- Email: [email protected]
Minimally invasive removal or revision of lumbar spinal fixation
The Spine Journal 4 (2004) 701–705
Case Studies
Minimally invasive removal or revision of lumbar spinal fixation Anthony A. Salerni, MD* Orthopedic Professional Association, 14 Maple Street, Suite 100, Gilford, NH 03249, USA Received 22 August 2003; accepted 4 March 2004
Abstract
BACKGROUND CONTEXT: There are both absolute and relative indications for the removal of pedicle screw fixation in the lumbar spine. Whatever the reasons are, removal of this hardware has required a surgical dissection that has been generally as extensive as the one used for their initial placement. These dissections are always disabling in the short term. In fact, the magnitude of this disabling pain can be significant enough so as to effectively eliminate screw removal as a logical treatment option for many conditions where indications for removal are only relative. Percutaneous pedicle screw fixation has served to amplify the stakes associated with this dilemma. In fact, this new technique makes the need for a less invasive method of pedicle screw removal greater now than ever. PURPOSE: This paper describes a minimal access surgical technique for pedicle screw construct removal that employs the tubular retractor system that was originally developed for microendoscopic discectomy. STUDY DESIGN: This case study represents a summary of the surgical experience gained from the first 10 patients to have undergone removal or revision of pedicle screw constructs by this minimally invasive method. METHODS: A retrospective analysis of pre- and postoperative clinical data was gathered from the hospital records. Surgical times and blood loss were also extracted from these records. The procedure is described in detail. Interpretation of the surgical parameters and clinical effects are discussed. RESULTS: Six patients presented with a radiculopathy secondary to a misdirected pedicle screw. Two of these patients were admitted for simple removal. The four remaining patients who had undergone percutaneous pedicle screw fixation developed acute radicular pain from a misdirected screw. These patients underwent revision of their constructs by this method. Screws were also removed unilaterally in four other patients as the initial phase to revision or additional surgery. All procedures were performed through 16 mm tubular retractors. Operative time averaged 33 minutes for the group, and it ranged between 22 and 40 minutes. Hospital length of stay averaged 1 day for the group. Hospital stay averaged only 0.8 hospital days for the patients in whom screw removal was the primary goal. At 1 month after surgery no patient felt limited by incisional pain. No complications occurred. CONCLUSIONS: Unlike most other minimal access surgical procedures, the learning curve for this procedure appears to be relatively flat. Removal of pedicle screw fixation in the manner described proved to be simple and straightforward. The benefits are dramatic and immediate. It is possible to complete the procedure within minutes, and the pain produced is best described as inconsequential. This minimally invasive technique radically alters both the intraoperative and postoperative courses for those who face pedicle screw removal. The disadvantages associated with the standard open approach are reduced to the production of mild short-term discomfort and an exposure to the potential risks of brief anesthesia and the possibility of a surgical infection. Considering that hospital stay should be limited to 1 day or less and that surgical times are less than 1 hour, minimally invasive removal or revision of hardware should reduce overall costs significantly. 쑖 2004 Elsevier Inc. All rights reserved.
Keywords:
PLIF; Minimally invasive; Interbody fusion; Lumbar; Percutaneous surgery
FDA device/drug status: not applicable. Nothing of value received from a commercial entity related to this research. 1529-9430/04/$ – see front matter doi:10.1016/j.spinee.2004.03.023
쑖 2004 Elsevier Inc. All rights reserved.
