Review Article

Degenerative Rotator Cuff Tears:

Refining Surgical Indications
Based on Natural History Data

Abstract
Jay D. Keener, MD Degenerative rotator cuff tears are the most common cause of
Brendan M. Patterson, MD shoulder pain and have a strong association with advanced aging.
Considerable variation exists in surgeons’ perceptions on the
Nathan Orvets, MD
recommended treatment of patients with painful rotator cuff tears.
Aaron M. Chamberlain, MD Natural history studies have better outlined the risks of tear
enlargement, progression of muscle degeneration, and decline in the
function over time. This information combined with the known factors
potentially influencing the rate of successful tendon healing such as
age, tear size, and severity of muscle degenerative changes can be
used to better refine appropriate surgical indications. Although
conservative treatment can be successful in the management of
many of these tears, risks to nonsurgical treatment also exist. The
application of natural history data can stratify atraumatic
From the Shoulder and Elbow
Service, Department of Orthopaedic
degenerative tears according to the risk of nonsurgical treatment and
Surgery, Washington University, St. better identify tears where early surgical intervention should be
Louis, MO. considered.
Dr. Keener or an immediate family
member has received royalties from
Genesis, Shoulder Innovations, and
Imascap; serves as a paid consultant
to Arthrex; and has received research
or institutional support from the
R otator cuff disease is the most
common cause of shoulder dis-
ability and is especially prevalent in
Additionally, male sex, dominant
arm, history of heavy labor, certain
acromial characteristics, and genetic
National Institutes of Health (NIAMS
and NICHD) and Zimmer Biomet. the aging population.1 Many authors factors correlate with rotator cuff
Dr. Patterson or an immediate family suggested that rotator cuff disease tears.9,10
member is an employee of Disk- is a natural aging phenomenon, Although the natural history of
Criminator. Dr. Chamberlain or an given the strong association with age degenerative rotator cuff tears has
immediate family member serves as a
paid consultant to Arthrex, DePuy, and the fact that most tears are recently been better defined, many
and Zimmer Biomet and has received asymptomatic.2,3 Cadaveric and in unanswered questions remain re-
research or institutional support from vivo imaging studies have shown the garding the risk factors for disease
Zimmer Biomet. Neither Dr. Orvets nor rates of asymptomatic rotator cuff progression, in particular pain de-
any immediate family member has
received anything of value from or has tears to increase proportionally with velopment. Natural history studies
stock or stock options held in a age, with 20% of patients in their are fundamental for developing
commercial company or institution sixties and up to 80% of patients appropriate treatment algorithms.
related directly or indirectly to the older than 80 years having tears.2 Despite the high prevalence of rotator
subject of this article.
Yamaguchi et al4 found that patients cuff pathology, substantial contro-
Supported by NIH grant: R01-051026. with a painful cuff tear at the age of versy exists regarding the optimal
J Am Acad Orthop Surg 2019;27: 66 years or older have a 50% chance management of symptomatic rotator
156-165 of having a contralateral rotator cuff cuff disease.11 Trends in nonsurgical
DOI: 10.5435/JAAOS-D-17-00480 tear that is often unknown to the management and rotator cuff repair
patient. Asymptomatic tears develop have varied markedly over time.12
Copyright 2018 by the American
Academy of Orthopaedic Surgeons. pain in approximately 30% to 40% Further complicating the matters, the
of patients within 2 to 5 years.5-8 symptom duration does not correlate

156 Journal of the American Academy of Orthopaedic Surgeons

Copyright Ó the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited.
 Jay D. Keener, MD, et al

