation or rapid culture (either SI or SVC) with five or f e w e r loci of infected cells.
Performance C o m p a r i s o n of the Suspension-Infection (SI) Test and Conventional Culture a a Trial 1"
SI Result + -
Conventional Culture Result + 107 8
0 415 b
SI Result
+ -
Trial 2c
Conventional Culture Result + 174 3
0 357
aBoth tests were performed using mink lung cells. ~530specimens. Sensitivity = 107/107 + 8 = 93.0%; specificity = 415/415 + 0 = 100%;positive predictive value = 107/107 + 0 = 100%;and negative predictive value = 415/415 + 8 = 98.1%. '534 specimens. Sensitivity = 174/174 + 3 = 98.3%; specificity = 357/357 + 0 = 100%;positive predictive value = 174/174 + 0 = 100%; and negative predictive value = 357/357 + 3 = 99.2%.
676
F.B. Johnson and E.M. Visick
By these criteria, 15 (13%) of the 115 true-positive specimens were low-level specimens. Hence, the SI method detected 7 (47%) of 15 of the low-level specimens and SVC detected 9 (60%) of 15 of them. In trial 2, 534 additional samples were tested in the same manner as in trial 1, except SVCs were not included, and each culture was inoculated with 0.5 ml of sample instead of 0.1 ml. Of 177 true-positive specimens (overall isolation rate of 33.1%), 174 were detected in overnight SI cultures for a sensitivity value of 98.3% (Table 1). In this trial, 58 of the 177 true-positive specimens (32.8%) were low-level specimens. Hence, the SI method detected 55 (94.8%) of 58 of the low-level specimens. The much better performance of SI in trial 2 in detecting low-level positive specimens was attributed to the increased inoculum size.
Time to Test Result The culture results obtained in trial 2 were plotted to show the difference in time required to obtain a positive or negative test result between the SI and conventional culture methods (Figure 1). The SI test, concluded in 1 day, detected 98.3% of the positive specimens and 100% of the negative specimens. By conventional culture, only 30% of the positive specimens were detected by day 1. By day 2, 67% of the positive specimens had been detected, 84% by day 3, 90% by day 4, 95% by day 5, and 100% by day 6. Incubation was continued to day 7, but all of the negative specimens were detected by day 6.
Herpes S i m p l e x Virus Serotyping It was assumed that the SI technique detected both HSV type-1 and type-2 strains because the total number of isolates detected by SI approached the total number of isolates detected by conventional culture. However, it was important to demonstrate directly the recovery of both types. Isolates were serotyped by request of the submitting physician, thus random isolates were subjected to typing. Of
TABLE 2
DISCUSSION The results of this study show that (a) detection of HSV in clinical samples is not significantly different whether using the SI test or SVC, (b) the sensitivity of SI can exceed 98%, (c) the SI test is a rapid test when compared with conventional culture, and (d) both HSV-1 and HSV-2 strains are detected by the SI method. It was also shown that the amount of inoculum used influences the number of positive specimens detected, hence, the sensitivity of the test. This finding was expected considering that many clinical specimens contain low amounts of infectious virus and increasing the inoculum size should augment the probability of including virus in the inoculum. In this study, experiments were not conducted using inocula >0.1 ml in SVC, and direct comparisons between SI and SVC with large inocula were not made. However, it is likely that the sensitivity value of SVC in our hands would increase somewhat with inoculum size as was found with SI. Studies published by others using SVC for the detection of HSV have indicated sensitivities of 70%100% compared with CPE detection in conventional culture (Espy et al., 1991; Gleaves et al., 1985b; Peterson et al., 1988; Woods and Mills, 1988). Because HSV-1 and HSV-2 appear to attach to different host cell receptors (Vahlne et al., 1979 and 1980) and it has not previously been shown that MV1 Lu cells in suspension display both types of receptors, it was of significance to determine whether clinical strains of both HSV-1 and HSV-2 were detected by the SI technique. The results revealed that strains of both types were isolated. There is no indication that SI, or the stain reagents used in this method, fail to detect any clinical HSV strains of either type because of serotypic differences. Fewer manipulations and procedural steps in a clinical test for an infectious agent may result in many benefits, such as, less hands-on time required to
Performance Comparison of Shell-Vial Culture and Conventional Culture in Mink Lung Cellsa
Shell-Vial Culture Result + -
82 strains that were typed, 35% were type-1 strains and 65% were type-2 strains (Table 3).
Conventional Culture Result +
-
109 6
0 415
aSensitivity = 109/109 + 6 = 94.8%; specificity~ 415/415 + 0 = 100%;positivepredictive value = 109/109 + 0 = 100%;and negative predictive value = 415/415 + 6 = 98.6%.
