- Email: [email protected]
Evolution of lesions of the labrum-ligament complex in posttraumatic anterior shoulder instability: A prospective study
Evolution of lesions of the labrum-ligament posttraumatic anterior shoulder instability: study Peter Habermeyer, Co/mar, France
MD,
Pascal
The aim
was
to evaluate
of this study
intraarticular Ninety-one were
Gleyze,
the evolution
disease in posttraumatic with postiraumatic arthroscopically
of
shoulder instabi/i+ shoulder instability
patients
examined
MD, and Markus
The intraarticular
disease
recorded on a special documentation sheet (containing 67 descriptive items). The patients were divided into 5 subgroups: first-time dislocation (n = 91; first or second recurrence (n = 12); 3 to 5 recurrences (n = 23); 6 or more recurrences (n = 32); and chronic subluxafions (n = 15). A// data were examined statistically. Each lesion was correlated with stage of evolution, age, and number of recurrences. The most frequent lesions were regrouped into “lesion families. n The initial and most constant lesion was the periosteal disinsertion of the anteroinferior labrum (single lesion). The was
labral
detachment
disinsertion lesion).
With
altered
the detached
lesion].
The
tive process, destruction
additional fourth which
damage
succeeded
in a second ligament
recurrences, structures
stage
saw
stabilizing into
4 stages
reveals
structures.
complex
however,
(quadruple progressively that
for arthroscopic Surg
(triple
site and
A pathophysiological
is proposed,
a precise therapeutic strategy b&zation. (J Shoulder Elbow
damage
of the degenera-
at the insertion
that recurrences
by the (double
mechanisms
tissue
the extension
led to failure
stage
complex
stress
through
of the labrum-ligament
lesion). This stu+ cation
was
of the glenohumeral
classifiwould shoulder
permit sta-
1999;8:66-74.)
P
osttraumatic shoulder instability was among the first glenohumeral diseases to be diagnosed (true dislocation) and treated (manual reduction). During the past decade the classic understanding1~5,8~14~*0,*5,26 of pathologic changes in intraarticular shoulder disease with recurrent anterior dislocation has been enhanced by arthroscopic findings.3~4~‘6~18,‘9~*4 Until recently, nearly complete catalogues of lesions associated with shoulder instability were regularly pubFrom ATOS Praxisklinik, Heidelberg; Clinique du Diaconat, Colmar; and Orthoptidische Universittitsklinik Heidelberg. Reprint requests: Peter Habermeyer, MD, ATOS Praxisklinik Heidelberg, BismarckstraOe 9-l 5, 691 15 Heidelberg, Germany. Copyright 0 1999 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/99/$8.00 + 0 32/l/91900
66
Rickert, MD,
complex in A prospective
Heidelberg,
Germany,
and
lished.2~16~27 However, new arthroscopic findings have shown the necessity of evaluating the relations of these lesions to one another and their links to what could be called the pathogenesis of anteroinferior shoulder instability.-/,’ 2 Working from Johnson’s classification15 of Bankart lesions, we made it the purpose of this study to describe intraarticular disease in posttraumatic shoulder instability and to examine the evolution of the labrum-ligament complex in relation to the number of recurrences. MATERIALS AND METHODS A prospective series of 91 therapeutic arthroscopic procedures in the shoulder (91 patients; October 1992 to November 1993) was conducted. All patients had posttraumatic, unidirectional, anteroinferior shoulder instability. Patients with multidirectional shoulder instability, defined according to Walch and Mole,27 and patients with general joint laxity were excluded. Clinical groups Patients with true dislocation were classified into 4 groups based on the number of dislocations that had occurred in their individual histories. Patients in group 1 (n = 9) were bein treated for their first dislocation of the shoulder. The in 1 ication for arthroscopic shoulder stabilization in these patients was the high risk of recurrence that they had as a result of participating in professional overhead and contact sports. Their mean age was 31 years (range 23 to 48 years, SD k7.5 years), and the average length of their preoperative histories was 0.8 months (range 0 to 3 months, SD *l month). Patients in group 2 (n = 12) had either 1 or 2 recurrences of shoulder dislocation. They had a mean age of 28 years (range 17 to 40 years, SD +6.8 years) and preoperative histories that lasted on average 29.1 months (range 1 to 194 months, SD +58 months). Patients in group 3 (n = 23) each had 3 to 5 recurrences of shoulder dislocation. They had a mean age of 28.5 years (range 17 to 50 years, SD +8.8 years) and preoperative histories that lasted on average 48.9 months (range 6 to 240 months, SD 249 months). Patients in group 4 (n = 32) each had more than 5 recurrences of shoulder dislocation. Their mean age was 28.3 ears (range 19 to 52 years, SD +6.9 years), and they ha J an average preoperative history of 79.3 months (range 4 to 288 months, SD *74 months). Patients in group 5 (n = 15) had chronic subluxations as defined by Blazina and Satzman.6 Their mean age was 30 years (range 20 to 52 years, SD +lO years), and
I Shoulder Volume
Elbow Surg 8, Number
Habermeyer,
DOCUMENTATION Name
SHEET
- SHOULDER
Age
atraumatlc 0
voluntary 0
first bme dlslocabon 0
IO
II 0
lnvoluntaly 0
Rickert
67
Ill 0
non domtnant 0
ant 0
post. 0
1nf 0
regular 0 elongated 0 scarred 0 lnterligamental tear 0 ligament - llmbus tear 0 scarred I” ectop~c poslbon0 H A.G L lesson 0
S LA P LESION mlsslng Type 1 Type II Type Ill Type IV Andrew3 Lesmn
GLEN00 regular 0 G LA D lesion 0 glencud nm defect Grade I 0 Gradell 0 Gradelll 0
-
mulbdlredlonal0
FINDINGS
0 0 0
regular elongated nmtured
SC
LIGAMENTS Type (Mwgan) synovitis
IO
elongated
:
hypoplastic
0
LONG HEAD regular synonbs witaltear compk?le tear subluxation luxation
0 0 0 0
female0
genera1 JOlIlt lkaty 0 hetory (month)
dominant 0
ARTHROSCOPIC
0 0 0 0 0 0 0 0 0 0 0
male 0
(years)
no of recurrences _
I& 0
HILL-SACHS-LESION. mlSSl”g 0 Type I (chondral) 0 Type II (osteochondral)O 0 Type 111(bony)
and
INSTABILITY
SUrflaflE
regular nxsslng hypoplasbc degeneratwe flap tear longltudlnal tear scarred at ongwla1site scarred m ectopc pc&on 8” contwty wth the IGHL not m wntmody wth IGHL Buford Complex
Gleyze,
1
OF BICEPS: 0 0 0 0 0 0
II0
Ill0
MGHL
IV0
IGHL
JOINT SIDE PARTfAL CUFF TEAR (Classify by Snyder) regular 0 Type 1 (5 1 all. wT!dbs> Type II : (I- 2 cm. fikw dmn@im) Type Ill 0 (Z-3cm,fragmefwmn) TypeN 0 (‘3an.cotwetetaar)
THERAPY.
Figure
1 Documentation
sheet
for shoulder
instability.
they had an average preoperative history of 29.8 months (range 3 to 96 months, SD +29 months). A routine clinical examination for instability was done preoperatively. In addition, Hawkins’ load and shift test13 was performed with the patients under anesthesia. In every case an x-ray instability series was evaluated with regard to anterior rim fracture (Bankart fracture), Hill-Sachs lesion, and lenohumeral index. Patients with a Bankart fracture or a 7racture of the greater tuberosity were excluded.
rior pouch in combination with a disruption of the ligamentous and labral structures, which support the anterior translation of the humerus from the glenoid. In other cases in which the ligaments were attenuated or atrophied, we looked for signs of acute trauma such as blood and blood products to verify the cause of trauma.15 This preliminary arthroscopic examination was performed in all patients by a single surgeon (PH) before any measures were attempted to repair the labrum-ligament complex.
