Chapter 35 Minimally Invasive Percutaneous Lumbar Fusion Technique

The development of minimally invasive techniques to address disease pathologies among a number of surgical specialties in place of traditional open surgery has become a noteworthy trend. Examples include the laparoscopic approach to gallbladder disease as well as knee and shoulder arthroscopy as a successor to open surgery for meniscal and rotator cuff disorders, respectively.

This shift toward minimally invasive or minimal-access technologies has also taken place over the last decade in the specialty of spinal surgery. The goal of the procedure remains the same: to achieve outcomes equivalent to or better than that of traditional open surgery while trying to reduce overall postoperative pain, intraoperative blood loss, hospital stay, and surgical scarring.

It is estimated that more than 200,000 lumbar procedures are performed yearly for underlying discogenic disease and instability [1]. The degenerative process, often a result of normal aging, leads to degenerative disc disease, spondylolisthesis, and segmental instability. The traditional open posterior lumbar fusion yields an acceptable efficacy with a high fusion rate [2]. However, the exposure that is needed to achieve optimal visualization of the transverse processes, lamina, and facet joints often requires a midline incision of more than 5 to 7 cm, depending on the number of levels involved. Disruption and retraction of the deep paraspinal musculature has been shown to cause localized denervation, which may lead to continued back pain and spasm after the procedure [3]. Multiple studies in this area have shown deleterious histologic effects, including regional ischemia from increased intramuscular pressure secondary to the use of retractor blades [4,5].

Less invasive techniques and instrumentation have been and continue to be developed to address the effects of extensive soft tissue disruption. In 1995, Mathews and Long [6] published their experience with the use of percutaneous pedicle screws. Unfortunately, they noted a high nonunion rate. Lowery and Kulkarni [7] published a review of eight cases in which they utilize a similar technique but with the use of rods instead of longitudinal connectors; they reported a 96% fusion rate for a mini–open anterior approach supported with a minimally invasive posterior fusion. Advances to the procedure described by Harms and Rollinger [8] have produced a method of achieving a solid arthrodesis through a posterior interbody approach.

Subsequently, the transforaminal lumbar interbody fusion (TLIF) approach was developed as a means of addressing the pathology in the disc space while minimizing the risk of iatrogenic injury to nerve roots and soft tissue. Foley and colleagues [11] described a minimally invasive method of performing TLIF that has become popular with trained surgeons [9–11]. The TLIF procedure continues to evolve as a minimal access procedure that is ideal for a patient who has clinically defined symptoms that correlate with results of diagnostic studies related to radiculopathy pain secondary to degenerative disc disease, recurrent herniated nucleus pulposus, or segmental instability. The approach allows for adequate access to the affected facet joint and lamina to perform a laminotomy and discectomy. Furthermore, through a mini–open and percutaneous approach, restoration of disc height, sagittal balance, and stabilization of the vertebral segments can be effectively performed. A study by Schwender and colleagues [12] reported clinically significant improvements in visual analog scale and Oswestry Disability Index scores along with a 100% fusion rate in a cohort of patients who underwent a minimally invasive TLIF procedure [12].

At the time of this writing, several minimally invasive systems are available for this purpose. Each of the systems features a top-loading polyaxial screw, with a unique rod delivery system that has traditionally been a challenge for minimally invasive systems. Most systems allow for compression and distraction. The Pathfinder Minimally Invasive Spine Instrumentation System (Abbott Spine, Austin, TX), PIVOT (Globus Medical, Audubon, PA), and VIPER (Depuy Spine, Raynham, MA) allow for a true percutaneous approach with the use of cannulated pedicle screws that are delivered over a guidewire. These systems allows for performance of a posterolateral fusion. TLIF can also be performed through a mini–open approach. The Atavi Atraumatic Spine Surgery System by Endius (Zimmer Spine, Minneapolis, MN) utilizes an illuminated expandable retractor system that allows for direct visualization and offers the surgeon flexibility in performing single or multiple procedures including interbody and posterolateral fusions. The SEXTANT System (Medtronic Sofamor Danek, Memphis, TN) is a minimally invasive system that requires three incisions to complete a fusion. The pedicle screws are placed through separate incisions as in the previously described systems. The rod is delivered through the third incision by means of a patented arc delivery system. TLIF can be performed with a mini-open approach using retractor tubes. In early 2007, the U.S. Food and Drug Administration (FDA) approved the Silverbolt MIS [minimally invasive screw] System by VertiFlex (VertiFlex, Inc., San Clemente, CA) for marketing and sales in the US. This system, designed by spine surgeons in conjunction with engineers, is a platform for percutaneous and mini–open approaches. The Silverbolt was built to address the difficult challenges of minimally invasive spine surgery including fusion at L5-S1.

