Outcomes of Oncologic Total Humeral Reconstruction

Discussion

Limb salvage surgery has replaced amputation as a means of treatment for most patients with primary and secondary malignancies of the extremities. Although the humerus is the most common area for these malignancies in the upper extremity, rarely is a total humeral resection necessary for local control. The results of the current study show that complications associated with the procedure are high, survival is poor, and function can be limited; however, this procedure offers an alternative to amputation and has the potential to preserve a functional hand.

Common complications following a total humeral replacement have previously been reported to be instability of the shoulder component, ulnar fracture, infection and radial nerve palsy (4, 9-11). The magnitude of the soft-tissue resection, in addition to the need for systemic treatment such as chemotherapy or radiation therapy, likely accounts for the high risk of dislocation and infection. In the series by Bernthal et al. (10) the authors noted a high rate of dislocation that they felt could have been potentially reduced by the use of a reverse total shoulder. In the current study, we had no cases of postoperative infection, and our only case of subluxation/dislocation was in a patient with a reverse total shoulder arthroplasty; however, in this particular shoulder, instability was likely secondary to polyethylene wear. It also should be noted that this patient had the longest survival as well, and the high rate of perioperative mortality likely accounts for our low rate of dislocation and infection.

There are various advantages and disadvantages to consider when choosing between an APC and an endoprosthesis. In the shoulder and elbow, the use of both an APC and endoprosthesis have been shown to have acceptable functional outcomes (12, 13). There is limited data on the use of a APC of the entire humerus, and the data on the use of endoprostheses focuses on implants that are no longer commercialized in the United States for off the shelf use (10), or implants which are not available for use in the United States (3, 4, 9). In the United States, currently only the Zimmer/Biomet (Warsaw, IN, USA) Segmental Revision System (SRS) is available for off-the-shelf implantation of a whole humerus metallic replacement. Although the proximal humeral portion of this system has had a reliable track record with good functional outcomes (12, 14), the distal portion of the SRS system, namely the Nexel Total Elbow, one study has reported to be associated with a high rate of early failure secondary to its design (15). Based on this, our institutional preference is to utilize a total humeral APC (when possible) as it allows for the use of different implants for the shoulder and elbow based on surgeon preference and implant track record. Currently we consider the use of an SRS only in patients with limited life expectancy (<6 months) secondary to the risk of early failure of the elbow component.

One finding in our study was the high rate of mortality observed in our patient population following these procedures, with three patients dying of disease within two months of the procedure. This is likely due to our use of these procedures in patients with metastatic disease compared to other large series (3, 9, 10). In addition to the poor survival, our functional outcome scores are lower than those previously reported. We are unable to directly compare differences between studies, however this could be due to the high rate of elbow contractures possibly due to the effects of adjuvant treatments. Although our complication rate was high, only one patient had a revision surgery. That being said, the rate of implant “failure” was higher than other studies and likely is secondary to the use of a competing risk analysis, versus other series which only use a Kaplan–Meier method, which does not account for death due to disease prior to implant failure (3, 10).

Study limitations. This is a small, retrospective study with limited available data, which constrained the analysis. Nevertheless, this is a rare procedure, and the prospective nature of our Total Joint Registry Database helps to limit recall bias. This study spans a long period of time over which surgical, medical imaging and reconstruction techniques have evolved. Multiple surgeons were involved in the tumor resection and reconstruction and due to the heterogenous nature of the tumors, there was no standard approach to each case.

Overall, the results of the current study show that total humeral resection and reconstruction is a demanding procedure with a high rate of complications, and relatively poor functional outcome scores. This procedure does provide an option for extremity reconstruction and a functional hand for a patient population with very limited alternative options. Our current preference is to perform an APC whenever possible, as it allows off-the-shelf customization of the implants and overcomes the current lack of endoprosthetic options.