[Biomechanical comparative study of three types of osteosynthesis in the treatment of supra and intercondylar fractures of the humerus in adults].

PURPOSE OF THE STUDY: Supra and intercondylar fractures are the most common fractures of the distal end of the humerus in adult. An osteosynthesis consisting of a plate is the treatment of choice. But location and type of plate always remain open for debate. The authors present the results of an in vitro biomechanical study, which compared the stiffness of three types of osteosynthesis commonly used in these fractures.

MATERIAL AND METHODS: The devices were on one hand the premolded lateral plate of Lecestre and Dupont (Howmedica) used alone or in conjunction with a medial 1/3 tubular plate of the AO group, and on the other hand a posterior plate, of which we are developing a new model, the Lambda plate (Protek) "Y"-shaped, monoblock, flat and molded on the humerus during operation. The study compared these different methods of fixation on fresh human humeri. In a first part, the posterior plate was compared to the single lateral one; in a second part, the posterior plate was compared to the coupled lateral and medial plates. The three plates were made of identical material. Both studies used eight pairs of bones with supra and intercondylar fractures realized by sawing. Each bone of a pair was fixed with one of the two types of device. Each humerus underwent different loading forces; sagittal bending (anterior and posterior) and torsion. The displacements were recorded using a calibrated measuring device coupled to two displacement sensors. The stiffness was calculated on force/displacement curves.

RESULTS: In anterior bending, the Lambda posterior fixation was significantly stiffer than the single lateral one (p < 0.05) (239 +/- 109 versus 129 +/- 65 N/mm), and was not significantly different of the bilateral fixation (229 +/- 93 versus 224 +/- 108 N/mm). In posterior bending, the mean stiffness of the Lambda fixation was not significantly different to that of the lateral fixation (91 +/- 27 versus 91 +/- 52 N/mm), and less than that of the bilateral one (130 +/- 39 versus 170 +/- 70 N/mm), but not significantly. In torsion the mean stiffness of the Lambda fixation was superior to that of the lateral one (146.75 +/- 50.66 versus 119.75 +/- 58.8 Nm/rad), and bilateral one (233.31 +/- 107.47 versus 212.31 +/- 113.55 Nm/rad), but again not significantly.

DISCUSSION: The ideal osteosynthesis for the fractures of the humeral distal doesn't exist, because the bone undergoes antero-posterior and posteroanterior cyclical forces during elbow flexion. Therefore the best device should be placed on both sides of the bone, but anatomical reasons make this location impossible. According to our study, we think the "less worst" device is the posterior one using the Lambda plate. Its symmetrical design allows a best loading distribution on the two columns. Its thickness compensates for its posterior location and the short interval between two holes allows to put many screws (4 to 6) into the epiphysis. All supra and intercondylar fractures, comminuted or not, can be treated with this material. The single lateral device isn't still enough. Its stiffness mainly depends on the orientation of the oblique screw in the medial column. But the design of this column doesn't always allow for an optimal location of the screw, which is the reason of several failures. The bilateral device gives a stiffness, comparable to the posterior one, but doesn't allow as many screws as the Lambda plate in the epiphysis. Therefore, it can't be used in very distal fractures.

CONCLUSION: In spite of the progress of the material and the accuracy of the indications, the treatment of supra and intercondylar fractures of the distal end of the humerus is always a difficult problem. The devices have to be as stable as possible to allow an early motion. The best one should be placed on both sides of the distal humerus, because of the sagittal cyclic forces it undergoes, but this location is anatomically impossible. Our study concludes that the device using the Lambda plate i