Abstract
We retrospectively reviewed the medical records, operative reports, and preoperative and postoperative radiographs of thirty-nine patients who had been managed operatively for malunion of a fracture of the proximal aspect of the humerus. The malunions were categorized according to the presence of osseous abnormalities, including malposition of the greater or lesser tuberosity (type I; twenty-eight patients), incongruity of the articular surface (type II; twenty-six patients), and malalignment of the articular segment (type III; sixteen patients). Soft-tissue abnormalities, such as soft-tissue contracture, a tear of the rotator cuff, and impingement, were also recorded. At an average of forty-four months (range, twelve to fifty-three months) postoperatively, the patients were assessed for pain relief, the range of motion of the shoulder, and the ability to perform activities of daily living.The result was satisfactory for twenty-seven patients (69 per cent) and unsatisfactory for the remaining twelve (31 per cent) at the latest follow-up evaluation. Of the twenty-seven patients who had a satisfactory result, twenty-six (96 per cent) had had complete operative correction of all osseous and soft-tissue abnormalities. Of the twelve patients who had an unsatisfactory result, four had had complete operative correction of these abnormalities (p < 0.0001).Twenty-six patients (67 per cent) had incongruity of the glenohumeral joint at the time of presentation. Twenty-three of these patients had the incongruity corrected with prosthetic arthroplasty (twenty-two) or arthrodesis of the glenohumeral joint (one); the result was satisfactory for seventeen (74 per cent). In contrast, the result was unsatisfactory for all three patients in whom the incongruity had not been corrected at the time of the operation (p = 0.01).Eleven patients had malposition of the greater or lesser tuberosity but a congruent joint surface preoperatively. Ten patients in this group were managed with either osteotomy of the tuberosity or acromioplasty, and nine of them had a satisfactory result at the latest follow-up evaluation. The result was unsatisfactory for one patient who was managed with only correction of a soft-tissue contracture (that is, no treatment of the malposition) (p = 0.05).Both osseous and soft-tissue abnormalities were identified as the cause of pain and stiffness in patients who had malunion of a fracture of the proximal aspect of the humerus. We concluded that operative management of these patients is successful only if all osseous and soft-tissue abnormalities are corrected at the time of the operation.
Malunion of a fracture of the proximal aspect of the humerus can result from inadequate operative reduction, loss of operative reduction, or non-operative treatment of a displaced fracture. Many patients who have a malunion of the proximal aspect of the humerus are seen because of pain and stiffness of the shoulder, which often cause severe functional impairment. Disruption of the normal anatomical relationships between the tuberosities, the humeral head, and the humeral shaft, as well as incongruity of the joint, can result in a decrease in the range of motion and strength of the shoulder. Soft-tissue scarring and tears of the tendons of the rotator cuff also contribute to stiffness and loss of strength in these patients. Operative treatment of malunion of the proximal aspect of the humerus can be challenging. The purpose of the present retrospective study was to analyze the results of such treatment.
*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.
†Shoulder and Elbow Service, Department of Orthopaedic Surgery, University of Pennsylvania Health System, One Cupp Pavilion, 39th and Market Streets, Philadelphia, Pennsylvania 19104. Please address requests for reprints to Dr. Iannotti.
‡California Pacific Medical Center, Clay at Buchanan, San Francisco, California 94120.
From 1984 to 1995, forty-seven patients were managed operatively for malunion of the proximal aspect of the humerus at the University of Pennsylvania Health System or at the California Pacific Medical Center. The indication for operative management was severe pain or loss of function, or both, that was unresponsive to non-operative treatment, including non-steroidal anti-inflammatory medication and physical therapy. Five patients who had degenerative osteoarthrosis or inflammatory arthritis of the glenohumeral joint that had been present before the initial injury and three patients who were followed for less than twelve months were excluded from the study. The remaining thirty-nine patients form the study group (Table I).
