Superior capsule reconstruction for reinforcement. before arthroscopic rotator cuff repair improves cuff integrity

Transcription

1 1 Superior capsule reconstruction for reinforcement 2 before arthroscopic rotator cuff repair improves cuff integrity 3 4 Teruhisa Mihata MD, PhD a,b,c, Thay Q Lee PhD b, Akihiko Hasegawa MD, PhD a, 5 Kunimoto Fukunishi MD a, Takeshi Kawakami MD, PhD a, Yukitaka Fujisawa MD, PhD a, 6 Yasuo Itami MD a, Mutsumi Ohue MD c, Munekazu Doi MD, PhD a, Masashi Neo MD, PhD a 7 8 a Department of Orthopedic Surgery, Osaka Medical College, Japan 9 b Orthopaedic Biomechanics Laboratory, VA Long Beach Healthcare System and UC Irvine, 10 California, USA 11 c Department of Orthopedic Surgery, Katsuragi Hospital, Japan 12 1

2 13 Introduction 14 Symptomatic rotator cuff tears compromises shoulder function even in daily life. 15 When non-operative treatment neither relieves shoulder pain nor recovers function, 16 arthroscopic repair is a good option for reparable rotator cuff tears. In most cases, 17 arthroscopic rotator cuff repair provides satisfactory clinical results 1-4. However, 18 postoperative re-tear of the repaired tendon makes significantly poorer outcome and less 19 satisfaction To obtain better clinical outcome, the factor that contribute postoperative re-tear of 21 the repaired rotator cuff tendons has been investigated 3,6-21. Patient age 13,16,17,20,21, gender 13, 22 bone mineral density 13, initial tear size 7,9,13,16,17,19-21, muscle atrophy 10,11 and fatty 23 infiltration 6,8,10,13-17,19,21 of supraspinatus or infraspinatus muscles, preoperative tendon 24 length 15 and tendon thickness 20 of the torn rotator cuff, tendon retraction 9,14,19,20, 25 acromiohumeral distance 13,14,19, duration of symptoms before surgery 9, and repair 26 method 3,12,16-18 were reported to be risk factor for re-tear after rotator cuff repair. Surgeons 27 can improve repair technique to make secure fixation of the rotator cuff tendons, however, 28 patient's related factors are hard to control for the treatment. Therefore, no previous report 29 has shown zero percent of failure rate after rotator cuff repair even though repair technique 30 has been developed to improve biomechanical characteristics of repairing method. 2

3 31 Most of risk factors for rotator cuff re-tear are related to tendon and muscle 32 degeneration due to age, overuse before initial tear, and disuse after initial tear 6,8-11,13-17, For the repair of severely degenerated rotator cuff tendons, high tension is needed to fix the 34 torn tendons to the greater tuberosity due to large defect of torn tendons, shortening of the 35 tendon tissue, and severe retraction. Also such degenerated tendons are weaker and thinner 36 tissue than native tendon, resulting in lower tensile strength. Therefore, when the severely 37 degenerated tendons are repaired, higher tension in the weaker tendon tissue makes high rate 38 of re-tear rotator cuff tendons. 39 Recently, superior capsule reconstruction (SCR) has been developed for the 40 treatment of irreparable rotator cuff tears Biomechanical and clinical studies have shown 41 that SCR restored superior shoulder stability and improved shoulder function More 42 recently, we developed a new surgical treatment, SCR for reinforcement of rotator cuff repair, 43 for severely degenerated but reparable rotator cuff tears. In this surgery, SCR may improve 44 superior shoulder stability, which decreases risk of subacromial abrasion after surgery, and 45 increased tendon strength by increasing tendon thickness. Consequently, SCR for 46 reinforcement may decrease re-tear rate of the repaired rotator cuff tendons. In this study, we 47 investigated whether SCR for reinforcement before arthroscopic rotator cuff repair (ARCR) 48 improves cuff integrity, especially in the case of severely degenerated supraspinatus tendon 3

4 49 tear. Our hypothesis was that SCR for reinforcement prevented postoperative re-tear after 50 ARCR Materials and Methods 54 We retrospectively reviewed our database of rotator cuff tears, which was collected 55 prospectively. The patients signed an informed consent form approved by the Institutional 56 Review Board at our university (Osaka Medical College, No. Clinical-254 (1854)). From to 2016, 452 consecutive shoulders with rotator cuff tears for which conservative 58 treatment had failed underwent arthroscopic surgery by a single surgeon (T.M.). Since 2013, 59 surgical indication for the treatment of rotator cuff tears has been determined by assessing the 60 severity of degeneration in the torn supraspinatus tendon using preoperative MRI. We 61 evaluated (1) muscle degeneration (MILD degeneration: fat area in the he supraspinatus fossa 62 is less than muscle area, SEVERE degeneration: fat area in the he supraspinatus fossa is more 63 than muscle area) (Figure 1A) (2) tendon degeneration (MILD degeneration: slightly thin, or 64 slight fatty degeneration in the tendon part, SEVERE degeneration: severely thin and fatty 65 degeneration in the tendon part or no tendon) (Figure 1B), and (3) tendon retraction (MILD 66 retraction: the torn tendon edge is located on the greater tuberosity or in the lateral half of the 67 humeral head, SEVERE retraction: the torn tendon edge is located in the medial half of the 4

