The azithromycin MIC was 32 mg/liter, similar to the MIC90
of clinical isolates from patients with disseminated MAC disease (15
). Azithromycin mean pharmacokinetic parameter estimates (± standard deviation [SD]) for the central compartment were a total clearance of 24.1 ± 2.41 liters/h and a volume of 290 ± 41.7 liters. However, when the concentrations of azithromycin achieved in the central compartment were related to the MIC, as shown in , the exposures achieved with each dose rarely exceeded the MIC, except at very high doses, such as those equivalent to 16,000 mg a day in patients. However, the concentrations of azithromycin achieved within macrophages were several thousandfold higher than in the central or peripheral compartment (). demonstrates that the ratio of the intracellular to the extracellular concentrations varied with time and drifted down 3.9-fold from the start of the dosing interval until the 4-h time point, after which the ratios remained stable. Thus, the relationship between intracellular and extracellular azithromycin concentrations was nonlinear. The azithromycin pharmacokinetics within the macrophages were best described by a two-compartment model with an absorption rate constant of 3.18 ± 3.83/h, a mean total clearance of 0.0012 ± 0.0007 liter/h, a volume of the central compartment of 0.00004 ± 0.00007 liter, an intercompartmental clearance of 18.53 ± 23.62 liters/h, and a larger volume of the peripheral compartment of 0.02 ± 0.01 liter. Thus, the azithromycin clearance from macrophages was 20,000 times slower than from the central compartment, which, together with higher drug penetration, accounts for the high intracellular drug exposures. The azithromycin exposures achieved within macrophages are shown in .
Azithromycin exposures achieved in the central compartment and within infected macrophagesa
Fig 1 Ratio of the intracellular azithromycin concentration to that in the central compartment. At each of the time points when simultaneous drug concentrations were measured, azithromycin concentrations were much higher inside macrophages for all systems. (more ...)
On day 0 of the dose-effect study, the starting inoculum in each HFS was 5.5 log10 CFU/ml. On subsequent days, the Emax was 0.87 ± 0.13 log10 CFU/ml on day 4, with the highest Emax of 2.11 ± 0.26 log10 CFU/ml achieved on day 7, after which the Emax slowly started to decrease through days 14 to 28, when it fell to 1.53 ± 0.31 log10 CFU/ml (). The day 7 dose-response curve, based on intramacrophage drug concentrations, was described by an EC50 that was an area under the concentration-time curve from 0 to 24 h (AUC0-24)/MIC ratio of 17,288 ± 565 (r2 = 0.95). If central compartment concentrations, analogous to “serum” concentrations, had been used, the EC50 would have been an AUC0-24/MIC ratio of only 2.11 ± 0.26. The full exposure-effect curve at the end of the study (day 28) was described by the following equation: effect (log10 CFU/ml) = 7.60 − 1.53 × AUC0-24/MIC2.86/(12,9582.86 + AUC0-24/MIC2.86), where all AUC/MIC exposures are intramacrophage concentrations.
Fig 2 Azithromycin dose-response curves in the hollow-fiber system. Given that extracellular concentrations did not exceed the MIC in most cases but the drug nevertheless could kill MAC, we chose to use the intracellular concentrations for the dose-response (more ...)
The decrease in maximal kill after 7 days was due to emergence of drug resistance in a portion of the total population. On day 4, the drug-resistant subpopulation was below the limits of detection. However, by day 10, a resistant subpopulation could be demonstrated. The relationship between drug exposure and the size of the drug-resistant population was best depicted by the inverted U-shaped curve (). However, when the same cultures were examined for drug resistance on agar supplemented with the efflux pump inhibitor, there were no drug-resistant colonies on the agar. Thus, the entire drug-resistant subpopulation that arose early during therapy could be accounted for by efflux pump induction.
Fig 3 Relationship between the azithromycin-resistant subpopulation and azithromycin exposure on day 10 of the dose-effect study. The relationship between the size of the low-level-resistant subpopulation and drug exposures (AUC/MIC) was described by the familiar (more ...)
