General Discussion | Basic Q & A | Success Rates | The Flavors | Hardware | History of Fusion Technique
Warning: Fusion of the lumbar spine, or any level of the spine for that matter, is an end-of-the road treatment intervention for back and/or leg pain. It is imperative the patient try every other form of care first, before going down this road—it is really a big deal and should not be taken lightly!
Why do I say this? Because the discs of the lumbar spine can be like a house of cards: if you topple one (or fuse one), then the others may fall (may need to be fused later). For example, if you fuse L4/5, then the levels above and below (L3/4 and L5/S1 respectively) will experience abnormally high biomechanical stress which may well lead to their untimely demise secondary to painful degenerative change (the same process that most likely cause the original level to be fused) and need for subsequent fusion at the adjacent level. Then, with the passage of more time, undo biomechanical stress will affect the levels adjacent to the second fusion procedure, which in turn puts undue biomechanical stress on those levels, which may result in their untimely demise and need for a third fusion surgery--and the cycle continues (some clinicians call this the "domino effect".
General Discussion: For patients who are at the end of the road with regard to pain and can't stand it any longer (as over 700,000 Americans were in 2004 ), the next big dilemma is choosing which "flavor" of fusion to try, for their are half-a-dozen or so techniques that are all supported/semi-supported by the medical literature. These techniques include posterolateral fusion (PLF); posterior lumbar interbody fusion (PLIF); anterior lumbar interbody fusion (ALIF); transforaminal lumbar interbody fusion (TLIF); a mixture of ALIF, PLIF, and PLF; extreme lateral interbody fusion (XLIF) and artificial disc replacement (ADR).
The decisions aren't over yet: After choosing which technique to employ, there are more decisions to be made, such as whether or not to use posterior instrumentation (which also comes in different flavors),what interbody grafting (cage) to use (they come in different flavors) and/or which, if any, biologic to use as an augmentation (i.e., do you use rhBMP-2, OP-1 [aka rhBMP-7], demineralized bone matrix, biphasic calcium phosphate, or tricalcium hydroxyapatite vs. iliac crest autograft(ICAG) or allograft. Although there is a lot to learn about fusion, the patient should not take this homework lightly, for (as mentioned above) this is a very invasive procedure and sometimes horrible things can occur (retrograde ejaculation, severe epidural fibrosis, arachnoiditis, cauda equina syndrome, severe chronic radiculopathy etc.).
Patient Selection: Although the surgical technique used to accomplish fusion is important [18, 19, 20], it is not the most important factor for obtaining a successful fusion surgery. The most important factor, indisputably, is patient selection—make damn sure you are a perfect candidate for fusion before trying it , because in a fairly recent study by Bono et al. , certain lumbar diagnoses did much better with regard to successful fusion and patient outcome when compared to others . Many spine surgeons agree that a lot of the "bad reputation" that fusion has comes from poor patient selection.
Another important factor for achieving a successful fusion is permanent work/play restrictions. The patient must understand that there are certain permanent restrictions that he/she must follow after their procedure, in order to mitigate undue stress upon the adjacent segments (remember adjacent segments already have natural undue stress placed upon them because of the altered biomechanics following the index fusion).
For example, patient should be restricted from heavy lifting; repetitive bending, twisting and/or stooping; prolonged sitting, standing, or walking; and should use common sense when it comes to selecting hobbies/recreational activity. Failure to do so (depending on patient genetics for collagen and proteoglycans) may well lead to an untimely wearing out of the adjacent unfused disc(s), which in turn will lead to adjacent level fusion-- the odds for success of a second fusion surgery typically are not very good.
Some Basic Questions and Answers:
Before I get too deep, let's talk about some basics. And here is my usual admonition: if you're not familiar with the anatomy of the spine and disc, please go to my anatomy page and study-up, for I will assume that you have a basic understanding of the anatomy and physiology of the lumbar spine. And it is also important to understand that two adjacent vertebrae and the disc they share (for example, L4 and L5) are collectively called a motion segment  or motor unit . I will use the term motion segment.
Why am I having a fusion? Although lumbar fusion can be used to manage rare conditions, such as traumatic fractures, tumors or infection of the spine, typically it is offered secondary to degenerative disease of the vertebra and/or disc which has/have become symptomatic– and I mean severely symptomatic with Oswestry Disability Index Scores well over the 40s. Surgeons often grouped (diagnosed) these degenerative conditions broadly into categories such as spondylolisthesis/spondylolysis, revision surgery, discogenic pain, degenerative disc disease, or chronic low back pain.However, there is a call by the research community  for surgeons to more precisely diagnose patients prior to surgery, for there is evidence that the more specific subdiagnoses have statistically different outcome results (see more below). One commonality that all diagnoses share is the fact that the natural movement between motion segments is thought to be responsible for the genesis of the patients chronic, intractable back/and/or leg pain. And by eliminating this natural movement, via the fusion procedure, the source of the pain will be eliminated.
