Unicompartmental knee arthroplasty has been designed to replace only one tibiofemoral compartment of the diseased knee, most commonly the medial compartment. Popularity and usage in the 1980s faded, following reports of high early conversion rates to total knee arthroplasty1,2. However, by the late 1990s, interest in unicompartmental knee arthroplasty heightened due to the many potential advantages, such as bone preservation, reduced operating-room time, better postoperative range of motion, improved gait, and greater patient satisfaction. Despite this interest, considerable issues still exist today, including the problem of early failure of both the femoral component3-8 and the tibial component4,9,10.
It has been well documented that early failures are primarily due to inaccurate positioning of components, which leads to overcorrection or undercorrection of the final alignment of the limb11-14. Malalignment of the limb is associated with increased polyethylene wear15, disease progression to the opposite compartment15-18, and implant loosening19. Malalignment of the femoral component is known to cause femoral fracture20 and tibial component loosening21. Excessive posterior slope (>7°) of the tibial component is associated with tibial component loosening22, rupture of the anterior cruciate ligament22, and increased bone stresses23. Thus, while unicompartmental knee arthroplasty has provided the opportunity for many clinical advantages, the technical difficulties associated with accurately aligning the implanted components have prohibited widespread clinical acceptance.
Computer-assisted surgical navigation has been introduced as a way of improving accuracy in less invasive procedures. The technical advantage of computer-assisted surgery is accurate intraoperative feedback of the positioning of the cutting jigs relative to the host bones. When augmented with the use of computer-assisted surgical navigation, unicompartmental knee arthroplasty produced more accurate and reproducible alignment than could be achieved with conventional unicompartmental knee arthroplasty24,25 or minimally invasive unicompartmental knee arthroplasty26. However, the use of navigation has yet to be associated with an improvement in clinical outcome after unicompartmental knee arthroplasty27. In addition to providing passive navigation, robotic systems may serve as a delivery tool for a surgical procedure that has been planned preoperatively in three-dimensional space. Robotics can be classified as either active or passive. In an active system, the surgeon positions the robot by means of a referencing procedure and then supervises the actions of the robot without the ability to modify (with the exception of interrupting) the procedure online. In a passive robotics system, the surgeon maintains control of the resection, but the robot restricts the volume in which the cutting tool can be maneuvered. The difference between robotic systems and computer-assisted surgery systems is that the surgeon can ignore the feedback and cut without restriction with use of a computer-assisted surgery system, whereas a robotic system constrains the motion of the cutting tool to only the planned volume of resection. Robotically assisted orthopaedic surgery has the potential to achieve levels of accuracy, precision, and safety not possible with computer-assisted orthopaedic surgery. It also brings a level of surgical participation that allows the surgeon to operate the system and not just evaluate the information. Its overall acceptance will be driven by reproducible accuracy, acceptance of added time to the surgical procedure, integration into the surgical flow, cost, and patient success. A review of the literature indicates that improving the accuracy of the surgical technique at the time of the index procedure can minimize each of the modes of implant failure, potentially improving the results of present and future procedures.
Due to the nature of the technology, most robotic systems in knee surgery are currently not minimally invasive. This is often due to the requirement of immobilizing the bones, the exposure needed to register the surfaces to be cut, and the bulkiness of the robotic equipment. Currently, the only minimally invasive robotic system used for knee surgery is a passive robotic-arm system called a Tactile Guidance System (TGS; MAKO Surgical Corporation, Fort Lauderdale, Florida). This system is currently used for implantation of unicompartmental knee arthroplasty components. The platform allows preoperative planning with an ability to adjust the plan intraoperatively. The robotic arm, which is under direct surgeon control, gives real-time tactile feedback to the surgeon as the procedure is performed. This surgical platform blends surgical interactive robotics, computer-assisted planning, and guidance with an intelligent bone cutting and/or shaping tool through use of minimally invasive techniques.
Preoperative computed tomographic scans are made of the patient, with slices acquired through the hip and ankle at 5-mm increments and through the knee joint at 1-mm increments. A plan is then created with use of three-dimensional reconstructions of the tibia and the femur in combination with computer-aided design models of the implant components. This allows for the preoperative fine-tuning of the femoral and tibial implant positions, including coronal and sagittal alignment, overall limb alignment, identification of gross anatomical deformities (e.g., cyst or vacuole), overlapping of the components throughout flexion, relative varus-valgus intercomponent alignment in the coronal plane, and component-to-bone positioning of the tibial component. The robotic technology consists of a robotic arm and a standard optical infrared camera (Fig. 1). Standard bicortical surgical navigation markers are placed in the femur and the tibia and are also mounted on the robotic arm. This system does not require rigid fixation of the robot to the patient. The surgeon utilizes a leg holder to hold the leg stable during resection, but the surgeon is still able to position the knee optimally to ensure access to the targeted surfaces.
As with traditional navigation, the hip center is obtained by conducting several rotations of the hip in an increasing circumduction fashion. The ankle malleoli, the femoral epicondyles, and the center of the tibia (the base of the anterior cruciate ligament) are identified. The landmarks are verified against the computed tomographic plan. Under computer-screen guidance, various landmarks and cortical surfaces are digitized to register the actual bone geometry to the virtual three-dimensional reconstruction. Once the bone geometry is registered, the virtual modeling of the patient's knee and intraoperative tracking allow real-time adjustments so that correct knee kinematics and soft-tissue balancing can be obtained. First, osteophytes interfering with medial collateral ligament function are removed and capsular adhesions interfering with knee function are relieved. As one of the indications for a unicompartmental knee arthroplasty is a correctible deformity, removal of these impediments makes it possible to achieve correct limb kinematics and tissue tension during passive manipulation of the knee throughout the full range of motion with an applied valgus stress. The three-dimensional poses of the femur and the tibia are captured throughout the range of motion with the medial collateral ligament properly tensioned. This provides for correct bone spacing (extension and flexion gaps) during implant planning such that after resection and component implantation, knee mechanics will be properly restored throughout the range of motion of the knee. The articular surfaces of the components are then planned to fill that space throughout the range of motion, ensuring the correct soft-tissue balance.
With each planned position of the implants, a visual representation of the tightness or looseness of the knee is displayed at every captured pose angle. The distance between planned components is displayed through a bar graph depicting this gap distance at the flexion angle captured (Fig. 2). Each bar represents knee tightness or looseness at one knee pose. For example, if four poses were captured at 10°, 40°, 60°, and 90° of flexion, four gap numbers would be calculated, and four bars would be displayed on the graph. The blue bars represent the flexion angles at which the components, if implanted, would have a gap between them. Orange bars represent the opposite. If the components were implanted as planned and the leg brought to a pose represented with an orange bar, the knee would be tight at that flexion angle (90°). By adjusting the implant plan, the flexion-extension gaps can be planned a priori to attain the desired postoperative values.
The end of the robotic arm is equipped with a burr that is used to resect the bone. The robotic arm facilitates controlled bone resection by applying stereotactic boundaries to a cutting burr. Stereotactic, or haptic, boundaries are virtual walls created by the software and implemented through the robotic arm hardware to restrict the cutting tip to within a predefined resection volume. This haptic boundary is not restricted to any particular geometry and is defined by the shape of the implant and the depth of resection. If the surgeon attempts to move the cutting burr past the predefined volume while the cutting burr is set to the burring mode, the robotic arm applies a force simulating contact with a rigid wall, thereby confining the tip to the correct region in space. While inside the volume of bone to be resected, the robotic arm operates without offering any resistance. Thus the robotic arm effectively acts as a three-dimensional virtual instrument set that precisely executes the preoperative plan, giving visual, tactile, and auditory real-time feedback (Fig. 3). The knee may move during the burring process, with the virtual haptic volume moving with it accordingly. However, if the motion is too fast, the safety features will turn off the motor to the burr. The visualization on the computer screen allows the surgeon to concentrate on the haptic feel and visual boundaries, devoid of direct visual inspection of the joint while burring. Permanent graphical feedback on the navigation screen depicts the actual achieved cavity compared with the planned cavity, specifically based on preoperative planning. Once both cavities have been prepared, femoral and tibial trial components are inserted and a complete flexion-extension arc is performed. Computerized simulation of the implants in situ shows the actual overlapping of the implant components, giving the surgeon feedback about the current limb alignment and knee gap kinematics. Finally, the implant components are cemented (Fig. 4) and a final range of motion is taken to compare the original, trial, and final implant kinematics and knee alignment. Typical radiographs are shown in Figure 5.
The goal of this new robotic-arm-guided unicompartmental knee arthroplasty procedure is to accurately and reproducibly resect bone relative to the preoperative plan to facilitate the implantation of a medial unicompartmental prosthesis. Robotic arm technology has the added advantage of permitting intraoperative control of the preparation of the osseous surfaces directly through the tactile guidance of a cutting burr rather than through use of optical guidance and a cutting block through which a traditional saw blade is manipulated. Because the bones to be prepared with this robotic arm technology are currently tracked with optical markers, the same intraoperative drawbacks of traditional navigation apply—namely, increased setup and operative time in addition to the requirement of maintaining a line of sight between the camera and the infrared tracking markers.
The integration of robotics into the orthopaedic workflow will allow surgeons to maximize the success of the critical steps of their surgical performance. The coupling of reproducible accuracy and a quick learning curve will allow surgeons with low-volume experience to achieve the level of success seen after years of mastering a technique. In addition, robotic-assisted orthopaedic surgery combines minimally invasive techniques with a virtual visualization system to achieve a surgical plan that has been identified and confirmed prior to cutting. Robotics has advanced the field of computer assistance to the next level of surgical guidance. Appropriate patient selection and preoperative planning allow the surgeon to rapidly confirm the plan intraoperatively and modify it to the patient's kinematics and soft-tissue balance to obtain the most appropriate mechanical alignment and implant positioning. This platform will allow multiple applications in the knee to be robotically assisted to improve the accuracy of the procedure. It will open up multiple applications to other joints as well as to the spine. The ability to attach specific cutting instruments to the robotic arm will improve our ability to cut bone accurately, facilitating further improvements in implant design and surgical technique. 