Introduction Non-invasive prediction of load bearing capacity is an important issue in the advanced clinical treatment of distraction osteogenesis in order to define the appropriate point of time for the removal of the external fixateur. Therefore, non-invasive stiffness measurements were recommended as a promising tool due to the high correlation between strength and various kinds of stiffness: Torsional, bending and compressive.

However, previous experiments only analysed the relationship between a single type of stiffness. This approach neglects the multi-dimensional characteristics of bone loading in compression, bending and torsion.

This study investigates how compressive, bending (ap and ml) and torsional stiffness are related to the torsional load bearing capacity of healing callus tissue using a common set of bone regenerate samples of sheep treated with distraction osteogenesis. In addition, this study compares the evolution of the various kinds of stiffness.

This study provides insight into how the various stiffness modes are suited to predict the load bearing capacity by in-vivo stiffness measurement.

Material and Methods Mid-diaphyseal osteotomies were performed in 26 right tibiae of mature, female domestic sheep. Tibiae were then stabilized using an external half-ring Ilizarov fixator. After a 4-day latency period the tibiae were distracted at a rate of 1.25 mm per day in two increments for 20 days. As a result of a parallel study, the callus was treated with different combinations of growth factors and carrier material resulting in four treatment groups plus a contralateral control group. The sheep were sacrificed and the tibiae were harvested on the 74th day.

The ends of the tibiae were embedded in PMMA and mounted to a sequence of special custom made jigs for compressive testing, 4-point-bending and torsion in a material testing machine.

Stiffness was calculated by regression of the initial linear part of the load-displacement curves.

In a final experiment, the specimens were loaded in torsion until failure to record the ultimate torsional moment.

Results Torsional stiffness exhibits the highest correlation with the ultimate torsional moment (r2 = 0.77), while the ones for compressive (r2 = 0.60) and bending (ap (r2 = 0.70); ml (r2 = 0.66)) are only slightly lower.

Discussion This ex-vivo study in sheep shows that torsional, bending (ap and ml) and compressive stiffness measurements are all suitable means to predict the load bearing capacity of healing callus tissue. Our results show that torsional stiffness measurements perform slightly better than compressive and bending stiffness measurements. However, further studies are necessary to underline the superior performance of torsional stiffness measurements, since the sheep-tibiae were failed by applying torsional stress.