ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Jan. 1995, p. 264–267 0066-4804/95/$04.0010 Copyright q 1995, American Society for Microbiology
Vol. 39, No. 1
Amikacin Levels in Bronchial Secretions of 10 Pneumonia Patients with Respiratory Support Treated Once Daily versus Twice Daily ´ ,1* H. GEORGES,1 J. M. JACQUIER,1 O. LEROY,1 C. BEUSCART,1 C. SANTRE D. BUGUIN,2 AND G. BEAUCAIRE1 Intensive Care and Infectious Diseases Unit1 and Department of Microbiology,2 Centre Hospitalier Tourcoing, Lille University Medicine School, F-59208 Tourcoing, France Received 22 March 1994/Returned for modification 23 June 1994/Accepted 24 October 1994
In this study, concentrations of amikacin in blood and bronchial secretions of 10 patients with mechanical ventilation for acute respiratory failure due to pneumonia were measured. One-half of the patients received amikacin twice daily, and the others received once-daily administration. Concentrations in bronchial secretions of the patients treated twice daily ranged from 3 to 4 mg/liter, i.e., they were similar to those in previously published reports. Peak concentrations in bronchial secretions occurred between 3 and 4 h after the onset of infusion, and they reached 4.8 6 2.6 mg/liter on day 1 and 4.0 6 2.7 mg/liter on day 3. For the patients treated with amikacin once daily, concentrations in bronchial secretions were more than twofold higher, above 8 mg/liter for 12 h. Peak concentrations in bronchial secretions occurred between 3 and 4 h after the onset of infusion and reached 13.6 6 9.3 mg/liter on day 1 and 10.4 6 3.5 mg/liter on day 3. These concentrations are higher than the MICs for less sensitive bacterial strains, such as Acinetobacter spp. and Pseudomonas aeruginosa. lowing conventional clinical, radiological, biological, and bacteriological diagnosis criteria were used to define bronchopneumonia: fever above 388C, a peripheral leucocyte count of .10,000/mm3, purulent endotracheal secretions, and pulmonary infiltrates on chest radiographs that were not otherwise explainable. Patients with previous renal insufficiency, defined as a serum creatinine concentration greater than 180 mmol/ liter, and those with a lack of bronchial secretions were excluded from the study. No patient had previously received amikacin or other aminoglycosides within 7 days of entry into the study. The patients were divided into two groups by using an open randomization schedule, and they received amikacin either once daily (od group) or twice daily (bid group). The patients in the once-daily treatment group received 15 mg of amikacin per kg of body weight every 24 h, and the twice-daily treatment group received 7.5 mg of amikacin per kg every 12 h. Both groups were administered amikacin as an intravenous infusion given over 30 min. The antibiotic treatment was performed in an empirical way, and amikacin was commonly administered with beta-lactams to achieve broad-spectrum coverage, including coverage for gram-negative strains, and highlevel bactericidal activity in mechanically ventilated patients with life-threatening disease. Bronchial secretions were collected by endotracheal suction in the usual manner (Vygon 534.16). No nebulization or antibacterial filters were used. Blood and bronchial secretion samples were collected at 0, 1, 3, 4, 6, 9, 12, 15, 18, and 24 h after the start of amikacin infusion on days 1 and 3. Samples were frozen at 2208C until completion of the study. Bronchial secretions were diluted in an equivalent sterile water volume and then centrifuged after incubation at 378C for 18 h to provide better viscosity. Cell smears were made to exclude hemorrhagic samples. Measurements of concentrations in bronchial secretions were performed on the supernatant. Amikacin concentrations in the serum and bronchial secretions were determined by a fluorescence polarization immunoassay (Abbott TDX). The assay sensitivity was 0.8 mg/liter, and the intrarun coefficient of variation was between 3 and 7% in the concen-
