Reduction of lumbar spondylolisthesis using a CDI ... - Springer Link

1 downloads 0 Views 637KB Size Report
Reduction of lumbar spondylolisthesis using a CDi pedicle screw system. Received: 21 July 1994. Abstract To evaluate reduction methods using two of the.
Arch Orthop Trauma Surg (1995) 114:188-193

© Springer-Verlag 1995

Y. A o t a . K. Kumano • S. Hirabayashi • Y. Ogawa

Reduction of lumbar spondylolisthesis using a CDi pedicle screw system

Received: 21 July 1994

A b s t r a c t To evaluate reduction methods using two of the several types of Cotrel-Dubousset instrumentation (CDI), 39 patients with low-grade (Meyerding grades I and II [12]) lumbar spondylolisthesis were followed clinically and radiologically for a minimum of 6 months. Our surgical methods were classified into three types as follow: reduction by 2-level assembly using double-threaded screws (type I); reduction by single-level assembly using the double-threaded screws (type II); and reduction by single-level assembly using tulip screws (type III). We found that satisfactory reduction could be achieved by single-level assembly using either the double-threaded or tulip screw without resection of disc material or the posterior leverage method. We found that the tulip screw was superior in obtaining stable reduction. We discuss the mechanism of reducing spondylolisthesis in comparison with that of previously reported methods.

maintain accurate positioned control of the motion segment in anatomical alignment. Pedicle screws have not, however, eliminated the need for weight-bearing support in the anterior column. A high rate of instrumentation failure was recognized in our experience with CDI with posterolateral fusion (PLF) from 1986 to 1988. We have therefore selected posterior lumbar interbody fusion (PLIF) for anterior support since 1988. Although PLIF with a pedicle screw system is supposed to afford rigid fixation, it is not sufficient. Brantigan and Steffee reported that in their experience of 697 pedicle screw/VSP cases, the failure rate of the interbody graft was 30% [3]. We find the reduction method important in reducing the failure rate. The surgical methods of CDI used were classified into three types. In this report, we describe our surgical technique and results. We discuss the mechanism of reduction as well as the advantages and disadvantages of each type of pedicular screw used in CDI with a view to reducing the failure rate.

Introduction In patients with Meyerding grades I and II spondylolisthesis, satisfactory reduction can be achieved through either the anterior or posterior approach. Although the anterior approach presents the advantage of being non-tranmatic to the cauda equina and nerve roots, the decompression achieved is often inadequate. In addition, when the left sympathetic trunk is injured, retrograde ejaculation may occur [16]. We selected the posterior approach because the decompression should be extensive enough to free the nerve roots completely, particularly in patients with spinal canal stenosis due to spondylolisthesis [11]. Use of a pedicle screw system has enabled surgeons to reduce degenerative deformities of the lumbar spine and Y. Aota (N?I) Miura Municipal Hospital, Department of Orthopedic Surgery, Miura City, Kanagawa, Japan Y. Aota - K. Kumano - S. Hirabayashi • Y. Ogawa Department of Orthopedic Surgery, Kantoh Rohsai Hospital, Kawasaki, Japan

Materials and methods Clinical materials Fifty-four patients who had undergone surgery involving CDI at Kantoh Rohsai Hospital for low-grade (grades I - II) spondylolisthesis were followed prospectively from August 1986 to May 1993. Since October 1988, PLIF was performed for anterior support after reduction in 38 patients. In the present study, these 38 patients were reviewed (12 male and 26 female). All but one patient were treated by single-stage surgery. One patient developed nonunion of PLIF and underwent second-stage reduction surgery by another method using CDI. Therefore, this patient was counted as two cases. In total, 39 cases were evaluated in this follow-up study by radiological and clinical examination. Their ages at surgery ranged from 16 to 78 years (mean 55.2 years). The numbers of patients with dysplastic spondylolisthesis, isthmic spondylolisthesis, degenerative spondylolisthesis, and postsurgery spondylolisthesis were 2, 17, 19, and 1, respectively. The affected level was L4-5 and L5-S1 in 27 and 12, respectively. The patients were followed for 6-48 months (mean 23 months). They were divided into three groups. The first one underwent type I surgery between 1988 and 1989 (n = 6); the second underwent type II surgery between 1989 and 1991 (n = 23); and the third underwent type III surgery after June 1991 (n --- 10) (Table 1).

189 Table 1 Treatment data for 39 cases Case no.

Age (years)

Sex

Affected level

Type a

Percent slipping before surgery / after surgery / follow-up

Group I 1 2 3 4 5 6

17 49 46 48 57 49

F F F M F M

5/S 5/S 5/S 5/S 4/5 4/5

Dy Is Is Is De De

48 /21 /41 c 28 /17 /17 23 /17 /21 21 /12 /26 c 19 / 7 / 7 13 / 0 / 0

Mean

25.3/12.3/18.7

Slip angle b before surgery / after surgery / follow-up

55 -18 10 -6 5 2

/ 19 /-17 / 10 /-9 / 8 /-11

8.0/

/ 34 c /-17 / 9 /-6 / 7 /-13

0.0/

lnstrumentation failure

Clinical results

Follow-up (months)

Breakage

G F P F G E

31 31 44 16 20 24

Loosening

2.3

27.7 (n = 6)

Group II 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

53 49 46 66 56 69 49 61 69 23 51 74 73 57 59 63 61 62 64 51 53 50 44

F M F F M F F M M M F M F F F F F F F F F M M

5/S 5/S 4/5 4/5 5/S 5/S 4/5 4/5 4/5 4/5 5/S 4/5 4/5 4/5 5/S 4/5 4/5 4/5 4/5 4/5 4/5 4/5 4/5

Is Is De De Is Is Is De De Is Is De De De Is De De De De Is De Is Is

36 34 30 30 29 27 25 25 23 23 23 23 22 21 19 18 15 15 14 13 12 12 11

/5 /4 /5 /0 /17 /3 /0 /11 / 0 /0 /0 / 7 /0 /-3 /0 /0 /0 /0 / 0 /0 /-5 /-5 /0

/5 /4 /5 /0 /17 /8 /0 / 9 /17 c /0 /21 c /10 /0 /-3 /0 /0 /0 /0 /21 ° /0 /-5 /5 /0

21.7/ 1.7/ 5.0

Mean

1 / 1 7 /-3 12 / 2 9 / 0 -5 1-4 -13 /-14 9 / 0 -3 /-3 9 /10 - 4 /-11 -1 / - 9 3 /-8 6 / 0 / 3 /-3 5 /-I0 1 5 / 3 / - 2 /-10 -5 /-8 4/1 -12 / - 6 0 /-7 -5 / -6 7/5

1 3 2 0 0 -9 0 -3 6 /-11 / -1 /-7 0 /-4 /-13 5 /-16 /-10 /4 / -7 /-6 / -4 /5

Breakage

Breakage Loosening

Breakage

F E E E E G G E P E E G E G E F F E F G G E G

1.7 / -3.5 / -2.8

43 14 34 30 25 22 44 14 27 18 15 41 26 22 36 25 24 13 48 38 24 13 7 26.2 (n = 23)

Group III 30 31 32 33 34 35 (15) 36 37 38 39 Mean

16 49 56 73 57 71 73 78 56 55

F F F F F M M M F F

5/S 4/5 4/5 4/5 4/5 4/5 5/S 4/5 4/5 4/5

Dy Is Is De De Po Is De De De

28 / 2 / 0 27 0 /0 26 0 /5 21 8 /10 19 7 /7 17 8 /8 17 2 /2 13 9 /10 13 0 /0 12 0 /2 19.3

3.6/ 4.4

a Dy, Is, De, or Po indicates dysplastic, isthmic, degenerative, or postsurgery spondylolisthesis, respectively. b A positive number indicates degrees of kyphosis; a negative number indicates lordosis in the affected level.

10 / - 9 6 / 0 / 2 /-9 7 / -8 14 / - 3 6 / 0 / 4/3 - 5 /-13 -8 /-9 10 / 2

/ -9 0 /-5 / -6 / 2 0 /3 /-11 /-8 / 1

4.6 / -4.6 / -3.3

Loosening

F E E F G P E F G G

12 18 18 19 8 15 6 6 17 15 13.4 (n = 10)

c Evident loss of reduction was recognized in cases 1, 4, 15, 17, and 25. Case 15 underwent second-stage reduction with the tulip screw and posterolateral fusion (case 35). The others developed a solid stable fusion in the slipped position

190 Surgical methods After administration of general endotracheal anesthesia, the patient was placed in a prone position on a Hastings' frame [6], so that the lumbar spine was lordotic and the spondylolisthesis partially reduced. Insertion of the pedicle screws was directed medially to form a triangular construction in the axial plane, which increases the pull-out strength of the bone-implant interface, as described in an experimental study by Ruland et al. [13]. After inserting all pedicle screws, complete laminectomy of the slipped vertebra, including facetectomies, was performed. Then the bilateral roots were decompressed from the central zone to the foraminal zone under direct vision. After the pre-bent rods were connected to the screws, reduction was gradually performed by these assemblies without using the posterior leverage method. The nerve roots must be watched carefully during reduction, and reduction must be stopped when impingement by a bony structure or undue stretching of the nerve roots is observed. W h e n reduction was obtained, PLIF was done using autologous bone. Finally, a transverse traction device (DTT) was connected between the two rods. The quadrilateral construction made in the frontal plane enhanced the stability of the implant. A lateral X-ray view should be obtained to confirm satisfactory reduction. Postoperatively, the patient was allowed to walk with a brace 1 week after surgery. Methods of reduction

lip angle

Type I method: Reduction with a 2-level assembly, using double-threaded screws Figure 1 illustrates reduction of spondylolisthesis at L5. The pedicle screw inserted in L4, the sacral screw and the rod have already been fixed. The double-threaded screw inserted into the slipped vertebra was linked to a pre-bent rod with the interposition of a rod-connector. As the nut of the double-threaded screw was tightened, the L5 body was pulled posteriorly.

Type H method: Reduction with a single-level assembly using double-threaded screws Figure 2 illustrates the reduction of spondylolisthesis at L5. The essential difference to reduction with a 2-level assembly is that the L5 body remains independent of the L4 body. As the nut of the double-threaded screw is tightened and the L5 body is pulled posteriorly, the rod-connector slides along the rod to the sacrum, correcting the kyphotic deformity (arrow).

Type II1 method: Reduction by tulip screws The tulip screw has an open body, within which a rod is placed and firmly secured by a plug. Figure 3 shows reduction using this assembly. The pre-bent rods were fixed in the tulip screws in L5. Reduction was then performed by pressing the two ends of the rod into the bodies of the L4 screws using a double rod-introductor attached to the Chopin plate. Radiographic analysis The preoperative, postoperative, and follow-up radiographs were analyzed with respect to percentage of slipping and the degree of the slip angle measured by Boxall's method [2] on standing lateral radiographs (Fig. 4). The raw data are presented in Table 1. Assessment of fusion was performed from standard anteroposterior and lateral radiographs. Special attention was paid to any evidence of continuity of bone trabeculation of PLIF. Solid bone fusion was determined by continuity of bone trabeculation with mobility of less than 5 ° in functional lateral radiographs. Instrumentation failure was read from anteroposterior and lateral radiographs with respect to bending, breakage, loosening, or migration of screws or rods.

3 Fig. 1 Type I. Reduction of spondylolisthesis of L5 by the 2-level assembly using double-threaded screws. The pedicle screw in L4, the sacral screw, and the rod have already been fixed. The doublethereaded screw inserted into the slipped vertebra was linked to a prebent rod with an interposing rod-connector. By tightening the nut of the double-threaded screw, the L5 body could be pulled posteriorly Fig. 2 Type II. Reduction of spondylolisthesis of L5 by the singlelevel assembly using double-threaded screws. As the L5 body was pulled posteriorly by tightening the nut of the double-threaded screw, the kyphotic deformity could be corrected by sliding the rod-connector along the rod (arrow). The degree of reduction was dependent on the distance between the rod-connector and the base of the screw (A) Fig.3 Type III. Reduction of spondylolisthesis of LA with tulip screws. Reduction was performed by pressing the two ends of the rod into the bodies of L4 screws using the double rod-introductor attached to the Chopin plate Fig.4 Boxall's method of measuring radiographic data. Percentage of slipping was calculated by measuring the distance (B) from a line on the posterior aspect of the lower vertebra to a line drawn on the posterior portion of the upper vertebra and was determined by B/A x 100. The slip angle was measured in degrees between a line drawn parallel to the inferior aspect of the upper vertebra and a line drawn perpendicular to the posterior aspect of the body of the lower vertebra

Clinical examination The surgical outcome was evaluated according to Kawabata et al.'s criteria [10] as follows: excellent (no symptoms and normal objective findings); good (marked improvement and no disabilities); fair (some residual symptoms and abnormal objective findings); and poor (no improvement or deterioration). The results are listed in Table 1.

191

Fig.5 Seventeen-year-old woman with dysplastic spondylolisthesis at L5 and dysfunction of the urinary bladder (case 1). Left Lateral standing X-ray before treatment. Middle Immediately after type I surgery, insufficient reduction was noted. Removal of the instrument was performed 31 months later. Although loss of reduction and breakage of L5 screws were recognized postoperatively, a solid fusion was obtained (right)

Results The operative outcome in lumbar fusion surgery is usually evaluated by determining the proportion of satisfactory clinical results and of fusion success. The clinical satisfactory results (good to excellent) in each group were 50%, 78%, and 60%, respectively. In all cases except case 15, solid bone fusion was obtained. N o n u n i o n o f P L I F was recognized radiographically in case !5, and second-stage reduction surgery was performed by the type III method. This is the only case in which poor clinical results were due to nonunion of PLIF. Comparison of each group by Kawabata's criteria was difficult because the clinical results were mainly dependent upon the severity o f the preoperative status in the other cases. Instrumentation failure was recognized in cases 1, 4, 7, 15, 17, 25, and 33 radiographically. Although reduction was lost in cases 1, 4, 15, 17, and 25, it was maintained in cases 7 and 33. Fusion was obtained in a slipped position in cases 1, 4, 17, and 25 without a worsening of clinical results within the follow-up period. The results of reduction and maintenance by the surgical groups are as follows. In group I, reduction was not sufficient in any case except case 6. Statistical analysis by Bonferroni's method revealed that although there was no significant difference in postoperative slip angle between either

Fig. 6 Sixty-nine-year-old man with degenerative spondylolisthesis at L4 (case 15). Preoperative lateral X-ray (top left). Immediately after type II reduction, slight kyphotic deformity remained due to insufficient sliding of the rod-connector (top right). Nonunion of the grafted PLIF bone was recognized 5 months after surgery. Second-stage surgery was performed for bilateral sciatica 29 months after surgery. Before the second-stage operation, evident loss of reduction and breakage of bilateral L5 screws were recognized (bottom left). A solid bone fusion was obtained without remission of sciatica 15 months after the second-stage surgery by type III reduction (case 35) (bottom right)

of the two groups, the percentage of slipping after surgery was significantly greater in group I than in either group II or group III (p < 0.05). In addition, loss of reduction was rec-

192

Fig.7 Sixteen-year-old woman with dysplastic spondylolisthesis at L5 (case 30). Preoperative lateral (left), postoperative lateral (middle), and follow-up lateral (right) X-rays. Note maintenance of reduction and a solid bone fusion at follow-up

ognized in cases 1 and 4 (2/6). Therefore, type I reduction was the least adequate method in this series (Fig. 5). Comparing group II with group III, there were no significant statistical differences in the postoperative percentage of slipping or slip angle by Bonferroni's method. There was also no difference in the loss of reduction in percentage of slipping or slip angle by Wilcoxon's test and chi-square test. Loss of reduction, however, was recognized in three cases in group II (cases 15, 17, 25; Fig. 6) but not in group III (Fig. 7). Despite sufficient reduction in percentage of slipping, reduction of the slip angle was not satisfactory in case 15 of group II. This was due to insufficient sliding of the connector (Fig. 6). Radiographic analysis showed that type III reduction affords the most stable fixation in this series.

Discussion Since the body's center of gravity lies anterior to the lumbosacral joint [1] and because the lumbar spine is subjected to shear stress due to the inclination of the superior surface of S 1 [9], the lumbar vertebrae are predisposed to slip forward and rotate anteriorly. Spondylolisthesis is a kyphotic deformity due to anterior translation and may be coupled with anterior rotation. In Meyerding grades I and II spondylolisthesis, anterior translation is the main component. Since anterior translation is almost always associated with varying degrees of anterior rotation due to disc space narrowing secondary to degeneration, reduction of spondylolisthesis using only posterior translation is inadequate. Posterior rotation of the slipped vertebra plays a particularly important role in decreasing shear stress, especially at the affected level, and in promoting spinal fusion. To reduce the failure rate of interbody graft and/or instrumentation, the most important factor is posterior rotation at the affected level. As for the methods of reduction, complete reduction is never achieved by the 2-level assembly. Obviously,

linkage with the upper vertebra interferes with reduction (Fig. 1). When tulip screws are used, reduction is obtained by inducing lordosis at the involved level (Fig. 3), by which simultaneous posterior translation and posterior rotation are obtained. While the double-threaded screw method is similar to the tulip screw method in that the kyphosis is corrected by inducing lordosis at the involved level, the rod-connector in the former method must slide properly along the rod to achieve complete reduction (Fig. 2). In practice, smooth sliding of the connector is not possible because the rod is pre-bent and the surface of the CDI rod is diamond cut. As for reduction with the double-threaded screw, excessive curvature of the rod also prevents the rod-connector from sliding smoothly along the rod, which results in incomplete reduction. The degree of reduction depends on the distance between the rod-connector and the base of the screw (A in Fig. 2). It is essential to insert the double-threaded screw deep enough into the vertebral body and to fully bend the rod lordotically. This rule also applies to reduction using tulip screws. Thus, we recommend using the tulip screw method rather than the double-threaded screw method for the following three reasons: the technique is easier and does not require the use of a rod-connector; the connection between the rod and the pedicle screw is stronger; better reduction can be obtained by smooth posterior translation and rotation of the slipped vertebra. In comparison with CDI, a screw and plate system in the reduction of spondylolisthesis may be more troublesome, because it is difficult to apply posterior rotation to the slipped vertebra with this instrumentation. In Steffee's procedure [17], reduction is performed by the posterior leverage method with a persuader, which was originally described by SchSllner [15]. In this method, posterior translation is first obtained with a persuader, and the reduction is then maintained with temporary plates, which are later replaced by pre-bent plates to maintain lordosis.

193 This method probably yields more effective reduction, described. Reduction was obtained by posterior translaparticularly for severely slipped vertebra as in Meyerding t i o n a n d posterior rotation of the slipped vertebra using IV spondylolisthesis. This maneuver, however, presents a these screws without requiring manipulation in the ingreat risk o f injury to the nerve roots because it involves traspinal canal space. Thus, the risk of injury to nerve manipulation within the spinal canal. Since satisfactory structures is lower than with the screw and plate system. reduction can be achieved without a persuader in Meyerd- We consider tulip screws superior to double-threaded ing grades I and II spondylolisthesis, this method should screws in terms o f obtaining complete and stable reducbe reserved only for severer grades. tion. Using the preoperative traction method, Scaglietti et al. confirmed that the application o f axial distraction in addition to posterior rotation is effective in grade III and IV References cases [14]. Edwards et al. described the concept of gradual 1. Amundson G, Edwards CC, Garfin SR (1992) Spondylolisthereduction by instrumentation [1, 4]. T h e y also considered sis. Rothman RH, Simeon FA (eds) The spine, 3rd ed. Saunsimultaneous application of the three corrective forces of ders, Philadelphia, pp 913-969 distraction, posterior translation, and posterior rotation in 2. Boxall D, Bradford DS, Winter RB, Moe JH (1979) Managea modular spinal system. However, application of a disment of severe spondylolisthesis in children and adolescents. J Bone Joint Surg [Am] 61:479-495 traction force posteriorly along a distraction device placed 3. Brantigan JW, Steffee AD (1993) A carbon fiber implant to aid at the lamina presents a clear disadvantage in terms of reinterbody fusion: two-year clinical results in the first 26 paduction, since a distraction force applied posteriorly to the tients. Spine 18:2106-2117 facet joint increases kyphosis as the axis o f rotatory dis4. Edwards CC (1990) Prospective evaluation of a new method for complete reduction of L5-S 1 spondylolisthesis using corplacement is near the facet joint [19]. Scaglietti et al. noted rective forces alone. Orthop Trans 14 : 549 a loss of reduction postoperatively in some cases despite 5. Fujiya M, Saita M, Kaneda K, Abumi K (1990) Clinical study satisfactory reduction during surgery. Therefore, the dison stability of combined distraction and compression rod instrutraction force should be applied anteriorly to the facet mentation with posterolateral fusion for unstable degenerative spondylolisthesis. Spine 15 : 121 6-1222 joint rather than posteriorly. 6. Hastings DE (1969) A simple frame for operations on the lumKaneda et al. advocated applying the distraction force bar spine. Can J Surg 12:250-253 using a distraction rod hooked onto the lamina at the in7. Hirabayashi K, Ikeda K, Tsuchida Z et al. (1972) Spondylolyvolved level to decompress the nerve roots in the foramisis and spondylolisthesis: pathological condition and therapeunal zone rather than to achieve complete reduction [8]. tic indication. J Jpn Orthop Assoc 46 : 675-694 8. Kaneda K, Satoh S, Nohara Y, Oguma T (1985) Distraction rod Since pedicular kinking o f the nerve roots is aggravated instrumentation with posterolateral fusion in isthmic spondywith the progression of spondylolisthesis and narrowing lolisthesis: 53 cases followed for 18-89 months. Spine 10: of the disc space [7], K a n e d a proposed that decompres383-389 sion could be obtained by widening the disc space and en9. Kapandji 1A (1982) The lumbar vertebral column. In: Kapandji IA (ed) The physiology of the joints; Vol 3, 2nd ed. Churchill larging the foramen with the distraction rod system. H o w Livingstone, New York ever, in most o f Kaneda's patients, no effect o f distraction 10. Kawabata M, Yamazaki N, Nihei R (1973) Surgical treatment could be observed radiographically at the final follow-up of lumbar disc herniation (in Japanese). Seikeigeka 24: despite g o o d clinical results [5]. Steffee and Sitkowski 875-880 11. Kumano K, Hirabayashi S, Takahashi S, Ishii J, Miyashita H also r e c o m m e n d e d applying a distraction force between (1989) Cotrel-Dubousset pedicle screw fixation for unstable the upper and lower screws to obtain intraforaminal delumbar spine due to spondylolisthesis. Sixth proceedings of the compression o f the nerve roots and to maintain the disinternational congress on Cotrel-Dubousset instrumentation, pp traction effect o f PL1F [18]. In Steffee's patients, in con141-146 trast to Kaneda's, the distraction effect was observed post12.Meyerding HW (1943) Spondylolisthesis: surgical treatment and results. J Bone Joint Surg 25 : 65-77 operatively. We consider that a distraction force which en13.Ruland CM, MaCafee PC, Warden KE, Cunningham BW larges the foraminal canal is not essential for decompres(1991) Triangulation of pedicular instrumentation: a biomesion, since we perform complete facetectomy to free the chanical analysis. Spine 16 : 270-276 nerve roots, but that P L I F is essential to maintain reduc14. Scaglietti O, Frontino G, Bartolozzi P (1976) Technique of anatomical reduction of lumbar spondylolisthesis and its surgition. cal stabilization. Clin Orthop 117:164-175 Steffee reported that the compression force applied by 15. Sch611ner D (1990) One stage reduction and fusion for spondyinstruments upon grafted P L I F bone has a favorable effect lolisthesis. Int Orthop 14 : 145-150 on spinal fusion [18]. To apply a compression force to the 16. Selby DK, Henderson RJ, Blumenthal S, Dossett D (1988) Angrafted bone, the screw and rod system is easier to use terior lumbar fusion. In: Canther JC (ed) Lumbar spine surgery, 2nd edn. Williams & Wilkins, Baltimore, pp 248-260 than the screw and plate system. 17. Steffee AD, Sitkowski DJ (1988) Reduction and stabilization of grade IV spondylolisthesis. Clin Orthop 227 : 82-89 18. Steffee AD, Sitkowski DJ (1988) Posterior lumbar interbody fusion and plates. Clin Orthop 227 : 99-102 Conclusions 19. White AA III, Panjabi MM (1990) The problem of clinical instability in the human spine: a systematic approach. In: White The surgical method of reducing spondylolisthesis using AA III, Panjabi MM (eds) Clinical biomechanics of the spine. either double-threaded screws or tulip screws in C D I was Lippincott, Philadelphia, pp 277-378