Background

Reconstruction of large bony defects in the femur and tibia during revision knee replacement remains a challenging clinical problem. Smaller bone defects have been traditionally and effectively treated with limited quantities of morselized cancellous bone graft,^3,²⁷ cement augmented with screw fixation,^1,^11,^28,²⁹ or modular augments attached to revision prosthetic implants.^4,¹⁴ Large or massive bone defects require more extensive reconstructive efforts and have been managed traditionally with the use of large structural allografts,* impaction bone grafting techniques with or without mesh augmentation,^17,^32,^33,³⁶ fabrication of custom prosthetic components,⁸ or specialized hinged knee components.¹⁵ Despite the multitude of utilized treatment methods, the best reconstructive technique for bone defects during revision knee replacement has not been clearly determined.⁸

Preoperative Planning

In revision knee arthroplasties with bone deficiency, it is helpful to determine several important variables to establish the requirements of reconstruction. Good quality anteroposterior, lateral, and patellar views are usually sufficient for assessment of bone loss in the large majority of cases; however, tomograms are occasionally helpful. It is often stated that the magnitude of bone loss observed on preoperative radiographs vastly underestimates the true magnitude of bone loss discovered intraoperatively. This tenet is particularly true of bone loss associated with osteolysis secondary to wear debris. If a lateral radiograph of the knee obtained before the original arthroplasty was performed is unavailable, a lateral radiograph of the opposite knee is valuable to determine the appropriate anteroposterior dimension of the knee. A true lateral radiograph, obtained with fluoroscopically positioning, may occasionally be useful for accurate assessment of the anteroposterior dimensions of the femur.

Bone loss in the four primary areas of typical occurrence assessed. The severity and location of bone loss and the quality of remaining bone should be assessed. The location of the joint line is marked and noted. The optimal joint line position is roughly 2 cm below the origin of the medial collateral ligament and 2.5 cm below the prominence of the lateral epicondyle. In some knees, particularly those with flexion instability, the prosthetic joint line may actually be lower than the anatomic joint line, which usually suggests that additional distal femoral resection will be required during the revision procedure to balance flexion and extension spaces.

Bone Loss Assessment and Classification

The critical step in determining the appropriate reconstruction method in revision total knee replacement (TKR) is to accurately determine the quantity, location, and extent of bone loss. This is done after meticulous removal of failed tibial and femoral implants, with careful attention to existing bone preservation. Once the components have been removed, it is important to determine whether the defects are contained or uncontained (segmental). In addition, the location of supportive bone that surrounds the bone loss is essential and will dictate the type and size of augmentation that is required. Smaller contained defects can be treated with cement fill with screw augmentation or with morselized allograft fill, particularly in older patients. However, larger uncontained defects typically require larger reconstructive measures such as modular block augments, bulk allograft, or highly porous metal metaphyseal cones.

A common system of categorizing bone defects in revision knee arthroplasty is the Anderson Orthopaedic Research Institute Bone Defect Classification.⁸ This system permits communication and comparison of knees between different institutions and also allows preoperative and postoperative classification and management recommendations for specific bone defect severities. In this bone defect classification, type 1 defects describe only minor and contained cancellous bony defects within either tibial plateau, type 2A defects include moderate to severe cancellous and/or cortical bone defects of only one tibial plateau, type 2B defects consist of moderate to severe cancellous bone defects of both tibial plateaus and/or segmental cortical defects of one tibial plateau, and type 3 defects describe combined cavitary and segmental bone loss in both tibial plateaus.

Reconstruction With Cement and Screws

Cement used as a reconstructive augment has the attraction of being simple, inexpensive, and efficient, as the revision knee arthroplasty is already utilizing the material for fixation in most instances. This reconstruction method is typically indicated for smaller contained defects measuring less than 5 mm in depth,^1,¹¹ although some authors have advocated its use in larger defects with excellent clinical results.^28,²⁹ When cement is used for defects in revision knee arthroplasty, augmentation with bone screws is typically recommended to enhance the biomechanical properties of the construct (Fig. 130-1). In addition, if the patient is young and active, it may be more advantageous to utilize morselized allograft to restore bone stock in these types of defects.

Figure 130-1 A, Anteroposterior (AP) and (B) lateral radiographs of contained medial tibial bone defect secondary to osteolysis. C, AP and (D) lateral radiographs of the reconstruction utilizing cement fill with screw augmentation at 12 months’ follow-up. Note the screw head well below and without contact with the tibial tray.

Surgical Technique

The surgical technique begins with tibial or femoral provisional or freshening cuts. Once these are performed, a more accurate assessment of the defect is possible. Meticulous débridement of the defect is performed with removal of all fibrous tissue that would impede adequate interdigitation of the cement and create suboptimal fixation. Sclerotic bone surfaces are frequently encountered in revision surgeries; these must be roughened with a small drill or a burr. Once the defect is clearly delineated and prepared, the location of remaining bone is identified for adequate screw fixation. Sloping surfaces are converted to step-shaped to minimize the quantity of shear forces acting on the cement, as cement is known to be much more biomechanically stable and supportive in compression. If the defect is of minimal depth, it may be filled with cement alone during cementation of the standard revision tibial or femoral components. If defects are larger, or if the surgeon is uncertain, reinforcement with screw augmentation is recommended. Once the bone defect has been adequately prepared for cement, titanium, self-tapping cancellous bone screws are placed into the host metaphyseal bone and are advanced so that the heads are positioned below the level of the eventual tibial tray or femoral component. Titanium screws are typically used to prevent galvanic corrosion that occurs with dissimilar metals, because a majority of tibial baseplates are composed of titanium. Once the screw is in position, the trial components are inserted to ensure that there is no contact of the screw head with the prosthesis. Once the cement has been mixed and is in a doughy state, it is placed into the defect and around the screw heads and is pressurized by hand. The final prosthesis is then placed and cement is allowed to cure with removal of the excess.

Clinical Results

Satisfactory midterm results have been reported with the use of screws and cement for bone defects in TKA. Ritter and associates ²⁸ reported on 57 TKRs with large (9 ± 5 mm) medial tibial defects reconstructed with screws and cement at an average of 6.1 years’ follow-up.²⁸ Although nonprogressive radiolucent lines were common and were seen in 27%, no cases of tibial component loosening, component failure, or cement failure were reported. In a subsequent report by the same authors, 125 TKAs that utilized screws and cement to fill large medial tibial defects secondary to severe varus deformities were reported at a mean of 7.9 years’ follow-up.²⁹ The authors reported two failures that occurred as the result of medial tibial collapse at 5 and 10 years, respectively, but no other failure or loosening was observed in the remainder of the cohort. However, this was a series of primary knee arthroplasties without the typical stem extensions used in revision knee arthroplasty to augment fixation and prevent medial collapse in the setting of bone deficiency and suboptimal bone quality. Therefore, in smaller and contained defects such as those encountered in revision knee arthroplasty, particularly in older or less active patients, the use of screws and cement is a viable and successful method of reconstruction that is inexpensive, relatively simple, and efficient.

Reconstruction With Morselized Allograft

Bone loss in revision knee arthroplasty can be treated reliably and successfully with morselized cancellous allograft and has an established clinical track record.* This method typically is reserved for contained defects (Fig. 130-2) and is particularly attractive for younger patients, in whom restoration of deficient bone stock is a priority given the potential for future reconstructive surgeries. Biologically, morselized cancellous allograft appears to incorporate similarly to cancellous autograft, albeit at a much slower rate. It is also beneficial to have a well-vascularized recipient bed to facilitate incorporation of the allograft bone; if a highly sclerotic defect is encountered, it may be beneficial to burr away the sclerotic bone to underlying cancellous and vascular bone, or conversely to use another reconstruction method such as a block augment. Furthermore, if the defect is large and segmental, although some authors have reported adequate results with impaction allografting,^17,¹⁸ reconstruction with more robust structural augments such as metal blocks, bulk allograft, or metaphyseal porous metal cones typically will produce more biomechanically stable constructs.

Figure 130-2 A, Anteroposterior (AP) and (B) lateral radiographs of failed total knee replacement (TKR) with severe femoral bone loss in the lateral condyle secondary to osteolysis. C, Intraoperative picture of impaction grafting using the intramedullary reamer to facilitate compaction of morselized allograft into the defect. D, Removal of the reamer after morselized allograft compaction demonstrating complete fill and reconstitution of the contained defect. E, AP and (F) lateral radiographs of the revision TKR in parts A and B demonstrating the reconstituted lateral femoral defect bypassed with an intramedullary stem.

Surgical Technique

As with all reconstructive techniques, the surgical technique of utilizing morselized allograft to fill contained defects requires meticulous débridement of the defect with careful attention to removal of all fibrous tissue. Careful attention is paid to preparation of a vascular bed for allograft incorporation to host bone and long-term bone reconstitution. Once the defect is adequately débrided, prepared, and confirmed to be contained with supporting peripheral structures, the morselized allograft can be inserted into the defect. It is also preferential to grind up any larger pieces into a fine morselized consistency to facilitate the development of biologic and structural properties. It is helpful to place an adequately sized reamer or trial stem into the medullary canal to impact the morselized allograft material around the reamer; this facilitates compaction of the graft to optimize its ability to provide structural support (see Fig. 130-2C and D). Once this is complete, the reamer is removed, and the final implant is inserted with an intramedullary stem for supplemental support.

Clinical Results

Midterm results are available for the technique of impaction allograft reconstruction in revision total knee arthroplasty. Lotke and associates ¹⁸ prospectively studied the midterm results of 48 consecutive revision TKAs with substantial bone loss treated with impaction allograft. At an average follow-up of 3.8 years, no mechanical failures of the revisions were reported, and all radiographs demonstrated incorporation and remodeling of the bone graft. Six complications were reported among the 42 revisions available for follow-up (14%): 2 periprosthetic fractures, 1 early infection salvaged with irrigation and antibiotics, 1 late infection resulting in fusion, and 2 patellar clunk syndromes. Although the authors concede that the technique is time consuming and technically demanding, they advocate impaction grafting for bone loss in revision TKA.¹⁸ Whiteside and colleagues³⁷ reported on 63 patients who underwent revision knee arthroplasty using morselized cancellous allograft to fill large femoral and/or tibial defects. Firm seating of the components on a rim of viable bone and rigid fixation with a medullary stem were achieved in all cases. Fourteen reoperations occurred, and a biopsy specimen taken from the central portion of the allograft revealed evidence of active new bone formation. Evidence of healing, bone maturation, and formation of trabeculae was observed on all radiographs at 1-year follow-up. Two patients in this series required revision surgery for aseptic loosening, and the authors believed that both had greatly improved bone stock, so new implants could be applied with minor additional grafting.³⁷