The effect of electrical stimulation therapies on spinal fusion: a cross-disciplinary systematic review and meta-analysis of the preclinical and clinical data

Free access

OBJECTIVE

Nonunion is a common complication of spinal fusion surgeries. Electrical stimulation technologies (ESTs)—namely, direct current stimulation (DCS), capacitive coupling stimulation (CCS), and inductive coupling stimulation (ICS)—have been suggested to improve fusion rates. However, the evidence to support their use is based solely on small trials. Here, the authors report the results of meta-analyses of the preclinical and clinical data from the literature to provide estimates of the overall effect of these therapies at large and in subgroups.

METHODS

A systematic review of the English-language literature was performed using PubMed, Embase, and Web of Science databases. The query of these databases was designed to include all preclinical and clinical studies examining ESTs for spinal fusion. The primary endpoint was the fusion rate at the last follow-up. Meta-analyses were performed using a Freeman-Tukey double arcsine transformation followed by random-effects modeling.

RESULTS

A total of 33 articles (17 preclinical, 16 clinical) were identified, of which 11 preclinical studies (257 animals) and 13 clinical studies (2144 patients) were included in the meta-analysis. Among preclinical studies, the mean fusion rates were higher among EST-treated animals (OR 4.79, p < 0.001). Clinical studies similarly showed ESTs to increase fusion rates (OR 2.26, p < 0.001). Of EST modalities, only DCS improved fusion rates in both preclinical (OR 5.64, p < 0.001) and clinical (OR 2.13, p = 0.03) populations; ICS improved fusion in clinical studies only (OR 2.45, p = 0.014). CCS was not effective at increasing fusion, although only one clinical study was identified. A subanalysis of the clinical studies found that ESTs increased fusion rates in the following populations: patients with difficult-to-fuse spines, those who smoke, and those who underwent multilevel fusions.

CONCLUSIONS

The authors found that electrical stimulation devices may produce clinically significant increases in arthrodesis rates among patients undergoing spinal fusion. They also found that the pro-arthrodesis effects seen in preclinical studies are also found in clinical populations, suggesting that findings in animal studies are translatable. Additional research is needed to analyze the cost-effectiveness of these devices.

ABBREVIATIONS CCS = capacitive coupling stimulation; DCS = direct current stimulation; ICS = inductive coupling stimulation; PEMF = pulsed electromagnetic field.

OBJECTIVE

Nonunion is a common complication of spinal fusion surgeries. Electrical stimulation technologies (ESTs)—namely, direct current stimulation (DCS), capacitive coupling stimulation (CCS), and inductive coupling stimulation (ICS)—have been suggested to improve fusion rates. However, the evidence to support their use is based solely on small trials. Here, the authors report the results of meta-analyses of the preclinical and clinical data from the literature to provide estimates of the overall effect of these therapies at large and in subgroups.

METHODS

A systematic review of the English-language literature was performed using PubMed, Embase, and Web of Science databases. The query of these databases was designed to include all preclinical and clinical studies examining ESTs for spinal fusion. The primary endpoint was the fusion rate at the last follow-up. Meta-analyses were performed using a Freeman-Tukey double arcsine transformation followed by random-effects modeling.

RESULTS

A total of 33 articles (17 preclinical, 16 clinical) were identified, of which 11 preclinical studies (257 animals) and 13 clinical studies (2144 patients) were included in the meta-analysis. Among preclinical studies, the mean fusion rates were higher among EST-treated animals (OR 4.79, p < 0.001). Clinical studies similarly showed ESTs to increase fusion rates (OR 2.26, p < 0.001). Of EST modalities, only DCS improved fusion rates in both preclinical (OR 5.64, p < 0.001) and clinical (OR 2.13, p = 0.03) populations; ICS improved fusion in clinical studies only (OR 2.45, p = 0.014). CCS was not effective at increasing fusion, although only one clinical study was identified. A subanalysis of the clinical studies found that ESTs increased fusion rates in the following populations: patients with difficult-to-fuse spines, those who smoke, and those who underwent multilevel fusions.

CONCLUSIONS

The authors found that electrical stimulation devices may produce clinically significant increases in arthrodesis rates among patients undergoing spinal fusion. They also found that the pro-arthrodesis effects seen in preclinical studies are also found in clinical populations, suggesting that findings in animal studies are translatable. Additional research is needed to analyze the cost-effectiveness of these devices.

ABBREVIATIONS CCS = capacitive coupling stimulation; DCS = direct current stimulation; ICS = inductive coupling stimulation; PEMF = pulsed electromagnetic field.

In Brief

The authors conducted a systematic review and meta-analysis of the available preclinical and clinical literature to determine the effects of electrical stimulation therapies on spinal fusion. The findings of this work—that electrical stimulation devices may produce clinically significant increases in arthrodesis rates among patients undergoing spinal fusion—substantially add to the literature on the effect of these therapies on spinal fusion. This is the largest and most comprehensive study to date on the effect of electrical stimulation therapies on spinal fusion.

Each year, approximately 400,000 Americans undergo a spinal fusion operation for the treatment of neck or back pain, radiculopathy, and/or myelopathy.77 These operations account for the highest aggregate hospital cost of any surgical procedure in America, estimated at $13 billion in 2011.89 Consequently, demonstration of clinical efficacy is paramount given increasing scrutiny of cost-effective care. Prior studies have suggested that clinical improvement following spinal fusion surgery is often in accordance with the radiological success of fusion, as defined by continuous bony union across the fusion site.3,59,87 For this reason, emphasis has been placed on reducing the rates of nonunion, or pseudarthrosis, which are reported to be as high as 81% in some small series.9,20,29,32,64,82 Interventions to accomplish this goal include preoperatively addressing risk factors (e.g., diabetes, chronic steroid use, and cigarette use)51 and improving operative technique (e.g., adequate decortication, removal of interposing soft tissues, and sufficient bone graft).12 Additionally, new technologies are continuously being investigated to enhance the fusion rate, including the use of recombinant human growth factors (e.g., bone morphogenetic protein–2),27,43,76 mesenchymal stem cells,79 novel bone graft substitutes,31,73 and dynamic instrumentation.65,91 Postoperative electrical stimulation therapy has also been suggested as an attractive adjuvant therapy to enhance or accelerate bony union.4,49

The use of electrical stimulation therapy to induce fusion has been investigated clinically since at least 1812, when Birch successfully treated a patient with tibial nonunion using “[s]hocks of electric fluid . . . passed [daily] through the space between the ends of the bones both in direction of the length of the limb and that of its thickness.”50 A considerable body of evidence has since been generated to support the general concept that electrical energy influences living bone (as well as other biological tissues).6 Notably, in the 1950s, Fukada and Yasuda described the piezoelectric effect of bone, defined as the generation of electric potentials in bone subjected to mechanical stresses. Using a custom galvanometer, they documented an electrical potential across the stressed bone, with the compressed bone being electronegative and the side under tension being electropositive.44 Subsequently, Friedenberg and Brighton described the bioelectric potentials in bone, in which areas of bone undergoing active repair or growth are electronegative relative to areas at rest.40,41 Therapeutic electrical stimulation devices are based on these biophysical principles—namely, that the external application of an electrical stimulus can stimulate bone growth through the induction of a negative bioelectric potential.

There are currently 3 types of electrical stimulation therapies used in spinal fusion: direct current stimulation (DCS), capacitive coupling stimulation (CCS), and inductive coupling stimulation (ICS), also known as pulsed electromagnetic field (PEMF) therapy (Fig. 1). Conventionally, DCS involves the implantation of cathodes (negative electrodes) into the prospective fusion mass and an anode (positive electrode) into the adjoining soft tissue. A continuous electrical current between 5 and 20 μA is then delivered to the fusion site via a subcutaneously implanted electric generator; the lifetime of this current is dictated by the charge size of the implanted battery, although most devices operate for a minimum of 6 months.16,39,42 CCS, in contrast to DCS, is completely noninvasive and employs two capacitive plates placed on the skin on opposite sides of the fusion site. Alternating current is applied to the plates, setting up an oscillating electric field (1–100 mV/cm). As the battery pack is external, it may be replaced and recharged, allowing for continuous use (24 hours/day) until there is radiological confirmation of fusion. Lastly, ICS employs electromagnetic coils placed over the fusion site. Alternating current applied to these coils induces an electromagnetic field covering the fusion site.83 Compared to CCS, ICS devices require shorter daily usage, with only 30 minutes to 2 hours of continuous use required per day until radiological confirmation of fusion is established. The mechanisms of action and the relative technical advantages and disadvantages of these 3 therapies are summarized in Table 1.1,7,8,10,11,16,17,19,21,23,30,37,66,85,88,92,93

FIG. 1.
FIG. 1.

Conceptual illustrations of the 3 types of electrical stimulation therapies used in spinal fusion. A: Posterolateral L3–5 inter–transverse process spinal fusion using bone graft, without electrical stimulation. B–D: Same procedure illustrating postoperative adjuvant therapy with DCS (B), CCS (C), or ICS (D), sometimes referred to as PEMF. In B, the electric generator is typically implanted subcutaneously. In C and D, the electric generators are externally located. Copyright Ethan Cottrill. Published with permission. Figure is available in color online only.

TABLE 1.

Mechanisms of action and relative technical advantages and disadvantages of the 3 types of electrical stimulation therapies used in spinal fusion surgery

Electrical Stimulation TherapyMechanisms of ActionRelative Technical AdvantagesRelative Technical Disadvantages
DCS1) Electrochemical (faradic) reaction at the cathode lowers the oxygen tension & raises the pH, favoring net bone formation:7,11,162 H2O + 4 e + O2= 4 OH. 2) Generation of H2O2 at the cathode stimulates macrophage secretion of vascular endothelial growth factor, a potent angiogenic agent involved in bone healing.21 3) Upregulation of osteoinductive growth factors (e.g., BMP-2, 6, & 7).371) Continuous, focal delivery of direct electric current for the life of the battery/device. 2) Virtually 100% pt compliance (requires no further action from the pt following implantation).1) Risks associated w/ the surgical implantation of a medical device (e.g., infection, immune reaction, protrusion causing discomfort, device breakage or malfunction, release of toxic substances). 2) May require additional training by the surgeon for correct implantation. 3) May require an additional surgery for removing the electrical generator following successful fusion. 4) May be incompatible w/ MRI (contrast w/ CCS & ICS, which involve external, removable devices).
CCS1) Direct activation of osteoblastic membrane-bound voltage-gated Ca channels leads to an increase in cytosolic Ca2+, inducing downstream phospholipase-A2− & calmodulin-mediated bone formation.8,17,23,30,92 2) Upregulation of osteoinductive growth factors (e.g., BMP-2, 4, 5, 6, & 7; & TGF-β1).88,931) Noninvasive & theoretically painless. 2) Relatively lightweight & discreet; wearable 24 hrs/day.1) Requires active pt participation, w/ recommended usage of 24 hrs/day throughout the duration of therapy. 2) Possible device-related medical complications, including electric shock, burns, & immune reaction to the cutaneous electrodes (& undetermined carcinogenicity & mutagenicity).
ICS1) Direct release of Ca2+ from intracellular stores, inducing downstream calmodulin-mediated bone formation.17,23,92 2) Modulation of osteoblastic PTH signaling at the plasma membrane, reducing inhibitory effects on collagen synthesis.19,66 3) Upregulation of osteoinductive growth factors (e.g., BMP-2 & 4; TGF-β1; & FGF-2).1,10,851) Noninvasive & theoretically painless. 2) Recommended usage is 30 mins to 2 hrs per day (compare to CCS).1) Requires active pt participation throughout the duration of therapy. 2) Heavier & more obtrusive than CCS devices.
BMP = bone morphogenetic protein; Ca = calcium; FGF = fibroblast growth factor; pt = patient; PTH = parathyroid hormone; TGF = transforming growth factor.

