28S Proceedings of the NASS 22nd Annual Meeting / The Spine Journal 7 (2007) 1S–163S

data to support the need for prospective cost-effectiveness studies for the sur-

gical management of appropriately selected patients suffering from FLSS.

FDA DEVICE/DRUG STATUS: This abstract does not discuss or include

any applicable devices or drugs.

doi: 10.1016/j.spinee.2007.07.066

2007 NASS Outstanding Paper Award – Basic Science

Mouse GDF-5 Protein and DNA Therapy Potentiates Intervertebral

Disc Cell Aggregation and Chondrogenic Gene Expression

Min Cui, MD, PhD, Yuqing Wan, PhD, D. Greg Anderson, MD, Frank

H. Shen, MD, Brian M. Leo, MD, Cato T. Laurencin, MD, PhD,

Gary Balian, PhD, Xudong Li, MD, PhD; University of Virginia School of

Medicine, Charlottesville, VA

BACKGROUND CONTEXT: Growth and differentiation factor-5 (GDF-

5) deficient mice showed abnormalities in intervertebral disc (IVD) struc-

ture and extracellular matrix gradient. Adenoviral-mediated GDF-5 deliv-

ery can promote the growth of rabbit disc cells.

PURPOSE: The aim of the present study was to investigate the effect of

recombinant protein and genomic forms of GDF-5 on the metabolism of

intervertebral disc cells.

STUDY DESIGN: The effects of recombinant protein GDF-5 and geno-

mic GDF-5 in mouse intervertebral disc cells will be evaluated in vitro.

METHODS: Mouse disc cells in vitro were treated with recombinant

GDF-5 protein. Likewise, the mouse GDF-5 gene was cloned into an ex-

pression vector and was used to transfect mouse disc cells in vitro. Proteo-

mic and genomic GDF-5 therapy was then assessed by measuring cell

proliferation, proteoglycan production, and extracellular gene expression.

RESULTS: Biochemical assays revealed an elevated GAG/DNA ratio in

mouse IVD cells cultured in the presence of various concentrations of

mGDF-5 protein. Real-time RT-PCR demonstrated that GDF-5 protein in-

creased the expression of collagen type II and aggrecan genes in a dose

dependent manner while decreased MMP-3 gene expression. Immunohis-

tochemistry showed increased aggregates formation in mouse IVD cells

treated with mGDF-5 compared to the control group. The murine GDF-

5 gene was successfully cloned into an expression plasmid vector and

GDF-5 protein production was confirmed by Western blot analysis. Type

II collagen and aggrecan gene expression were significantly increased in

the cells Nucleofected with the GDF-5 plasmid compared with cells

Nucleofected with control plasmid.

CONCLUSIONS: This is the first report to clone the murine GDF-5 gene

and use the Nucleofection method to transfer DNA into IVD cells. Our

data suggests that both recombinant protein and DNA forms of GDF-5

can improve extra cellular gene expression in mouse IVD cells. Our work

using gene therapy in the treatment of degenerative disc disease with

a novel ex vivo gene transfer technique may benefit patients with clinically

relevant axial spine pain.

FDA DEVICE/DRUG STATUS: This abstract does not discuss or include

any applicable devices or drugs.

doi: 10.1016/j.spinee.2007.07.067

Thursday, October 25, 200710:33–11:23 AM

General Session: Complications

56. Access Complications and Revisions in a Large Series of Patients

Undergoing Total Disc Replacement with the ProDisc L Device

Salvador Brau, MD1; 1Los Angeles, CA, USA

BACKGROUND CONTEXT: With 2 devices for total disc replacement

(TDR) approved for clinical use, it is important to know what the

expected incidence of complications and anterior revision for this proce-

dure might be.

PURPOSE: This study is intended to show what this incidence is in

a large series of patients in whom the data was collected with up to 4 years

follow up.

STUDY DESIGN/SETTING: A concurrent database was maintained on

all patients undergoing TDR with the ProDiscL device from Jan., 2002

to Dec., 2006. All approach complications were tabulated. Patients return-

ing for anterior revision were also followed.

PATIENT SAMPLE: All 405 patients that had the approach performed by

the author during this time frame were included.

OUTCOME MEASURES: All approach complications encountered and

revisions performed were considered to have affected the outcome and re-

ported here.

METHODS: 405 patients had 541 devices inserted at all levels of the

lumbar spine from L1 to S1. 297 patients had a single level TDR with

most at L4-5 or L5-S1 but some at L3-4 and 2 at L2-3. 83 had 2 level disc

replacement, 22 had TDR at 3 levels, 2 had TDR at 4 levels (L2 to S1)

and 1 had TDR from L1 to L5 with a fusion at L5-S1. 8 patients had a hy-

brid construct with a fusion at L5-S1 and TDR at L4-5. 2 of these patients

also had a TDR at L3-4 as part of a 3 level construct. 3 patients had

a TDR at the level above a prior anterior interbody fusion (2 at L3-4

and 1 at L2-3).

RESULTS: There were 3 left iliac vein lacerations repaired surgically.

