- Email: [email protected]
Arthroscopic treatment of internal rotation contracture and glenohumeral dysplasia in children with brachial plexus birth palsy
J Shoulder Elbow Surg (2010) 19, 102-110
www.elsevier.com/locate/ymse
Arthroscopic treatment of internal rotation contracture and glenohumeral dysplasia in children with brachial plexus birth palsy Scott H. Kozin, MDa,*, Matthew J. Boardman, DOb, Ross S. Chafetz, PT, DPT, MPHc, Gerald R. Williams, MDd, Alexandra Hanlon, PhDe a
Department of Orthopaedic Surgery, Temple University & Hand Surgeon, Shriners Hospital for Children, Philadelphia, PA b Department of Orthopaedic Surgery, Philadelphia College of Osteopathic Medicine, Philadelphia, PA c Shriners Hospital for Children, Philadelphia, PA d Thomas Jefferson University, Philadelphia, PA e Department of Public Health, Temple University, Philadelphia, PA Background: The purpose of the study was to assess the ability of arthroscopic anterior release, þ/- tendon transfers to maintain shoulder joint alignment in children with brachial plexus palsy, and to assess their outcome after arthroscopic reduction. Methods: Forty-four patients underwent arthroscopic release, þ/- tendon transfers to realign a dysplastic glenohumeral joint in children with brachial plexus palsy. Twenty-eight children underwent isolated release and 16 children underwent concomitant tendon transfers. MRI and clinical measurements were used to assess outcome at 1-year follow-up. Results: There was a significant improvement (P < .001) in both retroversion from 34 (15) to -19 (13), and percentage of the humeral head anterior to the middle of the glenoid fossa (PHHA) from 19% (12%) to 33% (12%), at 1 year. Passive external rotation increased from 26 (20) degrees to 47 (17) degrees (P < .001). Active elevation increased from 112 (28) degrees to 130 (38) (P ¼ .008) degrees. Patients that underwent tendon transfers obtained greater active elevation, 147 (9) degrees compared to 119 (6) degrees. Mallet aggregate and domain scores also demonstrated statistically significant improvements. Conclusions: Our results after arthroscopic release þ/- tendon transfers are encouraging with improvements in joint alignment and clinical evaluations following surgery. The clinical improvements paralleled the MRI corrections. Importantly, superior outcomes were associated with better preoperative clinical and MRI status. This indicates that early recognition of glenohumeral dysplasia and timely intervention results in better shoulder motion and improved joint alignment. Level of Evidence: 4. Ó 2010 Journal of Shoulder and Elbow Surgery Board of Trustees.
Investigation performed at Shriners Hospital for Children, Philadelphia, Pennsylvania. This retrospective chart review study was conducted under exempt status Temple University IRB #10950. *Reprint requests: Scott H. Kozin, MD, Shriners Hospital for Children, 3551 North Broad Street, Philadelphia, PA 19140. E-mail address: [email protected] (S.H. Kozin).
Incomplete recovery after brachial plexus birth palsy often results in decreased movement and muscle imbalance about the shoulder, as rotator cuff and deltoid innervation is incomplete. The internal rotators overpower the external
1058-2746/2010/$36.00 - see front matter Ó 2010 Journal of Shoulder and Elbow Surgery Board of Trustees. doi:10.1016/j.jse.2009.05.011
Shoulder arthroscopy in children with brachial plexus birth palsy rotators, which results in an internal rotation contracture. This constant position of internal rotation leads to early glenohumeral joint deformity by 6 months of age and advanced deformity by 2 years, which is characterized by increasing glenoid retroversion and posterior humeral head subluxation (a.k.a. glenohumeral dysplasia).2-4,9,10,14,17,23,24,27,28,32 Children with an established internal rotation contracture and glenohumeral joint deformity are unlikely to regain optimum shoulder function without intervention. The treatment is controversial. Tendon transfers about the shoulder result in better motion, but fail to realign the glenohumeral joint.5,15,30 The inability to affect joint alignment lends concern about potential long-term joint sequelae, and may explain the loss in clinical improvement over time.12,19 Open or arthoscopic anterior capsulectomy and subscapularis release can reduce the glenohumeral joint and promote remodeling over time.11,22,24,26 The purpose of the study was to report our experience using arthroscopic anterior release, þ/ tendon transfers to maintain shoulder joint alignment in children with brachial plexus palsy. An additional goal was to assess our patient’s outcomes after arthroscopic reduction.
Materials and methods Patient population This study is a retrospective chart review of 44 children that underwent arthroscopic anterior release, partial subscapularis tenotomy, and þ/ tendon transfers in an attempt to realign a dysplastic glenohumeral joint and improve motion in children with brachial plexus injury between 2002 and 2006. Patients available for review had pre-operative and post-operative imaging and clinical measurements (Table I). Of the 44 patients, 28 children underwent isolated release and 16 underwent concomitant tendon transfers. The choice between isolated arthroscopic release and release plus con-commitment tendon transfer was based upon multiple parameters. Factors included age of the patient, parental decision, and social circumstances. Children less than 3 years of age were usually treated with arthroscopic release alone, and those older underwent con-commitment tendon transfers. However, parents that felt the possibility of additional surgery was unacceptable or those families that lived a far distance and had difficulty returning to hospital may choose to undergo a combined procedure. There were 28 girls and 16 boys with an average age of 2.7 years (range, 0.9-8.4) (Table I). The brachial plexus palsy involved C5 and C6 in 36 children and C5, C6, and C7 in 8 children. Previous surgeries included 6 children that underwent nerve grafting/transfers and 3 children that had prior botulinum injection. A previously reported subgroup of 13 patients that underwent post-operative imaging in their spica cast to assess glenohumeral reduction was included.24
Clinical measures Licensed occupational therapists performed clinical measurements in the clinic. Measurements were obtained with a goniometer and
103 Table 1
Patient Demographics and Surgical Procedures All subjects
Gender Involved side Level of Injury Age at Surgery )
With Transfers
Male Female Right
28 16 24
8 8 8
Left C5/C6
20 36
8 14
C5/C6/C7 Average
8 2.6
2 3.4
Without Tendon Transfer 20) 8 16) 12 22) 6 2.3))
Not statistical different using Chi-square test. statistically different using independent t-test.
))
included both passive external rotation with the arm adducted and the shoulder stabilized, as well as active elevation of the shoulder. For global shoulder function, the Mallet classification system was used (Figure 1).6,15,30,31 The Mallet scale assesses global abduction, global external rotation, hand-to-mouth, hand-to-neck, and hand-to-spine active range of motion. Patients are scored on an ordinal scale from 1 (no function) to 5 (normal function). The Mallet classification has been shown to be a reliable instrument for assessing upper extremity function in children with brachial plexus birth palsy.1 Mallet scores were only recorded for age appropriate patients that could comply with testing via verbal cueing or enticement (n ¼ 21).
Surgical technique20,
22-24
Arthroscopy is performed in a lateral decubitus position using a 2.7-mm arthroscope. A posterior portal is established through the posterior soft spot after localization with a spinal needle and joint insufflation with saline. An anterior portal is made under direct visualization using an 18-gauge spinal needle inserted through the rotator interval, inferior to the biceps tendon. The rotator interval, anterior capsule, anterior ligaments, and subscapularis tendon are identified. An electrocautery is introduced through the anterior portal. The thickened superior glenohumeral ligament, middle glenohumeral ligament, and upper 1-half to 2-thirds of the subscapularis are released. This release starts superiorly in the rotator interval with the thickened superior glenohumeral ligament, followed by the middle glenohumeral ligament. The upper, intra-articular portion of the subscapularis tendon is then released. As the release continues inferiorly, the tendinous portion of the subscapularis transitions into a more muscular portion. This transition occurs near the reflection of the anterior band of the inferior glenohumeral ligament. At this point, the release becomes isolated to the capsule, with preservation of the inferior and lateral muscular portions of the subscapularis. The electrocautery is removed and exchanged for an arthroscopic punch. The inferior glenohumeral ligament is then released to a point slightly posterior to the mid-portion of the axillary pouch. The axillary nerve is protected by passing the punch onto the extra-articular surface of the inferior capsule with the jaws closed. This presumably dissects the nerve away from the capsule before performing the capsulotomy. The arthroscopic equipment is
104
S.H. Kozin et al.
Figure 1 A 4-year-old female (patient #37) with initial C5, C6, and C7 injury. Spontaneous recovery began at 5 months. No previous surgery. Status post right arthroscopic anterior release demonstrating Mallet parameters for shoulder function. (A) Abduction (grade IV); (B) external rotation (grade IV); (C) hand to neck (grade IV); (D) hand to mouth (grade IV); (E) hand to spine (grade II).
removed from the joint and the glenohumeral joint is manipulated into external rotation, both with the arm at the side and with the arm at 90 of elevation. Marked improvement is external rotation is noted, often with a palpable clunk associated with glenohumeral joint reduction. Failure to achieve joint reduction or passive external rotation less than 45 with the arm in adduction requires additional arthroscopic release of the axilary pouch and/or tight subscapularis. In children with concomitant tendon transfers, the posterior portal is extended into a posterior axillary incision. The latissimus dorsi and teres major tendons are released from the humerus with a periosteal sleeve. The axillary nerve is carefully identified and protected. The tendons are transferred to the superior-posterior rotator cuff and humerus following arthroscopic release. The child was placed in a shoulder spica cast with the glenohumeral joint positioned in 45 -60 of external rotation. The amount of abduction varies according to whether or whether or not tendon transfers were performed. The arm is positioned in 30 -40 of abduction after isolated release and 100 -120 of abduction
after release combined with tendon transfer. In both instances, external rotation and anterior translation is applied to reduce the glenohumeral joint. Casting is continued for 3 weeks after isolated release and 4-5 weeks after tendon transfer.
Magnetic resonance imaging The majority of the children underwent magnetic resonance imaging at our institution. In these cases, all preoperative and postoperative imaging was done with the same 1.5 Tesla LX platform imaging unit (GE Medical Systems, Milwaukee, WI). A GPFlex (GE Medical Systems) shoulder coil was used. Axial T2* gradient echo (256/ 128 Nex 2), axial T2 gradient with fat saturation (256/128 Nex 2), axial T@ fast spin echo with fat saturation (256/192 Nex 2), and T1 axial (256/192 Nex 2) images were acquired. The children were sedated and monitored by electrocardiography, oxygen-saturation measurements, and observation.
Shoulder arthroscopy in children with brachial plexus birth palsy
105
Figure 2 Normal, pre- and postoperative shoulder MRIs of 4-year-old female with right brachial plexus palsy (patient #37). (A) Normal left shoulder (PHHA ¼ 44%, version ¼ 14 ); (B) pre-operative right shoulder with pseudeglenoid (PHHA ¼ 16.7%, version ¼ 39 ); (C) postoperative right shoulder (PHHA ¼ 37.2%, version ¼ 16 ).
In all patients, preoperative shoulder MRI scans and postoperative MRI scans of the affected shoulder were obtained to assess the quality of the glenohumeral joint reduction (Figure 2). For the preoperative imaging study, the unaffected shoulder is placed in neutral rotation at the side while the affected shoulder rests in varying degrees of internal rotation dependent upon the degree of contracture. Imaging measurements and data collection were performed by one of us (MJB) who was blinded to all clinical information gathered in the study. An axial image inferior to the coracoid apophysis and spinoglenoid notch was selected to standardize measurements.14 Using this image slice, the glenoscapular angle (the degree of version) and percentage of the humeral head anterior to the middle of the glenoid fossa (PHHA) was calculated for all scans.2,3,14,16,29,32 The IMPAX for orthopedics (Agfa-Gevaert Group, Morstel, Belgium) software tool was used to measure the radiographic parameters. A line was constructed along the coronal axis of the scapular body that connected the medial margin of the scapula and middle of the glenoid fossa. This line was extended through the humeral head. In dysplastic joints with a humeral head articulating with a posterior articular concavity (a.k.a. pseudoglenoid), this line generally passed through the apex of the transition from native to pseudoglenoid (not the midpoint of the native glenoid).14-16,22 A line was then drawn along the labral surfaces of the glenoid connecting the anterior and posterior margins. If the labral margins were not identifiable, the cartilaginous glenoid surface was used to calculate version. In cases with a pseudoglenoid, the posterior articular concavity was used to measure glenoid version.14,20 Ninety degrees was subtracted from the angle formed by the coronal axis of the scapular body and the labral surfaces to calculate glenoid version.32 Negative values indicated retroversion, while positive values reflected glenoid anteversion. The deformed glenoid shape was also defined using the Classification of Glenohumeral Deformity Scale (GDS), which is a 4-point ordinal scales as follows: (1) concentric, round humeral head centered on concave glenoid of matching curvature; (2) flat, near complete loss of glenoid curvature; (3) biconcave, central apex defining anterior and posterior aspects of glenoid, with same version relative to scapula; (4) pseudoglenoid, central apex defining anterior and posterior aspects of glenoid, but retroversion of posterior aspect is increased relative to scapular center-line.21,23
The PHHA was calculated by making a measurement at the widest portion of the humeral head, perpendicular to the line along the axis of the scapula.15,32 The width of the humeral head projected anterior to the axis line was divided by the total width of the head. A lower PHHA indicated more posterior subluxation. The average measurements on the uninvolved images were compared to the average measurements on the involved sides. In addition, the preoperative and postoperative measurements on the involved side were compared to each other.
Statistical analysis Mean and standard deviations were computed for each of the continuous outcome measures at baseline, postoperative, and at 1-year following surgery. Paired t tests were used to compare changes from preoperative to 1-year follow-up for continuous variables of external rotation, active elevation, retroversion, PHHA, and aggregate Mallet score. As the individual Mallet domains are nonparametric, the sign rank test was used for comparisons. A one-way repeated measures ANOVA was used to evaluate statistically significant changes over time on the basis of baseline, immediate postoperative, and 1 year following surgery MRI outcome measures. Spearman Rho coefficients were used to evaluate relationships between independent variables of age at surgery, gender, side of injury, level of injury, tendon transfers, and preoperative clinical and MRI measurements on the dependent variables of 1-year external rotation, active elevation, PHHA, retroversion, and deformity scales. General linear models were used to evaluate significant correlates of active elevation among the independent variables of gender, age at injury, level of injury, concomitant transfers, pre-operative active elevation, and preoperative PHHA. All significant variables were included in the final model. To interpret findings in the presence of the significant interaction between pre-operative abduction and pre-operative PHHA, separate models are presented for groups defined by subjects having pre-operative abduction < 100 and >100 . Statistical significance was taken at the 0.05 level, and analyses were performed using SPSS and SAS 9.1 (SAS Institute Inc.).
106
S.H. Kozin et al.
Table II Preoperative and one-year postoperative MRIs and passive external rotation and active abduction and mallet scale
PHHA Retroversion Deformity ER ROM ABD ROM Mallet Abduction Mallet External Rotation Mallet Hand to Neck Mallet Hand to Spine Mallet Hand to Mouth Mallet Total
PreOp Mean (degrees)
Post-OP Mean (degrees)
p-value
19% 12% 34 15 2.9 1.0 26 20 112 28 3.9 0.4 2.1 0.5
33% 12% 19 13 1.9 0.4 47 17 130 38 3.9 0.4 3.7 0.6
0.001 0.001 0.001 0.001 0.008 1 < 0.001
2.4 0.7
3.8 0.4
< 0.001
2.3 0.6
2.3 0.7
2.1 0.3
3.4 0.7
< 0.001
12.7 1.6
17.1 1.4
< 0.001
< < < <
1
PHHA percentage of the humeral head anterior to the middle of the glenoid fossa; Retroversion, the degree of glenoid version; ER, external rotation; ABD, abduction; ROM, range of motion.
Results MRI and clinical data at 1 year (n [ 44) The average time for clinical evaluations was 1.0 0.3 years from date of surgery. The clinical data demonstrated a significant improvement (P < .001) in external rotation and active elevation (P ¼ .008) from pre-operative to 1year measurements (Table II) (Figure 4). Passive external rotation increased from -26 (20) degrees prior to surgery to þ47.0 (17) degrees at 1-year (P < .001). Active elevation improved from 112 (28) degrees before surgery to 130 (38) degrees at follow-up (P ¼ .008). Average aggregate Mallet scores improved significantly (P < .001) from 12.7 (þ1.6) prior to surgery to 17.1 (1.4) at 1-year follow-up (Table III). Specifically, statistical improvements were found in external rotation, 2.1 to 3.7; hand to neck, 2.4 to 3.8; and hand to mouth, 2.1 to 3.4. Changes in Mallet score for abduction and hand to spine were not significant (Table III). The average time of follow-up MRI was 1.2 0.3 years from date of surgery. Assessing MRI data, there was a significant improvement (P < .001) in both retroversion and PHHA from preoperative to 1 year following surgery (Table II) (Figure 3). For average retroversion, preoperative measurements improved significantly from -34 (15) degrees to -19 (13) degrees. For average PHHA, preoperative measurements improved from 19% (12%) to 33% (12%). The Glenoid Deformity Scale improved significantly (P < .001) from a preoperative average score of 2.9 (1.0) to 1.98 (0.4). Of the 19 patients with a type-4
pseudoglenoid deformity prior to surgery, only 1 had a persistent pseudoglenoid at 1-year follow-up. Using a Spearman Rho correlation, MRI and clinical outcomes at 1-year were evaluated further for significant prognostic indicators among preoperative clinical and MRI status, gender, level of injury, and presence of tendon transfers. The preoperative status was strongly associated with a positive outcome in external rotation (P ¼ .031), active elevation (P ¼ .022), PHHA (P ¼ .008), and retroversion (P ¼ .027) for all dependent variables except the deformity scale. For example, having more external rotation before surgery was associated with more external rotation after surgery. Similarly, having less subluxation and less retroversion was associated with better joint alignment after surgery. Regarding active elevation at follow-up, the preoperative degree for active elevation and having a concomitant tendon transfer (P ¼ .010) were associated with a better active elevation. There was an inverse relationship between PHHA and active elevation at follow-up (P ¼ .021). Using an ANCOVA model (Table III), the combination of a concomitant tendon transfer and less PHHA before surgery was statistically associated with better active elevation in patients with preoperative abduction <100 . The least squares adjusted mean values for active elevation in patient with pre-operative abduction <100 at 1 year was 135 (P ¼ .058) degrees for patients with concomitant tendon transfer, compared to 101 for those without concomitant tendon transfer. For subjects with equal to or greater than 100 abduction preoperatively, only having concomitant tendon transfer was statistically significant, with a least-squares means of 159 compared to 128 (P ¼ .016).
Discussion Glenohumeral dysplasia following brachial plexus birth palsy occurs early and frequently.7,8 Van der Sluijs et al28 imaged 16 children (17 shoulders) less than 1 year of age. In children less 5 months of age, a normal shoulder was found in 5 out of 7 cases; however, in children older than 5 months of age, a normal shoulder was seen in only 2 out of 10. Prevention of deformity is the mainstay of early management. Maintenance of passive external rotation prevents an internal rotation contracture, which leads to posterior humeral head subluxation and glenoid retroversion. The gradual loss of passive external rotation is a sign of impending glenohumeral dysplasia.15,21,32 Once external rotation becomes less than neutral with the scapula stabilized, glenohumeral deformity is evident in the majority of shoulders.14,21,32,33 Tendon transfers about the shoulder without open reduction have been shown to improve motion, but have a negligible effect on correction of any underlying deformity.5,15,18,30 At best, tendon transfers may halt the progression of deformity that is associated with muscular
Shoulder arthroscopy in children with brachial plexus birth palsy
107
Figure 3 A 6-year-old female (patient #3) with initial C5 and C6 injury. Spontaneous recovery began at 4 months. No previous surgery. Status post left arthroscopic anterior release and tendon transfers demonstrating Mallet parameters for shoulder function. (A) Abduction (grade IV); (B) hand to neck (grade IV from grade II preoperative); (C) hand to mouth (grade IV from grade III pre-operative); (D) hand to spine (grade II with no change from pre-operative).
imbalance. This inability to improve joint alignment may explain the loss in clinical improvement over time and raises concern about potential long-term joint sequelae.19 Open or arthoscopic anterior capsulectomy and
subscapularis release can reduce the glenohumeral joint and promote remodeling over time.11,22,24,26 However, the role of joint reduction remains controversial as complications have been reported.11,26 Open reduction is associated with
108
S.H. Kozin et al.
Figure 4 A 6-year-old female (patient #3). Pre- and postoperative MRIs following left shoulder arthroscopic anterior release and tendon transfers. (A) Pre-operative left shoulder with pseudeglenoid (PHHA ¼ 24.6%, version ¼ 37 ). (B) Postoperative left shoulder with marked improvement (PHHA ¼ 36.8%, version ¼ 4 ).
Table III
ANCOVA model results for abduction at one year
All subjects (n ¼ 42)
Means Squared
F-value
p-value
Transfer Preoperative ABD Preoperative PHHA abd)PHHA
8156 253 6750 4445
8.91 0.28 7.38 4.86
0.005 0.602 0.01 0.033
Abduction <100 (n ¼ 17) Transfer 4645.62 Preoperative PHHA 9376.05
4.25 8.59
0.0582 0.011
Abduction > 100 (n ¼ 27) Transfer 5467.85 Preoperative PHHA 17.85
6.67 0.02
0.016 0.8839
excessive external rotation and dramatic loss of internal rotation and midline function. Van der Sluijs et al26 performed open release in 19 children. Eight developed a ‘‘severe, functionally disturbing external rotation contracture of the shoulder.’’ However, open reduction has been shown to promote joint remodeling and thus remains a sensible treatment modality, especially in the young child.11 Pearl in 2003 published an arthroscopic technique for contracture and subscapularis release in children with brachial plexus birth palsy.20 The goal was to restore glenohumeral alignment and re-balance the joint using a less invasive procedure than formal open reduction. Forty-one children underwent arthroscopic release of the anterior capsule and subscapularis tendon. The mean age of the children was 3.5 years. Eighteen children were treated with arthroscopic release alone, while 23 children also underwent concomitant tendon transfer. The arthroscopic contracture release effectively restored passive
external rotation in 40 of the 41 children. The single patient that did not achieve external rotation was 12 years of age with advanced glenoid deformity. The status of the glenohumeral joint, however, was not evaluated after surgery. Pedowitz et al24 assessed the ability of arthroscopic release to reduce glenohumeral joint subluxation. Twentytwo children with an average age of 3.9 years underwent preoperative magnetic resonance imaging (MRI), arthroscopic surgery with or without tendon transfers, and postoperative imaging in their spica cast. Prior to surgery, the involved shoulder preoperative mean PHHA was 15.6% 13.5% and the mean glenoid version was e37 15 . After surgery and within the cast, the mean PHHA corrected to 46.9% 11.2% and the mean glenoid version improved to e8 8 (P < .001). The immediate improvement in glenoid version was primarily attributed to reduction of the humeral head from the pseudoglenoid onto the native glenoid and secondary to the fast remodeling and pliable pediatric cartilage.25 Pearl et al22 also reported follow-up on the first 33 children that underwent arthroscopic surgery. Nineteen children (all <3 years of age) underwent isolated arthroscopic release and 14 children (mean age, 6.7 years) underwent concomitant tendon transfer. Improved external rotation was noted in all children, except 1 child that was 12 years old. Passive external rotation increased between 60 and 80 dependent upon the procedure. Minimal improvement in mean active elevation was noted, even in the tendon transfer cohort. Four children (21%) that were treated with arthroscopic release alone required repeat surgery with an additional tendon transfer. Although internal rotation was not measured consistently prior to surgery, substantial reduction was note following surgery ranging from ‘‘moderate to severe.’’ MRI follow-up at 2
Shoulder arthroscopy in children with brachial plexus birth palsy years revealed ‘‘marked remodeling’’ in 12 out of 15 children with a pseudoglenoid deformity. Our report attempted to combine MRI and clinical findings in 44 patients. Twenty-eight children underwent isolated arthroscopic releases and 16 children underwent concomitant tendon transfers. In addition, a subgroup of 13 patients that underwent postoperative imaging in their spica cast was included to compare their current status of glenohumeral reduction.24 Similar to Pearl et al,22 we noted a significant improvement in external rotation both passive (26 to þ47 ) and active via the Mallet parameters that assess external rotation (external rotation, 2.1-3.7; hand to neck, 2.4-3.7; and hand to mouth, 2.1-3.4). In contrast to Pearl et al,22 we noted a significant improvement in active elevation (112 -130 ), especially in those patients undergoing concomitant tendon transfers compared to those children with isolated release (150 compared to 118 ). Our results of increased elevation may be multi-factorial, including patient selection and differences in surgical technique, immobilization, and rehabilitation. The Mallet score for abduction did not change; however, this scoring system does not discriminate once active elevation is greater than 90 . With regards to internal rotation, we previously noted a loss in internal rotation and midline function after complete arthrocopic release of the subscapularis24; therefore, we preserve the inferior and lateral muscular portion of the subscapularis. This technical modification can routinely be accomplished in younger children with less deformity and stresses the necessity of early diagnosis.14,24,32 Importantly, the Mallet parameter for internal rotation assesses the ability to place the hand on spine, and we noted no statistical change; however, our pre-operative score averaged 2.3, and a grade 2 implies inability to perform this motion. Clinically, the loss of internal rotation is seen by the loss of midline function, not the inability to place the hand on the spine. Recent reports of long-term results after tendon transfers report diminished motion over time. Kirkos et al13 report mean follow-up of 30 years in 10 patients, 8 of which had upper brachial plexus injuries. The authors assessed preoperative deformity on plain films only, and the tendon transfer technique was designed to improve external rotation and not abduction. Eight of the 10 patients had deterioration of external rotation between the 10- and 30-year follow-up. Gilbert et al6 reported long-term follow-up after isolated latissimus dorsi transfer. Children with isolated C5-C6 palsy had the greatest gains in abduction and external rotation; however, deterioration of abduction began at 6 years despite preservation of external rotation. In our series, we obtained and confirmed joint reduction and transferred both the latissimus dorsi and teres major tendons to improve abduction and external rotation. Nonetheless, longer-term follow-up is required to insure improvement in motion. We also acknowledge the weaknesses of our retrospective study, including the lack of randomization and relatively small sample size.
109
Conclusion Our results after arthroscopic release with or without tendon transfers are encouraging, with improvements in imaging studies and clinical evaluations following surgery. The glenoid version and PHHA statistically improved after surgery consistent with better glenohumeral joint alignment. Remarkably, 18 out of 19 children with a pseudoglenoid deformity corrected to a concentric or flat configuration at follow-up with resolution of the pseudoglenoid component. The clinical improvements paralleled the MRI corrections. Parameters for external rotation significantly improved after arthroscopic surgery with clinical enhancements in objective measures and functional Mallet scores. The addition of tendon transfers at the time of arthroscopic release resulted in even better active elevation; however, this combination risks impairment of midline function, which must be discussed with the family prior to surgery. Importantly, superior outcomes were associated with better preoperative clinical and MRI status. This indicates that early recognition of glenohumeral dysplasia and timely intervention will result in better shoulder motion and improved joint alignment. Our current practice is to begin passive shoulder range of motion shortly after birth to preserve external rotation. Passive range is combined with scapular stabilization to maximize glenohumeral joint motion. Loss of external rotation beyond neutral requires imaging to assess the glenohumeral joint. MRI findings consistent with an aligned joint warrants ongoing therapy, usually following botulinum toxin to the internal rotators. MRI findings consistent with a dysplastic joint, requires arthroscopic release. In children less than 3 years of age, isolated release is performed. In older children, arthroscopic release is usually combined with transfer of the latissimus dorsi and/ or teres major tendons dependent upon the strength of internal rotation.
References 1. Bae DS, Waters PM, Zurakowski D. Reliability of three classification systems measuring active motion in brachial plexus birth palsy. J Bone Joint Surg Am 2003;85:1733-8. 2. Beischer AD, Simmons TD, Torode IP. Glenoid version in children with obstetric brachial plexus palsy. J Pediatr Orthop 1999;19:359-61. 3. Bokor DJ, O’Sullivan MD, Hazan GJ. Variability of measurement of glenoid version on computed tomography scan. J Shoulder Elbow Surg 1999;8:595-8. 4. Dunkerton MC. Posterior dislocation of the shoulder associated with obstetric brachial plexus palsy. J Bone Joint Surg Br 1989;71: 764-6. 5. El-Gammal TA, Saleh WR, El-Sayed A, Kotb MM, Imam HM, Fathi NA. Tendon transfer around the shoulder in obstetric brachial
110
6. 7.
8. 9.
10.
11.
12. 13.
14.
15.
16. 17.
18.
19.
20.
plexus paralysis: Clinical and computed tomographic study. J Pediatr Orthop 2006;26:641-6. Gilbert A, Whitaker I. Obstetrical brachial plexus lesions. J Hand Surg [Br] 1991;16:489-91. Greenwald AG, Schute PC, Shiveley JL. Brachial plexus birth palsy: a 10-year report on the incidence and prognosis. J Pediatr Orthop 1984;4:689-92. Hardy AE. Birth injuries of the brachial plexus: incidence and prognosis. J Bone Joint Surg Br 1981;63:98-101. Hoffer MM, Phipps GJ. Closed reduction and tendon transfer for treatment of dislocation of the glenohumeral joint secondary to brachial plexus birth palsy. J Bone Joint Surg Am 1998;80:997-1001. Hoffer MM, Wickenden R, Roper B. Brachial plexus birth palsies. Results of tendon transfers to the rotator cuff. J Bone Joint Surg Am 1978;60:691-5. Hui JH, Torode IP. Changing glenoid version after open reduction of shoulders in children with obstetric brachial plexus palsy. J Pediatr Orthop 2003;23:109-13. Jackson ST, Hoffer MM, Parrish N. Brachial-plexus palsy in the newborn. J Bone Joint Surg Am 1988;70:1217-20. Kirkos JM, Kyrkos MJ, Kapetanos GA, Haritidis JH. Brachial plexus palsy secondary to birth injuries. Long-term results of anterior release and tendon transfers around the shoulder. J Bone Joint Surg [Br] 2005; 87:3231-5. Kozin SH. Correlation between external rotation of the glenohumeral joint and deformity after brachial plexus birth palsy. J Pediatr Orthop 2004;24:189-93. Kozin SH, Chafetz RS, Barus D, Filipone L. MRI and clinical findings before and after tendon transfers about the shoulder in children with residual brachial plexus birth palsy. J Shoulder Elbow Surg 2006;15: 554-61. Mintzer GM, Waters PM, Brown DJ. Glenoid version in children. J Pediatr Orthop 1996;6:563-6. Moukoko D, Ezaki M, Wilkes D, Carter P. Posterior shoulder dislocation in infants with neonatal brachial plexus palsy. J Bone Joint Surg Am 2004;86:787-93. Newman CJ, Morrison L, Lynch B, Hynes D. Outcome of subscapularis muscle release for shoulder contracture secondary to brachial plexus palsy at birth. J Pediatr Orthop 2006;26:647-51. Pagnotta A, Haerle M, Gilbert A. Long-term results on abduction and external rotation of the shoulder after latissimus dorsi transfer for sequelae of obstetric palsy. Clin Orthop 2004;426:199-205. Pearl ML. Arthroscopic release of shoulder contracture secondary to birth palsy: an early report on findings and surgical technique. Arthroscopy 2003;19:577-82.
S.H. Kozin et al. 21. Pearl ML, Edgerton BW. Glenoid deformity secondary to brachial plexus birth palsy. J Bone Joint Surg Am 1998;80:659-67. 22. Pearl ML, Edgerton BW, Kazimiroff PA, Burchette RJ, Wong K. Arthroscopic release and latissimus dorsi transfer for shoulder internal rotation contractures and glenohumeral deformity secondary to brachial plexus birth palsy. J Bone Joint Surg Am 2006;88:564-74. 23. Pearl ML, Edgerton BW, Kon DS, Darakjian AB, Kosco AE, Kazimiroff PB, et al. Comparison of arthroscopic findings with magnetic resonance imaging and arthrography in children with glenohumeral deformities secondary to brachial plexus birth palsy. J Bone Joint Surg [Am] 2003;85:890-8. 24. Pedowitz DI, Gibson B, Williams GR, Kozin SH. Arthroscopic treatment of posterior glenohumeral joint subluxation. J Shoulder Elbow Surg 2007;16:6-13. 25. Pirini S, Zeznik L, Hodges D. Magnetic resonance imaging study of congenital clubfoot treated with the Ponseti method. J Ped Orthop 2001;21:719-26. 26. van der Sluijs JA, van Ouwerkerk WJ, de Gast A, Nollet F, Winters H, Wuisman PI. Treatment of internal rotation contracture of the shoulder in obstetric brachial plexus lesions by subscapularis tendon lengthening and open reduction: early results and complications. J Pediatric Orthop B 2004;13:218-24. 27. van der Sluijs JA, van Ouwerkerk WJ, de Gast A, Wuisman PI, Nollet F, Manoliu RA. Retroversion of the humeral head in children with an obstetric brachial plexus lesion. J Bone Joint Surg Br 2002;84: 583-7. 28. van der Sluijs JA, van Ouwerkerk WJ, de Gast A, Wuisman PI, Nollet F, Manoliu RA. Deformities of the shoulder in infants younger than 12 months with an obstetric lesion of the brachial plexus. J Bone Joint Surg Br 2001;83:551-5. 29. van der Sluijs JA, van der Meij, Verbeke J, Manoliu RA, Wuisman PI. Measuring secondary deformities in children with obstetric brachial plexus lesion: reliability of three methods. J Pediatric Orthop B 2003; 12:211-4. 30. Waters PM, Bae DS. Effect of tendon transfers and extra-articular softtissue balancing on glenohumeral development in brachial plexus birth palsy. J Bone Joint Surg [Am] 2005;87:320-5. 31. Waters PM, Peljovich AE. Shoulder reconstruction in patients with chronic brachial plexus birth palsy. A case control study. Clin Orthop 1999;364:144-52. 32. Waters PM, Smith GR, Jaramillo D. Glenohumeral deformity secondary to brachial plexus birth palsy. J Bone Joint Surg Am 1998; 80:668-77. 33. Zancolli EA, Zancolli ER Jr. . Palliative surgical procedures in sequelae of obstetric palsy. Hand Clin 1988;4:643-69.
www.elsevier.com/locate/ymse
Arthroscopic treatment of internal rotation contracture and glenohumeral dysplasia in children with brachial plexus birth palsy Scott H. Kozin, MDa,*, Matthew J. Boardman, DOb, Ross S. Chafetz, PT, DPT, MPHc, Gerald R. Williams, MDd, Alexandra Hanlon, PhDe a
Department of Orthopaedic Surgery, Temple University & Hand Surgeon, Shriners Hospital for Children, Philadelphia, PA b Department of Orthopaedic Surgery, Philadelphia College of Osteopathic Medicine, Philadelphia, PA c Shriners Hospital for Children, Philadelphia, PA d Thomas Jefferson University, Philadelphia, PA e Department of Public Health, Temple University, Philadelphia, PA Background: The purpose of the study was to assess the ability of arthroscopic anterior release, þ/- tendon transfers to maintain shoulder joint alignment in children with brachial plexus palsy, and to assess their outcome after arthroscopic reduction. Methods: Forty-four patients underwent arthroscopic release, þ/- tendon transfers to realign a dysplastic glenohumeral joint in children with brachial plexus palsy. Twenty-eight children underwent isolated release and 16 children underwent concomitant tendon transfers. MRI and clinical measurements were used to assess outcome at 1-year follow-up. Results: There was a significant improvement (P < .001) in both retroversion from 34 (15) to -19 (13), and percentage of the humeral head anterior to the middle of the glenoid fossa (PHHA) from 19% (12%) to 33% (12%), at 1 year. Passive external rotation increased from 26 (20) degrees to 47 (17) degrees (P < .001). Active elevation increased from 112 (28) degrees to 130 (38) (P ¼ .008) degrees. Patients that underwent tendon transfers obtained greater active elevation, 147 (9) degrees compared to 119 (6) degrees. Mallet aggregate and domain scores also demonstrated statistically significant improvements. Conclusions: Our results after arthroscopic release þ/- tendon transfers are encouraging with improvements in joint alignment and clinical evaluations following surgery. The clinical improvements paralleled the MRI corrections. Importantly, superior outcomes were associated with better preoperative clinical and MRI status. This indicates that early recognition of glenohumeral dysplasia and timely intervention results in better shoulder motion and improved joint alignment. Level of Evidence: 4. Ó 2010 Journal of Shoulder and Elbow Surgery Board of Trustees.
Investigation performed at Shriners Hospital for Children, Philadelphia, Pennsylvania. This retrospective chart review study was conducted under exempt status Temple University IRB #10950. *Reprint requests: Scott H. Kozin, MD, Shriners Hospital for Children, 3551 North Broad Street, Philadelphia, PA 19140. E-mail address: [email protected] (S.H. Kozin).
Incomplete recovery after brachial plexus birth palsy often results in decreased movement and muscle imbalance about the shoulder, as rotator cuff and deltoid innervation is incomplete. The internal rotators overpower the external
1058-2746/2010/$36.00 - see front matter Ó 2010 Journal of Shoulder and Elbow Surgery Board of Trustees. doi:10.1016/j.jse.2009.05.011
Shoulder arthroscopy in children with brachial plexus birth palsy rotators, which results in an internal rotation contracture. This constant position of internal rotation leads to early glenohumeral joint deformity by 6 months of age and advanced deformity by 2 years, which is characterized by increasing glenoid retroversion and posterior humeral head subluxation (a.k.a. glenohumeral dysplasia).2-4,9,10,14,17,23,24,27,28,32 Children with an established internal rotation contracture and glenohumeral joint deformity are unlikely to regain optimum shoulder function without intervention. The treatment is controversial. Tendon transfers about the shoulder result in better motion, but fail to realign the glenohumeral joint.5,15,30 The inability to affect joint alignment lends concern about potential long-term joint sequelae, and may explain the loss in clinical improvement over time.12,19 Open or arthoscopic anterior capsulectomy and subscapularis release can reduce the glenohumeral joint and promote remodeling over time.11,22,24,26 The purpose of the study was to report our experience using arthroscopic anterior release, þ/ tendon transfers to maintain shoulder joint alignment in children with brachial plexus palsy. An additional goal was to assess our patient’s outcomes after arthroscopic reduction.
Materials and methods Patient population This study is a retrospective chart review of 44 children that underwent arthroscopic anterior release, partial subscapularis tenotomy, and þ/ tendon transfers in an attempt to realign a dysplastic glenohumeral joint and improve motion in children with brachial plexus injury between 2002 and 2006. Patients available for review had pre-operative and post-operative imaging and clinical measurements (Table I). Of the 44 patients, 28 children underwent isolated release and 16 underwent concomitant tendon transfers. The choice between isolated arthroscopic release and release plus con-commitment tendon transfer was based upon multiple parameters. Factors included age of the patient, parental decision, and social circumstances. Children less than 3 years of age were usually treated with arthroscopic release alone, and those older underwent con-commitment tendon transfers. However, parents that felt the possibility of additional surgery was unacceptable or those families that lived a far distance and had difficulty returning to hospital may choose to undergo a combined procedure. There were 28 girls and 16 boys with an average age of 2.7 years (range, 0.9-8.4) (Table I). The brachial plexus palsy involved C5 and C6 in 36 children and C5, C6, and C7 in 8 children. Previous surgeries included 6 children that underwent nerve grafting/transfers and 3 children that had prior botulinum injection. A previously reported subgroup of 13 patients that underwent post-operative imaging in their spica cast to assess glenohumeral reduction was included.24
Clinical measures Licensed occupational therapists performed clinical measurements in the clinic. Measurements were obtained with a goniometer and
103 Table 1
Patient Demographics and Surgical Procedures All subjects
Gender Involved side Level of Injury Age at Surgery )
With Transfers
Male Female Right
28 16 24
8 8 8
Left C5/C6
20 36
8 14
C5/C6/C7 Average
8 2.6
2 3.4
Without Tendon Transfer 20) 8 16) 12 22) 6 2.3))
Not statistical different using Chi-square test. statistically different using independent t-test.
))
included both passive external rotation with the arm adducted and the shoulder stabilized, as well as active elevation of the shoulder. For global shoulder function, the Mallet classification system was used (Figure 1).6,15,30,31 The Mallet scale assesses global abduction, global external rotation, hand-to-mouth, hand-to-neck, and hand-to-spine active range of motion. Patients are scored on an ordinal scale from 1 (no function) to 5 (normal function). The Mallet classification has been shown to be a reliable instrument for assessing upper extremity function in children with brachial plexus birth palsy.1 Mallet scores were only recorded for age appropriate patients that could comply with testing via verbal cueing or enticement (n ¼ 21).
Surgical technique20,
22-24
Arthroscopy is performed in a lateral decubitus position using a 2.7-mm arthroscope. A posterior portal is established through the posterior soft spot after localization with a spinal needle and joint insufflation with saline. An anterior portal is made under direct visualization using an 18-gauge spinal needle inserted through the rotator interval, inferior to the biceps tendon. The rotator interval, anterior capsule, anterior ligaments, and subscapularis tendon are identified. An electrocautery is introduced through the anterior portal. The thickened superior glenohumeral ligament, middle glenohumeral ligament, and upper 1-half to 2-thirds of the subscapularis are released. This release starts superiorly in the rotator interval with the thickened superior glenohumeral ligament, followed by the middle glenohumeral ligament. The upper, intra-articular portion of the subscapularis tendon is then released. As the release continues inferiorly, the tendinous portion of the subscapularis transitions into a more muscular portion. This transition occurs near the reflection of the anterior band of the inferior glenohumeral ligament. At this point, the release becomes isolated to the capsule, with preservation of the inferior and lateral muscular portions of the subscapularis. The electrocautery is removed and exchanged for an arthroscopic punch. The inferior glenohumeral ligament is then released to a point slightly posterior to the mid-portion of the axillary pouch. The axillary nerve is protected by passing the punch onto the extra-articular surface of the inferior capsule with the jaws closed. This presumably dissects the nerve away from the capsule before performing the capsulotomy. The arthroscopic equipment is
104
S.H. Kozin et al.
Figure 1 A 4-year-old female (patient #37) with initial C5, C6, and C7 injury. Spontaneous recovery began at 5 months. No previous surgery. Status post right arthroscopic anterior release demonstrating Mallet parameters for shoulder function. (A) Abduction (grade IV); (B) external rotation (grade IV); (C) hand to neck (grade IV); (D) hand to mouth (grade IV); (E) hand to spine (grade II).
removed from the joint and the glenohumeral joint is manipulated into external rotation, both with the arm at the side and with the arm at 90 of elevation. Marked improvement is external rotation is noted, often with a palpable clunk associated with glenohumeral joint reduction. Failure to achieve joint reduction or passive external rotation less than 45 with the arm in adduction requires additional arthroscopic release of the axilary pouch and/or tight subscapularis. In children with concomitant tendon transfers, the posterior portal is extended into a posterior axillary incision. The latissimus dorsi and teres major tendons are released from the humerus with a periosteal sleeve. The axillary nerve is carefully identified and protected. The tendons are transferred to the superior-posterior rotator cuff and humerus following arthroscopic release. The child was placed in a shoulder spica cast with the glenohumeral joint positioned in 45 -60 of external rotation. The amount of abduction varies according to whether or whether or not tendon transfers were performed. The arm is positioned in 30 -40 of abduction after isolated release and 100 -120 of abduction
after release combined with tendon transfer. In both instances, external rotation and anterior translation is applied to reduce the glenohumeral joint. Casting is continued for 3 weeks after isolated release and 4-5 weeks after tendon transfer.
Magnetic resonance imaging The majority of the children underwent magnetic resonance imaging at our institution. In these cases, all preoperative and postoperative imaging was done with the same 1.5 Tesla LX platform imaging unit (GE Medical Systems, Milwaukee, WI). A GPFlex (GE Medical Systems) shoulder coil was used. Axial T2* gradient echo (256/ 128 Nex 2), axial T2 gradient with fat saturation (256/128 Nex 2), axial T@ fast spin echo with fat saturation (256/192 Nex 2), and T1 axial (256/192 Nex 2) images were acquired. The children were sedated and monitored by electrocardiography, oxygen-saturation measurements, and observation.
Shoulder arthroscopy in children with brachial plexus birth palsy
105
Figure 2 Normal, pre- and postoperative shoulder MRIs of 4-year-old female with right brachial plexus palsy (patient #37). (A) Normal left shoulder (PHHA ¼ 44%, version ¼ 14 ); (B) pre-operative right shoulder with pseudeglenoid (PHHA ¼ 16.7%, version ¼ 39 ); (C) postoperative right shoulder (PHHA ¼ 37.2%, version ¼ 16 ).
In all patients, preoperative shoulder MRI scans and postoperative MRI scans of the affected shoulder were obtained to assess the quality of the glenohumeral joint reduction (Figure 2). For the preoperative imaging study, the unaffected shoulder is placed in neutral rotation at the side while the affected shoulder rests in varying degrees of internal rotation dependent upon the degree of contracture. Imaging measurements and data collection were performed by one of us (MJB) who was blinded to all clinical information gathered in the study. An axial image inferior to the coracoid apophysis and spinoglenoid notch was selected to standardize measurements.14 Using this image slice, the glenoscapular angle (the degree of version) and percentage of the humeral head anterior to the middle of the glenoid fossa (PHHA) was calculated for all scans.2,3,14,16,29,32 The IMPAX for orthopedics (Agfa-Gevaert Group, Morstel, Belgium) software tool was used to measure the radiographic parameters. A line was constructed along the coronal axis of the scapular body that connected the medial margin of the scapula and middle of the glenoid fossa. This line was extended through the humeral head. In dysplastic joints with a humeral head articulating with a posterior articular concavity (a.k.a. pseudoglenoid), this line generally passed through the apex of the transition from native to pseudoglenoid (not the midpoint of the native glenoid).14-16,22 A line was then drawn along the labral surfaces of the glenoid connecting the anterior and posterior margins. If the labral margins were not identifiable, the cartilaginous glenoid surface was used to calculate version. In cases with a pseudoglenoid, the posterior articular concavity was used to measure glenoid version.14,20 Ninety degrees was subtracted from the angle formed by the coronal axis of the scapular body and the labral surfaces to calculate glenoid version.32 Negative values indicated retroversion, while positive values reflected glenoid anteversion. The deformed glenoid shape was also defined using the Classification of Glenohumeral Deformity Scale (GDS), which is a 4-point ordinal scales as follows: (1) concentric, round humeral head centered on concave glenoid of matching curvature; (2) flat, near complete loss of glenoid curvature; (3) biconcave, central apex defining anterior and posterior aspects of glenoid, with same version relative to scapula; (4) pseudoglenoid, central apex defining anterior and posterior aspects of glenoid, but retroversion of posterior aspect is increased relative to scapular center-line.21,23
The PHHA was calculated by making a measurement at the widest portion of the humeral head, perpendicular to the line along the axis of the scapula.15,32 The width of the humeral head projected anterior to the axis line was divided by the total width of the head. A lower PHHA indicated more posterior subluxation. The average measurements on the uninvolved images were compared to the average measurements on the involved sides. In addition, the preoperative and postoperative measurements on the involved side were compared to each other.
Statistical analysis Mean and standard deviations were computed for each of the continuous outcome measures at baseline, postoperative, and at 1-year following surgery. Paired t tests were used to compare changes from preoperative to 1-year follow-up for continuous variables of external rotation, active elevation, retroversion, PHHA, and aggregate Mallet score. As the individual Mallet domains are nonparametric, the sign rank test was used for comparisons. A one-way repeated measures ANOVA was used to evaluate statistically significant changes over time on the basis of baseline, immediate postoperative, and 1 year following surgery MRI outcome measures. Spearman Rho coefficients were used to evaluate relationships between independent variables of age at surgery, gender, side of injury, level of injury, tendon transfers, and preoperative clinical and MRI measurements on the dependent variables of 1-year external rotation, active elevation, PHHA, retroversion, and deformity scales. General linear models were used to evaluate significant correlates of active elevation among the independent variables of gender, age at injury, level of injury, concomitant transfers, pre-operative active elevation, and preoperative PHHA. All significant variables were included in the final model. To interpret findings in the presence of the significant interaction between pre-operative abduction and pre-operative PHHA, separate models are presented for groups defined by subjects having pre-operative abduction < 100 and >100 . Statistical significance was taken at the 0.05 level, and analyses were performed using SPSS and SAS 9.1 (SAS Institute Inc.).
106
S.H. Kozin et al.
Table II Preoperative and one-year postoperative MRIs and passive external rotation and active abduction and mallet scale
PHHA Retroversion Deformity ER ROM ABD ROM Mallet Abduction Mallet External Rotation Mallet Hand to Neck Mallet Hand to Spine Mallet Hand to Mouth Mallet Total
PreOp Mean (degrees)
Post-OP Mean (degrees)
p-value
19% 12% 34 15 2.9 1.0 26 20 112 28 3.9 0.4 2.1 0.5
33% 12% 19 13 1.9 0.4 47 17 130 38 3.9 0.4 3.7 0.6
0.001 0.001 0.001 0.001 0.008 1 < 0.001
2.4 0.7
3.8 0.4
< 0.001
2.3 0.6
2.3 0.7
2.1 0.3
3.4 0.7
< 0.001
12.7 1.6
17.1 1.4
< 0.001
< < < <
1
PHHA percentage of the humeral head anterior to the middle of the glenoid fossa; Retroversion, the degree of glenoid version; ER, external rotation; ABD, abduction; ROM, range of motion.
Results MRI and clinical data at 1 year (n [ 44) The average time for clinical evaluations was 1.0 0.3 years from date of surgery. The clinical data demonstrated a significant improvement (P < .001) in external rotation and active elevation (P ¼ .008) from pre-operative to 1year measurements (Table II) (Figure 4). Passive external rotation increased from -26 (20) degrees prior to surgery to þ47.0 (17) degrees at 1-year (P < .001). Active elevation improved from 112 (28) degrees before surgery to 130 (38) degrees at follow-up (P ¼ .008). Average aggregate Mallet scores improved significantly (P < .001) from 12.7 (þ1.6) prior to surgery to 17.1 (1.4) at 1-year follow-up (Table III). Specifically, statistical improvements were found in external rotation, 2.1 to 3.7; hand to neck, 2.4 to 3.8; and hand to mouth, 2.1 to 3.4. Changes in Mallet score for abduction and hand to spine were not significant (Table III). The average time of follow-up MRI was 1.2 0.3 years from date of surgery. Assessing MRI data, there was a significant improvement (P < .001) in both retroversion and PHHA from preoperative to 1 year following surgery (Table II) (Figure 3). For average retroversion, preoperative measurements improved significantly from -34 (15) degrees to -19 (13) degrees. For average PHHA, preoperative measurements improved from 19% (12%) to 33% (12%). The Glenoid Deformity Scale improved significantly (P < .001) from a preoperative average score of 2.9 (1.0) to 1.98 (0.4). Of the 19 patients with a type-4
pseudoglenoid deformity prior to surgery, only 1 had a persistent pseudoglenoid at 1-year follow-up. Using a Spearman Rho correlation, MRI and clinical outcomes at 1-year were evaluated further for significant prognostic indicators among preoperative clinical and MRI status, gender, level of injury, and presence of tendon transfers. The preoperative status was strongly associated with a positive outcome in external rotation (P ¼ .031), active elevation (P ¼ .022), PHHA (P ¼ .008), and retroversion (P ¼ .027) for all dependent variables except the deformity scale. For example, having more external rotation before surgery was associated with more external rotation after surgery. Similarly, having less subluxation and less retroversion was associated with better joint alignment after surgery. Regarding active elevation at follow-up, the preoperative degree for active elevation and having a concomitant tendon transfer (P ¼ .010) were associated with a better active elevation. There was an inverse relationship between PHHA and active elevation at follow-up (P ¼ .021). Using an ANCOVA model (Table III), the combination of a concomitant tendon transfer and less PHHA before surgery was statistically associated with better active elevation in patients with preoperative abduction <100 . The least squares adjusted mean values for active elevation in patient with pre-operative abduction <100 at 1 year was 135 (P ¼ .058) degrees for patients with concomitant tendon transfer, compared to 101 for those without concomitant tendon transfer. For subjects with equal to or greater than 100 abduction preoperatively, only having concomitant tendon transfer was statistically significant, with a least-squares means of 159 compared to 128 (P ¼ .016).
Discussion Glenohumeral dysplasia following brachial plexus birth palsy occurs early and frequently.7,8 Van der Sluijs et al28 imaged 16 children (17 shoulders) less than 1 year of age. In children less 5 months of age, a normal shoulder was found in 5 out of 7 cases; however, in children older than 5 months of age, a normal shoulder was seen in only 2 out of 10. Prevention of deformity is the mainstay of early management. Maintenance of passive external rotation prevents an internal rotation contracture, which leads to posterior humeral head subluxation and glenoid retroversion. The gradual loss of passive external rotation is a sign of impending glenohumeral dysplasia.15,21,32 Once external rotation becomes less than neutral with the scapula stabilized, glenohumeral deformity is evident in the majority of shoulders.14,21,32,33 Tendon transfers about the shoulder without open reduction have been shown to improve motion, but have a negligible effect on correction of any underlying deformity.5,15,18,30 At best, tendon transfers may halt the progression of deformity that is associated with muscular
Shoulder arthroscopy in children with brachial plexus birth palsy
107
Figure 3 A 6-year-old female (patient #3) with initial C5 and C6 injury. Spontaneous recovery began at 4 months. No previous surgery. Status post left arthroscopic anterior release and tendon transfers demonstrating Mallet parameters for shoulder function. (A) Abduction (grade IV); (B) hand to neck (grade IV from grade II preoperative); (C) hand to mouth (grade IV from grade III pre-operative); (D) hand to spine (grade II with no change from pre-operative).
imbalance. This inability to improve joint alignment may explain the loss in clinical improvement over time and raises concern about potential long-term joint sequelae.19 Open or arthoscopic anterior capsulectomy and
subscapularis release can reduce the glenohumeral joint and promote remodeling over time.11,22,24,26 However, the role of joint reduction remains controversial as complications have been reported.11,26 Open reduction is associated with
108
S.H. Kozin et al.
Figure 4 A 6-year-old female (patient #3). Pre- and postoperative MRIs following left shoulder arthroscopic anterior release and tendon transfers. (A) Pre-operative left shoulder with pseudeglenoid (PHHA ¼ 24.6%, version ¼ 37 ). (B) Postoperative left shoulder with marked improvement (PHHA ¼ 36.8%, version ¼ 4 ).
Table III
ANCOVA model results for abduction at one year
All subjects (n ¼ 42)
Means Squared
F-value
p-value
Transfer Preoperative ABD Preoperative PHHA abd)PHHA
8156 253 6750 4445
8.91 0.28 7.38 4.86
0.005 0.602 0.01 0.033
Abduction <100 (n ¼ 17) Transfer 4645.62 Preoperative PHHA 9376.05
4.25 8.59
0.0582 0.011
Abduction > 100 (n ¼ 27) Transfer 5467.85 Preoperative PHHA 17.85
6.67 0.02
0.016 0.8839
excessive external rotation and dramatic loss of internal rotation and midline function. Van der Sluijs et al26 performed open release in 19 children. Eight developed a ‘‘severe, functionally disturbing external rotation contracture of the shoulder.’’ However, open reduction has been shown to promote joint remodeling and thus remains a sensible treatment modality, especially in the young child.11 Pearl in 2003 published an arthroscopic technique for contracture and subscapularis release in children with brachial plexus birth palsy.20 The goal was to restore glenohumeral alignment and re-balance the joint using a less invasive procedure than formal open reduction. Forty-one children underwent arthroscopic release of the anterior capsule and subscapularis tendon. The mean age of the children was 3.5 years. Eighteen children were treated with arthroscopic release alone, while 23 children also underwent concomitant tendon transfer. The arthroscopic contracture release effectively restored passive
external rotation in 40 of the 41 children. The single patient that did not achieve external rotation was 12 years of age with advanced glenoid deformity. The status of the glenohumeral joint, however, was not evaluated after surgery. Pedowitz et al24 assessed the ability of arthroscopic release to reduce glenohumeral joint subluxation. Twentytwo children with an average age of 3.9 years underwent preoperative magnetic resonance imaging (MRI), arthroscopic surgery with or without tendon transfers, and postoperative imaging in their spica cast. Prior to surgery, the involved shoulder preoperative mean PHHA was 15.6% 13.5% and the mean glenoid version was e37 15 . After surgery and within the cast, the mean PHHA corrected to 46.9% 11.2% and the mean glenoid version improved to e8 8 (P < .001). The immediate improvement in glenoid version was primarily attributed to reduction of the humeral head from the pseudoglenoid onto the native glenoid and secondary to the fast remodeling and pliable pediatric cartilage.25 Pearl et al22 also reported follow-up on the first 33 children that underwent arthroscopic surgery. Nineteen children (all <3 years of age) underwent isolated arthroscopic release and 14 children (mean age, 6.7 years) underwent concomitant tendon transfer. Improved external rotation was noted in all children, except 1 child that was 12 years old. Passive external rotation increased between 60 and 80 dependent upon the procedure. Minimal improvement in mean active elevation was noted, even in the tendon transfer cohort. Four children (21%) that were treated with arthroscopic release alone required repeat surgery with an additional tendon transfer. Although internal rotation was not measured consistently prior to surgery, substantial reduction was note following surgery ranging from ‘‘moderate to severe.’’ MRI follow-up at 2
Shoulder arthroscopy in children with brachial plexus birth palsy years revealed ‘‘marked remodeling’’ in 12 out of 15 children with a pseudoglenoid deformity. Our report attempted to combine MRI and clinical findings in 44 patients. Twenty-eight children underwent isolated arthroscopic releases and 16 children underwent concomitant tendon transfers. In addition, a subgroup of 13 patients that underwent postoperative imaging in their spica cast was included to compare their current status of glenohumeral reduction.24 Similar to Pearl et al,22 we noted a significant improvement in external rotation both passive (26 to þ47 ) and active via the Mallet parameters that assess external rotation (external rotation, 2.1-3.7; hand to neck, 2.4-3.7; and hand to mouth, 2.1-3.4). In contrast to Pearl et al,22 we noted a significant improvement in active elevation (112 -130 ), especially in those patients undergoing concomitant tendon transfers compared to those children with isolated release (150 compared to 118 ). Our results of increased elevation may be multi-factorial, including patient selection and differences in surgical technique, immobilization, and rehabilitation. The Mallet score for abduction did not change; however, this scoring system does not discriminate once active elevation is greater than 90 . With regards to internal rotation, we previously noted a loss in internal rotation and midline function after complete arthrocopic release of the subscapularis24; therefore, we preserve the inferior and lateral muscular portion of the subscapularis. This technical modification can routinely be accomplished in younger children with less deformity and stresses the necessity of early diagnosis.14,24,32 Importantly, the Mallet parameter for internal rotation assesses the ability to place the hand on spine, and we noted no statistical change; however, our pre-operative score averaged 2.3, and a grade 2 implies inability to perform this motion. Clinically, the loss of internal rotation is seen by the loss of midline function, not the inability to place the hand on the spine. Recent reports of long-term results after tendon transfers report diminished motion over time. Kirkos et al13 report mean follow-up of 30 years in 10 patients, 8 of which had upper brachial plexus injuries. The authors assessed preoperative deformity on plain films only, and the tendon transfer technique was designed to improve external rotation and not abduction. Eight of the 10 patients had deterioration of external rotation between the 10- and 30-year follow-up. Gilbert et al6 reported long-term follow-up after isolated latissimus dorsi transfer. Children with isolated C5-C6 palsy had the greatest gains in abduction and external rotation; however, deterioration of abduction began at 6 years despite preservation of external rotation. In our series, we obtained and confirmed joint reduction and transferred both the latissimus dorsi and teres major tendons to improve abduction and external rotation. Nonetheless, longer-term follow-up is required to insure improvement in motion. We also acknowledge the weaknesses of our retrospective study, including the lack of randomization and relatively small sample size.
109
Conclusion Our results after arthroscopic release with or without tendon transfers are encouraging, with improvements in imaging studies and clinical evaluations following surgery. The glenoid version and PHHA statistically improved after surgery consistent with better glenohumeral joint alignment. Remarkably, 18 out of 19 children with a pseudoglenoid deformity corrected to a concentric or flat configuration at follow-up with resolution of the pseudoglenoid component. The clinical improvements paralleled the MRI corrections. Parameters for external rotation significantly improved after arthroscopic surgery with clinical enhancements in objective measures and functional Mallet scores. The addition of tendon transfers at the time of arthroscopic release resulted in even better active elevation; however, this combination risks impairment of midline function, which must be discussed with the family prior to surgery. Importantly, superior outcomes were associated with better preoperative clinical and MRI status. This indicates that early recognition of glenohumeral dysplasia and timely intervention will result in better shoulder motion and improved joint alignment. Our current practice is to begin passive shoulder range of motion shortly after birth to preserve external rotation. Passive range is combined with scapular stabilization to maximize glenohumeral joint motion. Loss of external rotation beyond neutral requires imaging to assess the glenohumeral joint. MRI findings consistent with an aligned joint warrants ongoing therapy, usually following botulinum toxin to the internal rotators. MRI findings consistent with a dysplastic joint, requires arthroscopic release. In children less than 3 years of age, isolated release is performed. In older children, arthroscopic release is usually combined with transfer of the latissimus dorsi and/ or teres major tendons dependent upon the strength of internal rotation.
References 1. Bae DS, Waters PM, Zurakowski D. Reliability of three classification systems measuring active motion in brachial plexus birth palsy. J Bone Joint Surg Am 2003;85:1733-8. 2. Beischer AD, Simmons TD, Torode IP. Glenoid version in children with obstetric brachial plexus palsy. J Pediatr Orthop 1999;19:359-61. 3. Bokor DJ, O’Sullivan MD, Hazan GJ. Variability of measurement of glenoid version on computed tomography scan. J Shoulder Elbow Surg 1999;8:595-8. 4. Dunkerton MC. Posterior dislocation of the shoulder associated with obstetric brachial plexus palsy. J Bone Joint Surg Br 1989;71: 764-6. 5. El-Gammal TA, Saleh WR, El-Sayed A, Kotb MM, Imam HM, Fathi NA. Tendon transfer around the shoulder in obstetric brachial
110
6. 7.
8. 9.
10.
11.
12. 13.
14.
15.
16. 17.
18.
19.
20.
plexus paralysis: Clinical and computed tomographic study. J Pediatr Orthop 2006;26:641-6. Gilbert A, Whitaker I. Obstetrical brachial plexus lesions. J Hand Surg [Br] 1991;16:489-91. Greenwald AG, Schute PC, Shiveley JL. Brachial plexus birth palsy: a 10-year report on the incidence and prognosis. J Pediatr Orthop 1984;4:689-92. Hardy AE. Birth injuries of the brachial plexus: incidence and prognosis. J Bone Joint Surg Br 1981;63:98-101. Hoffer MM, Phipps GJ. Closed reduction and tendon transfer for treatment of dislocation of the glenohumeral joint secondary to brachial plexus birth palsy. J Bone Joint Surg Am 1998;80:997-1001. Hoffer MM, Wickenden R, Roper B. Brachial plexus birth palsies. Results of tendon transfers to the rotator cuff. J Bone Joint Surg Am 1978;60:691-5. Hui JH, Torode IP. Changing glenoid version after open reduction of shoulders in children with obstetric brachial plexus palsy. J Pediatr Orthop 2003;23:109-13. Jackson ST, Hoffer MM, Parrish N. Brachial-plexus palsy in the newborn. J Bone Joint Surg Am 1988;70:1217-20. Kirkos JM, Kyrkos MJ, Kapetanos GA, Haritidis JH. Brachial plexus palsy secondary to birth injuries. Long-term results of anterior release and tendon transfers around the shoulder. J Bone Joint Surg [Br] 2005; 87:3231-5. Kozin SH. Correlation between external rotation of the glenohumeral joint and deformity after brachial plexus birth palsy. J Pediatr Orthop 2004;24:189-93. Kozin SH, Chafetz RS, Barus D, Filipone L. MRI and clinical findings before and after tendon transfers about the shoulder in children with residual brachial plexus birth palsy. J Shoulder Elbow Surg 2006;15: 554-61. Mintzer GM, Waters PM, Brown DJ. Glenoid version in children. J Pediatr Orthop 1996;6:563-6. Moukoko D, Ezaki M, Wilkes D, Carter P. Posterior shoulder dislocation in infants with neonatal brachial plexus palsy. J Bone Joint Surg Am 2004;86:787-93. Newman CJ, Morrison L, Lynch B, Hynes D. Outcome of subscapularis muscle release for shoulder contracture secondary to brachial plexus palsy at birth. J Pediatr Orthop 2006;26:647-51. Pagnotta A, Haerle M, Gilbert A. Long-term results on abduction and external rotation of the shoulder after latissimus dorsi transfer for sequelae of obstetric palsy. Clin Orthop 2004;426:199-205. Pearl ML. Arthroscopic release of shoulder contracture secondary to birth palsy: an early report on findings and surgical technique. Arthroscopy 2003;19:577-82.
S.H. Kozin et al. 21. Pearl ML, Edgerton BW. Glenoid deformity secondary to brachial plexus birth palsy. J Bone Joint Surg Am 1998;80:659-67. 22. Pearl ML, Edgerton BW, Kazimiroff PA, Burchette RJ, Wong K. Arthroscopic release and latissimus dorsi transfer for shoulder internal rotation contractures and glenohumeral deformity secondary to brachial plexus birth palsy. J Bone Joint Surg Am 2006;88:564-74. 23. Pearl ML, Edgerton BW, Kon DS, Darakjian AB, Kosco AE, Kazimiroff PB, et al. Comparison of arthroscopic findings with magnetic resonance imaging and arthrography in children with glenohumeral deformities secondary to brachial plexus birth palsy. J Bone Joint Surg [Am] 2003;85:890-8. 24. Pedowitz DI, Gibson B, Williams GR, Kozin SH. Arthroscopic treatment of posterior glenohumeral joint subluxation. J Shoulder Elbow Surg 2007;16:6-13. 25. Pirini S, Zeznik L, Hodges D. Magnetic resonance imaging study of congenital clubfoot treated with the Ponseti method. J Ped Orthop 2001;21:719-26. 26. van der Sluijs JA, van Ouwerkerk WJ, de Gast A, Nollet F, Winters H, Wuisman PI. Treatment of internal rotation contracture of the shoulder in obstetric brachial plexus lesions by subscapularis tendon lengthening and open reduction: early results and complications. J Pediatric Orthop B 2004;13:218-24. 27. van der Sluijs JA, van Ouwerkerk WJ, de Gast A, Wuisman PI, Nollet F, Manoliu RA. Retroversion of the humeral head in children with an obstetric brachial plexus lesion. J Bone Joint Surg Br 2002;84: 583-7. 28. van der Sluijs JA, van Ouwerkerk WJ, de Gast A, Wuisman PI, Nollet F, Manoliu RA. Deformities of the shoulder in infants younger than 12 months with an obstetric lesion of the brachial plexus. J Bone Joint Surg Br 2001;83:551-5. 29. van der Sluijs JA, van der Meij, Verbeke J, Manoliu RA, Wuisman PI. Measuring secondary deformities in children with obstetric brachial plexus lesion: reliability of three methods. J Pediatric Orthop B 2003; 12:211-4. 30. Waters PM, Bae DS. Effect of tendon transfers and extra-articular softtissue balancing on glenohumeral development in brachial plexus birth palsy. J Bone Joint Surg [Am] 2005;87:320-5. 31. Waters PM, Peljovich AE. Shoulder reconstruction in patients with chronic brachial plexus birth palsy. A case control study. Clin Orthop 1999;364:144-52. 32. Waters PM, Smith GR, Jaramillo D. Glenohumeral deformity secondary to brachial plexus birth palsy. J Bone Joint Surg Am 1998; 80:668-77. 33. Zancolli EA, Zancolli ER Jr. . Palliative surgical procedures in sequelae of obstetric palsy. Hand Clin 1988;4:643-69.