The Truth About Facet Cysts

One very painful degenerative condition of the spine that I commonly see in my clinic is the facet cyst.   These are also referred to as synovial cysts as they originate from the synovium (lubricating tissue that lines the opposing surfaces of a joint) of the spinal facet joints.  While facet cysts theoretically can originate at any level of the spine, they most commonly occur in the lower lumbar spine (i.e. L4/5 and L5/S1.)   These cysts typically grow from the medial (inner) aspect of the joint were they then begin to compress the traversing nerve root as it passes nearby (see figure 1.)  This compression can be quite severe and, when coupled with the movement of the joint, can lead to excruciating pain (at the time of surgery I’ll often see a dent on the nerve root from the severe pressure of the cyst.)  After far lateral disc herniations, this is probably the most painful non-traumatic pathology that I encounter in my clinic. 

 Sagittal facet cystAxial facet cyst

Figure 1: Sagittal (left) and axial MRI showing large facet cyst (red arrows) emanating from the left L4/5 facet joint.  Note how the cyst originates from the medial (inner) aspect of the joint and severely narrows the right side of the spinal canal.)  

First and foremost, facet cysts are not cancerous. Many patients equate the term “cyst” with cancer so the first thing they want to know is whether or not they have cancer in their spine (reason #146 NOT to read your MRI reports because those terms you don’t understand will FREAK YOU OUT unnecessarily.)   Facet cysts are benign and simply indicate advanced arthritis of the joint from which they emanate.

The treatment of facet cysts is a bit controversial and can provoke colorful discussions at meetings of spine surgeons.  Most of the controversy stems from the question of how aggressive to be with initial surgical treatment of the cysts.  More on that later.   For now, let’s review the treatment options of facet cysts here:

  1. Watchful waiting.  This usually isn’t a good option for patients.  First of all, most patients with this condition are in so much pain when they walk in that they want surgery yesterday.  If I even begin to discuss waiting another 6-8 weeks for spontaneous resolution of their pain they look at me like I’m crazy and immediately begin looking for another doctor to provide another opinion.  The problem with watchful waiting is that unlike a herniated disc fragment which can be reabsorbed by the body, facet cysts typically don’t involute spontaneously. To be fair, though, I have seen cases where the compressed nerve becomes accustomed to the irritation and becomes less painful over time without intervention.  Usually, however, an invasive treatment is needed to take care of the problem.
  2. Interventional pain management. Here’s one of those areas of controversy; even my partner and I don’t always agree on the utility of this treatment for facet cysts.  The thought is that a pain management physician can guide a needle to the cyst using fluoroscopic guidance and then aspirate the fluid within the cyst to decompress it and thus take the pressure off the adjacent nerve.  There are two problems with this approach in my opinion.  First, if you were to come into the OR and watch me resect one of these cysts you’d see that 90% of the time the cyst is filled with a thick gelatinous substance rather than a thin fluid that can be aspirated.  It’s easy to see how this needle aspiration could fail.  Second, cysts that are simply aspirated will often recur.  Without removing the cyst wall the underlying structure of the cyst remains, only to fill up with fluid again later.  Now, I never discourage a patient from trying cyst aspiration or an epidural steroid injection prior to committing to surgery.  I certainly have seen it work.  It’s just important to understand that aspirating the cyst either isn’t possible or doesn’t permanently ablate the cyst so often any symptom improvement achieved with interventional pain management is short lived.
  3. Surgery.  When watchful waiting and/or interventional pain management treatments have been tried unsuccessfully (or if the patient is just too miserable), surgery should be considered.  There is some debate on how aggressive that surgical treatment should be right off the bat, though.  Some surgeons will jump straight to recommending a fusion for a patient with a facet cyst.  The thought is that the presence of a facet cyst suggests that the facet joint is structurally incompetent and that without a fusion these cysts will only recur.  There is some truth to this.  I quote up to a 20-30% rate of recurrence for facet cysts without fusion.  Typically, though, I like to avoid a fusion initially if possible.  Thus, my algorithm for the surgical treatment of facet cysts is as follows (see flow chart):
      1. MRI shows facet cyst that correlates with patient’s pain.  Patient has failed conservative management.
      2. If the facet cyst occurs in the setting of a spondylolisthesis (which they often do) I’ll usually offer a fusion right away (XLIF at L4/5 and above, ALIF at L5/S1.) 
      3. If there is no concurrent spondylolisthesis I’ll check flexion-extension Xrays to evaluate for instability (I may also just do this for the patient WITH spondylolisthesis if for some reason I’m trying to spare them a fusion.)  If the motion segment in question is unstable then the patient gets a fusion. If there is no instability I’ll do a minimally-invasive laminectomy to resect the cyst.
      4. If the cyst recurs after laminectomy the patient gets a fusion. 
Facet cyst algorigth

