Functional anatomy of the shoulder.

OBJECTIVE: Movements of the human shoulder represent the result of a complex dynamic interplay of structural bony anatomy and biomechanics, static ligamentous and tendinous restraints, and dynamic muscle forces. Injury to 1 or more of these components through overuse or acute trauma disrupts this complex interrelationship and places the shoulder at increased risk. A thorough understanding of the functional anatomy of the shoulder provides the clinician with a foundation for caring for athletes with shoulder injuries. DATA SOURCES: We searched MEDLINE for the years 1980 to 1999, using the key words "shoulder," "anatomy," "glenohumeral joint," "acromioclavicular joint," "sternoclavicular joint," "scapulothoracic joint," and "rotator cuff." DATA SYNTHESIS: We examine human shoulder movement by breaking it down into its structural static and dynamic components. Bony anatomy, including the humerus, scapula, and clavicle, is described, along with the associated articulations, providing the clinician with the structural foundation for understanding how the static ligamentous and dynamic muscle forces exert their effects. Commonly encountered athletic injuries are discussed from an anatomical standpoint. CONCLUSIONS/RECOMMENDATIONS: Shoulder injuries represent a significant proportion of athletic injuries seen by the medical provider. A functional understanding of the dynamic interplay of biomechanical forces around the shoulder girdle is necessary and allows for a more structured approach to the treatment of an athlete with a shoulder injury.

Winner of the AlbertTrillat Young Investigator Award. The effects of aggressive notchplasty on the normal knee in dogs.

We assessed the possible association between an aggressive intercondylar notchplasty and histopathologic, radiographic, and gait changes to the knee. Three groups of six adult greyhounds were observed for 6 months. Group I dogs had a sham operation. Group II dogs had a 4-mm notchplasty of the lateral femoral condyle where it articulates with the lateral tibial spine. Group III dogs had a 7- to 8-mm notchplasty of the lateral femoral condyle to simulate the long-term effects of an overly aggressive notchplasty. Force plate gait analyses were not significantly different for any dogs at 3 and 6 months. Histopathologic studies (hematoxylin and eosin and safranin O stains) revealed notchplasty area remodeling with a thin layer of lamellar bone covered by fibrous connective tissue. Both Group II and III dogs had significant loss of lateral femoral condyle and trochlear groove articular surface proteoglycans. The radiographic notch width index remained unchanged throughout the study for Group I; the indexes increased immediately after surgery in Groups II and III because of the notchplasty, but after 6 months these values returned to near-preoperative measurements. An aggressive intercondylar notchplasty caused articular cartilage histopathologic changes at 6 months consistent with those found in knees with early degenerative arthritis. Significant refilling of a non-impinged notchplasty occurred by 6 months after surgery. Our results raise concern about the effects of aggressive intercondylar notch widening in humans.

Injuries to the posterolateral aspect of the knee. Association of anatomic injury patterns with clinical instability.

Seventy-one consecutive patients with posterolateral knee injuries had clinical stability testing abnormalities documented prospectively. We compared these findings with the incidence and patterns of their injuries documented at surgery. An abnormal reverse pivot shift test was associated with injury to the fibular collateral ligament (P = 0.01), popliteal components (P = 0.01), and midthird lateral capsular ligament (P = 0.02). An abnormal posterolateral external-rotation test at 30 degrees of flexion was associated with injury to the fibular collateral ligament (P = 0.0001) and lateral gastrocnemius tendon (P = 0.01). An abnormal adduction test at 30 degrees of flexion was associated with injury to the posterior arcuate ligament (P = 0.02). The results of this study should alert the clinician to the possibility of injury to a specific anatomic structure when the corresponding clinical stability test is abnormal. Because the fibular collateral ligament was injured in only 23% of the knees in this large series of patients, we recommend that an injury to the fibular collateral ligament not be the sole determining factor in making the diagnosis of posterolateral injuries. The wide array of injuries to many individual anatomic components that we found indicates the complexity of injuries to the posterolateral aspect of the knee.

The posterolateral aspect of the knee. Anatomy and surgical approach.