5 68 humeral head or on the glenoid) (Figure 1C). In patients classified with " SEVERE " 69 degeneration or retraction in two or three categories of the supraspinatus degeneration, 70 arthroscopic SCR was performed (193 shoulders). When the patient had none or only one of " 71 SEVERE " degeneration or retraction, arthroscopic rotator cuff repair was chosen ( shoulders). Reducibility of the torn rotator cuff tendons was assessed during diagnostic 73 arthroscopy. In patients with " SEVERE " degeneration or retraction in none or one category 74 of the supraspinatus degeneration in the preoperative MRI, all torn tendons were reducible. 75 When the patients with " SEVERE " degeneration or retraction in two or three categories of 76 the supraspinatus degeneration in the preoperative MRI was judged as irreducible tear during 77 diagnostic arthroscopy, SCR was performed alone (156 shoulders). When the torn tendon 78 could reach to the original footprint in the patients with " SEVERE " degeneration or 79 retraction in two or three categories of the supraspinatus degeneration in the preoperative 80 MRI, arthroscopic SCR was performed for reinforcement, after which the torn tendon was 81 repaired over the fascia lata graft (ARCR with SCR: 37 shoulders). To assess the benefit of 82 SCR for reinforcement, the results of ARCR with SCR were investigated in this study. Three 83 patients in ARCR with SCR, who had full- range of motion at 6 months after surgery, 84 stopped our examination by the patients' related reason before 1 year after surgery. The 85 remaining 34 shoulders were enrolled in the study (Figure 2). Average age was 69.1 years 86 (40-80 years). Average tear size was 2.2cm (2-4 cm) and the number of torn tendons was two 5

6 87 (supraspinatus and infraspinatus) in 29 shoulders and three (supraspinatus, infraspinatus, 88 subscapularis) in 5 shoulders. The average time to final follow-up was 24 months (range, to 40 months). 90 For the control data, the results after ARCR alone in 91 consecutive patients with 91 medium (1-3 cm) to large (3-5 cm) rotator cuff tears (mean age, 63.6 years; range years) were used. All ARCR for the control study were performed from 2006 to 2009, when 93 we had not started SCR for reinforcement of ARCR. At that time, our surgical indication for 94 rotator cuff tear was determined based on arthroscopic findings. For reducible tear, in which 95 the torn tendon can reach to the original footprint, arthroscopic rotator cuff repair was 96 performed regardless of quality of the torn rotator cuff tendon and muscle. When the torn 97 tendon cannot reach to the original footprint (irreducible tear), we chose arthroscopic SCR 98 without rotator cuff repair or arthroscopic partial repair. Average tear size was 2.5cm ( cm) and the number of torn tendons was one (supraspinatus) in 32 shoulders, two 100 (supraspinatus and infraspinatus) in 50 shoulders, two (supraspinatus, subscapularis) in shoulders and three (supraspinatus, infraspinatus, subscapularis) in 1 shoulder. The average 102 time to final follow-up was 37 months (range, 18 to 66 months) Patient Assessment 6

7 105 Shoulder elevation, external rotation at side, and internal rotation were measured 106 actively by a single surgeon, before surgery; at 3, 6, and 12 months after surgery; and at the 107 final follow-up. Internal rotation was measured as the highest vertebral body that the patient 108 was able to reach with the thumb of the affected arm. All patients were assessed 109 preoperatively by using the scoring systems of the shoulder index of the American Shoulder 110 and Elbow Surgeons (ASES, a 100-point scoring system) and the Japanese Orthopaedic 111 Association (JOA, a 100-point scoring system) and were reassessed at the time of the final 112 follow-up. Postoperative complication rate has been also recorded. Magnetic resonance 113 imaging (MRI) was performed with a 1.5-T closed-type scanner (MRT-2000/V2, Toshiba, 114 Tokyo, Japan) before surgery and at the final follow-up after surgery. Oblique coronal, 115 oblique sagittal, and axial T2-weighted MR images were acquired for structural and 116 qualitative assessment of the rotator cuff and repair integrity after surgery. Preoperative 117 rotator cuff muscle quality was evaluated using Goutallier grading 8. Repair integrity was 118 classified into 5 categories by Sugaya s classification 4. Type I indicated a repaired cuff that 119 had sufficient thickness, with a homogeneously low intensity in each image; type II had 120 sufficient thickness associated with a partial high-intensity area; type III had insufficient 121 thickness without discontinuity; type IV revealed the presence of a minor discontinuity in 122 more than one slice of each image; and type V revealed the presence of a major discontinuity 123 in each image. Type IV and type V ratings in Sugaya s classification indicated postoperative 7

8 124 retear. Cuff repair integrity in all postoperative MRIs was evaluated by three experienced 125 shoulder surgeons to eliminate observer bias. Sugaya type, which was selected by two or 126 three shoulder surgeons, was determined as the current result of cuff repair integrity in each 127 MRI. There was no case in which each of the three shoulder surgeons selected different 128 Sugaya type Surgical Technique of SCR for reinforcement of ARCR 131 Preparation 132 All procedures were performed under general anesthesia in the lateral decubitus 133 position. Three portals were typically required for arthroscopic SCR. A posterior portal was 134 established for initial assessment of the glenohumeral joint. An anterior portal through the 135 rotator interval was established as the working portal for the treatment of intra-articular 136 lesions, such as labral tear and biceps tear, or subluxation. The arthroscope was then removed 137 from the glenohumeral joint and redirected into the subacromial space. A lateral portal was 138 also established. Any pathological bursal tissue that impeded clearance of the space was 139 removed. Arthroscopic subacromial decompression was performed to create a flat acromial 140 undersurface. Bony spurs in the inferior part of the acromioclavicular joint and at the distal 141 end of the clavicle were removed. The superior glenoid and rotator cuff footprint of the 142 greater tuberosity were debrided to expose cortical bone. If the subscapularis tendon tear was 8

9 143 reparable it was completely repaired with fully threaded titanium suture anchors (diameter, mm; Corkscrew FT Suture Anchor, Arthrex, Naples, FL) Choosing the graft size and harvesting the fascia lata 147 Appropriate graft size is the most important point in this surgery. The size of the 148 "superior capsular defect" was evaluated by using a measuring probe in both the 149 anteroposterior (from the anterior edge to the posterior edge of the torn tendon) and 150 mediolateral (from the superior edge of the glenoid to the lateral edge of the greater 151 tuberosity) directions at 30 of shoulder abduction. The optimal graft length in the 152 anteroposterior direction is exactly the same as the length of the defect without partial repair 153 of the torn infraspinatus tendon. The graft length in the mediolateral direction is 15 mm 154 longer than the distance from the superior edge of the glenoid to the lateral edge of the 155 greater tuberosity to give a 15-mm footprint on the superior glenoid. 156 A vertical skin incision is made over the lateral thigh, beginning the greater 157 trochanter of the femur. In the SCR for reinforcement of ARCR, one single layer of 158 autologous fascia lata is used (graft thickness is 1 to 3 mm). All fatty tissue should be 159 removed from the graft Graft attachment 9

10 162 All soft tissues were removed on the superior glenoid and greater tuberosity to 163 expose cortical bone. One or two fully threaded titanium suture anchors (diameter, 4.5 mm; 164 Corkscrew FT Suture Anchor, Arthrex, Naples, FL), each with two No. 2 Fiberwire 165 non-absorbable sutures (Arthrex), were inserted into the superior glenoid. All Fiberwires 166 from the superior glenoid were placed through the fascia lata in a mattress fashion outside the 167 body. Then the graft was inserted into the subacromial space via the lateral portal (Figure 3A). 168 We usually use a 5-cc syringe as a cannula. When the medial edge of the graft reached to the 169 superior glenoid, all Fiberwires were tied (Figure 3B). 170 To attach the lateral side of the fascia lata to the rotator cuff footprint on the greater 171 tuberosity, one or two fully threaded titanium suture anchors (diameter, 4.5 mm; Corkscrew 172 FT Suture Anchor, Arthrex, Naples, FL) were inserted at the medial edge of the footprint. All 173 Fiberwires from the medial row on the greater tuberosity were placed through the fascia lata 174 graft in a mattress fashion, and tied using a non-sliding knot (RC knot 3 ) (Figure 3B). After 175 the knots were tied for the graft attachment, all sutures were not cut because the suture limbs 176 would be used to repair the torn tendon over the graft (Figure 3B). The torn tendons were 177 repaired on the fascia lata graft by using the compression double-row technique, which is a 178 combination of the conventional double-row technique and the suture bridge 3,29, or the 179 transosseous equivalent rotator cuff repair (Figure 3C). The Fiberwires of medial anchors, 180 which were used for graft fixation to the greater tuberosity, were placed approximately 10 to 10

