Rotator Cuff Tear: Imaging

Introduction physicians (i.e., radiologists and orthopedic

surgeons).
The shoulder pain represents a complex problem Snyder’s arthroscopic classification of RCTs
that should be associated with several anatomical is a good example of usable MRI classification
situations in which rotator cuff tears (RCTs) and permits to evaluate RCTs also on the MR
describe the failure response to a new or chronic images on the basis of the partial bursal or articu-
biomechanical stimulus. An experienced clini- lar/full-thickness lesion, the extent of the lesion,
cian usually makes the diagnosis of RCTs, but and the number of involved tendons (Snyder
the choice of a correct surgical treatment must 2003; Millstein and Snyder 2003).
take count of the possible contributing factors,
which may be efficiently verified with MRI
(Sherman 1997). MR Sequences Technique
Findings at MR imaging may also help assess
surgery contraindications, and the radiologist Imaging of RCTs may be performed either with-
should be aware of findings that suggest neuro- out contrast (non-arthrographic) or with direct
logic anomalies of the shoulder, deltoid muscle contrast injection in the joint (direct MR arthrog-
abnormalities (e.g., atrophy), or tendon retraction raphy) or indirect ev contrast injection (indirect
that may have an effect on the return of the activ- MR arthrography). The choice of the technique
ity. The MR report should be built on the basis of may be based on the clinical suspect, because the
clinical implications of the findings, taking into conventional non-arthrographic MR showed low
account functional anatomy and the anatomic sensibility on the assessment of partial-thickness
outcome after cuff repair, and classify and define lesions. This is more evident when MR is per-
imaging findings that will assist in choosing a formed on a low-field dedicated open magnet, in
repair technique (Goutallier et  al. 2003): so the which a lower spatial resolution represents a crit-
morphological findings of an MR image have to ical limit. Improvements in MRI field strength,
be interpreted in the more wide view of biome- gradients, sequences, and coils permit at present
chanical imaging. Even if this approach is univer- time to increase the accuracy of the without con-
sally recognized, at this time a real MR trast MRI, and the exams may be considered the
classification of RCTs does not exist, and during goal standard in patients in which there is no
the years, several arthroscopic classifications instability component (direct MR arthrography
were adapted to MRI in order to avoid discrep- must be considered as primary exam) (Magee
ancy and confusion in the language used between and Williams 2006; Dinnes et al. 2003; Foad and

© Springer International Publishing AG, part of Springer Nature 2018 23

C. Chillemi et al., Arthroscopic Transosseous Rotator Cuff Repair,
https://doi.org/10.1007/978-3-319-76153-4_3
24 3  Rotator Cuff Tear: Imaging

Wijdicks 2012). The MR protocol is usually A  correct measure considers the margins of the
based on T1-, fat-suppressed proton density- lesion for its distal to the proximal side. In order to
andT2-weighted sequences. Additional imaging underestimate the lesion length, the limit of the
in the abduction external rotation (ABER) posi- proximal margin of the lesion must be chosen on
tion of the shoulder was reported to improve sen- the basis of tendon signal, as a tendon higher
sitivity, but the shape of the MR coil of some signal on T2 or PD fat saturated images may
constructors does not permit to perform correct represent a degenerated or myxoid tendon that the
positioning of the arm. surgeon will recognize only during arthroscopy
Recent studies demonstrate the usefulness of procedure (debridement for more consistent
3D sequences in the assessment of shoulder margins for repair). A correct measure procedure
lesions, above all in the labral tears (Lee et  al. permits to distinguish small (<1  cm), medium
2014; Choo et  al. 2012). At our institution we (1–3  cm), large (3–4  cm), or massive (>4  cm)
routinely use 3D sequences in the study of post- (Figs. 3.1, 3.2, and 3.3) lesions. In partial tears the
operative shoulder, because this permits to reduce identification of the actual tendon profile result is
scan time, reduce slice thickness, and obtain
images on post-processing phase evaluation in
several planes, not only oriented on glenoid axis
but also on tendons planes. In the evaluation of
preoperative shoulder, this may represent a future
approach, and the use of high-field magnet will
permit to perform this technique also without the
use of contrast (direct or indirect).

MR Imaging of Rotator Cuff Tears

MR shoulder study provides evidence about rota-

tor cuff tear depth or thickness, dimension, and
shape. Crucial other information regard the cora- Fig. 3.1 Coronal fat saturated proton density image
coacromial arch and coracohumeral space, ten- shows a small lesion as focal area of high intensity signal
don retraction, the extension to other tendon or in the insertional region of the supraspinatus tendon
structures of articular space, and muscle trophism
that may lead to a successful or unsuccessful
surgery.

