Stereotactic Surgery and Long-Term Maintenance of Cranial Implants in Research Animals

THOMAS W. GARDINER, PHD,1 AND LINDA A. TOTH, DVM, PHD2

Abstract ͉ Most neuroscience research is performed by using anesthetized animals or tissue samples obtained from animals that have been euthanatized. However, study of many important issues requires the use of animals that are alert and capable of engaging in behavior. Various methods have been used to humanely perform neuroscience experiments that involved unanesthetized animals. These techniques often involve surgical implantation of an apparatus that permits direct manipulation of brain tissue or measurement of neurochemicals or neuronal activity in conscious animals. We describe here common surgical techniques used to prepare animals for long-term neuroscience studies, discuss several issues related to short- and long-term postoperative care of animals with implants, and offer suggestions that veterinary and research personnel can use to prevent or mitigate some common problems that may develop when preparing and maintaining animals for these studies.

Most neuroscience research is performed by using anesthetized animals or tissue samples obtained from animals that have been euthanatized. However, the study of many important issues sometimes requires the use of animals that are alert and capable of engaging in behavior. Studies that address how the brain controls processes such as learning, sleep, or movement may be meaningful only if brain function is evaluated while the animal performs the behavior of interest. In addition, anesthesia can profoundly influence the brain’s responses to some experimental treatments. For example, amphetamine and other drugs can affect the nervous system differently in anesthetized animals as compared to unanesthetized animals (1). Learning how various drugs modulate brain function may require collect- ing data from unanesthetized animals. Various methods have been used to humanely perform neuroscience experiments on unanesthetized animals. These techniques often involve surgical implantation of an apparatus that permits direct manipulation of brain tissue or measurement of neurochemicals or neuronal activity in conscious FIG. 1. Photograph of an unrestrained rat with an implant for animals (Figure 1). To perform these types of studies, an ani- electrophysiologic recording. The electrical connector in the headpiece mal is anesthetized, and a headpiece apparatus is permanently permits the attachment of wires that conduct electrical signals to ampli- attached to its skull. The animal then is allowed a period of fiers and data collection equipment. The headpiece was surgically attached to the skull of the rat while it was anesthetized, and experimen- postsurgical recovery before collection of data actually begins. tal recordings can later be obtained without causing pain or discomfort Design of a headpiece apparatus depends on the type of study to the conscious rat. that will be performed. For example, a headpiece might contain electrodes that record electrical activity from single neurons in specific brain regions, electrodes that measure insertion of electrodes or other probes into the brain. In fact, electroencephalographic activity or chemical changes in the they sometimes are so relaxed that they fall asleep during the brain, or cannula tubes that enable collection of cerebrospinal surgical procedure (2). Such observations confirm that physi- fluid or administration of drugs and other substances directly cal penetration of brain tissue is not painful and that despite into specific brain regions. the strange or unnatural appearance of headpieces, long-term Permanently implanted headpieces allow experimental mea- experimental manipulations of the central nervous system can surements to be taken or manipulations to be performed be performed without causing stress or discomfort to the ani- without causing pain or distress to conscious animals. In fact, mals. To accomplish this, however, surgical implantation must in many instances, such as during measurement of neuronal be performed correctly, and the headpiece must be maintained activity, animals seem to be unaware of the experimental ma- properly after surgery. nipulation. Verbal reports from human patients undergoing We describe here common surgical techniques used to pre- neurosurgery reinforce this impression. Surgical treatment of pare animals for long-term experimental study of the brain. We some disorders requires that patients communicate with the also address several issues related to short- and long-term post- neurosurgeon during surgery so that their sensory responses operative care of animals with implants, including procedures and motor or cognitive states can be evaluated as the surgical that can be used to keep the brain surface healthy and protected procedures are performed. Anesthesia for these patients con- in animals implanted with headpieces that incorporate remov- sists only of locally acting agents applied to the scalp, yet these able electrodes. Although our examples reported here focus on patients report a complete lack of sensation associated with implants designed for electrophysiologic recording from individual neurons, the surgical procedures and considerations for Department of Anatomy and Neurobiology1, University of Tennessee, Memphis, postoperative care apply equally well to the implantation of other TN 38163; Department of Infectious Diseases2, St. Jude Children’s Research types of devices, such as microdialysis probes, cannulas used for Hospital, Memphis, TN 38105 drug infusions, and electrodes used for stimulating specific ar-

56 CONTEMPORARY TOPICS © 1999 by the American Association for Laboratory Animal Science Volume 38, No. 1 / January 1999 FIG. 2. Diagram of an implant for use in unanesthetized rats. The electrophysiologic recording electrode is attached to the skull. Two small mounting screws are anchored in the skull, and acrylic cement is used to bond the headpiece apparatus to the screws and directly to the cranium. The specific design of the headpiece depends on the nature of the experiment. In this example, the recording electrode is fixed only to the upper platform of the headpiece apparatus. Large screws positioned in the threaded posts in the base of the headpiece permit the upper platform to be moved in small increments, resulting in gradual insertion of the electrode into the brain so that neurons located at various depths can be monitored during subsequent experimental sessions. eas of the brain. We have not attempted to review all available modifications of these techniques, and we do not believe that any specific method described here is innately superior to alter- native procedures used by other researchers. Rather, our FIG. 3. A rodent stereotactic apparatus. The micromanipulator attached intention is to provide personnel who evaluate and manage ani- to the stereotactic frame can be moved in 3 dimensions by turning the mals used for long-term brain studies with an overview of these calibrated thumbscrews on each axis and, thus, can be used to accu- commonly used surgical techniques and to supply suggestions rately position an electrode or other device within a target structure in for managing or avoiding various problems. the brain during stereotactic surgery.

Principles of Stereotactic Surgery Techniques for recording the activity of individual neurons involve inserting fine-tipped microelectrodes into the brain and positioning their active recording tips in close proximity to single neurons so that ongoing changes in electrical activity can be measured. By preparing animals with appropriate cranial implants, researchers can use this approach to measure neuronal activity in conscious, active animals. One type of headpiece (Figure 2) is designed to permit permanent implantation of an array of electrophysiologic recording electrodes in the brain of a rat, and illustrates the basic features of most permanently attached headpieces. The surgical procedure first involves exposing the surface of the skull by incising and re- tracting the overlying skin and fascia. Next, small screws are placed in the skull to provide a supporting base for the headpiece assembly. A hole is then drilled through the skull over FIG. 4. Anesthetized rat positioned in a stereotactic apparatus. The skull the target location in the brain, and the dura is incised. The is anchored between the ear bars, and the incisor bar is used to adjust electrode or other device that is to be implanted is attached to the forward angle of the snout and to hold the skull firmly in position. a micromanipulator and slowly inserted to a depth consistent with the location of the intended target structure in the brain. sitioning, animals are placed in a stereotactic frame during sur- Acrylic cement is used to secure the device in the correct posi- gery. The probe is then inserted into position using a finely tion. Some electrode assemblies, including the one illustrated, calibrated micromanipulator. Some stereotactic frames are in- allow the electrode to be gradually advanced through the brain tended only for use with a specific species (Figure 3), whereas to permit evaluation of various neurons during sequential re- other designs include adaptors that accommodate animals rang- cording sessions. Other assemblies secure the electrode tip in a ing in size from rodents to primates. Two components of the permanent position at the time of the initial surgery, permit- stereotactic apparatus, the ear bars and the incisor bar, an- ting repeated studies at a single recording site. chor the skull, controlling the angle of the skull and its Most implantation procedures are performed by use of ster- position (Figure 4). The ear bars, which are inserted first, eotactic techniques, which allow an experimental device or are placed in the ear canals and clamped firmly against the probe to be implanted precisely into a region of interest in the skull to secure it in a horizontal plane. An important consid- brain. To achieve the necessary accurate three-dimensional po- eration in the choice of ear bars for stereotactic surgery is

Volume 38, No. 1 / January 1999 CONTEMPORARY TOPICS © 1999 by the American Association for Laboratory Animal Science 57 FIG. 6. Landmarks frequently used for stereotactic surgical procedures in rats. Skin overlying the skull has been retracted and the fascia removed (left). The skull has been completely cleaned and dried in preparation for attachment of the headpiece apparatus, and 2 landmarks, bregma and lambda, have been exposed. Bregma is usually defined as the point of intersection between the midline suture and the suture separating the frontal and parietal bones, whereas lambda is the intersection of the midline suture and the suture separating the parietal and occipi- FIG. 5. Examples of ear bars with pointed (left) and blunted (right) tips. tal bones. These landmarks can be used to determine the precise location of target structures in the brain. Shallow grooves forming a grid pattern the shape of their tips (Figure 5). Ear bars with relatively across the top of the cranium (right) will promote bonding of the cement to the cranium and greatly reinforce the strength of the headpiece pointed tips are somewhat easier to properly position in the ear attachment to the skull. canals and are often preferred for that reason. Pointed ear bars are acceptable for use in anesthetized animals that will be euthanatized without being allowed to recover from anesthesia, but containing ointments or solutions to the skin margins and skull they are not recommended for survival procedures, because the surface during surgery can reduce the amount of general anes- long narrow tips can potentially injure the tympanic membrane thetic that must be administered. Reducing the need for or damage the delicate bones that surround the ear canal (3). injectable or inhalant anesthetics through use of local anesthe- Blunt-tipped ear bars should be used for long-term studies (i. e., sia reduces the likelihood of an inadvertent overdose or other animals will be allowed to recover from anesthesia), because they anesthesia-related complications, particularly during prolonged are far less likely to produce injury. procedures. Lidocaine can be applied periodically to the skin or After the ear bars are inserted, the incisor bar is placed in the the skull surface during lengthy procedures. Ointments contain- mouth and positioned behind the incisors. This bar can be raised ing lidocaine can also be used to provide pain relief at the incision or lowered to adjust the angle of the head. A clamp is then placed site after surgery. over the snout to secure the head in position. Incisor bars can After securely positioning the animal’s head in the frame accommodate many species of animals used in stereotactic pro- with the ear and incisor bars, precise measurements that are cedures. However, for cats and primates, infraorbital clamps, fundamental to the success of the neurosurgical procedure are which are secured against the inferior orbital bones of the face, made using landmarks on the skull surface or other reference are generally used instead of snout clamps to stabilize the posi- points on the cranium. One important reference point is the tion of the head. interaural line, an imaginary line that passes through the brain As for other surgical procedures, adequate anesthesia is es- at the midpoint of the external auditory meatus (4); the sential for stereotactic surgery. However, two aspects of anesthesia interaural line is operationally defined as being located at the for stereotactic procedures specifically merit mention. One is level of the tips of the ear bars. The point at which the interaural related to the use of ear bars, which exert pressure against the line crosses the midsaggital plane is defined as stereotactic zero. skull and thereby cause a noxious stimulation that may require Any structure in the brain can be localized by determining ap- greater anesthetic depth than would otherwise be necessary. propriate three-dimensional distances from stereotactic zero, Removing ear bars from animals at the end of surgery causes an which is often used as a reference point for locating brain struc- abrupt decrease in this aversive stimulus and, in our experience, tures in non-rodent species. can result in a major increase in the depth of anesthesia. This Two other landmarks are also commonly used as reference possibility can be particularly important after prolonged peri- points for locating target structures in non-rodent and rodent ods of anesthesia and can be critical if problems with excessive species. Four bones (the frontal bone, the two parietal bones, anesthetic depth develop for unrelated reasons during surgery. and the occipital bone) form the dorsal surface of the skull. The In such cases, the surgical team should be particularly attentive midline intersection of the suture lines between the frontal and to careful monitoring of anesthetic depth and must be ready to parietal bones is defined as bregma, whereas lambda is defined provide physiologic support to the animal when the ear bars are as the midpoint of the best-fitting curve passing along the lamb- removed. A second consideration related to anesthesia for ster- doid suture, which is located between the parietal and occipital eotactic surgery is that local anesthesia should not be overlooked bones (5) (Figure 6). In rats, bregma is generally the more accu- as a valuable complement to general anesthesia. Indeed, many rate reference point for locating forebrain structures, whereas stereotactic procedures in human neurosurgery are completed lambda or the interaural line are the more accurate points for with use of local anesthesia alone. In animals, applying lidocaine- localizing structures in the brainstem (6,7).

58 CONTEMPORARY TOPICS © 1999 by the American Association for Laboratory Animal Science Volume 38, No. 1 / January 1999 To accurately position an electrode or other device in the should be carefully evaluated. brain, the skin and fascia overlying the skull are incised and re- Another critical factor for achieving long-term attachment of tracted, and relevant landmarks are located. Appropriate a headpiece is to maximize adherence of acrylic cement to the measurements are then made relative to these landmarks (i. e., bone. This goal is attained primarily by proper preparation of a certain distance anterior or posterior to bregma and lateral to the skull surface. Thorough cleaning and drying of the cranium the midline), using distances determined from information sup- is essential to promote adherence of the cement. Incomplete plied in atlases (4,5,8,9) or obtained from research articles. In cleaning or drying of the bone surface allows a fluid or tissue research articles, target locations are typically defined by their interface to obstruct crevices to which the acrylic cement should three-dimensional coordinates. For example, a lesion might be bond, preventing secure attachment of the headpiece to the skull. made or a probe implanted in the brain of a rat at a position 2.5 Before cement is applied, the fascia overlying the skull must be mm posterior to bregma, 4.0 mm lateral to midline, and 7.5 mm removed completely by cleaning the bone with dry sterile gauze ventral to the dura. Brain coordinates for specific target loca- pads or cotton-tipped swabs. A blunt probe can also be used to tions vary considerably depending on the sex, age, weight, and lightly scrape the surface of the skull. However, rubbing or scrap- strain of the animal. For example, in Wistar rats, lambda is lo- ing too vigorously will result in removal of the periosteum, which cated 0.4 mm caudal to the interaural line in juveniles, 0.3 mm can thereby contribute to bone degeneration and eventual de- rostral to the interaural line in young adults, and 0.6 rostral to tachment of the headpiece. The skull surface must remain healthy the interaural line in mature rats (7). Although lambda is lo- and viable if it is to provide a firm anchor for the headpiece. cated 0.3 mm rostral to the interaural line in hooded and Wistar After the fascia is removed from the skull, blood often seeps rats, it is 0.7 mm rostral to this line in Sprague-Dawley rats of a from intracranial blood vessels onto the surface of the bone. similar weight (7). Coordinates provided in stereotactic atlases All such bleeding must be stopped before the acrylic cement is are determined by evaluation of animals that are specifically de- applied. This step is essential for success of the procedure, be- fined in terms of strain and physical characteristics. If animals cause the methyl methacrylate solvent in the cement can cause used in an experiment differ in terms of strain, weight, age, or additional dilatation of these superficial vessels, thereby pro- sex from those used to develop an atlas, the atlas should be con- moting bleeding that results in blood located between the bone sidered to provide only approximate values. Accurate selection and the overlying cement. This situation interferes with adher- of initial coordinates for preliminary studies may be improved ence of the cement to the bone. Occlusion of these superficial by review of the relevant literature. Obviously, a critical determi- vessels is easily accomplished by using the tip of a surgical probe nant for success of intracranial surgery is the assurance that the and a firm rubbing motion to compress the bone overlying each correct target structure is affected or measured. bleeding vessel and occlude the pores through which blood is seeping. After all bleeding has been stopped, the surface of the Factors Contributing to Longevity bone should be thoroughly cleaned with sterile physiologic saline solution and dried with sterile gauze pads. In rats, we use a of an Implanted Headpiece sterile swab to apply a small amount of 95 or 100% ethanol to In addition to accurate targeting the location of interest, an- the surface of the bone. The ethanol absorbs any fluid remain- other determinant of successful studies is whether the implanted ing on the bone surface and then quickly evaporates, leaving a headpiece will remain firmly attached to the skull for the in- dry porous surface that will bond well with the cement (11). tended duration of the experiment. Required durations can However, because ethanol can stimulate bleeding at wound range from weeks to months for some types of experiments, such margins, it should be carefully applied only to the bone. Oth- as those involving the study of animals trained to perform a be- ers advocate applying dilute hydrogen peroxide as an aid for havioral task (10). Headpieces typically are attached by using drying the skull (3). stainless-steel screws to anchor the headpiece to the skull. These Applying a solution of acidulated fluoride to the skull can screws are then covered with acrylic cement that binds to the strengthen the bone and provide improved retention of head- bone. Many investigators use sterile bone cement (e. g., Surgical pieces in adult rats (12). After the skull is thoroughly cleaned, Simplex P bone cement, Howmedica International, Rutherford, acidulated fluoride solution can be sterilized by passage through NJ), particularly for surgeries on primates. However, nonsterile a 22-␮m filter attached to a syringe. The sterilized solution can cement is commonly used in surgeries on other species without then be applied to the skull by dripping it directly on the skull associated infections. Factors that are critical for promoting long- after passage through the filter or by using a cotton-tipped swab. term attachment of the headpiece include meticulous attention The solution should remain on the skull for several minutes. We to sterile technique, proper preparation of the skull surface, and typically treat the skull in this manner when implanting head- secure anchoring of the cement and screws to the skull. pieces in rats and kittens, and we have not seen any disadvantages The use of good sterile technique is essential for achieving to use of acidulated fluoride solution. long-term attachment of a headpiece. Obviously, introducing Binding of the cement to the skull can be promoted by mak- infection into the brain can have disastrous consequences for ing shallow grooves in the bone to increase the three-dimensional the animal and the experiment. However, other important but surface available for adherence of the dental cement (Figure 6). perhaps less obvious problems are associated with infection of Any type of instrument with a sharp cutting edge that will easily the bony calvarium. These include periosteal damage, bone de- penetrate the surface of the bone can be used. We have used generation, and pain and discomfort of the animal. Bone various dental scalers or curettes. Excessive pressure should be degeneration (osteoporosis) can cause the entire skull to be- avoided when etching these grooves, because torque applied come thin, soft, and spongy within several weeks after surgery. If against the ear bars can easily damage the ear canals of a small degeneration is severe or develops near the anchoring screws, animal. If the available instrument does not easily form grooves the headpiece will become loose or may become dislodged. Move- in the bone, depth of the grooves can be increased by tracing ment of a headpiece after several weeks of stable attachment them several times with the tip of a probe. In rats, grooves are suggests bone degeneration as a major contributing factor, al- typically located 1 to 2 mm apart. Excessively close spacing of though severe osteoporosis can cause headpiece detachment grooves should be avoided to ensure that most of the perios- much sooner. In such cases, inspection of the cranium may re- teum remains viable. Although small screws are commonly used veal thin or pliable bone under the headpiece. If there is evidence to anchor the headpiece to the skull, in rodents, grooves alone of such problems, the sterile technique used by the surgeon can be used to effectively secure the headpiece without requir-

Volume 38, No. 1 / January 1999 CONTEMPORARY TOPICS © 1999 by the American Association for Laboratory Animal Science 59 ing screws (13). This emphasizes the value of good adherence in whatever form it may take, will likely remain solidly attached of acrylic cement to the bone surface for promoting long-term for a long period. Common causes for headpiece detachment stability of the headpiece. The relatively thin bones and flat skulls are improper preparation of the bone surface, which prevents of rats limit the mechanical benefit of the supporting screws. optimal adherence of the acrylic cement, and degeneration of Nonetheless, we see no advantage to omitting screws in rats and the bone under the headpiece, which can be caused by infec- prefer to combine use of screws and grooves to maximize the tion or mechanical damage to the periosteum (e. g., thermal long-term stability of attached headpieces in rodents. damage, excessive scraping). However, headpiece detachment Proper insertion of screws also promotes long-term stability can also result if an animal catches the headpiece on a compo- of the headpiece. To insert screws into the skulls of small ani- nent of the cage and loosens a solidly-implanted device while mals such as rats, holes that are slightly smaller in diameter than struggling to free itself. In our experience, such problems are the screws themselves are drilled in the skull surface (11). Self- unusual, but repeated unexplained difficulties in maintaining tapping screws that will fit tightly are then advanced through long-term implantation of headpieces may require consideration these holes into the skull. In animals with relatively thick skulls, of such possibilities. Housing for animals with headpieces should (e. g., primates), it is a common procedure to initially drill the be carefully evaluated and modified when necessary to reduce holes and then to thread them, using a thread tap, before insert- the likelihood of such events. These precautions are particularly ing the screws. In cats and primates, curvature of the skull and important if the headpiece includes protrusions that can become relatively thick bones enhance the effectiveness of the support- caught in grill tops, open flooring, squeeze devices, or other ing screws, and in these species, making grooves in the surface caging components. of the cranium probably adds little in terms of mechanical advantage. Because the integrity of the periosteum is potentially Factors Related to Postoperative Healing compromised when making grooves, we do not use this tech- After headpiece construction is complete, careful attention nique during surgery on non-rodent species. to closure of the incision will facilitate healing. Cement applied The final step in forming a secure foundation for the im- to the headpiece adjacent to skin margins should be smooth. planted headpiece is to apply a thin layer of cement over the Rough or pitted cement in this location can provide a pocket screws and surface of the skull. The first layer must have a fluid for bacterial growth and can irritate and abrade the adjacent consistency that will permit the cement to flow readily into all skin. Obtaining a smooth surface at the borders of the head- available crevices on the skull. Once applied, this initial layer piece can sometimes be difficult, particularly if the surgeon is must be permitted to set completely before subsequent layers relatively inexperienced at applying acrylic cement. We have of cement are applied. Manipulations performed before the found that a smooth surface can be obtained routinely by using cement sets completely can loosen the cement from the skull. a fluid mixture of the solvent and cement that has not begun to Complete hardening may require 5 to 15 min depending on set; this mixture is then applied in a thin, superficial layer along the brand of cement that is used. If the cement does not bind the headpiece margin. Rough areas can also be smoothed using securely to the skull, the headpiece will often be dislodged a dental drill to remove sharp edges and uneven contours. within a few days after surgery or become dislodged early dur- Because of the intervening headpiece, the margins of the in- ing experimental manipulations. cision cannot be apposed in a normal manner, creating a special An important feature of acrylic cement is that the chemical problem for healing. To heal properly, the skin must adhere to reaction associated with hardening generates a substantial the headpiece during the healing process. Good apposition of Њ amount of heat. Temperatures as high as 70 to 100 C have been the skin to the headpiece improves the appearance of the ani- recorded in in vitro studies and during hip replacement surger- mal, probably promotes comfort, and also reduces the likelihood ies in humans and dogs (14–16). Products vary in the rate at that microorganisms on the skin surface will penetrate the peri- which they cure and, therefore, in the rate at which they gener- osteum and potentially undermine the stability of the headpiece. ate heat. A simple way to assess the amount of heat generated After a smooth layer of cement has been applied on the head- from a particular product is to hold a small amount of partially- piece adjacent to the skin margin, skin surrounding the cured cement in the palm of the closed hand. In many cases, the headpiece should be positioned to naturally conform with the cement will be too warm to be held comfortably during curing. contour of the cement. In our experience, this is most easily In that case, the rate of cement application during surgery should accomplished by initially removing a relatively small section of be adjusted accordingly. High temperatures produced during skin from the designated location for the headpiece so that the curing can cause tissue damage and result in development of a exposed surface of the skull is actually somewhat smaller than reactive fibrous layer between the cement and bone (14). In ste- the area the headpiece is expected to cover. During surgery, ex- reotactic procedures, thermal damage to the periosteum or the cessive skin can be retracted to provide ample space for headpiece brain has been associated with rapid application of cement and construction. For rats and kittens, we have found that two 2.5-in the subsequent generation of excessive heat (17). In addition, wire eyelid retractors, positioned at right angles to each other, methyl methacrylate solvents for acrylic cement can cause hy- make excellent skin retractors for this purpose. After headpiece potension, formation of emboli, and anaphylactic reactions construction is complete, the retractors are removed and the skin during orthopedic and cardiovascular procedures in humans is positioned loosely around the headpiece. Excessive skin can then (18–21). These potential complications warrant careful moni- be carefully removed with sharp scissors, retaining an amount of toring of the animal during application of the cement, despite skin that generously apposes and slightly overlies the headpiece. the apparently innocuous and perhaps tedious nature of this Sutures can be used to close the rostral and caudal aspects of the aspect of the surgical procedure. Reported risks to surgical per- incision, if necessary. Attention to these details is important, be- sonnel while using acrylic cement include contact dermatitis, cause, in our experience, skin surrounding the headpiece often vapor permeation of soft contact lenses, irritation of the respira- recedes during the first few days after surgery. This recession may tory tract and eyes, and headaches (22–25). These risks can be result from drying or devascularization of the skin margins during reduced by appropriate ventilation and the use of scavenging surgery or might simply be the result of muscle or jaw movements systems (23–25). that pull the skin ventrally and away from the headpiece. Regard- In summary, if sterile technique is appropriately used, the skull less of the cause of the skin recession, achieving adequate is properly cleaned, and acrylic cement infiltrates crevices of the coverage of the skull and good apposition of the skin to the head- bone surface and hardens completely, an implanted headpiece,

60 CONTEMPORARY TOPICS © 1999 by the American Association for Laboratory Animal Science Volume 38, No. 1 / January 1999 thereby permitting electrophysiologic recording of various regions of a given target structure. Surgical procedures used to construct a headpiece that will permit the use of removable electrodes are identical in many respects to those described previously; the primary difference is that a hollow chamber, rather than an electrode assembly, is attached to the skull of the animal (Figure 7).This chamber provides access to the brain via a hole that is drilled through the skull inside the perimeter of the chamber. The chamber also protects the exposed brain between experiments. During experiments, an external micromanipulator is mounted onto the chamber, and an electrode is gradually inserted into the brain. These micromainpulators are functionally similar to those found on stereotactic frames, but typically they permit finer control of electrode movement, attach securely onto the recording chamber, and can be con- trolled remotely. Because the use of removable electrodes requires prolonged exposure of the brain surface, special procedures are needed to prevent infection or desiccation of the exposed brain and dura mater. A number of methods have been developed to achieve these goals, but to our knowledge, their relative effectiveness has not been compared or documented empirically. However, a method of chamber maintenance that we and others have used with satisfactory results is to fill the recording chamber with sterile isotonic saline solution that contains a broad-spectrum antibiotic. This solution is replaced at regular (usually daily) in- tervals. The solution can be removed from the chamber FIG. 7. Diagram of an implant for electrophysiologic recording in pri- conveniently with a sterile Pasteur pipette attached to a vacuum mates, using removable electrodes. A cylindrical recording chamber is device. The chamber is flushed repeatedly with sterile physiologic attached to the skull, and a hole is drilled through the skull inside the saline solution, and the chamber interior can be cleaned by us- chamber (A). During recording sessions, a micromanipulator is attached ing a sterile cotton-tipped swab to remove debris adhering to to the chamber and used to insert a recording electrode into the brain the interior wall. After thorough irrigation is completed, the (B). Between experimental sessions, the recording chamber is typically chamber is refilled with an antibiotic-containing solution. We filled with a sterile physiologic saline solution that contains a broad- typically use a solution of chloramphenicol (0.5 ml of solution spectrum antibiotic solution and then is sealed. of 100 mg of chloramphenicol/ml), which has a relatively broad antimicrobial spectrum and, in our experience, is not associated piece is facilitated by carefully contouring the incision to permit with adverse effects. A sterile cap is then used to securely seal a narrow margin of skin to remain over the edge of the head- the chamber. When recordings are made, the chamber is com- piece at the conclusion of surgery. During surgery, protective monly flushed with physiologic saline solution before recording ointment can be applied to skin margins to prevent excessive is initiated and after completion of recording. drying of the skin and to promote dermal viability. Despite the use of antibiotics in the recording chamber, every All wounds are likely to heal most quickly with least risk of effort must be made to prevent contamination of the chamber infection if they are kept clean. This can be promoted by using a interior. This requires the use of sterile materials and aseptic sterile cotton-tipped swab to liberally apply antibiotic ointment technique during flushing and filling of the chamber as well as beneath and over the wound margin at the end of surgery. In during experimental procedures. To reduce the likelihood that our experience, healing seems to be facilitated when antibiotic the electrode will introduce microorganisms into the brain, many ointment is applied daily for 7 days after surgery. Systemic anti- investigators sanitize the electrode before use by immersing it in biotics can also be used before and after surgery. Chronic a 70% solution of ethanol and rinsing it with sterile saline solu- recurrence of crusty scabs around the headpiece can result from tion. This method seems to work well, but we prefer to sterilize poor apposition of the skin to the headpiece, removal of exces- the electrode by immersing the shaft for at least 10 min in a sive amounts of skin during surgery, or skin abrasion caused by solution of chlorine dioxide prepared according to rough regions of cement. The last problem can be alleviated by manufacturer’s instructions. Other sterilizing solutions might anesthetizing the animal and applying an additional thin coat also be satisfactory, provided the electrode is thoroughly rinsed of cement to the headpiece or smoothing the existing cement after sterilization and the sterilizing solution does not interact with a dental drill. Applying antibiotic ointment to the incision chemically with the electrode insulation. site for the next 1 or 2 days will help keep the skin margins clean Piercing the dura with the electrode can occasionally cause a and pliable and appears to promote healing and good apposi- slight behavioral response in a clinically normal animal, but in tion of the skin with the headpiece. most cases, animals, like human neurosurgery patients, do not appear to notice electrode penetration. This situation can change markedly if infection develops and the dura becomes inflamed. Long-Term Maintenance of In such cases, electrode penetration of the dura can elicit a vig- Recording Chambers orous behavioral response from the animal. Infections can Neuronal recordings are often collected from conscious ani- develop if microorganisms that are not sensitive to the antibi- mals by using recording electrodes that are inserted and otic are inadvertently introduced into the chamber. In such removed on a regular (usually daily) basis. An important ad- instances, swabs of the skull surface inside the chamber can be vantage of this system over permanently implanted electrodes submitted for bacterial culture, and the antibiotic susceptibility is that the electrodes can be inserted at multiple locations, of the infecting organism can be determined so that an appro-

Volume 38, No. 1 / January 1999 CONTEMPORARY TOPICS © 1999 by the American Association for Laboratory Animal Science 61 priate antibiotic regimen (systemically or locally administered) Proper application of these principles in experimental animals can be initiated. If a fungal organism is identified, a solution of will permit neuroscientists to humanely study a wide range of chlorine dioxide applied directly into the chamber can be a very important problems that could not otherwise be investigated. effective treatment (26). This solution should be replaced daily for at least 3 days. Recording sessions should be discontinued if Acknowledgments infection is detected, because electrode penetration can poten- The authors thank Drs. Peggy Danneman, Tim Mandrell, and tially transfer the microorganism into brain parenchyma, thereby Amy L. B. Frazier for assistance with this manuscript. This study promoting development of encephalitis. was supported, in part, by NS-26429, RR-08864, CA-21765, and Infections associated with contamination of the recording the American Lebanese Syrian Associated Charities (ALSAC). chambers can occur occasionally, but proper chamber maintenance should make them rare occurrences. If an animal has recurrent infections, procedures used for chamber maintenance References should be reviewed. Methods other than those described here 1. Kelland, M. D., A. S. Freeman, and L. A. Chiodo. 1989. can also be used to achieve successful long-term maintenance of Chloral hydrate anesthesia alters the responsiveness of iden- recording chambers. For example, instead of using a sterile sa- tified midbrain neurons to dopamine agonist line solution to fill the chamber, some investigators apply administration. Synapse 3:30–37. antibiotic ointment directly to the dura as a protective coating. 2. Tsoukatos, J., Z. H. T. Kiss, K. D. Davis, R. R. Tasker, and In rats, a thin layer of sterile silastic can be placed in the base of J. O. Dostrovsky. 1997. Patterns of neuronal firing in the the chamber to act as a protective barrier while still permitting human lateral thalamus during sleep and wakefulness. Exp. electrode penetration (10). Similar procedures have been used Brain Res. 113:273–282. in cats (27). Other protocols may be completely acceptable, pro- 3. Crawley, J. N., C. R. Gerfen, R. McKay, M. A. Rogawski, D. viding they maintain a sterile environment in the chamber R. Sibley, and P. Skolnick (eds.). 1998. Current protocols interior and do not irritate the dura or underlying brain tissue. in neuroscience on CD-ROM. Unit 7.2: Microdialysis in ro- A final procedure associated with long-term maintenance of dents. John Wiley & Sons, Inc., New York. recording chambers is the infrequent or periodic need to re- 4. Kruger, L., S. Saporta, and L. W. Swanson. 1995. Photo- move granulation tissue from the surface of the dura. This tough graphic atlas of the rat brain: the cell and fiber architecture fibrous material develops along the margins of the craniotomy illustrated in three planes with stereotaxic coordinates. p. and can gradually advance over the intended recording target, 1–23. Cambridge University Press, Cambridge. where it may damage or deflect delicate electrodes as they pass 5. Paxinos, G. and C. Watson. 1986. The rat brain in stereo- through the dura. This granulation tissue can be removed by a taxic coordinates. p. viii–xi, Academic Press, New York. procedure commonly known as dural scraping in which sterile 6. Slotnick, B. M. and D. L. Brown. 1980. Variability in the technique and careful dissection are used to debride the dura stereotaxic position of cerebral points in the albino rat. of an anesthetized animal. The frequency with which dural Brain Res. Bull. 5:135–139. scraping must be performed may be reduced if a trephine is 7. Paxinos, G., C. Watson, M. Pennisi, and A. Topple. 1985. used to make small openings in the skull only as they are Bregma, lambda and the interaural midpoint in stereotaxic needed, as opposed to drilling a single large opening in the surgery with rats of different sex, strain and weight. J. skull at the time of chamber attachment. The trephining pro- Neurosci. Meth. 13:139–143. cedure can easily be performed repeatedly on an anesthetized 8. Shek, J. K., G. Y. Wen, and H. M. Wisniewski. 1986. Atlas of animal, and the use of multiple small openings that are made the rabbit brain and spinal cord. p. ix–xiii, Karger Press, only as they become needed to accommodate new target lo- New York. cations usually reduces problems associated with excessive 9. Franklin, K. J. B. and G. Paxinos. 1997. The mouse brain dural granulation tissue. in stereotaxic coordinates. p. ix– xi, Academic Press, New York. Conclusions 10. Gardiner, T. W. and S. T. Kitai. 1992. Single-unit activity in One of the primary goals of neuroscience research is to char- the globus pallidus and neostriatum of the rat during per- acterize brain function, under conditions that are as normal as formance of a trained head movement. Exp. Brain Res. possible. For many types of studies, causing pain or distress to 88:517–530. the subject during the experiment will prevent collection of 11. Parry, T. J. and J. G. McElligott. 1993. A method for re- interpretable data. The ability to perform experiments in con- straining awake rats using head immobilization. Physiol. scious animals without inflicting pain or distress is as important Behav. 53:1011–1015. to the research effort as it is to the animal and animal care 12. Yamamoto, B. K. and C. L. Kutscher. 1981. The use of fluo- personnel. The use of proper surgical techniques and appro- ride for the prevention of chronic intracranial implant priate postoperative care can result in long-lasting implants that dislodgement. Pharmacol. Physiol. Behav. 15:663–664. cause minimal animal discomfort. Interestingly, the same prin- 13. Frommer, G. 1985. Personal communication. ciples that we have outlined in this review as being important 14. Bloch, B., J. K. Haken, and G. Hastings. 1970. Evaluation for the implantation and maintenance of headpieces in ani- of cold-curing acrylic cement for prosthesis stabilization. mals are also viewed as central to the success of reconstructive Clin. Orthop. Related Res. 72:239–241. cranial surgery using acrylic implants in humans. Important con- 15. Homsy, C. A., H. S. Tullos, and J. W. King. 1969. Evalua- siderations for reducing complications and achieving successful tion of rapid-cure acrylic compound for prosthesis long-term prosthetics in humans include obtaining clean bone stabilization. Clin. Orthop. Related Res. 67:169–171. surfaces that form a continuous bond with acrylic cement, prop- 16. Meyer, P. R., E. P. Lautenschlager, and B. K. Moore. 1973. erly anchoring the implant to the bone, smoothly contouring On the setting properties of acrylic bone cement. J. Bone the implant, eliminating crevices between the bone and acrylic, Joint Surg. 55A:149–156. and preserving well-vascularized skin adjacent to the implant 17. Hamlen, H. J. and E. Olson. 1995. Methylmethacrylate-as- (20). Standard postoperative measures include prophylactic ad- sociated thermal injury during cranioplasty in a nonhuman ministration of antibiotics and irrigation of the implant (20). primate. Contemp. Topics Lab. Anim. Sci. 34:73–75.

62 CONTEMPORARY TOPICS © 1999 by the American Association for Laboratory Animal Science Volume 38, No. 1 / January 1999 18. Kepes, E. R., P. S. Underwood, and L. Becsey. 1972. Intra- 23. Wesley, R. E. and J. D. Brinsko. 1992. Toxicity of methyl operative death associated with acrylic bone cement. JAMA. methacrylate monomer in orbital and cranial surgery. Ann. 222:576–577. Ophthalmol. 24:307–309. 19. Svartling, N., A.-M. Lehtinen, and L. Tarkkanen. 1986. The 24. Walton, N. Y., J. Kang, and D. M. Treiman. 1994. An inex- effect of anaesthesia on changes in blood pressure and pensive method for control of methyl-methacrylate fumes plasma cortisol levels induced by cementation with in the laboratory. Contemp. Topics Lab. Anim. Sci. 33:62– methylmethacrylate. Acta Anaesth. Scand. 30:247–252. 64. 20. Persing, J. A., A. J. Cronin, J. B. Delashaw, M. T. Edgerton, 25. Howmedica Surgical Simplex® P Radiopaque Bone Ce- S. L. Henson, and J. A. Jane. 1987. Late surgical treatment ment. 1997. Howmedica International, Rutherford, NJ. of unilateral coronal synostosis using methyl methacrylate. 26. Lee, G. E., P. J. Danneman, R. D. Rufo, R. P. Kalesnykas, J. Neurosurg. 66:793–799. and V. E. Eng. 1998. Use of chlorine dioxide for antimicro- 21. Duncan, J. A. T. 1988. Intra-operative collapse or death re- bial prophylactic maintenance of cephalic recording devices lated to the use of acrylic cement in hip surgery. Anaesthesia in rhesus macaques (Macacca mulatta). Contemp. Topics 44:149–153. Lab. Anim. Sci. 37:59–63. 22. Mattia, M. A. 1983. Hazards in the hospital environment: 27. McElligott, J. G. and B. D. Waterhouse. 1977. Microelec- anesthesia, gases and methylmethacrylate. Am. J. Nurs. trode implantation technique for use in the awake and 83:73–77. unrestrained cat. Physiol. Behav. 18:163–167.