Visceral and vascular complications resulting from anterior lumbar interbody fusion

Transcription

1 J Neurosurg (Spine 1) 91:60 64, 1999 Visceral and vascular complications resulting from anterior lumbar interbody fusion VISWANATHAN RAJARAMAN, M.D., F.R.C.S., ROY VINGAN, M.D., PATRICK ROTH, M.D., ROBERT F. HEARY, M.D., LISA CONKLIN, R.N., AND GEORGE B. JACOBS, M.D. Division of Neurosurgery, New Jersey Medical School, Newark, New Jersey; and Department of Neurosurgery, Hackensack University Medical Center, Hackensack, New Jersey Object. The literature on abdominal and general surgery related complications following anterior lumbar interbody fusion (ALIF) is scant. In this retrospective review of 60 patients in whom ALIF was performed at their institutions between 1996 and 1998, the authors detail the associated complications and their correlation with perioperative factors. The causes, strategies for their avoidance, and the clinical course of these complications are also discussed. Methods. The study group was composed of 31 men and 29 women whose mean age was 42 years (range years). The preoperative diagnosis was discogenic back pain in 33 patients (55%); failed back syndrome in 11 (18.3%); pseudarthrosis in five (8.3%); postlaminectomy syndrome in four (6.6%); spondylolisthesis in three (5%); burst fracture in two (3.3%); and malignancy in two (3.3%). A retroperitoneal approach to the spine was used in 57 of the 60 patients. One interspace was exposed in 28 patients (46.6%), two in 28 (46.6%), and three in four (6.6%). Discectomy and interbody fusion in which the authors placed titanium cages or bone dowels was performed in 56 patients and corpectomy with instrumentation in four. Seven (11.6%) of 60 patients had undergone previous abdominal surgery and 29 (48.3%) had undergone previous spinal surgery. The follow-up period averaged 12 4 months (mean standard deviation). Twenty-four general surgery related complications occurred in 23 patients (38.3%), including sympathetic dysfunction in six; vascular injury in four; somatic neural injury in three; sexual dysfunction in three; prolonged ileus in three; wound incompetence in two; and deep venous thrombosis, acute pancreatitis, and bowel injury in one patient each. There were no deaths. The incidence of complications was not associated with underlying diagnosis (p 0.1), age (p 0.5), previous abdominal or spinal surgery (p 0.1), or the number of levels exposed (p 0.1). Conclusions. This report provides a detailed analysis of the general surgery related complications following ALIF. Although many of these complications have been recognized in the literature, the significance of sympathetic dysfunction appears to have been underestimated. The high incidence of complications in this series likely reflects the strict criteria. Many of these complications were minor and resolved over time without long-term sequelae. KEY WORDS lumbar fusion postoperative complications retroperitoneal approach venous injury sympathetic dysfunction A N anterior approach to the lumbar spine was first described for the management of spondylolisthesis and Pott s disease in the 1930s. 14,20,24 Subsequently, posterior and anterior lumbar interbody fusion (ALIF) were used by orthopedic and neurological surgeons for the management of degenerative and traumatic spinal instability. 4,7,12,25 In the last few years there has been a rapid increase in the use of ALIF for the management of discogenic pain, degenerative instability, and pseudarthrosis. Its popularity is due, in large measure, to the recent availability of interbody fusion adjunctive devices such as cages and bone dowels. In the ALIF procedure a general or vascular surgeon usually creates the exposure, and in most instances, their involvement is restricted to the operating room and the early postoperative period. Therefore, to be able to counsel patients preoperatively and provide optimum follow-up care, 60 the spine surgeon should be aware of the potential abdominal and general surgical complications in the perioperative period. The purpose of this paper is to increase awareness of the possible general surgery related complications, to document their incidence, causes, and clinical course, and to outline avoidance strategies by providing a retrospective review of consecutive cases of ALIF performed at our institutions. Clinical Material and Methods We first began to perform ALIF by using cages and bone dowels in September 1996 at our two institutions. A retrospective study of 60 consecutive anterior approaches for ALIF, performed by experienced staff surgeons for four neurosurgeons, was undertaken. Computerized databases, hospital records, operative notes, and office charts

2 Anterior lumbar interbody fusion of all cases treated at our institutions from July 1996 to June 1998 were reviewed for all vascular, visceral, autonomic, and other complications related to the surgical approach used in these operations. The data recorded and transferred into a customized database included demographics, diagnosis, previous spinal and/or abdominal surgery, operative approach, number of levels exposed, procedure performed, implants used, peri- and/or postoperative complications, length of hospital stay (LOS), and followup course. Complications related to the spinal surgery, bone graft, and donor graft site were not included. Operative Approach With the patient supine, a retroperitoneal approach to the lumbar spine was made through a left paramedian vertical or oblique skin incision below the umbilicus. A lumbar support was placed to maintain lordosis. The subcutaneous tissues were divided using electrocautery to expose the anterior rectus sheath, and the sheath was divided longitudinally along the direction of its fibers and the epigastric vessels divided as needed. The muscle was retracted medially and the posterior rectus sheath and the transversalis fascia divided as required. The peritoneum and its contents were retracted medially by blunt dissection to expose the iliopsoas muscle and the lumbar spine. The ureter was swept medially along with the peritoneum and retracted medially. An Omni/radiolucent self-retaining retractor was then placed to keep the abdominal contents retracted. The left common iliac artery and vein were identified and traced to their bifurcation. The iliolumbar vein and middle sacral vessels were ligated and divided as needed. The distal aorta, inferior vena cava, and the iliac vessels were mobilized to the right, exposing the lumbar spine and disc spaces. Lumbar segmental arteries and veins were then ligated and divided as needed. During the neurosurgical portion of the procedure, hand-held retractors were used to retract vascular structures, and the pressure on the vessels was released at periodic intervals. Electrocautery was avoided during the dissection of the prevertebral soft tissue. On completion of the ALIF, the general surgeon returned to perform the closure. Iliac crest autografts, when used for fusion, were obtained either via a separate skin incision (right or left side) or via a separate fascial incision from within the abdominal wall. A lateral position and a muscledividing incision were used in performing corpectomy. A transperitoneal approach to the lumbar spine was used in three patients. Clinical Data The study group comprised 31 men and 29 women with a mean age of 42 years (range years). The frequency of different preoperative diagnoses is detailed in Table 1. Degenerative conditions of the lumbar spine accounted for more than 90% of cases. Nearly half the study population (29 of 60 patients) had undergone previous lumbar spinal surgery, with an average of 1.9 operations per patient. Seven patients (11.6%) had undergone previous abdominal surgery that included cholecystectomy, appendectomy, gastric bypass, hernia repair, and cesarean section. A retroperitoneal approach to the spine was performed in 57 of 60 patients. One interspace was exposed in 28 (46.6%), two in 28 (46.6%), and three in four (6.6%) of the 60 patients. TABLE 1 Preoperative diagnoses in 60 patients who underwent ALIF Diagnosis No. of Patients (%) discogenic back pain 33 (55) failed back syndrome 11 (18.3) pseudarthrosis 5 (8.3) postlaminectomy syndrome 4 (6.6) spondylolisthesis 3 (5.0) burst fracture 2 (3.3) malignancy 2 (3.3) Discectomy and interbody fusion was performed in 56 patients, and four patients underwent corpectomy. The number of different lumbar segments fused and the type of implant used are detailed in Table 2. A concomitant posterior anterior approach consisted of placement of instrumentation and posterolateral fusion in three patients and decompressive surgery in one patient. Four patients who did not improve following ALIF underwent a delayed procedure for posterior fusion. The average LOS for all patients was 4.5 days (range 2 21 days). The follow-up period averaged 12 4 months (mean standard deviation). Six-month follow-up data were available for 54 (90%) of the 60 patients. Statistical Analysis The differences observed between the groups were analyzed using the chi-square test and the Wilcoxon signed rank sum test as appropriate. Significance was established at a probability value of Results The incidences of the different complications are detailed in Table 3. There were no perioperative deaths in this series. The overall incidence of general surgical complications was 40%. Intraoperative complications included one major and three minor venous injuries. One minor vessel injury occurred at the time of graft placement, and the rest occurred when dissection was being performed to obtain exposure. All venous injuries were repaired using No. 6-0 prolene sutures. One patient sustained a colonic injury as the peritoneum was being bluntly dissected to define the retroperitoneal space. The 3-mm hole in the sigmoid was immediately recognized, the operative was field isolated, and the laceration was repaired. In the early postoperative period, one patient underwent reoperation for closure of an abdominal wound dehiscence. Another patient sustained a superficial wound dehiscence that was allowed to heal by secondary intention. For three patients the LOS was prolonged ( 7 days) because of persistent ileus. In one patient who developed ileus there was laboratory evidence of acute pancreatitis. In the late postoperative period (2 months postsurgery) one patient required placement of a filter in the vena cava and a short course of heparin therapy to treat deep vein thrombosis. There was symptomatic evidence of sympathetic nerve damage in six patients. Postoperatively in two patients, the right lower extremity was believed to be 61

3 V. Rajaraman, et al. TABLE 2 Lumbar segments and implants used for ALIF in 60 patients* Type of Implant Lumbar No. of BAK Bone Femoral segments Patients Cage Dowels Ring Other L L L5 S L L4 S L3 S1 3 3 L L total * BAK = Bagby and Kuslich. The other types of instrumentation included one Harms cage, two methyl methacrylate and Z plates, and two iliac crest and Kaneda devices. Surgery was abandoned in one patient. warmer than the left. Four other patients complained of temperature variation, dysesthesias, discoloration, and/or swelling of the left leg/foot in the early follow-up period. Of the three men with sexual dysfunction, two experienced retrograde ejaculation and one impotence. Two of three patients who suffered injury to the ilioinguinal nerve reported numbness in the medial upper thigh, whereas the third patient experienced dysesthetic pain. The incidence of complications was not related to age (p 0.5), and there was no significant association between the incidence of complications and underlying diagnosis (p 0.1), previous abdominal or spinal surgery (p 0.1), or number of levels exposed (p 0.1). Discussion In our study we found a high incidence of general surgery related complications following ALIF. We were surprised by these findings because all the surgical exposures were performed by experienced general or vascular surgeons. The most likely explanation is that many of the complications included in our analysis, such as intraoperative minor vascular injury, sympathetic neural injury presenting as short-lived temperature variation in the lower extremity, and prolonged ileus, have not been reported by other authors who have studied general surgery related complications following ALIF. In several clinical series the authors have addressed the issue of functional 10 and radiographic 16,26 outcome following ALIF. However, there are no comprehensive studies of general surgery related complications following ALIF. Faciszewski and coworkers 5 have retrospectively reviewed 1223 cases in which anterior spinal fusion was performed in the thoracic and lumbar spine and reported an overall complication rate of 40.5%. However, of the 492 patients in whom lumbar surgery was performed, only 38 underwent ALIF alone, and the rest underwent combined anterior posterior fusion procedures. The authors grouped their complications into those common to both anterior and posterior surgeries and those specific to anterior surgery alone. However, the authors did not separately report the complications encountered in the ALIF group, TABLE 3 Complications following ALIF in 60 patients Complication No. of Patients (%) sympathetic dysfunction 6 (10) vascular injury 4 (6.6) somatic neural injury 3 (5.0) sexual dysfunction 3 (5.0) prolonged ileus 3 (5.0) wound incompetence 2 (3.3) deep vein thrombosis 1 (1.6) pancreatitis 1 (1.6) bowel injury 1 (1.6) total 24 (40) and furthermore, they excluded somatic and sympathetic neural injury if the symptoms lasted fewer than 6 months. Tiusanen, et al., 26 have reported a detailed clinical outcome study in a series of 83 patients. In neither their abstract nor the conclusions do the authors report the incidence of complications. An analysis of Table 7 in their paper reveals an 80% general surgery related complication rate. Complications included unintentional sympathectomy (43%), retrograde ejaculation (in 24% of males), venous injury (18%), deep vein thrombosis (5%), and infection (3%). McAfee, et al., 18 have reported on the early postoperative complications following endoscopic ALIF in a highly selected group of 22 patients enrolled in the Bagby and Kuslich investigative device exemption study protocol. The single complication that was reported in this study was an iliac vein laceration that required conversion to an open minilaparotomy procedure. Baker and associates 1 have reported a 15% incidence of vascular complications following ALIF in 85 patients. They found an increased incidence of venous injury when the retroperitoneum was approached via a small paramedian incision as compared with a longer flank incision. In a number of individual case reports rare, specific complications have been highlighted. 3,11,13 In our report we provide a more complete analysis of general surgery related complications following ALIF. Intraoperative Complications The leading intraoperative complication in this series was vascular injury. A frequency rate of 6.66% is in keeping with that reported in the literature. However, some authors have reported an incidence rate as high as 18%. 1 Arterial injury is rare and usually presents as thrombotic occlusion 17,22 or intraoperative or delayed massive hemorrhage. Most venous injuries are related to the exposure rather than to graft placement. 1 Three instances of venous injury, in this series, occurred in patients undergoing onelevel fusion and one in the course of a two-level fusion procedure. One patient undergoing reexploration of a previous ALIF, for repositioning/removal of a suboptimally placed cage, suffered a major venous injury because of dense adhesions between the posterior vessel wall and the instrumentation. Surgery in this case had to be abandoned after the patient suffered significant blood loss. Late venous thrombosis on the affected side is a real concern, and close monitoring in the postoperative period is mandatory. 62

4 Anterior lumbar interbody fusion During the follow-up period, none of our patients who sustained venous injury developed evidence of venous thrombosis on Doppler study. In view of this potentially dangerous complication, Watkins 28 has warned that only one attempt at anterior surgery is possible at each level. Peritoneal perforation during retroperitoneal exposure is not uncommon. The true rate of its occurrence is difficult to discern in this retrospective study. In several of the operative reports we reviewed, there was mention of immediate closure of these defects when recognized. More important, however, is the recognition of bowel perforation, a rare complication. Our patient with this complication was a 38- year-old woman with no previous abdominal surgery. The injury to the bowel was immediately recognized and repaired. She suffered no postoperative complication and made an excellent outcome. In the presence of significant contamination of the wound, however, the fusion operation should be abandoned after bowel closure. The degree of adherence of the bowel to the peritoneum and the retroperitoneal structures cannot be predicted in advance. One should anticipate such problems in all cases with a history of intraabdominal sepsis, malignancy, previous abdominal surgery, or radiotherapy. Early Postoperative Complications Postoperative ileus following retroperitoneal exposure usually lasts 2 to 3 days. 23 Most of our patients were able to tolerate fluids by the 2nd postoperative day. Prolonged ileus is often related to multiple previous abdominal surgeries, extensive dissection, retroperitoneal hematoma formation, major intraoperative fluid shifts, or excessive narcotic use. None of the three patients in this series who developed prolonged ileus had undergone previous abdominal surgery or suffered significant blood loss. These patients were treated with intravenous fluids, nasogastric suction, and intestinal motility enhancers until unequivocal bowel sounds returned. Spontaneous resolution occurred in every case by the 10th postoperative day. One patient developed chemical pancreatitis, the exact cause of which remains unclear in the absence of any predisposing factors for pancreatitis. We assume it was related to the inadvertent mechanical injury to the pancreas by faulty retractor placement. The rate of abdominal wound dehiscence after retroperitoneal exposure, as documented in the general and vascular surgery literature, is 0 to 2.3%. 2 This complication is related to faulty surgical technique, in the absence of systemic metabolic derangements precluding normal wound healing. Late Postoperative Complications Deep venous thrombosis is reported to occur at a rate of 7 to 8% after surgery performed in the vicinity of major blood vessels in the abdomen. 6 In addition to routine prophylactic methods such as intraoperative sequential compressive boots, early postoperative mobilization, and use of low-dose heparin, delicate handling of the vascular structures intraoperatively is vital for the prevention of this complication. Although some authors have routinely used Steinmann pins to keep the vascular structures retracted, 1,7,9,26 it is desirable to use hand-held malleable retractors for vascular retraction. Periodic release of pressure when surgery is not being performed near these structures cannot be overemphasized. Injury to the lumbar sympathetic chain located on the lateral aspect of the vertebral bodies has been recognized in the past. Characteristically, the patient complains of a cold limb or foot on the side contralateral to the approach. This is because the affected side is warmer due to loss of sympathetic vasoconstriction. In all such cases, the distal pulses should be checked to exclude the possibility of arterial compromise. 17,22 Surprisingly, of our six patients with this complication, two sustained sympathetic dysfunction on the right side. The exact cause of this was not apparent. Review of the operative notes and postoperative radiographs did not reveal excessive dissection on the right side. In the four patients in this series with lower-extremity temperature variations (two on the right side and two on the left), symptoms resolved in three within 4 months of their surgery. One patient continued to experience intermittent swelling and discoloration at 6 months. One patient experienced persistent, nondisabling dysesthesias of the left foot and another burning paresthesia of the left thigh. Some authors regard sympathetic nerve injury as an unavoidable surgery-related complication of an anterior approach. 7,9,28 In most patients there are no long-term sequelae, but a minority suffers from prolonged dysesthetic pain. 19 Bilateral damage may cause ejaculatory disturbance. Sexual dysfunction was a surgery-related complication of ALIF in 9.6% of the men in this series. The reported incidence of this complication varies from 0.42 to 20%. 8,15,27 Some of these reports are anecdotal, 8 lacking even the numbers of male and female patients in the study population. In more detailed studies 12,27 there has been a reported 5 to 22% incidence of retrograde ejaculation and complete or partial resolution of the symptom in one third of the patients over 1 to 2 years. In the two patients with retrograde ejaculation in our series, near-complete resolution of the symptom occurred in one patient by 15 months follow-up examination. Retrograde ejaculation is believed to be caused by damage to the superior hypogastric plexus of the sympathetic system located in front of the L-5 and S-1 vertebral bodies. As this plexus is formed by contributions from the paramedian lumbar sympathetic chains, bilateral damage to the sympathetic chains may also cause ejaculatory disturbances. Johnson and McGuire 15 and Tiusanen, et al., 27 have summarized the pathophysiological mechanisms. There is no anatomical reason for impotence to result from a standard ALIF procedure, because the parasympathetic plexus responsible for erection is located deep in the pelvis. 15 In most instances, impotence is believed to have a psychogenic origin. Our patient was a 36-year-old man who underwent two-level anterior fusion to treat a work-related injury; he was involved in compensation litigation. Injury to the ilioinguinal and the lateral branch of the iliohypogastric nerves has rarely been reported subsequent to retroperitoneal exposure for ALIF. These patients present with numbness or paresthesia affecting the medial upper thigh. In two of three patients in this series these symptoms completely resolved within 6 months. One patient suffered dysesthesia that has only partially resolved 4 months postoperatively. The most common site of injury to these nerves is at their course over the iliac crest. This is in contrast to femoral and genitofemoral nerve injury that occurs at the level of the psoas muscle

5 V. Rajaraman, et al. In summary, the literature regarding surgery-related complications of ALIF has largely focused on issues relating to the graft, instrumentation, and fusion outcomes, and in few reports have the authors analyzed the general surgery related complications in any great detail. Significant numbers of general surgery related complications were recorded by Tiusanen, et al., 26 but were never highlighted in their paper. However, they suggested that ALIF be performed as a last-resort salvage procedure as a part of circumferential fusion because of the complication rate and because results of ALIF were not superior to those of conventional posterolateral fusion. Unintended sympathectomy has been accepted as inevitable, and in none of the large series have the authors reported on the long-term disability associated with this type of neural injury. As we have pointed out, a certain number of these patients may suffer longterm pain syndromes, and at least some of the cases of ejaculatory disturbance may be due to bilateral damage caused by inadequate attention to the sympathetic chain. When one combines the morbidity and complication rates related to graft harvest, instrumentation, neural injury, and fusion success, it becomes clear that ALIF in which cages are used, although apparently technically simple to perform, should not be undertaken without serious consideration of options. Conclusions In our retrospective review we have demonstrated a 40% incidence of general surgery related complications following retroperitoneal exposures for ALIF. We believe that this rate reflects a true incidence and is not related to a learning-curve effect, because the frequency of these complications was the same during both year-long periods in this study ( and ). Many of these complications are self limited and resolve over time. The spine surgeon should be aware of these complications to counsel properly his/her patients preoperatively and to optimize postoperative care. References 1. Baker JK, Reardon PR, Reardon MJ, et al: Vascular injury in anterior lumbar surgery. Spine 18: , Carlson MA: Acute wound failure. Surg Clin North Am 77: , Chan FL, Chow SP: Retroperitoneal fibrosis after anterior spinal fusion. Clin Radiol 34: , Crock HV: Anterior lumbar interbody fusion: indications for its use and notes on surgical technique. Clin Orthop 165: , Faciszewski T, Winter RB, Lonstein JE, et al: The surgical and medical perioperative complications of anterior spinal fusion surgery in the thoracic and lumbar spine in adults. A review of 1223 procedures. Spine 20: , Farkas JC, Chapuis C, Combe S, et al: A randomised controlled trial of a low-molecular-weight heparin (Enoxaparin) to prevent deep-vein thrombosis in patients undergoing vascular surgery. Eur J Vasc Surg 7: , Flynn JC, Hoque AM: Anterior fusion of the lumbar spine. Endresult study with long-term follow-up. J Bone Joint Surg (Am) 61: , Flynn JC, Price CT: Sexual complications of anterior fusion of the lumbar spine. Spine 9: , Fraser RD, Gogan WJ: A modified muscle-splitting approach to the lumbosacral spine. Spine 17: , Gill K, Blumenthal SL: Functional results after anterior lumbar fusion at L5 S1 in patients with normal and abnormal MRI scans. Spine 17: , Graham JM, Kozak JA, Reardon MJ: Rectus sheath hematoma after anterior lumbar fusion. Spine 16:1377, Inoue S, Watanabe T, Hirose A, et al: Anterior discectomy and interbody fusion for lumbar disc herniation. A review of 350 cases. Clin Orthop 183:22 31, Isiklar ZU, Lindsey RW, Coburn M: Ureteral injury after anterior lumbar interbody fusion. A case report. Spine 21: , Ito H, Tsuchiya J, Asami G. A new radical operation for Pott s disease. J Bone Joint Surg 16: , Johnson RM, McGuire EJ: Urogenital complications of anterior approaches to the lumbar spine. Clin Orthop 154: , Loguidice VA, Johnson RG, Guyer RD, et al: Anterior lumbar interbody fusion. Spine 13: , Marsicano J, Mirovsky Y, Remer S, et al: Thrombotic occlusion of the left common iliac artery after an anterior retroperitoneal approach to the lumbar spine. Spine 19: , McAfee PC, Regan JR, Zdeblick T, et al: The incidence of complications in endoscopic anterior thoracolumbar spinal reconstructive surgery. A prospective multicenter study comprising the first 100 consecutive cases. Spine 20: , McMahon SB: Mechanisms of sympathetic pain. Br Med Bull 47: , Mercer W: Spondylolisthesis: with a description of a new method of operative treatment and notes of ten cases. Edinburgh Med J 43: , Papastefanou SL, Stevens K, Mulholland RC: Femoral nerve palsy. An unusual complication of anterior lumbar interbody fusion. Spine 19: , Raskas DS, Delamarter RB: Occlusion of the left iliac artery after retroperitoneal exposure of the spine. Clin Orthop 338:86 89, Sicard GA, Reilly JM, Rubin BG, et al: Transabdominal versus retroperitoneal incision for abdominal aortic surgery: report of a prospective randomized trial. J Vasc Surg 21: , Speed K: Spondylolisthesis. Treatment by anterior bone graft. Arch Surg 37: , Stauffer RN, Coventry MB: Anterior interbody lumbar spine fusion. Analysis of Mayo clinic series. J Bone Joint Surg (Am) 54: , Tiusanen H, Seitsalo S, Osterman K, et al: Anterior interbody lumbar fusion in severe low back pain. Clin Orthop 324: , Tiusanen H, Seitsalo S, Osterman K, et al: Retrograde ejaculation after anterior interbody lumbar fusion. Eur Spine J 4: , Watkins R: Anterior lumbar interbody fusion surgical complications. Clin Orthop 284:47 53, 1992 Manuscript received November 3, Accepted in final form April 22, Address reprint requests to: Viswanathan Rajaraman, M.D., Division of Neurosurgery, New Jersey Medical School, 90 Bergen Street, DOC Building, Suite 7300, Newark, New Jersey rajaraman@sprintmail.com. 64