Use of Electrical Bone Stimulation in Spinal Fusion Scott D. Hodges, DO, Jason C. Eck, DO, MS, and S. Craig Humphreys, MD Abstract The rate of pseudarthrosis after spi- nal fusion depends on many factors, including surgical approach, num- ber of levels fused, type of bone graft material, use of instrumentation, and number of previous surgeries. Also, obtaining a solid spinal fusion often is hampered by patient traits, such as excessive weight, smoking, and use of steroids or NSAIDs. 1 Ad- ditionally, the radiographic criteria used to assess the presence or ab- sence of fusion vary, thus affecting the consistency of reported results. The incidence of pseudarthrosis af- ter spinal fusion may be as high as 35%, but not all patients are clini- cally symptomatic. 2 The high rate of pseudarthrosis has forced surgeons to identify factors contributing to nonunion and to seek methods to improve their ability to achieve solid fusion. Although advances in pa- tient selection and surgical tech- niques have helped decrease the rate of pseudarthrosis, further improve- ment is necessary. For almost five decades, various types of electrical stimulation have been investigated and used clinical- ly in the management of nonunion of long bones, but only recently have they been used for spinal fusion. The three major techniques used to de- liver electrical stimulation to bones aredirect current (DC) delivered through surgically implanted electrodes, puls- ing electromagnetic fields (PEMFs), and capacitively coupled electrical stimulation. Each method has differ- ent mechanisms of action as well as varying amounts of supporting lab- oratory data and relevant clinical stud- ies regarding safety and efficacy. Under- standing the scientific data regarding electrical stimulation can help de- termine whether to use these tech- niques and in which patients electri- cal stimulation may improve fusion success. Mechanism of Action In the 1950s, Yasuda et al 3 studied electrical stimulation to promote bone growth in a rabbit model. Increased callus formation was reported at the site of an implanted cathode. In the 1960s, it was found that electrical potentials are generated when bone is compressed, which leads to bone deposition by the piezoelectric ef- fect 4 (Fig. 1). This theory states that loading of bone creates a negative potential in the compression area and a positive potential in surrounding areas. The small current from these electrical potentials stimulates osteo- genesis at the site of compression. These findings led researchers to in- vestigate the efficacy of applying elec- trical energy to enhance osteogen- esis. The exact mechanism of action of electrical stimulation is incomplete- ly understood. Some studies have Dr. Hodges is Orthopaedic Spine Surgeon, Cen- ter for Sports Medicine and Orthopaedics, and Director of Research, Foundation for Research, Chattanooga, TN. Dr. Eck is Orthopaedic Sur- gery Intern, Memorial Hospital, York, PA, and Research Associate, Center for Sports Medicine and Orthopaedics, Foundation for Research. Dr. Humphreys is Orthopaedic Spine Surgeon, Cen- ter for Sports Medicine and Orthopaedics, Foun- dation for Research. Reprint requests: Dr. Hodges, Suite 303, 605 Glenwood Avenue, Chattanooga, TN 30404. Copyright 2003 by the American Academy of Orthopaedic Surgeons. Spinal fusion is commonly done to manage deformity, restore stability, and eliminate excessive motion at specific spinal levels. Pseudarthrosis limits the clinical success of spinal fusion. Three types of electrical stimulation, which is used to manage non- union in long bones, recently have been applied in an attempt to enhance the rate of spinal fusion. Direct current electrical stimulation is internal and thus eliminates dependence on patient compliance. Pulsed electromagnetic fields and capacitively coupled electrical stimulation are external techniques that require patient compli- ance but do not have the increased risk associated with implantable devices. Firm conclusions about efficacy are difficult to establish because of inconsistencies in both determining a reliable, reproducible end point for fusion and in incorporating the effect of patient parameters. Most data indicate a positive effect for use of direct cur- rent stimulation, but further studies are necessary to determine its appropriateness as an adjuvant to spinal fusion. J Am Acad Orthop Surg 2003;11:81-88 Perspectives on Modern Orthopaedics Vol 11, No 2, March/April 2003 81 shown that fibrocytes and collagen fibers grow along the gradient of an applied electrical current. 