Medial Collateral Ligament Reconstruction

Abstract

HISTORICAL PERSPECTIVE Medial elbow instability may be the result of one traumatic event or the cumulative effect of repetitive microtrauma. The anterior bundle of the ulnar collateral ligament (UCL) is the major stabilizing structure resisting valgus force at the elbow (1–7). In most cases isolated trauma to the ligament rarely leads to symptomatic instability. However, overhead or throwing athletes place repetitive high valgus stress on the elbow, which may produce symptoms of UCL insufficiency, resulting from a tear or attenuation of the ligament. Injury to this ligament occurs most commonly in baseball pitchers (8–12), but it has also been reported in several other sports. During the overhead throwing motion, valgus stress on the medial elbow begins during arm cocking and acceleration phase (Fig. 1), with the elbow flexed between 90° and 120°. This position is followed by a rapid extension, with a velocity of greater than 2,300°/s over 30 to 40 ms, to 25° of flexion at ball release. The average angular velocity over this arc of motion has been shown to be 5,000°/s, with peak elbow acceleration of 500,000°/s2 (13,14). Werner et al. (15) estimated the tremendous tensile force concentrated on the medial side of the elbow at nearly 290 N. The ultimate tensile forces of the UCL at resisting valgus torque have been reported to be 33 Newton-meters (N-m) (16). The estimated static demand on the UCL during pitching has been reported as 35 N-m (13) and may exceed the tensile strength of the ligament (17). The UCL is partly shielded from excess medial tensile forces through the combined efforts of the joint capsule, medial musculotendinous structures, geometric congruity of the articulation, and the dynamic compression of this highly congruous joint. However, the UCL is indeed the primary stabilizer against valgus stress, and with the tremendous forces generated at the medial elbow to throw an object repetitively at high speed or for long distances, it is easy to understand that the UCL is vulnerable to injury in overhand throwers.FIG. 1.: Valgus stress placed on the ulnar collateral ligament during arm cocking and acceleration phase in baseball pitching.INDICATIONS AND CONTRAINDICATIONS A history of pain associated with repetitive overhead throwing activities is the most common presentation. Definitive surgical treatment is necessary for athletes with an acute complete disruption of the ulnar collateral ligament or patients with recurrent pain secondary to an attenuated UCL and instability who show little or no improvement with conservative management and desire to return to play (9,11,18). Symptoms of ulnar nerve impairment occur in more than 40% of patients with UCL insufficiency (9). Currently, we believe that stabilization and ulnar nerve transposition are indicated for patients with ulnar nerve motor impairment or subluxation or both. If joint stability and pain relief are the main goals, nonoperative treatment usually suffices, particularly in nonthrowing athletes. In noncompetitive, occupational circumstances, reconstruction should not be performed until a trial of work modification and conservative measures has been attempted. If, however, the UCL injury is associated with gross instability, as seen with fractures of the coronoid, radial head, or retracted avulsion of the common flexor tendon, consideration should be given to early repair or reconstruction of the ligament combined with the associated repairs or fixation as indicated. PREOPERATIVE PLANNING A detailed history and physical examination are the keys to making the correct diagnosis of UCL injuries. The majority of these injuries are on a continuum of a valgus extension overload syndrome. When taking the history, it is important to ascertain what sport the athlete participates in, as well as the location of the elbow pain (i.e., medial, lateral, posterior, or a combination of these areas). It is important to record previous elbow injuries and treatment. The remainder of the history includes such information as frequency, duration, relation of symptoms to arm movement or position, and level of sports participation. Excessive repetitive stress, as seen in the late cocking and acceleration phases of pitching, can create overuse-related osteoligamentous injuries. These injuries occur on a spectrum of pathologies known as valgus extension overload (19) that have been reviewed and classified (20) (Table 1). The process begins with excessive tension overload on the medial aspect of the elbow, resulting in UCL injury, posteromedial impingement (Fig. 2) of the olecranon, followed by lateral compression injury to the radiocapitellar joint.Table 1: Valgus extension overload syndromeFIG. 