Coracoclavicular ligament reconstruction using the lateral
half of the conjoined tendon
Sonya M. Sloan, MD,a Jeffrey E. Budoff, MD,a,b John A. Hipp, PhD,a and Lyndon Nguyen, MS,a
Houston, TX
To preserve the coracoacromial (CA) ligament, we
have used the lateral half of the conjoined tendon as
an autograft source to replace the coracoclavicular
(CC) ligament. The purpose of this study is to compare
the ultimate tensile strength of the lateral 12 mm of the
conjoined tendon with that of the CA ligament and the
CC ligament. Eight paired cadaveric male shoulders
were tested to tensile failure with a custom pneumatic
testing apparatus. Although the lateral 12 mm of the
conjoined tendon was stronger than the CA ligament,
this difference was not statistically significant (P ⫽
.37). However, the intact CC ligament (621 ⫾ 209 N)
was approximately 250% stronger than either the lat-
eral 12 mm of the conjoined tendon (265 ⫾ 79 N, P
⬍ .001) or the CA ligament (246 ⫾ 69 N, P ⬍ .001).
We believe that the lateral half of the conjoined ten-
don is a viable alternative autograft source for recon-
struction of the CC ligament in cases of symptomatic
acromioclavicular joint dislocation. Though not as
strong as the native CC ligaments, the conjoined ten-
don is slightly stronger than the commonly used CA
ligament. (J Shoulder Elbow Surg 2004;13:186-90.)
A cromioclavicular (AC) dislocations, initially de-
fined by Allman1 as a grade III AC joint separation,
occur in conjunction with disruption of the coracocla-
vicular (CC) ligament complex. The first surgical treat-
ment of an AC joint dislocation was an arthrodesis
performed by Cooper8 in 1861. Since then, surgical
reconstruction of the CC ligament complex has been
accomplished by a number of different techniques
including the use of hardware fixation from the clav-
icle to the acromion,26,28 hardware fixation from the
clavicle to the coracoid,5,17 and reconstruction with
autologous semitendinosus tendon15 or the cora-
coacromial (CA) ligament.34 Transfer of the coracoid,
From the Department of Orthopaedic Surgery, Baylor College of
Medicine,a and Houston VAMC.b
Reprint requests: Jeffrey E. Budoff, MD, 6550 Fannin St, Suite
2525, Houston, TX 77030 (E-mail: jbudoff@bcm.tmc.edu).
Copyright © 2004 by Journal of Shoulder and Elbow Surgery
Board of Trustees.
1058-2746/2004/$35.00 ⫹ 0
doi:10.1016/j.jse.2003.12.002
186
relying on the dynamic function of the distally at-
tached conjoined tendon to stabilize the distal clavi-
cle, has also been reported.4,10,12,16
Today, one of the most widely used surgical pro-
cedures for the reconstruction of a high-grade AC
joint separation is the Weaver-Dunn technique.34 Re-
ported in 1972, the Weaver-Dunn procedure uses the
CA ligament as a proximally based transfer to recon-
struct the CC ligament complex. Though often effec-
tive, this procedure involves the sacrifice of the CA
ligament.
The CA ligament functions as a secondary, static
stabilizer against anterior-superior humeral head mi-
gration.* Recent literature documents complications
associated with CA ligament resection in patients with
irreparable rotator cuff tears. Consequently, in that
situation, its preservation has been noted to be desir-
able.†
To avoid complications associated with resection
of the CA ligament, we have used the lateral half of
the conjoined tendon as a proximally based transfer
for CC ligament reconstruction, especially in cases of
concomitant rotator cuff injury and/or surgery in
paraplegics. It is also an advantageous autograft
source in revision situations in which the CA ligament
has already been harvested or in cases in which an
excessive amount of distal clavicle has been resected
and a longer length of graft tissue is required. A
similar technique was originally reported by Var-
gas32 in 1942, who credited de Carvalho with the
idea.
The conjoined tendon takes origin from the cora-
coid immediately anterior to the CA ligament. As the
musculocutaneous nerve enters the conjoined tendon
in its muscular medial side, the lateral half, which is
entirely tendinous, can be easily harvested without
putting this nerve at risk. Moreover, a conjoined
tendon graft can be harvested to be significantly
longer than the CA ligament.
A literature search failed to note any study exam-
ining the properties of the conjoined tendon as a graft
source. The purpose of this study was to compare the
ultimate tensile strength (UTS) of the intact CC liga-
*References 2, 3, 6, 7, 9, 11, 13, 20, 24, 27, 33, 35, 36.
