Appraisal of literature
All studies identified by the above search were case series and case reports. Inclusion criteria in the series was generally poorly described. Study number were small with the largest series being only 37 patients. The tears were assessed by a number of modalities with no standard assessment utilised. Patient reported outcome measures were inconsistently used. As a consequence of this heterogeneity of patient groups, imaging and outcomes reported, reliable synthesis of data is not possible. As such the data is reported in a descriptive manner. Four distinct entities were identified. Acute tears, chronic tears, partial tears (either acute of chronic) and snapping triceps.
The triceps brachii is composed of three muscle bellies. The long head arises off the infraglenoid tubercle of the scapula, the medial head off the posterior aspect of the distal to the spiral groove and the lateral head off the lateral intermuscular septum and the posterolateral aspect of the humerus above the spiral groove. The triceps inserts as a bilaminate tendon over a wide area onto the tip of the olecranon. The average triceps width at its distal insertion, including the tendon and lateral expansion, has been reported as 40.6 mm. With the tendon being a mean width of 30.6 mm and the distance from its medial edge to the ulna nerve at 10.2 mm. Its footprint has a mean length of 22.5 mm and a width of 22.7 mm, giving the footprint an area of 466.2 mm2. In around half of specimens a discrete tendinous portion of the medial triceps deep to the long and lateral head has been reported in cadaveric dissection, with the long and lateral portion forming a tendon superficial to it. The mean dimensions of this medial head insertion were 16 mm × 4 mm, with a mean area of insertion of 44 mm2. This study also demonstrated a second variant in the other half of specimens. This had a common combined tendon insertion, but still with the medial fibres deep to the long and lateral head.
Partial tears can be located on the superficial tendon only (combined lateral or long head)[6-8] but have also been described for the deep portion of the tendon in isolation (medial head)[5,9,10] the central one third of the tendon in isolation, the lateral portion in isolation, the lateral and central portion combined or involving the medial and long head insertions with the lateral portion intact[10,11].
Triceps tendon ruptures most commonly affect men in the 4th and 5th decades of life. From our pooled results we observed the mean age at rupture to be 47.5 years with a range from 12 to 75 years[1,3,5,8-28]. There was an 11:1 ratio of males to females (n = 170)[1,3,5,8-29]. Three paediatric cases have been reported in the literature in 12 and 16 year olds[11,22,26].
Triceps tendon ruptures are reported to be more common in systemic disease or entities such as chronic renal failure with secondary hypo-parathyroidism, rheumatoid arthritis and diabetes. Localised infection has also been reported as a cause of rupture. Interestingly chronic pre-existing posterior elbow pain may be present in up to a third of the patients and with histology demonstrating chronic tendinopathy features in patients with pre-existing pain. In traumatic ruptures, one study observed intra-articular fractures or collateral ligament ruptures in 3 of 6 patients (50%) and advised arthroscopic assessment of the joint at the same time as repair. Concurrent olecranon fractures and ulna neuropathy secondary to haematoma have also been observed with traumatic ruptures. Bilateral tears have been reported in patients with a history of anabolic steroid use[26-28].
Mechanism of injury
The two described mechanisms for acute tears of the triceps are direct contact trauma, such as a fall or hitting fixed resistance with the posterior elbow[5,14,19,21,25]. Weightlifting[1,5,17,19,23-25,27,30] was the most common sport associated with acute tears and was often associated with a history of steroid use[1,23,30]. American football[3,29] and general sports injuries, as well as direct lacerations have all also been reported as mechanisms.
History and examination form the mainstay of diagnosis. Patients typically present with ecchymosis, pain, swelling, extension lag and a decreased active range of motion[10,19,21,22,31]. Palpable defects are commonly found and present in up to 80% of patients[10,21,22,31]. Partial tears may be easily missed as patients may have a good range of active motion, but do typically present with reduced power on extension of the elbow[5,6,9,15,18,20,31].
