SPINE Volume 34, Number 13, pp E473–E477 ©2009, Lippincott Williams & Wilkins
Epidural Spinal Cord Compression With Neurologic Deficit Associated With Intrapedicular Application of Hemostatic Gelatin Matrix During Pedicle Screw Insertion Jacob M. Buchowski, MD, MS, Keith H. Bridwell, MD, Lawrence G. Lenke, MD, and Christopher R. Good, MD
Study Design. Case report. Objective. In order to demonstrate the dangers of intrapedicular application of a hemostatic gelatin matrix to decrease blood loss during pedicle screw insertion, we present 2 patients who—as a result of inadvertent extravasation of the matrix into the spinal canal— developed epidural spinal cord compression (ESCC) requiring emergent decompression. Summary of Background Data. Variety of hemostatic agents can control bleeding during pedicle screw insertion. We have often used a hemostatic gelatin matrix to decrease bleeding from cannulated pedicles by injecting the material into the pedicle after manually palpating the pedicle. Methods. Medical records and radiographic studies of 2 patients with AIS who underwent surgical treatment of their deformity and developed a neurologic deficit due to extravasation of FloSeal were reviewed. Results. A 15 year-old male underwent T4 to L2 posterior spinal fusion (PSF). During pedicle screw insertion, a change in NMEPs and SSEPs was noted. A wake-up test confirmed bilateral LE paraplegia. Screws were removed and no perforations were noted on manual palpation. MRI showed T7 to T10 ESCC. He underwent a T5 to T10 laminectomy and hemostatic gelatin matrix noted in the canal and was evacuated. He was ambulatory at 2 weeks and by 3 months he had complete recovery. The second patient was a 15 year-old female who underwent T4 to L1 PSF. Following screw insertion, deterioration in NMEPs and SSEPs was noted. Screws were removed and SCM data returned to baseline. Except for 3 screws that had an inferior breach (Left T7 and Bilateral T8), screws were reinserted and remainder of the surgery was uneventful. Postoperative examination was normal initially but 2 days later, she developed left LE numbness/weakness. Implants were removed and MRI showed T4 to T9 ESCC.She underwent a left (concave) T4 to T9 hemilaminectomy. Hemostatic gelatin matrix was noted and was evacuated.
From the Department of Orthopaedic Surgery, Washington University, St. Louis, MO. Acknowledgment date: November 11, 2008. Revision date: January 12, 2009. Acceptance date: January 15, 2009. The manuscript submitted does not contain information about medical device(s)/drug(s). No funds were received in support of this work. No benefits in any form have been or will be received from a commercial party related directly or indirectly to the subject of this manuscript. The devices described in this report are FDA-approved, but not for this indication. Address correspondence and reprint requests to: Jacob M. Buchowski, MD, MS, Department of Orthopaedic Surgery, Washington University in St. Louis, 660 S. Euclid Ave., Campus Box 8233, St. Louis, MO 63110; E-mail: buchowskij@wustl.edu
Six weeks following surgery, she had a complete neurologic recovery. Conclusions. The use of a hemostatic gelatin matrix to decrease bleeding from cannulated pedicles during pedicle screw insertion can result in inadvertent extravasation into the spinal canal resulting in ESCC even in the absence of an apparent medial pedicle breach. Given the dangers associated with the technique, we recommend that gelatin matrix products be used judiciously during pedicle screw insertion. Key words: hemostatic gelatin matrix, epidural spinal cord compression (ESCC), extravasation. Spine 2009;34: E473–E477