The medical and operative reports as well as the preoperative and postoperative radiographs were retrospectively reviewed. The average age of the patients at the time of presentation was fifty-three years (range, twenty-seven to seventy-seven years). There were nineteen men and twenty women. Twenty-one fractures were on the left side, and eighteen were on the right. Twenty-three fractures (59 per cent) were in the dominant extremity.
The mechanism of the initial injury was a fall from a standing height for thirty patients (77 per cent) and a motor-vehicle accident for the remaining nine (23 per cent).
The fractures were defined, according to Neer's classification7,8 of fractures of the proximal aspect of the humerus, after a review of the medical and operative records or, when available, the radiographs made at the time of the initial injury. Eight patients (21 per cent) had a two-part fracture of the surgical neck, eight had a three-part fracture of the greater tuberosity, six (15 per cent) had a four-part fracture, five (13 per cent) had a two-part fracture of the greater tuberosity, three (8 per cent) had a four-part anterior fracture-dislocation, three had a humeral head-splitting fracture, two (5 per cent) had a two-part fracture-dislocation of the greater tuberosity, two had a four-part posterior fracture-dislocation, one (3 per cent) had a two-part fracture of the lesser tuberosity, and one had a two-part fracture-dislocation of the lesser tuberosity.
The initial treatment included immobilization in a sling followed by range-of-motion exercises begun three to four weeks later for thirty patients (77 per cent), internal fixation with plates and screws for three (8 per cent), closed reduction followed by immobilization in a sling for two (5 per cent), internal fixation with flexible intramedullary devices for two, and prosthetic replacement of the humeral head for two. The average time from the initial injury to the time that the patient was seen for treatment of the malunion was twenty-nine months (range, two to 270 months). The main symptoms at the time of presentation were pain and stiffness of the shoulder.
On the basis of the physical findings, radiographs, and operative reports, the malunions were defined according to the presence of osseous abnormalities (malposition of the greater or lesser tuberosity, incongruity of the articular surface, and malalignment of the articular segment) and soft-tissue abnormalities (soft-tissue contracture, a tear of the rotator cuff, and impingement). In many patients, the osseous abnormalities could not be adequately quantitated on the preoperative radiographs because there were combined deformities. For example, in the presence of varus malalignment of the articular segment due to malunion of the surgical neck, concomitant malposition of the greater tuberosity is difficult to assess radiographically. This difficulty is due to the abnormal position of the fragment that is normally used to quantitate displacement (the rest of the proximal aspect of the humerus).
Identification and quantification of the soft-tissue abnormalities were also difficult preoperatively. Lack of motion could presumably result from capsular contracture, extracapsular contracture, osseous impingement, or pain. Furthermore, small full-thickness tears of the rotator cuff and high-grade partial defects (those involving more than 50 per cent of the thickness of the cuff) were difficult to identify on the preoperative radiographs, especially when the original fracture had been treated operatively. Consequently, the intraoperative findings were primarily used to categorize the osseous and soft-tissue abnormalities.
The osseous abnormalities were categorized according to the following criteria. Malposition of the greater or lesser tuberosity (type-I malunion) was assessed on the preoperative radiographs as well as intraoperatively for all patients. The greater tuberosity was considered to be displaced and malpositioned if it was located one centimeter or more posterior or superior to its anatomical position1 (Fig. 1). The lesser tuberosity was considered to be malpositioned if it was located one centimeter or more medial to its anatomical position.
Incongruity of the articular surface (type-II malunion) was also assessed on the preoperative radiographs as well as intraoperatively for all patients. The etiology of such incongruity was often multifactorial and included avascular necrosis of the humeral head, step-off of the articular surface secondary to a split fracture of the humeral head, subluxation or dislocation of the glenohumeral joint, and posttraumatic osteoarthrosis. All patients who had avascular necrosis or posttraumatic osteoarthrosis also had malalignment of the fracture site (step-off of the articular surface or subluxation of the glenohumeral joint). We considered an intra-articular incongruity or step-off of five millimeters or more to be a malunion (Figs. 2-A, 2-B, and 2-C).
Malalignment of the articular segment (type-III malunion), which was also assessed with use of the preoperative radiographs and intraoperative findings for all patients, was defined as 45 degrees or more of rotational deformity of the articular segment with respect to the humeral shaft in any of three planes: coronal (varus or valgus angulation), sagittal (anterior or posterior angulation), or axial (malrotation about the longitudinal axis of the humeral shaft). Varus-valgus and anterior-posterior malalignment were quantitated on the preoperative radiographs by measurement of the angle subtended by the longitudinal axes of the two fragments (Figs. 3-A, 3-B, 3-C, 3-D and 3-E). Axial malalignment was assessed intraoperatively because it could not be quantitated preoperatively. To determine humeral malalignment intraoperatively, the humerus was rotated until the articular surface was centered in the glenoid fossa and then the angle between vertical (perpendicular to the floor) and the forearm was measured with the elbow flexed to 90 degrees. Angles created by external rotation of the humerus from the vertical position were considered to be positive, and those created by internal rotation of the humerus were considered to be negative. For the purpose of this study, angles of 90 degrees or more or of 0 degrees or less were indicative of malalignment.
As mentioned previously, the soft-tissue abnormalities were also difficult to assess preoperatively. Therefore, rotator-cuff tears and capsular contractures were only suspected on the basis of physical findings when there was a symmetrical decrease in active and passive motion such that flexion was to 90 degrees or less, external rotation was to 0 degrees or less, and internal rotation was to the posterolateral aspect of the buttock or caudad to that level. This suspicion was verified by examination with the patient under anesthesia as well as by the intraoperative findings. Soft-tissue contracture was considered to be the cause of stiffness of the glenohumeral joint if the decreased range of motion documented on outpatient evaluation was verified on examination with the patient under anesthesia and no osseous impingement was seen intraoperatively.
Because of the difficulty associated with interpreting postoperative imaging studies of patients who have a distorted anatomy of the proximal aspect of the humerus as a result of malunion, we analyzed the rotator cuff on the basis of the intraoperative findings. The size of any defect of the rotator cuff observed intraoperatively was quantitated by measuring its dimensions in both the anterior-posterior and the medial-lateral direction.
Subacromial impingement was difficult to diagnose preoperatively. Without exception, the patients with this condition had pain at the extremes of motion. In addition, motion of the glenohumeral joint was moderately restricted, but the restriction was not severe enough to fit our strict criteria for soft-tissue contracture. Impingement was indicated by the presence of at least 120 degrees of flexion, at least 30 degrees of external rotation, a malunion pattern that could theoretically lead to abnormal subacromial contact (superior malposition of the greater tuberosity or varus malalignment), displacement of the greater tuberosity of less than 1.5 centimeters, radiographic evidence of narrowing of the supraspinatus outlet (acromial spurring), and pain at the extreme of passive flexion of the glenohumeral joint that was relieved after subacromial injection of a local anesthetic.
According to our categorization of malunions, eight patients (21 per cent) had a malunion without any associated soft-tissue abnormality and thirty-one (79 per cent) had combined osseous and soft-tissue abnormalities. Of the eight patients who had only osseous abnormalities, two had isolated malposition of a tuberosity, one had isolated incongruity of the articular surface, one had isolated malalignment of the articular segment, and four had combined osseous deformities (Table I).
The osseous abnormality was isolated malposition of the greater or lesser tuberosity in nine (29 per cent) of the thirty-one patients who had combined osseous and soft-tissue abnormalities; isolated incongruity of the articular surface in two (6 per cent); incongruity of the articular surface and malalignment of the articular segment in five (16 per cent); malposition of the tuberosity and incongruity of the articular surface in eight (26 per cent); malposition of the tuberosity and malalignment of the articular segment in one (3 per cent); and malposition of the tuberosity, incongruity of the articular surface, and malalignment of the articular segment in six (19 per cent). The soft-tissue abnormality was capsular contracture in twenty-five (81 per cent) of the thirty-one patients, a torn rotator cuff in fifteen (48 per cent), and impingement in two (6 per cent).
The operative results were analyzed in terms of the correction of the osseous and soft-tissue abnormalities. The criteria that were used to determine the adequacy of the correction of the osseous abnormality depended on the abnormality. Correction of malposition of the greater or lesser tuberosity was considered to be adequate if it included acromioplasty or osteotomy of the tuberosity and reduction to within five millimeters of the anatomical position. Osteotomy and reduction was used to treat more than 1.5 centimeters of displacement, and acromioplasty was used to treat 1.0 to 1.5 centimeters of displacement. Correction of incongruity of the articular surface was considered to be adequate if arthroplasty or arthrodesis of the glenohumeral articulation was performed (Figs. 2-A, 2-B, and 2-C). Correction of malalignment of the articular segment in the sagittal or coronal plane was considered to be adequate if the articular segment was realigned to within 10 degrees of the longitudinal axis of the humeral neck during osteotomy or joint replacement (Figs. 3-A, 3-B, 3-C, 3-D and 3-E). Correction of malalignment of the articular segment in the axial plane was considered to be adequate if the resultant articular retroversion, measured intraoperatively as described earlier, was between 20 and 50 degrees. If there were combined osseous deformities, operative treatment was considered adequate only if all of the deformities were corrected adequately.
The criteria that were used to determine the adequacy of operative correction of the soft-tissue abnormality also depended on the specific abnormality. In the presence of clinically suspected soft-tissue contracture, osseous impingement was ruled out by intraoperative inspection. A capsular release was considered adequate if the release, performed either alone or in combination with lengthening of the subscapularis (in patients who had 0 degrees or less of external rotation), resulted in at least 130 degrees of flexion and at least 30 degrees of external rotation, as measured intraoperatively. Capsular release was performed without lengthening of the subscapularis in sixteen patients, and it was done with lengthening in six. All releases were performed by means of an open technique with circumferential release of the capsule.
Correction of a tear of the rotator cuff was considered to be adequate if the tear was repaired. Repair of the rotator cuff was combined with an acromioplasty if anterior acromial spurring was evident on the preoperative radiographs or intraoperatively. Impingement was considered to be adequately corrected if an acromioplasty was performed. An isolated acromioplasty was done in two patients who had clinical evidence of impingement, as determined with the criteria described earlier, and who were seen to have anterior acromial spurring on radiographs. Operative treatment of combined deformities was considered adequate only if all of the deformities were corrected.
The patients were evaluated, both at the time of presentation and at the latest follow-up evaluation, according to a grading system derived from the American Shoulder and Elbow Surgeons' Shoulder Evaluation Form4,5. The average time from the operation to the latest follow-up evaluation was forty-four months (range, twelve to fifty-three months). Subjective data included the preoperative and postoperative levels of pain and function and the overall level of satisfaction at the time of follow-up. Objective data included the active range of motion on physical examination.
The level of pain was graded by the patients on a visual-analog scale that ranged from 0 to 5 points; 0 points indicated disabling pain and 5 points, the absence of pain. The patients rated function by answering a questionnaire regarding their ability to perform activities of daily living with use of the extremity on the involved side. The ability to perform each activity was rated on a scale of 0 to 4 points; 0 points indicated complete inability to perform the task and 4 points, the ability to perform it without difficulty. Specific questions included the ability to use a back pocket, put on a coat, wash the contralateral axilla, eat with a utensil, comb the hair, use the arm at shoulder level, carry ten pounds (4.5 kilograms) with the arm at the side, dress with use of both hands, push or pull with the injured limb, use the hand overhead, throw, perform usual work activities, and perform usual sports activities. The functional capacity of each patient was determined by totaling the numerical responses and calculating them as a percentage of the maximum possible score. Active forward elevation and external rotation were measured preoperatively and at the latest follow-up examination.
The result was considered satisfactory if the subjective pain grade was 4 or 5 points (slight or no pain), active forward elevation was to at least 90 degrees, and the functional capacity of the affected shoulder was at least 50 per cent of normal.
Statistical analysis was performed with the two-tailed Student t test for numerical data and with chi-square analysis for non-parametric data. The level of significance was p = 0.05.