* Orthopedic Professional Association, 14 Maple Street, Suite 100, Gilford, NH 03249, USA. Tel.: (603) 528-9100; fax: (603) 524-5743. E-mail address: [email protected] (A. A. Salerni)
702
A. A. Salerni / The Spine Journal 4 (2004) 701–705
Introduction Removal or revision of lumbar pedicle screw fixation can be clinically beneficial for a number of situations. The advantage of doing so, however, must be weighed against the cost of putting the patient through an operation that generally requires a dissection at least as extensive as the one required to fashion the construct in the first place. For such situations as radicular pain from a malpositioned screw, however, there may be no way to avoid a second disabling operation accompanied by a lengthy recovery phase. In other situations the cost–benefit ratio may lead the patient and surgeon to conclude that the procedure is best avoided. Although this may be the correct decision for most, the cost–benefit equation used to make this decision effectively eliminates a treatment option that could potentially be useful for some individuals. Regardless of the potential benefits, the pain associated with this procedure has always been significant. Consequently, a procedure that can minimize dissection would have enormous advantages. The tools designed for MED [1] are well suited for this application. Materials and methods The medical charts of the first 10 consecutive patients to have had removal of a lumbar pedicle screw were used to gather data for this retrospective analysis. The relevant clinical and surgical data for the patient population are summarized in Table 1. Six patients presented with a radiculopathy secondary to a misdirected pedicle screw. This was a lateterm development in two of these patients, even though their fusions had been documented to be solid. These two patients were admitted for simple removal. The four remaining patients, who had undergone percutaneous pedicle screw fixation, developed acute radicular pain from a misdirected screw. These patients underwent revision of their constructs by the method to be described. Screws were also removed unilaterally in four other patients as the initial phase to revision or additional surgery. One of these patients had a solid two-level lumbar interbody fusion, and he was admitted
for unilateral removal of his construct to gain access for a planed foramenotomy as a treatment for his radicular pain. Results All procedures were performed through 16 mm tubular retractors. Operative time averaged 33 minutes for the group (range, 22 to 40 minutes). Hospital length of stay averaged 1 day for the group. Hospital stay averaged only 0.8 hospital days for the patients in whom screw removal was the primary goal. At 1 month after surgery no patient felt limited by incisional pain. No complications occurred. Technique Although the screw systems and reasons for removal or revision of pedicle screws and rods vary, the manners by which they are removed do not. The procedure is carried out with loupe magnification and a headlight. With the patient in the prone position on rolls, a C-arm fluoroscope is moved to the anteroposterior (AP) view. In many cases, the long axis of the screw and the long axis through the center of the polyaxial screw head will define the same line. In this case the fluoroscope is repositioned to obtain an image that is shot directly down the long axis of the screw to be manipulated. In most cases this will be close to the AP projection. However, because some techniques require a more lateral insertion point, the projection may be considerably oblique. Either way, a spinal needle is inserted directly over the screw head, as seen on this projection. It is advanced parallel with the beam to the screw head, which, when reached, will provide a tactile signal distinct for metal. If the polyaxial screw head was not centered over the long axis of the screw when it was locked to the rod, the two extended lines defined by these respective axes will intersect with the skin at different points (Fig. 1). A point on the skin between the two extended lines is chosen as the entry point for the spinal needle with this presentation so that the tools necessary for removal of the locking device and the pedicle screw can be positively engaged through the same incision. A 14 mm or 16 mm MetRx tube (Medtronic Sofamor Danek, Memphis, TN) is
Table 1 Clinical and surgical data for patient population Sex
Age
Diagnosis
Operation
Levels
Complications
OR time (minutes)
LOS (days)
Result
M M F M M M F F M M
40 42 63 32 48 31 43 48 66 40 Average age, 45
Screw malposition Screw malposition Junctional stenosis Failed back Failed back Screw malposition Screw malposition Screw malposition Foramenal stenosis Screw malposition
Redirection Removal Removal Revision Revision Redirection Redirection Removal Removal Removal
L5–S1 L5–S1 L4–S1 L5–S1 L3–L5 L5–S1 L5–S1 L5–S1 L4–S1 L5–S1
None None None None None None None None None None
30 38 25 40 40 34 35 30 40 22 Average time, 33
1 1 2 1 1 1 1 1 1 0 Average LOS, 1.0
Improved Improved Improved Unchanged Unchanged Improved Improved Improved Improved Improved
LOS⫽hospital length of stay; OR⫽operating room.
A. A. Salerni / The Spine Journal 4 (2004) 701–705
Fig. 1. Ideally, a path directly along the long axis of the screw best facilitates its removal. However, the trajectory that allows for precise engagement of the screw head may be considerably different from this for polyaxial pedicle screws. Choosing a surgical trajectory between these two ideal paths will allow enough freedom to positively engage both the screw and rod locking mechanism for most systems.
centered over this needle and the edges of the skin on the superior and caudal edges of the tube are marked. This demarcates the maximum extents of the required incision. The tube and needle are removed and a saggitally oriented line is drawn with a length that is represented by the diameter of the tube (Fig. 2). An incision is made in the skin, and the fascia is opened with electrocautery. The smallest MetRx dilator is passed through the muscle and held against the screw head. Fluoroscopy is used if the screw head is not readily contacted. While firmly holding the small dilator against the screw head, successively larger dilators are passed over this. A MetRx tube is passed over the last dilators and ultimately over the screw head. The dilators are calibrated to aid in selecting a tube of proper length. Choosing a tube slightly longer than the distance from skin edge to screw head allows the screw head to be captured in the tube by simply applying downward pressure on the external end of the tube. Inspection after the dilators are removed will show the muscle- and scar-covered screw head projecting up into the tube. This tissue is easily removed with a disc ronguer. Cautery can be helpful, but it must be set to low power settings. Sweeping the tube along the construct will allow much of the tissue over the rod to be removed as well. The tools specific to the system to be manipulated can be inserted down the tube to unlock the rod from the screw. Once this is accomplished, another MetRx tube is then directed to the next screw to be removed. It is often useful to have access to both ends of the rod during disengagement from the screw heads so each MetRx tube is left in its position as the operation proceeds. Unlike discectomy, foramenotomy or fusion, the tubes do not need to be locked in
703
Fig. 2. The tube, which has been cut away in this figure, is centered on the needle that defines the optimum trajectory for the approach. The diameter of the tube in contact with the skin determines the maximum length of the incision necessary for engagement of the construct. It will be possible to remove many single-level constructs through one such incision.
