Abstract
We report a five year military experience with anterior retroperitoneal spine exposure combining vascular and neurosurgical spine teams. From August 2005 through April 2010 (56 months), hospital records from a single institution were retrospectively reviewed. Complications, estimated blood loss, transfusions, operative time and length of stay were documented. Eighty-four patients with lumbar spondylosis underwent primary (63, 75%) or secondary exposure (21, 25%) of a single- (66, 79%) or multilevel disc space (18, 21%). Median operative time and estimated blood loss were 127 minutes (range, 30–331 minutes) and 350 mL (range, 0–2940 mL). The overall complication rate was 23.8%. Postoperative complications included six blood transfusions (7%), three patients with retrograde ejaculation (3.57%) or surgical site infection; two with a prolonged ileus (2.38%) or ventral hernia and one each with a bowel obstruction (1, 1.19%), deep venous thrombosis or lymphocele. All-cause mortality was 1%. In conclusion, a team approach can minimize complications while offering the technical benefits and durability of an anterior approach to the lumbar spine.
Introduction
Advances in spinal instrumentation and the introduction of artificial disc replacement have resulted in an increased need for anterior exposure of the lumbar spine. 1 Although lumbar interbody fusion can be performed anteriorly (ALIF) or posteriorly (PLIF) a total disc replacement (TDR) can only be performed using an anterior approach. In a military population, TDR is occasionally preferred over ALIF on the theoretical advantage of decreasing pain yet preserving mobility, flexion and rotation in this physically active population. Since TDR requires a wider exposure for device insertion and precision midline placement, the ability to avoid vessel injury is a major determinant when opting for the procedure. Although vascular injuries ranging from 1 to 15% are considered to be the most common complication, the anterior approach is also associated with potential injury to the bowel, nerve and ureters underscoring the benefit of a team approach. 1
The purpose of this study is to review contemporary operative management focusing on lessons learned with a combined vascular and neurosurgical spine approach to anterior spine exposure and compare the incidence of common complications in the civilian literature to our military population.
Methods
Consecutive patients who underwent anterior retroperitoneal exposure of the lumbar spine (ARES) by a combined vascular and neurosurgical spine team at a single institution from August 2005 to April 2010 (56 months) were retrospectively reviewed. Indications for intervention included degenerative disc disease, radiculopathy, vertebral body fractures, herniated disc, pseudoarthrosis and spondylolisthesis with or without spondylolysis.
Extracted data included demographics, indication, procedure, level, estimated blood loss (EBL), complications, length of hospital stay and all-cause mortality. A primary exposure (first time) or secondary exposure (return for re-operation) was also documented. All patients were followed by both surgical teams until discharge. Complications included vascular injury (hemorrhage with transfusion, patch or graft repair), postoperative hemorrhage (transfusion or return or re-operation), hernia, ileus, bowel obstruction, venous thromboembolism, surgical site infection lymphocele and retrograde ejaculation.
Descriptive statistics are used to report the demographics and complication rates for the study group. Continuous data are presented as median (range) for non-parametric data. Data were analyzed using Mann-Whitney tests to determine statistical significance and P values <0.05 were considered significant. Statistical analysis was performed with SPSS 15.0 (SPSS Inc, Chicago, IL, USA). This study was approved (351286-1) by the Institutional Review Board (IRB) at Walter Reed Army Medical Center.
Results
Indications for performing a spine procedure in 84 patients
Surgical procedures performed among 84 patients
ALIF, anterior lumbar interbody fusion; PLIF, posterior lumbar interbody fusion; TDR, total disc replacement
Levels of exposure in 84 patients with spinal surgery
Median operative time was 127 minutes (range 30–331 minutes). Median blood loss was 350 mL (range, 0–2940 mL). There was an overall complication rate of 23.8% (20). Multilevel exposures had a higher (431 versus 337 mL, P = 0.017) median blood loss than single exposures. There were no vascular injuries per se (patch or graft repair); however, blood transfusions were required in six (7%) patients who represented some form of venous bleeding during the exposure. One aortic injury with immediate hemorrhagic death occurred secondary to device misplacement during TDR and was not related to the exposure phase of the case. Spine exposure was aborted in only two (2%) patients, one who became difficult to ventilate and one during the second level of a multilevel re-do exposure secondary to venous bleeding and adhesions. Both patients were markedly obese with a body mass index (BMI) > 40.
Complications seen in 84 patients that had a spine procedure
Discussion
This study demonstrates an acceptable early complication rate (24%) associated with a combined surgical team in a military setting. Since the overall complication rate for open anterior lumbar exposures has been estimated at 30–40%, 1 we feel our findings are important when placed in the context of the modern literature. Our rate of vascular injury, while low, still remains the most common complication as reflected in 7% transfusion requirements for the study group. This underscores the need for a team that is capable of controlling pelvic venous bleeding with expertise in vessel mobilization to ensure safe device insertion for spinal fusion or TDR.
Our operative approach is performed using a left paramedian incision followed by a retroperitoneal dissection to the lumbosacral spine. Placement of the incision is dependent on the level of spine exposure necessary; for L5–S1 exposure the incision extends from the level of the umbilicus inferiorly for approximately 6 cm. For L4–L5 exposure that same incision is adjusted to extend slightly above the umbilicus. The peritoneal contents and ureter are then mobilized medially to visualize the left psoas muscle and overlying genito-femoral nerve. Subsequent mobilization of the left common iliac vein and artery and/or aorta and vena cava, depends on the disk space(s) requiring exposure. For an L5–S1 exposure we sweep laterally using kitners and sponge sticks the left common iliac vein while ligating small venous branches that insert medially. Excessive force at the iliocaval junction is avoided by dividing the iliolumbar vein and sweeping the common iliac vein medially for the L4–L5 exposure. Silk ties may slip resulting in iliolumbar vein retraction into the deep retroperitoneum and excessive hemorrhage from the posterolateral portion of the iliac vein therefore prolene suture ligatures are favored. The midline middle sacral vessels are ligated with ties and manually loaded clips. Bipolar electrocautery is preferred but used minimally. For TDR, a wider exposure is necessary for midline precision and careful mobilization is crucial. An Omni-Retractor (Omni-Tract Surgical, St Paul, MN, USA) is used for difficult dissection but the lower profile Endo-ring retractor (Medtronic Spine, Memphis, TN, USA) is much more expedient (Figure 1).
(a) The Endo-ring retractor (Medtronic, Memphis, TN, USA) can be used as a hand-held vein retractor or mounted on a frame. (b) The aluminum radiolucent blade, a detachable handle and a stability pin is assembled and threaded into the vertebral body after achieving sufficient midline exposure. (c) Use the mallet to start (but not pound) the threading of the pin
Vascular injury is the most common intraoperative complication and the incidence ranges from 1 to 15%, depending on the various definitions of injury. 1 We defined a vascular injury in terms of blood loss or the need to perform a formal repair with a patch or graft. Interestingly, there were no patch or graft repairs and therefore no major venous injuries in the 84 cases. However, there is a learning curve associated with the degree of tension that can be placed on the iliac vein before a suture may be necessary to control hemorrhage. In this series sutures were usually placed before bleeding occurred as in division of the iliolumbar branch complicating the data interpretation regarding suture repair (before or after hemorrhage). We opted to use transfusion for the surrogate of a minor venous injury requiring lateral venorrhaphy. Blood transfusions were required in only six (7%) patients placing our complication rate for vascular injury well within the range reported in the modern literature.2–4
In this series, we found that a few well placed sutures have no consequence to luminal size as we report only one case of DVT without thromboembolism. Although we did not routinely screen patients for DVT, however, all were treated with DVT prophylaxis and there were no other cases of postoperative limb edema. The DVT in this report was probably the result of sponge stick compression during venorrhaphy. The follow-up duplex ultrasound showed complete thrombus resolution without stenosis or limb edema within three months following early systemic anticoagulation and compression therapy.
We reported no cases of arterial thrombosis. Despite what might be considered too much lateral retraction based on changing evoked potentials, we have observed no cases of arterial thrombosis. Brau et al.
5
examined 1315 patients and reported the incidence of iliac artery thrombosis as 0.45%.
5
Based on this series, we recommend if somatosensory evoked potentials (SSEP) change after the Endo-Ring retractors are in position that the TDR or ALIF is completed expeditiously. In our experience once the retractors are removed the baseline is quickly re-established without sequalae. Chiriano et al.
6
reported a 3% rate of life-threatening vascular injury and Hans et al.
7
reported five of 560 (1%) patients with ARES developed significant arterial complications.6,7 Consistent with contemporary literature this study reports similar findings.1,5,7 The one arterial injury we report was unrelated to the exposure; however, it is important to emphasize that the instrumentation has significant potential complications and it is strongly recommended that all surgeons remain close during the procedure and that a team approach is employed when retractors need re-adjustment to accommodate bulky instruments (Figure 2).
(a) A hand-held renal vein retractor (black arrow) may be used inferiorly to provide additional pelvic exposure. The Endo-ring retractors are positioned medially, laterally and superiorly. (b) The Endo-ring (lateral, black arrow) illustrates how these low profile radiolucent retractors protect the iliac vessels during the discectomy (white arrow) and interbody fusion
Venous injury is considered more likely to occur with secondary exposure due to adhesions. In this series, 25% of the study group was undergoing a secondary exposure. Although the median EBL in the primary group was lower (337 mL) than in the revision cohort (431 mL) this was not statistically significant (P = 0.18) and Hamden et al. 1 and Chiriano et al. 6 have reported this finding previously.1,6 EBL was however significantly higher in the multilevel surgery group (719.4 mL) than in the single-level surgical cohort (227.96 mL) (P = 0.017). We find there is much more potential for venous bleeding during an L4–L5 exposure and appropriate caution should be exercised. At this level iliac vessels need much more mobility and occasionally the iliac artery and iliac vein are completely separated. In one report 23% required lateral venorrhaphy. 6 Although some bleeding is osseous and related to the discectomy and placement of the prosthesis, the EBL is essentially related to the exposure and it is important for the anesthesia team to be prepared with cross matched blood and cell saver support.