The result was satisfactory for twenty-seven patients (69 per cent) and unsatisfactory for twelve (31 per cent) at the latest follow-up evaluation (Table I). Of the twenty-seven patients who had a satisfactory result, twenty-six (96 per cent) had had adequate correction of all osseous and soft-tissue abnormalities intraoperatively. Of the twelve patients who had an unsatisfactory result, eight had had incomplete correction of these abnormalities (p < 0.0001, chi-square analysis).
Of the twenty-six patients who had had incongruity of the glenohumeral joint preoperatively, twenty-three were managed with either prosthetic arthroplasty (twenty-two) or arthrodesis of the glenohumeral joint (one). Seventeen (74 per cent) of these twenty-three patients had a satisfactory result. The unsatisfactory result in the remaining six patients was due to pain (one); lack of motion (one); pain and poor function (one); or pain, lack of motion, and poor function (three). In contrast, all three patients in whom the incongruity had not been corrected at the time of the operation had an unsatisfactory result (p = 0.01, chi-square analysis), which was due to pain and poor function (one); pain and lack of motion (one); or pain, lack of motion, and poor function (one).
Eleven patients had had isolated malposition of the greater or lesser tuberosity but a congruent joint surface preoperatively; ten of these patients were managed with either osteotomy of the greater tuberosity or acromioplasty. Nine of the ten had a satisfactory result at the latest follow-up evaluation. One patient, who had only manipulation under anesthesia to treat a soft-tissue contracture but no correction of the malposition, had an unsatisfactory result (p = 0.05, chi-square analysis).
The overall average pain score was 1.6 points (range, 0 to 3 points) preoperatively and 3.9 points (range, 0 to 5 points) at the latest follow-up evaluation. The average score for the patients for whom the correction was considered adequate was 4.2 points (range, 1 to 5 points), which represented an increase from an average of 1.6 points (range, 0 to 3 points) preoperatively. In contrast, the average score for the patients for whom the correction was not considered adequate was 2.8 points (range, 0 to 4 points) compared with an average of 1.4 points (range, 0 to 3 points) preoperatively (p = 0.0006, Student t test).
The overall average score for function was 41 per cent (range, 0 to 81 per cent), compared with the status on the normal contralateral side, before the operation and 64 per cent (range, 21 to 100 per cent) at the latest follow-up evaluation. The average score for the patients for whom the correction was considered adequate was 70 per cent (range, 27 to 100 per cent) at the latest follow-up evaluation compared with 41 per cent (range, 0 to 81 per cent) preoperatively. The average score for the patients for whom the correction was not considered adequate was 42 per cent (range, 21 to 69 per cent) at the latest follow-up examination compared with 40 per cent (range, 29 to 69 per cent) preoperatively (p = 0.0004, Student t test).
The average range of active forward elevation of the injured shoulder was 82 degrees (range, 15 to 165 degrees) preoperatively and 102 degrees (range, 30 to 160 degrees) at the latest follow-up evaluation. The average ranges were 77 degrees (range, 15 to 135 degrees) and 109 degrees (range, 30 to 160 degrees) for the patients for whom the correction was considered adequate and 96 degrees (range, 70 to 165 degrees) and 81 degrees (range, 40 to 125 degrees) for the patients for whom the treatment was not considered adequate. There was a significant difference between the two groups with regard to the preoperative values (p = 0.02, Student t test).
Nine of the twelve patients who had an unsatisfactory result had a total of thirteen subsequent operative procedures for treatment of the malunion. All of the patients were evaluated retrospectively, and the scores for pain, range of motion, and functional capacity were determined after the index treatment for the malunion. Only two patients who had a subsequent procedure had improvement at the latest follow-up evaluation compared with the status at the time of the index operation.
Of the seven patients in whom the initial fracture had been treated operatively, five had a satisfactory result after treatment of the malunion. We found no significant differences in the preoperative and postoperative scores for pain, range of motion, or functional capacity between the patients who had been managed operatively for the initial fracture and those who had been managed non-operatively.