Thirty cadaveric knees were dissected to obtain a detailed understanding of the anatomic structures of the posterolateral aspect of the knee, and a dependable surgical approach to evaluate injuries to these structures was developed and used on 71 consecutive patients who were operated on for posterolateral knee injuries. Three fascial incisions and one lateral midcapsular incision were used to provide surgical access. The following individual anatomic structures were identified: the layers of the iliotibial tract, long and short heads of the biceps femoris muscle, fibular collateral ligament, midthird lateral capsular ligament, fabello-fibular ligament, posterior arcuate ligament, popliteus muscle complex, lateral coronary ligament, and posterior capsule. This study increased our understanding of the individual anatomic structures and the relationships between these components. The surgical approach provided for the evaluation of these anatomic structures should aid the surgeon in properly assessing the injuries before surgical repair. This information should also stimulate more anatomic, biomechanical, and clinical studies of the posterolateral aspect of the knee.

Surgical correction of medial subluxation of the patella.

We evaluated the results of a surgical procedure to correct medial subluxation of the patella in 63 patients (65 knees), most of whom had undergone a lateral retinacular release. We performed a direct repair or a reconstruction of the lateral patellotibial ligament using locally available tissue such as strips of iliotibial band or patellar tendon. Followup averaged 53.7 months (range, 24 to 99). Outcome was based on the examiner's inability to clinically reproduce the patient's painful medial subluxation and on the patient's general impression of his or her improved functional status. Forty-four patients (68%) reported improvement in their functional levels and 49 (75%) reported that they were subjectively improved by the procedure. Overall, 50 patients (80%) had a rating of good or excellent. Six knees required a second surgical reconstruction because of failure to improve or because of a reinjury. Analysis of overall clinical outcome revealed no significant relationships based on the patient's age at the time of the initial procedure, sex, or length of followup (P > 0.10). Reconstitution of the lateral patellotibial ligament effectively corrected medial subluxation of the patella and long-term results of this salvage procedure were satisfactory.

The biceps femoris muscle complex at the knee. Its anatomy and injury patterns associated with acute anterolateral-anteromedial rotatory instability.

We dissected 30 cadaveric knees to provide a detailed anatomic description of the biceps femoris muscle complex at the knee. The main components of the long head of the muscle are a reflected arm, a direct arm, an anterior arm, and a lateral and an anterior aponeurosis. The main components of the short head of the biceps femoris muscle are a proximal attachment to the long head's tendon, a capsular arm, a confluens of the biceps and the capsuloosseous layer of the iliotibial tract, a direct arm, an anterior arm, and a lateral aponeurosis. We examined 82 consecutive, acutely injured knees with clinical signs of anterolateral-anteromedial rotatory instability for the incidence and anatomic location of injuries to the biceps femoris muscle. Injuries to components of that muscle were identified in 59 (72%) of these knees; 29 knees (35.4%) had multiple components injured. There were 3 injuries to the long head of the biceps femoris muscle (all in the reflected arm) and 89 to the short head. A statistically significant correlation (P = 0.01) was found between increased anterior translation with the knee at 25 degrees of flexion as demonstrated by the Lachman test and injury to the biceps-capsuloosseous iliotibial tract confluens. Additionally, adduction laxity at 30 degrees of flexion correlated with a Segond fracture (P = 0.04). These data establish, in part, the relationship of the biceps femoris complex injury to anterior translation instability.

Reliability of the clinical assessment in predicting the cause of internal derangements of the knee.

The reliability of the clinical assessment, which comprised history, physical examination, and plain radiographs, was determined by comparing the initial preoperative diagnosis with the postoperative diagnosis as determined arthroscopically, and by comparing the results of the clinical evaluation with published reports of arthrography, computed tomography (CT), and magnetic resonance imaging (MRI). The study group included 206 patients (216 knees) scheduled for arthroscopic surgery for suspected internal derangements. The primary preoperative clinical diagnosis was correct in 175 knees (81%), with an overall accuracy of 93%, sensitivity of 89%, and specificity of 94%. The most common preoperative diagnoses were medial and lateral meniscal tears. Results of the clinical assessment were comparable or superior to published reports of arthrography, CT, and MRI. Thus, it may be unnecessary to routinely use these costly special studies to determine the need for arthroscopic surgery. We conclude that a thorough clinical assessment can provide sufficient information for the surgeon to make a definitive primary preoperative diagnosis, and that arthroscopy should not be performed without first completing a complete preoperative examination.

