Dynamics of Clarithromycin and Azithromycin Efficacies against Experimental Haemophilus influenzae Pulmonary Infection

Abstract

The dynamics of clarithromycin and azithromycin efficacy against pulmonary Haemophilus influenzae infection in rats were evaluated. Efficacy was measured by reduction in pulmonary H. influenzae burden on days 3 and 7 postinoculation. Clarithromycin therapy was effective on day 3 or 7 of therapy, while azithromycin was effective on day 7 but not on day 3 of therapy. Both macrolides produced marked efficacy against all six strains of H. influenzae tested, including four strains for which MICs were above the susceptible breakpoint (8 μg/ml) concentration of clarithromycin. The two macrolides demonstrated markedly different pharmacokinetic characteristics, with clarithromycin present in both blood and tissue, while azithromycin was concentrated primarily in tissue. During pulmonary infection in rats, H. influenzae was found in both intracellular locations and an extracellular location in the lung. Blood concentrations of clarithromycin and azithromycin approximated human pharmacokinetics, and the blood concentrations for either macrolide rarely exceeded MICs for H. influenzae. At dosages producing blood concentrations similar to values achieved clinically, clarithromycin produced efficacy on day 3 of therapy, while both clarithromycin and azithromycin were equally effective on day 7. The different dynamics of clarithromycin and azithromycin suggest that length of therapy should be considered as a key parameter in evaluations of drug efficacy.

The macrolides clarithromycin and azithromycin exhibit markedly different pharmacokinetics. In clinical studies, clarithromycin achieved high concentrations in both tissue and plasma, with the highest levels in alveolar macrophages and epithelial lining fluid, while concentrations in plasma were maintained (7, 18). In contrast, azithromycin achieved very high concentrations in alveolar macrophages and low or undetectable levels in epithelial lining fluid and plasma (5, 7). Haemophilus influenzae colonizes and multiplies in both intracellular and extracellular locations, presenting both components to an infection (12).

The macrolide antibiotics clarithromycin and azithromycin are commonly used for the treatment of respiratory tract infections. Clarithromycin and azithromycin are active against respiratory pathogens such as H. influenzae, Streptococcus pneumoniae, Mycoplasma pneumoniae, and Legionella species (3, 22, 23, 25). Clarithromycin and azithromycin MICs for H. influenzae strains vary, but H. influenzae strains present as a unimodal population to both drugs (9). The serum concentrations of clarithromycin and azithromycin are often less than the MIC for H. influenzae. Clarithromycin and azithromycin demonstrate good clinical efficacy against H. influenzae and other pulmonary pathogens, due in part to the high tissue concentrations of the macrolides (2, 8). There have been no clinical studies correlating efficacy of clarithromycin or azithromycin to the MICs for H. influenzae in pulmonary infection.

The pharmacology and efficacies of clarithromycin and azithromycin at clinical dosages against a pulmonary H. influenzae infection in rats were investigated. Different schedules and dosages of clarithromycin and azithromycin were used to produce serum concentrations similar to those achieved clinically. The efficacy of clarithromycin and azithromycin was determined by the reduction of H. influenzae in lung tissue. Efficacy was determined at an interim period (day 3) and at the conclusion of therapy (day 7). The goal was to determine the dynamics of the relationship between macrolide efficacy, tissue concentration, and MICs for H. influenzae.

MATERIALS AND METHODS

Bacteria.

H. influenzae clinical isolates were from the Abbott Laboratories culture collection. The cultures were maintained at 37°C and 5% CO₂ in Haemophilus test medium. Following overnight incubation, cultures were diluted to produce an inoculum of 10⁶ to 10⁷ CFU per animal.

MIC determinations.

MICs for H. influenzae isolates were determined by broth microdilution by methods described by the National Committee for Clinical Laboratory Standards (NCCLS) (20). The tests were done in Haemophilus test medium incubated at 36°C in ambient air for 24 h, except for the test with strain 3643, which required 5% CO₂ for growth. MICs were determined visually.

Pharmacokinetics.