Although prior reviews have described the effects of electrical stimulation therapies on spinal fusion, none to date have systematically evaluated both the preclinical and clinical literature of all 3 available technologies. In this article, we perform such a review as a means of compiling the current evidence and validating the translatability of results achieved using these technologies in animal models. We set out to evaluate the available English-language literature for all 3 technologies, asking of each one: 1) To what degree does the technology improve bony fusion in animal models? 2) To what degree does the technology facilitate bony fusion in humans? Additionally, we report the results of a meta-analysis of the available clinical studies to provide an estimate of the overall effect at large and in subgroups.

Methods

Electronic Literature Search

A systematic review of the literature was performed using PubMed, Embase, and Web of Science databases. The search query was designed to obtain all of the available in vivo data (preclinical and clinical) examining the effect of electrical stimulation therapies on spinal fusion. The query for the PubMed database was as follows: (spinal fusion[mesh] OR spine fusion*[tw] OR spinal fusion*[tw] or spinal arthrodes*[tw] OR cervical fusion*[tw] OR lumbar fusion*[tw] OR lumbosacral fusion*[tw] OR interbody fusion*[tw] OR posterolateral fusion*[tw] OR cervical arthrodes*[tw] OR lumbar arthrodes*[tw] OR lumbosacral arthrodes*[tw] OR interbody arthrodes*[tw] OR posterolateral arthrodes*[tw]) AND (electric stimulation[mesh] OR electric stimulation therapy[mesh] OR electromagnetic fields[mesh] OR “electrical stimulation”[tw] OR “pulsed electromagnetic field*”[tw] OR “electromagnetic pulsing*”[tw] OR “magnetic fields*”[tw] OR “direct current stimulation*”[tw] OR “bone growth stimulation*”[tw] OR “electrical current*”[tw] OR “capacitively coupl*”[tw] OR “capacitive coupl*”[tw] OR “capacitive stimulat*”[tw] OR “inductively coupl*”[tw] OR “inductive coupl*”[tw] OR “inductive stimulat*”[tw]). This query was stylistically modified for use in the Embase and Web of Science databases. The bibliographies of the included studies were also queried for additional sources.

Included studies were preclinical or clinical peer-reviewed publications with full English-language text availability that evaluated the effects of one or more electrical stimulation therapies on spinal fusion. We defined electrical stimulation as the therapeutic use of electromagnetic energy (including direct current, capacitive coupling, and inductive coupling) with the expressed intent of promoting bony fusion after instrumented or noninstrumented spinal fusion. Studies were excluded if they examined a surgical model other than spinal fusion or if they mixed the results of spinal fusion with other surgical models. Eligible studies were screened against these criteria by two reviewers (E.C. and Z.P.); a third reviewer (A.K.A.) served as a referee, resolving any discrepancies between the first two reviewers. Critical Appraisal Checklists obtained from the Joanna Briggs Institute at The University of Adelaide were used to assess the quality of the clinical studies included in the meta-analysis.71 Because preclinical studies are all classified as level of evidence V, a similar appraisal was not conducted for them. Additionally, the QUOROM (Quality of Reporting of Meta-analyses) checklist was used for this systematic review and meta-analysis.70

Data Extraction

Studies meeting the inclusion criteria were reviewed to extract details regarding the type of electrical stimulation, specifications of the electrical therapy, means of determining bony fusion, and the overall fusion rate at last follow-up. For preclinical studies, we also recorded details about the animal species and surgical model employed. For clinical studies, we included details on the patient demographics and the surgical approach.

For both preclinical and clinical studies, the primary endpoint was the fusion rate at last follow-up. In preclinical studies, we defined this as the total number of levels fused divided by the total number of levels included in the prospective fusion mass. In clinical studies, we defined the fusion rate as that derived from the proportion of patients experiencing a successful radiological fusion at the last follow-up visit. The definition and method of assessment of fusion were recorded for each study.

Statistical Analysis

Statistical meta-analyses were performed using R version 3.4.2 (The R Foundation for Statistical Computing). Separately for the preclinical and clinical studies, we generated mean fusion rates and odds ratios using the Freeman-Tukey double arcsine transformation, a previously established method for normalizing proportions with variance stabilization.38 A random-effects meta-analysis was then employed to give a pooled estimate of the effect of electrical stimulation on fusion rates. We elected to forego a numbers-needed-to-treat analysis based on these results, as prior reports have demonstrated such estimates to be commonly misleading.81 Using this methodology, we also performed subgroup analyses of the clinical data based on smoking status, surgical history (index vs revision procedure), use of interbody devices, region fused, type of bone graft, use of instrumentation, and number of levels fused. For all analyses, an α of 0.05 was used as the definition of statistical significance.

Results

Our search identified 340 unique articles, and 47 of these met our inclusion criteria (Fig. 2). After reviewing the full texts, we included 17 preclinical studies15,22,25,28,35–37,46,48,52,54–56,67,75,86,94 and 16 clinical studies.5,14,26,34,47,53,57,60,63,68,69,72,74,78,84,90 Among the 14 excluded articles, the reasons for exclusion were lack of full-text availability (n = 10), surgical model other than spinal fusion (n = 2), and lack of an English-language translation (n = 2). Of these 33 articles, 11 preclinical (257 animals; 273 levels) and 13 clinical (2144 patients) studies were ultimately included in the meta-analysis. The clinical studies were deemed to have sufficient quality to be included in the meta-analysis (Critical Appraisal Checklists). The included articles are summarized in Tables 2 and 3, as well as in Supplemental Tables 1 and 2. Supplemental Fig. 1 plots these studies by year of publication, illustrating the dearth of recent studies.

FIG. 2.
FIG. 2.

Diagram of the consolidated standards of reporting trials for article selection.

TABLE 2.

Descriptive summaries of the identified preclinical studies (n = 17)

Authors & Year (LOE)*Animal & Surgical ModelStudy GroupsPostop Definition of Fusion OutcomeFusion Rate
DCS
Nerubay et al., 1986 (V)Pig (n = 20): L5–6 posterior facet joint fusion w/ iliac crest autograftA. DCS via Osteostim model S11; constant 20-μA current (n = 9)NAFusion rate not reported; however, a “significant increase of osteoblastic activity with bone formation” favoring the experimental group was observed at 2 mos postop (p = 0.037)
B. Control; implantation of the Osteostim device, w/o an active power source (n = 11)
Kahanovitz & Arnoczky, 1990 (V)Dog (n = 4): L1–2 & L4–5 posterior facet joint fusion w/ local autograftA. DCS; 10 μA (n = 2; each animal had either L1–2 or L4–5 electrically stimulated)Histological & radiographic evidence of complete bony fusion across the graft & both facets (defined as 1 level) at 12 wks postopA. 100% (4/4 levels)
B. Control; implantation of electrodes, w/o an active power source (n = 2; each animal had L1–2 or L4–5 electrically stimulated)B. 0% (0/4 levels)
Bozic et al., 1999 (V)Rabbit (n = 53): L3–4 posterolateral inter–transverse process fusion w/ bone graftA. Coralline HA w/ autologous bone marrow aspirate & an implanted DCS device (SpF-100, Electro-Biology, Inc.) (100 μA) (n = 15)Blind manual palpation of fusion segment (graded as fused or not) at 8 wksA. 87% (13/15 levels) (significantly higher in group A than C)
B. Coralline HA w/ autologous bone marrow aspirate & an implanted DCS device (SpF-XLIIb, Electro-Biology, Inc.) (40 μA) (n = 12)B. 50% (6/12 levels)
C. Coralline HA w/ autologous bone marrow aspirate (n = 12)C. 25% (3/12 levels)
Toth et al., 2000 (V)Sheep (n = 22): L4–5 discectomy & interbody fusion w/ Ti cage (Bagby & Kuslich) & iliac crest autograftTi interbody cage (Bagby & Kuslich) filled w/ iliac crest autograft & an implanted DCS device for 4 mosA. 100 μA (SpF-100, Electro-Biology, Inc.) (n = 8)Histological evidence of fusion defined as a continuous bony bridge from superior to inferior vertebrae at 4 mosA. 100% (8/8 levels)
B. 40 μA (SpF-XLII, Electro-Biology, Inc.) (n = 7)B. 71% (5/7 levels)
C. Control; 0 μA (n = 7)C. 29% (2/7 levels); each group statistically different (Fisher’s exact test, p < 0.009)
Dejardin et al., 2001 (V)Dog (n = 5): L1–2 & L4–5 facet joint fusion w/ local autograftPostop constant DCS for 12 wksA. 10 μA/cm (n = 1.5 animals; 3 levels)Histological & radiographic evidence of complete bony fusion across the graft & both facets (defined as 1 level) at 12 wksA. 100% (3/3 levels)
B. 4 μA/cm (n = 2 animals; 4 levels)B. 100% (4/4 levels)
C. 0.83 μA/cm (n = 1.5 animals; 3 levels)C. 100% (3/3 levels)
France et al., 2001 (V)Rabbit (n = 34): L5–6 posterolateral inter–transverse process fusion w/ iliac crest autograftA. Postop local DCS (60 μA) for 5 wks (n = 12)NAFusion rate not reported; however, via manual palpation scores & biomechanical testing, there was no statistically significant difference btwn the 3 groups
B. Postop local DCS (20 μA) for 5 wks (n = 8)
C. Control; implantation of stimulator but no current (0 μA) (n = 14)
Cook et al., 2004 (V)Nonhuman primate (n = 22): L5–6 anterior lumbar interbody fusion w/ iliac crest autograft & either a femoral allograft ring or Ti alloy fusion cageA. Ti alloy cage w/ autograft & postop local DCS (SpF, Electro-Biology, Inc.) (19.6 μA/cm2) for 26 wks (n = 4)CT evidence of fusion defined as “bridging callus with trabeculations” at 26 wksA. 100% (4/4 levels)
B. Ti alloy cage w/ autograft & postop local DCS (SpF, Electro-Biology, Inc.) (5.4 μA/cm2) for 26 wks (n = 7)B. 86% (6/7 levels)
C. Ti alloy cage w/ autograft alone (7)C. 71% (5/7 levels)
D. Femoral ring allograft w/ autograft (n = 4)D. 75% (3/4 levels) (no statistical significance btwn groups)
France et al., 2006 (V)Rabbit (n = 25) (nicotine & control models): L5–6 posterolateral inter–transverse process fusion w/ iliac crest autograftA. Postop local DCS (SpF-100, Electro-Biology, Inc.) (100 μA) & continuous dose of nicotine via transdermal patch (10.5 mg) for 5 wks (n = 9)Evidence of fusion by manual palpation, graded 3 separate times at 5 wks postop; fusion rate was calculated as the ratio of fused segments to total segments tested × 100A. 85.3% ± 12.7% (approximately 7/9 levels) (significantly greater than that of group C)
B. Postop continuous dose of nicotine via transdermal patch (10.5 mg) for 5 wks (n = 8)B. 66.7% ± 7.2% (approximately 5/8 levels) (significantly greater than that of group C)
C. Control; neither nicotine nor stimulator (n = 8)C. 37.5% ± 12.5% (approximately 3/8 levels)
Fredericks et al., 2007 (V)Rabbit (n = 5): L4–5 posterolateral inter–transverse process fusion w/ iliac crest autograftA. Postop local DCS (SpF-100, Electro-Biology, Inc.) (100 μA) for 28 days (n = 2)Radiographic evidence of bilat bridging fusion at 4 wksA. 100% (2/2 levels)
B. Control; no electrical stimulation (n = 3)B. 33% (1/3 levels)
MacEwan et al., 2016 (V)Goat (n = 2): L4–5 interbody fusion w/ local autograft & instrumentation permitting DCS (anodized Ti rods & pedicle screws)A. Constant DCS (Varta Microbattery, Inc.) (40 μA) routed directly through the Ti rods & pedicle screws (novel “osteogenic spinal instrumentation”) (n = 1)Micro-CT evidence of fusion, defined as solid bridging of the vertebral bodies at 3 mosA. 100% (1/1 level)
B. Control; same system w/o DCS (n = 1)B. 0% (0/1 level)
Cho et al., 2019 (V)Rat (n = 60); L4–5 posterolateral inter–transverse process fusion w/ tubular nickel-Ti (nitinol) mesh containers filled w/ iliac crest autograftA. Nitinol container filled w/ autograft & constant DCS (Cybermedic, Inc.) (100 μA) for 8 wks (n = 20)Evidence of fusion by manual palpation & micro-CT, defined as no intersegmental motion & continuous bridging bone at 8 wksA. 100% (20/20 levels)
B. Nitinol container filled w/ autograft & pulsed DCS (Cybermedic, Inc.) (100 μA, 100 Hz, 200 μsec) for 8 weeks (n = 20)B. 100% (20/20 levels)
C. Control: nitinol container filled w/ autograft alone (n = 20)C. 70% (14/20 levels) (significantly less than A & B; p < 0.01)
ICS
Kahanovitz et al., 1984 (V)Dog (n = 2): L2–4 posterior lumbar fusion w/ local autograft & instrumentationA. Postop local ICS (positive pulse: 200 μsec @ 1.1 mV/cm; negative pulse: 200 μsec @ 9.6 mV/cm; repeated for 5 msec @ 12 Hz); 12 hrs of stimulation/day (n = 1)NAFusion rate not reported; however, radiographically & histologically “no demonstrable objective differences” btwn the groups observed at 15 wks postop
B. Control: no postop ICS (n = 1)
Guizzardi et al., 1994 (V)Rat (n = 6): L4–6 posterolateral inter–transverse process fusion w/ local autograftA. Postop ICS (via coils fixed to the outside of the animals’ cages); mean 2.5 mV for 12 hrs/day (n = 3)NAFusion rate not reported; however, histological analysis showed an acceleration of bony callus formation in the experimental group
B. Control; no postop ICS (n = 3)
Kahanovitz et al., 1994 (V)Dog (n = 12); L1–2 & L4–5 posterior facet joint fusion w/ local autograft & wire to secure the graft in placePostop local ICS (pulse burst frequency: 1.5 Hz; pulse burst duration: 30 msec)A. 30 mins/day for 12 wks (n = 4)Histological & radiographic evidence of complete bony fusion across graft & both facets (defined as 1 level) at 12 wksA. 0% (0/8 levels)
B. 60 mins/day for 12 wks (n = 4)B. 0% (0/8 levels)
C. Control; no postop ICS (n = 4)C. 0% (0/8 levels)
Ito et al., 1997 (V)Dog (n = 26): L5–6 posterolateral inter–transverse process fusion w/ iliac crest autograftPostop local ICS (American Medical Electronics, Inc.) (positive pulse: 195 μsec; negative pulse: 65 μsec; peak change in positive flux density: 9 T/sec; peak change in negative flux density: 3 T/sec; pulses/burst: 99; burst interval: 670 μsec)A. w/ Steffee (variable screw placement plate) (n = 6)NAFusion rate not reported; however, the authors noted no statistically significant effect of pulsed electromagnetic field on the BMD or biomechanical properties of the fusion mass at 24 wks postop
B. w/o Steffee plate (n = 7)
Control; no postop local ICSC. w/ Steffee plate (n = 6)
D. w/o Steffee plate (n = 7)
Glazer et al., 1997 (V)Rabbit (n = 20): L5–6 posterolateral inter–transverse process fusion w/ iliac crest autograftA. Postop external ICS (Orthofix, Inc.) (via coils fixed to the outside of the animals’ cages); peak change in positive flux density: 9 T/sec; peak change in negative flux density: 3 T/sec; pulse burst: 26 msec; burst interval: 670 msec; 4 hrs/day for 6 wks (n = 10)Radiographic evidence of bilat fusion, defined as the presence of a continuous trabecular pattern across the fusion mass at 6 wksA. 80% (8/10 levels)
B. Control; no postop external ICS (n = 10)B. 60% (6/10 levels) (p > 0.05)
Zhuo et al., 2018 (V)Rabbit (n = 32): L5–6 posterolateral inter–transverse process fusion w/ bone graftA. Postop ICS (CMF SpinaLogic device, DJO Global; affixed to the cage) for 30 mins/day for 8 wks & nano-HA–coated biphasic Ca phosphate bone graft substitute (n = 8)Radiographic evidence of bilat fusion at 9 wksA. 100% (8/8 levels)
B. Postop ICS (CMF SpinaLogic device affixed to the cage) for 30 mins/day for 8 wks & biphasic Ca phosphate bone graft substitute (n = 8)B. 62.5% (5/8 levels)
C. Nano-HA–coated biphasic Ca phosphate bone graft substitute alone (n = 8)C. 75% (6/8 levels)
D. Biphasic Ca phosphate bone graft substitute alone (n = 8)D. 38% (3/8 levels) (significant improvement w/ nano-HA–coated biphasic Ca phosphate, p < 0.05)
BMD = bone mineral density; cont’d = continued; HA = hydroxyapatite; LOE = level of evidence; NA = not applicable; Ti = titanium.

Each preclinical study was classified as level V evidence according to guidelines of the North American Spine Society.

TABLE 3.

Descriptive summaries of the identified clinical studies (n = 16)

Authors & Year, Study Design (LOE)*Inclusion CriteriaSurgical ModelStudy Groups (no. of pts)Definition of Fusion OutcomeFusion Rate (pts) (p value)
DCS
Nerubay & Katznelson, 1984, case series (IV)Pediatric pts w/ spondylolisthesisPosterior lumbar fusion w/ iliac crest autograft w/ or w/o instrumentationOsteostim model S11; direct current of 20 μA over 4 cathodes (5 μA/cathode); 105-mA/hr battery capacity for at least 6 mos (n = 5)Radiographic evidence of solid bony fusion at 1 yr100% (5/5)
Kane, 1988, nonrandomized, multicenter (III)NAPosterior lumbar fusion w/ autograftDCS via Osteostim HS1 (BGS Medical Corp.); 20 μA over 4 cathodes; active for 22 wks (n = 82)NA91.5% (75/82) (p = 0.02)
Historical control: no DCS (n = 159)80.5% (128/159)
Kane, 1988, prospect RCT nonblinded (II)Pts w/ difficult-to-fuse spines: failed prior fusion, grade II or worse spondylolisthesis, multilevel fusion, &/or other high-risk medical condition (e.g., gross obesity)Posterior lumbar fusion (1–4 levels fused) w/ autograftDCS via Osteostim HS1; 20 μA over 4 cathodes; active for 22 wks (n = 31)Radiographic evidence of fusion at 18 mos81% (25/31) (p = 0.026)
Matched control: no DCS (n = 28)54% (15/28)
Kane, 1988, nonrandomized, multicenter (IV)NAPosterior lumbar fusion w/ autograftDCS via Osteostim HS1; 20 μA over 4 cathodes; active for 22 wks (n = 116)Radiographic evidence of fusion (no time point given)93% (108/116)
Meril, 1994, retro comparative cohort (III)Adult pts undergoing primary or revision anterior or posterior lumbar interbody fusionSingle-level or multilevel anterior or posterior lumbar interbody fusion w/ or w/o instrumentation & w/ autograft &/or allograftDCS device (EBI Medical Systems); 20 μA over 2 cathodes; active for 24 wks (n = 122)CT evidence of fusion, defined as unequivocal incorporation of the graft into the adjacent vertebral endplates on at least half of the curved coronal views, at ≥21 mos93% (113/122) (p < 0.001)
Control: no DCS (n = 103)75% (77/103)
Rogozinski & Rogozinski, 1996, prospect comparative study (III)Adult pts undergoing instrumented lumbosacral PLFLumbosacral posterolateral instrumented fusion w/ autograft (not interbody)DCS via SpF-2T (EBI Medical Systems); 20 μA over 2 cathodes w/ 220-mA/hr battery (n = 53)Radiographic evidence of fusion, defined as mature trabeculated bone across instrumented levels w/ no movement on stress views & no loss of fixation, at ≥19 mos96% (51/53) (p >0.05)
Control: no DCS (n = 41)85% (35/41)
Tejano et al., 1996, retro case series (IV)Pts w/ difficult-to-fuse spines: multilevel fusion, previous failed fusion, &/or grade II or worse spondylolisthesisPosterior or posterolateral lumbar fusion w/ iliac crest autograft w/o instrumentationDCS device (EBI Medical Systems); 20 μA over 4 cathodes for a minimum of 24 wks (n = 118)Radiographic evidence of fusion, w/ no motion btwn vertebrae at ≥2 years92% (108/118)
Kucharzyk, 1999, prospect comparative series (II)Pts w/ difficult-to-fuse spines: multiple prior spine surgeries, failed prior fusion, segmental instability, spinal stenosis, &/or spondylolisthesisPosterolateral lumbar fusion w/ Rogozinski hardware (early pedicle screw system) & autograftDCS device (model not given) (n = 65)Radiographic evidence of fusion, defined as the presence of bridging trabeculae, consolidation of bone graft, no pseudarthrosis lines, absence of spine-specific pain, & no failure of instrumentation, at a mean 3.8 yrs of follow-up95% (62/65) (p > 0.05)
Control: no DCS (n = 65)86% (56/65)
Jenis et al., 2000, RCT (II)Adult pts undergoing primary or revision lumbar or lumbosacral PLFPrimary or revision lumbar or lumbosacral PLF; single-level or multilevel w/ instrumentation & iliac crest autograftDCS via SpF-2T; 20 μA over 2 cathodes (n = 17)Radiographic evidence of fusion, defined as solid arthrodesis w/ trabecular bridging bone at 1 yr59% (10/17) (p > 0.05)
Control: no DCS (n = 22)82% (18/22)
Welch et al., 2004, retro case-control study (III)Pts w/ difficult-to-fuse spines: age >65, rheumatoid arthritis, prior failed fusion, infection, &/or immunosuppressedPosterior cervical fusion from the occiput to C3 w/ instrumentation & autograft &/or allograftDCS via SpF-2T (EBI Medical Systems); 20 μA over 2 cathodes w/ 220-mA/hr battery (n = 16)Radiographic evidence of fusion, defined as the incorporation of osteosynthetic bone into the lateral masses, facet joints, & lamina; the absence of movement btwn fused segments on dynamic studies; & the preservation of implant integrity, at a mean of 19 mos94% (15/16)
Andersen et al., 2009, RCT (II)Pts w/ difficult-to-fuse spines: age >60 eligible for spinal fusionPosterolateral lumbar fusion (1–4 levels) w/ local autograft &/or allograft, w/o instrumentationDCS via SpF-XL IIb stimulator (BiometSpine)100 μA (n = 8)CT evidence of fusion, defined as “a continuous bony bridge either between the transverse processes or at the lateral side of the facet joints on at least 1 side or a bilateral fusion of the facet joints,” at 24 mos50% (4/8)
40 μA (n = 40)32.5% (13/40)
Control: no DCS (“dummy electrodes” implanted) (n = 36)33% (12/36) (all groups, p > 0.05)
CCS
Goodwin et al., 1999, double-blind RCT (I)Adult pts undergoing lumbar spinal fusion1- or 2-level lumbar fusion (anterior interbody, posterior interbody, &/or posterolateral) w/ autograft &/or allograft & any type of internal fixation except interbody fusion cages; w/ or w/o instrumentationCCS via Bioelectron device (early generation of current Biomet Orthopak [Zimmer Biomet]); instructed to use 24 hrs/day until healing occurred, or for 9 mos (average actual = 15.7 hrs/day) (n = 85)Radiographic evidence of fusion, w/ ≥50% integration of interbody & vertebrae (interbody fusion) or uninterrupted bone masses on AP & lat radiographs (PLF), at 12 mos90.6% (77/85) (p > 0.05)
Control: inactive stimulator (n = 94)81.9% (77/94)
ICS
Mooney, 1990, double-blinded RCT (II) (<80% of cohort included due to noncompliance)Adults pts undergoing index interbody fusion via anterior or posterior approach1- or 2-level anterior or posterior lumbar interbody fusion w/ autograft &/or allograft, w/ or w/o instrumentationICS via custom brace (specifications not provided); 8 hrs/day (n = 64)Radiographic evidence of fusion, defined as >50% assimilated at ≥12 mos92% (59/64) (p < 0.005)
Control: nonfunctioning brace (n = 53)68% (36/53)
Marks, 2000, retro comparative (III)Adult pts w/ discogenic low-back painAnterior lumbar interbody &/or PLF w/ autograft &/or allograft, w/ or w/o instrumentationICS via Spinal-Stim (Orthofix, Inc.) for ≥4 hrs/day while awake beginning 2 days postop (n = 42)Radiographic evidence of fusion, defined as incorporation of the graft w/ no radiolucency btwn the graft & vertebral bone & no motion at each level, at a mean of 15.6 mos97.6% (41/42) (p < 0.001)
Control: no ICS (only thoracolumbosacral brace) (n = 19)52.6% (10/19)
Jenis et al., 2000, RCT (II)Adult pts undergoing primary or revision lumbar or lumbosacral PLFPrimary or revision lumbar or lumbosacral PLF; 1-level or multilevel w/ instrumentation & iliac crest autograftICS via Spinal-Stim model 8212 (Orthofix, Inc./American Medical Electronics, Inc.) w/in 30 days of op; at least 2 hrs/day on at least 90% of the 150 consecutive treatment days after op (actual average = 182.7 ± 59 days) (n = 22)Radiographic evidence of fusion, defined as solid arthrodesis (AP views) w/ trabecular bridging bone at 1 yr64% (14/22) (p > 0.05)
Control: no ICS (n = 22)82% (18/22)
Bose, 2001, retro cohort (IV)Pts w/ difficult-to-fuse spines (poorly defined): herniated nucleus pulposus, degenerative disc disease, spondylolisthesis, spinal stenosis, &/or prior failed fusionPLF w/ instrumentation & autograft &/or allograftICS via Spinal-Stim (Orthofix, Inc.) w/in 4 wks of op; recommended usage 4 hrs/day until fusion; Boston Overlap Brace for support (n = 48)Radiographic evidence of fusion, defined as 2-point bridging, no radiolucency, & intact hardware, at ≥6 mos (mean 16 mos)98% (47/48)
Linovitz et al., 2002, RCT (I)Adult pts undergoing index noninstrumented PLF of 1 or 2 levels btwn L3 & S1Index posterolateral lumbar 1- or 2-level fusion (btwn L3 & S1) w/o instrumentation, either w/ autograft alone or in combination w/ allograftICS via SpinaLogic, w/in 30 days of op; 30-min treatment per day for 9 mosRadiographic evidence of fusion, defined as continuity of fusion mass, at 12 mos63% (66/104) (p > 0.05)
Control: w/o active PEMF therapy49% (48/97)
Foley et al., 2008, RCT nonblinded for FDA IDE (II)Pts w/ difficult-to-fuse spines: smokers &/or undergoing multilevel fusion w/ allograftACDF w/ allograft & plateICS via CervicalStim (Orthofix, Inc.) w/in 7 days postop; recommended use 4 hrs/day for 3 mosRadiographic evidence of fusion, defined as ≥50% bony bridging through surfaces of the graft-vertebra interface, no radiolucency at any portion of the graft-vertebra junction, & ≤4° of motion btwn adjacent fused vertebrae at 12 mos92.8% (116/125) (p > 0.05)
Control: no ICS86.7% (104/120)
Coric et al., 2018; retro multicenter cohort compared to historical data from trial for initial FDA approval (III)Pts w/ difficult-to-fuse spines: age ≥65, active smoker, multilevel fusion, prior failed fusion, diabetic, &/or osteoporoticSingle- or multilevel ACDF w/ no restrictions on the interbody implant, graft material, or surgical procedureICS via CervicalStim (Orthofix, Inc.); 4 hrs/day for 3–6 mosRadiographic evidence of fusion, defined as the presence of continuous bridging bone on plain films, at 12 mos92.6% (201/217) (p < 0.05)
Historical controls: no ICS82.6% (76/92)
ACDF = anterior discectomy and fusion; AP = anteroposterior; FDA = Food and Drug Administration; IDE = investigational device exemption; PLF = posterolateral fusion; prospect = prospective; RCT = randomized controlled trial; retro = retrospective.

Levels of evidence classified according to guidelines of the North American Spine Society.

Overall Effect of Electrical Stimulation Technologies on Spinal Fusion

In the preclinical literature, the mean fusion rates were higher among animals treated with electrical stimulation therapy (77.7%) than among controls (42.0%). Across all studies, the use of electrical stimulation produced a nearly fivefold increase in the odds of a successful fusion (OR 4.79 [95% CI 2.51–9.16], p < 0.001) (Table 4). In the clinical literature, electrical stimulation similarly was shown to produce higher rates of fusion versus controls in which no electrical stimulation therapy was administered (84.9% vs 73.4%, respectively), although the overall effect was smaller than in the preclinical literature (OR 2.26 [95% CI 1.48–3.44], p < 0.001) (Table 4). Figure 3A illustrates the random-effects meta-analysis of the fusion rates from all clinical studies.

TABLE 4.

Mean fusion rates and odds ratios for the preclinical and clinical studies determined by random-effects meta-analysis*

Fusion Rate (no. fused/total)
Type of ESTType of Study & Authors & YearStimulation GroupControl GroupCochran’s QOR (95% CI) & p Value
Preclinical
 Kahanovitz & Arnoczky, 19904/40/45.455.64 (2.64–12.06); p < 0.001
 Bozic, 199919/2711/26
 Toth et al., 200013/152/7
 Cook et al., 200410/118/11
 France et al., 20067/98/16
 Fredericks et al., 20072/21/3
 MacEwan et al., 20161/10/1
DCS Cho et al., 201940/4014/20
 Overall (95% CI)87.6% (74.2–96.5%)45.3% (30.0–61.1%)
Clinical
 Kane, 1998208/229143/18713.602.13 (1.08–4.21); p = 0.03
 Meril, 1994113/12277/103
 Rogozinski & Rogozinski, 199651/5335/41
 Kucharzyk, 199962/6556/65
 Jenis et al., 200010/1718/22
 Andersen et al., 200917/4812/36
 Overall (95% CI)82.2% (65.8–94.1%)73.9% (61.7–84.4%)
Clinical02.12 (0.87–5.21); p > 0.05
CCS Goodwin et al., 199977/8577/94
 Overall (95% CI)90.6% (88.3–95.8%)81.9% (72.6–89.1%)
Preclinical
 Kahanovitz et al., 19940/160/80.033.08 (0.88–10.72); p > 0.05
 Glazer et al., 19978/106/10
 Zhuo et al., 201813/169/16
 Overall (95% CI)47.8% (1.1–97.8%)35.7% (4.9–75.9%)
Clinical
ICS Mooney, 199059/6436/5316.172.45 (1.20–4.99); p = 0.014
 Marks, 200041/4210/19
 Jenis et al., 200014/2218/22
 Linovitz, 200266/10448/97
 Foley et al., 2008116/125104/120
 Coric et al., 2018201/21776/92
 Overall (95% CI)86.0% (74.2–94.6%)71.2% (56.2–84.1%)
Preclinical
All Overall (95% CI)77.7% (54.2–94.3%)42.0% (27.5–57.2%)6.154.79 (2.51–9.16); p < 0.001
Clinical
 Overall (95% CI)84.9% (76.8–91.4%)73.4% (65.4–80.8%)29.922.26 (1.48–3.44); p < 0.001
Boldface type indicates statistical significance.

Only studies reporting the fusion rates for both the intervention (i.e., electrical stimulation) and control groups were included in the meta-analysis.

For preclinical studies, the fusion rate was defined as the number of bilateral vertebral levels fused divided by the total number of levels attempted. For clinical studies, the fusion rate was defined as the number of patients experiencing successful fusion divided by the total number of patients undergoing surgery. For the analysis of preclinical data, where the fusion rate was 0 in some cases, delta was set to 0.5 (Haldane-Anscombe correction). For clinical studies, where the fusion rate was always greater than 0, delta was set to zero.

FIG. 3.
FIG. 3.

Forest plots demonstrating random-effects meta-analysis of the fusion rates from all clinical studies (A), only clinical studies examining the effect of DCS on spinal fusion (B), and only clinical studies examining the effect of ICS on spinal fusion (C). Only studies reporting fusion rates for intervention (i.e., electrical stimulation) and control groups are included. Figure is available in color online only.

Effect of DCS on Spinal Fusion

Eleven preclinical and 9 clinical studies investigating the effect of DCS on spinal fusion were identified, and 8 preclinical and 6 clinical studies were included in the meta-analysis.

Preclinical Data

The preclinical studies (Table 2) involved rat (n = 1), rabbit (n = 4), dog (n = 2), pig (n = 1), sheep (n = 1), goat (n = 1), and monkey (n = 1) spinal fusion models. All surgical models involved one-level fusions of the lumbar spine, with 3 using posterior facet joint fusion, 5 using posterolateral inter–transverse process fusion, and 3 using interbody fusion. Among these studies, 11 used autograft, 1 used allograft, and 1 used synthetic bone graft; 3 of the studies employed instrumentation in the fusion construct. All but one study used implantable electrodes in the fusion beds. The remaining study routed electrical current through pedicle screws and rods.67

The reported fusion rates ranged between 70% and 100% for the treatment group and between 0% and 73% for controls (Supplemental Fig. 2A). On meta-analysis, the mean fusion rate was found to be significantly higher in DCS-treated levels than in controls (OR 5.64 [95% CI 2.64–12.06], p < 0.001) (Table 4).

Clinical Data

Nine clinical studies examined the effects of DCS on spinal fusion: 8 studies in adult cohorts and 1 study in a pediatric cohort (Table 3). Four studies examined its use in patients with difficult-to-fuse spines using the following definitions: 1) age > 60 years;5 2) multiple prior spine surgeries, failed prior fusion, segmental instability, spinal stenosis, and/or spondylolisthesis;60 3) multilevel fusion, failed prior fusion, and/or grade II or worse spondylolisthesis;84 and 4) age > 65 years, presence of rheumatoid arthritis, failed prior fusion, infection, and/or immunosuppression.90 One study was restricted to index procedures, while 8 included both index or revision procedures. Only 1 study employed interbody fusion; the remaining 8 used solely posterior/posterolateral fusion. The spinal segments investigated were cervical in 1 study and lumbar/lumbosacral in 8. Six studies used autograft only, and the remaining 3 used autograft and/or allograft. Instrumentation was placed in all patients in 4 studies and in some patients in 2 studies; 3 studies used in situ fusion only (Table 3).

The fusion rate ranged from 35% to 96% for treated patients and from 33% to 86% in controls (Supplemental Fig. 3A). In the meta-analysis, patients treated with DCS were found to have a significantly higher fusion rate at the last follow-up than the control patients (OR 2.13 [95% CI 1.08–4.21], p = 0.03) (Table 4 and Fig. 3B).

Effect of CCS on Spinal Fusion

No preclinical studies were found that described the use of CCS in a spinal fusion model, and only 1 clinical study met our inclusion criteria (Table 3). In that double-blind randomized controlled trial, Goodwin et al. examined the use of CCS in 179 adults undergoing one- or two-level fusions in which one of the following techniques was used: anterior lumbar interbody fusion, posterior lumbar interbody fusion, or posterolateral lumbar fusion.47 All patients were instructed to use the stimulation device (precursor of the Biomet OrthoPak [Zimmer Biomet]) for 24 hours/day for 9 months or until fusion was confirmed radiologically. At the 12-month follow-up visit, no significant difference in fusion rates was detected between CCS-treated (90.6%) and control (81.9%) patients (Table 4).

Effect of ICS on Spinal Fusion

Thirteen total studies—6 preclinical and 7 clinical—describing the results of ICS met our inclusion criteria. Of these, 3 preclinical and 6 clinical studies were included in the meta-analysis.

Preclinical Data

Preclinical studies described the effects of ICS in dog (n = 3), rabbit (n = 2), and rat (n = 1) models using posterior facet fusion (n = 2) or posterolateral inter–transverse process fusion (n = 4) of the lumbar spine. Four studies involved one-level procedures, whereas 2 involved multilevel fusions (≥ 2 levels). All studies used either autograft (n = 5) or synthetic bone graft (n = 1); 2 studies used instrumentation (Table 2).

The fusion rate varied widely across studies, ranging from 0% to 81% in treated groups and from 0% to 60% in controls (Supplemental Fig. 2B). In the aggregate, the included studies failed to show a significant difference in fusion rates between ICS-treated animals and controls (OR 3.08 [95% CI 0.88–10.72], p > 0.05) (Table 4).

Clinical Data

All clinical studies investigating the effect of ICS on spinal fusion examined adult patients (Table 3). Three of the studies examined the effects in only patients with difficult-to-fuse spines, defined by the studies as 1) patients with a herniated nucleus pulposus, degenerative disc disease, spondylolisthesis, spinal stenosis, and/or those who had undergone a prior failed fusion;14 2) those who smoked and/or were undergoing multilevel fusion with an allograft;34 or 3) those who were age ≥ 65 years, actively smoked, were undergoing multilevel fusion, had undergone a prior failed fusion, had diabetes, and/or had osteoporosis.26 Two studies restricted patients to those without a history of spine surgery, while the remaining 5 included both index and revision procedures. Three studies involved only posterior/posterolateral fusion procedures, 3 involved only interbody procedures, and 1 involved either type of procedure. The spinal segments investigated were cervical (n = 2) and lumbar/lumbosacral (n = 5). Autograft alone was used in 1 study, allograft alone in 1 study, and autograft and/or allograft in 5 studies. Instrumentation was used in all patients in 4 studies, some patients in 2 studies, and none of the patients in 1 study.

Fusion rates varied between 63% and 98% in the ICS group and between 49% and 87% in the control group (Supplemental Fig. 3C). Patients receiving ICS were found to have significant improvements in overall fusion rate relative to control patients (OR 2.45 [95% CI 1.20–4.99], p = 0.014) (Table 4 and Fig. 3C).

Subanalysis of Clinical Data

On meta-analysis, patients receiving some form of electrical stimulation were found to have a 126% increase in the odds of a successful fusion by last follow-up compared to controls (Fig. 3A). Table 5 summarizes the subgroup meta-analyses of the clinical data. The variables investigated include those listed as characteristic of patients with difficult-to-fuse spines, patients with a history of smoking, those undergoing revision surgery, those in whom interbody fusion is performed, and those undergoing a multilevel fusion, as well as the surgical level that was treated, the type of graft material used, and whether instrumentation was placed.

TABLE 5.

Random-effects subgroup meta-analysis of the clinical data*

Fusion Rate (no. fused/total)
VariableType of ESTAuthors & YearStimulation GroupControl GroupCochran’s QOR (95% CI) & p Value
Studies limited to difficult-to-fuse spinesDCSAndersen et al., 200917/4812/36
Kucharzyk, 199962/6556/653.212.14 (0.94–4.86); p > 0.05
Kane, 198825/3115/28
Overall (95% CI)73.3% (31.3–98.8%)59.2% (24.9–89.0%)
ICSFoley et al., 2008116/125104/120
Coric et al., 2018201/21776/920.252.34 (1.33–4.10); p = 0.003
Overall (95% CI)92.4% (89.4–95.0%)84.6% (79.5–89.1%)
AllOverall (95% CI)82.5% (64.1–95.2%)70.8% (52.2–86.3%)3.592.18 (1.43–3.32); p < 0.001
DCSMeril, 199485/9242/59
Rogozinski & Rogozinski, 199624/2614/184.442.46 (0.71–8.55); p > 0.05
Jenis et al., 20005/108/13
Overall (95% CI)83.1% (62.5–96.6%)70.5% (60.9–79.3%)
SmokerICSMooney, 199024/2712/20
Marks, 200018/190/38.054.48 (0.45–44.26); p > 0.05
Jenis et al., 20006/128/13
Overall (95% CI)80.3% (55.2–96.6%)47.0% (20.2–74.7%)
AllOverall (95% CI)82.2% (68.6–92.5%)62.5% (49.3–74.8%)12.502.84 (1.00–8.11); p = 0.05
DCSMeril, 199426/2814/20
Rogozinski & Rogozinski, 199627/2721/231.453.79 (0.99–14.53); p = 0.05
Jenis et al., 20006/78/9
Overall (95% CI)94.1% (82.8–99.6%)81.9% (67.1–93.0%)
NonsmokerICSMooney, 199035/3724/33
Marks, 200023/2310/166.373.66 (0.34–39.8); p > 0.05
Jenis et al., 20007/108/9
Overall (95% CI)91.5% (74.4–99.6%)71.7% (59.9–82.2%)
AllOverall (95% CI)93.1% (85.0–98.2%)77.0% (67.3–85.4%)7.813.58 (1.09–11.8); p = 0.04
Index surgery (no prior back surgery)DCSMeril, 1994101/10969/92
Rogozinski & Rogozinski, 199632/3424/270.503.69 (1.69–8.07); p = 0.001
Overall (95% CI)92.4% (87.6–96.2%)80.1% (66.1–91.1%)
CCSGoodwin et al., 199977/8577/9402.12 (0.87–5.21); p > 0.05
Overall (95% CI)90.6% (82.3–95.8%)81.9% (72.6–89.1%)
ICSMooney, 199059/6436/53
Marks, 200038/386/149.785.52 (1.17–25.95); p = 0.03
Linovitz et al., 200266/10448/97
Overall (95% CI)88.5% (60.6–100%)55.0% (40.7–69.0%)
AllOverall (95% CI)90.0% (78.9–97.3%)69.2% (55.6–81.2%)11.143.24 (1.69–6.21); p < 0.001
Revision surgery (prior back surgery)DCSMeril, 199412/138/11
Rogozinski & Rogozinski, 199619/1911/140.246.56 (0.98–44.0); p = 0.05
Overall (95% CI)95.7% (83.4–100)74.2% (56.4–88.6%)
ICSMarks, 20003/44/500.75 (0.03–17.51); p > 0.05
Overall (95% CI)75% (19.4–99.4)80% (28.4–99.5%)
AllOverall (95% CI)92.0% (75.2–99.7%)74.4% (58.4–87.6%)1.583.68 (0.72–18.73); p > 0.05
Posterior or posterolateral fusionDCSKane, 1988208/229143/187
Kucharzyk, 199962/6556/65
Andersen et al., 200917/4812/3611.341.77 (0.78–4.01); p > 0.05
Jenis et al., 200010/1718/22
Rogozinski & Rogozinski, 199651/5335/41
Overall (95% CI)79.4% (56.7–95.1%)73.6% (57.4–87.1%)
ICSJenis et al., 200014/2218/22
Linovitz et al., 200266/10448/973.940.95 (0.22–4.11); p > 0.05
Overall (95% CI)63.3% (54.8–71.4%)64.7% (32.7–90.6%)
AllOverall (95% CI)77.1% (56.3–92.6%)74.7% (60.8–86.4%)16.571.51 (0.82–2.79); p > 0.05
Interbody fusionDCSMeril, 1994113/12277/10304.24 (1.88–9.54); p < 0.001
Overall (95% CI)92.6 (86.5–96.6%)74.8% (65.2–82.8%)
ICSFoley et al., 2008116/125104/120
Coric et al., 2018201/21776/92
Mooney, 199059/6436/535.683.54 (1.71–7.31); p = 0.001
Marks, 200019/206/14
Overall (95% CI)92.3% (89.6–94.7%)74.1% (59.9–86.2%)
AllOverall (95% CI)92.3% (90.0–94.4%)74.8% (64.4–84.0%)6.083.56 (2.08–6.11); p < 0.001
Single-level fusionDCSKane, 198814/1610/16
Meril, 199485/9349/730.054.96 (2.32–10.63); p < 0.001
Rogozinski & Rogozinski, 199616/1618/20
Overall (95% CI)92.0% (84.6–97.2%)72.7% (56.3–86.4%)
ICSMooney, 199043/4629/40
Marks, 200018/186/121.328.77 (1.70–45.28); p = 0.01
Overall (95% CI)95.2% (87.0–99.5%)64.1% (42.7–82.8%)
AllOverall (95% CI)93.1% (88.5–96.6%)69.6% (58.8–79.4%)1.685.56 (2.91–10.64); p < 0.001
Multilevel (≥2) fusionDCSKane, 198811/155/12
Meril, 199423/2426/280.343.40 (1.15–10.0); p = 0.03
Rogozinski & Rogozinski, 199635/3717/21
Overall (95% CI)89.2% (76.1–97.5%)74.6% (45.5–95.0%)
ICSMooney, 199016/187/13
Marks, 200023/244/70.349.46 (2.16–41.43); p = 0.003
Overall (95% CI)91.4% (81.4–97.8%)54.6% (34.0–74.4%)
AllOverall (95% CI)90.4% (83.4–95.6%)68.0% (46.3–86.2%)1.884.86 (2.03–11.62); p < 0.001
Cervical fusionICSFoley et al., 2008116/125104/120
Coric et al., 2018201/21776/920.252.34 (1.33–4.10); p = 0.003
Overall (95% CI)92.4% (89.4–95.0%)84.6% (79.5–89.1%)
Lumbar or lumbosacral fusionDCSKane, 1988208/229143/187
Meril, 1994113/12277/103
Rogozinski & Rogozinski, 199651/5335/41
Kucharzyk, 199962/6556/6513.602.13 (1.08–4.21); p = 0.030
Jenis et al., 200010/1718/22
Andersen et al., 200917/4812/36
Overall (95% CI)82.2% (65.8–94.1%)73.9% (61.7–84.4%)
CCSGoodwin et al., 199977/8577/9402.12 (0.87–5.21); p > 0.05
Overall (95% CI)90.6% (82.3–95.8%)81.9% (72.6–89.1%)
ICSMooney, 199059/6436/53
Marks, 200041/4210/19
Jenis et al., 200014/2218/2215.902.84 (0.76–10.72); p > 0.05
Linovitz et al., 200266/10448/97
Overall (95% CI)81.6% (59.9–96.0%)62.2% (47.6–75.7%)
AllOverall (95% CI)82.9% (72.3–91.4%)70.9% (61.6–79.4%)29.682.25 (1.34–3.80); p = 0.002
AutograftDCSKane, 1988208/229143/187
Meril, 199451/5335/41
Kucharzyk, 199962/6556/658.862.03 (0.73–5.65); p > 0.05
Jenis et al., 200010/1718/22
Overall (95% CI)89.4% (79.7–96.2%)80.4% (75.1–85.2%)
ICSMooney, 199023/2514/19
Marks, 200019/205/1110.072.88 (0.28–29.58); p > 0.05
Jenis et al., 200014/2218/22
Overall (95% CI)84.0% (63.6–97.0%)68.9% (49.5–85.3%)
AllOverall (95% CI)87.4% (78.9–93.9%)78.5% (72.0–84.4%)19.392.14 (0.85–5.37); p > 0.05
AllograftICSMooney, 199025/2716/22
Marks, 200011/114/72.282.86 (1.18–6.95); p = 0.02
Foley et al., 2008116/125104/120
Overall (95% CI)92.8% (88.3–96.2%)76.7% (59.2–90.4%)
With instrumentationDCSMeril, 199424/2451/63
Rogozinski & Rogozinski, 199651/5335/41
Kucharzyk, 199962/6556/659.152.25 (0.50–10.1); p > 0.05
Jenis et al., 200010/1718/22
Overall (95% CI)91.4% (77.7–98.9%)83.2% (77.6–88.1%)
ICSMooney, 199044/4828/39
Marks, 20009/101/1
Jenis et al., 200014/2218/227.371.92 (0.94–3.93); p > 0.05
Foley et al., 2008116/125104/120
Coric et al., 2018201/21776/92
Overall (95% CI)88.5% (81.4–94.1%)82.4% (77.3–86.9%)
AllOverall (95% CI)89.8% (83.8–94.6%)82.8% (79.3–86.1%)16.441.94 (1.01–3.73); p = 0.05
Without instrumentationDCSKane, 1988208/229143/187
Meril, 199489/9826/405.922.64 (1.20–5.81); p = 0.02
Andersen et al., 200917/4812/36
Overall (95% CI)75.9% (45.3–96.3%)59.3% (34.1–82.2%)
ICSMooney, 199015/168/14
Marks, 200032/329/188.077.71 (0.86–69.38); p > 0.05
Linovitz et al., 200266/10448/97
Overall (95% CI)88.2% (54.7–99.9%)50.4% (41.9–58.9%)
AllOverall (95% CI)82.2% (63.7–95.0%)56.0% (40.7–70.7%)14.033.01 (1.56–5.84); p = 0.001
EST = electrical stimulation technology.Boldface type indicates statistical significance.

Only studies reporting the fusion rates for both the intervention (i.e., electrical stimulation) and control groups were included in the meta-analysis.

These studies included only patients with known risk factors for pseudarthrosis: 1) age > 60 years;5 2) failed prior fusion, grade II or worse spondylolisthesis, multilevel fusion, and/or other high-risk medical condition (e.g., gross obesity) (select cohort within study);57 3) multiple prior spine surgeries, failed prior fusion, segmental instability, spinal stenosis, and/or spondylolisthesis;60 4) age ≥ 65 years, active smoker, multilevel fusion, prior failed fusion, diabetic, and/or osteoporotic;26 or 5) active smokers and/or multilevel fusion.34

Notably, one or more electrical stimulation therapies resulted in statistically significant increases in the fusion rates compared to no stimulation in the following subgroups: patients with difficult-to-fuse spines, smokers, nonsmokers, patients undergoing index procedures, and those undergoing interbody fusions, single-level fusions, multilevel fusions, cervical fusions, lumbar/lumbosacral fusions, fusions with allograft alone, fusions with instrumentation, and fusions without instrumentation. In contrast, significant differences could not be detected between the fusion rates of patients receiving electrical stimulation therapy and controls in the following subgroups: revision surgery, posterior/posterolateral fusion subgroups, and autograft alone (Table 5).

Discussion

Spinal fusion is performed in the treatment of spinal pathologies of hundreds of thousands of Americans annually. Although most patients experience good outcomes, many experience nonunion, which can be associated with pain, persistent neurological compromise, and need for revision surgery.24 One class of surgical adjuvant therapies designed to avoid this outcome is electrical stimulation, including DCS, CCS, and ICS. In the current article, we have reported on the results of a systematic review and meta-analysis that evaluated the existing preclinical and clinical literature with the goal of addressing two questions: 1) To what degree does the technology improve bony fusion in animal models? 2) To what degree does the technology facilitate bony fusion in humans? We found that both DCS and ICS lead to significant improvements in fusion rates in humans and that DCS also produces significant increases in fusion rates in preclinical studies. Considering all electrical stimulation modalities as a whole, we found that electrical stimulation can significantly increase fusion rates among patients undergoing open fusion operations for a range of spinal pathologies. Subanalyses suggested that this effect persists in patients with difficult-to-fuse spines, smokers, those undergoing index procedures, and those undergoing interbody fusion. Further analyses investigating the effects based on the number of levels fused and whether instrumentation was used suggested that these variables do not alter the fusion benefits of electrical stimulation devices (Table 4).

The merits of any technology can be winnowed down to two questions: 1) Does it work? 2) Is it an economical means of achieving the goal? For medical technologies, we also consider the safety of the technology. Our review mainly addresses the first of these questions—namely, whether electrical stimulation is an effective means of promoting bony fusion. On the whole, our results suggest that the answer to this question is yes, as the results of our pooled analysis demonstrated significantly higher odds of fusion in patients treated with electrical stimulation (OR 2.26, p < 0.001). However, more in-depth investigation suggests that the majority of these results are driven by DCS and ICS. Between these 2 technologies, though, there appears to be no difference in efficacy.

This brings us to considering the questions of economics and safety profiles: 1) Are ICS and DCS economical means of promoting bony fusion? 2) Are they safe? The latter is most easily answered as both implantable DCS and noninvasive ICS devices have been approved under the relatively stringent FDA premarket approval process (class III devices) based on results of randomized controlled trials.5,34,47,53,57,72 The former question, i.e., whether the DCS and ICS devices are economical, is one that is harder to answer.

At present there are no high-quality studies evaluating the cost-effectiveness of electrical stimulation devices in the spine literature. However, back-of-the-envelope calculations are possible using estimates of device cost, pseudarthrosis rates, and cost of revision surgery. Prior studies of pseudarthrosis have found that the direct surgical costs of a revision operation are approximately $21,113 ± $11,895 for cervical operations58 and $28,069 ± $2508 for lumbar operations.2 To a gross approximation, this reduces to $25,000 per reoperation. Based on the present results, the approximate pseudarthrosis rate among patients receiving electrical stimulation therapy is 15%, compared to 27% in the control population. Of these patients, approximately half may require surgical revision for pseudoarthrosis.24,45,61,62,80 Accounting for this, the cost of surgical revisions for pseudarthrosis averaged across patients receiving electrical stimulation therapy is $1875, compared to $3375 for controls. For patients receiving electrical stimulation therapy, though, the cost of the stimulation device is an estimated additional $4000–$5000. Therefore, from a strictly financial standpoint, electrical stimulation devices may be a cost-ineffective means of improving fusion rates, except in patients with a high risk of nonunion.

The overall risk of pseudarthrosis among patients with difficult-to-fuse spines—commonly defined as those in whom prior fusion has failed, smokers, and those undergoing multilevel fusion procedures—has been reported to exceed 40%.5,13,18,57,62 Accordingly, the cost of revision operations averaged across these patients may exceed $10,000, suggesting that the use of electrical stimulation devices in this patient population may be cost-effective. Consistent with this, Medicare—the largest single insurer in the US—covers these devices only for patients with a history of multilevel fusion or a history of one or more prior failed fusion operations. This analysis does not consider the effect of nonunion on indirect costs, namely, days of lost work and decreased quality of life; however, it is likely that consideration of indirect costs will only increase the cost-effectiveness of these stimulation devices. Additional, high-quality investigation is warranted to evaluate this point.

Study Limitations

There are several limitations to this study. First, we were forced to exclude 6 preclinical and 3 clinical studies from the meta-analysis as they lacked control groups for estimation of odds ratios. This produces a potential selection bias that may limit the generalizability of the data. Second, the results of the study are based on a combination of prospective and retrospective studies. Retrospective studies are limited in the quality of the data they provide, which consequently limits the generalizability of the results of the present study. Nonetheless, we evaluated the quality of the clinical studies included in the meta-analysis (using the Critical Appraisal Checklists) and deemed each to have sufficient quality to be included in the present review. Additionally, although we provide estimates of the overall effect of electrical stimulation therapies at large and in subgroups, the heterogeneity of the included studies prevents us from answering the following questions: 1) Which patients will benefit most from electrical stimulation technologies? 2) For how long should treatment be continued? Furthermore, the definition of fusion varied between studies, suggesting that the outcome may have been distinct across studies, which would limit the validity of our meta-analysis. We describe this heterogeneity by presenting the definition and method of assessment of spinal fusion used by each study. All clinical studies employed plain radiograph– or CT scan–based radiological assessment, both of which are considered valid techniques in the clinical literature. Although a CT scan provides higher-resolution imaging and is therefore often considered the gold standard for fusion assessment, relative to standard radiography it is more expensive, exposes the patient to high radiation levels, and often provides no additional information.33 Nevertheless, the different assessment modalities impart heterogeneity to the results, which we attempted to address by employing random-effects versus fixed-effects models. Lastly, we pooled the results of several different electrical stimulation technologies. Though our results suggest that ICS and DCS have similar effects, they employ distinct technologies and have widely different patient compliance levels given that the latter is an implanted device, whereas the former is a wearable device. It is therefore possible that limitations in patient compliance among the ICS group limited the ability of our analysis to see differences in efficacy between the technologies. Given these limitations, it is apparent that future studies are necessary to directly compare the effectiveness of these different electrical stimulation technology modalities.

Conclusions

Here we report the results of the first systematic review and meta-analysis analyzing the effectiveness of electrical stimulation devices on spinal fusion in the preclinical and clinical literature. We found that these devices lead to significant increases in fusion rates, with a nearly fivefold increase in the odds in preclinical studies and a more than twofold increase in clinical studies. Subanalysis suggested that among the clinical population, DCS and ICS lead to significant decreases in pseudarthrosis rates, whereas CCS does not. Additional research is needed to analyze the cost-effectiveness of electrical stimulation devices to identify those patients in whom these devices are likely to be not only practically effective but also cost-effective.

Disclosures

Dr. Goodwin reports being a consultant for Augmedics and ROM3, having an ownership stake in AOSpine/NREF, and receiving textbook royalties from Kendall Hunt. Dr. Theodore receives royalties from DePuy and Globus Medical Inc.; he is a consultant for Globus. Dr. Witham reports receiving non–study-related grant support from the Gordon and Marilyn Macklin Foundation and Eli Lilly and Company; he is a consultant for DePuy Synthes Spine. Dr. Sciubba is a consultant for Baxter, DePuy Synthes, Globus, K2M, Medtronic, NuVasive, and Stryker. Mr. Cottrill reports receiving non–study-related grant support from the National Institute on Aging (F30AG063445).

Author Contributions

Conception and design: Sciubba, Cottrill, Pennington. Acquisition of data: Cottrill, Pennington. Analysis and interpretation of data: all authors. Drafting the article: all authors. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Sciubba. Statistical analysis: Cottrill, Pennington, Ahmed. Administrative/technical/material support: Sciubba, Theodore, Witham. Study supervision: Sciubba, Theodore, Witham.

Supplemental Information

Online-Only Content

Supplemental material is available with the online version of the article.

References

  • 1

    Aaron RKCiombor DMKeeping HWang SCapuano APolk C: Power frequency fields promote cell differentiation coincident with an increase in transforming growth factor-beta(1) expression. Bioelectromagnetics 20:4534581999

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 2

    Adogwa OOwens RKarikari IAgarwal VGottfried ONBagley CA: Revision lumbar surgery in elderly patients with symptomatic pseudarthrosis, adjacent-segment disease, or same-level recurrent stenosis. Part 2. A cost-effectiveness analysis: clinical article. J Neurosurg Spine 18:1471532013

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3

    Adogwa OParker SLShau DMendelhall SKCheng JAaronson O: Long-term outcomes of revision fusion for lumbar pseudarthrosis: clinical article. J Neurosurg Spine 15:3933982011

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4

    Aleem ISAleem IEvaniew NBusse JWYaszemski MAgarwal A: Efficacy of electrical stimulators for bone healing: a meta-analysis of randomized sham-controlled trials. Sci Rep 6:317242016

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 5

    Andersen TChristensen FBEgund NErnst CFruensgaard SØstergaard J: The effect of electrical stimulation on lumbar spinal fusion in older patients: a randomized, controlled, multi-center trial: part 2: fusion rates. Spine (Phila Pa 1976) 34:224822532009

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 6

    Balint RCassidy NJCartmell SH: Electrical stimulation: a novel tool for tissue engineering. Tissue Eng Part B Rev 19:48572013

  • 7

    Baranowski TJBlack J: The mechanism of faradic stimulation of osteogenesis in Blank MFindl E (eds): Mechanistic Approaches to Interactions of Electric and Electromagnetic Fields with Living Systems. Boston: Springer1987 pp 399416

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8

    Blackwell KARaisz LGPilbeam CC: Prostaglandins in bone: bad cop, good cop? Trends Endocrinol Metab 21:2943012010

  • 9

    Blumenthal SLBaker JDossett ASelby DK: The role of anterior lumbar fusion for internal disc disruption. Spine (Phila Pa 1976) 13:5665691988

  • 10

    Bodamyali TBhatt BHughes FJWinrow VRKanczler JMSimon B: Pulsed electromagnetic fields simultaneously induce osteogenesis and upregulate transcription of bone morphogenetic proteins 2 and 4 in rat osteoblasts in vitro. Biochem Biophys Res Commun 250:4584611998

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 11

    Bodamyali TKanczler JMSimon BBlake DRStevens CR: Effect of faradic products on direct current-stimulated calvarial organ culture calcium levels. Biochem Biophys Res Commun 264:6576611999

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 12

    Boden SDSumner DR: Biologic factors affecting spinal fusion and bone regeneration. Spine (Phila Pa 1976) 20 (24 Suppl):102S112S1995

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13

    Bolesta MJRechtine GR IIChrin AM: Three- and four-level anterior cervical discectomy and fusion with plate fixation: a prospective study. Spine (Phila Pa 1976) 25:204020462000

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 14

    Bose B: Outcomes after posterolateral lumbar fusion with instrumentation in patients treated with adjunctive pulsed electromagnetic field stimulation. Adv Ther 18:12202001

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 15

    Bozic KJGlazer PAZurakowski DSimon BJLipson SJHayes WC: In vivo evaluation of coralline hydroxyapatite and direct current electrical stimulation in lumbar spinal fusion. Spine (Phila Pa 1976) 24:212721331999

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 16

    Brighton CTAdler SBlack JItada NFriedenberg ZB: Cathodic oxygen consumption and electrically induced osteogenesis. Clin Orthop Relat Res (107):2772821975

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17

    Brighton CTWang WSeldes RZhang GPollack SR: Signal transduction in electrically stimulated bone cells. J Bone Joint Surg Am 83:151415232001

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 18

    Brown CWOrme TJRichardson HD: The rate of pseudarthrosis (surgical nonunion) in patients who are smokers and patients who are nonsmokers: a comparison study. Spine (Phila Pa 1976) 11:9429431986

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 19

    Cain CDAdey WRLuben RA: Evidence that pulsed electromagnetic fields inhibit coupling of adenylate cyclase by parathyroid hormone in bone cells. J Bone Miner Res 2:4374411987

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 20

    Calandruccio RABenton BF: Anterior lumbar fusion. Clin Orthop Relat Res 35:63681964

  • 21

    Cho MHunt TKHussain MZ: Hydrogen peroxide stimulates macrophage vascular endothelial growth factor release. Am J Physiol Heart Circ Physiol 280:H2357H23632001

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 22

    Cho PGJi GYHa YLee HYShin DA: Effect of the type of electrical stimulation on spinal fusion in a rat posterolateral spinal fusion model. Spine J 19:110611202019

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 23

    Choi YHChoi JHOh JWLee KY: Calmodulin-dependent kinase II regulates osteoblast differentiation through regulation of Osterix. Biochem Biophys Res Commun 432:2482552013

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 24

    Chun DSBaker KCHsu WK: Lumbar pseudarthrosis: a review of current diagnosis and treatment. Neurosurg Focus 39(4):E102015

  • 25

    Cook SDPatron LPChristakis PMBailey KJBanta CGlazer PA: Direct current stimulation of titanium interbody fusion devices in primates. Spine J 4:3003112004

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 26

    Coric DBullard DEPatel VVRyaby JTAtkinson BLHe D: Pulsed electromagnetic field stimulation may improve fusion rates in cervical arthrodesis in high-risk populations. Bone Joint Res 7:1241302018

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 27

    Cottrill EAhmed AKLessing NPennington ZIshida WPerdomo-Pantoja A: Investigational growth factors utilized in animal models of spinal fusion: systematic review. World J Orthop 10:1761912019

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 28

    Dejardin LMKahanovitz NArnoczky SPSimon BJ: The effect of varied electrical current densities on lumbar spinal fusions in dogs. Spine J 1:3413472001

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 29

    Epstein NEpstein JABenjamin VRansohoff J: Traumatic myelopathy in patients with cervical spinal stenosis without fracture or dislocation: methods of diagnosis, management, and prognosis. Spine (Phila Pa 1976) 5:4894961980

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 30

    Feyen JHMDi Bon Avan der Plas ALöwik CWGMNijweide PJ: Effects of exogenous prostanoids on the proliferation of osteoblast-like cells in vitro. Prostaglandins 30:8278401985

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 31

    Fischer CRCassilly RCantor WEdusei EHammouri QErrico T: A systematic review of comparative studies on bone graft alternatives for common spine fusion procedures. Eur Spine J 22:142314352013

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32

    Flynn JCHoque MA: Anterior fusion of the lumbar spine. End-result study with long-term follow-up. J Bone Joint Surg Am 61:114311501979

  • 33

    Fogel GRToohey JSNeidre ABrantigan JW: Fusion assessment of posterior lumbar interbody fusion using radiolucent cages: x-ray films and helical computed tomography scans compared with surgical exploration of fusion. Spine J 8:5705772008

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 34

    Foley KTMroz TEArnold PMChandler HC JrDixon RAGirasole GJ: Randomized, prospective, and controlled clinical trial of pulsed electromagnetic field stimulation for cervical fusion. Spine J 8:4364422008

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 35

    France JCNorman TLBuchanan MMScheel MVeale MAckerman ES: Direct current stimulation for spine fusion in a nicotine exposure model. Spine J 6:7132006

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 36

    France JCNorman TLSantrock RDMcGrath BSimon BJ: The efficacy of direct current stimulation for lumbar intertransverse process fusions in an animal model. Spine (Phila Pa 1976) 26:100210082001

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 37

    Fredericks DCSmucker JPetersen EBBobst JAGan JCSimon BJ: Effects of direct current electrical stimulation on gene expression of osteopromotive factors in a posterolateral spinal fusion model. Spine (Phila Pa 1976) 32:1741812007

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 38

    Freeman MFTukey JW: Transformations related to the angular and the square root. Ann Math Stat 21:6076111950

  • 39

    Friedenberg ZBAndrews ETSmolenski BIPearl BWBrighton CT: Bone reaction to varying amounts of direct current. Surg Gynecol Obstet 131:8948991970

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 40

    Friedenberg ZBBrighton CT: Bioelectric potentials in bone. J Bone Joint Surg Am 48:9159231966

  • 41

    Friedenberg ZBHarlow MCHeppenstall RBBrighton CT: The cellular origin of bioelectric potentials in bone. Calcif Tissue Res 13:53621973

  • 42

    Friedenberg ZBZemsky LMPollis RPBrighton CT: The response of non-traumatized bone to direct current. J Bone Joint Surg Am 56:102310301974

  • 43

    Fu RSelph SMcDonagh MPeterson KTiwari AChou R: Effectiveness and harms of recombinant human bone morphogenetic protein-2 in spine fusion: a systematic review and meta-analysis. Ann Intern Med 158:8909022013

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 44

    Fukada EYasuda I: On the piezoelectric effect of bone. J Phys Soc Jpn 12:115811621957

  • 45

    Girardo MBettini NDema ECervellati S: Uninstrumented posterolateral spinal arthrodesis: is it the gold standard technique for I degrees and II degrees grade spondylolisthesis in adolescence? Eur Spine J 18 (Suppl 1):1261322009

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 46

    Glazer PAHeilmann MRLotz JCBradford DS: Use of electromagnetic fields in a spinal fusion. A rabbit model. Spine (Phila Pa 1976) 22:235123561997

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 47

    Goodwin CBBrighton CTGuyer RDJohnson JRLight KIYuan HA: A double-blind study of capacitively coupled electrical stimulation as an adjunct to lumbar spinal fusions. Spine (Phila Pa 1976) 24:134913571999

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 48

    Guizzardi SDi Silvestre MGovoni PScandroglio R: Pulsed electromagnetic field stimulation on posterior spinal fusions: a histological study in rats. J Spinal Disord 7:36401994

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 49

    Haglin JMJain SEltorai AEMDaniels AH: Bone growth stimulation: a critical analysis review. JBJS Rev 5:e82017

  • 50

    Hartshorne E: On the causes and treatment of pseudarthrosis, and especially of that form of it sometimes called supernumerary joint. Am J Med 1:1431841

    • Search Google Scholar
    • Export Citation
  • 51

    Hofler RCSwong KMartin BWemhoff MJones GA: Risk of pseudoarthrosis after spinal fusion: analysis from the Healthcare Cost and Utilization Project. World Neurosurg120: e194e2022018

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 52

    Ito MFay LAIto YYuan MREdwards WTYuan HA: The effect of pulsed electromagnetic fields on instrumented posterolateral spinal fusion and device-related stress shielding. Spine (Phila Pa 1976) 22:3823881997

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 53

    Jenis LGAn HSStein RYoung B: Prospective comparison of the effect of direct current electrical stimulation and pulsed electromagnetic fields on instrumented posterolateral lumbar arthrodesis. J Spinal Disord 13:2902962000

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 54

    Kahanovitz NArnoczky SP: The efficacy of direct current electrical stimulation to enhance canine spinal fusions. Clin Orthop Relat Res (251):2952991990

    • Search Google Scholar
    • Export Citation
  • 55

    Kahanovitz NArnoczky SPHulse DShires PK: The effect of postoperative electromagnetic pulsing on canine posterior spinal fusions. Spine (Phila Pa 1976) 9:2732791984

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 56

    Kahanovitz NArnoczky SPNemzek JShores A: The effect of electromagnetic pulsing on posterior lumbar spinal fusions in dogs. Spine (Phila Pa 1976) 19:7057091994

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 57

    Kane WJ: Direct current electrical bone growth stimulation for spinal fusion. Spine (Phila Pa 1976) 13:3633651988

  • 58

    Kim EJChotai SWick JBStonko DPSivaganesan ADevin CJ: Patient-reported outcomes and costs associated with revision surgery for degenerative cervical spine diseases. Spine (Phila Pa 1976) 43:E423E4292018

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 59

    Kornblum MBFischgrund JSHerkowitz HNAbraham DABerkower DLDitkoff JS: Degenerative lumbar spondylolisthesis with spinal stenosis: a prospective long-term study comparing fusion and pseudarthrosis. Spine (Phila Pa 1976) 29:7267342004

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 60

    Kucharzyk DW: A controlled prospective outcome study of implantable electrical stimulation with spinal instrumentation in a high-risk spinal fusion population. Spine (Phila Pa 1976) 24:4654691999

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 61

    Lamberg TSRemes VMHelenius IJSchlenzka DKYrjönen TAOsterman KE: Long-term clinical, functional and radiological outcome 21 years after posterior or posterolateral fusion in childhood and adolescence isthmic spondylolisthesis. Eur Spine J 14:6396442005

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 62

    Leven DCho SK: Pseudarthrosis of the cervical spine: risk factors, diagnosis and management. Asian Spine J 10:7767862016

  • 63

    Linovitz RJPathria MBernhardt MGreen DLaw MDMcGuire RA: Combined magnetic fields accelerate and increase spine fusion: a double-blind, randomized, placebo controlled study. Spine (Phila Pa 1976) 27:138313892002

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 64

    Loguidice VAJohnson RGGuyer RDStith WJOhnmeiss DDHochschuler SH: Anterior lumbar interbody fusion. Spine (Phila Pa 1976) 13:3663691988

  • 65

    Lopez AJScheer JKDahdaleh NSPatel AASmith ZA: Lumbar spinous process fixation and fusion: a systematic review and critical analysis of an emerging spinal technology. Clin Spine Surg 30:E1279E12882017

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 66

    Luben RACain CDChen MCYRosen DMAdey WR: Effects of electromagnetic stimuli on bone and bone cells in vitro: inhibition of responses to parathyroid hormone by low-energy low-frequency fields. Proc Natl Acad Sci U S A 79:418041841982

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 67

    MacEwan MRTalcott MRMoran DWLeuthardt EC: Novel spinal instrumentation to enhance osteogenesis and fusion: a preliminary study. J Neurosurg Spine 25:3183272016

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 68

    Marks RA: Spine fusion for discogenic low back pain: outcomes in patients treated with or without pulsed electromagnetic field stimulation. Adv Ther 17:57672000

  • 69

    Meril AJ: Direct current stimulation of allograft in anterior and posterior lumbar interbody fusions. Spine (Phila Pa 1976) 19:239323981994

  • 70

    Moher DCook DJEastwood SOlkin IRennie DStroup DF: Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement. Quality of Reporting of Meta-analyses. Lancet 354:1896–19001999

    • Search Google Scholar
    • Export Citation
  • 71

    Moola SMunn ZTufanaru CAromataris ESears KSfetcu R: Chapter 7: Systematic reviews of etiology and risk in Aromataris EMunn Z (eds): Joanna Briggs Institute Reviewer’s Manual. Adelaide: The Joanna Briggs Institute2017 (https://reviewersmanual.joannabriggs.org/) [Accessed June 18 2019]

    • Search Google Scholar
    • Export Citation
  • 72

    Mooney V: A randomized double-blind prospective study of the efficacy of pulsed electromagnetic fields for interbody lumbar fusions. Spine (Phila Pa 1976) 15:7087121990

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 73

    Morris MTTarpada SPCho W: Bone graft materials for posterolateral fusion made simple: a systematic review. Eur Spine J 27:185618672018

  • 74

    Nerubay JKatznelson A: Clinical evaluation of an electrical current stimulator in spinal fusions. Int Orthop 7:2392421984

  • 75

    Nerubay JMarganit BBubis JJTadmor AKatznelson A: Stimulation of bone formation by electrical current on spinal fusion. Spine (Phila Pa 1976) 11:1671691986

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 76

    Poorman GWJalai CMBoniello AJWorley NMcClelland S IIIPassias PG: Bone morphogenetic protein in adult spinal deformity surgery: a meta-analysis. Eur Spine J 26:209421022017

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 77

    Rajaee SSBae HWKanim LEDelamarter RB: Spinal fusion in the United States: analysis of trends from 1998 to 2008. Spine (Phila Pa 1976) 37:67762012

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 78

    Rogozinski ARogozinski C: Efficacy of implanted bone growth stimulation in instrumented lumbosacral spinal fusion. Spine (Phila Pa 1976) 21:247924831996

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 79

    Salamanna FSartori MBrodano GBGriffoni CMartini LBoriani S: Mesenchymal stem cells for the treatment of spinal arthrodesis: from preclinical research to clinical scenario. Stem Cells Int 2017:35370942017

    • Search Google Scholar
    • Export Citation
  • 80

    Slizofski WJCollier BDFlatley TJCarrera GFHellman RSIsitman AT: Painful pseudarthrosis following lumbar spinal fusion: detection by combined SPECT and planar bone scintigraphy. Skeletal Radiol 16:1361411987

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 81

    Smeeth LHaines AEbrahim S: Numbers needed to treat derived from meta-analyses—sometimes informative, usually misleading. BMJ 318:154815511999

  • 82

    Stauffer RNCoventry MB: Anterior interbody lumbar spine fusion. Analysis of Mayo Clinic series. J Bone Joint Surg Am 54:7567681972

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 83

    Stippick TWSheller MR: Combined magnetic fields provide robust coverage for interbody and posterolateral lumbar spinal fusion sites. Med Biol Eng Comput 54:1131222016

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 84

    Tejano NAPuno RIgnacio JMF: The use of implantable direct current stimulation in multilevel spinal fusion without instrumentation. A prospective clinical and radiographic evaluation with long-term follow-up. Spine (Phila Pa 1976) 21:190419081996

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 85

    Tepper OMCallaghan MJChang EIGaliano RDBhatt KABaharestani S: Electromagnetic fields increase in vitro and in vivo angiogenesis through endothelial release of FGF-2. FASEB J 18:123112332004

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 86

    Toth JMSeim HB IIISchwardt JDHumphrey WBWallskog JATurner AS: Direct current electrical stimulation increases the fusion rate of spinal fusion cages. Spine (Phila Pa 1976) 25:258025872000

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 87

    Tsutsumimoto TShimogata MYoshimura YMisawa H: Union versus nonunion after posterolateral lumbar fusion: a comparison of long-term surgical outcomes in patients with degenerative lumbar spondylolisthesis. Eur Spine J 17:110711122008

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 88

    Wang ZClark CCBrighton CT: Up-regulation of bone morphogenetic proteins in cultured murine bone cells with use of specific electric fields. J Bone Joint Surg Am 88:105310652006

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 89

    Weiss AJElixhauser AAndrews RM: Characteristics of operating room procedures in U.S. hospitals, 2011: Statistical Brief #170 in Healthcare Cost and Utilization Project (HCUP) Statistical Briefs. Rockville, MD: Agency for Healthcare Research and Quality (US)2014

    • Search Google Scholar
    • Export Citation
  • 90

    Welch WCWillis SLGerszten PC: Implantable direct current stimulation in para-axial cervical arthrodesis. Adv Ther 21:3894002004

  • 91

    Yoshihara H: Rods in spinal surgery: a review of the literature. Spine J 13:135013582013

  • 92

    Zayzafoon M: Calcium/calmodulin signaling controls osteoblast growth and differentiation. J Cell Biochem 97:56702006

  • 93

    Zhuang HWang WSeldes RMTahernia ADFan HBrighton CT: Electrical stimulation induces the level of TGF-beta1 mRNA in osteoblastic cells by a mechanism involving calcium/calmodulin pathway. Biochem Biophys Res Commun 237:2252291997

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 94

    Zhuo XLi CLi BLi ZLv HHuang J: Effects of combined magnetic fields treatment and nano-hydroxyapatite coating on porous biphasic calcium phosphate bone graft in rabbit spinal fusion model. Spine (Phila Pa 1976) 43:E625E6332018

    • Crossref
    • Search Google Scholar
    • Export Citation

If the inline PDF is not rendering correctly, you can download the PDF file here.

Article Information

Contributor Notes

Correspondence Daniel Sciubba: Johns Hopkins School of Medicine, Baltimore, MD. dsciubb1@jhmi.edu.INCLUDE WHEN CITING Published online October 8, 2019; DOI: 10.3171/2019.5.SPINE19465.Disclosures Dr. Goodwin reports being a consultant for Augmedics and ROM3, having an ownership stake in AOSpine/NREF, and receiving textbook royalties from Kendall Hunt. Dr. Theodore receives royalties from DePuy and Globus Medical Inc.; he is a consultant for Globus. Dr. Witham reports receiving non–study-related grant support from the Gordon and Marilyn Macklin Foundation and Eli Lilly and Company; he is a consultant for DePuy Synthes Spine. Dr. Sciubba is a consultant for Baxter, DePuy Synthes, Globus, K2M, Medtronic, NuVasive, and Stryker. Mr. Cottrill reports receiving non–study-related grant support from the National Institute on Aging (F30AG063445).