- Email: [email protected]
Clinical evaluation of the improved streptex method for grouping streptococci
DIAGNMICROBIOLINFECTDIS 1984;2:113-118
113
Clinical Evaluation of the Improved .Streptex Method for Grouping Streptococci Elaine Bixler-Forell, William J. Martin, and Max D. Moody The improved Streptex method for serogrouping streptococci incorporates a new extraction enzyme and a simplified procedure requiring no centrifugation: A total of 114 clinical isolates of ~hemolytic streptococci were serogrouped from primary plates, isolation plates, and Todd-Hewitt broth cultures using this system. Results were compared to those of the heat extraction Lancefield precipitin method. An additional 33 stock culture isolates of related streptococcal species and 5 strains of Listeria were serogrouped to assess the specificity of the test. Agreement between the two methods was 82.5% with primary plates and 96.5% with both isolation plates and broth cultures. Four isolates from three different serogroups were nongroupable by the Lancefield method, but did agglutinate in specific Streptex antisera; therefore, the enzyme extraction procedure appeared more sensitive than the heat extraction method. Streptex accurately grouped five isolates of 7-hemolytic group B streptococci, but failed to detect antigen in 33% of the group D streptococcal extracts tested. In addition, cross-reactions were observed with strains of a-hemolytic streptococci. Streptex produced fewer ambiguous results and required fewer repeat tests. When used with isolation plates or broth cultures, Streptex is both sensitive and specific for the grouping of 13-hemolytic streptococci of groups A, B, C, F, and G.
INTRODUCTION Groups A, B, C, D, F, and G constitute more than 99% of ~-hemolytic streptococcal infections in humans (Facklam, 1980a). Indeed, all 13-hemolytic streptococci isolated from clinical specimens in which the infection is suspected to be of streptococcal origin should be serologically grouped (Finegold et al., 1966; Rotta and Facklam,
1980). For best results, serologic grouping of streptococci requires that the group-specific carbohydrate structural component of the cell wall be available as free antigen. Lancefield (1933, 1938) found that these antigens could be extracted in soluble form and identified by precipitation reactions with homologous antisera. This method requires initial isolation of the organism followed by a time-consuming multistep grouping procedure (Rantz and Randall, 1955). The Streptex method (Wellcome Diagnostics, Research Triangle Park, NC) offers the same serogrouping capabilities, but employs
From the Department of Pathology, Clinical Microbiology, UCLA Medical Center, Los Angeles, California {E.B.-F.); the Clinical Microbiology Laboratory, Departments of Pathology and Medicine, Tufts-New England Medical Center, Boston, Massachusetts {W.J.M.};and Wellcome Diagnostics, Research Triangle Park, North Carolina (M.D.M.). Address reprint requests to: Elaine Bixler-Forell, Supervisor, Bacteriology, UCLA Clinical Laboratories, CHS A2-155, Los Angeles, CA 90024. Received April 26, 1983; revised and accepted December 8, 1983.
© 1984 Elsevier Science Publishing Co., Inc. 52 Vanderbilt Avenue, New York, NY 10017
0732-8893/84/$03.00
114
E. Bixler-Forell et al.
a simplified enzyme extraction and slide agglutination method that requires no centrifugation. This study reports the feasibility of grouping streptococci from primary culture plates as well as from isolation plates and broth cultures.
MATERIALS AND METHODS Organisms One hundred fourteen strains of [3-hemolytic streptococci were isolated from a variety of clinical specimens: throat, wound, urine, genital tract, and miscellaneous sources, including blood culture. Each isolate was grouped from the following sources: 1) 16to 24-hr primary culture plates (5% sheep blood agar or polymyxin-nalidixic blood agar); 2) 24-hr isolation plates (5% sheep blood agar); and 3) 4- to 6ohr Todd-Hewitt broth (3.0 ml) (BBL, Cockeysville, MD) cultures. If visible turbidity was not obtained in 4-6 hr, broth cultures were reincubated for a total of 24 hr. In addition, 33 stock culture isolates of viridans group streptococci, ~/-hemolytic group B streptococci, group D enterococci and nonenterococci, and five stock culture isalates of Listeria were tested using growth from 24-hr isolation plates and 4- to 6-hr or 24-hr Todd-Hewitt broth cultures. All cultures were incubated at 35°C in 8-10% CO2. Growth of B-hemolytic streptococci on primary plates was assigned a purity index ranging from 1 to 5 based on one 13-hemolytic streptococcal colony per colonies of other organisms numbering 1) 100 or more; 2) 50-100; 3) 20-50; and 4] 5-20. Pure or almost pure cultures of B-hemolytic streptococci were assigned a purity index of 5.
