Comparison of Ofloxacin and Norfloxacin Concentration in Prostatic Tissues in Patients Undergoing Transurethral Resection of the Prostate

J. Chen, R.R.L. Chen, and H.S. Huang

COMPARISON OF OFLOXACIN AND NORFLOXACIN CONCENTRATION IN PROSTATIC TISSUES IN PATIENTS UNDERGOING TRANSURETHRAL RESECTION OF THE PROSTATE

Jun Chen, Russel Rhei-Lon Chen,1 and Ho-Shiang Huang

Background and purpose: To compare the concentrations of two fluoroquinolones, (J Formos Med Assoc ofloxacin (OFLX) and norfloxacin (NFLX), in the prostate glands of patients who 2001;100:548–52) underwent transurethral resection of the prostate (TUR-P) after oral ingestion of both drugs for surgical prophylaxis. Key words: Methods: Ten patients with clinical symptoms of benign prostatic hyperplasia under- ofloxacin going TUR-P received 200 mg of both drugs per os simultaneously 2 hours before norfloxacin surgery. The concentrations of the drugs in the serum and prostate at the time of prostatitis surgery were measured by high performance liquid chromatography. Patients’ clinical minimum inhibitory characteristics were evaluated, including findings from transrectal ultrasonography of concentration the prostate, prostate specific antigen concentration, renal function tests, and post- prostatectomy operative status. Results: Two hours after administration, the mean serum concentration of OFLX was 4.14 ± 0.64 mg/L (range 0.27–6.37) and of NFLX was 1.10 ± 0.22 mg/L (range 0.02–2.1). The concentration of ORLX in prostatic tissue was 4.10 ± 0.79 µg/g (range 1.70–6.37) and of NFLX was 2.22 ± 0.57 µg/g (range 0.63–4.35). The ratio of the prostatic tissue concentration (P) to the serum concentration (S) was 2.11 for OFLX and 5.71 for NFLX. The concentrations of both drugs exceeded the minimum inhibitory concentration (MIC) for most gram-negative organisms, but NFLX may be unable to

exceed the MIC90 of Enterobacter cloacae in some individuals. Leukocytosis and spiking fever developed after TUR-P in two of the 10 patients. Conclusions: Concentrations of OFLX were higher in both serum and prostatic adenoma tissues than those of NFLX (p < 0.001), while NFLX had better penetration into the prostate (P/S ratio) (p < 0.001). The results of this study indicated that the concentrations of both of these drugs exceeded the MIC for most gram-negative organisms.

Antibacterial chemoprophylaxis for urinary tract infec- of a single intravenous dose of pefloxacin (800 mg), tion (UTI) during and after transurethral resection of the prostatic tissue concentrations were found to be well prostate (TUR-P) has long been controversial. Through above the MICs of pefloxacin for the bacteria commonly administration of broad-spectrum antibiotics over a short causing acute and chronic prostatitis [4]. preoperative period or as a single preoperative dose, the Ofloxacin (OFLX) is a pyridonecarboxylic acid postoperative rates of UTI have decreased to less than derivative with a broad spectrum of antibacterial activ- 10% [1, 2]. The newer quinolones almost completely ity in vitro and in vivo [5]. OFLX has been shown to be cover the spectrum of bacteria causing UTI and show a good choice in the treatment of chronic bacterial good results in the treatment of uncomplicated and prostatitis [6]. Norfloxacin (NFLX) is a lipid-soluble complicated UTI [3]. After preoperative administration weak organic acid that binds to plasma proteins to a low

Departments of Urology and 1Pharmacy, College of Medicine, National Taiwan University, Taipei. Received: 16 January 2001. Revised: 5 February 2001. Accepted: 10 July 2001. Reprint requests and correspondence to: Dr. Jun Chen, Department of Urology, National Taiwan University, 7 Chung-Shan South Road, Taipei, Taiwan.

