To determine union rate in complicated nonunions of the scaphoid treated with a vascularised bone graft.

Vascularised bone grafting for scaphoid nonunions (1–2 ICSRA, Zaidemberg technique) has shown initial enthusiasm. Its usefulness has been challenged in cases where the proximal pole of the scaphoid is avascular. Complicated nonunions where the proximal pole is highly likely to be avascular occur in revision surgery and proximal pole nonunions.

Fourteen patients were retrospectively followed up. Eight had nonunion following previous scaphoid surgery (two previous ORIF, two previous nonvascular grafting, and four with two previous surgeries). Six patients had no previous surgery for a proximal pole nonunion of 12.5 months’ duration. All patients were male with an average age of twenty-four. Delay from fracture to vascularised bone grafting was twenty months. Graft harvesting was done according to the Zaidemberg technique by two orthopaedic surgeons. CT-scan was used to confirm union in all patients except two who were lost of the follow-up. Twelve patients were followed up by an independent surgeon at a postoperative minimal period of four months. Functional status was assessed with the DASH questionnaire and follow x-rays were performed to determine the presence of degenerative changes.

Union was confirmed by CT-scan in eleven of twelve followed patients (92%) at an average time of six months following vascularised graft. Radio-scaphoid osteoarthritis was seen in the one patient that didn’t achieve union.

This series suggests that the Zaidemberg graft is useful and may be proposed in situations of revision surgery and proximal pole non-unions. We achieved a high union rate in these complicated nonunions even though there was high likelihood that the proximal pole was avascular. This study stresses the importance of protective immobilization until documented union by CT-scan in this difficult subset of patients.

Lacerations of the FDP tendon in zone one may be reattached to bone with a modified Bunnell pullout suture or with suture anchors. Eleven cadaveric fingers were submitted to cyclical testing of five hundred cycles with either a modified Bunnell pullout suture of 3-0 polypropylene or a micro-Mitek suture anchor with 3-0 Ethibond. Gap formation was 6.6mm in the modified Bunnell group and 2.0mm in the micro-Mitek group (p< 0.001). Load to failure was 37.6N in the pullout group and 28.5N in the anchor group (p< 0.005). Gap in the pullout group and low failure load in the anchor group are of concern.

Distal zone one FDP tendon lacerations are usually re-attached to bone by a modified Bunnell pullout suture of 3-0 polypropylene. This treatment may lead to moderate to severe losses of DIP joint motion in up to 50% of patients. Suture anchors have recently been introduced as a fixation alternative. Cyclical testing simulating five days of a passive mobilisation protocol was used to compare the Micro-Mitek anchor to the modified-Bunnell pullout suture in FDP tendon fixation.

Eleven cadaveric fingers FDP tendons were repaired to bone using a modified Bunnell pullout suture of 3-0 polypropylene or a micro-Mitek anchor with 3-0 Ethibond. Testing was done from 2N to 15N at 5N/sec, for a total of five hundred cycles. Gap formation at the tendon bone interface was measured. Load-to-failure was performed on all specimens.

No specimens failed during cyclic testing. Gap formation was 6.6mm (SD 1.2, range 4.9–8.2mm) and 2.0mm (SD = 0.4, range 1.7–2.7mm) for the pullout technique and the micro-Mitek anchor repair respectively (p< 0.001). Load to failure data was 37.6N (SD 4.7, range 31.8–45.1N) for the pullout group and 28.5N (SD 4.0, range 21.8–33.4N) for the micro-Mitek group (p< 0.005).

This data suggests that both fixation techniques may be adequate to sustain five days of simulated passive rehabilitation therapy. Significant gap formation in the modified Bunnell pullout group is of concern although this needs to be correlated in the clinical setting. The lower failure rate of the micro-Mitek group may leave a narrow margin of safety for passive rehabilitation.

The behaviour of two different methods of reattachment of the flexor digitorum profundus tendon insertion was assessed. Cyclical testing simulating the first 5 days of a passive mobilisation protocol was used to compare the micro Mitek anchor to the modified-Bunnell pull-out suture. Twelve fresh-frozen cadaveric fingers were dissected to the insertion of the FDP tendon. The FDP insertion was then sharply dissected from the distal phalanx and repaired using one of two methods: group 1 -modified Bunnell pullout suture using 3/0 Prolene; group 2 micro Mitek anchor loaded with 3/0 Ethibond inserted into the distal phalanx. Each repaired finger was mounted on to a material testing machine using pneumatic clamps. We cyclically tested the repair between 2N and 15N using a load control of 5N/s for a total of 500 cycles. Gap formation at the tendon bone interface was measured every 100 cycles.

No specimens failed during cyclical testing. After 500 cycles, gap formation of the tendon-bone interface was 6.6mm (SD = 1.2mm), and 2.1 mm (SD = 0.3mm) for the pullout technique and the micro Mitek anchor repair respectively. Concerns related to suture anchors, such as anchor failure or protrusion, joint penetration, and anchor-suture junction failure, were not encountered in this study.

Cyclical loading results suggest that the repair achieved with both methods of fixation is sufficient to avoid failure. However, significant gap formation at the tendon-bone interface in the modified Bunnell group is of concern, suggesting it may not be the ideal fixation method.