place. In fact, because stripping the scar tissue covering the construct and rod extraction generally can require extensive pivoting of the tubes, keeping them free floating offers significant time saving advantage. Additional screws in the construct are addressed successively in the same manner, and there may be a tube in place for each. When the rod is unlocked from all the screw heads, it can be extracted. Even if all the scar tissue has been removed from the tops of the screw heads and rod, it still can take a significant force to disengage the rod. Once loose, however, it is relatively straightforward to leaver one end of the rod up through a tube for one-level constructs. Rods placed in multilevel constructs can be more difficult to remove. The longer the rod, the less apt it is to make the acute turn needed for extraction out an end tube. In situations such as this, a more oblique entry point is chosen and another small diameter MetRx tube is positioned on the rod at the top of the construct. The skin entry point is far enough cephalad so that the path of the tube makes a fairly obtuse angle with the coronal plane. Once the rod is disengaged from the screw head, it is just a matter of pushing the rod out the tube along this relatively flat trajectory. Once the rod has been removed, extraction of the screws is simply a matter of engaging the screw head with the system-specific screwdriver through the MetRx tubes stationed over each pedicle screw. If removal of the hardware was the objective, the procedure is complete, so after screw removal the tubes would be removed and each wound would be individually closed. In other cases, the surgical objective will be to revise rather than remove the hardware. Most often the goal will be to reposition a misdirected screw. In this case, the surgical plan proceeds exactly as for removal until the rod has been unlocked from all screw heads. The procedures diverge at this point and instead of disengaging and removing the rod, it is slid along the screw heads until it is free of the
704
A. A. Salerni / The Spine Journal 4 (2004) 701–705
malpositioned screw. This screw can be replaced through the tube using standard methods. After the new screw has been satisfactorily placed, the removal sequence is reversed. The rod is slid back to engage the new screw head and all connections are retightened using the proper amount of torque. Removal of the tubes and closure of the incisions complete the procedure.
S1 screw was determined to be satisfactory. However, the L5 screw was shown to have breached the inferomedial cortex of its pedicle by approximately 3 mm (Fig. 3). As a remedy, the patient underwent minimally invasive removal of this hardware. Surgical time was 22 minutes, and blood loss was less than 10 cc. The patient was discharged the same day free of leg pain.
Illustrative cases
Case 2
Case 1
A 26-year-old man had fully recovered from posterolateral fusion and pedicle fixation at L5–S1 for back pain associated with a spondylolysis in 1996. He returned to heavy manual labor for 3 years when he developed acute low back pain when he twisted while lifting an 80-pound weight at shoulder level in January 2001. Despite extensive amounts of land- and aquatic-based physical therapy, his back pain persisted. He had also been aggressively treated with various injections and medication by pain management with no significant improvements. He was sent for surgical evaluation after a discogram was found to reproduce his typical pain at the L5–S1 level. Discs above this were noted to be structurally normal. X-rays showed a good lateral fusion mass. The hardware was in good position. Magnetic resonance imaging was degraded by artifact, but the saggital views defined loss of water content of the L5–S1 disc. Discs above this were normal. The various treatment scenarios were presented to the patient. He chose to avoid anterior lumbar interbody fusion and instead opted for a minimally invasive posterior lumbar interbody fusion [2], along with minimally invasive removal and reinsertion of pedicle screw fixation on one side (Fig. 4). Removal of the hardware took 30 minutes and was followed by a minimally invasive posterior lumbar interbody fusion through the incision used for the iliac crest graft and through which he was also reinstrumented. The patient was discharged the following morning. His incisions are shown in Fig. 5.
A 40-year-old man had previously undergone L5–S1 anterior lumbar interbody fusion with unilateral percutaneous pedicle screw fixation on the right for a work-related injury. He improved considerably after this procedure and was able to return to a Department of Labor classification medium duty work capacity 6 months after the procedure. Followup X-rays, including flexion and extension views, were consistent with a solid fusion. Approximately 18 months after surgery, the patient was involved in a motor vehicle accident after which he began to experience right leg pain radiating in an L5 distribution. Because of the persistence of this new symptom, magnetic resonance imaging was obtained to rule out herniation of the L4–L5 disc. No evidence of a new disc problem was found. Metallic artifact impaired the accuracy of the interpretation at the operated segment and, although no specific abnormality could be identified, a high-resolution, thin-cut computed tomography scan was obtained to check the integrity of the fusion and the exact positions of the screws. Computed tomography demonstrated trabecular bone that was contiguous with both end plates growing through and between the interbody cages. Position of the
Discussion
Fig. 3. Curiously, the violation of the lateral recess by this L5 screw did not produce symptoms in this patient for more than a year after the initial surgery.
