The Spine Journal

-

(2013)

-

Clinical Study

Adult thoracolumbar and lumbar scoliosis treated with long vertebral fusion to the sacropelvis: a comparison between new hybrid selective spinal fusion versus anterior-posterior spinal instrumentation Mitsuru Yagi, MD, PhDa,b,*, Ravi Patel, MDa, Thomas W. Lawhorne, MDa,c, Matthew E. Cunningham, MD, PhDa, Oheneba Boachie-Adjei, MDa,c a Spine and Scoliosis Service, Hospital for Special Surgery, 535 East 70th St, NY 10021, USA Department of Orthopedic Surgery, National Hospital Organization Murayama Medical Center, Tokyo, Japan c Department of Orthopedics, Weill Medical College of Cornell University, 520 East 70th St, Starr Pavilion, 2nd Floor, NY 10065, USA b

Received 14 March 2012; revised 11 May 2013; accepted 24 June 2013

Abstract

BACKGROUND CONTEXT: Combined anteroposterior spinal fusion with instrumentation has been used for many years to treat adult thoracolumbar/lumbar scoliosis. This surgery remains a technical challenge to spine surgeons, and current literature reports high complication rates. PURPOSE: The purpose of this study is to validate a new hybrid technique (a combination of single-rod anterior instrumentation and a shorter posterior instrumentation to the sacrum) to treat adult thoracolumbar/lumbar scoliosis. STUDY DESIGN: This study is a retrospective consecutive case series of surgically treated patients with adult lumbar or thoracolumbar scoliosis. PATIENT SAMPLE: This is a retrospective study of 33 matched pairs of patients with adult scoliosis who underwent two different surgical procedures: a new hybrid technique versus a thirdgeneration anteroposterior spinal fusion. OUTCOME MEASURES: Preoperative and postoperative outcome measures include self-report measures, physiological measures, and functional measures. METHODS: In a retrospective case-control study, 33 patients treated with the hybrid technique were matched with 33 patients treated with traditional anteroposterior fusion based on preoperative radiographic parameters. Mean follow-up in the hybrid group was 5.3 years (range, 2–11 years), compared with 4.6 years (range, 2–10 years) in the control group. Operating room (OR) time, estimated blood loss, and levels fused were collected as surrogates for surgical morbidity. Radiographic parameters were collected preoperatively, postoperatively, and at final follow-up. The Scoliosis Research Society Patient Questionnaire (SRS-22r) and Oswestry Disability Index (ODI) scores were collected for clinical outcomes. RESULTS: Operating room time, EBL, and levels fused were significantly less in the hybrid group compared with the control group (p!.0001). The postoperative thoracic Cobb angle was similar between the hybrid and control techniques (p5.24); however, the hybrid technique showed significant improvement in the thoracolumbar/lumbar curves (p5.004) and the lumbosacral fractional curve (p!.0001). The major complication rate was less in the hybrid group compared with the control group (18% vs. 39%, p5.01). Clinical outcomes at final follow-up were not significantly different based on overall SRS-22r scores and ODI scores. CONCLUSION: The new hybrid technique demonstrates good long-term results, with less morbidity and fewer complications than traditional anteroposterior surgery select patients with

FDA device/drug status: Not applicable. Author disclosures: MY: Nothing to disclose. RP: Nothing to disclose. TWL: Nothing to disclose. MEC: Nothing to disclose. OB-A: Royalties: Depuy (C), K2M (C); Consulting: K2M (C), Depuy (C), Osteotech (B), Trans 1 (C); Speaking and/or Teaching Arrangements: K2M (C), Trans 1 (C); Research Support (Investigator Salary, Staff/Materials): K2M (C), Depuy (C) and Osteotech (B). 1529-9430/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.spinee.2013.06.090

The disclosure key can be found on the Table of Contents and at www. TheSpineJournalOnline.com. This study was approved by the institutional review board of the hospital. * Corresponding author. Adult and Pediatric Spine Surgery, Hospital for Special Surgery, 535 East 70th St, New York, NY 10021, USA. Tel.: (212) 606-1948; fax: (212) 794-2562. E-mail address: [email protected] (M. Yagi)

2

M. Yagi et al. / The Spine Journal

-

(2013)

-

thoracolumbar/lumbar scoliosis. This study received no funding. No potential conflict of interestassociated bias existed. Ó 2013 Elsevier Inc. All rights reserved. Keywords:

Anteroposterior spinal fusion; Long posterior spinal instrumentation; Anterior instrumentation; Adult idiopathic degenerative scoliosis

Introduction The correction of adult spine deformity often involves a long posterior spinal fusion to the sacrum. This surgery remains a technical challenge to spine surgeons, and current literature reports complication rates as high as 30% to 80% [1–7]. Failure of fixation at the lumbosacral junction for many reasons, including poor sacral bone stock, insufficient fixation, and pseudoarthrosis, have been implicated in a poor outcome [1–3,5,8–11]. Different instrumentation and fixation techniques have been proposed, including anterior column support, iliac fixation, and stopping constructs at L5 when possible, but the ideal solution to this challenging surgery remains controversial [12,13]. Another frequent complication is the development of proximal junctional kyphosis (PJK) in nearly 40% of patients at 5 years of follow-up [14]. Currently, a common technique for obtaining long-segment spinal fusion to the sacrum uses anterior column support with cages and bone grafting, coupled with a long posterior instrumentation, including fixation to the pelvis. A combined anteroposterior procedure confers several advantages, which include better correction of deformity in severe curves, reduction of pseudoarthrosis, and less likelihood of curve progression [1,2,5,15–19]. Efforts have been made to develop techniques that can obtain a similar correction but require less dissection and shorter fusions. Our hybrid surgical technique uses the concept of a long anterior instrumented fusion to correct the deformity, coupled to a short posterior instrumented fusion to the sacrum to stabilize the lumbopelvic junction [20]. Anterior instrumentation has been shown to afford comparable correction over shorter segments than that of the posterior approach. Anterior spinal fusion, therefore, saves a mean of 2.5 levels in pursuit of obtaining overall correction [21–25]. Other benefits of anterior instrumentation include better correction of hypokyphosis, greater rate of spontaneous coronal correction, and improved rotational correction of apical vertebra [21,26,27]. The mechanical advantages of anterior spinal fusion are attributable to stronger forces against the curvature, the release of an anterior longitudinal ligament with complete discectomy, and enhancement of arthrodesis with application of a morselized rib graft [25,28]. Although initial studies on anterior spinal fusion were associated with a greater rate of rod breakage and frequent pseudoarthrodesis, a growing body of literature indicates that, with improved instrumentation and technique, anterior spinal fusion can be safe and effective in a select group of the population [24,25,27–31]. There is a practical limitation to anterior rods, however, in that instrumentation

to the pelvis is not possible because of bifurcation of the great vessels. For this reason, a hybrid solution was suggested. Briefly, anterior rods were used for thoracolumbar and lumbar stabilization, and shorter posterior rods were used to connect this fusion segment to the sacrum with the use of pedicle screws. The purpose of this study is to assess radiological and clinical outcomes in patients requiring the hybrid instrumentation with a similar cohort undergoing traditional anterior release and long posterior fusion to the sacropelvis.

Materials and methods This is a retrospective study of 33 paired adult patients with scoliosis who underwent either a hybrid construct (study group) or a more conventional technique that involved anterior release with long posterior instrumentation to the sacrum (control group). All surgeries were performed by the senior author at a tertiary referral center. Inclusion criteria included adults with thoracic and thoracolumbar/ lumbar curves up to 80
requiring long fusion (sacrum to above T12); nonprogressive thoracic deformity with flexibility greater than or equal to 30%; lumbosacral fractional curve with segmental instability, stenosis, or significant facet arthrosis, or degenerative disc disease requiring lumbosacral fixation; and nonosteoporotic spine amenable to anterior screw fixation. Clinical examination and investigations were done to rule out any other cause of scoliosis. Exclusion criteria included clinical diagnosis of osteoporosis, cases of revision surgery, and availability of match group data. Patients in the two groups were matched for age and coronal curve type/magnitude by an investigator who was blinded to the postoperative results (Table 1). Baseline radiographic parameters were similar in the hybrid group versus the control group: thoracic cobb, 32614
versus 39616
(p5.06); thoracolumbar cobb, 50613
versus 55616
(p5.18); and lumbosacral fractional cobb, 2369
versus 24610
(p5.74; Table 2). Demographics of the hybrid and control groups were similar. Average age of the hybrid group was 56 years (range, 35–78 years) at the time of surgery compared with 57 years (range, 31–79 years) in the control group (p5.3). Gender distribution was also similar in the hybrid versus control groups (98% female vs. 96% female, p5.66). All patients had a minimum of 2 years of followup (mean follow-up, 5.1 years; range, 2–11 years). Mean follow-up in the hybrid group was 5.3 years (range, 2–11 years) compared with 4.6 years (range, 2–10 years) in the

M. Yagi et al. / The Spine Journal

-

(2013)

-

3

Table 1 Preoperative patient data (matched for age and curve magnitude) Group 1: hybrid

Group 2: control

Age CPL SVA No. (y) Th (
) T/TL (
) LS (
) (mm) (mm) TK (
) LL (
)

CPL SVA No. Age (y) Th (
) T/TL (
) LS (
) (mm) (mm) TK (
) LL (
)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

57 47 74 58 66 47 42 65 53 67 55 50 49 35 68 45 48 63 63 58 56 50 57 78 54 50 53 60 48 61 58 57 62

23 32 18 33 34 55 40 10 23 20 25 25 54 29 36 31 38 36 4 36 29 29 45 43 25 37 24 33 18 70 20 52 19

34 62 50 66 39 56 49 22 54 39 47 50 62 46 83 53 52 56 35 52 53 42 27 46 46 60 48 62 50 77 55 35 40

16 22 28 36 21 20 24 14 36 24 28 20 13 16 35 33 22 25 35 28 28 27 0 20 22 35 21 30 25 13 20 17 12

0.8 1 1.8 2 3.9 1.9 3.4 0.8 4 3 3.6 0.9 0.6 2.3 0.7 3.1 0.5 2 4.7 1.2 4.5 3.1 2.7 0 2 2.9 0 1 5 1.3 2.8 2.7 0.7

3.3 1 0.2 1.2 2.7 2 1.6 6 3.1 6 2.3 3.4 1.8 1 5 6.7 0 4.9 10.5 2.2 0.1 3 0.3 0.9 1.7 0.5 2.6 0.5 2.4 3 2 0.3 5.5

21 37 50 58 61 35 18 56 34 6 40 19 50 46 38 22 3 25 9 50 22 22 49 31 55 32 15 47 37 32 75 50 34

18 54 65 45 63 64 42 42 73 33 45 23 52 86 39 43 9 43 1 55 30 20 62 34 65 51 45 55 50 28 53 60 32

4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

61 51 79 53 61 50 31 50 58 63 61 44 47 38 75 59 53 63 62 57 56 45 55 76 60 58 51 60 57 54 69 54 62

27 45 35 35 27 56 53 20 24 52 52 25 44 33 2 35 14 46 50 20 56 33 41 35 59 46 46 55 28 83 24 56 22

39 62 60 66 40 62 24 24 52 38 38 55 78 68 74 66 65 50 44 53 62 56 36 66 68 77 71 66 60 76 52 62 13

11 22 15 39 13 33 14 19 11 22 22 15 35 32 32 39 32 22 21 15 33 22 22 39 29 21 35 34 31 12 11 33 8

3 1.6 4.5 3 3 5 3.5 3.5 3.5 2 2 3.5 0.5 5 4.5 3 2.1 2 4.5 4.3 5 3.5 2 3 2 3 5 5 2 0 3.5 4 0

5 1 8.1 4.4 5 7 3.5 5 5.5 0 0 4.5 4.9 2.8 8.5 4.4 1.2 5 9 6.7 7 3 0 4.4 3 2 1.1 7 3 2.2 5.5 1.1 0

22 34 60 2 22 68 18 79 52 52 52 55 82 8 80 2 50 77 50 77 68 43 52 2 22 13 24 68 71 66 52 60 70

45 22 50 7 45 46 46 66 44 49 49 50 81 30 50 7 54 52 49 65 46 33 49 7 46 20 15 46 0 55 44 70 66

Th, thoracic; T/TL, thoracic/thoracolumbar; LS, lumbosacral; CPL, coronal plumb line; SVA, sagittal vertical axis; TK, thoracic kyphosis (T2–T12); LL, lumbar lordosis (L1–S1).

control group (p5.12). As a surrogate for morbidity and extent of the procedure, data were collected for surgical time, estimated blood loss, and total number of motion segments fused for both groups. Additional clinical data were collected regarding major complications and reoperations. Radiographic evaluation used standardized standing 36-inch anteroposterior and lateral scoliosis radiographs taken Table 2 Comparison of operative data

Total fused vertebra (N) Anterior Posterior OR time (h) Estimated blood loss (L) Total screws (N)

Hybrid

Control

p Value

6.761.2 6.761.2 3.861.8 7.361.1 2.160.7 13.762.2

14.661.3 6.862.3 14.661.3 8.461.0 3.661.3 28.664.2

!.0001* .41 !.0001* !.0001* !.0001* !.0001*

OR, operating room. Note: The values of the hybrid group were compared with the corresponding values of the control group. * Statistically significant.

preoperatively, during the immediate postoperative period, and at the final follow-up visit. Patients were asked to stand naturally with their shoulders flexed forward approximately at 30
so that their upper thoracic vertebral bodies could be visualized on the lateral radiograph. The end plates at the proximal junction had to be visible clearly for study inclusion. Coronal plane measurements were made using the cobb technique of the thoracic, thoracolumbar/lumbar, and lumbosacral curves. Kyphosis was measured from T2–T12, and lordosis from T12–S1, using the Cobb method. Coronal balance was measured as the distance between the C7 plumb line to the center of the sacrum (center sacral vertical line [CSVL]). Sagittal spine alignment was measured from the C7 plumb line to the posterosuperior margin of the sacrum (sagittal vertical axis [SVA]). Proximal junctional kyphosis was measured using the cobb angle from the caudal end plate of the upper instrumented vertebrae to the cephalad end plate of the second supra-adjacent vertebrae. Abnormal PJK was defined by previously published criteria: a proximal junctional angle more than 10
and a proximal junctional cobb

4

M. Yagi et al. / The Spine Journal

10
greater than the preoperative measurement [32]. In terms of imbalance, a CSVL more than 2.5 cm was considered coronal imbalance and an SVA more than 4.0 cm was considered decompensation or sagittal imbalance [17,33].

-

(2013)

-

significant. All analyses were performed with the Statistical Package for the Social Sciences (SPSS, version 21, Chicago, IL, USA). Results

Patient outcomes The Scoliosis Research Society Patient Questionnaire (SRS-22r) and the Oswestry Disability Index (ODI) were used to evaluate patient outcomes at the final follow-up. Completed questionnaires were available for all patients. Surgical technique Hybrid construct A detailed technique has been described previously [20]. As an anterior procedure, a standard thoracoabdominal approach was used for exposure. After exposure, complete discectomies were performed, with preservation of only the posterior annulus and/or the posterior longitudinal ligament. After disc space preparation, single- or dual-rod segmental instrumentation is placed, typically ending at the L3 or L4 vertebral bodies. After rod rotation and correction of the deformity, structural grafts, typically a Harms titanium cage with a mixture of autograft and allograft, are placed in the intervertebral spaces. Last, segmental compression completes the correction and anchors the structural grafts. Uninstrumented anterior fusions were also performed at the levels distal to the instrumentation, in all cases including L5–S1. The posterior fusion was stopped proximal to the pelvis without iliac screws because of commonly associated implant pain [4]. Typically, anterior instrumentation extended from the lower thoracic levels to the mid lumbar levels and was combined with posterior instrumentation, extending from L3/L4–S1. A standard posterior spinal fusion with pedicle screw instrumentation was then performed, typically from the mid lumbar spine to the sacrum. No pelvic instrumentation was used. Anterior release/long posterior fusion (control group) The third-generation anteroposterior spinal fusion was performed in a similar manner as described earlier. Extensive release of anterior longitudinal ligaments and preparation of intervertebral discs was accomplished through an anterior approach. In contrast to the hybrid technique, no instrumentation was used. Subsequently, a longer posterior instrumentation was required to correct thoracic and thoracolumbar curves [11,34–36]. Pelvic fixation was used in almost all cases. Statistical analysis Statistical analyses were performed using a Student twotailed t test for comparing the means and the Pearson chisquare test for comparing categorical data. A p value less than .05 with a confidence interval of 95% was considered

Surgical time/blood loss Surgical time was significantly shorter in the hybrid group, mainly because of the smaller posterior procedure (7.361.1 hours vs. 8.461.0 hours, p!.0001). Similarly, blood loss was significantly less in the hybrid group (2.160.7 L vs. 3.661.3 L, p!.0001; Table 2). Total fusion levels In the hybrid group, the mean number of levels fused anteriorly was 6.761.2 (range, 3–8 levels), and the mean number of levels fused posteriorly was 3.861.8 (range, 1–9 levels). In the control group, the mean number of levels fused anteriorly was 6.862.3 (range, 3–8 levels), and the mean number of levels fused posteriorly was 14.661.3 (range, 13–16 levels). The total number of motion segments fused was decreased significantly in the hybrid group compared with the control group, again secondary to the shorter length of the posterior procedure (6.761.2 levels vs. 14.661.3 levels, p!.0001). There was no significant difference in the number levels fused anteriorly between the control and hybrid groups (p5.4). On average, the hybrid procedure saved 7.9 motion segments. Of 33 hybrid patients with posterior fusions to the sacrum, 6 patients had proximal instrumented level at L2, 15 at L3, 5 at L4, 1 at L5, 1 at T9, 1 at T10, 2 at T11, and 2 at T12. All the control group posterior fusions extended to between T2 and T5. Coronal correction The mean thoracic curve before surgery was 39616
for the control group and 32614
for the hybrid group (p5.06; Table 3). At final follow-up, the average correction of the thoracic curve was not significantly different (29621% for the control group vs. 36624% for the hybrid group, p5.24). The thoracolumbar/lumbar curve before surgeryaveraged 36624
for the control group and 50613
for the hybrid group (p5.18). At last follow-up, the percent correction of the thoracolumbar/lumbar curve obtained in the hybrid group was significantly better than the control group (61621% vs. 45625%, p5.004). The mean lumbosacral curves were also similar preoperatively in the two groups (control vs. hybrid, 24610
vs. 2369
, respectively; p5.74). At final follow-up, the percent correction of the lumbosacral curve in the hybrid group was significantly better than the control group (67621% vs. 41619%, p!.0001). Sagittal plane correction Preoperative thoracic kyphosis (T2–T12) was significantly greater in the control group than the hybrid group

M. Yagi et al. / The Spine Journal

-

(2013)

-

5

Table 3 Preoperative and postoperative comparison of coronal curve correction Control group Curve type

Hybrid group

Immediate Final Immediate Final Preoperative (
) postoperative (
) follow-up (
) Correction (%) Preoperative (
) postoperative (
) follow-up (
) Correction (%) p1* p2y

Thoracic 39616 TL/L 55616 L/S 24610

25612 29615 1469

26612 31615 1568

29621 45625 41619

32614 50613 2369

20612 17610 765

22614 20611 764

36624 61621 67621

.06 .24 .18 .004z .74 !.0001z

TL/L, thoracolumbar/lumbar; L/S, lumbosacral. Note: Percent collection was defined at the final follow-up. * The preoperative values were compared between the hybrid and the control groups. y The values for the percentage correction between hybrid and control groups. z Statistically significant.

(control vs. hybrid, 47625
vs. 36617
, respectively; p5.02). At final follow-up, mean kyphosis remained significantly greater in the control group than the hybrid group (47618
vs. 39614
, p5.03; Table 4). However, the incidence of abnormal PJK was significantly greater in the control group (9 of 32 patients, 28%) than the hybrid group (3 of 32 patients, 9%) (p5.02, Table 5). Lumbar lordosis was not significantly different in the two groups preoperatively (control, 45619
; hybrid, 40624
; p5.2). Postoperatively, lumbar lordosis was not significantly different initially in the control versus hybrid groups (51611
vs. 47610
, p5.09). However, at final follow-up, lumbar lordosis was better maintained in the control group compared with the hybrid group (50611
vs. 43615
, p5.02; Table 4). Overall balance In the hybrid group, the mean CSVL shift measured 2.2 cm (range, 0–4.7 cm) before surgery and 1.9 cm (range, 0–4.7 cm) at the final follow-up. A CSVL shift greater than 25 mm was found in 14 patients (44%) before surgery, Table 4 Preoperative and postoperative comparison of sagittal curvature

Kyphosis (T2–T12) Preoperative Immediate postoperative Final follow-up p1y p2z Lordosis (T12–S1) Preoperative Immediate postoperative Final follow-up p1y p2z

Hybrid group (
)

Control group (
)

36617 40615 39614 .13 .18

47625 45614 47618 .4 .5

.02 .09 .03

45619 47610 43615 .3 .3

40624 51611 50611 .02x .02x

.2 .09 .02

p3*

* The values of hybrid group were compared with the corresponding values of the control group. y The immediate postoperative values were compared with corresponding preoperative values. z The final follow-up values were compared with the corresponding values 2 years postoperative. x Statistically significant.

10 of whom were corrected to normal coronal balance. However, 8 patients with normal coronal alignment increased to an abnormal shift; therefore, a total of 12 patients (38%) in postoperative evaluation had radiographic coronal imbalance. In the hybrid group, the mean preoperative SVA measured 0.9 cm (range, 0–10.5 cm) before surgery and 3.0 cm (range, 0–11.5 cm) at the final follow-up. Of the four patients who had a preoperative SVA greater than 4.0 cm, two corrected to within the normal range. An additional 8 patients with normal sagittal balance preoperatively demonstrated an SVA of greater than 4 cm postoperatively for a total of 10 patients (31%) with radiographic sagittal imbalance (Table 6). In the control group, the mean CSVL was 3.1 cm (range, 0–4.5 cm) before surgery and 2.9 cm (range, 0–8.5 cm) at the final follow-up. Of 22 patients (69%) with preoperative coronal imbalance (O25 mm), 13 of them remained decompensated and 9 were corrected. Five additional patients experienced worsening of coronal balance; therefore, a total of 18 patients (56%) had coronal decompensation postoperatively. The mean SVA measured 3.4 cm (range, 0–8.5 cm) preoperation and 2.0 mm at the final follow-up (range, 0–12.2 cm). Of 18 patients who had sagittal imbalance (O40 mm) before surgery, 12 of them were corrected to normal sagittal alignment. An additional three patients with an initially normal SVA value increased to greater than 4 cm at the final follow-up (Table 6). Patient satisfaction There was no significant difference in the average score on the SRS-22r instrument between the hybrid and control Table 5 Incidence of proximal junctional kyphosis

Hybrid Control p2z

Preoperative PJ angle (
)

Postoperative PJ angle (
)

Incidence of PJK

p1*

565 464 .2

866 1169 .07

3/32 (9%) 9/32 (28%) .02y

.01y !.0001y

PJ, proximal junction; PJK, proximal junctional kyphosis. * The values of the hybrid group were compared with corresponding values of the control group. y Statistically significant. z The values of final follow-up were compared with corresponding preoperative values.

6

M. Yagi et al. / The Spine Journal

-

(2013)

-

Table 6 Pre- and postoperative comparison of coronal and sagittal balance

Coronal balance Preoperative Postoperative p3x Sagittal balance Preoperative Postoperative p3x

Hybrid (cm)

Control (cm)

Hybrid imbalance

Control imbalance

p1*

p2y

2.261.4 1.961.3 .17

3.161.4 2.962.0 .33

14 (44%) 12 (38%) .5

22 (69%) 18 (56%) .15

.01z .02z

.002z .03z

0.963.4 364.1 .04

3.463.3 2.063.8 .11

5 (16%) 10 (31%) .06

18 (56%) 9 (28%) .0004

.004z .3

.0001z .7

* The values of the hybrid group were compared with corresponding values of the control group. y The values of hybrid group were compared with corresponding values of control group. z Statistically significant. x The values of final follow-up were compared with corresponding preoperative values.

groups. Thirty of 33 patients (91%) in the hybrid group completed a postoperative SRS-22r instrument with an average score of 3.960.5 points compared with 26 of 33 patients (79%) in the control group with an average score of 3.960.7 points (p5.25). Furthermore, no statistical difference was present in the subgroup analysis (Table 7). Similarly, there was no significant difference in postoperative ODI scores between the hybrid and control groups (25616% vs. 27617%, p5.3).

(highest level, T9), typically one to two levels below the anterior upper instrumented vertebrae. In one case, the posterior instrumentation extended one level proximal to the anterior instrumentation. The number of revision surgeries required in the control group was significantly greater than that in the hybrid group (12 of 33 patients versus 6 of 33 patients, p5.03) despite comparable follow-up periods. Discussion

Complications Major complications in both groups included junctional kyphosis, pulmonary embolism, instrument failure, deep infections, and pseudoarthrosis. The rate of complication was significantly lower in the hybrid group (6 of 33 patients, 18%) than in the control group (13 of 33 patients, 39%; p5.01). In the control group, the major complications were PJK/degeneration requiring proximal extension (N57), deep infection (N52), neurologic injury (N51), pseudoarthrosis (N51), prominent iliac screw requiring removal (N51), and pulmonary embolism (N51). In the hybrid group, the complications were PJK/degeneration (N53), pseudoarthrosis (N52), and deep infection (N51). In the patients who had pseudoarthrosis in the middle of the construct had revision posterior extension fusion. The posterior instrumentation was carried up to the lower thoracic spine Table 7 Comparison of postoperative SRS-224 and ODI score

SRS-22r Function Pain Self-image Mental status Satisfaction Total ODI

Hybrid

Control

p Value

4.11 3.63 3.71 3.93 4.53 3.96 25.1

4.12 3.71 3.81 3.75 4.53 3.92 27.2

.21 .42 .47 .27 .49 .21 .32

SRS-22r, Scoliosis Research Society Patient Questionnaire; ODI, Oswestry Disability Index. Note: The values of the hybrid group were compared with corresponding values of the control group.

This study expands on a previous report of the ‘‘hybrid fixation’’ technique for fusions in adult deformity to the sacrum. The hybrid technique has been developed in an attempt to reduce the morbidity of traditional anteroposterior spinal fusion for adult deformity. Traditionally, an anterior approach has often been necessary preceding a long posterior spinal fusion. Although studies have proved the benefits of anteroposterior spinal fusion in achieving superior clinical outcome, the surgery is frequently performed in two separate stages, with predictable morbidity from both thoracotomy and long posterior dissection [16–18,36–38]. The addition of anterior instrumentation to treat thoracic and thoracolumbar spinal deformities has recently demonstrated a reduced rate of instrument failure and the added benefit of better spinal correction with fewer levels fused [21,24]. The hybrid construct uses the benefits of anterior instrumentation in patients who are already undergoing thoracotomy. As a result, a similar degree of overall correction is feasible with shortened posterior instrumentation, thereby reducing surgical morbidity. During the perioperative period, estimated surgical blood loss and surgical time were reduced significantly in the hybrid group. Although additional time was required to place the anterior instrumentation, this was more than offset by the dramatically shorter posterior construct required. On average, nearly eight motion segments were saved from fusion using the hybrid technique compared with the traditional anteroposterior technique. With the hybrid construct, there was an instrumentation overlap of one to four lumbar levels, and most anterior instrumentation did not extend above T10. Because thoracic and thoracolumbar curves were corrected

M. Yagi et al. / The Spine Journal

through anterior instrumentation, only short posterior instrumentation was needed for distal fusion. With a short posterior arthrodesis, the hybrid construct saved a significant amount of levels in the proximal direction. Although the benefits of saving distal fusion levels (mostly lumbar) have been documented, the advantages of saving thoracic levels are unclear [39–41]. Preservation of thoracic movement, however, may have a positive impact on global balance as well as demonstrate beneficial correlation between mobile discs and reduction of neck pain. These concepts were not pursued in detail for this study. Despite a reduction in motion segments fused, curve correction was efficacious with the hybrid technique. On average, the thoracolumbar/lumbar curve improved by 61% and the lumbosacral fractional curve improved by 67%, both significantly better than that achieved with the traditional anteroposterior technique. These results are similar to previous reports describing long spinal fusion to the sacrum [1,3,5,11]. Sagittal alignment and curvature is equally important to coronal curve measurement in long-term patient outcomes. Thoracic kyphosis was worse in the control group than hybrid group preoperatively (47
vs.36
). Postoperatively, there was not a significant difference in thoracic kyphosis from preoperative values. At final follow-up, thoracic kyphosis in the hybrid group did not increase significantly, with only a slight increase from 36
to 39
, despite fusion that stopped at the lower thoracic spine. In addition, the anterior instrumentation used did not result in any loss of lumbar lordosis (45
preoperatively vs. 43
at final follow-up, p5.3). However, lumbar lordosis at final follow-up was significantly less than that achieved in the traditional anteroposterior group (43
vs. 50
, p5.02), perhaps indicating that the longer posterior instrumentation in the control group does help to create and maintain lordosis. Sagittal and coronal balance was significantly worse in the control group preoperatively. Postoperatively, coronal balance was significantly improved in the hybrid group compared with the control group; however, this could be attributed to the preoperative differences between groups. However, the hybrid group did seem more prone to radiographic sagittal decompensation than the control group. In the hybrid group, 16% of patients were classified as having sagittal decompensation preoperatively, which increased to 31% (p5.06) at final follow-up. To date, anterior rod instrumentation has not been commonly associated with sagittal decompensation [21,25,29,42]. In our study, the preserved thoracic movement resulting from shorter instrumentation may have contributed to a greater increase in sagittal imbalance. However, despite the higher incidence of sagittal decompensation, more than 4 cm postoperatively in the hybrid group, there was a significant decrease in the incidence of PJK according to the criteria described by Kim et al. [14]. At final follow-up, abnormal PJK was present in 9% of the hybrid group compared with 28% in the control group (p5.02). This rate of PJK compares favorably with the reported incidence of 39% in adult

-

(2013)

-

7

deformity by 8 years of follow-up [14]. We hypothesize that the shorter posterior instrumentation used in the hybrid technique allows for better preservation of the posterior tension band (musculature, ligaments, joint capsule) at the junctional levels, which helps to prevent PJK. Historically, anterior instrumentation was associated with a greater rate of instrument failure and pseudoarthrosis compared with posterior instrumentation [21,43]. With an average follow-up of more than 5 years, the hybrid construct shows a decreased rate of revision procedures compared with the traditional anteroposterior group (18% vs. 36%). High rates of complications with long posterior spinal fusion are well documented in the literature, including difficulty in obtaining a fusion at the lumbosacral junction [1,2,4–6,17]. Interestingly, not a single hybrid construct used pelvic fixation; however, not one construct developed evidence of a pseudoarthrosis at L5/S1 at final follow-up. We do not know which variable contributes to this high fusion rate, but it is conceivable that the shorter posterior instrumentation places decreased stress on the lumbosacral junction and the sacral instrumentation. The weakness of this study is that patient outcome data are somewhat limited by the availability of only postoperative SRS-22r and ODI data. Preoperative date would be useful in helping to gauge improvement from each type of intervention. From the time the majority of the index procedures in this study were performed (1996–2005), other techniques for treating adult deformity have been developed. The use of all pedicle screw fixation techniques have increased, with a subsequent decrease in the need for anterior procedures, even in adult deformity [44,45]. The results of this study must be viewed in the context that they are not compared with the newer techniques of allposterior pedicle screw constructs. On the other hand, posterior spinal fusion has mainly two potential problems: one is the failure of fixation at the lumbosacral junction, and the other is that pseudoarthrosis has been implicated in poor outcome. Another frequent complication is the development of PJK in nearly 40% to 60% of patients. Currently, a common technique for obtaining long-segment spinal fusion to the sacrum uses anterior column support with cages and bone graft, coupled with a long posterior instrumentation, including fixation to the pelvis. A combined anteroposterior procedure confers several advantages, including better correction of deformity in severe curves, reduction of pseudoarthrosis, and less likelihood of curve progression. Efforts have been made to develop techniques that can obtain a similar correction but require less dissection and shorter fusions. We have believed that this procedure enable us the sufficient curve correction within a short segment, overcome the practical limitation of anterior rods construct, and will reduced potential segmental posterior pedicle screw construct-related complications (PJK and pseudoarthrosis at the lumbosacral junction). The low complication rate and low revision rate of this procedure benefit patients and reduce the cost of the treatment.

8

M. Yagi et al. / The Spine Journal

In addition, an extensive thoracolumbar approach was used in both groups of patients in this study. This study did not quantify directly the morbidity of a thoracolumbar approach in terms of patient satisfaction with appearance, abdominal bulges, persistent pain, hernias, and so forth. It is known from other work that there is significant longterm morbidity in the thoracolumbar approach [46]. The other potential weakness of this study is small sample size and the heterogeneity of the patient population. The wide range of follow-up (2–11 years) and age distribution (range, 31–76 years) makes the population less homogenous, but we believe the study population is well controlled in terms of age, gender, curve size, diagnosis, and so forth. Major vascular injury is one of the most devastating complications of anterior spinal fusion, whereas this never occurs in a posterior spinal fusion. The reported incidence of major vascular injury is 3% to 5% [47,48]. In the current study, although we did not experience major vascular injury in both groups, the small sample size of the study makes it less efficient in understanding the true incidence of such a catastrophic complication in the study population. The same can be said for proximal junctional failure with neurologic deficit. Proximal junctional failure is one of the most devastating complications recently recognized as a junctional kyphosis at the upper instrumented vertebrae. The reported incidence of proximal junctional failure is about 5% in posterior spinal fusion for adult spinal deformity [32,49,50]. Additional studies of a large population are needed to validate the benefit of the hybrid technique for patients with adult spine deformity.

Conclusion This retrospective analysis demonstrates that the hybrid construct, a partially overlapped anterior instrumented fusion with a short posterior fusion, may be a valid alternative in select patients with thoracolumbar/lumbar scoliosis. Based on our retrospective analysis of 66 patients, the overall curve correction was improved in the hybrid group, with fewer levels fused, decreased blood loss, and fewer revision procedures, mostly as a result of decreased proximal junctional disease. Additional studies in a large, long-term, multicenter population are needed to address the benefits and the potential serious complications of a partially overlapped anterior instrumented fusion with a short posterior fusion on patient outcome. References [1] Balderstrom RA, Winter RB, Moe JH, et al. Fusion to the sacrum for nonparalytic scoliosis in the adult. Spine 1986;11:824–9. [2] Boachie-Adjei O, Dendrinos GK, Ogilvie JW, et al. Management of adult spinal deformity with combined anterior-posterior arthrodesis and Luque-Galveston instrumentation. J Spinal Disord 1991;4: 131–41.

-

(2013)

-

[3] Devline VJ, Boachie-Adjei O, Bradford DS, et al. Treatment of adult spinal deformity with fusion to the sacrum using CD instrumentation. J Spinal Disord 1991;4:1–14. [4] Emami A, Deviren V, Berven S, et al. Outcome and complications of long fusion to the sacrum in adult spine deformity. Spine 2002;27:776–86. [5] Kostuik JP, Hall BB. Spinal fusion to the sacrum in adults with scoliosis. Spine 1983;8:489–500. [6] Kuklo TR, Bridwell KH, Lenke LG, et al. Minimum 2-year analysis of sacropelvic fixation and L5-S1 fusion using S1 and iliac screws. Spine 2001;26:1976–83. [7] Glazer PA, Colliou O, Lotz JC, et al. Biomechanical analysis of lumbosacral fixation. Spine 1996;21:1211–22. [8] Leong JCY, William WL, Yinggang Z, et al. Comparison of the strengths of lumbosacral fixation achieved with techniques using one and two triangulated sacral screws. Spine 1998;23:2289–94. [9] Linville DA, Bridwell KH, Lenke LG, et al. Complication in the adult deformity patients having combined surgery. Spine 1999;24: 355–63. [10] Saer EH, Winter TB, Lonstein JE. Long fusion to the sacrum in adults with nonparalytic scoliosis: an improved method. Spine 1990;15:650–3. [11] Weistroffer JK, Perra JH, Lonstein JE. Complications in long fusions to the sacrum for adult scoliosis: a minimum five year analysis of fifty patients. Spine 2008;33:1478–83. [12] Bridwell KH, Edwards CC, Lenke LG. The pros and cons to saving the L5–S1 motion segment in a long scoliosis fusion construct. Spine 2003;28:S234–42. [13] Islam NC, Wood KB, Transfeldt EE, et al. Extension of fusions to the pelvis in idiopathic scoliosis. Spine 2001;26:166–73. [14] Kim YJ, Bridwell KH, Lenke LG, et al. Proximal junctional kyphosis in adult spinal deformity after segmental posterior spinal instrumentation and fusion: minimum five-year follow-up. Spine 2008;33: 2179–84. [15] Byrd A, Scoles PV, Winter RB, et al. Adult idiopathic scoliosis treated by anterior and posterior spinal fusion. J Bone Joint Surg Am 1987;69A:843–50. [16] Lonstein JE, Akbarnia BA. Operative treatment of spinal deformities in patients with cerebral palsy or mental retardation. J Bone Joint Surg Am 1983;65:43–55. [17] McMaster MJ. Anterior and posterior instrumentation and fusion of thoracolumbar scoliosis due to myelomeningocele. J Bone Joint Surg Br 1987;69:20–5. [18] Shufflebarger HL, Grimm JO, Bui V, et al. Anterior and posterior spinal fusion: ‘‘staged’’ versus same-day. Spine 1991;16:930–3. [19] Spivak JM, Neuwirth MG, Giordano CP, et al. The perioperative course of combined anterior and posterior spinal fusion. Spine 1994;19:520–5. [20] Boachie-Adjei O, Charles G, Cunningham ME. Partially overlapping limited anterior and posterior instrumentation for adult thoracolumbar and lumbar scoliosis: a description of novel spinal instrumentation: ‘‘the hybrid technique’’. HSS J 2007;3:93–8. [21] Betz RR, Harms J, Clements DH, et al. Comparison of anterior and posterior instrumentation for correction of adolescent thoracic idiopathic scoliosis. Spine 1999;24:225–39. [22] Hammerberg KW, Zielke K. VDS instrumentation for idiopathic thoracic curvatures. Paper presented at: 1985 annual meeting of the American Academy of Orthopaedic Surgeons. Las Vegas, NV, January 22-29, 1985. [23] Kostuik JP, Carl A, Ferron S. Anterior Zielke instrumentation for spinal deformity in adults. J Bone Joint Surg Am 1989;71:898–912. [24] Lowe TG, Betz RR, Lenke LG, et al. Anterior single-rod instrumentation of the thoracic and lumbar spine: saving levels. Spine 2003;28:S208–16. [25] Majd ME, Castro FP, Holt RT. Anterior fusion for idiopathic scoliosis. Spine 2000;25:696–702.

M. Yagi et al. / The Spine Journal [26] Lenke LG, Betz RR, Harms J, et al. Spontaneous lumbar curve coronal correction after selective anterior or posterior thoracic fusion in adolescent idiopathic scoliosis. Spine 1999;24:1663–71. [27] Kaneda K, Shono Y, Satoh S, et al. New anterior instrumentation for the management of thoracolumbar and lumbar scoliosis. Spine 1996;21:1250–62. [28] Luk KD, Leong JC, Reyes L, et al. The comparative results of treatment of idiopathic thoracolumbar and lumbar scoliosis using the Harrington, Dwyer, and Zielke instrumentations. Spine 1989;14:275–80. [29] Sweet FA, Lenke LG, Bridwell KH, et al. Prospective radiographic and clinical outcomes and complications of single solid rod instrumented anterior spinal fusion in adolescent idiopathic scoliosis. Spine 2001;26:1956–65. [30] Johnson CE II, Turi M, Richards BS. Treatment of idiopathic scoliosis with anterior TSRH instrumentation. Orthop Trans 1994;18:118–9. [31] Kaneda K, Shono Y, Satoh S, et al. Anterior correction of thoracic scoliosis with Kaneda anterior spinal system: a preliminary report. Spine 1997;22:1358–68. [32] Yagi M, King A, Boachie-Adjei O. Incidence, risk factors and classification of proximal junctional kyphosis: surgical outcomes review of adult idiopathic scoliosis. Spine 2011;36:60–8. [33] Jackson RP, McManus AC. Radiographic analysis of sagittal plane alignment and balance in standing volunteers and patients with low back pain matched for age, sex and size: a prospective controlled clinical study. Spine 1994;19:1611–8. [34] Allen BL, Ferguson RL. A pictorial guide to the Galveston LRI pelvic fixation technique. Contemp Orthop 1983;7:51–61. [35] Allen BL, Ferguson RL. The Galveston technique of pelvic fixation with 1-rod instrumentation of the spine. Spine 1984;9:388–94. [36] Gau YL, Lonstein JE, Winter RB, et al. Luque-Galveston procedure for correction and stabilization of neuromuscular scoliosis and pelvic obliquity: a review of 68 patients. J Spinal Disord 1991;4:399–410. [37] Brown JC, Swank S, Specht L. Combined anterior and posterior spine fusion in cerebral palsy. Spine 1982;7:570–3. [38] Floman Y, Micheli LJ, Penny N, et al. Combined anterior and posterior fusion in seventy-three spinally deformed patients: indications, results and complications. Clin Orthop 1982;164:110–22.

-

(2013)

-

9

[39] Spencer DL, DeWald RL. Simultaneous anterior and posterior surgical approach to the thoracic and lumbar spine. Spine 1979;4:29–36. [40] Cochran T, Irstam L, Nachemson A. Long-term anatomic and functional changes in patients with adolescent idiopathic scoliosis treated by Harrington rod fusion. Spine 1983;8:576–84. [41] Ginsburg HH, Goldstein L, Haake PW, et al. Longitudinal study of back pain in postoperative idiopathic scoliosis: long-term follow-up. Paper presented at: the 30th annual meeting of the Scoliosis Research Society. Asheville, NC, September 13-16, 1995. [42] Smith JA, Deviren V, Berven S, et al. Does instrumented anterior scoliosis surgery lead to kyphosis, pseudoarthrosis, or inadequate correction in adults? Spine 2002;27:529–34. [43] Trammell TR, Benedict F, Reed D. Anterior spine fusion using Zielke instrumentation for adult thoracolumbar and lumbar scoliosis. Spine 1991;16:307–16. [44] Good CR, Lenke LG, Bridwell KH, et al. Can posterior-only surgery provide similar radiographic and clinical results as combined anterior (thoracotomy/thoracoabdominal)/posterior approaches for adult scoliosis? Spine 2010;35:210–8. [45] Rose PS, Lenke LG, Bridwell KH. Pedicle screw instrumentation for adult idiopathic scoliosis: an improvement over hook/hybrid fixation. Spine 2009;34:852–7. [46] Kim YB, Lenke LG, Kim YJ, et al. The morbidity of an anterior thoracolumbar approach: adult spinal deformity patients with greater than five-year follow-up. Spine 2009;34:822–6. [47] Fantini GA, Pappou IP, Girardi FP, et al. Major vascular injury during anterior lumbar spinal surgery: incidence, risk factors, and management. Spine 2007;32:2751–8. [48] Wood KB, Devine J, Fischer D, et al. Vascular injury in elective anterior lumbosacral surgery. Spine 2010;35:S66–75. [49] Hart RA, McCarthy I, Ames CP, et al. Proximal junctional kyphosis and proximal junctional failure. Neurosurg Clin N Am 2013;24: 213–8. [50] Hostin R, McCarthy I, O’Brien M, et al. Incidence, mode, and location of acute proximal junctional failures following surgical treatment for adult spinal deformity. Spine 2012 Sep 13. [Epub ahead of print].