2.: Posteromedial impingement resulting from attenuation of the ulnar collateral ligament accentuated by excessive valgus stress and internal rotation of the humerus.A focused physical examination of the elbow includes all the elements of any joint examination—inspection, palpation, motion measurement, stability testing, strength testing, and neurovascular testing—and special tests that are specific to the elbow joint. On physical examination a patient with a torn UCL will usually have soft tissue swelling over the medial flexor mass in an acute injury and may experience tenderness to palpation over the UCL complex, depending on the degree of inflammation at the time of examination. Such inflammation may also irritate or compress the ulnar nerve locally as it transverses the cubital tunnel (21). A positive Tinel's sign along the cubital tunnel is an indication of this focal nerve irritability. A careful neurologic examination of the upper extremity should be performed. Elbow flexion deformities are often found but usually do not adversely affect performance because the throwing motion involves elbow positions between 120° and 20° of flexion (15). Special test that are specific to evaluate UCL injuries are the O'Brien “milking maneuver” (Fig. 3) and the valgus laxity test described by Jobe (22) (Fig. 4). If posterior or posteromedial elbow pain exists with throwing, further investigation for osteophytes or olecranon fossa scarring should be undertaken using the combination of computed tomography (CT) scan, magnetic resonance imaging (MRI), and elbow arthroscopy to fully evaluate the extent of posterior impingement resulting from the valgus extension overload (19,20). Arthroscopic evaluation can also be helpful when the cause of medial elbow pain is in question. Medial joint line opening of greater than 2 mm to valgus stress may indicate UCL insufficiency but should not be used as sole the criterion for ligament reconstruction. Ellenbecker et al. (23) have shown that laxity does not necessarily correlate with symptoms.FIG. 3.: “Milking maneuver” usually elicits pain at the site of the ulnar collateral ligament.FIG. 4.: Valgus laxity test.Diagnostic studies begin with plain roentgenograms and stress films to rule out fracture, infection, and other pathologic conditions. Radiographic tests are helpful if positive, but a negative study should not rule out the diagnosis of UCL insufficiency. Standard radiographs (anteroposterior, true lateral, cubital tunnel, and axial views) may identify ossification within the UCL, loose bodies in the posterior compartment, marginal osteophytes about the radiocapitellar or the ulnohumeral articulations, olecranon and condylar hypertrophy, or osteochondral lesions of the capitellum. Stress radiographs, using gravity or specialized instrumentation (Fig. 5), can show excessive medial joint line opening, indicating ligament laxity. CT scans are most useful for evaluating fractures, loose bodies, and osseous changes seen with osteochondritis dissecans. Recently, magnetic MRI and saline magnetic resonance arthrography have become the standard to investigate soft-tissue conditions around the elbow (Fig. 6), specifically UCL injuries (24,25) (Fig. 7).FIG. 5.: Valgus stress radiograph showing increased opening of the medial joint line.FIG. 6.: Magnetic resonance imaging scan of normal ulnar collateral ligament. Arrowhead represents normal fat signal lateral to the ligament. Open arrows represent intact ligament.FIG. 7.: Magnetic resonance imaging scan showing torn ulnar collateral ligament. Arrows represent torn ends of ligament.TECHNIQUE Surgical Anatomy The medial epicondyle of the humerus begins to ossify at approximately 5 years of age, with blood supply generated by several anastomoses of the inferior ulnar collateral artery and posterior ulnar recurrent artery. The surgical anatomy involved in the reconstruction of the ulnar collateral ligament focuses on key landmarks surrounding the medial epicondyle. The location of the antebrachial cutaneous nerve lying within the superficial fascia of flexor mass, distal to the medial epicondyle, must be identified. The medial epicondyle serves as a source of attachment of three layers (Fig. 8). The superficial layer is composed of the common flexor tendons of the pronator teres, flexor carpi radialis, palmaris longus, and the flexor carpi ulnaris, all of which arise from the medial supracondylar ridge. The intermediate layer consists of the deeper fibers of the flexor tendon, primarily those of the deep humeral origin of the pronator teres. These fibers arise from the base of the medial epicondyle as well as from the anterior bundle of the UCL itself. The deepest layer of the medial elbow is made up of the medial collateral ligament and the medial elbow capsule. The anterior