Figure 1 Figure 2

A, Anterior cable intact tear. Typical appearance of degenerative rotator cuff

tear. Tear involves the supraspinatus and anterior infraspinatus within the rotator
crescent. The anterior attachment of the supraspinatus is intact, preventing
severe retraction of the supraspinatus tendon. * = biceps tendon. B, Anterior
Insertional rotator cuff anatomy cable disrupted tear. This degenerative cuff tear involves the anterior
according to Mochizuki in an supraspinatus tendon uncovering the biceps tendon. More severe retraction of
illustration of the superior right the supraspinatus muscle is seen. * = biceps tendon.
proximal humerus. GT = greater
tuberosity, HH = humeral head, ISP-I
= infraspinatus insertion, LT = lesser than previously thought. The inser- search. Kim et al16 mapped the
tuberosity, SSP-I = supraspinatus tion extends a mean of 12.6 mm in common locations of asymptomatic
insertion. (Reproduced with
permission from Mochizuki, T,
the anterior to posterior direction and symptomatic full-thickness de-
Sugaya, H, Uomizu, M, et al: into a tapered insertion that is much generative cuff tears in 272 patients
Humeral insertion of the smaller laterally than medially. The using ultrasonography, measuring
supraspinatus and infraspinatus: infraspinatus tendon curves anteri- the distance from the anterior tear
New anatomical findings regarding
the footprint of the rotator cuff. J
orly covering a mean of 32.7 mm of margin to the biceps tendon. Only
Bone Joint Surg Am 2008;90: the superior and posterior greater 33% of tears involved the most
962-969.) tuberosity in the sagittal plan (Figure anterior aspect of the supraspinatus
1). Previous studies of the infra- tendon (Figure 2). The most com-
spinatus tendon footprint cited a mon location of tears involved an
with the rotator cuff tear severity or
range of 16 to 29 mm in the anterior area 13 to 17 mm posterior to the
other patient factors.13 Clinical
to posterior dimension. Mochizuki biceps tendon. Small full-thickness
guidelines provided by the American
et al highlighted the morphology tears most commonly involved an
Academy of Orthopaedic Surgeons
of infraspinatus tendon as curving area 15 mm posterior to the biceps
on the management of rotator cuff
superiorly onto the greater tuberosity tendon, suggesting that degenerative
disease are largely weak or inconclu-
sweeping lateral the supraspinatus tears most commonly originate here.
sive because of lack of high-quality
insertion. These findings are clinically This region correlates with the in-
evidence. Their recommendations
relevant in that many tears classified fraspinatus and supraspinatus junc-
highlight the need for further research
as isolated to the supraspinatus ten- tion commonly described as the
to better define the natural history of
don based on previous anatomic rotator crescent that is bordered by
rotator cuff disease and the results of
definitions actually also involve the the rotator cable.15 Isolated small-
both surgical and nonsurgical inter-
infraspinatus tendon. and medium-sized tears with or
vention to further refine treatment
Degenerative rotator cuff tears without disruption of the anterior
recommendations.14
develop from age-related changes supraspinatus tendon were compared
that may be directly related to the by Namdari et al.17 They found that
Tear Characteristics— poor vascularity of the rotator cres- anterior supraspinatus disrupted tears
Degenerative Cuff Disease cent. Along a continuum, tears likely had a larger tear size and possessed
progress from tendinopathy to par- greater supraspinatus muscle degen-
The insertional anatomy of the rota- tial to full-thickness tears over time. eration. However, no differences in
tor cuff must be considered when Previous theories suggesting that baseline or postsurgical functional
discussing degenerative rotator cuff degenerative tears begin at the ante- outcomes (ASES; 92 versus 93) or
tears. Mochizuki et al15 showed that rior supraspinatus tendon insertion healing rates (93% versus 86%) for
the supraspinatus footprint on the adjacent to the biceps tendon have anterior supraspinatus intact versus
greater tuberosity was much smaller been challenged by prospective re- disrupted tears were seen.17

March 1, 2019, Vol 27, No 5 157

Copyright Ó the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited.
Rotator Cuff Surgical Indications

Disruption of the rotator cuff ten- versus partial thickness) and hand Multiple other studies with variable
don is felt to lead to muscular fatty dominance were associated with a methodologies have attempted to
infiltration (FI) and atrophy, sec- greater enlargement risks; however, define the tear enlargement risks in
ondary to tendon retraction and subject age, sex, and baseline tear size both asymptomatic and painful
impaired force transmission. Rotator were not. The risks of tear enlargement shoulders. Moosmayer et al6 pro-
cuff tear size and location are directly at 2 and 5 years were 22% and 50%, spectively followed 50 full-thickness
related to the patterns of fatty muscle respectively, for full-thickness tears degenerative rotator cuff tears for 3
degeneration. Kim et al18 demon- and 11% and 35%, respectively, for years and showed that symptoms
strated that fatty degeneration was partial-thickness tears. This data sug- developed in 36% of patients. The
nearly exclusive to full-thickness tears. gest that although tear progression is symptomatic group had a larger
Thirty-five percent of the shoulders common, the timeline is relatively increase in tear size, a greater pro-
had evidence of fatty degeneration on slow. Keener et al20 investigated gression of muscle degeneration, and
ultrasonography, and tears with fatty tear progression in anterior supra- more frequent biceps pathology on
degeneration had a greater width and spinatus intact compared with that follow-up imaging compared with
length than those without. In this in disrupted tears in 139 patients the asymptomatic group. In patients
cohort, disruption of the anterior with minimum 2-year follow-up. They with symptomatic full-thickness tears
supraspinatus insertion (anterior cable) found no statistical difference in the managed nonsurgically, using ultra-
was the most important predictor of risk of enlargement, time to enlarge- sonography, Safran et al7 demon-
supraspinatus muscle degeneration, ment, or magnitude of enlargement strated that 49% of tears increased in
whereas larger tear size was the best between groups. However, a trend size within 2 years. The only variable
predictor of infraspinatus muscle exists towards greater enlargement associated with increased pain was
degeneration. risks in the cable-disrupted tears (67% tear enlargement at the time of follow-
vs. 52%, P = 0.09) versus intact tears. up. Using MRI, Maman et al8 retro-
The factors important for pain spectively reviewed 59 patients with
Natural History development in shoulders with rotator cuff tears, with imaging
asymptomatic tears are not clearly a different time points. They found
Tear Enlargement and Pain defined.21 Interestingly, in patients that age greater than 60 years, full-
Progression presenting with a painful rotator cuff thickness tears, and FI were all
Defining the risks of tear progression tear, disease severity does not cor- associated with tear progression. Tear
and symptom development is critical relate well with VAS pain scores. enlargement occurred in 19% at 18
in developing treatment algorithms Both Mall et al19 and Keener et al20 months and at 48% with follow-up
and can be best understood by look- have prospectively demonstrated beyond 18 months. Moosmayer
ing at natural history studies. The that tear enlargement is a risk factor et al22 reported the findings of a
natural history of degenerative rota- for pain development; however, selected subgroup of 49 subjects with
tor cuff tears has been recently absolute correlations do not exist. In small- and medium-sized full-
defined in asymptomatic patients the study by Keener et al,20 46% of thickness tears, who were followed
followed prospectively. Given that asymptomatic shoulders developed for a mean of 8.8 years and managed
rotator cuff disease is often bilateral, pain over a 5-year period. Although nonsurgically. The authors excluded
screening the contralateral shoulder the risk of pain development was 23 shoulders requiring surgery dur-
in patients with a painful rotator cuff approximately 70% more likely if ing the study period. This study
tear provides a valuable study group enlargement occurred, 37% of the demonstrated great variability in
where no treatment intervention is newly painful shoulders did not the magnitude of tear enlarge-
needed.4,19 Keener et al5 examined the enlarge and 38% of enlarged tears ment (mean tear width increase was
risks of cuff tear enlargement and pain remained asymptomatic. Pain de- 8.3 mm). One-third of the tears in-
development in 224 asymptomatic velopment was associated with clin- creased greater than 10 mm (half of
shoulders with known full-thickness ically notable decline in shoulder these were greater than 20 mm).
tears (118), partial-thickness tears (56), function. There are clearly factors Progression of degenerative muscle
and intact rotator cuffs (50) followed other than tear enlargement that changes was noted in half the sub-
annually with ultrasonography and play a role in pain development in jects. Shoulder function remained
clinical examination (Table 1). At 5 shoulders with degenerative tears, stable if the tear size increase was
years of follow-up, the risk of tear and it is important to remember that less than 10 mm. Declining ASES
enlargement of 5 mm or greater was other potential pain generators may and Constant scores were noted
49%. Tear severity (full thickness play a role in symptom onset. with tear enlargement greater than