Lastly, and here’s where it gets really controversial, whenever I do an XLIF or ALIF on a patient with a facet cyst I no longer directly resect the facet cyst.  I rely completely on indirect decompression to take care of the cyst.  By using the appropriately sized spacer to restore disc height and correct spondylolisthesis, the facet cyst is essentially stretched out so that no longer compresses the adjacent nerve (see figure 2.)  It’s also believed that by eliminating the motion at the facet joint with a fusion the facet cyst can then spontaneously resolve.  It’s like magic.  This always makes the patient nervous though.  They always want to know why I’m not cutting that painful cyst out to be sure it’s off the nerve.  What I tell them is that I know, after looking the data on all my patients, that in just about every case the height restoration provided by the spacer is enough to decompress the nerve (there’s less than a 5% chance of failing indirect decompression in our experience.)  Also, these cysts are usually densely adherent to the underlying nerve and dura.  Dissecting these cysts off of the dura can be quite treacherous and the rate of dural tear and cerebrospinal fluid (CSF) leak in these cases is not insignificant (5-10% in our hands).  Thus, why would I subject the patient to the risk of CSF leak (not to mention the risk of spending the extra time under anesthesia that it takes to do the laminectomy) that is greater than the risk of them failing indirect decompression?  As we’ve discussed previously: believe in indirect decompression!  


Figure 2: Preoperative (left) and postoperative MRI images showing a patient with a left L4/5 facet cyst (red arrow) and spondylolisthesis that was treated with XLIF and percutaneous pedicle screw fixation (NO direct cyst resection.)  Patient’s leg pain was relieved immediately after surgery.  Note on the postop MRI (taken about 3 months after surgery) that the facet cyst has disappeared.

Thanks for reading!

J. Alex Thomas, M.D.

XLIF: explained

In this post I’m going to give you a step-by-step description of the extreme lateral interbody fusion (XLIF) procedure from patient positioning to skin closure.  While I will occasionally do XLIF as a standalone procedure (when only a spacer is inserted without any posterior instrumentation inserted at the back of the spine) I almost always insert pedicle screws after the XLIF is complete.  The information about XLIF is quite detailed and thus the post gets quite long so I’ll explain the pedicle screw portion of the case in a future post. Keep in mind that this is how I do XLIF in my OR for a standard one-level case.  While the general steps of the procedure are the same no matter where you get your XLIF, some variation may occur so if your surgeon does it a bit differently that doesn’t mean he’s doing it incorrectly. My hope is that this explanation will be helpful to you if you’re considering the procedure or if you’ve already decided to undergo XLIF and want to know what exactly is going to happen to you while you’re asleep.  This is a long, detailed post so hang in there.

Step 1: Induction of anesthesia and placement of neuromonitoring leads.  After you’re brought into the OR you’ll slide off your bed and over to the OR table.  The people who typically will be in the room with you will be myself, my physician assistant Jack, an OR nurse, a surgical technician (the person who handles all of the surgical instruments), the anesthesiologist or nurse anesthetist (CRNA), an X-ray technician (to run the fluoroscopy machine called a C-arm) and a representative from the company who manufactures the spacers and screws that I’ll implant in your spine.  Once you’re on the OR table you’ll be put to sleep and then intubated (when a breathing tube is inserted into your windpipe so that a machine can breathe for your during surgery.)  An extra IV and a Foley catheter (inserted into your bladder to collect urine) may be inserted as well.   After all of this is done we will then insert small needle electrodes into the major muscles of your legs (see figure 1.)  These electrodes are then connected to a computer system that then monitors the status of the nerve signals to these muscles during surgery.  As we’ll discuss in a later post, one of the major complications that may occur during XLIF is an injury to one of the nerves of the lumbar plexus that provides motor and sensory function to your legs.  These electrodes allow me to monitor the function of these muscles, and thus the nerves of the lumbar plexus, so that I know I’m not injuring any of them.  More on this later.  (Some patients occasionally ask me why there are little blood spots on their legs after surgery…the placement of these electrodes is why.)

IMG 4033

Figure 1: small needle electrodes are inserted into the major muscle groups of the legs to allow for monitoring of the nerves of the lumbar plexus.

