Ability of the VITEK(R) 2 system to detect group B streptococci with reduced penicillin susceptibility (PRGBS)

Sir, Group B Streptococcus (Streptococcus agalactiae, GBS) is the leading cause of neonatal sepsis and meningitis and an important pathogen among elderly people and those suffering from underlying medical disorders.1,2 The highest GBS mortality and morbidity result from invasive infections in neonates.1,2 Approximately 5% of GBS-infected infants die and survivors often suffer from severe neurological sequelae.2 Intrapartum antibiotic prophylaxis has been recommended by the CDC2 and is prescribed for pregnant women who have GBS isolated from vaginal specimens. Since the introduction of prophylaxis, the rate of GBS infection during the first post-natal week has decreased. Penicillins are the first-line agents in the prophylaxis and treatment of GBS infections because all clinical GBS isolates have been considered to be uniformly susceptible to β-lactams, including penicillins.2,3 However, we identified and characterized several GBS isolates demonstrating reduced penicillin susceptibility (PRGBS) through acquisition of multiple mutations in the penicillin-binding protein 2X (pbp2x) gene,4 and similar isolates were reported in the USA,5 Canada6,7 and Japan.8 After our research was published, EUCAST (http://www.eucast.org/clinical_breakpoints/) defined a clinical penicillin MIC breakpoint for Streptococcus groups A, B, C and G, together with the penicillin MIC resistance breakpoint (>0.25 mg/L). The EUCAST breakpoint is higher than the breakpoint for penicillin susceptibility set by the CLSI (≤0.12 mg/L). Until recently, PRGBS were isolated from respiratory specimens, blood, decubitus ulcers and adult hip-joint fluid,4–8 with no report of PRGBS isolated from neonates or vaginal specimens of pregnant women. The isolation rate of PRGBS from various sources is approximately 2.3% in Japan.8 The MICs of penicillin G for PRGBS (0.25–1 mg/L) are near the breakpoint set by the CLSI (≤0.12 mg/L). Therefore it is unclear whether automated susceptibility testing machines such as VITEK® 2 can detect PRGBS accurately. Because the VITEK® 2 system is widely used in clinical laboratories in Japan, we used this system as an example in order to evaluate the ability of automated susceptibility testing machines to detect PRGBS. The MICs of penicillin G were determined for 28 PRGBS using the agar dilution method as per CLSI recommendations.3Streptococcus pneumoniae ATCC 49619 was used as a quality control for MIC measurements. It was confirmed that these PRGBS harboured the amino acid substitutions in pbp2x genes, as described previously.4,8 We performed the determination of the MICs of penicillin G for 28 PRGBS three times using the VITEK® 2 compact system with AST-P546 cards (bioMérieux Clinical Diagnostics, Marcy l'Étoile, France) in accordance with the manufacturer's instructions. The results of the comparison between the MICs of penicillin G for 28 PRGBS, as determined by agar dilution and the VITEK® 2 system, are shown in Table 1. Although the MICs determined by the agar dilution method were 0.25–1 mg/L [above the breakpoint (≤0.12 mg/L) set by the CLSI], the MICs determined by the VITEK® 2 system were ≤0.12–1 mg/L. The MICs determined by the VITEK® 2 system were ≤0.12 mg/L in 38 instances (38/84, 45.2%; 84 instances = 28 strains × 3 times). The number of strains for which the MICs determined by the VITEK® 2 system were ≤0.12 mg/L at least two of three times was 13 (13/28, 46.4%). Comparison between MICs of penicillin G for 28 PRGBS determined by agar dilution and by VITEK® 2 ND, not determined. Comparison between MICs of penicillin G for 28 PRGBS determined by agar dilution and by VITEK® 2 ND, not determined. In this study, we investigated the ability of the VITEK® 2 system to detect PRGBS. It detected only half of the PRGBS in this study. Automated susceptibility testing machines such as VITEK® 2 are used in clinical settings worldwide, and these results suggest that many PRGBS may be misclassified as ‘susceptible’ to penicillin G. We recently revealed that PRGBS tends to be resistant to fluoroquinolones and macrolides, in addition to having reduced penicillin susceptibility,9 indicating that the classification of susceptibility to penicillin G is very important. The worldwide misclassification of PRGBS as ‘susceptible’ to penicillin G is undesirable and hinders attempts to clarify the clinical significance of reduced susceptibility to penicillin G. The MICs of penicillin G for PRGBS (0.25–1 mg/L) are near the ‘susceptible’ breakpoint (≤0.12 mg/L) set by the CLSI, while the MICs of oxacillin (2–8 mg/L) and ceftizoxime (4–128 mg/L) for PRGBS are higher than those of penicillin-susceptible GBS.4 However, the VITEK® 2 system AST-P546 cards for S. agalactiae do not include MIC determinations of oxacillin or ceftizoxime. We believe that inclusion of these MICs would enable more accurate detection of PRGBS by automated susceptibility testing machines. Moreover, it would be better for these machines to contain systems to alert operators to PRGBS-suspicious isolates when the MICs of penicillin G indicate a range near the susceptibility breakpoint, e.g. at 0.12 mg/L. Previously we reported that disc diffusion methods using oxacillin, ceftizoxime and ceftibuten were useful for detecting PRGBS.10 The disc diffusion method for detecting PRGBS does not require expensive or specialized equipment. Therefore, prior to any improvements in automated susceptibility testing machines, the disc diffusion method for detecting PRGBS will be useful for clinical microbiological laboratories worldwide. This study was supported by grants H21-Shinkou-Ippan-008 and H24-Shinkou-Ippan-010, from the Ministry of Health, Labor and Welfare, Japan, and, in part, by a research grant for medical science from the Takeda Science Foundation (2012). The third grant covered the cost of editing by Editage, as mentioned in the Transparency declarations section. The authors have no conflicts of interest to declare. The manuscript was edited by Editage, a language editing company. We thank Kumiko Kai, Yoshie Taki and Yumiko Yoshimura for technical assistance.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.