A wide variety of systemic and topical hemostatic agents are frequently used in spinal surgery to control intraoperative and postoperative bleeding. These agents range from systemic agents such as -aminocaproic acid (EACA), Aprotinin, desmopressin, and tranexamic acid, as well topical agents such as bone wax, hemostatic sponges, gelatin matrix products (usually mixed with thrombin), and fibrin sealants.1–15 These agents can be useful throughout the length of surgery, but particularly during the portions of the procedure during which there is increased blood loss such as during pedicle screw insertion. We have often used hemostatic gelatin matrix (e.g., FloSeal [Baxter, Inc., Deerfield, IL] and Surgiflow [Johnson and Johnson, Somerville, NJ]) to decrease bleeding from cannulated pedicles by injecting the hemostatic matrix material into the pedicle after manually palpating the pedicle to make sure that no perforations were created either during initial cannulation of the pedicle or subsequently during screw tract preparation before screw insertion (e.g., tapping of the screw tract). In order to demonstrate the dangers of this technique, we present 2 patients who—as a result of inadvertent extravasation of the matrix into the spinal canal— developed epidural spinal cord compression (ESCC) requiring emergent decompression. Case Reports Case 1 A 15-year-old male presented with progressive adolescent idiopathic scoliosis unresponsive to bracing. He had a Lenke 1AN curve with a right thoracic curve measuring approximately 47° from T4 –T12. Given the magnitude E473
E474 Spine • Volume 34 • Number 13 • 2009
of the deformity and its progressive nature, the patient was taken to the operating room for a posterior spinal fusion with pedicle screw-based instrumentation from T4 to L2. Spinal cord monitoring consisting of neurogenic mixed evoked potentials (NMEPs) and somatosensory-evoked potentials (SSEPs) was used during the procedure. Baseline spinal cord monitoring data were unremarkable both before and following exposure of the spine. Following exposure of the spine, pedicle screws were inserted. No wide releases or Ponte osteotomies were done. The pedicle screws were inserted using the modified free-hand technique. Briefly, anatomic landmarks were used to determine the starting point, which was then made with a burr. The pedicle probe was then used to create a tract through the pedicle into the vertebral body using the straight-ahead trajectory. A flexible ball-tipped sounder was used to make sure that no perforations were created by carefully palpating the pedicle medially, laterally, superiorly, and inferiorly. If no perforations were noted, commercially available hemostatic gelatin matrix was then injected into the pedicle using a syringe with an attached applicator tip in order to decrease the bleeding from the cannulated pedicle. The appropriately-sized tap was then used to prepare the screw tract. The flexible ball-tipped sounder was then again used to palpate the pedicle, and if no perforations were noted, additional hemostatic matrix material was injected and then the appropriately-sized pedicle screw was inserted. On average, approximately 1.0 to 1.5 mL of hemostatic matrix material was used per level. Pedicle screws were inserted beginning on the left side at L2 and progressing proximally. As the upper thoracic pedicle screws were being inserted, a change in NMEPs and SSEPs was noted by the spinal cord monitoring team and the surgeon was notified. An intraoperative Stagnara wake-up test was then performed and confirmed the presence of bilateral lower extremity paraplegia. All pedicle screws were removed and the flexible ball-tipped sounder was used to palpate all screw tracts. No pedicle perforations were noted. An emergent MRI was obtained and revealed evidence of epidural spinal cord compression from T7 to T10 (Figure 1A). The patient was taken to the operating room emergently and underwent a T5 to T10 laminectomy. In agreement with the preoperative study, intraoperative exploration revealed the presence of hemostatic gelatin matrix material in the spinal canal from T7 to T10 which was evacuated. Fortunately, with appropriate intervention, the patient was ambulatory at 2 weeks and by 4 months he had complete recovery of neurologic function. A CT scan obtained following the decompression demonstrated that despite anatomic placement of the left T10 pedicle screw, a perforation in the medial cortex of the pedicle was created during screw insertion (Figure 1B). It is likely that this perforation allowed the hemostatic matrix material to enter the spinal canal.