There is a growing body of evidence that the extensive muscular dissection used for posterior lumbar fusions and the placement of lumbar pedicle screws produce significant clinical disadvantages [3–9]. This awareness has at least partly fueled an active search for newer, less traumatic methods [10,11]. The fruits of this labor have produced a commercially available percutaneous pedicle screw system [12]. Despite this, there has been no technical parallel governing pedicle screw removal. Although reports of clever surgical methods for extraction of broken screws abound [13–15], the technique of hardware exposure has remained unchanged. There are both absolute and relative indications for removal of pedicle screw constructs. However, removal of lumbar pedicle screw systems have required dissections at least as extensive as the traditional methods initially required for their insertion. Regardless of the various indications, the process of removal
A. A. Salerni / The Spine Journal 4 (2004) 701–705
705
Fig. 4. The patient whose intraoperative X-rays are displayed in this figure had persistent back pain that was discogram positive after solid posterolateral fusion and fixation for spondylolysis. He underwent removal of his hardware on the left using this minimally invasive method. This construct was then revised to an interbody fusion with iliac crest autograft and repeat fixation using another minimally invasive method. The X-ray on the left shows the tubular access port that was used to access the disc. The film on the right shows the completed interbody fusion and new titanium pedicle screw fixation.
has, consequently, been both labor intensive for the surgeon and, at the very least, disabling for patients. In many cases these disadvantages are substantial enough so as to nullify screw removal as a logical treatment option. Framed in the vernacular, the endeavor becomes a long run for a short slide. The minimally invasive technique described in this paper changes all this. The disadvantages associated with the standard open approach are reduced to the production of short-term mild discomfort and patient exposure to the potential risks of a brief period of anesthesia and surgical infection. Considering
Fig. 5. This photograph demonstrates the dramatic difference in incision sizes. The smaller incision above the posterior iliac spine on the left was used to remove the patient’s hardware. The larger one just below and lateral to this was used for graft harvesting, discectomy, fusion and insertion of new pedicle screw fixation.
that hospital stay should be limited to 1 day or less and that surgical times are less than 1 hour, minimally invasive removal or revision of hardware should reduce overall costs significantly. References [1] Foley KT, Smith MM. Microendoscopic discectomy. Tech Neurosurg 1997;3:301–7. [2] Salerni AA. A minimally invasive approach for posterior lumbar interbody fusion. Neurosurg Focus 2002;13(6):1–5. [3] Kawaguchi Y, Matsui H, Gejo R, Tsuji H. Back muscle injury after posterior lumbar spine surgery: a histological and enzymatic analysis. Spine 1996;21:941–4. [4] Kawaguchi Y, Yabuki S, Styf J, et al. Back muscle injury after posterior lumbar spine surgery topographic evaluation of intramuscular pressure and blood flow in the porcine back muscle during surgery. Spine 1996;21:2683–8. [5] Mayer TG, Vanharanta H, Gatchel RJ, et al. Comparison of CT scan muscle measurements and isokinetic trunk strength in postoperative patients. Spine 1989;14:33–6. [6] See DH, Kraft GH. Electromyography in paraspinal muscles following surgery for root compression. Arch Phys Med Rehabil 1975;56:80–3. [7] Sihoven T, Herno A, Paljarva L, Airaksinen O, Partanen J, Tapaninaho A. Local denervation atrophy of paraspinal muscles in postoperative failed back syndrome. Spine 1993;18:575–81. [8] Styf JR, Willen J. The effects of external compression by three retractors on pressure in the erector spine muscles during and after posterior lumbar spine surgery in humans. Spine 1998;23:354–8. [9] Wetzel FT, LaRocca H. The failed posterior lumbar interbody fusion. Spine 1991;16:839–45. [10] Lowery GL, Kulkarni SS. Posterior percutaneous spine instrumentation. Eur Spine J 2000;9(suppl 1):S126–30. [11] Muller A, Gall C, Marz U, Reulen HJ. A keyhole approach for endoscopically assisted pedicle screw fixation in lumbar spine instability. Neurosurgery 2000;47(1):85–95. [12] Foley KT, Gupta SK, Justis JR. Percutaneous pedicle screw fixation of the lumbar spine. Neurosurg Focus 2001;10(4):1–8. [13] DiLorenzo N, Conti R, Romoli S. Retrieval of broken pedicle screws by “friction” technique. Technical note. J Neurosurg 2000;92(1 suppl): 114–116. [14] Duncan JD, MacDonald JD. Extraction of broken pedicle screws: technical note. Neurosurgery 1998;42(6):1399–400. [15] Miyamoto K, Shimizu K, Kouda K, Hosoe H. Removal of broken pedicle screws. Technical note. J Neurosurg 2001;95(1 suppl):150–1.
Case Studies
Minimally invasive removal or revision of lumbar spinal fixation Anthony A. Salerni, MD* Orthopedic Professional Association, 14 Maple Street, Suite 100, Gilford, NH 03249, USA Received 22 August 2003; accepted 4 March 2004
Abstract
BACKGROUND CONTEXT: There are both absolute and relative indications for the removal of pedicle screw fixation in the lumbar spine. Whatever the reasons are, removal of this hardware has required a surgical dissection that has been generally as extensive as the one used for their initial placement. These dissections are always disabling in the short term. In fact, the magnitude of this disabling pain can be significant enough so as to effectively eliminate screw removal as a logical treatment option for many conditions where indications for removal are only relative. Percutaneous pedicle screw fixation has served to amplify the stakes associated with this dilemma. In fact, this new technique makes the need for a less invasive method of pedicle screw removal greater now than ever. PURPOSE: This paper describes a minimal access surgical technique for pedicle screw construct removal that employs the tubular retractor system that was originally developed for microendoscopic discectomy. STUDY DESIGN: This case study represents a summary of the surgical experience gained from the first 10 patients to have undergone removal or revision of pedicle screw constructs by this minimally invasive method. METHODS: A retrospective analysis of pre- and postoperative clinical data was gathered from the hospital records. Surgical times and blood loss were also extracted from these records. The procedure is described in detail. Interpretation of the surgical parameters and clinical effects are discussed. RESULTS: Six patients presented with a radiculopathy secondary to a misdirected pedicle screw. Two of these patients were admitted for simple removal. The four remaining patients who had undergone percutaneous pedicle screw fixation developed acute radicular pain from a misdirected screw. These patients underwent revision of their constructs by this method. Screws were also removed unilaterally in four other patients as the initial phase to revision or additional surgery. All procedures were performed through 16 mm tubular retractors. Operative time averaged 33 minutes for the group, and it ranged between 22 and 40 minutes. Hospital length of stay averaged 1 day for the group. Hospital stay averaged only 0.8 hospital days for the patients in whom screw removal was the primary goal. At 1 month after surgery no patient felt limited by incisional pain. No complications occurred. CONCLUSIONS: Unlike most other minimal access surgical procedures, the learning curve for this procedure appears to be relatively flat. Removal of pedicle screw fixation in the manner described proved to be simple and straightforward. The benefits are dramatic and immediate. It is possible to complete the procedure within minutes, and the pain produced is best described as inconsequential. This minimally invasive technique radically alters both the intraoperative and postoperative courses for those who face pedicle screw removal. The disadvantages associated with the standard open approach are reduced to the production of mild short-term discomfort and an exposure to the potential risks of brief anesthesia and the possibility of a surgical infection. Considering that hospital stay should be limited to 1 day or less and that surgical times are less than 1 hour, minimally invasive removal or revision of hardware should reduce overall costs significantly. 쑖 2004 Elsevier Inc. All rights reserved.
Keywords:
PLIF; Minimally invasive; Interbody fusion; Lumbar; Percutaneous surgery
FDA device/drug status: not applicable. Nothing of value received from a commercial entity related to this research. 1529-9430/04/$ – see front matter doi:10.1016/j.spinee.2004.03.023
쑖 2004 Elsevier Inc. All rights reserved.
* Orthopedic Professional Association, 14 Maple Street, Suite 100, Gilford, NH 03249, USA. Tel.: (603) 528-9100; fax: (603) 524-5743. E-mail address: [email protected] (A. A. Salerni)
702
A. A. Salerni / The Spine Journal 4 (2004) 701–705
Introduction Removal or revision of lumbar pedicle screw fixation can be clinically beneficial for a number of situations. The advantage of doing so, however, must be weighed against the cost of putting the patient through an operation that generally requires a dissection at least as extensive as the one required to fashion the construct in the first place. For such situations as radicular pain from a malpositioned screw, however, there may be no way to avoid a second disabling operation accompanied by a lengthy recovery phase. In other situations the cost–benefit ratio may lead the patient and surgeon to conclude that the procedure is best avoided. Although this may be the correct decision for most, the cost–benefit equation used to make this decision effectively eliminates a treatment option that could potentially be useful for some individuals. Regardless of the potential benefits, the pain associated with this procedure has always been significant. Consequently, a procedure that can minimize dissection would have enormous advantages. The tools designed for MED [1] are well suited for this application. Materials and methods The medical charts of the first 10 consecutive patients to have had removal of a lumbar pedicle screw were used to gather data for this retrospective analysis. The relevant clinical and surgical data for the patient population are summarized in Table 1. Six patients presented with a radiculopathy secondary to a misdirected pedicle screw. This was a lateterm development in two of these patients, even though their fusions had been documented to be solid. These two patients were admitted for simple removal. The four remaining patients, who had undergone percutaneous pedicle screw fixation, developed acute radicular pain from a misdirected screw. These patients underwent revision of their constructs by the method to be described. Screws were also removed unilaterally in four other patients as the initial phase to revision or additional surgery. One of these patients had a solid two-level lumbar interbody fusion, and he was admitted