The intercondylar fossa of the normal canine stifle an anatomic and radiographic study.

The intercondylar fossa (ICF) in dogs consists of a cranial outlet, intercondylar shelf, caudal arch, caudal outlet, a medial wall, and a lateral wall. The normal cranial outlet is bell-shape and, in mixed-breed dogs (mean body weight 19.2 kg, N = 21), measured 5.8 mm cranially, 8.1 mm centrally, and 10.3 mm caudally. The ICF is oriented 12 degrees from the dorsal plane of the femoral diaphysis and obliqued 7 degrees, proximolateral to distomedial, in the sagittal plane. To adjust for dog size, a fossa width index (FWI) was calculated by dividing the cranial outlet width by the distance between epicondyles. The normal FWI as determined in this study was 0.18 cranially, 0.25 centrally, and 0.32 caudally. The fossa height index 0.31. Contact between the ICF and the cranial cruciate ligament began at about 115 degrees of extension. The contact area moved cranially in the intercondylar fossa as the stifle was extended. Evaluation of the ICF can be performed radiographically but positioning is critical.

The effect of intercondylar notchplasty on the normal canine stifle.

This study evaluated the effect of notchplasty (enlargement of the intercondylar fossa) in stable and unstable canine stifles. Bilateral notchplasty and unilateral cranial cruciate ligament (CrCL) transection were performed in 6 dogs. Exercise, consisting of walking 1.5 miles three times a week, began 1 month after surgery and continued until euthanasia 6 months after surgery. Evaluation methods included orthopedic examination, serial radiographs, thin section radiography, histopathology, and gross pathology. Notchplasty in the stable stifle did not cause lameness beyond 3 weeks, joint instability, or degenerative joint disease. In the stable stifle, smooth resurfacing of the notchplasty site with fibrous and osseous tissue occurred. Stifles with notchplasty and CrCL transection exhibited persistent lameness, instability, and degenerative joint disease. In CrCL deficient stifles osteophytes formed within the notchplasty site, resulting in a rough surface. Our observations indicated significant refilling in notchplasties of both stable and unstable stifles (P < .05). However, the intercondylar fossa (ICF) width 6 months after notchplasty was significantly smaller in unstable stifles compared with stable stifles (P < .05) indicating that greater refilling of the notchplasties occurred in the unstable stifles. In clinical cases, notchplasty should be larger than the desired final result to accomodate the partial refilling that occurs even in stable stifles.

Arthroscopically treated tears of the glenoid labrum. Factors influencing outcome.

Labral tears and associated glenohumeral ligament injuries were classified and treated arthroscopically in 83 patients. Transverse labral tears were the most common. Subtle increased glenohumeral translation was noted in 24% of patients. Glenohumeral ligament injury was present in 58% of the shoulders, with the middle and superior ligaments or middle ligaments alone most commonly involved. These ligament injuries were classified into different groups based on severity. Partial excision of the torn labrum was the primary treatment in all patients and was combined with glenohumeral ligament repair in 9 patients. Seventy-eight patients were available for subjective follow-up evaluation (average, 3.2 years), while 53 patients were objectively evaluated (average followup, 2.8 years). There were 14% excellent, 71% satisfactory, and 15% poor results. The main factor associated with poor results was a grade III glenohumeral ligament injury in which only partial excision of the torn labrum was performed. There was a positive correlation between injury to both the superior and middle glenohumeral ligaments and increased anterior translation demonstrated by preoperative stability testing at 60 degrees of humeral abduction in neutral rotation. We believe glenohumeral ligament injury is often associated with labral tears and arthroscopic partial labral excision is only effective in those patients who have grade I or II (less severe) glenohumeral ligament injuries.

The quantitative measurement of normal passive medial and lateral patellar motion limits.