bundle of the UCL is composed of an anterior and posterior band and is the principal static stabilizing ligament to valgus stress, whereas the common flexor tendon functions as a dynamic stabilizer to such stress. Davidson et al. (26) described the functional anatomy of the flexor muscle groups in relation to the ulnar collateral ligament, showing the flexor carpi ulnaris and the flexor digitorum superficialis muscles to be the predominant musculotendinous unit overlying the ulnar collateral ligament. The UCL complex is deep to the common flexor tendon and courses from the anteroinferior surface of the medial epicondyle to the sublime tubercle on the medial aspect of the coronoid process of the ulna. This ligament is the primary medial stabilizer of the elbow (4).FIG. 8.: Three layers of the medial elbow. I = superficial layer: common flexor tendons; II = intermediate layer: deeper fibers of flexor tendons, primarily those of the pronator teres; III = deep layer: ulnar collateral ligament and capsule.The goal of medial collateral ligament reconstruction is to replicate the restraining effects of the anterior bundle of the UCL. The ideal reconstruction should restore the biomechanical properties of the native ligament. A successful reconstruction should respect three principles: (1) restore the native anterior band's tensile properties over a full arc of elbow flexion, (2) provide adequate tensile strength, and (3) utilize a material that can undergo repetitive stresses without significant creep or stress relaxation. The tension generated in the UCL complex over the arc of elbow flexion has been previously studied and is related to the unique anatomic features of the anterior bundle, which is the primary restraint to valgus stress (3). This anterior bundle is functionally divided into separate anterior and posterior bands. The isometric fibers of the anterior bundle lie between its anterior one third and posterior two thirds (27). Under valgus load, the anterior band is taut while the elbow is flexed from 0° to 85°, and the posterior band becomes taut at 55° of flexion and remains taut throughout the remainder of flexion. In this way, one or both of the two bands of the anterior bundle are taut throughout the full arc of flexion. A surgical reconstruction of the medial collateral ligament should restore this balance of tension throughout the full arc of flexion without limiting motion of the elbow. Another important principle of medial collateral ligament reconstruction is that the reconstructed restraint to valgus torque must exceed the tensile forces generated at the medial elbow during the acceleration phase of pitching. Shortly before maximal shoulder external rotation, a large valgus torque is produced at the elbow and is counterbalanced by a varus torque estimated to be between 52 and 76 N-m (average: 64 N-m) (13). An in vitro study by Morrey and An (3) showed that the UCL contributes approximately 54% of the resistance to valgus torque. If one accepts these findings, it can be estimated that the varus torque generated by the UCL is between 28 and 41 N-m. This value approaches the findings of Dillman et al. (16) that the failure load of the UCL is 32 N-m. With such a narrow margin for failure, the importance of adequate tensile strength in the medial collateral ligament reconstruction becomes obvious. Finally, because a gradual increase in the resting length of the medial construct would adversely affect a thrower's performance, a reconstruction of the UCL should utilize a material that can undergo repetitive stresses without undergoing significant creep. The significance of creep and stress relaxation has been demonstrated in other ligament reconstruction procedures and should be considered as it applies to UCL reconstruction (28,29). For example, during anterior cruciate ligament reconstruction, the initial force applied to tension the graft decreases over time as a result of stress relaxation. This viscoelastic phenomenon reduces the patellar tendon graft forces to approximately 80% of their initial value after 1 h (30). Therefore, the amount of tension remaining in the graft after 1 h depends on the viscoelastic characteristics of the graft. Finally, if a reconstruction involves allograft tissue, the viscoelastic effect known as preconditioning must be considered. After prolonged periods of inactivity (hours), allograft tissues imbibe fluid, which is then extruded during the first few applications of force to the tissue. Therefore, the first few cycles after inactivity demonstrate greater stiffness and hysteresis energy loss than the subsequent cycles. Although no study has compared the viscoelastic properties of differing materials in UCL reconstruction, clinical studies have demonstrated good and excellent