Booth and Lotke were the first to describe the use of antibiotic-impregnated cement spacers during the interim period before second-stage reimplantation of an infected TKA [9]. They had successful results and reported infection control rates of 96% [9]. Spacers are classified as static or articulating , i.e., mobile [43]. A static spacer is a block of antibiotic-impregnated cement inserted between the femur and the tibia to maintain the joint space and to act as a local drug delivery vehicle . Better clinical results have been reported with static spacers than treatment regimens not involving any spacer [10]. Potential benefits, as mentioned previously, include prevention of soft tissue contracture, enhanced soft tissue healing, and increased patient comfort between stages [12]. In addition, static spacers provide some stability to the limb and allow the use of large doses of antibiotics [19]. Proponents o f their use argue that they are more effective at delivering antibiotics than articulating spacers [11]. However, increasing the dose of antibiotics may not necessarily reduce the incidence of recurrent infection [19]. Springer et al. reported a 9% reinfection rate in their patient cohort of static spacers, using 4.0 g of vancomycin and 4.8 g of gentamicin per batch of Simplex P (Stryker) cement [44]. Disadvantages of static spacers on the other hand include spacer migration, soft tissue injury, bone erosion, quadriceps shortening, knee stiffness, and a more challenging second-stage reimplantation secondary to difficult exposure and tissue scarring [45].

Several studies have shown equivalent clinical outcomes regardless of the type of spacer used, with marginal increases in the range of motion with articulating spacers [11, 14, 15]. Emerson et al. reported similar reinfection rates with static and articulating spacers: 7.7% (2 of 26 knees) and 9.1% (2 of 22 knees), respectively, (P = 0.8) [13]. A recent meta-analysis of seven level-III comparative studies demonstrated a reinfection rate of 12% for static spacers and 7% for articulating spacers, which was not statistically significant (P = 0.2) [45]. The ultimate range of motion post-reimplantation was found to be better in patients treated with articulating spacers versus static spacers (101° and 91°, respectively, p = 0.0002). However, clinical and functional outcomes were similar among patients in both groups. There also appeared to be no difference in the complication rates when either surgical technique was used [45]. In a retrospective study comparing patients who received either a static (25 patients) or articulating spacer (15 patients), Fehring et al. [12] failed to identify a difference in Hospital for Special Surgery knee scores (83 and 84 points). However, there was a slight improvement in the range of motion in articulating spacers in contrast to static ones (105° versus 98°). Unexpected bone loss was evident during the second-stage reimplantation in 15 of the 25 patients with static spacers [12]. It is worth noting though that not all static spacers are the same and that most of the bone loss was encountered with the traditional early block-type spacers [46] (Fig. 18.1). A newer endoskeleton type of spacers may cause less appreciable bone loss, owing to a better load distribution [47]. In a recent study of four patients, Yoo et al. [47] described a novel technique using an endoskeleton-type static spacer . This consisted of an antibiotic-impregnated cement intramedullary nail, which can easily be fashioned intraoperatively using a straight thoracic tube and a Steinmann pin. They reported excellent outcomes with no bone loss and suggested that this technique could be an alternative to articulating spacers in patients with significant bone loss [47]. There appears to be a consensus that some type of stem on the tibial side or both sides of a static spacer provide better stability and prevent spacer migration to a greater degree than spacers without a stem (Fig. 18.2). Static spacers should also maximally cover the surface of the bone and maximize ligament tension without overtly impinging on the collateral ligaments, posterior neurovascular structures, or the extensor mechanism (Fig. 18.3). The use of a cast may also be prudent to prevent patient non-compliance and avoiding all even accidental attempts at flexion. This strategy is particularly important in cases of significant bone loss, such as infection following revision surgery as articulating spacers will not achieve appropriate stability and static spacers will be at higher risk of migration or dislodgement.

Given some of the previously listed disadvantages of the static spacers, articulating spacers were introduced. They allow for limited weight bearing and knee ROM between stages. In addition, they maintain the joint space, facilitate reimplantation, and decrease some of the bone loss encountered with the early block-type static spacers. Since articulating spacers are limited in their fixation, there are multiple scenarios where certain bone and soft tissue deficiencies may preclude their use (Table 18.2). Three variations exist: cement-on-cement, cement-on-polyethylene, and metal-on-polyethylene [19].