Aminoglycosides are widely used for the treatment of gramnegative bronchopneumonia. A critical factor in their effectiveness is their ability to achieve therapeutic concentrations at the site of infection. Although concentrations in bronchial secretions do not necessarily reflect parenchymal or bronchial wall concentrations, these measurements provide information on blood-bronchioalveolar barrier. Previous studies revealed a nonsaturable transport system for aminoglycosides which involves a passive mechanism (8). Thus, high serum antibiotic concentrations should theoretically lead to high concentrations in bronchial secretions. Endotracheal administration of aminoglycosides has been postulated to provide therapeutic concentrations in the treatment of gram-negative pneumonias, but to date no study has exhibited superiority of this administration method compared with usual regimens (1, 14). Numerous recent in vitro and clinical studies have shown that once-daily aminoglycoside administration is as effective as and less toxic than twice-daily administration in the case of severe gramnegative infections (12). A previous clinical study conducted with 91 patients hospitalized in an intensive care unit demonstrated that once-daily amikacin administration was highly effective against gram-negative infections, particularly in the case of 45 patients with bronchopneumonia (4). We proposed that these results could be related to higher in situ amikacin concentrations in the patients treated once daily. In this study, we performed serial measurements of concentrations in serum and bronchial secretions of patients treated with amikacin twice daily and once daily. From November 1990 to July 1991, 13 patients admitted to an intensive care unit for acute respiratory failure related to bronchopneumonia were enrolled in a prospective study. The protocol was approved by the Ethics Committee of the Lille University Medicine School. Informed consent was not deemed to be mandatory. All patients were mechanically ventilated through a nasotracheal tube upon admission. The fol-
* Corresponding author. Phone: (33) 20764697. Fax: (33) 20254291. 264
VOL. 39, 1995
NOTES
265
TABLE 1. Clinical data, amikacin treatment groups, and bacteriological outcomes for patients Group
Patient
Sexa
Age (yr)
Microorganism
Dosage (mg)
ATB 1b
ATB 2c
od
1 2 3 4 5
M F F F M
55 76 66 82 46
S. pneumoniae No pathogen No pathogen S. marcescens No pathogen
1,000 1,000 1,000 500 600
ATM CTX CTX IMP CAZ
VAN
bid
1 2 3 4 5
M F M M M
55 44 78 41 51
H. influenzae S. pneumoniae S. pneumoniae S. pneumoniae S. pneumoniae
500 500 400 500 500
CAZ CAZ CAZ CTX CTX
a b c
VAN
M, male. F, female. ATB 1, combined antimicrobial agent 1. ATM, aztreonam; CAZ, ceftazidime; CTX, cefotaxime; IMP, imipenem-cilastatin. ATB 2, combined antimicrobial agent 2. VAN, vancomycin.
tration range of 0 to 30 mg/liter. Results were expressed in milligrams per liter of serum or bronchial secretion. Samples of blood and bronchial secretions (obtained by bronchial aspiration or with a fiber-optic distal protected brush) were cultured under aerobic and anaerobic conditions. A microorganism was considered the etiologic agent of the pneumonia only if it was isolated from the lower respiratory secretions obtained from bronchial aspirate (105 CFU/ml) and/or a fiber-optic distal brush with a cutoff point of 103 CFU/ml or if it was isolated from blood. Pharmacokinetic analysis was performed with a pharmacokinetics software program (SIPHAR) to determine areas under the curve (AUCs). AUCs from time zero to the last measured concentration were computed by using a linear trapezoidal rule. Data were expressed as means 6 standard deviations. AUCs for each patient of the two groups were compared by means of the Wilcoxon rank-sign test. A P value of ,0.05 was considered statistically significant. A total of 3 patients, 2 with hemorrhagic bronchial secretions and 1 with renal insufficiency, were excluded before day 3 of the study; 10 patients completed the trial. General clinical data are included in Table 1. The mean patient age was 59.4 6 14.3 years (range, 41 to 82 years). Six male patients and four female patients were enrolled. Renal function was normal at
entry and remained normal during the entire study in every patient. Eight patients had community-acquired bronchopneumonia, and two had nosocomial infections. Seven patients had bacteriologically proven bronchopneumonia, and the isolated pathogens were Streptococcus pneumoniae for five patients, Haemophilus influenzae for one patient, and Serratia marcescens for one patient. The following antimicrobial agents were combined with amikacin: ceftazidime (six patients), cefotaxime (three patients), imipenem (one patient), aztreonam (one patient), and vancomycin (two patients). All patients were ultimately cured according to clinical definitions. Concentrations of amikacin in bronchial secretions of 10 patients could be effectively measured on day 1, and those in secretions of 8 patients could be effectively measured on day 3. Thirty bronchial samples could not be assayed or were excluded for technical reasons, and 160 concentrations in bronchial secretions (84.2%) were actually measured. Mean values for concentrations in serum and bronchial secretions are summarized in Table 2. For bid group patients on day 1, the mean peak serum amikacin concentration was 20.2 6 9.3 mg/liter (range, 10.3 to 36.4 mg/liter), and the mean peak concentration in bronchial secretions, occurring 3 h after the onset of infusion, was 4.8 6 2.6 mg/liter (range, 1.1 to 8.3 mg/liter). On day 3, the mean
TABLE 2. Mean amikacin concentrations, AUC values, and ratios of concentrations in bronchial secretions to concentrations in serum for the two treatment groups Resulta Day 1 Treatment group
Once daily Twice daily
Concn in bronchial secretions (mg liter21)
Day 3
Serum AUC0–24 (mg z h liter21)
Secretion AUC0–24 (mg z h liter21)
46.5 6 9.3 13.6 6 9.3
230 6 113d
153 6 99e
20.2 6 9.3
131 6 41
Peak concn in serum (mg liter21)
4.8 6 2.6
d
60 6 38
e
Peak concn ratio (%)b
Secretion AUC0–24 (mg z h liter21)
30.6 66 6 23 37.2 6 7.6 10.4 6 3.5
223 6 85f
151 6 85g
40.5 81 6 16
29.4 46 6 20 20.2 6 5.3
140 6 40
79 6 45
31.3 57 6 20
Except those for peak concentration ratios, data are means 6 standard deviations. Ratio of peak concentration in bronchial secretions to peak concentration in serum. c Ratio of bronchial secretion AUC to serum AUC. d P 5 0.21. e P 5 0.21. f P 5 0.18. g P 5 0.13. a b
Concn in bronchial secretions (mg liter21)
Serum AUC0–24 (mg z h liter21)
AUC ratio (%)c
Peak concn in serum (mg liter21)
4.0 6 2.7
f
g
Peak concn ratio (%)b
AUC ratio (%)c
266
NOTES
ANTIMICROB. AGENTS CHEMOTHER.
FIG. 1. Amikacin pharmacokinetics for the bid treatment group. Mean concentrations in serum (`) and in bronchial secretions (1) on day 1 (plus standard deviations) are shown.
FIG. 2. Amikacin pharmacokinetics for the od treatment group. Mean concentrations in serum (`) and in bronchial secretions (1) on day 1 (plus standard deviations) are shown.
peak concentrations were 20.2 6 5.3 mg/liter in serum and 4.0 6 2.7 mg/liter in bronchial secretions. Mean trough concentrations in serum were 0.8 6 0.7 mg/liter and 0.7 6 0.6 mg/liter on day 1 and day 3, respectively. Figure 1 shows that the mean concentrations in bronchial secretions remained constant (3 to 4 mg/liter) over the 24-h monitoring period and that they were permanently below 5 mg/liter. There was no cumulative effect between day 1 and day 3. Important individual variations in the bronchial peak concentrations were noted (range, 1.1 to 8.3 mg/liter). As shown in Table 2, the mean ratio of peak concentrations in bronchial secretions to mean peak concentrations in serum (BC max/SC max) was 29.4% on day 1 and 31.3% on day 3 (range, 12 to 51.5%). The mean bronchial secretion AUC from 0 to 24 h (AUC0–24) was 59.9 mg z h/liter on day 1 and 79.4 mg z h/liter on day 3. The mean ratio of bronchial secretion AUC0–24 to serum AUC0–24 was 46% on day 1 and 57% on day 3 (Table 2). For od patients on day 1, the mean peak concentration in serum was 46.5 6 9.3 mg/liter, and the mean peak concentration in bronchial secretions, occurring 3 h after the onset of infusion, was 13.6 6 9.3 mg/liter. On day 3, the corresponding results were 37.2 6 7.6 mg/liter and 10.4 6 3.5 mg/liter, respectively. Despite the high peak concentrations in serum, mean trough concentrations were low: 1.1 6 1.1 mg/liter on day 1 and 2.1 6 1.2 mg/liter on day 3. Mean concentrations in bronchial secretions were above 8 mg/liter for 12 and 16 h after the onset of infusion on days 1 and 3, respectively (Fig. 2). There was no cumulative effect, as was previously observed with the bid treatment group. Peak concentrations in serum ranged from 26.2 to 59.0 mg/liter, and peak concentrations in bronchial secretions ranged from 9.1 to 26.1 mg/liter. The BC max/SC max mean ratio was 30.6% on day 1 and 40.5% on day 3 (range, 13.3 to 46.6%). The mean bronchial secretion AUC0–24 was 153 mg z h/liter on day 1 and 181.1 mg z h/liter on day 3. The mean ratio of bronchial secretion AUC to serum AUC was 66% on day 1 and 81% on day 3. The serum amikacin concentrations noted for the patients in