A bony fleck proximal to the olecranon is commonly identified on lateral radiographs[10,12,17,23,25,27] and is strongly suggestive of a triceps avulsion injury The bony fleck may also be demonstrated on ultrasound (US)[7,8]. This radiological sign reflects the intra-operative observation that 33%-73% of patients have avulsion fractures off the olecranon. The remainder of patients have a rupture at the bone tendon junction[19,25]. Triceps ruptures can be missed on X-ray[22,25] and both US and magnetic resonance imaging (MRI) have been used to diagnosis complete tears and partial tears[7,17,19,25,27]. US has been reported to be as accurate as MRI for both complete and partial ruptures, including identifying the location of partial rupture[8,22].
Partial tears: Good results of non-operative treatment even in patients with high functional demands have been reported for the treatment of partial tears. Treatment involves physiotherapy, bracing and avoidance of heavy lifting, pushing or resisted extension for a period of up to 12 wk[27,31]. Harris et al reported good results in a single high demand patient with bilateral tears treated with a conservative regime. Platelet rich plasma (PRP) injection has also been advocated for the treatment of partial tears[18,29]. Cheatham et al reported the results of a single patient treated with PRP with resolution of pain and return to gym 4 mo after a PRP and physiotherapy regime.
Acute tears: Early primary repair is indicated for complete acute ruptures. Surgery is preferably preformed within the first three weeks of injury. Numerous surgical techniques have been reported. These include suture anchors in the triceps footprint[5,6,21,25,32], bone tunnels[10,14,20,22,31,33] or a combination of both tunnels and anchors[15,19,33]. Suture anchors repairs have also been performed arthroscopically[5,6,32].
Bava et al reported the results of 5 male patients treated with an open suture anchor technique. They reported a mean American Shoulder and Elbow Surgeons elbow score of 99.2 at an average of 32 mo follow-up. Lempainen et al also reported excellent results in three athletes treated with an open suture anchor technique for acute tendon rupture. van Reit and Neuerman report on the largest series of open bone tunnel repairs. van Reit et al reported the results of 14 patients treated for acute tendon rupture with a bone tunnel technique. Triceps strength was noted to be 4/5 or 5/5 on manual testing in all examined subjects post operatively. Isokinetic testing of ten patients showed that peak strength was, on the average, 82% of that of the untreated extremity. Three re-ruptures were reported in their series. van Riet et al reported the results of six patients treated with a bone tunnel technique and reported good to excellent results at 12 mo post op with no cases of re-rupture. Kokkalis et al, Paniago et al and Paci et al, all report good results utilising a combined bone tunnel and suture anchor technique in three separate case reports with no adverse advents reported.
Techniques utilising arthroscopic repairs for acute injuries have been reported by Athwal et al and Ng et al. Athwal et al reported good results for the arthroscopic repair of two acute tears treated with an arthroscopic technique at 2 years’ follow-up with respect DASH and Mayo elbow scores. Ng et al did not report outcomes for their technique.
Because of the heterogeneity of types of tear, repair techniques, and outcome measures it is impossible to determine superiority of one technique over another. Generally surgery has been reported to give good improvement in the Morrey score, Mayo, ASES, Oxford elbow scores and DASH[6,9,14,21,29]. It is reported to significantly reduce pain and improve muscle strength[6,10,19]. Our pooled results have demonstrated that in three studies, all patients achieved full extension in some studies with flexion to at least 110 degrees[10,14,15], whereas others achieve an average loss of 7-10 degrees of extension[10,19]. Three of 14 (21%) re-ruptures were reported in van Reit’s study, one suture anchor pullout was been reported and needed revision, and one asymptomatic partial re-rupture was detected on post-operative MRI. Otherwise no re-ruptures were reported or commented on.
