Outcomes After Limited or Extensive Bursectomy

During Rotator Cuff Repair: Randomized Controlled

Trial
Ji-Hoon Nam, M.D., Sanghyun Park, M.D., Hwa-Ryeong Lee, D.P.T., and
Sae Hoon Kim, M.D., Ph.D.

Purpose: To evaluate the effects of extensive bursectomy (EB) and limited bursectomy (LB) during arthroscopic rotator
cuff repair. Methods: In the EB group (n ¼ 39), subacromial bursae were thoroughly removed from anterior to posterior
and lateral to medial. In the LB group (n ¼ 39), bursectomy was minimized to allow torn cuff visualization and tendon
repair. Visual analog scale pain scores, passive forward flexion, external rotation at the side (ER), and internal rotation at
the back were measured at 5 weeks and 3, 6, and 12 months after surgery. At each time point, bursal thickness was
measured and repair integrity was assessed by sonography or magnetic resonance imaging. Results: The analysis
included 36 patients in the LB group and 35 in the EB group. Group visual analog scale pain scores were not significantly
different at any time (P > .05 for all). Forward flexion and internal rotation at the back showed no intergroup difference
during follow-up. However, ER was significantly better in the LB group at 6 months and 1 year postoperatively (31  15
vs 22  16 [P ¼ .020] and 40  19 vs 27  20 [P ¼ .009], respectively). Integrity failures were not significantly
different at 5 weeks and at 3, 6, and 12 months (P > .05 for all). Marked bursal thickening (>2 mm) was more frequently
observed in the EB group (18 of 32 in the LB group and 27 of 32 in the EB group) at 6 months (P ¼ .014).
Conclusions: EB during arthroscopic rotator cuff repair appears to have no benefit in terms of reducing pain. More
adhesions in the subacromial space after EB may result in slower motion recovery, especially in terms of ER. The extent of
bursectomy did not affect tendon integrity. However, marked bursal thickening was more frequently observed in the EB
group. Level of Evidence: Level I, randomized controlled study.

T he subacromial bursa (SAB), which is the largest

bursa in the human body,1,2 is located under the
coracoacromial arch and deltoid muscle. The inferior
tendon and muscle and the coracoacromial arch during
shoulder movement, and thus, the SAB contributes to
lubrication of the shoulder joint.1,3,4 For this reason, the
aspect of the SAB contacts the greater tuberosity of the SAB is considered one of the most important tissues
humerus and rotator cuff muscles. The main function of associated with discomfort and restriction of shoulder
the SAB is to reduce friction between the rotator cuff movement.2,5-7
During arthroscopic rotator cuff repair, subacromial
bursectomy is required for visualization. Furthermore,
From the Department of Orthopedic Surgery, Seoul National University
some authors are of the opinion that the SAB may be
College of Medicine, Seoul National University Hospital, Seoul, Republic of
Korea. pathologic tissue and that its removal is likely to reduce
This study was funded by the Grant-Aid-Program of Seoul National symptoms and enhance outcomes,8,9 and this point of
University Hospital (No. 06 2015 1370; $10,000; original source: Jeil Phar- view is supported by previous reports that inflamma-
maceutical, Seoul, Republic of Korea). The authors report that they have no tory proliferation of the SAB present in a rotator cuff
conflicts of interest in the authorship and publication of this article. Full
tear is related to the degree of pain.8,10 In addition, the
ICMJE author disclosure forms are available for this article online, as
supplementary material. SAB is known to have a significantly richer supply of
Received January 1, 2018; accepted June 26, 2018. free nerve fibers than other tissues, and nociceptive
Address correspondence to Sae Hoon Kim, M.D., Ph.D., Department of information relayed by these fibers may be responsible
Orthopaedic Surgery, Seoul National University Hospital, 101 Daehak-ro, for pain generation.11,12 Furthermore, bursal specimens
Jongno-gu, Seoul 110-744, Republic of Korea. E-mail: drjacobkim@gmail.
from patients with subacromial bursitis or rotator cuff
com
Ó 2018 by the Arthroscopy Association of North America disease have shown increases in cytokines, including
0749-8063/171560/$36.00 interleukin (IL) 1, IL-6, tumor necrosis factor, cyclo-
https://doi.org/10.1016/j.arthro.2018.06.056 oxygenase (COX) 1, COX-2, matrix metalloproteases

Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol -, No - (Month), 2018: pp 1-8 1
2 J-H. NAM ET AL.

(MMPs), and substance P.9,12,13 Therefore, subacromial Surgical Technique and Group Allocation
bursectomy is considered important for reducing All surgical procedures were performed by a single
inflammation in rotator cuff disease.9 surgeon (S.H.K.). Surgery was performed with patients
Other authors have emphasized the biological role of under general anesthesia with no additional nerve
bursal tissue and recommended avoiding unnecessary block. All patients underwent all-arthroscopic repair of
SAB removal.14 In a histologic study, vascularized a full-thickness rotator cuff tear by the single-row repair
connective tissue covering rotator cuff tears and technique in the lateral decubitus position.
proliferating cells in fragmented tendons were found to Initially, an arthroscope was introduced into the gle-
exhibit more features of repair than of degeneration or nohumeral joint through a standard posterior viewing
necrosis, and thus, it was suggested that the SAB may portal, and intra-articular pathology was examined
play an important role in tendon reconstitution and carefully, particularly biceps status. To address biceps
remodeling.14 Furthermore, growth factors and cyto- pathology, biceps tendon debridement was performed
kines, such as basic fibroblast growth factor, trans- for tears of less than 50% of the tendon width. Biceps
forming growth factor b, and vascular endothelial tenotomy or subpectoral tenodesis was performed in
growth factor, have been detected in SAB tissues.15,16 patients with tears greater than 50% of the tendon
The purpose of this prospective randomized study was width. After completion of the intra-articular proced-
to evaluate the effects of extensive bursectomy (EB) ures, the arthroscope was moved to the subacromial
and limited bursectomy (LB) during arthroscopic space. Acromioplasty was not performed in any case
rotator cuff repair. The hypothesis was that clinical included in this study.
outcomes, primarily visual analog scale (VAS) pain Patients were randomly allocated (by permuted-block
scores, would be different in the EB and LB groups. randomization) to either the EB group (n ¼ 39) or LB
group (n ¼ 39). In the EB group, the SAB was thor-
Methods oughly removed with a shaver from anterior to poste-
rior and lateral to medial. Viewing and working portals
Inclusion and Exclusion Criteria were switched to enable complete bursa removal. Bursa
This prospective, randomized controlled study (Korea on the medial side of the rotator cuff tendon and
Centers for Disease Control & Prevention, registry No. muscle and in the anterior and posterior gutter area
KCT0002860, http://nih.go.kr/NIH_NEW/intro.html) was completely removed in all patients. In the LB
was performed after we obtained approval from our group, bursectomy was minimized to enable torn cuff
institutional review board (Seoul National University visualization and tendon repair.
Hospital Institutional Review Board No. H-1305-610- Rotator cuff repair was performed in a standardized
492). Written informed consent was obtained from all manner in all study subjects. In each case, the greater
participating patients. The study was double blinded, tuberosity was prepared using a motorized burr to
that is, patients and the single evaluator were unaware cause surface bleeding. Anchors were inserted at the
of group assignments. lateral footprint margin through an accessory portal
Sample size analysis was performed before the study created by a small stab incision. A triple-loaded all-
using previously reported VAS pain scores (the primary suture anchor, Y-Knot RC (ConMed Linvatec, Largo,
study outcome). These previous reports showed that FL), was used in all cases. Loaded sutures were passed
the VAS pain scores of patients after rotator cuff repair through the tendon at 0.5 to 1 cm medial to its edge
are normally distributed and exhibit a standard devia- using a flexible suture passer (Expressew; DePuy
tion of 2.17 For a mean VAS score difference of 1.5 Mitek, Raynham, MA). All knots were tied securely
between groups,18 29 subjects were needed per group using a self-locking, sliding knot. The number of an-
to reject the null hypothesis (power, 0.8; type I error, chors used depended on tear size: usually 1 for small
.05). Thus, on the basis of a predicted dropout rate of tears, 2 or 3 for medium to large tears, and 4 or more
25%, 78 patients were required. for massive tears.
Patients were enrolled prospectively from April 2015
to April 2016. The following inclusion criteria were Rehabilitation
applied: (1) a full-thickness rotator cuff tear was pre- The same rehabilitation protocol was used in the 2
sent, as verified by preoperative magnetic resonance groups. Immobilization was maintained with a sling
imaging (MRI) and confirmed at surgery, and (2) and abduction pad for 5 weeks in all patients. After
complete repair was possible arthroscopically. The weaning from brace use, patients were instructed on
exclusion criteria applied were previous ipsilateral how to conduct passive assisted stretching exercises and
shoulder surgery, incomplete repair, rheumatoid active exercises (forward elevation, external rotation,
arthritis, parkinsonism, Alzheimer disease, refusal to and internal rotation before active exercises). Return to
participate, and the need for an additional procedure sports was not allowed until approximately 5 months
such as acromioplasty or SLAP repair. after surgery.
 EXTENSIVE VERSUS LIMITED BURSECTOMY 3

Outcome Evaluation primary outcome at each of these time points. Passive

In accordance with our clinic’s protocol, all data (at ranges of motion (ROMs)dforward flexion (FF),
baseline and at 5 weeks and 3, 6, and 12 months after external rotation at the side (ER), and internal rotation
surgery) were prospectively collected by a clinical at the back (IR)dwere also evaluated as secondary
researcher (H-R.L.) unaware of study details and group outcomes at 3, 6, and 12 months after surgery. Three
assignments (Fig 1). The VAS pain score was the patients in the LB group and 4 patients in the EB group

Fig 1. Study flow diagram.

(MRI, magnetic resonance
imaging; ROM, range of
motion; Sono, sonography;
VAS, visual analog scale.)
4 J-H. NAM ET AL.

were lost to follow-up; thus, 36 patients in the LB group image analysis, bursal thickness was measured and
and 35 in the EB group were included in the clinical repair integrity was assessed. Bursal thickness was
outcome analysis. measured in 1 location in the middle of the repaired
Pain was assessed by patients using a 10-cm-long VAS tendon area, usually above the supraspinatus tendon.
pain ruler. Patients located an indicator on the front of Marked bursal thickening was defined as a measured
the ruler with facial expressions and were not able to bursal thickness of greater than 2 mm either by so-
see the VAS scores on the other side.19 Possible VAS nography or by MRI (Fig 2).21-23 Image evaluation was
scores ranged from 0 to 10, where 10 indicated greatest conducted by a musculoskeletal radiologist blinded to
pain. patient information and not otherwise involved in the
Passive ROM was measured using a goniometer with study.
the spine in a straight position. Passive FF was
measured in degrees between the arm and thorax in Statistical Analysis
the scapular plane. Passive ER was measured in degrees Statistical analysis was performed using IBM SPSS
between the thorax and forearm, with the patient’s arm Statistics software (version 21.0; IBM, Armonk, NY).
in an adducted position and the elbow flexed to 90 . Group demographic data were compared using the
Passive IR was measured using the vertebral level Student t test and the c-square test. The paired t test
reached by the thumb tip. The vertebral levels were and the Student t test were used to perform intragroup
numbered serially as follows: 12 for the 12th thoracic and intergroup comparisons of outcome variables at
vertebra, 13 for the first lumbar vertebra, 17 for the different times. Statistical significance was accepted for
fifth lumbar vertebra, and 18 for any level below the P < .05.
sacral region.20
Routine postoperative sonography was performed at Results
5 weeks, 3 months, and 6 months and MRI was per- Three patients in the LB group and 4 patients in the
formed at approximately 1 year after surgery. During EB group were lost to follow-up, and these patients