†References 3, 7, 13, 18, 20, 27, 29, 33, 35, 36.
J Shoulder Elbow Surg
Volume 13, Number 2
Figure 1 Biomechanical testing apparatus.
ment, the CA ligament, and the lateral half of the
conjoined tendon.
MATERIALS AND METHODS
Eight fresh-frozen male cadaveric shoulders (four right
and four left) were used. The mean age of the specimens
was 65 years, with a range from 64 to 68 years. Each
specimen was free from injury and disease, with no known
musculoskeletal disorders. The specimens were fresh-fro-
zen, stored at 0°C, and thawed for 24 hours at room
temperature before testing. Dissection of each shoulder was
completed with careful attention to preserving the conjoined
tendon, the CA ligament, and the CC ligament. After the
aforementioned structures were identified, the skin, subcu-
taneous tissue, and surrounding muscles were removed.
Each specimen was embedded in an open rectangular
mold by use of Smooth Cast 320 (Smooth On Inc, Easton,
Pa), a liquid polymer. The proximal clavicles were similarly
potted, by use of 2-inch tubular polyvinyl chloride pipe. A
ferrous angle iron was secured to the potted clavicle with a
threaded screw directly over the CC ligament. The distal
clavicle was then secured by an angle iron and clamp for
mounting purposes. The specimens were continuously hy-
drated by use of 0.9% normal saline solution at room
temperature. The AC ligament was sharply transected be-
fore each experimental procedure to simulate damage
caused by an AC joint dislocation. The biomechanical
testing apparatus was modified from that used by Mota-
medi et al22 and Harris et al14 (Figure 1). Biomechanical
testing was conducted with a custom pneumatic testing
apparatus. Each specimen was fixed to an x-y floating
platform to allow self-alignment to the load that was placed
at a right angle to the x-y plane. A pneumatic cylinder
(Parker Hannifin Corporation, Cleveland, Ohio) and load
call (MUSE Measurements Corporation, Sandmas, Calif),
centered directly above the graft, applied a uniaxial load
perpendicular to the anatomic position of the clavicle at a
constant rate of 31.15 N/s (7 lb/s). To ensure that all slack
was removed from the graft, a preload of 5 N was applied
before testing. All specimens were loaded until failure.
Sloan et al 187
Experimental procedure
Intact CC ligament testing. Each embedded specimen
was secured, and the load cell was applied in the direction
of the conoid ligament. The UTS was recorded for each
intact CC ligament complex. Testing then followed for CA
ligament reconstruction and then conjoined tendon recon-
struction.
CA ligament transfer. The entire CA ligament was taken
off its insertion on the undersurface of the acromion. Its
length was measured to the nearest millimeter so that the
conjoined tendon could be cut to the same length. The CA
ligament was prepared as a proximally based transfer
attached to the coracoid as performed during the Weaver-
Dunn technique.34 No. 5 Ethibond sutures (Ethicon, Inc,
Somerville, NJ) were placed in a Bunnell-type weave with
two large Keith needles simultaneously passed through the
ligament to ensure that the needle tips did not cut the
sutures. In all reconstructions, great care was taken to start
all sutures 10 mm from the tip of the coracoid, thereby
leaving a 10-mm “free length” of unreinforced soft tissue.
The sutures were tied through drill holes in the distal clavi-
cle. The apparatus was then secured and tested as above.
Conjoined tendon transfer. Twelve millimeters or ap-
proximately half of the lateral conjoined tendon was cut to
the same length as the CA ligament and sutured with No. 5
Ethibond in an identical fashion to the CA ligament recon-
struction. The 10-mm “free length” was observed in all
specimens. The sutures were tied through drill holes in the
distal clavicle, and the apparatus was secured and tested in
the same manner as for the CA ligament reconstruction.
The UTS of all specimens was noted, as was the mode of
failure (avulsion of the tendon from the coracoid substance,
avulsion fracture of the coracoid process, or intrasubstance
soft-tissue failure).
Data analysis
The data were statistically analyzed by use of paired t
tests and 1-way analysis of variance tests.
RESULTS
The UTS for the intact CC ligament complex aver-
aged 621 ⫾ 209 N. The UTS of the CA ligament
averaged 246 ⫾ 69 N. The UTS of the conjoined
tendon averaged 265 ⫾ 79 N.
The intact CC ligament complex had a greater
ultimate tensile load to failure than both the CA liga-
ment and the lateral 12 mm of the conjoined tendon
(P ⬍ .001). The lateral 12 mm of the conjoined
tendon tended to be stronger than the CA ligament,
but this difference was not statistically significant (P ⬎
.37) (Table I, Figure 2). There was no statistically
significant difference in mode of failure between any
of the structures tested (P ⫽ .12) (Table II).
