Posterior-only correction of Scheuermann kyphosis using pedicle screws: economical optimization...

ORIGINAL ARTICLE

Posterior-only correction of Scheuermann kyphosis using pediclescrews: economical optimization through screw density reduction

Eyal Behrbalk • Ofir Uri • Ruth M. Parks •

Michael Paul Grevitt • Marcus Rickert •

Bronek Maximilian Boszczyk

Received: 14 July 2013 / Revised: 30 May 2014 / Accepted: 14 July 2014

� Springer-Verlag Berlin Heidelberg 2014

Abstract

Introduction Posterior-only approach using pedicle

screws’ fixation has emerged as the preferred surgical

technique for Scheuermann kyphosis (SK) correction.

Insertion of multiple pedicle screws while increasing sta-

bility increases also the risk of complications related to

screw malpositioning and surgical cost. The optimal screw

density required in surgical correction of SK remains

unclear. This study compares the safety and efficacy of low

screw density (LSD) versus high screw density (HSD)

technique used in posterior-only correction of SK.

Methods Twenty-one patients underwent surgical cor-

rection of SK between 2007 and 2011 and were reviewed

after a mean of 29 months. HSD technique (i.e., 100 % of

available pedicles, averaged 25.2 ± 4 screws) was used in

10 cases and LSD technique (i.e., 54–69 % of available

pedicles in a pre-determined pattern, averaged 16.8 ± 1.3

screws; p \ 0.001) was used in 11 cases. Kyphosis cor-

rection was assessed by comparing thoracic kyphosis,

lumbar lordosis and sagittal balance on preoperative and

postoperative radiographs. Cost saving analysis was per-

formed for each group.

Results Preoperative thoracic kyphosis, lumbar lordosis

and sagittal balance were similar for both groups. The

average postoperative kyphosis correction was similar in

both HSD and LSD groups (29� ± 9� vs. 34� ± 6�,

respectively; p = 0.14). Complication occurred in four

patients (19 %) in the HSD group and in two patients (9 %)

in the LSD group (p = 0.56). Three patients required re-

operation. Compared to HSD using LSD saves 4,200£ per

patient in hardware and 88,200£ for the entire cohort.

Conclusion LSD technique is as safe and effective as

HSD technique in posterior-only correction of SK.

Implant-related cost could be reduced by 32 %.

Keywords Scheuermann kyphosis � Economical

optimization � Screw density � Pedicular screw � Cost

reduction � Correction

Introduction

Surgical correction of Scheuermann kyphosis (SK) is an

uncommon and challenging procedure. According to the

Scoliosis Research Society records less than 1 % of

spine surgeries performed by its members are carried out

for the treatment of SK [1]. The natural history of SK

seems to follow a benign course in the majority of cases,

with settling of symptoms and stabilization of the

deformity at skeletal maturity [2–4]. However, both eti-

ology and natural history of SK are not fully understood

and, therefore, the best practice strategy for the treatment

of SK remains unclear. Despite controversies, surgical

correction is usually reserved to skeletally mature

patients with severe back pain and curves over 75� [5–

7]. The first case series of SK correction was described

in 1975 by Bradford et al. [8]; deformity correction and

In this investigation the high screw density cohort is contributed by

Mr. Sm Mehdian.

E. Behrbalk (&) � R. M. Parks � M. P. Grevitt � M. Rickert �B. M. Boszczyk

The Centre for Spinal Studies and Surgery, Queen’s Medical

Centre, Nottingham, UK

e-mail: eyalbehrbalk@yahoo.com

O. Uri

Royal National Orthopaedic Hospital NHS Trust Brockley Hill,

Stanmore, Middlesex, UK

123

Eur Spine J

DOI 10.1007/s00586-014-3472-y

fusion were performed through a posterior-only approach

with Harrington instrumentation. Unfortunately, due to

unacceptably high complications rate and correction loss

with this technique, a combined approach of anterior

release followed by posterior instrumented spinal fusion

has been considered as the gold standard for SK cor-

rection for many years [3, 9]. The introduction of the

pedicle screw (which enables strong three column fixa-

tion) into spine surgery over the past years resulted in

growing evidence that posterior-only correction is com-

parable to combined anterior–posterior approach for the

treatment of SK [10–14]. Posterior-only approach short-

ens surgical time, minimizes exposure and blood loss

and may reduce postoperative complications [11, 15] and

has emerged as the preferred approach by many sur-

geons [2, 3, 16]. Using multiple pedicle screws increases

fixation stability, but may also increase the risk of

complications associated with screw malpositioning. The

optimal pedicle screw density (i.e., number of pedicle

screw used/number of available pedicle insertion sites)

required to maintain a stable fixation following correc-

tion of spinal deformities remains controversial. Previous

studies which evaluated the correlation of fixation

anchors’ density with curve correction and fixation sta-

bility in scoliosis surgery, showed no conclusive findings

[17, 18]. Nevertheless, the higher failure rate associated

with kyphosis correction compared to other spinal

deformities [1] may deter surgeons from using less than

the maximal fixation screws available.

To our knowledge, no previous study has evaluated

the influence of pedicle screw fixation density on defor-

mity correction and fixation stability in the treatment of

SK. The purpose of the current study was to evaluate the

safety and efficacy of low pedicle screws density (LSD)

technique in correcting SK through a posterior-only

approach and to compare the results to those of correc-

tion using high screw density technique (HSD) along

with determining the potential for implant-related cost

savings.

Patients and methods

Study design

The study was conducted at our spine surgery unit in our

university affiliated teaching hospital. We retrospectively

reviewed and analyzed data of patients who underwent

surgical correction for severe SK between January 2007

and January 2011. This investigation fulfills the criteria of

service evaluation of our institution and as such does not

require formal ethics committee appraisal.

Patient population

Medical records and imaging of 21 patients (average age

21 ± 7 years, 86 % males) who underwent surgical cor-

rection of advanced symptomatic SK in our hospital were

reviewed after a minimum follow-up of 24 months. Clin-

ical data were obtained prospectively and included detailed

medical history, physical examination and imaging of the

entire spine (i.e., high quality anterior–posterior and lateral

radiographs and MRI scan). All the patients had persistent

symptoms of severe back pain affecting activities of daily

living, not improving with non-operative treatment of at

least 12 months. No patient had neurological deficit. Pre-

operative imaging demonstrated thoracic kyphosis C60� in

all the patients (average 75� ± 9�, range 61�–96�). Ante-

rior wedging C5� at C3 consecutive thoracic vertebras,

endplate irregularities and/or Schmorl nodules were also

found in all the patients. No concomitant spinal pathology

(e.g., congenital anomaly, previous surgery, coronal plane

deformity C10�) was found in preoperative MRI scans of

the study cohort. All procedures were performed through a

posterior-only approach using pedicle screws. In ten

patients, pedicle screws were inserted into all available

pedicles in the fused segment (‘‘high screw density’’

technique), i.e., 100 % of available pedicles were used. In

11 patients pedicle screw density was reduced by avoiding

placement in all pedicles across the apex of deformity

(‘‘low screw density’’ technique), i.e., 65 ± 5 % of pedi-

cles were used (range, 54–69 %). The average number of

pedicle screws used per patient in the HSD group was

25.2 ± 4 compared to 16.8 ± 1.3 screws per patient which

were used in the LSD group (p \ 0.01). The average

number of levels fused per patient was similar for both

HSD and LSD groups (12.6 ± 2 vs. 13 ± 0.4, respec-

tively; p = 0.52). Naturally, the screws’ density in the

HSD group was significantly higher than in the LSD group

(100 ± 0 % vs. 65 ± 5 %, respectively, p \ 0.01).

Surgical technique

All corrections were performed by one of three senior spine

consultants team in our spine unit, using a similar surgical

technique. A posterior-only approach using pedicle screws

fixation was performed in all the cases. The superior and

inferior ends of the curve were determined preoperatively

by the Cobb method. Patients were placed prone on the

operating table with appropriate padding and support. A

standard posterior midline approach was used. The para-

spinal muscles were sub-periosteally elevated and the

posterior elements of the fusion segments were exposure.

Mono axial pedicle screws’ system was used in all the

procedures. Deformity correction and fusion were carried

Eur Spine J

123

out from the first thoracic vertebra involved in the kyphosis

to the lumbar vertebra below the first lordotic disc [19].

Pedicle screws were inserted using a free-hand technique

or under controlled insertion with the help of image

intensifier. Screw insertion point was at the junction of the

transverse process and lateral margin of the superior facet

joint in the lower thoracic and lumbar spine and at the

junction of the superior margin of the transverse processes

with the middle of the inferior facet joint in the upper

thoracic spine. To enable deformity correction, an apical

Ponte osteotomy and posterior column shortening were

carried out [20, 21]. Deformity correction was performed

utilizing two 6-mm titanium rod with or without cross-

links. Somatosensory and neurogenic motor-evoked

potentials were monitored throughout the procedure. The

posterior elements were decorticated to bleeding bone and

the bone graft obtained was placed along the length of the

instrumentation to facilitate fusion.

Screw placement patterns

HSD technique was used by one consultant and his team

in 10 patients and included screws’ insertion into all

available pedicles in the fused segment (i.e., 100 % of

pedicles were used). The fused levels in these patients

were T2-L2 in seven patients, T1-L2 in one patient, T2-

L3 in one patient and T7-L1 in one patient. LSD

technique was used by two consultants and their team in

11 patients and included: (1) screws’ insertion into all

four pedicles at the two most cephalad and two most

caudal levels of the fused segment, (2) screws at the

remaining levels were inserted either into alternate pedi-

cles of each level (pattern A, used in 6 patient), or into

both pedicles of every other level (pattern B, used in 5

patients). The fused levels in the LSD group were T3-L3

in eight patients, T1-L2 in one patient, T3-L2 in one

patient and T1-L1 in one patient. Figures 1, 2, and 3 show

the screw insertion patterns used in this study.

Postoperatively, Patients were encouraged to sit and

stand within the first 24 h after surgery under supervision

of a physiotherapist and were advanced to assisted ambu-

lation as tolerated on the second postoperative day. Return

to unrestricted activity was allowed at 3–6 months pro-

vided that radiographs confirmed fusion.

Outcome measures

Kyphosis correction was assessed by comparing preoper-

ative unassisted standing lateral radiographs of the entire

spine with postoperative radiographs at the latest follow-up

of at least 24 months. Radiographic parameters of kyphosis

correction were measured digitally by two spine fellows

and one orthopedic specialist not involved in the opera-

tions, using PACS (picture archiving and communication

Fig. 1 25-year-old male with advanced symptomatic Scheuermann

kyphosis, unresponsive to non-operative treatment. A surgical

correction through a posterior-only approach using 100 % pedicle

screw density technique was performed. Preoperative (a) and post-

operative standing lateral radiographs at 30-month follow-up (b) are

presented

Eur Spine J

123

system, General Electric, Health care systems, Easton

Turnpike Fairfield, US) and included: (1) thoracic kypho-

sis, defined as T5-T12 Cobb angle, (2) lumbar lordosis,

defined as L1-S1 Cobb angle, (3) sagittal balance, defined

as the displacement (measured in mm) of C7 plumb line in

relation to the superior posterior endplate of S1 (i.e.,

considered positive if anterior and negative if posterior).

Sagittal balance correction was defined as the overall shift

of C7 plumb line in relation to the posterior–superior

endplate of S1 (i.e., if the sagittal balance was changed

from ?20 mm before surgery to -10 mm after surgery, the

‘‘correction’’ was considered as 30 mm).

Fig. 2 A 19-year-old male with advanced symptomatic Scheuermann

kyphosis, unresponsive to non-operative treatment. A surgical

correction through a posterior-only approach using low pedicle screw

density technique (pattern A) was performed. Preoperative (a) and

postoperative standing lateral radiographs at 24-month follow-up

(b) are presented

Fig. 3 A 20-year-old male with advanced symptomatic Scheuermann

kyphosis, unresponsive to non-operative treatment. A surgical

correction through a posterior-only approach using low pedicle screw

density technique (pattern B) was performed. Preoperative (a) and

postoperative standing lateral radiographs at 24-month follow-up

(b) are presented

Eur Spine J

123

Data Analysis

Continuous parameters are described as the mean and the

standard deviation (SD) and categorical parameters are

described as proportions. Comparison between the groups

is presented with 95 % confidence intervals (CI) around the

difference. Unpaired two-tailed t test was used to compare

preoperative and postoperative values of continuous

parameters between the groups. Categorical parameters

were compared using two-tailed Fisher’s exact test. Sta-

tistical analysis was performed using SPSS for Windows

software (version 16.0; IBM, Chicago, Illinois). A proba-

bility of \0.05 was considered statistically significant.

Results

Both HSD and LSD groups were similar in terms of

patients’ age, gender, preoperative radiographic parameters

(e.g., thoracic kyphosis, lumbar lordosis and sagittal bal-

ance) and postoperative follow-up. The average thoracic

kyphosis was similar in early postoperative radiographs

and at the latest follow-up for both the HSD group

(42� ± 7� and 43� ± 9�, respectively; p = 0.78) and the

LSD group (44� ± 9� and 44� ± 8�, respectively;

p = 0.99). At the latest follow-up, thoracic kyphosis cor-

rection was 29� ± 9� in the HSD group and 34� ± 6� in

the LSD group (p = 0.14). Lumbar-lordosis correction was

15� ± 11� in the HSD group and 23� ± 14� in the LSD

group (p = 0.16). Sagittal balance correction was similar

for both HSD and LSD as well (16 ± 12 mm and

23 ± 15 mm, respectively; p = 0.25). Comparison of

patients’ characteristics and clinical data between the HSD

and LSD groups is summarized in Table 1.

We further analyzed LSD group based on the two dif-

ferent patterns of pedicle screws’ insertion, pattern A and B

as described earlier. Preoperative patients’ characteristics

and radiographic parameters of deformity were similar for

both groups. Postoperative outcomes of kyphosis correc-

tion were similar for both groups as well (Table 2).

Complications occurred in 6 of 21 patients in our cohort

(28 %). Four complications occurred in the HSD group

Table 1 Comparison of

patients’ characteristics and

clinical data between the high

screw density and the low screw

density groups

Values are presented as

mean ± SD and (range)a Defined as the ratio between

the number of screws used and

the number of available pediclesb Anterior displacement of C7

plumb line in relation to the

posterior–superior sacral corner

was considered positive and

posterior displacement as

negativec Comparison was performed

using unpaired two-tailed t testd Comparison was performed

using two-tailed Fisher’s exact

test

High screw

density

N = 10

Low screw density

N = 11

Difference

(95 % CI)

p value

Age at surgery, years 22 ± 8 (15–34) 19 ± 6 (12–33) 3 (-3, 9) 0.34c

Gender, N (%)

Males 8 (80 %) 10 (90.9 %) 10.9 % (-21, 42) 0.58d

Females 2 (20 %) 1 (9.1 %)

Number of levels fused 12.6 ± 2.0 (7–14) 13.0 ± 0.4

(12–14)

0.4 (-1, 2) 0.52c

Number of screws inserted 25.2 ± 4.0

(14–28)

16.8 ± 1.3

(14–18)

8.4 (6, 11) \0.01c

Screw densitya, % 100 ± 0

(100–100)

65 ± 5 (54–69) 35 (32, 38) \0.01c

Postoperative follow-up,

months

32 ± 11 (24–56) 26 ± 6 (24–42) 6 (-2, 14) 0.13c

Thoracic kyphosis, degrees

Preoperative 72 ± 7 78 ± 9 6 (-1, 13) 0.10c

Early postoperative 42 ± 7 44 ± 9 2 (-5, 9) 0.58c

Latest follow-up 43 ± 9 44 ± 8 1 (-7, 9) 0.79c

Correction 29 ± 9 34 ± 6 5 (-2, 12) 0.14c

Lumbar lordosis, degrees

Preoperative 71 ± 8 73 ± 12 2 (-7, 11) 0.66c

Latest follow-up 56 ± 10 50 ± 14 6 (-5, 17) 0.27c

Correction 15 ± 11 23 ± 14 8 (-4, 19) 0.16c

Sagittal balanceb, mm

Preoperative 27 ± 28 20 ± 24 7 (-17, 31) 0.54c

Latest follow-up 11 ± 21 -3 ± 25 14 (-7, 35) 0.18c

Correction 16 ± 12 23 ± 15 7 (-5, 19) 0.25c

Eur Spine J

123

(19 %) and two (9 %) in the LSD group (p = 0.56).

Complications in the HSD group included: (1) one patient

had screw penetration into T9 disc space that was detected

on postoperative spine radiograph. The patient remained

asymptomatic and no further intervention was required, (2)

another patient developed pseudo-arthrosis at T7-8 level

with breakage of both fixation rods at 6-month follow-up

and correction loss of 15�, subsequently. Nevertheless, the

patient had only marginal symptoms and declined further

surgery, (3) a third patient developed unilateral three-level

screw loosening (T3–T5 levels) at 3-month follow-up,

without correction loss. The patient underwent revision

surgery with instrumented fusion after infection free

interval of 5 months and normalization of all inflammatory

markers. At latest follow-up no recurrence of infection

noted, radiographs showed good bone fusion mass, (4) the

forth patient had unilateral screw loosening at L1–L2 levels

detected on 12-month radiographs after complete bone

union occurred. The patient remained asymptomatic and no

further intervention was required. Complications in the

LSD group included: (1) one patient developed proximal

junctional kyphosis of 20� at T2–T3 level at 1.5 months

follow-up. The patient was re-operated and T3 Smith-

Petersen osteotomy with fusion extension to T1 level was

carried out. Following this procedure, the patient

developed spastic paraplegia, which resolved over a

6-month time, (2) another patient developed acute deep

wound infection which was treated by surgical debride-

ment 2 weeks postoperatively and intravenous antibiotics.

Further debridement with metalwork removal was carried

out 4 weeks later on due to persistent infection. The patient

underwent second stage revision with instrumentation after

clearance of the infection (5 months after the index pro-

cedure) without recurrence of infection at the latest follow-

up.

Discussion

The first case series reporting on outcomes of surgical

treatment for SK was published in 1975 by Bradford et al.

[8] who reviewed 22 patients following posterior-only

correction and fusion with Harrington instrumentation.

Thoracic kyphosis was corrected from an average of 72�before surgery to 47� at 35-month follow-up. Correction

loss of 21� in average was noticed in 16 patients (72 %).

Larger and more rigid sagittal curves were found to

increase the risk for correction loss and implant failure.

Combined anterior–posterior fusion was recommended for

these deformities and remained the treatment of choice for

Table 2 Subgroup analysis of

the low screw density group

based on screw insertion pattern

Values are presented as

mean ± SDa Pedicle screws were inserted

into all four pedicles at the two

most cephalad and two most

caudal levels of the fused

segment and into either the

opposite pedicle of each of the

remaining levels (Pattern A), or

into both pedicles of every other

level of the remaining levels

(Pattern B)b Defined as the ratio between

the number of screws used and

the number of available pediclesc Anterior displacement of C7

plumb line in relation to the

posterior–superior sacral corner

was considered positive and

posterior displacement as

negatived Comparison was performed

using unpaired two-tailed t teste Comparison was performed

using two-tailed Fisher’s exact

test

Low screw

density pattern Aa

N = 6

Low screw

density pattern Ba

N = 5

Difference

(95 % CI)

p value

Age at surgery, years 17 ± 3 21 ± 7 4 (-3, 11) 0.23d

Gender, N (%)

Males 6 (100 %) 4 (80 %) 20 % (-22, 62) 0.45e

Females 0 (0 %) 1 (20 %)

Number of levels fused 12.8 ± 0.4 13.2 ± 0.5 0.4 (-0.2, 1) 0.17d

Number of screws inserted 16.6 ± 1.5 17.0 ± 1.2 0.4 (-1.4, 2.3) 0.64d

Screw densityb, % 65 ± 6 64 ± 4 1 (-6, 8) 0.75d

Postoperative follow-up, months 26 ± 3 28 ± 7 2 (-5, 9) 0.54d

Thoracic kyphosis, degrees

Preoperative 79 ± 10 78 ± 9 1 (-12, 11) 0.86d

Early postoperative 42 ± 11 45 ± 8 3 (-10, 16) 0.62d

Latest follow-up 43 ± 9 45 ± 7 2 (-9, 13) 0.69d

Correction 36 ± 7 33 ± 5 3 (-5, 11) 0.44d

Lumbar lordosis, degrees

Preoperative 74. ± 15 72 ± 9 2 (-15, 19) 0.80d

Latest follow-up 52 ± 17 47 ± 10 5 (-14, 24) 0.57d

Correction 22 ± 17 25 ± 13 3 (-18, 24) 0.75d

Sagittal balancec, mm

Preoperative 26 ± 29 14 ± 15 12 (-20, 44) 0.42d

Latest follow-up -2 ± 35 -3 ± 7 1 (-35, 37) 0.95d

Correction 28 ± 16 17 ± 13 11 (-9, 31) 0.24d

Eur Spine J

123

many years [3, 5, 9]. With the advancement in spine sur-

gery and the introduction of pedicle screw, which enables

solid three column fixation, the posterior-only approach

(using pedicle screws) has emerged as the preferred sur-

gical technique for the treatment of SK [2, 3, 16, 18].

Multiple pedicle screws are usually required when

posterior-only approach is used in correction of spinal

deformities. While improving construct stability, HSD also

increases the risk of complications related to screw mal-

positioning, increased time of surgery and raises the sur-

gical cost. The optimal pedicle screw density required to

maintain stability in surgical correction of spinal defor-

mities remains a topic of debate. In theory, for flexible

deformities, one might consider a construct build on strong

foundation screw only (e.g., six screws in three cranial

vertebras and six screws in three caudal ones); this low SD

construct (*25 % SD) might achieve the same clinical and

radiological results as our constructs. Clements et al. [22]

found direct correlation between HSD and major curve

correction in 292 patients following scoliosis correction.

Quan et al. [23] on the other hand found no correlation

between pedicle screws’ density and the magnitude of

coronal and sagittal curves correction in 49 patients fol-

lowing scoliosis correction. SK has been associated with

higher risk of postoperative complications including cor-

rection loss [1, 3, 5] suggesting that HSD may be required

to protect and maintain correction when operating on

patients with this complex deformity. Supporting that, is

Lee et al. [11] case series of 39 patients reported on 18

patients who underwent surgical correction of SK through a

posterior-only approach with 100 % pedicle screws density

technique. Kyphosis was corrected from 84� before surgery

to 40� at a mean of 32 months after surgery. Correction loss

in these patients was 2.2� and no complication occurred.

Kyphosis was corrected from 84� before surgery to 40� at a

mean of 32 months after surgery. Correction loss in these

patients was 2.2� and no complication occurred. A recent

study by Koller et al. [24] followed 111 patients for 2 years

after SK deformity correction, performing anterior release

and posterior segmental fusion using pedicle screws. Their

study reported on 87.3 ± 12.9 % screw density construct

(closer to our HSD group), but with only 8 vertebrae

involved in fusion (our study involved 12.6 ± 2 vertebrae

fusion), the preoperative TK as well as amount of defor-

mity correction at last follow-up was similar in their study

and ours (67.2 ± 12.2� with 30.1 ± 13.5� correction,

75 ± 9� with 31.6� correction, respectively). Interestingly,

although their construct was significantly shorter than ours

they found no correction loss over time and no correlation

between the level of the last instrumented vertebrae and

rate of distal junctional kyphosis.

Whether a 100 % pedicle screws density and long

constructs are indeed necessary in SK correction remains

unclear. Financial constraints, therefore, demand evalua-

tion of the necessity of maximal implant usage. To our

knowledge, no previous study has evaluated the outcome of

LSD technique in the treatment of SK.

The purpose of our study was to evaluate the safety and

efficacy of LSD technique in correcting SK through a

posterior-only approach. The findings of our study show

that posterior-only correction of SK using less pedicle

screw (i.e., 54–69 % of available pedicles) was as safe and

effective as correction using 100 % of available pedicles.

All sagittal correction parameters measured (e.g., thoracic

kyphosis, lumber lordosis and sagittal balance) were sim-

ilar for the HSD group (with an average of 25.2 screws

used per patient) and the LSD group (with an average of

16.8 screws used per patient) at a mean postoperative

follow-up of 32 months and 26 months, respectively. The

pattern of pedicle screws’ insertion in our LSD group did

not seem to affect the correction outcomes. Taking into

account these similar clinical results and the high cost of

each pedicle screw (around £300 per screw), the LSD

technique seems to be more cost effective.

Six of our patients (28 %) presented postoperative

complications (four implant-related complications, one

junctional kyphosis and one deep infection) and three of

them (14 %) required re-operation. Bradford et al. [8] in the

first report of SK correction through a posterior-only

approach described postoperative complications in 18 of

their 22 patients including correction loss, implant failure

and infection. Coe et al. [1] in a multicenter analysis of 683

procedures for SK correction reported on postoperative

complication rate of 15 % in patients who were operated on

through a posterior-only approach. Dennis et al. [25]

reported on postoperative development of proximal junc-

tional kyphosis (of 20� in average) in 20 of their 67 patients

(30 %) who underwent posterior-only correction of SK. Lee

et al. [11], on the other hand, had no postoperative com-

plications in their cohort of 18 patients following posterior-

only correction of SK. The relative high complication rate

found in our study and others [1, 2, 8, 25] highlights that

despite great advances in spine surgery over the past dec-

ades, surgical correction of SK remains a complex and

challenging procedure which should be performed only by

experienced surgeons after careful consideration.

Hardware cost–benefit analysis: UK price of one mono

axial pedicular screw and its cap is 350–600£ (average

500£). Two rods cost 400–600£ (average 520£). Hardware

cost of the typical HSD construct calculated with the

average hardware price amounts to 13,120£ (25.2 screws);

in the LSD construct this amounts to 8,920£ (16.8 screws).

Assuming even pricing across all implants used, this

amounts to 4,200£ (32 %) saving in hardware alone per

case in the LSD group and a total of 88,200£ for all 21

patients of the study cohort.

Eur Spine J

123

Our study has several potential limitations. First, the

cohort numbers are relatively small. However, given that

surgical correction of SK is a complex and uncommon

procedure we believe that these limitations are inevitable.

Second, the average sagittal curve correction in our study

was *30� (29� and 34� in the HSD and LSD groups,

respectively). Whether our results remain unchanged with

greater overall correction (e.g., kyphosis correction of

*40� as reported by Lee et al. [11] ) has to be further

evaluated. Third, this study focused on the radiological

outcome of SK correction regardless of clinical outcomes.

Assuming that radiological outcome of a surgical proce-

dure is an accurate reflection of the clinical results may be

misleading.

In conclusion, our findings suggest that LSD technique

using 54–69 % of available pedicles is as safe and effective

as HSD technique using 100 % of available pedicles for

posterior-only correction of SK, while incurring significant

cost savings.

Conflict of interest None.

References

1. Coe JD, Smith JS, Berven S, Arlet V, Donaldson W, Hanson D,

Mudiyam R, Perra J, Owen J, Marks MC, Shaffrey CI (2010)

Complications of spinal fusion for scheuermann kyphosis: a

report of the scoliosis research society morbidity and mortality

committee. Spine 35:99–103. doi:10.1097/BRS.0b013e3181

c47f0f

2. Tsirikos A, Jain A (2011) Scheuermann’s kyphosis; current

controversies. J Bone Joint Surg Br 93:857–864

3. Wood KB, Melikian R, Villamil F (2012) Adult Scheuermann

kyphosis: evaluation, management, and new developments. J Am

Acad Orthop Surg 20:113–121. doi:10.5435/JAAOS-20-02-113

4. Poolman R, Been H, Ubags L (2002) Clinical outcome and

radiographic results after operative treatment of Scheuermann’s

disease. Eur Spine J 11:561–569

5. Lowe TG, Kasten MD (1994) An analysis of sagittal curves and

balance after Cotrel-Dubousset instrumentation for kyphosis

secondary to Scheuermann’s disease: a review of 32 patients.

Spine 19:1680–1685

6. Lenke L (2004) Kyphosis of the thoracic and thoracolumbar spine

in the pediatric patient: prevention and treatment of surgical

complications. Instr Course Lect 53:501

7. Lamartina C (2010) Posterior surgery in Scheuermann’s kypho-

sis. Eur Spine J 19:515–516

8. Bradford DS, Moe JH, Montalvo F, Winter R (1975) Scheuer-

mann’s kyphosis. Results of surgical treatment by posterior spine

arthrodesis in twenty-two patients. J Bone and Joint Surg

57:439–448

9. Bradford DS, Ahmed KB, Moe J, Winter R, Lonstein J (1980)

The surgical management of patients with Scheuermann’s

disease. J Bone Joint Surg 62:705–712

10. Sturm PF, Dobson JC, Armstrong GW (1993) The surgical

management of Scheuermann’s disease. Spine 18:685–691

11. Lee SS, Lenke LG, Kuklo TR, Valente L, Bridwell KH, Sides B,

Blanke KM (2006) Comparison of Scheuermann kyphosis cor-

rection by posterior-only thoracic pedicle screw fixation versus

combined anterior/posterior fusion. Spine 31:2316–2321. doi:10.

1097/01.brs.0000238977.36165.b8

12. Geck MJ, Macagno A, Ponte A, Shufflebarger HL (2007) The

Ponte procedure: posterior only treatment of Scheuermann’s

kyphosis using segmental posterior shortening and pedicle screw

instrumentation. J Spinal Disord Tech 20:586–593

13. Koptan WM, ElMiligui YH, ElSebaie HB (2009) All pedicle

screw instrumentation for Scheuermann’s kyphosis correction: is

it worth it? Spine J 9:296–302

14. Johnston CE, Elerson E, Dagher G (2005) Correction of adoles-

cent hyperkyphosis with posterior-only threaded rod compression

instrumentation: is anterior spinal fusion still necessary? Spine

30:1528–1534

15. Graham EJ, Lenke LG, Lowe TG, Betz RR, Bridwell KH, Kong

Y, Blanke K (2000) Prospective pulmonary function evaluation

following open thoracotomy for anterior spinal fusion in ado-

lescent idiopathic scoliosis. Spine 25:2319–2325

16. Lowe TG, Line BG (2007) Evidence based medicine: analysis of

Scheuermann kyphosis. Spine 32:S115–S119. doi:10.1097/BRS.

0b013e3181354501

17. Chen J, Yang C, Ran B, Wang Y, Wang C, Zhu X, Bai Y, Li M

(2013) Correction of Lenke 5 adolescent idiopathic scoliosis

using pedicle screw instrumentation: does implant density influ-

ence the correction? Spine

18. Clements DH, Betz RR, Newton PO, Rohmiller M, Marks MC,

Bastrom T (2009) Correlation of scoliosis curve correction with

the number and type of fixation anchors. Spine 34:2147–2150.

doi:10.1097/BRS.0b013e3181adb35d

19. Arlet V, Schlenzka D (2005) Scheuermann’s kyphosis: surgical

management. Eur Spine J 14:817–827

20. Grevitt M, Kamath V, Avadhani A, Rajasekaran S (2010) Cor-

rection of thoracic kyphosis with Ponte osteotomy. Eur Spine J

19:351–352

21. La Maida GA, Misaggi B (2012) Posterior only treatment of adult

thoracic kyphosis with multiple Ponte osteotomies and pedicle

screw instrumentation. Eur Spine J 21:1–5

22. Clements DH, Betz RR, Newton PO, Rohmiller M, Marks MC,

Bastrom T (2009) Correlation of scoliosis curve correction with

the number and type of fixation anchors. Spine 34:2147–2150

23. Quan GM, Gibson MJ (2010) Correction of main thoracic ado-

lescent idiopathic scoliosis using pedicle screw instrumentation:

does higher implant density improve correction? Spine

35:562–567. doi:10.1097/BRS.0b013e3181b4af34

24. Koller H, Juliane Z, Umstaetter M, Meier O, Schmidt R, Hitzl W

(2014) Surgical treatment of Scheuermann’s kyphosis using a

combined antero-posterior strategy and pedicle screw constructs:

efficacy, radiographic and clinical outcomes in 111 cases. Eur

Spine J 23:180–191

25. Denis F, Sun EC, Winter RB (2009) Incidence and risk factors for

proximal and distal junctional kyphosis following surgical treat-

ment for Scheuermann kyphosis: minimum five-year follow-up.

Spine 34:E729–E734

Eur Spine J

123