158 Journal of the American Academy of Orthopaedic Surgeons

Copyright Ó the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited.
 Jay D. Keener, MD, et al

Table 1
Natural History Studies of Untreated Rotator Cuff Tears
Imaging
Study Study Group Duration of F/U Modality Results

Moosmayer 50 asymptomatic full- 3 yr Ultrasonography 36% of shoulders developed pain.

et al6 thickness tears and MRI Painful shoulders had greater enlargement
(10.6 mm versus 3.3 mm).
Increased rate of progression of advanced
atrophy in the symptomatic group (35% versus
12%).
Increased rate of fatty muscle degeneration in the
symptomatic group (35% versus 4%).
Maman 49 nonsurgically Minimum 6 mo, MRI F/U longer than 18 mo associated with greater tear
et al8 managed with range of progression (48% versus 19%).
symptomatic 7–58 mo Age .60 years associated with tear progression
rotator cuff tears (54% versus 17%).
24% of full-thickness tears and zero partial-
thickness tears developed muscle atrophy.
Risk of tear enlargement associated with the
presence of fatty muscle infiltration.
Keener Asymptomatic Mean 5.1 yr Ultrasonography 46% developed pain (50% full-thickness tears,
et al5 patients with 118 46% partial-thickness tears, and 28% controls).
full-thickness tears, Tear enlargement occurred in 61% of full-
56 partial-thickness thickness tears, 44% of partial-thickness tears,
tears, and 50 and 14% of controls.
controls Tear enlargement associated with hand
dominance.
Tear enlargement associated with pain
development.
Tear enlargement associated with cuff muscle
degeneration.
Tear size, age, and sex not correlated with
enlargement.
Safran et al7 61 symptomatic full- 2-3 yr Ultrasonography 49% of tears enlarged, 43% of tears were stable,
thickness rotator 8% of tears decreased in size.
cuff tears in 51 Correlation of tear enlargement and pain
patients aged development.
,60 yr Age, prior trauma, initial tear size. and bilateral
tears were not associated with tear
enlargement.
Fucentese 24 symptomatic full- Mean 42 mo Initial MR Mean Constant score was 75 at F/U.
et al24 thickness rotator arthrogram, 11 patients had no tear or smaller tear at F/U.
cuff tears in F/U MRI 9 patients had no change in tear size.
patients aged 6 patients (25%) had increased tear size.
,65 yr who Progression of FI from zero to 14% of cohort, but
declined surgery none were advanced.

FI = fatty infiltration, F/U = follow-up

20 mm and progression of muscle of advanced fatty muscle degenera- outcomes and lower tendon healing
degeneration. tion compared with smaller tears. rates following surgery (Figure 3). A
However, the timeline for the pro- recent report prospectively examined
gression of muscle degeneration and the risks of fatty muscle degenera-
Progression of Muscle the risk factors for these changes have tion progression using ultrasonogra-
Degeneration not been well-defined. Muscle de- phy.23 In a cohort of 156 full-thickness
It has been well recognized that larger generative changes are thought to tears (the majority being small or
and more chronic rotator tears are be clinically relevant because they medium sized), 55% of tears had some
associated with a greater likelihood have been linked to poorer clinical degree of fatty muscle degeneration

March 1, 2019, Vol 27, No 5 159

Copyright Ó the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited.
Rotator Cuff Surgical Indications

Figure 3

T1-weighted parasagittal MRI images demonstrating rotator cuff muscle health. A, Right shoulder. All cuff muscles healthy.
B, Right shoulder. Supraspinatus (thin arrow) with Goutallier grade II changes (fatty change noted but more muscle than fat).
Infraspinatus (thick arrow) with grade III changes (equal muscle and fat). Teres minor (*) with advanced fatty infiltration. C,
Left shoulder. Supraspinatus (thin arrow) and infraspinatus (thick arrow) with grade IV changes (more fat than muscle).

during a follow-up period of 6.0 years. outcomes is debated, it is generally scopic cuff repair. They reported
The presence of muscle degeneration felt that better and more consistent complete healing in 71% of patients,
was linked to older age and larger tear clinical results are obtained follow- noting age to be strongly correlated
size at baseline. Progression of muscle ing a successful tendon healing. with tendon healing. Patients with a
degeneration was more common in Therefore, identification of factors healed repair were on average 10
tears that enlarged (43% versus 20%). that better predict and surgical strat- years younger (57.8 6 9.4 years)
Progression of fatty muscle changes in egies that improve tendon healing is than those with a failed repair (68 6
enlarged tears was more common in important in identifying the optimal 7.6 years). Furthermore, for patients
tears that were larger at enrollment surgical candidates. Park et al25 younger than 55 years, the healing
(13.0 versus 10.0 mm) and in tears reported on 339 patients undergoing rate was 95%, and for patients older
with a greater magnitude of enlarge- arthroscopic rotator cuff repair for than 65 years the healing rate
ment (9.0 versus 5.0 mm), and when small- and medium-sized tears. They dropped to 43%.27 Oh et al28
the anterior supraspinatus was torn found patient age, tear size, and FI reported on 187 patients undergoing
(53% versus 17%). Although con- of the cuff muscles to be important arthroscopic or mini-open rotator
siderable variability exists in temporal risk factors in the development of cuff repair with a minimum follow-
progression of muscle changes com- recurrent rotator cuff tears. Multi- up of 1 year. The average age of
pared with enlargement events, the ple other studies have consistently patients with an intact repair CT
median time from an enlargement shown that patient age, tear size, arthrogram was 58 years compared
event to the progression of muscle and fatty muscle infiltration are key with 63 years for patients with a
degeneration was 1.0 and 1.1 years factors in predicting tendon healing retear. They also found that tendon
for the supraspinatus and infra- following rotator cuff repair. retraction and FI of the infraspinatus
spinatus, respectively. Patient age plays an important role were risk factors for poor healing.
in cuff tendon healing following sur- This finding highlights the influence
gery. Early literature from Harryman of age and other intrinsic tear char-
Factors Affecting Rotator et al26 revealed a strong correlation acteristics on rotator cuff healing.
Cuff Healing with rotator cuff healing and patient In multiple studies, tear size has
age, with older patients more likely been shown to influence healing rates
When considering surgical treatment, demonstrating recurrent defects. after repair. Galatz et al29 reported
it is paramount to understand factors Boileau et al27 reviewed the healing on the structural integrity of large
that influence rotator cuff healing. rates and functional outcome of and massive rotator cuff tears after
Although the correlation between 65 consecutive patients with full- arthroscopic cuff repair. Recurrent
successful tendon healing and clinical thickness tears treated with arthro- tendon defects were identified in

160 Journal of the American Academy of Orthopaedic Surgeons

Copyright Ó the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited.
 Jay D. Keener, MD, et al

