a Departamento de Ortopedia, Santa Casa de São Paulo, São Paulo, SP, Brasil
b Serviço de Ortopedia e Traumatologia, Hospital Israelita Albert Einstein, Universidade Federal de São Paulo (Unifesp), São Paulo, SP,Brasil
c Instituto de Ortopedia, Hospital das Clínicas, Faculdade de Medicina, Universidade de São Paulo (USP), São Paulo, SP, Brasil
d Escola Paulista de Medicina, Universidade Federal de São Paulo (Unifesp), São Paulo, SP, Brasil
Introduction
Rotator cuff injuries are very common in orthopedic practice,with prevalence between 5% and 33%. They frequently occurin the elderly population1,2,3,4and reach 22% among patientsover the age of 65 years.5Individuals over the age of 50 yearsare more commonly affected by chronic injuries of greaterseverity as a result of the degenerative process, while theyoung population (< 40 years) has injuries that are predomi-nantly of traumatic etiology.
While some authors have demonstrated good resultsregarding improvement of pain, function and quality of lifethrough conservative methods7,8or through surgical treat-ment, including by means of acromioplasty and debridementalone,9,10others have argued that surgical repair of rotatorcuff injuries leads to better and longer-lasting results.2,3,4,11Performing surgical procedures on elderly people may bea challenge. In 1995, Hattrup and Scottsdale2suggestedthat patients aged 65 years or over present a significantchance of having larger lesions, which potentially addsdifficulty to their repair. The bone quality is lower andthis results in osteoporosis of the greater tubercle, sub-chondral cystic degeneration and irregularity of the corticalbone, and it may complicate anchor fixation. Furthermore,elderly people frequently present comorbidities (diabetes mel-litus, rheumatoid arthritis or renal diseases), which maydiminish the healing response and compromise surgicalmanagement.
The long period between the initial presentation of symp-toms and the time when arthroscopic surgical repair of therotator cuff injury is performed is a negative prognostic fac-tor for the clinical results.13In addition, the size of the lesion,retraction, fatty degeneration and the quality of the tendonalso influence the healing.
Some authors have believed that these patients' advancedage interferes with the healing of rotator cuff lesion suturing,such that young patients tend to have more favorable evolu-tion. Nonetheless, arthroscopic repair on rotator cuff injuriesin elderly people produces significant functional improvementand positive changes to quality of life,16-20with advantages inrelation to open surgery and mini-open procedures, becauseof the small incision, absence of damage to the deltoid mus-cle, lower pain during the postoperative period and shorterhospital stay. Moreover, especially, it enables diagnosis andtreatment of associated lesions through viewing the joint withminimal tissue trauma.21-23Thus, Charousset et al.15andGrondel et al.17suggested that age alone is unrelated to poorresults from injury repairs.
The present study had the objective of evaluating theresults from arthroscopic surgical treatment of rotator cufflesions in patients aged 65 years or over.
Sample and methods
A retrospective analysis was conducted on 168 shoulders in163 patients aged 65 years or over who presented complete tearing of the rotator cuff and underwent arthroscopic surgicaltreatment performed by the shoulder and elbow group of ourservice between September 1998 and August 2009.
The inclusion criteria were that the patients should be 65years of age or over, with complete tearing of the rotator cuffthat was arthroscopically sutured in an operation that tookplace at least 1 year earlier. The following were consideredto be exclusion criteria: age of less than 65 years with eithercomplete or incomplete tearing; open procedures; open proce-dures; and postoperative length of follow-up less than 1 year.
Out of the 168 shoulders evaluated, three were excludeddue to loss from the follow-up and two because of death, forreasons unrelated to the treatment. The remaining 163 werestratified according to age, with division into three groups:65-69 (49.1%); 70-74 (26.4%) and 75 or over (24.5%).
There were 63 male patients (38.7%) and 100 femalepatients (61.3%). Their mean age was 70 years and 10 months(range: 65-83). The length of time with pain, from the startof the symptoms until the surgical treatment ranged from 2days to 240 months, with mean of 22 months and 25 days.The dominant side was affected in 108 cases (66.3%). Only 62patients (38%) reported having histories of trauma, while 26(16%) said that the start of their pain and its worsening wererelated to exertion of some type. Among the patients evalu-ated, 45 (27.6%) said that they practiced sports in which theyused their upper limbs.
All the patients underwent the surgical procedure in the"deckchair" position, under general anesthesia in associa-tion with anesthetic block of the brachial plexus. We madean initial arthroscopic inspection of the joint and identifiedthe associated lesions. Eighty-three patients (50.9%) presentedlesions in the long head of the biceps brachii muscle. Amongthese, the tendon was absent in 17 cases (10.4%) and it wasfound to be adhering to the bicipital groove in two (1.2%). Inthe other cases in which we observed dislocation or injuryaffecting more than 50% of the diameter of the tendon of thelong head of the biceps, we performed tenotomy (20 cases;12.3%). We performed tenotomy in association with tenode-sis in 44 cases (27%), before repairing the rotator cuff. We alsofound SLAP lesions (superior labrum, anterior and posterior)in four cases, Bankart lesions in three, calcareous tendinitisof the tendon of the supraspinatus muscle in one, adhesivecapsulitis in two and shoulder arthrosis in 12 (Table 1). Whenthe surgeon deemed it necessary, these lesions were treatedduring the same surgical procedure (Table 2).
Following this, we dealt with the subacromial space andperformed bursal debridement and mobilization of the ten-dons. Acromioplasty was performed in 145 cases (89%) andresection of the distal portion of the clavicle (Mumford), whichwas indicated in the cases of pain in the acromioclavicularjoint, was performed in 54 cases (33.1%). We then inducedbleeding in the bone bed of the greater tubercle of the humerusand sutured the lesion. The number of tendon-to-tendon ortendon-to-bone stitches ranged from 1 to 12 (mean of 5). Sutur-ing anchors were used in 158 cases, with a range from 1 to 5(mean of 2).
The rotator cuff lesions were classified as small, found in 26cases (16%); medium, 43 cases (26.4%); large, 25 cases (15.3%);or extensive, 69 cases (42.3%), in accordance with the classifi-cation of Hawkins.24Those that involved at least two tendons
were also considered to be extensive, as defined by Gerberet al.25
From analysis on the tendons affected, 100% of the casespresented involvement of the tendon of the supraspinatusmuscle; 43.6% the tendon of the infraspinatus muscle; and33.1% the tendon of the subscapularis muscle.
After the operation, the patients were immobilized for 6-8weeks by means of a functional sling. According to the sizeof the lesion and the retraction of the tendons, passive exer-cises such as pendulum exercises were allowed. If the lesionwas extensive, the shoulder was kept immobilized for at least4 weeks. Passive elevation was started in the fourth week andactive elevation in the sixth week. Muscle strengthening exer-cises were only allowed starting 4 months after the operation.
The mean length of postoperative follow-up was 50 monthsand 18 days, with a range from 12 to 144 months. During thisperiod, the patients were evaluated by means of the methodof the University of California in Los Angeles (UCLA).
In the statistical analysis, the SPSS software (StatisticalPackage for the Social Sciences), version 17.0, was used toobtain the results, and 95% intervals were considered to bestatistically significant (p < 0.05).
Results
The mean UCLA score among the 163 patients evaluated was33.6 points (range: 11-35). 80.4% of the cases were consideredto have attained excellent results, 16% were good and 3.6%were unsatisfactory: three fair and three poor.
The evaluation of the clinical results did not show anystatistically significant correlation with age progression, suchthat we found that 97.6% of the results in the age groupbetween 65 and 69 years were good or excellent; 95.4%between the ages of 70 and 74 years; and 95% aged 75 years orover (p = 0.49).
The size of the lesion did not show any statistically signifi-cant correlation with the postoperative clinical result (p = 0.86),or with age (p = 0.67). However, when associated with trauma,it was statistically significant. The larger lesions were gener-ally associated with the presence of trauma (p < 0.001).
The tendon of the infraspinatus was affected in 43.6% ofthe cases and there was a slight increase in incidence withage progression (p = 0.31).
The time that elapsed between the start of symptoms andthe surgical procedure showed a statistically significant rela-tionship (p < 0.027) with the postoperative results, given thatthe greater the time that elapsed between the symptoms andthe surgery was, the worse the final results were.
The incidence of complications was 6.1% (10 cases): onecase of adhesive capsulitis, one of tendinitis of the biceps, twoof limitation of range of motion regarding medial rotation,two of painful renewed tearing, two of non-painful renewedtearing with functional impotence, one of arthropathy of therotator cuff and one of pain in the acromioclavicular joint.
Discussion
In the light of increasing longevity among the populationand higher levels of physical activity, surgical repair of symp-tomatic lesions of the rotator cuff that do not presentimprovement through conservative treatment has come tobe considered to be a treatment option for patients aged 65years and over. Many forms of intervention have been advo-cated and there is some controversy regarding the type ofsurgical approach, i.e. whether to repair the lesion or justto perform debridement in isolation, among elderly patients.Some pioneering studies have demonstrated good resultsfrom subacromial decompression and arthroscopic debride-ment, which lead to temporary pain relief, but have inferiorand less long-lasting results than lesion repair.6,9,10,27Our find-ings from arthroscopic repair were good or excellent in 96.4%of the cases (157 shoulders) and were in line with some reportsin the literature, in which repairs to rotator cuff lesions per-formed arthroscopically have been indicated as consistentlybetter.
In 1995, Hattrup and Scottsdale2found a clear associa-tion between the size of the lesion and the patient's age.The presence of large and extensive lesions was significantlygreater among the older patients. Our study demonstratedthis trend, since the incidence of extensive lesions was 37.5%among patients up to 69 years of age; 44.2% among those aged70-74 years and 50% among those aged 75 years or over. Thelesions found in elderly patients are generally of degenerativeetiology, rather than of traumatic etiology; the musculature is found to have atrophied and the tendons have becomethinned and present poor quality for suturing (Figs. 1 and 2).
Although we did not find any studies in the literature cor-relating the size of the rotator cuff lesion with traumaticepisodes, trauma was reported by 62 patients (38%) and waspresent in 12 cases of large lesions (19.4%) and 37 of exten-sive lesions (59.7%). Thus, this study presented a statisticallysignificant correlation between histories of trauma and inci-dence of large and extensive lesions (p < 0.01). There was alsoa slight increase in the incidence of such lesions in relation toage progression (p = 0.11).
In an imaging study conducted by Boileau et al.,14thehealing rate for arthroscopically repaired lesions of thesupraspinatus among 65 patients aged 29-79 years was 71%.Age was an influential factor, given that patients over the ageof 65 years had a healing rate of 43% (p = 0.001). Nonetheless,despite the tendency toward worse results among patientsover the age of 65 years, which was also described by Hat-trup and Scottsdale,2Charousset et al.15and Lam and Mok,18our findings did not show any significant correlation betweenadvancing age groups and the final functional clinical result.
Boileau et al.14reported that the size of the lesion anddelamination of the tendon of the infraspinatus and/or sub-scapular muscle significantly impaired healing subsequentto arthroscopic suturing (p = 0.02). Charousset et al.15alsoreported that the size of the lesion generally correlated withthe prognosis. However, our study demonstrated that the sizeof the lesion was not related to the final results, and this find-ing was in line with that of Bittar,28who reported that 83%of the results from arthroscopic repair of extensive lesions inpatients of mean age 67 years were satisfactory. It is also inagreement with the results from Gartsman et al.21and Stoll-steimer and Savoie,29who showed that all sizes of lesionscould be treated arthroscopically, independent of age group.There were no differences between the results relating tosmall, medium-sized or large lesions.
Although Boileau et al.14did not find any correlationbetween healing and the duration of preoperative pain, ourstudy showed that the greater the time that elapsed betweenthe start of pain and the surgery was, the worse the final func-tional results were, with statistical significance (p = 0.027). Thestudy by Lam and Mok18also suggested that patients pre-senting symptoms for times longer than 34 months had a highlikelihood of worse functional results through surgical treat-ment. Ellman et al.26and Flurin et al.13reported that the longerthe time interval between preoperative pain and surgery was,
the larger the size of the lesion would be and, consequently,the more difficult it would be to repair it, with a worse prog-nosis.
The decision regarding which surgical technique to use(open, mini-open or totally arthroscopic) is generally basedon the surgeon's preference and experience. In an analysis onthe results obtained from arthroscopic repair of rotator cuffinjuries, Checchia et al.23concluded that this technique ismore efficient, both in relation to the results obtained (93.7%were excellent or good, according to the UCLA scores) andin relation to prevention of complications (9.2%). In relationto postoperative complications and treatment failures, ourresults are similar to those in the literature, with an incidenceof 6.1%. In a review of 40 articles on the results from openrepair of rotator cuff injuries conducted in 1997, Mansat et al.30found a complication rate of 10.5%. In 1992, Curtis et al.31reported a complication rate of 4.8% from arthroscopic pro-cedures and 8.8% from the mini-open technique. In contrast,in 1998, Berjano et al.32reported rates of 10.6% from arthro-scopic procedures and 5.3% from mini-open procedures. In2001, Grondel et al.17reported a complication rate of 6%, fromanalysis on extensive lesions repaired using the arthroscopicand mini-open techniques. On the other hand, from exami-nation of 263 patients who underwent arthroscopic repair ofrotator cuff lesions in 2007, Brislin et al.33reported a compli-cation rate of 10.6% (28 cases). Joint stiffness occurred morefrequently (23 cases) and this was unrelated to the extentof the lesion. In 2010, Verma et al.20also reported a similarcomplication rate (7.7%) and stated that the complication ratewas no greater among elderly individuals and therefore wasunrelated to age progression.
Among our patients, we found one case of adhesive cap-sulitis, which underwent serial block of the suprascapularnerve, which led to remission of the condition (UCLA 34);one case of tendinitis of the biceps, which was treated usingnon-steroidal anti-inflammatory drugs (UCLA 35); two casesof limitation of the range of motion in medial rotation, whichwere treated by means of physiotherapy (UCLA 13 and 33,respectively); two cases of painful renewed tearing: the firstunderwent new surgical repair and evolved with UCLA 33 andthe second was kept under expectant management, becauseof the patient's satisfaction (UCLA 29); two cases of non-painful renewed tearing with functional impotence, whichwere treated using physiotherapy, both with UCLA 33; one caseof arthropathy of the rotator cuff, which underwent reverseshoulder arthroplasty (UCLA 34); and one case of pain in theacromioclavicular joint, which underwent corticoid infiltra-tion and evolved with UCLA 35.
The incidence of dehiscence of suturing of rotator cuffinjuries after open repair is around 13-68%.34Slabaugh et al.35reported that this rate is probably greater than what is citedin the literature, because most of these cases are asymp-tomatic and do not imply any pain or significant functionalloss. In 2005, Ozbaydar et al.36used magnetic resonance imag-ing to evaluate fatty degeneration and recurrence of tearingof the rotator cuff subsequent to arthroscopic repair and
observed renewed tearing in 31.8%, even though 90.9% ofthe patients were satisfied. The extent of the fatty degener-ation did not present any difference from before to after theoperation. In 2010, Godinho et al.37evaluated the functionaland anatomical results from arthroscopic surgical treatmentof complete tearing of the rotator cuff in 100 patients ofmean age 60 years, and investigated the correlation betweenultrasound images and the Constant-Murley index. Theyfound that there was no statistically valid correspondencebetween them. They reported that 67% of the shoulders stud-ied presented excellent or good results, although 30% of thempresented ultrasound reports showing renewed tearing. Inour service, we did not routinely perform postoperative mag-netic resonance imaging examinations. In our study, theseexaminations were requested for two symptomatic casesthat showed dehiscence of the suture. Another two cases ofasymptomatic dehiscence were observed in patients who hadundergone the examination through requests from anotherservice.
In our sample, we found four cases of dehiscence of thesuture, but only two of these shown any symptoms.
All the rotator cuff lesions in our study were repairedcompletely, independently of the size of the lesion, or anyretraction or fatty degeneration. In no case were subacromialdecompression and debridement alone performed. Althoughwe are aware that dehiscence of the suture probably occursoften, we believe that good results can be expected throughthis technique.
It is important to emphasize that our study had the largestsample in the literature relating to arthroscopic repair of rota-tor cuff lesions among this age group. However, this studypresents limitations due to its retrospective nature, absence ofa control group and short minimum follow-up of 12 months,although the maximum recovery is achieved by 6-9 monthsafter the operation and from the twelfth month onwards, thereare almost no further changes.20Another limitation was thelack of analysis on the degree of fatty degeneration of mus-cles, as described by Goutallier et al.,38given that not all themagnetic resonance examinations of our patients included T1sagittal slices for complete analysis as described by Melladoet al.39in 2005.
Conclusion
Arthroscopic treatment of rotator cuff lesions in patients overthe age of 65 years presents good or excellent results in 96.4%,when evaluated using the UCLA functional method, with a lowcomplication rate.
Given that the objective of surgery is to improve pain andfunction, we believe that our study strongly suggests thatindications for repairing rotator cuff injuries among elderlypeople are valid, considering that advanced age does not influ-ence good postoperative clinical evolution. Likewise, the sizeof lesion is unrelated to the postoperative prognosis, eventhough large and extensive tearing is associated with the pres-ence of trauma. We can also conclude that the earlier thatthe surgical treatment is implemented, the better the finalfunctional results will be.
Conflicts of interest
The authors declare no conflicts of interest.
1. Ellman H, Kay SP, Wirth M. Arthroscopic treatment offull-thickness rotator cuff tears: 2- to 7-year follow-up study.Arthroscopy. 1993;9(2):195-200.2. Hattrup SJ, Scottsdale A. Rotator cuff repair: relevance ofpatient age. J Shoulder Elb Surg Am. 1995;4(2):95-100.3. Melillo AS, Savoie FH, Field LD. Massive rotator cuff tears:debridement versus repair. Orthop Clin N Am.1997;28(1):117-24.4. Montgomery TJ, Yerger B, Savoie FH. Management of rotatorcuff tears: a comparison of arthroscopic debridement andsurgical repair. J Shoulder Elb Surg Am. 1994;3(2):70-8.5. Fehringer EV, Sun J, VanOeveren LS, Keller BK, Matsen FA.Full-thickness rotator cuff tear prevalence and correlationwith function and co-morbidities in patients sixty-five yearsand older. J Shoulder Elb Surg. 2008;17(6):881-5.6. Gartsman G. Massive, irreparable tears of the rotator cuff.Results of operative débridement and subacromialdecompression. J Bone Joint Surg Am. 1997;79(5):715-21.7. Itoi E, Tabata S. Conservative treatment of rotator cuff tears.Clin Orthop Relat Res. 1992;(275):165-73.8. Ghroubi S, Chaari M, Elleuch H, Guermazi M, Baklouti S,Elleuch MH. Functional and quality of life outcome of noneoperated rotator cuff tears. Ann Readapt Med Phys.2008;51(9):714-21.9. Zvijac JE, Howard JL, Lemak LJ. Arthroscopic subacromialdecompression in the treatment of full thickness rotator cufftears: a 3 to 6 year follow-up. Arthroscopy. 1994;10(5):518-23.10. Rockwood CA Jr, Williams G Jr, Burkhead WZ Jr. Débridementof degenerative, irreparable lesions of the rotator cuff. J BoneJoint Surg Am. 1995;772(6):857-66.11. Weber SC. Arthroscopic debridement and acromioplastyversus mini-open repair in the management of significantpartial-thickness tears of the rotator cuff. Orthop Clin N Am.1997;28(1):79-82.12. Brewer BJ. Aging of the rotator cuff. Am J Sports Med.1979;7(2):102-10.13. Flurin PH, Landreau P, Gregory T, Boileau P, Lafosse L, Guillo S,et al. Cuff integrity after arthroscopic rotator cuff repair:correlation with clinical results in 576 cases. Arthroscopy.2007;23(4):340-6.14. Boileau P, Brassart N, Wakinson DJ, Charles M, Hatzidakis AM,Krishnan SG. Arthroscopic repair of full-thickness tears of thesupraspinatus: does the tendon really heal? J Bone Joint SurgAm. 2005;87(6):1229-40.15. Charousset C, Bellaiche L, Kalra K, Petrover D. Arthroscopicrepair of full-thickness rotator cuff tears: is there tendonhealing in patients aged 65 year or older? Arthroscopy.2010;26(3):302-9.16. Worland RL, Arredondo J, Angles F, Lopez-Jimenez F. Repair ofmassive rotator cuff tears in patients older than 70 years. JShoulder Elb Surg. 1999;8(1):26-30.17. Grondel RJ, Savoie FH, Field LD. Rotator cuff repairs in patients62 years of age or older. J Bone Joint Surg Am. 2001;10(2):97-9.18. Lam F, Mok D. Open repair of massive rotator cuff tears inpatients aged sixty-five years or over: is it worthwhile? JShoulder Elb Surg. 2004;13(5):517-21.19. Rebuzzi E, Coletti N, Schiavetti S, Giusto F. Arthroscopicrotator cuff repair in patients older than 60 years.Arthroscopy. 2005;21(1):48-54.20. Verma NN, Bathia S, Baker CL, Cole BJ, Nicholson GP, RomeoAA, et al. Outcomes of arthroscopic cuff repair in patientsaged 70 years or older. Arthroscopy. 2010;26(10):1273-80.21. Gartsman GM, Khan M, Hammerman SM. Arthroscopic repairof full-thickness tears of the rotator cuff. J Bone Joint SurgAm. 1998;80(6):832-40.22. Burkhart SS. A stepwise approach to arthroscopic rotator cuffrepair based on biomechanical principles. Arthroscopy.2000;16(1):82-90.23. Checchia SL, Santos PD, Miyazaki AN, Fregoneze M, Silva LA,Ishi M, et al. Avaliac¸ão dos resultados obtidos na avaliac¸ãoartroscópica do manguito rotador. Rev Bras Ortop.2005;40(5):229-38.24. Hawkins RJ, Misamore GW, Hobeika PE. Surgery for fullthickness rotator-cuff tears. J Bone Joint Surg Am.1985;67(9):1349-55.25. Gerber C, Fuchs B, Hodler J. The results of repair of massivetears of rotator cuff. J Bone Joint Surg Am. 2000;82(4):505-15.26. Ellman H, Hanker G, Bayer M. Repair of the rotator cuff:end-result study of factors influencing reconstruction. J BoneJoint Surg Am. 1986;68(8):1136-44.27. Kempf JF, Gleyze P, Bonnomet F, Walch G, Mole D, Frank A,et al. A multicenter study of 210 rotator cuff tears treated byarthroscopic acromioplasty. Arthroscopy. 1999;15(1):55-66.28. Bittar ES. Arthroscopic management of massive rotator cufftears. Arthroscopy. 2002;18(9 Suppl 2):104-6.29. Stollsteimer GT, Savoie FH 3rd. Arthroscopic rotator cuffrepair: current indications, limitations, techniques andresults. Instr Course Lect. 1998;47:59-65.30. Mansat P, Cofield RH, Kersten TE, Rowland CM. Complicationsof rotator cuff repair. Orthop Clin N Am. 1997;28(2):205-13.31. Curtis AS, Snyder SJ, Del Pizzo W, Friedman MJ, Ferkel RD,Karzel RP. Complications of shoulder arthroscopy.Arthroscopy. 1992;8(3):395.32. Berjano P, Gonzalez BG, Olmedo JF, Perez LA, Munilla MG.Complications in arthroscopic shoulder surgery. Arthroscopy.1998;14(8):785-8.33. Brislin KJ, Field LD, Savoie FH. Complications afterarthroscopic rotator cuff repair. Arthroscopy. 2007;23(2):124-8.34. Jost B, Pfirrmann CW, Gerber C, Switzerland Z. Clinicaloutcome after structural failure of the rotator cuff repairs.Bone Joint Surg Am. 2000;82(3):304-14.35. Slabaugh MA, Nho SJ, Grumet RC, Wilson JB, Romeo AA,Verma NN, et al. Does the literature confirm superior clinicalresults in radiographically healed rotator cuffs after rotatorcuff repair? Arthroscopy. 2010;26(3):393-403.36. Ozbaydar MU, Tonbul M, Yalaman O. The results ofarthroscopic repair of full-thickness tears of the rotator cuff.Acta Orthop Traumatol Turc. 2005;39(2):114-20.37. Godinho GG, Franc¸a FO, Freitas JMA, Watanabe FN, Nobre LO,Almeida Neto MA, et al. Avaliac¸ão da integridade anatômicapor exame de ultrassom e funcional pelo índice de Constant& Murley do manguito rotador após reparo artroscópico. RevBras Ortop. 2010;45(2):174-80.38. Goutallier D, Postel JM, Bernageau J, Lavau L, Voisin MC. Fattymuscle degeneration in cuff ruptures - pre and postoperativeevaluation by CT scan. Clin Orthop Relat Res. 1994;(304):78-83.39. Mellado JM, Calmet J, Olona M, Esteve C, Camins A, PalomarLP, et al. Surgically repaired massive rotator cuff tears: MRI oftendon integrity, muscle fatty degeneration, and muscleatrophy correlated with intraoperative and clinical findings.AJR Am J Roentgenol. 2005;184(5):1456-63.
