This chapter reviews the history of patient-specific instrumentation (PSI) and then outlines the design rationale, surgical techniques, and results seen in kinematically aligned (KA) total knee arthroplasty (TKA). The goals of KA TKA are to (1) restore the native femoral and tibial articular surfaces, (2) restore the native knee and limb alignment, and (3) reestablish the native tibial compartment forces and laxities of the knee. Achieving KA with PSI is made possible using techniques and technologies developed over the past decade. Using an accurate and reliable PSI system, coupled with secondary caliper checks, enables successful kinematic balancing in every patient. PSI offers advantages in surgical precision, potentially improved operative time, and advanced preoperative planning. It also obviates the need for additional pinholes or intraoperative registration otherwise seen in navigated knee replacements.
History of patient-specific instrumentation
PSI for TKA has been available for over 10 years with often mixed results. Many early versions of this technology relied on magnetic resonance imaging (MRI) modeling of the distal femur and proximal tibia. This yielded an inexact delineation between subchondral bone and cartilage. It also led to intraoperative uncertainty on the part of the surgeon during its application, because of a less reproducible positioning of the guide on cartilage. Furthermore, most early systems did not come with three-dimensional (3D) models of the distal femur and proximal tibia, etchings on the models, or the ability to attach extramedullary alignment rods to verify registration of the guide to the articular surface. In one instance, a manufacturer uses axial imaging of the knee in conjunction with standing full-length X-rays to properly align the hip-knee-ankle (HKA) axis. Flexion contractures of the knee and rotation of the limb make this type of system prone to error. Newer designs often use computed tomography (CT) scans for computer modeling, which has yielded greater predictability in its application once remaining cartilage is removed. , Studies have shown CT-scan modeling to be dimensionally more accurate than MRI. Nonetheless, the ultimate goals of any PSI system remain unchanged: allow preoperative planning, enhance surgical workflow, and increase precision and reproducibility.
Design rationale of modern patient-specific instrumentation systems
In designing the ideal PSI system, we must consider limitations of former versions. As many have failed to achieve the goals of accuracy and reliability, there have been numerous reports from authors who discourage the routine use of PSI during TKA. , Many of these unfavorable accounts relate to the inaccuracy of the underlying PSI system itself. It should be stressed that not all PSI systems are created equal. Unique limitations exist for each system that may not translate to other manufacturers. These inaccuracies can stem from a number of potential issues. However, based on our extensive experience, we have concluded that several factors can adversely affect the quality of the PSI product.
First, as alluded to previously, MRI-based PSI cutting blocks tend to be less accurate than CT-based ones. MRI scans are less accurate for determining joint line definition because cartilage margins can only be estimated, leading to potential reconstruction errors, cutting block mismatch, and lack of accuracy. MRI scans are also more expensive, require longer scan times, are more susceptible to patient motion artifact, and are often adversely affected by periarticular hardware. The MRI-based PSI blocks themselves rest on cartilaginous surfaces, which are inherently softer and less precise. Conversely, CT-based cutting blocks facilitate more reproducible positioning and bone resections once cartilage and synovium are surgically removed with Bovie cautery. In addition, large osteophytes and prominent bony landmarks can serve as contact points for accurate CT-based PSI block placement with minimal intersurgeon variability.
Second, the modern design advocated by the lead authors uses 3D-printed bone models. Exact replicas of the distal femur and proximal tibia are provided sterile to the surgeon for intraoperative use ( Fig. 6.1 ). These yield direct visualization of intended resections, angles, and component positions before making the actual bone cuts. The 3D models with etchings define contact points for cutting blocks. The intended resections can be confirmed visually on the bone and cross-referenced with the model. In addition, alignment slots exist for secondary check of accurate cutting block placement by adding extramedullary alignment rods. Therefore, in addition to the unambiguous registration of these guides to the bone, these resection etchings and alignment slots serve as secondary checks of planned resections, before actually completing the bone cuts.
Patient-specific instrumentation in the kinematic total knee arthroplasty: surgical technique
In kinematic balancing of TKA, the first objective is to restore the native prearthritic femoral and tibial articular surfaces. , This begins with precise “anatomic resurfacing” of the distal and posterior femur. This restores the native femoral-tibial flexion-extension axis and the patellofemoral flexion extension axis of the knee. , The advantage of PSI remains that preoperative planning can accommodate any surgeon preferences, which can be adjusted to address any degree of deformity ( Fig. 6.2 ). In cases performed by the lead authors, plans are made for 7-mm distal femoral and 6-mm posterior bony femoral resections. If we assume 2 mm for average cartilage thickness on the distal and posterior femur, this corresponds to a total cartilage and bone thickness of 9 mm distally and 8 mm posteriorly, which are the exact thicknesses of the femoral implant used by the authors distally and posteriorly, respectively. The amount of bone resection needs to match the thickness of the implants, minus 1 mm adjusting for sawblade kerf. Immediate feedback is provided by cutting and then subsequently measuring resected bone with a caliper following each stepwise cut performed to ensure 100% accuracy before moving on to the next step. This is a CT-based PSI system so the bone resections are exact and the calipered measurements are solely affected by cartilage, if any remaining.
A CT scan is performed according to a specific algorithm (@Medacta International) that acquires 0.5 to 1 mm thick axial images of the knee and 2 mm thick axial images of the hip and ankle using a 512 × 512 matrix. Other available PSI-based implants are commercially available and readily adaptable to kinematic principles. However, unlike another PSI system on the market that relies on a full-length CT scan of the lower extremity, this system uses a limited and low-dose CT scan of the knee, hip, and ankle. Images are uploaded through a secure server to a dedicated engineering team in Switzerland who create the PSI plan based on our kinematic knee preferences ( Fig. 6.2 ). The primary goal is to resurface the distal femur, and bone resections are set to match the exact thickness of the femoral implant (9-mm distal femur and 8-mm posterior femur). Therefore, the PSI-CT bone resections are set at 7 mm of bone from the distal femur and 6 mm of bone from the posterior femoral condyle. These cuts assume 2 mm of native cartilage thickness, which will resurface the femur to its native position respecting the three kinematic axes about the femur. Tibial resection depth is set at 6 to 8 mm depending on surgeon preference and patient variables. An 8-mm resection of bone results in a total resection of approximately 10 mm (assuming 2 mm of cartilage thickness) which exactly replicates the thickness of the tibial implant with the thinnest 10-mm polyethylene insert. A more conservative 6-mm cut on the tibia may be preferred to allow for the possibility of a recut if necessary, to balance the knee. The ultimate goal on the tibial resection is to evenly resect the medial and lateral tibial plateaus to resurface the tibia, while preserving host bone and ending up as close as possible to a 10-mm polyethylene. In the section later we cover the critical evaluation of resections and balancing of the flexion and extension gaps. Once the femur is anatomically resurfaced, the corrections needed to balance the knee are performed with tibial recutting guides. In our experience, resecting 2 mm less on the planned tibial resection is appropriate for (1) surgeons new to this technique; (2) severe deformity; (3) significant valgus deformity; and (4) significant bone loss (may reduce by 4 mm in select cases of significant tibial bone loss). Fortunately, each case is critically evaluated by a dedicated PSI engineering team. If anything unusual is identified by the engineer, the surgeon is alerted and can adjust the plan based on abnormal anatomy, retained hardware, or bone loss, before creating the final PSI cutting guides. Final approval of the operative plan is done by the surgeon on a 3D preoperative planner and the cutting blocks and bone models are then printed on a 3D printer and shipped ( Fig. 6.1 ). Typical time from CT scan to delivery to the operating room is 3 weeks.
Femoral resections with caliper verification
With KA, the anatomical placement of the femoral component is critical to the overall balancing of the knee. It determines the native arc of flexion and extension of the tibia on the femur and the native arc of flexion and extension of the patella on the femur. , In conventional KA TKA instruments, variability exists for the distal femoral starting drill hole, through which an intramedullary guide determines sagittal alignment of the femoral component. Instead, PSI guides allow very little variability in this plane of motion and can restore the native sagittal plane in a reproducible fashion. ,
The knee joint is exposed using the surgeon’s approach of choice. We have used standard medial parapatellar, minimedial parapatellar, subvastus, and midvastus approaches with these CT PSI blocks. After adequate femoral exposure, the surgeon will hold up the plastic femoral bone model, paying specific attention to the four etch marks representing the contact points for the femoral PSI guide ( Fig. 6.3 ). Using Bovie in combination with a smoke evacuator and extra suction, we remove all of the soft tissue at these four contact points to expose subchondral bone. The Bovie will remove cartilage, synovium, and any other soft tissue while leaving the underlying subchondral bone footprint. It is imperative that surgeons do not remove bone with a rongeur or curette as these blocks are designed to sit on bone, and often large osteophytes are used as hard contact points. The surgeon should then position the femoral guide on the femoral bone model to see how the guide fits. Registration of the guide to the bone typically involves engaging the anterior footprints and then flexing the guide down such that the distal footprints are contacting bone. The femoral guide is then placed on the distal femur and secured with two parallel and one cross pin at the anterior surface of the femur. Care must be taken to ensure that all four contact points are flush to bone. If any contact point is not flush to bone, the pins and the guide must be removed and then repositioned.
Anterior reference holes for component rotation are drilled to establish placement of the 4-in-1 guide. The PSI femur guide has a number of built-in secondary references that allow the surgeon to critically evaluate the intended resections before performing the definitive cuts. Fig. 6.4 illustrates the anterior reference slot on the PSI block to ensure proper anterior resection and to avoid anterior notching. Fig. 6.5 illustrates the visual comparison of the intended distal femoral resection and the etched resection on the plastic bone model. Once proper block position and placement are confirmed, the surgeon performs the distal femoral resections through the distal cutting slot of the CT PSI femoral guide. Distal resections are identified as medial and lateral with a marker. Resections are then measured by the surgeon and recorded on our worksheet. In our experience, these resections are often exact or within 1 mm of intended thickness and small corrections are easily attainable. The surgeon must assess whether there is complete cartilage loss (2 mm), partial cartilage wear (1 mm), or no cartilage loss in evaluating the thickness of the resections. We are targeting matched resections of the distal femur while accounting for differential cartilage wear patterns. If medial or lateral distal condyle is underresected by 1 mm, we run the saw blade through the plastic PSI distal cutting slot and resect 1 mm of additional bone from the underresected distal condyle. In the case of relative overresection of one condyle, we have 1- and 2-mm washers that can be placed on the affected distal condyle before chamfer cuts to ensure proper resurfacing of the distal femur. For example, if one condyle is overresected by 1 mm, we place a 1-mm washer on this side of the 4-in-1 cutting guide to appropriately resurface the distal femur ( Fig. 6.6 ). This results in the chamfer cuts being 1-mm proud and, once cemented, a 1-mm thicker cement mantle on the overresected distal femur surface and appropriate contact with remaining distal femur cuts. Again, our goal is to resurface the femur, preserve the three native kinematic axes around the knee, and avoid changing the joint line.