for unilateral removal of his construct to gain access for a planed foramenotomy as a treatment for his radicular pain. Results All procedures were performed through 16 mm tubular retractors. Operative time averaged 33 minutes for the group (range, 22 to 40 minutes). Hospital length of stay averaged 1 day for the group. Hospital stay averaged only 0.8 hospital days for the patients in whom screw removal was the primary goal. At 1 month after surgery no patient felt limited by incisional pain. No complications occurred. Technique Although the screw systems and reasons for removal or revision of pedicle screws and rods vary, the manners by which they are removed do not. The procedure is carried out with loupe magnification and a headlight. With the patient in the prone position on rolls, a C-arm fluoroscope is moved to the anteroposterior (AP) view. In many cases, the long axis of the screw and the long axis through the center of the polyaxial screw head will define the same line. In this case the fluoroscope is repositioned to obtain an image that is shot directly down the long axis of the screw to be manipulated. In most cases this will be close to the AP projection. However, because some techniques require a more lateral insertion point, the projection may be considerably oblique. Either way, a spinal needle is inserted directly over the screw head, as seen on this projection. It is advanced parallel with the beam to the screw head, which, when reached, will provide a tactile signal distinct for metal. If the polyaxial screw head was not centered over the long axis of the screw when it was locked to the rod, the two extended lines defined by these respective axes will intersect with the skin at different points (Fig. 1). A point on the skin between the two extended lines is chosen as the entry point for the spinal needle with this presentation so that the tools necessary for removal of the locking device and the pedicle screw can be positively engaged through the same incision. A 14 mm or 16 mm MetRx tube (Medtronic Sofamor Danek, Memphis, TN) is
Table 1 Clinical and surgical data for patient population Sex
Age
Diagnosis
Operation
Levels
Complications
OR time (minutes)
LOS (days)
Result
M M F M M M F F M M
40 42 63 32 48 31 43 48 66 40 Average age, 45
Screw malposition Screw malposition Junctional stenosis Failed back Failed back Screw malposition Screw malposition Screw malposition Foramenal stenosis Screw malposition
Redirection Removal Removal Revision Revision Redirection Redirection Removal Removal Removal
L5–S1 L5–S1 L4–S1 L5–S1 L3–L5 L5–S1 L5–S1 L5–S1 L4–S1 L5–S1
None None None None None None None None None None
30 38 25 40 40 34 35 30 40 22 Average time, 33
1 1 2 1 1 1 1 1 1 0 Average LOS, 1.0
Improved Improved Improved Unchanged Unchanged Improved Improved Improved Improved Improved
LOS⫽hospital length of stay; OR⫽operating room.
A. A. Salerni / The Spine Journal 4 (2004) 701–705
Fig. 1. Ideally, a path directly along the long axis of the screw best facilitates its removal. However, the trajectory that allows for precise engagement of the screw head may be considerably different from this for polyaxial pedicle screws. Choosing a surgical trajectory between these two ideal paths will allow enough freedom to positively engage both the screw and rod locking mechanism for most systems.
centered over this needle and the edges of the skin on the superior and caudal edges of the tube are marked. This demarcates the maximum extents of the required incision. The tube and needle are removed and a saggitally oriented line is drawn with a length that is represented by the diameter of the tube (Fig. 2). An incision is made in the skin, and the fascia is opened with electrocautery. The smallest MetRx dilator is passed through the muscle and held against the screw head. Fluoroscopy is used if the screw head is not readily contacted. While firmly holding the small dilator against the screw head, successively larger dilators are passed over this. A MetRx tube is passed over the last dilators and ultimately over the screw head. The dilators are calibrated to aid in selecting a tube of proper length. Choosing a tube slightly longer than the distance from skin edge to screw head allows the screw head to be captured in the tube by simply applying downward pressure on the external end of the tube. Inspection after the dilators are removed will show the muscle- and scar-covered screw head projecting up into the tube. This tissue is easily removed with a disc ronguer. Cautery can be helpful, but it must be set to low power settings. Sweeping the tube along the construct will allow much of the tissue over the rod to be removed as well. The tools specific to the system to be manipulated can be inserted down the tube to unlock the rod from the screw. Once this is accomplished, another MetRx tube is then directed to the next screw to be removed. It is often useful to have access to both ends of the rod during disengagement from the screw heads so each MetRx tube is left in its position as the operation proceeds. Unlike discectomy, foramenotomy or fusion, the tubes do not need to be locked in
703
Fig. 2. The tube, which has been cut away in this figure, is centered on the needle that defines the optimum trajectory for the approach. The diameter of the tube in contact with the skin determines the maximum length of the incision necessary for engagement of the construct. It will be possible to remove many single-level constructs through one such incision.