11 mm medial to the lateral edge of the torn rotator cuff tendon in a mattress fashion using a 182 suture shuttle or suture-passing device and tied using a non-sliding knot (RC knot 3 ). For the 183 compression double-row repair, two fully threaded titanium suture anchors (diameter, mm; Corkscrew FT Suture Anchor, Arthrex) with two No.2 Fiberwires were placed at 5 to mm inferior to the highest tip of the greater tuberosity. The Fiberwires of lateral anchors were 186 placed approximately 10 mm medial to the lateral edge of the rotator cuff tear using either a 187 suture shuttle or suture-passing device and tied with simple stitch before knot tying in the 188 medial row. Then, the conventional double-row repair was completed with knot tying for the 189 medial row. The Fiberwires in the medial row were not cut because the suture limbs would be 190 used to create suture bridges. One limb from the medial row anchor was fixed at the lateral 191 row anchor by tying the medial limb with a Fiberwire from the lateral anchor. Next, another 192 suture limb from the medial row anchor was tied to the first medial suture limb, which had 193 been fixed at the lateral row anchor, with a nonsliding knot, thereby completing suture 194 bridges using two medial suture limbs. Suture bridges were generated from the remaining 195 medial and lateral suture limbs in the same way used to create the first suture bridges. For the 196 transosseous equivalent rotator cuff repair, the limbs from medial row anchors were fixed by 197 using two SwiveLocks (Arthrex), which were inserted at 5 to 10 mm inferior to the highest 198 point of the greater tuberosity

12 200 Postoperative Protocol 201 Overall postoperative protocols for ARCR and SCR for reinforcement of ARCR 202 were same. An abduction sling (Block Shoulder Abduction Sling, Nagano Prosthetics & 203 Orthotics Co. Ltd. Osaka, Japan) were used for 4 weeks after surgery. After the 204 immobilization period, passive and active-assisted exercises were initiated to promote 205 scaption. Eight weeks after the surgery, patients began to perform exercises to strengthen the 206 rotator cuff and the scapula stabilizers. Full activities were allowed at 6month if patients had 207 sufficient range of motion and muscle strength. Physical therapists assisted all patients Statistical Analysis 210 By using t- and chi-square tests, we compared the American Shoulder and Elbow 211 Surgeons (ASES) score, active shoulder range of motion (ROM), and cuff tear integrity 212 (Sugaya MRI classification) between ARCR with and without SCR as well as between before 213 surgery and at final follow-up. A significant difference was defined as P <

13 214 Results 215 Cuff Integrity 216 At three months after ARCR with SCR, MRI showed type I cuff integrity in shoulders (79%) (Figure 4) and type II cuff integrity in 7 shoulders (21%). Three of seven 218 shoulders in type II at 3 months after ARCR with SCR became type I at 6 months, and other of 7 shoulders in type II at 3 months changed to type I at 1 year. At 1 year after ARCR with 220 SCR, 33 of 34 shoulders (97%) demonstrated type I cuff integrity, which represents the torn 221 tendon could heal with thick and good quality. Neither type III, which represents thin tendon 222 part or partial thickness of re-tear after ARCR, nor type IV/V, which is re-tear after ARCR, 223 were not found after ARCR with SCR. 224 For the ARCR alone, 4 of 91 shoulders (4%) revealed a re-tear (type V) at 1 year 225 after surgery. Three type V shoulders had large tear with goutallier 3 or 4, and the remaining 226 one type V shoulder had a very thin and degenerated tissue in the torn tendon before surgery. 227 One patient with type IV cuff integrity at 6 months after surgery became type III at 1 year, on 228 the other hand, the other type IV at 6 months became type V at 1 year. Seven shoulders (8%) 229 had type III cuff integrity, and 12 shoulders (13%) were classified as type II at 1 year after 230 ARCR. The remaining 68 shoulders (75%, 68/91) had type I cuff integrity after ARCR at year after ARCR

14 233 Functional Outcome 234 ASES and JOA scores increased significantly after ARCR both with (ASES: preoperatively to 92 postoperatively, JOA: 60 to 94) and without SCR (ASES: 41 to 91, JOA: to 96) (P < ). There was no significant difference in ASES and JOA scores between 237 ARCR with SCR and ARCR alone both before and after surgery. Postoperative ASES and 238 JOA scores in the groups with a healed repair (ARCR with SCR: 92 in ASES and 94 in JOA, 239 ARCR alone: 92 in ASES and 97 in JOA) were significantly better than that in the group with 240 a re-tear after ARCR (60 in ASES and 80 in JOA) (P < ). The shoulders with re-tear 241 after ARCR had significantly less postoperative ASES and JOA scores than the healed cases 242 after ARCR (P < in both scores) and ARCR with SCR (P < in ASES and P = in JOA score) Shoulder Range of Motion 246 The shoulder range of motion improved significantly both after ARCR with SCR: by for active elevation (P < ), by 10 for active external rotation (P = ), and by vertebral bodies of active internal rotation (P < ), and after ARCR alone : by 36 for 249 active elevation (P < ), by 8 for active external rotation (P = 0.006), and by vertebral bodies of active internal rotation (P < ). Postoperative active ROM did not 251 differ significantly between ARCR with and without SCR. Postoperative active elevation in 14

15 252 the shoulders with a healed repair (ARCR with SCR: 163, P = 0.003; ARCR alone: 171, P 253 < ) was significantly better than that in the shoulder with a re-tear after ARCR (120 ) Severity of the Rotator Cuff Tear 256 The Goutallier grade of the supraspinatus was significantly higher for ARCR with 257 SCR (mean, 2.8) than for ARCR alone (mean, 2.1) (P < ). There was no significant 258 difference in the number of torn tendons (P =.45), tear size (P =.10), preoperative (P =.06) 259 and postoperative acromiohumeral distance (P =.23), and the Goutallier grade of the 260 infraspinatus (P =.16), teres minor (P =.07) and subscapularis (P =.24) between ARCR with 261 SCR and ARCR alone. Acromiohumeral distance significantly increased after surgery in both 262 groups (P <.0001) (Figure 5) Concomitant Pathologies and Surgeries 265 There was no significant difference in the number of subscapularis repair (P =.45), and 266 biceps pathology and treatment (P =.16) between ARCR with SCR and ARCR alone. 267 Acromioplasty was performed in all patients Complications 15

16 270 There were no surgical complications, including nerve injury, infection, or suture anchor 271 problems both after ARCR with SCR and ARCR alone. In the group of ARCR with SCR, no 272 patient had any symptom with the harvest site at the final follow up Discussion 276 To obtain better clinical outcome after ARCR, postoperative re-tear of the repaired 277 rotator cuff tendons need to be prevented 1-5. Severe degeneration and retraction of the torn 278 rotator cuff tendon is reported to correlate with re-tear rate after rotator cuff 279 repair 6,8-11,13-17, Gladstone et al reported that preoperative muscle atrophy (P =.001) and 280 fatty infiltration (P =.008) of the supraspinatus were significantly correlated to postoperative 281 cuff integrity 7. Kim et al showed severity of the torn tendon retraction is correlated with 282 re-tear rate after arthroscopic rotator cuff repair (P < 001) 9. For severely degenerated and 283 retracted rotator cuff tears, we developed the superior capsular reconstruction for 284 reinforcement of repaired tendon. Current study showed that 97% shoulders demonstrated 285 type I cuff integrity, which represents the torn tendon could heal with thick and good quality, 286 and no partial or complete re-tear of the repaired tendon although all shoulders had severe 287 degeneration and/or retraction. These rates were significantly better than arthroscopic rotator 16

17 288 cuff repair alone. Therefore, we believe that the superior capsule reconstruction is a great 289 option for rotator cuff repair to improve tendon quality and prevent postoperative re-tear. 290 For the severe degenerated rotator cuff tears, patch augmentation may be one of the 291 surgical options to prevent postoperative rotator cuff re-tear. Several studies investigated the 292 postoperative cuff integrity after rotator cuff repair with patch augmentation. But the re-tear 293 rate is relatively high (10 to 62%) 30-35, and some reports showed no improvement of 294 postoperative cuff integrity even after patch augmentation compared with rotator cuff repair 295 alone 33,35. Furtherore, Flury et al concluded that SSP tear repair with patch augmentation does 296 not benefit patients in terms of reducing the risk of a recurrent tendon defect or improving 297 shoulder function up to 24 months after surgical repair 35. The reason why the re-tear rate did 298 not decrease in some studies may be location of the graft attachment. In all studies the patch 299 graft was attached on the rotator cuff tendons, while we attached the patch graft under the 300 repaired rotator cuff tendons to reconstruct the superior capsule. Previous biomechanical 301 studies showed that the superior capsule reconstruction improved shoulder superior stability 302 to intact level 27. Also current study showed no re-tear patient after arthroscopic rotator cuff 303 repair with superior capsule reconstruction for reinforcement, meaning that we have not seen 304 re-tear patient since Therefore, we recommended attaching the patch graft (superior 305 capsule reconstruction) to prevent re-year after rotator cuff repair, especially for severely 306 degenerated reducible tear. 17