Detection of HSV by Suspension-Infection
SUSPE NSION-IN F E CTION :
CONVENTIONAL CULTURE:
B
677
[~
number
~ percent I00
number
percent
074)
LU > m
I--
80
0
0. u) 6 0 z IJJ u uJ 4 0 o.
20
I
2
T IME
3
4
5
6
7
(DAYS)
POSTINOCULATION
FIGURE 1 Number and cumulative percentage of specimens observed to be positive by conventional culture by day after inoculation compared with 1-day result by suspension-infection culture. A total of 177 specimens were found positive by conventional culture after 6 days of incubation, whereas 174 (98.3%) of them were detected by the suspension-infection method in 1 day.
perform the test, less opportunity for experimental error, and less exposure of laboratory personnel to infectious organisms. The SI technique for the detection of HSV has these advantages. It also has the advantage of being more rapid than conventional culture while approaching culture sensitivity. Its advantages over SVC are that it does not require the time-consuming centrifugation step, fewer technical manipulations are necessary, and it does not require the use of shell vials, which are more difficult to open and close without contaminating laboratory surfaces than are screw-capped tubes. TABLE 3
It has been argued, in effect, that when optimally sensitive cells are used in conventional culture, detection of most positive samples is obtained early in the incubation period, making conventional culture a relatively rapid and inexpensive test (Ginnow Johnston and Siegel, 1990). It must be recognized that even under such conditions, however, typically only
673
A Rapid Culture Alternative to the Shell-Vial Method for the Detection of Herpes Simplex Virus F. Brent Johnson and Elizabeth M. Visick
The rapid test for detection of herpes simplex virus (HSV) in clinical specimens based on infection of cells in suspension (SI test) was compared to the shell-vial culture (SVC) method and conventional culture. Mink lung cells were used throughout the study. Detection of HSV was not significantly different whether using SI or SVC. The sensitivity of SI in detecting HSV, when compared with conventional culture, was 93.0%
using 0.1 ml inocula and 98.3% using 0.5 ml inocula. The time to obtain a final result with both SI and SVC was 1 day compared with 1-7 days by conventional culture. The SI method detected both HSV type-1 and HSV type-2 clinical isolates. The SI technique is a simple method for the rapid detection of HSV and can yield diagnostic results with a minimum of technical manipulations.
INTRODUCTION
the detection of infected cells hours or days prior to the appearance of virus-induced cytopathic effects. Furthermore, centrifugation of the culture, which is contained in shell vials or tissue culture plates, enhances virus replication and shortens the time to a positive culture result (Gleaves et al., 1985b; Oefinger et al., 1988; Pruneda and Almanza, 1987; Tse et al., 1989). Such a procedure is often referred to as spin amplification or the shell-vial centrifugation culture (SVC). SVC has been used for the detection of cytomegalovirus (CMV) (Gleaves and Meyers, 1987; Gleaves et al., 1984 and 1985a; Paya et al., 1988; Swensen and Kaplan, 1987) and HSV (Gleaves et al., 1985b; MacDonald et al., 1988; Oefinger et al., 1988; Peterson et al., 1988; Salmon et al., 1986; Woods and Mills, 1988) as well as Chlamydia trachomatis, the original application for SVC. Disadvantages of SVC include the requirement for advance preparation of shell-vial monolayer cultures. This requires an advance estimation of the number of cultures likely to be submitted to the laboratory for diagnosis, or requires a 1-day delay after receipt of the specimens to prepare the monolayers, or results in the use of older monolayers that may not be as sensitive to virus replication as are freshly prepared monolayers. Another disadvantage with SVC is the time-consuming centrifugation step itself. Recently, a rapid culture alternative to SVC was described (Luker et al., 1991) that involves infecting the culture cells before attachment while they are still in suspension by mix-
Herpes simplex virus (HSV) is among the easiest of viruses to isolate in the diagnostic virology laboratory; however, in the past several years, laboratories have been challenged with the task of detecting HSV in clinical specimens as rapidly as possible. Conventional culture has been the standard method of isolation because of its sensitivity, but, even so, attempts have been made to optimize culture conditions to make it more rapid and sensitive by employing the most sensitive cells (Espy et al., 1991; Ginnow Johnston and Siegel, 1990; Ginnow Johnston et al., 1990; Peterson et al., 1988). Although advances have identified cell types that are more sensitive to HSV than others, conventional culture remains a relatively slow method. In recent years, the general approach to obtain more rapid results, to immunostain infected cells with fluorescent antibody or immunoperoxidase stains, has been refined. These techniques enable From the Department of Microbiology, Brigham Young University, Provo; and MicroVir Laboratories, Orem, Utah, USA. Address reprint requests to Dr. F.B. Johnson, 887 WIDB, Department of Microbiology,Brigham Young University, Provo, UT 84602, USA Received 12 December 1991; revised and accepted 14 February 1992. © 1992Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010 0732-8893/92/$5.00
674
ing the virus inoculum with the cells and then allowing the cells to settle through the cell culture medium without centrifugation. The cells then form a monolayer that is immunostained after 18-24 hr and is assessed for infected cells. This technique was referred to as the suspension-infection (SI) method. The present study compared the performance of the SI and SVC methods in a clinical virology laboratory for the detection of HSV and referenced the results obtained by these methods to conventional culture results.