Rotator Cuff Tear: Dimension

The goal of a correct MR evaluation of RCTs is to

find correspondence between arthroscopic
classification of a tear (i.e., Snyder classification)
and the morphological or signal abnormalities on
images. In our experience the slice thickness and
spatial resolution (defined as size of the imaging
voxels for field of view) represent the basis of a
well-performed MR study, above all in the Fig. 3.2 Coronal fat saturated proton density image
evaluation of partial tears. For complete lesion, the shows a medium lesion as focal area of high intensity sig-
extension of tendon loss must be measured. nal in the insertional region of the supraspinatus tendon
MR Imaging of Rotator Cuff Tears 25

a b

Fig. 3.3 (a) Coronal fat saturated proton density image lesion, but some tissue with intermediate signal is evident
shows a large lesion as wide area of high intensity signal in the lesion (arrows in a and b): this represents a scar tis-
in the pre-insertion region of the supraspinatus tendon. In sue/fibrosis of subacromial bursa that must be distin-
(b) coronal TSE T2-weighted image shows the same guished from the actual lesion (arrowheads on a and b)

critical in order to underestimate the lesion Rotator Cuff Tear: Shape

(Aliprandi et al. 2013; Waldt et al. 2007; Magee
et al. 2003). In this way the presence of articular The shape of RCTs is another important aspect
fluid represents a significant support because it for a correct choice of surgical technique. The
provides the right contrast between the low tendon geometry is classified on the basis of shape on
signal and the high signal of a recent lesion tendon surface; in particular we can find three
(articular or bursal side) that “paint” the lesion as basic shapes:
minus profile area. A correct measure in length
includes both sides of the minus area, distinguishing –– Crescent, in which the tear has a notch shape,
from minimal (<1  cm), fraying of some fibers without great retraction or partial to bone
(<2  cm), severe (<3  cm), or very severe (>4  cm healing because of minimal traction on tear
with or without flap). Some difficulties may be –– L-shaped or reverse L-shaped in which the
experienced when the lesion shows a granulation longitudinal component of the tear divides the
tissue as healing attempt. Although this tissue may tendon fibers on long axis with a transverse
fill the profile defect, its signal remains continually component orientated anteriorly or posteriorly
high on T2- or also on T1-weighted images and –– V- or U-shaped, in which the humeral head is
permits a quite correct estimation of defect length. exposed by the tendon with greater retraction
About the lesion depth, an unenhanced MR may of the lesion
underestimate the lesion, but the limit of 50% of
tendon diameter may represent a useful cutoff As the capability of the arthroscopy to rec-
value. If a precise measurement is required, an MR ognize shape and dimension of the tear dimin-
arthrography is suggested, in particular the indirect ished in massive tears, preoperative MRI may
approach (with ev contrast), because in a direct be useful to recognize also the shape. We rec-
arthrography the partial lesion of the bursal side is ommended to evaluate the RCT also on axial
not demonstrated. plane (not only coronal); this permits to compare
26 3  Rotator Cuff Tear: Imaging

the arthroscopic view with MR images and obtain poor surgical outcome (Iannotti 1994; Romeo
additional data as the integrity of coracohumeral et al. 1999). When a SLAP lesion is suspected, an
ligament or anterior supraspinatus tendon fibers MRI arthrogram may help not only to evaluate
that may be important in the surgical evaluation the anchor integrity but also a proper evaluation
of interval competence (Burkhart et al. 2001). of interval lesion in order to exclude an intermit-
tent subluxation (if need of US integration), espe-
cially with anterior supraspinatus or superior
 otator Cuff Tear: Other Structures’
R subscapularis tendon tears (Farin et al. 1995).
Involvement