5,6 Addi- tionally, increased DNA and col- lagen synthesis by fibroblasts occurs in the presence of electrical currents. These results help explain how the induction of an electrical current can stimulate both the growth and di- rection of growth of new cells and matrix. This preclinical information led to the clinical use of electrical current to induce osteogenesis and improve fracture healing. 7,8 Others reported increased oxy- gen consumption and the produc- tion of hydroxyl radicals through an oxidative-reduction reaction at the site of applied cathodes. 9-11 These re- actions decrease the oxygen tension and raise the local pH at the site of the applied electrical current 9 (Fig. 1). The combination of low pO 2 and increased pH is more advantageous for osteogenesis because it also oc- curs in physeal plates and fracture callus. 10,11 This suggests that the ap- plication of current also may pro- duce an environment conducive to bone growth. Application of electrical current also can induce osteogenesis through increases in production and activation of biomechanical media- tors such as cyclic adenosine mono- phosphate (cAMP). 12,13 The activa- tion of cAMP leads to numerous metabolic reactions, including carti- lage and bone formation. Further- more, intracellular calcium 14,15 and prostaglandin E 2 16 increase after elec- trical stimulation. In an attempt to further describe the biochemical pathway used in sig- nal transduction by bone cells sub- jected to an electrical field, Lorich et al 17 investigated the effect of a ca- pacitively coupled electrical field on rat calvarial bone cells and mouse MC3T3-E1 bone cells. Various signal transduction inhibitors were used to determine the biochemical pathway. The inhibitors and their mechanisms of action are shown in Table 1. Each experimental run was done on four groups of bone cell cultures: (1) con- trol, (2) stimulated, (3) control plus inhibitor, and (4) stimulated plus in- hibitor. Electrically stimulated bone cells had statistically significant in- creases in DNA content of 22.2% (P = 0.03) to 45.3% (P = 0.001) com- pared with the control groups. Indo- cin (4 µg/mL; P = 0.01), verapamil (20 µmol; P = 0.005), W-7 (1 µmol; P = 0.001), and 4-bromophenacyl bro- mide (10 µmol; P = 0.03) completely blocked stimulation of bone cells but had no effect on control cells. How- ever, neomycin did not block the stimulation of bone cells. The au- thors determined that the signal trans- duction mechanism involves trans- membrane calcium translocation in response to voltage-gated channels, activation of phospholipase A 2 in the cell membrane, and an increase in prostaglandin E 2 . W-7, the cal- modulin antagonist, blocked stimu- lation of bone cells, revealing that calmodulin activation also was a nec- essary component of the pathway. If an inhibitor blocked any of these components, the bone cells showed no response to electrical stimulation. Because the control group exhibited the same amount of growth regard- less of the presence of the inhibitors, this pathway was thought to be nec- essary only for the response of bone cells to the electrical stimulation. There appear to be differences be- tween the pathways for electrical and mechanical stimulation of bone growth. Brighton et al 18 showed that the inositol phosphate cascade is crucial to the mechanical stimula- tion of bone cells. Blocking this cas- cade completely eliminated the ef- fects of mechanical stimulation on bone growth. However, in the study by Lorich et al, 17 blocking the cas- cade with neomycin had no effect on the electrical stimulation of bone cells. Figure 2 provides an overview of the proposed signal transduction pathways in both electrical and me- chanical stimulation on bone growth + + + + + + pO2 pH −−−−−− Tension Compression Figure 1 Effects of the piezoelectric effect. Compressive stress causes a negative elec- trical potential at the area of compression and a positive potential in surrounding ar- eas, which leads to a decrease in pO 2 and an increase in local pH. Table 1 Signal Transduction Inhibitors and Their Corresponding Action 17 Inhibitor Action Neomycin Blocks inositol phosphate pathway by inhib- iting phospholipase C−mediated hydrolysis of phosphatidylinositol 4,5-bisphosphate (PIP 2 ) Indocin Inhibits prostaglandin synthesis Verapamil Calcium channel blocker N-(6-aminohexyl)-5-chloro- 1-naphthalenesulfonamide hydrochloride (W-7) Calmodulin antagonist 4-Bromophenacyl bromide Inhibits phospholipase A 2 during the synthe- sis of prostaglandin E 2 Use of Electrical Bone Stimulation in Spinal Fusion 82 Journal of the American Academy of Orthopaedic Surgeons and shows the variations in the two pathways. Electromagneticfieldsgenerateelec- trochemical gradients in the tissue fluid surrounding cells that are weak- er than the gradient needed to cause cellular depolarization. Luben et al 19 hypothesized that the weak gradi- ent depends on a series of amplifi- cation mechanisms to generate cel- lular effects. Monolayer osteoblast cell cultures were exposed to an elec- trical gradient. After electrical stim- ulation, cell cultures were treated with hormones—parathyroidhormone(PTH), osteoclast-activating factor (OAF), and 1,25-dihydroxyvitamin D 3 —and as- sayed to evaluate cAMP accumula- tion, adenylate cyclase activity, and collagen synthesis. Increased levels of cAMP are present in bone cells after exposure to PTH. 20 However, the PTH- and OAF-stimulated in- creases in cAMP were markedly re- duced when osteoblast cell cultures were stimulated with electromag- netic fields. 