The Pennington's modification of the Kessler repair: unexpected observations through 3-D modelling

Objectives/Interrogation: The modified Kessler repair as described by Pennington is possibly the most widely used method for tendon repair. It is based on the creation of loops within the tendon, that tighten when a longitudinal force is applied. In his original paper, Pennington used a 'finger and twine' model to demonstrate this concept. He observed that in this configuration the longitudinal part of the string was superficial to the transverse part. He translated this configuration to his tendon suture technique and stated that "placement of the transverse part of the suture superficial to the longitudinal part would result in such a loop configuration". However, on dynamic fluoroscopic studies we observed that in the suture configuration obtained using the Pennington's technique did not reveal the expected behaviour. The suture was seen to fail without demonstrating the grasping effect of the loops around the tendon substance. Our aim was to perform a detailed analysis of the suture configuration and its behaviour under traction.

Methods: We recreated three models

1. The finger and twine model described in Pennington's original paper.
2. Tendon suture configuration described by Pennington on a transparent silicone cylinder in his original paper.
3. Three-dimensional computer generated of the suture configurations of the finger and twine model and the Pennington's suture technique to analyse the differences.
4. Solid models of the two configurations using malleable material.

Based on these models we qualitatively analysed the fundamental topological differences between the loop configuration seen in the finger and twine model and the actual configuration produced by the suturing technique.

Results and Conclusions: We found that the suture configuration described by Pennington although resembled the loop configuration seen in the 'finger and twine' model, there were some marked differences. The 'finger and twine' model demonstrated the topological properties which allowed the twine to tighten around the fingers, however when it was translated on to the tendon, the intended configuration was lost. The sole condition of placement of transverse part of the suture superficial to the longitudinal part, as stated by Pennington, was not sufficient for formation of true loops, resulting in the observed behaviour of the repair under longitudinal strain. Through the use of computer modelling, we were able to analyse and point out the fundamental topological conditions that were necessary form effective loop configurations.