When we assess a supraspinatus tendon tear on Rotator Cuff Tear: Tendon Retraction
MR, we have to consider that the rotator cuff has
a critical importance in the proper centering of The degree of tendon retraction is a crucial aspect
the humeral head in the glenoid fossa through the for surgical planning as the tendon insertion has to
passive or active action of the muscles; this is evi- be free of tension. Patte classification (Patte 1990)
dent in the anteroposterior stability provided by permits to classify the position of tendon stump
the transverse couple forces of the subscapularis near bony insertion or at glenoid level. A proper
and infraspinatus tendons. In this manner it is evaluation of tendon retraction may be performed
crucial to recognize, on presurgical MR, the after tear shape assessment because deep U-shaped
extension of the lesion on the rotator interval lesion may not represent a true retraction, but a
structures or subscapularis tendon, as this may dimension feature of the tear that may be mobile
effect both the surgical approach (deltopectoral) during arthroscopic suture (Burkhart et al. 2001).
and the decision to take a conservative approach
in “working shoulders” with isolated full-­
thickness tear of the supraspinatus tendon Rotator Cuff Tear: Muscle Atrophy
(Goutallier et  al. 2003; Aluisio et  al. 2003;
Parsons et al. 2002). To determine the extension Trophism evaluation of rotator cuff muscles with
of the lesions, the MR sagittal plane may be con- MRI is a critical information for the surgical out-
sidered as the main plane to evaluate. We suggest come because repairing a rotator cuff tear with
to perform TSE T2 weighted without fat satura- atrophic muscles will not result in recovery of
tion in order to obtain not only a proper contrast articular function (Post et al. 1983). Three meth-
between lesion areas and tendon intensity but ods may be used on preoperative MR images, to
also between interval ligaments and fat. It is estimate muscle trophism:
important to recognize a proper position of biceps
sulci centered on axial images because a partial –– Scapular ratio or Thomazeau method, in
extra or intra rotation may affect the evaluation of which, in the sagittal plane corresponding at
subscapularis tendon or biceps pulley (artifacts). the level larger supraspinatus fossa (Fig. 3.4)
We can consider topographic classification Patte (Thomazeau et  al. 1996), a ratio between
classification with six segments (Patte 1990). For cross-sectional area of the supraspinatus
the subscapularis lesions, Lafosse classification muscle and the area of the supraspinatus fossa
(Lafosse et al. 2007) permits to obtain data also (occupation ratio) may be calculated. In stage
on the involvement of biceps pulley for 1, corresponding to normal or slight atrophy,
Habermeyer classification for interval lesions occupation ratio is included from 1.00 to 0.60
(Habermeyer et al. 2004). Tear of the long head value; in stage 2, corresponding to moderate
of biceps tendon also influences the stability of atrophy, occupation ratio is included from
the shoulder and the postsurgical comfort. A 0.60 to 0.40 value; in stage 3, corresponding
proper evaluation of the tendon signal and loca- to severe atrophy, occupation ratio is lesser to
tion on presurgical MR images permits to reduce 0.40 value.
Coracoacromial and Coracohumeral Spaces 27

Fig. 3.5 Coronal fat saturated proton density image

shows superficial fraying of the bursal side of the inser-
Fig. 3.4 Sagittal T1-weighted images used for
tional region of the supraspinatus. Note the contrast
Thomazeau method in which internal area of the supraspi-
between subacromial fluid and tendon surface (arrow)
natus is divided by the external area value. In this patient
which permits the delineation of fibers disrupted
the ratio is 0.54 corresponding to moderate atrophy. Note
that the signal of muscle belly is normal without evident
fat infiltration (higher signal areas)
this space the supraspinatus glides with the help
of the subacromial-subdeltoid bursa that reduces
–– Tangent sign method, in which, on the same the friction. The space is normally virtual, and its
level of supraspinatus fossa used for cross-­ reduction causes anterior impingement. A cause
section method, a line is drawn through the of impingement may be a particular shape of the
superior borders of the scapular spine and the acromion process, as type 2 (curved downward)
superior margin of the coracoid process. A or type 3 (hooked downward anteriorly),
normal supraspinatus muscle should cross especially in association with osteophytes or
superior to the line. enthesophyte spurs from the coracoacromial
–– Goutallier’s classification using CT scan or ligament. MR images can depict the acromion
also MR scans (Goutallier et al. 1994; Fuchs types, defining the inclination on the frontal plane
et  al. 1999), in which, on the same level of (lateral downsloping) or in the sagittal plane
supraspinatus fossa used for others methods, a (anterior downsloping) (Bigliani et  al. 1991;
ratio between fatty tissue and muscle fibers is Toivonen et al. 1995; Yazici et al. 1995). When an
assessed as semiquantitative method. In impingement is suspected on clinical tests, it is
particular the muscle trophism may be staged important to remember that MR is performed
in four levels: stage 0 or normal muscle, stage with the arm in neutral position (adduction-quite
1 or some fatty streaks, stage 2 or less than internal rotated), so in order to recognize contact
50% fatty muscle atrophy, stage 3 with 50% areas with the acromion, indirect signs must be
fatty muscle atrophy, and stage 4 with more found in the images. The supraspinatus tendon
than 50% fatty muscle atrophy. profiles are the first markers of tendon integrity,
as the tendon wear appears as light irregularities
(Fig. 3.5) on humeral or bursal face. This aspect
Coracoacromial and Coracohumeral represents an ideal condition when images are
Spaces performed with high-field scanner (1.5T and
higher), but may be difficult to recognize in low-­
The superior aspect of the glenohumeral joint field scanners, as the spatial resolution (the size
includes the lateral third of the acromion at the of the imaging voxels) of images is lower. Indirect
top, the coracoacromial ligament at the middle signs as inflammatory reaction of bursa, or ten-
third, and the coracoid process at the lower third; don signal changes, may help to suspect an
this space is defined as coracoacromial arch. In impingement in different joint rotation grades.
28 3  Rotator Cuff Tear: Imaging

