TABLE 3

Population parameter estimates for imipenem plus aminoglycoside combination models against four A. baumannii strains

ParameterSymbol (unit)Mean value (SE%) for strain with drugsd
ATCC 19606; IPM+AGS (TOB, ISE, and AMK)FADDI-AB016; IPM+AGS (ISE)FADDI-AB014; IPM+AGS (ISE)FADDI-AB034; IPM+AGS (TOB)
Log10 initial inoculumLogCFU07.99 (2.3); 6.87 (1.6)a6.71 (3.8)6.71 (3.8)6.62 (2.8)
Log10 maximum population sizeCFUmax9.65 (2.3)9.57 (1.1)9.30 (6.7)9.41 (1.5)
Replication rate constantk21 (h−1)50 (fixed)50 (fixed)50 (fixed)50 (fixed)
Mean generation time
    IPMs AGSsk12,SS−1 (min)96.4 (11.2)87.0 (5.8)65.9 (31.0)56.0 (4.8)
    IPMr AGSik12,RI−1 (min)1,010 (4.0)226 (11.9)196 (31.3)812 (6.3)
    IPMi AGSrk12,IR−1 (min)96.4 (11.2)87.0 (5.8)65.9 (31.0)56.0 (4.8)
Log10 mutation frequency
    IPMLogMUT,IPM−3.16 (5.8)−1.14 (14.3)−1.28 (69.2)−3.55 (4.6)
    AGSLogMUT,AGS−5.16 (2.5)−3.99 (4.3)−4.37 (6.7)−7.20 (1.9)
Killing by imipenem
Maximum killing rate constantKmax,IPM (h−1)2.48 (12.4)1.32 (17.8)1.84 (19.4)2.31 (10.5)
Imipenem concn causing 50% of Kmax,IPM
    IPMs AGSsKC50,SS,IPM (mg/liter)Mono, 1.01 (17.7); combo (AGS ≥ 4 mg/liter), 0.971 (18.2)0.380 (17.8)0.332 (74.5)20.7 (19.3)b
    IPMr AGSiKC50,RI,IPM (mg/liter)Mono, 16.5 (9.0); combo, 13.2 (5.5)20.5 (7.3)23.4 (10.8)84.0 (8.7)b
    IPMi AGSrKC50,IR,IPM (mg/liter)Mono, 1.76 (11.4); combo, 0.422 (16.3)1.30 (27.7)4.15 (39.7)Mono, 22.4 (13.9)b combo, 6.42 (8.8)b
Killing by aminoglycosides
Maximum killing rate constant
    IPMs AGSsKmax,SS,AGS (h−1)5.42 (13.7)4.48 (22.7)5.84 (16.9)7.67 (12.1)
    IPMr AGSiKmax,RI,AGS (h−1)1.38 (19.8)1.32 (13.6)1.46 (24.6)3.71 (9.95)
    IPMi AGSrKmax,IR,AGS (h−1)5.42 (13.7)4.48 (22.7)4.72 (26.3)7.67 (12.1)
Aminoglycoside concn causing 50% of Kmax,AGS
    IPMs AGSsKC50,SS,ISE (mg/liter)22.1 (44.2)0.415 (27.7)0.216 (37.3)
KC50,SS,TOB (mg/liter)7.54 (11.5)2.87 (21.5)
KC50,SS,AMK (mg/liter)18.2 (12.5)
    IPMr AGSiKC50,RI,ISE (mg/liter)52.5 (6.8)7.35 (19.4)9.24 (24.1)
KC50,RI,TOB (mg/liter)69.1 (7.9)98.1 (7.3)
KC50,RI,AMK (mg/liter)75.2 (7.4)
    IPMi AGSrKC50,IR,ISE (mg/liter)147 (8.9)190 (8.5)178 (35)
KC50,IR,TOB (mg/liter)123 (5.82)120 (8.35)
KC50,IR,AMK (mg/liter)243 (5.7)
Mean turnover time for hypothetical signal molecules (= 1/kout,sig)MTT (h)0.942 (14.1)0.753 (32.9)0.794 (35.8)0.072 (31)
Maximum inhibition by hypothetical signal moleculesImax,sig120.992 (9.64)0.984 (18.4)0.986 (21.4)0.980 (20)
Log10 of hypothetical signal molecule concn at 50% of max effectLog10,IC50,sig7.61 (1.5)8.03 (2.6)8.01 (2.9)8.10 (2.2)
Hill coefficient
    IPMHillIPM5.0 (fixed)c2.0 (fixed)c5.0 (fixed)c4.04 (13.5)
    AGSHillAGS2.71 (12.6)1.0 (fixed)1.0 (fixed)1.0 (fixed)
SD of residual error on log10 scaleSDCFU0.401 (7.3)0.481 (11.0)0.287 (15.6)0.383 (9.6)
  • a Initial inocula from two different experiments. All other model parameter estimates were assumed to be the same at both initial inocula.

  • b The final model for isolate FADDI-AB034 used the same parameter estimate for KC50,SS,IPM and KC50,RI,IPM with and without the aminoglycoside, since allowing for different parameter estimates yielded no improvement. The model benefitted significantly from using different estimates for KC50,IR,IPM for monotherapy and combination therapies.

  • c We used only three imipenem concentrations in each of the static-concentration time-kill studies. Therefore, precise estimation of the Hill coefficients was difficult. It was, however, beneficial to include a Hill coefficient for imipenem. After evaluation of different Hill coefficient values, the Hill coefficients were fixed at the values shown here.

  • d SE%, relative standard error, in percent; mono, monotherapy; combo, combination therapy.