this study were similar to those reported by other investigators. Since most gram-negative bacteria are inhibited by amikacin concentrations of 1 to 4 mg/liter, all the patients, particularly those in the od treatment group, exhibited a high serum amikacin concentration-to-MIC ratio. For the od treatment group, mean peak concentrations in bronchial secretions were more than twice those observed for the bid treatment group. These peak concentrations were obtained between 3 and 4 h after the start of infusion, and thus they occurred later than the peak concentrations in serum. For the conventional twice-daily administration group, mean concentrations in bronchial secretions approached 3 to 4 mg/ liter, and they were continuously below 5 mg/liter. These results for the bid treatment group are consistent with previous observations with amikacin (6, 8). The reported MICs of amikacin for Acinetobacter spp. and Pseudomonas aeruginosa are usually greater than 6 mg/liter. Thus, these low amikacin levels could be inadequate to cure some gram-negative bronchopneumonias. As previously reported in the literature (7), there was no cumulative effect between day 1 and day 3. In the case of one patient administered amikacin twice daily, concentrations in bronchial secretions obtained on day 6 confirmed these results. For the group administered amikacin once daily, mean concentrations in bronchial secretions were greater than 10 mg/ liter at 12 h after the last dose of amikacin, and they remained above 5 mg/liter for 16 h after the last dose. These concentrations would be expected to inhibit in situ most gram-negative organisms, including Acinetobacter spp. and P. aeruginosa. Garraffo et al. demonstrated, in a study involving six healthy volunteers, that the concentration-dependent bactericidal activity of amikacin may be improved by increasing in vivo concentrations (10). For the od treatment group, concentrations in bronchial secretions were about three- or fourfold higher than the MICs for most gram-negative strains. Regarding P. aeruginosa and Enterobacter cloacae, Garraffo et al. demonstrated that the length of time for which amikacin levels were main-
VOL. 39, 1995
tained above the MICs was an important parameter for bactericidal activity. In bronchial secretions from patients administered amikacin once daily, amikacin levels were above 8 mg/ liter for more than 12 h. Several clinical studies, some including critically ill patients, have shown that once-daily amikacin treatment can be as effective as the usual regimen (3, 11, 16). In a population of 45 patients with gram-negative bronchopneumonias, once-daily amikacin treatment could achieve a high clinical success rate, above 70%; moreover, Acinetobacter spp. and P. aeruginosa behaved in the same way as did other gram-negative strains (4). These high success rates with gram-negative bronchopneumonias, even when Acinetobacter spp. and P. aeruginosa were involved, could be related to the high concentrations in bronchial secretions of the patients treated once daily with amikacin. The ratio of concentrations in serum to concentrations in bronchial secretions (BC max/SC max) remained relatively constant at about 30%, and it involved a nonsaturable transport system that was a purely passive mechanism. These results are consistent with previous results obtained with twice-daily regimens (6, 8). In another study conducted with 12 tracheostomized patients, Even et al. noted bronchial secretion-toserum ratios between 15 and 30%, which also occurred in the case of patients with impaired renal function (9). Few investigators have used AUCs to evaluate blood-bronchioalveolar aminoglycoside transport in clinical situations. Klastersky studied the penetration of netilmicin into bronchial secretions after intermittent or continuous infusion in tracheostomized patients. There was no difference between the two regimens, and the AUC ratio was approximately 20% (15). As indicated in Table 2, the bronchial secretion AUC-to-serum AUC ratio appeared more important for the od treatment group than it did for the bid treatment group, but the results did not reach statistical significance. Individual variations in concentrations in bronchial secretions were very important. These variations may be explained by large interindividual variations in peak concentrations in serum. Many factors contribute to these variations in the case of critically ill patients, including hemodynamic instability, variations in the volume of distribution, and