The goal in revision total knee arthroplasty (TKA) is to recreate a stable joint that is positioned and oriented close to the normal anatomic axis in all planes. This becomes a more difficult task in the revision setting secondary to bone and soft tissue loss. The abnormalities present during revision surgery may result from a combination of different factors: primary disease deformity, infection, osteolysis, aseptic loosening, implant removal complications, and concomitant systemic disease. To anatomically recreate the knee joint, the surgeon may be required to use augments of a biologic or mechanical nature to compensate for the bone or soft tissue loss.

In the normal tibia, the cancellous and cortical bone of the proximal tibia buttresses and supports the overlying articular cartilage. The stiffness of the cancellous bone decreases distally as the stiffness of the cortical shell increases.¹ The joint load is transmitted through the articular cartilage to the cancellous and cortical bone beneath. In the setting of primary knee arthroplasty, the combination of plastic, metal, and polymethylmethacrylate transmits the load directly to the underlying cancellous and cortical bone. In the revision setting, there is loss of the strongest supporting bone, and transmission of forces to the remaining subsurface could lead to early failure.²

Stems in primary TKA were introduced with early hinged prosthesis designs with the goals of resisting torsional and bending stresses at the bone-implant interface due to increased intraprosthetic constraint.³ Revision TKA components are commonly stemmed to protect the limited autogenous bone stock that is remaining. This bone may be directly under the component or under the cement, metal augments, or structural bone graft. When one is using large volumes of morselized or structural grafts, one may want to protect the graft from significant load.⁴ Conclusively, revision components without stems can place abnormal stresses on the normal bone by their constrained design nature. Joint loads are several times body weight. A stemmed component can transfer these loads if it is composed of materials that can withstand the stresses imposed on them.⁵ If the stem fails to transfer the load, then the remaining cancellous bone experiences load beyond its ultimate strength, and this will lead to a loss in fixation.⁶

A stem’s purpose, therefore, is to transmit force away from the joint line and, in so doing, lessen the stresses placed on the joint.⁷,⁸ Stems perform this function by being rigid and by being attached to a solid femoral component or tibial base plate. Brooks et al have shown that, in the varus-deficient proximal tibia, the addition of a metal-backed component decreases stress and allows for a more uniform distribution of force across the proximal tibia.⁹ Because these components are more rigid than the remaining cancellous bone, force is transmitted through them and onto the stem or onto the remaining tibial cortical rim. Bartel and associates have shown by finite element analysis that stresses on the cancellous bone beneath prostheses of conventional design can be diminished if a metal tray and a central peg are used.⁵ Lewis et al found that tibial post designs provided the lowest stresses on host bone.¹⁰ Once a stem (or post) length reaches 70 mm, the axial load at the joint line can be reduced by 23% to 39%.⁷,⁸,⁹ The bending moment carried by the stem can be variable, as fixation of the stem occurs distally.⁹ Addition of a central post and stem to the tray, however, increases the stiffness of the component and, in doing so, decreases the bending moment.⁸ The force is then returned to the bone at the metadiaphyseal or diaphyseal area, depending on the geometry, size, length, position, and composition of the stem. Bourne and Finlay demonstrated in a fresh cadaveric strain gauge study that loss of proximal cortical tibial contact resulted in a 33% to 60% decrease in strain values.¹¹ When stems up to 15 cm long are evaluated, marked stress shielding of the proximal tibial cortex and doubling of the strain located at the tip is noted. Additional potential advantages of the use of stems in revision TKA include an increased surface area for fixation, assistance with component alignment, and delivery of antibiotic cement into the intramedullary canals.