94% of shoulders. Although ad- Similar to tendon injuries elsewhere, for the patients undergoing surgical
vances in surgical techniques have it makes biologic sense that the heal- treatment. Long-term follow-up con-
improved the healing rates of ing environment is optimal in the tinued to demonstrate statistically
arthroscopic cuff repair, tear size acute setting before chronic degener- significant improvements in ASES
continues to be one of the primary ative changes have occurred. While and Constant scores for patients
determinants of successful tendon this concept is common dogma, undergoing surgery; however, these
healing. Park et al25 studied a large research supporting this notion is improvements were not considered
cohort of patients with small- to limited. Tan et al33 studied the effects clinically relevant at 5-year follow-up
medium-sized rotator cuff tears and of recent trauma and tendon healing because the differences failed to reach
found that patients with tears .2 cm in 1,300 patients undergoing arthro- the minimal clinically important dif-
had a healing rate of 65% compared scopic cuff repair. No notable differ- ference threshold.38 The crossover rate
with a healing rate of 89% in pa- ence was observed in the healing increased to 24% by 2 years, with 12
tients with tears #2 cm. More recent rates between patients who reported of 51 patients initially randomized to
literature from Tashjian et al30 has shoulder pain secondary to a specific physical therapy undergoing surgical
shown that tear retraction plays an event compared with tears with a repair. The authors found that 37% of
important role in tendon healing more insidious pain onset. In pa- the patients treated with physical
following repair of degenerative cuff tients reporting traumatic event, de- therapy had greater than 5 mm
tears. They investigated 42 patients laying surgery by more than 24 increase in tear size on 5-year follow-
undergoing arthroscopic rotator cuff months correlated with decreased up ultrasonography. The authors
repair and reported an overall heal- tendon healing. Other studies have supported an initial trial of physical
ing rate of 86%. Tendon healing was demonstrated a benefit of earlier therapy for small- to medium-sized
seen in 92% of tears when the surgical management of traumatic rotator cuff tears; however, they cau-
musculotendinous junction was lat- rotator cuff tears. Petersen and tioned that without surgery, many
eral to the glenoid compared with a Murphy34 investigated 36 shoulders tears have an increased risk of
healing of 56% in tears with with acute rotator cuff tears and enlargement and decreased function.
retraction of the musculotendinous found improved functional out- Kukkonen et al39 performed a ran-
junction medial to the glenoid. comes for patients who underwent domized controlled trial comparing the
Both FI and muscular atrophy are surgery in less than 4 months after outcomes of physical therapy, acro-
well-established risk factors influencing injury compared with those who mioplasty and physical therapy, or
tendon healing following rotator cuff underwent surgery after 4 months. rotator cuff repair for 160 patients
repair. Park et al25 found grade 2 and with full-thickness, degenerative rota-
higher fatty degeneration of the infra- tor cuff tears. They found no notable
spinatus to be an independent risk Clinical Outcomes—Higher difference in the functional scores or
factor for recurrent tears. Chung et al31 Level Evidence patient satisfaction between groups at
reviewed the results of 272 patients 2 years. The authors recommend
undergoing arthroscopic cuff repair. Multiple prospective studies have physical therapy as the preferred initial
Increased FI of the supraspinatus, in- documented excellent outcomes fol- treatment for isolated supraspinatus
fraspinatus, and subscapularis was lowing surgical repair of degenerative tears, with a caveat that many tears
associated with decreased tendon rotator cuff tears.35,36 More recently, treated without repair increase in size
healing. Furthermore, with multivari- several prospective randomized trials at short-term follow-up. Lambers
ate analysis, increased FI of the infra- have compared surgical versus non- Heerspink et al40 conducted a pro-
spinatus was an independent risk surgical management of rotator cuff spective randomized trial investigating
factor for recurrent tendon defects tears. Moosmayer et al37 performed a surgical versus nonsurgical treatment
following repair. Kim et al32 also randomized trial comparing physical in 56 patients with degenerative
found FI of the infraspinatus to be an therapy with surgical management of rotator cuff tears. At 1-year follow-
independent risk factor for a recurrent traumatic and degenerative rotator up, a notable improvement was
cuff tear with multivariate analysis of cuff tears. The authors originally observed in VAS pain scores for pa-
132 patients following repair of full- reported the outcomes of 103 pa- tients treated with cuff repair; how-
thickness cuff tears. tients with a minimum follow-up of ever, no notable difference was seen
Rotator cuff tears resulting from 1 year. In patients with small- to in Constant scores between groups. A
acute injuries are thought to be more medium-sized rotator cuff tears, they subgroup analysis of healing data
likely to heal than degenerative rota- found markedly greater improve- within revealed that patients with
tor cuff tears after surgical repair. ments in ASES and Constant scores intact repairs demonstrated notable

March 1, 2019, Vol 27, No 5 161

Copyright Ó the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited.
Rotator Cuff Surgical Indications

Table 2
Treatment Recommendations Based on Patient and Tear Characteristics
Risks of Tear
Enlargement/
Muscle
Degeneration
Progression Patient and Tear Characteristics Treatment Recommendation

Low risk Partial-thickness tears Maximize conservative treatment, surgery if

Large tears with advanced muscle changes persistently asymptomatic.
Degenerative tears in patients aged .65–70 yr
Atraumatic full-thickness tears less than 15 mm in size
with an intact anterior cable
Medium risk Age under 62-65 years Informed discussion of surgical and
Atraumatic full-thickness tears .15 mm nonsurgical options warranted.
Anterior cable disruption Consider surveillance exams with successful
Acute on chronic tears–preserved function conservative treatment.
High risk Acute traumatic full-thickness tears Strong consideration for early surgical repair.
Acute on chronic tears with new pseudoparalysis or
profound external rotation weakness
Minimal muscle degenerative changes
Age compatible with healing

improvements in pain and functional discussion of the risks and benefits of progression of arthritic changes in
outcome compared with patients both surgical and nonsurgical treat- patients with degenerative rotator
treated nonsurgically. ment. Conservative treatment is a cuff tears followed prospectively.
Although these studies suggest an well-accepted treatment method for Certain tear characteristics warrant
advantage of surgery over conservative atraumatic full-thickness rotator cuff consideration for either close sur-
management of rotator cuff tears at tears in the short term.41 In many veillance or recommendation for
short-term follow-up when patients are patients with notable medical co- early surgical intervention. Based on
randomized at baseline presentation, morbidities or advanced age, this the natural history data, tears can be
the magnitude of the clinical relevance may well be the preferred treatment. stratified according to the short-term
of these findings is challenged by However, we are faced with a chal- risks of developing features previ-
defined minimal clinically important lenge when deciding the best treat- ously shown to adversely affect the
difference thresholds. Importantly, ment for younger patients with healing rates and clinical outcomes of
these studies do not make a distinction full-thickness tears that possess a surgery (Table 2). These risks must
between tears with varying risk factors high likelihood of disease progres- be taken in the context of specific
for progression and do not adequately sion. Natural history studies have patient (age) and tear-related factors
address the potential downsides of better clarified tears with a higher (size, cable integrity, and fatty mus-
nonsurgical treatment over time. These risk of progression, highlighting an cle degeneration) already present at
include the tear enlargement risks, opportunity to refine surgical in- the time of clinical presentation.
degenerative muscle change progres- dications. Delayed surgical inter- Shoulder pain severity is highly
sion, and the subsequent deleterious vention in higher risk tears may variable across tear sizes, and pain
effects of cuff repair healing seen with allow tear enlargement and/or the can often be managed, at least short
advancing age. development of irreversible muscle term, conservatively. We suggest
changes, which, when combined that anatomic and patient-related
with the deleterious effects of aging, features that affect surgical results
Redefining Surgical will decrease the rate of successful are better objective criteria than pain
Indications Based on tendon healing.42 Many studies have severity for consideration for early
Natural History Data demonstrated both increased tear surgical intervention.
size and progression of muscle Low-risk tears include those with
Informed decision making for the degeneration within 2 years. Addi- low risk of tear enlargement and
management of degenerative rotator tionally, Chalmers et al43 have progression of muscle degeneration
cuff tears should entail a complete documented increased radiographic or tears with poor healing capacity.

162 Journal of the American Academy of Orthopaedic Surgeons

Copyright Ó the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited.
 Jay D. Keener, MD, et al

In these tears, there are lower risks of Figure 4

short-term tear progression or the
optimal window for surgery has been
missed. These include tears at both
ends of the disease spectrum: partial-
thickness rotator cuff tears and larger
tears with advanced fatty muscle
degeneration (grade III/IV Goutallier
changes) and/or proximal humeral
migration. Atraumatic tears in pa-
tients older than 65 to 70 years,
although still reasonable surgical
candidates, have a lower healing rate
and consideration for initial conser-
vative treatment is warranted.
Included in this group are atraumatic
full-thickness tears up to 15 mm in
size with an intact anterior supra-
spinatus tendon and healthy muscles
because the short-term risks of tear
progression are relatively low. There
is time to maximize conservative
treatment without affecting the re-
sults of later surgery if conservative
treatment fails.
Medium-risk tears include those
with moderate risk of short-term tear
progression in individuals with good
healing capacity (age under 62-65
years). These include atraumatic full- Acute on chronic rotator cuff tear, right shoulder. A, Coronal T2-weighted MRI
thickness tears 15 mm or larger and image. Large retracted tear of the supraspinatus tendon (thin arrow). B, Coronal
tears with disruption of the anterior T2-weighted MRI image. Retracted and kinked infraspinatus tendon (thin arrow).
Intramuscular edema noted within the infraspinatus (thick arrow). C, Parasagittal
supraspinatus tendon, as well as T2-weighted MRI image. Perimuscular edema noted within the supraspinatus
previously painful shoulders with and infraspinatus muscles. D, Parasagittal T1-weighted MRI image. Grade III
recent trauma (acute on chronic Goutallier fatty changes within the supraspinatus (thin arrow). Grade I/II
tears). These risks are amplified if Goutallier fatty changes within the infraspinatus (thick arrow).
there is already early fatty muscle
degeneration because these tears tendon retraction was not reliable in of tendon healing because of their
possess a greater risk of progression distinguishing between acute and acuity. In these shoulders, surgical
of muscle degeneration over time. chronic tears. For medium-risk tears, intervention has the greatest poten-
Acute on chronic rotator cuff tears an informed discussion of treatment tial to interrupt the natural history of
represent a unique challenge in dis- options with the patient is warranted. an untreated cuff tear. Included are
tinguishing whether a preexisting Surveillance physical or imaging ex- acute traumatic full-thickness tears,
tear was present, and if so, how much aminations, such as ultrasonography, especially those 15 mm or larger, in a
of the tear represents acute enlarge- should be considered to assess po- previously healthy shoulder. Also
ment (Figure 4). One study suggested tential tear progression with success- included are acute on chronic tears
MRI features that may distinguish ful conservative treatment. Patients with a dramatic loss of function such
acute from chronic tears.44 Acute should be counseled not to ignore an as pseudoparalysis and/or profound
tears have less muscle fatty degen- increase in shoulder weakness because external rotation weakness. An impor-
eration, often possessed a wavy or this may herald tear enlargement. tant consideration for these tears
kinked appearing central tendon, High-risk tears represent those with is the quality of the involved mus-
and are often associated with peri- the greatest risk of disease progres- cles, which should possess minimal
muscular edema. The amount of sion or tears that possess a high rate fatty degenerative changes. Strong

March 1, 2019, Vol 27, No 5 163

Copyright Ó the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited.
Rotator Cuff Surgical Indications