Morbid obesity (BMI > 35) is an accepted risk factor for vascular injury and Garg et al. 4 reported a significant correlation between EBL and BMI. 4 In this study we aborted two cases secondary to obesity and subsequent hemorrhage or difficulty with ventilation. One subsequently developed a bowel obstruction and ventral hernia. Therefore, surgeons should carefully consider the indications when consenting patients with morbid obesity.
Retrograde ejaculation is more prevalent with dissection at the L4–L5 level during dissection near the left common iliac artery and the L1 vertebral body involving the sympathetic chain ganglia. The literature reports this incidence of retrograde ejaculation rate ranging 1–25% with 7% reported in the team approach.2,3,8 In this series a 3.6% rate of retrograde ejaculation appears to be consistent with prior reports.2,3
With the technique described there were no cases of ureteral injury in this study. Ureteral injury is reported to range from 0.3 to 8.0%. 8 We did not encounter any special situations and did not employ any special maneuvers to mobilize and sweep the ureter to the side.2,3,9,10 Once a blunt hand dissection is complete, a Deaver retractor is placed into the retroperitoneum, and used to sweep the ureter medially. At this point the Deaver is replaced with a splanchnic blade (Omni-Tract Surgical) or the Endo-ring (Medtronic Spine) and the ureter remains protected during the procedure.
DVT and pulmonary emboli that occur after 1–2% of major reconstructive spine operations are most common after anterior lumbar surgery with a 5% rate of incidence. 10 We occasionally observe transient changes in left leg SSEP with prolonged retraction but in our experience those brief changes have not been associated with either a DVT or neurologic deficit. Given our experience we advocate for expeditiously completing the operation with emphasis on focused attention to minimize vascular injury by avoiding additional manipulation.
Artificial discs require greater side-to-side exposure of the disc spaces. This often requires further mobilization of the left common iliac vein. In our cohort, 40 artificial discs were placed of which 39 were Charite (Depuy Spine Inc, Raynham, MA, USA) and one was a Maverick (Medtronic Spine Inc). These were common for primary (33/63, 52.4%), or revised reconstructions (7/21, 33%). In comparison, an ALIF requires much less exposure than TDR (precise midline alignment) and therefore TDR may result in a greater incidence of venous laceration when compared with ALIF. The incidence of complications that require revision or removal of Charité TDR is uncertain. Lemaire et al. 11 reported 107 patients after a minimum of 10-year follow-up with no explants. 11 In the American multi-institutional trial, the revision and removal rates of 205 patients at two years were 2.4 and 1%, respectively.12,13 In the military we have yet to revise or explant any of the Charite discs. At the current time our neurosurgeons prefer ALIF and are no longer performing TDR.
The average length of hospital stay for our patients was 4.6 days (range, 1–32 days). This compares favorably in comparison to PLIF (mean, 5–7 days) and ALIF (mean, 4.3 days) in civilian reports. 14 Accordingly, ALIF patients regardless of single- or multilevel exposure seem to return to activities of daily living in 10–14 days while PLIF patients may require up to 6–8 weeks.15,16
Finally, in vascular surgery there is a national trend to employ endovascular techniques when feasible and this practice may, over time, diminish the current understanding of retroperitoneal anatomy. Therefore, these cases provide an important opportunity to teach military residents important techniques in hemorrhage control. These skills will be crucial when confronted with major pelvic venous injury on a modern battlefield and potentially save countless young lives.
Study limitations
The most obvious limitation is the retrospective design of the study. Furthermore, conclusions are broadly limited secondary to the very selective nature of this young healthy cohort in a military setting. This obviously will not represent the average American citizen undergoing a similar procedure. Despite these limitations, the findings of this study are novel and have merit in a unique population allowing for conclusions and comparisons to be drawn regarding the outcome in this military setting. This is also the largest series of TDR in a military population ever to be reported. Further randomized control trials are required to establish the long-term impact of our findings.
Conclusion
This is the largest series of anterior spine exposures in a military population. This study shows ARES has a very acceptable if not superior complication rate when compared with the contemporary civilian literature. Short hospital stays and low complication rates, with similar mortality were observed. A team effort employing a vascular and spine surgeon have demonstrated a safe, effective way to expeditiously approach the lumbar spine and should be considered as the demand for spinal instrumentation increases.
Footnotes
Acknowledgements
The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Department of the Army or the Department of Defense. The authors are employees of the US government. This work was prepared as part of their official duties and, as such, there is no copyright to be transferred.