GLOBUS MEDICAL, INC.
`EXHIBIT 1019
`IPR2015-to be assigned
`(Globus v. Bonutti)
`
`Page 1 of 8
`
`

`
`CLINICAL ORTHOPAEDICS AND RELATED RESEARCH
`Number 300. pp. 45-5!
`© I994 J. B. Lippincott Company
`
`Anterior Lumbar Fusion Options
`
`Technique and Graft Materials
`
`JEFFREY A. KOZAK, M.D.,* ALAN E. HEILMAN, M.D.,*
`AND JOHN P. O’BRIEN, PH.D., F.R.C.S.(ED.), F.A.C.S., F.R.A.C.S.**
`
`The biomechanical requirements of the ideal lum-
`bar interbody fusion are important in the surgical
`treatment of the vertebral endplates and graft se-
`lection. A series of 45 patients with anterior lum-
`bar fusion were reviewed retrospectively. The fu-
`sion was performed with a composite construct
`consisting of femoral allograft cross-sections and
`allograft cancellous bone. Preparation of vertebral
`endplates consisted of scraping the endplate until
`punctate bleeding appeared. The major part of the
`endplate was preserved to withstand interspace
`distraction and stability. A 97% fusion rate was
`achieved based on flexion and extension analysis
`with a six- to 12-month follow-up period. A final
`average interspace distraction of 2.39 mm was re-
`ported. The graft should consist of a rigid structure
`that exceeds physiologic loads anticipated in the
`postoperative period, and the composition to pro-
`mote arthrodesis. Anterior lumbar fusion with fem-
`oral allograft is an excellent procedure, but long-
`term further follow-up studies and statistical data
`are essential.
`
`The debate regarding the ideal operation to
`achieve lumbar fusion continues among spi-
`nal surgeons. Traditional techniques of in
`situ posterolateral arthrodesis with iliac graft
`have produced marginal arthrodesis rates."
`To improve the fusion rate and achieve a bio-
`mechanically superior fusion construct, con-
`siderable attention has been directed toward
`
`lumbar interbody fusion procedures.
`
`From ‘Joe W. King Orthopedic lnstitute, Houston,
`Texas.
`‘”“ The London Clinic. London, England.
`Reprint requests to Jeffrey A. Kozak. M.D.. 6560 Fan-
`nin, Suite 2090. Houston, TX 77030.
`Received and accepted: September I l. l99|.
`
`45
`
`The advantages of anterior lumbar fusion
`in comparison with posterior lumbar inter-
`body fusion are many, including ease of dis-
`section, reduced operative time and blood
`loss, noninterference with the potentially
`painful posterior elements of the lumbar
`spine, and avoidance of scarring within the
`spinal canal. In addition, the disk can be re-
`sected in its entirety, advantageous from a
`structural and biochemical perspective.
`This report reviews the requirements of an
`ideal anterior interbody construct and the
`most suitable graft materials within the auto-
`genie and allogenic category.
`
`ANTERIOR INTERBODY CONSTRUCT
`
`BIOMECHANICAL REQUIREMENTS
`
`An excellent review of the biomechanics of
`
`lumbar fusion has been provided by Evans.“
`The compressive forces across the grafted in-
`terspaces should be less than that required to
`induce failure of the graft construct, the graft
`should be able to transmit force without sig-
`nificant motion so that immediate mechani-
`cal load transfer is achieved, and the tech-
`nique should induce arthrodesis as quickly as
`possible with minimal to no morbidity asso-
`ciated with its use.
`
`SURGICAL TREATMENT or THE VERTEBRAL
`Bony ENDPLATE
`
`Any discussion of anterior interbody graft-
`ing technique must address the issue of end-
`plate preparation. If one studies the biome-
`chanical structure of the vertebral body, it is
`
`Page 2 of 8
`Page 2 of 8
`
`

`
`46
`
`Kozak et al.
`
`Clinical Orthopaedics
`and Related Research
`
`readily apparent that the greatest strength is
`present in the cortical endplate. In the center
`of the vertebral body, there is a very sparse
`cancellous bone, which has far fewer load-
`bearing characteristics than the vertebral
`endplates.
`Two basic techniques of endplate prepara-
`tion during interbody fusion are available.
`The first involves purposeful endplate cavita-
`tion to provide an optimal bleeding bed of
`cancellous bone in keeping with routine or-
`thopaedic principles. The earliest manifesta-
`tion of this approach was the “mortise” tech-
`nique, in which a trough was created in the
`subjacent vertebral bodies and a large block
`of bone was laid across this trough. In fact,
`this was the technique used by the Mayo
`Clinic during the 1960s, producing the re-
`ported poor results.” Multiple modifications
`of this approach have been discussed, the
`most recent of which has been popularized by
`Crock and co-workers.“"° This technique in-
`volves partial endplate cavitation with the cre-
`ation of two adjacent circular holes (each ori-
`ented parasagittally) across the prepared disk
`space and endplates. Subsequently, two cy-
`lindrical grafts are impacted into these cavi-
`ties. The result obviously involves endplate
`violation. If the graft is able to withstand the
`applied load and does not cavitate into the
`surrounding vertebral bodies, then an ideal
`condition for arthrodesis is present. Vertebral
`body cancellous bone, however, has minimal
`compressive strength, and it is hypothesized
`that graft cavitation into the vertebral body is
`likely, with subsequent increase in interspace
`motion reducing union potential.
`The second technique involves complete
`or near complete endplate preservation. It is
`conceptually attractive to preserve the end-
`plate to achieve maximum strength of the
`bone adjacent to the implanted graft. lfmaxi—
`mum stability of the interspace is desired,
`then a precisely cut graft with endplate preser-
`vation is mandatory. The disadvantage, how-
`ever, is that the endplate is minimally vascu-
`larized and the recipient bed is far less vascu-
`lar than the previously described technique.
`
`One could argue, on a physiologic basis, that
`union would be less likely to occur. If end-
`plate preservation is chosen, then punctate
`bleeding can be obtained by partial endplate
`resection back to this bleeding surface. There-
`fore, adequate vascularity is supplied to the
`graft while providing a mechanical bed for
`stability.
`There appears to be a trade-off. The sur-
`geon can set the stage for graft stability and
`interspace distraction with endplate preserva-
`tion, or choose to cavitate the endplate to
`maximize graft exposure to bleeding cancel-
`lous bone.
`
`GRAFT SELECTION
`
`The ideal graft for lumbar interbody fusion
`should provide an osteogenic, immunologi-
`cally equivalent matrix, provide immediate
`mechanical stability, and be technically easy
`to modify into appropriate size and shape.
`The first issue to be considered regarding
`graft selection is that of allograft versus auto-
`graft bone. Currently, the “gold standard”
`with regard to grafting procedures is that pro-
`vided by autogenous bone graft in any form.
`Many surgeons continue to stress the impor-
`tance of using autogenous bone products in
`the performance of a lumbar interbody fu-
`sion.” There is, however, an abundance of
`evidence to suggest that the use of allograft
`bone in lumbar interbody fusion provides a
`union rate at least equal to that provided by
`autogenous bone.‘ "'8 Given the immense ad-
`vantage of avoiding donor site problems, this
`is certainly a critical advantage to consider. It
`would seem that the routine principles of on-
`lay autogenous bone grafts to provide long
`bone fracture union may very well not be ap-
`plicable in the lumbar spine. Rather, the large
`available grafting service and the multiple
`marrow elements available within the verte-
`
`bral bodies only millimeters from the inter-
`body graft may very well allow the use of allo-
`graft materials.
`If, in fact, the goal is to use an interbody
`graft that can tolerate loads that exceed those
`
`Page 3 of 8
`Page 3 of 8
`
`

`
`Number 300
`March. 1994
`
`which are to be transmitted in the postopera-
`tive period, the use of cortical bone products
`becomes an important consideration. Esti-
`mates of lumbar in viva loads have ranged
`from 750 to 1600 lb for static loads” to 2000
`
`loads.° The compressive
`lb for high-level
`strength of iliac allograft products ranges
`from 396 to 1475 lb, whereas the compres-
`sive strength of femoral cortical rings is in
`excess of 15.175 lb.” From an autograft
`standpoint. the only available cortical bone is
`that of the fibula. as popularized by Wat-
`kins.” The donor site morbidity is pro-
`longed, however. A wide variety of allograft
`cortical bone products are available within
`the United States from a variety of manufac-
`turers. The strength advantage of cortical
`bone is immediately obvious, although one
`would hypothesize that the union rate asso-
`ciated with cortical bone is certainly less than
`that of cancellous products.
`Lastly. the critical issue of using a compos-
`ite construct must be addressed. It is unlikely
`that all of Evans‘ ideal principles will ever be
`achieved by a single product. either autograft.
`allograft. or artificial. Rather. the use of a
`composite construct with endplate preserva-
`tion will likely represent the best alternative.
`In fact. tricortical iliac bone is a composite
`structure consisting of both cortical and can-
`cellous bone. Autograft fibula is a composite
`structure as well. although the ratio of corti-
`cal to cancellous bone is likely excessive to
`produce a satisfactory union rate. The use of
`full-thickness allograft femoral rings supple-
`mented with cancellous bone products in the
`intramedullary canal ofthe femur and the pe-
`riphery ofthe interspace surrounding the fem-
`oral ring has been recently described.” The
`composite technique provides the advantage
`of rigid cortical bone to provide sufficient
`graft strength while leaving sufficient room in
`the remainder of the interspace to pack other
`bone products and promote arthrodesis. One
`may choose iliac autograft. iliac allograft, or
`demineralized bone products in conjunction
`with the cortical products. If iliac autograft is
`chosen. a nice graft can be obtained quickly
`
`Materials of Anterior Lumbar Fusion
`
`47
`
`with minimal morbidity using an acetabular
`reamer. With a combination of autograft and
`allograft, one should capitalize on the me-
`chanical strength of the allograft and biologic
`strength of the autograft. Long-term studies
`are pending to determine whether the union
`rate of allograft/autograft combinations justi-
`fies the increased morbidity.
`
`MATERIALS AND METHODS
`
`In one auth0r‘s practice. the first consecutive 45
`patients who had anterior lumbar fusion using
`femoral cortical allograft were retrospectively re-
`viewed. Vertebral endplate preparation consisted
`of cartilaginous endplate resection and minimal
`resection of the cortical endplate until punctate
`bleeding points were encountered. After maximal
`interspace distraction. a full-thickness femoral al-
`lograft was modified. with an oscillating saw. to
`the exact interspace dimensions. The femoral in-
`tramedullary canal was packed with iliac cancel-
`lous allograft and the construct was impacted into
`the interspace. A 6.5-mm AO cancellous screw
`with washer then was inserted into the surround-
`ing vertebral body in an oblique fashion so that the
`washer served as a buttress to prevent anterior
`graft expulsion. Finally. the periphery ofthe inter-
`space surrounding the femoral graft was packed
`with cancellous products as well
`(Figs.
`l and
`2).Twenty—one patients carried a diagnosis of disk
`disruption syndrome. I2 patients had postlaminec-
`tomy syndrome. and ll patients had the proce-
`dure to repaira previously unsuccessful posterolat-
`eral fusion exhibiting a mobile nonunion. There
`were 37 one-level fusions and eight two-level fu-
`sions. Requirements for inclusion into the study
`were adequate radiographs with llexion and exten-
`sion analysis at a minimum ofsix months afterthe
`procedure.
`In addition. the first 25 ofthe above-mentioned
`47 patients were available for a I2-month follow-
`up review. These patients were evaluated much
`more carefully with respect to interspace distrac-
`tion and long-term radiographic behavior of the
`femoral
`construct.
`Specifically.
`preoperative
`standing radiographs were carefully measured to
`assess both anterior and posterior intervenebral
`height. This same measurement was continued at
`the one-. three-. six-. and l2-month follow-up ra-
`diographs and appropriate adjustments for magni-
`fication were used. Distraction achieved at
`the
`time ofthe procedure and the long-term behavior
`ofthis distraction during the first I2 months after
`surgery was assessed. At the six- and 12-month
`follow-up examinations.
`flexion and extension
`
`Page 4 of 8
`Page 4 of 8
`
`

`
`43
`
`Kozak et al.
`
`Clinical Orthopaedics
`and Related Research
`
`
`
`Cancellous
`bone chips
`
`arthrodesis.
`
`FIG. 1. The femoral allograft cross-sec-
`tion cut to the exact dimensions of the
`interspace. The intramedullary canal is
`packed with bone products to promote
`
`Femoral cortical graft
`
`films were obtained to assess the issue of inter-
`space stabilization.
`
`RESULTS
`
`Taking the group as a whole, 45 ofthe 47
`patients had a full radiographic series, includ-
`ing flexion and extension films at the six-
`month postoperative interval. Two patients
`had moved out of state and described an ex-
`
`cellent clinical result without physician visits
`elsewhere. but could not return to the office
`
`for radiographic evaluation and were elimi-
`nated from the study.
`Assessing these results. a solid radiographic
`fusion with complete stability on flexion and
`
`extension analysis was achieved in 84% of the
`patients (Fig. 3). Thirteen percent of the pa-
`tients exhibited a radiolucent line on one or
`both sides of the femoral construct. but ar-
`
`throdesis was suggested by bridging bone ei-
`ther anterior or posterior to the femoral con-
`struct. and complete stability was present on
`flexion and extension analysis. Only one pa-
`tient in the study exhibited a mobile non-
`union, for a nonunion rate of 2%.
`
`Interesting trends were noted with regard
`to the first group of 25 patients in whom the
`issue of interspace distraction was carefully
`assessed. The group as a whole illustrated an
`average of 2.39 mm immediate interspace
`distraction that was maintained over time in
`
`
`
`Posterior
`longitudinal
`ligament
`
`FIG. 2. The composite graft struc-
`ture after placement into the inter-
`space.
`
`Cancellous
`bone chips
`
`A0 cancellous
`screw and washer
`
`bone chips
`
`\\
`
`Page 5 of 8
`Page 5 of 8
`
`

`
`Number 300
`March. 1994
`
`Materials of Anterior Lumbar Fusion
`
`49
`
`FIG. 3. Preoperative (left) and nine-month postoperative (right) lateral lumbar radiograph of one pa-
`tient in the series with an L5-S1 post laminectomy syndrome. Note the obvious solid arthrodesis with
`bridging bone anterior and posterior to the femoral graft.
`
`subsequent follow-up examinations. There
`was a tremendous skew of data in these pa-
`tients, with the final results ranging from 0.9
`mm interspace distraction to -0.6 mm (a neg-
`ative number indicating decreased interspace
`height at follow-up examination as compared
`with preoperative values). All of the femoral
`constructs remained intact and exhibited no
`
`evidence of collapse. It was clear that the pri-
`mary cause of interspace collapse was cavita-
`tion of the femoral construct
`into the
`
`surrounding vertebral endplates and adjacent
`vertebral bodies.
`
`COMPLICATIONS
`
`Complications were few in the series.
`There were two vascular injuries. The first
`was an avulsion of the iliolumbar vein during
`an L4—L5 exposure with subsequent 1500-cc
`blood loss. No transfusion was required and
`there were no long-term sequelae. Secondly,
`there was a small 3-mm vena caval tear dur-
`
`ing an L4—L5 exposure with a subsequent
`200 cc blood loss that was repaired primarily.
`
`One interesting complication was that of a
`rectus sheath hematoma, which occurred
`
`during an episode of coughing three days
`after surgery, requiring surgical drainage and
`ligation of the inferior epigastric vessels.
`There were no deep vein thromboses, pulmo-
`nary emboli, deaths,
`superficial or deep
`wound infections. Retrograde ejaculation
`was denied by all men.
`
`DISCUSSION
`
`The solid arthrodesis rate of 84% in this
`
`study is comparable to other techniques of an-
`terior interbody fusion.‘”‘8"°"2'”"7 Whether
`the femoral cortical allograft technique pre-
`sented represents an advantage over previ-
`ously described techniques is dependent on
`the classification of the patients who exhib-
`ited incomplete cortical arthrodesis but stabil-
`ity on flexion and extension analysis. If these
`patients are included in the successful group,
`then the fusion rate of the technique at six
`months would, in fact, be 97%. Interestingly,
`careful review of the initial 25 patients would
`
`Page 6 of 8
`Page 6 of 8
`
`

`
`50
`
`Kozak et al.
`
`Clinical Orthopaedics
`and Related Research
`
`
`
`FIG. 4. Flexion and extension lateral lumbar radiographs taken only three weeks after surgery. Reason-
`able interspace stability is present before arthrodesis with the discussed technique.
`
`indicate that the above-mentioned radiolu-
`
`cencies resolve with time. In fact, all of the
`above-mentioned radiolucencies surround-
`
`ing femoral grafts in interspaces stable on
`flexion and extension analysis were exhibited
`by patients recently operated on with a mini-
`mum follow-up period ofsix to nine months.
`None of the patients with a minimum follow-
`up period of one year exhibited such signs,
`and it is hypothesized that these radiolucent
`defects tend to consolidate with time.
`
`The authors hypothesize that the described
`technique of femoral Cortical interbody fu-
`sion with endplate preservation is in keeping
`with many of the requirements described by
`Evans. That is, the femoral cortical construct
`is certainly rigid enough to withstand the
`transmitted loads and provide immediate
`stabilization of the interspace. ln fact. flexion
`and extension films have been obtained on a
`
`small number of these patients as early as
`three weeks after surgery. and minimal to no
`motion is appreciated (Fig. 4).
`The advantage of an operative technique
`that uses only allograft is immense with re-
`
`gard to both savings of time and postopera-
`tive donor site morbidity. The expense of
`these bone products is considerable, how-
`ever, and should be considered in the final
`assessment.
`Ifinstrumentation is used in the involved
`
`interspace, either anteriorly or posteriorly, a
`less rigid construct could be used because the
`construct would be expected to bear a far
`smaller load and a lower Collapse rate ofthe
`graft would be noted. Therefore, the ratio of
`cortical to cancellous products in an inter-
`body construct should vary in any given sit-
`uation and would certainly be minimized if
`associated instrumentation was used. The au-
`thors stress that the use of the above-men-
`
`tioned technique has been designed with the
`concept ofa single anterior interbody fusion
`in the lower lumbar spine," and have not at-
`tempted its use in multilevel lumbar degener-
`ative disk disease. Rather, they fully advocate
`the use of simultaneous anterior and poste-
`rior procedures in this situation."
`The use ofa femoral construct and avoid-
`
`ance ofgraft collapse is critical in the authors’
`
`Page 7 of 8
`Page 7 of 8
`
`

`
`Number 300
`March, 1994
`
`Materials of Anterior Lumbar Fusion
`
`51
`
`2 P
`
`8”
`
`I0.
`
`ll.
`
`I2.
`
`I3.
`
`. Bradford, F. K., and Spurling, R. G.: The Intervene-
`bral Disc. Springfield. Charles C. Thomas, I945.
`Brown, W.: Personal communication, 1989.
`Crock, H. V., and Bedbrook, G.: Practice of Spinal
`Surgery. New York, Springer Verlag, I983.
`Dennis, S., Watkins, R.. Landaker, S., Dillin, W.,
`and Springer, D.: Comparison of disc space heights
`after anterior
`lumbar
`interbody fusion. Spine
`l4:876, I989.
`Evans, J. H.: Biomechanics of lumbar fusion. In
`Lin, P. M., and Gill. K. (eds.): Lumbar lnterbody
`Fusion. Rockville, Aspen Publishers, I989, p. I2.
`Flynn, J. C.. and Hogue, A.: Anterior fusion of lum-
`bar spine end result with long-term follow-up. J.
`Bone Joint Surg. 6IA:I I43. I979.
`Freebody, D., Bendall, R.. and Taylor, R. D.: Ante-
`rior transperitoneal lumbar fusion. J. Bone Joint
`Surg. 53A:6l7. l97I.
`Freivalds, A., Chalfin, D. B., Garg. A., and Lee,
`K. S.: A dynamic biomechanical evaluation of lift-
`ing maximal acceptable loads. J. Biomech. I7:25I.
`I984.
`Fujimaki, A., Crock, H. V., and Bedbrook, G. M.:
`The results of I50 anterior lumbar fusion operations
`performed by two surgeons in Australia. Clin.
`Onhop. I65:l64, I982.
`Gill. K.: Technique and complications of anterior
`lumbar interbody fusion. In Lin, P. M., and Gill. K.
`(eds.): Lumbar lnterbody Fusion. Rockville, Aspen
`Publishers, I989. p. I01.
`Goldner. J. L.. Urbaniak, J. R.. and McCollum,
`D. E.: Anterior disc excision and interbody spinal
`fusion for chronic low back pain. Orthop. Clin.
`North Am. 2:543, 1971.
`lnone. S. 1., Watanabe, T.. Hirose, A., Tanaka, T.,
`Matsui, N., Saegusa. 0.. and Sho. E.: Anterior disc-
`ectomy and interbody fusion for lumbar disc hernia-
`tion: A review of 350 cases. Clin. Orthop. I83:22,
`I984.
`Kozak, J. A., and O'Brien. J. P.: Simultaneous com-
`bined anterior and posterior fusion. independent
`analysis of a treatment for the disabled low back
`pain patient. Spine. 15:322. I990.
`Morales. R. W., Pettine, K. A., and Salib, R. M.: A
`biomechanical study of bone allografts used in lum-
`bar interbody fusions. Presented at the North Ameri-
`can Spine Society. Keystone. Colorado. Aug.
`I.
`1991.
`Salib, R.: Personal communication. I989.
`Staffer. R. N., and Coventry, M. B.: Anterior inter-
`body lumbar spine fusion. J. Bone Joint Surg.
`54A:756, I972.
`Tan. S. B., Kozak, J. A., and Graham, M. J.: A modi-
`lied technique of anterior lumbar fusion with femo-
`ral cortical allograft. J. Onhop. Surg. Techn. 5:83.
`I990.
`Watkins. R. G.: Anterior lumbar interbody fusion:
`Surgical technique. In Lin. P. M.. and Gill. K. (eds.):
`Lumbar lnterbody Fusion. Rockville. Aspen Pub-
`lishers. I989. p. I07.
`
`15.
`
`16.
`I7.
`
`I8.
`
`19.
`
`Page 8 of 8
`Page 8 of 8
`
`minds. The avoidance of graft collapse is ne-
`gated, however, if the graft collapses into the
`surrounding vertebral bodies. Dennis et al.’
`studied this issue in 31 consecutive patients
`who had undergone anterior lumbar inter-
`body fusion using both iliac allograft and au-
`tograft bone materials. They concluded that
`100% of the patients illustrated disk space
`height collapse to at least the preoperative
`value. Furthermore, 46% of the patients were
`narrower
`than their preoperative height.
`Watkins” also noted that autogenic iliac
`crest tricortical grafts routinely “exhibit par-
`tial resorption, loss of height, and settling be-
`fore healing.” It should be noted that an aver-
`age interspace distraction of 2.8 mm was
`achieved in this series, which is at least a con-
`ceptual advantage over iliac grafts in anterior
`lumbar fusion. As previously mentioned,
`however, there were cases of decreased inter-
`space height that were attributed to graft cavi-
`tation through the endplate and into the
`surrounding vertebral bodies. The ideal posi-
`tion and shape of the rigid interbody con-
`struct has yet to be determined. Endplate
`strength is currently evaluated in the labora-
`tory. Preliminary results would indicate that
`the maximal endplate strength is peripheral,
`near the ring apophysis. Therefore, a more
`peripherally situated cortical construct will
`likely represent the ideal situation to promote
`maximal interspace stabilization.
`It should be noted that the use of cortical
`
`bone in lumbar interbody spine fusion is not
`an original idea from the authors. The use of
`the femoral cortical construct in anterior in-
`
`terbody fusion was conceived and originally
`used by Salib” and Brown.’ Furthermore,
`Watkins has used fibular bone products for
`many years.”
`
`REFERENCES
`
`I. Bartelink. D. L.: The role of abdominal pressure in
`relieving the pressure on the lumbar intervertebral
`discs. J. Bone Joint Surg. 39B:7l8, I957.