Detection of decreased penicillin susceptibility in viridans group streptococci

Pathology (1998) 30, pp. 188-191

DETECTION OF DECREASED PENICILLIN SUSCEPTIBILITY IN VIRIDANS GROUP STREPTOCOCCI SUDHA POTTUMARTHY AND ARTHUR

J.

MORRIS

Microbiology Laboratory, Green Lane and National Women's Hospitals, Auckland, New Zealand

Summary One hundred consecutive clinically significant viridans group streptococcal isolates had their susceptibility to penicillin determined by the penicillin E-test method. The ability of penicillin 2 and 10 unit disks and the oxacillin 1 JIg disk to detect reduced peniCillin susceptibility, ie; MIC 2: 0.25 JIg/ml, in viridans group streptococci was determined by comparing the zone diameters against the penicillin E-test MICs. The sensitivity, specificity and predictive values of previous, existing and proposed interpretative criteria to detect decreased penicillin susceptibility were determined. Thirtyseven per cent of the isolates had reduced susceptibility to penicillin. The previous 1993 NCCLS interpretative criteria for the penicillin 10 unit disk, ie; resistant :s; 27 mm failed to detect 16 of 37 (43%) isolates with reduced penicillin susceptibility. The 1 /19 oxacillin disk using existing meningococcal interpretative criteria, ie; resistant :s; 10 mm, failed to detect 11 of 37 (40%) isolates with reduced penicillin susceptibility. When the oxacillin 1 JIg disk pneumococcal interpretative criteria were used, ie; resistant :s; 19 mm, all the isolates with reduced penicillin susceptibility were detected but 42 of 63 (67%) susceptible isolates were misclassified as resistant. Based on our data, we set new interpretative criteria to detect all isolates with decreased penicillin susceptibility for each of the three disks. Using our proposed zone diameters to detect decreased penicillin susceptibility of :s; 27 mm for the penicillin 2 unit disk, :s; 35 mm for the penicillin 10 unit disk, and :s; 17 mm for the oxacillin disk 34 (54%), 44 (70%),and 21 (33%) of the 63 susceptible iso~ lates, respectively, were misclassified as having decreased penicillin susceptibility. Our data show that the oxacillin 1 /1g disk is able to detect decreased susceptibility to peniCillin in viridans group streptococci with greater specificity than either penicillin 2 or 10 unit disks. Key words: Penicillin susceptibility, viridans group streptococci, disk diffusion method. Accepted 18 December 1997

INTRODUCTION Viridans group streptococci are a part of the normal oropharyngeal and gastrointestinal flora in humans. Despite their low intrinsic virulence they can cause serious disease eg; sepsis syndrome in neutropenic patients, deep seated pyogenic infections and endocarditis. 1- 3 Increasing resistance of viridans group streptococci to penicillin has been reported recently from many parts of the world. 4- 6 Concern has been expressed on how this emerging resistance will

affect chemoprophylaxis against endocarditis and therapy for neutropenic sepsis. 4 ,5 Until recently the NCCLS published interpretative criteria for pe~~cillin disk diffusion testing for viridans group streptocOCCl. Recent NCCLS documents, however, recommend that a penicillin MIC should be determined for viridans group streptococci isolated from sterile body sites. 8,9 This change probably reflects experience within the NCCLS, that penicillin disk testing does not provide an accurate enough indication of a given isolate's susceptibility as determined by a MIC. We are not aware, however, of published data demonstrating this. Screening Streptococcus pneumoniae with the oxacillin 1 Ilg disk to detect reduced penicillin susceptibility is established. 9 ,10 The 1 Ilg oxacillin disk has also been shown to have good discriminatory power in detecting reduced penicillin susceptibility in Neisseria meningitidis. 11 Although the reference method for MIC determination for viridans streptococci is the broth micro dilution method,12 MICs determined by penicillin E-test (AB BIODISK, Solna, Sweden) have been shown to correlate well with the reference method with 97.4% and 92.3% being within ± 1 dilutionY' 14 We have therefore tested local viridans group streptococcal isolates by the disk diffusion method using both penicillin 2 unit and 10 unit disks and the oxacillin 1 Ilg disk. The disk testing results were correlated with penicillin E-test MICs to determine if any interpretative criteria could be derived that would reliably detect decreased penicillin susceptibility in these isolates. MATERIALS AND METHODS Isolates

