When and How to Convert a UKA to a TKA
Wesley G. Lackey, MD; Jeffery L. Pierson, MD; Michael E. Berend, MD; and Kiley Hewson, PA-C
Unicompartmental knee arthroplasty (UKA) has been shown to be an excellent treatment for anteromedial osteoarthritis of the knee, with survivorship similar to that of total knee arthroplasty (TKA) in mid- and long-term studies. [1-5] However, with increasing numbers of joint arthroplasties being performed, and a growing percentage of UKA among them, the number of UKA revision is increasing proportionally.
With an understanding of the failure mechanisms of UKA, the principles of knee arthroplasty (sizing, rotation, ligament balancing, joint line), and the management of bone defects, the orthopaedic surgeon can successfully convert a failed UKA to a TKA.
When to Convert
- Find the reason for failure. When considering revision of a UKA, approach the patient in the same manner as you would a patient with a problematic TKA. A TKA is rarely revised for “unexplained pain.” Similarly, a UKA should not be revised to a TKA unless a definitive reason for failure is found. 
- Suspect infection. Any painful arthroplasty should create a high index of suspicion for infection. Workup to rule out infection after UKA should proceed in the same manner as for a TKA, with serum studies, aspiration, cell count, and cultures. Of note, the rate of periprosthetic joint infection (PJI) has been shown to be lower in UKA than in TKA,  and synovial fluid white blood cell counts indicative of infection have been shown to have a higher threshold in UKA than in TKA. 
- Address retained cement fragments. A common reason for early reoperation in UKA is retained cement fragments, which is why care should be taken during a primary UKA to meticulously remove all cement fragments when using cemented implants. If retained cement fragments cause pain or irritation, they should be addressed arthroscopically.
- “Physiological radiolucency” is not failure. Radiographs can be misinterpreted in regards to radiolucent lines and component position. The narrow, sharply defined “physiological radiolucency” after UKA has been defined in the literature and has been shown not to cause pain or failure.  This is distinct from the poorly defined, wide, and/or progressive radiolucencies associated with infection, loosening, or tibial collapse, which are the most common causes of mid to late failure in some studies. 
- Why has the UKA failed? Reasons for revision of a UKA to TKA include infection, wear, instability, component loosening, and progression of arthritis in other compartments. [10,11] Again, the workup for the cause of pain or instability should follow a similar algorithm to the workup for a problematic TKA.
- Never revise a UKA to a UKA. When failure of a UKA is diagnosed, we believe the surgeon should always convert to TKA (or antibiotic spacer in cases of infection). UKA should never be revised to UKA unless the components are stable and in good position. Example cases include exchanging an undersized bearing or addressing a dislocated bearing that was dislodged by retained posterior osteophytes.
How to Convert
Once the need for revision of UKA to TKA is established, the principles of knee arthroplasty should be employed. Factors to consider include:
- Bone loss
- Joint line
- Component sizing and rotation
Most revisions can be balanced with primary TKA implants and most series show survivorship similar to that of primary TKA. [12-21]
The complexity of the failure mechanism is usually predictive of the complexity of the revision, with medial tibial collapse with osteonecrosis requiring the most complex revisions. Increasing bone loss increases the need for screws and cement, augments, stemmed components, and occasionally constrained components (Figure 1).
Figure 1. Medial tibial osteonecrosis with collapse.
Removal of components should be performed with the same techniques as in TKA; however, it is often beneficial to leave components in place initially.
- For tibial preparation, the resection depth should be based on the un-resurfaced lateral tibial plateau. Remove the tibial component first, especially if it is loose (Figure 2). Then ignore the medial side temporarily and, using an extramedullary tibial cutting guide, cut the tibia perpendicular to the long axis at a depth based on the lateral (Figure 3).
- Next, size the tibial component and determine if a medial augment is necessary. If so, mark the midline of the tibial plateau anteriorly and posteriorly based on the flexion axis of the knee. Perform a free-hand saw cut vertically in the midline (a few millimeters), and then flatten the medial side to make room for the appropriate augment. Smaller defects can be filled with screws and extra cement, which has shown to be clinically effective and economical (Figure 4). 
- The distal femoral cut can be referenced with an intramedullary cutting guide based at the level of the intact UKA femoral component (Figure 5, 6). For anterior referencing, the AP and epicondylar axes are marked with cautery or methyline blue. The intramedullary guide is placed and a preliminary anterior cut is made with the component in place. The femur is sized prior to removal of the femoral component.
- In posterior referencing, the intramedullary distal cutting guide is placed and the lateral distal cut is made. A partial medial cut is made to mark the level of resection (Figure 6). The femoral sizing guide is placed using the anatomic rotational axis as marked (Whiteside’s line and epicondylar axis) (Figure 7).
- As mentioned above, the principles of TKA apply, including component sizing, position, and rotation, as well as ligament balancing. The final components are then cemented into place. No constraint is required (Figure 8). Post-operative radiographs are taken to assess the screws and cement (Figure 9).
Figure 2. Removing a loose tibial component.
Figure 3. Tibial cut referencing intact lateral side.