Blood loss was 350, 400 and 400 cc’s respectively and all had the proce-

dure completed and had no post operative sequelae. There were no arterial

injuries or thrombosis. There were no ureteral injuries and no reported

cases of retrograde ejaculation among the males, who represented just over

50% of this cohort. There were no reports of neurologic deficit that could

be associated with the approach. There were 4 cases of prolongued ileus

needing nasogastric decompression and 2 hernias have been reported. 3 pa-

tients have had documented deep vein thrombosis treated with anticoagu-

lation and resolved. 3 patients returned for revisions. One patient with a 3

level TDR had the device at L5-S1 repositioned 1 week post implantation

and remains well after 2 years. A single level TDR at L5-S1 was removed

8 months post op because of partial subsidence and the level fused. An-

other patient with an L3-4, L4-5 TDR had a traumatic event 10 months

post op and the polyehtylene insert at the L4-5 disc was partially dis-

lodged. The device was removed at L4-5 with fusion at this level and at

L5-S1, which had been traumatized as well. The device at L3-4 remained

in good position.

CONCLUSIONS: An incidence of vascular injury of 0.7% with no other

significant complications noted indicates that TDR can be carried out with

safety and confidence, especially since the early anterior revision rate is

also 0.7%. Removal of a device at L4-5 presented a significant approach

challenge, but once exposure was completed, the device was relatively

easy to remove from an antero-lateral approach without the need to per-

form an osteotomy to get the keeled plates out. After removal of the poly-

ethylene insert with distraction, there was sufficient room to remove the

plates after separating them from the end plates.

FDA DEVICE/DRUG STATUS: ProDisc L: Approved for this indication.

doi: 10.1016/j.spinee.2007.07.069

57. Direct Lateral Lumbar Transpsoas Interbody Fusion:

Complications and Patient Morbidity

Reginald Knight, MD1, Paul Schwaegler, MD2, David Hanscom, MD2,

Jeffery Roh, MD1; 1Orthopedics International Spine, Kirkland, WA, USA;2Orthopedics International Spine, Seattle, WA, USA

BACKGROUND CONTEXT: MIS management of degenerative lumbar

conditions can be divided on the basis of implant shape, rectangular

versus bulleted, or XLIF versus direct lateral (DLIF). Both procedures

strive to support the anterior column, indirectly decompress neural ele-

ments and promote spinal fusion. MIS techniques are touted as less

morbid and therefore an improvement in patient care. However, the lack

29SProceedings of the NASS 22nd Annual Meeting / The Spine Journal 7 (2007) 1S–163S

of surgeon familiarity with the expectations of this procedure has lim-

ited its use.

PURPOSE: Review comparisons of DLIF versus XLIF and the early com-

plication rate associated with the transpsoas technique.

STUDY DESIGN/SETTING: A non-randomized prospective review

within private practice setting.

PATIENT SAMPLE: May 2004 to December 2006, 58 patients were

treated with MIS transpsoas lumbar interbody fusion. 43 female and 15

males with mean age of 61 years. XLIF or DLIF was chosen on the basis

of surgeon preference. Procedure breakdown: levels - 38 single, 19 two, 1

three; 39 XLIF, 19 DLIF. Mean follow-up 15 months.

OUTCOME MEASURES: Procedural differences in patient morbidity

and complications was assess via hospital and office chart review.

METHODS: Following exhaustive non-operative care, patient education

and informed consent surgical candidates with mechanical low back pain,

radiculopathy and or neurogenic claudication were offered surgery. Pa-

tients were prospectively enrolled on a consecutive basis. Students t-test

was used for analysis. Complications were designated as surgical or med-

ical, sub-acute or chronic (lasting O1 year).

RESULTS: 13/58 (22%) experience complications (3 medical (5%):

UTI, MI, dementia; 10 surgical (17%): 6 meralgia paresthetica (10%),

2 L4 nerve root injuries (3.4%-chronic), 1 persistent psoas spasm

(1.7%), 1 loss of fixation (1.7%). Due to surgeon preference for staged

procedures hospital stay was not a valid data point. There were no sig-

nificant differences between sex, lumbar level fused or implant shape.

Significant differences were seen between 1 and 2 level procedures

(see table). During course of study 4 transpsoas cases were aborted with-

out complication.

CONCLUSIONS: Transpsoas approaches can be associated with signifi-

cant postoperative complications (22%). Most complications were sub-

acute in duration and did not appear to effect ultimate patient satisfaction.

3.4% involve significant injury to neural structures despite the use of

neurophysiologic monitoring.

FDA DEVICE/DRUG STATUS: This abstract does not discuss or include

any applicable devices or drugs.

doi: 10.1016/j.spinee.2007.07.070

58. The Role of TceMEPs in Detection of Iatrogenic Spinal Nerve

Root Deficit during Instrumented Lumbosacral Fusion

Bikash Bose, MD, FACS, FICS1, Anthony Sestokas, PhD2,

Daniel Schwartz, PhD2; 1Christiana Care Health System, Newark, DE,

USA; 2Bala Cynwyd, PA, USA

BACKGROUND CONTEXT: The complication rate of lumbar spine

fusion with instrumentation ranges from 1-33%. Several monitoring tech-

niques, including spontaneous electromyography (EMG) and spontaneous

EMGþtranscranial electrical motor evoked potential (tceMEP), have been

used in an attempt to prevent/limit the neurologic injuries that can occur

during nerve decompression or placement of instrumentation. Each ap-

proach offers some benefit in predicting postoperative deficits, but infor-

mation is limited about their specificity and sensitivity.