Surgical
Prospective
technique
Surgical procedure consisted of a glenohumeral arthroscopy with the patient in the lateral position and the affected arm under twin traction.11 Portal insertion was identical in every patient and begun in the classic posterior manner. A standard 30” scope was used. An anterior portal permitted the insertion of a probe. Visualization of anterior elements, in particular the junction between the inferior glenohumeral ligament (IGHL) and the glenoid labrum, was achieved with an associated anterosuperior portal. Healing of the anterior structures was documented in cases in which scar tissue formation between the labrum and the glenoid rim was seen. Blood staining and localized synovitis with capillary injection were enerally found in such cases. Pat 1 ologic distension or elongation of the glenohumeral ligaments was obvious in patients who had a large ante-
form
Information pertaining to all anatomic structures that could possibly be visualized during arthroscopy of the shoulder was recorded on a s ecial form (Figure 1). The statistically relevant items are s R own in Table I. To attain an optimal statistical exploitation of the data, the status of the body of an anatomic structure was separated from the status of its attachments. Analysis of each structure’s body (glenoid labrum, IGHL, biceps tendon, etc) should not be confused with analysis of its attachments (labral insertions, IGHL insertions, etc). This analytic form was conceived to discriminate data maximally while heeding the statistical imperative to define a normal value for each variable.9 Procedure
First, the surgeon’s ed and
the
rate
consistency
of statistically
and aberrant
reliability were testvalues was extract-
68
Habermeyer,
Table
I Results
Gleyze,
in detailed
and
Rickert
J Shoulder Elbow Surg January/February I999
numbers Group Total
Descriptives Patients Age* Preoperative history7 Right-side cases Left-side cases Labrum Attachments Normal Torn Scar No data Bodies Normal Flap Hypoplastic Absent No data IGHL Attachments Normal Lig-lab Lig-limb pomf;r Ectop ins Bodies Normal Scar Hypotrophic Absent No data Tension Normal Distended Glenoid tear Humeral tear No data LHB Attachments Normal SLAP I SLAP II SLAP Ill SLAP IV Andrews No data Glenoid rim defect Regular Grade I Grade II Grade Ill No data Hill-Sachs lesion Missing Type I (chondral) Type II (osteochondral)
Group 2 (1 - 2 recurrences)
91 29 51 48 43
0,8
12: 31 10
9 31
Group 3 (3-5 recurrences)
;:
6 3
49 12 11
79 15 17
30 10 5
0 7 5 0
2
:, 2 0
4 13 11 4
3 6 5 1
72 5 2 11 1
11 0
70 2 1 17 1 20
11 1 0 0
ti 5
: 0
in years.
labral,
anterior,
tear;
posterior
15 0 1
25 1
ii
11 :
50 0
: 0 8
:,
11 i 4 i
16 14 1 1 0
10 4 1
5 7 3 0 0
32 0
:,
9 1 1
7 12 3 0 1
13 17 1
69 2 12 1 5 2 0
18 1 3 1
27 0 2 0 3 0 0
9 0
72 2
19 0 1
27 1
14 0
i
Fi 0
?I 0
5 3 5 10 0
4 9 5 14 0
14 0 0
/&-limb, lesion.
:,
: 0
1’; 30 0
length in months. Ligament-labrum
26 1
10 12 0
29
tMean lig-lab,
17 1 0 5 0
; 7
:
2 7
*Mean
4;
Group 5 (subluxations)
29 5 7
ii
0
Group 4 (> 5 recurrences)
:;
52 34
$vdg~~bonyl
and
1
(1 dislocafion)
ligament-limbus
tear;
Ecfop
ins,
ectopic
insertion;
K-/B, long
head
of biceps;
z
: : 0
:,
SLAP, superior
Habermeyer,
J Shoulder Elbow Surg Volume 8, Number I
Table
II Pathologtc
findings
expressed
Total (n = 91)
as a percentage Group 1 (1 dislocation)
Gleyze,
and Rickert
69
(%) Group 2 (1 - 2 recurrences)
Group 3 (3 - 5 recurrences)
Group 4 (> 5 recurrences)
Group 5 (subluxations)
Labrum Attachment Body IGHL
78 20
2
Attachment Body Tension LHB
44 43 57
22 22 33
Attachment
24
22
IHB,
tong
head
75 19
73 27
%Z
65 57 65
47 50 59
z; 66
33
22
16
40
0
a
of biceps.
ed. Second, an analysis of all data recorded on the prospective form was performed. For each clinical group, we tried to establish interrelation among all the variables. This entailed, for example, searching for a correlation between a detachment of the labrum and a distension of the IGHL’s body after the first dislocation. The next step consisted of analyzing the outcome over time of each lesion in order to outline a dynamic comprehension of its evolution as a function of the number of recurrences. In this part of the analysis the subluxation group (group 5) was excluded because it constituted an isolated population. Factorial analysis was done to sort out and organize the great amount of collected data. Anatomic structures were divided into 4 sets that we named “lesion families.” Statistical
70 26
100
methods
An information database was devised (4th Dimension, Apple Computers) to obtain the best data exploiiation. Descriptive statistics were used to apply average, standard deviation, and standard error values to quantitative values.22 Factorial analysis was preceded by pertinence and homogeneity testing of collected data (Kayser equations) that ensured the reliability of the single surgeon’s arthroscopic descriptive reports (total pertinence = 0.8; P = .OOOl, close to a maximum considering individual anatomic variability).23 The analytic procedure included 3 successive steps (first, data matrix; second, correlation matrix; third, factor extraction) that allowed extraction of 4 factors. Each anatomic criterion was statistically correlated to 1 predominant factor (multiple regression). This put us in a osition to classify all anatomic criteria into 4 groups (fami Pies), with each group corresponding to 1 of the 4 factors individualized by statistical analysis. These 4 families should be understood as a group of statistical variables (anatomic criteria) that act identically or sufficiently alike to such a degree that it becomes possible to postulate the existence of links (or laws) among them .9~21 (It is important to recall here the difference between correlation, ie, an existing link, and causality, ie, the relation between cause and effect.) In searching for these correlations, or links, we used simple and multiple regression tests. Comparative studies were performed through variance analysis and chi-square testing.22,23 Causality can be demonstrated only by a randomized prospective study.
Among the structures grouped in family 1 were the labral insertion into the glenoid, the IGHL insertion into the labrum, and the long biceps insertion into the labrum. The labral body, the IGHL body, and the capsule were grouped in family 2; the long bite s body and the rotator cuff were placed in family 3; and t Re humeral head and the glenoid were assigned to family 4. RESULTS
The presentation that follows is limited to the evolution of the anterior labrum-ligament complex lesions and to the role of the long biceps in influencing the treatment outcomes of the anterior structures. Accompanying Hill-Sachs lesions and lesions of the glenoid were also analyzed and are mentioned when statistically relevant. Quantified
results
Detailed information corresponding to the various lesions is shown in Table I. Accumulated percentages represent the incidence rates of the lesions, independent of their type or degree (Table II). Results
of correlation
Group 1: First dislocation (n = 9). Labrum. In 89% (8) of the group 1 cases, the labral to glenoid attachments were ruptured (Figure 2). Only 22% (2) of the labra healed with scarring in an ectopic position below the glenoid rim. The most favorable period for primary healing seems to have ended around the sixth week (P < .05). The bodies of 22% of the labra appeared abnormal (2 anteroinferior flaps). /G/-IL. In 22% (2) of the group 1 cases, the junction between the IGHL and the labrum-limbus complex (periosteum of the glenoid rim situated under the labrum) was injured (Figure 3). The substance of the inferior glenohumeral ligament showed irregular alterations (eg, tissue disruption) in 22% (2) of the group 1 cases. In 33% (3) of these cases, the IGHL was distended. Group 2: One or two recurrences (n = 12). labrum. All group 2 cases (100%) had a tear at the attachment site of the labrum to the glenoid. In 42% (5) of these
70
Habermeyer,
Gleyze,
and
Rickert
J Shoulder
Elbow
Surg
January/February
attachment
Figure I6,
Long
lesion of labrum --I-
2 Succession of attachment lesion and biceps tendon; lux., dislocation; rec.,
lesion of labrum -c-
body lesion recurrences.
cases, the labra appeared to have healed in an ectopic position. No lesions of the labral substance were found, however, in this group. This might be explained by the lesser traumatic force necessary to cause a recurrence of dislocation. /GM The detachment of the IGHL from the labrumlimbus complex correlated with a detachment of the labrum from the glenoid (P = .05) and with the presence of a lesion on the long biceps insertion (P = .02). The IGHL appeared pathologic in 42% (5) of the group 2 cases, which correlated with the existence of a labral detachment from the glenoid rim (P = .05). The incidence of lesions associated with tension strains was 42% (4 body distensions, 1 tear on the glenoid side). These lesions correlated with a lesion of the deep surface of the rotator cuff (f = .04). Group
3: Three
to five
recurrences
(n = 23).
labrum.