Single and multiple oral-dose pharmacokinetic trials were conducted in 225- to 250-g Sprague-Dawley rats. In the multiple-dose trials, clarithromycin was administered at 12-h intervals, with a total of five dosages. Azithromycin was administered every 24 h, with a total of three dosages; the first dose was administered at double strength. The total number of dosages administered in the pharmacokinetic trial represented the midpoint of the efficacy trials. The dosages of each macrolide were selected to simulate human dosages and pharmacokinetics. Plasma and lung samples were collected 0.5, 1, 2, 3, 6, 12, and 24 h after the last dose. Three rats were assayed at each time point. The entire lung tissues were removed, rinsed free of blood with 0.9% NaCl, and homogenized thoroughly.

Drug concentrations in lung homogenate and plasma were determined by bioassay with Micrococcus luteus as the detecting organism as reported previously (10). M. luteus ATCC 9341 was used as the assay organism. Twenty milliliters of this standardized inoculum containing 4.9 × 10¹⁰ CFU was used to inoculate a liter of antibiotic medium no. 11 (Difco Laboratories, Detroit, Mich.). One hundred milliliters of seeded agar was poured into each 243 by 243 by 18 mm plastic dish (Nunc, Roskilde, Denmark). Thirty-six discs were placed on each plate. Each standard solution of spiked plasma or lung homogenate was tested in duplicate on each plate, and a disc for each sample was placed on three plates. The daily results for the test samples were considered valid only if the calculated results for the quality control samples were within 13% of the absolute concentration. The plates were incubated at 30°C for 18 h.

Zones of inhibition were measured with an image analyzer (Optomax). The mean of the duplicate zones of inhibition for each standard on each plate was used to calculate the standard curve. The standard curve was calculated by linear regression techniques and the following equation: y = a + bx, where y is the zone of inhibition and x is the log of the concentration of a standard. The potency of a sample was determined by substituting the zone of inhibition for y and solving for x, the log of the concentration of the unknown.

Drug efficacy trials.

Sevoflurane (Abbott Labs, North Chicago, Ill.) anesthesia was used while the rat was intubated with an 18-gauge stainless steel gavage needle. Rats were inoculated intratracheally with 0.1 ml containing 10⁶ to 10⁷ CFU of H. influenzae in 2% molten agarose cooled to approximately 37 to 40°C. In treatment groups, oral clarithromycin and azithromycin therapy was initiated on day 0, with the initial dose administered 5 h postinfection. Clarithromycin was then administered twice daily (b.i.d.) for days 0 to 6. Azithromycin was administered once daily (QD), at double-strength dosage on day 0 and at single-strength dosage on days 1 to 6. Dosages were selected to model clinical pharmacokinetics. A 2-log reduction (99%) in bacterial burden was selected as a standard for clinical response. This standard was based on observations that correlated improvement in clinical appearance of the rats with a >2-log reduction in bacterial burden. There were 10 rats in each treatment group.

Cytology.

For determination of H. influenzae location in pulmonary tissue, rats were infected as described above. At day 3, lung tissue was harvested, and samples of right and left lungs were used for cytological evaluation. Lung was sectioned transversely, and imprints of the cut surface were made on glass microscope slides. Slides were allowed to air dry and then were stained with Wright’s-Giemsa and examined microscopically.

Determination of drug efficacy.

Drug efficacy was determined by reduction in pulmonary bacterial burden. Lungs were harvested 12 to 24 h after the last dosage on day 3 or 7 postinfection. The lung tissue was homogenized in a sterile tissue homogenizer (Tekmar, Cincinnati, Ohio) and then diluted in sterile water. Bacterial burden was determined by dilution plating of the lung homogenates on chocolate agar, which was incubated at 37°C in 5% CO₂ for 24 h to allow visualization of H. influenzae colonies. The dilution factor in water (1/50 to 1/500,000) ensured that residual drug was present only in sub-MICs. Inhibitory activity of residual clarithromycin or azithromycin at subinhibitory concentrations cannot be ruled out but appears unlikely. Individuals with a reduction of greater than 2 logs (99%) of H. influenzae burden were designated as responders to specify a more rigorous biologic standard than simple statistical significance.

Statistics.

For the drug efficacy trials presented in Tables 2 and 3, statistical significance between mean bacterial count values was determined by Kruskall-Wallis test. Pairwise comparisons were performed by the Wilcoxon rank sum test when the preliminary test indicated significant differences at the 0.05 level (16).

TABLE 2.

Log burden of H. influenzae in rat lung at day 3 of therapy with clarithromycin or azithromycin

Drug/dosagea	H. influenzae 43095b (% respondersc)	H. influenzae 1435b (% respondersc)	H. influenzae 3556b (% respondersc)	H. influenzae 3598b (% respondersc)	H. influenzae 3558b (% respondersc)
Clarithromycin	MIC = 4	MIC = 8	MIC = 16	MIC = 16	MIC = 16
150 mg/kg	0.55 ± 1.10d (100%)	0.91 ± 1.43d (100%)	1.26 ± 1.88d (90%)	0.90 ± 1.56d (100%)	NDe (100%)
100 mg/kg	1.99 ± 1.75d (90%)	2.58 ± 1.64d (90%)	3.44 ± 2.20d (78%)	0.58 ± 1.15d (100%)	1.03 ± 2.17 (80%)
Azithromycin	MIC = 2	MIC = 2	MIC = 4	MIC = 4	MIC = 4
50/25 mg/kg	4.29 ± 1.88d (56%)	2.10 ± 1.99d (90%)	5.89 ± 1.01 (11%)	4.72 ± 0.63 (33%)	3.82 ± 1.70 (60%)
25/12.5 mg/kg	5.70 ± 1.42 (13%)	3.89 ± 1.73d (60%)	6.36 ± 1.03 (11%)	5.59 ± 0.79 (17%)	5.10 ± 0.49 (0%)
Untreated	6.10 ± 0.79	6.22 ± 0.68	6.95 ± 0.63	6.66 ± 0.86	6.33 ± 0.73

^aClarithromycin dosage, b.i.d. on days 0 to 2; azithromycin dosage, QD on days 1 to 2; 2× strength dosage on day 0. 

^bMean log count ± standard deviation. 

^cPercent responders, percentage of rats with >2-log reduction in H. influenzae burden. 

^dSignificantly lower than untreated controls (P <0.01). 

^eND, none detected. 

TABLE 3.

Log burden of H. influenzae in rat lung at day 7 of therapy with clarithromycin or azithromycin

Drug/dosagea	H. influenzae 43095b (% respondersc)	H. influenzae 1435b (% respondersc)	H. influenzae 3556b (% respondersc)	H. influenzae 3643b (% respondersc)	H. influenzae 3558b (% respondersc)
Clarithromycin	MIC = 4	MIC = 8	MIC = 16	MIC = 32	MIC = 16
150 mg/kg	0.54 ± 0.36d (100%)	0.38 ± 0.38d (100%)	NDd,e (100%)	NDd,e (100%)	NDd,e (100%)
100 mg/kg	1.20 ± 0.50d (100%)	0.57 ± 0.38d (100%)	0.57 ± 1.20d (100%)	NDd,e (100%)	1.00 ± 0.00d (100%)
Azithromycin	MIC = 2	MIC = 2	MIC = 4	MIC = 4	MIC = 4
50/25 mg/kg	0.40 ± 0.40d (100)	0.59 ± 0.59d (100%)	NDd,e (100%)	NDd,e (100%)	NDd,e (100%)
25/12.5 mg/kg	0.96 ± 0.53d (100%)	6.61 ± 0.53 (20%)	4.53 ± 2.78 (40%)	NDd,e (100%)	0.60 ± 0.40d (100%)
Untreated	6.58 ± 0.20	7.63 ± 0.26	6.73 ± 0.32	6.57 ± 0.19	7.23 ± 0.15

^aClarithromycin dosage, b.i.d. on days 0 to 6; azithromycin dosage QD on days 1 to 6; 2× strength dosage on day 0. 

^bMean log count ± standard deviation. 

^cPercent responders, percentage of rats with >2-log reduction in H. influenzae burden. 

^dSignificantly lower than untreated controls (P < 0.01). 

^eND, none detected. 

RESULTS

Potency results.

Susceptibility testing was performed by NCCLS methods, and the interpretive criteria contained in NCCLS document M7-A4 were applied. Strains 43095 (MIC = 4) and 1435 (MIC = 8) were determined to be clarithromycin susceptible, strains 3556, 3598, and 3558 (MIC = 16) were clarithromycin intermediate, and strain 3643 (MIC = 32) was determined to be clarithromycin resistant. All six strains (MIC = 2 or 4) were azithromycin susceptible (Tables 2 and 3).

Time course of infection.

In a time course trial, H. influenzae 43095 produced a pulmonary infection of rats that was maintained for at least 7 days (data not shown). An inoculum of 6.54 logs of H. influenzae delivered by intratracheal tube yielded a mean burden of 6.95 ± 0.28 (mean ± standard deviation) logs of H. influenzae per rat lungs at 1 h postinfection and 7.12 ± 0.37 logs at day 1, 6.91 ± 0.21 logs at day 3, and 5.77 ± 0.65 logs at day 7. There was no mortality during the 7 days of infection (data not shown).

Imprints of lung from rats infected with H. influenzae were highly cellular and composed of 50% neutrophils, 40% monocytes and alveolar macrophages, and 10% lymphocytes and plasma cells. H. influenzae was located both extracellularly and intracellularly within phagosomes of neutrophils and macrophages (data not shown), similar to previously reported observations (12).

Single-dose pharmacokinetics.

Single oral dosages of 150 mg of clarithromycin per kg of body weight produced a maximum concentration of drug in serum (C_max) of 4.99 μg/ml and an area under the concentration-time curve (AUC) of 32.5 μg × h/ml (Table 1). Single oral dosages of 100 mg of clarithromycin/kg produced a plasma C_max of 1.98 μg/ml and an AUC of 18.96 μg × h/ml. Lung concentrations of clarithro- mycin were approximately 30-fold greater than plasma concentrations based on AUC_0–24 values. Plasma concentrations of clarithromycin exceeded the MIC for only one isolate, H. influenzae 43095 (MIC = 4), and not for strains for which MICs were 8, 16, or 32 μg/ml.

TABLE 1.

Plasma and lung concentrations of clarithromycin and azithromycin following oral dosing

Drug and dosea	Plasma concn			Lung concn
	Cmax (μg/ml)	Cmin (μg/ml)b	AUC0–24 (μg · h/ml)	Cmax (μg/g)	Cmin (μg/g)	AUC0–24 (μg · h/ml)
Clarithromycin
Rat
150 mg/kg	4.99	0.49	32.54	87.8	23.1	913.0
100 mg/kg	1.98	0.03	18.96	43.4	3.8	646.8
Human (26)
500 mg	2.4		18.9c	NDd	ND	ND
Azithromycin
Rat
50 mg/kg	0.90	0.05	6.80	33.7	5.1	563.7
25 mg/kg	0.74	0.06	6.50	22.7	1.4	397.6
12.5 mg/kg	0.15	<0.05	1.41	6.76	3.3	116.9
Human (26)
500 mg	0.4		3.4	ND	ND	ND

^aClarithromycin dosage, b.i.d. in rat and human; azithromycin dosage, QD in rat and human. 

^bC_min values obtained at 12 h for clarithromycin and 24 h for azithromycin. 

^cThe AUC of the active 14-hydroxy metabolite of clarithromycin was 6.0 μg × h/ml in addition to the 18.9 μg × h/ml concentration of the parent. 

^dND, not determined. 

Single oral dosages of azithromycin at 50 mg/kg produced plasma C_max values of 0.90 μg/ml and an AUC value of 6.8 μg × h/ml. Single oral dosages of azithromycin at 25 or 12.5 mg/kg yielded plasma C_max values of 0.74 and 0.15 μg/ml and AUC values of 6.50 and 1.41 μg × h/ml. The lung AUC_0–24 values of azithromycin were approximately 60- to 80-fold higher than the plasma concentrations and approximately twofold lower than clarithromycin concentrations in the lungs. The concentrations of azithromycin in plasma did not exceed the MIC for any strain of H. influenzae.

Multiple-dose pharmacokinetics.

Multiple oral dosages (every 12 h [q12h] × 5) of 150 mg of clarithromycin/kg produced a plasma C_max of 7.0 μg/ml and an AUC of 152.7 μg × h/ml (Fig. 1A). Multiple clarithromycin dosages of 100 mg/kg produced C_max and AUC values of 5.5 μg/ml and 86.6 μg × h/ml. Lung concentrations of clarithromycin were substantially greater than plasma concentrations at each dosage. The concentrations of clarithromycin in lung tissue exceeded H. influenzae MICs for the entire dosing interval (12 h) following either the 100 or 150 mg/kg dose (Fig. 1A). Concentrations of clarithromycin in plasma exceeded the MIC only for H. influenzae 43095 (MIC = 4 μg/ml) and not those of other strains for which MICs were 8, 16, or 32 μg/ml. Multiple oral dosages produced higher concentrations of plasma and lung than did single dosages in the rat. This observation is similar to clinical observations (11).

FIG. 1.

(A) Plasma and lung concentrations of clarithromycin in rats following multiple oral dosages at 150 or 100 mg/kg, b.i.d. × 5. Clarithromycin was administered every 12 h, with a total of five dosages over 48 h. The MIC at which 90% of the isolates are inhibited (MIC₉₀) for H. influenzae (16 μg/ml) is indicated in relation to concentrations in plasma and lungs. (B) Concentrations of azithromycin in plasma and lungs in rats following multiple oral dosages at 50 mg/kg on day 0 and 25 mg/kg on days 1 and 2 or 25 mg/kg on day 0 and 12.5 mg/kg on days 1 and 2. Azithromycin was administered every 24 h, with a total of three dosages over 48 h. The MIC₉₀ for H. influenzae (4 μg/ml) is indicated in relation to concentrations in plasma and lungs.

Oral dosages of azithromycin at 50 mg/kg on day 0 and 25 mg/kg on days 1 and 2 yielded plasma C_max and AUC values of 1.12 and 13.9 μg × h/ml, respectively (Fig. 1B). Azithromycin at 25 mg/kg on day 0 and 12.5 mg/kg on days 1 to 2 produced C_max and AUC values of 0.44 μg/ml and 4.88 μg × h/ml, respectively. Multiple dosages of azithromycin also produced higher concentrations in plasma and tissue than did single oral dosages (13). Concentrations of azithromycin in the lungs were also markedly higher than those in plasma. For both dosages of azithromycin, concentrations in the lungs exceeded the MICs for H. influenzae for the entire 24-h dosage interval (Fig. 1B). Similar to clarithromycin, the azithromycin concentrations in plasma exceeded only the MIC for H. influenzae 43095 (MIC = 1 μg/ml) for a brief interval and did not exceed MICs of 2 or 4 μg/ml for isolates of H. influenzae.

Efficacy testing results at day 3 of therapy.

Similar levels of infection were achieved by the different H. influenzae isolates as shown by the log CFU counts recovered from lungs of untreated animals. Clarithromycin therapy at 150 mg/kg b.i.d. for 3 days yielded significant efficacy, with 5.7- to 6.3-log reductions in H. influenzae lung burden (90 to 100% response rate) against five different isolates (Table 2). At 100 mg/kg b.i.d., clarithromycin yielded 3.5- to 6.1-log reductions in H. influenzae lung burden (78 to 100% response rate). Clarithromycin produced equally good efficacy against H. influenzae isolates for which MICs were 4 to 16. For example, at 100 mg/kg b.i.d., clarithromycin produced a 5.3-log reduction (80% response rate) against H. influenzae 3558 (MIC = 16) and a 4.1-log reduction (90% response rate) against H. influenzae 43095 (MIC = 4).

The efficacy of azithromycin was variable at day 3. At 25 mg/kg QD (50 mg/kg QD day 0), azithromycin yielded a 1- to 4-log reduction in H. influenzae lung burden (11 to 90% response rate) against five different isolates. At 12.5 mg/kg QD (25 mg/kg day 0), azithromycin yielded a 0.4- to 2.3-log reduction in H. influenzae lung burden (0 to 60% response rate). The efficacy of azithromycin was also unrelated to H. influenzae MIC. At 25 mg/kg, azithromycin produced a 2.5-log reduction (60% response rate) against H. influenzae 3558 (MIC = 4) and only a 1-log reduction (11% response rate) against H. influenzae 3556 (MIC = 2).

Efficacy testing results at day 7 of therapy.

For azithromycin, there was a general trend in improvement of efficacy, with 6 days of dosage compared to 3 days of dosage (Table 3 and Fig. 2). Six days of azithromycin at 50 and 25 mg/kg produced a 6- to 7-log reduction of H. influenzae lung burden, compared to a 1- to 4-log reduction produced by 3 days of therapy against five different isolates of H. influenzae. For example, the 50 and 25 mg/kg dosage showed an improvement from a 1.81-log reduction with 3 days of therapy to a 6.18-log reduction with 6 days of therapy versus therapy with H. influenzae 43095. Against H. influenzae 3556, the 50 and 25 mg/kg/day dosage produced a 1.06-log reduction on day 3 and a 6.73-log reduction on day 7. The only azithromycin therapy groups which did not show an improvement in efficacy with 6 days compared to 3 days of treatment were the 25 and 12.5 mg/kg dosage groups versus H. influenzae 1435 and 3556, which still yielded 1.02- and 2.2-log reductions in H. influenzae lung burden (less than 50% responders). The lack of efficacy of azithromycin 25 and 12.5 mg/kg dosage was not related to the MIC for the H. influenzae, as greater efficacy was obtained against H. influenzae 3643 and 3558 (MIC = 4), than against H. influenzae 1435 or 3556 (MIC = 2). Three days of clarithromycin therapy produced 3.51- to 6.06-log reductions in H. influenzae lung burden (response rates of 80 to 100%); 6 days of therapy produced 6.23- to 7.23-log reductions (100% response rate) in both the 150 and 100 mg/kg groups.

FIG. 2.

(A) Reductions of H. influenzae correlated to AUC/MIC ratios at day 3 of therapy with clarithromycin or azithromycin. (B) Reductions of H. influenzae correlated to AUC/MIC ratios at day 7 of therapy with clarithromycin or azithromycin.

DISCUSSION

Clarithromycin and azithromycin are macrolide antibiotics with contrasting pharmacokinetic profiles. Clarithromycin achieves higher overall concentrations in tissues and phagocytic cells and in extracellular locations such as epithelial lining fluid and plasma (7, 19). Azithromycin achieves higher tissue and plasma concentration ratios and has a longer half-life but sustains low to undetectable levels in extracellular locations such as epithelial lining fluid and plasma (4, 13). The contrasting pharmacokinetics of clarithromycin and azithromycin were the basis for the study of the dynamics of efficacy against H. influenzae, which has both an intracellular and an extracellular component in pulmonary infection.

The rat model of H. influenzae lung infection is a reasonable simulation of clinical infection. The infection is maintained for greater than 7 days, allowing for prolonged therapy. There is usually no death in untreated controls, which improves analysis of drug efficacy. Unlike humans, rodents do not produce the active 14-hydroxy metabolite of clarithromycin, which is a factor in clinical efficacy versus H. influenzae. The 14-hydroxy metabolite of clarithromycin is twice as potent as the parent compound versus H. influenzae and may account for up to 50% of the efficacy (15). Thus, the rodent model may underestimate clarithromycin efficacy against H. influenzae infection.

The difference in the dynamics of the response of H. influenzae lung infection to therapy with clarithromycin and azithromycin was striking (Fig. 2). Clarithromycin therapy produced high response rates and log reductions of total H. influenzae bacterial burden at both day 3 and day 7 of therapy. Azithromycin therapy produced high response rates and log reductions of H. influenzae burden only at day 7 of treatment. The concentrations of clarithromycin and azithromycin achieved in the blood of rats were equal or greater than the clinical levels achieved in humans, and the overall lung concentrations of clarithromycin and azithromycin were well in excess of the MICs for H. influenzae. Therefore, the rats were not underdosed. The different dynamics of the efficacies for clarithromycin and azithromycin against H. influenzae may be due to the different pharmacokinetics of the drugs.

The concentration of antibiotic at the site of infection is a key consideration in the determination of therapeutic efficacy (14). The concentration of an antibiotic varies according to the pharmacokinetics of the drug in the specific intracellular and extracellular compartments at the infected site. H. influenzae is both an intracellular and an extracellular pathogen, and drug efficacy could be affected by the different locations of the bacteria. In clinical studies, clarithromycin exhibited a balanced partitioning between tissue and plasma, with high concentrations in alveolar macrophages, epithelial lining fluid, and plasma (7, 18). Concentration of clarithromycin at both intracellular and extracellular locations could have caused the rapid therapeutic efficacy demonstrated at day 3 postinfection. In clinical studies, azithromycin was shown to achieve high concentrations in alveolar macrophages but to have low to undetectable levels in epithelial lining fluid or plasma (5, 7). While it was not possible to differentiate drug efficacy against intracellular versus extracellular H. influenzae, low extracellular concentrations of azithromycin may account for the delay to demonstrate efficacy until day 7. An in vitro study by Scaglione et al. (24) demonstrated both intracellular and extracellular killing of bacteria by clarithromycin but only intracellular killing by azithromycin following exposure of cells and bacteria to physiological concentrations of drug. The different dynamics of clarithromycin and azithromycin efficacies against H. influenzae may be due to the different intracellular and extracellular partitioning of the two drugs.

Analysis of drug efficacy should take into account the location of infection and the required length of treatment (1). Macrolides are bacteriostatic drugs which generally do not exhibit concentration-dependent killing in vitro. For drugs of this type, the AUC/MIC ratio or the time of dosing interval for which drug concentration exceeds the MIC (t > MIC) is considered critical for efficacy (6, 10). These analyses are usually based on concentrations of drug in the blood. However, for H. influenzae infection, the maximal concentrations of clarithromycin or azithromycin in the blood generally do not exceed the MIC for H. influenzae. Tissue concentrations of antibiotics in micrograms per gram of tissue are often related to MICs in micrograms per milliliter of broth for bacteria. This represents a simplistic view of drug dynamics, since the potency of a drug in a gram of tissue in the body does not necessarily correlate to potency in a milliliter of broth in the incubator. Significant proportions of drug in tissue may be bound or inactive. Thus, correlations of drug concentrations in tissue (micrograms/gram) to MICs (micrograms per milliliter) must be made with caution.

The two macrolides demonstrated different dynamics of efficacy, with time of therapy as the key variable (Fig. 2). Length of therapy is an important but complex component of efficacy. During multiple-dose therapy, drug concentrations approach steady-state concentrations at different rates, depending on half-life. Different partitioning of drug into tissue will also potentially influence efficacy. Azithromycin efficacy improved dramatically from day 3 to day 7 of therapy, while clarithromycin efficacy was close to maximal by day 3. Since these two macrolides have identical mechanisms of action, the factors influencing the dynamics of efficacy are likely due to pharmacodynamic differences.

At dosages producing clinical blood concentrations, clarithromycin demonstrated efficacy against all H. influenzae strains, regardless of MIC. The efficacy of clarithromycin in this rat model against H. influenzae strains for which MICs were 16 and 32 suggests that the current NCCLS breakpoint of 8 μg/ml does not correlate to therapeutic efficacy. When a population analysis is performed with either clarithromycin or azithromycin MICs for H. influenzae, a unimodal population is found (17). No mechanistically based resistance for H. influenzae against macrolides is known. The NCCLS susceptibility breakpoint for clarithromycin (8 μg/ml) arbitrarily bisects the H. influenzae population at approximately 92.5% of the isolates. In contrast, disk diffusion tests of clarithromycin applying the NCCLS breakpoint of >13 mm (18) accurately identify a unimodal susceptible population (21). Better agreement between efficacy results and the two NCCLS methods would be achieved with a clarithromycin breakpoint of 16 μg/ml.

In summary, the dynamics of clarithromycin and azithromycin efficacies against H. influenzae pulmonary infection were demonstrated to be different. Clarithromycin therapy produced efficacy at both day 3 and day 7 of a treatment course, while azithromycin therapy produced reliable efficacy at day 7. The different dynamics of clarithromycin and azithromycin efficacies could be due to the intracellular and extracellular partitioning of the drugs relative to pathogen location. Clarithromycin concentrates in both intracellular and extracellular locations, which could lead to a rapid response against H. influenzae in all pulmonary locations. The extracellular concentrations of azithromycin are typically low, which could account for the delayed response observed due to a potential need for drug to diffuse out of cells to extracellular locations. The macrolides produced reductions in bacterial burden against H. influenzae strains independent of MIC. The dynamic analysis of clarithromycin and azithromycin efficacies at days 3 and 7 of therapy suggests that length of therapy should be considered as a key parameter in evaluations of drug efficacy.