place. In fact, because stripping the scar tissue covering the construct and rod extraction generally can require extensive pivoting of the tubes, keeping them free floating offers significant time saving advantage. Additional screws in the construct are addressed successively in the same manner, and there may be a tube in place for each. When the rod is unlocked from all the screw heads, it can be extracted. Even if all the scar tissue has been removed from the tops of the screw heads and rod, it still can take a significant force to disengage the rod. Once loose, however, it is relatively straightforward to leaver one end of the rod up through a tube for one-level constructs. Rods placed in multilevel constructs can be more difficult to remove. The longer the rod, the less apt it is to make the acute turn needed for extraction out an end tube. In situations such as this, a more oblique entry point is chosen and another small diameter MetRx tube is positioned on the rod at the top of the construct. The skin entry point is far enough cephalad so that the path of the tube makes a fairly obtuse angle with the coronal plane. Once the rod is disengaged from the screw head, it is just a matter of pushing the rod out the tube along this relatively flat trajectory. Once the rod has been removed, extraction of the screws is simply a matter of engaging the screw head with the system-specific screwdriver through the MetRx tubes stationed over each pedicle screw. If removal of the hardware was the objective, the procedure is complete, so after screw removal the tubes would be removed and each wound would be individually closed. In other cases, the surgical objective will be to revise rather than remove the hardware. Most often the goal will be to reposition a misdirected screw. In this case, the surgical plan proceeds exactly as for removal until the rod has been unlocked from all screw heads. The procedures diverge at this point and instead of disengaging and removing the rod, it is slid along the screw heads until it is free of the
704
A. A. Salerni / The Spine Journal 4 (2004) 701–705
malpositioned screw. This screw can be replaced through the tube using standard methods. After the new screw has been satisfactorily placed, the removal sequence is reversed. The rod is slid back to engage the new screw head and all connections are retightened using the proper amount of torque. Removal of the tubes and closure of the incisions complete the procedure.
S1 screw was determined to be satisfactory. However, the L5 screw was shown to have breached the inferomedial cortex of its pedicle by approximately 3 mm (Fig. 3). As a remedy, the patient underwent minimally invasive removal of this hardware. Surgical time was 22 minutes, and blood loss was less than 10 cc. The patient was discharged the same day free of leg pain.
Illustrative cases
Case 2
Case 1
A 26-year-old man had fully recovered from posterolateral fusion and pedicle fixation at L5–S1 for back pain associated with a spondylolysis in 1996. He returned to heavy manual labor for 3 years when he developed acute low back pain when he twisted while lifting an 80-pound weight at shoulder level in January 2001. Despite extensive amounts of land- and aquatic-based physical therapy, his back pain persisted. He had also been aggressively treated with various injections and medication by pain management with no significant improvements. He was sent for surgical evaluation after a discogram was found to reproduce his typical pain at the L5–S1 level. Discs above this were noted to be structurally normal. X-rays showed a good lateral fusion mass. The hardware was in good position. Magnetic resonance imaging was degraded by artifact, but the saggital views defined loss of water content of the L5–S1 disc. Discs above this were normal. The various treatment scenarios were presented to the patient. He chose to avoid anterior lumbar interbody fusion and instead opted for a minimally invasive posterior lumbar interbody fusion [2], along with minimally invasive removal and reinsertion of pedicle screw fixation on one side (Fig. 4). Removal of the hardware took 30 minutes and was followed by a minimally invasive posterior lumbar interbody fusion through the incision used for the iliac crest graft and through which he was also reinstrumented. The patient was discharged the following morning. His incisions are shown in Fig. 5.
A 40-year-old man had previously undergone L5–S1 anterior lumbar interbody fusion with unilateral percutaneous pedicle screw fixation on the right for a work-related injury. He improved considerably after this procedure and was able to return to a Department of Labor classification medium duty work capacity 6 months after the procedure. Followup X-rays, including flexion and extension views, were consistent with a solid fusion. Approximately 18 months after surgery, the patient was involved in a motor vehicle accident after which he began to experience right leg pain radiating in an L5 distribution. Because of the persistence of this new symptom, magnetic resonance imaging was obtained to rule out herniation of the L4–L5 disc. No evidence of a new disc problem was found. Metallic artifact impaired the accuracy of the interpretation at the operated segment and, although no specific abnormality could be identified, a high-resolution, thin-cut computed tomography scan was obtained to check the integrity of the fusion and the exact positions of the screws. Computed tomography demonstrated trabecular bone that was contiguous with both end plates growing through and between the interbody cages. Position of the
Discussion
Fig. 3. Curiously, the violation of the lateral recess by this L5 screw did not produce symptoms in this patient for more than a year after the initial surgery.