MATERIALS A N D METHODS Specimens Clinical specimens were gathered from hospitals and clinics in northern and central Utah. The specimens were primarily from female genital tract sources collected from pubescent and gravid females, although specimens from genital lesions in males and nongenital sites in both males and females were submitted. Specimens were transported in Richards liquid transport medium or on swabs in transtubes such as Virocult and Culturette (Johnson, 1990). Specimens were received in the laboratory and processed within 1-2 days of collection.
Conventional Culture The method and conditions used for conventional culture previously described (Johnson et al., 1985) were modified in the present study as follows: Culture was performed in mink lung cells (MV 1 Lu, ATCC CCL 64; American Type Culture Collection, Rockville, MD) contained in flat-sided cell culture tubes (110 x 16 ram; Nunc, Roskilde, Denmark). All cell cultures were prepared in house and were used within 2 days of preparation. The cell culture medium was Dulbecco modified minimal essential medium (Sigma Chemical Company, St Louis, MO) with 2% fetal bovine serum (HyClone Laboratories, Logan, UT), 0.11% sodium bicarbonate, 2 mM HEPES (N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid) buffer, and 50 p~g of gentamicin per milliliter. Cultures were inoculated with 0.1-0.5 ml (depending on the experiment) of specimen clarified by centrifugation (60 g for 5 min), virus adsorption was allowed to proceed for 60 rain, and then the cultures were renewed with 2 ml of fresh medium. They were incubated at 36°C and observed daily for cytopathic effect. Upon the appearance of cytopathic effect or at the end of 7 days, all cultures were immunostained to identify infected cells. The method for immunostaining, an immunoperoxidase method, was previously described (Luker et al., 1991).
F.B. Johnson and E.M. Visick
Suspension-Infection Culture Mink lung cells for the SI cultures were prepared by standard trypsinization of confluent monolayer cultures. The loosened cell sheets were suspended in cell culture medium (10 ml per 75-cm 2 flask) and then stored at 2°-8°C in flint glass vials (19 x 65 mm; American Scientific Products, McGaw Park, IL), as previously described (Luker et al., 1991). As the cells were stored, they settled to the bottom of the vials. Upon use, they were resuspended in the original medium in the storage vial by gentle pipetting. The method for SI culture consisted of combining in flatsided Nunc culture tubes 2 ml of Dulbecco modified minimal essential medium that contained 10% fetal bovine serum, 0.11% sodium bicarbonate, 2 mM HEPES buffer, penicillin (100 U/ml), streptomycin (100 ~g/ml), nystatin (6 ~g/ml), and amphotericin B (0.5 p,g/ml), 0.2 ml of cell suspension, and 0.1-0.5 ml (depending on the experiment) of patient specimen clarified by low-speed centrifugation (60 g for 5 min). After the caps were securely tightened, the contents of the tubes were mixed by two gentle inversions, whereupon they were incubated at 36°C for 18-24 hr with no further agitation. Following incubation, the cultures were fixed and stained by the immunoperoxidase method previously described (Luker et al., 1991). The criterion used for defining a positive culture was either the presence of at least one immunostained focus consisting of three or more cells with intact cell morphology or the presence of at least two individually stained cells located in different areas of the monolayer. These latter cells were also required to exhibit cell morphology typical of the cultured cell line. In practice, most positive cultures exhibited far more foci than required by the minimum criterion.
