Introduction

Direct skeletal attachment of prosthetic limbs, commonly known as osseointegration (“OI”), is being investigated by our team with the goal of safely introducing this technology into the United States for human use. OI technology allows for anchorage of prosthetic devices directly to bone using an intramedullary stem. For OI to be effective and secure, bone ingrowth and remodeling around the implant must be achieved. Physicians need an effective way to measure bone remodeling in order to make informed decisions on prescribed loading. This work describes methodology that was developed that utilizes computed tomography (CT) imaging as a tool for analyzing bone remodeling around an osseointegrated implant.

Background

Synthetic interbody spinal fusion devices are used to restore and maintain disc height and ensure proper vertebral alignment. These devices are often filled with autograft bone to facilitate bone bridging through the device while providing mechanical stability. Nonporous polyetheretherketone (PEEK) devices are widely used clinically for such procedures.¹Trabecular Metal devices are an alternative, fabricated from porous tantalum. It was hypothesized that the porous Trabecular Metal device would better maintain autograft viability through the center of the device, the ‘graft hole’ (GH).

Aim: To measure the percentage increase in length of the donor graft during rehabilitation from 0 degrees to 120 degrees of elbow flexion and to compare this with the end range strength.

Method: During the troids procedure four metal skin-clips were inserted at the proximal and distal margins of the proximal and distal tibialis posterior tendon weaves creating three intervals for measurement. Lateral x-rays of the humerus with the tube distance at 100cms were taken after five weeks of plaster immobilisation before elbow flexion commenced at a maximum rate of 15 degrees per week. X-rays were repeated when 60 degrees and 120 degrees of flexion obtained and when possible six months post surgery. Elbow torque was measured by the Troidometer throughout the range of motion at similar time intervals. Interval measurement was by a Vidar VRX 12 digital scanner. The Paired T test was used for statistical analysis.

Results: Sixteen arms (nine patients) were entered into the study but complete rehabilitation data were available from only 12 arms and late data from only four. There was a 12.3% average increase between 0 degrees and 60 degrees (range six to 20.6 degrees) and a further 3% increase from 60 degrees to 120 degrees (range −6 degrees to 21 degrees). The most stretch occurred in the distal segment and in bilateral arms the percentage of stretch was similar for each arm. There was no correlation between the percentage of stretch and end range torque or lag.

Conclusions: The Troids transfer restores elbow extension for tetraplegics but an extensor lag often develops which is thought to be from stretching of the donor tendon graft. This study confirmed that tendon stretch occurs but there was wide variation among individuals although similar for each arm in bilateral procedures. An average of 75% of stretch occurred during the 0 degrees to 60 degrees mobilisation. We concluded that tendon stretch is inevitable but is not a major contributor to end range weakness or lag.

We reviewed the results of reconstruction of 97 upper limbs in a consecutive series of 57 tetraplegic patients, treated from 1982 to 1990. Of these, 49 had functional and eight had cosmetic reconstructions. The principal functional objectives were to provide active elbow extension, hook grip, and key pinch. Elbow extension was provided in 34 limbs, using deltoid-to-triceps transfer. Hook grip was provided in 58 limbs, mostly using extensor carpi radialis longus to flexor pollicis longus transfer, and key pinch in 68, mostly using brachioradialis to flexor pollicis longus transfer. Many other procedures were employed. At an average follow-up of 37 months, 70% had good or excellent subjective results, and objective measurements of function compared favourably with other series. Revisions were required for 11 active transfers and three tenodeses, while complications included rupture of anastomoses and problems with thumb interphalangeal joint stabilisation and wound healing. We report a reliable clinical method for differentiating between the activity of extensor carpi radialis longus and brevis and describe a successful new split flexor pollicis longus tenodesis for stabilising the thumb interphalangeal joint. Bilateral simultaneous surgery gave generally better results than did unilateral surgery.