Another controversial aspect in revision total knee surgery deals with the type of fixation and ability to maintain normal axial and sagittal limb alignment with canal-filling stems. The revision knee surgeon must be aware of the potential malalignment issues that can occur with canal-filling cementless stems. On the tibial side, valgus bowing of the tibial diaphysis is not uncommon. Thus, when a canal-filling diaphyseal-engaging stem is used, axial malalignment can ensue (Fig. 13.2). In addition to the potential for axial malalignment, anteromedial overhang of the tibial tray may occur with the potential for postoperative anteromedial knee pain.

Hicks et al. [8] noted significant variability in the location of the tibial canal to the tibial plateaus. In their cadaveric review, they found that the intramedullary canal center was usually anterior and medial to the tibial plateau. This study highlighted the need for offset stems in revision total knee arthroplasty, especially if engaging the diaphysis of the tibia (Fig. 13.3).

Canal-filling stems can also have an effect on alignment on the femoral side. A canal-filling femoral stem can lead to anterior displacement of the femoral component. Such displacement overstuffs the patellofemoral space with the potential for limiting motion and increasing incidence of anterior knee pain. In addition, such anterior displacement by definition increases the flexion gap, which can lead to flexion instability (Fig. 13.4).

Strategies to prevent such translation include offset stems or stem bolts, which can move the stem anteriorly or posteriorly as necessary to prevent sagittal malalignment. Alternatively, a narrow cemented stem can be placed posteriorly in the canal limiting this effect. (Fig. 13.5).

A final alignment issue that can affect implant position occurs when the shaft of the femur is slightly lateral to the condylar bone. If one uses a canal-filling stem in this situation, lateral shift of the implant occurs. This helps patellar tracking . However, the eccentric lateral box position can compromise distal femoral bone stock.

Many of the described axial and sagittal malalignment issues can be handled in one of two ways. A narrow cemented stem can be used in most situations to prevent the previously mentioned malalignment issues. The stem is simply placed eccentrically in the canal to prevent malalignment. Care must be taken to ensure an adequate cement mantle. Alternatively, most manufacturers now offer adjustable offset stems that can compensate for limb malalignment and implant malposition that can occur when using straight canal-filling stems.

An emphasis on preoperative planning to recognize potential malalignment issues will help the revision surgeon have the necessary equipment available at the time of revision.

From the previous discussion, one can surmise that cemented fixation has certain advantages over its cementless stem counterpart. In the laboratory less micromotion has been reported. In addition, the limb alignment and implant position problems noted previously rarely occur with cemented constructs. To determine what type of stem fixation is best for the revision knee patient, it is also important to review the literature to date on this subject.

Although no prospective study comparing cemented versus cementless stems in revision total knee arthroplasty is available, proponents of each method have reported their results [9–12]. Murray et al. [9] reported the clinical and radiographic results of 40 patients who underwent cemented long-stem revision total knee arthroplasty at an average follow-up of nearly 5 years. Only one patient had asymptomatic radiographic loosening of the femoral component, while no tibial component was categorized as loose. Mabry et al. [13] reported excellent long-term results with cemented modular stems with 5- and 10-year implant survivorship free of revision for aseptic failure of 98% and 92%, respectively.

Bertin et al. [12] first described the use of juxta-articular cementing with the use of long uncemented stems in their analysis of 53 revision total knees. At a follow-up of only 18 months, 18% had complete radiolucent lines at the femoral bone-cement interface, while 21% of the tibial implants had a complete radiolucent line at the tibial bone-cement interface. The widths of the radiolucent zones were not thought to be progressive by the authors. Thin white lines were frequently seen around the cementless stems in Bertin’s study. Of the 73 stems with radiopaque lines next to the stems, 18 were tightly approximated to the stem, 40 were parallel within a few millimeters of the implant, while six had divergent sclerotic lines . The authors believed that these lines did not imply loosening but needed to be followed longer to establish their significance as the follow-up in this study was extremely short.

Haas et al. [10] reviewed 65 patients who underwent revision total knee surgery for aseptic loosening. Each patient had cement used on the cut surfaces in the metaphyseal region of the femur and tibia along with a cementless fluted stem. While the Knee Society Radiographic Scoring System was used to determine location of radiolucent lines, implants were not categorized according to Knee Society guidelines as stable, possibly loose requiring close follow-up, or loose. Radiolucent lines at the bone-cement interface were noted in 33% of the femoral implants and 64% of the tibial implants. Most were 1–2 mm and nonprogressive. Complete radiolucent lines at the bone-cement interface were noted in 7% of the tibial components and 1% of the femoral components.

Peters et al. [11] in 1997 reviewed 57 revision total knee arthroplasties performed for aseptic failures. Eighteen tibial stems and 34 femoral stems were used in this group of revisions. Thirty-two of these stems were cemented, and 20 were cementless. Adequate radiographs were available for only 39 of these 52 implants (75%). In Peters’ series, radiolucent lines were more prevalent adjacent to press-fit femoral stems compared with cemented constructs (p < 0.02). There was no significant difference in the total number of radiolucent lines around cemented and cementless tibial stems in their series (p = 0.73).