Abstract
Background:
Pseudoparalysis is defined as active forward flexion less than 90° with full passive motion. There is controversy about the ideal surgical management of a massive rotator cuff tear with pseudoparalysis.
Purpose/Hypothesis:
The purpose of this study was to prospectively analyze the ability to reverse pseudoparalysis with an arthroscopic rotator cuff repair (ARCR). The hypothesis was that in the absence of substantial glenohumeral arthritis, preoperative fatty infiltration of grade 3 or higher and an acromiohumeral interval (AHI) of less than 7 mm would not prevent reversal of pseudoparalysis with an ARCR.
Study Design:
Case series; Level of evidence, 4.
Methods:
A prospective multicenter study of ARCR performed for preoperative pseudoparalysis was conducted. The minimum follow-up was 1 year. The mean patient age was 63 years, and pseudoparalysis was present for a mean of 4.2 months preoperatively. Preoperative radiographic evaluation included plain film evaluation of the AHI and Hamada classification and MRI evaluation of fatty degeneration and rotator cuff retraction. Functional outcome was determined by the Simple Shoulder Test (SST), American Shoulder and Elbow Surgeons (ASES) Shoulder Score, visual analog scale (VAS), and subjective shoulder value (SSV).
Results:
Of the 58 patients enrolled, 56 had at least 1 year of follow-up. Mean active forward flexion improved from 47° preoperatively to 159° postoperatively (P < .001). Statistically significant improvements were seen in the SST (from 2.8 preoperatively to 10.1 postoperatively), SSV (from 28 to 83), ASES Shoulder Score (from 37 to 88), and VAS (from 5.7 to 1.1) (P < .001). Pseudoparalysis was reversed in 53 of 56 patients (95%). There was no difference in the rate of reversal of pseudoparalysis between those patients with an AHI of less than 7 mm (88.2%) and those with an AHI of 7 mm or more (96.9%) (P =.289). Pseudoparalysis was reversed in all 8 of the patients with fatty degeneration of grade 3 or higher in 1 or more of the rotator cuff muscles.
Conclusion:
ARCR can lead to reversal of preoperative pseudoparalysis in patients with minimal preoperative glenohumeral arthritis. ARCR is a viable first line of treatment for patients with pseudoparalysis in the absence of advanced glenohumeral arthritis.
While many patients with a massive rotator cuff tear are able to maintain overhead motion through balanced force couples,2,4 a subset of patients are severely debilitated by pseudoparalysis of the shoulder, the inability to actively raise the affected arm above shoulder level.
Some authors have suggested that patients with pseudoparalysis secondary to massive rotator cuff tear require a reverse shoulder arthroplasty (RSA) to regain active motion above shoulder level.29,30 However, RSA has been associated with high complication rates, 29 and there are concerns about the long-term survivorship of the implant. 15
An arthroscopic approach to managing pseudoparalysis may be desirable if it can reliably restore function while avoiding the potential morbidity of a joint replacement (particularly RSA). Two recent studies have reported that a primary arthroscopic rotator cuff repair (ARCR) of a massive rotator cuff tear can reverse pseudoparalysis.6,24 These concluded that an ARCR should be the primary procedure of choice in many cases of pseudoparalysis, but both studies were retrospective and in each the surgeries were performed by a single surgeon.
The purpose of this study was to prospectively analyze the ability to reverse pseudoparalysis with an ARCR. We also sought to determine whether specific preoperative radiographic factors were associated with recovery of pseudoparalysis after an ARCR. The hypothesis was that preoperative fatty infiltration of grade 3 or higher and an acromiohumeral interval of less than 7 would not prevent reversal of pseudoparalysis with an ARCR.
Methods
Study Population
This was a prospective multicenter study performed from February 2012 through February 2013. Institutional review board approval was obtained before the study commenced. Inclusion criteria included a massive rotator cuff tear, preoperative pseudoparalysis, and an ARCR. Exclusion criteria included revision rotator cuff repair, passive restriction of forward flexion, preoperative neurologic injury, a shoulder arthroplasty, and preoperative anterior and/or posterior shoulder instability. The minimum follow-up was 1 year. A massive RCT was defined as 5 cm or larger in diameter at its greatest dimension 7 or a complete tear of 2 tendons. 10 Pseudoparalysis was defined as active forward flexion 90° or less with full passive forward flexion,22,29 and the inability to hold the arm at 90°.
Radiographic Analysis
All patients underwent a standard preoperative radiographic evaluation consisting of plain radiographs and magnetic resonance imaging (MRI). Images were reviewed by 2 musculoskeletal radiologists at a single institution.
Plain radiographic evaluation included the acromiohumeral interval (AHI) and Hamada classification on an anterior-posterior radiograph of the glenohumeral joint. The AHI was measured as the smallest distance from the inferior surface of the acromion to the superior aspect of the humerus. Measurements were also classified as <7 mm or ≥7 mm based on previous studies.19,23 The radiographs were then graded according to the Hamada classification.16,28 In the Hamada classification, the AHI is maintained in grade 1. In grade 2, the AHI is narrowed. Grade 3 includes acetabularization (concave deformity of the acromion undersurface) in addition to the grade 2 narrowing. Grade 4 includes narrowing of the glenohumeral joint. Grade 5 involves humeral head collapse.
MRI evaluation included the degree of fatty degeneration and the tangent sign. Fatty degeneration was graded according to Goutallier et al 14 (as adapted by Fuchs et al 9 ). The degree of preoperative fatty degeneration by itself was not considered a contraindication to repair. 3 The tangent sign was graded as negative or positive based on whether the supraspinatus muscle crossed or did not cross the tangent line. The tangent line extended from the tip of the coracoid to the superior aspect of the scapular spine on the lateral-most T1 sagittal MRI image in which the scapular spine and scapular body were in contact (Figure 1). 26 A positive tangent sign has previously been associated with reparability of rotator cuff tears. 18
Sagittal magnetic resonance imaging demonstrates a positive tangent sign, represented by the line extending from the coracoid to the scapular spine. This patient had a complete repair, and his pseudoparalysis was reversed.
Surgical Technique
A total of 8 surgeons at 8 centers performed the ARCRs in this study. The indication for surgery was pseudoparalysis in the absence of medical contraindication. Nonoperative treatment was not required since previous study has indicated that longer duration of symptoms is related to a lower likelihood of recovery from pseudoparalysis after an ARCR. 6 The choice of surgery type (ARCR vs arthroplasty) was left to the discretion of each individual surgeon. All surgeons completed the same advanced shoulder arthroscopy fellowship such that the technique was relatively standardized. Tear dimensions were measured with a calibrated probe at the time of surgery after identification of the tear margins. Tendon involvement as a percentage of the anterior to posterior width of each tendon was quantified after the scapular spine was delineated.
The biceps tendon, if present, underwent tenotomy or tenodesis. The subscapularis was then evaluated and repaired if the tear was 30% or more of the upper subscapularis tendon. In the case of retracted subscapularis tears, a 3-sided arthroscopic release was performed, preserving the comma tissue, which connects the upper subscapularis to the anterior supraspinatus tendon. After repair of the subscapularis tendon, the posterosuperior rotator cuff was mobilized and repaired. Repairs were performed with double-loaded threaded suture anchors (eg, 5.5-mm BioComposite Corkscrew; Arthrex Inc). A suture-bridging double-row repair was performed with medial mattress sutures medially and knotless anchors laterally (eg, BioComposite SwiveLock C; Arthrex Inc) when there was sufficient tendon mobility and tendon length. Single-row repairs were performed by use of simple sutures with a 6-throw surgeon’s knot. Advanced mobilization techniques were performed when there was otherwise insufficient tendon mobility to perform a tendon to bone repair. These included an isolated anterior interval slide in continuity, an isolated posterior interval slide, and a double interval slide.20,21 A complete repair was defined as complete tendon coverage of the native rotator cuff footprint in the anterior to posterior dimension. Acromioplasty and distal clavicle excision were left to surgeon discretion.
Postoperatively, patients were immobilized in a sling for 6 weeks without dedicated physical therapy. At 6 weeks postoperatively, the sling was discontinued, and passive forward flexion and passive external rotation were allowed. At 3 to 4 months postoperatively, active forward flexion and passive internal rotation were allowed, and strengthening was initiated. Return to full activity was allowed at 6 to 12 months, including all sports activities without restriction.
Functional Outcome
A standardized preoperative evaluation was performed for each patient followed by postoperative assessments at 3, 6, and 12 months after surgery.
The physical examination at each visit included range of motion and manual strength assessment. Passive and active forward flexion and external rotation with the arm at the side were measured with a goniometer. Active internal rotation was estimated to the nearest spinal level. External rotation strength and abduction strength were graded manually from 1 to 5 according to the method of the Medical Research Council. Recovery from pseudoparalysis after ARCR was defined as restoration of active forward flexion greater than 90°.
At each visit, functional status was determined by the Simple Shoulder Test (SST) 13 and the American Shoulder and Elbow Surgeons (ASES) Shoulder Score. 25 Additional data collected included the subjective shoulder value (SSV) and pain graded as 0 to 10 on a visual analog scale (VAS).11,31 At the 12-month follow-up, postoperative patient satisfaction (yes or no), return to normal sports or work activity (yes or no), and any additional surgery or postoperative complications were recorded.
Statistical Analysis
All statistical analysis was conducted by a trained statistician using SAS 9.2 (SAS Institute). Continuous data were described by means and standard deviations. A t test or Wilcoxon rank sum was performed (depending on variable distribution) to analyze the difference in pre- and postoutcome scores for range of motion, strength, pain, SST score, ASES Shoulder Score, and SSV score. Two-tailed P values of less than .05 were considered significant.
Results
Study Population
A total of 58 patients met the study criteria during the study period. No patients declined to participate in the study. Complete functional outcome data were available for 56 (96.6%) patients at a mean of 13.6 months postoperatively. Preoperative radiographs were available for 49 (87.5%) and a preoperative MRI scan was available for 51 (91.1%) of these 56 patients.
The mean age (±SD) was 62.6 ± 9.0 years at the time of surgery. Pseudoparalysis had been present for a mean of 3.9 ± 5.6 months before surgery and was traumatic in origin in 45 (80.4%) patients (Table 1).
Operative Characteristics of Rotator Cuff Tears and Repair Technique
Data correspond to Goutallier grade 0, 1, 2, 3, and 4, respectively. Data were available for 51 patients.
The AHI averaged 8.1 mm before surgery. The interval was less than 7 mm in 34.7% of patients. According to the Hamada classification, the shoulders were grade 1 in 75.5%, grade 2 in 20.4%, grade 3 in 2.0%, and grade 4 in 2.0%. The tangent sign was positive in 57.4% of patients. The majority of patients had grade 1 or 2 fatty degeneration of the rotator cuff muscles. Eight of the 51 (15.7%) patients with an MRI had fatty degeneration of grade 3 or greater in 1 or more rotator cuff muscles.
Tear Pattern and Surgical Technique
The mean tear size was 4.9 cm from medial to lateral and 4.6 cm from anterior to posterior. All patients had a 100% tear of the supraspinatus tendon. The infraspinatus was also torn in 51 (91.1%) patients and averaged 78% ± 30% of the anterior to posterior length of the tendon insertion. The subscapularis was torn in 40 (71.4%) patients and averaged 65% ± 28% of the tendon insertion. The teres minor was torn in 6 (10.3%) patients.
Advanced mobilization techniques were required in 23 (41.1%) patients, including 13 isolated anterior interval slides, 2 isolated posterior interval slides, and 8 double interval slides. A complete repair was achieved in 91.1% of patients, including 92.6% (25/27) of patients with a positive tangent sign. A mean of 3.6 ± 1.1 anchors were used to repair the posterosuperior rotator cuff and 1.2 ± 1.0 anchors were used to repair the subscapularis. A double-row repair was used to repair the posterosuperior rotator cuff in 42 patients (75%) and the subscapularis tendon in 11 patients (19.6%). Associated procedures included an acromioplasty in 33 (58.9%) patients, distal clavicle excision in 2 (3.6%) patients, biceps tenodesis in 22 (39.3%) patients, and a biceps tenotomy in 27 (48.2%) patients.
Functional Outcome
Pseudoparalysis was reversed in 53 patients (94.6%) at final follow-up. The majority of patients recovered within 6 months, with a recovery rate of 87.3% (48/55) at that time point. The 2 patients who were lost to final follow-up had reversal of pseudoparalysis at 6 months postoperatively. There was no difference in the rate of reversal of pseudoparalysis between those patients with an AHI of less than 7 mm (88.2%) and those with an AHI of 7 mm or more (96.9%) (P = .289). Pseudoparalysis was reversed in the 8 patients with fatty degeneration of grade 3 or more in 1 or more of their rotator cuff muscles. The 3 patients who did not have reversal of pseudoparalysis had fatty degeneration of grade 2 or lower. All 5 patients (8.9%) who had a partial repair had reversal of pseudoparalysis. Fifty-three patients (94.6%) were satisfied with the surgery and 52 (92.9%) returned to previous work or activity. Two of the 3 patients who did not recover active forward flexion were satisfied with the procedure. Postoperative pain scores in these 3 patients were 0, 1, and 0, respectively. The functional outcome of the cohort is summarized in Table 2.
Results of Arthroscopic Rotator Cuff Repair for Pseudoparalysis a
Results are reported as mean ± SD unless otherwise indicated. NA, not applicable.
Complications
Two patients (3.6%) had a postoperative complication, which entailed 1 revision repair and 1 ulnar nerve lesion. The patient requiring revision repair fell 6 weeks after the initial surgery, retore his rotator cuff, underwent revision repair, and regained active forward flexion. The ulnar nerve lesion resolved nonoperatively. There were no instances of postoperative deep wound infection.
Discussion
Pseudoparalysis of the shoulder is severely disabling and is considered by some to require RSA. The current study, however, demonstrates that an ARCR can lead to predictable reversal of pseudoparalysis in appropriately selected patients. Our findings support the hypothesis that pseudoparalysis can be reversed with an ARCR even in the setting of an AHI less than 7 mm or grade 3 fatty degeneration of the rotator cuff. Pseudoparalysis was reversed with an ARCR in 94.6% of patients, with a rate of 100% for the patients with grade 3 or higher fatty degeneration and 88.2% for patients with an AHI of less than 7 mm. Our findings add to the growing evidence that ARCR should be the first-line treatment for pseudoparalysis in the absence of substantial glenohumeral arthritis.
To our knowledge, no studies of open rotator cuff repair have focused exclusively on reversal of pseudoparalysis. In 1986, Ellman et al 8 reported that preoperative abduction less than 100° was associated with a 9 times greater risk for an unsatisfactory result after open rotator cuff repair. Cofield et al 5 also reported that postoperative range of motion after an open rotator cuff repair was related to preoperative range of motion; only 27% of the massive rotator cuff tears in their study had an excellent or satisfactory result with an open repair. The investigators concluded, “It is clear that these standard techniques have deficiencies when used for the repair of massive tears, and introduction of newer, experimental repair methods for improving the functional outcome in this group may be justified.”8(p76)
Denard et al 6 retrospectively reported on 39 patients who underwent a primary ARCR for pseudoparalysis. Their 90% rate of reversal was comparable with that seen in the current study. While Denard et al did not specifically exclude patients based on fatty degeneration or a decreased AHI, the investigators did not perform a radiographic analysis. Oh et al 24 retrospectively reported a 75.9% rate of reversal of pseudoparalysis with an open repair or ARCR in 35 patients. The investigators excluded patients with “severe glenohumeral arthritis,” but many patients had advanced fatty degeneration. Preoperatively there was grade 3 or higher fatty degeneration in the supraspinatus in 88.6% of patients, in the infraspinatus in 41.9%, and in the subscapularis in 31.4%. While our rate of advanced fatty degeneration was lower (16%), together our findings and those of Oh et al show that fatty degeneration grade 3 or higher does not preclude reversal of pseudoparalysis with an ARCR. Although it has been shown that such fatty degeneration lowers healing rates after rotator cuff repair,12,14 this does not appear to prevent an improvement in function. In fact, in the study by Oh et al, complete rotator cuff healing was only 33%. Perhaps sufficient partial healing occurs in the majority of cases to rebalance force couples and provide a stable fulcrum for overhead range of motion. 1
One concern about ARCR for pseudoparalysis is the technical complexity of the repair and therefore whether the results are reproducible. In the studies by Denard et al 6 and Oh et al, 24 repairs were performed by a single surgeon (distinct from one another). In the current study, repairs were performed by 8 different surgeons. Our results therefore demonstrate that the ability to reverse pseudoparalysis with an ARCR is not limited to the abilities of a few surgeons. Rather, our results suggest that with proper training, many surgeons can attain reversal of pseudoparalysis.
We achieved a complete repair in 91% of patients, yet many of these tears would have been considered irreparable by other authors. Kissenberth et al 18 reported inability to achieve a complete repair in 14 of 17 (82%) patients with a positive tangent sign of the supraspinatus. However, the investigators specifically noted that they did not repair the subscapularis tendon, and they did not describe any attempt at interval slides. In contrast, we repaired the subscapularis in 71% of patients and performed interval slides in 41%. These techniques may account for our high rate of complete repairs despite a positive tangent sign in 57%. Regardless, a complete repair may not be necessary to reverse pseudoparalysis. Iagulli et al 17 previously demonstrated that partial repairs and complete repairs of a massive rotator cuff tear achieved similar postoperative functional outcomes. In our series, all 5 patients who had a partial repair had reversal of pseudoparalysis.
Werner et al 29 reported on 17 patients with pseudoparalysis and an “irreparable” rotator cuff tear who were treated with RSA before any attempt at rotator cuff repair. The investigators defined “irreparable” as chronic pseudoparalysis, an AHI of less than 7 mm, or grade 3 or higher fatty degeneration of the supraspinatus or infraspinatus. Forward flexion improved from 43° preoperatively to 103° postoperatively. The complication rate was 47%, and 18% of patients required a reoperation. While none of their contraindications to repair were considered as such in our study, Werner et al did not characterize the degree of glenohumeral arthritis so it is possible that several of their patients also had advanced rotator cuff arthropathy. In contrast, Mulieri et al 22 specifically reported the results of primary RSA for 34 patients with an “irreparable” rotator cuff tear without glenohumeral arthritis. These investigators excluded patients with substantial glenohumeral arthritis (Hamada grade ≥4), so their cohort appears similar to ours radiographically. The investigators considered pseudoparalysis (present in 33/34 patients) alone as one of their criteria for “irreparable.” They noted preoperative to postoperative improvements in ASES Shoulder Scores (from 33.6 to 77.5), SST scores (from 1.4 to 6.3), pain (from 6.6 to 1.7), and forward flexion (from 54° to 136°). While these represented functional improvements, none of the postoperative values exceeded our results despite similar preoperative status (ASES, from 36.8 preoperatively to 87.5 postoperatively; SST, from 2.8 to 10.1; pain, from 5.7 to 1.1; and forward flexion, from 47° to 159°).
Considering these comparative results, we believe that ARCR should be the first-line treatment for many patients with pseudoparalysis. While we acknowledge that RSA is likely more predictable in patients with advanced glenohumeral arthritis or those with anterosuperior escape (which were excluded from this study), our results as well as those of others show that regardless of a decreased AHI or fatty degeneration, an ARCR can be highly successful in reversing pseudoparalysis if there is minimal preoperative glenohumeral arthritis. This is particularly the case in the setting of acute tears. The majority of our patients had a short duration of pseudoparalysis and traumatic origin of injury, which likely affected our high rate of success. While technically demanding, this approach is associated with a much lower complication rate than RSA and avoids the concerns about implant longevity. Finally, as previously demonstrated, ARCR does not compromise the ultimate outcome of RSA in the event that reversal is not obtained. 27
The major strengths of the current study are the prospective design, cohort size, and high percentage of follow-up. The major limitations are the short-term follow-up and lack of a comparative group or identification of factors associated with success or failure. Several concomitant procedures were performed at the time of ARCR, so we cannot determine how these procedures (eg, distal clavicle excision or biceps tenodesis) influenced our results compared with the rotator cuff repair itself. We did not compare the results of patients with an ARCR to the results of those who underwent RSA during the study period, nor did we evaluate the efficacy of nonoperative management. The choice to perform an ARCR was based on surgeon preference, so there is potential for selection bias. Similarly, because only 3 patients did not have reversal of pseudoparalysis, we could not analyze factors that are associated with failure of ARCR for pseudoparalysis. We did not assess healing, so we cannot comment on the anatomic success of our repairs. Our goal was simply to assess functional outcome, as we believe that this is the most important factor in whether surgery for pseudoparalysis can be considered a success. In addition, as noted previously, complete healing is not necessary for reversal of pseudoparalysis. 24 Finally, most of our patients had pseudoparalysis for a short duration (mean, 3.9 months) and had a traumatic origin. Previous study has suggested that a longer duration of pseudoparalysis is a negative prognostic factor for recovery. 6
Conclusion
ARCR can lead to reversal of preoperative pseudoparalysis in patients with minimal preoperative glenohumeral arthritis. ARCR is a viable first line of treatment for patients with pseudoparalysis in the absence of advanced glenohumeral arthritis.
Footnotes
One or more of the authors has declared the following potential conflict of interest or source of funding: P.J.D. and P.N. are consultants for Arthrex Inc. P.C.B. and S.S.B. are consultants for and receive royalties from Arthrex Inc. C.R.A. is an employee of Arthrex Inc as medical director of education; he was not employed by Arthrex at the time of the study. Research funding for costs related to IRB approval were received from Arthrex Inc.