Step 2: Patient positioning and taping.  Now that you’re all hooked up to the anesthesia machine and the neuromonitoring system we’ll put you into the position that you’ll be in for the entirety of the case.  Several of us will all grab hold of the sheet you’re laying on and then will put you into the lateral decubitus position which basically is how you’d look if you were asleep on your side on a park bench.  Unless there’s something unusual about your anatomy I’ll position you RIGHT side up so that your XLIF incision will be on your RIGHT flank (regardless of which side your leg pain is worse on.)  XLIF surgeons will debate about which side up is best for the procedure.  In my hands I think that right side up is safer (honestly, because that’s what I’m familiar with after hundreds of XLIFs) but a lot of surgeons go left side up because that’s the way that XLIF is classically taught.   In my opinion it really doesn’t matter which side is up as long as it’s the side that your surgeon is comfortable with.  Even though you’re on a heavily padded OR table we’ll pad critical pressure points (like your armpit and knees) so that you don’t get pressure injuries during surgery.  We’ll then flex your knees and hips slightly (to relax the psoas muscle and the nerves of the lumbar plexus within so that they’re easier to navigate during the docking of the retractor.  See below.) and then secure you to the table with tape.  It’s lot of tape actually (see figure 2.)  It’s critical that I maintain a perfectly perpendicular trajectory to the side of spine during the XLIF procedure.  If you’re not secured to the table well enough you may slowly roll to one side or the other which may put you at risk when suddenly you’re not in the position that I expect you to be in. The multiple passes of tape across your hips, legs and chest prevent this rolling.  Of course we’ll also confirm throughout the case that you’re still positioned correctly by checking an image with the C-arm machine. 

IMG 3947\IMG 3959Figure 2: after positioning the patient with the right side up, we then secure the patient to the OR table with multiple passes of 3-inch tape.  Lots of tape.  

Step 3: X-ray confirmation and patient marking. Now that you’re positioned on the OR table with your left side down and your right side up I’ll then obtain an image of your spine using the C-arm machine and then plan an incision on your right flank centered over the level of the spine that we’re treating.  First we’ll shoot an AP X-ray shot (front to back) with the C-arm to confirm that your spine is not rotated at all.  Again I want to be certain that I’m approaching your spine at a 90-degree angle (directly lateral) so we’ll correct any rotation we see on that first X-ray.  Once we’ve ensured that you’re not rotated I’ll then bring the C-arm to a lateral view (side view) to localize the correct level where we’re working and then mark the boundaries of the disc space.  If we’re working at L4/5 for example (where I do in the vast majority of cases) I’ll mark the endplates of the L4 and L5 vertebral bodies as well as the front and back boundaries of the disc space.  I’ll then plan a 3cm incision centered over this marked disc space (see figure 3.)  I’ll then box in the area of the incision, as well as an area over your lumbar area where I’ll eventually make small incisions for the screws, with some preliminary sterile sticky dressings called “10-10s” (see figure 4.)

IMG 3969

Figure 3: Marking the boundaries of the disc space at L4/5.  Note the C-arm monitor in the background. 

IMG 3973

Figure 4: The boundaries of the disc space between the L4 and L5 vertebral bodies.  I’ll typically center my incision right over the disc space. 

Step 4: Patient prepping, draping and time out.  After you’re positioned and we’ve planned our incisions the OR nurse will then use iodine solutions to “scrub and paint” any areas of skin boxed in by the 10-10s (see figure 5.)  During this time I’m scrubbing my own hands and will come in to get my surgical gown and gloves on.  After this Jack (my PA) and I will cover you in drapes so that only your flank and lumbar areas are exposed.  I’m draping not just for the XLIF but also for the placement of pedicle screws (see figure 6.)  At this time we’ll also check to be certain that you have “4 twitches” which indicates how well your muscles respond to stimulus.  We don’t want you to have any muscle relaxing agent on board here so that it doesn’t interfere with the accuracy of the neuromonitoring system.  After all this is done and prior to making an incision we’ll do the OR time out where the entire team confirms that we’ve got everything in the OR ready to do the case correctly (including confirming the identity of the patient as well as the correct level/side of the procedure.). You can never be too careful.

IMG 3976

Figure 5: the OR nurse does a “scrub and paint” prep to remove any skin bacteria or other contaminants. 

IMG 3981

Figure 6: all set up and waiting for the OR time out.

Step 5: Incision and exposure.  After the time out we’re ready to start the case.  The flank incision is made first with the scalpel and then carried down through the underlying fat with Bovie cautery.  I’ll come down with the Bovie until I arrive at the external layer of fascia enveloping the muscular layers of the abdominal wall (the three layers are: external oblique, internal oblique and transversalis muscles, see figure 7). I carefully incise the external fascial layer and then use the scissors to bluntly dissect (not cut) through the muscular walls until I can pop through the inner layer of fascia and into the retroperitoneal space.  This is the space behind the cavity containing your abdominal contents.  I’ll confirm that I’m within the retroperitoneum by palpating the surface of the psoas muscle-it has a very distinctive feel as it rolls under my fingertip. Once I’m certain I’m where I’m supposed to be, in the retroperitoneum and on the surface of the psoas, I’ll proceed with traversing the psoas.