Case 2 A 15-year-old female presented with progressive adolescent idiopathic scoliosis unresponsive to bracing. She had a Lenke 1CN curve with a right thoracic curve measuring 57° from T5 to T10 and a compensatory curve measuring 49° from T10 to L4 (which corrected to 12° on side bending). She was taken to the operating room for a posterior spinal fusion with pedicle screw-based instrumentation from T4 to L1. Spinal cord monitoring consisting of neurogenic mixed evoked potentials (NMEPs) and somatosensory-evoked potentials (SSEPs) was used during the procedure. As in case 1, baseline spinal cord monitoring data were unremarkable both before and following exposure of the spine. Following exposure of the spine, pedicle screws were inserted bilaterally using the modified free-hand technique described above using a hemostatic gelatin matrix to decrease the bleeding from the cannulated pedicles. The periapical screws in the concavity of the curve (T6, T7, and T8 on the left) were inserted in a juxtapedicular manner given the small size of the pedicles at these levels. No wide releases or Ponte osteotomies were done. Following screw insertion, a deterioration in NMEPs and SSEPs was noted. The screws were removed and spinal cord monitoring data returned to baseline. Each screws tract was palpated and 3 screw tract were found to have an inferior breach: T7 on the left and T8 bilaterally. There were no perforations in the medial cortex or the ventral lamina at any of the instrumented levels. With the exception of these 3 screw tracts, the screws were reinserted and the remainder of the surgery was uneventful. Postoperative examination was normal initially following surgery, however, 2 days following the procedure the patient developed left lower extremity numbness/ weakness. The patient was emergently taken to the operating room where she underwent implant removal. She subsequently underwent an MRI which demonstrated evidence of epidural spinal cord compression from T4 to T9 (Figure 2). Given these results, she returned to the operating room, where she underwent a hemilaminectomy from T4 to T9 on the left (i.e., the concavity of the curve). Hemostatic gelatin matrix material was noted and evacuated and pedicle screws were reinserted. Despite perioperative left lower extremity weakness and numbness, by 6 weeks after surgery, the patient had a complete neurologic recovery. Discussion Various systemic and topical hemostatic agents are available. Systemic agents include: -aminocaproic acid (EACA), Aprotinin, desmopressin, and tranexamic acid, all of which have been used successfully in spinal operations that carry a risk of significant bleeding (such as thoracolumbar reconstructive surgery).3–11 These agents may result in complications (including deep venous thrombosis, pulmonary embolism, renal failure, and severe bradycardia),3–11 are expensive, and often are unnecessary in smaller spinal surgeries where the probabil-
Epidural Spinal Cord Compression due to Hemostatic Gelatin Matrix • Buchowski et al E475
Figure 1. Fifteen-year-old boy presented with progressive adolescent idiopathic scoliosis. He underwent a posterior spinal fusion with from T4 to L2. During pedicle screw insertion, deterioration in spinal cord monitoring occurred and bilateral lower extremity paraplegia was noted on intraoperative Stagnara wake-up test. MRI revealed evidence of epidural spinal compression from T7 to T10 as seen on T2-weighted images (A). Given these findings, patient underwent an emergent T5 to T10 laminectomy with eventual complete neurologic recovery. A CT scan obtained following the decompression demonstrated that despite appropriate placement of the left T10 pedicle screw, a perforation in the medial cortex of the pedicle was created during screw insertion (B) and it is likely that this perforation allowed the hemostatic gelatin matrix material to enter the spinal canal.
ity of major bleeding is low. Compared to systemic hemostatic agents, topical hemostatic agents have a greater role in spinal surgery as they can be very useful in smaller procedure where systemic hemostatic agents have a limited role. A wide variety of these hemostatic agents exist and most can be used to control intraoperative bleeding during spinal surgery.1,2 The most common topical hemostatic agents used in spine surgery include bone wax (Ethicon, Norderstedt, Germany), hemostatic sponges, gelatin matrix products (typically mixed with thrombin), and fibrin sealants.3,12 Hemostatic sponges may be either gelatin-based (e.g., Gelfoam [Pharmacia and Upjohn, Kalamazoo, MI]), collagen-based (e.g., Instat [Johnson and Johnson, Somerville, NJ]), or oxidized cellulose-based (e.g., Surgicel [Johnson and Johnson, Somerville, NJ]). Gelatin matrix products include
FloSeal (Baxter, Inc., Deerfield, IL) and Surgiflow (Johnson and Johnson, Somerville, NJ). Fibrin sealants, also known as fibrin glues, such as Tisseal (Baxter Healthcare Corp., Westlake Village, CA) consist of fibrinogen and thrombin that form a fibrin clot.1,3,13–15 In our practice, we have found hemostatic gelatin matrix products (FloSeal [Baxter, Inc., Deerfield, IL] and Surgiflow [Johnson and Johnson, Somerville, NJ]) to be particularly helpful in achieving hemostasis during surgery, especially during pedicle screw insertion. We have found that after thoroughly palpating the pedicle using a flexible ball-tipped sounder to make sure that no perforations were created during the pedicle screw tract preparation, injection of a hemostatic gelatin matrix into the cannulated pedicle—both after initial cannulation of the pedicle and after the pedicle has been tapped— controls
E476 Spine • Volume 34 • Number 13 • 2009
Figure 2. Fifteen year-old girl presented with progressive adolescent idiopathic scoliosis. She underwent a posterior spinal fusion with from T4 to L1. Following pedicle screw insertion, deterioration in spinal cord monitoring occurred. Screws were removed and spinal cord monitoring data returned to normal. With the exception of 3 screws (T7 on the left and T8 bilaterally) which had an inferior breach, screws were reinserted and the remainder of the surgery was uneventful. Two days following surgery, the patient developed left lower extremity numbness/weakness. She was emergently taken to the operating room where she underwent implant removal. She subsequently underwent an MRI which demonstrated evidence of epidural spinal cord compression from T4 to T9 as seen on T2-weighted images. Given these results, she underwent a hemi-laminectomy from T4 to T9 on the left (i.e., the concavity of the curve). Hemostatic gelatin matrix material was noted and evacuated and pedicle screws were reinserted. Fortunately, the patient had a complete neurologic recovery.
the blood loss associated with pedicle screw insertion. Although the technique has proven to be helpful in reducing blood loss associated with pedicle screw insertion, it’s not without its dangers. Gelatin matrix material swells by up to 20% following application (the maximum swell volume is usually achieved within 10 minutes of application), and, therefore if the material is placed in an area where such swelling can exert pressure on the neural elements, potential problems may develop. As prior studies have demonstrated, even in the best of hands, thoracic pedicle screw insertion is not trivial and anywhere from 3% to 42% of pedicle screws will have a detectable breach on CT analysis.16 –29 Although these perforations are often of little clinical significance in that they rarely lead to a neurologic deficit, such perforation can lead to development of a neurologic deficit or loss of fixation. Furthermore, if the hemostatic gelatin matrix technique described in this report is used in the setting of a medial pedicle breach, the hemostatic matrix material may be injected into the spinal canal potentially leading to compression of the neural elements, which if severe enough, can result in a neurologic deficit. As the pedicles in the concavity of the curve tend to be smaller in nature and more challenging to instrument than pedicles in a “straight” spine and as the spinal cord drifts away from the midline toward the pedicles in the concavity of the
curve, given the proximity of the spinal cord medial perforations of the pedicles in the concavity of the curve and extravasation of hemostatic matrix material through those pedicles (or through the foramens as likely occurred in case 2) is more dangerous and more likely to lead to development of a neurologic deficit (compared to, for example, pedicles in the convexity of the curve). It is precisely for these reasons that we carefully palpate the pedicle using a flexible ball-tipped sounder after each step of pedicle screw tract preparation before injecting the hemostatic gelatin matrix into the cannulated pedicle, especially in the convexity of the curve. Although palpation of the pedicle with ball-tipped sounder is able to detect most pedicle perforations, the technique is not perfect: in fact, a study by Lehman RA et al suggest that palpation of the pedicle may miss a pedicle perforation 19% to 50% of the time depending on surgeon experience.30 No pedicle breach was identified on pedicle palpation in case 1 in this report, demonstrating that even in the absence of a pedicle breach significant enough to be detected on manual palpation, the material can be injected into the spinal canal, especially, if it is injected into the pedicle under pressure. Indeed, a CT scan obtained following the decompression demonstrated a small perforation in the medical cortex of the left T10 pedicle despite anatomic screw trajectory/ insertion. Even though this perforation was not significant enough to be noticeable on manual palpation of the pedicle, we postulate that the hemostatic gelatin matrix entered the spinal canal through the perforation. Case 2 demonstrates that even in the absence of a medial breach (which would obviously be the most common pathway by which the material could enter into the spinal canal), the hemostatic gelatin matrix may reach the spinal canal and cause inadvertent epidural spinal cord compression. We suspect that the material entered the spinal through the foramen at least at 1 level where a screw was placed in a juxtapedicular manner. Although epidural spinal cord compression caused by the gelatin matrix material may lead to symptoms shortly after injection as occurred in case 1, symptoms may also develop in a delayed fashion (likely due to continued expansion of the matrix material over time, although typically maximum swell volume is achieved within 10 minutes of application) as demonstrated in case 2. In summary, the risk of creating a breach within the pedicle during pedicle screw insertion ranges from 3% to 42% and anywhere from 19% to 50% of these perforations may not be detected intraoperatively by palpation. If such a breach is not discovered and hemostatic gelatin matrix is used during pedicle screw insertion, inadvertent epidural spinal cord compression may occur, especially if the material is injected under pressure. Nevertheless, despite these risks, the use of these materials can be very helpful as pedicle screw insertion can often be associated with significant blood loss as high as approximately 200 mL per screw in some patients. Given the dangers, however, if one decides to use these hemostatic gelatin matrix
Epidural Spinal Cord Compression due to Hemostatic Gelatin Matrix • Buchowski et al E477
products during pedicle screw insertion we recommend that they be used judiciously. More specifically, we recommend limiting the amount of the gelatin matrix injected to approximately 0.5 to 1.0 mL per pedicle. Not using anything at all to control bleeding during pedicle screws insertion or using alternatives such as bone wax or gel foam to plug the pedicle tract superficially is also an option. In addition, injecting the material under pressure (i.e., pressurizing the pedicle) should be avoided and careful attention must be paid when the material is used during juxtapedicular screw insertion. Key Points ● Significant blood loss may occur during pedicle screw tract preparation. Injection of a hemostatic gelatin matrix (e.g., FloSeal [Baxter, Inc., Deerfield, IL] and Surgiflow [Johnson and Johnson, Somerville, NJ]) into the cannulated pedicle after thoroughly palpating the pedicle using a flexible balltipped sounder controls the blood loss associated with pedicle screw insertion. ● Although the technique has proven to be helpful in reducing blood loss associated with pedicle screw insertion, it’s not without its dangers as based on prior studies, the risk of creating a breach within the pedicle during pedicle screw insertion ranges from 3% to 42% and 19% to 50% of these perforations may not be detected intraoperatively by palpation, raising the possibility that the matrix material may be injected into the spinal canal. ● To demonstrate the dangers associated with the technique, we present 2 cases where the intrapedicular application of hemostatic gelatin matrix resulted in epidural spinal cord compression and neurologic deficit. In both cases, an urgent decompression and evacuation of the matrix material was required, but with the appropriate treatment, both patients fully recovered.
References 1. Block JE. Severe blood loss during spinal reconstructive procedures: the potential usefulness of topical hemostatic agents. Med Hypotheses 2005;65: 617–21. 2. Renkens KL Jr, Payner TD, Leipzig TJ, et al. A multicenter, prospective, randomized trial evaluating a new hemostatic agent for spinal surgery. Spine 2001;26:1645–50. 3. Szpalski M, Gunzburg R, Sztern B. An overview of blood-sparing techniques used in spine surgery during the perioperative period. Eur Spine J 2004; 13(suppl 1):18 –27.
