Rotator Cuff is Antagonist Following RTSA
1
ABSTRACT: 1
Introduction: There is disagreement regarding whether, when possible, the rotator cuff should be 2
repaired in conjunction with reverse total shoulder arthroplasty (RTSA). Therefore, we 3
investigated the effects of rotator cuff repair in RTSA models with varying magnitudes of humeral 4
and glenosphere lateralization (HLat & GLat). 5
Methods: Six fresh-frozen cadaveric shoulders were tested on a validated in-vitro muscle driven 6
motion simulator. Each specimen was implanted with a custom adjustable, load-sensing RTSA 7
after creation of a simulated rotator cuff tear. The effects of 4 RTSA configurations (0&10mm x 8
HLat & GLat) on deltoid force and joint load during abduction with/without rotator cuff repair 9
were assessed. 10
Results: Increasing HLat and GLat significantly affected deltoid force (-2.5±1.7%BodyWeight 11
(%BW), p=0.016 & +7.7±5.6%BW, p=0.016). Rotator cuff repair interacted with HLat & GLat 12
(p=0.005) such that with no HLat, GLat increased deltoid force without cuff repair (8.1±5.1%BW, 13
p=0.012) and this effect was increased with cuff repair (12.8±7.8%BW, p=0.010), but addition of 14
HLat mitigated this effect. Rotator cuff repair increased joint load (+11.9±5.1%BW, p=0.002), as 15
did GLat (+13.3±3.7%BW, p<0.001), and these interacted such that increasing GLat markedly 16
increases cuff repair's negative effects (9.4±3.2%BW, p=0.001 vs 14.4±7.4%BW, p=0.005). 17
Conclusion: Rotator cuff repair, especially in conjunction with GLat, produces an antagonistic 18
effect that increases deltoid and joint loading. The long-term effects of this remain unknown; 19
however, combining these factors may prove undesirable. HLat improves joint compression 20
through deltoid wrapping and increases the deltoid’s mechanical advantage. Therefore, it could be 21
used in place of rotator cuff repair, thus avoiding its complications. 22
Level of Evidence: Basic Science Study 23
Rotator Cuff is Antagonist Following RTSA
2
Keywords: Reverse Total Shoulder Arthroplasty, RTSA, RSA, shoulder, rotator cuff, simulator, 24
cuff tear arthropathy 25
26
Rotator Cuff is Antagonist Following RTSA
3
INTRODUCTION: 27
Reverse Total Shoulder Arthroplasty (RTSA) is primarily indicated for the treatment of rotator 28
cuff tear arthropathy or massive rotator cuff tears that are deemed irrepairable9,15,17,20. Despite 29
these indications, it is often possible to repair portions of the subscapularis and infraspinatus/teres 30
minor; however, there is disagreement regarding whether these tissues should be repaired as their 31
effects on RTSA biomechanics and outcomes remain unclear6,7,10,16. Additionally, the indications 32
for RTSA have expanded to include surgical conditions with an intact rotator cuff, such as the 33
management of A2, B2, or C glenoid erosions8,22. As such, the surgeon has the option to preserve 34
or release the rotator cuff in these scenarios. 35
Some have advocated repair of these tissues on the basis that they increase RTSA stability and 36
decrease the incidence of dislocation6,10,21,23 but clinical series by Clark et al. and Wall et al. have 37
disputed this effect7,24. However, a review by Wall et al. did find that repair of the subscapularis 38
may still be warranted because it significantly improves post-operative internal rotation24. In 39
contrast, Boulahia et al.5 suggested that subscapularis repair may detrimentally affect external 40
rotation through antagonistic loading against the already weakened posterior cuff; however, this 41
has not been confirmed in other reports, which specifically investigated subscapularis repair7,24. 42
Although to date the discussion of whether to repair the rotator cuff has primarily focused on 43
post-operative joint stability, the potential effect on muscle and joint loading must also be 44
considered. As is the case in the native glenohumeral joint, the concentric loads applied by a 45
repaired rotator cuff can be expected to counter the deltoid’s eccentric joint loads. However, in an 46
in-vitro study of one RTSA implant configuration, Ackland et al. demonstrated that the function 47
of the subscapularis is markedly shifted towards adduction – especially early in motion – compared 48
to its native role1. This finding suggests that repair of the subscapularis may resist abduction and 49
Rotator Cuff is Antagonist Following RTSA
4
thus increase muscle and joint loading, but it is unclear if this change in function holds across the 50
full range of RTSA configurations used clinically. 51
With the conflicting clinical information and the relative paucity of biomechanical evidence in 52
mind, we sought to investigate the effects of rotator cuff repair, or preservation in glenoid erosion 53
cases, on functional shoulder outcomes and joint kinetics. As well, we investigated how these 54
