JBJS | Glenohumeral Deformity Secondary to Brachial Plexus Birth Palsy

The Journal of Bone & Joint Surgery, Volume 80, Issue 5

Articles | May 01, 1998

Glenohumeral Deformity Secondary to Brachial Plexus Birth Palsy

PETER M. WATERS, M.D.†; GARTH R. SMITH, M.D.†; DIEGO JARAMILLO, M.D.†, BOSTON, MASSACHUSETTS

View Disclosures and Other Information

Investigation performed at Children's Hospital, Harvard Medical School, Boston

The Journal of Bone & Joint Surgery. 1998; 80:668-77

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

text A A A

Abstract

Ninety-four patients who had brachial plexus birth palsy were entered into a prospective study to evaluate the association between persistent palsy, age-related musculoskeletal deformity, and functional limitations. Of these patients, forty-two had either computerized tomography or magnetic resonance imaging to assess the presence and degree of incongruity of the glenohumeral joint, deformity of the humeral head, and hypoplasia of the glenoid as part of the preoperative planning for a reconstructive operation. Functional ability was rated with use of the classification of Mallet, on a scale of 1 to 5.The mean glenoscapular angle (the degree of retroversion of the glenoid) on the affected side was -25.7 degrees compared with -5.5 degrees on the unaffected side. Twenty-six (62 per cent) of the forty-two shoulders had evidence of posterior subluxation of the humeral head, with a mean of only 25 per cent (range, 0 to 50 per cent) of the head being intersected by the scapular line. Progressive deformity was found with increasing age (p < 0.001).The natural history of untreated brachial plexus birth palsy with residual weakness is progressive glenohumeral deformity due to persistent muscle imbalance. The status of the glenohumeral joint must be addressed when the choice between tendon transfer and humeral derotation osteotomy for reconstruction of the shoulder is considered for these patients.

Figures in this Article

Introduction

Introduction | Materials and Methods | Results | Discussion | References

Although many patients who have brachial plexus birth palsy can be expected to have complete or nearly complete recovery of motor strength about the shoulder^9,11,14,25, some have persistent weakness. Most commonly, this consists of weakness of the upper part of the trunk with limited abduction and external rotation of the shoulder. Operative procedures to improve the function of the shoulder currently consist of either tendon transfer or humeral osteotomy, but there are no clear guidelines for the selection of patients. There also are no specific recommendations regarding the timing of the operative intervention or the choice of the procedure. Although glenohumeral deformity has been recognized previously in patients who have long-standing muscle imbalance, no investigation, to our knowledge, has specifically related the patient's age or muscle function to the extent of the deformity.

The purposes of the current study were to evaluate the severity of the glenohumeral deformity and the functional limitations of patients who had residual brachial plexus birth palsy; to determine, with use of computerized tomography or magnetic resonance imaging, if there was a relationship between the patient's age and the specific deformity; and to assess the value of these two imaging modalities in helping the surgeon to decide whether tendon transfer or humeral osteotomy is the appropriate operative procedure.

*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.

†Departments of Orthopedic Surgery (P. M. W. and G. R. S.) and Radiology (D. J.), Children's Hospital, 300 Longwood Avenue, Boston, Massachusetts 02115. The e-mail address for Dr. Waters is waters@a1.harvard.tch.edu. Please address requests for reprints to Dr. Waters.

*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1 Schematic drawing showing the method of measuring the glenoscapular angle (glenoid vision) and the percentage of posterior subluxation of the humeral head. To measure the glenoscapular angle, a line is drawn parallel to the scapula and a second line is drawn tangential to the joint. The latter line connects the anterior and posterior margins of the glenoid. On magnetic resonance imaging scans, the cartilaginous margins are used. On computerized tomographic scans, the osseous margins are used. The intersecting line connects the center point of the first line (approximately the middle of the glenoid fossa) and the medial aspect of the scapula. The angle in the posterior medial quadrant is measured with a goniometer (arrow). Ninety degrees then is subtracted from this measurement to determine glenoid version. The percentage of posterior subluxation is measured by defining the percentage of the humeral head that is anterior to the same scapular line. The greatest circumference of the head is measured as the distance from the scapular line to the anterior portion of the head. This ratio (the distance to the anterior aspect of the humeral head [AB] divided by the circumference of the humeral head [AC], multiplied by 100) is the percentage of subluxation.

View Large | Download Slide(.ppt)
Add to My JBJS

Figs. 2-A through 2-F: Radiographic types as determined with use of computerized tomographic or magnetic resonance imaging scans. Fig. 2-A: Magnetic resonance imaging scan of a type-I (normal) glenohumeral joint (less than a 5-degree difference in glenoid version compared with that on the normal, contralateral side). The bisecting line extending from the spine of the scapula through the humeral head is outlined. The angle in the posterior medial quadrant is indicated by the arrow. Ninety degrees is subtracted from this measurement to determine glenoid version.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-B Computerized tomographic scan of a glenohumeral joint with type-II deformity (on the right). The deformity is minimum (more than a 5-degree difference in glenoid version compared with that on the normal, contralateral side and no evidence of posterior subluxation of the humeral head).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-C Magnetic resonance imaging scan of a glenohumeral joint with type-III deformity. There is moderate deformity of the glenoid with posterior subluxation of the humeral head (less than 35 per cent of the head is anterior to the scapular line). The scapular line and the tangential line indicating the anterior and posterior cartilaginous margins of the glenoid are shown. The angle in the posterior medial quadrant is indicated by the arrow.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-D Magnetic resonance imaging scan of a glenohumeral joint with type-IV deformity. There is progressive deformity (a false glenoid) and subluxation.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-E Computerized tomographic scan of a glenohumeral joint with grade-V deformity (on the right). There is associated progressive deformity of the humeral head with flattening.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-F Computerized tomographic scan of a glenohumeral joint with type-VI deformity (on the right). There is infantile dislocation of the humeral head.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 3 Graph showing increasing radiographic deformity (progressive glenoscapular retroversion and posterior subluxation of the humeral head) with increasing age. The most severe deformity (type V) is noted in adolescents.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 4 Graph showing the glenoscapular angles (retroversion), according to age. The angle is increased in older children, who have longer-standing muscle imbalance.

