Keywords
shoulder - joint instability - recurrence - Bankart lesions - arthroscopy
Introduction
A posterosuperior lesion at the humeral head associated with glenohumeral dislocation
was described by Hill et al in 1940[1] and, since then, it has gained increasing importance in the evaluation of anterior
shoulder instability. Its 47% incidence after first dislocations can increase up to
84% in clinically recurring anterior dislocations, [2]
[3] and its participation in these lesions and in treatment failures after the isolated
repair of an essential Bankart lesion is widely recognized.
Burkhart et al[4] described the term engaging Hill-Sachs for impaction fractures that, for their depth and orientation, allow their engagement
at the glenoid border during abduction and external rotation (ER). Yamamoto et al[5] improved this concept through the glenoid track, which allowed, through a preoperative
computed tomography (CT) scan, the evaluation of the risk of a Hill-Sachs lesion engaging
at the glenoid border.
The relevance and the diagnosis of Hill-Sachs lesion evolved along with its therapeutic
approach. In this context, Wolf et al[6] described an arthroscopic technique of lesion filling with the posterior capsule
and the infraspinatus tendon to render it extra-articular and prevent engaging. This
technique was called remplissage, a French term for filling.
The present study aims to:
-
Evaluate the functional outcome of the remplissage technique;
-
Evaluate the healing of the capsulotenodesis at the Hill-Sachs defect;
-
Evaluate the degree of fatty infiltration in the infraspinatus muscle in the postoperative
period;
-
Evaluate the external rotation force of the shoulder in the postoperative period.
We hypothesized that the remplissage technique promotes a satisfactory healing of
the posterior capsulotenodesis and provides excellent clinical results with no major
repercussions on the range of motion of the shoulder and on the strength and quality
of the infraspinatus muscle.
Material and Methods
Selection of the Patients
This was an analytical, retrospective study of individuals submitted to an arthroscopic
repair surgery of the Bankart lesion associated with the remplissage technique for
the treatment of anterior shoulder instability.
Between 2009 and 2015, 34 patients (34 shoulders) were submitted to an arthroscopic
repair of anterior shoulder instability with the remplissage technique by surgeons
from the Shoulder Group of the Hospital Ortopédico in Belo Horizonte, in the state
of Minas Gerais, Brazil. From these patients, nine were not located or did not agree
to participate in the present study. The minimum follow-up period was of 12 months,
and the maximum follow-up period was of 84 months, with an average of 33.3 months.
The mean age at the time of the procedure was 32.2 years old (ranging from 18 to 51
years old), and the dominant limb was approached in 60% of the cases. Regarding gender,
20% of the patients were female. The mean age at the first instability event was 24 ± 6
years old, and 88% of the patients had presented with at least one dislocation episode
(requiring a medical reduction). In addition, 24% of the patients presented recurrent
subdislocations. Among the patients able to inform the number of dislocation episodes,
the average was two episodes before the procedure. The mean interval between the first
instability event and the procedure was 5 years.
The enrolled patients presented with recurrent shoulder anterior dislocation of traumatic
origin, and a Bankart lesion associated with a Hill-Sachs lesion evidenced by a magnetic
resonance imaging (MRI) exam, with a minimum Hardy index of 20% in a CT scan. Patients
with associated rotator cuff lesions, as well as with fractures at the humeral proximal
third or with bony Bankart lesion > 20% and postsurgical follow-up period of < 1 year
were excluded from the trial.
Surgical Technique
Under general anesthesia and brachial plexus block, the patient is placed in contralateral
recumbency, with the trunk in a 30° posterior inclination, and the affected limb in
50° abduction and 15° anterior flexion, under 7 kg vertical axillary and fixed longitudinal
traction.[7]
The shoulder joint is accessed through the traditional posterior portal, 2 cm medial
and 2 cm inferior to the posterolateral angle of the acromion. The visualization of
the Hill-Sachs lesion requires the removal of the vertical traction and the increase
of the longitudinal traction to open the space between the deltoid and the greater
tubercle of the humerus. A percutaneous portal is created distally to the lateral
border of the acromion and its location is determined with a puncture needle, positioning
it in the central point of the Hill-Sachs lesion. The area is opened with a bone shaver
blade without deepening the lesion. Next, one or two 5-mm anchors are positioned,
according to the size of the bone defect ([Fig. 1]). We use preferentially two anchors, one at the superomedial margin, and the other
at the inferomedial aspect of the lesion, separated by a minimal distance of ∼ 1 cm.