The labral detachment rate was 70% (16) in the group 3 cases. There appeared to be 8 cases of labral healing, all in ectopic sites. Chronic lesions were more frequent (22%, 5 cases) than acute lesions (4%, 1 case). The presence of chronic lesions correlated with a labral detachment (P = .02) and with an IGHL injury (P < .045). The quality of the labral body (eg, disappearance) was related in the group 3 cases to the time interval from the first dislocation (P = .045). /G/-/L. Detachment of the IGHL from the glenoid rim and the scapular neck, or especially the simultaneous detachment of the IGHL from the labral margin and from the sublabral periosteum, was the most remarkable element noted in the group 3 cases. Such lesions were found in 65% (15) of the group 3 cases compared with 8% (1) of the group 2 cases (Table II). The ligamentous aspect in the third stage was linked to
of glenoid
body
7 999
attachment lesion of LB ---c
labrum
and
attachment
lesion
of long
biceps.
labral degenerative changes (P = .04). There was evidence of chronic change, including scar tissue formation and distension lesions, in 78% of the IGHL bodies. In 8.6% of the IGHL bodies, ligamentous tears were found at the glenoid insertion site (x 86.6%). The presence of these lesions correlated with the length of preoperative history (P = .04), with the existence of a lesion of the deep part of the rotator cuff (P c .04), and with an injury of the biceps tendon (P = .02). Group
4:
More
than
five
recurrences
(n
=
32).
Labrum. The rate of isolated detachment among group 4 cases was 41% (13 cases). Addition of 1 1 cases that healed in an extraanatomic position to the 13 cases that healed with isolated detachment yields a stable rate of alterations at the attachment site of 75% (24 cases). Labral degenerative changes were proportional to the number of recurrences (P = .02). /GM. Of the IGHL attachments in group 4 cases, 47% had abnormal features (21% [7] had abnormal attachments, 25% [8] had ectopic scarring). The presence of these features correlated with the length of preoperative history (P = .Ol) and with the status of the glenoid labrum (P = .OOl). In 50% (16 of 32) of these cases, the IGHL had a pathologic appearance. Such an appearance correlated with the status of the labral body (P = .05) and with a tear at the ligamentous attachment site (P = .OOl). The IGHL was distended in 53% (17) of the cases in group 4. One case had a tear on the glenoid side, and another case had a tear on the ligament’s humeral attachment (C 59%). Group S: Subluxations (n = 15). The incidence rates of various lesions in group 5 were compared with the incidence rates of the same lesions in group 2 (1 or 2 recurrences). Group 2 data were selected for this com-
Habermeyer,
J Shoulder E/bow Surg Volume 8, Number J
Gleyze,
and
Rickert
lux.,
dislo-
71
60% 50% 40% 30% 20% 0% I 1. - 2. ret
1. Iux ’ distension
of IGHL
lesion
I
3 - 5. ret of IGHL
body
> 5 rec.
lesion at IGHLinsertlon I
Figure cation;
3 Progression rec., recurrences
of IGHL
attachment
lesions.
Distension
parison because the duration of symptoms in group 2 patients was nearly equal to the duration of symptoms in group 5 patients. fabrum. The rate of labral detachment in the subluxation group was 73% (1 1 cases) (Figure 4). In 40% (6 cases) a completely detached labrum was seen during the preliminary arthroscopic examination. In 20% (3 cases) the labrum appeared to be healing correctly and in 13% (2) the labrum had healed in an extraanatomic position. The frequency of labral body lesions was proportional to the duration of clinical symptoms (P = .02) and higher than the frequency of labral body lesions in group 2 (27% vs 0%). /GM. The rate of attachment abnormalities was 53% (8 cases). In 4 cases the IGHL was completely detached and in 4 cases there was ectopic scarring near the insertion site. This detachment rate correlated with the length of preoperative history (P = .Ol ) and with the presence of labral hypoplasia or the labrum’s complete absence (P = .OOl). In 33% of the group 5 cases, the IGHL body was injured. The presence of such a lesion correlated with the presence of an unhealed IGHL detachment. In 66% of these cases, there was distension or tearing of the glenohumeral ligament. This was considerably higher than the rate for group 2, in which 42% of the cases had lesions associated with tension strains. Analysis global
of the biceps tendon series (91 cases)
attachment
lesions
in the
In the course of the first true dislocation, 22% (2 cases) had a superior labrum, anterior, and posterior (SLAP II) lesion just under the bicipital attachment site. In these cases the superior labrum-biceps complex was
and
macroscopic
changes
of IGHL
body.
displaced onto the glenoid surface, but could be relocated by probe. After the first and second recurrences, the frequency of bicipital attachment injury increased to 33%; the rate decreased to 22% after 3 to 5 recurrences and 16% after more than 5 recurrences. The presence of bicipital attachment lesions correlated with the occurrence of an IGHL detachment after the first and second recurrences (P = .006), with the degree of IGHL distension (P = .04), and with patient age (P = .02). This last relation introduces a degenerative factor to the lesion’s pathogenesis. Indeed, bicipital attachment lesions were commonly found in patients older than 35 years. DISCUSSION
Four anatomically individualized lesion families can be defined on the basis of this study with statistical factorial analysis (or multidimensional analysis). The first lesion family is that of the attachments or “anatomic hinges.” The family includes the attachment of the anteroinferior labrum onto the glenoid, or anterior hinge #l; the attachment of the IGHL onto the labrum-limbus complex, or anterior hinge #2; and the attachment of the biceps tendon onto the superior labrum, or the superior hinge. The term “hinge” refers to the point of convergence of strains; it is a fragile point of passage of tensile forces. The first of these 3 points to collapse is the labral attachment. Later, in a second phase, the IGHL and bicipital attachments also yield. The anteroinferior labrum seems to function as a “security lock.” Once it is detached, the biceps tendon and IGHL attachments are no longer protected. The labral and IGHL attachments constitute the ante-
72
Habermeyer,
Gleyze,
and
Rickert
labrum insertion labrum body
Figure 4 Comparison Ins., Insertion; IB, Long
J Shoulder Elbow Surg January/February I999
IGHL
of incidence of lesions in group biceps tendon; rec., recurrences
rior hinges, and they are injured before the capsular and ligamentous structures are injured. The biceps tendon appears to be a superior hinge, and it is injured before the lesions of the biceps tendon body and the rotator cuff are iniured (P = .05). The 3 hinges are the first elements to collapse (acute lesions of the attachment). Injury to the hinges influences the outcome over time of the other anterior structures (chronic lesions of the corresponding bodies). The second lesion family is that of the anterior structures, the labral-ligamentous complex. Included in this family are the labral body and the IGHL body (appearance, tensile status). The analytic criteria of the IGHL body are representative of the capsule’s status (appearance, distension). The chronic involvement of the labral body and the IGHL body is made manifest by distended tissue.15 Damage to the labral and IGHL bodies occurs after the collapse of the anterior hinges and constitutes the moment when the disorder becomes chronic. In our opinion, the third recurrence constitutes the “point of no return” in the degeneration of anterior structures. After the third recurrence, tissues progressively and irreversibly lose their mechanical properties. The third lesion family consists of associated lesions, including the long biceps body and the rotator cuff. The fourth lesion family is formed by the humeral head, the glenoid, and other bony parts (Table I). Although the results of the analysis of associated lesions and bony parts are not shown in Table I, these results were factored into the statistical analysis of anteroinferior structures (multiple correlations to be decomposed). We have not been able to take severity of trauma into account despite the need to have such a factor in an anatomic correlation model reliable enough to authorize a new therapeutic strategy. Severity of trauma is certainly an aggravating factor, but we support
ins.
IGHL IGHL
distension LB ins.
body
2 (1 to 2 recurrences)
and
group
5 (chronic
subluxation).
the theory of lesion succession. This theory maintains that before the third recurrence, successive ruptures of the hinges (strain points) take place with acute involvement of anterior structures. The third recurrence represents the “point of no return”; it is the border between reversible and irreversible damage. After the third recurrence the hinge zones have collapsed and the plastic deformation of anterior elements has become irreversible. On the basis of this study we propose the following chronologic classification of lesions in posttraumatic anterior shoulder instability. The set of observed lesions, the outcome over time, the number of recurrences, and the pathophysiologic explanations found in the literature seem to support our analysis of lesions in anterior shoulder instability. We are now able to think in terms of a pathophysiologic and chronologic classification of the lesions involved in anterior shoulder instability. The classification is chronologic in the sense that a time factor is integrated into the analysis. It is the number of recurrences, however, that will determine the outcome of the lesions. Figure 5 illustrates the stages of lesion evolution. We postulate 4 stages. Each stage is defined by the presence of a new lesion. The first stage is marked by an isolated simple lesion at the labral attachment, that is, an isolated labral detachment with a present periosteal hinge (Bankart lesion). The first anterior hinge is the only structure to collapse in the first stage. The second stage is marked by the appearance of a double lesion on the IGHL attachment; it is a combined labral and IGHL detachment lesion. The IGHL detachment is deformed by a continuous sublabral periosteal extension (Perthes lesion) 10,20. This stage represents the disruption of the second anterior hinge. The third stage is marked by a triple lesion and progressive degenerative
Habermeyer,
J Shoulder Elbow Surg Volume 8, Number J
Stage
-
1
Gleyze,
and Rickert
73
-24..
Normal
Labrum
i3/
ISOLATED Bankart
LESION Lesion Isolated labral detachment wth a present periostal hinge
9 Labral
Stage
Attachment
Rupture
I
2
DOUBLE
LESION Perthes
/
7 IGHL
Attachment
Rupture
\
TRIPLE
Double Lesion with a labral remnant
Triple lesion and labral degeneration -fibrous scarring - distention - HAGL
Triple lesion and labral degeneration - hypoplasla - labral absence
Superglenoid
Figure
LESION
5 Chronologic
Degeneration
classification
changes of the detached structures below the level of the glenoid. At the very sites where the 2 previous lesions might have healed, the formation of fibrous scarring occurs. The fourth stage is marked by a quadruple lesion. The degenerative process is extended above the level of the glenoid and progressive disappearance of the labrum-ligament complex begins. CONCLUSION The statistical analysis of different anatomic lesions led us to formulate a chronologic classification based on dynamic pathophysiologic reasoning rather than catalog complex descriptions of multiple arthroscopic findings in the unstable shoulder. The dynamic progression of these lesions seems to suggest a time-dependent, recurrence-dependent worsening of labral and ligamentous pathology in dislocation and subluxation disorders. This is a preliminary study that attempts to explain a pathophysiologic mechanism. Because we did not consider patients’ anatomic and clinical outcomes, it is not possible to unfold the therapeutic implications of this study. By delineating the appearance of an additional lesion with each recurrence, we hope to have created
of successive
lesions
into 4 stages.
an explanation of intraarticular disease in posttraumatic anteroinferior shoulder instability. This should prove useful in planning more precise therapeutic strategies that are better adapted to modern surgical techniques, particularly arthroscopy. REFERENCES 1
Adams JC Recurrent Surg 1948,308.26-38
2
Adolfsson L, lysholm J Arthroscopy and stablllty anterior shoulder mstablllv Arthroscopy 1989,5
3
Andrews JR, Angelo RL Shoulder arthroscopy for the throwing athlete In. Paulos iE, Tlbone JE, editors Operative technique in shoulder surgery Gaithersburg, MD Aspen,1 991 p 79-84. Arnoczky SP, Altchek DW, O’Brien SJ Anatomy of the shoulder In McGlnty JS, editor Operative arthroscopy New York Raven Press; 1991 p 425-42 Bankart AS Recurrent or habitual dlslocatlon of the shoulder joint. Br Med J 1923,2 1 132-3 Blazma ME, Saizman JS Recurrent anterior subluxatlons of the shoulder In athletics A dlstlnct entl% [abstract] J Bone Jolnt Surg 1969,5 1 A. 1037-8
4
5 6
7
8
dlslocatlon
of the shoulder
J Bone
joint
testing for the 3 1.520
Coudane H, Mole D, Sommelet J, Schmitt D Hfstolre naturelle des lesions du bourrelet glenoidlen Apropos de 52 cas J Med iyonnalse 1987,1424 79-85 Gallle WE RecurrIng dlslocatlon of the shoulder J Bone Joint Surg 1948,30B.9-18
74
Habermeyer,
Gleyze, and Rickert
I Shoulder
Elbow
January/February
Gleyze P la don&e en recherche cllnlque expiirlmentale. Mbmoire de DEA Unlt& fondement des sciences du CNRS (UPR2.56). Strasbourg Dept de blostatlstique de CUnlversltb L Pasteur; 1992
18
Nevlaser TJ. The ALPSA lesion, a cause of anterior the shoulder Arthroscopy 1993;9 17-2 1
19
10
Habermeyer P. Operative arthroskopie beI schultermstabllitat In. Habermeyer P, Schweiberer L, editors Schulterchirurgie Munchen, Urban & Schwarzenberg Verlag; 1996 p. 287
20
Nevlaser Tj. The GLAD lesion. another cause der paIn Arthroscopy 1993;9 22-3 Perthes G Uber operationen be) habitueller Deutsch Z Chir 1906,56: 149-5 1
11
Habermeyer P, Wledemann E Arthroscopic three point Bankart suture repair In Resch H, Beck E, edrtors. Arthroscopy of the shoulder. Berlin. Springer Verlag; 1992. Hawkins RJ Controversy In antenor shoulder instability Clan Orthop 199 1;272.1.52-6 1 Hawklns RJ, Bokor DJ. ClInIcal evaluation of shoulder problems In Rockwood CA, Matsen FA, editors. The shoulder Volume 1 Phlladelphla Saunders, 1998 p 164-97 Joessel G. Topographisch-chirurgische anatomle fur studlerende und drzte Bonn Max Cohen; 1884. Johnson LL Shoulder arthroscopy In Johnson LI, editor Arthroscop~c surgery principles and practice St Louis Mosby-Year Book, 1986.
9.
12 13
14 15
16. 17
21
22
A
special
Professor several and I
thank Mansat hundred
publishing
you
to
Michel
left this authors staff.
Mansat,
position
in January
all over
the continent
Bonne
of anterior
of
shoul-
schulterluxatlon.
AG l’approche systkmique, la conceptlon cybern& la notlon de dlscontinult& dans la m&thologle de la blom&dlcale. Syst&mes naturels, syst&mes artlflciels. Vallon, France, Mllieux, 1986 D, lellouch J t’8ssal therapeutlque chez l’homme Paris Flammarlon Mbdeclne Sciences, 198 1
Senghas RE Statistics In the Journal of Bone and Jomt Surgery SuggestIons for authors J Bone Joint Surg 1992;74A 3 19-20
24
Snyder lesions
2.5
Townley CO The capsular mechanism in recurrent drslocation of the shoulder J Bone Joint Surg 1950,32A 370-80. Turkel SJ Stablllzlng mechanrsms preventrng anterior dislocation of the glenohumeral jornt J Bone Jolnt Surg 198 1,63A. 1208-l 7
27.
our
InstabIlIty
23
26
Johnson LL. An arthroscoplst’s perspective of anatomy and pathology. Clan Orthop 1987,223 1 13-2.5 Johnson LL Diagnostic and surgical arthroscopy of the shoulder St LOUIS Mosby-Year Book; 1993 p 300-l 1
Petrovic tlque et recherche Champ Schwartz 2nd ed
Surg I999
as well
Ferkel RD, Friedman 1990;6 274-9
RJ SLAP
Walch G, Mole D lnstablllti& et luxatlons de l’kpaule [artrculatlon glkno-hum&ale) Encycl M&d Chir App’ locomoteur 1991,14037AlO 14
faithful 1999
SJ, Karzel RP, DelPlzzo W, of the shoulder Arthroscopy
Editor with
in Europe
the appreciation
as the
members
for and
many
years.
gratitude
of
of the USES editorial
chance! I
MD,
Pascal
The aim
was
to evaluate
of this study
intraarticular Ninety-one were
Gleyze,
the evolution
disease in posttraumatic with postiraumatic arthroscopically
of
shoulder instabi/i+ shoulder instability
patients
examined
MD, and Markus
The intraarticular
disease
recorded on a special documentation sheet (containing 67 descriptive items). The patients were divided into 5 subgroups: first-time dislocation (n = 91; first or second recurrence (n = 12); 3 to 5 recurrences (n = 23); 6 or more recurrences (n = 32); and chronic subluxafions (n = 15). A// data were examined statistically. Each lesion was correlated with stage of evolution, age, and number of recurrences. The most frequent lesions were regrouped into “lesion families. n The initial and most constant lesion was the periosteal disinsertion of the anteroinferior labrum (single lesion). The was
labral
detachment
disinsertion lesion).
With
altered
the detached
lesion].
The
tive process, destruction
additional fourth which
damage
succeeded
in a second ligament
recurrences, structures
stage
saw
stabilizing into
4 stages
reveals
structures.
complex
however,
(quadruple progressively that
for arthroscopic Surg
(triple
site and
A pathophysiological
is proposed,
a precise therapeutic strategy b&zation. (J Shoulder Elbow
damage
of the degenera-
at the insertion
that recurrences
by the (double
mechanisms
tissue
the extension
led to failure
stage
complex
stress
through
of the labrum-ligament
lesion). This stu+ cation
was
of the glenohumeral
classifiwould shoulder
permit sta-
1999;8:66-74.)
P
osttraumatic shoulder instability was among the first glenohumeral diseases to be diagnosed (true dislocation) and treated (manual reduction). During the past decade the classic understanding1~5,8~14~*0,*5,26 of pathologic changes in intraarticular shoulder disease with recurrent anterior dislocation has been enhanced by arthroscopic findings.3~4~‘6~18,‘9~*4 Until recently, nearly complete catalogues of lesions associated with shoulder instability were regularly pubFrom ATOS Praxisklinik, Heidelberg; Clinique du Diaconat, Colmar; and Orthoptidische Universittitsklinik Heidelberg. Reprint requests: Peter Habermeyer, MD, ATOS Praxisklinik Heidelberg, BismarckstraOe 9-l 5, 691 15 Heidelberg, Germany. Copyright 0 1999 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/99/$8.00 + 0 32/l/91900
66
Rickert, MD,
complex in A prospective
Heidelberg,
Germany,
and
lished.2~16~27 However, new arthroscopic findings have shown the necessity of evaluating the relations of these lesions to one another and their links to what could be called the pathogenesis of anteroinferior shoulder instability.-/,’ 2 Working from Johnson’s classification15 of Bankart lesions, we made it the purpose of this study to describe intraarticular disease in posttraumatic shoulder instability and to examine the evolution of the labrum-ligament complex in relation to the number of recurrences. MATERIALS AND METHODS A prospective series of 91 therapeutic arthroscopic procedures in the shoulder (91 patients; October 1992 to November 1993) was conducted. All patients had posttraumatic, unidirectional, anteroinferior shoulder instability. Patients with multidirectional shoulder instability, defined according to Walch and Mole,27 and patients with general joint laxity were excluded. Clinical groups Patients with true dislocation were classified into 4 groups based on the number of dislocations that had occurred in their individual histories. Patients in group 1 (n = 9) were bein treated for their first dislocation of the shoulder. The in 1 ication for arthroscopic shoulder stabilization in these patients was the high risk of recurrence that they had as a result of participating in professional overhead and contact sports. Their mean age was 31 years (range 23 to 48 years, SD k7.5 years), and the average length of their preoperative histories was 0.8 months (range 0 to 3 months, SD *l month). Patients in group 2 (n = 12) had either 1 or 2 recurrences of shoulder dislocation. They had a mean age of 28 years (range 17 to 40 years, SD +6.8 years) and preoperative histories that lasted on average 29.1 months (range 1 to 194 months, SD +58 months). Patients in group 3 (n = 23) each had 3 to 5 recurrences of shoulder dislocation. They had a mean age of 28.5 years (range 17 to 50 years, SD +8.8 years) and preoperative histories that lasted on average 48.9 months (range 6 to 240 months, SD 249 months). Patients in group 4 (n = 32) each had more than 5 recurrences of shoulder dislocation. Their mean age was 28.3 ears (range 19 to 52 years, SD +6.9 years), and they ha J an average preoperative history of 79.3 months (range 4 to 288 months, SD *74 months). Patients in group 5 (n = 15) had chronic subluxations as defined by Blazina and Satzman.6 Their mean age was 30 years (range 20 to 52 years, SD +lO years), and
I Shoulder Volume
Elbow Surg 8, Number
Habermeyer,
DOCUMENTATION Name
SHEET
- SHOULDER
Age
atraumatlc 0
voluntary 0
first bme dlslocabon 0
IO
II 0
lnvoluntaly 0
Rickert
67
Ill 0
non domtnant 0
ant 0
post. 0
1nf 0
regular 0 elongated 0 scarred 0 lnterligamental tear 0 ligament - llmbus tear 0 scarred I” ectop~c poslbon0 H A.G L lesson 0
S LA P LESION mlsslng Type 1 Type II Type Ill Type IV Andrew3 Lesmn
GLEN00 regular 0 G LA D lesion 0 glencud nm defect Grade I 0 Gradell 0 Gradelll 0
-
mulbdlredlonal0
FINDINGS
0 0 0
regular elongated nmtured
SC
LIGAMENTS Type (Mwgan) synovitis
IO
elongated
:
hypoplastic
0
LONG HEAD regular synonbs witaltear compk?le tear subluxation luxation
0 0 0 0
female0
genera1 JOlIlt lkaty 0 hetory (month)
dominant 0
ARTHROSCOPIC
0 0 0 0 0 0 0 0 0 0 0
male 0
(years)
no of recurrences _
I& 0
HILL-SACHS-LESION. mlSSl”g 0 Type I (chondral) 0 Type II (osteochondral)O 0 Type 111(bony)
and
INSTABILITY
SUrflaflE
regular nxsslng hypoplasbc degeneratwe flap tear longltudlnal tear scarred at ongwla1site scarred m ectopc pc&on 8” contwty wth the IGHL not m wntmody wth IGHL Buford Complex
Gleyze,
1
OF BICEPS: 0 0 0 0 0 0
II0
Ill0
MGHL
IV0
IGHL
JOINT SIDE PARTfAL CUFF TEAR (Classify by Snyder) regular 0 Type 1 (5 1 all. wT!dbs> Type II : (I- 2 cm. fikw dmn@im) Type Ill 0 (Z-3cm,fragmefwmn) TypeN 0 (‘3an.cotwetetaar)
THERAPY.
Figure
1 Documentation
sheet
for shoulder
instability.
they had an average preoperative history of 29.8 months (range 3 to 96 months, SD +29 months). A routine clinical examination for instability was done preoperatively. In addition, Hawkins’ load and shift test13 was performed with the patients under anesthesia. In every case an x-ray instability series was evaluated with regard to anterior rim fracture (Bankart fracture), Hill-Sachs lesion, and lenohumeral index. Patients with a Bankart fracture or a 7racture of the greater tuberosity were excluded.
rior pouch in combination with a disruption of the ligamentous and labral structures, which support the anterior translation of the humerus from the glenoid. In other cases in which the ligaments were attenuated or atrophied, we looked for signs of acute trauma such as blood and blood products to verify the cause of trauma.15 This preliminary arthroscopic examination was performed in all patients by a single surgeon (PH) before any measures were attempted to repair the labrum-ligament complex.
Surgical
Prospective
technique
Surgical procedure consisted of a glenohumeral arthroscopy with the patient in the lateral position and the affected arm under twin traction.11 Portal insertion was identical in every patient and begun in the classic posterior manner. A standard 30” scope was used. An anterior portal permitted the insertion of a probe. Visualization of anterior elements, in particular the junction between the inferior glenohumeral ligament (IGHL) and the glenoid labrum, was achieved with an associated anterosuperior portal. Healing of the anterior structures was documented in cases in which scar tissue formation between the labrum and the glenoid rim was seen. Blood staining and localized synovitis with capillary injection were enerally found in such cases. Pat 1 ologic distension or elongation of the glenohumeral ligaments was obvious in patients who had a large ante-
form
Information pertaining to all anatomic structures that could possibly be visualized during arthroscopy of the shoulder was recorded on a s ecial form (Figure 1). The statistically relevant items are s R own in Table I. To attain an optimal statistical exploitation of the data, the status of the body of an anatomic structure was separated from the status of its attachments. Analysis of each structure’s body (glenoid labrum, IGHL, biceps tendon, etc) should not be confused with analysis of its attachments (labral insertions, IGHL insertions, etc). This analytic form was conceived to discriminate data maximally while heeding the statistical imperative to define a normal value for each variable.9 Procedure
First, the surgeon’s ed and
the
rate
consistency
of statistically
and aberrant
reliability were testvalues was extract-
68
Habermeyer,
Table
I Results
Gleyze,
in detailed
and
Rickert
J Shoulder Elbow Surg January/February I999
numbers Group Total
Descriptives Patients Age* Preoperative history7 Right-side cases Left-side cases Labrum Attachments Normal Torn Scar No data Bodies Normal Flap Hypoplastic Absent No data IGHL Attachments Normal Lig-lab Lig-limb pomf;r Ectop ins Bodies Normal Scar Hypotrophic Absent No data Tension Normal Distended Glenoid tear Humeral tear No data LHB Attachments Normal SLAP I SLAP II SLAP Ill SLAP IV Andrews No data Glenoid rim defect Regular Grade I Grade II Grade Ill No data Hill-Sachs lesion Missing Type I (chondral) Type II (osteochondral)
Group 2 (1 - 2 recurrences)
91 29 51 48 43
0,8
12: 31 10
9 31
Group 3 (3-5 recurrences)
;:
6 3
49 12 11
79 15 17
30 10 5
0 7 5 0
2
:, 2 0
4 13 11 4
3 6 5 1
72 5 2 11 1
11 0
70 2 1 17 1 20
11 1 0 0
ti 5
: 0
in years.
labral,
anterior,
tear;
posterior
15 0 1
25 1
ii
11 :
50 0
: 0 8
:,
11 i 4 i
16 14 1 1 0
10 4 1
5 7 3 0 0
32 0
:,
9 1 1
7 12 3 0 1
13 17 1
69 2 12 1 5 2 0
18 1 3 1
27 0 2 0 3 0 0
9 0
72 2
19 0 1
27 1
14 0
i
Fi 0
?I 0
5 3 5 10 0
4 9 5 14 0
14 0 0
/&-limb, lesion.
:,
: 0
1’; 30 0
length in months. Ligament-labrum
26 1
10 12 0
29
tMean lig-lab,
17 1 0 5 0
; 7
:
2 7
*Mean
4;
Group 5 (subluxations)
29 5 7
ii
0
Group 4 (> 5 recurrences)
:;
52 34
$vdg~~bonyl
and
1
(1 dislocafion)
ligament-limbus
tear;
Ecfop
ins,
ectopic
insertion;
K-/B, long
head
of biceps;
z
: : 0
:,
SLAP, superior
Habermeyer,
J Shoulder Elbow Surg Volume 8, Number I
Table
II Pathologtc
findings
expressed
Total (n = 91)
as a percentage Group 1 (1 dislocation)
Gleyze,
and Rickert
69
(%) Group 2 (1 - 2 recurrences)
Group 3 (3 - 5 recurrences)
Group 4 (> 5 recurrences)
Group 5 (subluxations)
Labrum Attachment Body IGHL
78 20
2
Attachment Body Tension LHB
44 43 57
22 22 33
Attachment
24
22
IHB,
tong
head
75 19
73 27
%Z
65 57 65
47 50 59
z; 66
33
22
16
40
0
a
of biceps.
ed. Second, an analysis of all data recorded on the prospective form was performed. For each clinical group, we tried to establish interrelation among all the variables. This entailed, for example, searching for a correlation between a detachment of the labrum and a distension of the IGHL’s body after the first dislocation. The next step consisted of analyzing the outcome over time of each lesion in order to outline a dynamic comprehension of its evolution as a function of the number of recurrences. In this part of the analysis the subluxation group (group 5) was excluded because it constituted an isolated population. Factorial analysis was done to sort out and organize the great amount of collected data. Anatomic structures were divided into 4 sets that we named “lesion families.” Statistical
70 26
100
methods
An information database was devised (4th Dimension, Apple Computers) to obtain the best data exploiiation. Descriptive statistics were used to apply average, standard deviation, and standard error values to quantitative values.22 Factorial analysis was preceded by pertinence and homogeneity testing of collected data (Kayser equations) that ensured the reliability of the single surgeon’s arthroscopic descriptive reports (total pertinence = 0.8; P = .OOOl, close to a maximum considering individual anatomic variability).23 The analytic procedure included 3 successive steps (first, data matrix; second, correlation matrix; third, factor extraction) that allowed extraction of 4 factors. Each anatomic criterion was statistically correlated to 1 predominant factor (multiple regression). This put us in a osition to classify all anatomic criteria into 4 groups (fami Pies), with each group corresponding to 1 of the 4 factors individualized by statistical analysis. These 4 families should be understood as a group of statistical variables (anatomic criteria) that act identically or sufficiently alike to such a degree that it becomes possible to postulate the existence of links (or laws) among them .9~21 (It is important to recall here the difference between correlation, ie, an existing link, and causality, ie, the relation between cause and effect.) In searching for these correlations, or links, we used simple and multiple regression tests. Comparative studies were performed through variance analysis and chi-square testing.22,23 Causality can be demonstrated only by a randomized prospective study.
Among the structures grouped in family 1 were the labral insertion into the glenoid, the IGHL insertion into the labrum, and the long biceps insertion into the labrum. The labral body, the IGHL body, and the capsule were grouped in family 2; the long bite s body and the rotator cuff were placed in family 3; and t Re humeral head and the glenoid were assigned to family 4. RESULTS
The presentation that follows is limited to the evolution of the anterior labrum-ligament complex lesions and to the role of the long biceps in influencing the treatment outcomes of the anterior structures. Accompanying Hill-Sachs lesions and lesions of the glenoid were also analyzed and are mentioned when statistically relevant. Quantified
results
Detailed information corresponding to the various lesions is shown in Table I. Accumulated percentages represent the incidence rates of the lesions, independent of their type or degree (Table II). Results
of correlation
Group 1: First dislocation (n = 9). Labrum. In 89% (8) of the group 1 cases, the labral to glenoid attachments were ruptured (Figure 2). Only 22% (2) of the labra healed with scarring in an ectopic position below the glenoid rim. The most favorable period for primary healing seems to have ended around the sixth week (P < .05). The bodies of 22% of the labra appeared abnormal (2 anteroinferior flaps). /G/-IL. In 22% (2) of the group 1 cases, the junction between the IGHL and the labrum-limbus complex (periosteum of the glenoid rim situated under the labrum) was injured (Figure 3). The substance of the inferior glenohumeral ligament showed irregular alterations (eg, tissue disruption) in 22% (2) of the group 1 cases. In 33% (3) of these cases, the IGHL was distended. Group 2: One or two recurrences (n = 12). labrum. All group 2 cases (100%) had a tear at the attachment site of the labrum to the glenoid. In 42% (5) of these
70
Habermeyer,
Gleyze,
and
Rickert
J Shoulder
Elbow
Surg
January/February
attachment
Figure I6,
Long
lesion of labrum --I-
2 Succession of attachment lesion and biceps tendon; lux., dislocation; rec.,
lesion of labrum -c-
body lesion recurrences.
cases, the labra appeared to have healed in an ectopic position. No lesions of the labral substance were found, however, in this group. This might be explained by the lesser traumatic force necessary to cause a recurrence of dislocation. /GM The detachment of the IGHL from the labrumlimbus complex correlated with a detachment of the labrum from the glenoid (P = .05) and with the presence of a lesion on the long biceps insertion (P = .02). The IGHL appeared pathologic in 42% (5) of the group 2 cases, which correlated with the existence of a labral detachment from the glenoid rim (P = .05). The incidence of lesions associated with tension strains was 42% (4 body distensions, 1 tear on the glenoid side). These lesions correlated with a lesion of the deep surface of the rotator cuff (f = .04). Group
3: Three
to five
recurrences
(n = 23).
labrum.
The labral detachment rate was 70% (16) in the group 3 cases. There appeared to be 8 cases of labral healing, all in ectopic sites. Chronic lesions were more frequent (22%, 5 cases) than acute lesions (4%, 1 case). The presence of chronic lesions correlated with a labral detachment (P = .02) and with an IGHL injury (P < .045). The quality of the labral body (eg, disappearance) was related in the group 3 cases to the time interval from the first dislocation (P = .045). /G/-/L. Detachment of the IGHL from the glenoid rim and the scapular neck, or especially the simultaneous detachment of the IGHL from the labral margin and from the sublabral periosteum, was the most remarkable element noted in the group 3 cases. Such lesions were found in 65% (15) of the group 3 cases compared with 8% (1) of the group 2 cases (Table II). The ligamentous aspect in the third stage was linked to
of glenoid
body
7 999
attachment lesion of LB ---c
labrum
and
attachment
lesion
of long
biceps.
labral degenerative changes (P = .04). There was evidence of chronic change, including scar tissue formation and distension lesions, in 78% of the IGHL bodies. In 8.6% of the IGHL bodies, ligamentous tears were found at the glenoid insertion site (x 86.6%). The presence of these lesions correlated with the length of preoperative history (P = .04), with the existence of a lesion of the deep part of the rotator cuff (P c .04), and with an injury of the biceps tendon (P = .02). Group
4:
More
than
five
recurrences
(n
=
32).
Labrum. The rate of isolated detachment among group 4 cases was 41% (13 cases). Addition of 1 1 cases that healed in an extraanatomic position to the 13 cases that healed with isolated detachment yields a stable rate of alterations at the attachment site of 75% (24 cases). Labral degenerative changes were proportional to the number of recurrences (P = .02). /GM. Of the IGHL attachments in group 4 cases, 47% had abnormal features (21% [7] had abnormal attachments, 25% [8] had ectopic scarring). The presence of these features correlated with the length of preoperative history (P = .Ol) and with the status of the glenoid labrum (P = .OOl). In 50% (16 of 32) of these cases, the IGHL had a pathologic appearance. Such an appearance correlated with the status of the labral body (P = .05) and with a tear at the ligamentous attachment site (P = .OOl). The IGHL was distended in 53% (17) of the cases in group 4. One case had a tear on the glenoid side, and another case had a tear on the ligament’s humeral attachment (C 59%). Group S: Subluxations (n = 15). The incidence rates of various lesions in group 5 were compared with the incidence rates of the same lesions in group 2 (1 or 2 recurrences). Group 2 data were selected for this com-
Habermeyer,
J Shoulder E/bow Surg Volume 8, Number J
Gleyze,
and
Rickert
lux.,
dislo-
71
60% 50% 40% 30% 20% 0% I 1. - 2. ret
1. Iux ’ distension
of IGHL
lesion
I
3 - 5. ret of IGHL
body
> 5 rec.
lesion at IGHLinsertlon I
Figure cation;
3 Progression rec., recurrences
of IGHL
attachment
lesions.
Distension
parison because the duration of symptoms in group 2 patients was nearly equal to the duration of symptoms in group 5 patients. fabrum. The rate of labral detachment in the subluxation group was 73% (1 1 cases) (Figure 4). In 40% (6 cases) a completely detached labrum was seen during the preliminary arthroscopic examination. In 20% (3 cases) the labrum appeared to be healing correctly and in 13% (2) the labrum had healed in an extraanatomic position. The frequency of labral body lesions was proportional to the duration of clinical symptoms (P = .02) and higher than the frequency of labral body lesions in group 2 (27% vs 0%). /GM. The rate of attachment abnormalities was 53% (8 cases). In 4 cases the IGHL was completely detached and in 4 cases there was ectopic scarring near the insertion site. This detachment rate correlated with the length of preoperative history (P = .Ol ) and with the presence of labral hypoplasia or the labrum’s complete absence (P = .OOl). In 33% of the group 5 cases, the IGHL body was injured. The presence of such a lesion correlated with the presence of an unhealed IGHL detachment. In 66% of these cases, there was distension or tearing of the glenohumeral ligament. This was considerably higher than the rate for group 2, in which 42% of the cases had lesions associated with tension strains. Analysis global
of the biceps tendon series (91 cases)
attachment
lesions
in the
In the course of the first true dislocation, 22% (2 cases) had a superior labrum, anterior, and posterior (SLAP II) lesion just under the bicipital attachment site. In these cases the superior labrum-biceps complex was
and
macroscopic
changes
of IGHL
body.
displaced onto the glenoid surface, but could be relocated by probe. After the first and second recurrences, the frequency of bicipital attachment injury increased to 33%; the rate decreased to 22% after 3 to 5 recurrences and 16% after more than 5 recurrences. The presence of bicipital attachment lesions correlated with the occurrence of an IGHL detachment after the first and second recurrences (P = .006), with the degree of IGHL distension (P = .04), and with patient age (P = .02). This last relation introduces a degenerative factor to the lesion’s pathogenesis. Indeed, bicipital attachment lesions were commonly found in patients older than 35 years. DISCUSSION
Four anatomically individualized lesion families can be defined on the basis of this study with statistical factorial analysis (or multidimensional analysis). The first lesion family is that of the attachments or “anatomic hinges.” The family includes the attachment of the anteroinferior labrum onto the glenoid, or anterior hinge #l; the attachment of the IGHL onto the labrum-limbus complex, or anterior hinge #2; and the attachment of the biceps tendon onto the superior labrum, or the superior hinge. The term “hinge” refers to the point of convergence of strains; it is a fragile point of passage of tensile forces. The first of these 3 points to collapse is the labral attachment. Later, in a second phase, the IGHL and bicipital attachments also yield. The anteroinferior labrum seems to function as a “security lock.” Once it is detached, the biceps tendon and IGHL attachments are no longer protected. The labral and IGHL attachments constitute the ante-
72
Habermeyer,
Gleyze,
and
Rickert
labrum insertion labrum body
Figure 4 Comparison Ins., Insertion; IB, Long
J Shoulder Elbow Surg January/February I999
IGHL
of incidence of lesions in group biceps tendon; rec., recurrences
rior hinges, and they are injured before the capsular and ligamentous structures are injured. The biceps tendon appears to be a superior hinge, and it is injured before the lesions of the biceps tendon body and the rotator cuff are iniured (P = .05). The 3 hinges are the first elements to collapse (acute lesions of the attachment). Injury to the hinges influences the outcome over time of the other anterior structures (chronic lesions of the corresponding bodies). The second lesion family is that of the anterior structures, the labral-ligamentous complex. Included in this family are the labral body and the IGHL body (appearance, tensile status). The analytic criteria of the IGHL body are representative of the capsule’s status (appearance, distension). The chronic involvement of the labral body and the IGHL body is made manifest by distended tissue.15 Damage to the labral and IGHL bodies occurs after the collapse of the anterior hinges and constitutes the moment when the disorder becomes chronic. In our opinion, the third recurrence constitutes the “point of no return” in the degeneration of anterior structures. After the third recurrence, tissues progressively and irreversibly lose their mechanical properties. The third lesion family consists of associated lesions, including the long biceps body and the rotator cuff. The fourth lesion family is formed by the humeral head, the glenoid, and other bony parts (Table I). Although the results of the analysis of associated lesions and bony parts are not shown in Table I, these results were factored into the statistical analysis of anteroinferior structures (multiple correlations to be decomposed). We have not been able to take severity of trauma into account despite the need to have such a factor in an anatomic correlation model reliable enough to authorize a new therapeutic strategy. Severity of trauma is certainly an aggravating factor, but we support
ins.
IGHL IGHL
distension LB ins.
body
2 (1 to 2 recurrences)
and
group
5 (chronic
subluxation).
the theory of lesion succession. This theory maintains that before the third recurrence, successive ruptures of the hinges (strain points) take place with acute involvement of anterior structures. The third recurrence represents the “point of no return”; it is the border between reversible and irreversible damage. After the third recurrence the hinge zones have collapsed and the plastic deformation of anterior elements has become irreversible. On the basis of this study we propose the following chronologic classification of lesions in posttraumatic anterior shoulder instability. The set of observed lesions, the outcome over time, the number of recurrences, and the pathophysiologic explanations found in the literature seem to support our analysis of lesions in anterior shoulder instability. We are now able to think in terms of a pathophysiologic and chronologic classification of the lesions involved in anterior shoulder instability. The classification is chronologic in the sense that a time factor is integrated into the analysis. It is the number of recurrences, however, that will determine the outcome of the lesions. Figure 5 illustrates the stages of lesion evolution. We postulate 4 stages. Each stage is defined by the presence of a new lesion. The first stage is marked by an isolated simple lesion at the labral attachment, that is, an isolated labral detachment with a present periosteal hinge (Bankart lesion). The first anterior hinge is the only structure to collapse in the first stage. The second stage is marked by the appearance of a double lesion on the IGHL attachment; it is a combined labral and IGHL detachment lesion. The IGHL detachment is deformed by a continuous sublabral periosteal extension (Perthes lesion) 10,20. This stage represents the disruption of the second anterior hinge. The third stage is marked by a triple lesion and progressive degenerative
Habermeyer,
J Shoulder Elbow Surg Volume 8, Number J
Stage
-
1
Gleyze,
and Rickert
73
-24..
Normal
Labrum
i3/
ISOLATED Bankart
LESION Lesion Isolated labral detachment wth a present periostal hinge
9 Labral
Stage
Attachment
Rupture
I
2
DOUBLE
LESION Perthes
/
7 IGHL
Attachment
Rupture
\
TRIPLE
Double Lesion with a labral remnant
Triple lesion and labral degeneration -fibrous scarring - distention - HAGL
Triple lesion and labral degeneration - hypoplasla - labral absence
Superglenoid
Figure
LESION
5 Chronologic
Degeneration
classification
changes of the detached structures below the level of the glenoid. At the very sites where the 2 previous lesions might have healed, the formation of fibrous scarring occurs. The fourth stage is marked by a quadruple lesion. The degenerative process is extended above the level of the glenoid and progressive disappearance of the labrum-ligament complex begins. CONCLUSION The statistical analysis of different anatomic lesions led us to formulate a chronologic classification based on dynamic pathophysiologic reasoning rather than catalog complex descriptions of multiple arthroscopic findings in the unstable shoulder. The dynamic progression of these lesions seems to suggest a time-dependent, recurrence-dependent worsening of labral and ligamentous pathology in dislocation and subluxation disorders. This is a preliminary study that attempts to explain a pathophysiologic mechanism. Because we did not consider patients’ anatomic and clinical outcomes, it is not possible to unfold the therapeutic implications of this study. By delineating the appearance of an additional lesion with each recurrence, we hope to have created
of successive
lesions
into 4 stages.
an explanation of intraarticular disease in posttraumatic anteroinferior shoulder instability. This should prove useful in planning more precise therapeutic strategies that are better adapted to modern surgical techniques, particularly arthroscopy. REFERENCES 1
Adams JC Recurrent Surg 1948,308.26-38
2
Adolfsson L, lysholm J Arthroscopy and stablllty anterior shoulder mstablllv Arthroscopy 1989,5
3
Andrews JR, Angelo RL Shoulder arthroscopy for the throwing athlete In. Paulos iE, Tlbone JE, editors Operative technique in shoulder surgery Gaithersburg, MD Aspen,1 991 p 79-84. Arnoczky SP, Altchek DW, O’Brien SJ Anatomy of the shoulder In McGlnty JS, editor Operative arthroscopy New York Raven Press; 1991 p 425-42 Bankart AS Recurrent or habitual dlslocatlon of the shoulder joint. Br Med J 1923,2 1 132-3 Blazma ME, Saizman JS Recurrent anterior subluxatlons of the shoulder In athletics A dlstlnct entl% [abstract] J Bone Jolnt Surg 1969,5 1 A. 1037-8
4
5 6
7
8
dlslocatlon
of the shoulder
J Bone
joint
testing for the 3 1.520
Coudane H, Mole D, Sommelet J, Schmitt D Hfstolre naturelle des lesions du bourrelet glenoidlen Apropos de 52 cas J Med iyonnalse 1987,1424 79-85 Gallle WE RecurrIng dlslocatlon of the shoulder J Bone Joint Surg 1948,30B.9-18
74
Habermeyer,
Gleyze, and Rickert
I Shoulder
Elbow
January/February
Gleyze P la don&e en recherche cllnlque expiirlmentale. Mbmoire de DEA Unlt& fondement des sciences du CNRS (UPR2.56). Strasbourg Dept de blostatlstique de CUnlversltb L Pasteur; 1992
18
Nevlaser TJ. The ALPSA lesion, a cause of anterior the shoulder Arthroscopy 1993;9 17-2 1
19
10
Habermeyer P. Operative arthroskopie beI schultermstabllitat In. Habermeyer P, Schweiberer L, editors Schulterchirurgie Munchen, Urban & Schwarzenberg Verlag; 1996 p. 287
20
Nevlaser Tj. The GLAD lesion. another cause der paIn Arthroscopy 1993;9 22-3 Perthes G Uber operationen be) habitueller Deutsch Z Chir 1906,56: 149-5 1
11
Habermeyer P, Wledemann E Arthroscopic three point Bankart suture repair In Resch H, Beck E, edrtors. Arthroscopy of the shoulder. Berlin. Springer Verlag; 1992. Hawkins RJ Controversy In antenor shoulder instability Clan Orthop 199 1;272.1.52-6 1 Hawklns RJ, Bokor DJ. ClInIcal evaluation of shoulder problems In Rockwood CA, Matsen FA, editors. The shoulder Volume 1 Phlladelphla Saunders, 1998 p 164-97 Joessel G. Topographisch-chirurgische anatomle fur studlerende und drzte Bonn Max Cohen; 1884. Johnson LL Shoulder arthroscopy In Johnson LI, editor Arthroscop~c surgery principles and practice St Louis Mosby-Year Book, 1986.
9.
12 13
14 15
16. 17
21
22
A
special
Professor several and I
thank Mansat hundred
publishing
you
to
Michel
left this authors staff.
Mansat,
position
in January
all over
the continent
Bonne
of anterior
of
shoul-
schulterluxatlon.
AG l’approche systkmique, la conceptlon cybern& la notlon de dlscontinult& dans la m&thologle de la blom&dlcale. Syst&mes naturels, syst&mes artlflciels. Vallon, France, Mllieux, 1986 D, lellouch J t’8ssal therapeutlque chez l’homme Paris Flammarlon Mbdeclne Sciences, 198 1
Senghas RE Statistics In the Journal of Bone and Jomt Surgery SuggestIons for authors J Bone Joint Surg 1992;74A 3 19-20
24
Snyder lesions
2.5
Townley CO The capsular mechanism in recurrent drslocation of the shoulder J Bone Joint Surg 1950,32A 370-80. Turkel SJ Stablllzlng mechanrsms preventrng anterior dislocation of the glenohumeral jornt J Bone Jolnt Surg 198 1,63A. 1208-l 7
27.
our
InstabIlIty
23
26
Johnson LL. An arthroscoplst’s perspective of anatomy and pathology. Clan Orthop 1987,223 1 13-2.5 Johnson LL Diagnostic and surgical arthroscopy of the shoulder St LOUIS Mosby-Year Book; 1993 p 300-l 1
Petrovic tlque et recherche Champ Schwartz 2nd ed
Surg I999
as well
Ferkel RD, Friedman 1990;6 274-9
RJ SLAP
Walch G, Mole D lnstablllti& et luxatlons de l’kpaule [artrculatlon glkno-hum&ale) Encycl M&d Chir App’ locomoteur 1991,14037AlO 14
faithful 1999
SJ, Karzel RP, DelPlzzo W, of the shoulder Arthroscopy
Editor with
in Europe
the appreciation
as the
members
for and
many
years.
gratitude
of
of the USES editorial
chance! I