Fig 2. Bursal thickness was

measured during assess-
ment of rotator cuff repair
integrity. Sonography (A)
and magnetic resonance
imaging (B) of right shoul-
der. (C) Magnetic reso-
nance imaging of left
shoulder in another patient.
 EXTENSIVE VERSUS LIMITED BURSECTOMY 5

Table 1. Patient Demographic Characteristics and Surgical Data

Variable Limited Bursectomy (n ¼ 36) Extensive Bursectomy (n ¼ 35) P Value

Age, yr 59.7  6.1 60.7  5.4 .484
Sex 17 M and 19 F 20 M and 15 F .403
Laterality 25 right and 11 left 26 right and 9 left .650
Diabetes 3 4 .710
No. of involved tendons S in 9, SþI in 21, and SþIþSubS in 6 S in 3, SþI in 24, and SþIþSubS in 8 .176
No. of anchors/patient 1.9  0.7 2.0  0.7 .379
Biceps procedure None in 14, tenotomy in 3, and subpectoral None in 10, tenotomy in 1, and subpectoral .327
tenodesis in 19 tenodesis in 24
NOTE. Data are presented as mean  standard deviation or absolute number.
F, female; M, male; S, tears confined to 1 tendon (supraspinatus); SþI, supraspinatus tears with some degree of infraspinatus involvement;
SþIþSubS, tears involving 3 tendons (supraspinatus, infraspinatus, and part of subscapularis).

were excluded from the analysis. Finally, 71 patients subacromial space. In terms of tendon integrity, no
were included in the analysis (36 in the LB group and difference was observed between the 2 study groups.
35 in the EB group) (Fig 1). However, marked bursal thickening was more
Demographic characteristics and surgical data are frequently observed in the EB group, and this reached
shown in Table 1. Biceps procedures were performed in statistical significance at 6 months postoperatively,
47 cases (tenotomy in 4 and subpectoral anchor which was in accord with the ER limitation observed at
tenodesis in 43). Tear size was categorized according to 6 months. The minimal clinically important difference
the number of tendons involved; tears confined to 1 for ER has been proposed to be 11 ,24 and in our study,
tendon (supraspinatus) were observed in 12 cases, the intergroup difference observed at 1 year post-
supraspinatus tears with some degree of infraspinatus operatively (13 ) was greater than this value. However,
involvement were found in 45 cases, and tears at 6 months postoperatively, the difference was 9 . In
involving 3 tendons (supraspinatus, infraspinatus, and terms of bursal thickness, normally the SAB is rarely
part of the subscapularis) were observed in 14 cases thicker than 2 mm and greater thicknesses are regarded
(Table 1). as abnormal.21-23
Preoperative VAS pain scores were similar in the 2 The extent of bursectomy remains uncertain when
study groups, and no significant difference was performing rotator cuff repair. Some clinicians have
observed during follow-up (Table 2). Regarding post- claimed that the bursal tissue is inflamed in painful
operative passive ROMs, mean FF was similar in the 2 rotator cuff tears and that it should be removed to
groups throughout follow-up. However, the LB group improve symptom relief,9 whereas others have claimed
showed significantly better ER at 6 months and 1 year that the bursal tissue provides a gliding surface between
postoperatively and tended to show greater IR at the rotator cuff and the acromion and deltoid4 and that
6 months postoperatively (Table 3). At final follow-up removal of this tissue detrimentally affects repair
(1 year postoperatively), Constant scores, Simple because it contains vasculature and growth factors.2
Shoulder Test scores, and American Shoulder and Furthermore, it has been claimed that after the crea-
Elbow Surgeons scores failed to show any intergroup tion of a bleeding milieu by thorough bursectomy,
differences (Table 4). fibrosis could occur and cause ROM limitation. The
Sonography was performed at 5 weeks after surgery results of our study concur with this hypothesis.
in 69 cases, at 3 months in 66, and at 6 months in 64, Biomolecular studies have indicated that various cy-
and MRI was performed at 1 year in 41 cases. At tokines play roles in inflammation of the SAB.9,13,25,26
5 weeks and at 3 and 6 months after surgery, integrity In these studies, proinflammatory factors were
failures evaluated by sonography were not significantly
different. Similarly, no significant intergroup difference Table 2. Preoperative and Postoperative VAS Pain Scores
was evident by MRI at 1 year after surgery (Table 5).
Marked bursal thickening was more frequently VAS Pain Score
observed in the EB group at 6 months but not at 1 year Limited Extensive
after surgery (Table 6). Bursectomy (n ¼ 36) Bursectomy (n ¼ 35) P Value
Preoperative 4.8  2.5 4.4  1.9 .477
5 wk PO 2.7  1.6 2.7  2.0 .956
Discussion 3 mo PO 3.5  2.2 4.0  2.2 .338
The findings of this study suggest EB during rotator 6 mo PO 2.2  1.6 2.3  1.6 .827
cuff repair may not reduce pain and could lead to 1 yr PO 1.6  1.8 1.0  1.5 .173
external rotation deficits for up to 1 year after surgery, NOTE. Data are presented as mean  standard deviation.
which may be due to the formation of adhesions in the PO, postoperative; VAS, visual analog scale.
6 J-H. NAM ET AL.

