This study is a retrospective review of transmetatarsal amputation (TMA) outcomes in patients with diabetes and non-healing ulcers of the forefoot. All were treated by single stage TMA and insertion of antibiotic beads in the surgical wound. Healing time was approximately eighteen weeks with a failure rate (subsequent BKA) of 25%. This is in contrast to literature values of healing times (not isolated to diabetics) of twenty-eight weeks and BKA rates of 34–40%. The decreased morbidity associated with our surgical procedure may address the costly management of diabetic foot ulcers (presently estimated to be $600 million per year).

The purpose of this study was to review the outcome of transmetatarsal amputation (TMA) in diabetic patients as a single stage procedure using antibiotic pellets in the wound.

We report faster healing times and a decreased rate of subsequent below-knee amputation (BKA) when compared with related studies in the literature.

1) Our procedure may decrease morbidity in a problem (diabetic foot ulcer management) costing approximately $600 million per year. 2) This study uniquely addresses TMA in diabetics.

Mean healing time was eighteen weeks (range six to forty weeks). Patients with intact foot pulses or reconstructed vasculature had a mean recovery time of 12.5 weeks. Healing times for unreconstructable vasculature or documented deep infection were thirty weeks and twenty weeks respectively. The overall failure rate (BKA) was 25%. Neither vascular status nor the presence of deep infection predicted subsequent BKA.

Retrospective review of patient charts. Forty consecutive diabetic patients (mean age 58.3 yrs, range 40–77) with foot ulcers of > twelve weeks duration had TMA performed at a tertiary care center by one of three surgeons. Data tabulated included demographics, diabetes profiling, vascular interventions and follow-up parameters.

Diabetic foot ulcer morbidity is a significant cost burden to health care; despite this, salvage procedures for this problem are not well studied. Previous papers (not isolated to diabetics) report TMA healing times of twenty-eight weeks and subsequent BKA rates of 34–40%. Refinements of the TMA technique in diabetics may decrease early and late morbidity and thus address this costly problem.

Testing of cadaver ankle specimens was conducted to investigate the changes in kinematics with lateral ligament reconstructions. Testing included an intact condition, after injury at the ATFL and CFL sites, and separately a Brostrom repair and an anatomical gracil-lis graft reconstruction. Calcaneal range of motion was determined about the axis of applied moment in plantarflexion-dorsiflexion and in inversion-eversion directions. The injury and reconstructions were most sensitive during IE applied moment. Both reconstructions appeared to behave similar to intact motion. Failure of some Brostrom repairs however, suggest that the gracillis-graft reconstruction is initially a stronger repair.

Limited research has biomechanically investigated lateral ankle ligament reconstruction procedures. The objective of this study was to determine the changes in ankle kinematics with a dual ligament Brostrom repair and an anatomical gracillis graft reconstruction.

Seven cadaveric ankle specimens were tested independently in an intact condition, after an ATFL/CFL injury model, and two reconstructions. The anatomical graft reconstruction wove a gracillis tendon through the calcaneus and fibula to dually reconstruct the ATFL and CFL, and anchored to the talus. Moments were applied to the calcaneus for three cycles in plantarflexion-dorsiflexion (PD) and inversion-eversion (IE) while allowing unconstrained motion. Three dimensional motions of the calcaneus and tibia were optoelectronically tracked. Range of motion (ROM) was calculated about the axis of applied moment for the calcaneus with respect to the tibia.

The ROM increase from the intact condition with the injury model was only significant for IE (p=0.001). No significant differences were found between intact and any treatments in the PD configuration. In IE, both the graft reconstruction and the Brostrom repair were significantly different from the injury model (p=0.002 and p=0.015 respectively), where the gracillis reconstruction appears more similar to the intact condition. For two specimens the Brostrom repaired ATFL failed during applied inversion moment.

The injury and reconstructions were most sensitive during IE applied moment. Both reconstructions appeared to behave similar to the intact condition. Failure of some Brostrom repairs however, suggest that the gracillis-graft reconstruction is initially a stronger repair.

Funding: Workers Compensation Board of British Columbia