548 J Formos Med Assoc 2001 • Vol 100 • No 8 Ofloxacin and Norfloxacin Concentrations in Serum and Prostate Adenoma extent [7]. Because recurrent UTI in men was thought in a homogenizer, to which was added 3 mL of 0.1 M to be frequently associated with prostatic infection, phosphate buffered saline (PBS). The sample was then Sabbaj et al compared the efficacy of NFLX to co- homogenized for 3 minutes. After centrifuging at trimoxazole in male patients with recurrent UTI. They 3000 rps for 10 minutes, the supernatant was used for concluded that NFLX was more efficacious than co- analysis. A mixture of 200 µL of serum or tissue trimoxazole in eradicating bacteria in the prostate supernatant, 200 µL of internal standard solution (1.5 (93% vs 67%) [7]. This study evaluated the extent of µg pipemidic acid dissolved in PBS), and 5 mL of penetration of OFLX and NFLX, the two most widely dichloromethane were vortexed for 1 minute. The used fluoroquinolones in Taiwan, in serum and pros- mixture was then centrifuged at 3,000g for 10 minutes, tatic adenoma tissue 2 hours after oral administration. and 4 mL of the dichloromethane layer was transferred The effectiveness of these agents in the prevention of to a clean evaporative tube and evaporated under an postoperative UTI was also evaluated. air gas stream. The residue was dissolved in 200 µL mobile phase (15 mM phosphate buffer pH = 7.0 containing 20 mM tetrabutylam-monium hydrogen sulfate with acetonitril (10:1, v/v)) and treated in a atients and Methods thermomixer and with ultrasound. Of the final sample, P 20 µl was injected onto the HPLC column. A Hitachi F-1000 fluorescence-detector (Ibaragi, Japan) and a Sample collection LoiChrosper 100 RP-18 (150 X 4.6 mm, 5 µm) column Ten patients with benign prostatic hyperplasia were were employed for the analysis at a flow rate of admitted to National Taiwan University Hospital 1 mL/minute at 30°C. The retention time for pipemidic (NTUH) to undergo transurethral resection of the acid (internal standard) was 3.4 minutes, for OFLX was prostate. None of the patients had received OFLX or 5.14 minutes, and for NFLX was 9.5 minutes. Concen- NFLX within the preceding 72 hours. Every patient trations of OFLX and NFLX obtained from the received a single oral dose of OFLX (200 mg) and serum were expressed as mg per L and from prostatic NFLX (200 mg) simultaneously 2 hours before stan- specimens were expressed as µg per g of prostate dard TUR-P was performed. Before administering these tissue. Values were recorded as the mean of duplicate two drugs, 10 mL of venous blood was collected, and measurements for each specimen. Differences were another 5 mL of blood was collected 2 hours after assessed using a paired t-test, with a p value of less than administration. Blood cell count (CBC), urinary 0.05 indicating statistical significance. analysis, blood urea nitrogen (BUN), and creatinine concentration (Cr) were measured before and 3 days after TUR-P. Prostate-specific antigen (PSA) concentration was measured once before surgery (TANDEM- ε–PSA, Hybritech Inc, San Diego, CA, USA). Transrectal Results ultrasonography of the prostate (TRUS-P) was performed on every patient to measure the volume and The clinical characteristics of patients before and after estimate the weight of the whole prostate and adenoma TUR-P are listed in Table 1. The pathology of the (Bruel & Kjaer Model 8551 multiplane transducer). prostatic adenoma showed nodular hyperplasia in all During TUR-P, we selected a portion of the resected patients except patient 10, who had adenocarcinoma, prostatic adenoma strips (approximately 1 g) and though his PSA concentration was 1.0 and PSA density washed them thoroughly with physiologic saline in the was 0.05. Furthermore, pathology results indicated operating theater to avoid contamination by urine. that almost all patients (9/10) also had focal inflamma- These specimens and blood samples were then stored tory cell infiltration, with pictures of chronic prostatitis at –70°C until use. Because all patients had indwelling (Table 1). Two patients (4 and 9) experienced leuko- Foley catheters after TUR-P with 24-hour normal saline cytosis (WBC = 23,200 and 12,900, respectively) after irrigation, postoperative infection was defined as eleva- TUR-P, which was resolved by additional parenteral tion of white blood cell count combined with shift-to- antibiotic treatment. Patients 7 and 10 had positive pre- left in differential count in the CBC of patients who had TUR-P urine culture results, but they recovered well fever during the postoperative period. after TUR-P. Patient 1 experienced gross hematuria and dysuria 1 month after TUR-P, and cystoscopy Measurement of drug concentration revealed bladder neck contracture. Although this Concentrations of OFLX and NFLX were measured patient recovered smoothly after the first operation, by high-performance liquid chromatography (HPLC). the pathology from the late bladder neck resection Approximately 1 g of prostatic tissue sample was placed revealed acute prostatitis.

J Formos Med Assoc 2001 • Vol 100 • No 8 549 J. Chen, R.R.L. Chen, and H.S. Huang

Table 1. The clinical characteristics of patients undergoing transurethral resection of the prostate (TUR-P)

Patient Age BW Whole Adenoma Pathology PSA PSAD BUN Cr Outcome (kg) (g)* prostate (ng/mL) (mg/dL) (mg/dL) (g)*

1 73 63 22.4 7.5 Nodular 1.3 0.058 13.9 1.6 BN contracture hyperplasia + UTI 1 month later + prostatitis 2 62 68 26.2 14.4 Nodular 2.5 0.095 22 1.1 Smooth postoperative hyperplasia course + prostatitis 3 65 70 30.5 15.0 Nodular 3.7 0.121 24 2.2 Smooth postoperative hyperplasia course + prostatitis 4 58 73.5 51.0 43.5 Nodular 12.0 0.235 14.9 1.3 Leukocytosis after hyperplasia TUR-P + prostatitis 5 64 51.3 54.5 35.0 Nodular 5.6 0.103 11 1.0 Preoperative u/c: yeast-like hyperplasia organism; smooth postop- + prostatitis erative course 6 69 53 21.3 10.0 Nodular 3.8 0.178 17 0.9 Smooth postoperative hyperplasia course + prostatitis 7 52 84 21 11.5 Nodular 4.6 0.219 10.9 1.0 Preoperative u/c: E. Coli hyperplasia + prostatitis 8 61 50 44.6 32.2 Nodular 5.3 0.119 12 1.0 Smooth postoperative hyperplasia course 9 79 36.5 39.3 19.6 Nodular 5.5 0.140 19.7 1.1 Leukocytosis after hyperplasia TUR-P + prostatitis 10 69 80 20.0 13.9 Adenocar- 1.0 0.05 14 0.9 Smooth postoperative cinoma course; cerebellar infarction + prostatitis 2 years later

Mean 65.2 62.9 33.1 20.3 4.53 0.132 15.9 1.2 SEM 7.71 15.0 4.18 3.87 0.98 0.02 1.5 0.1

*Weight estimated by transrectal ultrasonography of the prostate. BW = body weight; PSA = prostate-specific antigen; PSAD = prostate- specific antigen density; BUN = blood urea nitrogen; Cr = creatinine concentration; BN = bladder neck; UTI = urinary tract infection; u/c = urine culture.

The serum concentrations of OFLX and NFLX before data from the mean, the P/S ratio would have been 1.45 oral administration were zero. Two hours after ± 0.16 for NFLX and 0.72 ± 0.04 for OFLX, reaching administration, the mean serum concentration of OFLX significance (p < 0.001). At the same time, the mean was 4.14 ± 0.64 mg/L and of NFLX was 1.10 ± 0.22 mg/L plasma concentrations would be 1.13 ± 0.21 mg/L for (Table 2). The concentrations of these two fluoro- NFLX and 4.99 ± 0.35 mg/L for OFLX (p < 0.001), and the quinolones in prostatic tissues were 4.10 ± 0.79 µg/g for concentrations in prostatic tissue would be 1.91 ± 0.41 µg/ OFLX and 2.22 ± 0.57 µg/g for NFLX (Table 2). OFLX g for NFLX and 3.61 ± 0.40 µg/g for OFLX (p = 0.009). concentrations were significantly higher than those of NFLX in both serum and prostatic tissue (p < 0.001). The ratio of the prostatic tissue concentration (P) to serum iscussion concentration (S) was 2.11 for OFLX and 5.71 for NFLX. D However, the difference in P/S ratio between OFLX and NFLX was not statistically significant (p = 0.175). The P/ The degree of transference of antimicrobial agents S ratios in Patients 7 and 10 were considerably different into prostatic fluid is influenced by several characteristics from those in other patients. If we deleted these outlying of the drug, and better penetration is seen after

550 J Formos Med Assoc 2001 • Vol 100 • No 8 Ofloxacin and Norfloxacin Concentrations in Serum and Prostate Adenoma

Table 2. Individual and mean (± SEM) ofloxacin (OFLX) and norfloxacin (NFLX) concentrations in the plasma and in prostate tissue 2 hours after the oral administration of these two drugs

Patient Prostatic tissue (µg/g) Plasma (mg/L) Prostate/serum concentration NFLX OFLX NFLX OFLX NFLX OFLX

1 1.026 3.190 0.490 4.050 2.094 0.788 2 2.504 2.962 1.740 4.630 1.439 0.640 3 2.142 2.917 1.910 3.630 1.121 0.804 4 1.344 2.782 1.060 4.910 1.268 0.567 5 1.413 3.921 0.910 5.490 1.553 0.714 6 1.864 4.128 1.490 6.260 1.251 0.659 7 0.630 1.696 0.020 0.270 31.50 6.281 8 4.345 6.109 2.100 6.370 2.069 0.959 9 0.661 2.897 0.780 4.600 0.847 0.630 10 6.270 10.375 0.450 1.140 13.93 9.101

Mean* 1.91 3.61 1.13 4.99 5.707 2.114 Range (0.66–4.35) (2.09–6.11) (0.49–2.1) (3.63–6.37) (0.85–2.1) (0.57–0.96) SEM* 0.41 0.40 0.21 0.348 3.125 0.954 p† p < 0.01 p < 0.001 p = 0.175

*Mean and SEM are calculated without the data from Patients 7 and 10. †Differences between NFLX and OFLX in prostatic tissue, plasma, and prostate/serum ratio were assessed using a paired t-test. administration of lipid-soluble bases with high pKa According to reports by Paton & Reeves and Aagaard, values, such as co-trimoxazole (pKa = 7.4) and newer maximum serum concentration can be generally ob- fluoroquinolones (NFLX: pKa1 = 6.2–6.4, pKa2 = 8.7–8.9; served 1 to 3 hours following administration of NFLX and

OFLX: pKa1 = 5.74, pKa2 = 7.9) [8]. Reviewing the literature, 1.1 to 1.4 hours following administration of OFLX [12, the concentration of OFLX demonstrable in prostatic 13]. Therefore, oral administration of antibiotic tissue varied between 0.41 and 8.02 mg/kg, and the 2 hours before surgery should be adequate to achieve concentration of NFLX was between 0.75 and 7.58 mg/kg maximal serum concentrations of these two drugs. Re- [3, 8]. The concentration of OFLX (3.61 ± 0.40) was viewing our results (Table 2), the plasma concentrations significantly higher than that of NFLX (1.91 ± 0.41 µg/ of OFLX and NFLX were, indeed, elevated to mL) after deleting the two extreme data points (Patients the expected levels (except in Patients 7 and 10) [3], but 7 and 10). The mean concentration of OFLX (3.61 ± 0.40 two patients (Patients 4 and 9) still experienced postop- mg/g; range 1.70–6.37 µg/g) in the prostate obtained in erative leukocytosis with fever. Postoperative urine and this study was in excess of the OFLX minimum inhibitory blood cultures were examined for these two patients but concentration (MIC) for most pathogens responsible for failed to find any pathogen. Although chemoprophylaxis a large percentage of prostatic infections: E. cloacae (MIC90, for UTI during and after TUR-P has been reported by µ µ 0.78 g/mL), E. coli (MIC90, 019 g/mL), and Proteus preoperative administration of a single oral dose of µ mirabilis (MIC90, 0.39 g/mL) [9]. Although the mean ciprofloxacin (500 mg) and the results have been prom- concentration of NFLX (1.91 ± 0.41 µg/mL; range 0.661– ising [4], the success rate of the present study was only 4.345 µg/mL) also exceeded the MIC for E. cloacae 80%. Fluoroquinolones used in this study have a relatively µ µ (MIC90, 0.78 g/mL), E. coli (MIC90, 0.39 g/mL), and P. short mean serum half-life (about 3 to 7 hours for NFLX µ mirabilis (MIC90, 0.39 g/mL), the prostatic concentra- and 5 to 7 hours for OFLX). A further 5 days of treatment tions of NFLX in Patients 7 and 9 were not above the MIC90 with these fluoroquinolones has been suggested by Grabe for E. cloacae [3]. et al to further reduce perioperative UTI rates [14]. With respect to gram-positive organisms, OFLX was Furthermore, adjusting the dosage according to the body present at concentrations above the MIC90 for Staphylococ- weight may be necessary as shown by Patients cus epidermidis (0.78 µg/mL) and Streptococcus faecalis (3.13 7 (84 kg) and 10 (80 kg), who had lower plasma concen- µg/mL), two of the most common causes of prostatitis trations of NFLX and OFLX. [10, 11]. However, the NFLX concentration in prostate Another possible reason why single preoperative did not exceed the NFLX MIC90 for S. epidermidis (3.13 NFLX and OFLX administration was unable to prevent µg/mL) and S. faecalis (50 µg/mL). The pseudomonads infection after TUR-P in this study was that 90% of were not well covered by either drug in this study. patients had chronic prostatitis (confirmed by their

J Formos Med Assoc 2001 • Vol 100 • No 8 551 J. Chen, R.R.L. Chen, and H.S. Huang pathology). Failure of antibiotic therapy in chronic bacte- 2. Naber KG: Antibakterielle Chemoprophylaxe bei rial prostatitis was due to the difficulty of eradicating transurethaler Resektion der Prostasta. Aktuelle Urologie protected bacterial microcolonies within an infection- 1987;18:34–7. induced altered microenvironment deep within the 3. Cunha BA: The fluoroquinolones for urinary tract prostate gland [15]. These protected small bacterial colo- infections: a review. Adv Ther 1994:11:277–96. 4. Giannopoulos A, Koratzanis G, Giamarellos-Bourboulis nies might be released from the prostate during TUR-P, EJ, et al: Pharmacokinetics of intravenously adminis- and result in infection after TUR-P. Furthermore, the tered pefloxacin in the prostate; perspectives for its MIC90s of NFLX and OFLX for Klebsiella pneumoniae, application in surgical prophylaxis. Int J Antimicrob Agents E. cloacae, Staphylococcus aureus, and S. epidermidis were 2001;17:221–4. shown to be at least four- to eight-fold higher in Taiwan 5. Seibert G, Limbert M, Klesel N: Comparison of the than in other countries [16]. Although we could not find antibacterial in vitro and in vivo activity of ofloxacin these pathogens in the pre- and/or postoperative cultures, (HOE 280, DL 8280) and nalidixic acid analogues. Eur J the development of resistance strains in Taiwan, where Clin Microbiol 1983;2:548–53. antimicrobial drugs are used with little restriction, may 6. Aagaard J, Knes J, Madsen PO: Prostatic tissue levels of also play a role in post-TUR-P infection. ofloxacin. Urology 1991;38:380–2. In conclusion, it is not surprising that OFLX reached 7. Sabbaj J, Hoagland VL, Cook T: Norfloxacin versus co- trimoxazole in the treatment of recurring urinary tract significantly higher concentrations in serum and pros- infections in men. Scand J Infect Dis Suppl 1986;48:48–53. tatic adenoma tissues than NFLX for the same dose, 8. Boerema JB, Bach D, Jol C, et al: Penetration of rufloxacin because oral bioavailability is estimated to be 90% for into human prostatic tissue and fluid. J Antimicrob Chemother OFLX and 35% to 45% for NFLX [12]. However, NFLX 1991;28:547–54. has a higher P/S ratio than OFLX. Both exceeded the 9. Sato K, Inoue Y, Fujii T, et al: Antibacterial activity of MICs for most gram-negative organisms but some indi- ofloxacin and its mode of action. Infection 1986;14(Suppl viduals may only achieve lower prostatic concentrations 4):S226–S230. against E. cloacae. The results of this study show that the 10. Wedren H: On chronic prostatitis with special studies success rate for eradication of postoperative infection of Staphylococcus epidermidis. Scand J Urol Nephrol 1989; after a single administration of OFLX (200 mg) and 123:1–36. NFLX (200 mg) orally was 80%. Comparing these results 11. Doble A: Chronic prostatitis. Br J Urol 1994;74:537–41. 12. Paton JH, Reeves DS: Fluoroquinolone antibiotics— with the rate of postoperative UTI and septic complica- microbiology, pharmacokinetics and clinical use. Drugs tion after a single dose of ciprofloxacin (less than 5%), it 1988;36:193–228. appears that a longer course of OFLX or NFLX treatment 13. Aagaard J: Ofloxacin in prostatitis. Drugs 1993;45 after TUR-P may further lower the complication rate. (Suppl):52. 14. Grabe M, Hellsten S, Forsgren A. The effects of a short ACKNOWLEDGMENTS: We gratefully acknowledge Mr. and a prolonged perioperative course of ciprofloxacin in Kau Yeong-Shun for his excellent technical assistance. TUR compared to a control group without antibiotics. International Symposium on New Quinolones, Geneva, 1986. [abstract no. 179] 15. Nickel JC, Downey J, Clark J, et al: Antibiotics pharma- eferences cokinetics in the inflamed prostate. J Urol 1995;153: R 527–9. 16. Chen YC, Chang SC, Hsu LY, et al: In vitro antimicro- 1. Childs SJ, Wells WG, Mirelman S: Antibiotic prophylaxis bial activity of fluoroquinolones against clinical for genitourinary surgery in community hospitals. J Urol isolates obtained in 1989 and 1990. J Formos Med Assoc 1983;130:305–8. 1993;92:1040–8.

552 J Formos Med Assoc 2001 • Vol 100 • No 8