Nineteen of the thirty-nine patients had been seen for treatment of the malunion less than one year after the fracture. Of these, sixteen had a satisfactory result at the latest follow-up evaluation. In contrast, of the twenty patients who had been seen one year or more after the fracture, only eleven (55 per cent) had a satisfactory result (p = 0.04, chi-square analysis).
Complications
Twelve complications in eleven patients were a result of the operative treatment. Three intraoperative fractures of the humeral shaft occurred during preparation of the canal for insertion of a humeral prosthesis; these were treated intraoperatively with insertion of a long-stem humeral prosthesis and cerclage wire.
Six complications developed in six patients during the early postoperative period. Two patients had instability of the humeral head component. One of them had posterior subluxation of a hemiarthroplasty component; this was treated with revision hemiarthroplasty with insertion of a component with a larger prosthetic head. The second patient had anterior subluxation of a hemiarthroplasty component; this was treated with revision hemiarthroplasty with placement of the humeral component in an increased degree of retroversion. Non-union of the osteotomy site and loosening of the humeral component developed in another patient, who had had an osteotomy to treat varus angulation of the surgical neck during total shoulder arthroplasty; this patient was one of the three who sustained an intraoperative humeral fracture at the time of insertion of the prosthetic component. A patient who had been managed with osteotomy and fixation with a flexible intramedullary rod for the treatment of varus malunion of the surgical neck was found to have avascular necrosis of the humeral head after removal of the rod, four months after the initial fixation. The avascular necrosis was treated with total shoulder arthroplasty. Early loosening of the humeral component developed in another patient and was treated with revision total shoulder arthroplasty. Periscapular muscle pain and weakness developed despite successful osseous union in one patient who had been managed with arthrodesis of the glenohumeral joint. Despite the pain and weakness, the patient had a satisfactory result and was employed as a manual laborer at the time of the latest follow-up evaluation.
An attritional tear of the rotator cuff developed as a late complication in three patients. The presenting symptoms were progressive loss of motion and strength. All three of these patients had had a torn rotator cuff at the time that the malunion was treated. Two of the three patients had an excellent result postoperatively, with improved functional capacity; both of these patients had no pain at the latest follow-up evaluation, but they did have marked functional limitations. The third patient had moderate relief of pain and moderate improvement in function postoperatively. This patient had repair of the rotator cuff because of progressive pain and disability, but there was a decrease in function at the latest follow-up evaluation.
The results of operative treatment of malunion of the proximal aspect of the humerus have been reported infrequently2,3,6,9-11. Habermeyer and Schweiberer retrospectively studied two groups of seventeen patients who had malunion of the proximal aspect of the humerus: one group was managed with prosthetic reconstruction, and the other group was managed with varus or valgus osteotomy, derotational osteotomy, or arthrodesis. Those authors noted that the results for the patients who had been managed with osteotomy were slightly better than those for the patients who had been managed with reconstruction, and they recommended osteotomy to treat malunion in the absence of posttraumatic osteoarthrosis.
Morris et al. followed six patients who had malunion of the greater tuberosity and decreased function of the shoulder. Those authors used computerized tomographic scanning preoperatively to determine the direction and magnitude of displacement of the greater tuberosity. Operative treatment consisted of osteotomy of the tuberosity and repositioning of the displaced fragment. Computerized tomographic scanning improved the accuracy of the preoperative assessment of the malposition of the greater tuberosity, and the osteotomy led to substantial improvement in the function of the shoulder.
In a study by Solonen and Vastamaki, seven patients had a corrective osteotomy to treat varus deformity that had resulted from malunion of a fracture of the surgical neck. The indication for the operation was a decreased range of motion. The result was satisfactory for five of the seven patients.
Tanner and Cofield compared the results of replacement of the humeral head in forty-nine shoulders that had an acute or old fracture or fracture-dislocation. Sixteen patients had a malunion as well as incongruity of the glenohumeral joint. Those authors found that patients who had had a late reconstruction had a better range of motion of the treated joint at the time of follow-up than those who had had reconstruction of an acute fracture. However, the patients who had had a chronic fracture had a greater number of complications, which Tanner and Cofield ascribed to difficulty with the performance of the operation, tissue-scarring, and distortion of the anatomy.
In a similar report, Norris et al. described late reconstruction after failed treatment of twenty-three three-part and four-part fractures. Their series included seventeen malunions that had been treated with either total shoulder arthroplasty or replacement of the humeral head. They found the results to be inferior to those after reconstruction of acute fractures. They also found late reconstruction to be technically demanding and to be associated with more complications than reconstruction of acute fractures.
Interpretation of these reports is difficult for several reasons. Most dealt primarily with correction of osseous malalignment or resurfacing of the articular surface. Although management of the periarticular soft tissues was often emphasized, none of the reports identified or quantitated the effect that soft-tissue abnormalities associated with malunion of the proximal aspect of the humerus had on the operative planning or the postoperative results.
Malunion of the proximal aspect of the humerus is almost always accompanied by some soft-tissue abnormality, such as soft-tissue contracture, a tear of the rotator cuff, or subacromial impingement. Furthermore, there is often a combination of malposition of the tuberosity, incongruity of the articular surface, and malalignment of the articular segment. We attempted to provide a method for systematic evaluation of both the osseous and the soft-tissue abnormalities associated with malunion of the proximal aspect of the humerus so that all potential sources of pain and dysfunction can be identified and treated.
There are several inherent limitations to the present study. It is a retrospective review of the cases of patients who were referred to one of two tertiary shoulder centers because of painful malunion of the proximal aspect of the humerus5. It may not be valid to compare our patients with patients who have had a malunion or a similar fracture treated acutely. Although we believe that the criteria for operative selection, the operative techniques, and the postoperative rehabilitation protocols were uniform, they were not controlled in a prospective manner; this complicates the interpretation of our results.
Preoperative imaging was not standardized in the present study. Although all patients had routine radiographs, including a trauma series8, neither computerized tomographic scanning nor magnetic resonance imaging was routinely performed. It is possible that computerized tomographic scanning, particularly with three-dimensional reconstruction, and magnetic resonance imaging may have assisted in the preoperative categorization of the malunions6. However, even for patients for whom computerized tomographic scans and magnetic resonance images were available, we relied on intraoperative assessment of the osseous and soft-tissue abnormalities.
The criteria that we used to define a satisfactory result (slight or no pain, at least 90 degrees of active forward elevation, and at least 50 per cent functional use of the arm compared with that on the normal side) were modest. Nevertheless, only twenty-seven (69 per cent) of the thirty-nine patients had a satisfactory result. Twenty-six (96 per cent) of these twenty-seven patients were thought to have had adequate correction of all osseous and soft-tissue abnormalities. The low percentage of satisfactory results in the present study, despite correction of all identified abnormalities, confirms that operative treatment of these complex injuries is challenging.
Our data indicate that a delay in the operative treatment of malunion of the proximal aspect of the humerus has a negative impact on outcome. Sixteen of the nineteen patients who had been managed less than one year after the injury had a satisfactory result compared with eleven (55 per cent) of the twenty patients who had been managed at least one year after the injury. This may be the result of many factors, including disuse atrophy and more mature soft-tissue scarring associated with prolonged malunion.
The small number of malunions in each category makes comparison of the data between categories difficult. However, the best results were seen in the group of ten patients who had an isolated malposition of the greater or lesser tuberosity and had all of the osseous and soft-tissue abnormalities adequately corrected at the time of the operation. Only two of the eleven patients who had malposition of a tuberosity had findings that fit the criteria for isolated subacromial impingement. Acromioplasty without osteotomy of the tuberosity led to a satisfactory result in both patients. However, neither of these patients had more than 1.5 centimeters of displacement of the tuberosity, which would have necessitated osteotomy of the tuberosity. The one patient in whom a soft-tissue contracture and more than 1.5 centimeters of displacement of the tuberosity were treated with capsular release without osteotomy of the tuberosity had a poor result.
Six of the twenty-two patients who had joint replacement to treat incongruity of the articular surface had an unsatisfactory result. Only two of these six patients had had adequate correction of all osseous and soft-tissue abnormalities intraoperatively. Three of these six patients had a hemiarthroplasty. The result was poor for all three of these patients because of continued pain, stiffness, and poor function. These patients may have had a better result if a total shoulder arthroplasty rather than a hemiarthroplasty had been performed.
The overall rate of complications in our series was high. Eleven patients (28 per cent) had at least one complication, and nine complications developed in eight of the patients who had a joint replacement. The intraoperative fractures and postoperative instability were most likely the results of technical difficulties due to soft-tissue contracture, muscular atrophy or weakness, and abnormal anatomical landmarks associated with malunion.
Operative treatment of malunion of the proximal aspect of the humerus is challenging and often requires correction of combined soft-tissue and osseous abnormalities. We have developed an algorithm to assist in the categorization of these difficult malunions and the planning for their operative treatment (Fig. 4). Adequate treatment of all osseous and soft-tissue abnormalities affords the best opportunity for a satisfactory postoperative result. Satisfactory results were not associated with the type of malunion or soft-tissue abnormality, but they were associated with the adequacy of the treatment. However, as our data indicate, even when all identified abnormalities are corrected, the result still may not be satisfactory. Furthermore, operative treatment of this difficult problem was associated with a rate of complications that approached 30 per cent.
Flatow, E. L.; Cuomo, F.; Maday, M. G.; Miller, S. R.; McIlveen, S. J.; and Bigliani, L. U.: Open reduction and internal fixation of two-part displaced fractures of the greater tuberosity of the proximal part of the humerus. J. Bone and Joint Surg.,73-A: 1213-1218, Sept. 1991.73-A1213
1991
Gerber, C.: Rekonstruktive Chirurgie nach fehlverheilten Frakturen des proximalen Humerus bei Erwachsenen. Orthopade,19: 316-323, 1990.19316
1990
[PubMed]
Habermeyer, P., and Schweiberer, L.: Korrektureingriffe infolge von Humeruskopffrakturen. Orthopade,21: 148-157, 1992.21148
1992
[PubMed]
Lazarus, M. D.; Chansky, H. A.; Misra, S.; Williams, G. R.; and Iannotti, J. P.: Comparison of open and arthroscopic subacromial decompression. J. Shoulder and Elbow Surg.,3: 1-11, 1994.31
1994
Leggin, B.; Shaffer, M.; Newman, R.; Iannotti, J.; Williams, G.; and Brenneman, S.: Reliability and validity of a shoulder outcome scoring system. Unpublished data.
Morris, M. E.; Kilcoyne, R. F.; Shuman, W.; and Matsen, F., III: Humeral tuberosity fractures. Evaluation by CT scan and management of malunion. Orthop. Trans.,11: 242, 1987.11242
1987
Neer, C. S., II: Displaced proximal humeral fractures. Part I. Classification and evaluation. J. Bone and Joint Surg.,52-A: 1077-1089, Sept. 1970.52-A1077
1970
Neer, C. S., II: Displaced proximal humeral fractures. Part II. Treatment of three-part and four-part displacement. J. Bone and Joint Surg.,52-A: 1090-1103, Sept. 1970.52-A1090
1970
Norris, T.; Green, A.; and McGuigan, F.: Late prosthetic shoulder arthroplasty for displaced proximal humerus fractures. J. Shoulder and Elbow Surg.,4: 271-280, 1995.4271
1995
Solonen, K. A., and Vastamaki, M.: Osteotomy of the neck of the humerus for traumatic varus deformity. Acta Orthop. Scandinavica,56: 79-80, 1985.5679
1985
Tanner, M. W., and Cofield, R. H.: Prosthetic arthroplasty for fractures and fracture-dislocations of the proximal humerus. Clin. Orthop.,179: 116-128, 1983.179116
1983
[PubMed]