Allograft prosthetic composite RSA surgery uses telescope and flange techniques

Proximal humeral bone loss is commonly encountered in the arthroplasty setting and this can provide a challenging problem for reconstructive shoulder surgeons.

Massive bone loss can be seen in posttraumatic cases, revision arthroplasty and following tumor resection. Lack of proximal bone can lead to problems related to poor soft tissue tensioning, which is paramount in the success of a reverse shoulder arthroplasty. If the bone loss is not addressed appropriately, this can subsequently lead to prosthetic joint instability, implant loosening and poor function.

There are various ways in which to address proximal humeral bone loss, including an endoprosthesis and allograft prosthetic composite (APC). The APC provides certain benefits which make it an appealing option, although there are certainly technical challenges with this procedure. The important advantages include re-establishing the deltoid wrapping effect — potentially enhancing stability, providing an attachment site for tendons and, although the literature is limited, higher postoperative functional scores compared with an endoprosthesis. Newer-generation endoprostheses seek to improve on all these aspects over traditional designs, so future comparative studies with APC will be helpful.

The purpose of this article is to provide some pearls and helpful tips in regard to performing an APC procedure with RSA.

When preparing for an APC, preoperative planning is essential. This typically includes critical assessment of radiographs of the involved shoulder, as well as obtaining ipsilateral and contralateral full-length humerus films to determine the premorbid bone length (Figures 1, 2). Preoperative labs and constant vigilance are important as the APC procedure should only be performed in the aseptic setting and many revision shoulders with bone loss may also have concomitant infection. Lastly, it is important to assess the integrity of the deltoid insertion radiographically as its absence will severely hamper the results of either APC or endoprosthesis, which is important for patient expectations.

We recommend a technique taught by Mark D. Lazarus MD, which uses an ipsilateral male allograft proximal humerus for female patients and an ipsilateral female proximal femur allograft for male patients to obtain a larger graft than the host bone. Frozen allograft is preferred vs. fresh allograft due to its cost and availability. Given that the articular surface of the allograft will be removed when placing the RSA prosthesis, there is no need to obtain a cartilage-friendly fresh allograft. We also request from the tissue supplier that the tendons stumps of the rotator cuff remain attached to the allograft.

An extensile deltopectoral incision is used, which incorporates the previous incision, if this is feasible. After deltoid mobilization laterally and pectoralis major/conjoint tendon mobilization medially, any remaining subscapularis and posterosuperior rotator cuff is mobilized and tagged for later repair. Distally, the brachialis is split and the radial nerve is identified. The prior implant is removed as is any cement, if necessary. Depending on the nature of the glenosphere, this is the point when it can be exchanged or left alone, if the glenoid base plate is stable. Given the typical soft tissue loss seen in a revision setting, as well as the higher instability risk, we recommend upsizing the glenosphere with lateral offset to alleviate these risks. The remaining humeral bone is debrided with care so that the deltoid insertion is left intact.

A key portion of the case involves measuring the desired length of the allograft. This is done most effectively with the telescope technique.

1. Place the implant trial into the remaining host bone.

2. The shoulder is reduced while holding the trial coapted to the glenosphere, as it will obviously not yet have axial or rotational stability.

3. Pull traction on the arm and the trial will "telescope" out of the native bone to the appropriate height.

4. The height of the stem at the appropriate tension is marked with a marking pen at the host bone/implant interface.

5. This length of the stem, above the mark, is the planned length of required allograft bone.

The allograft proximal humerus (Figure 3) is thawed on the back table as soon as its use is confirmed. Although fixation with a plate and unicortical screws, with or without a step-cut osteotomy (Figure 4), is an excellent option that avoids the need to identify the radial nerve for cable passing, the amount of surface area for bone integration is small and the step must be created perfectly for the proper rotation, length and alignment. We prefer the flange technique as the long flange of allograft bone provides an extensive surface for bony union.

As described earlier in the surgical planning section of this article, use a male proximal humerus for a typical female patient and use a female proximal femur for a male patient. This is important so the allograft flange will fit out and over the patient's native bone.

The radial nerve is identified and protected. This step may be already done as part of stem extraction.

The allograft is prepared on the back table with an anterolateral flange of bone intact (Figure 5). If the allograft bone does not come with tendon stumps, sutures for later capsular/rotator cuff repair are placed (Figure 6).

Two batches of cement are used to securely cement the implant into the allograft. The distance between the stem and the flange must approximate the cortical thickness of the remaining bone.

The APC is cemented in place and cables are used to secure the allograft flange before the cement hardens (Figure 7).

Fixation methods may vary. However, we recommend cable fixation as it provides enough axial and rotational stability when used with the telescoping technique. Prior to implantation of the APC, multiple cables (usually three) are passed with care to avoid iatrogenic injury to the radial nerve. The cables are provisionally tensioned. A cement restrictor is then placed in such a way as to ensure there is adequate length for the cement mantle distally. The remaining canal is cemented. The final allograft and long-stemmed implant are placed so that the anterior flange of the graft is on the anterior humerus. The implant is held in place until the cement cures. The cables are then final tensioned and clipped. At this point, the bearing surface can be trialed and final polyethylene that sets appropriate RSA tension can be placed. Lastly, the rotator cuff is reattached utilizing the shuttled sutures. Postoperative radiographs are made (Figures 8, 9).

Patients who undergo this surgery typically do not require formal physical therapy, although they wear a sling postoperatively. However, patients are taught pendulum exercises and are encouraged initially to perform these followed by supine, passive well-arm-assisted range of motion. The sling can be discontinued at the 6-week postoperative mark when active range of motion begins.

Performing RSA with APC has shown high rates of satisfactory outcomes with reliable bony union, while also preventing failure and instability. Overall complication rates in this challenging population remain high at about 20% with revision only required in 8% of patients. The dislocation rate in RSA with APC varies, but it is likely around 8%, as well. Possibly the largest concern is that of union and avoiding graft resorption. The bony healing rate is reliably more than 80% with time to union occurring at 3 to 9 months. The authors have found that utilizing the techniques described herein can improve the safety, reliability and outcomes in these difficult cases of massive proximal humeral bone loss.