One of the most controversial issues in total knee replacement is whether or not to resurface the patella. In order to determine the effects of different designs of femoral component on the conformity of the patellofemoral joint, five different knee prostheses were investigated. These were Low Contact Stress, the Miller-Galante II, the NexGen, the Porous-Coated Anatomic, and the Total Condylar prostheses. Three-dimensional models of the prostheses and a native patella were developed and assessed by computer. The conformity of the curvature of the five different prosthetic femoral components to their corresponding patellar implants and to the native patella at different angles of flexion was assessed by measuring the angles of intersection of tangential lines.

The Total Condylar prosthesis had the lowest conformity with the native patella (mean 8.58°; 0.14° to 29.9°) and with its own patellar component (mean 11.36°; 0.55° to 39.19°). In the other four prostheses, the conformity was better (mean 2.25°; 0.02° to 10.52°) when articulated with the corresponding patellar component. The Porous-Coated Anatomic femoral component showed better conformity (mean 6.51°; 0.07° to 9.89°) than the Miller-Galante II prosthesis (mean 11.20°; 5.80° to 16.72°) when tested with the native patella. Although the Nexgen prosthesis had less conformity with the native patella at a low angle of flexion, this improved at mid (mean 3.57°; 1.40° to 4.56°) or high angles of flexion (mean 4.54°; 0.91° to 9.39°), respectively. The Low Contact Stress femoral component had the best conformity with the native patella (mean 2.39°; 0.04° to 4.56°). There was no significant difference (p > 0.208) between the conformity when tested with the native patella or its own patellar component at any angle of flexion.

The geometry of the anterior flange of a femoral component affects the conformity of the patellofemoral joint when articulating with the native patella. A more anatomical design of femoral component is preferable if the surgeon decides not to resurface the patella at the time of operation.

Thoracolumbar junction of the spinal column is the common site of spinal trauma and is often complicated by neurological dysfunction. From 1997 to 2000, there were 12 patients surgically stabilized. 8 of them were victims of major trauma while the rest was after a trivial fall in osteoporotic spine. Lengthen of follow-up ranged from 6 to 42 months.

Ages of the patients in the major trauma group were from 22 to 65. Except the one who had anterior approach because of multiple level lesions, all fractures after major trauma were initially relocated and stabilized posteriorly. Subsequent anterior procedures were necessary in three of them because of significant residual spinal canal stenosis. All except one had satisfactory lower limb function on follow up. Two patients who were paralysed on admission were able to walk independently and 4 others had improved by at least one Frankel grade. Return of neurological function was usually observed within the first week after the procedure. Residual sphincter dysfunction was however, a common problem.

The management of four osteoporotic spinal fractures in thoracolumbar junction was more unpredictable. Patients were from 66 to 92 years old. Anterior decompression was often performed because of the presence of retropulsed fragment. Although some improvement of lower limb function could be achieved, rehabilitation in three of them was complicated by loss of reduction or failure of the implant.

Recovery of the neurological function in the lower limbs was found to have no correlation with the amount of stenosis of the spinal canal. Most of the damage occurred probably at the time of injury. The sphincter control was most difficult to rehabilitate after an insult to the conus medullaris.