19 The decrease in cAMP was the result of a decreased activa- tion of adenylate cyclase in the cell membrane by PTH and OAF. How- ever, total levels of adenylate cyclase activity were not affected by the elec- trical stimulation, which suggests that the decrease in activation of aden- ylate cyclase is caused by an effect on the membrane receptors for PTH and OAF, not for adenylate cyclase. Additionally, the inhibition of col- lagen synthesis generally found in the presence of PTH was eliminated by electrical stimulation. 21 Electrical stimulation had no effect on collagen synthesis resulting from 1,25-dihy- droxyvitamin D 3 , which acts through cytoplasmic effects, not through mem- brane receptors. Luben et al 19 con- cluded that the increased bone forma- tion generated from electrical stimulation resulted from inhibition of membrane receptors for PTH and OAF. Direct Current Stimulation DC stimulation involves implanting a cathode during fusion surgery and placing it in direct contact with the bone graft and decorticated verte- bral elements. The stimulation ap- pears to be most effective within 5 to 8 mm of the cathode. The cathode wire can be coiled to increase the bony surface area being stimulated. The cathode is connected by insulat- ed wire to a subcutaneously placed battery pack (Fig. 3). The electrical stimulation is generally continued for 6 to 9 months after fusion. Mag- netic resonance imaging (MRI) is not recommended after implantation of an internal stimulator. There are no data regarding the use of MRI with the leads only, but the large amount of artifact produced by the presence of the implant limits the usefulness of the images. There have been no reports of harm done to patients un- dergoing MRI with an internal stim- ulator in place. Electrically induced osteogenesis has been reported to exhibit a dose- response curve as a function of the amount of current delivered. A cur- rent between 5 and 20 µA produces the greatest amount of osteogenesis. Tissue necrosis is likely at currents >20 µA. 22 Animal Studies Nerubay et al 23 evaluated the ef- fect of DC stimulation on lumbar spine fusion in a swine model. Cur- rent was applied to the fusion site in 15 pigs; 15 control pigs received no stimulation. Radiographic and his- tologic results were reviewed post- operatively. At 2 months, there was a significant (P < 0.01) increase in fusion success in the group receiv- ing DC stimulation. Kahanovitz andArnoczky 24 report- ed similar results in a canine model. Twelve dogs underwent bilateral spine fusion at L1-L2 and L4-L5. A DC stimulator was implanted at each fac- et fusion site. Half of the devices Electrical stimulation Voltage-gated Ca 2+ channel Ca 2+ PTH receptor OAF receptor PGE 2 PGE 2 PLA 2 Verapamil Extracellular space Cell membrane Cytoplasm W-7 Activation of calmodulin Osteoblast proliferation Mechanical stimulation Activation of inositol phosphate cascade Neomycin (−) (−) (−) (−) (−) Ca 2+ Ca 2+ (−) Indocin BPB Figure 2 Diagram of the cellular effects of electrical and mechanical stimulation on bone cells. PGE 2 = prostaglandin E 2 , PLA 2 = phospholipase A 2 , BPB = 4-bromophenacyl bromide, W-7 = N-(6-aminohexyl)-5-chloro-1-naphthalenesulfonamide hydrochloride, Ca 2+ = cal- cium ion, − = inhibition of action, PTH = parathyroid hormone, OAF = osteoclast-activating factor. Scott D. Hodges, DO, et al Vol 11, No 2, March/April 2003 83 were activated; half acted as con- trols. At 12 weeks, all of the levels receiving electrical stimulation had evidence of solid fusion, whereas none of the control levels was fused (P < 0.004). Clinical Studies In 1988, Kane 25 investigated the efficacy of DC stimulation in spinal fusion. This prospective random- ized placebo-controlled study in- cluded a mixed group of 59 patients at high risk for nonunion who had either a history of failed prior fu- sion, grade II or worse spondylolis- thesis, or required multiple-level fu- sion. Radiographic evidence of fusion was assessed at 12 to 18 months postoperatively by the sur- geons and a blinded radiologist. There was a statistically significant (P = 0.026) increase in radiographic evidence of fusion in the stimulated patients (81% [25/31]) compared with the control patients (54% [15/ 28]). In 1994, Meril 26 compared inter- body spinal fusion rates in a pro- spective randomized study of pa- tients treated with and without DC stimulation. Patients underwent a mod- ified Crock procedure with allograft using either an anterior or posterior approach. The experimental group received 20 µA of DC for a mini- mum of 24 weeks.Atamean follow-up of 21 months, 93% of patients (113/ 122) in the experimental group had evidence of solid fusion, whereas only 75% of control patients (77/ 103) showed successful fusion (P = 0.0003). Importantly, the authors re- ported statistically significantly greater fusion success in various high-risk patient groups, including smokers (92% [85/92] versus 71% [42/59]; P = 0.001), noninstrumented cases (91% versus 65% [P = 0.0006]), and L4-L5 fusions (91% versus 59% [P = 0.003]). In 1996, Rogozinski and Rogozin- ski 27 investigated the effects of DC stimulation on instrumented spinal fusion. An initial series of 26 pa- tients received no stimulator, fol- lowed by a series of 42 patients who did. An additional 26 patients were randomized into the study (11 re- ceived a stimulator, 15 did not). All patients underwent posterolateral lumbosacral fusion using autolo- gous bone graft and pedicle screw instrumentation. Ninety-six percent of the experimental patients (51/53) had radiographic evidence of solid fusion versus 85% of the control pa- tients (35/41) at a mean follow-up of 20.5 months (P = 0.02). Fusion was determined radiographically by the surgeons. As in the study by Meril, 26 statistically significant im- provement in fusion rates was re- ported in high-risk patients, includ- ing those with previous surgery (100% versus 79% [P = 0.02]) and A B C Figure 3 A 46-year-old woman reported back pain and bilateral lower extremity pain. Her pain score was 9 of 10 on a visual analog pain scale, with an Oswestry score of 29 of 50, indicating severe disability in activities of daily living. Lumbar spine disk degeneration and collapse of L4-5 and L5-S1 were evident on MRI and radiographs. A, Anteroposterior radiograph showing collapse of the L5-S1 disk space. The patient underwent transforaminal lumbar interbody fusion of L4-5 and L5-S1 using iliac crest bone graft. B, At 5 months postopera- tively, the patient experienced continued soreness and felt as if the hardware were moving. Although radiographs showed the hardware to be in excellent position with evidence of fusion, 6 months postoperatively the patient still had pain. Reexploration revealed gross movement and pseudarthrosis. C, A DC bone stimulator (arrow) was implanted with application of autologous growth factors. At 6 weeks after the second procedure, the patient reported 90% improvement and requested to return to work. At 8-month follow-up, she reported mild to moderate low back pain, but radiographs revealed solid fusion, with hardware and stimulator in excellent position. Use of Electrical Bone Stimulation in Spinal Fusion 84 Journal of the American Academy of Orthopaedic Surgeons multiple-level fusions (95% versus 81% [P = 0.04]). In 1999, Kucharzyk 28 evaluated the efficacy of DC stimulation on a series of 130 nonrandomized patients undergoing posterior instrumented spinal fusion with posterolateral au- tologous bone graft. The first group of 65 patients received no stimula- tion; the second group of 65 patients did. Follow-up radiographs were eval- uated according to Dawson’s crite- ria (A0, no evidence of fusion, to A4, solid fusion with hypertrophy of the graft). 29 Clinical success was evalu- ated with the modified Smiley- Webster surgical rating scale. Fusion success was 95.6% in the experimen- tal patients versus 87% in the con- trol patients (P = 0.02). In patients involved with workers’ compensa- tion, fusion success was 93% in the experimental group versus 81% in the control group. Jenis et al 30 compared the efficacy of DC and PEMF stimulation on in- strumented lumbar fusion in a study of 61 patients randomized to one of three treatment protocols for post- operative fusion augmentation. All patients underwent a posterior lum- bar approach, with or without decom- pression, and bilateral lateral inter- transverse process fusion and insertion of pedicle screw–rod instrumenta- tion. The three treatment protocols for postoperative fusion augmenta- tion were PEMF, DC, or no stimula- tion. At 1-year follow-up, no signif- icant differences were evident among the three groups in terms of either radiographic evidence of fusion or bone density measurements of fu- sion mass. The authors concluded that electrical stimulation with DC or PEMF as an adjunct to postero- lateral instrumented lumbar fusion did not result in improved fusion rate or clinical outcome. Possible ex- planations for the lack of added ben- efit from the stimulation are the small number of patients in each group and a high fusion success rate (95.3%) in the control group. 