long-term results of 80% with autogenous tissue (palmaris longus or plantaris tendon) passed through bone tunnels (9). Although autogenous palmaris tendon graft is the standard tissue for reconstruction, alternative sources of tendon have been proposed and include the contralateral palmaris longus tendon, the plantaris tendon, and the extensor tendon to the fourth toe (9,31,32). Authors' Preferred Method Once the decision to perform medial collateral ligament reconstruction has been made, the first step is to select the graft source. The presence or absence of the palmaris longus tendon should be documented. When this tendon is available, it is the tissue of choice for reconstruction. Bilateral absence of the palmaris longus has been reported in approximately 6% to 25% of the general population (33–35). In revision cases where the palmaris longus may have been previously harvested, the autogenous plantaris tendon can be used. Also before surgery, the ulnar nerve should be palpated in elbow flexion and extension to confirm its location and to ensure that it does not subluxate anteriorly. The patient should be questioned regarding any previous surgical procedures to the involved extremity and the elbow should be inspected for any unexplained scars or obvious deformities. Ensuring proper patient positioning is an important first step when preparing to perform UCL reconstruction. The patient should be supine with the arm abducted to the side on a Parker arm board. External rotation of the humerus should afford the surgeon easy access to the posteromedial elbow. Placing a folded towel under the elbow can facilitate this position. UCL reconstruction should be performed under general endotracheal anesthesia with pneumatic tourniquet control for hemostasis. The tourniquet should be placed proximally enough on the upper arm so as to allow access to the arcade of Struthers, should the decision be made to transpose the ulnar nerve. The entire upper extremity is prepared and the arm is draped free so that the entire volar forearm to the palm is exposed. The position of the ulnar nerve should be confirmed by palpating the cubital tunnel in elbow flexion and extension one final time before making the skin incision. The surgeon may choose to use a marking pen to reference the landmarks of the medial elbow (medial epicondyle, ulnar nerve, medial olecranon). A curvilinear incision is then made centered over, or just slightly anterior to, the medial epicondyle. The length of the incision may vary according to the procedure performed (ulnar nerve transposition may require more proximal extension of the incision), but a length of 8 to 10 cm is generally sufficient. Care is taken to protect the medial antebrachial cutaneous nerve as it passes across the medial aspect of the proximal forearm (Fig. 9). If there has been a previous transposition of the ulnar nerve, the nerve is identified, dissected free, and protected before the procedure is continued. To expose the ulnar collateral ligament, a longitudinal split is made from the medial epicondyle distally, passing through the myofascia, the underlying flexor-pronator aponeurosis, and the muscle mass until the fibers of the UCL are visualized. Using this muscle-splitting approach (36,37), retraction of the flexor mass to both sides provides access and exposure to the anterior portion of the UCL complex (Fig. 10). Some deep fibers of the flexor-pronator mass originate off the anterior bundle and often an elevator can be used to lift these fibers away and expose the more distal aspect of the ligament.FIG. 9.: Superficial dissection. Care must be taken to identify the branches of the medial antebrachial cutaneous nerves in the fascia overlying the flexor muscles.FIG. 10.: Muscle-splitting approach to ulnar collateral ligament. “Safe zone”(inset) is identified 1.8 cm anterior to medial epicondyle and 1.0 cm distal to sublime tubercle of the ulna.Figure 10: ContinuedAfter the exposure is complete, the pathology in the ligament can be assessed. With a complete tear, the joint may be exposed and the injury becomes immediately obvious. Often the ligament's external surface appears normal, and in these situations a longitudinal incision is made, splitting the ligament in line with its fibers from its origin on the medial epicondyle to its insertion on the sublimus tubercle. This step will expose the undersurface of the ligament and can often reveal internal degeneration or partial detachment of the undersurface of the ligament. With the joint exposed, the elbow is flexed between 20° and 30° and a valgus stress is applied. With insufficiency of the ligament, the ulnohumeral articulation will gap open greater than 2 mm (Fig. 11).FIG. 11.: Anterior band of ulnar collateral ligament is split, exposing the