To quantify normal motion, medial and lateral passive patellar motion limits were measured in 67 high school athletes randomly selected from a group of 1340 athletes undergoing preseason physical examinations. Patellar displacement was measured at knee flexion angles of 0 degree and 35 degrees, using both a Patella Pusher (a hand-held force gauge) and a manual technique, and the results were compared. Demographic data and physical examination of the deceleration mechanism (Q angle, vastus medialis obliquus dysplasia, patella alta and baja, and valgus and varus alignment) were correlated with patellar motion limits. With the knee in extension, passive displacement of the patella averaged 9.6 mm medially and 5.4 mm laterally. In flexion, medial displacement averaged 9.4 mm and lateral displacement averaged 10.0 mm. No positive correlations were found between demographic data or deceleration mechanism examination parameters and patellar motion limits, suggesting that motion produced by the displacement force was limited by ligamentous restraints only. The clinical assessment of the passive limits of patellar motion should include examination at knee flexion angles of 0 degree and 35 degrees. The manually produced displacement was found to be more reproducible than displacement by the Patella Pusher (P < 0.05).

How iliotibial tract injuries of the knee combine with acute anterior cruciate ligament tears to influence abnormal anterior tibial displacement.

A knowledge of the patterns of injury to the components of the iliotibial tract allows a clearer interpretation of motion limits testing in patients with abnormal anterior tibial translation of the knee (anterior cruciate ligament-deficient knees). Eighty-two consecutive patients with acute knee injuries were classified as anteromedial-anterolateral rotatory instability (anterior cruciate ligament-deficient) based on the abnormal motion demonstrated by clinical examination tests for instability. At surgery, injuries to the intraarticular and extraarticular anatomic structures were identified and correlated to the abnormal grades of motion demonstrated by the knee motion limits examination. Tears of the anterior cruciate ligament occurred in 80 (98%) of the knees. However, the grade of abnormal motion demonstrated by the Lachman and pivot shift tests was quite variable. This variation did not correlate statistically with anterior cruciate ligament tears. Injuries to the anatomic components of the iliotibial tract were confirmed in 76 (93%) of the knees. These injuries correlated highly with variations in grades of abnormal motion detected by the following tests: lateral joint line opening at 30 degrees (r2 = 0.05); Lachman test (r2 = 0.08); pivot shift (r2 = 0.16); and anterior translation at 90 degrees of flexion (r2 = 0.34). Thus, injuries to the components of the iliotibial tract are thought to contribute to the variation in grades of abnormal motion observed in this complex subgroup of anterior tibial translation instabilities.

Distal femoral osteotomy for valgus deformity of the knee.

Distal femoral varus osteotomy and blade-plate fixation for valgus deformity of the knee proved effective in restoring axial alignment in 18 of 36 knees (34 patients). Patients were followed for an average of 5.4 years (range: 2 to 19 years). The osteotomies were performed on 14 men and 22 women (average age: 44 years; range: 14 to 77). The patients' average preoperative valgus deformity of the anatomical axis was 19.4 degrees (range: 8 degrees to 33 degrees). The surgical procedures performed were a medial closing wedge osteotomy (14 knees) and a lateral opening wedge osteotomy with bone grafting (22 knees). Postoperative correction of the anatomical axis averaged 3.8 degrees valgus (range: 8 degrees varus to 20 degrees valgus). Maximum improvement was reached within 6.3 months by patients who were less than 60 years old and within 5.1 months by patients who were more than 60 years old. Pain decreased or resolved in 21 of 35 knees (60%); activity level improved in 24 of 35 knees (69%). One patient was unavailable for follow up evaluation. Varus osteotomy in the distal femur was concluded to be an acceptable form of treatment in the valgus knee alone or associated with traumatic or osteoarthritis of the lateral compartment.

Accuracy of clinical isometry and preload testing during anterior cruciate ligament reconstruction.

To determine the effect of suture elongation on the accuracy of tension isometers, an anatomic specimen was used to measure peak loads on suture materials placed in extreme nonisometric positions. Accuracy and reproducibility were determined for three tension isometers, and the influence of suture elongation was determined for one of them. In one device, reproducible readings ranged from -12.3% to + 23.6% of mean static loads recorded by hydraulic testing equipment; average error approached 39.1%. Values tended to be overestimated if readings ranged as high as 32.5% over specific excursion ranges. Selected suture materials were subjected to these loads, using hydraulic testing equipment. Stainless steel exhibited extremely low levels of elongation but only at large diameters. Ethibond had similar low levels of elongation but only at large diameters. Monofilament nylon exhibited significant elongation that could cause erroneous readings and nonisometric placement. The measurement tolerances of a selected tension isometer should be considered by the surgeon performing anterior cruciate ligament (ACL) substitution. Several readings, both loading and unloading, can be used to determine preload and isometry. Material properties of the selected suture and the accuracy of the tension isometer should be considered when interpreting readings from isometric positioners. This brings into question the ability of currently available tension isometers to deliver readings within prescribed tolerances for either isometry of preload testing.