18 307 Superior capsule reconstruction was developed to improve shoulder function and 308 pain relief for irreparable rotator cuff tear. For the irreparable cases, thick graft (6 to 8mm) is 309 recommended to prevent graft tear after surgery 25. In the superior capsule reconstruction for 310 reinforcement of rotator cuff repair, the graft is just one layer of fascia lata and the thickness 311 is just 2 to 3mm because the torn tendon is repaired on the fascia lata graft. To make a thick 312 graft, autograft of fascia lata is a good material because the large size of fascia lata can be 313 harvested and the fascia lata can be united between layers. On the other hand, for the superior 314 capsule reconstruction reinforcement of rotator cuff repair, the other materials, such as human 315 dermal graft or porcine dermal graft, may be good candidates because 2 to 3 mm thickness is 316 enough. 317 Re-tear of the repaired tendon made worse ASES and JOA scores and decreased 318 active elevation compared with healed ARCR. This result is compatible with previous 319 studies 1-5 and it is necessary to carefully consider preventing re-tear after ARCR. Current 320 study showed that additional superior capsule reconstruction completely prevent re-tear after 321 ARCR, and postoperative functional outcome including ASES and JOA scores and active 322 elevation in ARCR with SCR reinforcement was significantly better than re-tear cases after 323 ARCR. Therefore, the superior capsule reconstruction is recommended to add to ARCR, 324 specifically in the severely degenerated rotator cuff tears. 18

19 325 In the SCR, the graft is attached medially to the superior glenoid, and laterally to the 326 greater tuberosity to improve superior stability in the glenohumeral joint, however, this 327 surgical technique may decrease shoulder range of motion when the graft tension is too tight. 328 Previous biomechanical study showed SCR decreased total rotational range of motion (the 329 sum of external rotation and internal rotation) by 5 to 20 compared with intact condition Current study showed a significantly decreased range of motion only in internal rotation after 331 ARCR with SCR compared with ARCR alone. But the amount of decreased internal rotation 332 was only 2 vertebral bodies, and all patients had no restriction in any activity. Furthermore, 333 there was no significant difference in active elevation and external rotation between ARCR 334 with SCR and ARCR alone. Therefore, SCR for reinforcement does not cause shoulder 335 stiffness after ARCR, although SCR may cause slightly decreased internal rotation, which is 336 no influence of any activity. 337 The current study had several limitations. First, Sugaya classification is subjective 338 evaluation of postoperative cuff integrity although this classification is widely used in 339 previous studies. Therefore, we judged the final result of Sugaya classification by using three 340 experienced shoulder surgeons' evaluation. Therefore, we believe that the final result of 341 Sugaya classification is reliable in this study. Second, the data of two groups in this study 342 were corrected in the different period to eliminate the patient selection bias. On the other 343 hand, the different period may affect clinical results. However, all surgeries were performed 19

20 344 by one experienced shoulder surgeon, who had had many experiences of shoulder surgery at 345 the first surgery in this study. And postoperative protocol did not change during this study 346 and postoperative physical therapy was performed by the same physical therapy team. 347 Therefore, we believe the effect of different time period of data collection is not significant in 348 this study Conclusions 351 SCR for reinforcement prevented postoperative re-tear after ARCR and improved 352 the quality of the repaired tendon on MRI. Furthermore, postoperative functional outcomes 353 were similar in patients who underwent ARCR alone and those who also underwent SCR, 354 even though degeneration of the torn tendons was greater in the latter group

21 356 Figure Legends 357 Figure 1 Severity of degeneration in the torn supraspinatus tendon was evaluated using 358 preoperative MRI. (A) Muscle degeneration (MILD degeneration: fat area in 359 the he supraspinatus fossa is less than muscle area, SEVERE degeneration: fat 360 area in the he supraspinatus fossa is more than muscle area). (B) Tendon 361 degeneration (MILD degeneration: slightly thin, or slight fatty degeneration in 362 the tendon part, SEVERE degeneration: severely thin and fatty degeneration in 363 the tendon part or no tendon). (C) Tendon retraction (MILD retraction: the torn 364 tendon edge is located on the greater tuberosity or in the lateral half of the 365 humeral head, SEVERE retraction: the torn tendon edge is located in the 366 medial half of the humeral head or on the glenoid) Figure 2 Surgical indication Figure 3 ARCR with SCR. (A) All Fiberwires from the superior glenoid were placed 371 through the fascia lata in a mattress fashion outside the body. Then the graft 372 was inserted into the subacromial space via the lateral portal. (B) When the 373 medial edge of the graft reached to the superior glenoid, all Fiberwires were 374 tied. All Fiberwires from the medial row on the greater tuberosity were placed 21

22 375 through the fascia lata graft in a mattress fashion, and tied using a non-sliding 376 knot. After the knots were tied for the graft attachment on the greater 377 tuberosity, all sutures were not cut because the suture limbs would be used to 378 repair the torn tendon over the graft. (C) The torn tendons were repaired on the 379 fascia lata graft by using the compression double-row technique, or the 380 transosseous equivalent rotator cuff repair Figure 4 Magnetic resonance imaging. (A) Coronal image before surgery. (B) Sagittal 383 image before surgery. (C) Coronal image at 3months after ARCR with SCR. 384 (D) Sagittal image at 3months after ARCR with SCR Figure 5 Plain X-ray. (A) Before surgery. (B) At 1year after ARCR with SCR Table 1 Change in cuff integrity until 1 year after surgery Table 2 Comparison of shoulder ranges of motion and functional outcomes between 392 ARCR with and without SCR