consideration for surgical repair is asymptomatic and symptomatic shoulders. 17. Namdari S, Donegan RP, Dahiya N, Galatz
J Bone Joint Surg Am 2006;88:1699-1704. LM, Yamaguchi K, Keener JD:
warranted for tears such as these in Characteristics of small to medium-sized
patients aged 65 years or younger. 5. Keener JD, Galatz LM, Teefey SA, et al: A rotator cuff tears with and without
prospective evaluation of survivorship of disruption of the anterior supraspinatus
asymptomatic degenerative rotator cuff tendon. J Shoulder Elbow Surg 2014;23:
tears. J Bone Joint Surg Am 2015;97:89-98. 20-27.
Summary
6. Moosmayer S, Tariq R, Stiris M, Smith HJ: 18. Kim HM, Dahiya N, Teefey SA, Keener JD,
The natural history of asymptomatic Galatz LM, Yamaguchi K: Relationship of
In recent years, an improved knowl- rotator cuff tears. J Bone Joint Surg Am tear size and location to fatty degeneration
edge of the natural history of de- 2013;95:1249-1255. of the rotator cuff. J Bone Joint Surg Am
generative rotator cuff tears has 2010;92:829-839.
7. Safran O, Schroeder J, Bloom R, Weil Y,
strengthened our understanding of Milgrom C: Natural history of nonoperatively 19. Mall NA, Kim HM, Keener JD, et al:
treated symptomatic rotator cuff tears in Symptomatic progression of asymptomatic
this disease process. When the risks of patients 60 years old or younger. Am J Sports rotator cuff tears: A prospective study of
disease progression are coupled with Med 2011;39:710-714. clinical and sonographic variables. J Bone
the known factors that influence 8. Maman E, Harris C, White L, Tomlinson
Joint Surg Am 2010;92:2623-2633.
tendon healing following rotator cuff G, Shashank M, Boynton E: Outcome of 20. Keener JD, Hsu JE, Steger-May K, Teefey
repair, we can better refine surgical nonoperative treatment of symptomatic SA, Chamberlain AM, Yamaguchi K:
rotator cuff tears monitored by magnetic Patterns of tear progression for
indications. The optimal manage- resonance imaging. J Bone Joint Surg Am asymptomatic degenerative rotator cuff
ment of tears must be individualized 2009;91:1898-1906. tears. J Shoulder Elbow Surg 2015;24:
based on the clinical presentation 1845-1851.
9. Yamamoto A, Takagishi K, Osawa T, et al:
and various patient- and tear-related Prevalence and risk factors of a rotator cuff 21. Dunn WR, Kuhn JE, Sanders R, et al:
tear in the general population. J Shoulder Symptoms of pain do not correlate with
factors. Elbow Surg 2010;19:116-120. rotator cuff tear severity: A cross-sectional
study of 393 patients with a symptomatic
10. Tashjian RZ, Farnham JM, Albright FS, atraumatic full-thickness rotator cuff tear. J
Teerlink CC, Cannon-Albright LA:
References Evidence for an inherited predisposition
Bone Joint Surg Am 2014;96:793-800.
contributing to the risk for rotator cuff 22. Moosmayer S, Gärtner AV, Tariq R: The
Evidence-based Medicine: Levels of disease. J Bone Joint Surg Am 2009;91: natural course of nonoperatively treated
1136-1142. rotator cuff tears: An 8.8-year follow-up of
evidence are described in the table of
tear anatomy and clinical outcome in 49
contents. In this article, references 39 11. Dunn WR, Schackman BR, Walsh C, et al: patients. J Shoulder Elbow Surg 2017;26:
Variation in orthopaedic surgeons’
and 40 are level I studies. References perceptions about the indications for
627-634.
5, 6, 15, 19, 20, 23, 32, 37, 41, 43, rotator cuff surgery. J Bone Joint Surg Am 23. Hebert-Davies J, Teefey SA, Steger-May K,
and 44 are level II studies. References 2005;87:1978-1984. et al: Progression of fatty muscle
degeneration in atraumatic rotator cuff
4, 7, 8, 13, 16–18, 21, 22, 33, 34, 38, 12. Varkey DT, Patterson BM, Creighton RA, tears. J Bone Joint Surg Am 2017;99:
and 42 are level III studies. Refer- Spang JT, Kamath GV: Initial medical 832-839.
management of rotator cuff tears: A
ences 1-3, 9, 11, 12, 25–31, 35, and demographic analysis of surgical and 24. Fucentese SF, Von Roll AL, Pfirrmann WA,
36 are level IV studies. nonsurgical treatment in the United States Gerber C, Jost B: Evolution of
Medicare population. J Shoulder Elbow nonoperatively treated symptomatic
References printed in bold type are Surg 2016;25:e378-e385. isolated full-thickness supraspinatus tears.
J Bone Joint Surg Am 2012;94:801-808.
those published within the past 5 13. Unruh KP, Kuhn JE, Sanders R, et al: The
years. duration of symptoms does not correlate 25. Park JS, Park HJ, Kim SH, Oh JH:
with rotator cuff tear severity or other Prognostic factors affecting rotator cuff
1. Moosmayer S, Smith H-J, Tariq R, Larmo patient-related features: A cross-sectional healing after arthroscopic repair in small to
A: Prevalence and characteristics of study of patients with atraumatic, full- medium-sized tears. Am J Sports Med 2015;
asymptomatic tears of the rotator cuff: An thickness rotator cuff tears. J Shoulder 43:2386-2392.
ultrasonographic and clinical study. J Bone Elbow Surg 2014;23:1052-1058.
Joint Surg Br 2009;91:196-200. 26. Harryman DT, Mack LA, Wang KY,
14. Pedowitz RA, Yamaguchi K, Ahmad CS, Jackins SE, Richardson ML, Matsen FA III:
2. Milgrom C, Schaffler M, Gilbert S, van et al: Optimizing the management of Repairs of the rotator cuff: Correlation of
Holsbeeck M: Rotator-cuff changes in rotator cuff problems. J Am Acad Orthop functional results with integrity of the cuff. J
asymptomatic adults: The effect of age, Surg 2011;19:368-379. Bone Joint Surg Am 1991;73:982-989.
hand dominance and gender. J Bone Joint
Surg Br 1995;77:296-298. 15. Mochizuki T, Sugaya H, Uomizu M, et al: 27. Boileau P, Brassart N, Watkinson DJ,
Humeral insertion of the supraspinatus and Carles M, Hatzidakis AM, Krishnan SG:
3. Tempelhof S, Rupp S, Seil R: Age-related infraspinatus: New anatomical findings Arthroscopic repair of full-thickness tears
prevalence of rotator cuff tears in regarding the footprint of the rotator cuff. J of the supraspinatus: Does the tendon really
asymptomatic shoulders. J Shoulder Elbow Bone Joint Surg Am 2008;90:962-969. heal? J Bone Joint Surg Am 2005;87:
Surg 1999;8:296-299. 1229-1240.
16. Kim HM, Dahiya N, Teefey SA, et al:
4. Yamaguchi K, Ditsios K, Middleton WD, Location and initiation of degenerative 28. Oh JH, Kim SH, Kang JY, Oh CH, Gong
Hildebolt CF, Galatz LM, Teefey SA: The rotator cuff tears: An analysis of three HS: Effect of age on functional and
demographic and morphological features of hundred and sixty shoulders. J Bone Joint structural outcome after rotator cuff repair.
rotator cuff disease: A comparison of Surg Am 2010;92:1088-1096. Am J Sports Med 2010;38:672-678.