medial joint. With valgus stress applied, opening of greater than 2 mm constitutes ligament injury.If necessary, posterior olecranon osteophytes can be removed at this time through a vertically oriented, posterior arthrotomy. Care must be taken to protect the ulnar nerve while a quarter-inch osteotome and a rongeur are used to remove the tip of the olecranon or debride posterior loose bodies. The arthrotomy is closed with absorbable suture. The anterior bundle's insertion site is then identified and converging 3.2-mm drill holes are made just anterior and posterior to the sublime tubercle, leaving a bone bridge between the two holes. The drill holes should be oriented at right angles to each other and care must be taken to stay approximately 5 mm distal to the articular surface. A curette is then used to connect the tunnels at their apex. To expose the medial supracondylar ridge, the flexor mass is split longitudinally at its anterior, proximal origin, and the humeral tunnels are drilled. Once this has been done, divergent holes can be drilled from distal to proximal, placing the apex at the ligament's origin on the medial epicondyle (Fig. 12).FIG. 12.: Drill hole placement in the sublime tubercle and medial epicondyle. Inset: drill hole placement in the anteroposterior and lateral views.Figure 12: ContinuedAttention should now be turned to harvesting the palmaris longus tendon graft. With the forearm fully supinated, the tendon should be identified superficially and just ulnar to the flexor carpi radialis. A 2-cm transverse incision is then made at the level of the distal flexor wrist crease directly over the tendon. A hemostat is used to spread the subcutaneous tissues, as the tendon is isolated from surrounding structures. The tendon is palpated proximally and a second transverse incision is made approximately 8 cm from the distal incision, again directly over the palmaris longus tendon. The tendon is transected distally and brought out through the proximal incision. Serial transverse incisions are made more proximally as needed, in a fashion similar to the first and second incisions, until the level of the musculotendinous junction. Usually three incisions are required. The tendon is divided at the musculotendinous junction to give a length of 15 to 20 cm. The incisions are irrigated and closed routinely. A 1-0 nonabsorbable braided suture is placed in one end of the graft using a locking stitch (Fig. 13).FIG. 13.: Harvesting the palmaris longus tendon using the three-incision technique.The graft is then passed through the ulnar tunnel using a hooked suture-passer and weaved through the humeral tunnels in a figure-of-eight fashion across the joint. With a standard-length autogenous graft, an effort is made to pull the excess graft end back into the ulnar tunnel, creating a three-ply reconstruction. Care is taken to bury the tendon ends in the tunnels. The elbow is then held in approximately 60° of flexion, the medial elbow joint is compressed by applying a varus stress to the elbow, and the graft is pulled taut. The sutures on the humeral side are sewn to the tissue of the medial intermuscular septum, while the ulnar side is sutured to the remaining tagged leaflets of the native UCL that were previously split. The graft is further tensioned and secured by suturing the strands together near the mouth of the humeral tunnel and near the sublime tubercle of the ulna, while also incorporating the remnants of the native UCL into the new graft, itself providing additional strength (Fig. 14). The elbow is ranged to verify isometricity and to check for abrasion of the graft on the joint line.FIG. 14.: Three-ply reconstruction of the ulnar collateral ligament. Tension sutures are placed proximally and distally through the graft and soft tissue.The wound is thoroughly irrigated and the myofascia at the split in the flexor-pronator mass are then sewn side-to-side with absorbable suture. The tourniquet is then released, hemostasis is obtained with electrocautery, and the wound is closed in layers. The skin is re-approximated in a subcuticular fashion using 3-0 nylon suture. A sterile dressing is applied, followed by a well-padded posterior splint with the elbow in 90° of flexion and neutral rotation, leaving the wrist and hand free. Other Methods In a cadaveric study, Hechtman et al. (32) have described a reconstruction of the UCL of the elbow using bone anchors. These authors attempted to develop a reconstruction technique that would lessen the amount of soft tissue dissection as well as the potential for ulnar nerve complications compared with the bone-tunnel technique described by Jobe et al. (11). Using a loop of a palmaris longus tendon graft secured over a bony trough by four bone anchors, biomechanical load-to-failure testing of each reconstruction method was compared with the original strength of each elbow's native UCL. When the elbows underwent a single valgus load at 30° of flexion, the intact ligament was significantly stronger (p < 0.0005) than both reconstruction methods. The ratio of reconstructed over intact ligament strength was slightly higher in the bone-tunnel group compared with the bone-anchor group, but this was not found to be statistically significant (76% vs. 64%, respectively). Clinical results using the bone anchor technique will be necessary before its widespread use as an alternative to the clinically proven bone-tunnel reconstruction methods. To develop a reconstruction method that would facilitate more constant graft tensioning, Altchek et al. (38) have described a construct that combines bone-tunnel and bone-bridge suture security (Fig. 15). Using the same ulnar tunnels as those described by Jobe et al. (11), a palmaris longus autograft is passed through the bone and the free ends are docked into a single tunnel drilled into the medial humeral epicondyle. This technique allows each graft end to be tensioned individually. Fixation on the humeral side depends on suture tied over a bone bridge. Clinical results using this procedure have approached those obtained by Conway et al. (9). As described, this produces a two-ply graft across the medial joint.FIG. 15.: “Docking” technique using suture security.ANATOMIC INTERFERENCE RECONSTRUCTION Most recently, we have conducted a cadaveric study using interference fixation to secure a three-ply palmaris longus tendon graft at the origin and insertion of the native UCL anterior bundle. The procedure is described as follows. The native UCL complex is exposed through the flexor-pronator mass using the authors' preferred muscle-splitting approach described above. The anterior bundle of the UCL is sharply incised along its longitudinal axis, with care taken not to disrupt the native ligament's origin and insertion. The joint line of the medial elbow is exposed and injury to the ligament is assessed. Once the decision to proceed with reconstruction of the UCL has been made, the sublime tubercle is identified and a guide pin is inserted into the bone at the insertion site of the midpoint of the anterior bundle. The guide pin is angled distally 45° to the long axis of the native ligament, at a distance 5 mm from the articular surface. A 4-mm cannulated drill bit is inserted over the guide pin and a tunnel is drilled into the ulna to a depth of 2.5 cm. Attention is then turned to the medial epicondyle and a guide wire is inserted into the point of origin of the anterior bundle in a retrograde fashion. The surgeon should confirm that the insertion site is medial on the epicondyle at the ligament's native origin. The pin should be parallel to the fibers of the anterior bundle as it is inserted. The pin should be directed so that proximally it exits the epicondyle anterior to the intermuscular septum and clear of the ulnar nerve. The same 4-mm cannulated drill bit is inserted over the guide pin and the humeral tunnel is drilled. The palmaris longus tendon is harvested from the ipsilateral forearm as described above. The tendon is folded over itself to create a three-ply graft. A 2-0 monofilament suture is used to secure one end of the graft into a single bundle using a standard whipstitch. Using a notched guide wire, the secured end of the graft is inserted into ulnar bone tunnel to its full depth. The surgeon must ensure that the tendon is fully seated within the tunnel. A 5 × 20-mm soft tissue interference screw is inserted over the guide wire. With the ulnar side secure, a 2-0 monofilament suture is used to independently secure the proximal free ends of the graft. A suture-passer is passed through the medial epicondyle from proximal to distal and is used to pull the suture ends into the humeral tunnel. The graft is then tensioned as the proximal ends of the ligament are pulled in a colinear fashion. The elbow is flexed and extended to confirm a physiologic construct. Graft tightening as the elbow is flexed to 90° confirms appropriate tension. A guide pin is passed into the humeral tunnel alongside the graft and a 5 × 20-mm soft tissue interference screw is used to secure the reconstruction as the elbow is maintained in 60° of flexion and a varus force is applied (Fig. 16). A 1-0 absorbable suture is then used to oversew the remaining native UCL into the reconstructed graft.FIG. 16.: Anatomic interference reconstruction (AIR) technique.COMPLICATIONS Complications occur in approximately 9% of patients (31). Superficial infections involving the graft harvest site occurred in approximately 5% of patients, all resolving with oral antibiotics. Postoperative problems involving the