CA ligament length varied from 34 to 41 mm, with
a mean of 38 mm. CA ligament width ranged from
24 to 33 mm and averaged 27.6 mm. The length of
the lateral side of the conjoined tendon devoid of any
and all muscle fibers, proximal to its separation into
188 Sloan et al J Shoulder Elbow Surg
March/April 2004

gration, a greater length of graft source, and the

strength of the lateral 12 mm of the conjoined tendon
is at least equivalent to that of the CA ligament. No
donor-site morbidity, weakness, fatigue, cramping, or
musculocutaneous nerve injury has been experienced
in association with this technique. Although it is ac-
knowledged that violation of the CA ligament can
also be avoided by using allograft, the purpose of this
report is to present an alternative autograft for recon-
struction of the CC ligaments. Use of the lateral half of
the conjoined tendon may be considered when an
allograft is not desired.
The ultimate failure load for the intact CC ligament
(621 ⫾ 209 N) was comparable to that noted in
studies by Harris et al14 (500 ⫾ 134 N) and Mota-
medi et al22 (725 ⫾ 231 N). However, we found the
UTS of the CA ligament to be somewhat higher (246
Figure 2 Comparison of ultimate tensile failures. ⫾ 69 N) than that measured by Harris et al14 (145 ⫾
107 N).
The ultimate failure load for the conjoined tendon
Table I UTS reconstruction (265 ⫾ 79 N) was equivalent to the
strength of a CC screw measured by Harris et al14
UTS (N)
(229 ⫾ 99 N) but was less than that measured by
Structure (mean ⴞ SD)
Motamedi et al22 (390 ⫾ 254 N). It was also lower
Intact CC ligament 590.4 ⫾ 191.8 than augmentations performed with braided poly-
Conjoined tendon 265.3 ⫾ 78.9 dioxanone (677 N ⫾ 115 N) or with polydioxanone
CA Ligament 246.3 ⫾ 69.4 placed through (986 ⫾ 391 N) or around (763 ⫾
218 N) the clavicle.22
The recent trend has been toward preservation of
Table II Mode of failure the CA ligament, because of its function as a second-
ary, static stabilizer of the humeral head against
Structure Midsubstance Bone Insertional superior migration in the presence of rotator cuff
CC ligament 3 1 4
deficiency.‡ Should a patient later have a large rota-
CA ligament 3 0* 4 tor cuff tear develop, this loss of the CA ligament may
Conjoined tendon 2 2 4 predispose to anterosuperior instability and a de-
creased ability to actively elevate the arm, especially
*CA ligament was not tested because of prior bone fracture at clavicle. in those who undergo arthroplasty.3,13,20,27,33,35,36
As the prevalence of rotator cuff tears increases with
its two muscles (the long head of the biceps and age— up to approximately 51% in individuals aged
coracobrachialis), ranged from 55 to 67 mm and greater than 80 years19,21,23,25,30,31—rotator cuff
averaged 61 mm. However, the overall conjoined dysfunction is common enough that it may be antici-
tendon length proximal to the myotendinous junction pated to occur in a significant percentage of our
measured approximately 5 to 7 inches. surgical population.
When the CA ligament is being used as a graft
DISCUSSION source, obtaining adequate graft length for CC liga-
ment reconstruction can occasionally be problematic.
In this study two potential autograft sources for CC A more than ample length of graft can routinely be
ligament reconstruction were compared with regard harvested from the conjoined tendon, even in cases of
to UTS as well as length and failure type. We found prior excessive distal clavicle resection.
the lateral 12 mm of the conjoined tendon to be From a biomechanical testing standpoint, where
slightly stronger than the entire CA ligament, although one places suture into soft-tissue reconstructions is
this difference did not reach statistical significance significant, as sutures artificially reinforce the length
with the numbers tested. Clinical advantages of using of tissue in which they lie. As the free interval (the
the conjoined tendon instead of the CA ligament as a tissue length devoid of suture) becomes shorter, there
proximally based transfer for CC ligament reconstruc- is less unreinforced tissue length available to fail, and
tion include retention of the CA ligament as a passive
stabilizer against anterior-superior humeral head mi- ‡References 2, 3, 6, 7, 9, 11, 13, 20, 24, 27, 33, 35, 36.