Shell-Vial Culture Shell vials (screw-capped polypropylene vials, 57 x 16.5 mm; Sarstedt, Newton, SC) containing monolayers of mink lung cells on 12-mm coverslips were used for SVC. The vials were seeded with cells in our laboratory and used within 24 hr of seeding. The growth media were decanted prior to inoculation of the cultures with 0.1 ml of sample. The vials were then centrifuged for 60 min at 700 g at room temperature. After centrifugation, 1 ml of growth medium (described above in methods for suspensioninfection culture) was added to each vial, and the vials were incubated for 16-24 hr at 36°C. The culture media were then decanted, and the monolayers washed with 1 ml of phosphate-buffered saline (PBS) and then fixed in 1 ml of cold acetone. After airdrying, the cultures were immunostained with fluorescein-conjugated type-common HSV antibody
675
Detection of HSV by Suspension-Infection
(Dynatech Diagnostics, South W i n d h a m , ME) and evaluated by fluorescence microscopy.
Performance of S u s p e n s i o n - I n f e c t i o n and S h e l l - V i a l Culture
Herpes S i m p l e x Virus S e r o t y p i n g Specimens being cultured for routine evaluation by SI were occasionally a c c o m p a n i e d b y a request for HSV serotyping should the culture exhibit a positive result. In such cases, duplicate tubes were inoculated with specimen. At 16-24 hr, one tube was stained by the i m m u n o p e r o x i d a s e m e t h o d and evaluated for the presence of infected cells. If the culture was positive, then the second tube was incubated further until 3 + to 4 + cytopathic effects (CPEs) developed. The infected, d e t a c h e d cells were h a r v e s t e d by centrifugation of the culture tube (60 g for 5 min), washed with 2 ml of PBS, and recentrifuged. The infected cell pellets were s u s p e n d e d in 0.1 ml PBS, spotted in two wells each on teflon-coated glass slides, allowed to air-dry, fixed in cold acetone, and t h e n airdried again. Specimens prepared in this m a n n e r were stained with type-specific m o n o c l o n a l antibodies (Syva Corporation, Palo Alto, CA), one well with type-1 reagent and one well with type-2 reagent. The stained cells were e v a l u a t e d b y fluorescence microscopy.
Data A n a l y s i s The SI trial 1 data and the SVC data (Tables 1 and 2) were analyzed by the chi-squared test of independence.
TABLE I
RESULTS
Clinical specimens (n -- 530) in trial 1 w e r e tested in parallel for HSV b y conventional culture, SI, and SVC. Mink lung cells of identical passage levels were u s e d in all tests. Each culture was inoculated with 0.1 ml of specimen. Of the 115 positive specimens (overall isolation rate of 21.7%), 107 were detected in overnight SI cultures for a sensitivity value of 93.0% (Table 1). Of the 115 positive specimens, 109 were detected by SVC for sensitivity value of 94.8% (Table 2). The data (numbers of test-positive and negative specimens) obtained in trial 1 (Table 1) b y the SI m e t h o d were c o m p a r e d with the data obtained b y the SVC m e t h o d (Table 2) for statistical significance of the two sensitivity values. A chisquared value of 0.304 was obtained, indicating no significant difference b e t w e e n results obtained by SI and SVC m e t h o d s at a 95% confidence level (critical value = 3.841). The six true-positive specimens not detected by SVC were also not detected b y SI. In addition, two more positive specimens were not detected b y SI. All these samples contained low a m o u n t s of HSV and were referred to as low-level specimens. Lowlevel specimens were defined as true-positive specimens detected by either conventional culture after 3 or more days of incubation or rapid culture (either SI or SVC) with five or f e w e r loci of infected cells.
Performance C o m p a r i s o n of the Suspension-Infection (SI) Test and Conventional Culture a a Trial 1"
SI Result + -
Conventional Culture Result + 107 8
0 415 b
SI Result
+ -
Trial 2c
Conventional Culture Result + 174 3
0 357
aBoth tests were performed using mink lung cells. ~530specimens. Sensitivity = 107/107 + 8 = 93.0%; specificity = 415/415 + 0 = 100%;positive predictive value = 107/107 + 0 = 100%;and negative predictive value = 415/415 + 8 = 98.1%. '534 specimens. Sensitivity = 174/174 + 3 = 98.3%; specificity = 357/357 + 0 = 100%;positive predictive value = 174/174 + 0 = 100%; and negative predictive value = 357/357 + 3 = 99.2%.
676
F.B. Johnson and E.M. Visick
By these criteria, 15 (13%) of the 115 true-positive specimens were low-level specimens. Hence, the SI method detected 7 (47%) of 15 of the low-level specimens and SVC detected 9 (60%) of 15 of them. In trial 2, 534 additional samples were tested in the same manner as in trial 1, except SVCs were not included, and each culture was inoculated with 0.5 ml of sample instead of 0.1 ml. Of 177 true-positive specimens (overall isolation rate of 33.1%), 174 were detected in overnight SI cultures for a sensitivity value of 98.3% (Table 1). In this trial, 58 of the 177 true-positive specimens (32.8%) were low-level specimens. Hence, the SI method detected 55 (94.8%) of 58 of the low-level specimens. The much better performance of SI in trial 2 in detecting low-level positive specimens was attributed to the increased inoculum size.
Time to Test Result The culture results obtained in trial 2 were plotted to show the difference in time required to obtain a positive or negative test result between the SI and conventional culture methods (Figure 1). The SI test, concluded in 1 day, detected 98.3% of the positive specimens and 100% of the negative specimens. By conventional culture, only 30% of the positive specimens were detected by day 1. By day 2, 67% of the positive specimens had been detected, 84% by day 3, 90% by day 4, 95% by day 5, and 100% by day 6. Incubation was continued to day 7, but all of the negative specimens were detected by day 6.
Herpes S i m p l e x Virus Serotyping It was assumed that the SI technique detected both HSV type-1 and type-2 strains because the total number of isolates detected by SI approached the total number of isolates detected by conventional culture. However, it was important to demonstrate directly the recovery of both types. Isolates were serotyped by request of the submitting physician, thus random isolates were subjected to typing. Of
TABLE 2
DISCUSSION The results of this study show that (a) detection of HSV in clinical samples is not significantly different whether using the SI test or SVC, (b) the sensitivity of SI can exceed 98%, (c) the SI test is a rapid test when compared with conventional culture, and (d) both HSV-1 and HSV-2 strains are detected by the SI method. It was also shown that the amount of inoculum used influences the number of positive specimens detected, hence, the sensitivity of the test. This finding was expected considering that many clinical specimens contain low amounts of infectious virus and increasing the inoculum size should augment the probability of including virus in the inoculum. In this study, experiments were not conducted using inocula >0.1 ml in SVC, and direct comparisons between SI and SVC with large inocula were not made. However, it is likely that the sensitivity value of SVC in our hands would increase somewhat with inoculum size as was found with SI. Studies published by others using SVC for the detection of HSV have indicated sensitivities of 70%100% compared with CPE detection in conventional culture (Espy et al., 1991; Gleaves et al., 1985b; Peterson et al., 1988; Woods and Mills, 1988). Because HSV-1 and HSV-2 appear to attach to different host cell receptors (Vahlne et al., 1979 and 1980) and it has not previously been shown that MV1 Lu cells in suspension display both types of receptors, it was of significance to determine whether clinical strains of both HSV-1 and HSV-2 were detected by the SI technique. The results revealed that strains of both types were isolated. There is no indication that SI, or the stain reagents used in this method, fail to detect any clinical HSV strains of either type because of serotypic differences. Fewer manipulations and procedural steps in a clinical test for an infectious agent may result in many benefits, such as, less hands-on time required to
Performance Comparison of Shell-Vial Culture and Conventional Culture in Mink Lung Cellsa
Shell-Vial Culture Result + -
82 strains that were typed, 35% were type-1 strains and 65% were type-2 strains (Table 3).
Conventional Culture Result +
-
109 6
0 415
aSensitivity = 109/109 + 6 = 94.8%; specificity~ 415/415 + 0 = 100%;positivepredictive value = 109/109 + 0 = 100%;and negative predictive value = 415/415 + 6 = 98.6%.
Detection of HSV by Suspension-Infection
SUSPE NSION-IN F E CTION :
CONVENTIONAL CULTURE:
B
677
[~
number
~ percent I00
number
percent
074)
LU > m
I--
80
0
0. u) 6 0 z IJJ u uJ 4 0 o.
20
I
2
T IME
3
4
5
6
7
(DAYS)
POSTINOCULATION
FIGURE 1 Number and cumulative percentage of specimens observed to be positive by conventional culture by day after inoculation compared with 1-day result by suspension-infection culture. A total of 177 specimens were found positive by conventional culture after 6 days of incubation, whereas 174 (98.3%) of them were detected by the suspension-infection method in 1 day.
perform the test, less opportunity for experimental error, and less exposure of laboratory personnel to infectious organisms. The SI technique for the detection of HSV has these advantages. It also has the advantage of being more rapid than conventional culture while approaching culture sensitivity. Its advantages over SVC are that it does not require the time-consuming centrifugation step, fewer technical manipulations are necessary, and it does not require the use of shell vials, which are more difficult to open and close without contaminating laboratory surfaces than are screw-capped tubes. TABLE 3
It has been argued, in effect, that when optimally sensitive cells are used in conventional culture, detection of most positive samples is obtained early in the incubation period, making conventional culture a relatively rapid and inexpensive test (Ginnow Johnston and Siegel, 1990). It must be recognized that even under such conditions, however, typically only