effects are influenced by changes in two geometrical implant parameters that have previously been 55
shown to have a strong influence on shoulder biomechanics. Specifically, we wanted to clarify 56
how rotator cuff repair affects active internal/external rotation range of motion and external 57
rotation strength, while also determining if it had a detrimental impact on deltoid and joint loading. 58
We hypothesized that rotator cuff repair would resist abduction thus increasing deltoid muscle 59
force requirements and the resulting joint load. As well, we hypothesized that glenosphere 60
lateralization would have no effect on internal/external rotation but would exacerbate the negative 61
effects of rotator cuff loading while humeral lateralization would improve internal/external 62
rotation and mitigate the effects of rotator cuff repair. 63
Rotator Cuff is Antagonist Following RTSA
5
MATERIALS & METHODS: 64
Instrumented RTSA Implant 65
In this in-vitro biomechanical study, it was possible to measure joint loads and investigate the 66
effects of systematic adjustments to implant geometry using a previously described custom 67
modular implant system with a built-in load sensor (Figure 1)14,19. Four combinations of humeral 68
and glenosphere lateralization were investigated (respectively: 0&0; 0&10; 10&0; 10&10mm) 69
where, when both variables are at 0mm, the configuration is considered to be neutral, 70
corresponding to a traditional Grammont-style implant. Specifically, neutral was defined as the 71
glenoid baseplate level with the inferior glenoid rim, the glenosphere center of rotation coincident 72
with the glenoid surface, neutral humeral version with a 155° head-neck angle, and a 12.5-mm 73
lateral offset between the humeral stem and deepest point of the cup. The 10mm offset 74
configurations were achieved by making mechanical adjustments to the custom implant without 75
altering the surgical fixation. To ensure accurate mechanical properties, commercially available 76
polyethylene humeral cups (38mm, Delta Xtend; DePuy, Warsaw, IN, USA) were used. The 77
glenosphere was custom fabricated to accommodate a 6 degree of freedom load cell (Nano25, 78
ATI-IA; Apex, NC, USA) that attached medially to a glenoid fixation baseplate, which was 79
recessed into the glenoid vault to allow neutral glenosphere positioning. 80
81
Active Motion Simulator & Specimen Preparation 82
Six fresh frozen cadaveric shoulders (60±21 years) without signs of cuff deficiency or prior surgery 83
were prepared and the humerus was transected distal to the deltoid tuberosity. To enable repeated 84
access to the glenohumeral joint throughout the testing protocol, the subscapularis muscle was 85
elevated from the scapula and reflected laterally without disrupting its insertion on the lesser 86
Rotator Cuff is Antagonist Following RTSA
6
tuberosity. A full thickness superior rotator cuff tear was simulated by releasing the entire 87
supraspinatus and upper portion of the infraspinatus from the greater tuberosity. The specimens 88
were then implanted with the above described custom adjustable RTSA implant positioned in the 89
neutral configuration including 0° retroversion relative to the trans-epicondylar axis and with 90
humeral distalization dictated by aligning the superior humeral cup with the superior aspect of the 91
greater tuberosity for all specimens. Following implantation, the three deltoid heads were sutured 92
at their insertion, and the subscapularis and inferior infraspinatus/teres minor musculotendinous 93
junctions were sutured across their width using a running locking stitch. Specimen preparation was 94
then completed as described by Giles et al,12 including fixation of optical trackers to the scapula 95
(Optotrak Certus, NDI, Waterloo, ON, Canada) and insertion of an instrumented intramedullary 96
humeral rod that could provide optical motion tracking and data regarding the loads applied to the 97
rod by the experimenter. These data were recorded using a six degree of freedom load cell (Mini45, 98
ATI-IA; Apex, NC, USA) interposed between the proximal rod, which was inserted into the 99
humeral canal, and the distal rod, which was mated to the testing simulator during passive testing 100
and fitted with brass weights – to simulate the mass of the transected distal arm – during active 101
motion testing. The scapula was cemented to the simulator, and all muscles were connected to 102
computer controlled pneumatic actuators through physiologically accurate lines of action. This 103
simulator produces accurate and repeatable motions along predefined glenohumeral rotation 104
profiles by using a previously validated multi-PID (Proportional-Integral-Differential) control 105
system, which employs real time feedback of the humerus and scapula orientation, to produce 106
independent loads for each muscle group 13. The simulator also continuously rotated the scapula 107
to maintain a physiologically accurate glenohumeral-to-scapulothoracic rhythm. 108
109
Rotator Cuff is Antagonist Following RTSA
7
Testing Protocol 110
Four RTSA configurations with 0 or 10mm of glenosphere lateralization and 0 or 10mm of 111
humeral lateralization were tested in random order. For each configuration, fully unconstrained 112
and simulator-guided muscle driven motions and strength tests were conducted. Fully 113
unconstrained active abduction was simulated from an adducted position (lowest level of 114
abduction without humeral cup impingement) to 90° of humerothoracic abduction at a rate of 1°/s. 115
To ensure proper joint loading was achieved, a physiologic glenohumeral-to-scapulothoracic 116
rhythm of 2:1 was maintained by rotating the scapula using feedback of the instantaneous level of 117
glenohumeral abduction. Additionally, to assess the effect of repairing the rotator cuff, active 118
motion trials were performed with and without (order of testing cuff repair was randomized) 119
subscapularis and infraspinatus/teres minor loading. 120
Simulator-guided muscle driven active internal and external rotational range of motion (IR/ER 121
ROM) was assessed in full adduction by attaching the distal humeral rod (with weights removed) 122
to a previously described guiding arc12 that constrained ad-abduction and flexion-extension while 123
leaving axial rotation free to vary. To realistically replicate the pattern of muscle forces present 124
during IR/ER motions, muscle loading ratios were calculated using previously reported pCSA 125
(physiological Cross-Sectional Area) data2 and EMG signals recorded during IR and ER motions11. 126
The resulting ratios were, for internal rotation: 1, 0.16, 0.16, 0, 0, and for external rotation: 0.32, 127
1, 0, 0.22, 0.3, for the subscapularis, infraspinatus/teres minor, anterior, middle, and posterior 128
deltoid, respectively. Maximum rotational ROM for each motion was recorded after the load of all 129
muscles was increased until the prime mover – the subscapularis for internal rotation and 130
infraspinatus/teres-minor for external rotation – was loaded to 50N. External rotation strength was 131
also assessed by applying the same ramp loading protocol while holding the humerus in neutral 132
Rotator Cuff is Antagonist Following RTSA
8
rotation using the simulator’s guiding arc and recording the maximum axial torque measured by 133
the humeral rod’s load cell. 134
135
Outcome Variables and Statistical Analyses 136
For active abduction, the effects of rotator cuff repair status (yes/no), and humeral and glenosphere 137
lateralization on resultant joint load, total deltoid muscle force, and joint load angle were assessed. 138
The force components recorded by the glenosphere load cell were transformed into a coordinate 139
system aligned with a standard glenoid coordinated system, which was coincident with the 140
glenosphere center. The resultant joint load was then calculated from these data and the joint 141
loading angle was found using the superiorly and laterally directed forces such that a 0° angle is 142
purely compressive, and positive angles are directed superiorly. The total deltoid force outcome 143
was calculated through the summation of the force on the three deltoid heads at each point in time. 144
Resultant joint load and total deltoid force data were subsequently expressed as a percentage of 145
the specimen’s total body weight (%BW). All outcomes for active abduction were evaluated every 146
7.5° of humerothoracic abduction (5° glenohumeral and 2.5° scapulothoracic) from 15° to 82.5° 147
as this was the range achieved by all conditions without implant impingement. Each active motion 148
outcome variable was statistically tested using a 4-way (cuff status, humeral lateralization, 149
glenosphere lateralization, abduction level) repeated-measures analysis of variance (RM-150
ANOVA) with follow-up analyses of interactions where appropriate and Bonferroni-corrected 151
pairwise comparisons with significance set at p<0.05. A power analysis for each outcome variable 152
indicated that testing of six shoulder specimen would enable a clinically meaningful change of 153
5%BW to be detected with ≥80% power. 154
Rotator Cuff is Antagonist Following RTSA
9
For active range of motion trials, the reported ROM value was taken as the amount of rotation 155
from the neutral position achieved during maximal muscle loading. Active external rotation 156
strength was recorded as the torque applied to the humeral rod in Newton-Meters (Nm) during 157
maximal muscle loading. For consistency, the ROM and strength outcomes were recorded after 158
holding the maximum muscle loads for 5 seconds. For each of these three outcomes, a 2-way 159
(humeral lateralization, glenosphere lateralization) RM-ANOVA was performed with analyses of 160
interactions where appropriate. 161
162
163