One hundred consecutive viridans group streptococcal isolates were collected between 3 hospitals and a community laboratory in Auckland from May 1996 until February 1997. Viridans group streptococci isolated from sterile sites and clinically significant isolates from non-sterile sites were included. Isolates were speciated in the referring laboratories according to each laboratory's protocol by either the API 20 STREP Kit (bio Merieux Vitek, Inc., Hazelwood, USA), or the Rapid ID 32 STREP Kit (bio Merieux Vitek, Inc., Hazelwood, USA) or the Microscan Walkaway40 (DADE International). Susceptibility testing

The inoculum was prepared by the direct colony suspension method from an overnight growth. The turbidity was adjusted to 0.5 McFarland's standard. For each organism two Mueller-Hinton agar plates supplemented with 5% defibrinated sheep blood were inoculated and disk diffusion testing with 2 and 10 unit penicillin disks, an oxacillin 1 pg disk and a penicillin E-test MIC were determined simultaneously. Plates were incubated at 35°C in 5% C02 atmosphere and read after 24 h. NCCLS methods for disk susceptibility testing were followed. 7- 9 E-test MrCs were

0031-3025/98/020188-04 © 1998 Royal College of Pathologists of Australasia

SCREENING FOR PENICILIN RESISTANCE IN IX-STREPTOCOCCI

189

Penicillin

E-test

MIC ~g/ml

0.008 0.016 0.023 0.032 0.047 0.064 0.094 0.125 0.19 0.25 0.38 0.5 0.75 2 3 4

1

1

6

1

1

1

1 2

2

1 2

433 1

: 3 1

2

322 2 1

2

1

1

1

2

311

3

1

1

1

1

1

: 2

.1 1

1

2

1 1

1

12 16 11

14 15 16 17 18 19 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 36 39 40 42 43 44 zone of inhibition, mm

Fig. 1 Scattergram representing penicillin E-test MICs versus penicillin 10 unit disk zone diameters derived from testing 100 viridans group streptococci. Solid horizontal line represents MIC interpretative criterion,12 dotted vertical line represents previous interpretative criterion7 and solid vertical line represents a proposed interpretative criterion, resistance :so 35 mm. Penicillin E-test

MIC IJglml

0.008 0.016 0.023 0.032 0.047 0.064 0.094 0.125 0.19 0.25 0.38 0.5 0.75 2 3 4 6 12 16

3 3 1 1

o

8

12 15 16 17 18 19 20 21

22 23 24 25 26 27 28 29 30 31 32 33 34 35

zone of inhibition, mm

Fig. 2 Scattergram representing penicillin E-test MICs versus penicillin 2 unit disk zone diameters derived from testing 100 vitidians group streptococci. Solid horizontal line represents MIC interpretative criterion,12 solid vertical line represents a proposed interpretative criterion, resistance :so 27 mm. determined in accordance with the manufacturer's instructions for use and interpretation. NCCLS criteria for interpretating penicillin MICs: sensitive :so 0.12 ,ug/ml, intermediate 0.25-2,ug/ml and resistant 2: 4 ,ug/ml were used. 12. 15 Penicillin MICs determined by E-test and the zone diameters obtained with the 3 disks were compared by scatter plots. Reduced penicillin susceptibility was defined as an MIC 2: 0.25 ,ug/mL The discriminatory power of each of the disks to detect deereased penicillin susceptibility was analysed by determining sensitivity, specificity and predictive values at previous and existing interpretative criteria. For the penicillin 10 unit disk, the previous NCCLS recommendation :so 27 mm as resistant, was evaluated. 7 For the oxacillin l,ug disk, the eurrent criteria, meningococcal :so 10 mm as resistant,and pneumococcal, :so 19 mm as resistant, were evaluated.