23 394 Table 3 Comparison of shoulder ranges of motion and functional outcomes between 395 healed and re-tear cases Table 4 Severity of rotator cuff tear Table 5 Concomitant pathologies and surgeries

24 401 References Gazielly DF, Gleyze P, Montagnon C. Functional and anatomical results after rotator 403 cuff repair. Clin Orthop Relat Res. 1994: Huijsmans PE, Pritchard MP, Berghs BM, van Rooyen KS, Wallace AL, de Beer JF. 405 Arthroscopic rotator cuff repair with double-row fixation. J Bone Joint Surg Am ;89: Mihata T, Watanabe C, Fukunishi K, Ohue M, Tsujimura T, Fujiwara K, Kinoshita M. 408 Functional and structural outcomes of single-row versus double-row versus combined 409 double-row and suture-bridge repair for rotator cuff tears. Am J Sports Med ;39: Sugaya H, Maeda K, Matsuki K, Moriishi J. Repair integrity and functional outcome 412 after arthroscopic double-row rotator cuff repair. A prospective outcome study. J 413 Bone Joint Surg Am. 2007;89: Kim HM, Caldwell JM, Buza JA, Fink LA, Ahmad CS, Bigliani LU, Levine WN. 415 Factors affecting satisfaction and shoulder function in patients with a recurrent rotator 416 cuff tear. J Bone Joint Surg Am. 2014;96: Barry JJ, Lansdown DA, Cheung S, Feeley BT, Ma CB. The relationship between tear 418 severity, fatty infiltration, and muscle atrophy in the supraspinatus. J Shoulder Elbow 419 Surg. 2013;22:

25 Gladstone JN, Bishop JY, Lo IK, Flatow EL. Fatty infiltration and atrophy of the 421 rotator cuff do not improve after rotator cuff repair and correlate with poor functional 422 outcome. Am J Sports Med. 2007;35: Goutallier D, Postel JM, Bernageau J, Lavau L, Voisin MC. Fatty muscle 424 degeneration in cuff ruptures. Pre- and postoperative evaluation by CT scan. Clin 425 Orthop Relat Res. 1994: Kim IB, Kim MW. Risk Factors for Retear After Arthroscopic Repair of 427 Full-Thickness Rotator Cuff Tears Using the Suture Bridge Technique: Classification 428 System. Arthroscopy. 2016;32: Liem D, Lichtenberg S, Magosch P, Habermeyer P. Magnetic resonance imaging of 430 arthroscopic supraspinatus tendon repair. J Bone Joint Surg Am. 2007;89: Jo CH, Shin JS. Cross-sectional area of the supraspinatus muscle after rotator cuff 432 repair: an anatomic measure of outcome. J Bone Joint Surg Am. 2013;95: Cho NS, Yi JW, Lee BG, Rhee YG. Retear patterns after arthroscopic rotator cuff 434 repair: single-row versus suture bridge technique. Am J Sports Med. 2010;38: Chung SW, Oh JH, Gong HS, Kim JY, Kim SH. Factors affecting rotator cuff healing 436 after arthroscopic repair: osteoporosis as one of the independent risk factors. Am J 437 Sports Med. 2011;39:

26 Kim JR, Cho YS, Ryu KJ, Kim JH. Clinical and radiographic outcomes after 439 arthroscopic repair of massive rotator cuff tears using a suture bridge technique: 440 assessment of repair integrity on magnetic resonance imaging. Am J Sports Med ;40: Meyer DC, Wieser K, Farshad M, Gerber C. Retraction of supraspinatus muscle and 443 tendon as predictors of success of rotator cuff repair. Am J Sports Med ;40: Park JS, Park HJ, Kim SH, Oh JH. Prognostic Factors Affecting Rotator Cuff Healing 446 After Arthroscopic Repair in Small to Medium-sized Tears. Am J Sports Med ;43: Lee YS, Jeong JY, Park CD, Kang SG, Yoo JC. Evaluation of the Risk Factors for a 449 Rotator Cuff Retear After Repair Surgery. Am J Sports Med. 2017;45: Hantes ME, Ono Y, Raoulis VA, Doxariotis N, Venouziou A, Zibis A, Vlychou M. 451 Arthroscopic Single-Row Versus Double-Row Suture Bridge Technique for Rotator 452 Cuff Tears in Patients Younger Than 55 Years: A Prospective Comparative Study. 453 Am J Sports Med. 2018;46: Shin YK, Ryu KN, Park JS, Jin W, Park SY, Yoon YC. Predictive Factors of Retear 455 in Patients With Repaired Rotator Cuff Tear on Shoulder MRI. AJR Am J Roentgenol ;210:

27 Le BT, Wu XL, Lam PH, Murrell GA. Factors predicting rotator cuff retears: an 458 analysis of 1000 consecutive rotator cuff repairs. Am J Sports Med. 2014;42: Deniz G, Kose O, Tugay A, Guler F, Turan A. Fatty degeneration and atrophy of the 460 rotator cuff muscles after arthroscopic repair: does it improve, halt or deteriorate? 461 Arch Orthop Trauma Surg. 2014;134: Mihata T, Bui CNH, Akeda M, Cavagnaro MA, Kuenzler M, Peterson AB, McGarry 463 MH, Itami Y, Limpisvasti O, Neo M, Lee TQ. A biomechanical cadaveric study 464 comparing superior capsule reconstruction using fascia lata allograft with human 465 dermal allograft for irreparable rotator cuff tear. J Shoulder Elbow Surg ;26: Mihata T, Lee TQ, Watanabe C, Fukunishi K, Ohue M, Tsujimura T, Kinoshita M. 468 Clinical results of arthroscopic superior capsule reconstruction for irreparable rotator 469 cuff tears. Arthroscopy. 2013;29: Mihata T, McGarry MH, Kahn T, Goldberg I, Neo M, Lee TQ. Biomechanical Role 471 of Capsular Continuity in Superior Capsule Reconstruction for Irreparable Tears of 472 the Supraspinatus Tendon. Am J Sports Med. 2016;44: Mihata T, McGarry MH, Kahn T, Goldberg I, Neo M, Lee TQ. Biomechanical Effect 474 of Thickness and Tension of Fascia Lata Graft on Glenohumeral Stability for Superior 27

28 475 Capsule Reconstruction in Irreparable Supraspinatus Tears. Arthroscopy ;32: Mihata T, McGarry MH, Kahn T, Goldberg I, Neo M, Lee TQ. Biomechanical Effects 478 of Acromioplasty on Superior Capsule Reconstruction for Irreparable Supraspinatus 479 Tendon Tears. Am J Sports Med. 2016;44: Mihata T, McGarry MH, Pirolo JM, Kinoshita M, Lee TQ. Superior capsule 481 reconstruction to restore superior stability in irreparable rotator cuff tears: a 482 biomechanical cadaveric study. Am J Sports Med. 2012;40: Hirahara AM, Adams CR. Arthroscopic Superior Capsular Reconstruction for 484 Treatment of Massive Irreparable Rotator Cuff Tears. Arthrosc Tech. 2015;4:e Mihata T, Fukuhara T, Jun BJ, Watanabe C, Kinoshita M. Effect of shoulder 486 abduction angle on biomechanical properties of the repaired rotator cuff tendons with types of double-row technique. Am J Sports Med. 2011;39: Gilot GJ, Alvarez-Pinzon AM, Barcksdale L, Westerdahl D, Krill M, Peck E. 489 Outcome of Large to Massive Rotator Cuff Tears Repaired With and Without 490 Extracellular Matrix Augmentation: A Prospective Comparative Study. Arthroscopy ;31:

29 Yoon JP, Chung SW, Kim JY, Lee BJ, Kim HS, Kim JE, Cho JH. Outcomes of 493 Combined Bone Marrow Stimulation and Patch Augmentation for Massive Rotator 494 Cuff Tears. Am J Sports Med. 2016;44: Lederman ES, Toth AP, Nicholson GP, Nowinski RJ, Bal GK, Williams GR, Iannotti 496 JP. A prospective, multicenter study to evaluate clinical and radiographic outcomes in 497 primary rotator cuff repair reinforced with a xenograft dermal matrix. J Shoulder 498 Elbow Surg. 2016;25: Sears BW, Choo A, Yu A, Greis A, Lazarus M. Clinical outcomes in patients 500 undergoing revision rotator cuff repair with extracellular matrix augmentation. 501 Orthopedics. 2015;38:e Lenart BA, Martens KA, Kearns KA, Gillespie RJ, Zoga AC, Williams GR. 503 Treatment of massive and recurrent rotator cuff tears augmented with a poly-l-lactide 504 graft, a preliminary study. J Shoulder Elbow Surg. 2015;24: Flury M, Rickenbacher D, Jung C, Schneider MM, Endell D, Audige L. Porcine 506 Dermis Patch Augmentation of Supraspinatus Tendon Repairs: A Pilot Study 507 Assessing Tendon Integrity and Shoulder Function 2 Years After Arthroscopic Repair 508 in Patients Aged 60 Years or Older. Arthroscopy. 2018;34: