Laparoscopic discectomy with anterior interbody fusion of L5-S1

D. Olsen, D. McCord, M. Law

Centennial Medical Center, Nashville, TN, USA

Received: 14 May 1996/Accepted: 15 June 1996

AbstractBackground:A laparoscopic approach to the spine for theperformance of a minimally invasive discectomy was firstdescribed in 1991. Since that time, a number of approachesto laparoscopic discectomy have appeared in the literature.Although these reports demonstrate the ability to approachthe spine through a laparoscopic technique, they do notaddress the issues of loss of disc space, lumbar instability,and the need for interbody fusion.Methods:Described is a technique of laparoscopic discec-tomy with interbody fusion that has been performed suc-cessfully in 75 patients. Although a carbon fiber implantwas utilized to aid in the fusion process, the technique canequally be performed using donor bone as the interbodysupport. In the 75 patients attempted, 73 procedures weresuccessfully completed via the laparoscopic approach. Onepatient was converted to an open anterior approach due toextensive pelvic adhesions from prior surgery. A secondpatients procedure was aborted after the diagnostic laparos-copy demonstrated dense presacral scarring from a previousgynecological procedure.Results:There were no major complications in the series.Two patients with high riding bladders sustained bladderlacerations that were recognized and repaired with simplesuture closure. There were no bowel injuries, and moreimportantly, no major vessel injury.The patients were discharged from the hospital on an aver-age within 36 hours, with a return to work averaging be-tween 2–4 weeks depending on the patients type of work.Using a modified pain score for evaluation, post operativepain was reduced by 75%.Conclusions:From this study, it is concluded that laparo-scopic discectomy with interbody fusion is not only fea-sible, but appears to give good results with follow up ex-tending out beyond two years. Issues regarding the use of

carbon fiber cages vs. bone and indications of the procedureare independent of the laparoscopic approach and are ad-dressed extensively in the orthopedic literature. It can beconcluded that when there is surgical indication for L5-S1discectomy, that a laparoscopic approach with interbodyfusion may become the procedure of choice.

Key Words: Laparoscopic discectomy — Interbody fusion— Laparoscopic spine access surgery

Introduction

An episode of severe, disabling lower back pain is estimatedto affect over 80% of the population sometime during theirlifetime. Even more significant, is that lower back problemsare the number one cause of disability for patients under theage of 45. They are estimated to cost 123 billion dollars ayear in health care costs. Only one third of the cost isincurred from healthy care expenditures, the remaining costfrom chronic life long disability! [11] Despite the magni-tude of the problem, the surgical treatment for lower backdisorders has been less than optimal. It has been associatedwith significant morbidity and less than favorable results [7,10]. Over the years, in an attempt to improve upon theirresults and minimize the morbidity, spine surgeons havesought new ways of surgically addressing lower back dis-orders. In 1991, a technique to access the disc space ante-riorly using laparoscopy was described in an attempt toachieve improved results and minimize the morbidity asso-ciated with discectomy. [6] Since that original report, therehave been a number of articles that describe similar tech-niques. These reports not only demonstrate the feasibility ofthe laparoscopic approach, but that the procedure can beperformed with good results and low morbidity [1, 8, 12].Although these early reports are intriguing, they are all lim-ited to simple discectomy, addressing only the herniateddisc. In an attempt to expand upon the versatility of thelaparoscopic approach, we developed a technique of access-ing the L5-S1 disc space to carry out a complete discectomywith interbody fusion utilizing standard orthopedic tech-

Presented at the annual meeting of the Society of American Gastrointes-tinal Endoscopic Surgeons (SAGES), Philadelphia, Pennsylvania, USA,14–17 March 1996

Correspondence to:Douglas O. Olsen, 300 20th Avenue North, Suite G-3,Nashville, TN 37203, USA

Surg Endosc (1996) 10: 1158–1163

SurgicalEndoscopy© Springer-Verlag New York Inc. 1996

niques. To date, the technique has been successfully per-formed on 75 patients with good results.