Lancefield Method All clinical isolates were grouped from 24-hr Todd-Hewitt broth (40 ml) cultures using the autoclave extraction and capillary tube precipitin method of Rantz and Randall (1955; Facklam, 1980b). Streptex Method The Streptex kit contains the following materials in sufficient quantity for 40 group identifications: 1) one bottle each of latex particles coated with rabbit antibody to streptococcal group antigens A, B, C, D, F, and G; 2) freeze-dried proteolytic enzyme obtained from cultures of Streptomyces griseus; 3) freeze-dried polyvalent positive control containing antigens from a representative strain of each streptococcal group (A, B, C, D, F, and G); 4) glass tile; and 5) disposable mixing sticks. All grouping procedures were done according to directions from the manufacturer. Rehydrated extraction enzyme (0.4 ml) was dispensed into clean 10 x 75 mm disposable glass tubes. Using a bacteriologic loop, growth from each primary and isolation plate was used to make a light suspension (~>McFarland 1.0] in the tube of enzyme solution. When Todd-Hewitt broth cultures were used, one to three drops of the culture were added to the enzyme solution using a Pasteur pipette. All inoculated tubes were incubated at 37°C in a water bath for 1 hr, and the resulting extract was tested immediately or after overnight storage at 4°C. After thorough mixing, a drop of each group-specific latex antiserum was transferred to a separate circle on the glass tile. Using a Pasteur pipette, one drop of culture extract was then placed in each circle containing antisera. The contents of each circle were mixed and spread to cover the complete area of the circle using a separate mixing stick. The slide was then manually rotated for a maximum of I min. A positive result was indicated by clearly visible clumping of the latex particles. For purposes
Clinical Evaluation of Streptex Method
115
of this study, agglutination intensity was recorded from 1 + to 4 +. In a negative reaction, the latex particles did not agglutinate and the milky appearance of the suspension remained essentially unchanged. Faint granularity was considered a negative result. Nonagreements
All discrepancies were repeated at least once by the Lancefield heat extraction and Streptex methods using growth obtained from the 24-hr isolation plates. If the disagreement was not resolved by this procedure, organisms were inoculated into 40 ml of Todd-Hewitt broth and incubated at 35°C for 4-7 days. Growth obtained by this procedure was heat extracted and the resultant antigen grouped using both the capillary tube precipitin method and the Streptex slide agglutination method. RESULTS Overall agreement between the two serogrouping methods is summarized in Table 1. All ~/-hemolytic group B streptococci were accurately grouped by Streptex. Discrepancies from isolation plates and broth cultures included four isolates that were nongroupable by the Lancefield method; of these four isolates, one was grouped by Streptex as group C, one as group F, and two as group G. The original Streptex grouping was confirmed by the Lancefield tube precipitin method for three of the four isolates when these organisms were grown in Todd-Hewitt broth for 4-7 days, thereby yielding large amounts of growth for heat extraction of antigen. Only the Streptex group F isolate remained nongroupable by the Lancefield method. Reactions in Streptex antisera were inconsistent when stock isolates of group D streptococci were used. Streptex accurately grouped 66.7% (12 of 18) and 61.1% (11 of 18) of the
TABLE 1. Comparison of Lancefield and Streptex Grouping Methods for Clinical and Stock Culture Isolates of Streptococci Streptex grouping
Lancefield grouping
Number of isolates
Primary plate
Isolation plate
Broth culture
A
21 37 17 10 12 17
20 35 12 7 8 12
21 37 17 10 12 13
21 37 17 10 12 13
114
94
110
110
B
C F G Nongroupable Total clinical isolates Agreement of clinical isolates
(%)
94/114 (62.5) ~-B °
5
~-D° ~-Da Total clinical and stock isolates Overall agreement (%)
7 11
~Stockisolates.