Rotator Cuff is Antagonist Following RTSA
10
RESULTS: 164
165
Active Abduction: 166
Deltoid Force 167
Results for total deltoid force indicated that rotator cuff loading (i.e. clinical rotator cuff repair or 168
intact anterior and posterior cuff) did not have a significant main effect (p=0.3) when assessed 169
across all implant configurations and abduction angles; however, in addition to the expected main 170
effect of changing abduction angle (p=0.001), both humeral and glenosphere lateralization 171
significantly affected deltoid force (p<0.021). Specifically, a 10mm increase in humeral 172
lateralization significantly decreased deltoid force (average±SD: 3±2%BW, p=0.016) – although 173
this decrease cannot be considered clinically meaningful – while a 10mm increase in glenosphere 174
lateralization significantly increased deltoid force (8±5%BW, p=0.016). Additionally, there was a 175
significant three-way interaction between cuff status, humeral lateralization, and glenosphere 176
lateralization (p=0.005, Figure 2) such that when there was no humeral lateralization, increasing 177
glenosphere lateralization significantly increased deltoid force without rotator cuff loading 178
(8±5%BW, p=0.012) and this effect was exacerbated when cuff loads were applied (13±8%BW, 179
p=0.010). However, this negative effect on deltoid force was mitigated by humeral lateralization 180
to the extent that deltoid force increases due to glenosphere lateralization were no longer 181
significant (p>0.05). 182
183
Resultant Joint Load 184
In contrast to the effects on total deltoid force, resultant joint load data does indicate that rotator 185
cuff loading has a significant main effect (p=0.002) as does abduction (p<0.001) and glenosphere 186
Rotator Cuff is Antagonist Following RTSA
11
lateralization (p<0.001), but not humeral lateralization (p=0.195). Repairing the rotator cuff 187
resulted in a 12±5%BW (p=0.002) increase in resultant joint load across all implant configurations, 188
while increasing glenosphere lateralization increased joint load by 13±4%BW (p<0.001). 189
Additionally, cuff status and glenosphere lateralization interacted (Figure 3) such that increasing 190
glenosphere lateralization significantly increases the negative effect of loading the rotator cuff 191
(Unloaded vs Loaded: 0mm=9±3%BW (p=0.001), 10mm=14±7%BW (p=0.005)). 192
193
Joint Load Angle 194
Analysis of the joint load angle data found that abduction (p<0.001), rotator cuff loading 195
(Unloaded vs Loaded: 40±14° vs 34±13°, p<0.001), and humeral lateralization (0 vs 10mm: 196
40±13° vs 33±14°, p<0.001; Figure 4) all have significant main effects such that load angle 197
becomes more compressive as these factors increase but glenosphere lateralization had no effect 198
(0 vs 10mm: 37±14° vs 37±13°, p=0.83). Both cuff status and glenosphere lateralization 199
significantly interact with abduction angle (p<0.001) such that their effects on joint load angle are 200
reduced as abduction progresses. Conversely, the effect of humeral lateralization was almost 201
completely constant (variation <0.7°) across abduction. 202
203
Active Internal-External Rotation: 204
Range of Motion & Strength 205
A two-way RM-ANOVA for active internal rotation found that neither implant variable had a 206
significant main effect (p>0.05) but humeral and glenosphere lateralization did produce a 207
significant cross-over interaction (p=0.002; Figure 5). Specifically, with 0mm of glenosphere 208
lateralization, increasing humeral lateralization by 10mm significantly increased range of motion 209
Rotator Cuff is Antagonist Following RTSA
12
(7±5°, p=0.019). However, with 10mm of glenosphere lateralization, the lateralized humeral 210
configuration produced less internal rotation compared to the neutral (i.e. 0mm) humeral 211
configuration (5±7°), but this difference was not significant (p=0.092). 212
Similar analysis of the active external rotation data found that it was significantly increased by 213
both humeral (10±6°, p=0.008) and glenosphere lateralization (10±9°, p=0.033) (Figure 5). 214
A two-way RM-ANOVA for active external rotation strength found that neither humeral 215
lateralization (p=0.99) or glenosphere lateralization (p=0.61) produced a significant main effect. 216
217
Rotator Cuff is Antagonist Following RTSA
13
DISCUSSION: 218
Although RTSA is primarily indicated as a treatment for rotator cuff tear arthropathy or massive 219
cuff tears, in many cases, portions of the degenerated infraspinatus and subscapularis are amenable 220
to repair despite having decreased contractile capabilities. Previous studies, focusing on the effects 221
of cuff repair on RTSA stability/dislocation and internal-external rotation range of motion5–222
7,10,21,23,24, have disagreed on whether these tissues should be repaired when possible. This study is 223
unique as it focuses on the kinetic effects of rotator cuff repair and how this is affected by RTSA 224
configuration. We have found that rotator cuff repair increases the demands on the deltoid during 225
abduction and this in turn increases joint loading. This effect is further exacerbated by glenosphere 226
lateralization. Conversely, humeral lateralization reduced the deltoid muscle’s required force, 227
which mitigated the increases in joint loading caused by rotator cuff repair. Furthermore, humeral 228
lateralization produced a more compressive joint load, which effectively stabilizes the joint, but 229
did so without increasing joint load magnitude. 230
Effects on Active Abduction 231
As expected, repair and loading of the rotator cuff significantly (p=0.001) shifted the joint 232
loading angle towards compression (~6°). Additionally, results indicate that increasing humeral 233
lateralization from neutral to 10mm produces a similar compressive shift in joint angle of 7° 234
irrespective of cuff repair status, but if combined with cuff repair, the loading angle can be shifted 235
~13° towards compression. Shifting of the joint load towards greater compression is desirable as 236
it is less challenging to glenosphere baseplate fixation, which is especially important to the 237
promotion of osseous integration in the early post-operative phase. These data also suggest that 238
both cuff repair and humeral lateralization are effective at stabilizing an RTSA by producing a 239
more compressive overall joint load, which increases concavity compression. However, while 240
Rotator Cuff is Antagonist Following RTSA
14
rotator cuff repair produces this affect through larger muscle and joint loading, humeral 241
lateralization causes the deltoid to wrap around the greater tuberosity to a greater extent and this 242
redirects its tension through the humeral head and articulation thus resulting in a more 243
compressively directed joint load without introducing additional force into the joint (Figure 6). 244
Therefore, with respect to joint load angle which affects functional stability and implant fixation, 245
repair of the rotator cuff and use of humeral lateralization are equally effective; however, these 246
factors must be viewed in light of the manner in which they produce these effects and their other 247
impacts on RTSA biomechanics. 248
As previously reported, these data demonstrate that glenosphere lateralization significantly 249
increases joint loads (14%BW) as a result of decreasing the deltoid’s mechanical advantage14,18,19. 250
Rotator cuff repair was also found to significantly increase joint load magnitude (12%BW). 251
Additionally, the interaction between these two factors meant that if the rotator cuff was repaired, 252
and the glenosphere was lateralized – for instance to avoid scapular notching – the joint load 253
increased by approximately 29%. The sensitivity of humeral polyethylene liners to increases in 254
load has not be clearly defined in the literature, an increase of this magnitude may lead to increased 255
wear3. 256
Although repairing the cuff caused a significant increase in joint load magnitude, this outcome 257
alone does not clearly elucidate the cuff’s role in abduction following RTSA, as this increase may 258
be attributable to the added rotator cuff loading, or, alternatively, could be an indicator that the 259
cuff acts as an adductor thus increasing the required deltoid force and thus the overall joint load. 260
This distinction can be made by looking at deltoid muscle force. Specifically, when assessing the 261
average effect across all RTSA configurations, these data suggest that the cuff is not an adductor 262
as there are no significant increases in deltoid force when the cuff is repaired. However, there is a 263
Rotator Cuff is Antagonist Following RTSA
15
significant interaction between cuff repair status, humeral and glenosphere lateralization such that 264
increases in deltoid force due to cuff repair (i.e. the slope of the lines in Figure 2) were exacerbated 265
when glenosphere lateralization was increased (~5%BW) irrespective of the level of humeral 266
lateralization. This demonstrates that glenosphere lateralization (through implant or bony means – 267
as in the BIO-RSA which uses a bone graft to achieve lateralization) causes the rotator cuff to act 268
as an adductor and antagonist to the deltoid. Therefore, the combination of these factors should be 269
employed with caution due to the ~13%BW increase in the deltoid force required to produce 270
motion. 271
Interestingly, the three-factor interaction in the deltoid force data also indicates that when the 272
glenosphere is lateralized, although humeral lateralization did not alter cuff repair’s effect of 273
increasing deltoid force, it did decrease the magnitude of deltoid load nearly equally for both cuff 274
repair states (~5%BW). This consistent decrease in deltoid force caused by adjusting humeral 275
lateralization indicates that it increases the deltoid mechanical advantage, thus decreasing the 276
deltoid’s required force, but it does not markedly change the antagonistic adduction effect of the 277
rotator cuff. Therefore, whereas the concurrent use of rotator cuff repair and glenosphere 278
lateralization should be considered cautiously, the addition of humeral lateralization in cases where 279
the glenosphere is lateralized can improve function and reduce deltoid fatigue irrespective of 280
rotator cuff repair status. 281
282
Effects on Internal-External Rotation 283
Evaluation of internal rotation range of motion using physiologically accurate muscle loading 284
ratios found that neither implant variable significantly affected motion but that they interacted with 285
one another in a cross-over pattern caused by changing combinations of impingement and soft 286
Rotator Cuff is Antagonist Following RTSA
16
tissue restraint. Specifically, IR could be increased by lateralizing the glenosphere (~7°) as a result 287
of decreased contact with the anterior glenoid and coracoid, or by lateralizing the humerus (~10°) 288
due to increases in the cuff’s IR moment arms. However, IR is reduced compared to these two 289
configurations (by ~2° and ~5°, respectively), when both parameters are lateralized, due to 290
overstuffing of the joint and over tensioning of the posterior soft tissues. As such, over tensioning 291
of the posterior soft tissues in the horizontal plane results in a posterior tether to internal rotation. 292
In the case of external rotation range of motion, both humeral and glenosphere lateralization 293
produced significant and equivalent increases (10°, respectively). These increases can be explained 294
similarly to those initially seen with internal rotation; however, in this case, rotation was not 295
limited by over tensioning of the anterior soft tissues when both lateralizations were applied as 296
was the case with internal rotation. Therefore, these results help to confirm the clinical findings of 297
Boileau et al. that BIO-RSA can increase motion by decreasing impingement4, but also supports 298
the value of humeral lateralization in increasing motion by improving the rotator cuff’s mechanical 299
advantage. However, it has been shown that these improvements in internal rotation can be 300
partially negated by combined lateralization. As such, it is important to specifically assess IR range 301
of motion when using lateralized glenospheres or a BIO-RSA and to appropriately manage the soft 302
tissues. If IR is restricted with lateralization, it may be prudent to release contracted posterior 303
capsular tissues. 304
External rotation strength is critically important to performing activities of daily living and thus 305
was of interest in this protocol; however, neither implant variable assessed in this study showed 306
any effect on strength. Despite not effecting external rotation strength in the tested neutral position, 307
it is possible that humeral lateralization increases external rotation strength with the arm in 308
externally rotated orientations by causing the greater tuberosity and cuff insertion to lay more 309
Rotator Cuff is Antagonist Following RTSA
17
eccentric relative to the cuff’s line of action, which should increase its mechanical advantage and 310
thus strength. However, further investigations are required to test this hypothesis. 311
312
Limitations & Future Work 313
The results of this study should be viewed in light of its limitations. First, as with all in-vitro 314
testing, the use of cadaveric specimens may affect results; specifically, in this study specimens did 315
not exhibit existing rotator cuff tears and thus the testing of lateralized RTSA configurations with 316
simulated rotator cuff repair did not result in the large passive joint forces that could occur 317
clinically due to tendon retraction. However, from a clinical standpoint, over tensioned soft tissues 318
often relax post-operatively and thus this study’s rotator cuff repair model can be considered 319
analogous to this relaxed state. Second, in conditions where rotator cuff repair was simulated, the 320
muscle loads used during testing were guided by previously reported EMG data for healthy 321
patients as none exist post-RTSA. The use of these healthy EMG signals may influence the 322
magnitude of the observed trends but because all conditions were tested with the same EMG 323
patterns, comparisons between implant configurations should still yield meaningful results. Third, 324
the current simulator is not able to simulate all shoulder muscle groups, specifically 325
humerothoracic muscles, thus measurements of active IR-ER rotation could only form a basis for 326
comparison between implant configurations and not to define the precise range of motion. Fourth, 327
the current protocol only assessed external rotation strength in neutral rotation, which did not show 328
an affect; however, further investigations in other orientations may have more comprehensively 329
described the effects of the tested conditions. Finally, we believe that further research is required 330
to elucidate the effects on active abduction of independently repairing the subscapularis and 331
Rotator Cuff is Antagonist Following RTSA
18
infraspinatus muscles, which would require simulation of humerothoracic muscles not currently 332
included in this model. 333
Rotator Cuff is Antagonist Following RTSA
19
CONCLUSION 334
Rotator cuff repair in the setting of RTSA produces an antagonistic effect that increases the deltoid 335
muscle’s work during elevation and this effect is exacerbated by increased glenosphere 336
lateralization. The combination of these rotator cuff forces and the increased demands on the 337
deltoid muscle significantly increases joint loads. The long-term effects of these changes remain 338
unknown but their magnitude is likely to be clinically meaningful. Humeral lateralization, 339
however, produces increased deltoid wrapping that results in improved joint compression and 340
stability – without the added muscle and joint loading associated with rotator cuff repair – and 341
increases the deltoid’s mechanical advantage which reduces the deltoid force required for 342
elevation. Therefore, humeral lateralization could be used in conjunction with rotator cuff repair 343
to maximize joint stability, or in its place to produce a similar stabilizing effect while decreasing 344
the demands on the deltoid and avoiding complications associated with rotator cuff repair. 345
346
Rotator Cuff is Antagonist Following RTSA
20
REFERENCES 347
1. Ackland DC, Roshan-Zamir S, Richardson M, Pandy MG. Moment arms of the shoulder 348
musculature after reverse total shoulder arthroplasty. J. Bone Joint Surg. Am. 349
2010;92(5):1221–1230. doi:10.2106/JBJS.I.00001 350
2. Aluisio F V, Osbahr DC, Speer KP. Analysis of rotator cuff muscles in adult human 351
cadaveric specimens. Am. J. Orthop. (Belle Mead. NJ). 2003;32(3):124–9. 352
3. Archard JF. Contact and Rubbing of Flat Surfaces. J. Appl. Phys. 1953;24(8):981. 353
doi:10.1063/1.1721448 354
4. Boileau P, O’Shea K, Moineau G, Roussane Y. Bony Increased-Offset Reverse Shoulder 355
Arthroplasty (BIO-RSA) for Cuff Tear Arthropathy. Oper. Tech. Orthop. 2011;21(1):69–356
78. doi:10.1053/j.oto.2010.11.003 357
5. Boulahia A, Edwards TB, Walch G, Baratta R V. Early results of a reverse design 358
prosthesis in the treatment of arthritis of the shoulder in elderly patients with a large 359
rotator cuff tear. Orthopedics. 2002;25(2):129–33. 360
6. Chalmers PN, Rahman Z, Romeo AA, Nicholson GP. Early dislocation after reverse total 361
shoulder arthroplasty. J. Shoulder Elbow Surg. 2014;23(5):737–44. 362
doi:10.1016/j.jse.2013.08.015 363
7. Clark JC, Ritchie J, Song FS, Kissenberth MJ, Tolan SJ, Hart ND, et al. Complication 364
rates, dislocation, pain, and postoperative range of motion after reverse shoulder 365
arthroplasty in patients with and without repair of the subscapularis. J. Shoulder Elbow 366
Surg. 2012;21(1):36–41. doi:10.1016/j.jse.2011.04.009 367
8. Denard PJ, Walch G. Current concepts in the surgical management of primary 368
glenohumeral arthritis with a biconcave glenoid. J. Shoulder Elbow Surg. 369
Rotator Cuff is Antagonist Following RTSA
21
2013;22(11):1589–98. doi:10.1016/j.jse.2013.06.017 370
9. Drake GN, O’Connor DP, Edwards TB. Indications for reverse total shoulder arthroplasty 371
in rotator cuff disease. Clin. Orthop. Relat. Res. 2010;468(6):1526–33. 372
doi:10.1007/s11999-009-1188-9 373
10. Edwards TB, Williams MD, Labriola JE, Elkousy HA, Gartsman GM, O’Connor DP. 374
Subscapularis insufficiency and the risk of shoulder dislocation after reverse shoulder 375
arthroplasty. J. Shoulder Elbow Surg. 2009;18(6):892–6. doi:10.1016/j.jse.2008.12.013 376
11. Escamilla RF, Yamashiro K, Paulos L, Andrews JR. Shoulder muscle activity and 377
function in common shoulder rehabilitation exercises. Sport. Med. 2009;39(8):663–685. 378
12. Giles JW, Boons HW, Elkinson I, Faber KJ, Ferreira LM, Johnson J a., et al. Does the 379
dynamic sling effect of the Latarjet procedure improve shoulder stability? A 380
biomechanical evaluation. J. Shoulder Elb. Surg. 2013;22(6):821–827. 381
doi:10.1016/j.jse.2012.08.002 382
13. Giles JW, Ferreira LM, Athwal GS, Johnson JA. Development and Performance 383
Evaluation of a Multi-PID Muscle Loading Driven In Vitro Active-Motion Shoulder 384
Simulator and Application to Assessing Reverse Total Shoulder Arthroplasty. J. Biomech. 385
Eng. 2014;136(12):121007. doi:10.1115/1.4028820 386
14. Giles JW, Langohr GDG, Johnson JA, Athwal GS. Implant Design Variations in Reverse 387
Total Shoulder Arthroplasty Influence the Required Deltoid Force and Resultant Joint 388
Load. Clin. Orthop. Relat. Res. 2015;473(11):3615–3626. doi:10.1007/s11999-015-4526-389
0 390
15. Gladstone JN, Bishop JY, Lo IKY, Flatow EL. Fatty infiltration and atrophy of the rotator 391
cuff do not improve after rotator cuff repair and correlate with poor functional outcome. 392
Rotator Cuff is Antagonist Following RTSA
22
Am. J. Sports Med. 2007;35(5):719–28. doi:10.1177/0363546506297539 393
16. Grassi FA, Zorzolo I. Reverse shoulder arthroplasty without subscapularis repair for the 394
treatment of proximal humeral fractures in the elderly. Musculoskelet. Surg. 395
2014;98(S1):5–13. doi:10.1007/s12306-014-0321-4 396
17. Harreld KL, Puskas BL, Frankle M. Massive rotator cuff tears without arthropathy: when 397
to consider reverse shoulder arthroplasty. J. Bone Joint Surg. Am. 2011;93(10):973–84. 398
18. Hoenecke HR, Flores-Hernandez C, D’Lima DD. Reverse total shoulder arthroplasty 399
component center of rotation affects muscle function. J. Shoulder Elb. Surg. 2014;1–8. 400
doi:10.1016/j.jse.2013.11.025 401
19. Langohr GDG, Giles JW, Athwal GS, Johnson JA. The effect of glenosphere diameter in 402
reverse shoulder arthroplasty on muscle force, joint load, and range of motion. J. Shoulder 403
Elbow Surg. 2014; doi:10.1016/j.jse.2014.10.018 404
20. Mulieri P, Dunning P, Klein S, Pupello D, Frankle M. Reverse shoulder arthroplasty for 405
the treatment of irreparable rotator cuff tear without glenohumeral arthritis. J. Bone Joint 406
Surg. Am. 2010;92(15):2544–56. doi:10.2106/JBJS.I.00912 407
21. Oh JH, Shin S-J, McGarry MH, Scott JH, Heckmann N, Lee TQ. Biomechanical effects of 408
humeral neck-shaft angle and subscapularis integrity in reverse total shoulder arthroplasty. 409
J. Shoulder Elbow Surg. 2014;23(8):1091–8. doi:10.1016/j.jse.2013.11.003 410
22. Sears BW, Johnston PS, Ramsey ML, Williams GR. Glenoid Bone Loss in Primary Total 411
Shoulder Arthroplasty: Evaluation and Management. J. Am. Acad. Orthop. Surg. 412
2012;20(9):604–613. doi:10.5435/JAAOS-20-09-604 413
23. Trappey GJ, O’Connor DP, Edwards TB. What are the instability and infection rates after 414
reverse shoulder arthroplasty? Clin. Orthop. Relat. Res. 2011;469(9):2505–11. 415
Rotator Cuff is Antagonist Following RTSA
23
doi:10.1007/s11999-010-1686-9 416
24. Wall B, Nové-Josserand L, O’Connor DP, Edwards TB, Walch G. Reverse total shoulder 417
arthroplasty: a review of results according to etiology. J. Bone Joint Surg. Am. 418
2007;89(7):1476–85. doi:10.2106/JBJS.F.00666 419
420
Rotator Cuff is Antagonist Following RTSA
24
421
FIGURE & TABLE LEGENDS 422
Figure 1 - Computer rendering of the custom instrumented RTSA composed of a glenosphere 423
fixation baseplate (i), a six degree of freedom load cell (ii), glenosphere lateralization spacer (iii), 424
custom 38mm glenosphere with hollow to accommodate ii & iii and a corresponding humeral cup 425
(Depuy) (iv), humeral head-neck angle component (v), and baseplate and fixation stem which 426
facilitates humeral lateralization (vi). 427
Figure 2 - Effects of rotator cuff repair status, HLat, & GLat on deltoid force (mean±SD). Note 428
that p-values indicate the effect of GLat with and without loading on the rotator cuff. 429
Figure 3 - Effects of rotator cuff repair status and GLat on joint load (mean±SD) across abduction. 430
Note that shaded areas represent ±1 Standard Error of the Mean and p-values are for effect of 431
rotator cuff repair status averaged across abduction 432
Figure 4 – Effects of rotator cuff repair status and HLat on joint loading angle (mean±SD). 433
Included p-values indicate the significant difference between the two levels of each factor averaged 434
across the levels of the other factor. Note that a 0° joint load corresponds to pure compression and 435
positive values are superiorly directed. 436
Figure 5 – Effect of HLat and GLat on internal and external rotation range of motion (mean±SD). 437
Note that * and † indicate the significant difference (p=0.008 & p=0.033) between the two levels 438
of the humeral and glenosphere lateralization factors averaged across the levels of the other factor 439
for external rotation range of motion. 440
Figure 6 – Illustration of the deltoid’s compressive effect on the glenohumeral joint for RTSA 441
configurations with (right) and without (left) humeral lateralization. Note that the white arrows 442
represent the deltoid muscle force direction acting at the level at which the tissue contacts the 443
greater tuberosity and the black arrows represent the force imparted on the humeral head. This 444
Rotator Cuff is Antagonist Following RTSA
25
imparted force is related to the component of the muscle forces perpendicular to the tangent of the 445
greater tuberosity’s surface at the point of contact. Also note the increase in deltoid moment arm 446
caused by humeral lateralization (i.e. left vs right pane). 447