Total knee arthroplasty (TKA) is a successful and reproducible surgery that improves pain, function, and limb alignment in knees that have developed arthritis and deformity. Most limb deformity in osteoarthritic knees arises intra-articularly, from asymmetric wear of the joint surfaces over time. Malalignment and associated ligament imbalance are most commonly corrected in TKA with appropriate bone resection and soft-tissue balancing. However, some patients with knee arthritis have an accompanying extra-articular deformity (EAD) of the femur and/or tibia. These deformities can be due to congenital conditions, trauma, metabolic disorders, or from previous surgery.

EAD poses technical challenges to TKA that must often be addressed with supplemental corrections of limb alignment, rotation, and soft-tissue balance at the time of knee replacement. These corrections can be made at the time of TKA or can be done as a separate procedure prior to TKA. Correction of EAD can be done either with modification of intra-articular bone cuts at the time of TKA or with extra-articular osteotomies, commonly at the site of deformity. Numerous techniques have been described for addressing EAD, and the recent popularization of surgical navigation has further expanded the toolkit available to surgeons. However, each EAD is fundamentally unique and often requires custom solutions tailored to the specific deformity. This chapter aims to characterize techniques for overcoming EAD in TKA by emphasizing principles that can be tailored by the surgeon to each patient’s specific deformity.

Developmental femoral shaft bowing in the coronal plane is the most common EAD encountered in candidates for TKA, especially in Asian and Indian populations.¹,² Posttraumatic femoral deformities often arise after femoral shaft fractures that heal in a malunited position. Many femoral fractures, especially those that do not undergo open reduction and internal fixation, heal with varying degrees of shortening, translation, rotation, and malalignment any point along the length of the femur, with deformities of the distal femur close to the site of the TKA being particularly difficult to manage. Previous corrective osteotomies can change the bony anatomy of the femur. Like fractures, these prior osteotomies are commonly close to the knee joint and involve hardware, making them more complex to manage. Standing 36-inch radiographs of the entire limb aid in evaluating translation and angulation but provide little information on rotational deformities. CT scans and intraoperative visualization remain the best tools for assessing malrotation.

Tibial deformities affecting TKA can also be present at any point along the length of the leg but are most commonly found at the tibial metaphysis. These are often a result of previous tibial plateau fractures or high-tibial osteotomies (HTOs). When assessing previous HTO, it is vital to recall and understand that proximal tibial slope and translation are frequently changed as a result of the osteotomy. For this reason, intraoperative visual referencing using the tibial plateau may be inaccurate and lead to incorrect final alignment or issues with the tibial stem component contacting the metaphyseal endosteal cortex. Offset tibial stems or well-designed implants with multiple smaller pegs may be useful in this situation.

As in the femur, hardware is often present in the proximal tibia and must be addressed in sequential or simultaneous fashion. The tibial tubercle and extensor mechanism are of vital importance to the success of a TKA, and their competency must be preserved. Previous injury, surgery, or fixation may have changed the orientation or integrity of these structures. Any evidence of extensor lag on examination or patella baja/alta on preoperative imaging should be identified and incorporated into the surgical plan. Additionally, hardware removal around the tibial tubercle at the time of surgery may create bone voids that act as stress risers, increasing the fracture risk of this structure.

Retained hardware is often a complicating factor in patients with EAD. Plates, screws, staples, and nails can be present both in the periparticular region or along the femoral or tibial shafts. There is no single correct way to deal with retained hardware, as each case poses unique issues and challenges. In general, if the hardware does not impede intraoperative instrumentation or implant insertion, it can be left in place and ignored (as long as it does not compromise potential future surgical options). If hardware interferes with TKA implantation, it must be removed.³ Acceptable options include removing hardware selectively or entirely. As a last resort, difficult-to-remove hardware can be burred or cut with appropriate metal-cutting tools.⁴ The surgeon can choose to perform hardware removal in staged fashion prior to TKA or simultaneously at the time of TKA. If there is a large burden of hardware that will leave significant stress risers or if there is concern for previous infection, staged removal and culture as needed is preferred. Otherwise, single-stage TKA is performed, and hardware obstructing insertion of the prosthesis is removed as it is encountered. In single-stage TKA, bone voids from removed hardware must be anticipated and addressed with bone graft and/or supplemental fixation such as stems, augments, and cones/sleeves.⁵,⁶

