Gap balancing typically occurs with extension balancing first to achieve stability [61]. As such, the tibial resection is paramount to balancing the knee as a varus resection causes excessive femoral component internal rotation, while the opposite is true for a valgus osteotomy [61]. Measured resection allows the osteotomies of the tibia and femur to occur prior to significant soft-tissue balancing. Instead of using tension, anatomic landmarks (e.g., surgical TEA, Whiteside’s line, posterior condylar axis) are used to judge femoral rotation. However, these landmarks may occasionally be difficult to assess [62–64], and rotational alignment is variable, especially when referencing from the surgical TEA [37, 65, 66]. Use of multiple landmarks may help reduce rotational malalignment [67]. Some claim gap balancing produces better rectangular balance in flexion and extension without sacrificing femoral rotational accuracy [61, 68]. Others claim no difference [59, 69]. Compared to the measured resection technique, gap balancing has less condylar lift off throughout ROM [67], restores the joint line more accurately [70], and produces less flexion contractures [71]. However, no differences in KSS [23, 24, 72] or functional quality of life [73] have been reported at 2 years. In addition, some have reported that gap balancing with tensor systems achieves adequate rectangular gaps in flexion and extension more frequently than measured resection [58]. Longer-term outcomes and survivorship have not been reported.

Midterm outcomes are similar between conventional and MIS TKAs. However, MIS has supported faster gains in the early recovery [79–83]. Skin incision length is approximately 4 cm shorter than conventional exposures [79, 80]. Radiographic malalignment in the coronal plane has been reported to be similar between the two techniques [79, 81, 83], but some conflicting reports exist [82]. In a randomized clinical trial (RCT) comparing three groups, computer-assisted surgery (CAS) quadriceps sparing (QS) MIS TKA, QS MIS TKA without navigation, and conventional TKA, Lin et al. [84] showed coronal alignment was similar among the groups. Others have supported CAS to improve alignment for MIS TKA [85]. In a meta-analysis, ROM was greater in the early postoperative period (6 days–3 months), but final ROM (>1 year postoperatively) has not been reported [79, 80, 82, 83]. Functional scores are similar between the two groups, while short-term KSS was higher for MIS groups [80, 83], and KSS was similar at 1 year [80, 83]. In a non-randomized study comparing 40 CAS MIS TKAs to 40 CAS TKAs, KSS were significantly better for MIS TKA in the short term, but KSS were similar at 5 years. Bonutti et al. [86] reported on their original 103 MIS TKA cohort showing survivorship of 97.1 % at a mean follow-up of 9 years. Neither long-term survivorship nor outcomes have been compared in a randomized controlled fashion.

Cruciate-retaining (CR) prosthetic designs are thought to maintain ROM by restoring normal knee kinematics and anatomic femoral rollback, while posterior-stabilized designs (PS) are thought to control ROM by allowing for an enhanced femoral-tibial articulation [87]. Short-term benefits of each design remain debated. However, many studies have shown increased survivorship with CR designs [88–90]. In a recent meta-analysis restricted to RCTs, Li et al. [91] showed higher ROM for PS TKAs, while KSS and survivorship were similar at 2 and 5 years postoperatively. However, Abdel et al. [88] showed superior survivorship for 5,389 CR TKAs at 15 years. Subanalysis for knees with a preoperative deformity (i.e., flexion and angular deformities >15°) revealed that 15-year survivorship was 90 % for CR TKAs but only 75 % for PS TKAs (p < 0.001). For knees without deformity, survivorship was 88 % for CR TKAs and 78 % for PS TKAs (p < 0.04). Of the 16,584 primary TKAs completed at the same institution between 1985 and 2005, CR designs had significantly better survivorship at 10 and 15 years when stratified by implant manufacturer and design [89].

Mobile-bearing (MB) TKAs were designed with the intention of reducing wear and thus revision procedures, as well as providing more natural kinematics [92]. In the largest meta-analysis to date, Moskal et al. [93] showed that the rates of aseptic loosening, periprosthetic joint infection, and revision for any cause were similar between MB and FB TKAs. Additionally, ROM, KSS, Hospital for Special Surgery (HSS) scores, and SF-12 scores were similar between the groups. Interestingly, a large series from the International Consortium of Orthopaedic Registries (ICOR) containing 258,190 FB CR TKAs and 61,426 MB CR TKAs showed worse survivorship for MB CR TKAs at yearly time points from years one to ten postoperatively [94]. Additionally, Li et al. [95] showed no difference in patient preference by meta-analysis.

All-polyethylene tibial components provide the benefit of not creating any backside wear and thus have a potential for increased survivorship. In a large meta-analysis and systematic review, survivorship was similar between all-polyethylene and metal-backed tibial components at 2 [96], 10 [96], and 15 years [89, 96]. On the other hand, the Swedish registry data of over 27,000 TKAs supported a higher survival rate and lower revision rate at 10 years for all-polyethylene tibial components [97]. Kremers et al. [89] showed that CR all-polyethylene tibial components had a survivorship superior to CR metal-backed tibial components. However, metal-backed tibial components were superior to the PS design. Of note, an early PS all-polyethylene tibial design (DePuy PFC; Warsaw, IN) showed high failure rates, which inherently skewed the results [89]. Additionally, all-polyethylene tibial components represent a significant cost savings; for community practices, negotiated rates were $3,035 and $2,078, respectively [98]. While all-polyethylene tibial design offers lower costs, greater polyethylene thickness for comparable bone loss, and less backside polyethylene wear, modularity allows intraoperative flexibility and the option of polyethylene exchange in the revision setting [99, 100].

Cement fixation remains the gold standard for TKA but may not improve survivorship. Early cementless designs certainly showed inferior survivorship [101]. However, innovations such as hydroxyapatite-coated technology were not used in early comparisons. More recent meta-analyses support similar rates of revision TKAs between cemented and uncemented components [102, 103]. In a meta-analysis of 3,568 TKAs, Mont et al. [103] showed similar survivorship of 95.6 % and 95.3 % at 10 years and 76 and 71 % at 20 years for cementless and cemented TKAs, respectively. Additionally, screw reinforcement did not extend survivorship of cementless TKA [103]. In a systematic review of 5 randomized controlled trials (RCTs) and 297 TKAs, Nakama et al. [102] showed aseptic loosening was twice as likely for cemented TKA at 2 years. Additionally, Voigt and Moiser [104] showed hydroxyapatite-coated tibial components had less migration by radiostereographic analysis compared to both porous-coated and cemented tibial components. Revision rates were similar at 2, 8, and 10 years [104]. For 80 bilateral TKAs (160 knees) in patients younger than 55, Kim et al. [105] randomized one knee to trabecular metal cementless TKA and one to a cemented construct. At 17 years of follow-up, survivorship for aseptic loosening or revision for any cause was similar, and there was no difference in KSS, UCLA activity scores, or ROM. Finally, in an RCT of 397 knees, Pulido et al. [106] showed that both cemented and uncemented highly porous metal tibias provided comparably durable fixation and reliable pain relief and restoration of function when compared with a traditional cemented modular tibia in TKA.