Robotic-assisted total knee arthroplasty: a game changer in alignment strategies

Robotic-assisted total knee arthroplasty: the advent of semiactive and fully active systems

The term "robot" was initially coined in 1920 by the Czech science-fiction writer Karel Capek, who described artificially humanoids repeatedly performing labor tasks [2]. In the operating room, robots were initially designed to minimize human error while maximizing operative accuracy [3]. Since the first reported robotic-assisted total knee arthroplasty (raTKA) in the early 2000s [5], the surgical tasks performed by robotic-assisted surgery platforms have rapidly evolved. Currently, all robotic technologies use dynamic referencing to assess intraoperative knee stability, alignment, and range of motion (ROM) while adjusting bone resection, ligament balance, and position of the prothesis [2, 6]. Based on how bone resections are conducted, robotic-assisted surgery platforms can be either classified as "fully active" (ie, a robotic arm performs autonomously preprogrammed bone resections on the patient) or "semiactive" (ie, the surgeon performs bone resection within a predefined zone that has been set in the preoperative plan while receiving intraoperative feedback) [1, 3, 7]. Such systems normally include a haptic interface by which the surgeon receives information about the forces and stress exerted on articular surfaces and can adjust his/her movements intraoperatively. Preoperative computed tomography (CT) scans or magnetic resonance imaging (MRI) enables the surgeon to generate a virtual three-dimensional (3D) model of the patient-specific bony anatomy. The 3D virtual model is employed to preplan bone cuts, implant size, as well as positioning, and is subsequently mapped intraoperatively to the patient's bone anatomy using navigational trackers, in a process known as registration [8, 9]. Another type of semiactive systems are handheld robotic burrs, which are manually controlled by the surgeon. Instead of a haptic interface, the robot follows the navigation field's burring tool trajectory, controlling the speed and exposure of the device to protect against iatrogenic ligament injuries [3, 10]. Contrary to haptic systems, this type of semiactive system is imageless (ie, it is not based on CT/MRI scans) and compatible with a wide range of prosthetic implants and brands [2, 10, 11]. Each robotic device is associated with several benefits and limitations (Table 1). While semiactive systems are the most frequently used today, the appearance of second-generation fully active systems might change the orthopedic landscape in the near future.

Clinical and functional outcomes of robotic-assisted total knee arthroplasty

In the orthopedic field, the superiority of raTKA over conventional TKA in improving implant positioning and lower limb alignment is currently well accepted [19–21], albeit some systematic reviews offer conflicting results. In an attempt to compare similar systems, Batailler et al [9] reviewed the functional outcomes of 14 studies performing a MAKO CT-based robotic arm-assisted TKAs. The authors claimed that patients treated with raTKA reported reduced postoperative pain and they required shorter hospital stays in comparison with conventional TKA-treated patients [9, 22, 23]. Furthermore, patient-reported functional scores were equal or slightly better in the raTKA cohort at 1 year after surgery. However, treatment improvement declined over time [9, 24]; the use of raTKA led in the mid-term to equivalent functional outcomes as conventional TKA, a trend also reported when using other robotic-assisted systems [3]. When assessing the complications associated with raTKA in a recent metaanalysis, Zhang et al [25] documented nearly identical complication rates for raTKA and conventional TKA. In addition, they reported a reduced deviation between the targeted prothesis position and the implanted one in the raTKA cohort [25]. An important shortfall of these reviews is the comparison of cohorts without taking into consideration the heterogeneity of limb alignment techniques or implant positioning methods. To improve the accuracy of interpreting raTKA outcomes, both criteria should be considered in future raTKA studies.

Is raTKA a cost-effective technique?