It is important to remember that even hardware distant from the joint can interfere with TKA if it impedes access to the intramedullary canal. Full-length femoral nails and distal interlocking screws can interfere with stemmed revision femoral components and can occasionally even interfere with primary implants. Femoral arthroplasty components from different manufactures each possess their own specific geometry. In general, cruciate-retaining and sacrificing, but not substituting, femoral components do not have a femoral box and rarely interfere with previous femoral nails. Conversely, cruciate-substituting and varus-valgus-constrained implants can have larger box geometries and may interfere with a previous femoral nail, even without placing a stem. Numerous references exist to help determine the exact geometry and dimensions of most implants on the market.⁷ Additionally, even short femoral nails or long-stemmed hip replacement components can affect intramedullary referencing even though their location can be quite proximal to the knee joint. If an intramedullary component at the hip is present at or distal to the femoral isthmus, using an intramedullary alignment jig may be inaccurate or impossible. In these cases, we do not recommend short intramedullary alignment rods as they do not provide accurate alignment. Extramedullary alignment jigs with intraoperative radiographic assistance, custom cutting jigs designed from preoperative three-dimensional imaging, or surgical navigation should be employed.

Lastly, an evaluation of retained hardware prior to TKA should always include attention to previous incisions and any soft-tissue defects. Previous incisions should be used when possible. Incisions can be lengthened or connected at nonacute angles as long as the flaps are sufficiently large. In general, choosing the most lateral of the previous incisions is least likely to compromise skin viability and healing. If soft-tissue redundancy allows, scars can be excised to promote more cosmetic rehealing, but tension-free approximation should not be compromised to accomplish this. Previous gastrocnemius rotational flaps or other soft-tissue abnormalities/defects should be identified, especially at the proximal-medial tibia where coverage is often the most tenuous. Plastic surgery consultation preoperatively may be helpful in determining the best plan for the surgical approach and exposure.

Preoperative planning begins with a thorough history and physical examination. The surgeon should understand the cause of the EAD from the history, particularly if prior trauma was involved. The physical examination should focus on clinical alignment, skin quality, prior incisions, muscle strength, and ligamentous assessment. A long-lasting deformity can result in excessive ligament stress and stretching that may not be evident on imaging. Deformity can also result in secondary patellar maltracking.

Preoperative imaging is critical in cases of deformity. A full-length hip-to-ankle standing radiograph is of vital importance in planning for a case with EAD (Fig. 46-1). The radiograph should be evaluated both for EAD, as well as intra-articular deformities such as asymmetric bone loss at the joint line, commonly seen in advanced osteoarthritis. Both types of deformities must be accounted for in planning for correction of alignment. Lateral X-rays should be viewed for deformities in the sagittal plane, and patellar integrity should be evaluated with a Merchant view of the knee. Supplemental advanced imaging such as a CT scan can be helpful for evaluating bone loss in three dimensions and is also sometimes required for various types of surgical navigation or patient-specific instrumentation.⁸

With previous surgeries and retained hardware, an index of suspicion for infection is warranted, and screening laboratory tests with possible aspiration prior to definitive surgery should be considered.

The spectrum of EAD and retained hardware often requires careful planning for specific instruments required during that case. When considering retained hardware, consider having broken hardware-removal trays, implant-specific (or universal) extraction devises, metal cutting burrs, and intraoperative fluoroscopy available. When considering EAD, consider having revision-constrained instrumentation with augments/cones/stems, extra-articular navigation, computer navigation, patient-specific instrumentation, saws, angle-measuring guides, plates, screws, nails, and custom implants available.⁹,¹⁰,¹¹