Materials and methods

The results from the first 75 patients in which a laparoscopic discectomywith interbody fusion was attempted were retrospectively reviewed. Indi-cations for the procedure (discectomy with fusion) included:

● degenerative segmental instability resulting in incapacitating lower backpain

● recurrent herniated disc disease with concomitant degenerative disease● degenerative spondylolisthesis● pseudoarthrosis with instability after a previous failed operative proce-dure

The selection criteria for offering patients the operative procedure wasbased strictly on indications for surgery, and no patient was refused surgerybecause of body habitus, history of previous abdominal surgery, or otherfactors that are often considered to be relative contraindications for lapa-roscopic abdominal surgery. All patients were approached with a co-surgeon approach, being evaluated and followed by both a spine and lap-aroscopic surgeon.

The technique consists of a four puncture laparoscopic approach, withtwo 5 mm trocars being places in the left and right flanks respectively, a 10mm trocar at the umbilicus for placement of the 10 mm 45 degree lapa-roscope, and a 15 mm trocar in the supra-pubic location for accessing thedisc space and carrying out the orthopedic portion of the procedure (Fig. 1).Initial access is gained using an open approach at the umbilicus and aftersafe abdominal access has been achieved the two 5 mm flank trocar areplaced under direct vision. All pelvic adhesions are release allowing boththe sigmoid colon and small bowel to be reduced from the pelvis. This is

facilitated by placing the patient in a steep Trendelenburg position. Toprevent the patient from sliding on the table, a role is placed above theshoulders and secured to the table with tape. Additionally, the patients armsare positioned across the chest to facilitate cross table x-rays of the pelvisrequired during the procedure. The sigmoid colon is retracted left lateral byplacing loop ligatures on the appendix epiploica as stay sutures. The su-tures are cut long, and brought through the skin using an Endo-close needle(U.S. Surgical, Norwalk, CT). The sutures are pulled tight and secured witha Kelly hemostat on the outer surface of the abdominal wall. After iden-tifying the sacral promontory with palpation, the peritoneum overlying thesacral promontory is divided in a longitudinal fashion, retracting the tissueslaterally to preserve the presacral nerve fibers. The middle sacral vesselsare identified, cauterized using bipolar cautery, and divided. With the discspace exposed, a 22 gauge spinal needle is passed through the anteriorabdominal wall in the suprapubic position and placed into the disc space.A cross table x-ray is performed to not only verify that the proper discspace has been exposed, but to give the appropriate angle of trajectory forplacement of the 15 mm supra-pubic trocar (Fig. 2). The 15 mm trocar isplaced only after the disc space has been initially accessed and verifiedwith cross table x-ray. Placement of this trocar is critical, since a directin-line approach must be achieved to carry out the work on the disc space(Fig. 3). If the angle of approach of the 15 mm trocar is off, problems canoccur gaining access to the depths of the disk space. With the disc spaceidentified, the annulus is divided over the disk space, actually cutting a‘‘window’’ out of the fibrous material. This aids in the visualization of thedisk space during the operative procedure. The disk material is initiallydisrupted utilizing graduated shaves. The depth of insertion can be deter-mined by the cross table lateral x-ray of the spine. With the disk materialdisrupted, the disk is removed with a variety of graspers and curets. Withthe majority of the disk material removed, the disk space is expanded byplacing sequentially larger spacers within the disk space. Care is taken notto over expand the disk space so as to cause pressure or tenting of the nerveroots. If there is a posterior fragment noted on preoperative MRI, it isremoved at this time. A final preparation of the end plates is performedunder direct visualization, utilizing rasps and curets to achieve good bleed-ing bone. Bone is harvested from the patients right iliac crest, and a stan-dard anterior interbody fusion is carried out. In our series of patients, acarbon fiber cage is used to achieve the stabilization of the interspace, butvirtually any device can be used, to include cadaver bone plugs. A finalx-ray is performed to verify the position of the fusion device, and theperitoneum closed over the operative site using chromic suture. The area ischecked for hemostasis, and the pelvis irrigated and aspirated clear. Theretraction stay sutures are released, and the 15 mm suprapubic trocar isremoved under direct vision making certain that no bladder injury hasoccurred. The fascia at the 15 mm and 10 umbilical trocar sites is closedunder direct vision prior to deflating the pneumoperitoneum. The skinincisions are closed with absorbable suture, and both the NG tube andFoley catheter removed in the operating room. Post operatively, the pa-tients are ambulated immediately upon recovering from anesthesia. Theyare started on clear liquids and advanced on their diets slowly over the next12–24 hours. If the patients are stable and demonstrate no evidence

Fig. 1. Diagram demonstrating the location of trocar placement for a L5-S1 discectomy. The 15 mm suprapubic trocar is not placed until the sacralpromontory has been exposed, and cross table x-ray performed.

Fig. 2. Picture of a cross table x-ray with needle in the L5-S1 disc spacedemonstrating how the needle determines the placement angle of the 15mm trocar.

1159

of ileus, they are discharged the evening of the 1st post operative day, tofollow up in 7–10 days in the office.

Results

In our series, 75 patients to date have undergone an attemptat laparoscopic discectomy with interbody fusion of the L5-S1 disc space. The study includes all patients who wereoffered the procedure between December of 1993 andMarch of 1996. No patients were excluded from the study.There where 38 males and 37 females in the series withheights and weights typical of a normal cross section of thepopulation (Table 1). Several patients in the series would beclassified ‘‘morbidly obese’’ by all criteria. We were suc-cessful in completing the laparoscopic procedure in 73 ofthe 75 patients (97%), with one aborted procedure and oneconversion. Complications included two bladder lacera-tions, one retrograde ejaculation, and one patient with anextruded posterior fragment for an overall morbidity of 5%(Table 2). The only return to the operating room in thegroup was the patient with the extruded posterior fragment,who subsequently recovered without any significant sequelafrom the complication. No patient suffered bowel injury,vascular injury, ureteral injury, graft extrusion/migration, ordeath. Operative times averaged 3 hours and 12 minutes,with minimal blood loss. No patient required any transfu-sion of blood products. Patients were discharge on the av-erage within 36 hours from the time of their operative pro-cedure. Follow up studies on 23 of the patients who havebeen followed up for more than two years show that 18/23patients had marked improvement in their symptoms, and5/23 had moderate improvement (Table 3). No patient ex-

perienced a lack of improvement in their symptomatology.Average pain score on a modified pain scale of 1–10showed a pre-operative score of 8.7, and a post-operativescore of 2.3

Discussion

Although a number of laparoscopic access techniques havebeen described for spine surgery, they have been limited intheir application by providing access for only a simply dis-cectomy. Although simple discectomy has been the main-stay of treatment for lower back pain, it is limited in itsapplication, allowing relief for only neurogenic pain from aherniated disc. [7, 11] Furthermore, there is concern thatremoving the majority of the disc material will in itself leadto problems with instability. [5] If degenerative disc diseaseis truly a dynamic process as previously suggested, [4] thenonly discectomy with fusion will halt the progression of thedisease and give the best long term results. Studies havesuggested that the combination of discectomy with fusiongives significantly better results than simple discectomyalone. [9]

Spine surgery itself is controversial, with many unan-swered questions that will take well designed prospectiverandomized studies to answer. Because of these controver-sies, we sought a technique that would allow a ‘‘standard’’approach to the disc space to allow more ‘‘traditional’’methods to be utilized for management of the interspace.Although we did use an investigational device for the ma-jority of our fusions, it is a device that has been designed foropen anterior interbody fusion, and had only minor modi-fications made to adapt to the laparoscopic approach. Wehave inserted the more conventional bone plug in place ofthe carbon fiber cage, demonstrating that this access tech-nique will allow the spine surgeon the ability to apply awide range of operative spine techniques to this minimallyinvasive anterior approach. By approaching the procedurein such a way, we can eliminate much of the criticism thathas been offered regarding other laparoscopic techniquesthat have utilized proprietary instrumentation. Since thistechnique is not dependent on the method in which the disc

Table 1.Patient statistics (75 patients)

Number ofpatients

Average heights(range)

Average weights(range)

Males 37 58109 185[(5839–6859) (120[–260[)

Females 38 5849 157[(48119–5899) (105[–250[)

Table 2.Complications (Morbidity 5%)

• Laceration to the bladder (2)–Both identified at the time of surgery–Both repaired without laparotomy

• Avulsion of posterior edge of L5 (1)–Fragment displaced posterior into canal–Required reoperation to remove fragment

• Retrograde ejaculation (1)

Table 3. 2 year follow-up

• Activity–18/23 patients had significant improvement– 5/23 patients had moderate improvement–No patient reported lack of any measurable response

• Pain (modified pain scale, 0–10)–Pre-op (8.7)–Post op (2.3)

Fig. 3. Diagram showing the required angle of trajectory of the 15 mmtrocar, and the relationship of this trocar to the bladder. If the bladder ishigh ridding, or the disc space unusually low, the bladder is at risk.

1160

space is handled, any critical analysis of the procedure canbe limited to the risks of the laparoscopic approach, sincethe issues related to the method of dealing with the spinalfusion have been and will continue to be discussed in thespine literature. The issue to discuss, therefore, is with re-gards to the risks of the laparoscopic approach. The majorcomplications that we experienced in our series of 75 pa-tients consisted of two laparoscopic complications, and onecomplication related to the actual spine procedure. Since thecomplication related to the spine procedure is generic to anyanterior approach, we will not consider it in this discussion.

From our series the major laparoscopic risk was injuryto the bladder. Because the supra-pubic 15 mm trocar iscritical in its placement (Fig. 3), the patient’s anatomy dic-tates where that trocar must be placed. If the patient has anunusually low lying disc space, or an abnormally high ridingbladder, the bladder comes into potential danger with theinsertion of that trocar. Because the bladder is an extra-abdominal organ, it often can not be easily seen during thedirect insertion of this trocar. The key point for the surgeonto remember, is to consider this risk every time the suprapubic trocar is inserted, and carefully inspect the site as thetrocar is removed. If the dome of the bladder has beenlacerated, it can be easily repaired without laparotomythrough a small extension in the supra-pubic incision. Ifthere is question regarding the safety of insertion of thistrocar it is conceivable that the supra-pubic trocar could beinserted using an open technique, sweeping the bladderdome downward.

Retrograde ejaculation could be considered a complica-tion of the spine procedure, but since it is intimately relatedto the access technique we should consider it in our discus-sion. Although we did experience one patient with retro-grade ejaculation, it is hard to say if our complication wasfrom the approach or inherent to the spine procedure. Thepatient in question had a marked spondylolisthesis whichwas reduced with the procedure, and it is conceivable that itwas the reduction of this step off that was the cause of theretrograde ejaculation and not the approach. Although theincidence of this complication was low in our series, it hasto be considered in every male patient who is a potentialcandidate for an anterior procedure. This is especially true ifthe patient is younger, and still desiring a family. The key toavoiding, or at least minimizing, this potential injury is toadopt the attitude that no structure should be divided in thepresacral space. Although this is not physically possible,taking this attitude and trying to ‘‘sweep’’ the majority ofthe tissues to the side will go a long way in preserving thepresacral fibers. All division of tissues should be done in alongitudinal fashion in the direction of the fibers, and theonly structure that routinely needs to be divided to gainaccess to the disc space is the middle sacral vessels. Theseare easily identified, and lie directly on top of the annularligament.

A potential complication that has to be considered ismajor vessel injury. Although we did not experience anyinjuries to the major vessels, every procedure has this risk ifthe surgeon is not cautious. The vessel that is at most risk isthe crossing left iliac vein (Fig. 4). This is the ‘‘lowest’’ ofthe vessels and often lies just above the L5-S1 disc space. Itis this vessel that precludes this approach to be used rou-tinely to approach the L4-L5 disc space, since the vein

almost always lies directly over the interspace. During lap-aroscopy with the patient in a Trendelenburg position, andwith 15 mm of pressure within the abdominal cavity, thevein is often collapsed to a point that it is difficult to see. Forthis reason, the surgeon has to be very careful during hisdissection to not divide or cut any thickened tissues that arein the region of the iliac vein. Small bleeding points are besthandled initially with simple pressure to attempt hemosta-sis, since a small hole or rent in a vein will often seal withthis approach. If a small bleeding point develops, and thesurgeon immediately attempts control with dissection,clamping, or cautery, the small hole or tear can become amajor hole which will require open laparotomy to repair.The arteries generally do not present a problem, since notonly are they out of the operative field, but they are mucheasier to identify, and much more forgiving during routinedissection.