IMG 8828

Figure 7: right flank incision showing the subcutaneous fat, the external fascia of the muscular abdominal wall and the oblique muscles within. 

Step 6: Dilation of psoas muscle and docking of XLIF retractor. Once I can feel the surface of the psoas muscle I’ll take the first of a series of 3 tubular dilators and place it on the surface of the muscle at the level of the spine that I’m targeting (see figure 8.)  Remember those monitoring leads we inserted into the leg muscles prior to the case?  This is where I’m going to need them.  The dilators are electrically stimulated (via a clip at the top) in such a way that they emit a small directional electrical charge.  Thus, when I pass the dilator through the psoas en route to the side of the disc space I will rotate the dilator in order to send out a directional electrical charge to each quadrant of the working area to hunt for the nerves of the lumbar plexus.  If the tip of the dilator is close to a nerve it will stimulate the nerve at a certain threshold and fire the muscles in the leg innervated by that nerve.    My goal is to land on the disc space in front of the nerves of the lumbar plexus so I only want to see stimulation when I’m stimulating towards the back of the patient (see figure 9.)  If I see stimulation when I have the tip of the dilator aiming towards the front of the patient then I know I’m not in the right place.  I know this is complicated and probably more than you need to know for your XLIF.  In my opinion, though, this process of traversing the psoas muscle with the dilators and using them to locate the nerves of the lumbar plexus is the crux of the procedure.  If done properly this makes XLIF one of the safest fusion procedures around and reduces your chance of a nerve injury to near zero. 

Once I’m happy with where I’ve placed the first dilator I’ll insert a K-wire through the dilator and into the disc space to anchor it in place (see figure 10.)  I’ll then use sequentially larger tubular dilators, also stimulated, to gently dilate the psoas muscle.  Once the dilation is complete I’ll insert the specialized MaXcess retractor over the dilators and attach it to a special arm that’s attached to the bed.  I remove the dilators and then insert light sources to illuminate the working area.  Lastly, I’ll use a handheld nerve stimulator to be sure there are no nerves traversing the working area that were somehow missed during the initial dilation.  Once I know the area is free of nerves I’ll insert a shim through the back retractor blade and into the disc space.  This secures the retractor to the disc space so that it won’t shift during the procedure (and also prevents any nerves of the lumbar plexus from creeping into the working area.)  Now the retractor is anchored to the spine, the working area is illuminated and free of any nerves and I’m ready to go (see figure 11.)  Of note, the back blade of the retractor (which should have the nerves of the plexus safely behind it) can also be stimulated.  Throughout the remainder of the case I’ll periodically stimulate this blade to check the health of the nerves behind it.  If it’s taking more and more stimulation to fire the nerve I can assume that the nerve is at risk of injury and I’ll take steps to mitigate these risks (like move more quickly or reposition the retractor.)  The point is that I’m constantly thinking about the nerves of the lumbar plexus for the entire time that the retractor is in place within the psoas muscle. 

IMG 3990Https static1 squarespace com static 54908575e4b06ab6fc17a15c t 5b22831f575d1f5fd7345d34 1528988450887 Designed NVM5 pdf

Figure 8: image on left shows insertion of the first of a series of 3 tubular dilators.  Note the cable attached to the dilator.  The tip of the dilator emits a directional electric charge that allows me to search for the nerves of the lumbar plexus.  Image on right shows the tip of the retractor on the side of the spine in front of the nerves of the lumbar plexus.  As the tip gets closer to a motor nerve the monitoring system will turn from green to yellow to red to indicate the proximity of the dilator to the nerve.  

Mendeley Desktop

Figure 9: the nerves of the lumbar plexus shown on the left side of the spine (this patient, unlike the patient in our images, is LEFT side up with the front of the spine at the top of the image and the head of the patient to the right.)  The goal is to dock IN FRONT of the motor nerves of the plexus, particularly the femoral nerve (see the small red circle on the L4/5 disc space.)  Note that the sensory nerves of the plexus, the genitofemoral nerve, the iliohypogastric nerve and the ilioinguinal nerve run on the surface of the psoas muscle or in the soft tissue of the retroperitoneal space and thus aren’t usually at risk while docking on the side of the disc space. (Source: Uribe et al, 2010)

Viewer v5 4 202 Internet Explorer

Figure 10: C-arm image showing the first dilator in place on the center of the L4/5 disc space.  At about this “50-yard line” of the disc space I’m usually far enough forward to be in front of the lumbar plexus.  It’s tough to see but a K-wire has also been inserted into the disc to hold the dilator in place.

