[¹⁸F]Ciprofloxacin, a New Positron Emission Tomography Tracer for Noninvasive Assessment of the Tissue Distribution and Pharmacokinetics of Ciprofloxacin in Humans

Abstract

The biodistribution and pharmacokinetics of the fluorine-18-labeled fluoroquinolone antibiotic [¹⁸F]ciprofloxacin in tissue were studied noninvasively in humans by means of positron emission tomography (PET). Special attention was paid to characterizing the distribution of [¹⁸F]ciprofloxacin to select target tissues. Healthy volunteers (n = 12) were orally pretreated for 5 days with therapeutic doses of unlabeled ciprofloxacin. On day 6, subjects received a tracer dose (mean injected amount, 700 ± 55 MBq, which contained about 0.6 mg of unlabeled ciprofloxacin) of [¹⁸F]ciprofloxacin as an intravenous bolus. Thereafter, PET imaging and venous blood sampling were initiated. Time-radioactivity curves were measured for liver, kidney, lung, heart, spleen, skeletal muscle, and brain tissues for up to 6 h after radiotracer administration. The first application of [¹⁸F]ciprofloxacin in humans has demonstrated the safety and utility of the newly developed radiotracer for pharmacokinetic PET imaging of the tissue ciprofloxacin distribution. Two different tissue compartments of radiotracer distribution could be identified. The first compartment including the kidney, heart, and spleen, from which the radiotracer was washed out relatively quickly (half-lives [t_1/2s], 68, 57, and 106 min, respectively). The second compartment comprised liver, muscle, and lung tissue, which displayed prolonged radiotracer retention (t_1/2, >130 min). The highest concentrations of radioactivity were measured in the liver and kidney, the main organs of excretion (standardized uptake values [SUVs], 4.9 ± 1.0 and 9.9 ± 4.4, respectively). The brain radioactivity concentrations were very low (<1 kBq · g⁻¹) and could therefore not be quantified. Transformation of SUVs into absolute concentrations (in micrograms per milliliter) allowed us to relate the concentrations at the target site to the susceptibilities of bacterial pathogens. In this way, the frequent use of ciprofloxacin for the treatment of a variety of infections could be corroborated.

Fluoroquinolone antibiotics are one of the most frequently prescribed classes of antibiotics that are increasingly used for the treatment of a variety of infections due to their broad antibacterial spectra and utility against a variety of infections (19). Within the next years, fluoroquinolones are expected to claim more than 30% of the $33 billion global antibiotics market. In 1987 the first oral broad-spectrum fluoroquinolone, i.e., ciprofloxacin, was approved by the Food and Drug Administration. Since its approval, ciprofloxacin has been extensively studied, more than 250 million patients have been treated worldwide, and its safety profile is well documented in more than 32,000 publications (28).

Only a few studies, however, have been performed to directly measure the pharmacokinetics (PKs) of ciprofloxacin in target organ tissues, such as the urinary and gastrointestinal tracts, lung tissue, and soft tissues (4, 8). This is because organ ciprofloxacin concentrations are usually difficult to measure by the standard methodology used for PK analysis (21, 22). Consequently, characterization of the organ ciprofloxacin distribution and ciprofloxacin PKs has mainly been restricted to indirect estimates of the drug concentrations in plasma or bodily secretions. However, such surrogates for the true concentrations in tissue compartments do not permit identification of subinhibitory concentrations in tissue, which result from impaired distributions from plasma to tissue and which might lead to therapeutic failure and/or the emergence of resistant bacterial strains. Thus, a method that allows direct quantification of drug distribution and determination of the PKs within body compartments that are difficult to access would be most welcome.

Positron emission tomography (PET) is a noninvasive nuclear imaging technique that allows the concentrations of drug molecules in the tissues of virtually all organs to be measured with good spatial and temporal resolutions (3, 13). PET possesses a sensitivity in the lower picomolar range but requires the drug of interest to be radiolabeled with an appropriate positron-emitting radioisotope, such as carbon-11 (¹¹C; half-life, 20.4 min) or fluorine-18 (¹⁸F; half-life, 110 min). Owing to its longer physical half-life, ¹⁸F is preferred for imaging for PK analysis since it allows longer imaging durations. However, a major limitation stems from the fact that many drugs do not contain fluorine in their molecular structures, and incorporation of ¹⁸F would therefore alter the PK and pharmacodynamic properties of the drug molecule.

Fluoroquinolones are fluorine-containing drugs, and ¹⁸F labeling therefore affords a radiotracer that is chemically identical to its unlabeled counterpart. A series of previous studies has used PET to describe the PKs of ¹⁸F-labeled fluoroquinolones (i.e., fleroxacin, lomefloxacin, and trovafloxacin) in the peripheral tissues of healthy volunteers and patients with bacterial infections (11, 12, 14, 25). PET imaging therefore has the potential to quantify by a noninvasive procedure the concentrations at the target site directly within the drug's biophase.

Recently, ciprofloxacin, the benchmark compound with which all other fluoroquinolones are compared, was radiolabeled with ¹⁸F (20). The present study was designed to describe the biodistribution and PKs in tissue of the newly developed radiotracer, [¹⁸F]ciprofloxacin (Fig. 1), for the first time in humans by a noninvasive procedure. Special interest was paid to characterization of the [¹⁸F]ciprofloxacin distribution to those organs usually not easily accessible to standard PK analysis.

FIG. 1.

Chemical structure of [¹⁸F]ciprofloxacin.

MATERIALS AND METHODS

The study was approved by the local ethics committee. All volunteers were given a detailed description of the study, and their written consent was obtained. The study was performed in accordance with the Declaration of Helsinki and the Good Clinical Practice Guidelines of the European Commission.

Healthy volunteers.

Twelve healthy male volunteers (mean age, 30 ± 8 [standard deviation {SD}] years; mean weight, 83 ± 12 kg; mean height, 180 ± 5 cm) were included in the present study. Each volunteer was subjected to a screening examination that included medical history, physical examination, 12-lead electrocardiogram, blood pressure, heart rate, complete blood count, urinalysis, urine drug screen, clinical blood chemistry, blood coagulation tests, HBs antigen test, and human immunodeficiency virus antibody test. Subjects were excluded if they had taken any prescription medication or over-the-counter drugs within the 2 weeks prior to the study or if they had undergone any diagnostic analysis with radioactive tracers or X rays during the 6 months preceding the study.

Study design and study medication.

This study was carried out as a descriptive, exploratory, single-center, nonrandomized trial. Subjects were orally administered unlabeled ciprofloxacin (Ciproxin 250-mg Filmtabletten; Bayer Austria, Vienna, Austria) twice daily over 5 days plus once on the study day. On the study day subjects additionally received [¹⁸F]ciprofloxacin as an intravenous bolus (injected dose, 700 ± 55 MBq, which corresponds to about 0.6 mg of unlabeled ciprofloxacin) 3 h after administration of the last oral dose of unlabeled ciprofloxacin.

Synthesis and quality control of [¹⁸F]ciprofloxacin.

[¹⁸F]ciprofloxacin (Fig. 1) was prepared by a two-step radiosynthesis method as described previously (20). The purified radiotracer was formulated in 7 ml of sterile phosphate buffer (0.2 M, pH 5.5 to 6.0). For intravenous administration, an aliquot of the product solution was diluted with physiological saline solution to an injection volume of 10 ml. Prior to injection into volunteers, quality control was performed for each batch of [¹⁸F]ciprofloxacin synthesized, as described previously (20). As assessed by analytical high-pressure liquid chromatography (HPLC) and thin-layer chromatography, the radiochemical purity of [¹⁸F]ciprofloxacin was >99% and the chemical purity exceeded 95%. The specific radioactivity of [¹⁸F]ciprofloxacin at the time of tracer injection was about 400 MBq/μmol. Therefore, an injected amount of 700 MBq of [¹⁸F]ciprofloxacin corresponded to approximately 1.8 μmol or 0.6 mg of unlabeled ciprofloxacin.

Study protocol.

Each subject received a standard oral treatment with 250 mg of unlabeled ciprofloxacin twice daily for 5 days before PET imaging. On day 6, prior to administration of the final oral dose, a venous catheter was placed in each arm (one for infusion of radiolabeled drug and one for blood sampling). Three hours later [¹⁸F]ciprofloxacin (700 ± 55 MBq) was administered as an intravenous bolus. Serial PET imaging and venous blood sampling were initiated at the start of the bolus infusion and were continued for 6 h. Due to the limited axial field of view (FOV) of the PET camera and the short physical half-life of ¹⁸F, detailed PK studies were performed for specific groups of organs in different sets of subjects. Extracranial organs were studied in eight subjects; the drug distribution in brain tissue was studied in four subjects. For the study of extracranial organs, two body regions were imaged; the first region included the heart, lung, and skeletal muscles (musculus triceps and musculus pectoralis); the second region included the liver, spleen, and kidney. Each subject was positioned supine on the imaging bed of the PET camera. In order to correct for the attenuation of photons by tissue, a transmission scan with two 400-MBq ⁶⁸Ge pin sources was recorded for 10 min prior to radiotracer injection. The imaging protocol consisted of the following frame sequence: during the first 30 min after injection, dynamic images of the selected organ groups were acquired (frame lengths, five times for 1 min each and five times for 5 min each). Then, a transmission scan followed by three consecutive whole-body scans (over four axial FOVs each, with a scan duration of 5 min per FOV) was performed. The subjects were then allowed to resume their usual activities. Approximately 2.5 and 5 h after tracer injection, the subjects were repositioned in the PET camera. Four whole-body scans (four axial FOVs of 6 min per FOV) were recorded after 2.5 h, and one whole-body scan (four FOVs of 10 min per FOV) was recorded after 5 h. Prior to each repositioning, a transmission scan was performed. PET images were acquired with an Advance PET scanner (General Electric Medical Systems, Wukesha, Wis.) with a transversal FOV of 55 cm and an axial FOV of 15 cm.

Data analysis.

Reconstruction of the PET data was performed by means of iterative reconstruction by the ordered subsets-expectation-maximization method with 28 subsets and two iterations. The loop filter (Gaussian) was set to a full width at half maximum (FWHM) of 4.3 mm, and a postfiltering algorithm of 6.00 mm of FWHM was applied. Attenuation correction was performed by using the manufacturer's segmentation algorithm for transmission data. For each of the selected organs, regions of interest (ROIs) were drawn in the reconstructed image which best represented the anatomy of the respective organ. For the heart, a ROI was drawn on a transaxial slice and placed over the left ventricular myocardium. For the lung, ROIs drawn on the left and right sides were pooled. The ROI for the liver did not include the gall bladder. These ROIs were then transferred to all other images of the time sequence; and radioactivity concentrations (in kilobecquerels per milliliter), corrected for radioactive decay from the time of tracer injection, were calculated. Since the density of all tissues studied (except for lung tissue) was approximately 1 g · ml⁻¹, concentrations expressed as kilobecquerels per milliliter were considered to be equal to concentrations expressed as kilobecquerels per gram of tissue. For lung tissue, radioactivity concentrations were corrected for a density of 0.26 ± 0.03 g · ml⁻¹ (15). Radioactivity concentrations were normalized to the injected radiotracer amount and expressed as standardized uptake values (SUVs). An SUV is defined as the local radioactivity concentration (in kilobecquerels per gram) divided by the amount radioactivity administered per gram of body weight (in kilobecquerels per gram). The radioactivity concentration data were combined to provide time-radioactivity curves (TACs) for the whole observation period. Since the whole-body scans of individual subjects were not acquired at exactly the same time points after tracer injection, TACs were aligned by linear interpolation. For each tissue studied, the TACs of all subjects were averaged. Error estimates were obtained by means of error propagation. Local variations in the radioactivity concentrations measured in the tissues studied, as well as intersubject differences upon calculation of the mean, were taken into account. The average TACs (average SUV ± SD) of the respective tissues were used to calculate the following PK parameters by using the Kinetica 2000 software package (version 3.0; InnaPhase Corporation, Philadelphia, Pa.): maximum concentration of drug in serum (C_max; in average SUVs ± SD), the time to C_max (T_max; in minutes), the terminal elimination half-life (t_1/2; in minutes), and the area under TAC from time zero to infinity (AUC_tot; in minutes). Plateau concentrations (C_plateau; i.e., the mean radioactivity concentration from 2 to 6 h after tracer injection) were calculated for the target tissues of antibiotic treatment (14).

Blood analysis.

Venous blood samples (9 ml) were collected at 1, 2, 5, 10, 20, 25, 30, 45, 60, 90, 120, 240, and 360 min following radiotracer injection. Plasma was obtained by centrifugation at 1,600 × g for 10 min. The radioactivity concentrations in aliquots of whole blood and plasma (1 ml each) were measured in a Packard Cobra II auto-gamma counter (Packard Instrument Company, Meriden, Conn.) that had been cross-calibrated with the PET camera. Radioactivity counting rates were corrected for decay from the time of tracer injection and normalized to the injected radiotracer amount in order to generate TACs for plasma and whole blood. Select plasma samples obtained at 1 to 120 min after tracer injection were analyzed for radiolabeled metabolites by HPLC. The HPLC system consisted of a Rheodyne 9125 injector equipped with a 1-ml sample loop (Rheodyne, Rohnert Park, Calif.), a Merck Hitachi D-6000 interface (Hitachi High Technologies, San Jose, Calif.), and a Merck Hitachi L-6220 pump. For detection, a Merck Hitachi L-4000 UV detector (wavelength, 280 nm) in series with a Packard Radiomatic Flo-one Beta Flow scintillation analyzer (Perkin-Elmer Life Sciences Inc., Boston, Mass.) were used. A Waters μBondapak C₁₈ column (300 by 3.9 mm; pore size, 10 μm) was eluted with a mixture of 10 mM aqueous phosphoric acid (A) and absolute ethanol (B) at a flow rate of 1.5 ml · min⁻¹. The following binary gradient time program was used: 0 to 5 min, A:B at 85:15 (vol/vol) isocratic; 5 to 9 min, A:B at 85:15 to 75:25; 9 to 12 min, A:B at 75:25 isocratic; 12 to 14 min, A:B at 75:25 to 85:15. On this system [¹⁸F]ciprofloxacin eluted with a retention time of about 7 min. Plasma samples (2 ml) were first counted in a gamma counter and then treated with methanol (4 ml) and a small amount of unlabeled ciprofloxacin hydrochloride. The samples were centrifuged (2,000 × g, 10 min) to precipitate plasma proteins. The supernatant was concentrated to dryness, redissolved in a mobile phase for analytical HPLC (1 ml), and counted in a gamma counter to determine the amount of radioactivity recovered. Then, the solution was injected into the HPLC system and the eluate was monitored for UV absorption and radioactivity.

RESULTS

All study procedures were well tolerated by all volunteers. There was no serious adverse event and no side effects from the administration of the study medication.

Distribution and PK parameters of [¹⁸F]ciprofloxacin in the brain.

There was a very low level of accumulation of [¹⁸F]ciprofloxacin in the central nervous system, with radioactivity concentrations below 1 kBq · ml⁻¹, which could not be reliably quantified. The distribution of the radioactivity mainly mirrored the tracer distribution in the brain vessels (Fig. 2).

FIG. 2.

PET images (coronal, sagittal, and transaxial slices) acquired 0 to 30 min after intravenous injection of 700 MBq of [¹⁸F]ciprofloxacin in one healthy male volunteer. Dark shades of gray represent high radioactivity concentrations. Radioactivity is present only in the vascular system, whereas intracerebral accumulation is minimal.

Distribution and PK parameters of [¹⁸F]ciprofloxacin in peripheral organs.

Following intravenous injection of [¹⁸F]ciprofloxacin, radioactivity uptake was recorded in several tissues. TACs for the tissues were generated for up to 6 h following tracer injection. The TACs of the tissues investigated are presented in Fig. 3; and the values calculated for the PK parameters (C_max, T_max, t_1/2, and AUC_tot) are shown in Table 1. In some tissues, such as the kidney, heart, and spleen, a rapid T_max (<13 min) followed by fast radiotracer washout (t_1/2s, 68, 57, and 106 min, respectively) was observed. Other tissues, including liver, muscle, and lung tissue, displayed prolonged radiotracer retention (t_1/2s, >130 min). The highest C_maxs of radioactivity were measured in the liver and kidney, with SUVs of 4.9 ± 1.0 and 9.9 ± 4.4, respectively. C_plateaus, expressed as SUVs and absolute concentrations (in micrograms per milliliter), are shown in Table 2. Figure 4 depicts the radioactivity accumulation in the liver and its subsequent excretion via bile and the bowels at 45, 160, 190, and 240 min after bolus administration.

FIG. 3.

Mean TACs for various organs of 12 healthy volunteers determined by PET imaging after injection of an intravenous bolus of [¹⁸F]ciprofloxacin. The concentrations in tissues are given as average SUVs ± SDs. For the spleen and kidney, TACs could be generated only for the first 30 min after injection, since radiotracer uptake was negligible in later PET images. For the sample sizes for various organs, see footnote a of Table 1.

TABLE 1.

Mean values of Pk parameters for select tissues determined by PET imaging after intravenous bolus injection of 700 MBq of [¹⁸F]ciprofloxacin in 12 healthy volunteers

Tissuea	Cmax (avg SUV ± SD)	Tmax (min)	T1/2 (min)	AUCtot (min)b
Liver	4.9 ± 1.0	12.5	142	718
Kidney	9.9 ± 4.4	3.5	68	442
Heart	1.8 ± 0.5	12.5	57	135
Spleen	1.2 ± 0.3	3.5	106	149
Pectoral muscle	0.7 ± 0.2	115.5	136	251
Triceps	0.7 ± 0.3	55.5	576	447
Lung	1.0 ± 0.3c	75.5	207	348

^aTissue sample sizes (n volunteers) were as follows: for liver, dynamic, n = 5; whole body, n = 6; for kidney and spleen, dynamic, n = 4; for heart, dynamic, n = 6; whole body, n = 7; for pectoral muscle, triceps, and lung, dynamic, n = 4; whole body, n = 8. Dynamic denotes the first 30 min of the TAC.

^bThe units for AUC_tot (minutes) are different from conventional units because radioactivity uptake by tissue was normalized for the injected dose per body weight and is presented as SUV.

^cLung radioactivity concentrations were corrected for a lung tissue density of 0.26 g·ml⁻¹ (15).

TABLE 2.

[¹⁸F]ciprofloxacin C_plateaus expressed as SUVs and absolute ciprofloxacin concentrations in select target tissues^a

Tissue	Cplateau
	Avg SUV ± SD	Avg μg · ml−1 ± SD
Liver	1.1 ± 0.3	3.4 ± 0.9
Lung	0.6 ± 0.1	1.7 ± 0.4
Pectoral muscle	0.5 ± 0.2	1.5 ± 0.5
Triceps	0.4 ± 0.2	1.3 ± 0.6
Kidneyb	3.1 ± 3.6	9.5 ± 10.8

^aThe absolute concentrations were estimated by multiplying the final oral dose of ciprofloxacin divided by the average body weight by the corresponding SUV, as described previously (14).

^bFor the kidney the radioactivity concentration at 30 min after tracer injection was used instead of C_plateau.

FIG. 4.

Coronal whole-body PET images obtained at 45, 160, 190, and 240 min after intravenous injection of 681 MBq of [¹⁸F]ciprofloxacin in one healthy male volunteer. Dark shades of gray represent high radioactivity concentrations. The progressing hepatobiliary and renal excretion of the radiotracer is shown over time.

Blood PKs of [¹⁸F]ciprofloxacin.

