Methods: We tested 76 XDR AB and 94 XDR PsA isolates by the checkerboard method using 9 wells for each drug combination, labeled the serum achievable concentration (SAC) wells. The results were correlated by time-kill assay utilizing the concentrations of drug from the fractional inhibitory concentration well. Clinical cure, microbiological clearance, and 30-day all-cause mortality were determined in patients with AB or PsA pneumonia or bacteremia treated with an evaluated regimen by checkerboard assay. Patients were grouped by their antimicrobial regimen being synergistic and additive (S/A) versus indifferent and antagonistic (I/A). A secondary analysis was performed grouping regimens with and without a SAC well. Fishers exact and Mann-Whitney U tests were utilized for analysis.
Results: Among AB isolates, minocycline-colistin was the most frequent combination to display an S/A effect at 7% and 79%, respectively. Among synergistic strains from this combination, bactericidal activity was achieved 100% of the time by the time-kill assay. Among PsA isolates, doripenem-amikacin was the most frequent combination to display an S/A effect at 8.5% and 39.4%, respectively. When evaluating clinical outcomes among all AB (n=18) and PsA (n=12) together, there were no differences between S/A vs I/A antibiotic regimens in clinical cure (38% vs 32%; p=1.00), microbiological clearance (63% vs 36%; p=0.24), or 30- day mortality (63% vs 50%; p=0.63). Likewise, there were no differences when comparing outcomes among those with and without a SAC drug regimen. However, when evaluating only AB, those with a SAC drug regimen compared to those without had improved microbiological clearance (88% vs 30%; p=0.025).
Conclusion: The checkerboard assay did not correlate with clinical outcomes for PsA, but for AB, microbiological clearance improved. A real time checkerboard platform for XDR AB may have clinical and microbiologic benefits. Real-time evaluation of the checkerboard platform is needed.
S. Pouch, None
K. Thomas, None
P. Pancholi, None
J. M. Balada-Llasat, None
See more of: Poster Abstract Session