Modified Arthroscopic Latarjet Procedure

Suture-Button Fixation Achieves Excellent

Remodeling at 3-Year Follow-up
Jian Xu,* MD, PhD, Haifeng Liu,* MD, Wei Lu,*y MD, PhD, Zhenhan Deng,* MD, PhD,
Weimin Zhu,* MD, PhD, Liangquan Peng,* MD, Kan Ouyang,* MD,
Hao Li,* MD, and Daping Wang,* MD, PhD
Investigation performed at Shenzhen Second People’s Hospital,
First Affiliated Hospital of Shenzhen University, Shenzhen, China

Background: Some studies have advocated the use of suture-button fixation during the Latarjet procedure to reduce complica-
tions associated with screw fixation. However, the sample size of these studies is relatively small, and their follow-up period is
short.
Purpose: To investigate the efficacy of the suture-button Latarjet procedure with at least 3 years of follow-up and remodeling of
the coracoid graft.
Study Design: Case series; Level of evidence, 4.
Methods: A total of 152 patients who underwent the suture-button Latarjet procedure between February 2013 and February 2016
were selected, and 128 patients who met the inclusion criteria were enrolled in this study. Preoperative and postoperative clinical
results were assessed. The position and healing condition of the coracoid graft and arthropathy of the glenoid and humeral head
were also assessed using radiography and 3-dimensional computed tomography (CT).
Results: The mean follow-up time was 40.3 6 5.8 months. There were 102 patients included in this study. The mean visual analog
scale score for pain during motion, the American Shoulder and Elbow Surgeons score, the Rowe score, and the Walch-Duplay
score were improved considerably. A total of 100 grafts achieved bone union. The overall absorption rate was 12.6% 6 4.3%.
Graft absorption mostly occurred on the edge and outside the ‘‘best-fit’’ circle of the glenoid. A vertical position was achieved
in 98 grafts (96% of all cases) immediately postoperatively, with the mean graft midline center at the 4 o’clock position. In the
axial view, CT showed that 89 grafts were flush to the glenoid, whereas 2 and 11 grafts were fixed medially and laterally, respec-
tively. In all cases, the bone graft and glenoid tended to extend toward each other to form concentric circles during the remodel-
ing process. During follow-up observations, the height of the 11 grafts that were positioned laterally (ie, above the glenoid level)
exhibited a wave-curved change. No arthropathy was observed in any patient.
Conclusion: Patient outcomes were satisfactory after the modified arthroscopic suture-button Latarjet technique. Graft absorp-
tion mostly occurred on the edge and outside the ‘‘best-fit’’ circle of the glenoid. The graft exhibited the phenomenon of ectatic
growing when it fused with the glenoid and finally remodeled to a new concentric circle with the humeral head analogous to the
original glenoid. Grafts positioned laterally did not cause arthropathy of the joints within the period of the study.
Keywords: shoulder; anterior instability; glenoid defect; Latarjet; outcome; suture button; remodeling

The arthroscopic Latarjet procedure has gradually become
popular in recent years.5,16,17,20,23 Metal screws are tradi-
tionally used to fix the coracoid process to the glenoid,
and screw fixation has been validated through clinical
and biomechanical data.11 Despite stable fixation, the com-
plications of traditional Latarjet screw fixation remain an
issue among surgeons.3 Hardware complications range
The American Journal of Sports Medicine
2020;48(1):39–47
DOI: 10.1177/0363546519887959
Ó 2019 The Author(s)
from 6.5% to 46%. Therefore, some studies have advocated
the use of suture-button fixation during the Latarjet proce-
dure to reduce the aforementioned complications.4,10,25
This modification is considered to improve the accuracy
of bone location and avoid complications, with no reported
neurological or implant complications at a mean follow-up
period of 14 months. Meanwhile, a biomechanical study
showed that the screw and suture-button fixation techniques
exhibit comparable biomechanical strength for coracoid bone
block fixation during the Latarjet procedure.21 However, the
midterm efficacy of the suture-button fixation technique is
still unproven. The current study aimed to analyze the

39
40 Xu et al
midterm clinical effects of a modified suture-button fixation
technique during the Latarjet procedure and to investigate
remodeling of coracoid grafts after performing this ‘‘flexible’’
fixation technique. We hypothesized that our technique
would exhibit excellent midterm outcomes, the glenoid and
graft would fuse with each other, and finally, the new glenoid
would remodel to a new concentric circle with the humeral
head analogous to the original glenoid.
METHODS
This retrospective study was approved by the ethical com-
mittee of Shenzhen Second People’s Hospital, and all
patients gave informed consent before surgical procedures.
Patient Data
A total of 152 patients who underwent suture-button fixa-
tion during the Latarjet procedure between February 2013
and February 2016 were evaluated for inclusion in the
study. The inclusion criteria were as follows: (1) a glenoid
defect .20%, (2) a glenoid defect .15% with an Instability
Severity Index Score .6, (3) contact sport athletes with
a glenoid defect .10% but \15%, and (4) failure after
Bankart repair. Meanwhile, the exclusion criteria were
as follows: (1) epilepsy, (2) incomplete follow-up data, and
(3) loss to follow-up (Figure 1). The glenoid bone defect
was calculated from anterior to posterior using a bare area
method on the en face view of 3-dimensional (3D) computed
tomography (CT). The depth of the Hill-Sachs lesion was
measured directly on axial and coronal CT scans to deter-
mine the largest lesion size. The Instability Severity Index
Score was calculated in accordance with the literature.6,22
Operative Techniques
All surgical procedures were performed by the senior
author (W.L.). The surgical technique was modified pri-
marily on the basis of a previously described technique.3,27
The modifications were as follows: (1) the coracoid graft
and conjoint tendon were prepared using a mini-open tech-
nique with an incision of 2.5 cm. The incision began from
1 cm under the coracoid process in the direction of the
axilla. The coracoacromial ligament and part of the pector-
alis minor muscle were first cut 1 cm from the border of the
coracoid. With the help of an oscillating saw, osteotomy of
the coracoid process was performed at its bend, so that it
measured approximately 20 mm long. A total of 2 bone tun-
nels were drilled with a distance of 6 mm in the cut bone
block along its axis. High-strength sutures made of ultra
high–molecular weight polyethylene fiber (Ultrabraid No.
y
Address correspondence to Wei Lu, MD, PhD, Department of Sports Medicine, Shenzhen Second People’s Hospital, First Affiliated Hospital of
Shenzhen University, No. 3002 Sungang West Road, Futian District, Shenzhen, Guangdong 518000, China (email: winerl@sina.com).
*Department of Sports Medicine, Shenzhen Second People’s Hospital, First Affiliated Hospital of Shenzhen University, Shenzhen, China.
Submitted April 24, 2019; accepted September 20, 2019.
One or more of the authors has declared the following potential conflict of interest or source of funding: This study was funded by the Health and Family
Planning Commission of Shenzhen Municipality (No. SZBC2017022) and the Guangdong Science and Technology Department (No. 2018A030310646).
AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD
and disclaims any liability or responsibility relating thereto.
The American Journal of Sports Medicine
2 white suture assembly; Smith & Nephew) were pulled
into the distal tunnel. Moreover, 3 high-strength sutures
were pulled into the central hole of a suture button (Endo-
button; Smith & Nephew) and then pulled together to the
proximal bone tunnel (Figure 2). (2) The anterior (includ-
ing part of the incision used for obtaining the graft), stan-
dard antelateral, and posterior portals were set (Figure 2).
(3) The glenoid was marked at the 4-o’clock position, and
then the subscapularis muscle was split from the posterior
to anterior direction until the anterior fascia of the subscapu-
laris muscle became visible. A switch stick was used to pro-
tect the axillary nerve from damage. The muscle was split
with a 1.5 cm–diameter window (Figure 3). (4) The glenoid
tunnel was drilled where the suture linked to the graft was
passed, and the graft was pulled to the glenohumeral joint
via the sutures. A knotless suture anchor for antirotation
(PushLock; Arthrex) was fixed to the glenoid to prevent rota-
tion of the graft (Figure 3).
Rehabilitation Protocol
The patient’s arm was immobilized in internal rotation in
a sling for 6 weeks. Rehabilitation protocols were standard-
ized. Pendulum exercises were performed several times
daily, beginning on postoperative day 1. No active exercise
or workout with weights or pulleys was allowed until 6
weeks postoperatively. Active forward flexion and passive
external rotation were allowed at 6 weeks postoperatively,
and active movement in all directions was allowed at 3
months postoperatively, except for biceps tendon contrac-
tion exercises. After 3 months, active biceps tendon con-
traction exercises were initialized and gradually
increased. Contact sports or ‘‘at-risk’’ activities were not
allowed for 6 months.
Evaluation of Clinical Efficacy
A comprehensive patient history on the cause of the initial
dislocation and the number of dislocations was recorded
preoperatively. Preoperative and postoperative clinical
results were assessed using a visual analog scale (VAS)
for pain and instability. Active shoulder ranges of motion,
including forward flexion, external rotation at the side,
and external and internal rotation at 90° of abduction,
were assessed preoperatively and at the last follow-up.
The Rowe score, American Shoulder and Elbow Surgeons
(ASES) score, and Walch-Duplay score were used for the clin-
ical assessment. Meanwhile, complications that occurred
intraoperatively and postoperatively were recorded and the
occurrence rate was calculated.
AJSM Vol. 48, No. 1, 2020 Midterm Outcomes of the Modified Suture-Button Latarjet Procedure
Figure 1. Patient selection in this study.
Figure 2. Arthroscopic approach and graft preparation: (A)
arthroscopic approaches (green spots) and skin incision
(green line) and (B) graft preparation including Endobutton
and suture placement.
Radiological Assessment
Radiography and 3D CT were performed routinely for the
preoperative evaluation of the glenoid and humeral bone
defects. The patients were routinely required to undergo
a radiological examination every 6 months. The evaluation
consisted of the following: (1) The graft’s vertical position
on the en face view and its horizontal position on the trans-
verse view of the 3D reconstruction image of the glenoid
were obtained. (2) The 3D volume in graft osteolysis was
calculated to measure and observe volume absorption.12
Meanwhile, the absorption rate was analyzed. (3) Graft
union with the glenoid was assessed using the method of
Hovelius et al.13-15 (4) For graft and graft-glenoid interface
remodeling, the dynamic changes of the bone block over
time, particularly when some of the bone blocks were
higher than the glenoid in the transverse plane, were
41
recorded. The glenoid-graft combination was regarded as
the entire object to be observed over time. (5) For gleno-
humeral joint arthropathy, humeral head degeneration
was observed as described by Samilson and Prieto.24
Reliability of Measurements
The intraobserver correlation coefficient of repeated CT
measurements ranged from 0.865 to 0.980 for all the meas-
urements. The interobserver correlation coefficient of CT
measurements ranged from 0.875 to 0.970 for all the meas-
urements. The results showed good to excellent reliability,
demonstrating that all the measurements can be per-
formed with good repeatability.
Statistical Analysis
A paired Student t test was used to compare the surface
area of the preoperative glenoid with that of coracoid
transfer and the reconstructed glenoid. A paired t test
was also used to assess the differences between preopera-
tive and postoperative range of motion measurements
and shoulder scores. Repeated-measures analysis of vari-
ance was used to analyze graft volume and height over
the glenoid level. The SPSS Statistics software package
(version 20.0; IBM) was used for all the statistical analy-
ses, and P \ .05 was considered statistically significant.
RESULTS
Patient Data
The mean follow-up time was 40.3 6 5.8 months. During
the follow-up period, 102 patients were included in the
study, comprising 88 male and 14 female patients with
42 Xu et al The American Journal of Sports Medicine

Figure 3. Bankart failure case that demonstrated subscapularis muscle split, bone tunnel preparation, and coracoid graft fixation.
(A) The glenoid was marked at the 4-o’clock position. (B) The subscapularis muscle was split from the posterior to anterior direc-
tion. (C) The anterior fascia of the subscapularis muscle could be visualized. (D) The muscle was split with a 1.5 cm–diameter
window. (E) The glenoid tunnel was drilled where the suture linked to the graft was passed. (F) The graft was pulled to the gle-
nohumeral joint through the sutures. (G) A knotless suture anchor was fixed to the glenoid to prevent rotation of the graft. (H) The
graft was fixed to the glenoid. GN, glenoid; RF, radiofrequency; SS, subscapularis; HH, humeral head.

85 left and 17 right shoulders. The mean age was 24.8 6 TABLE 1
4.8 years (range, 18-36 years). Furthermore, 59 patients Patient Dataa
had a bone defect of .20%, 25 had a bone defect of 15% to
20% and an Instability Severity Index Score .6, 12 had Parameter Value
a bone defect of 10% to 15% and played contact sports, and Age, mean 6 SD (range), y 24.8 6 4.8 (18-36)
6 had failed Bankart repair. The mean defect area of the Sex, male/female, n 88/14
shoulder glenoid in all the patients was 19.3% 6 5.7%, and Side, left/right, n 85/17
the mean number of dislocations were 8.8 6 5.3 (range, 4- Number of dislocations 8.8 6 5.3
30). In addition, 99 patients had a Hill-Sachs lesion to differ- Body mass index, kg/m2 24.3 6 5.1
ing extents. General information is provided in Table 1. Glenoid defect area, % 19.3 6 5.7
.20%, n 59
15%-20%, n 25
10%-14%, n 12
Subjective Pain Bankart failure, n 6
Hill-Sachs injury, n 99
The VAS scores for pain during motion decreased from
a mean of 3.3 6 1.2 (range, 0-6) preoperatively to 1.1 6 a
Data are reported as mean 6 SD unless otherwise indicated.
0.8 (range, 0-3) at the last follow-up (P \ .001). The
improvements in pain during motion were statistically sig-
nificant (Table 2). 172.1° 6 15.6°, 122.8° 6 17.1°, 73.0° 6 11.2°, 75.4° 6
16.1°, and 63.7° 6 13.2°, respectively. No significant differ-
ence was detected preoperatively and postoperatively (P .
Range of Motion .05) (Table 2).

The mean preoperative forward flexion, abduction, exter-

nal rotation at the side, external rotation at 90° of abduc-
tion, and internal rotation at 90° of abduction were
175.2° 6 17.8°, 125.2° 6 15.3°, 77.8° 6 15.3°, 78.3° 6
16.5°, and 65.6° 6 12.3°, respectively. Postoperatively,
the mean forward flexion, abduction, external rotation at
the side, external rotation at 90° of abduction, and internal
rotation at 90° of abduction at the last follow-up were
Functional Recovery
At the last follow-up, all patients had resumed normal life
without dislocations and were satisfied with the outcome.
Among these patients, 87 could resume preoperative or
intensive exercises. The ASES, Rowe, and Walch-Duplay
scores of the patients increased from 80.2 6 16.2, 40.2 6
AJSM Vol. 48, No. 1, 2020 Midterm Outcomes of the Modified Suture-Button Latarjet Procedure 43

TABLE 2
Functional Results Preoperatively and at Final Follow-upa

Parameter Preoperative Final Follow-up P Value

VAS for pain (during motion), mean 6 SD (range) 3.3 6 1.2 (0-6) 1.1 6 0.8 (0-3) \.001
Range of motion, deg
Forward flexion 175.2 6 17.8 172.1 6 15.6 .317
Abduction 125.2 6 15.3 122.8 6 17.1 .651
External rotation at the side 77.8 6 15.3 73.0 6 11.2 .231
External rotation at 90° of abduction 78.3 6 16.5 75.4 6 16.1 .343
Internal rotation at 90° of abduction 65.6 6 12.3 63.7 6 13.2 .375
ASES score 80.2 6 16.2 95.2 6 5.6 \.001
Rowe score 40.2 6 9.8 94.5 6 2.7 \.001
Walch-Duplay score 67.5 6 10.2 95.6 6 3.2 \.001

a
Data are reported as mean 6 SD unless otherwise indicated. Bold values are statistically significant. ASES, American Shoulder and
Elbow Surgeons; VAS, visual analog scale.

9.8, and 67.5 6 10.2 preoperatively to 95.2 6 5.6, 94.5 6 2.7, TABLE 3
and 95.6 6 3.2 postoperatively, respectively. The difference Graft Position in En Face and Axial Views
was statistically significant (P \ .05) (Table 2).
Location n (%)

Complications
No postoperative infection, axillary nerve injury, or vascu-
lar injury occurred in any patient. However, one 42-year-
old female patient experienced a redislocation due to
a fall that was manually reduced. She was able to resume
her normal daily activities. Moreover, 1 patient experi-
enced stiffness of the shoulder joint, and forward flexion
and abduction were less than 120° and 90°, respectively.
After physical therapy, the patient’s condition improved
considerably. The overall complication rate was 1.9%.
Imaging Assessment
Graft Position. A vertical position (between the 3 and 5
o’clock position) was achieved in 98 grafts (96% of all cases)
immediately postoperatively, with the mean graft midline
center at the 4 o’clock position. However, 2 grafts were
positioned superiorly, whereas another 2 grafts were posi-
tioned inferiorly (Table 3). In the axial view, CT scans
showed that 89 grafts were flush to the glenoid, and 2
and 11 grafts were fixed medially and laterally, respec-
tively (Figure 4).
Graft Healing and Absorption. A total of 100 grafts
achieved bone union; meanwhile, 2 grafts experienced
fibrous union. The graft volume initially decreased and
then increased. On average, the postoperative 6-month,
1-year, 2-year, and 3-year graft volumes changed to
87.3% 6 7.1%, 88.6% 6 5.2%, 93.2% 6 6.4%, and 92.7%
6 6.9%, respectively, compared with the volumes immedi-
ately postoperatively. Statistically significant differences
were found between the postoperative 6-month and 1-
year graft volumes and the postoperative 2-year and 3-
year graft volumes (P \ .05). However, the differences
between the postoperative 6-month and 1-year volumes
(ie, immediately postoperatively) and the postoperative 2-
year and 3-year volumes were not significant (P . .05).
En face view
3-5 o’clock 98 (96.1)
Superior to 2:30 2 (1.9)
Inferior to 4:20 2 (1.9)
Axial view
Flush 89 (87.3)
Medial 2 (1.9)
Lateral 11 (10.7)
All grafts were absorbed to a certain extent. The grafts
manifested osteolysis on the outer side of the ‘‘best-fit’’ cir-
cle of the glenoid. On average, osteolysis did not within the
first year after surgery. The graft volume stabilized 2 years
postoperatively (Figure 5).
Remodeling of Graft and Graft-Glenoid Interface. All
grafts remodeled to a steady state within 2 years. In the
en face view, the grafts exhibited growth superiorly, inferi-
orly, medially, and laterally. The interface gap between the
glenoid and graft was nearly filled with bone. The glenoid
and graft fused with each other and finally remodeled
analogous to the shape of the intact glenoid. The graft
and glenoid eventually formed congruent concavity with
the ipsilateral humeral head. In the axial view, among
the 89 grafts that were flush to the glenoid immediately
postoperatively, 37 grafts finally remodeled remaining
flush to the glenoid, and 52 grafts remodeled to a curved
shape congruent to the humeral head with a higher outer
edge (Figure 6). In addition, 2 grafts positioned medially
immediately postoperatively were found to be filled with
bone vertically and horizontally; however, they could not
be flush to the glenoid.
Moreover, 11 grafts positioned laterally evidently
remodeled. The mean height of the graft above the glenoid
level exhibited a wave-curved change. The mean height
immediately postoperatively was 5.8 6 3.8 mm. Then, it
decreased to 4.6 6 3.2 mm at 6 months postoperatively.
The height remained stable and slightly increased to 4.8
6 2.7, 5.2 6 2.6, and 5.1 6 2.3 mm at 1, 2, and 3 years
44 Xu et al The American Journal of Sports Medicine

Figure 4. Graft position in en face and axial views. (A) En face view showing the graft in the normal position. (B) En face view
showing the graft positioned superiorly. (C) En face view showing the graft positioned inferiorly. (D) Axial view showing that
the graft was flush to the glenoid. (E) Axial view showing that the graft was fixed too medially. (F) Axial view showing that the graft
was fixed too laterally.

year postoperatively and at 6 months, 2 years, and 3 years

postoperatively. The height at the interface of the glenoid
and graft decreased, but the height of the anterolateral
edge of the graft did not decrease and even increased in
9 cases, thereby forming an arc with the same circle as
the humeral head (Figure 8).
Glenohumeral Joint Arthropathy. In accordance with
the grade of Samilson and Prieto,24 all the patients showed
no degenerative changes, even the laterally positioned
grafts viewed on radiography and CT at the final follow-up.

DISCUSSION

The most important findings of this study are as follows:

Figure 5. Graft volume changed over the years at follow-up.
postoperatively, respectively (Figure 7). The height at 6
months postoperatively significantly differed from that
immediately postoperatively and at 2 and 3 years postoper-
atively (P \ .05). No significant difference was found
between the height immediately postoperatively and at 2
and 3 years postoperatively and between the height at 1
(1) Bone union was achieved at 6 months postoperatively
using the suture-button technique. (2) Absorption of the
graft mostly occurred on the edge and outside the ‘‘best-
fit’’ circle of the glenoid. The ‘‘new’’ glenoid tended to
form a pear shape on the en face view and remained stable
for 1 year. (3) The graft exhibited growth superiorly, infe-
riorly, medially, and laterally. The glenoid and graft fused
with each other, and finally, the new glenoid remodeled to
a new concentric circle with the humeral head analogous to
the original glenoid. Grafts positioned laterally did not
cause arthropathy of the joints at midterm.
AJSM Vol. 48, No. 1, 2020 Midterm Outcomes of the Modified Suture-Button Latarjet Procedure 45

Figure 6. Remodeling of graft and graft-glenoid interface. (A-D) En face view: From (A) immediately postoperatively to (B) 3 years
postoperatively, the glenoid and graft were fused and remodeled analogous to the shape of the intact glenoid. (C, D) Grafts also
showed that vertically remodeled, absorption occurred outside of the ‘‘best-fit’’ circle. (E-H) Axial view: From immediately post-
operatively to 3 years postoperatively, the glenoid and graft were remodeled to a congruent concavity with the ipsilateral humeral
head.

Figure 7. Mean height of the graft above the glenoid level
changed over the years.
In 2010, Boileau et al2 performed a suture-button
arthroscopic Latarjet technique, which not only made the
location of the graft position more accurate and achieved
a high healing rate but also avoided the complications of
screw fixation. Such complications have concerned sur-
geons for a long time.7-9,26 Our technique exhibits charac-
teristics similar to those of Boileau et al.1-3 In the
present study, we observed and analyzed the longer term
outcomes and remodeling of the suture-button Latarjet
technique using a relatively large sample.
Technique Modification
Our modification is based on the method proposed by Boil-
eau et al,2 who used the technique of a coracoid graft
obtained arthroscopically and generally required 6 to 7
approaches. Obtaining a coracoid graft arthroscopically is
actually the most time-consuming step; thus, we used the
method introduced by Young et al.28 Only a small incision
(2.5 cm) was required. In general, this step can be com-
pleted within 15 minutes after skillful training. Mean-
while, the graft is thoroughly decorticated by use of an
oscillating saw. This step may increase the healing rate
between the glenoid and graft. In addition, all the sutures
in the graft were placed into the cannula through an ante-
rior approach and in front of the subscapularis tendon,
thereby avoiding soft tissue embedding caused by pulling
the coracoid graft from anterior to posterior and saving
operative time. Meanwhile, the subscapularis tendon that
was split with a width of only 1.5 cm from the posterior
to anterior side was adopted, and a switching stick for axil-
lary nerve protection was used. This procedure may protect
the patient from experiencing axillary nerve injuries. The
overall excellent outcome in our case series proved the
reproducibility and accuracy of positioning the grafts by
the suture-button technique.
Graft Osteolysis and Remodeling
Similar to other studies,18 bone resorption occurred at dif-
ferent degrees in the current study’s group of patients.
However, we observed that bone resorption mostly
46 Xu et al The American Journal of Sports Medicine

Figure 8. Remodeling of the graft and graft-glenoid interface when positioned laterally. (A-D) En face view: (A) Immediately post-
operatively, (B) 1 year postoperatively, (C) 2 years postoperatively, and (D) 3 years postoperatively, the glenoid and graft fused
and remodeled to the shape of the intact glenoid without causing arthropathy. (E-H) Axial view: (E) The graft was fixed laterally,
(F) the height at the glenoid-graft interface decreased gradually at 1 year postoperatively, and the height of the anterolateral edge
of the graft did not decrease and even increased at (G) 2 years and (H) 3 years postoperatively, forming an arc with the same circle
as the humeral head.

occurred on the edge and outside of the ‘‘best-fit’’ circle of
the glenoid. Increased bone resorption may be contributed
to excessive biceps tendon activities within 6 months post-
operatively. Our observation showed that graft osteolysis
typically occurred at 3 months postoperatively. Active
biceps tendon contraction might affect healing of the graft
by placing tension on the graft and inducing osteolysis.
Thus, no biceps tendon active exercise was recommended
within this period. We also found that grafts united with
the glenoid at a mean of 6 months postoperatively. The
grafts remained stable without osteolysis at 1 year postop-
eratively, and then they gradually grew upward and down-
ward. Finally, they filled the gap and formed an arc-shaped
glenoid with the same circle as the original glenoid at 2
years postoperatively. This type of healing and remodeling
may be attributed to elastic fixation via the suture-button
technique.
The suture-button technique provides flexible fixation
while exhibiting biomechanical strength comparable to
screws, as proven by this study. Minimal displacement
might occur when grafts experience impingement pro-
duced by the humeral head. Considering the Wolff law,
bone proliferation will occur in an area with a rich blood
supply, which is possibly the reason why grafts that are
positioned laterally achieve more bone regeneration
because of the compressive stress stimulus. The area with-
out a stimulating compressive force would absorb to
a certain extent, which may explain why the upper part
of the graft absorbs more, whereas the part located in the
concentric circle of the humeral head absorbs less.
In contrast with suture buttons, screws achieve complete
tight fixation, which does not allow graft displacement.
Therefore, arthropathy might be easily produced by
impingement of the humeral head. No arthropathy was
observed in our study. The primary reason may be minimal
displacement of the graft, which produced no apparent com-
pression force on the humeral head, particularly in the lat-
erally positioned grafts. In addition, the grafts remodeled to
the final shape at a mean of 2 years, and all the patients had
no humeral head degenerative changes, thereby suggesting
that the suture-button technique is an effective method for
preventing joint arthropathy.
Bankart Repair
Bankart suture repair was not performed in the current
study’s group of patients. The literature19 indicates that
the Latarjet procedure combined with Bankart suture
repair will limit the shoulder’s external rotation, which is
inconsistent with the rehabilitation protocol for the early
recovery of external rotation. A total of 5 patients under-
went second-look surgery because of patients’ demand for
hardware removal, and the capsule was found completely
attached to the glenoid.
AJSM Vol. 48, No. 1, 2020 Midterm Outcomes of the Modified Suture-Button Latarjet Procedure
Limitations
Despite the merits shown in our study, it still has some
limitations. First, our technique was not compared with
other techniques, such as the open technique. Second,
our rehabilitation protocol is conservative. Other progres-
sive protocols may be developed. Third, more prospective
research should be performed to provide additional evi-
dence. Fourth, a longer follow-up is needed to determine
if degenerative changes develop.
CONCLUSION
The outcome was satisfactory after the suture-button Latar-
jet technique. Absorption of the graft mostly occurred on the
edge and outside the ‘‘best-fit’’ circle of the glenoid. Finally,
the ‘‘new’’ glenoid tended to form a pear shape on the en face
view at a mean of 2 years postoperatively. The graft
exhibited growth superiorly, inferiorly, medially, and later-
ally. Last, the glenoid and graft fused with each other, and
the new glenoid remodeled to a new concentric circle with
the humeral head analogous to the original glenoid. Grafts
positioned laterally caused no arthropathy of the joints at
the early follow-up. Considering that arthropathy may not
be detectable for many years after the procedure, long-
term detection should be conducted in the future.
REFERENCES
1. Boileau P, Gendre P, Baba M, et al. A guided surgical approach and
novel fixation method for arthroscopic Latarjet. J Shoulder Elbow
Surg. 2016;25(1):78-89.
2. Boileau P, Mercier N, Roussanne Y, Thelu CE, Old J. Arthroscopic
Bankart-Bristow-Latarjet procedure: the development and early
results of a safe and reproducible technique. Arthroscopy. 2010;
26(11):1434-1450.
3. Boileau P, Saliken D, Gendre P, et al. Arthroscopic Latarjet: suture-
button fixation is a safe and reliable alternative to screw fixation.
Arthroscopy. 2019;35(4):1050-1061.
4. Bonnevialle N, Thelu CE, Bouju Y, et al. Arthroscopic Latarjet proce-
dure with double-button fixation: short-term complications and learn-
ing curve analysis. J Shoulder Elbow Surg. 2018;27(6):e189-e195.
5. Cerciello S, Corona K, Morris BJ, Santagada DA, Maccauro G. Early
outcomes and perioperative complications of the arthroscopic Latar-
jet procedure: systematic review and meta-analysis. Am J Sports
Med. 2019;47(9):2232-2241.
6. Cunningham G, Benchouk S, Kherad O, Ladermann A. Comparison
of arthroscopic and open Latarjet with a learning curve analysis.
Knee Surg Sports Traumatol Arthrosc. 2016;24(2):540-545.
7. Domos P, Lunini E, Walch G. Contraindications and complications of
the Latarjet procedure. Shoulder Elbow. 2018;10(1):15-24.
8. du Plessis JP, Dachs RP, Vrettos BC, et al. The modified Latarjet pro-
cedure in female patients: clinical outcomes and complications.
J Shoulder Elbow Surg. 2018;27(1):e9-e15.
9. Frank RM, Gregory B, O’Brien M, et al. Ninety-day complications fol-
lowing the Latarjet procedure. J Shoulder Elbow Surg. 2019;28(1):88-
94.
For reprints and permission queries, please visit SAGE’s Web site at http://www.sagepub.com/journalsPermissions.nav.
47
10. Gendre P, Thélu CE, d’Ollonne T, Trojani C, Gonzalez JF, Boileau P.
Coracoid bone block fixation with cortical buttons: an alternative to
screw fixation? Orthop Traumatol Surg Res. 2016;102(8):983-987.
11. Giacomo GD, Costantini A, de Gasperis N, et al. Coracoid bone graft
osteolysis after Latarjet procedure: a comparison study between two
screws standard technique vs mini-plate fixation. Int J Shoulder
Surg. 2013;7(1):1-6.
12. Haeni DL, Opsomer G, Sood A, et al. Three-dimensional volume
measurement of coracoid graft osteolysis after arthroscopic Latarjet
procedure. J Shoulder Elbow Surg. 2017;26(3):484-489.
13. Hovelius L, Sandstrom B, Olofsson A, Svensson O, Rahme H. The
effect of capsular repair, bone block healing, and position on the
results of the Bristow-Latarjet procedure (study III): long-term follow-
up in 319 shoulders. J Shoulder Elbow Surg. 2012;21(5):647-660.
14. Hovelius L, Sandstrom B, Saebo M. One hundred eighteen Bristow-
Latarjet repairs for recurrent anterior dislocation of the shoulder pro-
spectively followed for fifteen years, study II: the evolution of disloca-
tion arthropathy. J Shoulder Elbow Surg. 2006;15(3):279-289.
15. Hovelius L, Sandstrom B, Sundgren K, Saebo M. One hundred eigh-
teen Bristow-Latarjet repairs for recurrent anterior dislocation of the
shoulder prospectively followed for fifteen years, study I: clinical
results. J Shoulder Elbow Surg. 2004;13(5):509-516.
16. Hurley ET, Lim FD, Farrington SK, Mullett H. Open versus arthroscopic
Latarjet procedure for anterior shoulder instability: a systematic review
and meta-analysis. Am J Sports Med. 2019;47(5):1248-1253.
17. Jiang C. Arthroscopic versus open Latarjet in the treatment of recurrent
anterior shoulder dislocation with marked glenoid bone loss: a prospec-
tive comparative study. Response. Am J Sports Med. 2018;46(5):P10.
18. Kee YM, Kim JY, Kim HJ, Sinha S, Rhee YG. Fate of coracoid grafts after
the Latarjet procedure: will be analogous to the original glenoid by
remodelling. Knee Surg Sports Traumatol Arthrosc. 2018;26(3):926-932.
19. Kleiner MT, Payne WB, McGarry MH, Tibone JE, Lee TQ. Biome-
chanical comparison of the Latarjet procedure with and without cap-
sular repair. Clin Orthop Surg. 2016;8(1):84-91.
20. Kordasiewicz B, Kicinski M, Malachowski K, Wieczorek J, Chaberek
S, Pomianowski S. Comparative study of open and arthroscopic cor-
acoid transfer for shoulder anterior instability (Latarjet): computed
tomography evaluation at a short term follow-up, part II. Int Orthop.
2018;42(5):1119-1128.
21. Provencher MT, Aman ZS, LaPrade CM, et al. Biomechanical com-
parison of screw fixation versus a cortical button and self-tensioning
suture for the Latarjet procedure. Orthop J Sports Med.
2018;6(6):2325967118777842.
22. Rouleau DM, Hebert-Davies J, Djahangiri A, Godbout V, Pelet S, Balg
F. Validation of the instability shoulder index score in a multicenter
reliability study in 114 consecutive cases. Am J Sports Med.
2013;41(2):278-282.
23. Ryliskis S, Siomin K. Arthroscopic versus open Latarjet in the treat-
ment of recurrent anterior shoulder dislocation with marked glenoid
bone loss: a prospective comparative study. Letter to the editor.
Am J Sports Med. 2018;46(5):P9-P10.
24. Samilson RL, Prieto V. Dislocation arthropathy of the shoulder. J
Bone Joint Surg Am. 1983;65(4):456-460.
25. Tytherleigh-Strong GM, Morrissey DI. Failed Latarjet procedure trea-
ted with a revision bone block stabilization using a suture-button fix-
ation. J Shoulder Elbow Surg. 2017;26(4):e102-e107.
26. Williams H, Evans JP, Furness ND, Smith CD. It’s not all about redis-
location: a systematic review of complications after anterior shoulder
stabilization surgery. Am J Sports Med. 2019;47(13):3277-3283.
27. Xu J, Liu H, Lu W, et al. Clinical outcomes and radiologic assessment
of a modified suture button arthroscopic Latarjet procedure. BMC
Musculoskelet Disord. 2019;20(1):173.
28. Young AA, Maia R, Berhouet J, Walch G. Open Latarjet procedure for
management of bone loss in anterior instability of the glenohumeral
joint. J Shoulder Elbow Surg. 2011;20(2)(suppl):S61-S69.