Two traditional methods of stem insertion have been used. Use of a cemented stem results in transmission of load closer to the joint line, as the stems are shorter (often 30 to 100 mm) and the force is transmitted along the bone-cement interface. The use of cement can fill metaphyseal voids between the stem and bone and help eliminate micromotion. This should decrease severe stress shielding.¹² Additional proposed advantages of the use of cemented metaphyseal stems include an increased freedom of anteroposterior and mediolateral component
placement, decreased end of stem pain, and the avoidance of anatomic abnormalities including a diaphyseal bow or deformity. Filling the intramedullary canal with cement can make future revisions more problematic, however, and may lead to further bone loss and destruction during cement removal.¹³ With diaphyseal engaging cementless stems (often greater than 150 mm in length), forces are transmitted to the tip of the stem where cortical bone contact occurs.¹³,¹⁴,¹⁵ Proposed advantages include their ease of removal (when using polished cementless stems lacking an ongrowth surface) and ability to assist with component alignment. Researchers have raised concern regarding possible proximal stress shielding with cementless stems; however, this is, more often than not, technique-dependent.¹¹ Stress shielding may actually be less if the stem is not anchored in cement.¹⁴ Cementless stem insertion may actually weaken bone due to excessive reaming or may possibly promote early loosening if the stem is undersized.¹⁵,¹⁶ Furthermore, there is the potential for increased end of stem pain.¹⁷,¹⁸,¹⁹ End of stem pain has been reported to have an incidence of up to 11% on the femoral and 14% on the tibial sides with the use of a cobalt-chrome diaphyseal engaging cementless stem.¹⁷ However, the development of titanium stems with slots or flutes has led to a decrease in end of stem pain due to a decrease in modulus of elasticity.¹⁸ Concerns of stress shielding and end of stem pain are technique-related and should not be indications to avoid the use of cementless stems. Biomechanically, several studies have indicated similar improvements in implant stability and stress distribution when using either cemented metaphyseal engaging or cementless diaphyseal engaging stems.²⁰,²¹

It is essential that the revision surgeon realize that stems are not a substitute for optimal component fit. They are simply an adjunct to relieve a portion of the excess stress seen by the components at the joint line. The type and size of stem are irrelevant if the juxtaarticular tissues are not adequately reconstructed. As stresses become greater or the soft tissues more compromised, the approach to stem fixation must be altered.¹⁶

There are a multitude of different stem geometries available. Use of larger diameter stems leads to increased load transmissions, but this is usually negated by the fact that most systems have a set diameter fixation point at the stem-component junction.²² In addition, the bending moment of the base plate is determined at this junction.²³ Longer stems in the knee result in more proximal bone shielding.⁸ This factor cannot be assessed in isolation, as shorter, wider stems impinge at their tips because of the conical shape of metaphyseal endosteum. The use of longer, thinner stems prevents tip impingement, as the stem can migrate in the sleeve of tubular diaphyseal endosteum. The contact area of the stem within the bone also determines how the load is transferred to the cortex. To complicate matters, the surface preparation of the stem may also alter fixation. Presently, most cementless stems are smooth or blasted without a porous coating. Stem composition is often titanium to decrease the modulus of elasticity and potential for end of stem pain and proximal stress shielding. Flutes have been added to the stems to aid in fixation and decrease stem stiffness. Flutes or splines on the stem engage in endosteal bone and, it is hoped, function to decrease rotational stresses at the joint line. The flutes may also act to decrease the modulus of elasticity of the stems and thus decrease the severity of proximal stress shielding (Fig. 67-1).

Preoperative planning is essential before knee revision arthroplasty. Full-length anteroposterior and lateral radiographs allow for complete assessment of the femur and tibia. Besides allowing determination of the position of the joint line, alignment of bony cuts, size and position of components, and need for augmentation, these radiographs permit assessment of the intramedullary canals to ensure that intramedullary alignment conforms to the mechanical axis orientation. It is critical to template the entry point of the femoral and tibial stem to optimize component alignment, to determine the ideal stem length
and mode of fixation, and to determine whether offset stems are required. Eccentric joint surfaces may require the use of offset stems or tibial housings to ensure proper alignment of the component while optimizing tibial plateau coverage and preventing implant overhang (Figs. 67-1 and 67-2). Canal assessment ensures that straight-stemmed components may be inserted and possibly determine the need for an osteotomy secondary to severe deformity. Presence of a severe diaphyseal deformity may influence stem selection as use of a short cemented stem may be preferred in this scenario. For cementless stems, stem length and width are estimated to obtain adequate endosteal press-fit. Stem length must be estimated with the component to account for each component’s respective housing. In general, longer stems are used to provide more rigid support, as their point of contact extends for a longer length along the endosteal diaphyseal surface. Length and degree of support cannot be assessed in isolation, as the extent of press-fit is a significant concomitant factor. Longer stems with tight diaphyseal endosteal cortical press-fit are chosen in cases of massive bony deficiencies (Fig. 67-3). Long stems may also be used to provide constrained component support when significant soft tissue imbalance or instability is present and more constrained implants are used (Fig. 67-4).