J Shoulder Elbow Surg
Volume 13, Number 2
the structure’s UTS increases. It may be possible to
manipulate soft-tissue strength by varying the length of
the free interval. In this study the starting point for the
suture was kept a constant 10 mm from the coracoid
insertion for all conjoined tendon and CA ligament
specimens tested. No failure occurred in the length of
tissue reinforced by suture. All soft-tissue failure invari-
ably occurred within the free interval between the
coracoid and the start of suture placement.
We readily acknowledge that this study has all of
the usual weaknesses of an in vitro single pull–to–
failure study. Although failure of the CC ligament
reconstruction may occur by repetitive submaximal
loads, we believe that the UTS is indicative of the
comparative strengths of these constructs. As a prac-
tical matter, fatigue testing requires multiple speci-
mens, which may have variable material properties
and may, therefore, be inconclusive because of a
higher SD. The CC ligament and its reconstructions
were tested only in the superior direction for their
ability to stabilize the distal clavicle, but we believe
that this is the most clinically relevant direction in this
situation. This direction was chosen because it is well
recognized as the direction in which the majority of
joint dislocations occur.
As two conjoined tendons failed by bony avulsion
of the coracoid, bony failure is probably more com-
mon in older cadaveric specimens than in younger
living patients. This study may have underestimated
the UTS of the lateral half of the conjoined tendon
because of age-related changes in bone density.
These two avulsions demonstrated a mean UTS of
approximately 324 N. It is also possible that younger
patients will have different soft-tissue loading charac-
teristics than older cadaveric specimens.
Augmentation of the CC reconstruction is per-
formed to protect the biological graft during healing.
Although augmentation may make subtle differences
in graft strength clinically irrelevant, it was not used in
this study to avoid obscuring strength differences be-
tween the graft sources. Finally, any biological graft
source may serve as a scaffold for future scar forma-
tion; therefore, the UTS at time zero may be less
relevant than future tissue remodeling and/or scar
formation, which may increase future graft resistance
to mechanical loading.
We conclude that conjoined tendon autograft may
be beneficial for several reasons:
1. It is at least as strong as the CA ligament.
2. There is an ample length of graft source, even in
cases of revision involving prior overly aggressive
distal clavicle resection.
3. It avoids the sacrifice of the CA ligament, a sec-
ondary, static stabilizer against superior humeral
head migration. It may be considered as an alter-
native autograft in cases of revision surgery, in
Sloan et al 189
patients with a coexisting rotator cuff tear, or in
patients who depend on their upper extremities for
transferring and ambulation (ie, paraplegics).
REFERENCES
1. Allman F Jr. Fractures and ligamentous injuries of the clavicle and
its articulation. J. Bone Joint Surg Am 1967;49:774-84.
2. Arntz C, Frederick F, Jackins S. Surgical management of complex
irreparable rotator cuff deficiency. J Arthroplasty 1991;6:363-
70.
3. Arntz C, Jackins S, Matsen F. Prosthetic replacement of the
shoulder for the treatment of defects in the rotator cuff and the
surface of the glenohumeral joint. J Bone Joint Surg Am 1993;
75:485-91.
4. Bailey RW, O’Connor GA, Titus PD, Baril JD. A dynamic method
of repair for acute and chronic acromioclavicular disruption. Am J
Sports Med 1976;4:58-71.
5. Bosworth BM. Acromioclavicular separation: new method of
repair. Surg Gynecol Obstet 1941;73:866-71.
6. Brown T, Bigliani LU. Complications with humeral head replace-
ment. Orthop Clinic North Am 2000;31:77-90.
7. Budoff J, Nirschl R, Guidi E. Debridement of partial-thickness tears
of the rotator cuff without acromioplasty: long-term follow-up and
review of literature. J Bone Joint Surg Am 1998;80:733-48.
8. Cooper ES. New method of treating longstanding dislocations of
the scapuloclavicular articulation. Am J Med Sci 1861;41:389.
9. Curl LA, Warren RF. Glenohumeral joint stability. Selective cutting
studies on the static capsular restraints. Clin Orthop 1996;330:
54-65.
10. Dewar FP, Barrington TW. The treatment of chronic acromio-
clavicular dislocation. J. Bone Joint Surg Br 1965;47:32-5.
11. Dines D, Bisson L. Modular total shoulder arthroplasty indications,
surgical considerations and results. Semin Arthroplasty 1997;8:
292-03.
12. Ferris BD, Bhamra M, Paton EF. Coracoid process transfer for
acromioclavicular dislocations. A report of 20 cases. Clin Orthop
1989;242:184-7.
13. Field L, Dines D, Zabinski S, Warren R. Hemiarthroplasty of the
shoulder for rotator cuff arthropathy. J. Shoulder Elbow Surg
1997;6:18-23.
14. Harris R, Wallace A, Harper G. Structural properties of the intact
and the reconstructed coracoclavicular ligament complex. Am J
Sports Med 2000;28:103-8.
15. Jones H, Lemos M, Schepsis A. Salvage of failed acromioclavic-
ular joint reconstruction using autogenous semitendinosus tendon
from the knee. Am J Sports Med 2001;29:234-7.
16. Katznelson A, Nerubay S, Friedlander OC. Dynamic repair of
acromio-clavicular dislocation. Acta Orthop Scand 1975;46:
199-204.
17. Kumar S, Sethi A, Jain AK. Surgical treatment of complete acro-
mioclavicular dislocation using the coracoacromial ligament and
coracoclavicular fixation: report of a technique in 14 patients.
J Orthop Trauma 1995;9:507-10.
18. Lee T, Black A, Tibone J, McMahon P. Release of the coracoacro-
mial ligament can lead to glenohumeral laxity: a biomechanical
study. J Shoulder Elbow Surg 2001;10:68-72.
19. Lehman C, Cuomo F, Kummer F, Zuckerman J. The incidence of
full thickness rotator cuff tears in a large cadaveric population.
Bull Hosp Jt Dis 1995;54:30-1.
20. Levine W, Pollock R. Prosthetic replacement in rotator cuff-defi-
cient patient. Semin Arthroplasty 1997;8:321-7.
21. Milgrom C, Schaffler M, Gilbert S, van Holsbeeck M. Rotator-cuff
changes in asymptomatic adults. The effect of age, hand domi-
nance and gender. J Bone Joint Surg Br 1995;77:296-8.
22. Motamedi A, Blevins F, Willis M, McNally T, Shahinpoor M.
Biomechanics of the coracoclavicular ligament complex and
augmentations used in its repair and reconstruction. Am J Sports
Med 2000;28:380-4.
190 Sloan et al
23. Needell SD, Zlatkin MB, Sher JS, Murphy BJ, Uribe JW. MR
imaging of the rotator cuff: peritendinous and bone abnormalities
in an asymptomatic population. AJR Am J Roentgenol 1996;166:
863-7.
24. Nirschl RP. Rotator cuff surgery. Instr Course Lect 1989;38:447-
62.
25. Ogata S, Uhthoff HK. Acromial enthesopathy and rotator cuff
tear: a radiologic and histologic postmortem investigation of the
coracoacromial arch. Clin Orthop 1990;254:39-48.
26. Phemister DB. The treatment of dislocation of the acromioclavic-
ular joint by the open reduction and threaded-wire fixation. J Bone
Joint Surg 1942;24:166.
27. Pollock R, Deliz E, McIlveen S, Flatow E. Prosthetic replacement
in rotator cuff-deficient shoulders. J Shoulder Elbow Surg 1992;
1:173-86.
28. Post M. Current concepts in the diagnosis and management of
acromioclavicular dislocations. Clin Orthop 1985;200:234-47.
29. Salter E, Nasca R, Shelley B. Anatomical observations on the
acromioclavicular joint and supporting ligaments. Am J Sports
Med 1987;15:199-206.
J Shoulder Elbow Surg
March/April 2004
30. Sher JS, Uribe JW, Posada A, Murphy BJ, Zlatkin MB. Abnormal
findings on magnetic resonance images of asymptomatic shoul-
ders. J Bone Joint Surg Am 1995;77:10-5.
31. Tempelhof S, Rupp S, Seil R. Age-related prevalence of rotator
cuff tears in asymptomatic shoulders. J Shoulder Elbow Surg
1999;8:296-9.
32. Vargas L. Repair of complete acromioclavicular dislocation utiliz-
ing the short head of the biceps. J Bone Joint Surg 1942;24:
772-3.
33. Watson M. Major ruptures of the rotator cuff. J Bone Joint Surg Br
1985;67:618-24.
34. Weaver J, Dunn H. Treatment of acromioclavicular injuries, es-
pecially complete acromioclavicular separation. J Bone Joint Surg
Am 1972;54:1187-94.
35. Wiley A. Superior humeral dislocation. A complication following
decompression and debridement for rotator cuff tears. Clin Or-
thop 1991;263:135-41.
36. Willams G, Rockwood C. Hemiarthroplasty in rotator cuff-defi-
cient shoulders. J Shoulder Elbow Surg 1996;5:362-7.