Because of the midline nature of the dissection, the ure-ters are not generally at risk. If the surgeon gets lost andventures off the midline, they could potentially be injured. Itis important for the surgeon to continually keep his land-marks in sight to avoid this possibility. The greatest dangerof getting off tract is in the heavier patient who has a lot ofpreperitoneal fat. The sacral promontory is not as obvious inthese patients, and the surgeon has to proceed much morecarefully to avoid from straying off the midline.

Fig. 4 Diagram showing the relationship of the great vessels to the lowerlumbar disc spaces.

1161

Avoiding bowel injury is best accomplished by cautiousdissection of adhesions and avoiding any manipulation ofthe bowel for retraction ‘‘off camera’’. In our technique,this is achieved by utilizing a static means of retraction thatdoes not rely on any mechanical retractors. All retraction ofthe intestine is accomplished using gravity, and stay suturesthat are placed on omentum, mesentery, or appendix epi-ploica. In this way, any tearing of tissue will not cause anyharm to the patient. An additional benefit from this tech-nique, is the ability to place multiple traction points withoutadditional trocar ports.

Summary

Although the topic of surgical treatment of lower back dis-orders is full of controversy and opposing opinions, we havedemonstrated a technique for an anterior surgical approachfor the L5-S1 interspace that has low morbidity, and allowsthe spine surgeon the ability to perform a complete discec-tomy along with an interbody fusion of the disc space. Thisincreases the surgical indications [2, 3] and allows a largergroup of patients the benefit of a laparoscopic approach.Although we relied on a carbon fiber cage for stabilizationof the fusion, the technique easily lends itself to more tra-ditional methods of fusion. The technique is not instrumen-tation dependent! When it is ultimately decided that a pa-tient is a candidate for surgical intervention, the anteriorlaparoscopic approach with fusion should be considered.

References

1. Cloyd DW, Obenchain TG, Savin M (1995) Transperitoneal laparo-scopic approach to lumbar discectomy. Surg Laparosc Endosc 5(2):85–89.

2. Crock HV (1991) Internal disc disruption. In: Frymoyer JW, DuckerTB, Hadler NM, Kostuik JP, Weinstein JN, Whitecloud TS (eds) TheAdult Spine, Principles and Practices. Raven Press. New York, pp2015–2026.

3. Hanley EN, Phillips ED, Kostuik JP (1991) Who should be fused? In:Frymoyer JW, Ducker TB, Hadler NM, Kostuik JP, Weinstein JN,Whitecloud TS (eds) The Adult Spine, Principles and Practices. RavenPress. New York. pp 1893–1917.

4. Kirkaldy-Willis WH, Farfan HF (1982) Instability of the lumbar spine.Clin Orthop 165: 110–123.

5. Mixter WJ, Barr JS (1934) Rupture of the intervertebral disc withinvolvement of the spinal canal. N Engl J Med 211: 210–215.

6. Obenchain TG (1991) Laparoscopic lumbar discectomy. J Laparoen-dosc Surg 1: 145–149.

7. Spengler DM, Frymoyer JW (1991) Lumbar discectomy: Indications& technique. In: Frymoyer JW, Ducker TB, Hadler NM, Kostuik JP,Weinstein HN, Whitecloud TS (eds) The Adult Spine, Principles andPractices. Raven Press. New York. pp 1785–1800.

8. Stein S, Slotman GJ (1994) Laser-assisted laparoscopic lumbar dis-kectomy. N J Med 91: 175–176.

9. Vaughn PA, Malcolm BW, Maistrelli GL (1988) Results of L4-L5 discexcision alone versus disc excision and fusion. Spine 13:690–695.

10. Weber H (1983) Lumbar disc herniation. A controlled, prospectivestudy with ten years of observation. Spine 8: 131–140.

11. Wood GW (1992) Lower back pain and disorders of the intervertebraldisc. In: Crenshaw AH (ed) Campbell’s Operative Orthopaedics 8th ed.Mosby Year Book Inc. St. Louis. pp 3715–3790.

12. Zelko JR, Misko J, Swanstrom L, Pennings J, Kenyon T (1995) Lap-aroscopic lumbar discectomy. Am J Surg 169: 496–498.

Discussion

Dr. Moviel:Who is in charge of these patients, and who isdoing the operations?

Dr. Olsen: I believe, and I think many of my colleaguesbelieve, that this is an operation that is a co-surgical proce-dure, a procedure that requires the specialities and expertiseof two separate surgical specialties: a laparoscopic surgeonto not only give the initial access, but to maintain the accessand, more specifically, to avoid any complications with theblood vessels, the bowel, bladder and other intra-abdominalstructures.

The spine portion of the procedure is appropriately dealtwith by the spine surgeon; someone who has been trainedand knowledgeable in all of the intricacies and issues relatedto the spine and that’s why I specifically avoided mention-ing the precise technique of inter-body fusion, and empha-sized that the access technique we did is applicable forvirtually any sort of anterior fusion technique that the spinesurgeon desires.

Dr. Slotman:Several institutions are doing laparoscopicdiscectomy as an outpatient surgery, and yet yours, 36 to 48hours or staying one or two nights overnight; is that pro-longed stay related to the laparoscopic procedure or to theaddition of hip and spine surgery?

Dr. Olsen:A lot of the extra time is due to my conservatism.You can discharge these patients in 24 hours, essentially onan outpatient basis. But when you are dealing with patientswho are morbidly obese and have had extensive abdominaladhesions, and you do a lot of adhesiolysis and mobilizationof the bowel, these patients may develop a slight degree ofileus. Until you have watched the patients for 24 hours, andallowed them to slowly progress their diets, you are notgoing to be able to know whether or not they’re going toexhibit problems and at least for myself, at this point intime, I have insisted on patients having an overnight stay;and at least for the first part of the next day, getting themback on their diets and making certain that they are up andabout and eating without any difficulty.

And that’s why our hospitalization time has been ap-proximately 36 hours.

Dr. Kim: What are your tips to reduce the incidence of aretrograde ejaculation?

Dr. Olsen:That particular complication, I think, would be acombination of the work by both the general surgeon andthe spine surgeon. It’s often difficult to determine whetheror not retrograde ejaculation is due to the presacral dissec-tion, or in our case where there was a severe spinal adhesisthat we reduced, which part of the operation was responsiblefor that. Most often, however, it’s going to be due to thepresacral dissection and that’s why I emphasized that you

Table 4.Results

• Operative time 3 hrs 12 min.• Discharged in 36 hrs• Return to “normal” activities 1–2 weeks• Successful in 97% patients

–One patient converted due to extensive adhesions–One patient aborted due to presacral scar

1162

want to do a longitudinal division of the structures, andsweep the structures laterally so that you preserve the pre-sacral plexus to minimize that potential risk. But the mostimportant thing is to counsel patients preoperatively aboutretrograde ejaculation being a potential complication. Aswhen we do colon work, and in so doing the low anteriorresection it behooves you to tell patients that there is apotential risk because it has been described and can occur.

Dr. Kim: Once you open the peritoneum, do you use elec-trocautery or something else?

Dr. Olsen: I use bipolar cautery until we are down to theannulus.

My spine surgeon likes to use monopolar, so once weget down to the annulus and I expose the area, he usesmonopolar to divide the annulus.

Dr. Kim: Don’t you think that electrocautery may damagesome nerve tissue there?

Dr. Olsen:Well, I use bipolar cautery and divide the tissuesin such a way that those nerves, I believe, are preserved.And I think our series demonstrates that that is the case.

Dr. Vignati: We are doing the same procedure through asmall peritoneal incision with a retroperitoneal exposure;and I was wondering if you have any comparison with thisapproach to the laparoscopic approach? And in addition,have you gone to any other levels through the laparoscopicapproach?

Dr. Olsen:To answer your second question, yes. We havebeen to other levels through the laparoscopic approach.

But I would not present that material at this time, be-cause I do not believe that that technique, given the expo-sure necessary to do a standard anterior interbody fusiontechnique, is yet worked out to the point that I feel com-fortable demonstrating and showing it.

The L5/S1, I believe, is very well worked out.

1163