branches of the medial antebrachial cutaneous nerve or the ulnar nerve are the most common complications after ulnar ligament reconstruction in conjunction with nerve transposition, and may occur in up to 25% of patients (9). Careful subcutaneous dissection may prevent inadvertent transection of cutaneous nerves causing local paresthesias or painful neuroma formation. Meticulous hemostasis may help prevent compression of the ulnar nerve secondary to hematoma formation. Segmental devascularization of the transposed ulnar nerve may play a major role in the occurrence of postoperative ulnar nerve palsy (39). Therefore, in patients requiring ulnar nerve transposition, gentle handling of the ulnar nerve and preservation of as much of its vascular leash as possible may limit the incidence of this problem. Currently, the ulnar nerve is only transposed if the patient has symptoms and documented motor ulnar neuropathy, or when pathology in the posterior compartment requires exposure through the cubital tunnel. Rupture or stretch of the reconstruction is possible but is uncommon if three strands of the graft cross the joint. If medial instability recurs, reoperation is indicated and a new autogenous graft is used to construct the deficient ligament. POSTOPERATIVE MANAGEMENT The postoperative management for reconstruction of the UCL begins with gentle handgrip exercises. The patient squeezes a sponge or soft ball as soon as it is comfortable to do so. Immobilization is discontinued at approximately 10 days, and active wrist, elbow, and shoulder range of motion exercises are initiated. No brace is used after surgery. After 4 to 6 weeks strengthening exercises are begun, avoiding valgus stress until postoperative 4 months. At 4 months the patient is usually allowed to toss a ball 30 to 40 feet, two to three times a week for about 15 min. At 5 months the patient may increase the tossing distance to 60 feet, and at 6 months the patient may perform throwing lightly from the wind-up. At 7 months a graduated program of range of motion, strengthening, and total-body conditioning exercises is performed. Throwers and pitchers are limited to throwing one-half speed, while gradually increasing the duration of their session to 25 to 30 min. Pitchers are permitted to throw from the pitching mound and progress to 70% of maximum velocity during the eighth or ninth month. Over the next 2 to 3 months, the duration of throwing sessions and velocity are slowly increased to simulate a game situation. Throwing in competition is permitted at 1 year if the shoulder, elbow, and forearm are pain free while throwing and have returned to normal strength and range of motion. Throughout the rehabilitation phase, careful supervision and focus on body and throwing biomechanics should be emphasized. In a professional pitcher, it may require more than 18 months to regain preoperative ability and competitive level, with relatively shorter periods required for other player positions or overhead sports (11). RESULTS Reconstruction of the UCL in overhead or throwing athletes unable to compete secondary to valgus instability of the elbow has successfully returned most patients to their previous level of participation (9,11,31). At the Kerlan-Jobe Orthopaedic Clinic, of 70 operations performed on 68 patients with valgus instability of the elbow, 56 athletes had a reconstruction of the torn or incomplete ligament using a free tendon graft (9). These patients were followed for 6.3 years after surgery. Sixty-eight percent of the reconstruction group returned to their previous level of participation. The mean time to return to competition was 12 months. Azar et al. (31) performed 78 reconstructions of the UCL in overhand athletes, and 59 patients were available for follow-up. Average follow-up was 35.4 months. The average time from surgery to the return to competitive throwing was 9.8 months, with professional baseball players requiring 1 year of rehabilitation before their return to the competitive arena. Forty-eight of the 59 patients with reconstruction returned to their previous level of competition or to higher levels. If sport-specific participation and player position are considered, it appears that baseball pitchers are the most difficult to return to their previous competitive level, with 62% reporting excellent results (9). The length of time from onset of symptoms to operation (acute vs. chronic) and the type of ligament injury (avulsion vs. attenuation) did not affect postoperative outcome. With increasing experience in the reconstruction of the UCL and the development of new techniques, we at the Kerlan-Jobe Orthopaedic Clinic believe that there is an increased ability to return the athlete to the previous level of participation for a prolonged period of time.