The mean TACs of [¹⁸F]ciprofloxacin in plasma and whole blood are depicted in Fig. 5. The profiles of both curves were almost congruent over the entire observation period. A t_1/2 of 225 min was calculated. Select plasma samples obtained at 1 to 120 min after radiotracer injection were analyzed after protein precipitation by radio-HPLC. The amount of radioactivity recovered by the assay used was approximately 80%. No radiolabeled metabolites were detected in plasma up to 2 h after tracer injection. Figure 6 shows a representative HPLC chromatogram of an extract of plasma collected 120 min following the administration of [¹⁸F]ciprofloxacin.

FIG. 5.

Mean TACs (average SUV ± SD) for plasma and whole blood from 12 healthy volunteers after intravenous administration of 700 ± 55 MBq of [¹⁸F]ciprofloxacin.

FIG. 6.

HPLC chromatogram of a plasma extract obtained at 120 min following injection of 681 MBq of [¹⁸F]ciprofloxacin into one healthy male volunteer. A Waters μBondapak C₁₈ HPLC column (300 by 3.9 mm; pore size, 10 μm) eluted with a mixture of 10 mM aqueous phosphoric acid and ethanol was used (see Materials and Methods for a detailed description of the HPLC system).

DISCUSSION

The present study was designed to evaluate the utility of [¹⁸F]ciprofloxacin and PET for the noninvasive measurement of the tissue ciprofloxacin distribution and PKs of ciprofloxacin in humans. Recently, a two-step procedure for the radiosynthesis of [¹⁸F]ciprofloxacin (Fig. 1) was developed (20). Ciprofloxacin represents an ideal candidate drug for labeling with ¹⁸F, since it contains fluorine in its native structure and since the majority of ciprofloxacin is excreted as unchanged drug (17). An important prerequisite for a PET tracer is metabolic stability, since PET measures the total radioactivity concentrations in tissue and is not able to discern different radiolabeled chemical entities, such as radiolabeled metabolites. Owing to the well-described metabolic stability of ciprofloxacin, the radioactivity within designated ROIs was expected to reflect mainly the unchanged parent molecule.

Following pretreatment with an oral dose of 250 mg of ciprofloxacin twice daily to simulate antibiotic treatment for moderate bacterial infections, a tracer dose of 700 MBq of [¹⁸F]ciprofloxacin, which contained about 0.6 mg of unlabeled ciprofloxacin, was given as an intravenous bolus. If dose linearity between the tracer and the therapeutic dose of ciprofloxacin is assumed, the distribution of the radiolabeled antibiotic in organ tissue was considered to be comparable to the distribution achieved after standard dosing (2). Due to exposure to high radioactive dose rates in the gastrointestinal tract and the difficulties with the preparation of an oral formulation of radiolabeled drug, PET tracers are usually not administered orally. Therefore, the intravenous route of administration was chosen for [¹⁸F]ciprofloxacin. A similar combination of oral predosing and intravenous radiotracer application was previously used in a PET-PK study with [¹⁸F]trovafloxacin (14).

The pattern of [¹⁸F]ciprofloxacin distribution observed in the present study was in accordance with the well-described tissue ciprofloxacin distribution (Fig. 4) (9, 26). Two different tissue compartments could be identified: a rapidly equilibrating compartment with fast radiotracer washout that included the heart, spleen, and kidney and a second compartment with prolonged radiotracer retention that comprised the liver, muscle, and lung tissue. Particularly high radioactivity concentrations, with up to 20-fold higher C_maxs compared with those in other tissues, were observed in the liver and the kidney, the two organs known to be responsible for more than 90% of drug elimination (9). Nonrenal clearance mechanisms account for approximately one-third of ciprofloxacin elimination and comprise a combination of hepatic metabolic degradation, biliary excretion, and transluminal secretion across the enteric mucosa (9). Nonrenal radiotracer clearance is reflected by high gastrointestinal tract radioactivity concentrations (Fig. 4). The relatively long t_1/2 (142 min) for liver tissue could be explained by a first-pass effect of one of the four ciprofloxacin metabolites, which has been described previously (17). Furthermore, renal clearance mechanisms are reported to account for approximately two-thirds of total ciprofloxacin elimination (9). In the present study, renal radiotracer clearance appeared to proceed faster than hepatobiliary clearance (t_1/2s, 68 and 142 min for the kidney and the liver, respectively), resulting in negligible radioactivity uptake by kidney tissue at later imaging time points (Fig. 4).

The [¹⁸F]ciprofloxacin concentrations were very low in the brain and could therefore not be reliably quantified. The radioactivity signal measured mainly represented scatter from surrounding blood vessels rather than radiotracer actually taken up by brain tissue (Fig. 2). The inability to quantify concentrations in brain tissue reflects the known poor penetration of fluoroquinolone antibiotics through the blood-brain barrier. As fluoroquinolones are known substrates of P-glycoprotein, a drug transporter expressed in high concentrations at the blood-brain barrier (10), the low level of uptake of the radiotracer by the brain observed in the present study could be explained by P-glycoprotein-mediated brain efflux. However, it should be noted that our PET study was performed with healthy volunteers. In patients, the integrity of the blood-brain barrier might be disrupted by disease processes such as inflammation that lead to altered uptake and potential neurological side effects (1). Importantly, no radioactivity uptake by bone was observed, which indicates the absence of in vivo defluorination of [¹⁸F]ciprofloxacin and the release of free [¹⁸F]fluoride, which is known to accumulate in bone. This finding demonstrates the high in vivo stability of [¹⁸F]ciprofloxacin.

When mean TACs for plasma and whole blood were compared, both profiles were nearly congruent over the entire observation period (Fig. 5). This finding indicates the uptake and retention of [¹⁸F]ciprofloxacin by cellular components of blood, which is consistent with the known accumulation of ciprofloxacin within neutrophil granulocytes and monocytes (5, 27). Since blood cells account for about 50% of the blood volume, the present curves indicate that approximately 50% of the total radioactivity in blood is bound to blood cells. When extracts of plasma obtained up to 2 h following tracer injection were analyzed by radio-HPLC, unmetabolized parent compound accounted for the entire radioactivity eluting from the HPLC column (Fig. 6). Even though the possibility cannot be excluded that very small amounts of radiolabeled metabolites, amounts below the limit of detection of the HPLC radioactivity detector (approximately 0.5 kBq), were present, these findings further corroborate the known metabolic stability of ciprofloxacin and the absence of circulating drug metabolites (17). Consequently, it can be assumed that the radioactivity measured in all body tissues (except the excretory organs) reflected unmetabolized [¹⁸F]ciprofloxacin.

Beyond the description of the distribution in tissue and the PKs of a drug, PET has the potential to quantify tissue drug concentrations. In the present study, it was assumed that the late part of the TACs for tissue (from 2 to 6 h after tracer injection) reflected the ciprofloxacin concentrations achieved after oral predosing with unlabeled ciprofloxacin. Moreover, it was assumed that the final oral dose of ciprofloxacin administered was completely absorbed. Therefore, the C_plateaus of radioactivity were multiplied by the final oral dose of ciprofloxacin per body weight, as described previously in a PET-PK study with [¹⁸F]trovafloxacin (14), in order to give an estimate of total tissue ciprofloxacin concentrations in micrograms per milliliter (Table 2). In the present study, ciprofloxacin concentrations ranged from 1.3 ± 0.6 μg · ml⁻¹ in skeletal muscle to 9.5 ± 10.8 μg · ml⁻¹ in kidney tissue after 5 days of pretreatment with a rather low dose of 250 mg of ciprofloxacin (Table 2).

It must be emphasized that PET measurements yield total radiotracer concentrations per milliliter of tissue and cannot differentiate between intra- and extracellular concentrations. Measurement of the total concentrations in tissue, however, may be misleading, since only the free, extracellular drug concentration exerts antibacterial activity in most cases. Intracellular ciprofloxacin kinetics have been studied previously and have demonstrated rapid antibiotic uptake, with an intracellular concentration/extracellular concentration (IC/EC) ratio of about 4, combined with a prolonged intracellular retention (16). Consequently, from the IC/EC ratio, the extracellular drug fraction can be estimated and can be used to correlate concentrations in tissue and antibacterial activity. Another suitable approach to the selective assessment of fluoroquinolone concentrations in different tissue compartments would be a combination of PET and the microdialysis technique. Microdialysis has previously been applied to the measurement of ciprofloxacin concentrations in the extracellular compartments of various tissues in humans (7, 8) and is based on sampling of the unbound fraction of drugs from the extracellular space of virtually all tissues with a semipermeable membrane located at the tip of a small probe. A comparison of the PKs of ciprofloxacin in skeletal muscle measured by microdialysis (7) with the TACs measured in the present PET study (Fig. 3) demonstrated a similar kinetic profile, with T_maxs attained within the first hour after drug administration. In future studies, a combination of both techniques could enable the in vivo measurement of the intracellular PKs of fluoroquinolones, which are relevant in the treatment of infections caused by intracellular bacterial pathogens.

When the estimated extracellular concentrations were correlated to the MICs at which 90% of isolates are inhibited (MIC₉₀s) for relevant extracellular bacteria, the concentrations of ciprofloxacin in the kidneys were sufficient for the optimum treatment of urinary tract infections caused by nearly all pathogens with the exception of Enterococcus spp., which are resistant to fluoroquinolones (18, 24). Compared to the concentrations in the kidney, our results show about 10-fold lower concentrations of ciprofloxacin in muscle and lung tissues. These values still reach excellent concentration/MIC ratios for relevant gram-negative bacteria, such as Haemophilus influenzae (MIC₉₀ = 0.02 μg · ml⁻¹), Moraxella catarrhalis (MIC₉₀ = 0.12 μg · ml⁻¹), and Klebsiella spp. (MIC₉₀ = 0.14 μg · ml⁻¹). However, they might be inadequate for the optimum treatment of soft tissue and respiratory tract infections caused by Pseudomonas aeruginosa (MIC₉₀ = 0.65 μg · ml⁻¹) and most gram-positive bacteria, such as Staphylococcus spp. and Streptococcus spp. (MIC₉₀s = 0.5 to 2 μg · ml⁻¹) (18, 24).

The results of the present study only describe the distribution of [¹⁸F]ciprofloxacin in healthy volunteers and do not take into account the distribution into infected tissues. A recent microdialysis study found no significant difference in the penetration of ciprofloxacin into inflamed and unaffected tissues in diabetic foot lesions (23). Another study, however, reported increased radiotracer uptake and retention by various types of infected tissues compared to that by healthy tissues by using a radiolabeled analogue of ciprofloxacin, i.e., [^99mTc]ciprofloxacin, and single-photon emission computed tomography (6). The authors attributed the observed increase in concentrations in infected tissues to specific binding of the radiotracer to bacterial cells and proposed the use of [^99mTc]ciprofloxacin as an infection imaging tracer (6). Owing to complex formation with bacterial DNA and DNA gyrase, [¹⁸F]ciprofloxacin might be retained by bacterial cells in vivo and might thus allow the imaging of localized infectious processes by PET. However, future studies on the PKs of [¹⁸F]ciprofloxacin in patients with bacterial infections need to be performed to confirm this assumption by comparing the amounts of radiotracer uptake by healthy and infected tissues.

In conclusion, the first application of [¹⁸F]ciprofloxacin in humans has demonstrated the safety and utility of this newly developed radiotracer for PET imaging of the PKs of ciprofloxacin and its distribution in tissue. The frequent use of ciprofloxacin monotherapy for the treatment of urinary tract infections was corroborated by relating the target site PKs in various organs, which are usually not accessible by standard methodologies, to the susceptibilities of bacterial pathogens. PET, whether it is used alone or in combination with other techniques, has the potential to become a powerful tool for the assessment of drug biodistribution and will exert an increasingly important impact on drug development in the future.