Quality Control Limits for Dilution and Disk Diffusion Susceptibility Tests of Trovafloxacin against Eight Quality Control Strains

Abstract

A 10-laboratory collaborative effort was designed to generate data to propose quality control limits for susceptibility tests of trovafloxacin. Broth microdilution, agar dilution, and disk diffusion tests were evaluated with eight different control strains. All tests were reproducible, and control limits are proposed.

For susceptibility testing of any new antimicrobial agent, it is necessary to have established quality control guidelines. We report herein the results of a collaborative study designed to determine the quality control limits for susceptibility tests of a new fluoroquinolone against eight standard quality control organisms. This protocol involved 10 participating facilities in contrast to the 5-laboratory protocol that has been used previously. The 10 participants are listed in the acknowledgments. The larger database should provide control limits that can be advanced with greater confidence because interlaboratory variability is usually the most important consideration. The drug that was evaluated by this expanded study protocol is trovafloxacin (CP-99,219), which has in vitro activity against gram-positive and gram-negative aerobic and anaerobic bacteria (1, 2, 4, 7, 8).

Dilution tests.

Broth microdilution tests were performed with Escherichia coli ATCC 25922, Staphylococcus aureus ATCC 29213, Enterococcus faecalis ATCC 29212, Pseudomonas aeruginosa ATCC 27853, Streptococcus pneumoniae ATCC 49619, and Haemophilus influenzae ATCC 49247. Broth microdilution trays were obtained from a common source and were prepared to contain twofold concentrations of trovafloxacin (0.001 to 16 μg/ml) diluted in each of five different lots of cation-adjusted Mueller-Hinton broth (CAMHB), representing three different manufacturers (Acumedia, BDMS, and Difco). The control drug, ofloxacin, was diluted (0.004 to 32 μg/ml) in one lot of CAMHB. For testing H. influenzae, Haemophilus test medium supplements were added to each lot of CAMHB, and 2 to 3% lysed horse blood was added for testing S. pneumoniae. Fifteen replicates of each strain were also tested by agar dilution in one laboratory in order to determine whether the same MIC control limits could be applied to agar dilution and broth microdilution tests.

Agar dilution tests were performed with Neisseria gonorrhoeae ATCC 49226. Agar dilution plates of six different lots of GC agar were prepared to contain twofold concentrations of trovafloxacin ranging from 0.06 to 0.0005 μg/ml. One lot was common to all laboratories; the remaining five lots were assigned to two laboratories each as unique lots. Ciprofloxacin, the control drug, was prepared in the same concentrations in one lot. All plates were supplemented with 1% XV supplement (PML Microbiologicals, Wilsonville, Oreg.).

Disk diffusion.

Disk diffusion tests were performed with E. coli ATCC 25922, S. aureus ATCC 25923, P. aeruginosa ATCC 27853, S. pneumoniae ATCC 49619, H. influenzae ATCC 49247, and N. gonorrhoeae ATCC 49226. Six lots of Mueller-Hinton agar (MHA) representing four manufacturers (Acumedia, BDMS, Difco, and Oxoid) were tested. Haemophilus test medium supplements were added to the MHA for testing H. influenzae, and 5% sheep blood was added for testing S. pneumoniae. Six lots of GC agar supplemented with 1% XV supplement were employed for testing N. gonorrhoeae. For each medium type, one lot was common to all 10 laboratories and five were assigned to 2 laboratories, each as unique lots. Trovafloxacin disk content of 10 μg was selected based on previous work showing poor trovafloxacin diffusion in agar and very small zones around 5-μg disks (2). Two lots of commercially prepared trovafloxacin 10-μg disks (Remel and Difco) were tested throughout.

Study design.

On each of 10 test days, each laboratory tested one microdilution tray (five trovafloxacin MICs) and two trovafloxacin disk test plates, one from the common lot and one from the unique lot of MHA (four zone diameter measurements). For the GC agar dilution tests, 30 separate inoculum preparations were tested on two series of trovafloxacin plates (one from the common lot and one from the unique lot). Trovafloxacin inhibitory-zone diameters on GC agar were measured on 20 plates of the unique lot and 10 plates of the common lot of GC agar, a total of 60 zone diameter measurements per laboratory.

All laboratories were instructed to follow the test procedures precisely as outlined in the National Committee for Clinical Laboratory Standards documents for dilution tests (5) and disk diffusion tests (6).

For all but one of the seven organisms tested by dilution methods, there was a unimodal distribution of trovafloxacin MICs (Table 1) with a well-defined mode and with 98.8 to 100% of MICs within a 3-dilution range (mode ± 1 doubling concentration). The exception was P. aeruginosa ATCC 27853, for which the trovafloxacin MIC distribution suggested a true mode halfway between the log₂ concentrations that were tested (0.5 and 1.0 μg/ml). In that situation, the control range is the midpoint mode ± 1.5 doubling concentrations, which is a 4-dilution range (0.25 to 2.0 μg/ml). The microdilution results for the control drug ofloxacin were all within the control ranges for E. faecalis ATCC 29212, P. aeruginosa ATCC 27853, E. coli ATCC 25922, and S. pneumoniae ATCC 49619. For H. influenzae ATCC 49247 and S. aureus ATCC 29213, 99.0% of ofloxacin test results were in the control range.

TABLE 1.

Distribution of trovafloxacin MICs for seven quality control strains, determined by the broth microdilution procedure^a

MIC (μg/ml)	No. of replicates
	E. coli ATCC 25922	S. aureus ATCC 29213	E. faecalis ATCC 29212	P. aeruginosa ATCC 27853	S. pneumoniae ATCC 49619	H. influenzae ATCC 49247	N. gonorrhoeae ATCC 49226
0.001
0.002							1
0.004	31	1				10	111
0.008	362	71				483	338
0.015	106	323			1	7	150
0.03	1	100	1		4
0.06		4	49		136
0.125		1	431		330
0.25			18	8	29
0.5				231
1.0				261
2.0
4.0

^aHorizontal lines delimit the proposed MIC control ranges for trovafloxacin.  

The agar dilution trovafloxacin MICs for the six organisms tested by broth microdilution were the same or within 1 dilution of the broth microdilution MIC modes. Thus, the MIC limits proposed for broth microdilution tests should be applicable to agar dilution tests as well.

The distributions of trovafloxacin inhibitory-zone diameters for six quality control organisms are given in Table 2. By using the median statistic of Gavan et. al. (3) and expanding the range by a millimeter when necessary to include 95% of the results, reasonable quality control ranges are proposed for five strains (Table 2). The quality control limits for other quinolones (5- or 10-μg disks) in testing N. gonorrhoeae ATCC 49226 are 9 to 11 mm. For 10-μg trovafloxacin disks, a 14-mm range was required to include 95% of zone measurements. This spread is partially related to the exceptionally large zone diameters observed with this potent drug. Unless a smaller disk content (e.g., 1 μg) is used for this organism, it would appear that the 14-mm range proposed herein (42 to 55 mm) is reasonable.

TABLE 2.

Zone diameter distributions for trovafloxacin and control drugs

Microorganism	Diam (mm)a
	Trovafloxacin		Ciprofloxacin QC range (% in range)	Ofloxacin QC range (% in range)
	Mode (range)	Proposed QC range (% in range)
E. coli ATCC 25922	32/33b (28–37)	29–36 (99.0)	30–40 (98.5)
S. aureus ATCC 25923	32 (28–37)	29–35 (97.5)	22–30 (100)
P. aeruginosa ATCC 27853	24 (21–28)	21–27 (99.7)	25–33 (100)
S. pneumoniae ATCC 49619	28 (24–34)	25–32 (97.5)		16–21 (99.5)
H. influenzae ATCC 49247	35 (31–40)	32–39 (98.0)		31–40 (98.5)
N. gonorrhoeae ATCC 49226	48 (40–57)	42–55 (95.2)		48–58 (96.5)

^aQC, quality control. 

^bEqual numbers of zones at 32 and at 33 mm. 

These proposed QC ranges have all been approved by the National Committee for Clinical Laboratory Standards subcommittee on antimicrobial susceptibility testing and will appear in future publications by that group.