137
110/114 (96.5) 5 3 9 127 127/137 (92.7)
110/114(96.5) 5 5 6 126 126/137(92.0)
116
E. Bixler-Forell et al.
TABLE 2. Intensity of Streptex Reactions with Agglutination of Only One Group-Specific Antiserum Intensity
Primary plates
4+ 3+ 2+ 1+
66 12 5 4
Total
87
(75.9) a (13.8) (5.7) (4.6)
Isolation plates 107 6 5 0
Broth culture
(90.7) (5.1) (4.2) (0)
79 29 9 0
118
(67.5) (24.8) (7.7) (0)
117
°All values represent number of isolates reacting at specified intensity, with percentage in parentheses.
group D streptococci from isolation plates and broth cultures, respectively. Two group D organisms showed strong cross-reactivity with Streptex group G antiserum. In specificity studies, five isolates of Listeria demonstrated no agglutination in Streptex antisera. However, 3 of 10 viridans group streptococci did cross-agglutinate in Streptex antisera, one to group C, one to group D, and one to group F. The intensity of the Streptex antisera reactions for both clinical and stock culture isolates when agglutination occurred with only one group-specific antiserum is shown in Table 2. Overall agglutination was strongest when growth was obtained from isolation plates. Most of the clinical isolates of ~-hemolytic streptococci were recovered as the predominant organism or in pure or almost pure culture with a purity index of 4 or 5 (Table 3). Moreover, no significant differences were demonstrated in the purity index distribution of isolates for any of the streptococcal groups. As the purity index rating of the primary plates increased, the number of discrepant results decreased. When there was nonagreement between Lancefield and Streptex results, the average purity index was 2.5. This figure was lower than the average index for both total isolates (3.5) and Lancefield and Streptex agreements (3.7). Streptex required 1.8% (2 of 114) and 3.5% (4 of 114) repeat testings from isolation plates and broth cultures, respectively, whereas the Lancefield method required 25.4% (29 of 114) repeat testings to confirm grouping results.
TABLE 3. Comparison of Purity Indices from Primary Plates for Agreement and Nonagreement between Streptex and Lancefield Methods Nonagreement a
Purity index
Number of isolates
Agreement
Positive b
1 2 3
15 19 11
10 (66.7) a 12 (63.2) 8 (72.7)
4 (26.7) 5 (26.3) 1 (9.1)
1 (6.7) 2 (10.5) 2 (18.2)
4
36
32 (88.9)
2 (5.6)
2 (5.6)
5
33
32 (97.0)
0
Total Average
114 3.5
94 (a2.5) 3.7
(O)
12 (10.5) 2.1
Negativec
1
(3.0)
8 (7.0) 3.0
aAverage purity index of all nonagreements = 2.5. bNonagreement between group-specific agglutination reactions in Lancefleld and Streptex antisera or polyagglutination in two or more Streptex antisera. ~Group-specific agglutination in Lancefield antisera and no agglutination in Streptex antisera. dAll values represent number of isolates, with percentage in parentheses.