Dose-scheduling study results are shown in , which demonstrates that the total microbial burden did not differ with the dosing schedule. Similarly, the dosing schedule did not affect the sizes of drug-resistant subpopulations. This means that both the azithromycin microbial kill and resistance suppression are most closely linked to the AUC0-24/MIC ratio. On day 7, a low-level azithromycin-resistant subpopulation was encountered in all HFSs treated with azithromycin, with reduction in the size of the resistant subpopulation in the presence of the efflux pump inhibitor thioridazine (). This phenotype did not grow on agar containing 256 mg/liter azithromycin. By day 28, however, an even larger proportion of the azithromycin-resistant subpopulation could be inhibited by thioridazine (). This means that progressively larger populations of intracellular MAC continued to be recruited for efflux pump induction, which was dependent on the drug exposure (). In addition, day 28 cultures revealed that the resistant subpopulations in several HFSs grew on agar with 256 mg/liter azithromycin, but the efflux pump inhibitor did not change the size, consistent with the emergence of another, more stable mechanism of drug resistance (). Taken together, these experiments suggest that efflux pumps play an early role in establishing low-level resistance, which gives the intracellular MAC survival advantage until the emergence of high-level resistance.
Effect of the dosing schedule on the total microbial population. There was virtually no difference in microbial kill by dosing schedule for each drug exposure examined, consistent with an AUC/MIC-linked effect.
Fig 5 Emergence of high-level drug resistance follows low-level resistance mediated by efflux pumps. (A) On day 7, a low-level drug-resistant subpopulation was encountered. The subpopulation is a phenotype that grows on agar supplemented with 3× MIC (more ...)
We looked for the early induction of three putative efflux pump genes, with results shown in . The figure demonstrates a 56-fold induction of MAV_3306, which encodes an ATP binding cassette (ABC) transporter, in a stepwise increase over 3 days, i.e., in a dose-dependent fashion with time as the “dose.” This suggests a recruitment process that takes several days, as opposed to an on-off “switch.” On the other hand, for MAV_1406, which encodes a major facilitator superfamily (MFS) pump, there was an abrupt induction noted at the 24-h time point, at which time it had reached the maximum induction, which was maintained throughout the 72 h of the experiment. In contrast to these pumps, MAV_1695 demonstrated no significant induction, suggesting either that the pump it encodes is not involved in azithromycin resistance in MAC or that it may be induced at a later stage. Overall, these data suggest induction of efflux pumps ahead of the early phenotypic reversible low-level resistance, itself preceding nonreversible resistance by up to 3 weeks: an “antibiotic resistance arrow of time” ().
Fig 6 Real-time PCR of three putative azithromycin efflux pumps. MAC was incubated with 16 mg/liter azithromycin. (A) MAV_1659 demonstrated no significant upregulation during the first 72 h of exposure to azithromycin. (B) MAV_1406 demonstrated no significant (more ...)
Antibiotic resistance arrow of time. The diagram depicts the proposed steps in the evolution of high-level antibiotic resistance in mycobacteria.
The two MAC efflux pump proteins whose genes were upregulated in the presence of azithromycin were examined for sequence conservation in other pathogenic mycobacteria. With regard to the MFS efflux pump encoded by MAV_1406, orthologs identified included MAP2516 of M. avium
, MTC Rv1258c, M. marinum
MMAR_4182, M. bovis
JTY_1291, and M. abscessus
MAB_1409c. The degrees of protein identity and the conserved amino acid motifs are shown in . There is remarkable conservation of protein secondary structure, which is closely linked to function. The MTC ortholog Rv1258c (0.73 identity with MAV_1406) is a well-known proton-dependent efflux pump whose substrates are macrolides and tetracyclines (1
). In regard to the second efflux pump, the ABC transporter encoded by MAV_3306, proteins with even higher identity were identified (all ≥0.90 identity). They were M. avium
MAP1198, M. marinum
MMAR_2279, M. ulcerans
MUL_1481, M. bovis
BCG BCG_1534, MTC Rv1473, and MLBr_01816 in the leprosy bacillus. Protein identity for the secondary structures is shown in . Given how well conserved these proteins are, we propose that they play roles in the other pathogenic mycobacteria similar to those in MAC.
Fig 8 Predicted protein secondary structure of the putative efflux pump encoded by MAV_3306 and its orthologs. Each red arrow represents an α-helix, while each blue arrow represents a β-strand. Highly conserved residues are shown in blue, while (more ...)
Fig 9 Predicted protein secondary structure of the putative efflux pump encoded by MAV_1406 and its orthologs. Each red arrow represents an α-helix, while each blue arrow represents a β-strand. Highly conserved residues are shown in blue, while (more ...)