View Larger

TABLE I THE MALLET CLASSIFICATION¹⁶

*A trumpet sign consists of abduction of the shoulder with simultaneous flexion of the elbow.

Functional Parameter	Class 1	Class 2	Class 3	Class 4	Class 5
Global abduction	None	<30 degrees	30—90 degrees	>90 degrees	Normal
External rotation	None	<0 degrees	0—20 degrees	>20 degrees	Normal
Hand-to-neck ability	None	Not possible	Difficult	Easy	Normal
Hand-to-mouth ability	None	Marked trumpet sign*	Partial trumpet sign*	<40 degrees of abduction	Normal
Internal rotation	None	Not possible	To S1	To T12	Normal

Add to My JBJS

TABLE I THE MALLET CLASSIFICATION¹⁶

*A trumpet sign consists of abduction of the shoulder with simultaneous flexion of the elbow.

Functional Parameter	Class 1	Class 2	Class 3	Class 4	Class 5
Global abduction	None	<30 degrees	30—90 degrees	>90 degrees	Normal
External rotation	None	<0 degrees	0—20 degrees	>20 degrees	Normal
Hand-to-neck ability	None	Not possible	Difficult	Easy	Normal
Hand-to-mouth ability	None	Marked trumpet sign*	Partial trumpet sign*	<40 degrees of abduction	Normal
Internal rotation	None	Not possible	To S1	To T12	Normal

View Larger

TABLE II MALLET CLASSIFICATION FOR THE FORTY-TWO PATIENTS*

Case	Functional Parameter
Global Abduction	External Rotation	Hand-to- Neck Ability	Hand-to- Mouth Ability
1	2	2	2	3
2	1	1	1	1
3	3	3	2	3
4	2	2	2	2
5	3	2	2	2
6	3	2	2	3
7	4	3	4	4
8	3	2	3	2
9	4	3	4	4
10	4	2	3	4
11	3	2	2	2
12	3	2	2	2
13	3	2	2	2
14	4	3	4	4
15	2	2	2	2
16	3	2	2	2
17	3	3	3	3
18	3	1	2	2
19	1	2	2	2
20	4	2	4	4
21	2	3	3	3
22	3	2	2	1
23	4	2	3	2
24	3	2	3	2
25	4	2	3	4
26	3	2	2	2
27	4	2	2	2
28	3	2	2	2
29	4	2	3	2
30	3	2	2	3
31	3	2	3	2
32	3	2	2	2
33	3	2	2	2
34	3	2	2	2
35	3	2	2	2
36	2	2	2	1
37	3	2	2	3
38	3	2	2	2
39	3	2	2	2
40	4	2	3	3
41	3	2	3	2
Mean	3.0	2.1	2.4	2.4

Add to My JBJS

TABLE II MALLET CLASSIFICATION FOR THE FORTY-TWO PATIENTS*

Case	Functional Parameter
Global Abduction	External Rotation	Hand-to- Neck Ability	Hand-to- Mouth Ability
1	2	2	2	3
2	1	1	1	1
3	3	3	2	3
4	2	2	2	2
5	3	2	2	2
6	3	2	2	3
7	4	3	4	4
8	3	2	3	2
9	4	3	4	4
10	4	2	3	4
11	3	2	2	2
12	3	2	2	2
13	3	2	2	2
14	4	3	4	4
15	2	2	2	2
16	3	2	2	2
17	3	3	3	3
18	3	1	2	2
19	1	2	2	2
20	4	2	4	4
21	2	3	3	3
22	3	2	2	1
23	4	2	3	2
24	3	2	3	2
25	4	2	3	4
26	3	2	2	2
27	4	2	2	2
28	3	2	2	2
29	4	2	3	2
30	3	2	2	3
31	3	2	3	2
32	3	2	2	2
33	3	2	2	2
34	3	2	2	2
35	3	2	2	2
36	2	2	2	1
37	3	2	2	3
38	3	2	2	2
39	3	2	2	2
40	4	2	3	3
41	3	2	3	2
Mean	3.0	2.1	2.4	2.4

View Larger

TABLE III DATA WITH REGARD TO RADIOGRAPHIC TYPE

*The values are given as the mean. †This measurement was determined by the percentage of the humeral head that was anterior to the bisecting (scapular) line. A value of 50 per cent indicates a centered humeral head and a value of 0, complete posterior dislocation.

Radio- graphic Type	No. of Patients	Age* (yrs.)	Glenoscapular Angle* (degrees)		Posterior Subluxation† (per cent)		Mallet Class*¹⁶
Affected Side	Normal Side	Affected Side	Normal Side	Global Abduction	External Rotation	Hand-to- Neck Ability	Hand-to- Mouth Ability
I	6	3.0	-6	-5	45	50	2.2	2.0	1.8	2.2
II	13	4.8	-16	-4	42	50	3.2	2.2	2.7	2.8
III	4	3.7	-29	-5	17	50	3.0	2.2	2.7	2.5
IV	8	3.4	-37	-9	9	49	3.5	2.0	2.4	2.5
V	4	11.3	-47	-6	4	48	3.0	2.0	2.2	2.0
VI	5	1.9	-46	-5	1	47	2.8	2.0	2.0	2.0
VII	2	13.5	-5	-2	40	50	3.5	2.0	3.0	2.5

Add to My JBJS

TABLE III DATA WITH REGARD TO RADIOGRAPHIC TYPE

Radio- graphic Type	No. of Patients	Age* (yrs.)	Glenoscapular Angle* (degrees)		Posterior Subluxation† (per cent)		Mallet Class*¹⁶
Affected Side	Normal Side	Affected Side	Normal Side	Global Abduction	External Rotation	Hand-to- Neck Ability	Hand-to- Mouth Ability
I	6	3.0	-6	-5	45	50	2.2	2.0	1.8	2.2
II	13	4.8	-16	-4	42	50	3.2	2.2	2.7	2.8
III	4	3.7	-29	-5	17	50	3.0	2.2	2.7	2.5
IV	8	3.4	-37	-9	9	49	3.5	2.0	2.4	2.5
V	4	11.3	-47	-6	4	48	3.0	2.0	2.2	2.0
VI	5	1.9	-46	-5	1	47	2.8	2.0	2.0	2.0
VII	2	13.5	-5	-2	40	50	3.5	2.0	3.0	2.5

Materials and Methods

Introduction | Materials and Methods | Results | Discussion | References

Between 1989 and 1995, ninety-four patients were seen by the senior one of us (P. M. W.) at Children's Hospital in Boston for the evaluation of brachial plexus birth palsy. All patients were enrolled in a prospective evaluation at the time of presentation. Forty-two of these patients had persistent weakness of the upper part of the trunk and functional limitations of the shoulder, and they were evaluated with either computerized tomography or magnetic resonance imaging as part of the preoperative planning for possible reconstruction of the shoulder. Twenty-three patients were girls and nineteen were boys; their mean age was 5.1 years (range, nine months to fourteen years). The left shoulder was affected in nineteen patients and the right shoulder, in twenty-three.

Twenty-six patients were evaluated with computerized tomography and sixteen, with magnetic resonance imaging. Initially, the preoperative protocol had called for computerized tomography; however, with improvements in magnetic resonance imaging for evaluation of the articular cartilage surfaces of the glenoid and the humeral head¹⁰, this technique came to be preferred for children who are less than five years of age because these structures are still incompletely ossified at this age.

The protocol for each modality included the imaging of both shoulders to facilitate comparison with the normal anatomy. High-resolution (four-millimeter-thick) axial and oblique coronal gradient-recalled echo images of the affected joint were made with magnetic resonance imaging, and three-millimeter-thick axial slices with two and three-dimensional reconstruction were made with computerized tomography. During the last year of the study, imaging was performed in a helical fashion, with reconstruction at one-millimeter intervals. Both imaging techniques were used to assess the presence and degree of incongruity of the glenohumeral joint, deformity of the humeral head, and hypoplasia of the glenoid.

Radiographic Evaluation

The magnetic resonance images of the children who were less than five years of age showed that, on the unaffected side, the normal glenoid was mostly cartilaginous and thus of high signal intensity. A normal physis was readily seen at the chondro-osseous junction. The glenoid labrum was of low signal intensity, similar to that of an adult.

On each scan, the glenoscapular angle (the degree of version of the glenoid) was measured on both sides for comparison. The technique described by Friedman et al.⁶ and Randelli and Gambrioli²³ was used to measure glenoid version both on computerized tomographic and magnetic resonance imaging scans. A line connecting the most lateral margin of the posterior aspect of the glenoid and the most medial margin of the anterior aspect of the glenoid was drawn. On the magnetic resonance imaging scans of the younger patients, the cartilaginous surfaces were used as the margins even though the osseous margins would have indicated more retroversion. On the computerized tomographic scans of the older patients, the osseous margins were used. A bisecting line connecting the center point of the first line (approximately the middle of the glenoid fossa) and the medial aspect of the scapula was constructed. The angle in the posteromedial quadrant was measured with a goniometer (Fig. 1). Ninety degrees then was subtracted from this measurement to determine the glenoscapular angle (glenoid version). A negative value indicated that the glenoid was retroverted, whereas a positive value indicated that it was anteverted¹⁸. The same scapular line was used to determine the percentage of the humeral head that was anterior to it in order to quantitate the degree of posterior subluxation of the head. The greatest circumference of the head was measured, as was the distance from the bisecting scapular line to the anterior portion of the head. This ratio (the distance to the anterior aspect of the humeral head divided by the circumference of the head, multiplied by 100) was the percentage of subluxation (Fig. 1).

Glenohumeral deformity and subluxation were graded radiographically. Type I indicated a normal glenoid (less than a 5-degree difference in retroversion compared with that on the normal, contralateral side [Fig. 2-A]); type II, minimum deformity (more than a 5-degree difference in retroversion compared with that on the normal side, with no posterior subluxation of the humeral head [Fig. 2-B]); type III, moderate deformity (posterior subluxation of the humeral head, defined as less than 35 per cent of the head anterior to the bisecting line [Fig. 2-C); type IV, severe deformity (a false glenoid [Fig. 2-D]); type V, severe flattening of the humeral head and glenoid, with progressive or complete posterior dislocation of the head (Fig. 2-E); type VI, dislocation of the glenohumeral joint in infancy (Fig. 2-F); and type VII, growth arrest of the proximal aspect of the humerus.

In the abnormal shoulders, there was thinning of the posterior aspect of the glenoid, which was poorly defined or not demonstrable. In the more severely affected shoulders, the glenoid cavity was poorly developed posteriorly and the posterior aspect of the labrum and the humeral head were posteriorly displaced (Fig. 2-C). The osseous glenoid was nearly normal in these young patients. On coronal images, eleven of twelve patients had caudad orientation of the glenoid. There was thinning of the superior aspect of the glenoid cartilage in the eight most severely affected patients and marked atrophy of the muscles of the shoulder girdle in all patients.

Computerized tomographic images of the patients who were more than five years of age demonstrated more severe deformity of the posterior aspect of the osseous glenoid (Fig. 2-E). In some instances, the articular surface became convex. The humeral head was less well developed, and it often was deformed in adolescent patients.

Functional Evaluation

On enrollment in the study, the forty-two patients were given a detailed physical examination, which included documentation of the time when the function of the biceps returned, strength-testing of all muscle groups in both upper extremities (with a grade of 1 to 5 being assigned), recording of the passive and active range of motion, and determination of the functional active range of motion of the shoulder and elbow joints. The functional ability of the involved shoulder, including global abduction, external rotation, the ability to bring the hand to the neck and to the mouth, and internal rotation, was graded with use of the Mallet classification system¹⁶. This system assigns a score (class number) of 1 to 5 for each of the five functional parameters, with class 1 indicating no function and class 5, normal function (Table I).

The data regarding the patient's age, the degree of glenohumeral deformity, and the functional status were analyzed with use of a simple association matrix as well as an analysis of variance.

Twenty-seven of the forty-two patients subsequently had operative intervention. Fifteen patients with a type-I, II, or III radiographic deformity had a latissimus dorsi and teres major tendon transfer (mean duration of follow-up, three years and three months; range, two to seven years), eight patients with a type-IV or V deformity had a humeral derotation osteotomy (mean duration of follow-up, six years; range, three to ten years), and four patients with a type-VI deformity had an open reduction and capsulorrhaphy for the treatment of infantile dislocation. The remaining fifteen patients were still being followed at the time of writing.

Results

Introduction | Materials and Methods | Results | Discussion | References

The forty-two patients had marked functional limitations, with lower-than-normal mean Mallet scores for global abduction (3.0), external rotation (2.1), and the ability to bring the hand to the neck (2.4) and to the mouth (2.4) (Table II). The internal rotation component of the Mallet system was not routinely recorded in this series and thus was not included in the statistical analysis.

Although the major purpose of this paper is to report the classification of the glenohumeral deformities, and although we realize that the postoperative results are preliminary, these patients had marked improvement in the Mallet functional scores. The patients who had had a latissimus dorsi and teres major tendon transfer had improvement in the mean scores for global abduction (from 2.9 to 3.9), external rotation (from 2.0 to 3.9), and the ability to bring the hand to the neck and to the mouth (from 2.2 to 4.0 for both parameters). Similar improvement was noted in the mean scores of the patients who had had a humeral osteotomy²⁷, with global abduction improving from 3.0 to 4.0; external rotation, from 2.0 to 3.7; and the ability to bring the hand to the neck and to the mouth, from 2.3 to 4.0 and from 2.1 to 3.6, respectively.

Radiographic Findings

Radiographically, six deformities were classified as type I; thirteen, as type II; four, as type III; eight, as type IV; four, as type V; five, as type VI; and two, as type VII. The mean ages of the patients who had type-I through type-IV deformities were comparable, but these patients were younger, on the average, than those who had a type-V or VII deformity (growth arrest). The mean age of the children who had a type-VI deformity (infantile dislocation) was less than two years (Table III).

Glenoid retroversion (the mean glenoscapular angle) on the affected side was markedly increased compared with that on the normal side (mean, 25 compared with 5 degrees). There was progressive retroversion with increasing subluxation (Table III).

Twenty-six (62 per cent) of the forty-two patients had evidence of posterior subluxation of the humeral head, with a mean of 25 per cent (range, 0 to 50 per cent) of the head being anterior to the intersecting scapular line compared with 49 per cent (range, 40 to 55 per cent) in the normal shoulders. There was increasing subluxation with increasing deformity. A mean of 45 per cent (range, 35 to 50 per cent) of the head was anterior to the intersecting scapular line in patients who had radiographic type-I deformity, 42 per cent (range, 35 to 50 per cent) in those who had type-II deformity, 17 per cent (range, 0 to 33 per cent) in those who had type-III deformity, 9 per cent (range, 0 to 30 per cent) in those who had type-IV deformity, 4 per cent (range, 0 to 11 per cent) in those who had type-V deformity, and 1 per cent (range, 0 to 5 per cent) in those who had type-VI deformity (Table III). The mean was 40 per cent (30 and 50 per cent) in the two patients who had type-VII deformity.

The Mallet functional assessment score was averaged according to each of the seven radiographic types. All types were associated with marked functional deficits in global abduction, external rotation, and the ability to bring the hand to the neck and to the mouth (Table III).

