BACKGROUND:
Rotator cuff disease is uncommon in primary glenohumeral osteoarthritis. Consequently, the prognostic implications of rotator cuff disease in patients undergoing prosthetic replacement for the treatment of primary glenohumeral osteoarthritis are uncertain. The purpose of this study was to report the effects of the condition of the supraspinatus tendon and the rotator cuff musculature on the results of shoulder arthroplasty in the treatment of primary osteoarthritis.

METHODS:
Five hundred and fifty-five shoulders in 514 patients who had an arthroplasty for the treatment of primary glenohumeral osteoarthritis as part of a multicenter study were evaluated. Forty-one shoulders had a partial-thickness tear of the supraspinatus, and forty-two had a full-thickness tear. Ninety shoulders had moderate (stage-2) fatty degeneration of the infraspinatus, and nineteen had severe (stage-3 or 4) degeneration. Eighty-four shoulders had moderate fatty degeneration of the subscapularis, and fifteen had severe degeneration. The influence of the condition of the supraspinatus tendon and the infraspinatus and subscapularis musculature on the postoperative outcome was evaluated with respect to the scores according to the system of Constant and Murley, active mobility, subjective satisfaction, radiographic result, and rate of complications.

RESULTS:
The shoulders were evaluated at a mean of 43.1 months postoperatively. With the numbers available, supraspinatus tears were not found to influence the postoperative outcome with respect to the total Constant score, active mobility, subjective satisfaction, radiographic result, or rate of complications. Additionally, the treatment of these tears did not markedly influence the outcome parameters. Conversely, both shoulders with moderate fatty degeneration and those with severe degeneration of the infraspinatus were associated with poorer results than those with no degeneration with respect to the total Constant score (p < 0.0005), active external rotation (p < 0.0005), active forward flexion (p = 0.001), and subjective satisfaction (p = 0.031). Similar although less dramatic results were seen with fatty degeneration of the subscapularis.

CONCLUSIONS:
This study demonstrates that minimally retracted or nonretracted rotator cuff tears that are limited to the supraspinatus tendon do not appreciably affect most shoulder-specific outcome parameters in shoulder arthroplasty performed for the treatment of primary osteoarthritis. Conversely, fatty degeneration of the infraspinatus and, less importantly, subscapularis musculature adversely affects many of these parameters.