Fatty Infiltration and Atrophy

of the Rotator Cuff Do Not Improve

After Rotator Cuff Repair and
Correlate With Poor Functional Outcome
James N. Gladstone,*† MD, Julie Y. Bishop,‡ MD, Ian K.Y. Lo,§ MD, and Evan L. Flatow,† MD
†
From the Shoulder Service, Department of Orthopaedic Surgery, Mount Sinai Medical Center,
‡
New York, New York, The Ohio State University, Department of Orthopaedic Surgery,
§
Columbus, Ohio, and the University of Calgary, Department of Surgery, Calgary,
Alberta, Canada

Background: The role of degenerative changes in rotator cuff musculature with respect to the functional outcomes of rotator
cuff repair have only recently been recognized and are still not well understood. In addition, the reversibility of these changes
with repair of the tendons is questionable.
Hypothesis: Poorer preoperative muscle quality negatively affects outcome, and a successful outcome (in terms of a healed
repair) might demonstrate improvements in fatty infiltration and muscle atrophy.
Study Design: Cohort study; Level of evidence, 2.
Methods: Thirty-eight patients (mean age, 62 years) were prospectively evaluated with preoperative and 1-year postoperative
clinical examination and appropriate magnetic resonance image sequencing to determine grades of muscle atrophy and fatty
infiltration of the supraspinatus and infraspinatus muscles. American Shoulder and Elbow Society (ASES), Constant, and pain
scores were determined as well as strength measurements. The retear rate and progression of muscle degeneration were also
evaluated. Independent predictors of outcome measurements and cuff integrity were determined.
Results: The overall clinical outcome, including ASES, Constant, and pain scores, improved significantly (P < .0001). Strength in for-
ward elevation improved significantly (P < .006), while external rotation strength did not. There was a strongly negative correlation
between muscle quality and outcome results in most cases. When the results were adjusted for multivariate effect, muscle atrophy
and fatty infiltration of the infraspinatus muscle were the only independent predictors of ASES and Constant scores (P < .03). Tear
size and rotator cuff healing did not play an independent role. Tear size, however, was the only independent predictor of ultimate cuff
integrity (P = .002). Both atrophy and fatty infiltration progressed significantly over the course of the study. In cases in which the ten-
don had re-torn, the progression was found to be more significant than when the repair proved successful (P < .003).
Conclusion: Muscle atrophy and fatty infiltration of the rotator cuff muscles, particularly of the infraspinatus, play a significant
role in determining functional outcome after cuff repair. Tear size appears to have the most influential effect on repair integrity. A
successful repair did not lead to improvement or reversal of muscle degeneration and a failed repair resulted in significantly more
progression. In general, healed repairs demonstrated minimal progression. These findings suggest that repairs should be per-
formed, if possible, before more significant deterioration in the cuff musculature in order to optimize outcomes, and that under-
standing the degree of muscle atrophy and fatty infiltration before surgery can help guide patient expectations.
Keywords: rotator cuff tears; rotator cuff repair outcomes; muscle atrophy; fatty infiltration

*Address correspondence to James N. Gladstone, MD, Department of

Orthopaedic Surgery, Mount Sinai Medical Center, 5 East 98th Street, Box
1188, New York, NY 10029 (e-mail: james.gladstone@mountsinai.org).
No potential conflict of interest declared.
The American Journal of Sports Medicine, Vol. 35, No. 5
DOI: 10.1177/0363546506297539
© 2007 American Orthopaedic Society for Sports Medicine
Rotator cuff repair is a successful procedure, both objectively
and subjectively, with regard to pain relief and functional
outcome. This appears to be the case even when the repair
itself subsequently fails. However, investigators have docu-
mented significantly better functional results when the
repair remains intact.7,11,12,18 Although variables such as
tear size, number of tendons involved, tear chronicity, and

719
720 Gladstone et al
muscle quality (fatty infiltration and muscle atrophy) have
been implicated as predictors of outcome and repair
integrity, relatively few studies have analyzed the direct
effect of these variables on outcomes. Goutallier et al9 were
one of the first groups to demonstrate a highly negative cor-
relation between fatty infiltration of the infraspinatus mus-
cle and outcome. Thomazeau et al18 and Gerber et al7 showed
a direct correlation between increasing muscle atrophy and
retearing of a repair. In these 2 studies, atrophy had a chance
to improve if the repair remained healed. Fatty infiltration,
on the other hand, never regressed, even with a successful
repair. To our knowledge, no study has used a regression
analysis model to determine the independent effect of vari-
ables such as initial tear size, muscle quality, and tendon
integrity on outcome.
In this study, we analyzed the effect of muscle quality on
measures of functional outcome including American
Shoulder and Elbow Society (ASES) and Constant scores,
pain, strength, and the structural integrity of the repair. In
addition, we evaluated the progression (or regression) of
fatty infiltration and muscle atrophy during the study
interval. We hypothesized that poorer preoperative muscle
quality would negatively affect outcome, and that a suc-
cessful outcome (in terms of a healed repair) might demon-
strate improvements in fatty infiltration and muscle
atrophy. A clearer understanding of these factors would
allow better preoperative counseling and decision-making
as to the most appropriate treatment plan.
MATERIALS AND METHODS
Patient Enrollment
Thirty-eight patients were enrolled in this study. The
patients were a subset of a larger prospective, Institu-
tional Review Board–approved study using postoperative
MRI to evaluate rotator cuff integrity at a minimum of 1
year after repair. The larger study was conducted to eval-
uate the functional outcome and repair integrity in open
rotator cuff repairs compared with arthroscopic repairs.3
Patients were included in this study if both the preoper-
ative and postoperative MRI scans permitted appropriate
evaluation of fatty infiltration and, in particular, were
medial enough on the T1-weighted sagittal oblique view
to allow determination of muscle atrophy. All patients
had full-thickness rotator cuff tears diagnosed by MRI
preoperatively and then confirmed at the time of surgery.
Patients had either an open rotator cuff repair (15
patients) or an arthroscopic repair (23 patients) per-
formed by one surgeon (E.L.F.), with the open subset
occurring before the use of arthroscopic repair tech-
niques. Patients with concomitant disorders such as
glenohumeral arthritis, fracture, or osteonecrosis were
excluded. All patients had complete clinical and MRI
examinations preoperatively and at minimum follow-up
of 1 year (range, 12 to 15 months), which was part of the
larger study protocol.
The American Journal of Sports Medicine
Outcome Assessment
The preoperative and postoperative evaluation consisted of
a patient-based questionnaire and physical examination
performed by an independent examiner (our shoulder fel-
low). The examination included measurement of range of
motion and strength testing with the use of a dynamome-
ter in both forward elevation and external rotation.
Forward elevation strength was assessed with the
patient’s arm abducted 90° in the scapular plane, with the
elbow extended and forearm pronated. Resistance was
applied at the wrist and the measurement recorded in
pounds. External rotation was measured with the arm at
the side, the elbow flexed 90°, and the wrist in neutral
rotation. Resistance was applied at the wrist and meas-
ured in pounds. These evaluations allowed the calculation
of several standard shoulder-specific scores, including the
Constant score, the ASES survey, and a visual analog scale
(VAS) pain score.
Radiographic Assessment
Because many of the patients brought their MRI scans to
their first office visit, only the postoperative MRI was
standardized. The protocol for the postoperative MRI con-
sisted of coronal images in both T1 and inversion recovery
(IR) format with 3-mm slices, sagittal oblique images in IR
format with 3-mm slices, and axial images in T1 format
with 4-mm slices. Rotator cuff integrity, fatty infiltration,
and muscular atrophy were evaluated and graded by 2
independent examiners and, if necessary, a consensus
agreement was reached. Any tendon defect on the T2-
weighted sequence that filled with fluid was considered
a retear. No attempt was made to compare defect size at
follow-up with the preoperative tear size.
Fatty infiltration was evaluated using the criteria estab-
lished by Goutallier et al9 on the basis of fatty streaks within
the muscle belly. This classification system, originally
described for CT, was applied to the T1-weighted coronal MRI
sequences in this study: grade 0, no fat; grade 1, thin streaks
of fat; grade 2, less fat than muscle; grade 3, equal amounts of
fat and muscle; and grade 4, more fat than muscle. Previous
studies have found that this staging system can be appropri-
ately applied to MRI.5,20 Additionally, for some of the statisti-
cal analyses, the grades were dichotomized into none to mild
fatty infiltration (grades 0 and 1) and moderate to severe fatty
infiltration (grades 2 through 4). This dichotimization has
clinical implications because studies have indicated that
grades 0 or 1 are found in both unaffected shoulders as well
as shoulders with rotator cuff tears, while grades 2 or higher
appear only in patients with rotator cuff tears.16,18,21
Muscle atrophy was graded according to the scale pro-
posed by Warner et al20 (Figure 1), based on an oblique
sagittal plane image where the coracoid and scapular
spine meet the scapular body. On this view, both the
supraspinatus and infraspinatus muscle can be seen.
Atrophy was graded as none, mild, moderate, or severe. For
Vol. 35, No. 5, 2007
Figure 1. Grading for muscle atrophy based on a T1-weighted
oblique sagittal plane magnetic resonance image. The amount
of muscle above or below a line drawn from the tip of the cora-
coid to the tip of the scapular spine determines the grade.
(Reprinted from Warner JJP, Higgins L, Parsons IM 4th, Dowdy
P. Diagnosis and treatment of anterosuperior rotator cuff tears.
J Shoulder Elbow Surg. 2001;10:37-46 with permission from
The Journal of Shoulder and Elbow Surgery Board of Trustees.)
statistical analysis, none was given a score of 0; mild, 1; mod-
erate, 2; and severe, 3. As part of the analysis, the grades
were also dichotomized to none/mild (grades 0 and 1) and
moderate/severe (grades 2 and 3).
Statistics
All statistical analyses were performed by a biostatistician.
Analysis was performed using Spearman’s correlation coef-
ficient to determine correlation of muscle atrophy and fatty
infiltration to outcome measures, strength, pain, tear size,
and cuff integrity. To determine independent predictors of
these measures, a stepwise linear regression analysis was
used. The effect of variables on cuff integrity was deter-
mined using a stepwise logistic regression analysis. A χ2 test
was used to examine muscle degeneration with respect to
healed and retorn tendons. Progression of fatty infiltration
and muscle atrophy was evaluated using the Wilcoxon
signed rank test. The McNemar test analyzed progression
based on dichotomized preoperative grades. In all analyses,
P < .05 was considered statistically significant.
Surgical Technique: Mini-Open
or Open Rotator Cuff Repair
Our surgical technique for open rotator cuff repairs has
been previously described2 and will not be restated except
as it relates to this study. Whether using the mini-open or
open technique, the rotator cuff tendon was repaired in a
similar fashion in all cases. Tear size was measured with a
ruler. Bursal, capsular, and interval releases for tendon
mobilization were performed with a Mayo scissors in an
attempt to obtain a repair without tension.
Once adequately mobilized, the rotator cuff was repaired
to the prepared greater tuberosity with sutures placed in a
modified Mason-Allen fashion and passed through bone
tunnels reinforced with Cuff Link bone tunnel devices
Fatty Infiltration and Atrophy After Rotator Cuff Repair 721
(Mitek Worldwide, Westwood, Mass) when bone quality
was compromised.
All patients used a sling for 6 weeks and were allowed
only passive range of motion during this time period. At 6
weeks, gradual full active motion was instituted, progress-
ing to resistive strengthening, which was continued for a
total of 3 to 4 months.
Surgical Technique: Arthroscopic Rotator Cuff Repair
Our technique for arthroscopic rotator cuff repair closely
parallels our open technique with respect to capsular and
interval releases for tendon mobilization.4 A radiofre-
quency probe was used to perform the arthroscopic
releases. Medial retraction, as well as the length of the tear
from anterior to posterior, was measured with use of a cal-
ibrated probe.
Once the rotator cuff was adequately mobilized, the ten-
don was repaired to the prepared greater tuberosity with
suture anchors placed laterally. The number of anchors
and sutures used depended on the size of the rotator cuff
tear. A medially placed tendon-transfixing device
(CuffTack, Mitek Worldwide or Suretac, Smith & Nephew,
Andover, Mass) was used to supplement fixation. We no
longer use these devices and now use 2 rows of anchors,
medial and lateral. An attempt was always made to fix the
cuff tendon to the lateral aspect of the tuberosity, if tension
allowed, to maximize the surface area for healing.
Postoperative care was similar to that described above
for open and mini-open repairs.
RESULTS
Overall Outcomes
Fifteen patients underwent open rotator cuff repair and 23
underwent arthroscopic rotator cuff repair, for a total of 38
patients. All patients were followed up a minimum of 1
year postoperatively (range, 12 to 15 months) with MRI
and a clinical examination. The mean patient age was 62
years (range, 36 to 78 years). The mean symptom duration
was 10.5 months (range, 2 weeks to 51 months). The aver-
age tear size was 2.59 cm (range, 0.5 to 5 cm).
There was a statistically significant improvement in the
ASES (from 47.3 ± 20.4 to 88.2 ± 13.8), Constant (58 ± 16.7
to 82 ± 13.8), and pain (5.6 ± 2.8 to 0.86 ± 1.7) scores (all
values mean ± standard deviation) in all patients (P <
.0001 for all). Forward elevation strength improved signif-
icantly from 7.65 ± 6.4 lb to 13.5 ± 8.1 lb (P < .0006) after
cuff repair, while external rotation strength increased, but
not significantly (P = .01), from 12.1 ± 8.9 lb to 15 ± 8.4 lb
(Table 1).
Correlation of Functional Outcome to Muscle Quality
Table 2 summarizes the correlation of fatty infiltration and
muscle atrophy with outcomes. These values are presented
as an overview. However, it must be noted that they are
722 Gladstone et al The American Journal of Sports Medicine

TABLE 1 repair, regardless of the degree of muscle atrophy or fatty

Functional Outcome Resultsa infiltration.
To determine independent predictors of functional outcome,
Measurement Preoperative Postoperative P value a stepwise linear regression analysis was performed. This
model takes into account the potential effect of the analyzed
ASES score 47.3 88.2 .0001
Constant score 58 82 .0001
variables on each other. The variables analyzed were fatty
Pain score 5.6 0.9 .0001 infiltration of the supraspinatus and infraspinatus, muscle
Forward elevation 7.7 13.5 .0006 atrophy of the supraspinatus and infraspinatus, tear size, and
strength (lb) whether the tendon had healed (intact or retorn). Table 3
External rotation 12.1 15 .01 summarizes the results. Infraspinatus atrophy is an inde-
strength (lb) pendent predictor of ASES score (P = .001), Constant score
a
(P = .033), and forward elevation strength (P = .04).
ASES = American Shoulder and Elbow Society. Supraspinatus atrophy independently predicts forward eleva-
tion (P = .013). Infraspinatus fatty infiltration independently
unadjusted results; that is to say, they do not take into predicted ASES score (P = .01) and external rotation strength
account the effect one has on the other. For example, in the (P = .004). Supraspinatus fatty infiltration also independently
unadjusted model, fatty infiltration of the infraspinatus affected external rotation strength (P = .04). Using this
(r = –.359, P < .03) and muscle atrophy of the supraspinatus regression analysis model, tear size and postoperative healing
(r = –0.402, P < .02) both appear to negatively influence the of the rotator cuff repair were not found to be independent
Constant score. This does not account for the interrelated- predictors of functional outcome or strength.
ness of these variables. To determine independent predictors
of outcome, a regression analysis is required (see below). Effect of Muscle Quality and
In almost every case, atrophy and fatty degeneration have Tear Size on Cuff Integrity
a strongly negative effect on assessment score, strength
parameters, and cuff repair integrity. There is also a strongly Overall, 61% of the rotator cuff repairs had healed at follow-
positive correlation with initial tear size. Table 2 shows that up and 39% were retorn. Preoperative variables of fatty
supraspinatus fatty infiltration did not affect the outcome infiltration and muscle atrophy were analyzed to determine
measures of ASES (r = –.231, P = .17) and Constant (r = –.236, their effect on postoperative cuff integrity.
P = .16) scores, while the 3 other variables, supraspinatus and Preoperative muscle atrophy (P = .001) and fatty infiltra-
infraspinatus muscle atrophy as well as infraspinatus fat, cor- tion (P = .008) of the supraspinatus and atrophy of the
relate significantly. There was no relationship between mus- infraspinatus (P = .041) were significantly correlated to post-
cle quality and pain scores (P > .2 in all cases); pain improved operative cuff integrity. Atrophy of the supraspinatus aver-
significantly in all cases with the performance of a rotator cuff aged 0.65 (range, 0 to 3, on a 4-point scale) in patients with

TABLE 2
Correlation of Functional Outcome, Tear Size, and Symptom Duration to Muscle Qualitya

ASES Score Constant Score Strength ER Strength FE

Muscle quality P value corr coeff P value corr coeff P value corr coeff P value corr coeff

Supraspinatus fat .17 –0.231 .16 –0.236 .007 –0.457 .012 –0.415
Infraspinatus fat .027 –0.364 .029 –0.359 .0001 –0.637 .001 –0.546
Supraspinatus atrophy .034 –0.354 .015 –0.402 .005 –0.482 .0001 –0.599
Infraspinatus atrophy .014 –0.401 .006 –0.44 .0001 –0.58 .001 –0.6

Symptom Duration Cuff Integrity Tear Size

P value corr coeff P value corr coeff P value corr coeff

Supraspinatus fat .3 –0.2 .005 –0.457 .0001 0.745

Infraspinatus fat .086 –0.33 .076 –0.3 .0001 0.712
Supraspinatus atrophy .057 –0.371 .001 –0.53 .0001 0.819
Infraspinatus atrophy .035 –0.399 .01 –0.425 .0001 0.651
a
Spearman’s rho correlation coefficients (corr coeff) and corresponding P values are listed to show any correlations between preoperative
fatty infiltration and muscular atrophy and the preoperative rotator cuff tear size and patient’s symptom duration, the postoperative
American Shoulder and Elbow Society (ASES), Constant, and strength scores, as well as postoperative rotator cuff integrity. Significant cor-
relations are in bold. ER, external rotation; FE, forward elevation.
Vol. 35, No. 5, 2007 Fatty Infiltration and Atrophy After Rotator Cuff Repair 723

TABLE 3
Independent Predictors of Functional Outcome, P Valuesa

Supraspinatus Infraspinatus

Outcome Atrophy Infiltration Atrophy Infiltration Tear Size Cuff Integrity

ASES score - - P = .001 P = .01 - - - -

Constant score - - P = .033 - - - - -
FE strength P = .013 - P = .04 - - - - -
ER strength - P = .04 - P = .004 - - - -
a
Results of stepwise linear regression analysis to determine independent predictors of functional outcome. Tear size and cuff integrity at
follow-up were not independent predictors of any of the outcome measures. ASES, American Shoulder and Elbow Society; FE, forward ele-
vation; ER, external rotation.

TABLE 4 100
Correlation of Muscle Quality and 90
Cuff Integrity at Follow-upa 80 70 67
70

Re-tear Rate%
Tendon Repair
60
50
Muscle Healed Retorn P value
40 29
Supraspinatus MA 0.65 1.73 .002 30 22
FI 0.57 1.73 .007 20
Infraspinatus MA 0.30 0.87 .012 10
FI 0.39 0.87 .076 0
None/Mild Mod/Severe None/Mild Mod/Severe
a FI FI MA MA
Comparison of preoperative muscle atrophy (MA) and fatty
infiltration (FI) in the group with a healed rotator cuff repair com- Figure 2. Effect of clinically insignificant versus significant mus-
pared with the retorn group. cle atrophy (MA) and fatty infiltration (FI) on retear rate (P < .05).

an intact cuff at follow-up versus 1.73 (range, 0 to 3) in those

with a retear (P = .002). Infraspinatus atrophy averaged 0.30
(range, 0 to 3) in the intact group, while those with a retear
averaged 0.87 (range, 0 to 3; P = .012). Fatty infiltration of the
supraspinatus was similarly greater in the retear group
(1.73; range, 0 to 4, on a 5-point scale) compared with those
with an intact cuff repair (0.57; range, 0 to 4; P = .007).
Although fatty infiltration of the infraspinatus was greater
in the retorn group (0.87; range, 0 to 4 vs 0.39; range, 0 to 4),
this only approached significance (P = .076) (Table 4).
The magnitude of effect on cuff integrity is highlighted
when the variables are further analyzed in terms of clini-
cally significant (> grade 2) versus clinically insignificant
(grades 0 to 1) preoperative muscle degeneration. A statis-
tically significant (P < .05) difference in healing rate was
found between the dichotomized groups. Of the 28 patients
with clinically insignificant fatty infiltration, 8 (29%) had
retorn, while 7 of the 10 patients with clinically significant
fatty infiltration had not healed (70%). With respect to pre-
operative muscle atrophy, those with no to minimal atro-
phy failed in 22% of cases (5 of 23 patients), whereas the
failure rate with moderate to severe atrophy increased to
67% (10 of 15 patients) (Figure 2).
Interestingly, a stepwise logistic regression analysis,
including the variables of muscle atrophy, fatty infiltra-
tion, and tear size, demonstrated that tear size (P = .002)
was the only independent predictor of postoperative rota-
tor cuff integrity.
Progression
No reversal of muscle atrophy or fatty infiltration was noted
during the time of this study. In fact, in all but 1 case, muscle
degeneration either remained the same or progressed. In each
category, supraspinatus atrophy (P < .001) and fatty infiltra-
tion (P = .001) as well as infraspinatus atrophy (P =0.015) and
fatty infiltration (P = .001), the progression was statistically
significant (Table 5). The greater the preoperative degree of
muscle degeneration, the more progression was noted at the
time of follow-up.
Twenty-three patients had no or mild muscle atrophy
and 8 of them (35%) progressed (7 progressed 1 grade and
1 patient progressed 2 grades). Of the 15 patients who had
moderate to severe atrophy preoperatively, 10 (67%) pro-
gressed from grade 2 to grade 3 (Figures 3, 4, and 5). Four
patients progressed in the supraspinatus, 2 in the infra-
spinatus, and 4 in both. The only exception was in 1 patient
whose atrophy improved from a moderate to mild grade
(grade 2 to 1) with a successful repair of the supraspinatus.
When we looked at the progression of fatty infiltration in
individual patients, none showed an improvement during the
preoperative to postoperative period. Of the 28 patients with
no or mild fatty infiltration, 7 (25%) progressed. Of the 10
patients with moderate to severe fatty infiltration preopera-
tively, 7 (70%) progressed (4 in both the supraspinatus and
infraspinatus and 3 in the infraspinatus alone). Four
724 Gladstone et al The American Journal of Sports Medicine

TABLE 5
Progression of Fatty Infiltration and Muscle Atrophy Overall, and Differentiating Those With
an Intact Versus Retorn Rotator Cuff

P value for Degree

of Progression,
Overall Healed Retorn Retorn vs. Intacta

Supraspinatus fatty infiltration Preoperative 1.03 0.57 1.73 .003

Postoperative 1.61 0.86 2.67
P value .001 .06 .007
Degree of progression 0.58 0.29 0.94
Infraspinatus fatty infiltration Preoperative 0.58 0.39 0.87 .001
Postoperative 1.28 0.52 2.33
P value .001 .16 .003
Degree of progression 0.7 0.13 1.46
Supraspinatus atrophy Preoperative 1.08 0.65 1.73 .31
Postoperative 1.47 0.79 2.33
P value .001 .02 .007
Degree of progression 0.39 0.14 0.60
Infraspinatus atrophy Preoperative 0.53 0.27 0.86 .001
Postoperative 0.53 0.30 1.67
P value .02 1.00 .014
Degree of progression 0.33 -0.05 0.80
Average tear size (cm) 2.8 2.1 3.8 .001
a
P value < .05 represents a significantly greater degree of progression in the retorn group than in the intact group.

100 Association of Fatty Infiltration,

90 Muscle Atrophy, and Tear Size
80 70 67
Progression %

70 The amount of preoperative fatty infiltration of the

60 supraspinatus correlated highly with the amount of preop-
50 35
erative muscle atrophy of the supraspinatus (r = .714, P <
40
25
.001) and the infraspinatus (r = .422, P < .009), and the
30 amount of fatty infiltration in the infraspinatus (r = .573,
20 P < .0001). Preoperative fatty infiltration of the infra-
10 spinatus correlated highly with muscle atrophy of the
0 infraspinatus (r = .881, P < .001), and also with atrophy of
None/Mild Mod/Severe None/Mild Mod/Severe the supraspinatus (r = .575, P < .0001). Finally, the amount
FI FI MA MA
of preoperative muscle atrophy of the supraspinatus and
Figure 3. Effect of clinically insignificant versus significant infraspinatus correlated significantly with each other (r =
muscle atrophy (MA) and fatty infiltration (FI) on progression .551, P < .0001).
(P < .05). Tear size correlated strongly with fatty infiltration of the
supraspinatus (r = .673, P < .001) and infraspinatus (r =
.728, P < .001) as well as with muscle atrophy of the
supraspinatus (r = .807, P < .001) and infraspinatus (r =
patients progressed 1 grade, while the other 3 progressed
.666, P < .001). When fatty infiltration and muscle atrophy
by 2 grades.
were dichotomized, there was a significant difference in
Overall, the amounts of fatty infiltration and muscular
mean tear size between the none to mild group and the
atrophy progressed regardless of cuff integrity. Although
moderate to severe group in each case.
the progression of fatty infiltration of both the supraspina-
tus and infraspinatus and atrophy of the supraspinatus
were found to be minimal when the cuff remained intact, DISCUSSION
there was no overall reversal. Nonetheless, progression
was significantly less in those tendons that healed than in Rotator cuff repairs typically lead to clinical improvement
those in which repair failed (atrophy of the infraspinatus, for patients from both a pain and functional standpoint,
P = .001; fatty infiltration of the supraspinatus P = .003 regardless of the integrity of the repair at follow-up.12
and infraspinatus P = .001). This did not apply in the case Nonetheless, numerous studies have shown even better
of supraspinatus muscle atrophy (P = .31) (Table 5). outcomes with a maintained repair.6,7,11,12,18 Historically,
Vol. 35, No. 5, 2007 Fatty Infiltration and Atrophy After Rotator Cuff Repair 725

Figure 4. Intact repair with no progression at 1-year follow-up. A, preoperative T1-weighted MRI coronal sequence through the
supraspinatus demonstrating grade 2 fatty infiltration. B, preoperative T1-weighted MRI sagittal sequence demonstrating mini-
mal muscle atrophy of the supraspinatus. C, no progression of supraspinatus fatty infiltration postoperatively. D, no change in
muscle atrophy postoperatively.

the reasons for a poorer outcome have focused on factors
such as retear rates, chronicity, and size of tear. Although
these factors are important, the inherent quality of the
muscle whose tendon is to be repaired may play as criti-
cal a role in determining the ultimate outcome of a cuff
repair. More recent studies have begun to analyze the
effect of fatty infiltration and atrophy of the rotator cuff
musculature on functional outcome after rotator cuff
repair.7,9,12,14,18,20 These studies have generally focused on
the effect of either fatty infiltration9 or muscle atro-
phy7,18,20 without considering the 2 together, or performing
a regression analysis to distinguish the independent pre-
dictors of functional outcome and repair integrity. In addi-
tion, if muscle degeneration is a determining factor in
functional outcome, its regression or progression with a
cuff repair would be of importance to document. Our study
sought not only to analyze the effect of muscle quality on
functional outcomes and structural integrity of the repair,
but also to determine the progression, or regression, of
fatty infiltration and atrophy in a group of prospectively
followed rotator cuff repairs. An enhanced understanding
of these factors may allow for better preoperative counsel-
ing of patients as to expected functional results, and per-
haps a clearer indication as to which severely involved
tendons should have a repair be attempted.
This study demonstrates, as have others,8-10,13,14 the crit-
ical role muscle quality plays in the outcome of rotator cuff
repairs. There is a strongly negative correlation between
increasing muscle degeneration and poorer functional
outcomes in both assessment scores and strength meas-
urements (Table 2). More interesting, perhaps, is the find-
ing that when a regression analysis model is used, both
fatty infiltration and muscle atrophy come out as inde-
pendent predictors of outcome, apparently overriding the
effect of tear size or repair integrity, both of which were
entered into the analysis as variables. While our analysis
and others15,17,18 have shown a very strong relationship
between fatty infiltration, muscle atrophy, and tear size,
these results demonstrate that they cannot be considered
interchangeable. Furthermore, these findings stress the
importance of appropriately performed MRI to best assess
the muscle degeneration in rotator cuff tears preoperatively.
726 Gladstone et al The American Journal of Sports Medicine

Figure 5. Failure of repair at 1 year with progression of mus-

cle changes. A, T2-weighted coronal sequence demonstrating
retear of the supraspinatus. B, T1-weighted coronal sequence
showing no preoperative fatty infiltration of the supraspinatus.
C, T1-weighted sagittal sequence showing mild supraspinatus
atrophy preoperatively. D, postoperative supraspinatus with
progression to grade 3 fatty infiltration. E, postoperative
supraspinatus with progression to moderate muscle atrophy.

Appropriate MRI requires a T1-weighted sequence in both
coronal and sagittal orientations, and cuts sufficiently
medial to evaluate muscle atrophy.
The infraspinatus independently predicted functional
outcome—atrophy predicted ASES and Constant scores,
while fatty infiltration also predicted ASES score.
Interestingly, the supraspinatus did not independently
predict outcome, and even when the individual unadjusted
results were considered, supraspinatus fatty infiltration
did not correlate with functional outcome scores.
Supraspinatus atrophy only independently affected for-
ward elevation strength, while supraspinatus fatty infil-
tration was an independent predictor of external rotation
strength along with fatty infiltration of the infraspinatus.
The seemingly greater importance of the infraspinatus
toward function may be because those tears that extend
into or involve the infraspinatus are much larger.
Additionally, an ineffective infraspinatus, either through
rupture or muscle degeneration, can offset the biomechan-
ics of the glenohumeral joint,12,19 thus leading to poorer
Vol. 35, No. 5, 2007
outcomes. Conversely, patients with a torn supraspinatus
and an intact anterior-posterior force-couple can maintain
good function and strength. Another possible explanation
for greater influence of the infraspinatus may be that
chronic retraction or atrophy of the supraspinatus can
cause traction on the suprascapular nerve and subsequent
degeneration of the infraspinatus, even in the absence of a
tear of the infraspinatus, as postulated by Albritton et al.1
Goutallier et al9 showed that degeneration of the infra-
spinatus was more of a negative predictor of outcome and
function than the supraspinatus, even after a successful
rotator cuff repair. Jost et al,12 when evaluating preopera-
tive and postoperative fatty infiltration of the infraspina-
tus and supraspinatus in 20 reruptures of the rotator cuff,
showed that only the amount of postoperative fatty infil-
tration of the infraspinatus was significantly correlated to
postoperative Constant scores.
Cuff integrity after rotator cuff repair was directly influ-
enced by the degree of muscle degeneration when individ-
ual correlations were examined. The only independent
predictor of healing of a repair, however, was preoperative
tear size. This finding is in keeping with the conclusions of
several historical studies11 emphasizing the correlation of
higher retear rates with a larger preoperative tear size.
These studies, however, did not examine the effect of mus-
cle degeneration on healing potential or functional out-
come. In more recent studies, the effect of muscle quality
on cuff integrity was evaluated, but regression analyses
were not performed to determine independent predictors.
Thomazeau et al18 found that muscle atrophy of the
supraspinatus correlated with the severity of rotator cuff
disease, and they reported higher retear rates with
increasing grades of atrophy. They concluded that preoper-
ative supraspinatus atrophy was the main predictive fac-
tor for a retear. Goutallier et al9 found that retears
correlated with fatty infiltration of both the supraspinatus
and infraspinatus, and that infraspinatus degeneration
had a highly negative influence on the outcome of
supraspinatus repairs. Interestingly, in our study infra-
spinatus fatty degeneration was the only variable that had
no effect on retear rate. Their supraspinatus retear rate
was 50% when the infraspinatus fatty degeneration was
stage 2 or greater. We found retear rates of 70% and 67%
when patients had stage 2 or greater fatty infiltration and
muscle atrophy, respectively.
If muscle degeneration equates negatively to functional
outcome and retear rates, then it would stand to reason
that progression of fatty infiltration and muscle atrophy
portends poorly for the patient’s well-being. Our study did
not show reversal of these characteristics (except for 1
case, 2.6%), even in the face of successful repairs. In fact,
each variable showed statistically significant progression
of degeneration during the course of the study. Higher
grades (2 and greater) of degeneration progressed more
significantly (70% vs 25%), and rotator cuffs that retore
also progressed more significantly than those that healed.
In general, there was minimal to no progression with suc-
cessful repairs. These findings are in contradistinction to
other studies that have shown improvement in muscle
Fatty Infiltration and Atrophy After Rotator Cuff Repair 727
atrophy7,18 and fatty infiltration9 in cases of successfully
repaired tendons. While Goutallier et al9 showed that the
fatty infiltration of the infraspinatus never regressed, even
with a healed repair, they showed that fatty degeneration
of the supraspinatus can reverse and improve, leading to a
recommendation to operate on wide tears before the
appearance of irreversible muscle damage. Gerber et al7
found that infraspinatus atrophy progressed and was
unaffected by repair integrity, while supraspinatus atro-
phy increased only if the repair failed. In those patients
with a successful repair, there was an improvement in the
cross-sectional area of the supraspinatus. Fatty infiltra-
tion, on the other hand, systematically progressed, but sig-
nificantly less so when the repair was successful. These
findings were consistent with ours. Thomazeau et al18 also
reported a reversal of supraspinatus atrophy in more than
half of their intact cuff repairs; however, no reversal was
seen in those with a failed repair. These data, along with
our results showing the higher rate of progression in those
with moderate to severe atrophy, support the belief that
rotator cuff repair should occur before irreversible histo-
logic and electromyographic changes occur.14 In addition,
an attempt at the strongest repair possible should always
be undertaken, as the progression rate is greater in cuff
muscles with failed repairs.
There are several limitations to this study, in particular
the small number of patients enrolled. Because many of
our patients came to their initial visit with MRI scans,
potential study participants were excluded because of
inadequate preoperative image sequences to allow for ade-
quate muscle belly assessment. We did not have a control
group in which only a wide bursectomy and débridement
without repair was performed. This would have allowed a