Procedures and technique

The techniques described here represent different methods of achieving the same surgical goals. The method chosen should be based on the specific anatomy of the patient and on the training and comfort level of the operating surgeon. It should be noted that there is a steep learning curve to performing minimally invasive lumbar fusion for surgeons who were trained to perform traditional open approaches. The concept of working in a constrained surgical channel when performing decompression through a mini–open approach can be a challenge for some surgeons and may initially take longer than the classic open approach. Three-dimensional visualization of anatomical structures is replaced with thought processes involving biplanar fluoroscopy. The ability to understand the orientation of pedicles is paramount because instrumentation systems are based on correct targeting. Disorientation of the surgeon in relation to the anatomy of the spine can lead to longer operating times, greater exposure to radiation, and, in the end, less than acceptable outcomes. Appropriate preoperative planning by the surgeon should include a surgical plan for method of approach, extent of decompression, and placement of internal fixation. Changes to spinal biomechanics and restoration of alignment must also be considered.

The techniques described herein utilize the VertiFlex Silverbolt MIS System and are based on the experience of the senior chapter author (JHS).

Fluoroscopic Pedicle Targeting

Accurate pedicle screw placement via a percutaneous approach relies heavily on good fluoroscopic images. Obtaining clear images begins with patient positioning. Obese patients, patients with congenitally small pedicles, and those with significant scoliotic deformity can present a challenge to achieving optimal visualization of the lumbar pedicles. When the patient is placed in the prone position, the iliac crests must be aligned in both the sagittal and coronal planes. All radiopaque structures should be placed outside the intended surgical field prior to sterile draping. An AP view should be obtained to localize the pedicles. Ideally, the image beam is directed perpendicular to the spine and centered. In cases involving the L4, L5, and S1 pedicles with noted scoliotic deformity, significant asymmetric disc space narrowing, or segmental instability, rotational distortion may be present. A modified Ferguson view can be employed, in which the fluoroscopic unit is angulated up to 30 degrees cephalad to match the inclination of the sacrum.

1. The patient’s written informed consent is obtained.

2. After the correct pedicle is identified, a Jamshidi needle is introduced, beginning at the outer-upper quadrant of the pedicle and advanced to just lateral of the medial pedicle wall in the inner quadrant (Fig. 35-1).

3. A lateral view is necessary to confirm correct placement within the pedicle. In order to overcome the effects of parallax, the fluoroscopy unit may have to be angled either cranially or caudally to be tangential to the end plates.

4. Once optimal visualization is achieved, the Jamshidi needle is advanced through the isthmus of the pedicle, staying within a zone between the superior and inferior margins (Fig. 35-2).

5. The tip of the needle is advanced through the pedicle isthmus to the vertebral body.

Figure 35–1 Anteroposterior view of pedicle illustrating needle entry point and path for needle advancement. 1, outer wall of pedicle; 2, midline of pedicle; 3, inner wall of pedicle.

Figure 35–2 Lateral view of pedicle corresponding to Figure 35,1, showing the path for needle advancement.1, beginning point of pedicle; 2, center of pedicle; 3, posterior margin of vertebral body.

The surgeon must be aware of differences in pedicle morphology in different disease states and among subpopulations. [19,22] Efficacy of the pedicle screw is highly dependent on proper placement, including matching the pedicle angle (Fig. 35-3). Care should be taken to avoid placing the needle too far medially and inferiorly.

Figure 35–3 Axial view of vertebral body illustrating the correct angle for pedicle screw placement.