Chronic tears: The management of chronic tendon ruptures is a challenge. Occasionally it is possible to repair the tendon using one of the techniques described above. For chronic ruptures with significant tendon retraction, reconstruction with a graft may be required. Grafts include achilles allograft[11,13] or ipsilateral semi-tendinous tendon[10,15], aconeus[10,11], Latisimus Dorsi, plantaris and palmaris longus. Aconeus and palmaris longus have also been used in augmentation of primary repairs. Sanchez-Sotelo et al reports the largest case series of chronic rupture reconstruction from the Mayo clinic. They utilised an achilles tendon allograft in three cases or an anconeus muscle flap in four cases. One rotation flap failed six months after operation. At an average of 33 mo follow-up the remaining six patients had no or slight pain, restoration of a functional arc of movement and normal or slightly decreased power of extension.
Petre et al reported the intact triceps tendon has a peak load to failure of 1741N, in comparision to peak load to failure of 317N, and 593N for a direct repair, and augmented repair tendons respectively. Comparison made between peak load to failure for a trans-osseous cruciate suture technique and a two bone tunnel and knotless sutures technique, showed a significant difference of 510N for the knotless technique and 283N for the cruciate technique. A similar peak load to failure of 317N was found for the same trans-osseous cruciate suture technique, but found to be less than the peak load of 593N when the repair was augmented with interwoven flexor carpi radialis. Interestingly Yeh et al observed no significant difference in the peak load to failure for an “anatomic” double row trans-osseous repair, the trans-osseous cruciate suture technique and a 2 suture anchor with Krakow-type whip-stitch technique.
The snapping triceps occurs in a younger population than triceps tendon rupture with a mean age of 32 years (n = 30), ranging from 14-65 years[36-42] and slightly reduced male to female ratio of 6.5:1[36-42].
Snapping triceps is a dynamic condition occurring during flexion or extension of the elbow and is characterised by a snap on both active and passive movement. The triceps can dislocate on either the medial or lateral side, but is much more common medially. It can be asymptomatic, cause snapping, elbow pain or ulna neuropathy if dislocating medially[42-44]. Snapping triceps has been demonstrated to be bilateral in some patients, but is not necessarily symptomatic on both sides. On the medial side, the snapping can be attributed to the ulna nerve, but snapping may still occur despite ulna nerve transposition. Snapping triceps may be associated with dislocation of the ulna nerve. Spinner et al reported all 17 patients in his series (100%) having concurrent dislocation of the ulna nerve with the snapping triceps. Spinner postulated that a snapping ulna nerve and snapping triceps could be differentiated by the angle at which the snapping occurred. The ulna nerve is thought to snap at 70-90 degrees of flexion, whereas the triceps is thought to snap at around 115 degrees of flexion.
Snapping triceps is thought to be due to the medial vector placed on triceps that can occur in cubitus varus and is not thought to be associated with rotational deformities or muscle activation patterns. This medial vector is a function of the T angle where the T angle is the angle between the subtended line of pull of triceps (humeral shaft with extended elbow) and the longitudinal line of proximal ulna.
On the medial side snapping triceps can be a complication of displaced supracondylar fractures[36,45], inherited as an accessory medial triceps or abnormal insertion[43,46], due to hypertrophy of the medial triceps in athletes[43,47], associated with hypermobility of the ulna nerve, associated with osseous abnormalities and as a complication of ulna nerve transposition[48,49]. On the lateral side it has been associated with a widened triceps tendon inserting more laterally and is treated with resection of the lateral edge.
US, MRI, CT and sonoelastography have been used for diagnosis[37-39,43,50]. Ultrasound is the imaging modality of choice of some as can be used as a dynamically to differentiate between a snapping medial triceps and a subluxing medial nerve.
Initially conservative treatment can be attempted with NSAIDS and avoidance of provoking activities for 3-6 mo. If this is unsuccessful surgery can be considered. Surgery can include resection of the triceps edge, transposition of the tendon, transposing an associated ulnar nerve and correction of cubitus varus[43,45]. Transposition involves transferring the medial third of the tendon to the lateral position[36,47].