There is a growing body of evidence that the extensive muscular dissection used for posterior lumbar fusions and the placement of lumbar pedicle screws produce significant clinical disadvantages [3–9]. This awareness has at least partly fueled an active search for newer, less traumatic methods [10,11]. The fruits of this labor have produced a commercially available percutaneous pedicle screw system [12]. Despite this, there has been no technical parallel governing pedicle screw removal. Although reports of clever surgical methods for extraction of broken screws abound [13–15], the technique of hardware exposure has remained unchanged. There are both absolute and relative indications for removal of pedicle screw constructs. However, removal of lumbar pedicle screw systems have required dissections at least as extensive as the traditional methods initially required for their insertion. Regardless of the various indications, the process of removal
A. A. Salerni / The Spine Journal 4 (2004) 701–705
705
Fig. 4. The patient whose intraoperative X-rays are displayed in this figure had persistent back pain that was discogram positive after solid posterolateral fusion and fixation for spondylolysis. He underwent removal of his hardware on the left using this minimally invasive method. This construct was then revised to an interbody fusion with iliac crest autograft and repeat fixation using another minimally invasive method. The X-ray on the left shows the tubular access port that was used to access the disc. The film on the right shows the completed interbody fusion and new titanium pedicle screw fixation.
has, consequently, been both labor intensive for the surgeon and, at the very least, disabling for patients. In many cases these disadvantages are substantial enough so as to nullify screw removal as a logical treatment option. Framed in the vernacular, the endeavor becomes a long run for a short slide. The minimally invasive technique described in this paper changes all this. The disadvantages associated with the standard open approach are reduced to the production of short-term mild discomfort and patient exposure to the potential risks of a brief period of anesthesia and surgical infection. Considering
Fig. 5. This photograph demonstrates the dramatic difference in incision sizes. The smaller incision above the posterior iliac spine on the left was used to remove the patient’s hardware. The larger one just below and lateral to this was used for graft harvesting, discectomy, fusion and insertion of new pedicle screw fixation.
that hospital stay should be limited to 1 day or less and that surgical times are less than 1 hour, minimally invasive removal or revision of hardware should reduce overall costs significantly. References [1] Foley KT, Smith MM. Microendoscopic discectomy. Tech Neurosurg 1997;3:301–7. [2] Salerni AA. A minimally invasive approach for posterior lumbar interbody fusion. Neurosurg Focus 2002;13(6):1–5. [3] Kawaguchi Y, Matsui H, Gejo R, Tsuji H. Back muscle injury after posterior lumbar spine surgery: a histological and enzymatic analysis. Spine 1996;21:941–4. [4] Kawaguchi Y, Yabuki S, Styf J, et al. Back muscle injury after posterior lumbar spine surgery topographic evaluation of intramuscular pressure and blood flow in the porcine back muscle during surgery. Spine 1996;21:2683–8. [5] Mayer TG, Vanharanta H, Gatchel RJ, et al. Comparison of CT scan muscle measurements and isokinetic trunk strength in postoperative patients. Spine 1989;14:33–6. [6] See DH, Kraft GH. Electromyography in paraspinal muscles following surgery for root compression. Arch Phys Med Rehabil 1975;56:80–3. [7] Sihoven T, Herno A, Paljarva L, Airaksinen O, Partanen J, Tapaninaho A. Local denervation atrophy of paraspinal muscles in postoperative failed back syndrome. Spine 1993;18:575–81. [8] Styf JR, Willen J. The effects of external compression by three retractors on pressure in the erector spine muscles during and after posterior lumbar spine surgery in humans. Spine 1998;23:354–8. [9] Wetzel FT, LaRocca H. The failed posterior lumbar interbody fusion. Spine 1991;16:839–45. [10] Lowery GL, Kulkarni SS. Posterior percutaneous spine instrumentation. Eur Spine J 2000;9(suppl 1):S126–30. [11] Muller A, Gall C, Marz U, Reulen HJ. A keyhole approach for endoscopically assisted pedicle screw fixation in lumbar spine instability. Neurosurgery 2000;47(1):85–95. [12] Foley KT, Gupta SK, Justis JR. Percutaneous pedicle screw fixation of the lumbar spine. Neurosurg Focus 2001;10(4):1–8. [13] DiLorenzo N, Conti R, Romoli S. Retrieval of broken pedicle screws by “friction” technique. Technical note. J Neurosurg 2000;92(1 suppl): 114–116. [14] Duncan JD, MacDonald JD. Extraction of broken pedicle screws: technical note. Neurosurgery 1998;42(6):1399–400. [15] Miyamoto K, Shimizu K, Kouda K, Hosoe H. Removal of broken pedicle screws. Technical note. J Neurosurg 2001;95(1 suppl):150–1.