Viewer v5 4 202 Internet ExplorerLateral Transpsoas Retractor Technology Neupsy Key

Figure 11: image on left shows C-arm image showing retractor docked over the L4/5 disc space.  A shim is inserted into the disc space to anchor the retractor in place and prevent any migration of the nerves of the lumbar plexus into the working area. Image on right is an illustration of what is seen on the C-arm image (courtesy of Nuvasive.)

Step 7: Discectomy and preparation of disc space.  Now I can look down the retractor at the side of the disc space.  I use surgical loupes to magnify the working area at the base of the retractor (see figure 12.)  I know that the area is free of any nerves so I can start to remove the disc without fear of injury to the nerves of the plexus.  I cut out the annulus of the disc space and then use a “box-cutter” to traverse the disc space and release the contralateral annulus (see figure 13.)  This contralateral release is key to getting maximum height restoration with your spacer later. I’ll then use curettes and other scrapers to clean all of the disc material out of the disc space and off of the vertebral endplates.  This is another key step.  If the endplates aren’t sufficiently debrided of disc material then new bony growth into the spacer and bone graft won’t occur.  Bad disc preparation=no fusion! 

IMG 4012

Figure 12: XLIF is done through a narrow, minimally-invasive corridor so I have to use surgical loupes to magnify the working area at the base of retractor.  

Viewer v5 4 202 Internet Explorer

Figure 13: A “box-cutter” is passed from right-to-left (left-to-right on this AP, or front-view, C-arm image) through the disc space to remove the bulk of the disc material and also to release the contralateral annulus of the disc space (notice how the cutter is protruding just a bit outside the left side of the disc space.)  This contralateral release is critical for getting adequate height restoration from the inserted spacer.

Step 8: Sizing of disc space and insertion of spacer. Ok so now the disc space is completely cleaned out.  I’m ready to insert the properly sized spacer (also known as a cage.)  We will first insert a trial to make sure the spacer is going to sit in the proper location within the disc space (see figure 14.)  I’ve already done measurements on preoperative images so I usually already know what size spacer I’m going to need but if there’s any question the appearance of the trial within the disc space (on the fluoroscopic image) will help me select the spacer size.  Once we’ve decided on the correct spacer height, length, width and lordosis (angulation) we’ll pack it with graft material and get ready to insert it into the prepared disc space.  I could spend an entire post talking about various graft materials.  Usually I use a product called Osteocel which is basically cadaveric bone fragments prepared in a way such that it contains stem cells to promote bone growth (see figure 15.)  I’ll then attach the spacer to an inserter and gently tap it into the disc space with a mallet.  I’ll check the final positioning of the spacer with front and side view fluoroscopic images (see figure 16.) If I’m happy with the placement of the spacer I’ll irrigate, look for any bleeding and then collapse and remove the retractor.  Generally the entire process, from initial docking of the retractor to the time I collapse and remove the retractor takes 10-15 minutes in my hands (see figure 17.)  There is very clear evidence in the literature that suggests that the longer that retractor is place in the psoas the higher the risk of injury to the lumbar plexus.  Speed matters in XLIF! 

Viewer v5 4 202 Internet Explorer

Figure 14: a trial spacer is inserted into the disc space to ensure  a) that the spacer will sit in a good position in the disc space, and b) that we’ve selected the spacer with the appropriate height, width, length and lordosis (usually we do this in advance based on the patient’s MRI but we double check here.)

IMG 3878

Figure 15: a PEEK spacer loaded with Osteocel and ready for insertion. PEEK isn’t visible on X-ray so small metal markers are embedded in the spacer to allow for visualization (notice the small metal bumps on the surface of the spacer.)

Viewer v5 4 202 Internet Explorer

Figure 16: AP C-arm image showing spacer in place, still attached to inserter.  Note the metal markers (black lines) that indicate the outer boundaries as well as the center of the spacer.

IMG 3880

Figure 17: intraoperative view looking down the retractor at spacer in its final position within the disc space.

Step 9: Drain placement and closure. After I remove the retractor I place a small drain into the psoas muscle where I was working.  This is an unusual step that not many surgeons do.  One minor complication that can occur after XLIF is numbness, tingling and even burning on the front of the thigh on the side of the access for XLIF.  Usually this is minor but in some cases it can be quite bothersome and can persist for several months.  This complication usually occurs as a result of stretching of one of the sensory nerves of the lumbar plexus.  When we looked at our data for 50 patients who didn’t have a drain placed versus 50 patients who did have a drain placed, we found that the patients with a drain had a 10x reduction in the incidence of postoperative thigh sensory disturbances (from 40% to 4%!)  Ever since we discovered this in our data we’ve always used drains in the psoas after XLIF.  After the drain is placed (which is removed the next day before the patient goes home) we’ll close the fascia of the muscular abdominal wall and then the dermis (the layer just below the skin) with absorbable sutures.  The skin is closed with small adhesive strips (see figure 18.) 