Table 3. Postoperative Serial ROM

Limited Bursectomy (n ¼ 36) Extensive Bursectomy (n ¼ 35) P Value

3 mo PO
Forward flexion,  111  25 110  28 .841
External rotation,  20  19 13  18 .156
Internal rotation (spine level reached) 14.5  3.0 14.4  2.8 .948
6 mo PO
Forward flexion,  139  25 128  26 .089
External rotation,  31  15 22  16 .020
Internal rotation (spine level reached) 11.4  2.6 12.6  3.0 .091
1 yr PO
Forward flexion,  154  17 152  17 .646
External rotation,  40  19 27  20 .009
Internal rotation (spine level reached) 9.4  2.4 10.0  2.9 .403
NOTE. Data are presented as mean  standard deviation.
PO, postoperative; ROM, range of motion.

targeted and IL-1b levels were found to be elevated in cuff repair.29 However, in this study we failed to detect
the bursal tissues of patients with rotator cuff disor- any difference between EB and LB in terms of rotator
ders.25,26 A correlation was reported between increased cuff healing.
IL-1b messenger RNA expression in bursae and pre- One should bear in mind that acromioplasty was not
operative ROM and Constant scores,25 and a relation performed in any case during our study, and recently,
was also reported between its expression and VAS pain suspicion was cast on the effect of acromioplasty.30,31 In
scores.26 Other cytokines, such as tumor necrosis factor, a previous study, bursectomy alone without acromio-
IL-1a, IL-6, MMP-1, MMP-9, COX-1, COX-2, and plasty was proposed to be effective in the treatment of
substance P, have also been reported to be increased in rotator cuff disease.8,32
the SAB in rotator cuff disorders.9,13 In line with these
findings, it was claimed in a previous study that simple Limitations
arthroscopic resection of the SAB appears to be as Regarding study limitations, first, LB and EB are
effective as subacromial decompression with acromio- subjective concepts and have no literature basis, and
plasty.27On the other hand, the SAB could provide a thus, the cutoff point used in this study may be debat-
smooth gliding surface between the coracoacromial able. As we verified previously, bursae in the LB group
arch and rotator cuff. Furthermore, the SAB may assist were saved as much as possible to visualize and repair
tendon healing by providing a healing milieu because torn rotator cuffs, whereas in the EB group, bursae
bursal tissue is rich in stem cells, growth factors, and were cleared to the anterior and posterior gutter areas
other biological factors.28,29 Tendons also contain stem and to the far medial part of the rotator cuff muscle and
cells, but their amounts and differentiation capacities from torn cuff areas. Second, rehabilitation is a concern
are considerably less than those in bursae. In addition, because controversy exists regarding the best time to
inflammatory cytokines expressed in bursal tissue could start passive ROM exercise after rotator cuff repair.33
aid tendon healing during the early reparative phase of Some authors have reported that early ROM exercise
does not affect healing rates as compared with delayed
rehabilitation and results in the early restoration of
Table 4. Functional Scores of Both Study Groups
ROM.34,35 Furthermore, it has been suggested recently
Limited Extensive
Bursectomy Bursectomy
(n ¼ 36) (n ¼ 35) P Value Table 5. Integrity of Rotator Cuff Tendons
Constant score
Cuff Integrity: Intact vs Defect, n
Preoperative 60.5  15.5 63.9  13.2 .326
Postoperative (1 yr) 76.6  9.4 77.1  10.1 .856 Limited Extensive
SST score Bursectomy Bursectomy P Value
Preoperative 4.4  2.8 4.9  2.7 .477 5 wk PO 35 vs 1 30 vs 3 .343
Postoperative (1 yr) 9.5  2.7 9.5  3.0 .962 3 mo PO 35 vs 1 27 vs 3 .323
ASES score 6 mo PO 32 vs 0 28 vs 4 .113
Preoperative 60.7  14.6 64.1  15.3 .349 1 yr PO 17 vs 3 18 vs 3 .948
Postoperative (1 yr) 87.8  12.2 91.1  11.3 .280 NOTE. Sonography was performed at 5 weeks, 3 months, and
NOTE. Data are presented as mean  standard deviation. 6 months and magnetic resonance imaging was performed at 1 year
ASES, American Shoulder and Elbow Surgeons; SST, Simple postoperatively.
Shoulder Test. PO, postoperative.
 EXTENSIVE VERSUS LIMITED BURSECTOMY 7

Table 6. Marked Bursal Thickening (>2 mm) in Both Groups 6. Kessel L, Watson M. The painful arc syndrome. Clinical
at Different Times After Surgery classification as a guide to management. Bone Joint J
1977;59:166-172.
Marked Bursal Thickening: Yes vs No, n 7. Neer C II. Anterior acromioplasty for the chronic
Limited Extensive impingement syndrome in the shoulder. 1972. J Bone Joint
Bursectomy Bursectomy P Value Surg Am 2005;87:1399-1399.
5 wk PO 21 vs 15 24 vs 9 .210 8. Henkus HE, de Witte PB, Nelissen RG, Brand R, van
3 mo PO 25 vs 11 26 vs 4 .141 Arkel ER. Bursectomy compared with acromioplasty in
6 mo PO 18 vs 14 27 vs 5 .014 the management of subacromial impingement syndrome:
1 yr PO 15 vs 5 17 vs 4 .719
A prospective randomised study. J Bone Joint Surg Br
NOTE. Sonography was performed at 5 weeks, 3 months, and 2009;91:504-510.
6 months and magnetic resonance imaging was performed at 1 year 9. Blaine TA, Kim YS, Voloshin I, et al. The molecular
postoperatively.
pathophysiology of subacromial bursitis in rotator cuff
PO, postoperative.
disease. J Shoulder Elbow Surg 2005;14:84S-89S.
10. Judge A, Murphy RJ, Maxwell R, Arden NK, Carr AJ.
Temporal trends and geographical variation in the use of
that delayed rehabilitation does not alter functional subacromial decompression and rotator cuff repair of the
outcomes or improve the tendon healing rates of large shoulder in England. J Bone Joint Br 2014;96:70-74.
tears.36-38 In our study an identical rehabilitation pro- 11. Soifer TB, Levy HJ, Soifer FM, Kleinbart F, Vigorita V,
tocol was applied in all study subjects, that is, 5 weeks Bryk E. Neurohistology of the subacromial space.
of immobilization with no passive ROM exercise fol- Arthroscopy 1996;12:182-186.
lowed by gradual rehabilitation, regardless of tear size. 12. Gotoh M, Hamada K, Yamakawa H, Inoue A, Fukuda H.
Third, no imaging follow-up was performed in some Increased substance P in subacromial bursa and shoulder
pain in rotator cuff diseases. J Orthop Res 1998;16:
cases (mostly for financial reasons), although a suffi-
618-621.
cient number were followed up to meet the sample size
13. Voloshin I, Gelinas J, Maloney MD, O’Keefe RJ,
requirements for the determination of primary out- Bigliani LU, Blaine TA. Proinflammatory cytokines and
comes. In addition, at the last imaging follow-up visits, metalloproteases are expressed in the subacromial bursa
MRI was used instead of sonography, which may have in patients with rotator cuff disease. Arthroscopy 2005;21:
influenced measured bursal thicknesses. However, 1076.e1-1076.e9.
group comparisons made at each time point were per- 14. Uhthoff HK, Sarkar K. Surgical repair of rotator cuff
formed using findings obtained using the same ruptures. The importance of the subacromial bursa. J Bone
modalities. Joint Surg Br 1991;73:399-401.
15. Sakai H, Fujita K, Sakai Y, Mizuno K. Immunolocaliza-
tion of cytokines and growth factors in subacromial bursa
Conclusions of rotator cuff tear patients. Kobe J Med Sci 2001;47:
EB during arthroscopic rotator cuff repair appears to 25-34.
have no benefit in terms of reducing pain and may 16. Yanagisawa K, Hamada K, Gotoh M, et al. Vascular
retard motion recovery by promoting adhesion forma- endothelial growth factor (VEGF) expression in the sub-
tion in the subacromial space. In terms of tendon acromial bursa is increased in patients with impingement
integrity, bursectomy extent did not affect the results. syndrome. J Orthop Res 2001;19:448-455.
However, marked bursal thickening was more 17. Han SS, Lee YH, Oh JH, Aminzai S, Kim SH. Randomized,
frequently observed in the EB group. controlled trial of multimodal shoulder injection or
intravenous patient-controlled analgesia after arthro-
scopic rotator cuff repair. Knee Surg Sports Traumatol
References Arthrosc 2013;21:2877-2883.
1. Kennedy MS, Nicholson HD, Woodley SJ. Clinical anat- 18. Tashjian RZ, Hung M, Keener JD, et al. Determining the
omy of the subacromial and related shoulder bursae: A minimal clinically important difference for the American
review of the literature. Clin Anat 2017;30:213-226. Shoulder and Elbow Surgeons score, Simple Shoulder
2. Codman EA. The shoulder: Rupture of the supraspinatus Test, and visual analog scale (VAS) measuring pain after
tendon and other lesions in or about the subacromial bursa. shoulder arthroplasty. J Shoulder Elbow Surg 2017;26:
Malabar, FL: RE Kreiger, 1984. 144-148.
3. Birnbaum K, Lierse W. Anatomy and function of the 19. Oh JH, Kim SH, Kim KH, Oh CH, Gong HS. Modified
bursa subacromialis. Acta Anat (Basel) 1992;145:354-363. impingement test can predict the level of pain reduction
4. Birnbaum K, Prescher A, Heller KD. Anatomic and after rotator cuff repair. Am J Sports Med 2010;38:
functional aspects of the kinetics of the shoulder joint 1383-1388.
capsule and the subacromial bursa. Surg Radiol Anat 20. Oh JH, Kim SH, Lee HK, Jo KH, Bin SW, Gong HS.
1998;20:41-45. Moderate preoperative shoulder stiffness does not alter
5. Hawkins RJ, Abrams JS. Impingement syndrome in the the clinical outcome of rotator cuff repair with arthro-
absence of rotator cuff tear (stages 1 and 2). Orthop Clin scopic release and manipulation. Arthroscopy 2008;24:
North Am 1987;18:373-382. 983-991.
8 J-H. NAM ET AL.

21. White EA, Schweitzer ME, Haims AH. Range of normal rotator cuff tears: A multicenter, randomized controlled
and abnormal subacromial/subdeltoid bursa fluid. trial. J Bone Joint Surg Am 2011;93:1953-1960.
J Comput Assist Tomogr 2006;30:316-320. 31. Gartsman GM, O’Connor DP. Arthroscopic rotator cuff
22. Tsai YH, Huang TJ, Hsu WH, et al. Detection of sub- repair with and without arthroscopic subacromial
acromial bursa thickening by sonography in shoulder decompression: A prospective, randomized study of one-
impingement syndrome. Chang Gung Med J 2007;30: year outcomes. J Shoulder Elbow Surg 2004;13:424-426.
135-141. 32. Donigan JA, Wolf BR. Arthroscopic subacromial decom-
23. Milgrom C, Schaffler M, Gilbert S, van Holsbeeck M. pression: Acromioplasty versus bursectomy alonedDoes
Rotator-cuff changes in asymptomatic adults. The effect of it really matter? A systematic review. Iowa Orthop J
age, hand dominance and gender. J Bone Joint Surg Br 2011;31:121.
1995;77:296-298. 33. Koo SS, Parsley BK, Burkhart SS, Schoolfield JD.
24. Muir SW, Corea CL, Beaupre L. Evaluating change in Reduction of postoperative stiffness after arthroscopic
clinical status: Reliability and measures of agreement for rotator cuff repair: Results of a customized physical ther-
the assessment of glenohumeral range of motion. N Am J apy regimen based on risk factors for stiffness. Arthroscopy
Sports Phys Ther 2010;5:98-110. 2011;27:155-160.
25. Ko JY, Wang FS, Huang HY, Wang CJ, Tseng SL, Hsu C. 34. Lastayo PC, Wright T, Jaffe R, Hartzel J. Continuous
Increased IL-1b expression and myofibroblast recruitment passive motion after repair of the rotator cuff. A pro-
in subacromial bursa is associated with rotator cuff lesions spective outcome study. J Bone Joint Surg Am 1998;80:
with shoulder stiffness. J Orthop Res 2008;26:1090-1097. 1002-1011.
26. Gotoh M, Hamada K, Yamakawa H, et al. Interleukin-1- 35. Arndt J, Clavert P, Mielcarek P, et al. Immediate passive
induced glenohumeral synovitis and shoulder pain in motion versus immobilization after endoscopic supra-
rotator cuff diseases. J Orthop Res 2002;20:1365-1371. spinatus tendon repair: A prospective randomized study.
27. Budoff JE, Nirschl RP, Guidi EJ. Current concepts review- Orthop Traumatol Surg Res 2012;98:S131-S138.
debridement of partial-thickness tears of the rotator cuff 36. Parsons BO, Gruson KI, Chen DD, Harrison AK,
without acromioplasty. Long-term follow-up and review Gladstone J, Flatow EL. Does slower rehabilitation after
of the literature. J Bone Joint Surg Am 1998;80:733-748. arthroscopic rotator cuff repair lead to long-term stiffness?
28. Song N, Armstrong AD, Li F, Ouyang H, Niyibizi C. J Shoulder Elbow Surg 2010;19:1034-1039.
Multipotent mesenchymal stem cells from human sub- 37. Chang K-V, Hung C-Y, Han D-S, Chen W-S, Wang T-G,
acromial bursa: Potential for cell based tendon tissue en- Chien K-L. Early versus delayed passive range of motion
gineering. Tissue Eng Part A 2014;20:239-249. exercise for arthroscopic rotator cuff repair: A meta-
29. Patel S, Gualtieri AP, Lu HH, Levine WN. Advances in analysis of randomized controlled trials. Am J Sports Med
biologic augmentation for rotator cuff repair. Ann N Y 2015;43:1265-1273.
Acad Sci 2016;1383:97-114. 38. Cuff DJ, Pupello DR. Prospective randomized study of
30. MacDonald P, McRae S, Leiter J, Mascarenhas R, arthroscopic rotator cuff repair using an early versus
Lapner P. Arthroscopic rotator cuff repair with and delayed postoperative physical therapy protocol.
without acromioplasty in the treatment of full-thickness J Shoulder Elbow Surg 2012;21:1450-1455.