Penetration of Rufloxacin into the Cerebrospinal Fluid in Patients with Inflamed and Uninflamed Meninges

Abstract

Forty-four patients scheduled for lumbar puncture (LP) were recruited to determine the level of penetration of orally administered rufloxacin into cerebrospinal fluid (CSF). The patients were divided into three clinical groups: those with normal CSF (groups A_1d and A_7d), those with aseptic meningitis (group B), and those with bacterial meningitis (group C). Members of group A_1d received a single 400-mg rufloxacin dose, while group A_7d, B, and C constituents had a multiple-dose regimen (one 400-mg dose, followed by one 200-mg dose daily for 6 days). LP was performed on group A_1d members 5 h after they had received treatment, while for group A_7d it was undertaken 5 h after administration of the last dose. For group B, LP was performed 5 h after the first and the last doses, whereas for group C it was undertaken after the first, fourth, and last doses. Concentrations of rufloxacin in simultaneously collected CSF and plasma samples were determined. Mean CSF/plasma rufloxacin concentration ratios ranged from 0.57 to 0.84, depending on the study group. A higher, but not statistically significant, degree of penetration into CSF was observed in patients with bacterial meningitis than in those with normal CSF or aseptic meningitis. These data indicate that rufloxacin diffuses efficiently into the CSF of patients with either inflamed or uninflamed meninges.

Although the introduction of antibiotics made bacterial meningitis curable, the morbidity and mortality associated with this disease remain unacceptably high. In a recent review of 493 episodes of bacterial meningitis in adults, the overall case fatality rate was 25% (6).

The penetration of an antibacterial drug into the cerebrospinal fluid (CSF) is an important characteristic in terms of its potential use in central nervous system (CNS) infections. In the presence of inflamed meninges, therapeutic concentrations of many antibiotics can be attained in CSF (2). However, in some persons with brain abscesses, early meningitis, or late-stage meningitis, only minor impairments of the blood-CSF barrier occur. Dexamethasone, recommended for use in adjunctive therapy for bacterial meningitis, reduces meningeal inflammation (25) and thereby decreases drug penetration into CSF (2, 23). To ensure successful therapy and to prevent relapses, therapeutic concentrations of any antibiotic proposed for the treatment of CNS infections should be achieved even in the presence of uninflamed meninges.

Fluoroquinolones are broad-spectrum antimicrobial agents with in vitro and in vivo bactericidal activity against susceptible organisms (10). The individual fluoroquinolones differ in their physiochemical properties, pharmacokinetic characteristics, and body fluid penetration abilities (10, 14). Their lipophilicity (octanol-buffer partition coefficient = 0.37 at pH 7) (1, 7) and the high apparent volume of distribution of 130 to 150 liters (12, 15, 26, 28) account for their high degree of penetration into CSF. However, analysis of comparative data on CSF transport and disposition of quinolones distinguishes between agents with high (e.g., pefloxacin) and low (e.g., ciprofloxacin) levels of availability to the CSF (11, 13, 21).

Rufloxacin is an orally administered fluoroquinolone that is effective against the majority of Enterobacteriaceae family members, Neisseria meningitidis, Haemophilus influenzae, and methicillin-sensitive staphylococci but has limited activity against streptococci and methicillin-resistant staphylococci and no activity against Pseudomonas aeruginosa and enterococci (16, 24, 27). It is rapidly absorbed by the gastrointestinal tract, with a maximal plasma concentration after 4 h (12) of 3 to 5 mg/liter (12, 15) and a plasma half-life of 35 h (12). During multiple-dose administration, the drug shows an accumulation ratio in plasma ranging from 2.4 to 3.5 (3, 15). The aim of this study was to investigate the penetration of rufloxacin into the CSF of patients with inflamed or uninflamed meninges.

MATERIALS AND METHODS

Study design.

The open, four-armed, nonrandomized study was conducted at the Clinic of Infectious Diseases, Perugia University, Perugia, Italy. After providing medical histories and being subjected to physical examinations, hospitalized patients were scheduled for lumbar puncture (LP) for diagnostic or follow-up reasons. On the basis of their CSF characteristics, patients were classified into three groups: those with normal CSF (groups A_1d and A_7d), those with aseptic meningitis (22) (clear CSF meningitis) (group B), and those with bacterial (purulent) meningitis (group C). The first rufloxacin dose was 400 mg; the subsequent six doses were 200 mg daily. CSF was collected 5 h ± 30 min after the initial dose and after the last dose (day 7) in groups B and C (as well as on at day 4 in the latter). For groups A_1d and A_7d, the samples collected on day 1 (group A_1d) and day 7 (group A_7d) were obtained from different subjects, since it is considered unethical to perform more than one LP on a patient who is not suspected of having meningitis. The decision to assign a patient to group B or C was made within 24 h of receipt of the results of the CSF examination. The study protocol, patient information sheet, and patient consent form were approved by the Ethics Committee of Regione Umbria.

Patients.

Forty-four hospitalized patients (31 men and 13 women) requiring LP were included in the study after providing written informed consent. Further details on these patients are given in Table 1. The exclusion criteria were as follows: a serum creatinine concentration of >2 mg/100 ml or a creatinine clearance of ≤40 ml/min, serum aminotransferase concentrations of more than twice the upper normal limit, a history of hypersensitivity to quinolones, or a history of convulsions or epilepsy. Also excluded were patients with concomitant diseases or conditions affecting the oral administration of rufloxacin or its absorption, as well as patients treated with other quinolones in the 7 days prior to study admission.

TABLE 1.

Characteristics of investigated patients

Characteristic	Value for group
	A1d(n = 10)	A7d(n = 10)	B (n = 13)	C (n = 11)
Mean body wt, kg (range)	65.6 (51–76)	65.2 (47–80)	61.1 (40–92)	72.2 (55–89)
Male/female ratio	8/2	9/1	7/6	7/4
Mean age, yrs (range)	30.4 (27–36)	39.4 (26–60)	32.5 (20–56)	53.2 (20–65)
Indication(s) for LP or meningitis etiology (n)	AIDS (4)	AIDS (6)	Etiology undetermined (7)	Streptococcus pneumoniae (5)
	HIVa (4)	Latent syphilis, AIDS (2)	Etiology undetermined, AIDS (2)	Etiology undetermined (2)
	Latent syphilis (1)	Latent syphilis (1)	Cryptococcus, AIDS (2)	L. monocytogenes (1)
	PMLb, AIDS (1)	AIDS, non-Hodgkin's lymphoma (1)	Mumps (1)	N. meningitidis (1)
			VZVc (1)	S. milleri (1)
CSF parameters, mean ± SD
Leukocytes (per mm3)	1.9 ± 1.5	1.3 ± 1.3	201.8 ± 204.4	1,998.6 ± 1,496.1
Neutrophils (%)	11.1 ± 33.3	0.0 ± 0.0	15.7 ± 33.4	97.4 ± 3.50
Protein (mg/dl)	54.6 ± 25.7	43.0 ± 16.3	71.0 ± 38.4	146.0 ± 69.60
CSF/plasma protein ratio (10−3)	6.0 ± 0.0	6.0 ± 0.0	10.0 ± 0.10	22.0 ± 0.10
CSF/plasma glucose ratio	0.56 ± 0.09	0.62 ± 0.31	0.50 ± 0.18	0.36 ± 0.11

^aHIV, human immunodeficiency virus. 

^bPML, promyelocytic leukemia. 

^cVZV, varicella-zoster virus. 

Drug.

Rufloxacin was supplied as 200-mg tablets by Mediolanum Farmaceutici, Milan, Italy. The drug was administered by the oral route on an empty stomach (1 h before or 2 h after meals), with the dosing schedule depending on the study group. Group A_1d patients received a single 400-mg dose, while patients of groups A_7d, B, and C were administered a 400-mg dose on day 1 and daily 200-mg doses from days 2 to 7. Rufloxacin was administered 5 h ± 30 min before the scheduled LP. Throughout the duration of the study, no quinolone antimicrobial agent other than rufloxacin was allowed. Theophylline, cimetidine, antacids, sucralfate, bismuth preparations, iron salts, and calcium carbonate were prohibited due to their interference with quinolone absorption (20). Nonsteroidal anti-inflammatory drugs, which are known to potentiate a possible proconvulsant effect of quinolones and to affect their penetration into CSF (20), were also prohibited. Comedication varied according to clinical necessity and consisted of antibiotics, antiviral and antiretroviral drugs (no protease inhibitors were used), sedatives, corticosteroids, antiemetic drugs, and antifungal and antineoplastic agents.

Clinical assessment.

Body temperature, vital functions, and neurologic symptoms were recorded daily for 7 days. All adverse events either reported by patients or observed by the investigator were also recorded and graded for severity and relationship to the study drug. Hematology and biochemistry studies and urinalysis were performed before the first and after the last rufloxacin dose. For group C patients, the above-listed laboratory tests were also undertaken on day 4, if deemed necessary.

Sample collection.

Simultaneous blood and CSF samples were collected at the following times: for group A_1d, 5 h ± 30 min after administration of the single dose of rufloxacin; for group A_7d, 5 h ± 30 min after administration of the last dose of rufloxacin; for group B, 5 h ± 30 min after administration of the day 1 and day 7 rufloxacin doses; and for group C, 5 h ± 30 min after administration of the day 1, day 4, and day 7 rufloxacin doses. To determine the rufloxacin concentration in a CSF sample, 1.0-ml aliquot of CSF was transferred into a sterile plastic Eppendorf tube and immediately stored at −20°C. Venous blood samples (5 ml) were drawn from the forearm through an indwelling butterfly needle into heparinized tubes and immediately centrifuged at 4,000 × g for 5 min. The supernatant plasma was frozen and stored at −20°C. Both CSF and plasma samples were kept frozen for no longer than 4 months. Separate aliquots of CSF were immediately analyzed as follows: microscopic examination, culture, leukocyte and erythrocyte counts, protein concentration (with determination of the plasma protein concentration for a simultaneously collected blood sample in order to calculate the CSF protein/plasma protein ratio), and glucose concentration (with determination of the glucose level for a simultaneously collected blood sample in order to calculate the CSF glucose/blood glucose ratio).

Rufloxacin assay.

Rufloxacin concentrations in CSF and plasma were determined by a modified version of the isocratic high-pressure liquid chromatographic technique described by Kisicki et al. (15). The chromatographic system consisted of a pump (Perkin-Elmer series 410 LC); an automatic sample injector (ISS-101, Perkin-Elmer; injection volume, 50 μl), a Suplex-pKb 100 column (150 by 4.6 [internal diameter] mm; particle size, 5 μm) (Supelco), a Suplex-pkb 100 precolumn (2 cm; Supelco); a spectrofluorimetric detector (Shimadzu RF-551); and an integrator (PCAX2 computer [Epson]; OMEGA-2 Analytical Workstation software [Perkin-Elmer]). The mobile phase was a mixture of acetonitrile and 0.01 M phosphate buffer, pH 2.8 (8:92), with a flow rate of 1 ml/min. The column effluent was monitored by a fluorescence detector operating at an emission wavelength of 521 nm and an excitation wavelength of 294 nm. The retention times of rufloxacin and the internal standard (ofloxacin) were 6 and 8 min, respectively. Plasma and CSF standard curves were linear in the concentration range considered. No interfering peaks were found at the retention times for rufloxacin and the internal standard. The lower limit of sensitivity was 0.1 μg/ml for both plasma and CSF. The coefficients of variation ranged from 1.4% (7.5 μg/ml) to 8.1% (0.25 μg/ml) for CSF and from 1.25% (15 μg/ml) to 6.38% (0.5 μg/ml) for plasma.

Pharmacokinetic and statistical analyses.

The degree of penetration of rufloxacin into CSF at different time points was estimated by calculating the ratio of the concentration of the drug in CSF to that in (hereafter termed the CSF/plasma percent ratio) plasma 5 h ± 30 min after rufloxacin administration on days 1, 4, and 7, according to the study group. The significance of differences in rufloxacin penetration (expressed as CSF/plasma percent ratios) between the study groups after the first drug administration and at steady state (day 7) was examined by one-way analysis of variance. The statistical significance of differences in CSF/plasma percent ratios for days 1 and 7 within groups was tested by paired t test. Statistical analyses were performed with the SAS System for Windows (release 6.12). Differences were deemed significant if the P value was <0.05 (two-sided t test).

RESULTS

Pharmacokinetics.

One patient in group C was withdrawn from the study on day 1, due to technical difficulties in performing LP within the required time interval after rufloxacin administration. This patient was not included in the pharmacokinetic assessment and was evaluated only in terms of drug safety. Two patients in group B dropped out of the study, after 3 and 6 days, due to the occurrence of serious adverse events. They were included in both the pharmacokinetic and safety assessments.

The ranges and the mean concentrations of rufloxacin in plasma and CSF and the CSF/plasma ratios for the study groups are reported in Table 2. As indicated by the CSF/plasma ratios, no significant differences were observed between groups in terms of rufloxacin penetration into CSF either on day 1 (P = 0.207) or on day 7 (P = 0.123). A trend toward a greater degree of penetration of the drug in group C patients was observed at both time points. The mean CSF/plasma ratios ranged from 0.57 to 0.61 for groups A_1d, A_7d, and B and from 0.72 to 0.84 for group C. For groups B and C, 7 and 15.3% increases respectively, in the CSF/plasma ratios were noted after multiple dosing compared with day 1 dosing only, but the difference did not reach statistical significance (P > 0.05). The individual rufloxacin concentrations in CSF and plasma following the first and last (at 7 days) rufloxacin doses are plotted in Fig. 1. Even though partial overlapping of the groups is evident, a close look at the figure will reveal that at day 1 only 20% of patients with normal CSF and 15.3% of those with aseptic meningitis showed a penetration of ≥0.75, whereas 40% of patients with bacterial meningitis showed a penetration of ≥0.75 (P = 0.626). The lines in Fig. 1 represent the CSF/plasma ratios. The individual rufloxacin concentrations in CSF and plasma at steady state (on day 7) show penetrations of ≥0.75 in 20 and 18% of patients in groups A_7d and B, respectively, and in 50% of the patients in group C (P = 0.348).

TABLE 2.

Concentrations of rufloxacin in CSF and plasma, and CSF/plasma concentration ratios

Group and sample	n	Rufloxacin concn (μg/ml), mean ± SD (range)		CSF/plasma ratio (%), mean ± SD (range)
		In CSF	In plasma
A1d, day 1	10	1.66 ± 0.61 (0.76–2.82)	3.02 ± 1.15 (1.61–5.37)	0.61 ± 0.31 (0.29–1.21)
A7d, day 7	10	1.97 ± 0.57 (1.06–2.77)	3.37 ± 1.19 (1.78–5.08)	0.60 ± 0.12 (0.48–0.84)
B
Day 1	13	1.98 ± 1.17 (0.20–4.97)	3.46 ± 1.78 (0.35–7.29)	0.57 ± 0.12 (0.41–0.82)
Day 7	11	3.16 ± 0.85 (1.81–4.97)	5.43 ± 2.23 (3.10–10.79)	0.61 ± 0.13 (0.46–0.84)
C
Day 1	10	2.93 ± 0.91 (1.94–4.35)	4.15 ± 1.24 (2.38–6.41)	0.72 ± 0.13 (0.55–0.97)
Day 4	10	2.90 ± 0.99 (1.58–4.99)	3.48 ± 0.83 (1.83–4.68)	0.84 ± 0.20 (0.49–1.11)
Day 7	10	2.47 ± 0.76 (1.59–3.94)	3.29 ± 1.05 (1.92–4.60)	0.83 ± 0.45 (0.36–1.99)

FIG. 1.

Double-logarithmic graph of CSF versus plasma concentrations of rufloxacin measured 5 h ± 30 min after administration of the drug on day 1 (open symbols) and on day 7 (closed symbols). Dotted lines represent percent penetration into CSF. mcg, micrograms.

Safety.

One or more clinically adverse events were reported in 23 patients, for a total of 31 events. CNS-related experiences were the most frequently reported adverse events (48%), followed by gastrointestinal disturbances (32%). In two patients of group B, the adverse events were serious and led to withdrawal of the participants from the study. They included generalized seizures followed by coma in one patient with undetermined meningitis and AIDS and a status epilepticus in a patient with severe cryptococcal meningitis, increased intracranial pressure, and vomiting. These two patients had basic illnesses which could themselves elicit generalized seizures. It is debatable whether rufloxacin could have contributed to the proconvulsant potential of these two CNS infections. Overall, for 9 of 31 adverse events, a causal relationship to rufloxacin treatment was considered possible.

No significant changes in biochemical or hematological variables were observed during rufloxacin treatment other than those reflecting the underlying meningitis or other concurrent medical conditions.