The currently used semiactive robotic systems offer surgeons greater technical reliability than manual TKA [3]. This, however, comes at the expense of a high economical investment. The major costs in raTKA, alongside equipment acquisition and maintenance, derive from preoperative imaging, implants, and operation consumables. In a recent publication, Chen et al [26] reported that the use of raTKA added additional costs of 6.1–12% to the total surgical expense in comparison with conventional arthroplasty [26]. In contrast, a recent economic analysis determined an increased average cost at 90 days after surgery in patients younger than 65 years of age undergoing conventional TKA compared with the raTKA cohort. The lower postoperative visit rate to skilled nursing facilities for raTKA patients (2.24% among all treated-patients vs 4.37% in the conventional TKA cohort) accounts for such differences [27], raising the question why raTKA was initially not deemed cost-effective. When we ask Lustig about this topic, he warns us, "when evaluating the costs of raTKA, one must evaluate not only the price of the equipment but also the benefits regarding the costs of the hospitalization, complications, and surgical revisions". Thus, only long-term studies considering such parameters will evaluate whether raTKA is a cost-effective technique or not.

Robotic-assisted total knee arthroplasty and functional alignment philosophy: an intertwined road

The promise held by the robotic systems of performing accurate bone cuts to the desired resection levels was undeniably attractive to knee surgeons. Hence, these systems quickly made their way into clinical practice, especially for those surgeons who are interested in functional alignment (FA). This emerging philosophy stems from kinematic alignment (KA) and, as its predecessor, it aims to restore the kinematics of the native preosteoarthritic knee by resurfacing the femorotibial joint while preserving the soft-tissue envelope. The key difference of FA over KA is that it attempts to realign the knee in all three axes (ie, coronal, sagittal, and axial), not only in extension but also in flexion, and takes into consideration the height and orientation of the joint line, the optimal trochlea positioning, and a personalized ligament balancing [21]. Functional alignment shares with KA an initial constitutional coronal extension alignment approach, but adds ligament stress values throughout the operative process to optimize implant position [28] (for further discussion on the types of alignment, see in Part 1 of this series). In the following sections, Lustig guides us step by step through the preoperative planning and surgical workflow of FA in TKA using a CT-based robotic device.

Personalized preoperative planning

A personalized preoperative plan according to FA alignment starts with 3D imaging of the patient's anatomy and the generation of a virtual plan of implant positioning.

Imaging and calculation of essential parameters

Imaging includes hip-to-ankle x-rays, CT scans, or MRIs. Essential parameters that need to be calculated are hip-knee-ankle (HKA) angle, mechanical lateral distal femoral angle (mLDFA), mechanical proximal tibial angle (mPTA), and posterior tibial slope (PTS) measurers.

Virtual implant positioning plan

When generating a patient-specific preoperative plan, the surgeon must bear in mind that FA considers the behavior of varus and valgus knees to be significantly different, which warrants a different approach for each. While we are describing a systematic approach for varus knees to be followed, detailed parameters for each joint morphotype are displayed in Table 2.

1. Reconstitution of native alignment through coronal alignment: The target coronal alignment guide is set within the limits of an HKA angle ranging from 174° to 180°.
2.  Implant adjustment to respect joint line height and obliquity: The main goal is to maintain the joint line obliquity (JLO) of the patient. It is preserved through orientation of the femoral and proximal tibial bone using the arithmetic hip-knee-angle (aHKA) method. It is calculated by subtracting the LDFA from the mPTA [29]. If the aHKA is negative, the constitutional alignment is varus, and valgus if the aHKA is positive. Importantly, the maximum alteration to JLO is set within the limits of ± 3 mm, and when larger gaps are present, the depth of the cuts should be adjusted by a combination of femoral and tibial alignment.
3.  Equal laxity of tibiofemoral compartments in flexion and extension: The extension and flexion gaps in the medial and lateral compartments should be equal to the global thickness of the implant. Balanced laxity is achieved prior to cuts being made by placing the implant in a position that fits the behavior of the patient's knee through an arc of flexion.
4.  Implant positioning and size: The size of the femoral component is selected using posterior referencing; in general, the smallest size is chosen and must not overhang the femur, notch the anterior femur, or overhang mediolateral bone edges, and should avoid overstuffing the patellofemoral joint. Its position is first matched around the distal femoral radius of curvature, restoring the native depth of the trochlear groove. In the coronal plane, the position of the femoral component is modified from a starting point of 0° to the mechanical axis to balance the extension gap, while the sagittal plane is positioned to optimize component sizing and to avoid femoral notch by flexing up to 10°. In the axial plane, the femoral implant is aligned to the transepicondylar axis with 3° of freedom to balance the flexion gap. The femoral resection should match LDFA in the coronal plane modified by the soft-tissue laxity profile. In the case of the tibia component, its position in the coronal plane is aligned to the tibial mechanical axis and modified to balance extension and flexion gaps by up to 6° of varus. Valgus tibial position should be avoided. In the sagittal plane, the tibial component position is set to match the patient's prearthritic posterior tibial slope, modified to balance the flexion gap if necessary. In the axial plane, the tibial component is positioned using the Akagi line. Tibial resection should match mPTA modified by soft-tissue laxity.
5.  Final limb sagittal alignment to achieve full extension: The objective is the total correction of any sagittal deformity, with a sagittal alignment set between 0–5°.
   