Clinical Evaluation of Streptex Method
117
DISCUSSION The accuracy of latex agglutination reagents has been previously reported as 97% for grouping of stock culture isolates of [3-hemolytic streptococci when growth was obtained from isolation plates or Todd-Hewitt broth cultures (Facklam et al., 1979). Data reported in this study using clinical isolates correlated well with these findings. Keville and Doern ~1982) reported that 91.7% of clinical isolates tested were correctly identified when growth from isolation plates was used as the source of antigen. These authors attributed this lower accuracy to the greater proportion of group F isolates included in their study [41% (7 of 17) of group F isolates gave no visible reaction with Streptex antisera]. In contrast, all 10 Lancefield group F isolates included in this study were correctly identified by Streptex. In a previous evaluation comparing the improved Streptex method to the Lancefield formamide extraction tube precipitin method, agreement of 92% was reported using growth from primary plates (Castle et al., 1982). In this study, when growth from primary plates was used agreement between the Lancefield heat extraction and Streptex methods was low (82.5%). This finding was probably due to either insufficient Inoculum or interfering cross-reactions caused by a mixed population of organisms: The lower average purity index from the primary plates giving discrepant results supports this conclusion (Table 3). Using the Streptex method, a purity index of ~>4 was necessary for reliable serogrouping of [3-hemolytic streptococci from primary plates. Agreement between the Streptex and Lancefield methods was excellent when isolation plates or broth cultures were used; however, organisms obtained from isolation plates generally yielded stronger, more definitive Streptex reactions than those obtained using either primary plates or broth cultures. Based on these results and the higher percentage of discrepant results obtained from primary plates, the best source of antigen was determined to be 24-hr 5% sheep blood agar isolation plates. The Streptex method appeared more sensitive as it was able to .serogroup four isolates that were nongroupable b y the Lancefield method. This increased sensitivity was confirmed when the Lancefield procedure, using unusually large amounts of growth from these initially nongroupable organisms, produced the same grouping results as Streptex using inocula of conventional size. Caution should be exercised when interpreting the serogroup reactions obtained with Steptex antisera and u-hemolytic streptococci because of 1) the cross-reactions observed with viridans group streptococci; 2) the possession by some group D organisms of an antigen that cross-reacts with group G antisera; and 3) the failure of Streptex antisera to yield a positive agglutination reaction with a significant percentage of group D extracts. From the important perspective of appropriate antimicrobial therapy, it is necessary to distinguish group D enterococci from nonenterococci. For this reason, we chose not to attempt routine serologic grouping of suspected group D isolates, but rather to identify these organisms with standard biochemical tests such as bile esculin hydrolysis and growth in 6.5% NaCI. Streptex reagents are more costly than those for the Lancefield procedure. Because the Streptex enzyme extraction is faster and simpler to perform than the Lancefield heat extraction procedure and fewer repeat tests are required to confirm serogroups of clinical isolates, this cost is offset by the savings in the technologist's time. The Streptex enzyme extraction method is both sensitive and specific for accurate serogrouping of B-hemolytic streptococci of groups A, B, C, F, and G. Moreover, when used with isolation plates or broth cultures, Streptex has shown excellent correlation with the conventional Lancefield heat extraction method.
118
E. Bixler-Forell et al.
REFERENCES Castle D, Kessock-Philip S, Easmon CSF (1982) Evaluation of an improved Streptex kit for the grouping of beta-haemolytic streptococci by agglutination. J Clin Pathol 35:719. Facklam RR (1980) Isolation and identification of streptococci. Atlanta: U.S. Department of Health, Education, and Welfare, Public Health Service, Centers for Disease Control. Facklam RR (1980) Streptococci and aerococci. In Manual of Clinical Microbiology. 3rd ed. Eds., EH Lennette, A Balows, WJ Hausler, and JP Truant. Washington, DC: American Society for Microbiology, pp 88-109. Facklam RR, Cooksey RC, Wortham EC (1979) Evaluation of commercial latex agglutination reagents for grouping streptococci. J Clin Microbiol 10:641. Finegold DS, Stagg ML, Kunz LJ (1966) Extra-respiratory streptococcal infections. Importance of the various serologic groups. N Engl J Med 275:356. Keville MW, Doern GV (1982) Comparison of the API 20S Streptococcus identification system with an immunorheophoresis procedure and two commercial latex agglutination tests for identifying beta hemolytic streptococci. J Clin Microbiol 16:92. Lancefield RC (1933) Serological differentiation of human and other groups of hemolytic streptococci. J Exp Med 57:571. Lancefield RC (1938) Micro precipitin technic for classifying hemolytic streptococci, and improved methods for producing antisera. Proc Soc Exp Biol Med 38:473. Rantz LA, Randall E (1955) Use of autoclaved extracts of hemolytic streptococci for serological grouping. Stanford Med Bull 13:290. Rotta J, Facklam RR (1980) Manual of Microbiological Diagnostic Methods for Streptococcal Infections and Their Sequelae. WHO/BAC 80.1. Geneva: World Health Organization.
113
Clinical Evaluation of the Improved .Streptex Method for Grouping Streptococci Elaine Bixler-Forell, William J. Martin, and Max D. Moody The improved Streptex method for serogrouping streptococci incorporates a new extraction enzyme and a simplified procedure requiring no centrifugation: A total of 114 clinical isolates of ~hemolytic streptococci were serogrouped from primary plates, isolation plates, and Todd-Hewitt broth cultures using this system. Results were compared to those of the heat extraction Lancefield precipitin method. An additional 33 stock culture isolates of related streptococcal species and 5 strains of Listeria were serogrouped to assess the specificity of the test. Agreement between the two methods was 82.5% with primary plates and 96.5% with both isolation plates and broth cultures. Four isolates from three different serogroups were nongroupable by the Lancefield method, but did agglutinate in specific Streptex antisera; therefore, the enzyme extraction procedure appeared more sensitive than the heat extraction method. Streptex accurately grouped five isolates of 7-hemolytic group B streptococci, but failed to detect antigen in 33% of the group D streptococcal extracts tested. In addition, cross-reactions were observed with strains of a-hemolytic streptococci. Streptex produced fewer ambiguous results and required fewer repeat tests. When used with isolation plates or broth cultures, Streptex is both sensitive and specific for the grouping of 13-hemolytic streptococci of groups A, B, C, F, and G.
INTRODUCTION Groups A, B, C, D, F, and G constitute more than 99% of ~-hemolytic streptococcal infections in humans (Facklam, 1980a). Indeed, all 13-hemolytic streptococci isolated from clinical specimens in which the infection is suspected to be of streptococcal origin should be serologically grouped (Finegold et al., 1966; Rotta and Facklam,
1980). For best results, serologic grouping of streptococci requires that the group-specific carbohydrate structural component of the cell wall be available as free antigen. Lancefield (1933, 1938) found that these antigens could be extracted in soluble form and identified by precipitation reactions with homologous antisera. This method requires initial isolation of the organism followed by a time-consuming multistep grouping procedure (Rantz and Randall, 1955). The Streptex method (Wellcome Diagnostics, Research Triangle Park, NC) offers the same serogrouping capabilities, but employs
From the Department of Pathology, Clinical Microbiology, UCLA Medical Center, Los Angeles, California {E.B.-F.); the Clinical Microbiology Laboratory, Departments of Pathology and Medicine, Tufts-New England Medical Center, Boston, Massachusetts {W.J.M.};and Wellcome Diagnostics, Research Triangle Park, North Carolina (M.D.M.). Address reprint requests to: Elaine Bixler-Forell, Supervisor, Bacteriology, UCLA Clinical Laboratories, CHS A2-155, Los Angeles, CA 90024. Received April 26, 1983; revised and accepted December 8, 1983.
© 1984 Elsevier Science Publishing Co., Inc. 52 Vanderbilt Avenue, New York, NY 10017
0732-8893/84/$03.00
114
E. Bixler-Forell et al.
a simplified enzyme extraction and slide agglutination method that requires no centrifugation. This study reports the feasibility of grouping streptococci from primary culture plates as well as from isolation plates and broth cultures.
MATERIALS AND METHODS Organisms One hundred fourteen strains of [3-hemolytic streptococci were isolated from a variety of clinical specimens: throat, wound, urine, genital tract, and miscellaneous sources, including blood culture. Each isolate was grouped from the following sources: 1) 16to 24-hr primary culture plates (5% sheep blood agar or polymyxin-nalidixic blood agar); 2) 24-hr isolation plates (5% sheep blood agar); and 3) 4- to 6ohr Todd-Hewitt broth (3.0 ml) (BBL, Cockeysville, MD) cultures. If visible turbidity was not obtained in 4-6 hr, broth cultures were reincubated for a total of 24 hr. In addition, 33 stock culture isolates of viridans group streptococci, ~/-hemolytic group B streptococci, group D enterococci and nonenterococci, and five stock culture isalates of Listeria were tested using growth from 24-hr isolation plates and 4- to 6-hr or 24-hr Todd-Hewitt broth cultures. All cultures were incubated at 35°C in 8-10% CO2. Growth of B-hemolytic streptococci on primary plates was assigned a purity index ranging from 1 to 5 based on one 13-hemolytic streptococcal colony per colonies of other organisms numbering 1) 100 or more; 2) 50-100; 3) 20-50; and 4] 5-20. Pure or almost pure cultures of B-hemolytic streptococci were assigned a purity index of 5.
Lancefield Method All clinical isolates were grouped from 24-hr Todd-Hewitt broth (40 ml) cultures using the autoclave extraction and capillary tube precipitin method of Rantz and Randall (1955; Facklam, 1980b). Streptex Method The Streptex kit contains the following materials in sufficient quantity for 40 group identifications: 1) one bottle each of latex particles coated with rabbit antibody to streptococcal group antigens A, B, C, D, F, and G; 2) freeze-dried proteolytic enzyme obtained from cultures of Streptomyces griseus; 3) freeze-dried polyvalent positive control containing antigens from a representative strain of each streptococcal group (A, B, C, D, F, and G); 4) glass tile; and 5) disposable mixing sticks. All grouping procedures were done according to directions from the manufacturer. Rehydrated extraction enzyme (0.4 ml) was dispensed into clean 10 x 75 mm disposable glass tubes. Using a bacteriologic loop, growth from each primary and isolation plate was used to make a light suspension (~>McFarland 1.0] in the tube of enzyme solution. When Todd-Hewitt broth cultures were used, one to three drops of the culture were added to the enzyme solution using a Pasteur pipette. All inoculated tubes were incubated at 37°C in a water bath for 1 hr, and the resulting extract was tested immediately or after overnight storage at 4°C. After thorough mixing, a drop of each group-specific latex antiserum was transferred to a separate circle on the glass tile. Using a Pasteur pipette, one drop of culture extract was then placed in each circle containing antisera. The contents of each circle were mixed and spread to cover the complete area of the circle using a separate mixing stick. The slide was then manually rotated for a maximum of I min. A positive result was indicated by clearly visible clumping of the latex particles. For purposes
Clinical Evaluation of Streptex Method
115
of this study, agglutination intensity was recorded from 1 + to 4 +. In a negative reaction, the latex particles did not agglutinate and the milky appearance of the suspension remained essentially unchanged. Faint granularity was considered a negative result. Nonagreements
All discrepancies were repeated at least once by the Lancefield heat extraction and Streptex methods using growth obtained from the 24-hr isolation plates. If the disagreement was not resolved by this procedure, organisms were inoculated into 40 ml of Todd-Hewitt broth and incubated at 35°C for 4-7 days. Growth obtained by this procedure was heat extracted and the resultant antigen grouped using both the capillary tube precipitin method and the Streptex slide agglutination method. RESULTS Overall agreement between the two serogrouping methods is summarized in Table 1. All ~/-hemolytic group B streptococci were accurately grouped by Streptex. Discrepancies from isolation plates and broth cultures included four isolates that were nongroupable by the Lancefield method; of these four isolates, one was grouped by Streptex as group C, one as group F, and two as group G. The original Streptex grouping was confirmed by the Lancefield tube precipitin method for three of the four isolates when these organisms were grown in Todd-Hewitt broth for 4-7 days, thereby yielding large amounts of growth for heat extraction of antigen. Only the Streptex group F isolate remained nongroupable by the Lancefield method. Reactions in Streptex antisera were inconsistent when stock isolates of group D streptococci were used. Streptex accurately grouped 66.7% (12 of 18) and 61.1% (11 of 18) of the
TABLE 1. Comparison of Lancefield and Streptex Grouping Methods for Clinical and Stock Culture Isolates of Streptococci Streptex grouping
Lancefield grouping
Number of isolates
Primary plate
Isolation plate
Broth culture
A
21 37 17 10 12 17
20 35 12 7 8 12
21 37 17 10 12 13
21 37 17 10 12 13
114
94
110
110
B
C F G Nongroupable Total clinical isolates Agreement of clinical isolates
(%)
94/114 (62.5) ~-B °
5
~-D° ~-Da Total clinical and stock isolates Overall agreement (%)
7 11
~Stockisolates.
137
110/114 (96.5) 5 3 9 127 127/137 (92.7)
110/114(96.5) 5 5 6 126 126/137(92.0)
116
E. Bixler-Forell et al.
TABLE 2. Intensity of Streptex Reactions with Agglutination of Only One Group-Specific Antiserum Intensity
Primary plates
4+ 3+ 2+ 1+
66 12 5 4
Total
87
(75.9) a (13.8) (5.7) (4.6)
Isolation plates 107 6 5 0
Broth culture
(90.7) (5.1) (4.2) (0)
79 29 9 0
118
(67.5) (24.8) (7.7) (0)
117
°All values represent number of isolates reacting at specified intensity, with percentage in parentheses.
group D streptococci from isolation plates and broth cultures, respectively. Two group D organisms showed strong cross-reactivity with Streptex group G antiserum. In specificity studies, five isolates of Listeria demonstrated no agglutination in Streptex antisera. However, 3 of 10 viridans group streptococci did cross-agglutinate in Streptex antisera, one to group C, one to group D, and one to group F. The intensity of the Streptex antisera reactions for both clinical and stock culture isolates when agglutination occurred with only one group-specific antiserum is shown in Table 2. Overall agglutination was strongest when growth was obtained from isolation plates. Most of the clinical isolates of ~-hemolytic streptococci were recovered as the predominant organism or in pure or almost pure culture with a purity index of 4 or 5 (Table 3). Moreover, no significant differences were demonstrated in the purity index distribution of isolates for any of the streptococcal groups. As the purity index rating of the primary plates increased, the number of discrepant results decreased. When there was nonagreement between Lancefield and Streptex results, the average purity index was 2.5. This figure was lower than the average index for both total isolates (3.5) and Lancefield and Streptex agreements (3.7). Streptex required 1.8% (2 of 114) and 3.5% (4 of 114) repeat testings from isolation plates and broth cultures, respectively, whereas the Lancefield method required 25.4% (29 of 114) repeat testings to confirm grouping results.
TABLE 3. Comparison of Purity Indices from Primary Plates for Agreement and Nonagreement between Streptex and Lancefield Methods Nonagreement a
Purity index
Number of isolates
Agreement
Positive b
1 2 3
15 19 11
10 (66.7) a 12 (63.2) 8 (72.7)
4 (26.7) 5 (26.3) 1 (9.1)
1 (6.7) 2 (10.5) 2 (18.2)
4
36
32 (88.9)
2 (5.6)
2 (5.6)
5
33
32 (97.0)
0
Total Average
114 3.5
94 (a2.5) 3.7
(O)
12 (10.5) 2.1
Negativec
1
(3.0)
8 (7.0) 3.0
aAverage purity index of all nonagreements = 2.5. bNonagreement between group-specific agglutination reactions in Lancefleld and Streptex antisera or polyagglutination in two or more Streptex antisera. ~Group-specific agglutination in Lancefield antisera and no agglutination in Streptex antisera. dAll values represent number of isolates, with percentage in parentheses.
Clinical Evaluation of Streptex Method
117
DISCUSSION The accuracy of latex agglutination reagents has been previously reported as 97% for grouping of stock culture isolates of [3-hemolytic streptococci when growth was obtained from isolation plates or Todd-Hewitt broth cultures (Facklam et al., 1979). Data reported in this study using clinical isolates correlated well with these findings. Keville and Doern ~1982) reported that 91.7% of clinical isolates tested were correctly identified when growth from isolation plates was used as the source of antigen. These authors attributed this lower accuracy to the greater proportion of group F isolates included in their study [41% (7 of 17) of group F isolates gave no visible reaction with Streptex antisera]. In contrast, all 10 Lancefield group F isolates included in this study were correctly identified by Streptex. In a previous evaluation comparing the improved Streptex method to the Lancefield formamide extraction tube precipitin method, agreement of 92% was reported using growth from primary plates (Castle et al., 1982). In this study, when growth from primary plates was used agreement between the Lancefield heat extraction and Streptex methods was low (82.5%). This finding was probably due to either insufficient Inoculum or interfering cross-reactions caused by a mixed population of organisms: The lower average purity index from the primary plates giving discrepant results supports this conclusion (Table 3). Using the Streptex method, a purity index of ~>4 was necessary for reliable serogrouping of [3-hemolytic streptococci from primary plates. Agreement between the Streptex and Lancefield methods was excellent when isolation plates or broth cultures were used; however, organisms obtained from isolation plates generally yielded stronger, more definitive Streptex reactions than those obtained using either primary plates or broth cultures. Based on these results and the higher percentage of discrepant results obtained from primary plates, the best source of antigen was determined to be 24-hr 5% sheep blood agar isolation plates. The Streptex method appeared more sensitive as it was able to .serogroup four isolates that were nongroupable b y the Lancefield method. This increased sensitivity was confirmed when the Lancefield procedure, using unusually large amounts of growth from these initially nongroupable organisms, produced the same grouping results as Streptex using inocula of conventional size. Caution should be exercised when interpreting the serogroup reactions obtained with Steptex antisera and u-hemolytic streptococci because of 1) the cross-reactions observed with viridans group streptococci; 2) the possession by some group D organisms of an antigen that cross-reacts with group G antisera; and 3) the failure of Streptex antisera to yield a positive agglutination reaction with a significant percentage of group D extracts. From the important perspective of appropriate antimicrobial therapy, it is necessary to distinguish group D enterococci from nonenterococci. For this reason, we chose not to attempt routine serologic grouping of suspected group D isolates, but rather to identify these organisms with standard biochemical tests such as bile esculin hydrolysis and growth in 6.5% NaCI. Streptex reagents are more costly than those for the Lancefield procedure. Because the Streptex enzyme extraction is faster and simpler to perform than the Lancefield heat extraction procedure and fewer repeat tests are required to confirm serogroups of clinical isolates, this cost is offset by the savings in the technologist's time. The Streptex enzyme extraction method is both sensitive and specific for accurate serogrouping of B-hemolytic streptococci of groups A, B, C, F, and G. Moreover, when used with isolation plates or broth cultures, Streptex has shown excellent correlation with the conventional Lancefield heat extraction method.
118
E. Bixler-Forell et al.
REFERENCES Castle D, Kessock-Philip S, Easmon CSF (1982) Evaluation of an improved Streptex kit for the grouping of beta-haemolytic streptococci by agglutination. J Clin Pathol 35:719. Facklam RR (1980) Isolation and identification of streptococci. Atlanta: U.S. Department of Health, Education, and Welfare, Public Health Service, Centers for Disease Control. Facklam RR (1980) Streptococci and aerococci. In Manual of Clinical Microbiology. 3rd ed. Eds., EH Lennette, A Balows, WJ Hausler, and JP Truant. Washington, DC: American Society for Microbiology, pp 88-109. Facklam RR, Cooksey RC, Wortham EC (1979) Evaluation of commercial latex agglutination reagents for grouping streptococci. J Clin Microbiol 10:641. Finegold DS, Stagg ML, Kunz LJ (1966) Extra-respiratory streptococcal infections. Importance of the various serologic groups. N Engl J Med 275:356. Keville MW, Doern GV (1982) Comparison of the API 20S Streptococcus identification system with an immunorheophoresis procedure and two commercial latex agglutination tests for identifying beta hemolytic streptococci. J Clin Microbiol 16:92. Lancefield RC (1933) Serological differentiation of human and other groups of hemolytic streptococci. J Exp Med 57:571. Lancefield RC (1938) Micro precipitin technic for classifying hemolytic streptococci, and improved methods for producing antisera. Proc Soc Exp Biol Med 38:473. Rantz LA, Randall E (1955) Use of autoclaved extracts of hemolytic streptococci for serological grouping. Stanford Med Bull 13:290. Rotta J, Facklam RR (1980) Manual of Microbiological Diagnostic Methods for Streptococcal Infections and Their Sequelae. WHO/BAC 80.1. Geneva: World Health Organization.