 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1996 The Journal of Bone and Joint Surgery, Inc. Arthroscopic Release for Chronic, Refractory Adhesive Capsulitis of the Shoulder*   Investigation performed at the Shoulder Service, Center for Sports Medicine, University of Pittsburgh, Pittsburgh
Idiopathic adhesive capsulitis usually responds to gentle physical therapy or, if that fails, to closed manipulation with the patient under anesthesia. In some patients, however, loss of motion may be refractory to either of these treatments and an operative release may be indicated. We are reporting on the technique and results of arthroscopic capsular release as a new alternative for the management of such patients. During a three-year period, we managed twenty-three patients who had idiopathic adhesive capsulitis that had failed to respond to physical therapy or closed manipulation. These patients had an arthroscopic anterior capsular release and received forty-eight hours of intensive physical therapy as inpatients. During the physical therapy, the patients received an interscalene regional analgesic with use of repeated nerve blocks or with a continuous infusion through an interscalene catheter. This was followed by a supervised outpatient physical-therapy program. Six patients also had an arthroscopic acromioplasty for the treatment of impingement. There were no complications related to any of the procedures. At a mean of thirty-nine months (range, twenty-four to sixty-four months) after the arthroscopic procedure, the improvement in the score of Constant and Murley averaged 48 points (range, 13 to 77 points). The mean improvement in motion was 49 degrees (range, 0 to 105 degrees) for flexion; 42 degrees (range, 10 to 80 degrees) and 53 degrees (range, 0 to 100 degrees) for external rotation in adduction and abduction, respectively; and eight spinous-process levels (range, three to fourteen levels) and 33 degrees (range, 30 to 60 degrees) for internal rotation in adduction and abduction, respectively. These gains in motion were all significant (p < 0.01) compared with the preoperative values and were within a mean of 7 degrees of the values for the contralateral, normal shoulder. We concluded that, in patients who have loss of motion that is refractory to closed manipulation, arthroscopic capsular release improves motion reliably with little operative morbidity.
Idiopathic adhesive capsulitis affecting the glenohumeral joint is believed to be a self-limited condition and can usually be treated successfully with physical therapy7,11,12,14,18,22,26,27,29,31,49-51,55,57. Some patients may, however, be unwilling to wait the one to three years often required for the resolution of symptoms12,25,27,28,31,55. Furthermore, some studies have demonstrated long-term residual pain and limitation of motion in most patients after non-operative treatment2,5,6,31,49-51. For these reasons, when little motion has been gained despite a prolonged supervised physical-therapy program, some authors have recommended closed manipulation with the patient under anesthesia13,21,28,38,39,41,42,47,49,57. Although successful for some patients13,42,47,49,51, manipulation has been associated with complications such as humeral fracture, neural injury, and dislocation of the shoulder10,25,30. For some patients who have failed to gain mobility after manipulation, an open soft-tissue release has been suggested, often in association with lengthening of the subscapularis14,23,25,36,37,42,46. This approach is associated with the morbidity of an extensive operative dissection and with the need to protect the repair of the subscapularis during initial postoperative motion. While some authors have suggested that arthroscopy is not useful in either the diagnosis or the treatment of a stiff shoulder38,39,41,42, others have recommended its use to help to delineate abnormalities, to document the results of closed manipulation, and to assist in distention of the contracted joint capsule3,19,44,48,53,56. Moreover, some investigators have proposed arthroscopically guided sectioning of the contracted capsule as a treatment for adhesive capsulitis14,48. During a three-year period, we prospectively evaluated the efficacy of arthroscopic release for patients who had chronic, refractory adhesive capsulitis. The purpose of the current paper is to describe the results of this treatment.
Selection of Patients All eighty-one patients were managed with a supervised physical-therapy program for at least four months. After completion of the program, forty-eight patients noted that the pain had decreased and the motion had improved to the point where they were comfortable sleeping at night and they could perform all of their daily work and recreational activities with slight or no limitations. Although we did not have an absolute criterion for acceptable improvement of motion in these patients, all achieved motion to within 80 per cent of that of the contralateral, normal extremity. The remaining thirty-three patients continued to have loss of both active and passive motion as well as pain for a mean duration of twelve months (range, four to forty-eight months), despite a supervised physical-therapy program that averaged six and one-half months (range, four to twelve months). All thirty-three patients had an attempted closed manipulation under anesthesia, and ten of them had demonstrable improvement in motion and decreased pain. In the remaining twenty-three patients, motion did not improve after manipulation under anesthesia. They had an arthroscopic release, and they form the basis of this study. The mean age of these patients was forty-eight years (range, twenty-seven to seventy-three years). Seven right and sixteen left shoulders were involved, in eleven men and twelve women.
Clinical Assessment The function of all shoulders was graded according to the 100-point scoring system of Constant and Murley8,9. This system combines assessments of subjective symptoms and objective findings. Pain is given a maximum of 15 points, with 15 representing no pain and 0, severe, constant pain. Activities of daily living are assigned a maximum of 20 points. Flexion, abduction, external rotation, and internal rotation are each given a maximum of 10 points (total, 40 points). Pain-free motion of the shoulder is given 25 points, and marked weakness is assigned 0 points. Six patients had painful flexion, which suggested concomitant impingement. This diagnosis was confirmed by a positive impingement sign, as described by Neer32 or by Hawkins and Hobeika17, and a positive impingement test32. The Neer impingement sign is considered positive when the examiner causes pain by elevating the humerus with one hand while depressing the scapula with the other hand. The impingement sign of Hawkins and Hobeika is considered positive when flexion to 90 degrees combined with internal rotation and horizontal adduction produces pain. An impingement test is considered positive when pain with elevation of the shoulder is eliminated after an injection of fifteen milliliters of 1 per cent Lidocaine (xylocaine) into the sub-acromial space through a posterior approach. Standard radiographs were made in the anteroposterior and axillary planes. For the six patients who had pain in the anterosuperior portion of the shoulder with active flexion, special radiographs of the acromion (a supraspinatus outlet radiograph1,35 and a caudal tilt radiograph45) were made in order to define the supraspinatus outlet. These patients were found to have either a type-II (curved) or a type-III (hooked) acromion according to the classification of Bigliani et al.1 Some patients who were referred to us had had either an arthrogram or a magnetic resonance image made; neither had demonstrated any intrinsic abnormality of the shoulder. These studies were not performed routinely for all of our patients. Finally, videotaped documentation of the function and the range of motion of the shoulder both preoperatively and postoperatively was available for all patients. All operative findings and procedures were recorded on videotape and photographs as well.
Operative Technique
Anesthesia and Postoperative Analgesia Toward the end of the series, we began to use an interscalene catheter to provide both anesthesia and postoperative analgesia4. The patients received a continuous infusion of 0.25 per cent bupivacaine at a rate of six milliliters an hour for forty-eight hours. All patients also had self-administered (patient-controlled) analgesia through an intravenous pump that was set to administer one milligram of morphine every eight minutes as needed, to a maximum of thirty milligrams in four hours.
Anterior Capsular Release Arthroscopy was performed with the patient seated in an upright beach-chair position54. An 18-gauge spinal needle was introduced posteriorly into the joint. As the humeral head was compressed tightly into the glenoid, this needle had to be inserted carefully over the humeral head in order to enter the joint. Sterile saline solution was then injected; however, the volume of the contracted joint usually permitted no more than ten milliliters. As it may be difficult to place the spinal needle into the contracted, small joint, it is important to confirm its intra-articular position by observing a retrograde flow of fluid out of the needle when the syringe is removed. Inflation of the joint with fluid was also helpful as increased fluid pressure pushed the humeral head away from the glenoid and reduced the risk of injury to the cartilage when the arthroscope was inserted. The arthroscope was introduced carefully into the joint, after the spinal needle had been oriented over the humeral head. Correct intra-articular placement of the arthroscope was confirmed by backflow of fluid out of the arthroscopic sheath. At this point, it was possible to see only the anterosuperior region of the joint. An 18-gauge spinal needle was then inserted anteriorly so that it entered the joint just underneath the biceps tendon. The needle was then replaced with a seven-centimeter arthroscopic shoulder cannula (Linvatec, Concept Arthroscopy, Largo, Florida). A 4.5-millimeter motorized shaver (full radius resector; Linvatec, Concept Arthroscopy) was used to debride the joint of inflamed synovial and scar tissue through the anterior portal. An arthroscopic electrocautery device with a hooked tip (Linvatec, Concept Arthroscopy) was then placed through the anterior cannula, and the anterior capsular scar just underneath the biceps tendon was divided. Division of the capsule was in its mid-portion and proceeded from superior to inferior until the interval between the superior border of the subscapularis tendon and the anterior border of the supraspinatus tendon (the rotator interval region of the capsule15) was encountered (Figs. 1-A and 1-B). As the thickened capsular tissue was divided, the rotator interval opened and the joint became more mobile, with the humeral head moving inferiorly and laterally. This allowed the arthroscope to be inserted farther into the joint anteriorly and inferiorly. After this region had been released fully and the tendon of the subscapularis was seen to be freely mobile with external rotation of the arm, the arthroscopic instruments were removed and a gentle manipulation was performed.
In eighteen patients, release of the rotator interval region of the capsule restored external rotation of the adducted shoulder without any need for manipulation. A gentle closed manipulation was performed to restore motion in other planes. In five patients, although release of the rotator interval restored external rotation in adduction the patient continued to have limitation of external rotation in abduction. These patients were found to still have a thick, contracted anterior portion of the capsule distal to the level of the proximal border of the subscapularis tendon. This portion of the capsule was then divided with the electrocautery device down to the anteroinferior portion of the glenoid (Fig. 1-C), and the shoulder was manipulated to regain external rotation in abduction.
Additional Procedures
Postoperative Treatment The patients were discharged on the afternoon of the second postoperative day, after the second physical-therapy session of that day. For the first two weeks, they attended daily sessions of supervised physical therapy and performed a home-exercise program consisting of pulley and cane-assisted motion in all planes. This was then reduced to supervised therapy two or three times a week for the next four weeks, after which the program was individualized according to each patient's progress. Patients were not given a sling support at any time, and they were encouraged to use the affected extremity for activities of daily living as soon as possible after the operation. Strengthening was begun as soon as pain and active motion of the shoulder allowed. Patients were encouraged to try to swim in a pool between two to four weeks after the operation.
Statistical Analysis
The mean duration of follow-up was thirty-nine months (range, twenty-four to sixty-four months). There were no complications related to the operation. Arthroscopy revealed a thick, shortened anterosuperior portion of the capsule in all twenty-three patients (Figs. 1-A, 1-B, and 1-C). Fourteen patients also had marked synovitis. Six patients had subacromial bursitis and a prominent anterior portion of the acromion. No patient had abnormalities of the articular cartilage, biceps tendon, or rotator cuff. The mean improvement in the score of Constant and Murley8,9 was 48 points (range, 13 to 77 points) (Fig. 2). All patients indicated that they had no pain or only occasional, mild discomfort when using the shoulder forcefully. Nine patients stated that they had participated in sports involving the upper extremity, such as swimming, tennis, golf, and softball, before the problem with the shoulder had begun. After the procedure, seven of these nine patients were able to return to their previous level of participation.
All improvements in range of motion were significant (p < 0.01) (Figs. 3-A, 3-B, 3-C, 3-D through 3-E). Flexion improved a mean of 49 degrees (range, 0 to 105 degrees); external rotation in adduction, a mean of 42 degrees (range, 10 to 80 degrees); external rotation in abduction, a mean of 53 degrees (range, 0 to 100 degrees); internal rotation in adduction, a mean of eight spinous-process levels (range, three to fourteen levels); and internal rotation in abduction, a mean of 33 degrees (range, 30 to 60 degrees). With the numbers available, we could not detect a significant difference between the postoperative range of motion and that of the contralateral, asymptomatic shoulder for flexion, although there remained a small difference (7 ± 9 degrees [mean and standard deviation]; p < 0.05) for external and internal rotation.
Although the etiology of primary adhesive capsulitis remains in question, it is generally believed that the condition is benign and that it has a self-limited course7,11,12,18,22,26,27,29,49,52,55,57. Nevertheless, some studies have shown long-term functional disability2,5,6,50,51. The duration of non-operative treatment for adhesive capsulitis has been debated, but most authors have recommended two to six months14,25,30,41,42,48. Neviaser and Neviaser42 suggested that patients in whom flexion was still limited to less than 90 degrees and those who had not had any improvement in motion were candidates for closed manipulation. We arbitrarily chose a four-month limit, as many of our patients were unwilling to continue therapy for a longer duration when no gains were being made. Furthermore, we had no specific criterion for loss of motion in order for our patients to have this procedure, although all of our patients had pain and considered the motion of the shoulder to be severely restricted. Forty-eight (59 per cent) of our original group of eighty-one patients responded to a therapy program, while ten of the remaining thirty-three had improvement after closed manipulation14,42. Although many of our patients did not have an arthrogram, all patients had limitation of both passive and active motion, indicating that soft-tissue contracture was a limiting factor. Furthermore, while it has been suggested that the arthrographic finding of a contracted axillary pouch is diagnostic of adhesive capsulitis, other authors have not found such a relationship with measured losses of motion20,40,43. We excluded all patients who had loss of motion because of an operation or who had had severe trauma such as a fracture, as these conditions may be due to a pathological process that is different than idiopathic adhesive capsulitis14,33. In addition, many such patients have an extra-articular component to the loss of motion14,23,48. Because our technique of arthroscopic release is only for capsular lesions, we would expect it to be difficult to improve motion in patients who have an extra-articular component. Only a few studies have actually documented the technique of measurement and the range of motion in patients who have adhesive capsulitis2,6,51. Our patients were evaluated prospectively by the same examiner, with use of the same techniques of measurement to assess the range of motion. Furthermore, all patients had videotaped documentation of the preoperative, intraoperative, and postoperative range of motion. Overall function of the shoulder was assessed with use of the scoring system of Constant and Murley8,9, which has been shown to be a sound method of evaluation for degenerative conditions of the shoulder. Postoperative control of pain has been shown to be critical to maintenance of gains in motion after closed manipulation for the treatment of adhesive capsulitis14,42,48. In the current series, although all patients ultimately had improvement it was our impression that regional anesthesia and postoperative analgesia afforded better relief of pain and comfort than general anesthesia and postoperative analgesia with narcotics. We now routinely use an interscalene block or an interscalene catheter and block for both anesthesia and postoperative control of pain. We have found, as have others4,14,24,48, that this technique is safe and well tolerated and that it usually reduces the need for narcotics while allowing intensive passive range-of-motion therapy in the immediate postoperative period. For patients who have adhesive capsulitis that remains refractory to closed manipulation, some surgeons have recommended an open release16,36,37,46. Ozaki et al.46, Neer et al.34,36,37, and others14,15,48 have pointed out that contracture and scarring of the rotator interval region of the capsule is a major component of refractory adhesive capsulitis of the shoulder. The major capsulo-ligamentous components are the coracohumeral and superior glenohumeral ligaments, and in the normal shoulder these structures are usually elastic and function only to limit excessive external rotation and inferior translation in the adducted arm15,16,36,37,46. Open release of the rotator interval region has been successful in the treatment of chronic, refractory adhesive capsulitis36,37,46. Occasionally, if the capsular contracture is more extensive, open operative release of the inferior portion of the capsule may be necessary, and in some patients a z-plasty lengthening of the subscapularis and the capsule in the coronal plane is performed14,23. Some authors39,41,43 have thought that arthroscopy is of no diagnostic or therapeutic use in the management of patients who have adhesive capsulitis of the shoulder. Others have thought that it may even be contraindicated because of the risk of iatrogenic chondral injury from insertion of an instrument into a tight shoulder joint41,42. Pollock et al.48 thought that arthroscopy was useful mainly after closed manipulation to treat concomitant lesions, but they also indicated that arthroscopic sectioning of the coracohumeral ligament might be helpful in some patients who have loss of external rotation that is refractory to non-operative treatment. They were, however, concerned about the risk of injury to the axillary nerve if capsular release was continued more inferiorly than the superior border of the subscapularis tendon. Harryman14 applied an arthroscopic release technique to a small number of patients who had primary adhesive capsulitis. Most surgeons have agreed that insertion of an arthroscope into a stiff shoulder with a limited joint volume is difficult3,38,39,41,42,56. Wiley56 recommended use of a smaller, 3.8-millimeter arthroscope. We were able to place a standard-size arthroscope into the joint of all of our patients without creating any chondral injuries. We did this by carefully guiding the arthroscope over the humeral head and beginning the division of the capsule in the rotator interval region just underneath the biceps tendon. This allowed the humeral head to gradually move inferiorly and laterally, creating more space for the arthroscope to be inserted into the anterior and inferior portion of the joint. In all of our patients, we found the rotator interval region of the capsule to be composed of dense, thick, unyielding scar tissue. In eighteen patients, section of the rotator interval region combined with gentle manipulation effectively restored motion in all planes; however, in the remaining five patients, additional division of the remaining anterior portion of the capsule was necessary. Our arthroscopic technique allowed us to perform this extensive release without the extensive dissection that would have been required for an open release; however, even after the release no patient had instability. We believe that this was because of articular compression through the remaining soft-tissue tension around the joint. In summary, patients who have adhesive capsulitis should be managed non-operatively with a supervised physical-therapy program. While most will have gains in motion and relief of pain to a level where they are satisfied, some will continue to have decreased motion and will be candidates for closed manipulation under anesthesia. In patients in whom closed manipulation fails, arthroscopy offers the added advantage of a precise, controlled capsular release as well as the treatment of concomitant lesions. Manipulation after such a release requires much less force in order to restore motion.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
¶Department of Orthopaedics, Northwest Private Hospital, Brickport Road, Burnie, Tasmania 7320, Australia.
  CiteULike  Connotea  Del.icio.us  Facebook  Technorati  Twitter What's this?
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |