1
`
`NUVASIVE 1067
`NuVasive, Inc. v. Warsaw Orthopedic, Inc.
`IPR2013-00206
`IPR2013-00208
`
`

`

`
`
`SPINE Volume 23, Number 13, pp 14761-1484
`@1998, Lippincott—Raven Publishers
`
`I Minimally Invasive Anterior ‘
`Retroperitoneal Approach to the
`Lumbar Spine
`
`Emphasis on the Lateral BAK
`
`|V|D,* John J. Regan, MD,T W. Peter Gels, MD,i
`Paul C. lVchfee,
`and Ira L. Fedder,
`|\/lD*
`
`Study Design. Eighteen patients with lumbar inste-
`bility from fractures, postlamlnectomy syndrome. or in-
`fection were treated prospectively with minimally Inva-
`sive retroperitoneal lumbar fusions.
`Objectives. To determine if interbody Booby and
`Kuslich fusion cages and femoral eiiogratt bone dowels
`can be inserted in a transverse direction via a lateral
`endOeoorplc retroperitonaal approach to achieve spinal
`stability.
`Summary of Background Data. Endoscopic spinal
`approaches have been used to achieve lower lumbar
`fusion when instrumentation is placed through a laparofi
`ecopio, transparitoneal route. However. complications of
`using this approach include postoperative intra-ahdomlv
`nal adhesione. retrograde ejaculation. great vessel ins
`jury, and implant migration. This study is the first clini-
`cal series investigating the use of the lateral
`ratroperitoneel minimally invasive approach for lumbar
`fusions from L1 to L5.
`Methods. Eighteen patients underwent anterior inter-
`body decompression and/or stabilization via endoscopic
`retroperitoneal approaches. In most oases. three 12vmm
`portals were used. Two parallel transverse interbody
`cages restored the neurotoranrlnai height and the da-
`eired amount of lumbar lordoeie was achieved by in-
`serting a larger anterior cage. distraction plug, or bone
`dowel.
`.
`, Results. The overall morbidity of the procedure was
`lower than that associated with traditional "open" retro-
`peritoneai or leparotomy techniques, with a mean
`length of hospital stay at 2.9 days (range. outpatient
`procedure to 5 days). The mean estimated lntraopera-
`tive blood loss was 205 cc (range, 25in00 cc). There
`were no cases of implant migration. significant subsi-
`dence, or pseudoerthrosis at mean follow-up examina-
`tion of 24.3 months (rapes, 12-40 months) after
`surgery.
`
`
`
`From the *Scoliosis and Spine Center, St. Josephs Hospital, Baltimore,
`Maryland, the tTexas Back Institute and the Institute for Spine and
`Biomedical Research, Plano, Texas, and the iMinimally Invasive Ser-
`vices Training Institute, St. Josephs Hospital, Baltimore, Maryland.
`Acknowledgment date: August 28, 1997.
`First revision date: October 23, 1997.
`Acceptance date: December 2, 1997.
`Device status category: 9.
`
`1476
`
`Conclusions. This preliminary study of 18 patients
`illustrates that endoscopic techniques can be applied
`effectively through a ratroperltoneal approach with the
`patient in the lateral position. Unlike the patients who
`had undergone transperitoneal procedures described in
`previous reports, in these preliminary 18 patients, there
`were no cases of retrograde ejaculation, injury to the
`great vessels, or implant migration. [Key words: endo»
`scopic retroperitoneal, minimally invasive retroperlto—
`neal lumber fusions, transverse axis BAK] Spine 1998;
`23:1476—1484
`
`The use of minimally invasive and endoscopic ap-
`proaches has been described for multiple abdominal pro-
`cedures, including cholecystectomyfgfi1’32 appendecto—
`my,28 colon resection,12 and Nisson fuudoplication,”
`Recently, increased attention has been paid to the use of
`these approaches with lumbar discectomyZS’26 and lum—
`bar anterior interbody arthrodesis.1’4’8 Most endoscopic
`approaches described thus far have been transperitoneal
`and have depended on C02 insufi’lation to provide work—
`ing space and to retract the small bowel out of the sur—
`gical field. Gaur6 and McDougall et all24 were the first to
`describe retroperitoncoscopy, an endoscopic retroperi-
`toneal approach for urologic procedures. The current
`report describes the natural transition toward retroperi-
`toneal minimally invasive endoscopic spinal surgery,
`which does not require C02 insufflation, Trendelenburg
`position, entrance into the peritoneum, or anterior dis-
`section near the great vessels to provide safe exposure for
`spinal surgery.
`I Materials and Methods
`
`Twelve minimally invasive retroperitoneal lumbar procedures
`Were performed at St. Josephs Hospital in Baltimore, Mary-
`land, and six were performed at Presbyterian Hospital of
`Plano, Texas, between March 1994 and September 1996.
`There were 6 female and 12 male patients, with a mean age of
`53.4 years (range, 31—76 years).
`The indications for surgery included 13 cases of degenera-
`tive conditions, three Cases of infections, one unstable burst
`fracture, and one case of a retroperitoneal neurofibroma in-
`
`MC0000006
`
` 2
`
`

`

`1477
`Endoscopic Retroperitoneal in terhody Fusions ° McAfee et al
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Figure 1. This 75-year-old man
`had back pain and right anterior
`thigh pain 2 years after he had
`undergone laminectomies from
`L3 to Sl with a posterolateral fu-
`sion from L4 to Si, The lateral (Al
`and anteroposterior (B)
`radio-
`graphs show “vacuum disk" sign
`at L3—L4 with lateral translation
`of the L3 vertebral body on L4.
`His characteristic pain was re-
`produced by an L3—L4 discogram
`performed by an independent ra-
`diologist. Lateral (C) and antero-
`posterior (D) radiographs were
`obtained after the procedure us—
`ing the endoscopic retroperito-
`neal approach was performed
`and a transversely oriented BAK
`fusion cage was inserted (15 mm
`in diameter and 24 mm lengthl.
`The patient's back and right leg
`pain resolved after surgery.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`volving the lumbosacral plexus. Ten of the 13 patients in the
`degenerative category had undergone previous destabilizing
`laminectomy procedures elsewhere before referral to the au-
`thors’ institutions. Flexion~ extension lateral radiographs dem—
`onstrated more than 3.5 mm of translation, and anteroposte—
`rior radiographs showed 10 degrees or more of scoliotic disc
`space collapse with “vacuum disc sign.” Positive discograms
`documented a provocative pain response at the unstable level
`(Figure 1). The one patient in the degenerative category who
`had not undergone previous destabilizing surgery had a mas-
`sive 1.2—L3 central disc herniation with left quadriceps weak~
`ness.
`
`The retroperitoneal approach proved to be very versatile in
`the range of vertebral levels addressed throughout the 18 cases.
`Four patients underwent procedures at Ll—LZ, seven patients
`at L3—L4, and two patients at L4—L5. There were four endo—
`scopic dccompressions and fusions at L2—L3. One patient with
`vertebral osteomyelitis underwent a decompressive procedure
`from L2 to L4.
`
`Surgical Technique. The approach is a combination of Video—
`assisted thoracoscopic and laparoscopic methods. The patient
`is put under general endotracheal anesthesia, then turned in the
`
`MCOOOOOO7
`
` 3
`
`

`

`
`1478 Spine ' Volume 23 ' Number 13 - 1998
`
`
`
`Figure 2. A view of a transparent optical trochar (Optiview, Ethi-
`con Endosurgery, Cincinnati, OH) that was used in dissecting the
`retroperitoneal space. Notice the "winged keel cutting edges,"
`which only wiil penetrate a fascial layer, such as the peritoneum,
`if the trochar is forcibly twisted backwards and forwards.
`
`lateral decubitus position on a radiolucent, graphite, Jackson
`Maximum lateral access table (0.5.1. Corporation, Union City,
`CA) made specifically for the endoscopic approach, with side
`rails designed to accommodate robotic arms7 and to facilitate
`c—arm fluoroscopy. A 1-cm incision is made at the anterior
`portion of the 12th rib for approaching from L1 or L2. Below
`L2, a lateral egarm fluoroscopic image is obtained, with a metal
`marker overlying the patient’s 5k in in the midaxillary line. This
`method optimizes the placement of the working portal directly
`over the unstable disc or vertebral segment. The three tech—
`niques used to dissect the rctropcritoncal space are: finger dis—'
`section, balloon insufflation, or the use of an optical, transpar-
`ent, dissecting trochar7 called an Optiview (Ethicon
`Fndosurgery, Cincinnati, OH; Figure 2).
`The 107mm laparoscope is inserted into the Optiview dis—
`secting trochar and refocused once the trochar enters the sub—
`cutaneous tissue. The trochar has two “winged keel” cutting
`surfaces that will not penetrate a fascia] layer such as the peri—
`toneum unless the trochar is twisted. Therefore, the three ab—
`dominal muscular layers overlying the peritoneum are pene—
`trated in sequence under direct visualization until the
`preperitoneal fat is encountered. The trochar is used to create a
`potential space that is superficial to the peritoneum until the
`laterally oriented fibers of the psoas major muscle are viewed.
`Usually, the genitofemoral nerve is visualized on the surface of
`the psoas muscle. At this juncture, a dissection balloon, such as
`that manufactured by Origin (Menlo Park, CA), can be filled
`with 1 liter of normal saline or air to dissect the rctropcritoncal
`layer, more correctly referred to as the retrotransversalis fascia.
`Alternatively, carbon dioxide insufflation can be forced into the
`rctropcritoncal cavity up to a pressure of 20 mm of mercury to
`create a working space to triangulate endoscopically.30 Once
`the retroperitoneal space is enlarged, at least three portals are
`used—working portal, for pituitary rongeur; curettes; a high—
`powered burr; or Kerrison rongeurs. A second portal is neces—
`sary for the ’l,O-mm laparoscope. A third portal is used for
`retraction of the psoas major muscle off of the spine in a pos-
`terior direction. The relatively avascular intervertebral discs are
`exposed first. Then, the respective midportions of the adjacent
`vertebral bodies are exposed, and the lumbar segmental vessels
`
`are ligated and divided. Occasionally, a fourth 10-mm portal is
`used for suctioning in highly vascular cases requiring corpecl‘oe
`mies for tumors or infections. Occasionally, for longer strut
`grafts or instrumentation, the 10—mm working portal is ex—
`tended in size as much as 5 cm, and an endoscopically assisted,
`mini~laparotomy type of rctropcritoncal exposure facilitates
`the corpectomy or spinal instrumentation. If the size of the
`working portal is extended, of course, the CO2 insufflation is
`lost, and the working space in the retroperitoneum has to be
`maintained by using retractors. This technique is advantageous
`because the spinal decompression can be accomplished without
`airtight seals, and because standard thoracoscopic instruments
`can be used on the lumbar spine. In other words, throughout
`the remainder of the procedure, spine instruments of heterog—
`enous shaft diameters can be used, and airtight seals around
`trocars are not required.
`Once the vertebral level is confirmed fiuoroscopically, the
`transversalis fascia, pcrinephric fascia, and rctropcritoncal
`contents are retracted anteriorly (Figure 3). Electrocautery is
`used to mark the intervertebral discs adjacent to the involved
`lesion. For example, for an L1 corpectomy, the Ll—LZ and the
`T’lZ—Iil, intervertebral disc spaces are marked. A left-sided ap~
`proach to the surgery is preferred to a right—sided approach,
`because it is easier to dissect the aorta off the spine than to
`dissect around the more friable inferior vena cava; this is par—
`ticularly true in cases of pyogenic vertebral osteomyelitis or
`cases of neoplasm that occur after radiation therapy with ret-
`roperitoneal fibrosis. The psoas muscle is retracted posteriorly,
`and the ureter is retracted anteriorly.
`If a corpectotny is being performed21 after the two adjacent
`discectomies, the surgeon must have access to three methods of
`hemostasis: 1) Endo-Avitene Microfibrillar Collagen (Alcon,
`lnc., Humacao, Pucrto Rico), 2) Gelfoam (Upjohn Corp.,
`Kalamazoo, MI) soaked in Thrombin (GenTrac Corp, Middle—
`town, \WI), and 3) bipolar endoscopic electrocautery. At this
`point, the segmental vessels are dissected from the underlying
`bone and elevated with a right—angled clamp. It is importanl to
`use two vascular clips or an endoloop on the high—pressure side
`of the vessels; the vessels are divided with endoshears. As a
`general rule, with any spine procedure the segmental vessels are
`ligated and divided in the anterior half of the vertebral body to
`allow collateral circulation to the neuroforamen and spinal
`cord to occur to its maximum potential. If the lesion is a tumor
`or infection, then a culture and a frozen section are obtain ed at
`this time in the procedure. A 45—degree, 47mm~wide endo—
`scopic Kerrison rongcur is used to resect the pedicle. Starting
`cephalad, the instrument is pointed caudad to protect the exit-
`ing spinal roots. Either Kaneda (Acromed Corp., Cleveland,
`OH) heavy—duty rongeurs or a high-powered, 5~mm burr, such
`as the Ziinmer (Wausau, IN) Ultra—power or Anspach with
`long extensions, can be used to hollow out the vertebral body.
`Curettes and small 2e3—mm Kerrison rongeurs are used to com~
`plctc the corpectomy. It is important to decompress the spinal
`canal all the way across to the base of the opposite pedicle.
`Decompression is accomplished only when the opposite pcdiclc
`is palpated or visualized. An autogenous iliac strut graft is
`ramped into place, filling the anterior portion of the corpec-
`tomy defect.
`Alternatively, if the patient only requires a discectomy or
`one-level fusion, an anterior interbody fusion can be accom—
`plished endoscopically. The disc space height is restored by
`using a distraction plug placed from the side. Optitnally, two
`
`MCOOOOOOB
`
` 4
`
`

`

`
`
`Endoscopic Retroperitoneal Interbody Fusions - McAfee et al 1479
`
`distraction plugs are tamped into the disc space: one anteriorly
`and one posteriorly. At this point, either a single—barrel or dou-
`ble-barrel drill tube is placed over the distraction plugs. The
`position of the distraction plugs is monitored with anteropos-
`terior and lateral fluoroscopy. The center of the distraction
`plugs will correspond with the center of the BAK interbody
`fusion cages or endoscopic bone dowels.1 The double-barrel
`tube is tamped into place to engage its teeth into the superior
`and inferior vertebral bodies to maintain the normal height of
`the disc space (luring the reaming and tapping of holes into the
`intervertebral endplates. The BAK fusion cage or laparoscopic
`bone dowels, which are composed of femoral allograft, are
`packed with autogenous iliac graft. The morselized iliac au-
`tograft can be harvested with minimally invasive techniques
`through a 12—mm incision by using a disposable T—shaped awl.
`After surgery, the patient is placed in a warm and form
`corset (interbody fusion) or a thoracolumbar sacral orthosis
`(after a corpectomy), until radiographic fusion is accom—
`plished. Intraoperatively, it is important to countersink the
`BAK fusion cages or laparoscopic bone dowel. The authors
`advocate packing additional bone graft superficial to the cage.
`At 3—6 months after surgery, arthrodesis can be confirmed if
`solid trabecular bone is observed to bridge one vertebrae to the
`adjacent level; this is the most reliable radiographic sign of a
`solid arthrodesis. Three-dimensional, computed—tomography
`reconstruction images of the bone within the cages and flexion—
`extension lateral radiographs also can provide useful informa—
`tion.
`
`femoral allograft. The long axes of the implants were
`positioned in a transverse direction.
`There were no cases of implant migration or pseudo-
`arthrosis. There were no cases of a radiolucent interface
`between the implant and the vertebral body. There were
`no cases of subsidence more than 1 mm, and there was
`trabccular bony bridging across the adjacent vertebrae
`laterally by 6 months after surgery.
`One additional patient did not undergo an endoscopic
`stabilization procedure. He was a 47—year—old man with
`an unstable burst fracture who had undergone left ante-
`ri or Kaneda instrumentation at the referring institution 3
`months earlier. He had had an incomplete neurologic
`deficit; a preoperative computed tomography scan had
`demonstrated continued right-sided cauda equina com—
`pression. A right-sided endoscopic decompression was
`performed at the authors’ institution, and additional sta-
`bilization was not required.
`For all 18 cases, the mean duration of the surgical
`procedure, including the harvesting of iliac crest bone
`autograft, was 115.2 minutes (range, 60 —260 minutes).
`The mean estimated blood loss, which, at the authors’
`institutions, is determined by the attending anesthesiol-
`ogist, was 205 cc (range, 25—1000 cc). The mean length
`of hospital stay was 2.9 days (range, outpatient proce—
`dure to 5 days).
`
`I Results
`
`The mean length of the postoperative follow—up period
`was 24.3 months (range, 12—40 months). Fourteen pa—
`tients underwent left-sided retroperitoneal approaches,
`and, in four patients, the pathology was addressed more
`easily on the right side. There were four patients who in
`whom a single incision was made of 5 centimeters or less.
`Fourteen patients had either three or four portals mea-
`suring approximately 12 mm in length. These fourteen
`patients had CO2 insufflation to assist the retrotransver-
`salis dissection. The patients with one incision of 5 cen-
`timeters or less had lesions compatible with infection or
`tumor, and the use of C02 insufflation was avoided to
`prevent pressurizing the tumor cells or bacteria systemi-
`cally into the patient’s bloodstream.9’10’11’15
`Fusions were performed in 15 of 18 cases by using
`structural bone graft and/or interbody fusion cages. A
`38—year—old radiologist’s wife with a neurofibroma aris-
`ing from the lumbosacral plexus adjacent to the left com—
`mon iliac vein did not demonstrate preoperative or in-
`traoperative instability; therefore, a fusion procedure
`was not indicated.
`
`Ten patients underwent fusion surgery with custom
`BAK interbody fusion cages. The long axes of the cages
`were in the transverse direction (Figure 4).4 In each case,
`the cages were packed with autogenous iliac bone graft
`harvested using a minimally invasive, T—shaped trochar,
`that was 10 mm in diameter. Four patients underwent
`placement of laparoscopic bone dowels fashioned from
`
`Complications
`There were three patients with postoperative complica—
`tions. Case 2 was a 71-year—old man on renal dialysis
`who presented with sepsis. Endoscopic, retroperitoneal,
`L3—l.4 discectomy; debridement; and fusion were per-
`formed to culture and manage an L3—L4 pyogenic osteo-
`Inyelitis. Six weeks after surgery, after treatment with
`intervenous antibiotics, the patient underwent posterior,
`segmental stabilization with Texas Scottish Rite Hospi-
`tal implants from L1—L5 for more definitive stabilization
`and fusion. The single—level, anterior, interbody, endo—
`scopic fusion was not believed to be adequate to prevent
`long-term lumbar kyphosis and instability.
`The second complication occurred intra operatively in
`Case 3, when a laparoscopic bone dowel partially frac-
`tured at the point of attachment of the driver into the
`central drilling peg in the femoral cortical allograft. The
`fractured piece of allograft was extremely small (approx—
`imately 0.5 cm X 0.5 cm X 0.5 cm), and the patient’s
`spinal stability was not jeopardized. There were no long-
`term sequelae, and the patient had a solid arthrodesis,
`which was facilitated by morselized iliac autograft
`placed in the central chamber of all laparoscopic bone
`dowels.
`
`The third complication occurred in a 69-year-old man
`who underwent transverse BAK cage insertion for post-
`laininectomy instability at L3—L4. The patient developed
`a hematoma in the psoas muscle at L3—L4 after surgery.
`This resulted in a temporary genitofemoral nerve palsy,
`which resolved spontaneously within 3 months.
`
`MCOOOOOOQ
`
` 5
`
`

`

`1480 Spine 0 Volume 23 0 Number 13 0 1998
`
`l Discussion
`
`Retropcritoneal lumbar fusion and stabilization offers
`several advantages over conventional anterior transperi-
`toneal laparoscopic approaches of the lumbar spine.8
`Retroperitoncal approaches obviate the risk of small
`bowel obstruction or postoperative intraperitoneal ad-
`hesions.16’17 Additionally, there should be a reduced risk
`of retrograde ejaculation, because the autonomic plexus
`is not dissected, in contrast to preliminary reports of
`transperitoneal laparoscopic approaches.13’22’30 The pa-
`tient is in the lateral decubitus position, which facilitates
`exposure of the lumbar spine, as gravity helps retract the
`abdominal contents anteriorly. With the straight, lateral
`position, as opposed to the supine Trendelenburg posi—
`tion required for transperitoneal laparoscopy, it is easier
`to get orthogonal to the disc space and spine with later—
`ally directed placement of interbody threaded fusion
`cages. The surgeon can use two longer cages in the trans-
`verse axis, with a larger—diameter cage anterior and a
`smaller-diameter cage posterior, thus “customizing” or
`“dialing in” the optimal degree of intervertebral lordosis
`(Figure 5). The same effect of altering the sagittal plane
`alignment was achieved in the majority of patients in this
`study by using distraction plugs of different diameters to
`adjust the degree of lordosis even before reaming the
`cage diameter. Two longer cages placed laterally via the
`retroperitoneal approach should be biomechanically
`more stable, because the surface area of the vertebral
`body cage contact area is greater than that achieved
`when using a transperitoneal approach. In addition, the
`anterior longitudinal ligament and posterior longitudi-
`nal ligament are not violated with the lateral retroperi—
`toneal approach. With the transperitoneal approach, if
`the surgeon reams, taps, or drills too deeply, the spinal
`canal contents are at risk. With the lateral retroperito-
`neal approach, however, the orthopedic drilling, ream-
`ing, tapping, and cage insertion are directed toward the
`contralateral psoas muscle instead of the neurologic
`structuress’zz
`In the report of the laparoscopic BAK study13 submit—
`ted to the Food and Drug Administration, the incidence
`of iatrogenic intraoperative disc herniation in patients
`undergoing surgery at one level was 2.8% (6 of 215
`patients) and that in patients undergoing transperito-
`neal, two-level BAK was 12% (3 of 25 patients). Overall,
`for BAK implants inserted via a straight anterior—to-
`
`
`posterior direction, the incidence of reoperation for iat-
`rogenic penetration or for pushing intervertebral disc
`material into the spinal canal was 2.3%. Furthermore,
`lateral retroperitoneal proceduresyobviate the need to
`dissect and mobilize the common iliac vein and artery, as
`is necessary with transperitoneal exposure of the L4 —L5
`intervertebral disc. The authors have found that it is eas~
`icr to make the long axis of two retroperitoneal cages or
`bone dowels parallel via a shotgun or double—barrel tro—
`char than via a transperitoneal independent trochar.
`Mayer18 reported on 20 patients who underwent ret-
`roperitoneal, microsurgical, anterior lumbar interbody
`fusion between January 1, 1995 and January 31, 1996.
`In this procedure, an extensive quadrilateral retraction
`frame was used, but anterior stabilization implants were
`not. All patients had undergone an additional posterior
`pedicle screw instrumentation procedure 1—2 weeks ear—
`lier.
`Boden et al2 described a video-assisted, lateral, inter-
`transverse—process arthrodesis in a rabbit and a non—
`human primate model. This was a posterior approach,
`which did not involve spinal stabilization.
`Ordway et al27 compared the biomechanical charac-
`teristics of a transversely oriented carbon-fiber cage with
`those of an anteriorly oriented cage in the bovine lumbar
`spine and found that, in most cases, the differences were
`not statistically significant. In addition, Ordway et al
`compared two anteriorly oriented cages with just one
`transversely oriented cage, which is a biased comparison.
`However, the carbon fiber cage has been studied exten—
`sively at the current authors’ laboratory as well3; it is
`basically an unreamed spacer that is not screwed into
`place. The BAK system is inherently more stable than the
`carbon fiber cage because it uses preinsertional distrac-
`tion via distraction plugs, which results in better liga-
`mentotaxis. The BAK involves a tap that cuts threads
`into the two adjacent vertebral end plates, and the BAK
`reduces the strain and micromotion on the bone graft
`contained within the cage more successfully than does
`the rectangular carbon-fiber design.14
`There are several potential disadvantages of the min-
`imally invasive, retroperitoneal approach. Particularly at
`L4 —L5, it may be necessary to remove part of the iliac
`crest or place the docking portal through the ili ac wing to
`be orthogonal to the L4 —I,5 disc space.22’30 In addition,
`a large mass of psoas muscle containing lumbosacral
`)
`
`Figure 3. A, Schematic diagram showing the orientation for the retroperitoneal approach. The ”X” marks the anterosuperior iliac spine,
`and the three portals are shown by black dots. B, Initial lateral radiograph of a 76-year-old, dialysis-dependent man with biopsy-confirmed
`osteomyelitis at L3—L4. Initially, he was treated at another institution with intravenous antibiotics and a thoracolumbosacral orthosis. When
`he arrived at the authors' institution, he was in septic shock. c, A sagittal magnetic resonance image revealing vertebral collapse with
`an associated retropulsed vertebral body fragment and thecal sac compression. D, A schematic diagram depicting the laparoscopic View
`through the transversalis fascia as the L3—L4 intervertebral disk is exposed. E, A laparoscopic view of the retroperitoneal space. The black
`arrows demonstrate the interval used to reach the vertebral body. The psoas major (Pl
`is posterior, and the ureter (U)
`is anterior.
`Perinephric fat (K) is dissected bluntly to expose the interval. F, A schematic diagram showing debridement of the infected granulomatous
`tissue with the anterior thecal sac exposed between the L3 and L4 vertebral bodies posterior to the pituitary rongeur. G, A laparoscopic
`photograph showing a pituitary rongeur debriding the infected L3—L4 intervertebral disc.
`
`mcoooom o 7
`
` 6
`
`

`

`
`Endoscopic Retroperitoneal Interbody Fusions - McAfee et al
`1481
`
` Retractor
`
`ACamera
`
`Endoshears
`
`
`
`,nephreniqv
`
`, Fat\\ ‘
`
`Rongeur
`
`Pituitary
`
`MCOOOOO11
`
` 7
`
`

`

`1482 Spine - Volume 23 - Number 13 - 1998
`
`
`
`Figure 4. Anteroposterior and
`lateral
`radiographs of
`two pa—
`tients obrained after endoscopic
`retroperitoneal BAK interbody iw
`sion was performed and instru-
`mentation for postlaminectomy
`instability was placed. A and B
`Show a one—cage technique. 0
`and D Show a two-cage tech-
`nique.
`
`nerve roots may need to be mobilized laterally. However,
`the authors of this study still prefer mobilizing the psoas
`muscle to mobilizing the common iliac vein and artery,
`as is necessary with transperitoneal approaches. The po—
`sition of the ureter constantly needs to be considered in
`transperitoneal and retroperitoneal minimally invasive
`approaches. One additional advantage of the retroperi-
`toneal approach is that spine surgeons probably will be
`more comfortable performing the approach without de—
`pending on a general surgical “access surgeon.”
`Obviously, it is difficult to prove statistically that the
`incidence of complications associated with the retroper—
`itoneal approach is lower than that associated with the
`intraperitoncal endoscopic spinal procedures, because
`even transperitoneal spinal fusion is only a recently de-
`scribed procedure and is still
`in its infancy. However,
`
`many general surgical, gynecologic, and urologic proce—
`dures in which the retroperitoneal approach is used have
`been reported showing fewer complications than their
`transperitoneal operative counterparts. Leverant et al17
`showed that intra—abdorninal adhesions do occur with
`
`laparoscopy. They found intraperitoneal adhesions in 79
`of 124 patients whose only prior surgery was laparos—
`copy; none of the 91 controls without prior laparosco—
`pies in that study had adhesions. lajer et al16 reported a
`1% incidence of hernias in trocar ports after abdominal
`laparoscopy. Hernias through trocar ports have not been
`described with retroperitoneal approaches.
`The authors of the current study have performed more
`than 150 endoscopic spinal procedures, and have had a
`patient with postoperative small bowel obstruction. This
`obstruction occurred in a patient 2 weeks after he had
`
`
`
`
`
`
`
`
`
`
`
`
`
`MCOOOOO12 I
`
` 8
`
`

`

`Endoscopic Retroperitoneal Interbody Fusions - McAfee et al
`
`1483
`
`Kyphosis
`
`Neutrai
`
`Lordos’is
`
`
`
`
`
`
`
`Retreperitoneal “Transverse Axis" Cages canoe used 7
`to "dial in" tne'deslred kyphosis or lordosis.
`
`'
`
`Figure 5. A schematic diagram illustrating how differential sizing
`of transversely oriented distraction plugs, interbody bone dowels,
`or fusion cages can "dial
`in" or adjust the desired amount of
`lumbar kyphosis or lordosis through a minimally invasive retroper—
`itoneal approach.
`
`undergone anterior transperitoneal laparoscopic L4—L5
`fusion, when the greater olmentum became adherent to
`the posterior peritoneum.
`Tiusanen et al34 reported an incidence of retrograde
`ejaculation of 5.9% as a complication of anterior inter-
`body fusions and found that it occurred only after trans—
`abdominal procedures. In the report of the first series of
`240 laparoscopic BAK interbody fusions and stabiliza—
`tions13 submitted to the Food and Drug Administration,
`there were 12 cases (5%) of retrograde ejaculation that
`occurred as a complication of laparoscopic procedures.
`Although the numbers are too small to analyze statisti—
`cally, there probably is a higher incidence of this compli—
`cation at LS—Sl exposure than at L4 —L5 exposure; it has
`been described to occur after anterior fusions to L4 and
`
`with periaortic lymph node dissection. Retroperitoneal
`exposure, either endoscopic or conventional, is associ-
`ated with a lower incidence of this postoperative compli—
`cation. ‘8730
`It is difficult to compare the morbidity of traditional
`versus minimally invasive, endoscopic, anterior, retro-
`peritoneal approaches in the orthopaedic literature di—
`rectly, because the length of hospital stay, operative time,
`and length of time out of work have not been reported.
`However, there are three studies of traditional, retroper—
`itoneal, anterior decompressions of the spine in which
`the current authors have participated and reported: those
`of McAfee et ad20 with 70 patients, McAfee19 with 185
`patients, and McAfee and Zdeblick23 with 23 patients.
`Overall, it is the authors’ impression that the incidence of
`complications and morbidity is much lower for endo-
`scopic procedures; in particular, associated medical peri—
`operative complications are reduced, including urinary
`tract infections, post-operative atelectasis, and pneumo—
`nia. However, further experience is clearly necessary to
`confirm this impression. The authors currently are par-
`ticipating in a prospective, multicenter trial by the Na-
`tional Institute of Health to investigate further the coni—
`
`plication rates of conventional versus minimally invasive
`techniques.
`In summary, in this report of 18 cases of minimally
`invasive, lateral retroperitoneal, lumbar procedures, the
`main advantages were:
`
`1 ) the length of stay was less than that associated with
`larger muscle-splitting “open approaches,”
`2) the need for mobilization of the great vessels was
`reduced compared with that of transperitoneal lapa-
`roscopic approaches,
`3) the procedure allowed for two larger cages to be
`inserted parallel to one another in a transverse direc-
`tion, horizontal, rather than perpendicular, to the spi—
`nal canal.
`
`4) by varying the diameter of the retroperitoneally
`placed interbody fusion cages, interbody allograft
`bone dowels, or distraction plugs, customization of
`the amount of lumbar lordosis was possible without
`increasing the risk of implant dislodgement or
`pseudoarthrosis.
`
`Eighteen cases is not a large series, but the results are
`favorable compared with preliminary results of alterna-
`tive techniques.
`
`References
`
`1. Bagby G. Arthrodesis by the distraction—compression
`methods using a stainless steel implant. Orthopaedics 1988,11:
`931—4.
`2. Boden SD, Moskovitz PA, Morone MA, Toeibitaby Y.
`Video—assisted lateral intertransverse process arthrodesis: Val—
`idation of a non-minimally invasive lumbar spinal fusion tech-
`nique in the rabbit and non—human primate (Rhesus) models.
`Spine 1996;21:2689—97.
`3. Brantigan jW, McAfcc PC, Cunningham BW, Wang H,
`Orbegoso CM. Interbody lumbar fusion using a carbon fiber
`cage implant versus allograft bone: An investigational study in
`the Spanish goat. Spine 1994;19:1436—44.
`4. Brodke DS, Dick JC, Kunz DN, McCabe R, Zdeblick TA.
`Posterior lumbar interbody fusion: A biomechanical compari—
`son including a new threaded cage. Spine 1997;22:26~31.
`5. Cook TA, Dehn TC. P