   

Registration, ligament balancing, and bone resection

1. Incision and registration: The surgical procedure starts with the removal of all osteophytes and adhesions followed by the registration of femoral and tibial landmarks.
2. Ligament balancing: To determine the "neutral" alignment before any bone cuts are performed, the effects of the preplanned implant position and bony cuts in both flexion (90°) and extension (0–10°) positions are displayed on the computer interface. Based on this feedback, the surgeon can correct the virtual position of the implants to correct any asymmetry between medial and lateral gaps. Any deviations of the desired extension space can be corrected by adjusting the femoral and tibial component position in the coronal plane within ± 3 mm variation from the joint line height. In contrast, any asymmetry in the flexion space is modified by rotating the femoral component of the implant (for specific details for varus or valgus phenotype, see Table 2).
3. Bony resection and implant positioning: once the knee is considered to be balanced, the bone cuts are performed with the robotic arm, the remaining osteophytes are removed, and the final prothesis is inserted

The promising clinical outcomes of functionally aligned total knee arthroplasty

With random clinical trials comparing functionally versus mechanically aligned TKA still ongoing [30], most current evidence about the clinical outcomes of FA arises from a handful of studies. For example, in a series of 110 consecutives TKAs, Shatrov et al [31] reported that kinematically aligned TKA failed to restore intraoperatively a balanced knee in 65.7% and 49.1% cases for the extension and flexion gap, respectively. Instead, better alignment and decreased bony resection were achieved when implementing FA principles. In addition, another study from the same authors comparing mechanical alignment (MA), FA and KA showed that a kinematically placed femoral component led to positioning considered unsafe in over 13% of cases compared to a functionally placed femoral component restoring most closely trochlear depth in all three positions of flexion [31]. The superiority of FA over KA in achieving excellent functional outcomes is also supported by Clark et al [32], who documented limb alignment and implant positioning in a series of 650 TKAs performed with an arm-assisted robot with a 2-year follow-up; with a mean aHKA range of 1.8° varus (6.5° varus–4.6° valgus), and a mean range coronal position of 2.4° (3.2° varus–7.9° valgus) and 4.2° varus (6.4° varus–3.7° valgus) for the femoral and tibial components, respectively; the authors described minimal deviations in the desired alignment targets. Moreover, patients treated with FA raTKA significantly improved knee ROM (105° flexion preoperatively vs 125° flexion postoperatively, P ˂ .001) and pain (Oxford Knee Scores of 22 preoperatively vs 77 postoperatively, P ˂ .001) 2 years after surgery. However, this is not a complication-free technique; among 650 surgeries, the authors reported over 40 complications, although most of them were minor in nature [32]. Long-term studies assessing the clinical outcomes and implant survivorship associated with FA TKA are required to establish whether or not this technique is superior to conventional TKA.

Conclusions and future perspectives

Robotic-assisted platforms are becoming increasingly popular, especially among the detractors of MA. The era of one-size-fits-all is reaching an end, and the use of patient-specific arthroplasty is booming at the moment. With the aid of robotic-assisted surgical platforms, surgeons can provide the patients with improved limb alignment and prothesis position without iatrogenic ligament injuries. As Lustig points out, "robotic-assisted surgical platforms have turned TKA into a more reliable and reproducible technique. The promising short-term results observed with raTKA, if consistent in the long run, together with the use of customized implants (see Part 3) are paving the way for personalized knee arthroplasty in the near future".