IMG 4027

Figure 18: we insert a small Blake drain into the psoas muscle where we were working to drain any residual blood and hopefully prevent any postoperative thigh symptoms.  

Step 10: Posterior instrumentation and fusion.  After the spacers are inserted we’ll place bilateral pedicle screws and perform the posterolateral bony fusion.  I think I’ve probably already exhausted you with the very detailed discussion above so I’m going to talk about this second half of the procedure in the next post.  It is worth noting that the two things we typically DON’T do at this point (which most other lateral access surgeons still do) are: 1) perform a laminectomy to directly decompress stenosis; and 2) flip the patient prone to place the pedicle screws.  First, remember I told you about the power of indirect decompression.  Now that you’re a believer in indirect decompression you know that a direct laminectomy is almost never necessary after XLIF.  I know from looking at my XLIF data over the past 5 years that (as of the posting of this post) out of 253 XLIF patients since I started keeping detailed records in 2014 only 3 have had to return to the OR for failure of indirect decompression and none have had to go back in the past two years.  That’s 1.2%.  Why would I subject you to nearly an extra hour of anesthesia time, not to mention the risks of me drilling around your nerve roots (my spinal fluid leak rate is probably 1-2% although I’ve never formally calculated it) when 99% of my patients don’t need direct decompression after XLIF??  Believe in indirect decompression! 

Regarding the placement of pedicle screws, the standard method is to finish the XLIF, take the drapes down, flip the patient face down onto another OR table and then re-prep and drape.  That takes at least 45 minutes with all hands on deck!  Instead, we’ve pioneered a strategy of lateral single-position surgery (LSPS) in which we keep the patient in the lateral position and place the screws that way (see figure 19.)  This has led to massive increases in our OR efficiency that I’m certain will translate into better patient outcomes (I’m involved in several studies to prove this in the literature.)  It seems obvious to just place screws in the lateral position after XLIF but believe it or not a lot of surgeons still feel like they have to reposition the patient prone to place screws safely.  Slowly but surely, through surgeon education and by publishing our data, we’re starting to convince surgeons of the value of LSPS.  More on this in future posts.

IMG 6280

Figure 19: placing pedicle screw fixation (yes, using a power drill) with the patient in the lateral decubitus position (patient is right side up with head towards the left of the image.)  By avoiding having to flip the patient prone and also by allowing members of the surgical team to work concurrently (notice my assistant Jack working in the front of the patient) this strategy of lateral single position surgery saves a tremendous amount of time under anesthesia for the patient. This likely translates into improved patient outcomes. 


Figure 20: pre- and post-operative images of a patient who underwent an L4/5 XLIF and percutaneous pedicle screw fixation for a spondylolisthesis.  This patient had complete resolution of their preoperative symptoms after surgery and went home after less than 24 hours in the hospital. 

Thanks for reading!  I know it was a detailed one!  I just want you to be educated as possible about XLIF if you’re considering it yourself!

J. Alex Thomas, M.D.


1. Uribe JS, Vale FL, Dakwar E: Electromyographic Monitoring and Its Anatomical Implications in Minimally Invasive Spine Surgery. 35:368–374, 2010.

Lateral ALIF is a true single-position strategy for lumbar fusions

I’m back!  I realize there’s been quite a delay since my last post and for that I apologize.   For over a year now I’ve been busy helping to develop a new retractor system for lumbar spine surgery.  This retractor allows surgeons to perform an anterior lumbar interbody fusion (ALIF) in the lateral position, a procedure we (perhaps not so creatively) call lateral ALIF (see figure 1).  Why does that matter?  In my opinion, an ALIF is the most powerful way to fuse a segment of the lumbar spine and correct spinal deformity (hint: it’s because ALIF allows you to insert the largest spacers!). One drawback of ALIF, though, is that since it’s traditionally performed in the supine position (with the patient laying on his back), if the surgeon wishes to place posterior instrumentation he has to close the incision on the front of the patient and then reposition the patient prone to get access to the back of the spine.  This process of repositioning can add nearly an hour of time to the procedure and may also increase risk to the patient.  By keeping the patient on his side, in a single position, the surgeon can harness the power of ALIF and then immediately be ready to place posterior instrumentation, all without having to stop to reposition the patient (see figure 2).  Now, a so-called 360-degree lumbar fusion that used to take 3 hours to perform can now be done in an hour.  This is good for my OR throughput but it’s GREAT for you, the patient, who will avoid that extra time under anesthesia.  This could potentially be the one of the most important innovations in spine surgery in years. 


Figure 1: The new Lateral ALIF retractor during a recent ALIF at L5/S1.  The patient is in the lateral decubitus position with their left side up.  The patient’s head is at the left of the image.


Figure 2: Lateral ALIF at L5/S1.  This image gives you an idea of the massive increase in efficiency you get without having to reposition the patient.  Here, I’ve already placed wires for insertion of percutaneous pedicle screws while my physician assistant Jack Bagley continues to close the abdominal lateral ALIF incision.  Patient is right side up with head towards the left of the image. 

The ALIF, first described in the 1930s, is the original interbody fusion in which bone graft is inserted into the cleaned out intervertebral disc (IVD) space to promote fusion and correct spinal deformity (in modern ALIF the bone graft is carried in a spacer or cage).  Since then, many other techniques have been developed to place spacers into the disc space via a posterior approach.   These other –IF procedures, such as posterior lumbar interbody fusions (PLIFs) or transforaminal interbody fusions (TLIFs) represent early attempts at a single position strategy.  These procedures allow surgeons to perform the three standard steps of a spinal fusion: 1) neural decompression (laminectomy or discectomy), 2) interbody fusion, and 3) placement of posterior instrumentation with the patient in the prone position.  Thus, traditional spine surgeons may say “Well I’ve been doing ‘single-position’ lumbar fusions for years.”  Indeed, TLIF and PLIF are the most common way to perform lumbar fusions these days.  The problem with TLIF and PLIF, though, is that in order to place spacers from behind, one or more nerves have to be retracted out of the way to sneak the spacer into the disc space.  That means that for PLIF and TLIF the surgeon is forced to use very small spacers (see figure 3).  Because you’ve read recent Spinal (con)Fusion posts, though, you know that I believe in the power of large intervertebral spacers.  Bigger is better and thus ALIF is a much more powerful technique for spinal fusion than PLIF or TLIF.  Now, by doing the ALIF in the lateral position I can have concurrent access to the back of the patient to perform a decompression and place pedicle screws without “flipping” the patient. 

Banners and Alerts and Inbox 32171 messages 7 unread

Figure 3: side-by-side comparison of various intervertebral spacers.  Notice how much larger the XLIF/ALIF spacer  is versus the much smaller TLIF or PLIF spacers. Size matters!

As you can probably tell, I’m very excited about this new retractor and surgical technique (yes, I’m biased.)  In future posts we’ll talk more about the details of a lateral ALIF procedure.  You’ll also see how well lateral ALIF at L5/S1 compliments the extreme lateral interbody fusion (XLIF) at L4/5 and above.  Lots more about single-position lateral surgery to come! 

Thanks for reading!

J. Alex Thomas, M.D.

Believe in Indirect Decompression!

One of the most common procedures that I book patients for is an extreme lateral interbody fusion (XLIF).  This is a minimally-invasive lumbar fusion procedure that has all of the benefits of the classic anterior lumbar interbody fusion (ALIF) without its downsides (all of the bad things that can occur by traversing someone’s abdomen to get to the spine.)   In previous posts we’ve alluded to various types of spinal deformity that can cause pain.  The one constant in all of these types of spinal deformity: stenosis.  Both XLIF and ALIF rely on an old orthopedic principle known as indirect decompression in which properly sized spacers are used to correct spinal deformity and thus correct stenosis. 

Recall from previous discussion that stenosis, or narrowing around nerve roots, typically results after years of spinal degeneration.  The resulting stenosis can be divided into two simplistic types (and these types are just the way that I think about it in my head, don’t go looking for them in textbooks!).  The first type of stenosis is structural stenosis.  Here, the basic anatomical components of the spine are generally unchanged in terms of their shape or volume; it’s just that these parts of the spine have collapsed onto nearby nerve roots thereby causing pain.  I believe that this is the most common form of stenosis by far.  One common example of structural stenosis occurs when the intervertebral disc (IVD) has degenerated and collapsed resulting in loss of foraminal height and foraminal stenosis.  Another example of structural stenosis is seen in spondylolisthesis when one vertebral body slides over the one below it causing a dynamic foraminal stenosis that worsens when the patient stands and loads their spine (more on this topic later.)  Lastly, it is my opinion that the central stenosis that causes neurogenic claudication in the elderly is a form of structural stenosis resulting from buckling of the ligamentum flavum (this is controversial as some believe that the body actually produces reduntant ligamentum flavum which would be more like the reactive stenosis discussed below.) In each of these cases, normal spinal structure and alignment has been lost resulting in stenosis and pain.

The other simplistic type of stenosis is reactive stenosis.  Here, there IS an increase in the shape or volume of a component of the spine, which results in stenosis and compression of nearby nerves. The most common example of this occurs when the facet joint degenerates and becomes larger as it becomes consumed by arthritis.  This leads to osteophyte (fancy word for bone spur) formation, which can cause nerve root compression and radiculopathy. 

To fix reactive stenosis the surgeon must perform a direct decompression procedure, a laminectomy or foraminotomy, to remove all excess bony overgrowth from around the nerves.  In fact, classically this is the way that all nerve compression (regardless of which type of stenosis is causing the compression) is relieved.  The typical neurosurgeon’s mentality (and I can say this because I’m one of them) is that the only way to know that a nerve is decompressed is to remove any overlying bone and actually see the nerve.  But what if this isn’t always necessary?  I believe that in most cases it’s NOT necessary (and trust me, it takes a huge leap of faith on the part of both the surgeon and the patient to come to this realization.)  Because most spinal stenosis is structural and not reactive, restoration of normal structure and alignment of the spine will relieve stenosis and pain without the extra time and risk of a laminectomy.

Recall that loss of normal spinal structure and alignment begins with degeneration and resulting collapse of the IVD.  So, to restore structure and alignment we go to the disc space (remember the post on spacers?)!  Indirect decompression is achieved when a properly sized spacer is inserted into a collapsed disc space to restore the height of the neural foramen (see figures 1 and 2).  This then relieves nerve root compression because the space around the nerve in the foramen is restored; I don’t have to do more work to remove bone that doesn’t need removing!  To be sure, this is a controversial topic.  I know plenty of very good spine surgeons who just don’t believe in indirect decompression and subject their patients to a concurrent laminectomy with every spinal fusion.  They’re paranoid (we surgeons are a VERY paranoid bunch, some are just more so than others) that if they don’t directly decompress the nerve and visually confirm that it’s decompressed then they may not relieve the patient’s pain.  I get it.  Like I said, it’s a leap of faith.  A laminectomy isn’t a benign procedure though.  There’s a 5-10% risk of dural tear and spinal fluid leak for starters.  Typically this is a minor complication but it can be catastrophic.  There’s also the risk of scar tissue formation around exposed nerve roots, which can lead to chronic pain after surgery.  Finally there’s just the risk of being under anesthesia for the extra time needed to perform the laminectomy.  Why would I subject my patients to these risks when I know that the indirect decompression achieved by the spacer will probably suffice?  Believe in indirect decompression!


Figure 1: A, preoperative image showing severe collapse of the IVD resulting in foraminal height loss and nerve root compression; B, postoperative image demonstrating restoration of disc space and foraminal height after insertion of a large intervertebral spacer.

T LUMBAR L 5 S 1 SPOT 4AP 5130

Figue 2: As above in figure 1, image on left is preoperative image demonstrating severely collapsed IVD at L5/S1 with resultant severe foraminal stenosis (pink outline).  The image on the right is a postoperative image after an L5/S1 anterior lumbar interbody fusion (ALIF) with significant increase in disc space, and thus, foraminal height.  This patient’s leg pain was relieved immediately after surgery WITHOUT laminectomy. 

Of course there are times when relying solely on indirect decompression may not be appropriate.  In cases of severe reactive stenosis in which, say, a nerve root is encased in bone (see figure 3), indirect decompression probably isn’t going to work no matter how large of a spacer you put in.  Also, in cases of large concurrent disc herniations or facet cysts (a type of reactive stenosis, I suppose, which I’ll discuss in a later post) I may also be forced to do a direct decompression.  The more cases I do, though, the more I’m surprised at what I can get away with in terms of avoiding a direct decompression.  These days I’ll typically assume that indirect decompression will work but explain to the patient that there is a very small chance that indirect decompression may fail and that we may have to do a small “second stage” laminectomy later.  How small of a chance you ask?  I went back and looked at the data for every one of my lumbar fusions performed since December 2013.  Of nearly 250 patients only 8 needed reoperation for failure of indirect decompression.  That’s a 3% risk.  I’d say those are pretty good odds in favor of direct decompression. 

Reactive stenosis

Figure 3: Severe “reactive” foraminal stenosis at L4/5 and L5/S1 resulting from severe bony overgrowth around nerve within the neural foramen (red arrows).  This patient failed indirect decompression and required a minimally-invasive foraminotomy a few months after his initial surgery.

Thanks for reading!

J. Alex Thomas, M.D.