30 Pulsing Electromagnetic Fields In PEMFs, an external coil generates an electromagnetic field that in turn induces an electrical current at the fusion site. This technique has gained popularity because of its noninvasive nature, but the device has to be worn 8 to 10 hours a day for 6 to 8 months, and its efficacy depends on patient compliance. Animal Studies In 1994, Kahanovitz et al 31 found no benefit from PEMFs on posterior lumbar spinal fusions in a canine model. Bilateral posterior facet fu- sions were done at L1-L2 and L4-L5 in 24 dogs. Eight received pulsing for 30 min/day and eight for 60 min/day; eight controls received no pulsing. No radiographic or histo- logic evidence of a difference in fusion mass between the experi- mental and control groups was ap- parent. Guizzardi et al 32 assessed the ben- efits of PEMF stimulation in relation to callus development after poste- rior spinal fusion in a rat model. Six rats received pulsing; six others served as controls. Two electromag- netic coils were fixed to the outside of each cage of the experimental rats, and they received an average in- duced electrical field of 2.5 mV for 18 h/day. The rats were sacrificed at 4 and 8 weeks postoperatively. At 4 weeks, the experimental rats had an accelerated bone callus organization and early replacement of the carti- lage tissue with bone. However, at 8 weeks, there was no substantial dif- ference between the experimental and control groups. The authors con- cluded that during the early stages, PEMF appears to accelerate the de- velopment of callus, favoring the transformation of cartilage to bone. However, it was not possible to de- termine if the initial histologic accel- eration resulted in a higher rate of fusion. In 1997, Glazer et al 33 investigated the use of electromagnetic fields for posterolateral spinal fusion in a rab- bit model. The experimental group re- ceived pulsingfor4h/dayfor 6 weeks. Radiographic evidence of pseudar- throsis decreased from 40% to 20% with PEMF stimulation, but this de- crease was not statistically signifi- cant because of the small number of animals in each group. Biomechani- cal studies showed 37% greater stiff- ness (P = 0.02), a 35% increase of the fusion mass (P = 0.04), and a 42% increase in the load to failure of the fusion mass (P = 0.02) in the exper- imental group compared with matched controls. Clinical Studies Simmons 34 applied PEMFs to 13 patients with failed posterior lum- bar interbody fusion at a mean of 40 months after index surgery. At the conclusion of the 4-month treatment period, there was an increase (P val- ue not reported) in bone mass in 85% of the patients (11/13) and a solid fusion in 77% (10/13). The pre- PEMF and post-PEMF radiographs were evaluated by an independent reviewer who provided an impres- sion of each spinal level, but specific criteria used to evaluate and quan- tify increased bone mass or solid fusion were not specified. In another clinical study, 35 all 195 patients undergoing lumbar inter- body fusion using either an anterior or posterior approach were divided into two groups. Patients in the ex- perimental group were instructed to wear a brace with electromagnetic coils at least 8 h/day. All patients were evaluated at a minimum of 12 months postoperatively. Radiograph- ic evidence showed a 92% fusion rate in the experimental group (90/ 98) versus 65% in the control group (63/97) (P < 0.005). Fusion was solid when it was more than 50% assimi- lated. This study also showed in- creased fusion rates for high-risk pa- tients, including smokers (89% versus Scott D. Hodges, DO, et al Vol 11, No 2, March/April 2003 85 60%), and for multiple-level fusions (89% versus 54%); however, P val- ues were not provided. Capacitively Coupled Electrical Stimulation Capacitively coupled electrical stim- ulation involves a small computer- controlled stimulator that delivers current through flexible cables to hy- drogel surface electrodes using a si- nusoidal waveform. The leads are placed on either side of the spine at the level of the center of the fusion mass. Patients are instructed to wear the stimulator 24 hours a day until healing has occurred. As with PEMFs, this technique is noninva- sive, but it also is dependent on pa- tient compliance. Capacitively coupled electrical stim- ulation has been successful in the treatment of nonunion of long bones. Scott and King 36 investigated the ef- ficacy of capacitively coupled elec- trical stimulation in 21 patients with an established nonunion in a long bone. Ten patients received stimula- tion; 11 served as controls. Patients in the experimental group were in- structed to wear the device for up to 6 months or until healing occurred. Nonunion was healed in 6 of the 10 experimental patients, but no heal- ing occurred in the control group. Goodwin et al 37 conducted a mul- ticenter randomized double-blind pro- spective study comparing a capaci- tively coupled electrical stimulation group with a control group. Of the initial 337 patients, 34 had not com- pleted treatment at the time of anal- ysis, 20 had missing data, and 63 had been eliminated from the study (either voluntarily or because of non- compliance, adverse reactions, wound infection, surgeon protocol violations, or relocation). Of the 220 patients with all clinical data, only 179 had completed radiographic data, leav- ing only 53% of the original patients for the final study group. The patients underwent lumbar fusion with posterior or anterior lumbar interbody fusion or postero- lateral fusion; received autograft, al- lograft, or a combination of both; and had instrumentation other than interbody cages. Patients were in- structed to use the stimulator 24 hours a day until healing occurred or for 9 months if healing had not yet occurred. Overall success was based on both radiographic and clin- ical success. Blinded radiographs were reviewed by independent radiolo- gists. Fusion success was contingent on the presence of mature-appearing uninterrupted bony masses bilater- ally at the fusion levels. Overall, there was an 84.7% combined clinical and radiographic success rate in the ex- perimental group versus 64.9% in the control group (P < 0.005). Eval- uated separately, the clinical success rates were 88.2% for the experimen- tal group and 75.5% for the control group (P = 0.034); radiographic suc- cess was 90.6% for the experimental group and 81.9% for the control group (not statistically significant). Greater clinical than radiographic success was evident when electrical stimulation was used. The greatest improvements were reported in pa- tients who had posterolateral fusion (P = 0.006) and internal fixation (P = 0.013). Although this study showed good overall success, analysis of the data for various subgroups revealed that patients with previous surgery, smokers, and those with herniated disks received no statistically signif- icant benefit from the stimulation compared with the placebo group. There also was no significant benefit from stimulation when the data were analyzed according to number of levels fused. Discussion Decreased medical funding and in- creased demand present a challenge to develop cost-effective techniques for spinal fusion. At prices ranging from $2,500 to $7,000, electrical stim- ulators are cost effective. However, each technique has benefits and lim- itations. DC stimulation has been more thoroughly evaluated for use with spinal surgery than have the other methods and has been effective in increasing fusion rates. The primary advantage of this technique is that, because of its surgical placement, it does not require patient compliance. However, an implanted device may add increased risk and discomfort, and the battery may leak or mal- function. Because the device is inter- nal, any malfunction would require another procedure to make repairs. There are no firm published esti- mates of rates of infection or device malfunction leading to surgical re- moval. The efficacy of PEMFs and capaci- tively coupled electrical stimulation cannot be verified because of lack of study data. These techniques do have the advantage of being exter- nally worn hardware, which elimi- nates the risk of infection and avoids potential removal. However, the de- vice is effective only when the pa- tient complies with instructions. No- tably lower success rates have been reported in patients who did not wear the device for the instructed amount of time. 37 This serious limi- tation should be considered, espe- cially in patients with a history of noncompliance. There is a major methodologic in- consistency in most clinical studies of the efficacy of electrical stimula- tion for enhancing spinal fusion. No universally agreed on and validated criteria exist to determine the pres- ence of either spinal fusion or pseud- arthrosis. Pseudarthrosis generally is defined as the failure of an at- tempted spinal fusion to fuse by 1 year postoperatively. Various crite- riahavebeenusedtodiagnosepseud- arthrosis, including evidence of motion on flexion/extension radio- Use of Electrical Bone Stimulation in Spinal Fusion 86 Journal of the American Academy of Orthopaedic Surgeons graphs, density of the fusion mass, radiographic evidence of bridging bony trabeculae, scintigraphy, and the presence of clinical symptoms. Each technique is subject to error and rates of false-negative results up to 29%. 38 The most accurate method of evaluating pseudarthro- sis is reexploration of the fusion mass, but even this has limitations. 39 Although reexploration was not possible for the clinical studies dis- cussed here, direct reexploration and histologic methods were used to check for pseudarthrosis in the animal studies. 24,32 Positive results with use of DC were reported by Kahanovitz and Arnoczky. 24 The efficacy of PEMFs was supported in the studies by Guizzardi et al 32 and Glazer et al, 33 but Kahanovitz et al 31 found no sig- nificant differences compared with control animals. Without a reliable method to assess the presence of sol- id fusion in the clinical setting, it is difficult to fully accept the results of many of the electrical stimulation studies. Currently, in most centers, electri- cal stimulation devices are used on an individual basis. Indications may in- clude patients who smoke, have di- abetes, metabolic bonedisease, or mild to moderate osteoporosis, or who are undergoing revision or multiple-level fusions. No strict criteria for device removal are suggested in the litera- ture. Implanted devices are generally removed only when the patient com- plains of discomfort, when there is de- vice malfunction, or to allow for fu- ture ability to use MRI. Removal of batteries is not recommended unless there is a device malfunction or other complication. Summary Although many studies of the use of electrical stimulation to enhance spi- nal fusion have shown promising results, they exhibit numerous limi- tations, including poor patient ran- domization, retrospective design, and, for some of them, financial sup- port from device manufacturers. The most crucial limitation is the lack of an accurate means of assessing the presence of solid fusion. Radio- graphic criteria used to demonstrate fusion are inconsistent; thus, any re- ported success achieved by the elec- trical stimulators graded by radio- graphic criteria must be cautiously interpreted. Also, no direct compar- isons of the three techniques have been made. As a result, the benefits and limitations of each must be weighed to determine appropriate indications and methodology. Fu- ture research should provide further insight into the specific mechanisms by which electrical stimulation re- sults in bone growth and thereby lead to further advances in these techniques. References 1. Dimar JR II, Ante WA, Zhang YP, Glass- man SD: The effects of nonsteroidal anti-inflammatory drugs on posterior spinal fusions in the rat. Spine 1996;21: 1870-1876. 2. Grubb SA, Lipscomb HJ: Results of lumbosacral fusion for degenerative disc disease with and without instru- mentation: Two- to five-year follow-up. Spine 1992;17:349-355. 3. Yasuda I, Noguchi K, Sata T: Abstract: Dynamic callus and electric callus. J Bone Joint Surg Am 1955;37:1292-1293. 4. Bassett CAL, Becker RO: Generation of electric potentials by bone in response to mechanical stress. Science 1962;137: 1063-1064. 5. Bassett CAL, Pawluk RJ, Becker RO: Effects of electric currents on bone in vivo. Nature 1964;204:652-654. 6. Frank C, Schachar N, Dittrich D, Shrive N, deHaas W, Edwards G: Electromag- netic stimulation of ligament healing in rabbits. Clin Orthop 1983;175:263-272. 7. Borsalino G, Bagnacani M, Bettati E, et al: Electrical stimulation of human fem- oralintertrochantericosteotomies: Double- blind study.Clin Orthop 1988;237:256-263. 8. Stefan S, Sansen W, Mulier JC: Experi- mental study on the electrical imped- ance of bone and the effect of direct current on the healing of fractures. Clin Orthop 1976;120:264-267. 9. Brighton CT, Friedenberg ZB: Electrical stimulation and oxygen tension. Ann N Y Acad Sci 1974;238:314-320. 10. Brighton CT, Heppenstall RB: Oxygen tension in zones of the epiphyseal plate, the metaphysis and diaphysis: An in vitro and in vivo study in rats and rabbits. J Bone Joint Surg Am 1971;53: 719-728. 11. Brighton CT, Krebs AG: Oxygen ten- sion of healing fractures in the rabbit. J Bone Joint Surg Am 1972;54:323-332. 12. Black J: Tissue response to exogenous electromagnetic signals. Orthop Clin North Am 1984;15:15-31. 13. Norton LA, Rodan GA, Bourret LA: Epiphyseal cartilage cAMP changes pro- duced by electrical and mechanical per- turbations. Clin Orthop 1977;124:59-68. 14. Brighton CT, Black J, Friedenberg ZB, Esterhai JL, Day LJ, Connolly JF: Amul- ticenter study of the treatment of non- union with constant direct current. J Bone Joint Surg Am 1981;63:2-13. 15. Wang Q, Zhong S, Ouyang J, et al: Osteogenesis of electrically stimulated bone cells mediated in part by calcium ions. Clin Orthop 1998;348:259-268. 16. Davidovitch Z, Shanfeld JL, Mont- gomery PC, et al: Biochemical media- tors of the effects of mechanical forces and electric currents on mineralized tissues. Calcif Tissue Int 1984;36(suppl 1):S86-S97. 17. Lorich DG, Brighton CT, Gupta R, et al: Biochemical pathway mediating the re- sponse of bone cells to capacitive cou- pling. Clin Orthop 1998;350:246-256. 18. Brighton CT, Fisher JR Jr, Levine SE, et al: The biochemical pathway mediat- ing the proliferative response of bone cells to a mechanical stimulus. J Bone Joint Surg Am 1996;78:1337-1347. 19. Luben RA, Cain CD, Chen MC, Rosen DM, Adey WR: Effects of electromag- netic stimuli on bone and bone cells in vitro: Inhibition of responses to par- athyroid hormone by low-energy low- frequency fields. Proc Natl Acad SciUSA 1982;79:4180-4184. 20. Peck WA: Cyclic AMP as a second mes- Scott D. Hodges, DO, et al Vol 11, No 2, March/April 2003 87 senger in the skeletal actions of para- thyroid hormone: A decade-old hy- pothesis. Calcif Tissue Int 1979;29:1-4. 21. Wong GL, Luben RA, Cohn DV: 1,25- dihydroxycholecalciferol and parathor- mone: Effects on isolated osteoclast-like and osteoblast-like cells. Science 1977; 197:663-665. 22. Friedenberg ZB, Andrews ET, Smolen- ski BI, Pearl BW, Brighton CT: Bone reaction to varying amounts of direct current. Surg Gynecol Obstet 1970;131: 894-899. 23. Nerubay J, Marganit B, Bubis JJ, Tad- mor A, Katznelson A: Stimulation of bone formation by electrical current on spinal fusion. Spine 1986;11:167-169. 24. Kahanovitz N, Arnoczky SP: The effi- cacy of direct current electrical stimu- lation to enhance canine spinal fusions. Clin Orthop 1990;251:295-299. 25. Kane WJ: Direct current electrical bone growth stimulation for spinal fusion. Spine 1988;13:363-365. 26. Meril AJ: Direct current stimulation of allograft in anterior and posterior lum- bar interbody fusions. Spine 1994;19: 2393-2398. 27. Rogozinski A, Rogozinski C: Efficacy of implanted bone growth stimulation in instrumented lumbosacral spinal fu- sion. Spine 1996;21:2479-2483. 28. Kucharzyk DW: A controlled prospec- tive outcome study of implantable elec- trical stimulation with spinal instru- mentation in a high-risk spinal fusion population. Spine 1999;24:465-469. 29. Dawson EG, Lotysch M III, Urist MR: Intertransverse process lumbar arthro- desis with autogenous bone graft. Clin Orthop 1981;154:90-96. 30. Jenis LG, An HS, Stein R, Young B: Prospective comparison of the effect of direct current electrical stimulation and pulsed electromagnetic fields on instru- mented posterolateral lumbar arthrod- esis. J Spinal Disord 2000;13:290-296. 31. Kahanovitz N, Arnoczky SP, Nemzek J, Shores A: The effect of electromagnetic pulsing on posterior lumbar spinal fu- sions in dogs. Spine 1994;19:705-709. 32. Guizzardi S, Di Silvestre M, Govoni P, Scandroglio R: Pulsed electromagnetic field stimulation on posterior spinal fusions: A histological study in rats. J Spinal Disord 1994;7:36-40. 33. Glazer PA, Heilmann MR, Lotz JC, Bradford DS: Use of electromagnetic fields in a spinal fusion:A rabbit model. Spine 1997;22:2351-2356. 34. Simmons JW: Treatment of failed poste- rior lumbar interbody fusion (PLIF) of the spine with pulsing electromagnetic fields. Clin Orthop 1985;193:127-132. 35. Mooney V:A randomized double-blind prospective study of the efficacy of pulsed electromagnetic fields for inter- body lumbar fusions. Spine 1990;15: 708-712. 36. Scott G, King JB:A prospective, double- blind trial of electrical capacitive cou- pling in the treatment of non-union of long bones. J Bone Joint Surg Am 1994; 76:820-826. 37. Goodwin CB, Brighton CT, Guyer RD, Johnson JR, Light KI, Yuan HA: A double-blind study of capacitively cou- pled electrical stimulation as an ad- junct to lumbar spinal fusions. Spine 1999;24:1349-1357. 38. Blumenthal SL, Gill K: Can lumbar spine radiographs accurately deter- mine fusion in postoperative patients? Correlation of routine radiographs with a second surgical look at lumbar fusions. Spine 1993;18:1186-1189. 39. McMaster MJ, Merrick MV: The scinti- graphic assessment of the scoliotic spine after fusion. J Bone Joint Surg Br 1980;62:65-72. Use of Electrical Bone Stimulation in Spinal Fusion 88 Journal of the American Academy of Orthopaedic Surgeons . electrical potentials are generated when bone is compressed, which leads to bone deposition by the piezoelectric ef- fect 4 (Fig. 1). This theory states that loading of bone creates a negative potential. a negative potential in the compression area and a positive potential in surrounding areas. The small current from these electrical potentials stimulates osteo- genesis at the site of compression. These. piezoelectric effect. Compressive stress causes a negative elec- trical potential at the area of compression and a positive potential in surrounding ar- eas, which leads to a decrease in pO 2 and