4. Bednar DA, Bednar VA, Chaudhary A, et al. Tranexamic acid for hemostasis in the surgical treatment of metastatic tumors of the spine. Spine 2006;31: 954 –7. 5. Bitan FD. Aprotinin in spine surgery: review of the literature. Orthopedics 2004;27(suppl 6):681–3. 6. Cole JW, Murray DJ, Snider RJ, et al. Aprotinin reduces blood loss during spinal surgery in children. Spine 2003;28:2482–5. 7. Hu SS. Blood loss in adult spinal surgery. Eur Spine J 2004;13(suppl 1):3–5. 8. Lentschener C, Cottin P, Bouaziz H, et al. Reduction of blood loss and transfusion requirement by aprotinin in posterior lumbar spine fusion. Anesth Analg 1999;89:590 –7. 9. Neilipovitz DT. Tranexamic acid for major spinal surgery. Eur Spine J 2004; 13(suppl 1):62–5. 10. Urban MK, Beckman J, Gordon M, et al. The efficacy of antifibrinolytics in the reduction of blood loss during complex adult reconstructive spine surgery. Spine 2001;26:1152– 6. 11. Okubadejo GO, Bridwell KH, Lenke LG, et al. Aprotinin may decrease blood loss in complex adult spinal deformity surgery, but it may also increase the risk of acute renal failure. Spine 2007;32:2265–71. 12. Schonauer C, Tessitore E, Barbagallo G, et al. The use of local agents: bone wax, gelatin, collagen, oxidized cellulose. Eur Spine J 2004; 13(suppl 1):89 –96. 13. Hida K, Yamaguchi S, Seki T, et al. Nonsuture dural repair using polyglycolic acid mesh and fibrin glue: clinical application to spinal surgery. Surg Neurol 2006;65:136 – 42. 14. Patel MR, Caruso PA, Yousuf N, et al. CT-guided percutaneous fibrin glue therapy of cerebrospinal fluid leaks in the spine after surgery. AJR 2000;175: 443– 6. 15. Patel MR, Louie W, Rachlin J. Postoperative cerebrospinal fluid leaks of the lumbosacral spine: management with percutaneous fibrin glue. Am J Neuroradiol 1996;17:495–500. 16. Boos N, Webb JK. Pedicle screw fixation in spinal disorders: a European view. Eur Spine J 1997;6:2–18. 17. Kim YJ, Lenke LG, Bridwell KH, et al. Free hand pedicle screw placement in the thoracic spine: is it safe? Spine 2004;29:333– 42. 18. Brown CA, Lenke LG, Bridwell KH, et al. Complications of pediatric thoracolumbar and lumbar pedicle screws. Spine 1998;23:1566 –71. 19. Cotrel Y, Dubousset J, Guillaumat M. New universal instrumentation in spinal surgery. Clin Orthop 1988;227:10 –23. 20. Hamill CL, Lenke LG, Bridwell KH, et al. The use of pedicle screw fixation to improve correction in the lumbar spine of patients with idiopathic scoliosis. Is it warranted? Spine 1996;21:1241–9. 21. Liljenqvist UR, Halm HF, Link TM. Pedicle screw instrumentation of the thoracic spine in idiopathic scoliosis. Spine 1997;22:2239 – 45. 22. Lonstein JE, Denis F, Perra JH, et al. Complications associated with pedicle screws. J Bone Joint Surg Am 1999;81:1519 –28. 23. West JL, Ogilvie JW, Bradford DS. Complications of variable screw plate pedicle screw fixation. Spine 1991;16:577–9. 24. Belmont PJ, Klemme WR, Dhawan A, et al. In vivo accuracy of thoracic pedicle screws. Spine 2001;26:2340 – 6. 25. Kim YJ, Lenke LG, Bridwell KH, et al. CT scan accuracy of “free hand” thoracic pedicle screw placement in pediatric spinal deformity. Poster presented at: Scoliosis Research Society Annual Meeting; September 2001; Cleveland, OH. 26. Suk SI, Kim WJ, Lee SM, et al. Thoracic pedicle screw fixation in spinal deformities: are they really safe? Spine 2001;26:2049 –57. 27. Suk SI, Lee CK, Kim WJ, et al. Segmental pedicle screw fixation in the treatment of thoracic idiopathic scoliosis. Spine 1995;20:1399 – 405. 28. Youkilis AS, Quint DJ, McGillicuddy JE, et al. Stereotactic navigation for placement of pedicle screws in the thoracic spine. Neurosurgery 2001;48: 771–9. 29. Gertzbein SD, Robbins SE. Accuracy of pedicular screw placement in vivo. Spine 1990;15:11– 4. 30. Lehman RA, Potter BK, Kuklo TR, et al. Probing for thoracic pedicle screw tract violation(s): is it valid? J Spinal Disord Tech 2004;17:277– 83.