age (2, 13). Other patient-related factors may enhance or reduce bloodbronchioalveolar transport, such as pulmonary perfusion, local inflammation, paO2 and pH, protein binding, and concentrations of protein and cells in bronchial secretions (5, 17). Moreover, certain sampling method could affect the results, including intrabronchial hemorrhages due to excessive aspiration of samples. In summary, many published reports have documented the fact that conventional twice-daily amikacin administration provides inadequate bronchial concentrations to inhibit gram-negative strains of Acinetobacter spp. and P. aeruginosa for which MICs are above 6 mg/liter. Our study reveals that therapeutic concentrations in bronchial secretions are obtained with once-
NOTES
267
daily administration regimens. These results should provide a good rationale for once-daily administration of aminoglycosides in the treatment of severe gram-negative bronchopneumonia. Further clinical trials are needed to confirm this hypothesis. This study was supported in part by a grant from Bristol-Myers Squibb Company. We are indebted to Paul Santella for his helpful assistance. REFERENCES 1. Aguillera, D., L. Holzapfel, D. Carre`re-Debat, D. P. Giudicelli, P. Granier, D. Gontier, et al. 1988. Evaluation du traitement par aminoside intratrache´al des pneumopathies nosocomiales sous ventilation me´canique. Reanim. Soins Intens. Med. Urg. 4:3–7. 2. Beaucaire, G., C. Beuscart, S. Dalmas, Kerreneur, O. Leroy, P. Lestavel, S. Moulront, C. Santre´, and C. Socolowski. 1991. Amikacine en dose unique journalie`re. Efficacite ´ et aspects pratiques. A propos de 91 patients de re´animation. Reanim. Soins Intens. Med. Urg. 7:352–358. 3. Beaucaire, G., O. Leroy, C. Beuscart, P. Karp, C. Chidiac, M. Caillaux, and the Study Group. 1991. Clinical and bacteriological efficacy, and practical aspects of amikacin given once daily for severe infections. J. Antimicrob. Chemother. 27(Suppl. C):91–103. 4. Beaucaire, G., O. Leroy, C. Beuscart, P. Lestavel, M. Caillaux, and C. Chidiac. 1990. Once-daily dosing amikacin (15 mg/kg) for treatment of severe pneumonias in 45 patients: preliminary results. Am. Rev. Respir. Dis. 141:PA 600. 5. Bergan, T., A. Engeset, and W. Olszewski. 1987. Does serum protein binding inhibit tissue penetration of antibiotics? Rev. Infect. Dis. 9:713–718. 6. Bergogne-Berezin, E. 1981. Penetration of antibiotics into respiratory tree. J. Antimicrob. Chemother. 8:171–174. 7. Bergogne-Berezin, E., P. Even, and G. Berthelot. 1978. Pharmacocine´tique de l’amikacine dans les se´cre ´tions bronchiques. Rev. Fr. Mal. Respir. 6:385– 392. 8. Dull, W. L., M. R. Alexander, and J. E. Kasik. 1979. Bronchial secretion levels of amikacin. Antimicrob. Agents Chemother. 16:767–771. 9. Even, P., E. Bergogne-Berezin, P. Reynaud, and G. Berthelot. 1979. Pharmacocine´tique de l’amikacine dans les se ´cre´tions bronchiques. Nouv. Presse Med. 8:3441–3444. 10. Garaffo, R., H. B. Drugeon, P. Dellamonica, E. Bernard, and P. Lapalus. 1990. Determination of optimal dosage regimen for amikacin in healthy volunteers by study of pharmacokinetics and bactericidal activity. Antimicrob. Agents Chemother. 34:614–621. 11. Giamarellou, H., K. Viallouros, G. Pettrikos, E. Moschovakis, E. Vavouraki, and D. Voutsinas, et al. 1991. Comparative kinetics and efficacy of amikacin administered once or twice daily in the treatment of system gram-negative infections. J. Antimicrob. Chemother. 27(Suppl. C):73–79. 12. Gilbert, D. N. 1991. Once-daily aminoglycoside therapy. Antimicrob. Agents Chemother. 35:399–405. 13. Hassan, E., and J. Ober. 1986. Predicted and measured aminoglycoside pharmacokinetic parameters in critically ill patient. Crit. Care Med. 14:394– 398. 14. Klastersky, J., F. Carpentier-Meunier, L. Kahan-Coppens, and J. P. Thys. 1979. Endotracheally administered antibiotics for gram-negative bronchopneumonia. Chest 75:586–591. 15. Klastersky, J. E., J. P. Thys, and G. Mombelli. 1981. Comparative studies of intermittent and continuous administration of aminoglycosides in the treatment of broncho-pulmonary infections due to gram-negative bacteria. Rev. Infect. Dis. 3:74–83. 16. Maller, R., B. Isaksson, I. Nilsson, and L. Soren. 1988. A study of amikacin given once versus twice daily in serious infections. J. Antimicrob. Chemother. 22:75–79. 17. Pennington, J. E. 1981. Penetration of antibiotics into respiratory secretions. Rev. Infect. Dis. 3:67–73.