Percutaneous Approach

1. The patient’s written informed consent is obtained.

2. The patient is placed prone on a radiolucent spine table following intubation. We prefer the use of the Jackson Spinal Table (Mizuho OSI, Union City, CA). This table allows for ease of patient positioning, reduces vena cava compression, and facilitates fluoroscopic imagery in multiple planes.

3. Appropriate antibiotic and deep vein thrombosis (DVT) prophylaxis should be a standard of practice.

4. Following standard skin preparation and sterile draping, AP fluoroscopic images should be obtained to locate the pedicles and relevant anatomy. This step can be performed with a spinal needle. Ideal placement is generally 3 to 4 cm off the midline.

5. A Jamshidi needle is placed using the fluoroscopic pedicle targeting technique previously described (Fig. 35-4).

6. Once placement of the Jamshidi needle within the vertebral body at the appropriate orientation is confirmed, a guidewire is passed through the cannulated center of the entry needle (Fig. 35-5). Care should be taken not to advance the guidewire to within 10 mm from the anterior wall of the vertebral body.

7. Following confirmation by lateral fluoroscopic images of the tip of the guidewire, the Jamshidi needle is withdrawn.

8. Skin and underlying fascia are dilated by means of sequential dilators to create a pathway for the pedicle screw (Fig. 35-6). The largest dilator is left in place to protect surrounding soft tissue and for measurement purposes.

9. A cannulated tap is advanced over the guidewire down to the pedicle. The size of the tap is at the surgeon’s discretion. Depth measurements located on the outside of the tissue dilator sleeve can help guide depth during tapping. Care should be taken to prevent the guidewire from advancing.

10. A pedicle screw of appropriate size and length is preloaded by the surgical staff. The slotted head of the polyaxial screw must be placed in alignment with the tower assembly to facilitate accurate placement of the rod. There are multiple handle options for the surgeon to choose from. Once the screw is secured, the tap and tissue dilator sleeve are withdrawn while the screwdriver and tower assembly are placed over the guidewire.

11. The pedicle screw is advanced with the polyaxial screwdriver until the appropriate depth is achieved (Fig. 35-7). AP and lateral fluoroscopic images are obtained to confirm the screw’s placement within the pedicle, orientation, and overall depth. Care should be taken to avoid advancing the screw head to bone, which would limit the ability to seat the rod.

12. The guidewire is withdrawn as the screw enters the pedicle. The screwdriver is withdrawn from the tower assembly.

Figure 35–4 (A) Insertion of Jamshidi needle. (B) Corresponding fluoroscopic anteroposterior view.

Figure 35–5 Insertion of guidewire following confirmation of correct needle placement within the pedicle.

Figure 35–6 Sequential dilators placed over the guidewire.

Figure 35–7 Pedicle screw load on tower assembly, placed over the guidewire.

Subsequent pedicle screws are placed with this same technique. It is important to note that all screw tower assemblies should line up in the same orientation and height before on the next step of the procedure (Fig. 35-8).

13. Once all towers are placed and confirmed, an alignment guide is placed over each tower. The guide is secured with the use of a universal locking tool (Fig. 35-9). The alignment guides can then be attached by being snapped together.

14. A rod measurement guide is placed on the top of the alignment guide to facilitate measurement of the rod (Fig. 35-10).

15. A tissue splitter is introduced into the screw tower assembly (Fig. 35-11).

16. A precontoured rod, as previously measured, is placed into the most caudal screw tower. The rod features a ball-end that locks into the slotted head of the pedicle screw. A pivot arm guides the rod to the opposite pedicle screw head (Fig. 35-12).

17. Once the rod is seated, a cap inserter is placed in the tower assembly that does not contain the rod pusher (Fig. 35-13).

18. Subsequent screw caps are now placed after the rod pusher is removed.

19. If compression or distraction is desired, a compression/distraction tool is available that seats on the alignment guide. Appropriate distraction or compression can be applied with a T-handle driver (Fig. 35-14).

20. Final tightening of the construct are performed with an antitorque stabilizer and torque driver (Fig. 35-15). The screws should be tightened to 55 to 65 pounds per square inch (psi) as indicated on the torque driver.

21. The alignment guides are removed with the universal locking driver. The screw tower assemblies are loosened and removed.

Figure 35–8 Alignment of screw tower assemblies.

Figure 35–9 Alignment guides placed and locked with use of universal locking tool.

Figure 35–10 Placement of the rod measurement guide.

Figure 35–11 The photographic view of inserting of the tissue splitter into the tower assembly (A) and the tissue splitter itself (B).

Figure 35–12 Intraoperative picture of deployment of precontoured rod with pivot arm (A) and fluoroscopic lateral view (B).

Figure 35–13 Cap inserter placed in the tower assembly.

Figure 35–14 Compression being applied to the construct.

Figure 35–15 Final tightening with antitorque handle.

Mini-Open Approach

A mini-open approach can be performed unilaterally for the purposes of decompression with or without interbody fusion. If interbody fusion is being considered, the incision site should be made just lateral to the pedicles to facilitate exposure to the disc space for cage placement. A table-mounted retractor such as the VertiFlex Oracle Expandable Retractor is used to create a working channel (Figs. 35-16 to 35-18):

An illuminated headlight and loupe magnification are beneficial to the surgeon during the initial part of the procedure.

Pedicle screws are placed with this mini-open, muscle-sparing approach under direct visualization.

Decortication and placement of bone grafting are performed in the lateral gutter. Pedicle screws are placed on the opposing side by means of the percutaneous technique already described.

Figure 35–16 Placement of Oracle Expandable Retractor (VertiFlex, Inc., San Clemente, CA) after tissue dilation.

Figure 35–17 Oracle Expandable Retractor (VertiFlex, Inc., San Clemente, CA) expanded to open the working corridor.

Figure 35–18 Intraoperative view after facetectomy and decompression.

This technique has been popularized for recurrent disc herniations, lateral recess stenosis, and degenerative disc disease with instability. If decompression through a laminotomy without TLIF is indicated, an incision slightly medial to the pedicles on the affected side can be performed with the described mini-open approach.

Another exposure variation is the midline incision with undermining of the subcutaneous tissue to the fascial layer on both sides. A bladed retractor system or microlaminectomy retractor can by utilized to achieve appropriate mini-open exposure followed by placement of pedicle screws under direct visualization.

Postoperative care

Post-operatively, patients should be monitored with frequent neurologic assessments. Treatment for postoperative pain varies according to the clinical practice standards of the surgeon. Many surgeons do prescribe a lumbar orthosis after fusion surgery. Ambulation usually begins on postoperative day 1. Criteria for discharge from the hospital should include adequate pain control with oral analgesic medication, safe ambulation with the use of an assistive device (if needed), and medical stability. The hospital stay for minimally invasive lumbar fusion varies from an average of 2 days to longer for patients who have additional medical co-morbidities.

The usual precautions during the acute postoperative phase include avoidance of heavy lifting, bending, and high-impact activities. Generally, patients should be encouraged to ambulate to avoid deconditioning. Nonsteroidal anti-inflammatory medications should be avoided because they are known to decrease the fusion rate. [23] Instructions for surgical wound care should be given to and understood by the patient prior to discharge. Patients should be told to contact the surgeon or support staff if they have persistent fevers, chills, motor weakness, increasing pain, headaches, or wound problems. Follow-up examinations to determine neurologic status and radiographs should be performed at intervals defined by the treating surgeon. Assessment of the patient’s complaints, pain level, function, and satisfaction should be ascertained at subsequent follow-up visits.

CASE STUDY 35.1 Anterior Displacement of L4 on L5 Corrected by Oracle Expandable Retractor with Polyetheretherketone (PEEK) Interbody Cage and Vertiflex Silverbolt Pedicle Screws

A 38-year-old woman presented to our spine surgery practice with a 5-month history of positionally dependent back pain with radiations including paresthesias to the bilateral lower extremity. Prior to this evaluation, her primary care physician had administered several local cortisone injections to the low back. Additional treatments included the use of nonsteroidal anti-inflammatory medications and at-home stretching exercises. A detailed history taken at the time of her evaluation revealed an average pain level of 8 out 10 on a visual analog scale. She reported that the symptoms were aggravated with sitting.

A focused physical examination of the back and lower extremities was conducted. It revealed mild tenderness over the midline, a slight list, and mild weakness of the right anterior tibialis anterior (4+/5). Straight-leg raising on the right side reproduced the left-sided low back symptoms. No long tract signs were identified. Standing radiographs, including flexion and extension views, were taken. The radiographs showed a grade I spondylolisthesis at L4-L5 with moderate loss of disc height (Fig. 35-19). Furthermore, a flexion-extension instability was noted.

Figure 35–19 Case Study 35.1: Preoperative lateral radiograph showing L4–L5 spondylolisthesis.

The patient was referred for MRI because some conservative measures had already been tried and had failed and because she presented with significant physical findings. MRI showed 6 mm of anterior displacement of L4 on L5 (Fig. 35-20). No definite pars defect was identified. A central disc herniation was present at L4-L5 with neural encroachment. Disc desiccation was also prominent at the L4-L5 level. Following a detailed review of the MRI findings, conservative and surgical treatment options were discussed with the patient. She elected to pursue surgical management of her problem secondary to increasing pain and limited ability to perform daily activities.

Figure 35–20 Case Study 35.1: Preoperative sagittal magnetic resonance image showing 6 mm of anterior displacement of L4 on L5.

Five weeks after initial presentation in our office, the patient underwent surgery. An L4-L5 left-sided hemilaminectomy (Gill procedure) along with interbody and intertransverse fusion with instrumentation was performed. A mini-open incision was utilized on the left, approximately 3.5 cm off the midline. Blunt finger dissection was carried down to identify the transverse process of L4 and L5. The Oracle Expandable Retractor by VertiFlex, Inc., was used to span the interval between the L4 and L5 pedicles. A facetectomy was performed on the left side. The disc space was distracted. Fluoroscopic images confirmed reduction of the spondylolisthesis. Following subtotal discectomy, a polyetheretherketone (PEEK) interbody cage was placed in the disc space. The transverse processes of L4 and L5 were then decorticated. VertiFlex Silverbolt pedicle screws were placed at L4 and L5 through the same mini-open incision. Bone graft was placed in the lateral gutter. A precontoured rod was placed and secured. Compression was placed across the screws prior to final tightening.

Attention was then turned to the right side. The pedicles of L4 and L5 were identified and marked with fluoroscopic guidance. A Jamshidi needle was placed at the same orientation as the L5 pedicle; placement and orientation were confirmed with biplanar fluoroscopy. A guidewire was placed through the needle and advanced through the pedicle into the vertebral body. The pedicle was tapped with a cannulated tap. A pedicle screw was placed percutaneously at the L5 level. The same procedure was carried out at the L4 level. The appropriate-size rod was then delivered to the screws. Final tightening of the screws was performed, and final fluoroscopic images were obtained (Fig. 35-21).

Figure 35–21 Case Study 35.1: Postoperative anteroposterior fluoroscopic image showing placement of pedicle screws and interbody cage.

Postoperatively the patient spent two uneventful days in the hospital before being discharged home. At the time of the patient’s first follow-up visit approximately 2 weeks after surgery, she had discontinued the use of all analgesic medications. She reported her overall pain level to be 2 on a scale of 10 (visual analog scale) and to be limited to the low back. She also reported complete resolution of her preoperative leg symptoms. Her physical examination revealed no neurologic deficits. Standing radiographs were obtained to confirm position of the instrumentation. The patient remained nearly asymptomatic at the time of her last follow-up visit.

CASE STUDY 35.2 Spondylolisthesis Grade I/II at L4-L5, Grade I Slip at L5-S1 Corrected by Oracle Expandable Retractor with Polyetheretherketone (PEEK) Interbody Cage and Vertiflex Silverbolt Pedicle Screws

A 66-year-old woman presented to our practice with a 14-month history of progressively increasing low back pain with radiations down the right leg. She described the symptoms as unremitting back and buttock pain with posterolateral leg radiation of pain to the foot. Onset seemed to be insidious because she knew of no preceding injury. The patient did admit to having previous episodes of low back pain prior to the current onset, however. A review of records of treatment she received prior to presenting in our office revealed conservative care by a pain management physician. Treatments included multiple epidural steroid injections, facet blocks, medial branch blocks, and sacroiliac joint injections. The patient reported experiencing relief initially for several months, but the effect gradually declined. Additional treatments included physical therapy and narcotic analgesic medications. The medical history of the patient was also positive for hypertension and obesity.