RESULTS One hundred viridans group streptococci were recovered from blood (69), pleural aspirates (six), peritoneal dialysis fluid (five), bone (four), intra-abdominal collections

(three), knee aspirates (three), cerebrospinal fluid (two) and others (eight) including aortic graft, eyes and wounds. Among them 25 were identified as S. mitis, 19 as S. milleri group (including S. anginosus eight, S. intermedius one), seven as S. sanguis, three as S. oralis, two as S. salivarius and 44 were not speciated. Thirty-seven isolates had reduced susceptibiliy to penicillin. Of the ten isolates resistant to penicillin, MICs ?: 4 {lg/ml, five were S. mitis and five were ul1speciated. Twenty-seven isolates had intermediate susceptibility to penicillin, MICs 0.25-2 {lg/ml; S. mitis (eight), S. oraUs (three), S. anginosus (two), S. sanguis (two), S. salivarius (one) and unspeciated (11). Scatter plots of penicillin E-test MIC versus zone diameters for the thrce disks are depicted in Figs. 1,2 and 3. The ability of the disks to detect decreased susceptibility to penicillin using previous and existing interpretative criteria varied considerably, see Table 1. The penicillin 10 unit disk, using NCCLS 1993 criteria, ie; resistant :s; 27 mm,7 had a sensitivity of 57% with 16 of 37 (43%) 'isolates with reduced penicillin susceptibility being misclassified

190

Pathology (1998), 30, May

POTTUMARTHY and MORRIS

Penicillin E-test 0.008 MIC 0.016

Ilg/ml 0.023

1: 3

0.032 4:

0.047 0.064

1:

0.094

1:

.

0.125 0.19 0.25

1

0.38

2

0.5

4

0.75

1

2

4

3

2

4

3

6

3

2

1

4

1

1:

2

1: 2

6' 3 1: 1

3 4

1:

12 16 6 8

9

10

11

12

13

14

15

16 17 18

19 20 21

22 23

24 25 26 27 28 29

zone of inhibition, mm

Fig. 3 Scattergram representing penicillin E-test MICs versus oxacillin 1 f,lg disk zone diameters derived from testing 100 viridans group streptococci. Solid horizontal line represents MIC interpretative criterion,12 dotted vertical line represents existing interpretative criterion for meningococci,l1 dashed vertical line represents existing interpretative criterion for pnuemococci,1O solid vertical line represents a proposed interpretative criteria, resistance :517 mm. TABLE

I

Sensitivity, specificity and predictive values of 2 and 10 unit penicillin disks and I pg oxacillin disk to detect decreased penicillin susceptibility in viridans group streptococci*

Interpretative criteria Previous and existing criteria Penicillin 10 unit disk :5 27 mmt Oxacillin 1 pg disk :5 10 mm:j: Oxacillin I pg disk :5 19 mm§ Proposed criteria Penicillin 10 unit disk :5 35 mm Penicillin 2 unit disk :5 27 mm Oxacillin I pg disk :5 17 mm

Sensitivity (%)

Specificity (%)

Positive predictive value (%)

57 70 100

98 95 33

95 90 47

79 85 100

100 100 100

30 46 67

46 52 64

100 100 100

* Decreased penicillin susceptibility defined as an B-test penicillin MIC of

Negative predictive value (%)

;,. 0.25 I'g/mI.

t 1993 NCCLS interpretative criteria for streptococci other than S. pneumoniae.7

*§ Interpretative criteria for N. meningitidis,'l Interpretative criteria for S. pneumoniae. 9

• 10

as susceptible and one of 63 (2%) susceptible isolates classified as resistant, see Fig_ 1. The oxacillin 1 p.g disk, using meningococcal criteria, ie; resistant :::s 10 mm 11 had a sensitivity of 70% with 11 of 37 (30%) of isolates with reduced penicillin susceptibility misclassified as susceptible and three of 63 (5%) susceptible isolates classified as resistant, see Fig. 3. The oxacillin 1 p.g disk, using pneumococcal criteria, ie; resistant ::5 19 mm9, 10 had a sensitivity of 100%, with 42 of 63 (67%) susceptible isolates misclassified as resistant, see Fig. 3. The specificity of proposed interpretative criteria, chosen as the smallest zone sizes with 100% sensitivity, for each of the three disks are given in Table 1. The penicillin 2 unit disk, resistant ::5 27 mm, had a specificity of 46% with 34 of 63 (54%) of susceptible isolates misclassified as resistant (Fig 2). The penicillin 10 unit disk, resistant :::s 35 mm, had a specificity of 30% with 44 of 63 (70%) of susceptible isolates misclassified as resistant (Fig O. The oxacillin 1 p.g disk, resistant :::s 17 mm, had a specificity of 67% with 21 of 63 (33%) of susceptible isolates misclassified as resistant (Fig. 3).

DISCUSSION Although penicillin-resistant isolates of viridans group streptococci were cultured from gingival flora of patients receiving penicillin prophylaxis in the early 1960s, resistance among clinically significant isolates has been uncommon until recently.s Three recent studies have reported 38%, 56% and 57% of blood culture isolates have reduced penicillin susceptibility.4-6 Of our isolates, 37% had reduced penicillin susceptibility, 27% had intermediate susceptibility and 10% were resistant. For S. mitis isolates, 13 of 25 (52%) had reduced susceptibility, five of which were resistant. These results parallel those recently reported by others.s Two of eight (25%) S. anginosus isolates had intermediate susceptibility to penicillin. Among viridans group streptococci the S. milleri group is noted to be the most susceptible to penicillin, but emerging resistance in this group has recently been reported, mostly in S. anginosus isolates.s,16 None of the previous or existing criteria rdiably detect reduced penicillin susceptibility in viridans group streptococci. Penicillin 10 unit and oxacillin 1 pg disk using

SCREENING FOR PENICILIN RESISTANCE IN IX-STREPTOCOCCI meningococcal criteria lacked sensitivity. The oxacillin 1 pg disk, using pneumococcal criteria, had maximum sensitivity but lacked specificity (Table 1). We evaluated the discriminatory power of the disks, at proposed interpretative criteria set to maximise the sensitivity. Using these criteria the oxacillin disk had a greater specificity, 67%, compared to that of penicillin 2 and 10 unit disks, 46% and 30% respectively (Table 1). The most recent NCCLS documents recommend determining penicillin susceptibility for viridans group streptococci isolated from sterile sites by an MIC method. 8, 9 No particular method is mandated by the NCCLS, indicating that laboratories have a choice of which method to use. We chose the E-test method because it is commonly used to determine an MIC as well as its simplicity and good correlation with the broth microdilution method.13,14 The need to determine penicillin MICs for this group of organisms is limited. For endocarditis the penicillin MIC is required because treatment regimens are based on MIC results? Determining a penicillin MIC by E-test or reference broth dilution method is therefore mandatory for endocarditis isolates. However, for febrile neutropenic patients in whom bacteremia due to viridans group streptococci is a leading cause, the relevance of reduced penicillin susceptibility for therapy is unknown. Finally, no NCCLS recommendations exist for testing penicillin susceptibility of viridans group streptococci from non-sterile sites, eg; S. milleri from intra-abdominal pyogenic infections. Although the determination of a penicillin MIC for viridans group streptococci isolated from sterile sites is not too labour-intensive or expensive it would be useful to be able to screen for reduced penicillin susceptibility. Analogous to penicillin susceptibility testing of S. pneumoniae, where penicillin MICs are determined for significant isolates, eg; those causing meningitis, and other isolates are screened with an oxacillin disk, determining penicillin MICs for viridans group streptococci could be restricted to endocarditis isolates. I? All other viridans group streptococci isolated from both sterile and non-sterile sites could initially be screened for reduced penicillin susceptibility, and penicillin MICs determined if they fail the screening test. If the need for a screening test is established our data suggest that the oxacillin 1 pg disk using new interpretative criteria is a better predictor of reduced susceptibility than either of the penicillin disks. Our suggested interpretative zone diameter for the oxacillin 1 Jig disk should be examined using a larger number of isolates and ideally compared with MICs determined by the reference method. In summary, we found that 37% of local viridans group streptococci had reduced penicillin susceptibility. Use of the penicillin 10 unit disk with NCCLS 1993, M2-AS, criteria is unreliable in detecting penicillin resistance. Our data support the removal by the NCCLS of the interpretative criteria for this disk for viridans streptococci. The oxacillin 1 pg disk using new interpretative criteria has a better discriminatory power than either penicillin 2 or 10 unit disks. ACKNOWLEDGEMENTS The authors wish to thank the following microbiology laboratories for providing the isolates: Auckland Hospital, Green Lane Hospital, Middlemore Hospital, and Diagnostic Laboratory. They also thank

191

Mr Terry Leydon, Product Manager, AB Biodisk, for donating the penicillin E-test strips, and Ms Sharon Paul, Charge Technologist, Microbiology, Green Lane Hospital, for her technical assistance. Address for correspondence: S.P., Microbiology Laboratory, Green Lane Hospital, Greenlane West, Auckland 3, New Zealand.

References

1. Bochud P-Y, Calandra T, Francioli P. Bacteremia due to viridans streptococci in neutropenic patients: a review. Am J Med 1994; 97: 256-64. 2. Singh KP, Morris A, Lang SDR, MacCulloch DM, Bremner DA. Clinically significant Streptococcus anginosus (Streptococcus milleri) infections: a review of 186 cases. NZ Med J 1988; 101: 813-6. 3. Wilson WR, Karchmer, AW, Dajani AS, Taubert KA, Bayer A, Kaye D, et al. Antibiotic treatment of adults with infective endocarditis due to streptococci, enterococci, staphylococci and HACEK microorganisms. JAMA 1995; 274: 1706-13. 4. Carratala J, A1caide F, Fernandez-Sevilla A, Corbella X, Linares J, Gudiol F. Bacteremia due to viridans streptococci that are highly resistant to penicillin: increase among neutropenic patients with cancer. Clin Infect Dis 1995; 20: 1169-73. 5. DOelTI GV, Ferraro MJ, Brneggemann AB, Ruoff, KL. Emergence of high rates of antimicrobial resistance among viridans group streptococci in the United States. Antimicrob Agents Chemother 1996; 40: 891-4. 6. McWhinney PHM, Patel S, Whiley RA, Hardie JM, Gillespie SH, Kibbler Cc. Activities of potential therapeutic and prophylactic antibiotics against blood culture isolates of viridans group streptococci from neutropenic patients receiving ciprofioxacin. Antimicrob Agents Chemother 1993; 37: 2493-5. 7. National Committee for Clinical Laboratory Standards. Peiformance standards for antimicrobial disk susceptibility tests-5th edn; approved standard. NCCLS document M2-A5. VilIanova, PA: NCCLS, 1993. 8. National Committee for Clinical Laboratory Standards. Peiformance standards for antimicrobial susceptibility testing-6th infonnational supplement. NCCLS document M100-S6. Villanova, PA: NCCLS, 1995. 9. National Committee for Clinical Laboratory Standards. Peiformance standards for antimicrobial disk susceptibility tests-6th edn; approved standard. NCCLS document M2-A6. Wayne, PA: NCCLS, 1997. 10. Swenson JM, Hill BC, Thornsberry C. Screening pneumococci for penicillin resistance. J Clin Microbiol 1986; 24: 749-52. 11. Campos J, Trnjillo G, Seuba T, Rodriguez A. Discriminative criteria for Neisseria meningitidis isolates that are moderately susceptible to penicillin and ampicillin. Antimicrob Agents Chemother 1992; 36: 1028-31. 12. National Committee for Clinical Laboratory Standards. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically--4th edn; approved standard. NCCLS document M7-A4. Wayne, PA: NCCLS, 1997. 13. Rindler JA, Brnckner DA. MIC testing of viridans streptococci using E-test as compared to a reference method. In: Program and abstracts of the 33rd Interscience Conference on Antimicrobial Agents and Chemotherapy 1993; p. 254. 14. Rosser SJ, Alfa MJ, Hoban S, Kennedy J, Kabani A, Harding GKM. E-test vs. agar dilution susceptibility testing for viridans streptococc. In: Program and abstracts of the 96th General Meeting for American Society for Microbiology 1996; abstract C-183: p.33. 15. National Committee for Clinical Laboratory Standards. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically-3rd edn; approved standard. NCCLS document M7-A3. VilIanova, PA: NCCLS, 1993. 16. Santar C, Canigia LF, Relloso S, Lanza A, Bianchini H, Smayevsky J. Species belonging to the "Streptococcus milleri" group: antimicrobial susceptibility and comparative prevalence in significant clinical specimens. J Clin Microbiol 1996; 34: 2020-2. 17. Jorgensen JH, Swenson JM, Tenover FC, Ferraro MJ, Hindler JA, Murray PR. Development of interpretative criteria and quality control limits for broth microdilution and disk diffusion antimicrobial susceptibility testing of Streptococcus pneumoniae. J Clin Microbiol 1994; 32: 2448-59.