Abstract
BACKGROUND: Glenoid loosening remains a concern in anatomical total shoulder replacement. Preoperative planning software allows optimization of the component positioning, but the target orientation remains unclear due to conflicting optimization priorities. Commonly, the component is aligned to the prescribed version and inclination that reflect the population's average anatomy. The freehand technique attempts to secure strong fixation by aiming to preserve the subchondral bone. This study compared the state of the subchondral plate after reaming and compared the results of these two techniques. METHODS: Two groups of shoulder computed tomography scans were assessed, 34 normal and 34 osteoarthritic. Preoperative planning software was used to place the glenoid component in prescribed adjusted angles, with neutral (0(o),0(o)), retroverted (-10(o),0(o)) and inclined alignment (0(o),10(o)). The computed tomography Hounsfield values at the virtually reamed surface were assessed to determine the percentage of the intended component-bone interface consisting of cortical bone, here termed "cortical bone seating". This was then compared to positioning the component using a freehand technique. RESULTS: The freehand technique improved cortical bone seating in the osteoarthritic group with a mean (standard deviation) of 53.3% (14.3), while neutral alignment resulted in 36.7% (10.8), retroversion 40.4% (13.1), and inclination 39.3% (13.5), P < .001. A similar trend was observed in the normal group. CONCLUSIONS: The freehand method resulted in significantly improved cortical bone seating compared to the prescribed adjusted angles. These findings question the use of a one-size-fits-all-orientation and suggest that applying a technique that aims for maximum cortical fixation (freehand) may reduce the risk of aseptic loosening.