A detailed examination identified a positive straight-leg raising test response on the right side and mild tenderness in the right sacroiliac region. No motor, reflex, or sensory deficits were apparent. Standing radiographs were obtained, which clearly documented a spondylolisthesis grade I/II at L4-L5, grade I slip at L5-S1 with disc space narrowing, and facet disease at both levels (Fig. 35-22). Flexion and extension radiographs confirmed segmental instability. Furthermore, MRI of the lumbar spine showed a right lateral disc protrusion at L4-L5 that was causing moderate right foraminal narrowing. A diffuse posterior disc bulge with mild foraminal narrowing was noted at L5-S1 with facet arthropathy at both levels (Fig. 35-23). The findings were reviewed at length with the patient. Surgical management was recommended in light of her complaints and the findings of the diagnostic workup. The patient consented to an L4-L5, L5-S1 laminectomy with interbody and intertransverse fusion with instrumentation, despite the risks, benefits, and alternatives presented to her.

Figure 35–22 Case Study 35.2: Preoperative lateral radiograph showing spondylolisthesis at L4–L5 and L5–S1 with disc space narrowing.

Figure 35–23 Case Study 35.2: Preoperative sagittal magnetic resonance image showing a diffuse posterior disc bulge with mild foraminal narrowing at L5–S1 with facet arthropathy at both levels.

After endotracheal intubation, the patient was placed on a Jackson Spinal Table. All bony prominences were well padded. The fluoroscope was brought into the operative field to identify landmarks. The pedicles were marked with use of an AP view. A skin incision approximately 5 cm in length was made over the pedicles on the patient’s right side. Blunt finger dissection was carried out to reach the pedicles and transverse processes. The Oracle Expandable Retractor was used to maintain a working corridor. A hemilaminectomy was performed at L4-L5 followed by a total facetectomy. Local bone was morselized and used to pack the polyetheretherketone (PEEK) interbody cage that was inserted after subtotal discectomy. Lateral fluoroscopic images confirmed appropriate placement of the cage. Attention was then turned to the L5-S1 level, where the same procedure was performed using the technique just described. Polyaxial pedicle screws were then placed at L4, L5, and S1 under direct visualization. A Woodson dissector along with fluoroscopy was used to confirm accurate placement of the pedicle screws. A precontoured rod was placed. Compression was applied prior to final tightening of the screws.

The percutaneous approach was used to deliver the pedicle screws to the opposite side. Final fluoroscopic images obtained prior to wound closure showed significant reduction in the preoperative spondylolisthesis at L4-L5 and at L5-S1 (Fig. 35-24). The patient reported immediate relief of her leg complaints.

Figure 35–24 Case Study 35.2: Final lateral (left) and anteroposterior (right) fluoroscopic views following minimally invasive TLIF with percutaneous pedicle screw placement at two levels.

Key points

Indications for minimally invasive lumbar fusion surgery mirror those for conventional, open surgery. Proper patient selection and surgical indications are necessary to avoid surgical failure.

The learning curve for this procedure can be steep for surgeons who are accustomed to performing open spinal surgery. Complexities can relate to exposures through a narrow working channel, having to rely on biplanar fluoroscopy instead of direct three-dimensional visualization of anatomy, and using minimally invasive instrumentation that the surgeon may be unfamiliar with.

Many minimally invasive systems are available. Each system features differences in design and instrumentation. However, the overall approach is the same. A thorough understanding of each patient’s anatomy and pedicle orientation is necessary to be able to safely direct pedicle screws utilizing a percutaneous approach

A method of percutaneous screw placement has been described. Variations of this method, including a mini-open approach, have been presented; choice of variation is based on the treating surgeon’s preference, training, and overall comfort level.