A practical rule may be used as help to distinguish

a normal from an abnormal tendon: the tendon
signal must be lower or quite similar to the
muscles in long repletion time sequences (as
­
T2-weighted). A raise of tendon signal represents
an abnormal orientation of the fibers in the space
for internal disruption or changes in tissutal
composition (as myxoid regress).
The same rules may be used for the identifica-
tion of the abnormalities of subscapular tendon
or in general for the evaluation of rotator interval.
This space is crucial for shoulder stability, and
MRI shows lower sensibility in assessment of
lesions in the same way of Walch’s “hidden”
lesions during arthroscopy (Walch et al. 1998). In
the same way of coracoacromial impingement,
an interval lesion must be indirectly suspected on
Fig. 3.6  Sagittal T2-weighted images show superior gle-
MRI observing in sagittal planes the coracohu- nohumeral ligament lesion confident with Habermeyer
meral ligament, the glenohumeral ligament, and group II interval lesion
the biceps tendon. The same rule of the raised
signal on T2-weighted applied on tendon or liga-
ments permits to suspect a lesion of the biceps
pulley, a partial lesion of free margin of the sub-
scapularis or supraspinatus tendons. Two classifi-
cations must be referred for lesions on this space
(mentioned above):

–– Habermeyer rotator interval lesion classifica-

tion (Habermeyer et al. 2004) with four groups
of different lesions of the subscapularis/supra-
spinatus and superior glenohumeral ligament
(Fig. 3.6)
–– Lafosse subscapularis lesion classification
(Lafosse et  al. 2007), with five types of
Fig. 3.7  Axial T2-weighted images show Lafosse III
subscapularis tendon lesion with or without subscapularis lesion (arrow) with long head biceps luxa-
biceps tendon luxation (Fig. 3.7) tion on medial side (arrowhead)
References 29