DISCUSSION

The results of the present study indicate that rufloxacin penetrates into the CSF of patients with inflamed or noninflamed meninges. The penetration of the drug into CSF, expressed as the ratio of the drug concentration in CSF to the plasma drug concentration, ranged from 0.57 to 0.84, depending on the study group. In patients with meningitis, a good penetration into CSF appeared to persist beyond the acute phase of illness; day 1 samples for groups B and C were taken on average 2.4 and 3.5 days, respectively, after hospital admission. CSF rufloxacin concentrations measured at day 7 were therefore for samples taken at a late stage of illness, when the meninges were healing. This was supported by a significant reduction of leukocyte counts C and protein content in the CSF noted on day 7. The observed increase in the CSF/plasma drug concentration ratio after multiple dosing, compared to day 1 only dosing, in groups B and C, although not statistically significant, suggests that the permeability of the meningeal barrier to rufloxacin is mainly dependent on the compound's lipophilicity rather than the state of the meningeal barrier itself. The CSF/plasma drug concentration ratios after the first administration of rufloxacin and at steady state in patients with uninflamed meninges were similar (groups A_1d and A_7d).

A higher degree of penetration was observed following multiple dosing and in patients with highly inflamed meninges, but the differences between study groups never reached statistical significance. The lack of a positive relationship between the CSF/plasma rufloxacin concentration ratio and indices of meningeal permeability in patients with purulent meningitis supports the observation that rufloxacin enters the CSF even in the presence of a normal blood-CSF barrier. Similar results have previously been obtained with ofloxacin (19) and rifampin (17). Passage of lipophilic antibacterial agents across capillary epithelial cells of the blood-CSF barrier is less dependent on the derangement of the barrier than is the transit of hydrophilic compounds (2). Rufloxacin penetration, expressed as the CSF/plasma drug concentration ratio measured during the acute stage of meningitis, far exceeded that of the more hydrophilic compound ciprofloxacin (29). These data agree with the findings of Jaehde et al. (13), who reported that the transport kinetics of various quinolones across the blood-brain barrier are highly inversely correlated with partition coefficients as a measure of their lipophilicity at physiologic pH.

Our study was not designed to determine the kinetics of rufloxacin CSF. A different study design, incorporating multiple plasma and CSF samplings from neurosurgical devices, would provide more pharmacokinetic data (18). Data on rufloxacin penetration into body fluids after a single oral administration have demonstrated a mean time to maximum drug concentration in plasma ± standard deviation of 3.83 ± 2.62 h (12) and mean times to maximum drug concentration in the extravascular compartment ranging from 3.5 to 11.8 h (26, 28). Especially for drugs with long elimination half-lives, a delayed time to peak concentration in CSF was observed (5). Since rufloxacin has good lipophilicity and the longest half-life of any quinolone currently available, higher concentrations in CSF would have been found if a longer sampling time had been used. As for other fluoroquinolones, rufloxacin is rapidly bactericidal, with the minimal bactericidal concentration almost always of the same magnitude as the MIC (16).

The concentrations of rufloxacin in the CSF were up to 25 times higher than the reported MICs for N. meningitidis and 6 times the MICs at which 90% of the organisms are inhibited for H. influenzae, several members of the family Enterobacteriaceae, and strains of Aeromonas and Plesiomonas spp. (16, 27), the last two being rarely involved in meningitis. Unfavorable CSF concentration-to-MIC ratios are found for Klebsiella spp., Enterobacter spp., Serratia marcescens, Pseudomonas spp., Listeria monocytogenes, and streptococci (8, 16, 27).

CNS-excitatory side effects (nervousness, sleep disturbances, hallucinations, seizures, etc.) observed following the administration of quinolones are related to the ability of these drugs to cross the blood-brain barrier (4, 5, 9). In our study, there was no evidence of any relationship between CNS-related side effects and CSF drug levels. In view of the random nature of the observed CNS side effects, one can conclude that they were unrelated to rufloxacin. The most frequent CNS-related adverse reaction was headache, which was reported in 10 patients, 4 of whom belonged to group A1d, i.e., who received only a single 400-mg dose of rufloxacin.

In conclusion, the results of the present study suggest that rufloxacin might be a useful agent for the treatment of CNS infections caused by highly susceptible pathogens, even in patients exhibiting a minimal impairment of the blood-CSF and blood-brain barriers. Penetration through noninflamed meninges suggests that the drug could be used for prophylaxis during neurosurgical procedures or to prevent relapses of acute bacterial meningitis. Prospective comparative clinical trials to investigate this aspect are warranted.