Anterior cervical discectomy and fusion (ACDF) is one of the most prevalent elective spinal surgeries for degenerative disc disease.1 With the increasing number of fusions being performed every year, new surgical materials (i.e., bone graft substitutes) have been introduced for the purpose of enhancing fusion and avoiding morbidity of autograft harvesting.2 During the past 20 years, bone graft substitute use has proliferated and the cost of these products has steadily increased. Thus, it is imperative to evaluate the radiographic and clinical outcomes associated with different bone graft substitutes (Table 1).
Bone graft substitutes have become an increasingly popular alternative to iliac crest autograft harvesting due to the morbidity; thus, effort has been spent in development and testing of these substitutes.3,4 In most cases, the aim is to produce a substance that improves fusion and therefore may be osteogenic (synthesizes bone), osteoconductive (promotes ingrowth of blood vessels and cells), or osteoinductive (promotes differentiation of stem cells to form osteoblasts).5
Four main classes of bone graft substitutes have been developed over the last 50 years including ceramic-based synthetic grafts, allografts, recombinant human bone morphogenetic proteins (BMPs), and mesenchymal stem cells. In addition, bone marrow aspirate can be used but requires processing including centrifuge and preparation for use. These bone graft substitutes vary with respect to their osteogenic, osteoconductive, and osteoinductive properties, as well as fusion rates. Many studies compare the efficacy of one bone graft substitutes to autograft, but no comparison of all substitutes available has been made. We aim to determine radiographic and clinical outcomes for these bone graft substitutes in single- or double-level ACDFs for degenerative disease.
MATERIALS AND METHODS
A comprehensive review of the literature was completed using the databases EMBASE and PUBMED. The search included the words: “anterior cervical discectomy and fusion” or “ACDF” and “allograft,” “mesenchymal stem cells,” “synthetic bone graft,” “bone morphogenetic protein,” “BMP,” “bone marrow aspirate,” “PEEK,” “ceramic-based synthetic.” Relevant abstracts were reviewed to determine whether inclusion criteria were met. After the articles were selected for inclusion, references within the articles were examined for inclusion. Only articles written in English were considered.
Criteria for Inclusion of Studies
Type of Intervention
ACDFs in which single- or double-level procedures were completed were reviewed. Discectomy alone, total disc replacements, and studies in which no structural graft (PEEK, allograft, or ceramic-based synthetic strut/ring) was used were excluded. Fusions in which no anterior plate and screw instrumentation were used were eliminated to reduce the confounding impact of noninstrumented versus instrumented fusion.6,7 Studies in which multilevel fusions could not be separated out from single- or double-level fusions were excluded. Studies were not included if no bone graft substitute was used.
Types of Studies
Randomized control trials and retrospective and prospective studies were included. Case reports, systematic review, literature reviews, and meta-analysis were excluded due to low level of evidence. Although bone graft substitutes in spine surgery have been around since 1958, the vast majority of commercially available products have been developed since 1990.8 Therefore, articles prior to 1990 were excluded. We included only publications with follow-up at least 6 months after surgery with radiographic analysis of fusion. Often clinical outcomes were reported and included. Minimum sample size of 20 subjects was required for inclusion.
Types of Patients
Patients enrolled in the studies had degenerative disc disease with resultant radiculopathy or myelopathy. Trauma, tumor, or infectious causes were excluded as fusion rates and clinical outcomes in these diseases can be significantly affected. Patients had to be 18 years of age or older with no maximum age. All sexes were included; however, animal studies were not. Smoking, older age, female sex, and high body mass index (BMI) were documented if noted in studies as these serve as potential risk factors for osteoporosis and poor fusion.9,10
Information from the studies was used to compare fusion rates and clinical outcomes for surgeries. Clinical outcomes were measured using the Oswestry Disability Index, Neck Disability Index (NDI), and visual analog scores (VAS). Fusion rates were evaluated by spinal x-rays or computed tomographies (CTs) and were recorded for all studies including how long postoperative films were taken. Details of studies were recorded on a predeveloped form. Article name, authors, number of levels operated, number of participants, study design, reason for surgery, instrumentation, type of bone graft, clinical outcome variables, and radiographic outcomes were recorded. Presence or absence of smoking, diabetes, BMI, or weight was also noted.
A decision tree of all found articles to the selected articles is shown. A table showing the articles analyzed in detail was constructed with type of study, number of participants, and graft used. A separate table was used to detail the results of each study. In addition, a table detailing the covariates that affect fusion was developed.
Fusion rates and clinical outcomes were compared between each bone graft substitute using the student t test with P values reported. Mean fusion rate, Oswestry Disability Index, NDI, and VAS pain scores were obtained. Mean with standard deviations (SDs) were calculated for each covariate. Student t test was computed to compare the covariates between each bone graft substitute.
Electronic searches of the databases identified 479 unique articles. A total of 181 publications were deemed not relevant by the title leaving 298 articles for screening in depth. After reviewing the abstracts, 224 publications were excluded. Upon reading the full text, 53 additional articles were excluded leaving 19 publications. Citations within these articles were reviewed and 316 new articles found. Two hundred eighty-eight articles were eliminated after reviewing the title, leaving 28 for review. Only three were deemed appropriate by inclusion criteria. Thus, 22 articles were included in the final review (Figure 1).
Three articles discussed the use of mesenchymal stem cells, six ceramic-based synthetics, six BMP, sixteen allografts (nine struts, six rings filled with other grafts, one cancellous filing for ceramic-based synthetic (CBS) ring) and one bone marrow aspirate. Peolyetheretherketone grafts were used as structural components filled with bone graft substitute as well as allograft and CBS rings were used in conjunction with other grafts or as stand-alone strut grafts. This is more than 22 as 7 studies looked at 2 or more grafts used together, and 5 studies had 2 arms comparing different grafts to each other. Nine studies had no comparison group and eight were compared to autograft alone. All comparison arms performed superior or equivalent to their autograft counterpart except ceramic-based synthetic grafts.
Twelve were retrospective reviews, eight nonrandomized prospective clinical trials, and two randomized control trials. All studies meeting criteria for analysis were written from 2003 to 2017, despite starting the database search at 1990. This was mainly due to lack of anterior plate and instrumentation use which became more prevalent after reports of its superiority over noninstrumented fusions in ACDFs starting around 2000. Most studies looked at both one- and two-level surgeries, with six looking at one-level and four two-level alone. The average sample size of participants receiving bone graft substitutes was 67.5 (SD 39.3) with a median of 62.5 (Table 2).
Imaging with x-ray and/or CT was completed in all studies at a minimum of 6 months after surgery to assess fusion and in some studies, subsidence.11 Clinical outcomes were available in most studies, typically standardized with NDI and VAS. Average length of follow-up reported in the studies was 16.9 months (SD 6.37) ranging from 6 to 24 months. Most studies had a fusion rate of 80% or more at last follow-up. One study12 using ceramic-based synthetic graft only had 62% fusion, although this group was only followed for 6 months. Another significant outlier was Kim et al that used allograft alone with only 31% fusion at 12-months post-op14 (Table 3).
Fusion was determined with plain x-rays for 11 studies, CT in 2 studies and both for 9 studies. All studies using BMP showed 100% fusion rate despite length of the study or other graft used.14–19 Ceramic-based synthetic grafts had the lowest fusion rate when used alone at 80.5% (16.8 SD).20–22 Zulkefli et al12 followed patients for only 6 months with a lower fusion rate of only 62%. When this study is excluded as an outlier the mean fusion rate was 86.6% (SD 14.0). When considering all studies in which ceramic-based synthetic grafts were used, fusion rate increased to 87.0% (SD 15.0) which holds with the findings that other bone graft substitutes resulted in higher fusion rates.20–23 Mesenchymal stem cells with allograft spacers or peolyetheretherketone (PEEK) cages had a fusion rate of 89.6% (SD 3.3).24–26 One study using bone marrow aspirate with allograft and ceramic-based synthetic showed a 100% fusion rate in a sample of 66 patients.27 Allograft when used alone resulted in an 87.3% fusion rate on average, although this is significantly decreased due to one outlier study with fusion rate of 31%.13,15,16,24,28–33 When the outlier study13 is removed, average fusion rate increases to 93.5% (SD 4.97). When considering all studies using allograft (minus outlier study), fusion rate increased to 94% (SD 5.3) which coincides with 100% fusion rate seen with BMP (Table 4).
There was significantly higher fusion rate with BMP compared to other bone graft substitutes (P < 0.001). Conversely, there was significantly worse fusion in ceramic-based synthetic grafts compared to other bone grafts (P < 0.001). There was no significant difference between fusion rate with mesenchymal stem cells compared to allograft (P > 0.05). There was no significant difference in clinic outcomes between different bone graft substitutes (P > 0.05), although dysphagia was significantly greater in studies using BMP (P < 0.001).
It is known that smoking and increased BMI can decrease the rate of bone fusion.9,26 These covariates and patient demographics, which may influence fusion were analyzed when available (Table 5). Smoking rate was presented in 13 or the 22 studies with an average of 29% smoking in these studies. No statistical difference in smoking was noted between any substrate group.
Mean age at intervention for all studies was 50.09 (SD 3.58). There was no statistical significant difference in mean ages between bone graft substitutes (P > 0.05). The average percentage of women in each study was equivocal at 49.57 (SD 10.99). The percentage of women was statistically higher in BMP studies compared to ceramic-based synthetics studies (P = 0.0017) and allograft (P = 0.002). All other comparisons showed no significant difference (P > 0.05).
BMI was recorded in 5 out of the 22 studies, with a mean BMI in allograft groups of 26.2, BMP 28.8, mesenchymal stem cell 28.5, and ceramic-based synthetic 24.75. There was no statistical significance between allograft and other bone graft substitutes, but BMI was significantly higher in mesenchymal stem cell studies compared to ceramic-based synthetics (P = 0.02).
There are an increasing number of bone graft substitutes available, even over the last few years. Given the rapid explosion of these substitutes, it becomes incredibly difficult to truly compare the efficacy of these products over the long term when new substitutes are rapidly entering the market. This systematic review shows that these bone graft substitutes result in similar clinical outcomes. BMP showed the highest fusion rate, however, had the highest rate of dysphagia and complications. Ceramic-based synthetics had the lowest fusion rates. Mesenchymal stem cells and allograft were similar in fusion rate, but the addition of mesenchymal stem cells to allograft did not increase fusion.
The results indicate that high osteoinductive properties in grafts (as seen in BMP) seems to be the most important factor in bone fusion in ACDF surgery for degenerative disease but can lead to significant complications due to promotion of inflammation and heterotopic ossification. Similarly, osteoconduction alone leads to fusion of vertebrae but at much lower rates and longer time frame as seen in ceramic-based synthetics. Hence, allograft and mesenchymal stem cells, which holds both moderate osteoconductive and osteoinductive properties seems to result in the best fusion rates with least complications. Thus, to differentiate between these two grafts, cost should be considered.
The average cost of each product from literature review is as follows: $80 to $400/cm3 for ceramic-based synthetics, $100 to $200/cm3 for DBM allograft and $700 to $1000 for allograft struts, $4300 weighted average for BMP, $500/cm3 of mesenchymal stem cells and $700 for bone marrow aspirate kits. For all bone grafts except allograft spacers and some ceramic-based synthetic spacers, a structural cage (PEEK) must be used and considered in the cost averaging $600.
Ceramic-based synthetics are the cheapest implants when no other structural spacer is used; however, they have significantly lower fusion rates. BMP had significantly higher fusion rates compared to other grafts, although it has significantly higher cost and complication rates. BMP, additionally, is not FDA approved for use in cervical spine due to these known complications with airway compromise and dysphagia. Mesenchymal stem cells did not increase fusion rates when added to allograft spacers, and thus added expense should be reconsidered given no statistically significant change in fusion or clinical outcome occurs. Allograft has the highest fusion rate for the relatively low cost compared to other bone graft substitutes with equivalent clinical outcomes. Bone marrow aspirate was only used in one study, but did achieve a fusion rate of 100%. This needs to be explored in more depth with other studies; however, given the need for a separate incision and the extra time in the operating room for extraction, it may not be efficacious. Autograft crest is in terms of financial cost the lowest as there are no added expenses, although the added morbidity from this separate incisional acquisition has led to the surfacing of these other grafts. It must be remembered that autograft bone also is the model in which these other grafts have been made and the decision to use autograft versus other bone graft substitutes should be weighed with risks and benefits in each surgeon's hand.
Of note, the Kim et al's study had a very low fusion rate compared to the other studies with allograft and from most published literature of fusion. As this was an excessive outlier, data were compared with and without this value. These patients were followed for a year, which typically shows fusion rates of more than 80% as seen in the other studies.20,34,35 No covariates were significantly different in this study and the standard Smith-Robinson approach with anterior plating was used for all surgeries. Kim et al used x-rays for final fusion analysis, which has been known to underestimate fusion compared to CT scans, although a majority of the studies used x-ray alone and had much higher fusion rates.14,36 No other significant findings explain the low fusion rate in this study; however, the autograft comparison group in this study similarly had a low fusion rate, albeit higher, at 65% which may indicate the internal validity or techniques used during surgery or fusion assessment were not to standard.
One study12 using ceramic-based synthetic graft only had 62% fusion, although this group was only followed for 6 months. Radiographic fusion typically occurs around 6 to 9 months postoperative,37 which may explain the lower fusion rate in this study and could artificially lower the average fusion rate of the ceramic-based synthetic grafts.
An additional caveat must be considered. The field of bone graft substitutes has vastly expanded in recent years. This has resulted in a vast array of inadequately studied substitutes that have entered the market. Although there is significantly more variety, the crowded market makes it difficult to study any one product in depth and with long-term follow-up. Therefore, it is difficult for health providers to make evidence-based decisions in choosing the optimal bone graft substitute. Unfortunately, several other drivers then become factors in influencing what substitute a surgeon may choose including relationships with industry, poor quality anecdotal evidence or advertising, availability at their institution, and cost.
This is a systematic review of the literature and thus no patient data were collected and the statistical power of the review will be limited to sample sizes and exposures present in the articles published. Few randomized control trials have been completed on some of the bone graft substitutes such as mesenchymal stem cells or bone marrow aspirate; thus, the level of evidence is lower in these manuscripts. Limitations may also occur from lack of details on covariates such as smoking and BMI in the publications. Limitations on assessing fusion rates on radiographs or CT, lack of agreement on criteria for fusion, and high interobserver and intraobserver reliability in assessing a fusion are all limitations in this study especially when assessing 22 different articles. Selection of studies, methods of analysis, and interpretations are all common sources of limitations for systematic reviews which this review does not escape.
Implications of Research
The clinical outcomes are similar across all substitutes; thus, we recommend institutions consider scaling back the use of more expensive bone graft substitutes such as BMP and mesenchymal stem cells for ACDF surgeries because allograft is a cheaper substrate with similar fusion rate. With these data, surgeons and hospitals can better determine, which bone graft substitutes should be used in routine degenerative ACDF surgeries. Follow-up research will be performed to determine whether different bone graft substitutes affect fusion in lumbar, thoracic, or posterior cervical surgeries including on-lay grafts as well. With the rising cost of healthcare and push to lower cost to patients and hospitals, determining the most efficacious and cost-effective bone graft substitute could be of great benefit.Key PointsCritical evaluation bone graft substitutes are necessary to optimize radiographic and clinical outcomes for ACDF in degenerative disease.Allograft alone has the lowest cost with similar fusion rates and clinical outcomes compared to other bone graft substitutes.A systematic literature review of 22 published articles reporting fusion rates in ACDF procedures using bone graft substitutes was conducted.
1. Oglesby M, Fineberg S, Patel A, et al. Epidemiological trends in cervical spine surgery for degenerative diseases between 2002 and 2009. Spine (Phila Pa 1976)
2. Morrow T, Felcone L. Defining the difference: what makes biologics unique. Biotechnol Healthc
3. Banwart J, Asher M, Hassanein R. Iliac crest bone graft harvest donor site morbidity: a statistical evaluation. Spine (Phila Pa 1976)
4. Pollock R, Alcelik I, Bhatia C, et al. Donor site morbidity following iliac crest bone harvesting for cervical fusion
: a comparison between minimally invasive and open techniques. Eur Spine J
5. Oryan A, Alidadi S, Moshiri A, et al. Bone regenerative medicine: classic options, novel strategies, and future directions. J Orthop Surg Res
6. Wang J, McDonough P, Endow K, et al. Increased fusion
rates with cervical plating for two-level anterior cervical discectomy and fusion
. Spine (Phila Pa 1976)
7. Chung C, Kim C. Anterior plating is better than the stand-alone cage in the restoration of segmental kyphosis. Spine J
8. Tilkeridis K, Touzopoulos P, Ververidis A, et al. Use of demineralized bone matrix in spinal fusion
. World J Orthop
9. Andersen T, Christensen F, Laursen M, et al. Smoking as a predictor of negative outcome in lumbar spinal fusion
. Spine (Phila Pa 1976)
10. Djurasovic M, Bratcher K, Glassman S, et al. The effect of obesity on clinical outcomes after lumbar fusion
. Spine (Phila Pa 1976)
11. Hakało J, Wroński J, Ciupik L. Subsidence and its effect on the anterior plate stabilization in the course of cervical spondylodesis. Part I: definition and review of literature. Neurol Neurochir Pol
12. Zulkefli A, Jeyasilan K, Zairul A, et al. Comparison of fusion
rate between autogenous tricortical iliac crest bone graft and hydroxyapatite block graft in anterior cervical discectomy and fusion
surgery. Malaysian Orthop J
13. Hayeri MR, Tehranzadeh J. Diagnostic imaging of spinal fusion
and complications. Appl Radiol
2009; online publication; Accessed December 7, 2017.
14. Kim S, Park K, Jung S, et al. An early comparative analysis of the use of autograft versus allograft
in anterior cervical discectomy and fusion
. Korean J Spine
15. Baskin D, Ryan P, Sonntag V, et al. A prospective, randomized, controlled cervical fusion
study using recombinant human bone morphogenetic protein
-2 with the cornerstone-SRTM allograft
ring and the ATLANTISTM
anterior cervical plate. Spine (Phila Pa 1976)
16. Vaidya R, Carp J, Sethi A, et al. Complications of anterior cervical discectomy and fusion
using recombinant human bone morphogenetic protein
-2. Eur Spine J
17. Vaidya R, Weir R, Sethi A, et al. Interbody fusion
and rhBMP-2 leads to consistent fusion
but early subsidence. J Bone Joint Surg Br
18. Buttermann G. Prospective nonrandomized comparison of an allograft
with bone morphogenic protein versus an iliac-crest autograft in anterior cervical discectomy and fusion
. Spine (Phila Pa 1976)
19. Stachniak J, Diebner J, Brunk E, et al. Analysis of prevertebral soft-tissue swelling and dysphagia in multilevel anterior cervical discectomy and fusion
with recombinant human bone morphogenetic protein
-2 in patients at risk for pseudarthrosis. J Neurosurg Spine
20. Lu D, Tumialan L, Chou D. Multilevel anterior cervical discectomy and fusion
with and without rhBMP-2: a comparison of dysphagia rates and outcomes in 150 patients. J Neurosurg Spine
21. Yi J, Lee G, Nam W, et al. A prospective randomized clinical trial comparing bone union rate following anterior cervical discectomy and fusion
using a polyetheretherketone cage: hydroxyapatite/B-tricalcium phosphate mixture versus hydroxyapatite/demineralized bone matrix mixture. Asian Spine J
22. Yang X, Liu L, Song Y, et al. Outcome of single level anterior cervical discectomy and fusion
using nano-hydroxyapatite/polyamide-66 cage. Indian J Orthop
23. Ahn J, Lee J, Kim J. Comparative study of clinical outcomes of anterior cervical discectomy and fusion
using autobone graft or cage with bone substitute. Asian Spine J
24. Chen J, Wu C, Lee S. The use of a hollow polymethylmethacrylate cervical spacer with plating in the treatment of single level cervical disc disease. Chang Gung Med J
25. McAnany S, Ahn J, Elboghdady I, et al. Mesenchymal stem cell allograft
as a fusion
adjunct in one- and two-level anterior cervical discectomy and fusion
: a matched cohort analysis. Spine J
26. Vanichkachorn J, Peppers T, Bullard D, et al. A prospective clinical and radiographic 12-month outcome study of patients undergoing single-level anterior cervical discectomy and fusion
for symptomatic cervical degenerative disc disease
utilizing a novel viable allogeneic, cancellous, bone matrix (trinity evolution™) with a comparison to historical controls. Eur Spine J
27. Eastlack R, Garfin S, Brown CMS. Osteocel Plus cellular allograft
in anterior cervical discectomy and fusion
: evaluation of clinical and radiographic outcomes from a prospective multicenter study. Spine (Phila Pa 1976)
28. Khoueir P, Oh B, DiRisio D, et al. Multilevel anterior cervical fusion
using a collagen-hydroxyapatite matrix with iliac crest bone marrow aspirate
: an 18-month follow-up study. Neurosurgery
29. Samartzis D, Shen F, Matthews D, et al. Comparison of allograft
to autograft in multilevel anterior cervical discectomy and fusion
with rigid plate fixation. Spine (Phila Pa 1976)
30. Suchomel P, Barsa P, Buchvald P, et al. Autologous versus allogenic bone grafts in instrumented anterior cervical discectomy and fusion
: a prospective study with respect to bone union pattern. Eur Spine J
31. Balabhadra R, Kim D, Zhang H. Anterior cervical fusion
using dense cancellous allografts and dynamic plating. Neurosurgery
32. Samartzis D, Shen F, Goldberg E, et al. Is autograft the gold standard in achieving radiographic fusion
in one-level anterior cervical discectomy and fusion
with rigid anterior plate fixation? Spine (Phila Pa 1976)
33. Rhee J, Patel N, Yoon S, et al. High graft resorption rates with dense cancellous allograft
in anterior cervical discectomy and fusion
. Spine (Phila Pa 1976)
34. Lee D, Cho J, Hwang C, et al. What is the fate of pseudarthrosis detected 1 year after anterior cervical discectomy and fusion
? Spine (Phila Pa 1976)
35. Miller L, Block J. Safety and effectiveness of bone allografts in anterior cervical discectomy and fusion
surgery. Spine (Phila Pa 1976)
36. Floyd T, Ohnmeiss D. A meta-analysis of autograft versus allograft
in anterior cervical fusion
. Eur Spine J
37. Park D, Rhee J, Kim S, et al. Do CT scans overestimate the fusion
rate after anterior cervical discectomy and fusion
? J Spinal Disord Tech