In-vitro release of vancomycin and netilmicin from bone cement

Letters t o t h e Editors

no known deleterious effects on cardiac potassium channels [lo]. Genetic polymorphism in the metabolism of terfenadine niight explain elevated serum concentrations of both terfenadine and its active metabolite, although this seems unlikely. In conclusion, in the presence of itraconazole, circulating intact terfenadine levels remain high and are potentially responsible for severe arrhythmia. The role of an increased serum concentration of the acid metabolite requires further investigation.

Olivier Lortholary, Nathulie R i z z o , Michel Tad, Jean-Yver Artigou, Johann Beaudreuil, Yves Cohen, Philippe Casasrus and Olivier Petitjean Avicenne Hospital, Bobigny, France

Acknowledgements The authors thank Marion Merrell Dow Research Institute, Kansas City, Missouri, for determining the serum concentrations of terfenadine and its acid metabolite.

References 1. MacConnell TJ, Stanners AJ. Torsades de pointes complicating treatment with terfenadine. BMJ 1991; 302: 1469. 2. Monahan BP, Ferguson CL, Killeavy ES, Lloyd BK, Troy J, Cantileiia LR. Torsades de pointes occurring in association with trrfenadine use. JAMA 1990; 264: 2788-90. 3. Zininiermanri M, Duruz H, Guinand 0, et al. Torsades de pointer after treatment with terfenadine and ketoconazole. Eur Heart J 1992; 13: 1002-3. 4. Honig PK, Woosley RL, Zarnani K, Coniier DD, Cantileria LR. Changes in the pharmacokinetics and electrocardiographic pharnracodynaniics olterfenadine with concomitant administration of erythromycin. Clin Pharmacol Ther 1992; 52: 231-8. 5. Woosley RL, Chen Y, FreitrianJP, Gillis RA. Mechanism of the cardiotoxic actions of terfenadine. JAMA 1993; 269: 1532-6. 6. Crane JK, Shih HT. Syncope and cardiac arrhythmia due to an interaction between itraconazole and terfenadine. Ani J Med 1993; 95: 445-6. 7. Pohjola-Sintonen S, Viitasalo M, Toivonen L, Neuvonen I? Itraconazole prevents terfenadine metabolism and increases risk of torsades de pointes ventricular tachycardia. Eur J Chi Pharniacol 1993; 45: 191-3. 8. Mathews D R , McNutt B, Okerholm R, Flicker M, McBride G. Torsades de pointes occurring in association with terfenadine use. JAMA 1991; 266: 2375-6. 9 Honig PK, Wortharn DC, Zaiiiani K, Mullin JC, Conner DP, Cantilena LR. The effect of fluconazole on the steadystate pharmacokinetics and electrocardiographic pharniacodynaniics of terfenadine in humans. Clin Pharniacol Ther 1993; 53: 630-6.

211

10. Crumb WJ Jr, Brown AM. Terfenadine blockade of a potassium current in human atrial niyocytes. Circulation 1993; 88: 1231. 11. LavrGsen K, van Houdt J, Meuldernians W, Janssens M, Heykants J. The interaction ofketoconazole, itraconazole and erythroniycin with the in vitro metabolism of antihistamines in human liver microsoiiies [abstract 95 1431. Annual Meeting of the European Academy of Allergology and Clinical Immunology, Rotterdam, The Netherlands, 1993. 12. Eller MG, Okerholm RA. Pharniacokinetic interaction between terfenadine and ketoconazole. Clin Pharniacol Ther 1991; 49: 130. 13. Honig PK, Wortham DC, Zamani K, Conner DP, Mullin JC, Cantilena LR. Terfenadine-ketoconazole interaction. Pharmacokinetic and electrocardiographic consequences. JAMA 1993; 269: 1513-8. 14. Honig PK, Wortharn DC, H d l R, Zarnani K, Smith JE. Cantilena LR. Itraconaeole affects single-dose terfenadine pharmacokinetics and cardiac repolarization pharmacodynamics. J C h i Pharniacol 1993; 33: 1201-6. 15. Paserchia LA, Hewett J, Woosley RL. Effects of ketoconazole on QTc. Clin Pharniacol Ther 1994; 55(2): 146. 16. Chen Y, Woosley RL. Krtoconazole blocks potassium currents in feline heart. Circulation 1993; 88(4, pt 2): 1-38. 17. Davies AJ, Harindra V, McEvan A, Ghose RR. Cardiotoxic effect with convulsions in terfenadine overdow. BMJ 1989; 298: 325.

In-vitro release of vancomycin and netilmicin from bone cement To the Editors: Despite all efforts, prosthetic infections still occur in 1 to 2% of cases. Once established, infection can be difficult to treat with systeniic antibiotics. Several studies have shown that gram-positive bacteria, especially Staphylocoicus aiueus, coagulase-negative staphylococci, Enterococcus faecalis and Corynebacteriaceae, are responsible for 71 to 84%) of such infections Ill. In the management of infected joint implants, revision arthroplasty with the use of antibiotic-loaded acrylic cement (ALAC) has proven to be superior to other methods of currently available treatment 121. Furthermore, ALAC is a useful adjunct in the nianagenient of chronic osteomyelitis and in the prevention of recurrent osteoniyelitis after an initial attack [ 3 ] .Increasing antibiotic resistance of bacteria that infect prosthetic joints has stimulated interest in the incorporation of more effective antiniicrobial agents into acrylic cement. Vanconiycin is effective against all staphylococci, and aniinoglycosides have a wider spectrum of activity against gram-negative bacteria [4].

Clinical M i c r o b i o l o g y and Infection, Volume 1 Number 3

212

Considering their broad antibacterial effects against both gram-positive and gram-negative bacteria, a study of the in-vitro elution of vancomycin and netilmicin from bone cement was carried out. A prospective in-vitro quantification was performed to measure the elution of vancomycin and netilmicin from bone cement. Vancomycin (500 mg) and netilmicin (250 mg) were admixed with 10 g of polymethylmethacrylate (PMMA) copolymer powder, alone and in combination, prior to the addition of the liquid polymer. This was followed, according to the manufacturer's specifications, by the addition of 5 mL of liquid methylmethacrylate. An exothermic polymerization reaction occurred, yielding a liquid cement that solidified over a period of 10 min at room temperature under aseptic conditions. From the polymerized cement, uniform cylindrical disks 6 m m in diameter, 9 m m thick and weighing approximately 500 mg were prepared. Each disk contained 25 mg of vancomycin and 12.5 mg of netilmicin alone or in combination. The disks were incubated in 10 mL tripticase soy broth (TSB) and human plasma at 36OC. Antibiotic concentrations were assayed by fluorescense polarization immunoassay (TDx, Abbott Laboratories, North Chicago, IL) on day 2 and weekly, then replaced and retested, for a period of 12 weeks. The lowest measurable level of drug concentration was defined as that which could be distinguished from 0 with 95% confidence; this was determined as 2.0 mg/L for vancomycin and 0.09 mg/L for netilmicin [ 5 ] . To compare the persistence of antibacterial activity around the ALAC disks, 1 mL of TSB or plasma was mixed with lo5 CFU/mL of a clinical isolate of S. aureus. The density of S. aureus suspension was determined spectrophotometrically and, subsequently, 100 pL of the suspension was transferred onto a TSB agar plate. After incubation at 36" C for 18 h, the plates

Table 1 Antibiotic concentrations released from antibioticloaded bone cement in tripticase soy broth (mg/L) Time Day 2 Week 1 Week 2 Week 3 Week 4 Week 5 Week 6 Week 7 Week 8 Week 9 Week 10 Week 11 Week 12

Vancomycin

Netilmicin

50.30 54.40 6.02 4.49 3.58 2.79 1.88 1.66 1.45 1.20 1.00 0.70 0.60

46.80 49.60 1.58 0.54 0.50 0.40 0.25 0.20 0.20 0.15 0.10 0.09 -

were examined for bacterial gowth. Between-group comparisons were made by Mann-Whitney Utest and p < 0.05 was considered significant. The in-vitro antibacterial activity of vancomycin and netilmicin, alone and in combination, in ALAC was assessed. All determinants were analyzed in duplicate (in TSB and in plasma). There was no significant difference in levels of netilmicin and vancomycin in either media samples. Antibiotic concentrations in TSB were higher than in plasma, but this was not statistically significant. Similarly, no differences were observed in weekly vancomycin and netilmicin levels in either TSB (Table 1) or plasma, alone or in combination. The release of vancomycin and netilmicin in combination did not differ from the release of each drug alone. These data indicate that the in-vitro release of neither parent antibiotic is reduced when mixed in cement. Vancomycin was detected in the test media during all 12 weeks of the study whereas netilmicin was detected for only 11 weeks of the study. Although initial rapid decreases in antibiotic concentration were seen during leaching from the cement, the antibacterial activity of vancomycin and netilmicin continued for 6 weeks and 3 weeks, respectively, during the monitoring period. O n the basis of the present and previous studies [6-81, all of the antibiotics tested remained stable throughout the polymerization process of the cement. Use of such antibiotic-cement mixtures in clinical practice appears to be effective in preventing prosthetic infections and, in addition, antibiotic-loaded PMMA cement may have a clinical role in the treatment of musculoskeletal sepsis caused by gram-positive and gram-negative bacteria, particularly in the presence of organisms resistant to the usual antibiotic agents. Hakan Leblebicioglu, Ahmet SaniG, Murat Giinaydin, I f a n Sencan, Nevzat Dabak and Yunus Nas Ondokuz Mayis University Medical School, Samsun, Turkey

Vancomycinfnetilmicin 74.52 64.85 8.03 3.45 1.62 1.52 1.so 1.25 1.20 1.00 0.80 0.60 0.60

37.73 39.20 1.79 0.75 0.49 0.31 0.31 0.25 0.25 0.20 0.15 0.10 -

Acknowledgements This study was presented as a poster at the 7th European Congress of Clinical Microbiology and Infectious Diseases, held in Vienna, Austria from March 26 to 30, 1995.

References 1. Kuechle DK, Landon GC, Musher D M , Noble PC. Elution of vancomycin, daptomycin and aniikacin from acrylic bone cement. Clin Orthop Re1 Res 1991; 264: 302-8.

Letters t o t h e E d i t o r s

2. Steinbrink K. T h e case for revislon arthroplasty u i n g antibiotic-loaded acrylic cement. Clin Orthop Re1 Res 1990; 261: 19-22. 3. Gerhart TN, Roux RD, Hanff PA, Horowitz GL, Kenshaw AA, Hayes WC. Antibiotic-loaded biodegradable bone crmcnt for prophylaxis and treatment of experimental osteomyelitis in rats. J Orthop Res 19?3; 11: 250-5. 4. Reese RE, Betts RE Antibiotic use. In: Reese KE, Betts RF, eds. A practical approach to infectious diseases, third edition. Boston: Little, Brown and Co, 1?91: 821-1007. 5. Abbott Laboratories Diagnostic Division. TDx" system assays I. North Chicago, IL: Abbott Laboratories, 1992. 6. Scyral P. Zannicr A, Argenson JN, Raoult D. The release in vituo of vancornycin and tobraniycin from acrylic cement. J Antimicrob Chernothrr 1?94; 33: 337-9. 7. Popharn GJ, Mangino P, Seligson D, Henry SL. Antibloticimpregnated beads. Part 11: Factors in antibiotic selection. Orthop Rev 1991; 20: 331-7. 8. Askew MJ, Kufel MF, Fleissner PK Jr, Gradisar IA Jr, Salstroni S, Tan JS. Effect of vacuum n i i x i q o n the mechanical properties of antibiotic-impregnated polyniethylmethacrylate bone cement. J Biomed Mater Res 1990; 24: 573-80.

Thrice-weekly maintenance therapy for cytomegalovirus retinitis in patients with acquired immunodeficiency syndrome (AIDS) To the Editors: Cytomegalovirus (CMV) retinitis is the most common localization of CMV disease in patients with acquired imniunodeficiency syndrome (AIDS) and is, in fact, the most frequent ocular opportunistic infection, often leading to a partial or total loss of vision [I]. Induction treatment for CMV disease is relatively well standardized (2 to 3 weeks of daily intravenous ganciclovir or foscarnet) and able to control the progression of the retinitis in approximately 60 to 80% of cases [2,3]. Without maintenance therapy, the relapse rate approaches 100% within 3 to 6 weeks. Even with maintenance therapy with ganciclovir (5 mg/kg/day) or foscarnet (120 nig/kg/day) daily for 5 daydweek, 50% of patients relapse after a median time of 8 to 16 weeks [4,5]. It has been suggested that doubling the dose of ganciclovir (to 10 nig/kg/day), but with thrice-weekly administration, may be as effective as the standard dosage of 5 days/week [6,7]. To follow is our report of the results of an open, prospective, clinical study of maintenance therapy, using the thrice-weekly approach, for CMV retinitis in AIDS patients. A total of 69 human immunodeficiency virus (HIV)-infected adults diagnosed with C M V disease between May 1988 and December 1993 were eligible for inclusion in the study. Those who had C M V retinitis (M = 57), with or without other localizations,

213

were given at least 2 weeks of intensive induction therapy with eithcr intravenous ganciclovir (5 ing/kg every 12 h), foscarnet (60 mg/kg every 8 h) cir both. O f these 57 patients, 33 successfully completed the induction treatment, gave their conscnt to receive intravenous maintenance therapy thrice weekly and were included in the study. Maintenance therapy was started with either ganciclovir (10 mg/kg/day) or foscarnet (1 00 nig/kg/day) or both, every Monday, Wednesday and Friday, for a mean period of 19 (range 6 to 78) weeks. The crude relapse rate was 51% (17/33) and the crude mortality rate was 66%)(22/33). Altogether, 27 patients received ganciclovir (14 relapses, 28 deaths), four received foscarnet (two relapses, two deaths), and two patients received both (one relapse, two deaths). The time to relapse of CMV retinitis, time to death or relapse and time to death are shown in Figure 3. The median time to relapse was 18 weeks, the median survival time free of CMV retinitis was 14 weeks and the median survival time was 34 weeks. Neutropenia occurred in eight patients (24%,),seven receiving ganciclovir and one receiving foscarnet, and anemia occurred in two patients (6%,),both of whom were receiving ganciclovir. No changes in therapy were required. The neutropenia was con trolled using granulocyte colony-stimulating factor and the anemia with blood transfusion when required. Treatment with either drug improved median survival time from the time of CMV diagnosis to 6 to 8 months to almost 1 year in those responding to antiCMV treatment. Maintenance therapy iq mandatory to delay relapse and progression to blindness. T h e median time to relapse has been reported to range from 2 to 4 weeks without maintenance therapy and from 8 to 19 weeks with daily or 5 days/week maintenance therapy [4,5]. Both ganciclovir and foscarnet appear to be equally effective in preventing relapses, although in at least one study, foscarnet significantly increased survival time (by around 4 months) coinpared with ganciclovir [4). Hall and colleagues [6] and Ganveg and coworkers [7]have reported that a total weekly dose of 30 mg/kg of ganciclovir, distributed in three, five or seven doses, has similar efficacy in preventing relapses. Our results are in agreement with these studies and emphasize that the thrice-weekly approach does not appear to be worse in ternis of survival. Ganciclovir therapy is often complicated by severe neutropenia, thereby requiring temporary or permanent discontinuation or niodification of dosage in a t least 25% of patients. In our study, eight (24%) of 29 patients developed moderate neutropenia that did not require discontinuation of therapy. Thus, the toxicity rate was similar to that seen with standard therapy (5 nig/kg/day for 5 days/week). At present, oral ganciclovir i s being evaluated as