Analysis of subjective knee complaints using visual analog scales.

A questionnaire using a system of visual analog scales was developed for analyzing subjective knee complaints. This system was tested on 117 consecutive patients who had undergone knee surgery and 65 patients at their initial office evaluation of a knee disorder. The validity of and patient affinity for this type of questionnaire was compared with that of three other established subjective evaluation methods. The visual analog scale system was shown to be valid and comparable to other methods while offering several advantages. It brought greater sensitivity and greater statistical power to data collection and analysis by allowing a broader range of responses than did traditional categorical responses. It removed bias that was introduced by examiner questioning, and it allowed graphic temporal comparisons. Most importantly, patient affinity was higher for this type of subjective evaluation than for other methods.

The stabilizing function of passive shoulder restraints.

UNLABELLED: The static restraints of the scapulohumeral joint provide stability for the humeral head in the glenoid cavity, limit extremes of motion of the glenohumeral joint, and guide positioning of the humerus during normal shoulder movement. Eleven fresh-frozen cadaver shoulders of unknown age were attached to a shoulder motion device that allowed measurement of motion in three planes with an accuracy to 0.5 degrees. Four shoulders underwent motion analysis and seven were used for strain gauge analysis of the static scapulohumeral ligamentous restraints. The results of the motion analysis demonstrated that any attempt at simple motion (flexion, extension, abduction, internal or external rotation) resulted in coupled motion in two additional planes. The strain gauge data, expressed as a percent of total tension for each ligament tested, demonstrated a reciprocal tension-sharing relationship among all ligament components and a transference of tension among these components when original and new joint positions were compared. These data provide an in vitro model of shoulder restraint function to explain primary restraint, tension sharing, and transference of tension functions in the in vivo scapulohumeral joint. CLINICAL RELEVANCE: These principles of shoulder function have application in the treatment of instability and frozen shoulder syndrome, and provide an in vitro model to better understand static restraint function in the throwing mechanism.

The anatomy of the extensor mechanism.

The complex function of the extensor mechanism demands knowledge of not only its anatomic structures but of their balanced relationships as well. An awareness of a layer concept of tissue orientation is helpful in understanding surgical access as well as the normal function of the decelerator mechanism. This awareness is a prerequisite to the evaluation and treatment of mechanical disorders.

Office evaluation and management of the symptomatic knee.

In summary, data recorded during the office evaluation should include a chief complaint, a subjective history, a functional history, motion and strength analysis, and objective examination measurements. Storage and access of these data may be categorized on the basis of probable etiology. Finally, the historic and examination criteria for conservative or surgical care may be reviewed periodically to evaluate the results of a specific treatment and changes may be made as necessary to improve patient satisfaction and improve the physician's diagnostic and therapeutic capabilities.

Isolated chondral fractures of the knee.

Eighteen patients with isolated chondral fractures of the knee were reviewed to reevaluate the clinical signs and symptoms. No other conditions had been present to account for the symptoms. Most of the patients had significant injuries often involving a flexed knee. All patients had had symptoms and physical findings strongly suggestive of meniscal abnormality. Roentgenographic examinations were noncontributory. The lesions occurred in four distinct patterns. All were full thickness. Many of the lesions occurred on the posterior aspect of a femoral condyle, a location where they could be easily overlooked if the condyle were not examined with the knee in extreme flexion. Abutment of the tibial spine against the medial femoral condyle may be one cause of these injuries. The incidence of isolated chondral fractures was found to be higher than previously reported. A negative arthroscopic meniscal examination in a patient with symptoms and physical findings suggestive of meniscal abnormality should alert the physician to the possibility of an isolated chondral fracture.