The anchors should not be positioned further than 2 cm apart from each other due to
the risk of inferior involvement of the teres minor.[8] Sutures are passed through the posterior capsule and the infraspinatus tendon with
a penetrator, the U-suture is prepared, and the repairs are left to be fixed at the
end of the procedure ([Fig. 2]). The anteroinferior and anterosuperior portals are prepared with the inside-out
method, through the rotator interval, at the superior border of the subscapular tendon
and at the anterior margin of the acromion adjacent to the bicipital long head tendon,
respectively, and sustained by 8.5 mm cannulae. The arthroscope is then introduced
from the posterior portal to the anterosuperior cannula, and a 7.5 mm cannula is introduced
at the posterior portal for instrumentation.
Fig. 1 Anchor positioning at Hill-Sachs lesion.
Fig. 2 U-suture prepared for capsulotenodesis.
The standard repair of the Bankart lesion is performed with anchors at 1-, 3- and
5-hour positions for the right shoulder, or the equivalent of 7-, 9- and 11-hour for
the left shoulder, which are introduced at the glenoid anterior border with a maximum
joint medialization of 2 mm. Then, a capsulotenodesis is performed at the Hill-Sachs
lesion, and the previously passed U-suture is fixed. The intra-articular view of the
defect filling of the Hill-Sachs lesion is then obtained (indirect suture visualization)
([Fig. 3]).
Fig. 3 Intra-articular view of the filling of the Hill-Sachs lesion defect.
Postoperative Period
Immobilization with an abduction cushion device for 6 weeks to maintain neutral rotation
avoids capsulotenodesis stress, releases active elbow, wrist and fingers flexion and
extension from the first postoperative day, and allows pendulum shoulder movements
from the 2nd week on. Physical therapy rehabilitation starts, and it comprises three phases: the
first one, for pain relief and passive and self-passive range of motion gain; in the
second, isometric muscle strengthening of the rotator cuff begins; finally, the third
phase consists of isotonic muscle strengthening. A progressive proprioception training
is performed since the first phase.
Clinical and Functional Evaluation
Clinical and Functional Evaluation
The clinical evaluation consisted of questions aimed to determine the demographical
and specific features of the clinical instability. The functional evaluation used
the Carter-Rowe[9] and the Walch-Duplay scores,[10] considering criteria such as: stability, assessing apprehension, as well as subdislocation
or the recurrence of dislocation; mobility, assessing the range of motion in anterior
elevation and in external and medial rotation; and function, inquiring about work
and/or sport limitations and pain. Using the predetermined protocol,[11] the ranges of motion were evaluated with a goniometer, and the ER force with the
elbow flexed at 90 degrees and the arm in a neutral position was assessed with a simple
kitchen scale.
Postoperative Image Evaluation
Postoperative Image Evaluation
Filling Index Score of Remplissage and Goutallier Evaluation
At the end of the evaluation, the patients were asked to undergo an MRI of the shoulders
to analyze the defect filling of the Hill-Sachs lesion and the fatty infiltration
of the infraspinatus muscle. The classification was based on the filling index score
of remplissage (FISOR),[12] which classifies the filling, evaluated at the axial and sagittal planes, in five
grades: complete filling (4 points), partial with small residual defect (3 points),
partial with large residual defect (2 points), minimal filling (1 point) and complete
filling failure (zero points). Lastly, the points from the axial and sagittal planes
were added, determining 5 outcomes groups: excellent (7-8 points), good (5-6 points),
regular (3-4 points), bad (1-2 points), and no filling (zero points).
Adaptation to the Filling Index Score of Remplissage
Since our evaluation considered patients submitted to an MRI with no contrast media
(complying with their request for not using these agents), with the arm in ER, and
only axial imaging sequence available, only the final FISOR classification methodology
was considered. The examination was performed with the arm in ER for infraspinatus
tendon relaxation, aiming to highlight a possible healing failure.
Regarding our evaluation of the healing of the capsulotenodesis, originally classified
in five groups, the clustering of low-grade filling (FISOR 0–2) and high-grade filling
(FISOR 3–4) was required to improve the statistical analysis.
Statistical Analysis
The variables were analyzed for their distribution type with the Kolmogorov-Smirnov
test; force and ER variables were compared with the two-tailed Student t-test. The
significance level was set at p < 0.05.
The present research was approved by the Ethics Committee under the number CAAE 62694016.7.0000.5125.
Results
Functional Results
The average Carter-Rowe score was 91 ± 14 (ranging from 50 to 100), with 88% of good
and excellent results[9], while the average Walch-Duplay score was 92 ± 10 (ranging from 70 to 100), with
92% of good and excellent results[10] ([Tables 1] and [2]).
Table 1
|
n
|
%
|
|
≤ 50 (bad)
|
1
|
4
|
|
51–74 (regular)
|
2
|
8
|
|
75–89 (good)
|
2
|
8
|
|
90–100 (excellent)
|
20
|
80
|
|
Total
|
25
|
100
|
Table 2
|
n
|
%
|
|
≤ 50 (bad)
|
0
|
0
|
|
51–74 (regular)
|
2
|
8
|
|
75–89 (good)
|
10
|
40
|
|
90–100 (excellent)
|
13
|
52
|
|
Total
|
25
|
100
|
The ER force showed an average difference of - 0.92 kg when compared with the contralateral
shoulder ([Table 3]).
Table 3
|
n
|
Average (kg)
|
Standard deviation
|
Average difference
|
95% CI
|
p-value
|
|
Operated side
|
25
|
8.96
|
2.67
|
−0.92
|
(−1.44; −0.40)
|
< 0.001
|
|
Non-operated side
|
25
|
9.88
|
2.88
|
|
|
|
Regarding the range of motion, there was an average difference of 10° when considering
a ER at 0° abduction (ER1) and of 10° when considering a ER at 90∘ abduction (ER2),
once again when compared with the contralateral shoulder ([Table 4]).
Table 4
|
n
|
Average (kg)
|
Standard deviation
|
Average difference
|
95% CI
|
p-value
|
|
ER1 operated side
|
25
|
77.04
|
14.6
|
−10.04
|
(− 14.74; −5.34)
|
< 0.001
|
|
ER1 contralateral
|
25
|
87.08
|
7.07
|
|
|
|
|
ER2 operated side
|
25
|
94.52
|
12.37
|
−10
|
(− 14.26; −5.74)
|
< 0.001
|
|
ER2 contralateral
|
25
|
104.52
|
14.61
|
|
|
|
Concerning the recurrence of instability, 4 patients (16%) were positive at the apprehension
clinical examination, and only 1 (4%) complained of a postoperative subdislocation
episode. None of the evaluated patients presented a postoperative dislocation episode.
Imaging Results
From the 25 evaluated patients, 16 complied with a control MRI. According to the adaptation
to the FISOR score, [12] all of them presented high-grade filling ([Fig. 4A] and [B]) ([Table 5]). Fatty infiltration was also evaluated by the Goutallier classification; 11 patients
showed no evidence of fatty infiltration (stage 0), and 4 presented evidence of fatty
streaks (stage 1).
Table 5
|
Filling
|
n
|
%
|
|
FISOR 0 (no filling)
|
0
|
0
|
|
FISOR 1 (low-grade filling)
|
0
|
0
|
|
FISOR 2 (low-grade filling)
|
0
|
0
|
|
FISOR 3 (high-grade filling)
|
4
|
25
|
|
FISOR 4 (high-grade filling)
|
12
|
75
|
Fig. 4 A, partial filling with small residual defect of the Hill-Sachs lesion on magnetic
resonance imaging; B, complete filling.
Discussion
The remplissage technique has been used in the approach to Hill-Sachs off-track lesions
with no major glenoid bony injuries, always associated with the arthroscopic repair
of a Bankart lesion. This technique is widely supported, both biomechanically[13] and clinically, with satisfactory functional results, low recurrence indexes, and
a mild reduction of ER, with no functional repercussion.[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21] These benefits and the complication index of < 1%,[14] excluding recurrences, which is considerably lower than the one observed in the
open procedures, make the remplissage a good option in the therapeutic armamentarium
of anterior instability.
Our group indicated the performance of the remplissage technique when the Hardy index
was ≥ 20%[22], the Hill-Sachs lesion was off-track, and the associated glenoid bone loss was < 25%.
Previous studies[6]
[15]
[16]
[17]
[18]
[19]
[20]
[21] cite recurrence rates ranging from 0 to 12%; a systematic review indicates an average
rate of 5.4%.[14] Comparison was difficult due to the different definitions of recurrence. Some authors
define treatment failure as dislocations, while others define it as a positive apprehension
test at the postoperative period. In our casuistry, 4 patients (16%) were positive
at the apprehension test, and only 1 (4%) had a postoperative subdislocation episode.
There were no dislocation recurrences in the evaluated individuals. All of the individuals
reported satisfaction with the outcome of the surgical treatment. Two patients presented
with adhesive capsulitis during the postoperative rehabilitation, with good evolution.
Regarding the objective functional evaluation, represented by functional scores, good
to excellent postoperative outcomes predominate in the literature, with mean Carter-Rowe
scores of 83 to 91 points.[15]
[16]
[17]
[18]
[19] Boileau et al[15] had 87% of good to excellent results in the Walch-Duplay score, and Merolla et al[17] presented 96% of good to excellent results for both scores. We had similar indexes,
of 88%, of good to excellent results at the Carter-Rowe score, and of 92% for the
Walch-Duplay score.
Another literary controversy is the impact of remplissage at the ER of the shoulder.
Previous studies showed a loss of 1.9° to 10° with statistical relevance,[15]
[16]
[17] while others demonstrated mild losses with no statistical relevance.[14]
[19]
[21] Once again, it is difficult to compare these trials due to the differences in the
measurement techniques, in the shoulder position, and in the contrast with the contralateral
side. As already mentioned, we obtained results with statistically significant losses
in ER, of approximately 10°, both in ER1 and ER2. It is worth mentioning that this
fact is not accompanied by damages at the objective (scores) or subjective (reported
satisfaction with the treatment) functional evaluations.
Since the approach to the infraspinatus tendon is a part of the surgical technique,
we have decided to evaluate the influence of its possible deficit on the global force
of ER of the shoulder, measured in ER1. The fatty infiltration of the muscle belly
was also evaluated with MRI exams.
At the evaluation of the ER force, there was a significant difference of ∼ 1 kg compared
with the measurement of the contralateral side, unlike the results obtained by Merolla
et al[17] and by Zhu et al,[18] who did not find significant differences in these measurements. Since the functional
scores did not include a force measurement, its comparison with their results would
not be relevant.
Image evaluation through MRI showed that all of the 16 patients who agreed with the
examination presented, at most, fatty streaks at the muscular belly (Goutallier 1);
this finding is consistent with the results from Park et al.[23] Regarding the filling of the Hill-Sachs lesion defect, the fact that all patients
presented good to excellent filling (high-grade filling) according to the adaptation
performed by our group makes a statistical analysis impractical, as well as the comparison
with function, range of motion, or force.
Rhee et al[12] concluded that FISOR is a useful measurement criterion to evaluate structural results
after the remplissage procedure. The original work reported 82.6% of good/excellent
results for 23 patients. Regular, poor or negative results were obtained in 17.4%
of the evaluated patients, with no clinical correlation. The FISOR classification
was not used in its original form at the present work due to variations in the MRI
exams, that is, examinations performed without contrast media and with the operated
arm in ER to relax the infraspinatus tendon and thus allow a better evaluation of
the filling of the Hill-Sachs defect. With this adaptation, we obtained 100% of good
and excellent results (high-grade filling), unlike the original work, as mentioned
above.
The limitations of the present study are the following: 1) regarding the imaging results,
the comparative evaluation of the healing and filling of the Hill-Sachs defect was
limited by the adaptation performed by our group in the original FISOR score. This
adaptation was required due to the different protocols employed to obtain the postoperative
images; 2) short minimum follow-up (12 months), lack of a control group, and a high
index of loss at follow-up.
Conclusion
Our experience with the remplissage technique presented good to excellent functional
results (observed in ∼ 90% of the patients) and good capsulotenodesis healing results.
The results suggest that there was no morphological or functional damage to the infraspinatus,
and that the mild loss of ER had no major functional repercussions. Further studies
are required, with higher numbers of patients and a more detailed investigation of
their sports activity, to draw more solid conclusions about these subjects.
