Repeated Failure of Fixation of an Allograft Humerus Reconstruction
In the 12 years since surgery and chemotherapy for a parosteal osteosarcoma, a young female patient has undergone multiple procedures for non-union at the allograft-host bone junctions in her left humerus. Will a revision fixation procedure finally allow for union of the bone?
Brandon J. Yuan, MD
In 2005, an 18-year-old, right-hand-dominant female patient underwent wide local resection and allograft reconstruction for a parosteal osteosarcoma of the left humeral diaphysis. Her initial postoperative radiographs are shown in Figure 1.
Figure 1. Radiographs following surgery for a parosteal osteosarcoma.
The patient completed a course of adjuvant chemotherapy and has had no evidence of recurrence or metastases since that time.
However, she has had multiple subsequent procedures on the left arm for non-union occurring at the allograft-host bone junctions. In 2006, she underwent revision internal fixation of the distal junction site (Figure 2).
Figure 2. Radiographs show revision internal fixation of the distal junction site.
Two years later, she underwent revision internal fixation of the proximal and distal junction sites due to failure of the proximal humeral locking plate at the proximal junction site (Figure3). Both sites were grafted with autograft bone. The patient did well for a number of years.
Figure 3. Radiographs show revision internal fixation of the proximal and distal junction sites due to failure of the proximal humeral locking plate at the proximal junction site.
Repeat Failure of Fixation at the Distal Junction Site
Now age 30, the patient presents with increasing deformity and pain in the left arm. Radiographs (Figure 4) show repeat failure of the distal junction between the allograft and the host bone. The long proximal humeral locking plate and the second 3.5-mm low-contact dynamic compression plate have failed. A preoperative CT scan demonstrates union of the proximal junction site. Her left hand functions normally and she is neurologically intact in the left upper extremity.
Figure 4. Radiographs show repeat failure of the distal junction between the allograft and the host bone.
Prior to any revision procedure, the surgeon must rule out underlying infection or endocrine abnormalities that may contribute to delayed bone healing. Preoperatively, this patient’s C-reactive protein (CRP) level, erythrocyte sedimentation rate (ESR), white blood cell (WBC) count, and endocrine labs (vitamin D, thyroid-stimulating hormone, parathyroid hormone, alkaline phosphatase, calcium, and phosphorus levels) were all within normal limits.
Options at this point include:
- Revision internal fixation with cancellous autograft
- Revision internal fixation with vascularized autograft
- Resection of the defect with either shortening of the humerus or maintenance of length with segmental bone reconstruction through induced membrane grafting or bone transport
In this particular case, the assumption can likely be made that the reason for continued non-union is the lack of biologic activity at the junction between the patient’s host bone and the allograft. In the orthopedic oncology literature, large bulk allografts have been used to successfully reconstruct diaphyseal defects, although the durability of the allograft remains a concern.
The primary solution for this particular problem requires the addition of biologic graft that will allow for improved odds of incorporation with host bone. However, the gap between the proximal host bone and the distal junctional non-union is large. Revision internal fixation of the junction site alone without addition of structural, biologically active graft is likely to fail, as it has the previous 3 times.
Vascularized structural autograft from the fibula or rib can be utilized to span large defects and can be augmented with internal fixation. In addition, the graft can be large enough to span the entire initial defect and allograft reconstruction in this case, ensuring incorporation of the vascularized graft into host bone at either end of the humerus.
Revision Fixation Procedure
After explaining the options to the patient, the decision was made for revision fixation with a medullary nail, followed by augmentation with an onlay vascularized fibular auotgraft.
- The patient’s prior anterolateral incision was utilized for the initial incision.
- The prior implants were removed. Intramedullary fixation required the removal of all prior broken screws in the distal allograft and in the distal native humerus. This left large holes in the distal humeral diaphysis (Figure 5).
- The distal humeral segment was rotated as needed to allow the distal interlocking bolts to pass through native bone in the distal humerus, where it was available.
- Next, a vascularized autograft fibula was placed anterolaterally and secured with 2 lag screws to the native humerus above and below the allograft (Figure 6).
- The prior holes in the humerus and the junction sites were grafted with iliac crest autograft (Figure 7).
Figure 5. Removal of all broken screws in the distal allograft and distal native humerus left large holes in the distal humeral diaphysis.
Figure 6. A vascularized fibula was placed anterolaterally and secured with 2 lag screws to the native humerus above and below the allograft.
Figure 7. An iliac crest autograft was used to fill holes in the humerus and at the junction site.
Postoperative radiographs (Figure 8) demonstrate anatomic alignment with maintenance of humeral length. The entire allograft segment is spanned by both the medullary nail and the autograft fibula.
Figure 8. Postoperative radiographs show anatomic alignment with maintenance of humeral length.
At 1 year after surgery, follow-up radiographs (Figure 9) demonstrate incorporation of the fibular autograft into the proximal and distal humerus, spanning the prior fracture site and allograft. Although the fracture at the previous junction site between the allograft and native humerus has not healed, the humerus is now effectively stabilized.
Figure 9. Radiographs show that 1 year after surgery, the fibular autograft has been incorporated into the proximal and distal humerus, spanning the prior fracture site and allograft.
- When considering the approach for non-union surgery, the reason for the lack of healing must first be ascertained. This allograft junction has been rigidly stabilized multiple times, always resulting in catastrophic implant failure. Clearly, the goal in this case is the addition of biologic supplementation to the non-union.
- Biologically inactive non-unions due to trauma, infection, soft tissue injury, radiation therapy, or allograft reconstruction (as in this case) may require advanced bone grafting techniques beyond simple cancellous autograft in select cases.
- In the setting of chronic osteomyelitis with a biologically inactive non-union, an approach that involves resection of the infection and non-union, followed by induced membrane bone grafting or bone transport, may be more appropriate. However in the absence of infection, the use of structural vascularized autograft allows for immediate stability and avoids the need for prolonged bone transport or large segment cancellous autografting.
Brandon J. Yuan, MD, is an Assistant Professor in the Division of Orthopedic Trauma, Mayo Clinic, Rochester, Minnesota.
Disclosures: Dr. Yuan has no disclosures relevant to this article.
- Aponte-Tinao LA et al. Should fractures in massive intercalary bone allografts of the lower limb be treated with ORIF or with a new allograft? Clin Orthop Relat Res. 2015 Mar;473(3):805-11
- Houdek MT et al. Complications and long-term outcomes of free fibula reconstruction following resection of a malignant tumor in the extremities. Plast Reconstr Surg. 2017 Feb;139(2):510e-519e