164 Journal of the American Academy of Orthopaedic Surgeons

Copyright Ó the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited.
 Jay D. Keener, MD, et al

29. Galatz LM, Ball CM, Teefey SA, Middleton 35. Cole BJ, McCarty LP, Kang RW, Alford W, 40. Lambers Heerspink FO, van Raay JJ,
WD, Yamaguchi K: The outcome and Lewis PB, Hayden JK: Arthroscopic rotator Koorevaar RC, et al: Comparing
repair integrity of completely cuff repair: Prospective functional outcome surgical repair with conservative
arthroscopically repaired large and massive and repair integrity at minimum 2-year treatment for degenerative rotator cuff
rotator cuff tears. J Bone Joint Surg Am follow-up. J Shoulder Elbow Surg 2007;16: tears: A randomized controlled trial. J
2004;86-A:219-224. 579-585. Shoulder Elbow Surg 2015;24:
1274-1281.
30. Tashjian RZ, Hung M, Burks RT, 36. DeFranco MJ, Bershadsky B, Ciccone J,
Greis PE: Influence of preoperative Yum JK, Iannotti JP: Functional outcome 41. Kuhn JE, Dunn WR, Sanders R, et al:
musculotendinous junction position on of arthroscopic rotator cuff repairs: A Effectiveness of physical therapy in
rotator cuff healing using single-row correlation of anatomic and clinical treating atraumatic full-thickness rotator
technique. Arthroscopy 2013;29: results. J Shoulder Elbow Surg 2007;16: cuff tears: A multicenter prospective
1748-1754. 759-765. cohort study. J Shoulder Elbow Surg
2013;22:1371-1379.
31. Chung SW, Oh JH, Gong HS, Kim JY, Kim 37. Moosmayer S, Lund G, Seljom U, et al:
SH: Factors affecting rotator cuff healing Comparison between surgery and 42. Kim HM, Caldwell JE, Buza JA, et al:
after arthroscopic repair. Am J Sports Med physiotherapy in the treatment of small and Factors affecting satisfaction and shoulder
2011;39:2099-2107. medium-sized tears of the rotator cuff: A function in patients with recurrent rotator
randomised controlled study of 103 cuff tear. J Bone Joint Surg Am 2014;96:
32. Kim DH, Jang YH, Choi YE, Lee HR, Kim
patients with one-year follow-up. J Bone 106-112.
SH: Evaluation of repair tension in
Joint Surg Br 2010;92:83-91.
arthroscopic rotator cuff repair. Am J
43. Chalmers PN, Salazar DH, Steger-May K,
Sports Med 2016;44:2807-2812.
38. Moosmayer S, Lund G, Seljom US, et al: et al: Radiographic progression of arthritic
33. Tan M, Lam PH, Le BT, Murrell GA: Tendon repair compared with changes in shoulders with degenerative
Trauma versus no trauma: An analysis physiotherapy in the treatment of rotator rotator cuff tears. J Shoulder Elbow Surg
of the effect of tear mechanism on cuff tears. J Bone Joint Surg Am 2014;96: 2016;25:1749-1755.
tendon healing in 1300 consecutive 1504-1514.
patients after arthroscopic rotator cuff 44. Loew M, Magosch P, Lichtenberg S,
repair. J Shoulder Elbow Surg 2016;25: 39. Kukkonen J, Joukainen A, Lehtinen J, Habermeyer P, Porschke F: How to
12-21. Mattila KT, Tuominen EK, Kauko T, discriminate between acute traumatic and
Äärimaa V: Treatment of nontraumatic chronic degenerative rotator cuff lesions:
34. Petersen SA, Murphy TP: The timing of rotator cuff tears: A randomized controlled An analysis of specific criteria on
rotator cuff repair for the restoration of trial with two years of clinical and imaging radiography and magnetic resonance
function. J Shoulder Elbow Surg 2011;20: follow-up. J Bone Joint Surg Am 2015;97: imaging. J Shoulder Elbow Surg 2015;24:
62-68. 1729-1737. 1685-1693.

March 1, 2019, Vol 27, No 5 165

Copyright Ó the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited.