Surg Neurol 1992;37:356-60

Cranioplasty Using Gentamicin-Loaded Acrylic Cement: A Test of Neurotoxicity Juan F. Ronderos, M.D., David A. Wiles, M.D., Francis A. Ragan, Ph.D., Colby W. Dempesy, Ph.D., Frank C. Culicchia, M.D., C. J. Fontana, and Donald E. Richardson, M.D. Department of Neurosurgery, Tulane University School of Medicine, and Department of Pathology, Louisiana State University School of Medicine, New Orleans, Louisiana

Ronderos JF, Wiles DA, Ragan FA, Dempesy CW, Culicchia FC, Fontana CJ, Richardson DE. Cranioplasty using gentamicin-loaded acrylic cement: a test of neurotoxicity. Surg Neurol 1992;37:356-60. Cranioplasty represents a formidable challenge for neurosurgeons, with a significant morbidity from both early and late wound infections. Polymethylmethacrylate (PMMA) is one of the most widely used materials in this setting. Despite the advantages of this material, such as ease of handling and inert biochemical properties, it is still a foreign body that is prone to infection. We present an animal model using a gentamicin-impregnated PMMA patch to assess the neurotoxicity as well as the efficacy of using this as an alternative material to lessen the infectious morbidity in this clinical setting. In part two of our experiment, we used a PMMA patch of similar weight and surface area in a physiological saline solution to determine the rate of gentamicin elution from the patch. The results obtained appear promising with no evidence of neurotoxicity and warrant further study to assess the clinical efficacy of PMMA in this setting. KEY WORDS: Polymethylmethacrylate; Gentamicin; Cranioplasty; Toxicity

Buchholz and Engelbrecht [2] first described the use of antibiotic-impregnated polymethylmethacrylate (PMMA) to control deep wound infection following total joint arthroplasty. Extensive testing has established that thermostable antibiotics will elute from P M M A during the first days of implantation at bactericidal concentrations [3]. Indeed, the antibiotic present in local bone and soft tissues greatly exceeds the concentration possible with systemic administration [3 ]. Bactericidal levels of an-

Address reprint requests to:Juan F. Ronderos, M.D., Department of Neurosurgery, Tulane University School of Medicine, 1430 Tulane Avenue, New Orleans, Louisiana 70112. Received March 1, 1991; accepted November 11, 1991.

© 1992 by ElsevierSciencePublishingCo., Inc.

tibiotic have been shown in serum for several days and in urine for several weeks following insertion [1 1]. T h e safety profile in similar uses has been quite satisfactory. The advantages of using antibiotic-loaded P M M A have been recognized and practically applied in the field of orthopedics. H o w e v e r , the field o f neurosurgery is confronted with similar complications o f infection with cranioplasty and yet has failed to adopt the use of antibiotic-loaded PMMA. T h e fear of antibiotic neurotoxicity and weakening of the P M M A compressive and diametral tensile strengths are the probable reasons. Studies addressing the strength of antibiotic-loaded P M M A have already appeared in the literature. Mixing acceptable concentrations of powdered antibiotic (except chloramphenicol and tetracycline [1,4]) to p o w d e r e d P M M A before the addition of the liquid catalyst does not affect the structural integrity of PMMA. N o decrease in the mechanical properties of P M M A up to a concentration of 2.0 g gentamicin per 40 g P M M A have been demonstrated [ 1,4-7,14]. Considering the acceptable strength characteristics of antibiotic-loaded PMMA, our study is based on determining the neurotoxicity associated with this application. We have conducted an initial study with dog cranioplasty using gentamicin-loaded PMMA. Gentamicin was chosen based on reports showing it to be the antibiotic with the highest sustained release from P M M A [11]. Gentamicin also is active against both gram-negative and gram-positive organisms, has a low rate o f allergic reaction, has free solubility in water, and is stable at relatively high temperatures, and resistance to it is rare [ 11]. We report that cranial repair with gentamicin-impregnated P M M A offers the same advantages as in arthroplasty and with no evidence suggestive of central nervous system (CNS) toxicity at concentrations used in this experiment. Materials and Methods Six mongrel dogs weighing 2 5 - 3 5 pounds were used. The operations were p e r f o r m e d using aseptic technique 0090-3019/92/$5.00

Neurotoxicity in Cranioplasty

Figure 1. Sketch demonstrating a typical craniectomy preparation used in

this study.

and standard operative protocol. The animals were first sedated with intramuscular sedation xylazine (1-2 mg/kg) and then anesthetized with a combination of intravenous pentobarbital (25 mg/kg) and inhalation anesthetics (halothane). Baseline samples of blood and urine were obtained. The head and neck were shaved, and standard preoperative skin preparation with povidone iodine solution was used. A curvilinear incision was made just anterior to the right ear and taken anteriorly toward the midline to the level between the eyes (Figure 1). This produced a flap, which exposed a large portion of the right hemisphere of the skull. A twist drill and bone rongeurs were used to fashion a craniectomy of approximately 4 x 4 cm, Care was taken to avoid any violation of the dura mater. When a tear occurred, it was repaired primarily with 4-0 interrupted braided nylon sutures. A circular PMMA patch was fashioned to fit the cranial defect and allowed to cool before permanent implantation. The PMMA patch was composed of gentamicin 1.5 g in powdered form (Elkins-Sinn, Inc., Cherry Hill, NJ) mixed with 30 g of PMMA powder. Liquid catalyst was then added to the mixture in the proportions provided by the cranioplasty kit (Codman and Shurtleff, Randolph, MA). The wound was closed in a layered fashion with strict adherence to hemostasis. Prior to the permanent insertion of the PMMA patch, a linear incision was made at the back of the neck at the level of the craniocervical junction. Using a muscle splitting technique, the lateral muscle masses were taken down to the level of the C1 lamina. A partial taminectomy was performed, the dura mater was incised, and a

Surg Neurol 1992;37:356-60


silastic catheter was inserted into the cisterna magna. A baseline cerebrospinal fluid (CSF) sample was obtained, and a subcutaneous reservoir was attached. The wound was then closed in layers. Postsurgical care was provided according to the guidelines described in "PHS Policy on Human Care and Use of Laboratory Animals."* All samples were collected at 2, 4, 24, 48, 168, and 336 hours postoperatively. Samples of blood were obtained using standard venipuncture techniques. Cerebrospinal fluid was obtained by a 25-gauge needle aspiration of the subcutaneous reservoir. Urine was obtained through intermittent catheterization or with manual bladder compression. Wound exudate samples were obtained by injecting 1 mL of physiologic saline subcutaneously over the patch site, waiting 5 minutes, then withdrawing as much saline as possible. Since this technique detects only a portion of the gentamicin available at the wound site, the levels obtained are clearly only a relative measurement of actual levels. At the final collection the animal was killed. In addition to the regular fluid collection samples, two portions of brain tissue were taken. A proximal sample was taken directly under the PMMA patch and a distal sample was taken at a point farthest from the PMMA patch. All procedures were performed under the supervision of the Tulane Vivarium. A cylindrical plug of the gentamicin-impregnated PMMA material was made approximating the mass and surface area of a cranioplasty patch in order to measure the gentamicin elution rate in vitro. This cylinder was placed in a slightly larger glass tube containing physiologic saline. By slowly rotating the glass tube in a rolling cannister, the washing action of body fluids over the patch in the in vivo experiment was approximated. Elution samples were obtained by inserting fresh saline samples of 1-mL volume in the tube at 0, 2, 4, 24, 48, and 168 hours; rotating the tube for 1 hour; then removing the saline for analysis. In the intervening "nonsampling" periods, saline continued to wash the rotating cylinder, but this fluid was discarded. In vitro and in vivo samples were measured using the EMIT assay technique (Syva Company, Palo Alto, CA). Blood samples were collected in tubes containing 15 mg EDTA solution, spun down, and plasma extracted for assay. Brain tissue samples (each approximately 0.5 rag) were homogenized in 2 mL of physiologic saline, spun down, and the supernatant extracted for assay. Cerebrospinal fluid, urine, and wound exudate samples were assayed without further processing. The smallest EMIT assay detection level was set at 0.1 ~g/mL of gentamicin. * This experimental procedure was performed under the supervision and guidance of the Advisory Committee for Animal Research at Tulane University Medical School.


Surg Neurol 1992;37:356-60

Ronderos et al

Gm in CSF

Gm in Plasma I




0.8 f









4 . 0 ( ~

0 2 4 1 2 7 1 4 HOUR DAY


2 4 1 HOUR

Gm in Urine




Gm in W.E. r










cin. The in vitro sample yielded over 300 /~g/mL in the first hour of saline washing; the average gentamicin content in the in vivo sample (CSF, urine, and wound exudate) reflected this elevation. Plasma levels remained modest throughout, suggesting that antibiotic dilution occurred quite rapidly in the circulation. Following this initial observation, the elution rate from the cylinder remained reasonably stable over the seven days of observation, dropping from 15 to 10 /~g/mL. This implied availability of antibiotic can easily account for the gentamicin levels in the in vivo fluids, once the initial surface excess was physiologically eliminated over the first day. In particular, the large wound exudate levels ofgentamicin, which are of the same magnitude as the initial in vitro concentrations, can be explained. Gentamicin levels of the proximal and distal brain tissue were not detectable (ie, g/mL). A few sporadic case reports have been reported of direct CNS toxicity from gentamicin intrathecal therapy. Toxicity was manifested as pathological plaques of inflammation in the parenchyma and meninges of the CNS. Toxicity was noted to be a function of both the concentration and the duration of gentamicin in the CNS. Short lived, high levels of gentamicin or chronic low concentrations appeared to have little toxic effects in the CNS [13]. Our studies demonstrate very little gentamicin crossing through the dura mater. Even when tears in the dura mater were primarily repaired, the CSF elevations in gentamicin concentration were very transient (1.0/.Lg/mL at 24 hours). Additionally, gentamicin in brain tissue at autopsy was not measurable, and no gross lesions were noted at autopsy. These factors support the absence of neurotoxicity with gentamicinimpregnated PMMA in cranioplasty. In conclusion, there appears to be no evidence to suggest any neurotoxic effects of gentamicin-impregnated PMMA. Wound exudate levels remained quite high for an acceptable length of time while CSF, plasma, and urine levels were not extremely high. No wound infections were encountered, and evidence from similar applications suggests that there is excellent perioperative wound prophylaxis [2,9]. At this point, long-term studies would further delineate any potential complications that may become apparent. Special thanks go to C. J. Fontana, Frank Chapple, D.V.M., and Tom Fairburn, who helped make this study possible.

References 1. Buchholz HW. Hip arthroplasty for total replacement. St, George Model, Hamburg, Germany. Presented at the Annual Meeting of the AAOS, San Francisco, California, March 6, 1971. 2. Buchholz HW, Engelbrecht H. Uber die Depotwirkung einiger Antibiotica bei Vermischung mit dem Kunstharz Palacos. Chirurg 1970;41:511. 3. Chapman MW, Hadley WK. The effect of PMMA and antibiotic combinations on bacterial viability. An in vitro and preliminary in vivo study. J Bone Joint Surg 1976;58A:76.


Surg N e u r o l 1992;37:356-60

4. Ger E, Dall D, Miles T, Forder A. Bone cement and antibiotics. S Afr MedJ 1977;51:276. 5. Lautenschlager EP, Marshall GW, Marks KE, Schwartz J, Nelson CL. Mechanical strength of acrylic bone cements impregnated with antibiotics. J Biomed Mater Res 1976;10:837. 6. Lee AJC, Ling RSM, Vangala SS. Some clinically relevant variables affecting the mechanical behaviour of bone cement. Arch Orthop Trauma Surg 1978;92:1. 7. Marks DE, Nelson CL, Lautenschlager EP. Antibiotic impregnated acrylic bone cement. J Bone Joint Surg 1976;58A:358. 8. Murray WR. Use of antibiotic-containing bone cement. Clin Orthop 1984;190:89-95. 9. Reichelt A, Brand A. Das verhalten der Blutsenkungsgeschwindigkeit nach der Implantation yon Totalendoprothesen des Huftgelenkes. Z Orthop 1975;113:900. 10. Rosenthal AL, Rovell JM, Girard AE. Polyacrylic bone cement

R o n d e r o s et al

containing erythromycin and colistin I. In vitro bacteriological activity and diffusion properties of erythromycin, colistin and erythromycin/colistin combination. J In Med Res 1976;4: 296-304. 11. Wahlig H, Dingeldein E. Antibiotics and bone cements. Acta Orthop Scand 1980;51:49-56. 12. Wahlig H, Dingeldein E, Bergmann R, Reuss K. The release of gentamicin from polymethylmethacrylate beads. J Bone Joint Surg 1978;60:270-5. 13. Watanabe I, Hodges G, Dworzack DL, Kepes JJ, Duensing GF. Neurotoxicity of intrathecal gentamicin: a case report and experimental study. Ann Neurol 1978;4:564-72. 14. Weinstein AM, Bingham D, Sauer B, Lunceford M. The effect of high pressure insertion and antibiotic inclusions upon the mechanical properties of polymethylmethacrylate. Clin Orthop 1976; 121:67.

Cranioplasty using gentamicin-loaded acrylic cement: a test of neurotoxicity.

Cranioplasty represents a formidable challenge for neuro-surgeons, with a significant morbidity from both early and late wound infections. Polymethylm...
411KB Sizes 0 Downloads 0 Views