Consequences of Delayed Ciprofloxacin and Doxycycline Treatment Regimens against Francisella tularensis Airway Infection

Abstract

This study examines the efficacy, bacterial load, and humoral response of extensively delayed ciprofloxacin or doxycycline treatments following airway exposure of mice to Francisella tularensis subsp. holarctica (strain LVS) or to the highly virulent F. tularensis subsp. tularensis (strain SchuS4). A delay in onset of both antibiotic treatments allowed the rescue of all LVS-infected animals. However, for animals infected with SchuS4, only ciprofloxacin was efficacious and prolongation of treatment rescued all animals.

TEXT

Francisella tularensis, a Gram-negative facultative intracellular bacterium, is the etiological agent of tularemia and is classified as a category A biological warfare agent by the CDC (5, 12). Human infections are mainly caused by F. tularensis subsp. holarctica (type B) or by F. tularensis subsp. tularensis (type A) (18). The latter, especially when inhaled, is highly infectious and virulent in humans (6, 18).

Antibiotics usually provide curative therapy for tularemia. Mortality depends on the infection type, patient health, and treatment onset. Aminoglycosides offer good bactericidal properties and low relapse rates, yet due to their toxicity and the need for parenteral administration, they are not the drug of choice for prophylactic treatment. According to the CDC guidelines, in cases of F. tularensis epidemics or a bioterror scenario, ciprofloxacin and doxycycline represent the treatments of choice (5). Tetracycline and doxycycline treatments have documented relapses (∼10%); however, others indicate tetracycline to be as effective as streptomycin when a prolonged treatment is given (7, 15). Ciprofloxacin and other quinolones have been shown to be effective against experimental tularemia (13, 14, 19), as well as in clinical treatment of type B tularemia (1, 2, 9, 16).

Efficacy of different ciprofloxacin and doxycycline treatment regimens against F. tularensis airway infection.

Mice (female BALB/c) were inoculated intranasally (i.n.) with 100 50% lethal doses (LD₅₀) of either F. tularensis LVS (10⁵ CFU) or SchuS4 (10² CFU), a dose that initiates a rapidly fatal disease, resulting in death of all animals within 5 to 7 days (3). Treatments with ciprofloxacin (50 mg/kg of body weight twice daily [b.i.d.], intraperitoneally [i.p.]) or doxycycline (40 mg/kg b.i.d., i.p.) were initiated at either 24, 48, or 72 h postinfection (p.i.) and continued for 7 and 14 days, respectively.

Ciprofloxacin or doxycycline treatment rescued all LVS-infected mice regardless of initiation time of treatment (Table 1). In the case of delayed treatment starting 72 h p.i., clear clinical signs (weight loss, scruffy appearance, lethargy) were observed. At the end of the treatment, all LVS-exposed animals exhibited no clinical signs and significant anti-F. tularensis antibody titers were measured 45 days p.i. (Table 1). When ciprofloxacin treatment was initiated 72 h post-SchuS4 infection, ∼30% of the mice appeared to be morbid and died 4 to 5 days after cessation of treatment, while treatment starting 24 or 48 h p.i. rescued all animals. None of the SchuS4-infected mice that were treated with doxycycline 72 h following infection survived this airway infection (Table 1), while survival rates of mice treated 24 or 48 h p.i. were 90% and 30%, respectively. During the treatment periods described above, all mice survived irrespective of the antibiotic used. A similar systematic study performed by Russell et al. (14) found that following SchuS4 intraperitoneal exposure, treatment with either doxycycline or ciprofloxacin is efficient when initiated 24 h p.i. These results of treatment relatively early p.i. are in complete agreement with our results, although we used a different route of infection—airway exposure. However, when the onset of treatment was delayed to 48 or even 72 h p.i., a scenario which could be relevant to bioterror, we find a clear advantage of ciprofloxacin over doxycycline (Table 1).

Table 1.

Comparative efficacy of different regimens of treatment by ciprofloxacin and doxycycline, initiated at different times after airway (i.n.) infection of LVS or SchuS4^a

Treatment	Onset of treatment (h p.i.)	% Survivalb (n = 10)		Ab titer of surviving animalsc (GMT)
		LVS infection	SchuS4 infection	LVS infection	SchuS4 infection
None	NA	0	0	NA	NA
Ciprofloxacin (7 days treatment)	24	100	100	800	80
	48	100	100	840	200
	72	100	70	1,680	2,500
Doxycycline (14 days treatment)	24	100	90	12,600	7,500
	48	100	30	29,000	21,000
	72	100	0	29,000	NA

^aNA, not applicable.

^bSignificance versus the control was determined by the Fisher exact test (P < 0.05).

^cDifferent animals were used for monitoring survival and for immunological parameters. Antibody (Ab) titer limit of detection, <40. Antibody titers were calculated as reciprocal geometric mean titers (GMT), with geometric standard deviations not greater than 1.8.

We quantified the bacterial loads in the lungs, liver, and spleen of airway-infected mice during antibiotic treatment and following its cessation (as described in references 3 and 10). As seen in Fig. 1, in LVS-exposed mice, all organs tested were clear of bacteria 4 to 7 days after ciprofloxacin administration. Doxycycline reduced the bacterial load during the 14 days of treatment, albeit at a lower rate (Fig. 1). Unlike the case for ciprofloxacin, 2 days after cessation of doxycycline (day 19 p.i.) bacteria reemerged in all organs, reaching a level of approximately 10³ CFU in the lungs, liver, or spleen (note that the LD₅₀ for intranasal administration with LVS is 10³ CFU and that for intraperitoneal administration is ∼1 CFU). The bacterial load in the organs remained for 7 days posttreatment, declining gradually below the limit of detection (Fig. 1). No morbidity was observed, and all animals survived.

Fig 1.

Bacterial counts of LVS during and following antibiotic treatment. LVS bacterial counts in the lungs (A) and liver (B) during and after antibiotic treatment with ciprofloxacin (gray circles) or doxycycline (black rectangles). Arrows indicate the onset and termination of the indicated treatment. The limit of detection was 5 CFU/organ. Values represent averages and standard deviations of results for three animals per time point.

The bacterial loads of SchuS4 in organs from animals that were treated with ciprofloxacin or doxycycline 72 h p.i. were determined immediately after cessation of treatment and 2 to 3 days thereafter (Fig. 2 and 3). Ten days p.i. (the last day of ciprofloxacin treatment), all the animals were devoid of bacteria (Fig. 2). However, 3 days later (13 days p.i.) 30% of the animals exhibited considerable levels of bacteria in the lungs, liver, and spleen (ca. 10³ CFU, equivalent to 10³ LD₅₀ i.n. or i.p.), in agreement with the observation of survival shown in Table 1. Doxycycline treatment initiated 72 h post-SchuS4 infection was efficient as long as the animals received the antibiotic, yet the disease relapsed after treatment cessation. At this stage approximately 10² CFU were found in the lungs, spleen, and liver, which further increased to 10³ to 10⁴ CFU in all these organs 2 days later (Fig. 3A). These results are in accordance with the inability to rescue animals from SchuS4 i.n. infection if treatment is initiated 72 h p.i. (Table 1). The occurrence of infection relapse was documented in the treatment of human cases of tularemia, particularly when tetracycline or doxycycline was used (7, 11, 15, 16), underlying the poorly understood “depot effect” of F. tularensis infection (3, 4, 8, 17). Of note, SchuS4- and LVS-infected mice eventually exhibited similar levels of humoral response following doxycycline treatment, yet only LVS-infected animals survived the high level of reemerging bacteria (equivalent lethal doses of over 10 to 100 LD₅₀) (Fig. 1).

Fig 2.

Bacterial counts of SchuS4 after cessation of ciprofloxacin treatment. Counts of residual bacteria were determined when ciprofloxacin treatment was initiated 72 h p.i. and continued for 7 days (note that when treatment initiated 48 h p.i. with ciprofloxacin, bacteria were not detected in any organ after cessation of treatment; see text). Bacterial counts were measured in 10 animals, in the lungs (black bar), liver (gray bar), and spleen (white bar) at the end of the treatment (day 10 p.i.) and 3 days later (day 13 p.i.) in an additional 10 animals. The limit of detection was 5 CFU/organ. For survival data after the ciprofloxacin treatment regimen, see Table 2. Values represent averages and standard deviations of results for 3 to 10 (as indicated) animals per time point.

Fig 3.

Bacterial counts of SchuS4 after cessation of doxycycline treatment. Bacterial counts following SchuS4 infection when doxycycline treatment was initiated 72 h p.i. for a period of 14 days (A) or 21 days (B). Bacterial counts in the lungs (black bar), liver (gray bar), and spleen (white bar) were measured at the end of the treatment (day 17 and day 24 p.i., respectively) and 2 days later (day 19 and day 26 p.i., respectively). The limit of detection was 5 CFU/organ. For survival data after the doxycycline treatment regimen, see Table 2. Values represent averages and standard deviations of results for three animals per time point.

The data underline the importance of timely onset of antibiotic treatment as an efficient countermeasure against F. tularensis subsp. tularensis infection and clearly demonstrate the advantage of ciprofloxacin in cases of delayed onset of treatment.

Consequences of further extension of the period of antibiotic treatments of SchuS4-infected animals.

To reduce and eventually eliminate bacterial relapse after termination of treatment, we examined the extension of ciprofloxacin and doxycycline treatment periods in the SchuS4 airway infection murine model. An extended treatment consisting of 10 days of ciprofloxacin administration (instead of 7 days) increased survival (to 100%) even when treatment was initiated 72 h p.i., without any disease relapse (Table 2). Conversely, prolongation of doxycycline treatment to 21 days yielded a marginal improvement when the treatment was initiated 72 h p.i. (Table 2). Following the extended treatment, no bacteria were found (limit of detection, <5 CFU) in all inspected tissues (Fig. 3B); however, 2 days after cessation of treatment a relapse occurred, with bacterial counts reaching 10⁴ CFU in all organs.

Table 2.

Effect of duration of antibiotic treatment initiated 72 h after airway infection by SchuS4

Treatment	Onset of treatment (h p.i.)a	Duration of treatment (days)	% Survival of SchuS4 infectionb (n = 10)
None	NA	0	0
Ciprofloxacin	72	7	70
	72	10	100
Doxycycline	72	14	0
	72	21	10

^aNA, not applicable.

^bSignificance versus the control was determined by the Fisher exact test (P < 0.05) except for the 21-day doxycycline group.

The findings of this study establish that both ciprofloxacin and doxycycline are effective in preventing the development of tularemia in the mouse model following airway infection, but in the case of bacteriostatic antibiotics such as doxycycline there is a significant failure to achieve complete recovery unless treatment is initiated within a short time window (24 h p.i.). At later stages of the disease (72 h p.i), prolonging ciprofloxacin treatment (to 10 days) but not prolonging doxycycline treatment (to 21 days) appeared to be an effective strategy for successful therapy against the highly virulent F. tularensis SchuS4 strain. These observations have important implications for designing efficient therapeutic approaches for treatment of tularemia in various scenarios where early treatment is not possible.