PURPOSE: The purpose of this study was to compare spontaneous EMG

and spontaneous EMGþtceMEP monitoring in identifying evolving spinal

nerve root deficit during posterior lumbosacral decompression with inter-

body cage fusion.

STUDY DESIGN/SETTING: We retrospectively reviewed charts of pa-

tients who had undergone lumbosacral fusion surgery with instrumentation

performed by a single neurosurgeon at Christiana Care Hospital and mon-

itored by a single neuromonitoring practice.

PATIENT SAMPLE: Patients ranged in age from 20-70 years and demo-

graphics were fairly evenly split between men and women. All patients un-

derwent lumbar reconstructive surgery for spinal degenerative conditions.

OUTCOME MEASURES: Outcome measures included the number of

patients who sustained new onset postoperative lower extremity weakness,

EMG alerts, tceMEP alerts, and the correlation of alerts to postoperative

deficits.

METHODS: We reviewed charts of 342 patients who had undergone lum-

bosacral fusion with instrumentation performed by a single neurosurgeon

at Christiana Care Hospital. In all surgeries, a single professional-level

neuromonitoring practice monitored spinal nerve root function using spon-

taneous EMG. During the last 197 surgeries, spontaneous EMG monitoring

was augmented with tceMEP monitoring. Results of each monitoring ap-

proach were then correlated to the appearance of new postoperative

deficits.

RESULTS: Of 342 patients, 12 (3.5%) showed new onset postoperative

lower extremity weakness. Of these, sustained neurotonic EMG activity

was reported in 7 (58%). Six of 12 patients were monitored with tceMEPs

in addition to EMG activity. Persistent tceMEP amplitude attenuation

(O75%) occurred in all 6 patients (100%).

CONCLUSIONS: We conclude spontaneous EMG monitoring alone is

associated with a high rate of false negative results, allowing nerve root

deficits to develop and progress undetected during lumbosacral decompres-

sion with interbody cage fusion. Conversely, tceMEPs appear to be very

sensitive to evolving nerve root deficits. Used together, spontaneous

EMG and tceMEP monitoring appear to offer high sensitivity and specific-

ity in detecting developing nerve root deficits.

FDA DEVICE/DRUG STATUS: This abstract does not discuss or include

any applicable devices or drugs.

doi: 10.1016/j.spinee.2007.07.071

59. Prospective Study of Post-operative Lumbar Epidural

Hematoma: Incidence and Risk Factors

Mark Sokolowski, MD1, Timothy Garvey, MD1, John Perl, II, MD2,

Amir Mehbod, MD1, Woojin Cho1, Margaret Sokolowski, PhD3,

Ensor Transfeldt, MD1; 1Twin Cities Spine Center, Minneapolis, MN, USA;2Abbott Northwestern Hospital, MN, USA; 3Minneapolis, MN, USA

BACKGROUND CONTEXT: Postoperative epidural hematomas have

been identified in up to 100% of asymptomatic patients following open

microdiscectomy. No prospective study of postoperative hematoma and

thecal sac compression has been performed following lumbar decompres-

sion surgery with or without fusion.

PURPOSE: To determine the incidence of epidural hematoma with asso-

ciated thecal sac compression following lumbar decompression surgery,

and to identify risk factors correlated with postoperative hematoma

volume.

STUDY DESIGN/SETTING: A prospective study of patients undergoing

lumbar decompression surgery with or without associated fusion.

PATIENT SAMPLE: The study population consisted of 50 consecutive

patients undergoing lumbar decompression surgery with or without associ-

ated fusion at one institution.

OUTCOME MEASURES: Relative changes in pre- and post-operative

thecal sac cross sectional area (CSA) and hematoma volume were mea-

sured and calculated for each patient using digital imaging software.

METHODS: Pre-operative MRI and clinical data including 10 identified

pre- and post-operative risk factors were prospectively collected on 50 con-

secutive patients undergoing lumbar decompression surgery with or with-

out fusion. Postoperative MRIs were performed on all patients within 3-5

days of surgery. Relative thecal sac compression due to hematoma was cal-

culated at all levels where postoperative CSA was smaller than preopera-

tive. Volumes of compressive hematomas were calculated. Multivariate

analysis was performed to determine which risk factors correlated with

postoperative hematoma volume.

RESULTS: The incidence of soft tissue hematoma was 100%. Despite de-

compression, 64% of patients actually had a smaller thecal CSA at one or

more levels postoperatively secondary to hematoma formation. No patient

had new post-operative neurologic deficits. The mean number of levels de-

compressed was 2.0, with hematoma extension over a mean 2.4 levels.