UW Shoulder and Elbow Academy: The humeral component: what's in a stem?

One of the key elements in placing the humeral component is assuring the proper relationship of the articular surface to the tuberosity.

Here is the x-ray of a patient in whom the left humeral component was positioned elsewhere in a non-anatomic relationship with the greater tuberosity. This too-superior position resulted from the use of a stem that fit tightly in the diaphysis, preventing full seating.

We had the opportunity to perform a ream and run on this patient's contralateral shoulder using a smaller diameter stem and impaction grafting. The postoperative x-ray shows the desired relationship between the humeral prosthesis and the tuberosity.

It is of note that this positioning was achieved with a thinner stem than that used on the left side - a shorter stem was not needed. (See "Total shoulder and short stems: less is not more" at this link).

Some other differences in stem design are of note: the right sided prosthesis has a tapered metaphysis which enables impaction grafting; this is in contrast to the cylindrical stem used on the left which does not allow impaction grafting. Because of the ability to use impaction grafting, the right prosthesis does not need the bone ingrowth surface that is present on the left. Such a bone ingrowth surface can make revision complicated by risking tuberosity fracture.

It is of note that surgeons often use templating to determine the size of the prosthetic stem (see this link). The problem is that the medial-lateral width of the canal seen on the standard AP view is often greater than the AP width seen on the lateral view. Consider the two cross sections at the top of the figure below that demonstrate the elliptical nature of the canal. The bottom two cross sections show that reaming of the canal removes bone preferentially from the anterior and posterior aspects of the diaphysis.

This asymmetrical bone removal can create an endosteal notch distally, predisposing the humerus to fracture.

Here are some additional thoughts about fixation of the head component to the humeral bone

The methods by which the humeral head component is fixed to the humeral bone range from stemless and mini-stemmed to long stemmed and include fixation with cement, press fit, tissue ingrowth and impaction grafting.

There are at least three issues to consider. (1) The shape of the humeral canal is curved longitudinally and is elliptical in cross section; some canals are cylindrical while others are funnel shaped. (2) Fixation needs to be achieved without risk of loosening on one hand and without risk of fracture on the other. Intramedullary reaming to achieve a diaphyseal press fit preferentially removes bone from the anterior and posterior endosteal cortex; this endosteal notching at the prosthesis tip creates a risk of periprosthetic fracture. (3) Any shoulder arthroplasty is at potential risk - as high as 10% - for revision surgery to manage infection, loosening, malposition, fracture or instability, so that it is important that the humeral component be completely removable without seriously damaging the humeral bone stock.

We have found that these three issues are best addressed by a 120-140 mm long smooth (non-ingrowth) humeral stem that fits in the canal without diaphyseal reaming and that is fixed with impaction autografting using bone harvested from the resected humeral head rather than with cement. Cancellous allograft is used in revision cases or where the amount of bone available for grafting is insufficient

Because it is difficult to achieve secure impaction with a cylindrical humeral body, such as that shown below

we prefer a stem with a proximal flare: the metaphyseal portion of the prosthetic body is thicker in the anteroposterior and medial-lateral dimensions than the diaphyseal portion - a configuration that enables secure fixation without driving the stem tightly into the humeral cortex distally.

The effectiveness of this technique has been documented by a laboratory study showing the increment in quality of fit and fill and by clinical follow-up studies. This method of conservative reaming and broaching combined with impaction grafting avoids the most common problem with press-fit humeral components: too high positioning of the prosthesis because of incarceration of the distal end of the stem in the humeral diaphysis leaving the prosthesis prominent with respect to the tuberosities and the glenoid as shown in the cases below.

Tip incarceration and incomplete seating of the humeral component may result from choosing a stem sized based on the AP x-ray, which usually shows a wider canal width than that on the axillary because of the oval shape of the humeral canal. Obviously, the problem of a too high and tight stem fit cannot be solved by trying to hammer the prosthesis down further, rather our solution is to change to a smaller implant stem diameter and fill the void by impaction grafting. We like to say that the impaction grafting approach allows the surgeon to ‘get it wrong but still get it right’. In a way this is similar to the method used by Procrustes, the legendary Greek innkeeper, who fit his ‘guest’ to his bed, rather than fitting the bed to the guest. If the prosthesis sits too far distal, the problem is solved by adding more autograft. If it sits too far proximal, a smaller stem can be used. If the position of the prosthesis is not ideal, it can be fine-tuned by selective placement of the graft. If an impaction grafted, uncemented, non-porous coated stem needs to be revised, it can usually be disimpacted without concern about cement removal and without the need to perform a humeral osteotomy as may be necessary to remove a prosthesis with ingrowth surfaces, trabecular metal, or platform (modular) stems. In special situations it may be necessary to saw off the tip of the prosthesis to the necessary length and fix it with cement

We reserve the use of long stems for cases in which an area of cortical weakness needs to be bypassed as shown below.

We do not find systems with ‘platform’ stems attractive because many failures of anatomic arthroplasties are related to improper placement of the stem as shown below.

It seems doubtful that a platform design would facilitate revision of the cases shown below to a reverse.

As long as the stem is not cemented and does not have a bony ingrowth surface, conversion to a reverse is usually straightforward

While some systems offer adjustable neck shaft angles, we have found this added complexity to be unnecessary in that a standard 45 degree cut can accommodate normal, varus and valgus anatomy.