- Email: [email protected]
Arthroscopic debridement for dialysis shoulders
Arthroscopic Debridement for Dialysis Shoulders Koji Midorikawa, M.D., Masafumi Hara, M.D., Gen Emoto, M.D., Yozo Shibata, M.D., and Masatoshi Naito, M.D.
Purpose: To arthroscopically treat “dialysis shoulder,” severe shoulder pain in patients on long-term hemodialysis. This pain occurs only at rest such as during hemodialysis or while sleeping and is temporarily alleviated by assuming the sitting position or moving the shoulder joint. Limitations in range of motion and pain with overhead activity or the arm in the forward flexed position are not the patient’s chief complaints. Although frozen shoulder and impingement syndrome may be observed as complications, pain at rest is the most characteristic. Type of Study: Clinical research on arthroscopic debridement to treat dialysis shoulder. Methods: We performed arthroscopic debridement of 36 dialysis shoulders in 29 patients. Only complete debridement in the glenohumeral joint and subacromial bursa was performed, without invasion to the bone and ligaments. Results: The pain improved in 34 shoulders in 27 patients (94%) a mean of 29.8 months after surgery, showing satisfactory results. In this group, the mean Japan Orthopaedic Surgery Association shoulder score (maximum 100 points) was 66.4 points before surgery but increased to 86.6 points postoperatively, statistically significant by 2-group t test. Conclusions: We decided before beginning the study that no postoperative rehabilitation would be necessary. There were no complications and no need for further surgery, with all but 2 of the patients being satisfied with their postoperative condition. Key Words: Amyloid—Arthroscopy—Arthroscopic surgery—Hemodialysis—Shoulder—Subacromial bursa.
T
he incidence of bone and joint disorders in hemodialysis (HD) patients increases with the duration of HD therapy. Shoulder pain is observed in many patients, some of whom complain of severe shoulder pain only at rest such as during HD or at night. The nature of the pain and its occurrence time are very characteristic and differ markedly from that observed in other disorders. However, most patients leave this pain untreated partly because the mechanism of the pain is still unclear. Although there are various theories on its mechanism, none of them has been confirmed. Some surgical methods have also been reported, but none of them are established. In our
From the Department of Orthopaedic Surgery, School of Medicine, Fukuoka University, Fukuoka, Japan. Address correspondence and reprint requests to Koji Midorikawa, M.D., 7-45-1 Nanakuma, Jyonan-Ku, Fukuoka 8140180, Japan. E-mail: [email protected] © 2001 by the Arthroscopy Association of North America 0749-8063/01/1707-2590$35.00/0 doi:10.1053/jars.2001.23584
department, good results have been obtained by arthroscopic debridement for chronic rheumatoid arthritis (RA),1 in which joint pain is caused primarily by synovitis. We speculated that shoulder pain in patients on long-term HD is also caused by synovitis on the basis of findings of imaging studies and arthroscopic findings, and have applied arthroscopic debridement to this condition. We defined “dialysis shoulder” as severe shoulder pain in patients on long-term HD that occurs only at rest such as during HD or sleep at night and is temporarily alleviated by assuming the sitting position or moving the shoulder joint. Limitations in the range of motion (ROM) and pain with overhead activity or with the arm in the forward flexed position are not the patient’s chief complaints. Although impingement and capsulitis may occur in this condition, the chief complaint is pain at rest, and we selected our subjects using this as the primary criterion.2 In this report, patients with severe shoulder pain who underwent arthroscopic debridement are presented. Arthroscopic debridement was
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 17, No 7 (September), 2001: pp 685– 693
685
686
K. MIDORIKAWA ET AL.
performed in patients with the above-defined dialysis shoulder who complained of sleep disorders resulting from nocturnal shoulder pain or pain during HD. METHODS The study included 29 patients (36 shoulders) with dialysis shoulders. The patients in this sample group were between 3 and 56 months after surgery (average, 29.8 months), with follow-up care continuing in all cases indefinitely. There were 16 men (19 shoulders) and 13 women (17 shoulders) with a mean age of 56.8 years (range, 39 to 78 years). The mean HD history was 16.3 years (range, 6 to 25 years), and shoulder pain developed a mean of 12.8 years (range, 6 to 18 years) after initiation of HD. In 12 patients, the operated shoulder was the same side as the shunt. In 9 patients the operated shoulder was the opposite side to the shunt. The remaining 8 patients underwent bilateral operations. In these patients, improvement in pain was evaluated using the Japan Orthopaedic Surgery Association Shoulder (JOA) score. In the JOA score, 30 points are assigned to pain, 20 points to function, 30 points to the range of motion the joint (ROM), 5 points to radiographic findings, and 15 points to joint stability with a full mark of 100 points (Table 1). Operative Technique For the arthroscopic surgery, patients were placed in the lateral decubitus position and 2 to 3 kg of skin traction was applied to the arm in about 30° abduction of the shoulder. Because traction was carried out using a brace made of thick sponge, we could arthroscope the side with the shunt in place.3-5 Arthroscopic surgery consisted of synovectomy and complete debridement in the glenohumeral joint (GHJ) and the subacromial bursa (SAB). No procedures that change the anatomic architecture such as cutting of ligament or resecting of bone were performed. After the procedure, a sling was applied, but free use of the shoulder was allowed from the next day. In all cases, the synovial membrane biopsy specimens taken during arthroscopy were stained with Congo red and examined histopathologically by light microscopy and polarizing microscopy. Preoperative Evaluation Physical examination before surgery was measured using the JOA score, and the results were as follows:
Pain, average 5.6 points (range, 5-10 points); Function, average 17.0 points (range, 10.5-20 points); ROM, average 24.4 points (range, 13-30 points); Radiographic findings, average 4.7 points (range, 3-5 points); Joint stability, average 15 points. Plain radiographic examination, ultrasonography, arthrography of the GHJ, and magnetic resonance imaging (MRI) were all used (without exception) to observe the morphology in every painful shoulder. Plain radiographs showed small cyst-like shadows and moth-eaten images in the humeral head, but negligible arthritic changes (Fig 1). Ultrasonography showed a thickened rotator cuff, heterogeneous echoes in the rotator cuff, and fluid retention in the SAB (Fig 2). T2-weighted MRIs revealed scattered low-intensity areas in the rotator cuff, indicating degeneration, and high-intensity fluid retained in the GHJ and SAB as well as in the cyst in the humeral head (Fig 3). Arthrography showed enlargement of the capsule and irregularity of the inner wall (Fig 4). These findings were all common in varying degrees to all patients. In order to predict the outcome of surgery, a local anesthetic was injected into the GHJ and SAB before surgery and pain was evaluated. The analgesic effects differed among the patients. RESULTS The analgesic effects of the local anesthetic injection were varied and followed no common pattern. Typical arthroscopic findings were as follows. In the GHJ, synovial villi resembled ice-covered twigs that, on close examination, had no visible surface capillaries, a characteristic of this disease. In some areas, polyp-like proliferation of synovial villi and degeneration of articular cartilage were present. Avascular synovial proliferation was a common finding in all patients (Fig 5). Postoperative arthroscopy, following arthroscopic debridement, showed an intact subscapularis tendon and middle glenohumeral ligament in all patients (Fig 6). In the SAB, synovial proliferation similar to that in the GHJ, adhesion, and narrowing were observed to varying degrees in all patients (Fig 7). After debridement, there was a significant increase in space in the subacromial region (Fig 8). Success of the surgery focused on how successful we were at relieving the preoperative pain. Nocturnal pain and pain during HD rapidly disappeared after operation in 34 shoulders of the 27 patients. However, in the other 2 patients (2 shoulders) there was no improvement in pain, indicating a failure of the procedure. Therefore, these 2 patients were classified as
ARTHROSCOPIC DEBRIDEMENT FOR DIALYSIS SHOULDERS
687
TABLE 1. JOA Shoulder Score (total 100 points) Pain (30 points) None Tenderness or minimal pain in sports or heavy labor Minimal pain in ADL Moderate and tolerable pain (analgesic needed, occasional night pain) Severe pain (ADL limited, frequent night pain) Totally incapacitated because of pain Function (20 points) Total function (10 points) Strength in abduction (5 points, measured at 90° abduction or at possible level) Normal Excellent Good Fair Poor Zero Endurance (5 points, time in seconds holding 1-kg dumbbell horizontally with elbow extended and forearm pronated) More than 10 seconds More than 3 seconds More than 2 seconds Zero Activities of daily living (10 points) Making knot in back 1, 0.5, 0 Reaching opposite axilla 1, 0.5, 0 Open and close sliding door 1, 0.5, 0 Wearing jacket 1, 0.5, 0 Sleep on involved side 1, 0.5, 0 Self-hygienic care 1, 0.5, 0 Reaching side pocket (jacket) 1, 0.5, 0 Combing hair 1, 0.5, 0 Reaching overhead shelf 1, 0.5, 0 Reaching mouth 1, 0.5, 0 Subtract one point from above for each unable activity, specify: 1. 2. 3. Range of motion (30 points) Elevation (15 points) ⬎150° ⬎120° ⬎90° ⬎60° ⬎30° 0° External rotation (9 points) ⬎60° ⬎30° ⬎0° ⬎⫺20° ⬍⫺20° Internal rotation (6 points) Above T12 spinous process Above L5 spinous process Gluteal Below gluteal Radiographic evaluation (5 points) Normal Moderate changes of subluxation Advanced changes or dislocation Joint stability (15 points) Normal Slight instability or apprehension Severe instability or history or state of subluxation Relevant history or state of dislocation Remarks: Record range of motion and pain of elbow and hand disabilities, if present.
30 25 20 10 5 0
5 4 3 2 1 0 5 3 1 0
15 12 9 6 3 0 9 6 3 2 0 6 4 2 0 5 3 0 15 10 5 0
688
K. MIDORIKAWA ET AL. (mean, 20.2 points). With alleviation of pain, the score in function and ROM increased slightly (Table 2). The improvement achieved in the category of pain (5.6/30 to 23.4/30 points) is statistically significant by 2-group t test (P ⬍ .0001), underlining the benefit of the operative technique. Improvement in function (16.8/20 to 18.3/20 points) and ROM (24.3/30 to 25.2/30 points) were also statistically significant by 2-group t test (Fig 9). Comparison of Results Comparison between the no-response and goodresponse groups showed no definite differences in regard to age, duration of HD, the time of onset of shoulder pain, the relationship between the shunt side and pain side, nor the preoperative JOA score (Table 3). Comparison of Arthroscopic Findings The arthroscopic findings in all cases were characterized by synovial villi resembling ice-covered twigs with no visible vascularization. Comparison of Histopathologic Findings Histopathologic examination of the synovial tissue resected at arthroscopy showed amyloid deposition in all cases. The specimen shown in Fig 10, stained with Congo red, shows homogenous amorphous amyloid deposition lacking cell components in the synovial membrane. Under a polarizing microscope, yellowgreen fluorescence of amyloid was shown (Fig 11).
FIGURE 1. Plain radiographic examination showed small cyst-like shadows and moth-eaten images in the humeral head but negligible arthritic changes: (A) anteroposterior view (B) axial view.
the no-response group and compared with the goodresponse group (27 patients, 34 shoulders). Results in the Good-Response Group In the good-response group, the mean JOA score was 66.4 points before operation and increased to 86.6 points after operation. The increase in the score of pain (17.8 points) accounted for most of the increase
FIGURE 2. Ultrasonography showed a thickened rotator cuff (large arrows), heterogeneous echoes in the rotator cuff, and fluid retention in the SAB (small arrows).
ARTHROSCOPIC DEBRIDEMENT FOR DIALYSIS SHOULDERS
689
FIGURE 3. T2-weighted MRIs revealed scattered low-intensity areas in the rotator cuff, indicating degeneration, and high-intensity fluid retained in the GHJ and SAB as well as in the cyst in the humeral head: (A) coronal view (B) axial view.
DISCUSSION There have been some recent studies on surgery for shoulder pain in patients on long-term HD.5-10 The underlying disease varied among the reports. Some investigators6,7 use patients with impingement syndrome or stiff shoulder as the subjects, whereas our subjects did not present with such pain on motion or ROM, and so comparison of the 2 would be redundant.
FIGURE 4. Arthrography showed enlargement of the capsule and irregularity of the inner wall: (A) internal rotation position (B) external rotation position.
690
K. MIDORIKAWA ET AL.
FIGURE 5. Preoperative arthroscopic findings in the GHJ: avascular ice-covered twig-like synovial villi, characteristic of this disease, were observed. In some areas, polyp-like proliferation of synovial villi was also seen.
At our department, arthroscopic debridement is performed for chronic RA in which the primary cause of joint pain is synovitis.3 Shiota et al.11 reported that the ratio of chondroitin 6-sulfate to chondroitin 4-sulfate
FIGURE 6. Postoperative arthroscopic findings in GHJ: intact subscapular muscle (single arrow head) and middle glenohumeral ligament (double arrow heads) are shown.
FIGURE 7. Preoperative arthroscopic findings in SAB: synovial proliferation similar to that in the GHJ, adhesion, and narrowing were observed.
in the synovial fluid of HD patients was similar to that in RA patients, and this finding suggests that synovitis may be a factor in fluid retention in the joint of HD patients. We performed arthroscopic debridement for
FIGURE 8. Postoperative arthroscopic findings in SAB: after debridement, there was a significant increase in space in the subacromial region.
ARTHROSCOPIC DEBRIDEMENT FOR DIALYSIS SHOULDERS TABLE 2. JOA Shoulder Score: The Good-Response Group Average Scores (range)
Pain Function ROM Radiographic Stability Total
Preoperative
Postoperative
Gain
5.6 (5-10) 16.8 (10.5-20) 24.3 (13-30) 4.7 15 66.4 (47.5-74)
23.4 (10-30) 18.3 (12-20) 25.2 (15-30) Unchanged Unchanged 86.6 (65-99)
17.8 (5-25) 1.5 (0-3) 0.9 (⫺2-5) — — 20.2 (5-29.5)
this shoulder pain because imaging and arthroscopic findings suggested that this pain was caused by synovitis. Surgery for dialysis shoulder was performed when a specific symptom such as severe pain at rest was not resolved with conservative treatment. Complete arthroscopic debridement alone in the GHJ and SAB for dialysis shoulder was effective in 94% of the patients, and most patients were satisfied with the results. The use of local anesthetic injection as a predictor of postoperative outcome was not considered clinically helpful because the effects differed among the patients. In patients with recurrence after cutting of the transverse carpal ligament for carpal tunnel syndrome, recompression by regenerated ligament-like scar tissue is a problem.12-15 Similarly, regeneration of ligament-like tissue was also reported in patients with recurrence after cutting of the coracoacromial ligament.5,16 We considered these to be the same changes, and the recurrence rate after cutting of the coracoacromial ligament appears to be high. An arthroscopic procedure is less invasive than open surgery and is advantageous in HD patients who are susceptible to bacterial infection and whose wounds heal slowly. Our operative method is minimally invasive and does not destroy the anatomic structure and, therefore, can be readily performed
FIGURE 9. The results of the good-response group: the improvement achieved in the category of pain, function, and ROM were statistically significant.
691
TABLE 3. Comparison Between the Good-Response Group and No-Response Group
Patients Age Duration of HD Age at occurrence of pain Shunt side (except bilateral cases) Amyloid deposition
Good-Response
No-Response
27 pts, 34 shoulders 56.9 yr 16.1 yr
2 pts, 2 shoulders 54.5 yr 16 yr
12.7 yr
13.5 yr
12/19 cases (63.2%) Positive
1/2 cases (50%) Positive
repeatedly. On the other hand, arthrotomy is highly invasive and requires postoperative rehabilitation. Arthroscopic findings and histologic examination of the synovium would indicate that amyloid deposition in the synovial tissue may be the cause of pain. Yet, if this is the case, it is unclear as to why 2 of our patients who had similar findings did not respond to arthroscopic debridement. It appears that the cause of pain generated in long-term HD has not been clarified. With that thought in mind, this underscores the advantage of using a minimally invasive approach such as arthroscopy to treat the problem. Using arthroscopic debridement, there was a significant improvement in pain relief in our patients. SUMMARY (1) We propose a new definition of a subcategory of dialysis arthropathy of the shoulder, dialysis shoulder, for characteristic shoulder pain during HD or at night in patients on long-term HD therapy. (2) Arthroscopic examination of patients with dialysis shoulder showed characteristic synovial proliferation with the appearance of feather-like or ice-covered twigs. (3) Histopathologic examination revealed amyloid deposition in and on the surface of the synovial membrane and on the surface of the articular cartilage. (4) Arthroscopic
692
K. MIDORIKAWA ET AL.
FIGURE 10. With Congo red staining, homogenous amorphous amyloid deposition (arrow heads) lacking cell components is observed in the synovial membrane (original magnification ⫻200).
FIGURE 11. Under a polarizing microscope, yellow-green fluorescence of amyloid was noted (original magnification ⫻200).
ARTHROSCOPIC DEBRIDEMENT FOR DIALYSIS SHOULDERS debridement with complete synovectomy of the GHJ and SAB for dialysis shoulder yielded good results in 94% of shoulders.
8.
REFERENCES
9.
1. Midorikawa K, Takeshita M, Hara M, Kuroda K, Nakamura A, Shono E, Takagishi N. The short-term results of shoulder arthroscopic synovectomy in rheumatoid arthritis. Kyusyu RA 1989; 8:53-56 (English summary). 2. Midorikawa K, Shibata Y, Emoto G, Ogata K. Shoulder pain in patients on chronic hemodialysis—Histopathological evaluation. J Shoulder Elbow Surg 1997;6:244. 3. Midorikawa K, Hara M, Shibata Y, Emoto G, Ogata K. Arthroscopic surgery for dialysis arthropathy of the shoulder. Presented at the Combined Congress of the International Arthroscopy Association and the International Society of the Knee, Hong Kong, May, 1995. 4. Midorikawa K, Hara M, Shibata Y, Emoto G, Ogata K. Shortterm results of arthroscopic debridement for dialysis arthropathy of the shoulder. J Shoulder Elbow Surg 1995;4:S68 (suppl). 5. Midorikawa K, Shibata Y, Emoto G, Ogata K. The results of arthroscopic surgery for dialysis shoulders with severe pain. Orthop Trans 1997;21:287-288. 6. Konishiike T, Hashizume H, Nishida K, Inoue H, Nagoshi M. Shoulder pain in long-term hemodialysis patients. A clinical study of 166 patients. J Bone Joint Surg Br 1996;78:601-605. 7. Okutsu I, Hamanaka I, Ninomiya S, Tamai K, Takatori Y, Abe
10. 11. 12. 13.
14. 15. 16.
693
I. Shoulder impingement syndrome in hemodialysis patients— Endoscopic subcutaneous coracoacromial ligament release. Kansetsukyou 1989;14:55-62 (English summary). Shimizu T, Kurokawa M, Hirasawa T. Treatment of dialysis arthropathy in shoulder. Katakansetsu 1994;18:380-383 (English summary). Shiota E, Arizono T, Kunisaki Y. Surgical treatment for dialysis-related arthropathy in the shoulder. Seikeigeka to Saigaigeka 1994;43:1540-1542 (English summary). Takenaka R, Fukatsu A, Matsuo S, Ishikawa K, Toriyama T, Kawahara H. Surgical treatment of hemodialysis-related shoulder arthropathy. Clin Nephrol 1992;38:224-230. Shiota E, Maekawa M, Ohtani M. Analysis of synovial fluid components of hydrarthrosis in long-term hemodialysis patients. J Orthop Sci 1999;4:171-175. Botte MJ, Von Schroeder HP, Abrams RA, Gellman H. Recurrent carpal tunnel syndrome. Hand Clin 1996;12:731-743. Forman DL, Watson HK, Caulfield KA, Shenko J, Caputo AE, Ashmead D. Persistent or recurrent carpal tunnel syndrome following prior endoscopic carpal tunnel release. J Hand Surg 1998;23:1010-1014. Hunter JM. Recurrent carpal tunnel syndrome, epineural fibrous fixation, and traction neuropathy. Hand Clin 1991;7: 491-504. Jenny-Hagen, K., Sennwald, G.: Recurrent carpal tunnel syndrome—Caused by an inadequate surgical technique? Helv Chir Acta 1990;57:125-128. Nagishi M, Hashizume H, Masaoka S, Inoue H, Konishiike T. Surgical treatment for shoulder arthropathy induced by dialysis-related amyloidosis. Katakansetsu 1990;23:451-454 (English summary).
Purpose: To arthroscopically treat “dialysis shoulder,” severe shoulder pain in patients on long-term hemodialysis. This pain occurs only at rest such as during hemodialysis or while sleeping and is temporarily alleviated by assuming the sitting position or moving the shoulder joint. Limitations in range of motion and pain with overhead activity or the arm in the forward flexed position are not the patient’s chief complaints. Although frozen shoulder and impingement syndrome may be observed as complications, pain at rest is the most characteristic. Type of Study: Clinical research on arthroscopic debridement to treat dialysis shoulder. Methods: We performed arthroscopic debridement of 36 dialysis shoulders in 29 patients. Only complete debridement in the glenohumeral joint and subacromial bursa was performed, without invasion to the bone and ligaments. Results: The pain improved in 34 shoulders in 27 patients (94%) a mean of 29.8 months after surgery, showing satisfactory results. In this group, the mean Japan Orthopaedic Surgery Association shoulder score (maximum 100 points) was 66.4 points before surgery but increased to 86.6 points postoperatively, statistically significant by 2-group t test. Conclusions: We decided before beginning the study that no postoperative rehabilitation would be necessary. There were no complications and no need for further surgery, with all but 2 of the patients being satisfied with their postoperative condition. Key Words: Amyloid—Arthroscopy—Arthroscopic surgery—Hemodialysis—Shoulder—Subacromial bursa.
T
he incidence of bone and joint disorders in hemodialysis (HD) patients increases with the duration of HD therapy. Shoulder pain is observed in many patients, some of whom complain of severe shoulder pain only at rest such as during HD or at night. The nature of the pain and its occurrence time are very characteristic and differ markedly from that observed in other disorders. However, most patients leave this pain untreated partly because the mechanism of the pain is still unclear. Although there are various theories on its mechanism, none of them has been confirmed. Some surgical methods have also been reported, but none of them are established. In our
From the Department of Orthopaedic Surgery, School of Medicine, Fukuoka University, Fukuoka, Japan. Address correspondence and reprint requests to Koji Midorikawa, M.D., 7-45-1 Nanakuma, Jyonan-Ku, Fukuoka 8140180, Japan. E-mail: [email protected] © 2001 by the Arthroscopy Association of North America 0749-8063/01/1707-2590$35.00/0 doi:10.1053/jars.2001.23584
department, good results have been obtained by arthroscopic debridement for chronic rheumatoid arthritis (RA),1 in which joint pain is caused primarily by synovitis. We speculated that shoulder pain in patients on long-term HD is also caused by synovitis on the basis of findings of imaging studies and arthroscopic findings, and have applied arthroscopic debridement to this condition. We defined “dialysis shoulder” as severe shoulder pain in patients on long-term HD that occurs only at rest such as during HD or sleep at night and is temporarily alleviated by assuming the sitting position or moving the shoulder joint. Limitations in the range of motion (ROM) and pain with overhead activity or with the arm in the forward flexed position are not the patient’s chief complaints. Although impingement and capsulitis may occur in this condition, the chief complaint is pain at rest, and we selected our subjects using this as the primary criterion.2 In this report, patients with severe shoulder pain who underwent arthroscopic debridement are presented. Arthroscopic debridement was
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 17, No 7 (September), 2001: pp 685– 693
685
686
K. MIDORIKAWA ET AL.
performed in patients with the above-defined dialysis shoulder who complained of sleep disorders resulting from nocturnal shoulder pain or pain during HD. METHODS The study included 29 patients (36 shoulders) with dialysis shoulders. The patients in this sample group were between 3 and 56 months after surgery (average, 29.8 months), with follow-up care continuing in all cases indefinitely. There were 16 men (19 shoulders) and 13 women (17 shoulders) with a mean age of 56.8 years (range, 39 to 78 years). The mean HD history was 16.3 years (range, 6 to 25 years), and shoulder pain developed a mean of 12.8 years (range, 6 to 18 years) after initiation of HD. In 12 patients, the operated shoulder was the same side as the shunt. In 9 patients the operated shoulder was the opposite side to the shunt. The remaining 8 patients underwent bilateral operations. In these patients, improvement in pain was evaluated using the Japan Orthopaedic Surgery Association Shoulder (JOA) score. In the JOA score, 30 points are assigned to pain, 20 points to function, 30 points to the range of motion the joint (ROM), 5 points to radiographic findings, and 15 points to joint stability with a full mark of 100 points (Table 1). Operative Technique For the arthroscopic surgery, patients were placed in the lateral decubitus position and 2 to 3 kg of skin traction was applied to the arm in about 30° abduction of the shoulder. Because traction was carried out using a brace made of thick sponge, we could arthroscope the side with the shunt in place.3-5 Arthroscopic surgery consisted of synovectomy and complete debridement in the glenohumeral joint (GHJ) and the subacromial bursa (SAB). No procedures that change the anatomic architecture such as cutting of ligament or resecting of bone were performed. After the procedure, a sling was applied, but free use of the shoulder was allowed from the next day. In all cases, the synovial membrane biopsy specimens taken during arthroscopy were stained with Congo red and examined histopathologically by light microscopy and polarizing microscopy. Preoperative Evaluation Physical examination before surgery was measured using the JOA score, and the results were as follows:
Pain, average 5.6 points (range, 5-10 points); Function, average 17.0 points (range, 10.5-20 points); ROM, average 24.4 points (range, 13-30 points); Radiographic findings, average 4.7 points (range, 3-5 points); Joint stability, average 15 points. Plain radiographic examination, ultrasonography, arthrography of the GHJ, and magnetic resonance imaging (MRI) were all used (without exception) to observe the morphology in every painful shoulder. Plain radiographs showed small cyst-like shadows and moth-eaten images in the humeral head, but negligible arthritic changes (Fig 1). Ultrasonography showed a thickened rotator cuff, heterogeneous echoes in the rotator cuff, and fluid retention in the SAB (Fig 2). T2-weighted MRIs revealed scattered low-intensity areas in the rotator cuff, indicating degeneration, and high-intensity fluid retained in the GHJ and SAB as well as in the cyst in the humeral head (Fig 3). Arthrography showed enlargement of the capsule and irregularity of the inner wall (Fig 4). These findings were all common in varying degrees to all patients. In order to predict the outcome of surgery, a local anesthetic was injected into the GHJ and SAB before surgery and pain was evaluated. The analgesic effects differed among the patients. RESULTS The analgesic effects of the local anesthetic injection were varied and followed no common pattern. Typical arthroscopic findings were as follows. In the GHJ, synovial villi resembled ice-covered twigs that, on close examination, had no visible surface capillaries, a characteristic of this disease. In some areas, polyp-like proliferation of synovial villi and degeneration of articular cartilage were present. Avascular synovial proliferation was a common finding in all patients (Fig 5). Postoperative arthroscopy, following arthroscopic debridement, showed an intact subscapularis tendon and middle glenohumeral ligament in all patients (Fig 6). In the SAB, synovial proliferation similar to that in the GHJ, adhesion, and narrowing were observed to varying degrees in all patients (Fig 7). After debridement, there was a significant increase in space in the subacromial region (Fig 8). Success of the surgery focused on how successful we were at relieving the preoperative pain. Nocturnal pain and pain during HD rapidly disappeared after operation in 34 shoulders of the 27 patients. However, in the other 2 patients (2 shoulders) there was no improvement in pain, indicating a failure of the procedure. Therefore, these 2 patients were classified as
ARTHROSCOPIC DEBRIDEMENT FOR DIALYSIS SHOULDERS
687
TABLE 1. JOA Shoulder Score (total 100 points) Pain (30 points) None Tenderness or minimal pain in sports or heavy labor Minimal pain in ADL Moderate and tolerable pain (analgesic needed, occasional night pain) Severe pain (ADL limited, frequent night pain) Totally incapacitated because of pain Function (20 points) Total function (10 points) Strength in abduction (5 points, measured at 90° abduction or at possible level) Normal Excellent Good Fair Poor Zero Endurance (5 points, time in seconds holding 1-kg dumbbell horizontally with elbow extended and forearm pronated) More than 10 seconds More than 3 seconds More than 2 seconds Zero Activities of daily living (10 points) Making knot in back 1, 0.5, 0 Reaching opposite axilla 1, 0.5, 0 Open and close sliding door 1, 0.5, 0 Wearing jacket 1, 0.5, 0 Sleep on involved side 1, 0.5, 0 Self-hygienic care 1, 0.5, 0 Reaching side pocket (jacket) 1, 0.5, 0 Combing hair 1, 0.5, 0 Reaching overhead shelf 1, 0.5, 0 Reaching mouth 1, 0.5, 0 Subtract one point from above for each unable activity, specify: 1. 2. 3. Range of motion (30 points) Elevation (15 points) ⬎150° ⬎120° ⬎90° ⬎60° ⬎30° 0° External rotation (9 points) ⬎60° ⬎30° ⬎0° ⬎⫺20° ⬍⫺20° Internal rotation (6 points) Above T12 spinous process Above L5 spinous process Gluteal Below gluteal Radiographic evaluation (5 points) Normal Moderate changes of subluxation Advanced changes or dislocation Joint stability (15 points) Normal Slight instability or apprehension Severe instability or history or state of subluxation Relevant history or state of dislocation Remarks: Record range of motion and pain of elbow and hand disabilities, if present.
30 25 20 10 5 0
5 4 3 2 1 0 5 3 1 0
15 12 9 6 3 0 9 6 3 2 0 6 4 2 0 5 3 0 15 10 5 0
688
K. MIDORIKAWA ET AL. (mean, 20.2 points). With alleviation of pain, the score in function and ROM increased slightly (Table 2). The improvement achieved in the category of pain (5.6/30 to 23.4/30 points) is statistically significant by 2-group t test (P ⬍ .0001), underlining the benefit of the operative technique. Improvement in function (16.8/20 to 18.3/20 points) and ROM (24.3/30 to 25.2/30 points) were also statistically significant by 2-group t test (Fig 9). Comparison of Results Comparison between the no-response and goodresponse groups showed no definite differences in regard to age, duration of HD, the time of onset of shoulder pain, the relationship between the shunt side and pain side, nor the preoperative JOA score (Table 3). Comparison of Arthroscopic Findings The arthroscopic findings in all cases were characterized by synovial villi resembling ice-covered twigs with no visible vascularization. Comparison of Histopathologic Findings Histopathologic examination of the synovial tissue resected at arthroscopy showed amyloid deposition in all cases. The specimen shown in Fig 10, stained with Congo red, shows homogenous amorphous amyloid deposition lacking cell components in the synovial membrane. Under a polarizing microscope, yellowgreen fluorescence of amyloid was shown (Fig 11).
FIGURE 1. Plain radiographic examination showed small cyst-like shadows and moth-eaten images in the humeral head but negligible arthritic changes: (A) anteroposterior view (B) axial view.
the no-response group and compared with the goodresponse group (27 patients, 34 shoulders). Results in the Good-Response Group In the good-response group, the mean JOA score was 66.4 points before operation and increased to 86.6 points after operation. The increase in the score of pain (17.8 points) accounted for most of the increase
FIGURE 2. Ultrasonography showed a thickened rotator cuff (large arrows), heterogeneous echoes in the rotator cuff, and fluid retention in the SAB (small arrows).
ARTHROSCOPIC DEBRIDEMENT FOR DIALYSIS SHOULDERS
689
FIGURE 3. T2-weighted MRIs revealed scattered low-intensity areas in the rotator cuff, indicating degeneration, and high-intensity fluid retained in the GHJ and SAB as well as in the cyst in the humeral head: (A) coronal view (B) axial view.
DISCUSSION There have been some recent studies on surgery for shoulder pain in patients on long-term HD.5-10 The underlying disease varied among the reports. Some investigators6,7 use patients with impingement syndrome or stiff shoulder as the subjects, whereas our subjects did not present with such pain on motion or ROM, and so comparison of the 2 would be redundant.
FIGURE 4. Arthrography showed enlargement of the capsule and irregularity of the inner wall: (A) internal rotation position (B) external rotation position.
690
K. MIDORIKAWA ET AL.
FIGURE 5. Preoperative arthroscopic findings in the GHJ: avascular ice-covered twig-like synovial villi, characteristic of this disease, were observed. In some areas, polyp-like proliferation of synovial villi was also seen.
At our department, arthroscopic debridement is performed for chronic RA in which the primary cause of joint pain is synovitis.3 Shiota et al.11 reported that the ratio of chondroitin 6-sulfate to chondroitin 4-sulfate
FIGURE 6. Postoperative arthroscopic findings in GHJ: intact subscapular muscle (single arrow head) and middle glenohumeral ligament (double arrow heads) are shown.
FIGURE 7. Preoperative arthroscopic findings in SAB: synovial proliferation similar to that in the GHJ, adhesion, and narrowing were observed.
in the synovial fluid of HD patients was similar to that in RA patients, and this finding suggests that synovitis may be a factor in fluid retention in the joint of HD patients. We performed arthroscopic debridement for
FIGURE 8. Postoperative arthroscopic findings in SAB: after debridement, there was a significant increase in space in the subacromial region.
ARTHROSCOPIC DEBRIDEMENT FOR DIALYSIS SHOULDERS TABLE 2. JOA Shoulder Score: The Good-Response Group Average Scores (range)
Pain Function ROM Radiographic Stability Total
Preoperative
Postoperative
Gain
5.6 (5-10) 16.8 (10.5-20) 24.3 (13-30) 4.7 15 66.4 (47.5-74)
23.4 (10-30) 18.3 (12-20) 25.2 (15-30) Unchanged Unchanged 86.6 (65-99)
17.8 (5-25) 1.5 (0-3) 0.9 (⫺2-5) — — 20.2 (5-29.5)
this shoulder pain because imaging and arthroscopic findings suggested that this pain was caused by synovitis. Surgery for dialysis shoulder was performed when a specific symptom such as severe pain at rest was not resolved with conservative treatment. Complete arthroscopic debridement alone in the GHJ and SAB for dialysis shoulder was effective in 94% of the patients, and most patients were satisfied with the results. The use of local anesthetic injection as a predictor of postoperative outcome was not considered clinically helpful because the effects differed among the patients. In patients with recurrence after cutting of the transverse carpal ligament for carpal tunnel syndrome, recompression by regenerated ligament-like scar tissue is a problem.12-15 Similarly, regeneration of ligament-like tissue was also reported in patients with recurrence after cutting of the coracoacromial ligament.5,16 We considered these to be the same changes, and the recurrence rate after cutting of the coracoacromial ligament appears to be high. An arthroscopic procedure is less invasive than open surgery and is advantageous in HD patients who are susceptible to bacterial infection and whose wounds heal slowly. Our operative method is minimally invasive and does not destroy the anatomic structure and, therefore, can be readily performed
FIGURE 9. The results of the good-response group: the improvement achieved in the category of pain, function, and ROM were statistically significant.
691
TABLE 3. Comparison Between the Good-Response Group and No-Response Group
Patients Age Duration of HD Age at occurrence of pain Shunt side (except bilateral cases) Amyloid deposition
Good-Response
No-Response
27 pts, 34 shoulders 56.9 yr 16.1 yr
2 pts, 2 shoulders 54.5 yr 16 yr
12.7 yr
13.5 yr
12/19 cases (63.2%) Positive
1/2 cases (50%) Positive
repeatedly. On the other hand, arthrotomy is highly invasive and requires postoperative rehabilitation. Arthroscopic findings and histologic examination of the synovium would indicate that amyloid deposition in the synovial tissue may be the cause of pain. Yet, if this is the case, it is unclear as to why 2 of our patients who had similar findings did not respond to arthroscopic debridement. It appears that the cause of pain generated in long-term HD has not been clarified. With that thought in mind, this underscores the advantage of using a minimally invasive approach such as arthroscopy to treat the problem. Using arthroscopic debridement, there was a significant improvement in pain relief in our patients. SUMMARY (1) We propose a new definition of a subcategory of dialysis arthropathy of the shoulder, dialysis shoulder, for characteristic shoulder pain during HD or at night in patients on long-term HD therapy. (2) Arthroscopic examination of patients with dialysis shoulder showed characteristic synovial proliferation with the appearance of feather-like or ice-covered twigs. (3) Histopathologic examination revealed amyloid deposition in and on the surface of the synovial membrane and on the surface of the articular cartilage. (4) Arthroscopic
692
K. MIDORIKAWA ET AL.
FIGURE 10. With Congo red staining, homogenous amorphous amyloid deposition (arrow heads) lacking cell components is observed in the synovial membrane (original magnification ⫻200).
FIGURE 11. Under a polarizing microscope, yellow-green fluorescence of amyloid was noted (original magnification ⫻200).
ARTHROSCOPIC DEBRIDEMENT FOR DIALYSIS SHOULDERS debridement with complete synovectomy of the GHJ and SAB for dialysis shoulder yielded good results in 94% of shoulders.
8.
REFERENCES
9.
1. Midorikawa K, Takeshita M, Hara M, Kuroda K, Nakamura A, Shono E, Takagishi N. The short-term results of shoulder arthroscopic synovectomy in rheumatoid arthritis. Kyusyu RA 1989; 8:53-56 (English summary). 2. Midorikawa K, Shibata Y, Emoto G, Ogata K. Shoulder pain in patients on chronic hemodialysis—Histopathological evaluation. J Shoulder Elbow Surg 1997;6:244. 3. Midorikawa K, Hara M, Shibata Y, Emoto G, Ogata K. Arthroscopic surgery for dialysis arthropathy of the shoulder. Presented at the Combined Congress of the International Arthroscopy Association and the International Society of the Knee, Hong Kong, May, 1995. 4. Midorikawa K, Hara M, Shibata Y, Emoto G, Ogata K. Shortterm results of arthroscopic debridement for dialysis arthropathy of the shoulder. J Shoulder Elbow Surg 1995;4:S68 (suppl). 5. Midorikawa K, Shibata Y, Emoto G, Ogata K. The results of arthroscopic surgery for dialysis shoulders with severe pain. Orthop Trans 1997;21:287-288. 6. Konishiike T, Hashizume H, Nishida K, Inoue H, Nagoshi M. Shoulder pain in long-term hemodialysis patients. A clinical study of 166 patients. J Bone Joint Surg Br 1996;78:601-605. 7. Okutsu I, Hamanaka I, Ninomiya S, Tamai K, Takatori Y, Abe
10. 11. 12. 13.
14. 15. 16.
693
I. Shoulder impingement syndrome in hemodialysis patients— Endoscopic subcutaneous coracoacromial ligament release. Kansetsukyou 1989;14:55-62 (English summary). Shimizu T, Kurokawa M, Hirasawa T. Treatment of dialysis arthropathy in shoulder. Katakansetsu 1994;18:380-383 (English summary). Shiota E, Arizono T, Kunisaki Y. Surgical treatment for dialysis-related arthropathy in the shoulder. Seikeigeka to Saigaigeka 1994;43:1540-1542 (English summary). Takenaka R, Fukatsu A, Matsuo S, Ishikawa K, Toriyama T, Kawahara H. Surgical treatment of hemodialysis-related shoulder arthropathy. Clin Nephrol 1992;38:224-230. Shiota E, Maekawa M, Ohtani M. Analysis of synovial fluid components of hydrarthrosis in long-term hemodialysis patients. J Orthop Sci 1999;4:171-175. Botte MJ, Von Schroeder HP, Abrams RA, Gellman H. Recurrent carpal tunnel syndrome. Hand Clin 1996;12:731-743. Forman DL, Watson HK, Caulfield KA, Shenko J, Caputo AE, Ashmead D. Persistent or recurrent carpal tunnel syndrome following prior endoscopic carpal tunnel release. J Hand Surg 1998;23:1010-1014. Hunter JM. Recurrent carpal tunnel syndrome, epineural fibrous fixation, and traction neuropathy. Hand Clin 1991;7: 491-504. Jenny-Hagen, K., Sennwald, G.: Recurrent carpal tunnel syndrome—Caused by an inadequate surgical technique? Helv Chir Acta 1990;57:125-128. Nagishi M, Hashizume H, Masaoka S, Inoue H, Konishiike T. Surgical treatment for shoulder arthropathy induced by dialysis-related amyloidosis. Katakansetsu 1990;23:451-454 (English summary).