Modified Suture-Button Latarjet Procedure With

Coracoacromial Ligament and Pectoralis Minor

Preservation Achieves Good Clinical Outcomes at
2-Year Follow-Up: Case Series of Latarjet Technique
Wei Lu, M.D., Ph.D., Daqiang Liang, M.D., Yan Liu, M.D., Bing Wu, M.D., Ph.D.,
Zeling Long, M.D., Ph.D., Jingyue Su, M.D., Haifeng Liu, M.D., Ph.D., Xun Luo, M.D., Ph.D.,
and Zhenhan Deng, M.D., Ph.D.

Purpose: To evaluate whether the modified suture-button Latarjet procedure with coracoacromial ligament (CAL) and
pectoralis minor (PM) preservation could achieve excellent outcomes at the 2-year follow-up. Methods: During January
2019 to January 2021, the data of patients who underwent modified suture-button Latarjet with CAL and PM preser-
vation in our department were collected. The glenoid bone loss of these patients was greater than 20% or greater than
10% with high demands for exercise. Partial coracoid osteotomy was based on the results of a preoperative 3-dimensional
computed tomography evaluation of the glenoid defect area and corresponding coracoid process morphology. The pre-
operative and postoperative clinical results were assessed. The minimal clinically important difference (MCID) was used to
compare improvement in clinical outcomes. Graft-glenoid union and remodeling were assessed using postoperative 3-
dimensional computed tomography, and magnetic resonance imaging was performed to confirm the integrity of the
CAL and PM postoperatively. Results: In total, 35 patients were included in this study; the mean follow-up time was 26.9

1.9 months. No case of recurrent dislocation or sublaxity. Significant improvements were observed in mean visual
analog scale (VAS) scores for pain during motion, American Shoulder and Elbow Surgeons (ASES) score, Rowe score, and
Walch-Duplay score (P < .001). The percentage of patients achieving at least an MCID improvement in clinical outcomes
was VAS 85.71%, ASES 97.14%, Rowe 100%, and Walch-Duplay 97.14%. Thirty-three patients (94.3% of all cases) were
able to return to their preoperative sport levels, 34 grafts (97.1%) achieved bone union (1 soft union) in 6.3
2.2 months,
and the coracoid grafts restored 97.1
4.0% of the perfect-fitting circle at the last follow-up. Postoperative computed
tomography scan showed that 31 grafts (88.6%) were placed ideally in vertical view. In the axial view, 25 grafts (82.9%)
were flushed to the glenoid, whereas 1 and 5 grafts were fixed medially and laterally, respectively. The CAL and PM were
visualized postoperatively. No arthropathy was observed in any patient at the last follow-up. Conclusions: The modified
suture-button Latarjet procedure with CAL and PM preservation obtained good clinical and radiological results without
recurrence or complications. A substantial number of patients (>85%) achieved the MCID for the VAS, ASES, Rowe, and
Walch-Duplay scores. In addition, the malpositioned graft (17.1%) did not cause arthropathy of the joints at 2-year
follow-up. Level of Evidence: Level IV, retrospective case series.

See commentary on page 29

I n young patients with recurrent anterior shoulder

instability (RASI) and concomitant glenoid bone loss
(GBL), isolated soft-tissue repair results in a high
coracoid process (CP) and conjoint tendon transfer
Latarjet operation for the treatment of RASI, which
Walch later standardized. Because of its low recurrence
recurrence rate.1 In 1954, Latarjet first proposed the rate, high rate of satisfaction, and high rate of return to

From the Departments of Sports Medicine (W.L., D.L., B.W., Z.L., J.S., Address correspondence to Zhenhan Deng, Department of Orthopaedic
H.L., X.L.), Infection and Critical Care Medicine, and Nosocomial Infection Surgery, The First Affiliated Hospital of Wenzhou Medical University,
Prevention and Control (Y.L.), and Nosocomial Infection Prevention and Wenzhou, Zhejiang, 325000, China. E-mail: dengzhenhan@[Link]
Control (Y.L.), The First Affiliated Hospital of Shenzhen University, Shenzhen Ó 2024 The Author(s). Published by Elsevier Inc. on behalf of the
Second People’s Hospital, Shenzhen, Guangdong, China; and Department of Arthroscopy Association of North America. This is an open access article under
Orthopaedic Surgery, The First Affiliated Hospital of Wenzhou Medical the CC BY-NC-ND license ([Link]
University, Wenzhou, Zhejiang, China (Z.D.). 0749-8063/231709
Received December 6, 2023; accepted April 14, 2024. [Link]

Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 41, No 1 (January), 2025: pp 19-28 19
20 W. LU ET AL.

sports, the Latarjet procedure has been widely adopted Since 2013, on the basis of more than 8 years of
by surgeons globally.2 In 2007, the French surgeon experience in clinical and basic research on suture-
Laffosse successfully completed the Latarjet procedure button Latarjet,11-14 a novel modified suture-button
arthroscopically, which became a milestone from open arthroscopic Latarjet was proposed in 2021, including
to mini-invasive surgery.3 However, despite stable fix- partial coracoid osteotomy, mini-window splitting of
ation, the high rate of hardware complications during the subscapularis (SSC) (8-10 mm), and CAL and PM
screw-fixation Latarjet procedures remains an issue preservation, which resulted in a perfect-fitting circle
among surgeons.4 In 2012, Boileau et al.5 improved the (PFC) of glenoid restoration and has been discussed in
technique by fixing the coracoid graft (CG) with suture detail in a previous publication.15 This technique is
buttons to avoid screw-related complications, which further developed to “5P” key elements: partial coracoid
achieved better clinical and imaging results. These de- osteotomy, preservation of CAL, PM reconstruction,
cades of continuous refinement have led to the ongoing porthole SSC muscle split, and PFC glenoid recon-
optimization of Latarjet surgery. struction. The purpose of this study is to evaluate
However, surgeons are increasingly concerned about whether a modified suture-button Latarjet procedure
the loss of function in the coracoacromial ligament with CAL and PM preservation could achieve excellent
(CAL) and pectoralis minor (PM), which are usually outcomes at 2-year follow-up. We hypothesized that it
detached in traditional Latarjet procedures. A previous would also exhibit good clinical and radiographic out-
biomechanical study had proven that the CAL con- comes under the premise of CAL and PM preservation
tributes to glenohumeral stability, the loss of which and small split window of the SSC.
may result in humeral head superior migration.6 CAL
resection induces an increased glenohumeral trans- Methods
lation, especially in those with massive rotator cuff This study was approved by the Shenzhen Second
tears.7 In addition, as proposed by Beckers et al.,8 the People’s Hospital (no. 20210722001-FS02), and all pa-
Latarjet procedure alters scapulothoracic motion by tients provided informed consent before surgery.
releasing the PM insertion to the CP and by changing
the vector of the conjoint tendon. Scapular dyskinesis Patient Data
was found in 5 of 20 patients who underwent Latarjet Forty patients who underwent the modified suture-
procedure, and detachment of the PM might be the button Latarjet procedure with CAL preservation and
main reason that lead to dyskinesis.9 A computed to- PM reconstruction between January 2019 and January
mography (CT) scan study found that the symmetry of 2021 were included in this study. The GBL measure-
scapular position was lost initially after the Latarjet ments were based on preoperative CT scans. The
procedure, with a decrease of the alpha angle and Instability Severity Index Score was calculated in
scapular protraction.10 Therefore, preservation of these accordance with the literature.16
structures are important to achieve satisfied outcomes The inclusion criteria were as follows: (1) RASI with
and avoid potential complications. Some attempts to GBL >20%; (2) RASI with 15% < GBL < 20%, and
salvage these tissues,6,9 but there are no reports of Instability Severity Index Score score >6; (3) RASI with
further research outcomes. 10% < GBL < 15%, and competitive sports level; and

Fig 1. Schematic diagram of traditional Latarjet and modified suture-button Latarjet procedures. (A) The CAL and PM are
dissected after coracoid osteotomy in traditional Latarjet procedure. (B) In the modified suture-button Latarjet procedure, the
CAL is intact during the coracoid osteotomy, and the PM is sutured to the CP remains after that. (C) The triangular CG passes
very easily through the porthole SSC split window. (CAL, coracoacromial ligament; CG, coracoid graft; CP, coracoid process; GL,
glenoid; PM, pectoralis minor; RF, radiofrequency; SSC, subscapularis.)
 SUTURE-BUTTON LATARJET WITH 2-YEAR FOLLOW-UP 21

(4) revision surgery. The exclusion criteria were: (1) (1) A straight incision 25 mm inferior to the tip of the
epilepsy, (2) inability to complete follow-up or incom- CP was made. The deltoid muscle was separated
plete follow-up data, and (3) previous shoulder surgery, and then the CP was exposed. Once insertion of
except for Bankart repair. the CAL and PM on the CP was achieved, the PM
fibers were separated from the bone inward on
Preoperative Design the medial side of the CP by radiofrequency while
The principles of this new technique and the tradi- the CAL insertion was kept intact. Partial coracoid
tional Latarjet procedure are illustrated in Figure 1. osteotomywas performed according to the pre-
Three-dimensional computed tomography (3D CT) operative design. Then, the PM was sutured to the
reconstruction was performed preoperatively to outline distal end of the CP remains, which is equivalent
the glenoid and CP of the bilateral shoulder joints. to the CAL’ position. The tension of PM is small
According to the CT PICO (patient/population, inter- and the space under PM is large to avoid possible
vention, comparison and outcomes) method,17 the PFC compression by the brachial plexus-vascular
of the healthy side should be determined from the en sheath below.
face view and mirrored to the affected side. The pro- (2) The center fixation tunnel and side antirotation
portion of the GBL and the size and shape of the gle- tunnel on the CG were drilled and then high-
noid defect area (GDA) (the yellow arc area in Fig 2B) strength sutures and button were installed for
on the affected side should be measured based on the later use. Before the incision was closed, which
glenoid of the healthy side. The CP’s length, width, and can be the anterior portal, the PM was sutured
thickness were measured on 2-dimensional CT images, onto the remaining CP.
and a preoperative design of the partial coracoid (3) The posterior and anterolateral portals were
osteotomy level was made in each case.15 The CG was created, the shoulder joint was explored under
designed from the CP based on the size and shape of the arthroscopy, the glenoid tunnel was drilled ac-
GDA. Reference data for CG size (WL index): length cording to the preoperative design, and guidewire
(L) ¼ glenoid defect side length; width (W) ¼ the length was threaded.
between the apex of PFC and glenoid defect side; (4) An 8- to 10-mm window was split from the back
thickness of at least 6 mm (usually 2 mm larger than to the front at the lower two-thirds of the SSC
the real glenoid defect width, given possible bone lateral to the glenoid surface.
resorption) (Fig 2). (5) Under the guidance of the wire, the high-strength
sutures were passed on the CG through the split
Operative Technique and Rehabilitation Protocol of the SSC and pulled out from behind through
All the surgical procedures were performed by a se- the glenoid tunnel.
nior physician (W.L.). Detailed technical notes were (6) The high-strength wire was pulled at the back to
introduced in previous publications.15 The process make the CG adhere to the glenoid, and another
included the following steps:

Fig 2. Preoperative GDA and PFC measurement of the glenoid (patient/population, intervention, comparison, and outcome
[pico] method) and partial coracoid osteotomyelevel design of the modified suture-button Latarjet procedure. (A) Glenoid PFC
of the healthy side on the en face view. (B) GDA and WL index measurement. The blue line indicates the needed length (L) and
width (W) of the CG (WL index), and the yellow area indicates the GDA. (C) Partial coracoid osteotomy level design. Green lines
indicate the measurement of the length, width, and thickness of the CP. (D) Simulation of the CG matched to the GDA, and the
PFC is restored. (CG, coracoid graft; CP, coracoid process; GDA, glenoid defect area; PFC, perfect-fitting circle.)
22 W. LU ET AL.

button was inserted at the back and tie it for Table 1. Standard for Judging the Vertical Position of the
fixation. Graft
(7) The antirotation anchor was fixed on the glenoid Grade Graft’s Vertical Position
using a Pushlock anchor. Excellent Graft covers over 75% of the GDA
We name this technique as “LUtarjet technique” after Good Graft covers over 50%-75% of the GDA
its inventor, Dr. Wei Lu.15 The entire surgical process was Fair Graft covers 25%-50% of the GDA
Poor Graft covers less than 25% of the GDA
completed within an average of 60 minutes. Standard-
ized rehabilitation protocols were applied in accordance GDA, glenoid defect area.
with previous cases.14 The contact area between the CG
and the glenoid defect side is smaller than that in tradi-
tional coracoid osteotomy because partial coracoid absorption, remodeling, and graft-glenoid interface heal-
osteotomy is performed using the modified suture- ing. Magnetic resonance imaging (MRI) was also per-
button Latarjet technique. Therefore, our rehabilitation formed at POD 0 and the last follow-up to identify the
protocol was conservative. The arm was immobilized for integrity of the CAL and PM from the coronal view.
6 weeks in a neutral rotation sling. After 6 weeks, the The vertical position of the CG in the en face view and
sling was removed, and a return to everyday life was its horizontal position in the transverse view of the 3D
encouraged. No heavy repeated lifting was allowed for reconstructed image of the glenoid were obtained. On
the until postoperative month (POM) 6. the basis of the coverage of the CG on the GDA, we
developed a standard for judging the vertical position of
Evaluation of Clinical Efficacy the graft: Excellent: graft covers more than 75% of the
The patients’ history of the cause of the initial dislo- GDA; good: graft covers 50% to 75% of the GDA; fair:
cation and the number of dislocations were compre- graft covers 25% to 50% of the GDA; and poor: graft
hensively recorded after admission. Generalized laxity covers less than 25% of the GDA (Fig 3, Table 1).
was measured using the Beighton score.18 Preoperative The graft was considered too lateral if it went beyond
and postoperative clinical results were assessed using a the glenoid rim by more than 3 mm and it was judged to
visual analog scale (VAS) for pain. The American be medial if it was medial to the rim by more than 5
Shoulder and Elbow Surgeons (ASES), Rowe, and mm.19 Graft healing was assessed using imaging studies
Walch-Duplay scores were used for clinical assessment. performed at 2 years postoperatively. Healing of the graft
Complications that occurred intraoperatively and post- was graded as bony union, fibrotic union (radiolucent
operatively were recorded, and the rate of return to zone <5 mm), or migration (radiolucent zone >5
play, return to previous performance, and recurrence mm).17 Humeral head degeneration was evaluated using
rate were calculated during follow-up. the standard as described by Samilson and Prieto with 4
groups: (1) normal, (2) mild (osteophytes <3 mm on the
Radiologic Assessment humeral head), (3) moderate (osteophytes between 3
Radiography and 3D CT of both shoulders were per- and 7 mm), and (4) severe (osteophytes >7 mm).20
formed preoperatively to evaluate each patient’s GBL and
humeral bone defects. On the first postoperative day (POD Statistical Analysis
0) and at POMs 6, 12, and 24, a CT scan of the affected All the quantitative data were expressed as mean

shoulder was conducted to observe the CG’s position, standard deviation. The minimal clinically important

Fig 3. CG’s vertical position evaluation criteria on the en face view of the glenoid after modified suture-button Latarjet pro-
cedure. Excellent: graft covers more than 75% of the GDA (A); good: graft covers 50% to 75% of the GDA (B); fair: graft covers
25% to 50% of the GDA (C); and poor: graft covers less than 25% of the GDA (D). (CG, coracoid graft; GDA, glenoid defect area.)
 SUTURE-BUTTON LATARJET WITH 2-YEAR FOLLOW-UP 23

Table 2. Patient Data follow-up (P < .001; Table 3). The ASES score increased
from 75.9
5.6 preoperatively to 92.9
6.0 at the last
Parameter Value
follow-up (P < .001). The Rowe score increased from
Age, yr, mean
SD (range) 26.6
6.2 (16-36)
Gender, male/female, n 26/9 40.5
9.3 preoperatively to 93.6
5.6 at the last
Side, left/right, n 14/21 follow-up (P < .001). The Walch-Duplay score
Number of instability events, mean
SD 6.9
4.1 increased from 65.6
10.5 preoperatively to 94.1
5.8
Body mass index, mean
SD 23.4
3.7 at the last follow-up (P < .001). On the basis of the
Beighton score, mean
SD 3.7
2.3
standard deviation of the delta, the calculated MCID
GDA, % 18.5
4.1
>20%, n 10 values for clinical outcomes were 0.75 (VAS), 4.42
15%-20%, n 20 (ASES), 4.35 (Rowe), and 5.93 (Walch-Duplay). The
10%-14%, n 5 percentages of patients achieving the MCID were
Bankart failure, n 2 85.71% (VAS), 97.14% (ASES), 100% (Rowe), and
Hill-Sachs injury, n 28
97.14% (Walch-Duplay).
GDA, glenoid defect area; SD, standard deviation.
Recovery and Complications
All patients preoperatively participated in various
difference (MCID) values were calculated using athletic activities, including yoga, association basketball,
distribution-based methods, specifically by determining soccer, badminton, tennis, and swimming. Return to
half of the standard deviation of the observed change play is 100%, and return to previous performance was
scores (delta). Statistical analysis was conducted using 94.3% (33 cases) at the last follow-up.
SPSS, Version 16.0 software (IBM Corp., Armonk, NY). The average surgical duration was 59.5
12.7 mi-
Pairwise comparisons were performed using a paired t- nutes without intraoperative complications of vascular
test, and statistical significance was set at P < .05. or nerve injury. No recurrent dislocation, positive
apprehension signs, postoperative infection, axillary
Results nerve injury, or vascular injury occurred in any patient
at the last follow-up.
Patient Data
Forty patients were enrolled in this study. One patient Imaging Assessment
died of unrelated illness, and 4 patients were lost to
follow-up. Our final cohort included 35 patients aged
26.6
6.2 years were enrolled, and the rate of follow- CAL and PM Integrity. Postoperative MRI showed the
up was 87.5%. Among them, there were 26 male and 9 integrity of the preserved CAL and reconstructed PM in
female patients, and 14 left and 21 right shoulders were all patients (Fig 4).
included. The mean follow-up time was 26.9
1.9 Graft Position. In the en face view, excellent graft
months. All patients were diagnosed with RASI as the positioning was achieved immediately postoperatively
result of trauma or sports activities. The mean GDA was in 31 patients (88.6%), 2 in a good position, and 1 in a
18.5%
4.1% (range 11.5%-28.4%) of the entire fair position. However, one graft was placed separately
glenoid surface area, and the mean time from initial from the glenoid, covering less than 25% of the GDA,
dislocation to surgery was 27.7
14.6 months (range and ranked as a poor position (Table 4). In the axial
8.4-58.6 months). Twenty-eight patients had a view, 29 grafts were flushed to the glenoid, and 1 and
concomitant Hill-Sachs injury to the humeral head. 5 grafts were fixed medially and laterally, respectively
Two patients underwent revision surgery after Bankart (Table 4).
failure. General information is shown in Table 2.
Graft Healing and Remodeling. A total of 33 grafts
Clinical Assessment (94.2%) achieved bone union at POM 6, and finally, 34
The VAS scores for pain during motion decreased grafts (97.1%) achieved bone union at the last follow-
from 3.1
1.4 preoperatively to 0.97
0.82 at the last up. One patient experienced nonunion but no

Table 3. Functional Results at Preoperative and Final Follow-Up

Parameter Preoperative Final Follow-Up P Value MCID %MCID

VAS for pain (during motion) 3.1
1.4 0.97
0.82 <.001 0.75 85.71
ASES score 75.9
5.6 92.9
6.0 <.001 4.42 97.14
Rowe score 40.5
9.3 93.6
5.6 <.001 4.35 100
Walch-Duplay score 65.6
10.5 94.1
5.8 <.001 5.93 97.14
ASES, American Shoulder and Elbow Surgeons; MCID, minimal clinically important difference; %MCID, number of patients who reached
minimal clinically important difference; VAS, visual analog scale.
24 W. LU ET AL.

Fig 4. MRI shows the integrity of the CAL and PM at various times points in one case, indicating preservation of the CAL and PM
after modified suture-button Latarjet procedure (coronal view). (CAL, coracoacromial ligament; CG, coracoid graft; MRI,
magnetic resonance imaging; PM, pectoralis minor; POD 0, the first postoperative day; POM 24, postoperative 24 months; PRE,
preoperative.)

recurrent dislocation or positive apprehension sign, and Discussion

was satisfied with shoulder function. On the CT results, The most important findings of this study are as fol-
the grafts exhibited obvious growth inferiorly, medially, lows: (1) CAL and PM were preserved after our modi-
and laterally. The interface gap between the glenoid fied suture-button Latarjet procedure. (2) Nearly all
and the graft was nearly filled with bone. The patients achieved graft-glenoid bone union and re-
glenoid and graft fused and were finally remodeled ported satisfied outcomes. (3) The new glenoid
analogously to the shape of the intact glenoid (PFC) remodeled to a new concentric circle with the humeral
(Fig 5). One graft was not in an excellent position on head analogous to the original glenoid over time. (4)
POD 0, even covering less than 25% of the GDA, but Grafts that were malpositioned did not cause arthrop-
the gap between the graft and the glenoid was filled athy of the joints at 2-year follow-up. Therefore, our
with bone and tended to remodel to the PFC over hypothesis was corroborated.
time (Fig 6). The success of the Latarjet technique initially de-
The mean GDA was 18.4
4.1% preoperatively. The pends on the triple blocking mechanism. For more
graft restored 92.9
3.6% of the PFC postoperatively than 60 years, surgeons have continuously improved
compared with 81.5% preoperatively (P < .001), and the Latarjet technique, leading to its constant refine-
97.1% at final follow-up. Any 2 pairwise comparisons ment. However, surgeons, especially those who want
of the 3 are statistically significant (P < .05) (Fig 7). to perform the Latarjet but are worried about com-
plications, have long been concerned about the loss of
Glenohumeral Joint Arthropathy. No patient showed
the CAL and PM and the possible resulting superior
degenerative changes at the last follow-up on the basis
migration of the humeral head and scapular
of the Samilson and Prieto criteria,20 even those with
dyskinetic.6-9
graft positioned laterally (Fig 8).
Our improvement in the modified suture-button
Latarjet is based on the graft healing and remodeling
Table 4. Graft Position in En Face and Axial Views
pattern of suture-button fixation innovated by P. Boil-
eau. Previous work found that suture-button fixation
Location N (%) graft has the characteristics of self-remodeling, namely,
En face view absorption outside the circle. According to the Wolff
Excellent (>75% GDA) 31 (88.6%) law, bone proliferation will occur in an area with a rich
Good (>50%-75% GDA) 2 (5.7%)
Fair (25%-50% GDA) 1 (2.9%)
blood supply, which is possibly the reason why grafts
Poor (<25% GDA) 1 (2.9%) that are positioned laterally achieve more bone regen-
Axial view eration because of the compressive stress stimulus. The
Flush 29 (82.9%) area without a stimulating compressive force would
Medial 1 (6.0 mm, 2.9%) absorb to a certain extent, which may explain why the
Lateral 5 (4.6
1.3 mm, 14.3%)
upper part of the graft absorbs more whereas the part
PFC, perfect fitting circle.
located in the concentric circle of the humeral head
 SUTURE-BUTTON LATARJET WITH 2-YEAR FOLLOW-UP 25

Fig 5. Three cases are presented demonstrating 3D-rendered sagittal computed tomography scan views of the bony incorpo-
ration of the CG over time compared with the unaffected contralateral side (“Contralateral”) and the affected preoperative
glenoid (“Pre-Operative”) after modified suture-button Latarjet procedure. The avulsed bone block can be seen in case 3 at the
injury side preoperatively. The red circle indicates the PFC of the glenoid. Contralateral, preoperative mirror image of the normal
side. Pre-Operative, injury side with GBL preoperatively. (3D, 3-dimensional; CG, coracoid graft; GBL, glenoid bone loss; PFC,
perfect-fitting circle; POD 0, injury side at the first postoperative day. POM 6, injury side at 6 months after surgery. POM 12,
injury side at 12 months after surgery. POM 24, injury side at 24 months after surgery.)

absorbs less.14 On the basis of our previous findings, this The radiologic examination showed that the graft
portion of the graft that is outside of the PFC can be tends to remodel and expand to restore the PFC, even
preserved, and the only partial CG is needed, making though its initial size is larger or smaller or its position
the CAL preservation possible. Therefore, we proposed relative to the glenoid is imperfect. As detected by
keeping partial coracoid osteotomy as minimal as postoperative MRI scans, both CAL and PM have been
possible while meeting the requirements for restoring fully preserved as expected.
the PFC. Meanwhile, having a sufficient length of Another potential benefit of this technique is that its
remaining CP end facilitates the reconstruction of the osteotomized coracoid surface (lateral side) can be
PM, allowing for essentially tension-free suturing in its securely affixed directly to the healing facet of the
original position. glenoid without the need for coracoid freshening. Thus,
In clinical evaluation, we found that the pain and the possibility of insufficient coracoid freshening under
functional scores improved considerably at the last arthroscopy is avoided.
follow-up. Importantly, a substantial percentage of pa- We also proposed the concept of a GDA for RASI with
tients (>85%) achieved MCID for the VAS, ASES, GBL, designed for partial coracoid osteotomy according
Rowe, and Walch-Duplay scores. All patients could to the size and shape of the glenoid defect.15 The length
return to their daily life activities, and most could re- of the CG was in accordance with the length of the
turn to their preoperative sports levels. No recurrence, GDA, making it as perfect a match as possible. How-
severe complications, or degenerative changes were ever, the PICO method and glenoid index (WL index)
observed at the final follow-up. technique reported in the literature emphasize the
26 W. LU ET AL.

Fig 6. Remodeling of CG and graft-glenoid interface on computed tomography en face view of 3 representative good, fair, and
poor cases at POD 0 and POM 6, 12, and 24 after modified suture-button Latarjet procedure. The red circle indicates the PFC of
the glenoid. POD 0, the first postoperative day; POM, postoperative month. (CG, coracoid graft; PFC, perfect-fitting circle; POD 0,
injury side at the first postoperative day. POM 6, injury side at 6 months after surgery. POM 12, injury side at 12 months after
surgery. POM 24, injury side at 24 months after surgery.)

defect ratio,21,22 which is vital for surgical method se- remodeling may be attributed to a flexible fixation
lection; nonetheless, it has no specific reference value mechanism using the suture-button technique.14 There
for the harvest size of CG. was one case in a poor (<25% GDA) position in which
On the basis of the concept of the GDA, aiming to the graft departed from the glenoid at POD 0; however,
restore the PFC, we further redefined the standard after conservative rehabilitation, the bone tissue
vertical position of CG (higher than traditional tech- continued to grow, fill the gap, and tended to restore
niques). They were divided into excellent, good, fair, the PFC (Fig 6). Although remodeling is ongoing, the
and poor positions according to the extent to which the patient has already returned to daily activities with
graft covered the GDA. Moreover, the grafts gradually satisfactory functional outcomes at the last follow-up.
expanded during the 2-year follow-up, filling the gap This technique appears to successfully address 3
and forming the GDA. This type of healing and problems about which experts have been concerned.
 SUTURE-BUTTON LATARJET WITH 2-YEAR FOLLOW-UP 27

is difficult during the Latarjet procedure (Fig 1C).

Furthermore, the SSC is split by only 8 to 10 mm (or
left intact) on the lateral side of the glenoid, safe-
guarded by a switch stick, with minimal interference to
the axillary nerve.24

Limitations
First, the sample size was relatively small, with a short
follow-up time. Second, this technique was not
compared with other techniques, such as the traditional
screw or suture-button arthroscopic Latarjet technique.
Third, the measurement of these CG was not retained
and compared with traditional CG bone mass. Fourth,
the rate of lateral graft malposition is relatively high. Last
Fig 7. PFC restoration at PRE, POD 0, and POM 24 after but not least, the reliability of the classification of graft
modified suture-button Latarjet procedure. *P < .05, **P < .01, vertical position used in this study needs to be verified.
***P < .001. (PFC, perfect-fitting circle; PRE, preoperative; POD
0, first postoperative day; POM 24, postoperative 24 months.) Conclusions
The modified suture-button Latarjet procedure with
CAL and PM preservation obtained good clinical and
First, the CAL is preserved, avoiding the potential risk of radiologic results without recurrence or complications.
superior translation of the humeral head without CAL A substantial percentage of patients (>85%) achieved
shielding.6,7 Second, the PM is reconstructed without the MCID for the VAS, ASES, Rowe, and Walch-Duplay
tension, which retained the internal rotation tension of scores. In addition, the malpositioned graft (17.1%) did
the scapula while minimizing interference to the not cause arthropathy of the joints at 2-year follow-up.
brachial plexus and axillary vessels,23 which was
proved by our postoperative MRI results and no related Disclosures
complication was reported. The SSC muscle split is The authors (W.L. D.L., Y.L., B.W., Z.L., J.S., H.L.,
minimized, reducing the impact on the scapula.8-10 X.L., Z.D.) declare that they have no known competing
Third, the modified antirotation suture tunnel on financial interests or personal relationships that could
coracoid and the triangular arc-shaped CG made pass- have appeared to influence the work reported in this
ing through the SSC muscle split window easier, which paper.

Fig 8. Radiographs taken on the POD 0 and POM 24 show no degenerative changes of the shoulder joints of the surgical side
after modified suture-button Latarjet procedure (one case with graft positioned laterally). (POD 0, first postoperative day; POM
24, postoperative 24 months.)
28 W. LU ET AL.

Funding Statement 11. Deng Z, Liang D, Zhu W, et al. A pilot study of blood supply
This study was supported by the National Natural of the coracoid process and the coracoid bone graft after
Science Foundation of China (No. 82072515), Guang- Latarjet osteotomy. Biosci Rep 2019;39:BSR20190929.
12. Liang D, Liu H, Liang X, et al. Effect of modified arthro-
dong Basic and Applied Basic Research
scopic Latarjet on acromiohumeral distance at 5-year
Foundation (2023A1515220072, 2021A1515220030), follow-up. Orthop J Sports Med 2021;9:
and Start-up Funding for Talented Scientific Research 23259671211063844.
of the First Affiliated Hospital of Wenzhou Medical 13. Xu J, Liu H, Lu W, et al. Clinical outcomes and radiologic
University (2023QD026). assessment of a modified suture button arthroscopic
Latarjet procedure. BMC Musculoskelet Disord 2019;20:173.
14. Xu J, Liu H, Lu W, et al. Modified arthroscopic Latarjet
References procedure: Suture-button fixation achieves excellent
1. Rollick NC, Ono Y, Kurji HM, et al. Long-term outcomes remodeling at 3-year follow-up. Am J Sports Med 2020;48:
of the Bankart and Latarjet repairs: A systematic review. 39-47.
Open Access J Sports Med 2017;8:97-105. 15. Deng Z, Long Z, Lu W. LUtarjet-limit unique coracoid
2. van der Linde JA, Wessel RN, Trantalis JN, van den osteotomy Latarjet (with video). Burns Trauma 2022;10:
Bekerom MP. Review of Latarjet (1954) on the treatment tkac021.
of recurrent shoulder dislocations. J ISAKO 2018;3: 16. Rouleau DM, Hébert-Davies J, Djahangiri A, Godbout V,
242-248. Pelet S, Balg F. Validation of the instability shoulder index
3. Lafosse L, Lejeune E, Bouchard A, Kakuda C, Gobezie R, score in a multicenter reliability study in 114 consecutive
Kochhar T. The arthroscopic Latarjet procedure for the cases. Am J Sports Med 2013;41:278-282.
treatment of anterior shoulder instability. Arthroscopy 17. Hovelius L, Sandstrom B, Olofsson A, Svensson O,
2007;23:1242.e1-e5. Rahme H. The effect of capsular repair, bone block heal-
4. Boileau P, Saliken D, Gendre P, et al. Arthroscopic Latarjet: ing, and position on the results of the Bristow-Latarjet
Suture-button fixation is a safe and reliable alternative to procedure (study III): Long-term follow-up in 319
screw fixation. Arthroscopy 2019;35:1050-1061. shoulders. J Shoulder Elbow Surg 2012;21:647-660.
5. Boileau P, Gendre P, Baba M, et al. A guided surgical 18. Beighton P, Horan F. Orthopaedic aspects of the Ehlers-
approach and novel fixation method for arthroscopic Danlos syndrome. J Bone Joint Surg Br 1969;51:444-453.
Latarjet. J Shoulder Elbow Surg 2016;25:78-89. 19. Kraus TM, Graveleau N, Bohu Y, Pansard E, Klouche S,
6. Lee TQ, Black AD, Tibone JE, McMahon PJ. Release of the Hardy P. Coracoid graft positioning in the Latarjet proced-
coracoacromial ligament can lead to glenohumeral laxity: ure. Knee Surg Sports Traumatol Arthrosc 2016;24:496-501.
A biomechanical study. J Shoulder Elbow Surg 2001;10: 20. Samilson RL, Prieto V. Dislocation arthropathy of the
68-72. shoulder. J Bone Joint Surg Am 1983;65:456-460.
7. Wellmann M, Petersen W, Zantop T, Schanz S, Raschke MJ, 21. de Beer JF, Roberts C. Glenoid bone defectsdopen
Hurschler C. Effect of coracoacromial ligament resection on Latarjet with congruent arc modification. Orthop Clin
glenohumeral stability under active muscle loading in an North Am 2010;41:407-415.
in vitro model. Arthroscopy 2008;24:1258-1264. 22. Baudi P, Righi P, Bolognesi D, et al. How to identify and
8. Beckers J, Van Isacker T, Berghs B. Coracoid process calculate glenoid bone deficit. Chir Organi Mov 2005;90:
transfer for anterior shoulder instability: A pectoralis mi- 145-152.
nor sparing method. Acta Orthop Belg 2020;86:497-501. 23. Galvin JW, Romanowski JR, Boykin RE, Eichinger JK,
9. Carbone S, Moroder P, Runer A, Resch H, Gumina S, Lafosse L. Neurovascular compression after the Latarjet
Hertel R. Scapular dyskinesis after Latarjet procedure. procedure. Orthopedics 2015;38:e1164-e1168.
J Shoulder Elbow Surg 2016;25:422-427. 24. Liang X, Liang D, Qiu Z, et al. Modified Latarjet splitting
10. Cerciello S, Edwards TB, Cerciello G, Walch G. Scapular subscapularis muscle under arthroscopy: An anatomical
position after the open Latarjet procedure: Results of a study based on axillary nerve, glenoid, and subscapularis
computed tomography scan study. J Shoulder Elbow Surg muscle. Zhongguo Xiu Fu Chong Jian Wai Ke Za Zhi
2015;24:199-202. 2023;37:556-560.