Comparative evaluation of tigecycline susceptibility testing methods for carbapenem-resistant Klebsiella pneumoniae and Acinetobacter baumannii.

backgroundIn this study, we aimed to evaluate the performance of different tigecycline susceptibility testing methods for CRKP and CRAB using the broth microdilution (BMD) method as the gold standard, under different breakpoint criteria.

methodsA total of 53 CRKP and 47 CRAB clinical isolates, collected from December 2021 to January 2023, were tested for tigecycline susceptibility using six clinically common methods: Vitek2 (using the Vitek2 Compact system), the Kirby-Bauer (KB) disk diffusion susceptibility test, the modified KB (mKB) test with a buffer solution, the Epsilometer test (E-test), minimum inhibitory concentration test strips (MTS) and BMD (as the reference standard). The susceptibility test results were interpreted based on the standards issued by the European Committee on Antimicrobial Susceptibility Testing(EUCAST), the United States Food and Drug Administration (FDA), and the British Society for Antimicrobial Chemotherapy(BSAC). The performance of each method was assessed through essential agreement (EA), categorical agreement (CA), error rates and Kappa (κ) values.

resultsIn terms of methodology, the mKB significantly increased the inhibition zone diameter compared with the KB for both CRKP (3.2 ± 0.8 mm) and CRAB (3.6 ± 0.8 mm) (P < 0.05, respectively). Using BMD as the gold standard and following FDA criteria, the mKB demonstrated an accuracy rate significantly higher than the KB in interpreting CRKP susceptibility results (62.3% vs. 1.9%, P < 0.05). The agreement rate between the E-test and the MTS for CRKP was as high as 84.9%, indicating no statistically significant difference (P > 0.05). However, the E-test yielded significantly higher accuracy than the mKB (98.1% vs. 60.4%, P < 0.05). For CRAB, the accuracy of the E-test was lower than that for CRKP (68.1% vs. 98.1%). When compared with BMD, the E-test had a high EA value of 92.5% for CRKP and a lower EA value of 68.1% for CRAB. Under the FDA standard for CRKP, the E-test yielded a CA value of 98.1%, with very major error (VME) and major error (ME) rates of 0%, a minor error (mE) rate of 1.9% and a κ-value of 0.658, indicating substantial agreement. In contrast, when applied to CRAB, the performance of the E-test declined markedly, with a CA value of only 68.1%, VME and ME rates still at 0%, but a much higher mE rate of 31.9% and a κ-value of just 0.082, indicating poor agreement.

conclusionThe mKB improves the detection performance of the KB, suggesting that pH variation is one of the factors affecting the accuracy of KB-test-based TST. Among the five clinical verification methods included in this study, the E-test demonstrates the best overall performance under all interpretive criteria, making it more suitable for confirming the tigecycline susceptibility of CRKP. However, this method appears less reliable for CRAB, indicating the significant impact of interspecies differences on TST performance. Additionally, the verification methods aligned better with the FDA interpretive criteria. Future research should focus on methodological optimisation based on various influencing factors and on refining interpretive breakpoints specific to each method, which will be key to improving TST accuracy.