comparison of the progression of fatty infiltration and
muscular atrophy in muscles where no attempt at rotator
cuff repair was undertaken and those in which one was
performed, whether the repair ultimately failed or
remained intact. Furthermore, because the natural history
of the progression (or reversal) of muscle atrophy and fatty
infiltration is not fully understood, a follow-up time of 1
year may not be sufficient to observe the full effects. Also,
the quantification of atrophy and fatty infiltration was
based on interpretation of 2-dimensional magnetic reso-
nance images at 1 specific cut through the muscle belly
(either on coronal or sagittal sequences). This may not give
as accurate a measurement as a volumetric analysis.
Finally, the effect of stretching the muscle in the process of
repairing it may also affect its appearance on postopera-
tive MRI and has not been studied. We did not record the
size of the retear compared with the original tear. Looking
at this factor in relation to the preoperative amounts and
subsequent progression of fatty infiltration and muscular
atrophy may have provided useful information.
In conclusion, our analysis reveals that fatty infiltration
and muscle atrophy of the infraspinatus and supraspinatus
significantly affect the functional outcome after rotator cuff
repair even though pain is consistently relieved. Outcome
appears to be mostly affected by the condition of the infra-
spinatus, and the likelihood of a retear is affected by tear
728 Gladstone et al
size. Neither fatty infiltration nor muscular atrophy
reverses after surgery, even with a successful outcome, and
in moderate to severe cases progression is significantly
greater. In patients with a failed repair, progression of atro-
phy and fatty infiltration can be expected to increase more
significantly. Our study, as well as others, suggests there
may be a “point of no return” at which time the muscles
undergo irreversible change at an ultrastructural level. The
challenge ahead will be to figure out whether this point can
be determined, and thus allow for intervention in a timely
fashion, and to better understand the consequences of these
degenerative changes on ultimate outcome.
ACKNOWLEDGMENT
We sincerely appreciate the efforts and support provided to
us for the statistical analysis in this manuscript by our bio-
statistician, John Doucette, Associate Professor,
Department of Community Medicine at the Mount Sinai
School of Medicine.
REFERENCES
1. Albritton MJ, Graham RD, Richards RS 2nd, Basamania CJ. An
anatomic study of the effects on the suprascapular nerve due to retrac-
tion of the supraspinatus muscle after a rotator cuff tear. J Shoulder
Elbow Surg. 2003;12:497-500.
2. Arroyo J, Flatow E, Bigliani L. Rotator cuff disease. In: Baratz M,
Watson A, Imbriglia J, eds. Orthopedic Surgery: The Essentials. New
York, NY: Thieme Medical Publishers; 1999:275-287.
3. Bishop J, Klepps S, Lo IK, Bird J, Gladstone JN, Flatow EL. Cuff
integrity after arthroscopic versus open rotator cuff repair: a prospec-
tive study. J Shoulder Elbow Surg. 2006;15:290-299.
4. Flatow EL, Klepps S. Arthroscopic Mobilization and Rotator Cuff
Repair (Videotape). American Academy of Orthopaedic Surgeons,
70th Annual Meeting, New Orleans, Louisiana, February 2003.
5. Fuchs B, Weishaupt D, Zanetti M, Hodler J, Gerber C. Fatty degener-
ation of the muscles of the rotator cuff: assessment by computed
tomography versus magnetic resonance imaging. J Shoulder Elbow
Surg. 1999;8:599-605.
6. Gazielly DF, Gleyze P, Montagnon C. Functional and anatomical results
after rotator cuff repair. Clin Orthop Relat Res. 1994;304:43-53.
The American Journal of Sports Medicine
7. Gerber C, Fuchs B, Hodler J. The results of repair of massive tears of
the rotator cuff. J Bone Joint Surg Am. 2000;82:505-515.
8. Goutallier D, Bernageau J, Patte D. Assessment of the trophicity of
the muscles of the ruptured rotator cuff by CT scan. In: Post M,
Morrey BF, Hawkins RF, eds. Surgery of the Shoulder. St Louis, MO:
Mosby Year Book; 1990:11-13.
9. Goutallier D, Postel J, Bernageau J, Lavau L, Voisin MC. Fatty mus-
cle degeneration in cuff ruptures. Pre- and postoperative evaluation
by CT scan. Clin Orthop Relat Res. 1994;304:78-83.
10. Goutallier D, Postel JM, Lavau L, et al. CT evaluation of muscular
atrophy before and after repair of the rotator cuff. Fifth International
Conference of Surgery of the Shoulder. Paris, France; 1992:50.
11. Harryman DT II, Mack LA, Wang KY, Jackins SE, Richardson ML, Matsen
FA 3rd. Repairs of the rotator cuff. Correlation of functional results with
integrity of the cuff. J Bone Joint Surg Am. 1991;73:982-989.
12. Jost B, Pfirrmann CW, Gerber C. Clinical outcome after structural fail-
ure of rotator cuff repairs. J Bone Joint Surg Am. 2000;82:304-314.
13. Nakagaki K, Ozaki J, Tomita Y, Tamai S. Fatty degeneration in the
supraspinatus muscle after rotator cuff tear. J Shoulder Elbow Surg.
1996;5:194-200.
14. Nakagaki K, Ozaki J, Tomita Y, Tamai S. Function of supraspinatus
muscle with torn cuff evaluated by magnetic resonance imaging. Clin
Orthop Relat Res. 1995;318:144-151.
15. Nakagaki K, Ozaki J, Tomita Y, et al. Alterations in the supraspinatus
muscle belly with rotator cuff tearing: evaluation with magnetic reso-
nance imaging. J Shoulder Elbow Surg. 1994;3:88-93.
16. Pfirrmann CW, Schmid MR, Zanetti M, Jost B, Gerber C, Hodler J.
Assessment of fat content in supraspinatus muscle with proton MR
spectroscopy in asymptomatic volunteers and patients with
supraspinatus tendon lesions. Radiology. 2004;232:709-715.
17. Shimizu T, Itoi E, Minagawa H, Pradhan RL, Wakabayashi I, Sato K.
Atrophy of the rotator cuff muscles and site of cuff tears. Acta Orthop
Scand. 2002;73:40-43.
18. Thomazeau H, Boukobza E, Morcet N, Chaperon J, Langlais F.
Prediction of rotator cuff repair results by magnetic resonance imag-
ing. Clin Orthop Relat Res. 1997;344:275-283.
19. Thompson WO, Debski RE, Boardman ND 3rd, et al. A biomechani-
cal analysis of rotator cuff deficiency in a cadaveric model. Am J
Sports Med. 1996;24:286-292.
20. Warner JJP, Higgins L, Parsons IM 4th, Dowdy P. Diagnosis and
treatment of anterosuperior rotator cuff tears. J Shoulder Elbow Surg.
2001;10:37-46.
21. Zanetti M, Gerber C, Hodler J. Quantitative assessment of the mus-
cles of the rotator cuff with magnetic resonance imaging. Invest
Radiol. 1998;33:163-170.