Determination of the Postexposure Prophylactic Benefit of Oral Azithromycin and Clarithromycin Against Inhalation Anthrax in Cynomolgus Macaques

Abstract

Background

Sufficient and diverse medical countermeasures against severe pathogenic infections, such as inhalation anthrax, are a critical need. Azithromycin and clarithromycin are antimicrobials commonly used for both upper and lower respiratory infections. They inhibit protein synthesis by blocking the formation of the 50S ribosomal subunit. To expand the armamentarium, these 2 antibiotics were evaluated in a postexposure prophylactic model of inhalation anthrax in cynomolgus macaques.

Methods

This prophylaxis study had 4 test arms: azithromycin, clarithromycin, a levofloxacin control, and a placebo. Beginning 24 hours after exposure to a target challenge dose of 200 lethal dose 50 (LD₅₀) of Bacillus anthracis Ames spores, animals were treated orally until 30 days postchallenge and then observed until 75 days postchallenge.

Results

The test group that received clarithromycin had a survival rate of 67%. The test group that received azithromycin had a survival rate of 50%, but the peak azithromycin plasma levels achieved were <30 ng/mL—much lower than the expected 410 ng/mL. The levofloxacin positive control had a survival rate of 50%; all of the negative controls succumbed to infection.

Conclusions

The efficacy of clarithromycin prophylaxis was statistically significant compared with placebo, while azithromycin prophylaxis was indistinguishable from placebo. Given the low plasma concentrations of azithromycin achieved in the study, it is not surprising that half the animals succumbed to anthrax during the dosing period; the animals that survived beyond the time during which placebo control animals succumbed survived to the end of the observation period.

The goal of this study was to evaluate the prophylactic benefit of azithromycin and clarithromycin, administered orally in cynomolgus macaques, starting 24 hours post (lethal) challenge (PC) with Bacillus anthracis Ames spores. Azithromycin and clarithromycin are broad-spectrum semisynthetic derivatives of the macrolide erythromycin [1, 2]. Similar to other macrolide antibiotics, azithromycin and clarithromycin inhibit protein synthesis by blocking assembly of the 50S large ribosomal subunit and impeding extension of newly formed polypeptide chains [3–6].

Early in vitro studies revealed that clarithromycin and azithromycin were effective against both gram-positive and gram-negative microorganisms [7–9]. Both are commonly used against an array of upper respiratory tract infections, such as pharyngitis and sinusitis [9–14]. Azithromycin and clarithromycin are also used against lower respiratory tract infections, such as pneumonia and bronchitis [15–18].

Inhalation anthrax is a severe infection where, initially, the spores of B. anthracis are deposited in the lower respiratory tract of the lungs. From there, germinated spores migrate from the lungs and become centered in the mediastinal lymph nodes and spread out via the bloodstream to the rest of the body. Because B. anthracis spores can be produced and aerosolized to purposely infect people, effective antimicrobials should be made available to mitigate such an event. Antimicrobials have been shown to be effective against inhalation anthrax and have been recommended for use and stockpiled by the Centers for Disease Control and Prevention (CDC) against this infection [19, 20]. Azithromycin and clarithromycin were evaluated as antimicrobial postexposure prophylaxis (PEPAbx) against lethal levels of aerosolized B. anthracis spores in a cynomolgus macaque model. A description of this study and the degree of protection provided by each of these 2 antimicrobials follows.

METHODS

Challenge Material

Difco sporulation medium was used for this study. The medium was made of 40 g Bacto nutrient broth, 50 mL of 10% (weight to volume) potassium chloride (KCl), and 50 mL of 1.2% (weight to volume) magnesium sulfate heptahydrate. These were mixed with water for injection (WFI) for a final volume of 5000 mL and pH adjusted to 7.6 with 1.0 M sodium hydroxide. This mixture was autoclaved, then cooled; after cooling, 5 mL each of sterile 1 M calcium nitrate tetrahydrate, 0.1 M manganese(II) chloride tetrahydrate, and 1.0 mM iron(II) sulfate heptahydrate were added. Bacillus anthracis Ames spores from a seed stock culture obtained from BEI Resources were grown in the prepared Difco sporulation medium. The spore batches produced were combined at a concentration of 1.5 × 10¹⁰  colony forming units (CFU)/mL. Pooled spores were refrigerated until needed for challenges.

Test System

Twenty cynomolgus macaques (10 males and 10 females) aged under 4 years old and weighing 2.4 to 3.7 kg were obtained from Covance Research Products, Inc. (Alice, TX).

Test Articles

Azithromycin for injection was obtained from the University of Alabama Health Services Foundation HSF Clinic Pharmacy (Birmingham, AL). Clarithromycin for oral suspension and levofloxacin (control article) oral solution were both obtained from Harbin Discount Pharmacy in Birmingham, Alabama. Test and control articles were stored at room temperature. US Pharmacopeia (USP) WFI was obtained from Buler Schein Animal Health (Dublin, OH) and stored at room temperature.

Inhalation Challenge with Bacillus anthracis Ames Spores

Prior to the spore challenge, animals were anesthetized with Telazol (5 mg/kg intramuscularly [IM]) and positioned in a supine orientation in radial head-out plethysmography plenums. An aerosolized dose of 200 median lethal dose (LD₅₀) (ie, 200 × 61 800 CFU = 1.2 × 10⁷ CFU [21]) B. anthracis Ames spores was targeted for delivery to each animal via a mask placed over the nose. The nebulizer was filled with an appropriate amount of B. anthracis spores suspended in USP WFI and connected to the aerosol delivery line. The nebulizer spore suspension concentration was determined once from a pre-spray sample collected from the master suspension prior to the first challenge (described below).

A plethysmography device was used to determine the lung tidal volume for each animal and when the animal had breathed a total accumulated inhaled volume of approximately 7.0 L of aerosol containing the prescribed challenge dose of 200 LD₅₀. Once the target volume of 7.0 L was achieved, the B. anthracis Ames spore aerosol was directed away from that challenge mask and replaced with filtered air using the electronically actuated solenoid filter device. At this point, the nebulizer and liquid impinger were stopped, and a postspray nebulizer sample was collected for microbiological analysis.

Postchallenge Dose Confirmation

To ensure the spore concentration of the aerosol was at the expected level, impinger aerosol samples were taken, diluted in WFI, and plated on Tryptic Soy Agar plates. Mean plate counts were calculated to confirm the aerosol spore concentration. Measurement of the impinger aerosol spore concentration, the liquid volume, the sampling flow rate of the aerosolization unit, and the sampling time enabled the accurate calculation of the aerosol spore concentration. With a known aerosol spore concentration, the actual individual lethal aerosol doses of B. anthracis Ames spores for each animal were determined by the product of the spore concentration in the aerosol and the cumulative volume of aerosol breathed, as measured by plethysmography.

Administration of Antibiotics

On days 1 through 30 PC, chair-restrained animals were treated via nasogastric gavage starting at 24 ± 2 hours PC, and treatments continued as described below. A 5-French nasogastric catheter was used to deliver antimicrobial or vehicle. Following dose delivery, the catheter was flushed with up to 3 mL of USP WFI.

Each antibiotic formulation was inverted and vortexed or stirred prior to dosing. The most recent body weights for each animal were used to calculate the volume in milliliters of test article needed to administer the appropriate dose (mg/kg) to each animal.

• Group 1: USP WFI, 1.2 mL/kg was administered twice daily (∼12 ± 2 hours apart) for 30 days (days 1–30 PC).

• Group 2: Levofloxacin was administered twice daily (∼12 ± 0.5 hours apart) for 30 days (days 1–30 PC). The first dose each day was 15 mg/kg; the second dose was 4 mg/kg.

• Group 3: Azithromycin was administered once on day 1 at 7.5 mg/kg; subsequent doses of 3.8 mg/kg were administered once a day (∼24 ± 0.5 hours apart) for 29 days (days 2–30 PC).

• Group 4: Clarithromycin was administered at a dose of 30 mg/kg twice a day (∼12 ± 0.5 hours apart) for 30 days (days 1–30 PC).

After dosing with each test group from days 1 through 30 PC (phase I), the animals were observed from days 31 through 75 PC (phase II).

Bioanalytical Analysis of Plasma for Antibiotic Levels

Plasma levels (peak and trough) of levofloxacin, azithromycin, and clarithromycin were determined using high-performance liquid chromatography–tandem mass spectrometry (HPLC-MS/MS) methods [22, 23]. Plasma levels of levofloxacin, azithromycin, and clarithromycin in cynomolgus macaques were determined from blood samples taken just prior to treatment and at approximately 1, 8, 12, and 24 hours postdosing on days 1 and 29 PC.

Macroscopic and Microscopic Pathology

Animals found dead, euthanized because they were moribund, or euthanized as scheduled on day 75 PC were subjected to a complete gross postmortem examination (ie, necropsy). The gross examination included examination of external surfaces of the body, all orifices of the body, and cranial, thoracic, abdominal, and pelvic cavities and their contents.

Animals in all dose groups had tissues collected for microscopy and histopathology from the following organs: brain, kidney, liver, lung with mainstem bronchi, spleen, and mandibular, mediastinal, mesenteric, and bronchial lymph nodes. Gross lesions in these organs were also sampled and similarly evaluated. Samples of each tissue listed were sent for microbiological evaluation. The tissues were homogenized, vortexed, diluted, plated in triplicate on blood agar, and incubated for colony formation. Plates were analyzed for B. anthracis.

Statistical Analysis

For each group, the survival curves were estimated using the Kaplan–Meier method. Log-rank tests were then applied to each pairwise group to test for significant differences in their survival distributions (P < .05). A Bonferroni correction was applied to the calculated P values to account for the multiple log-rank tests [24, 25].

RESULTS

Study Metrics

Table 1 presents various study metrics for the PEPAbx study; it describes key elements of the study populations and design including starting mean body weights, average spore challenge, route of drug administration, time-to-death, and level of protection for levofloxacin, azithromycin, and clarithromycin.

Table 1.

Study Metrics for the PEPAbx Evaluation of Azithromycin and Clarithromycin Against Inhalation Anthrax in Cynomolgus Macaques Performed at Southern Research

			Mean Weights at Challenge, kg
Study Arma	n	% Male	Mean ± SD	Min−Max	Average Spore Challengeb (LD50)	Nonsurvivors Mean Time to Death, Days (Min−Max)	Number. of Survivors (% Survival)
USP WFI	4	50	2.7 ± 0.1	2.5 − 2.8	234	6 (4 − 9)	0 (0)
Levofloxacin	4	50	3.0 ± 0.4	2.7 − 3.5	124	39 (39)	2 (50)
Azithromycin	6	50	2.6 ± 0.3	2.4 − 3.1	109	5.3 (4 − 6)	3 (50)
Clarithromycin	6	50	2.7 ± 0.5	2.3 − 3.7	159	44 (36 − 52)	4 (66.7)

Abbreviations: LD₅₀, median lethal dose; SD, standard deviation; Min–Max, minimum–maximum; PEPAbx, antibiotic postexposure prophylaxis; USP WFI, US Pharmacopeia water for injection.

^aAntimicrobials and placebo administered via nasogastric tube.

^bAnimals challenged via pulmonary exposure.

Aerosolization

The spore master suspension was diluted to be 2.0 × 10⁹ CFU/mL. Each animal breathed a total accumulated inhaled volume of approximately 7.0 L of aerosol. Group 1 animals were exposed to an average spore dose of 1.5 × 10⁷ CFU, group 2 animals were exposed to an average dose of 7.7 × 10⁶ CFU, group 3 animals were exposed to an average dose of 6.7 × 10⁶ CFU, and group 4 animals were exposed to an average dose of 9.8 × 10⁶ CFU. Average LD₅₀ equivalents for each group are shown in Table 1.

Mortality/Time to Death

The key indicator of efficacy was survival following exposure. The Figure 1 Kaplan–Meier plot and Table 1 show the proportion of animals that survived and length of survival for the 3 antimicrobials—levofloxacin, clarithromycin, and azithromycin—and for placebo for days 0 through 75 PC. As illustrated in Figure 1, no animals given placebo survived, half of the animals in the positive control (2/4) and azithromycin (3/6) groups survived, and two-thirds (4/6) of those that received clarithromycin survived.

Figure 1.

Kaplan–Meier plots of the test groups that received azithromycin, clarithromycin, and levofloxacin. Abbreviation: USP WFI, Pharmacopeia water for injection.

During phase I (days 1–30 PC), 3 of the 6 animals given azithromycin either became moribund and were euthanized or succumbed to the infection—in the same time frame as the animals in the placebo group. The remaining 3 animals that received azithromycin survived to the end of the study. All animals that received either clarithromycin or levofloxacin survived phase I. During phase II (days 31–75 PC), 1 animal succumbed and 1 was euthanized for being moribund in the levofloxacin group on day 39 PC, leaving 2 that survived to day 75 PC. Also, 2 animals in the clarithromycin test group succumbed to infection during phase II: 1 on day 36 PC and 1 on day 52 PC. The remaining 4 animals survived to the study’s end on day 75 PC.

Statistical Evaluation of Mortality

As shown in Table 2, survival distributions were not statistically different between the 3 antimicrobial test groups. Animals given clarithromycin and those given levofloxacin had survival rates that were superior to that of USP WFI (0.0%), with respective P values of .006 and .040 after a Bonferroni correction for multiple comparisons. In contrast, the survival rate for the azithromycin group was not different from that for the USP-WFI negative control group after correction (P = .75).

Table 2.

Statistical Evaluation of Percent Survival of the Azithromycin and Clarithromycin Test Groups Compared With the Levofloxacin Positive Control and USP-WFI Negative Control Groups

Comparison Groupa	P Values for Survival Comparisons (Bonferroni Correction)
Azithromycin vs USP WFI	.753
Clarithromycin vs USP WFI	.006
Levofloxacin vs USP WFI	.040

Abbreviation: USP WFI, US Pharmacopeia water for injection.

^aStudy was not powered to compare outcomes between antimicrobial treatment groups.

Plasma Drug Levels

Mean peak levels of levofloxacin, azithromycin, and clarithromycin for days 1 and 29 PC are summarized in Table 3. On day 1 PC, peak levofloxacin plasma levels at 1 hour after dosing ranged from 1120 to 7010 ng/mL. Peak levels for day 1 exceeded the previously reported levofloxacin minimum inhibitory concentration (MIC) for B. anthracis Ames spores by 9.3 to 58 times [26]. Trough levels ranged from 87 to 304 ng/mL 12 hours after dosing. Peak and trough levels at the beginning and end of the treatment were similar.

Table 3.

Mean Plasma Peak Levels of Levofloxacin, Azithromycin, and Clarithromycin From Day 1 and Day 29 Postchallenge Samples for Each Animal and the Expected Level

Antimicrobial	Day 1 Postchallenge Level ± SD, ng/mL	Day 29 Postchallenge Level ± SD, ng/mL	Expected Level, ng/mL
Levofloxacin	2978 ± 2731	2843 ± 1050	6200a
Azithromycin	30 ± 12	29 ± 18	410b
Clarithromycin	3463 ± 1713	4578 ± 3210	3000–4000c

Abbreviations: MIC, minimum inhibitory concentration; PK, pharmacokinetic

^aExpected plasma level is based on maximum serum concentration (Cmax) levels observed from a PK study performed in cynomolgus macaques given orally administered levofloxacin [22].

^bPeak azithromycin levels were expected to be ∼410 ng/mL [26]. This is far lower than the azithromycin MIC for Bacillus anthracis Ames spores of 1000 ng/mL [26].

^cExpected plasma level is based on Cmax levels observed from a PK study performed in cynomolgus macaques given orally administered clarithromycin [23].

Only 2 animals that received azithromycin had detectable plasma levels 1 hour after the first dose on day 1: the first animal had a level of 21.9 ng/mL and the second had a level of 38.2 ng/mL. Although peak azithromycin levels were expected to be approximately 410 ng/mL [27], peak samples from day 29 for the 3 survivors were just 12.0, 47.0, and 27.5 ng/mL for reasons unknown. This is far lower than the azithromycin MIC for B. anthracis Ames spores of 1000 ng/mL [26].

On day 1 PC, peak plasma levels of clarithromycin ranged from 2100 to 6460 ng/mL in samples taken 1 to 8 hours after the first dose. Trough levels ranged from 366 to 1480 ng/mL 12 hours after dosing. On day 29 PC, peak levels at 1 hour after dosing ranged from 2030 to 10 700 ng/mL, and trough levels at 12 hours after dosing ranged from 241 to 7120 ng/mL. Peak and trough levels for days 1 and 29 PC exceeded the previously reported clarithromycin MIC for B. anthracis Ames spores of 120 ng/mL [26].

Pathology

All of the animals in the study that succumbed or were euthanized because they were moribund had both macroscopic lesions (Supplementary Tables 1 and 2) and microscopic lesions (Supplementary Tables 3 and 4) consistent with systemic B. anthracis infections [28, 29].

DISCUSSION

The positive control (levofloxacin) and negative control (placebo) arms in this study behaved as expected. All placebo-treated animals succumbed to inhalational anthrax within 9 days, while all levofloxacin-treated animals survived during the 30 days of antibiotic treatment. Two of 4 levofloxacin-treated animals succumbed to inhalational anthrax after cessation of treatment—both on day 39. This observation is similar to the study published in 1993 by Friedlander et al [20], where animals succumbed on days 36, 39, 42, 50, and 58 after cessation of various antibiotics, and may be due to latent germination of persistent spores in lung alveoli. The test article clarithromycin also resulted in complete survival during the 30-day treatment phase, with 2 of 6 animals succumbing to inhalational anthrax on days 36 and 52. The mean plasma drug levels for levofloxacin and clarithromycin were as expected and were 9- to 58-fold and 16- to 89-fold greater than the MIC for B. anthracis, respectively.

The test article azithromycin failed to protect 3 animals that succumbed within 6 days, similar to placebo controls, while the 3 remaining animals survived to the end of the observation period. Plasma antibiotic levels of azithromycin were largely undetected; only 2 of 6 animals had detectable levels on day 1 (mean, 30 ng/mL) and 3 of 3 had detectable levels on day 29 (mean, 28 ng/mL). The expected level of azithromycin was 410 ng/mL and the MIC is 1000 ng/mL; the reason for not achieving the target plasma dose is not known. Given the low plasma levels of azithromycin, it is not surprising that survival was essentially not distinguished from placebo controls.

Azithromycin and clarithromycin were also evaluated for postexposure prophylactic efficacy in mouse studies (unpublished data, Watkins 2013). In both studies, neither azithromycin nor clarithromycin protected mice when treated on days 1–30 PC. In Hatch et al [30], plasma drug levels for both drugs were below the target levels, although these authors suggested that this could have been a result of an additional freeze-thaw cycle for the plasma samples. Regardless, a lower-than-expected plasma antimicrobial level comports with lack of protection. Given a similar finding in this non-human primate (NHP) study for azithromycin, it is possible that pharmacokinetic differences in animals may preclude the evaluation of azithromycin for prophylaxis of inhalation anthrax.

Survival in NHPs given clarithromycin, but not azithromycin, was statistically different from that of untreated (USP WFI) controls. Even though the final survival rate for the azithromycin-treated animals was the same as that of the levofloxacin-treated animals (50%), statistical significance was not achieved since the log-rank test used for this analysis compared the entire survival experience for each group. More than simply comparing the final survival rates of each group at the end of the experiment, it also compares the whole survival curve of 1 group with another. Although the survival curve of the azithromycin group eventually diverged from that of the USP-WFI group, the log-rank test was not significant because the curves were initially similar.

Survival and survival probability in the clarithromycin- and levofloxacin-treated NHPs were similar. Levofloxacin is already Food and Drug Administration approved for the prophylaxis and treatment of anthrax, and clarithromycin is a promising candidate for PEPAbx following exposure to lethal doses of aerosolized B. anthracis spores.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Acknowledgments. The authors acknowledge Dr. Robert W. Watkins, Study Director, for his guidance of the performance of Study 11051.02.09 at Southern Research, Birmingham, AL, and for writing the study report used to produce this manuscript. They also thank Southern Research Institute for performing Study 11051.02.09, under National Institute of Allergy and Infectious Diseases (NIAID) contract no. N01-AI-30063, which was initiated on 6 June 2012 and completed 5 February 2013. They also thank Mr. Eric Chu of Leidos for providing statistical analysis and generation of the Kaplan–Meier curves used in this manuscript.

Disclaimer. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the CDC or the authors' affiliated institutions.

Financial support. This project was supported by the Centers for Disease Control and Prevention and the Office of the Assistant Secretary for Preparedness and Response.

Supplement sponsorship. This article appears as part of the supplement “Anthrax Preparedness,” sponsored by the Centers for Disease Control and Prevention.

Potential conflicts of interest. All authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.