Each radiographic type was analyzed with regard to age, Mallet score, and radiographic changes. Because the radiographic changes associated with type-VI deformity (infantile dislocation) or type-VII deformity (growth arrest) may be due to birth trauma and thus may not be secondary to persistent muscle imbalance, these patients were not considered in the age-related statistical analysis. In subgroups I through V, there was a significant association between increasing age and progressive deformity (radiographic type) (p < 0.001) (Fig. 3). The most severe deformity (type V) was noted in the older patients (mean age, 11.3 years) and was associated with the adolescent growth spurt. Similarly, there was a significant association between age and the glenoscapular angle (glenoid retroversion) (p < 0.04) (Fig. 4). There also was a trend, although it was not significant (p < 0.16) with the numbers available, toward an increase in subluxation of the humeral head with an increase in age.

The various radiographic parameters and the patients' ages were independently analyzed with regard to the functional parameters of the Mallet classification system. There was a significant association between global abduction and the glenoscapular angle (p < 0.01) and between global abduction and the radiographic type (p < 0.03). The other Mallet functional parameters were not found to be associated with age or with any of the radiographic parameters, with the numbers available. However, this was a preselected group of preoperative patients, all of whom had had poor Mallet scores at the time of the original evaluation.

Discussion

Introduction | Materials and Methods | Results | Discussion | References

As shown by previous studies, most patients with brachial plexus birth palsy who begin to recover in the first three months of life can be expected to have improvement to nearly normal function^{9,11,14,17,25}. However, those who have delayed recovery often fail to regain normal muscle strength and motion. The primary issues with regard to these patients concern the indications for an operation, the timing of such intervention, and the choice of procedure. The choices that are currently available include microsurgical nerve reconstruction in infants and secondary reconstruction with tendon transfers or osteotomy. There is still controversy concerning the optimum treatment of residual weakness and functional limitation^{1,4,7,8,12,20,22,30}. Knowledge of the natural history of brachial plexus birth palsy in children who have residual weakness is incomplete. In addition, we are aware of no randomized, prospective studies, in which the functional abilities of untreated patients have been compared with those of patients who have been managed operatively.

Computerized tomography has been used previously as a means of identifying posterior dislocation of the humeral head in patients who have brachial plexus birth palsy¹³. However, it has not been used to evaluate the severity of the glenohumeral subluxation or deformity, to provide a classification of these deformities, or to aid in decision-making. Magnetic resonance imaging has been employed more recently because of its usefulness in assessing the growth and articular cartilage surfaces of the glenoid and the humeral head^10,18.

The current study was designed to better define the natural history of glenohumeral deformity in children who have residual brachial plexus palsy and to provide better guidelines for the selection of a particular operative procedure. As would be expected in growing children who have persistent muscle imbalance, this study demonstrated various degrees of glenohumeral deformity.

Various skeletal deformities, such as glenoid hypoplasia, posterior subluxation and flattening of the humeral head, and hooking of the coracoid, have been described previously^{2,3,5,15,21,24}; however, these descriptions were based on plain radiographs and did not include information regarding the progression of the deformity. Furthermore, when operative intervention has been contemplated, the choice between tendon transfer and humeral osteotomy according to the degree of osseous deformity of the glenoid and the humeral head has not been clear. The scant information that is available has emphasized the presence or absence of deformity as a factor in decision-making. In younger children, these changes are confined to the posterior articular and physeal cartilage of the glenoid. The physeal abnormality of the glenoid as seen on magnetic resonance images suggests an impairment of growth of the posterior aspect of the glenoid. As seen in the current study, there are degrees of deformity of the posterior aspect of the glenoid, the glenohumeral joint, and the humeral head; it is not simply a matter of these deformities being present or absent. The operative decision may be clear at either end of the spectrum: patients who have type-I or II changes may be managed with a tendon transfer and those who have type-V changes may be managed with a humeral osteotomy. However, the intermediate types of deformity pose a more difficult problem because of the possibility of remodeling of the glenoid and the humeral head. Although joint-remodeling in children has been well studied in the hip, it has never been examined prospectively in the shoulder, to our knowledge. At present, it is impossible to predict accurately how much remodeling is possible in a given patient after reconstruction of the shoulder. A full discussion of the theoretical benefits of each operative procedure is beyond the scope of the present study, but examination of the deformity both before and after reconstruction to better define joint-remodeling seems beneficial. The development of a reproducible system for the grading of patients preoperatively also is important in the design of future prospective studies. Computerized tomography and magnetic resonance imaging both provide a reliable, reproducible means of evaluating the status of the glenohumeral joint and are valuable in the preoperative planning and postoperative follow-up of these patients. Magnetic resonance imaging appears to offer the most information regarding the physeal changes of the glenoid in younger patients, whereas the lower cost and the ease of multiplanar reconstruction make computerized tomography the preferable modality for older patients.

In examining the spectrum of deformity, we found a striking association between increased age and type-V changes (that is, severe deformity). The age-group in which this type of deformity was found corresponds to the adolescent growth spurt of the glenoid and the humeral head. By the time that type-V deformity is present, an osteotomy as a salvage procedure is the only option. In younger patients who have minimum (type-II or III) glenohumeral deformity, anterior release of the pectoralis major and transfer of the latissimus dorsi and teres major muscles to the insertion of the rotator cuff may correct the muscle imbalance. If done early enough, this may result in reconstruction of the glenohumeral joint. Whether glenoid osteotomy has a role in the treatment of type-III or IV deformity is unclear.

We could not detect a significant association between most of the parameters that were tested and the Mallet scores. (The only relationships that were found were between global abduction and the glenoscapular angle [p < 0.01] and between global abduction and the radiographic type [p < 0.03].) This was probably due to bias in patient selection and to the relatively small number of patients in the study. All forty-two patients had severe functional deficits and thus were considered to be candidates for an operation. Patients who had less severe deficits were not evaluated with computerized tomography or magnetic resonance imaging. The use of these modalities was a part of preoperative planning. The patients with type-I, II, or III deformity had a latissimus dorsi and teres major tendon transfer, those with type-IV or V deformity had a humeral osteotomy, and those with type-VI deformity had an open reduction and capsulorrhaphy of the dislocated glenohumeral joint. Whether the primary cause of the functional deficit was motor weakness or progressive osseous deformity due to the long-standing muscle weakness was difficult to determine, but probably the deficit was caused by a combination of these factors. Because the Mallet functional scores for all of our patients were relatively poor, this scoring system was not sufficiently sensitive to differentiate subtle differences in this series. In addition, there were more patients in the younger age-groups, which limited differentiation between radiographic groups with regard to age and function. Furthermore, the Mallet classification has its own limitations. The system is a measure of so-called functional ability, which is related to many parameters, including muscle strength, the active and passive range of motion, and the presence of contracture or osseous deformity. Also, it may be difficult to evaluate an uncooperative or young child's level of function in a reproducible manner. Although the Mallet classification is used frequently, it has not yet been validated for precision or accuracy.

The patients with type-VI changes (dislocation of the shoulder in infancy) represent a unique population. There are probably two subgroups of patients with this deformity: those who sustained it due to trauma during labor (in whom it may be noted at birth) and those who have severe muscle imbalance leading to early dislocation (in whom it may be noted by the age of six to nine months). The exact cause of the dislocation may be difficult to determine in a given patient and may be unclear depending on when the patient was first evaluated. Troum et al. recently discussed the theories regarding this injury; the most widely held theory was that severe muscle imbalance leads to early dislocation²⁶. Our study lends support to this hypothesis.

In the remote past, patients who had infantile dislocation were managed with resection of the humeral head¹⁹ or closed manipulation²⁸ and a cast³⁰. Disappointing results led to use of the subscapular and anterior capsular release described by Fairbank⁵ and to modification of the Fairbank procedure by Wickstrom et al., to include fixation with a Steinmann pin²⁹. More recently, Troum et al. advocated open reduction with use of a combined anterior release and posterior capsulorrhaphy²⁶.

It is important to be aware of the possibility of glenohumeral dislocation in infants. Severe osseous deformity may develop quickly in these patients, as it did in those who had profound glenoid deformity in the current series. Thus, the period of time in which an open reduction may be feasible is markedly limited.

In conclusion, our study with computerized tomography or magnetic resonance imaging as well as analysis of functional parameters demonstrated an association between persistent palsy and glenohumeral deformity in patients who had residual functional deficits secondary to persistent brachial plexus birth palsy. We believe that the information provided by these imaging modalities is useful in preoperative planning for these patients. Magnetic resonance imaging allows visualization of the articular and physeal cartilage and is currently our preferred tool for the imaging of younger patients. Computerized tomography is used in patients who are more than five years of age.

References

Introduction | Materials and Methods | Results | Discussion | References

Adler, J. B., and Patterson, R. L.: Erb's palsy. Long-term results of treatment in eighty-eight cases. J. Bone and Joint Surg.,49-A: 1052-1064, Sept. 1967.49-A1052 1967

Babbitt, D. P., and Cassidy, R. H.: Obstetrical paralysis and dislocation of the shoulder in infancy. J. Bone and Joint Surg.,50-A: 1447-1452, Oct. 1968.50-A1447 1968

Dunkerton, M. C.: Posterior dislocation of the shoulder associated with obstetric brachial plexus palsy. J. Bone and Joint Surg.,71-B(5): 764-766, 1989.71-B(5)764 1989

Egloff, D. V.; Raffoul, W.; Bonnard, C.; and Stadler, J.: Palliative surgical procedures to restore shoulder function in obstetric brachial palsy. Hand Clin.,11: 597-606, 1995.11597 1995 [PubMed]

Fairbank, H. A. T.: Birth palsy: subluxation of the shoulder-joint in infants and young children. Lancet,1: 1217-1223, 1913.11217 1913

Friedman, R. J.; Hawthorne, K. B.; and Genez, B. M.: The use of computerized tomography in the measurement of glenoid version. J. Bone and Joint Surg.,74-A: 1032-1037, Aug. 1992.74-A1032 1992

Gilbert, A.; Brockman, R.; and Carlioz, H.: Surgical treatment of brachial plexus birth palsy. Clin. Orthop.,264: 39-47, 1991.26439 1991 [PubMed]

Gilbert, A.; Razaboni, R.; and Amar-Khodja, S.: Indications and results of brachial plexus surgery in obstetrical palsy. Orthop. Clin. North America,19: 91-105, 1988.1991 1988

Greenwald, A. G.; Schute, P. C.; and Shiveley, J. L.: Brachial plexus birth palsy: a 10-year report on the incidence and prognosis. J. Pediat. Orthop.,4: 689-692, 1984.4689 1984

Gudinchet, F.; Maeder, P.; Oberson, J. C.; and Schnyder, P.: Magnetic resonance imaging of the shoulder in children with brachial plexus birth palsy. Pediat. Radiol.,25 (Supplement 1): 125-S128, 1995.25 (Supplement 1)125 1995

Hardy, A. E.: Birth injuries of the brachial plexus. Incidence and prognosis. J. Bone and Joint Surg.,63-B(1): 98-101, 1981.63-B(1)98 1981

Hentz, V. R., and Narakas, A.: The results of microneurosurgical reconstruction in complete brachial plexus palsy. Assessing outcome and predicting results. Orthop. Clin. North America,19: 107-114, 1988.19107 1988

Hernandez, R. J., and Dias, L.: CT evaluation of the shoulder in children with Erb's palsy. Pediat. Radiol.,18: 333-336, 1988.18333 1988 [PubMed]

Jackson, S. T.; Hoffer, M. M.; and Parrish, N.: Brachial-plexus palsy in the newborn. J. Bone and Joint Surg.,70-A: 1217-1220, Sept. 1988.70-A1217 1988

Liebolt, F. L., and Furey, J. G.: Obstetrical paralysis with dislocation of the shoulder. A case report. J. Bone and Joint Surg.,35-A: 227-230, Jan. 1953.35-A227 1953

Mallet, J.: Paralysie obstétricale du plexus brachial. Traitement des sequelles. Rev. chir. orthop.,58 (Supplement): 166-168, 1972.58 (Supplement)166 1972

Michelow, B. J.; Clarke, H. M.; Curtis, C. G.; Zuker, R. M.; Seifu, Y.; and Andrews, D. F.: The natural history of obstetrical brachial plexus palsy. Plast. and Reconstr. Surg.,93: 675-680, 1994.93675 1994

Mintzer, C. M.; Waters, P. M.; and Brown, D. J.: Glenoid version in children. J. Pediat. Orthop.,16: 563-566, 1996.16563 1996

Phelps, A. M.: Report of a case of congenital dislocation of the shoulder backward. Trans. Am. Orthop. Assn.,8: 239-245, 1896.8239 1896

Phipps, G. J., and Hoffer, M. M.: Latissimus dorsi and teres major transfer to rotator cuff for Erb's palsy. J. Shoulder and Elbow Surg.,4: 124-129, 1995.4124 1995

Pollock, A. N., and Reed, M. H.: Shoulder deformities from obstetrical brachial plexus paralysis. Skel. Radiol.,18: 295-297, 1989.18295 1989

Price, A. E., and Grossman, J. A.: A management approach for secondary shoulder and forearm deformities following obstetrical brachial plexus injury. Hand Clin.,11: 607-617, 1995.11607 1995 [PubMed]

Randelli, M., and Gambrioli, P. L.: Glenohumeral osteometry by computed tomography in normal and unstable shoulders. Clin. Orthop.,208: 151-156, 1986.208151 1986 [PubMed]

Sever, J. W.: Obstetrical paralysis. Report of eleven hundred cases. J. Am. Med. Assn.,85: 1862-1865, 1925.851862 1925

Specht, E. E.: Brachial plexus palsy in the newborn. Incidence and prognosis. Clin. Orthop.,110: 32-34, 1975.11032 1975 [PubMed]

Troum, S.; Floyd, W. E., III; and Waters, P. M.: Posterior dislocation of the humeral head in infancy associated with obstetrical paralysis. A case report. J. Bone and Joint Surg.,75-A: 1370-1375, Sept. 1993.75-A1370 1993

Waters, P. M.: Is biceps recovery a reliable prognosticator for brachial plexus birth palsy? Read at the Annual Meeting of the Pediatric Orthopaedic Society of North America, Phoenix, Arizona, May 15, 1996.

Whitman, R.: The treatment of congenital and acquired luxations at the shoulder in childhood. Ann. Surg.,42: 110-115, 1905.42110 1905 [PubMed]

Wickstrom, J.; Haslam, E. T.; and Hutchinson, R. H.: The surgical management of residual deformities of the shoulder following birth injuries of the brachial plexus. J. Bone and Joint Surg.,37-A: 27-36, Jan. 1955.37-A27 1955

Zancolli, E. A.: Classification and management of the shoulder in birth palsy. Orthop. Clin. North America,12: 433-457, 1981.12433 1981

Topics

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-D Magnetic resonance imaging scan of a glenohumeral joint with type-IV deformity. There is progressive deformity (a false glenoid) and subluxation.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-E Computerized tomographic scan of a glenohumeral joint with grade-V deformity (on the right). There is associated progressive deformity of the humeral head with flattening.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-F Computerized tomographic scan of a glenohumeral joint with type-VI deformity (on the right). There is infantile dislocation of the humeral head.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 4 Graph showing the glenoscapular angles (retroversion), according to age. The angle is increased in older children, who have longer-standing muscle imbalance.

View Larger

TABLE I THE MALLET CLASSIFICATION¹⁶

*A trumpet sign consists of abduction of the shoulder with simultaneous flexion of the elbow.

Functional Parameter	Class 1	Class 2	Class 3	Class 4	Class 5
Global abduction	None	<30 degrees	30—90 degrees	>90 degrees	Normal
External rotation	None	<0 degrees	0—20 degrees	>20 degrees	Normal
Hand-to-neck ability	None	Not possible	Difficult	Easy	Normal
Hand-to-mouth ability	None	Marked trumpet sign*	Partial trumpet sign*	<40 degrees of abduction	Normal
Internal rotation	None	Not possible	To S1	To T12	Normal

Add to My JBJS

TABLE I THE MALLET CLASSIFICATION¹⁶

*A trumpet sign consists of abduction of the shoulder with simultaneous flexion of the elbow.

Functional Parameter	Class 1	Class 2	Class 3	Class 4	Class 5
Global abduction	None	<30 degrees	30—90 degrees	>90 degrees	Normal
External rotation	None	<0 degrees	0—20 degrees	>20 degrees	Normal
Hand-to-neck ability	None	Not possible	Difficult	Easy	Normal
Hand-to-mouth ability	None	Marked trumpet sign*	Partial trumpet sign*	<40 degrees of abduction	Normal
Internal rotation	None	Not possible	To S1	To T12	Normal

View Larger

TABLE II MALLET CLASSIFICATION FOR THE FORTY-TWO PATIENTS*

Case	Functional Parameter
Global Abduction	External Rotation	Hand-to- Neck Ability	Hand-to- Mouth Ability
1	2	2	2	3
2	1	1	1	1
3	3	3	2	3
4	2	2	2	2
5	3	2	2	2
6	3	2	2	3
7	4	3	4	4
8	3	2	3	2
9	4	3	4	4
10	4	2	3	4
11	3	2	2	2
12	3	2	2	2
13	3	2	2	2
14	4	3	4	4
15	2	2	2	2
16	3	2	2	2
17	3	3	3	3
18	3	1	2	2
19	1	2	2	2
20	4	2	4	4
21	2	3	3	3
22	3	2	2	1
23	4	2	3	2
24	3	2	3	2
25	4	2	3	4
26	3	2	2	2
27	4	2	2	2
28	3	2	2	2
29	4	2	3	2
30	3	2	2	3
31	3	2	3	2
32	3	2	2	2
33	3	2	2	2
34	3	2	2	2
35	3	2	2	2
36	2	2	2	1
37	3	2	2	3
38	3	2	2	2
39	3	2	2	2
40	4	2	3	3
41	3	2	3	2
Mean	3.0	2.1	2.4	2.4

Add to My JBJS

TABLE II MALLET CLASSIFICATION FOR THE FORTY-TWO PATIENTS*

Case	Functional Parameter
Global Abduction	External Rotation	Hand-to- Neck Ability	Hand-to- Mouth Ability
1	2	2	2	3
2	1	1	1	1
3	3	3	2	3
4	2	2	2	2
5	3	2	2	2
6	3	2	2	3
7	4	3	4	4
8	3	2	3	2
9	4	3	4	4
10	4	2	3	4
11	3	2	2	2
12	3	2	2	2
13	3	2	2	2
14	4	3	4	4
15	2	2	2	2
16	3	2	2	2
17	3	3	3	3
18	3	1	2	2
19	1	2	2	2
20	4	2	4	4
21	2	3	3	3
22	3	2	2	1
23	4	2	3	2
24	3	2	3	2
25	4	2	3	4
26	3	2	2	2
27	4	2	2	2
28	3	2	2	2
29	4	2	3	2
30	3	2	2	3
31	3	2	3	2
32	3	2	2	2
33	3	2	2	2
34	3	2	2	2
35	3	2	2	2
36	2	2	2	1
37	3	2	2	3
38	3	2	2	2
39	3	2	2	2
40	4	2	3	3
41	3	2	3	2
Mean	3.0	2.1	2.4	2.4

View Larger

TABLE III DATA WITH REGARD TO RADIOGRAPHIC TYPE

Radio- graphic Type	No. of Patients	Age* (yrs.)	Glenoscapular Angle* (degrees)		Posterior Subluxation† (per cent)		Mallet Class*¹⁶
Affected Side	Normal Side	Affected Side	Normal Side	Global Abduction	External Rotation	Hand-to- Neck Ability	Hand-to- Mouth Ability
I	6	3.0	-6	-5	45	50	2.2	2.0	1.8	2.2
II	13	4.8	-16	-4	42	50	3.2	2.2	2.7	2.8
III	4	3.7	-29	-5	17	50	3.0	2.2	2.7	2.5
IV	8	3.4	-37	-9	9	49	3.5	2.0	2.4	2.5
V	4	11.3	-47	-6	4	48	3.0	2.0	2.2	2.0
VI	5	1.9	-46	-5	1	47	2.8	2.0	2.0	2.0
VII	2	13.5	-5	-2	40	50	3.5	2.0	3.0	2.5

Add to My JBJS

TABLE III DATA WITH REGARD TO RADIOGRAPHIC TYPE

Radio- graphic Type	No. of Patients	Age* (yrs.)	Glenoscapular Angle* (degrees)		Posterior Subluxation† (per cent)		Mallet Class*¹⁶
Affected Side	Normal Side	Affected Side	Normal Side	Global Abduction	External Rotation	Hand-to- Neck Ability	Hand-to- Mouth Ability
I	6	3.0	-6	-5	45	50	2.2	2.0	1.8	2.2
II	13	4.8	-16	-4	42	50	3.2	2.2	2.7	2.8
III	4	3.7	-29	-5	17	50	3.0	2.2	2.7	2.5
IV	8	3.4	-37	-9	9	49	3.5	2.0	2.4	2.5
V	4	11.3	-47	-6	4	48	3.0	2.0	2.2	2.0
VI	5	1.9	-46	-5	1	47	2.8	2.0	2.0	2.0
VII	2	13.5	-5	-2	40	50	3.5	2.0	3.0	2.5