Where is this pain-generating motion occurring? The motion that fusion tries to eliminate occurs specifically between three locations within the motion segment: the right facet, left facet, and/or the intervertebral disc. These locations are demonstrated in figure #1.
Figure #1 is a sagittal view (view from the side) of a motion segment (the L4/L5 motion segment).
The red and green circles indicate the three pivot points where all spine motion occurs.
Note that the facet joints are overlapped in this view, so these posterior pivot points cannot be individually ascertained (I put a green circle to indicate the left facet joint that is blocked by the right facet joint).
The goal of most fusion procedures is to "glue" these structures/pivot points together by removing the cartilage/disc, roughing up the bone, and then putting in the glue (the glue, of course, is ground-up bone) which will solidify (ossify or fuse) with the passage of time.
How does fusion work? fusion is designed to stop all the pain-causing-movement within a motion segment by "cementing" the adjacent vertebrae together, typically at and around the three pivot points as well as the adjacent transverse processes and spinous processes. Such permanent/successful fusion of the motion segment correlates with better clinical outcomes . Of course, surgeons don't actually use the kind of cement that is used for sidewalks or driveways; instead, ground-up human bone, which is given the name morselized bone, may be used. Or, the surgeon may use prefabricated bone in the form of plugs, strips, or just about any shape you can imagine.
Where does this human bone come from? It either comes from you (this is called autograft or autogenous bone graft) or from a donor (this is a polite way to say cadaver bone, which is called allograft or allogeneic bone graft)--the former methodology results in less complications and no chance of disease transmission [12, 13]. More specifically, in 1994, Jorgensen et al.  published a one-of-a-kind study that compared autograft to allograft in the same group of patients and discovered that the autograft was much better at creating a successful fusion. What they did is take 144 patients, all in whom suffered chronic intractable low back pain secondary to various degenerative disorders of the lumbar spine, and fuse them with a technique called posterolateral fusion (PLF). This procedure fuses the left and the adjacent right transverse processes. So what they did is fuse the left side with The various flavors of allograft and the right side with autograft which was taken from the patient's hip (iliac crest autograft) and then compared the two sides at the end of the follow-up interval (which was between 14-27 months). After analysis of the patients at follow-up, which included x-rays, there was no comparison: the side that had been fused with iliac crest autograft, was far more successful when compared to the side that was fused with the different flavors of allograft.
What is the surgical/technical goal of fusion? the goal of fusion is to oossify or to fuse the motion segment into one solid unit by using things such as intertransverse bone grafting, transpedicular instrumentation, interbody grafting and grafting of bone between adjacent lamina and facets. In other words, the goal is to fuse adjacent vertebrae together so that no motion at all can occur. Such successful arthrodesis has been demonstrated to be directly related to reduction in patient pain and an increase in patient function . However, it should be noted that there are papers demonstrating that the clinical outcome and patient satisfaction of fusion is not necessarily related to a successful fusion as some patients achieve an excellent result without solid fusion and some patients achieve a poor result with solid fusion [fn].
Who is the ideal candidate for fusion? there is an algorithm that must be strictly applied in order to ensure that the patient is a proper candidate for fusion:
#1: Exhaust All Other Treatment Options: because of the high complication rate and less than stellar outcome rates, fusion is a last-ditch effort to bring the patient back from a near total disability, up to a state where he/she can function at a higher level, such as return to some type of employment, and have some form of social and recreational life. Therefore, all of the less invasive treatments must be tried first, which include (1) all forms of conservative care (rest, bracing, exercise, physical therapy, gentle chiropractic care, acupuncture, and massage); (2) injective procedures (facet injections, rhizotomies, sacroiliac injections, intradiscal injections); (3) medication (ibuprofen, acetaminophen, Celebrex, Vicodin, Percocet, OxyContin, MS Contin, Lyrica/ Neurontin; and (4) perhaps the fringe treatments (prolotherapy, gel platelet therapy, SED).
#2: Prove It: There are ancillary diagnostic techniques that often must be completed in order to ensure that the patient is a proper candidate for fusion. These techniques all involve the use of an anesthetic, which is injected in and around the suspected pain generator. For example, if the facet joint is suspected to be the chronic pain generator, then by injecting it with an anesthetic (lidocaine), the patient should afford a few hours of marked pain relief. If so, then there is proof that the facet joint is a pain generator and (assuming facet blocks and denervation procedures have failed) the proposed fusion of that joint should eliminate the pain. Another example: Let's say the doctor suspects that there is stenosis within the L4/5 lateral recess and canal, which in turn is causing pain. Then, selective nerve root blocks (a.k.a. transforaminal epidural steroid injections) should be performed by a skilled pain management doctor. Again, if the patient affords pain relief for first few hours after the procedure, then it is logical to conclude that the stenotic compression of the thecal sac and exiting nerve root is in fact the pain generator for which decompression and fusion should help. Another example: for suspected disc related problems (i.e. symptomatic DDD, IDD, and/or symptomatic annular tear(s)), the "prove it" story is a little different. We use a technique called provocation discography to confirm that the disc is in fact the pain generator. I won't go into all the details [learn about it here], but know that it involves injecting contrast into the nucleus of the disc under fairly high pressure (50 PSI) and observing the effect. Ideally the patient should scream in pain and that pain should be his or her usual pain (that is called concordant pain). If that is the case (there is more to this testing), then there is evidence that the patient is suffering from discogenic pain and the planned fusion should diminish the pain.
#3: You Better Be Really Hurting: many surgeons are going to disagree with this criteria, but this is what I believe: Just because your back pain is preventing you from participating in some of the more arduous recreational activities (golf, skiing/snowboarding, snowmobiling, softball, bowling, etc.), that does not mean you are a candidate for fusion. YOU MUST BE REALLY SUFFERING to be a candidate. More specifically, you should have trouble sleeping, going out socially, working and performing any type of recreational activity. In other words, I would expect your Oswestry Disability Index (ODI) to be well above 50 and your visual analog scale (VAS) for back pain to be above five most of the time (even with medication). So don't be lured into fusion by some unscrupulous surgeon if your current Oswestry score is 20! Heck, 20 is already an outcome that is considered a fusion success , so the chances of affording further relief are not that great. A patient with a 20 ODI can live quite comfortably with pain medication and modification of activity.
The $64,000 question: What are the chances for "success"?
The answer to this question all depends how you define "success" and which of the thousands of published investigations (a.k.a. studies or papers) on fusion you want to believe.
I believe the best way to gauge the success of fusion (or of any treatment intervention for that matter) is to review a special type of study called a meta-analysis.
A meta-analysis is constructed by a research team by searching all the medical literature databases and then pulling all the papers that were published on the topic of interest – fusion in this case. The next step is the weeding out process. More specifically, the researchers will throw out all low quality studies and only use ones that meet certain gold standard criteria such as the AHRQ criteria, the Downs criteria or the Cochrane review criteria. Once the cream of the crop is collected, then you can "pool" all the data together in order to make one giant, statistically powerful study; i.e., the meta-analysis or the study of all the quality studies.
While there are several different meta-analysis studies on the subject of fusion, I'll take the most recent  to answer our question of whether or not fusion works and what are the success rates.
In 2010, Yajun et al.  published the results of their meta-analysis on the subject of lumbar fusion. After the weeding out process, the team pooled data from 770 patients, all in whom had undergone fusion (a variety of techniques were utilized from artificial disc replacement to ALIF to the posterior fusion procedures) and were available to report their outcome at a two-year time point. More specifically, this patient data came from four high-quality studies, which were all randomized controlled trials-- the pinnacle in the hierarchy of research quality-- which compared ADR to traditional fusion techniques of all flavors. After a painstaking statistical analysis of the data, the team made, among others, these conclusions:
At the two year time point (i.e., two years following the surgery), the average Oswestry score from those 770 patients was 32.2. This indicates that the typical functional outcome (i.e., how well the patient is functioning with regard to the activities of daily living as well as work function) was not so good. More specifically, these patients continued to suffer "moderate disability,” fusion surgeries notwithstanding. Here's the official description of what such a ODI score means:
21%-40%: moderate disability: The patient experiences more pain and difficulty with sitting, lifting and standing. Travel and social life are more difficult and they may be disabled from work. Personal care sexual activity and sleeping are not grossly affected and the patient can usually be managed by conservative means.
Visual Analog Scale Scores:
At the two year time point, the average VAS score was 3.4 on a scale of 0 to 10, where 0 indicates no pain at all and 10 indicates severe in-the-hospital type pain. Therefore, it would seem that the patients were not suffering too much pain, but were more functionally disabled.
So the bottom line of the question, "what are my chances for success?” seems to be this: the chances are fairly good for partial recovery a function and a good drop in the level of pain. That is, with the passage of time the amount of usual pain will be diminished to a point it can be managed with a simple pain medication and (importantly) the modification of activity. However, the average fusion patient continues to be moderately disabled to the point where arduous activities (sports or work) will be out of the question. Not the rosiest of outcomes if you ask me; however, probably much better than the alternative of suffering a great amount of average pain and functional disability.
The mean of the means is a powerful statistical number and gives us a good idea of what outcome can be expected from a procedure.
In this case, we learned that regardless of the type of fusion, there was no statistical difference between ADR and conventional fusion at the two-year or five-year time-point .
The Flavors Of Fusion: PLF, ALIF, PLIF, and TLIF.
What's the quickest way to start a brawl in a room full of spinal surgeons? Ask them "what is the best fusion technique for correcting spondylolisthesis?" Spinal surgeons are strongly opinionated when it comes to their craft and rarely agree that one technique is superior to another – theirs is always the correct way.
So, let us talk about the most common "flavors" of fusion. In order to unify the techniques, I will describe each one with regard to a very generic two-level lumbar fusion at the L4/5 motion segment.
#1) PLF: posterolateral fusion is the grandfather of fusion technique as it was developed just over 100 years ago. Although PLF can and does stand alone as a treatment intervention, it is often used in conjunction with the other flavors of fusion in order to afford greater stability and durability to the procedure. One of the criticisms of PLF is that it involves an extensive dissection (the stripping of muscle and fascia off of bone) of the adjacent transverse processes, facet(s) and sometimes lamina. However, this exposure is necessary so that the structures can be decorticated (have their outer layer ground off) in preparation for their cementing together with ground-up human bone (morselized bone). Also noteworthy is the fact that although the procedure can be supplemented with instrumentation (pedicle screws, fasteners and rods), it does not employ any intervertebral disc substitutes such as a cage or graft.
Figure #2: you are looking at a from-the-behind view (PA) of the lower back. The skin and muscle (pink dot) have been cut and pulled back exposing the bottom three vertebrae (L5, L4, and L3). Morselizel bone (green dot) has been added between the ground-down spinous processes, which will (with the passage of time) harden (ossifiy) into a solid mass (fusion). On the right, the transverse processes have been burred down (orange dot) in order to prepare them for the morselized bone phase of the PLF. The linear structure (blue dots) represents the supraspinous ligament that is covering the spinous processes.
#2) PLIF: in 1953, Cloward published a seminal paper that described a new methodology for the fusion of a motion segment, which was termed posterior lumbar interbody fusion (PLIF). PLIF is quite an invasive procedure as it necessitates a bilateral laminectomy as well as a partial bilateral facetectomy (basically the entire posterior arch of the vertebrae is moved). Then, in order to approach the disc, the entire thecal sac and exiting nerve root must be retracted for a prolonged period of time (this is one of the pitfalls of this procedure, for this retraction time can result in a fairly high rate of permanent nerve damage). Once the disc is removed, and interbody graft or cage (which is filled with autologous or allogenic bone) is inserted in its place and transpedicular instrumentation is attached to stabilize the motion segment while fusion occurs. Sometime, PLIF is augmented by PLF.
||#3) TLIF: The newest mainstream posterior procedure is called transforaminal lumbar interbody fusion (TLIF). This procedure, first reported on by Harms and Jeszenszky in 1998 , is arguably an important improvement on traditional PLIF, because it minimizes nerve root and thecal sac retraction/damage [8,9] and necessitates less osseous and soft tissue dissection. This is because the technique approaches the epidural space from a more posterolateral direction, taking out the facets on one side and only part of the lamina. As with TLIF, PLIF removes most of the intervertebral disc in order to make room for the insertion of a graft or cage which acts as a replacement. Transpedicular posterior instrumentation as well as intertransverse posterolateral fusion augments this procedure. The pitfalls of this technique are (1) there is a steep learning curve and (2) the workspace for the surgeon is limited compared to traditional PLIF, which removes both lamina and the spinous process.
#4) ALIF: There is an approach from the front (anterior/through the belly) that spares contact with the delicate posterior neural structures (i.e., the cauda equina, nerve roots, spinal nerve and dorsal root ganglia) altogether, which is called anterior lumbar interbody fusion (ALIF). While the risk of retractor damage to the thecal sac and traversing nerve roots is eliminated with ALIF, the technique is limited in that it cannot decompress the epidural space, so it cannot be used for stenosis (a very common condition), foraminal collapse, large disc herniation, recurrent disc herniation, or far lateral disc herniation. Furthermore, well reported on complications include male sterility (via retrograde ejaculation from damaged nerve fiber) and damage to the important autonomic nervous system (the sympathetic chain ganglia). Because of its limitations, it is often combined with posterior approaches.
#5) Honorable Mention: in 2004 the FDA approved the Charite™ lumbar artificial disc replacement (ADR or total disc replacement or TDR) for the treatment of symptomatic degenerative spine disease (DSD), which includes discogenic pain and facet syndrome; however, this methodology will not be discussed on this page.
Instrumentation and Fixation Systems: every one of the four flavors of fusion can be accompanied by a system of metal (pedicle screws, plates, cross bridges, fasteners, and rods) designed to immediately stabilize the vertebrae involved in the fusion. Typically, in the modern world, what is done is this: holes are carefully drilled in each of the four pedicles (remember in this paper we are talking about a L4/L5 fusion or a two-level fusion). Then titanium pedicle screws are screwed deep into the holes. Special fasteners are attached to the pedicle screws which support a lordotic shaped titanium rod on each side (kind of looks like vertical rails of a railroad tract). Once the system (this is called a "construct") is in the proper alignment, it is carefully tighten down. This tightening creates an immediate fusion as well as places the vertebrae and a slight extension which ensures the natural lordotic curve of the lumbar spine is maintained.
Grafts and Cages: another accessory procedure is the placement of a graft or cage between adjacent vertebra—this will take the place of the disc. In fact, these interbody devices are almost always used in this day and age with the PLIF, TLIF and ALIF procedures. The PLF procedure, however, does not involve the placement of an interbody device. This is a complicated subject and we will discuss it later on.
THE HISTORY OF FUSION: Draft.
1) Nachemson AF, Schultz AB, Berkson MH. Mechanical properties of human lumbar spine motion segments; influences of age, sex, distal level, and degeneration. Spine 1997;4(1):1-8.
2) Drum D. the vertebral motor unit and intervertebral foramen. In: The research stat of spinal manipulation therapy, M. Goldstein, Ed. US Department health education and welfare, HIH Bethesda, MD (1975).
3) Djurasovic M, Glassman SD, Dimar II Jr, et al. Does fusions status correlate with patient outcomes in lumbar spinal fusion? Spine 2011;36:404-409.
4) Deyo RA, Nachemson A, Mirza SK. Spinal-fusion surgery: the case for restraint. N Engl J Med 2004; 350:722-726.
5) Vamvanij V, Fredrickson BE, Thorpe JM, et al. Surgical treatment of intervertebral this disruption. J Spinal Discord 1998; 11:375-382. *
6) Bono CM, Lee CK. The influence of subdiagnosis on radiographic and clinical outcomes after lumbar fusion for Degenerative Disc Disorders: an analysis of literature from two decades. spine 2005; 30 (2): 227-234.
7) Harms JG, Jeszenszky D. Die posteriore, lumbale, interkorporelle Fusion in unilateraler transforaminaler Technik. Orthop Traumatol 1998; 10:90-102.
8) Fraser RD: Interbody, posterior, and combined lumbar fusion. Spine 1995; 20 (24 Suppl):167S-177S.
9) Stonecipher T, Wright S: Posterior lumbar interbody fusion with facet-screw fixation. Spine 1989; 14:468-471.
10) Yajun W, Yue Z, Xiuxin H, Cui C. A meta-analysis of artificial total disc replacement versus fusion for lumbar degenerative disc disease. Eur Spine J 2010; 19:1250-1261.
11) Glassman SD, Carreon LY, Djurasovic M, et al. "Lumbar Fusion Outcomes Stratified by Specific Diagnostic Indication: 2008 Outstanding Paper Award Runner up" Spine J 2009;9: 13-21.
12) Jorgenson SS, Lowe TG, France J, Sabin J. "A Prospective Analysis of Autograft Versus Allograft and Posterolateral Lumbar Fusion in the Same Patient. Spine 1994; 19 (18): 2048-2053.
13) Glassman SD, Howard JM, Sweet A, Carreon LY. "Complications and concerns with osteobiologics for spine fusion in clinical practice. Spine 2010; 35:1621-1628.