References Lafosse L, Jost B, Reiland Y, Audebert S. Bruno Toussaint

and Reuben Gobezie. J Bone Joint Surg Am.
2007;89:1184–93.
Aliprandi A, Sdao S, Cannaò PM, Khattak YJ, Longo
Lee YH, Kim AH, Suh JS.  Magnetic resonance visual-
S, Sconfienza LM, Sardanelli F.  Imaging of shoul-
ization of surgical classification of rotator cuff tear:
der pain in overhead throwing athletes. Sport Sci
comparison with three-dimensional shoulder mag-
Health. 2013;9:81–8. https://doi.org/10.1007/
netic resonance arthrography at 3.0 T.  Clin Imaging.
s11332-013-0151-z.
2014;38(6):858–63. https://doi.org/10.1016/j.
Aluisio FV, Osbahr DC, Speer KP.  Analysis of rotator
clinimag.2014.07.003.
cuff muscles in adult human cadaveric specimens. Am
Magee T, Williams D. 3.0-T MRI of the supraspinatus ten-
J Orthop. 2003;32(3):124–9.
don. AJR Am J Roentgenol. 2006;187:881–6.
Bigliani LU, Ticker JB, Flatow EL, Soslowsky LJ, Mow
Magee T, Shapiro M, Williams D.  Comparison of high-­
VC.  Relationship of acromial architecture and dis-
field strength versus low-field-strength MRI of the
eases of the rotator cuff [in German]. Orthopade.
shoulder. AJR Am J Roentgenol. 2003;181:1211–5.
1991;20(5):302–9.
Millstein ES, Snyder SJ.  Arthroscopic management
Burkhart SS, Danaceau SM, Pearce CE Jr. Arthroscopic
of partial, full thickness, and complex rotator cuff
rotator cuff repair: analysis of results by tear size
tears: indications, techniques, and complications.
and by repair technique—margin convergence ver-
Arthroscopy. 2003;19(Suppl 1):189–99. https://doi.
sus direct tendon-to-bone repair. Arthroscopy.
org/10.1016/j.arthro.2003.10.033].
2001;17(9):905–12.
Parsons IM, Apreleva M, Fu FH, Woo SL. The effect of
Choo HJ, Lee SJ, Kim OH, Seo SS, Kim
rotator cuff tears on reaction forces at the glenohu-
JH.  Comparison of three-dimensional isotropic
meral joint. J Orthop Res. 2002;20(3):439–46.
T1-weighted fast spin-echo MR arthrography with
Patte D. Classification of rotator cuff lesions. Clin Orthop
two-dimensional MR arthrography of the shoul-
Relat Res. 1990;(254):81–6.
der. Radiology.  2012;262(3):921–31. https://doi.
Post M, Silver R, Singh M.  Rotator cuff tear: diagno-
org/10.1148/radiol.11111261.
sis and treatment. Clin Orthop Relat Res. 1983;
Dinnes J, Loveman E, McIntyre L, Waugh N. The effec-
(173):78–91.
tiveness of diagnostic tests for the assessment of
Romeo AA, Hang DW, Bach BR Jr, Shott S. Re- pair of
shoulder pain due to soft tissue disorders: a systematic
full thickness rotator cuff tears: gender, age, and other
review. Health Technol Assess. 2003;7:iii.
factors affecting outcome. Clin Orthop Relat Res.
Farin PU, Jaroma H, Soimakallio S. Medial displacement
1999;367:243–55.
of the biceps brachii tendon: evaluation with dynamic
Sherman OH.  MR imaging of impingement and rotator
sonography during maximal external shoulder rota-
cuff disorders: a surgical perspective. Magn Reson
tion. Radiology. 1995;195(3):845–8.
Imaging Clin N Am. 1997;5(4):721–34.
Foad A, Wijdicks CA.  The accuracy of magnetic reso-
Snyder S.  Arthroscopic classification of rotator cuff
nance imaging and magnetic resonance arthrogram
lesions and surgical decision making. In: Wilkins LW,
versus arthroscopy in the diagnosis of subscapularis
editor. . Philadelphia: Shoulder Arthroscopy; 2003.
tendon injury. Arthroscopy. 2012;28:636–41.
p. 201–7.
Fuchs B, Weishaupt D, Zanetti M, Hodler J, Gerber
Thomazeau H, Rolland Y, Lucas C, Duval JM, Langlais
C.  Fatty degeneration of the muscles of the rotator
F. Atrophy of the supraspinatus belly: assessment by
cuff: assessment by computed tomography versus
MRI in 55 patients with rotator cuff pathology. Acta
magnetic resonance imaging. J Shoulder Elb Surg.
Orthop Scand. 1996;67(3):264–8.
1999;8(6):599–605.
Toivonen DA, Tuite MJ, Orwin JF.  Acromial structure
Goutallier D, Postel JM, Bernageau J, Lavau L, Voisin
and tears of the rotator cuff. J Shoulder Elb Surg.
MC. Fatty muscle degeneration in cuff ruptures. Pre-
1995;4(5):376–83.
and postoperative evaluation by CT scan. Clin Orthop
Walch G, Nove’-Josserand L, Boileau P, Levigne
Relat Res. 1994;304:78–83.
C.  Subluxations and dislocations of the tendon of
Goutallier D, Postel JM, Zilber S, Van Driessche
the long head of the biceps. J Should Elbow Surg.
S. Shoulder surgery: from cuff repair to joint replace-
1998;7(2):100–8.
ment an update. Joint Bone Spine. 2003;70(6):422–32.
Waldt S, Bruegel M, Mueller D, Holzapfel K, Imhoff AB,
Habermeyer P, Magosch P, Pritsch M, Scheibel MT,
Rummeny EJ, Woertler K. Rotator cuff tears: assess-
Lichtenberg S.  Anterosuperior impingement of
ment with MR arthrography in 275 patients with
the shoulder as a result of pulley lesions: a pro-
arthroscopic correlation. EurRadiol. 2007;17:491–8.
spective arthroscopic study. J Should Elbow Surg.
Yazici M, Kopuz C, Gulman B.  Morphologic variants
2004;13(1):5–12.
of acromion in neonatal cadavers. J Pediatr Orthop.
Iannotti JP.  Full-thickness rotator cuff tears: factors
1995;15(5):644–7.
affecting surgical outcome. J Am Acad Orthop Surg.
1994;2(2):87–95.