Tornier BluePrint

The Aequalis shoulder replacement (Tornier) is a popular, reliable system with runs on the board. Tornier have now developed a computerised tool called BluePrint that allows the surgeon to play with CT scans in two and three dimensions to best position the glenoid component within the bone

Why?

Two main reasons: minimisation of the glenoid face ream and optimisation of bone stock volume in the glenoid vault. Both have proven (or at least generally agreed) implications on longevity/ survivorship of the implant. Given that approximately 40% of osteoarthritic glenoids show posterior wear (Walch B or C - noting that the classification has recently been refined) means that a significant proportion of patients will require some adjustment in the “angle of attack” to correct for this when placing the critical guide pin that determines how the glenoid is prepared. This is much harder to do on the table with CT scans illuminated in the background than it is with a computerised system set up to make a custom guide (custom guides for guide pins being the most prominent of the set of “patient specific instrumentation” which has become the rage in arthroplasty surgery). Ultimately shoulder replacement surgery will progress to on-table computer guidance systems (Exactech provides the only navigated system currently available) but Tornier does not (yet) provide this option.

Although I feel compelled to follow the advice of the development team that researches, reports, and advances the shoulder system I use I think there are two counterarguments against a full-hearted adoption of BluePrint in its current guise. The first is obvious: BluePrint (and all the other competing systems out there) only takes data from CT scans to create patient-specific bone models but does not take account of the surrounding soft tissue. Soft tissue tension is critical for stability and function in shoulder replacements (more so in anatomic TSR than reverse TSR). Soft tissue restraints also play a role in the exposure of the glenoid which influences how the glenoid is attacked. I’ve seen enough live surgery events from experienced surgeons to know that I’m not the only one who (occasionally) modifies their angle of attack to accommodate for this (especially in the elderly patient with osteoporotic bone). To say that shoulder replacement surgery is a balance of getting the best compromise given a number of suboptimal variables would not be far off the truth. BluePrint is a tool that optimises bony variables but does not take into account soft tissue variables. 

The second problem is that BluePrint does not account for how the scapula sits in relation to the chest wall. In BluePrint the scapula and its articulation with the proximal humerus is taken in isolation for modelling purposes. How that articulation sits on the chest wall also has implications that we are only now starting to understand (eg glenoid notching in reverse TSR).

Another thing lurks in the background. The surgeon - and the orthopod in particular - prides himself on his spatial orientation, his technical ability. and the mastery of his tools. Years of practice hones what natural talent (or lack thereof) preceded it. Guided systems diminishes that role and takes away a skillset that future surgeons may never develop to full potential. 

Surgeons have bodged shoulder replacements into patients with generally good results since the mid 1970’s. If there is adequate bone stock, an active deltoid, and an intact rotator cuff then we’re good to go. Even better when the patient is elderly with low physical demands. The developments in shoulder arthroplasty over the past forty years have refined our approach (surgical technique and hardware configuration) based on a growing dataset of long term outcomes. There remain problem areas such as glenoid volume and wear patterns which programs like BluePrint attempt to address.

It is fair to say that orthopaedic surgery with its plethora of ever-evolving tools and hardware has a proclivity to yield to market forces that are not necessarily in the interest of the patient (or society which ultimately bears the cost of such developments). But it’s also fair to say that using a computer to create virtual models of a patient’s anatomy/ pathology, though still unproven, gives us a better handle of the task at hand. 

Another example of the endless march of the machine? Well, computers are more exacting when posed with a problem where there are known and quantifiable parameters.

No doubt computerised systems (with on-table navigation and robotic assistance as we are starting to see in knee replacement surgery) will become gold-standard practice in the not so distant future where people live longer, in better health, and with greater demands. Machine learning AI may account for - and teach us about - areas we currently have difficulty quantifying: optimal soft tissue balance, joint stability, and humeral version. The question is not whether we will see semi-autonomous (or even fully autonomous) robots in the surgical theatre but when.

End of the Anthropocene. 

Dawn of the Machine Age.

Computerised modelling like the Tornier’s Blueprint is a step in the right direction. But for now, the human brain, with all its foibles and fallibility, remains best placed to process the optimal (compromised) solution to accomodate all the (perceived) suboptimal variables. 

Watch this space.