Pulsatile Delivery of Clarithromycin Alone or in Combination with Amoxicillin against Streptococcus pneumoniae

Abstract

We evaluated pulsatile dosing of clarithromycin and amoxicillin alone or combined against Streptococcus pneumoniae with various susceptibilities. When combined, pulsatile amoxicillin with clarithromycin was superior to either 8- or 12-h dosing against the intermediate strain and was identical for the susceptible strain. Pulse dosing of antimicrobials warrants further investigation.

Pulsatile dosing is a new technology allowing drugs to be released from an oral formulation in a well-controlled fashion, resulting in specific concentrations of drug exposure early in the dosing interval followed by a dose-free period. Benefits of this may allow for the ease of medication delivery to patients as a once-daily oral administration. Recently, preliminary data from two laboratories have demonstrated that amoxicillin delivered as a pulsatile dose, four fractionated doses delivered over 6 h, is equal or superior to the equivalent amount of drug administered twice or thrice daily in vitro against amoxicillin- susceptible and intermediate-susceptible Streptococcus pneumoniae (3). The purpose of this study is to evaluate the pharmacodynamics and the effects of resistance development of clarithromycin in combination with amoxicillin delivered as pulsatile doses compared to an equivalent total dose delivered traditionally every 8 or 12 h against S. pneumoniae with various susceptibility profiles.

Two S. pneumoniae strains, American Type Culture Collection reference strain 49150 and clinical mefA-positive isolate 16891 (obtained from D. E. Low, Toronto, Canada), were evaluated. Clarithromycin analytical powder (Plantex Corporation, Hackensack, NJ) and amoxicillin analytical powder (Sigma, St. Louis, MO) were used throughout. Fresh stock solutions of drug were prepared on the first day of the experiment and stored at 2 to 8°C between dosage administration times.

Susceptibility testing was performed on both S. pneumoniae test strains using methods as outlined by CLSI (formerly NCCLS) guidelines modified to simulate experimental model conditions (6). All cultures were plated onto tryptic soy agar with 5% sheep blood and incubated at 37°C for 24 h in the presence of approximately 3% CO₂.

A previously described 250-ml one-compartment in vitro model was used to simulate amoxicillin (half-life [t_1/2] = 1 h) and clarithromycin (t_1/2 = 6 h) pharmacokinetics over the 72-h simulation at a starting inoculum of ∼1 × 10⁶ CFU/ml (1-3, 5).

Regimens evaluated include the total daily dose divided evenly and given as a pulsatile regimen (4 divided doses administered over the first 6 h) as well as every-8- and every-12-h dosings. The total daily dose of clarithromycin was 15, 7.5, or 3.75 μg/ml/day, with the least effective dose to be utilized. Amoxicillin was 30 μg/ml/day, which was already determined to be the least effective regimen (3). The least effective regimen was used in order to determine the activity of both drugs used together rather than any resulting kill attributed to a single agent. Total drug concentrations were simulated due to low protein binding of these compounds.

Samples (0.5 ml) from each experiment were obtained at 0, 1, 2, 4, 8, 24, 28, 32, 48, 56, and 72 h and used for the pharmacodynamic analysis. Each sample was processed as previously described (5). Antibiotic carryover was accounted for by using appropriate dilutions so that the antibiotic concentrations were below the MIC and/or by the addition of β-lactamase (Sigma-Aldrich, St. Louis, MO) to amoxicillin-containing samples. The total reduction in log₁₀ CFU/milliliter over 72 h was determined by plotting time-kill curves. Any potential changes in MICs were checked at 24, 48, and 72 h via susceptibility testing using Etest methodology (AB Biodisk, Solna, Sweden). Results from these experiments were compared to those of previous amoxicillin 30-μg/ml/day pulsatile experiments that were completed in our laboratory (3).

Samples (0.5 ml) were removed from each model at 0.5, 1, 2, 4, 6, 8, 24, 28, 32, 48, 56, and 72 h for determination of antibiotic concentrations. All samples were stored at −70°C until analysis. Concentrations of amoxicillin were determined as previously described (3) using 10, 5, 0.625, and 0.3125 μg/ml for the standard curve. Clarithromycin concentrations were determined by bioassay using Micrococcus luteus ATCC 9341 with concentrations of 2, 0.5, 0.25, 0.125, and 0.0625 μg/ml for the standard curve. The interday coefficients of variation are 3.6 to 5.6% and 3.1 to 6.8% for the clarithromycin and amoxicillin assays, respectively. Peak concentrations, elimination rates, half-lives, and time above the MIC were then calculated from the concentration-versus-time plots using PK Analyst software (Micromath, Salt Lake City, Utah). Based upon these pharmacokinetic parameters, the time the amoxicillin concentration was higher than its MIC for S. pneumoniae 16891 (T > MIC) was calculated for each 24-h dosing interval.

Time to 99.9% kill (T_99.9) was determined by visual inspection. Bacterial densities in log₁₀ CFU/milliliter at evaluated time points were compared by one-way analysis of variance with Tukey's post hoc test for multiple comparisons. A P value of ≤0.05 was considered significant.

S. pneumoniae ATCC 49150 was susceptible to both drugs, which had amoxicillin MICs of <0.12 and 0.032 μg/ml and clarithromycin MICs of 0.5 and 0.47 μg/ml via broth microdilution and Etest, respectively. S. pneumoniae 16891 demonstrated intermediate susceptibility to amoxicillin (MIC, 4 μg/ml) and resistance to clarithromycin at 8 μg/ml by both broth microdilution and Etest.

Clarithromycin and amoxicillin bioassay results are shown in Table 1, with %T > MIC results shown in Table 2. Due to the lengthy half-life of clarithromycin, accumulation resulted as noted by the 6-h concentration in Table 2.

TABLE 1.

Pharmacokinetic results^a

Drugd	kel (h−1)e	t1/2 (h)f	Concn after the first dose (μg/ml)
			Actual	Expected
Clarithromycin
15 μg P	0.11 ± 0.01	6.41 ± 0.67	3.69 ± 0.02	3.75
7.5 μg P	0.11 ± 0.01	6.26 ± 0.45	2.09 ± 0.04	1.88
3.75b μg P	0.11 ± 0.01	6.17 ± 0.31	0.95 ± 0.02, 1.70 ± 0.07c	0.94, 1.80c
3.75 μg Q12	0.12 ± 0.02	6.06 ± 1.2	1.48 ± 0.02	1.88
3.75 μg Q8	0.12 ± 0.03	6.18 ± 1.3	1.26 ± 0.06	1.25
0.5 μg CI	0.11 ± 0.01	6.40 ± 0.75	0.49 ± 0.02	0.5
Amoxicillin
30b μg P	0.73 ± 0.07	0.95 ± 0.10	9.2 ± 0.48	7.5
30 μg Q12	0.64 ± 0.02	1.08 ± 0.02	15.2 ± 0.23	15
30 μg Q8	0.63 ± 0.06	1.11 ± 0.06	12.5 ± 0.51	10
8 μg CI	0.75 ± 0.04	0.92 ± 0.04	7.7 ± 0.30	8

^aData are presented as means ± standard deviations.

^bTotal concentration of clarithromycin per milliliter in a 24-h dosing cycle.

^cConcentration of clarithromycin (3.75 μg P) at 6 h due to drug accumulation.

^dP, pulsatile dosing; Q12, every-12-h dosing; Q8, every-8-h dosing; CI, continuous infusion.

^eClarithromycin target, 0.1155; amoxicillin target, 0.693.

^fClarithromycin target, 6 h; amoxicillin target, 1 h.

TABLE 2.

Amoxicillin time above its MIC (versus a resistant strain; MIC = 4 μg/ml)^a

Dose and regimen	Actual (±SD) result		Targeted result
	%T > MIC0-6b	%T > MIC0-24c	%T > MIC0-6	%T > MIC0-24
30 μg/ml/day (4 pulses)	65.8 ± 5.5	21.9 ± 2.0	56.7	19.6
30 μg/ml/day (Q8)	30.0 ± 2.8	18.3 ± 4.4	21.7	16.3
30 μg/ml/day (Q12)	35.0 ± 1.0	16.3 ± 0.5	31.7	15.8
8 μg/ml (CI) (6 h)	95.0 ± 0.3	26.5 ± 0.0	100.0	29.2

^aData are presented as means ± standard deviations. Q12, every-12-h dosing; Q8, every-8-h dosing; CI, continuous infusion.

^bPercent time above MIC in the first 6 h of the dosing regimen.

^cPercent time above MIC in the 24 h of the dosing regimen.

The results from the in vitro pharmacodynamic models are shown in Fig. 1 and 2.

FIG. 1.

Clarithromycin pulse dosing against S. pneumoniae 16891. mcg, microgram.

FIG. 2.

Summary of results of amoxicillin combined with clarithromycin. Amoxicillin at 30 μg/ml/day plus clarithromycin at 3.75 μg/ml/day with dosing every 8 h (Q 8 h), dosing every 12 h (Q 12 h), or pulse dosing (Pulse) along with 6 h of continuous infusion of amoxicillin at 8 μg/ml and clarithromycin at 0.5 μg/ml against (A) S. pneumoniae 16891 and (B) S. pneumoniae ATCC 49150.

Initial results using clarithromycin pulse dosing with the 15- and 7.5-μg/ml/day regimens resulted in killing to detection limits (T_99.9 = 8 and 6 h, respectively) of 2 log₁₀ CFU/ml against S. pneumoniae 16891 (Fig. 1). We proceeded with the 3.75-μg/ml/day dose of clarithromycin, which resulted in regrowth to beyond initial inoculum for pulsing, and traditional regimens were given every 8 and 12 h. This dose was determined to be the least effective dose and was utilized with amoxicillin.

Contrary to the use of each agent alone, use of clarithromycin at 3.75 μg/ml/day (approximately equivalent to 600 mg orally once per day) and amoxicillin at 30 μg/ml/day (approximately equal to 1 g orally three times per day) in a pulsatile regimen resulted in significant killing against the amoxicillin-intermediate-susceptible and clarithromycin-resistant strain S. pneumoniae 16891 at the 72-h sample (T_99.9 = 6 h; P < 0.001). This result is in direct contrast to the same drug concentrations being administered either every 8 or every 12 h for the same organism as demonstrated in Fig. 2A (T_99.9 = 7 and 4 h; lost after 30 and 12 h, respectively). Combination pulsatile dosing demonstrated significant killing by 24 h compared to either the continuous infusion or dosing every 8 h (P < 0.001) and by 48 h compared to every-12-h dosing (P = 0.001). No change in the drug's MIC for the organism was found during any utilized regimen. Against the ATCC 49150 strain, killing to detection limits was demonstrated for all drug regimens (P > 0.05) (Fig. 2B).

A clarithromycin concentration of 0.5 μg/ml given as a 6-h continuous infusion (CI) was similar to the area under the concentration-time curve for the 3.75-μg/ml/day pulsatile regimen. Therefore, this concentration was combined with amoxicillin at 8 μg/ml delivered continuously over the first 6 h of the 24-h cycle. As shown in Fig. 2, CI over 6 h demonstrated results which were significantly inferior to those by either pulse (P < 0.01), traditional dosing every 8 h (P = 0.05), or traditional dosing every 12 h (P < 0.01) against S. pneumoniae 16891. Against S. pneumoniae 49150, Cha and Rybak (3) demonstrated that amoxicillin concentrations of 0.5 μg/ml given as either pulsatile or traditional dosing resulted in minimal regrowth. Therefore, killing to detection limits using amoxicillin at 8 μg/ml combined with clarithromycin at 0.5 μg/ml given by CI was expected. Due to minimal response by S. pneumoniae 16891 to the combination CI, CI models were not run against S. pneumoniae 49150.

Against S. pneumoniae 16891, amoxicillin remained above its MIC 65.8% ± 5.5% of the 6-h pulsatile dosing period; however, amoxicillin remained above its MIC only 21.9% ± 2.0% of the total 24-h dosing interval. This is in direct contrast with dosing every 8 or 12 h, which results in concentrations above the MIC of 18.3% ± 4.4% and 16.3% ± 0.5% of the 24-h interval. CI for 6 h resulted in the drug concentration being 95.0% ± 0.3% above its MIC for the 6-h infusion time; however, concentrations remained above the MIC for only 26.5% ± 0.0% of the total day. As for clarithromycin, at no time during any of the regimens was the concentration above the MIC.

Previously, Cha and Rybak demonstrated that use of amoxicillin alone resulted in a pulsatile killing response against S. pneumoniae 16891 compared to traditional dosing every 8 or 12 h. Although complete eradication of the organism was not proven, they were able to demonstrate an organism response contrary to more traditional dosing (3). This pulsatile response was revealed when S. pneumoniae was treated with traditional 8- or 12-h dosing and then switched to pulsatile at 48 h. The dosing change resulted in similar responses to previous pulsatile regimens, whereas when the treatment was switched from pulsatile dosing to traditional dosing, failure of the regimen was demonstrated.

This project evaluated a pulsatile dosing schedule utilizing the combination of a macrolide and beta-lactam antibiotic against an S. pneumoniae isolate which has decreased susceptibility to these antibiotics in comparison to more traditional two- or three-times-daily regimens.

Beta-lactam and macrolide antibiotics generally require the serum concentration to remain above the MIC for 40 to 50% of the 24-h dosing interval for efficacy; pulsatile dosing appears to violate this parameter (1). As Cha and Rybak previously demonstrated, optimizing the T > MIC of amoxicillin by using shorter dosing intervals or CI alone did not result in complete eradication of the intermediate-susceptible S. pneumoniae isolate (3). However, with the addition of a macrolide at concentrations which never exceed the MIC and at a suboptimal area under the concentration-time curve/MIC ratio of the organism, kill to detection limits was demonstrated. In addition, use of an antimicrobial agent with a long half-life such as clarithromycin in a pulsatile regimen will result in accumulation of drug.

At this time, the exact mechanism which explains why the pulsatile combination resulted in complete eradication of the organism despite the lack of T > MIC is unknown. Although administration of penicillins during the early growth phase of the organism provided enhancement of killing, speculation exists as to whether this mechanism is responsible for the overall effectiveness of pulsatile dosing administration (3, 4). One possibility would be synergistic activity of the two antimicrobials. By checkerboard analysis against the intermediate isolate, we found no change in amoxicillin MIC and a clarithromycin MIC of 4 μg/ml, which is still below levels utilized. As another possibility, Isbister et al. demonstrated an increased expression of RNA by S. pneumoniae when exposed to pulsatile doses of amoxicillin (J. D. Isbister, G. A. Molina, D. Lunsford, R. J. Guttendorf, E. M. Rudnic, R. Harding, and D. Barzaghi, Abstr. 44th Intersci. Conf. Antimicrob. Agents Chemother., abstr. F-1542, 2004) which may provide the basis of why pulsatile regimens result in an altered killing profile compared to that of more traditional dosing. However, examination of the mechanism and molecular changes resulting from pulsatile dosing is needed to fully appreciate why this treatment regimen is successful against S. pneumoniae.

Demonstration of a successful combination of a macrolide and beta-lactam against a nonsusceptible strain of S. pneumoniae may be an important finding for treatment of these pathogens. Of interest, there was no difference noted among each of the dosing regimens against the susceptible strain. Sun et al. (7) evaluated pulsatile dosing of amoxicillin and clarithromycin against three susceptible S. pneumoniae isolates in a murine pneumonia model. Their study was of similar design, and they were able to demonstrate no difference between regimens, which directly supports our results with control isolate 49150. The fact that these results demonstrate no pharmacodynamic difference between pulsatile dosing and the traditional dosing regimens for organisms which remain susceptible may translate to a convenient regimen designed as a pulsatile oral formulation providing once-daily antimicrobial administration for respiratory tract infections involving this pathogen, although further research is warranted.

Present applicability of these results in clinical practice is limited due to the nature of the experiments. Although the results appear promising, this study had limitations. First, only a limited number of organisms have been included. Additionally, we did not simulate clarithromycin's metabolite, which may have activity against S. pneumoniae. Finally, T_99.9 may not be the best endpoint for combinations of drugs with different primary pharmacodynamic predictive parameters.

In the current study, use of pulsatile dosing against S. pneumoniae with reduced susceptibility demonstrates superior reduction in bacterial concentration compared to that of more traditional two- or three-times-daily dosing. However, further research exploring the mechanism for pulsatile dosing and confirmation of these results are needed before applying this information clinically.