The use of custom implants for total knee arthroplasty

The paradigm shift

In the last two decades surgeons have progressively and, yet unconsciously, shifted from a systematic alignment approach for TKA to a more personalized alignment. But why the shift? Various studies recognized that the human morphology varies enormously [1–3], more than we previously thought, so applying a systematic approach of the same measurements and same implants to restore anatomy in all patients is inefficient. Very few knees can follow the equation of 90° + 90° = 180°, therefore we now know that this approach will not restore the native anatomy. In fact, many patients have a natural alignment outside of the ± 3° valgus or varus range and correcting this alignment to conform to 0° will only provoke more patient discomfort, pain, and reduced quality of life [4]. Moreover, another short fall in this approach is that the term 'biomechanical alignment' is a misconception, because simply measuring angles on a two-dimensional (2D) x-ray does not provide valid biomechanical insight into the functioning joint; a more geometrical construction in 3D must be adopted to appreciate each patient's natural alignment, anatomy, and biomechanics.

Accepting this shift to kinematic alignment (KA) and constitutional alignment (CA) with personalized planning and custom implants means that three main issues need to be addressed:

1. How much varus is too much and what is the acceptable limit? Several papers have tried to answer this question. One well-cited paper indicated that maintaining alignment within ± 3° improves patient outcomes and implant survival [5]. However, the short sight in this and other papers is that alignments were measured on short-film x-rays, while long-leg weight-bearing x-rays are needed to determine the full alignment.
2. Distinguishing between constitutional versus arthritic deformity. Previously, all plannings were based on 2D x-rays. However, 2D x-rays show only the global deformity, which is a combination of the constitutional and the arthritic deformity including bone wear, laxity, knee rotation, loss of flexion etc. If the surgeon cannot differentiate the constitutional deformity from the arthritic deformity, the final TKA alignment can hardly reproduce the native (prearthritic) alignment. Unfortunately, very little of the literature differentiates between arthritic and prearthritic deformity, because of the lack of tools.
   For example, if a patient has a 16° global varus alignment due to a constitutional tibial varus, a 90° tibial cut will result in a very asymmetrical resection between the medial and lateral plateaus and will remove a lot of bone laterally, resulting in lateral laxity, tightness on the medial side, and difficulty for balancing ligaments. In such a patient, a varus cut at 85° will provide a more symmetrical resection with well-balanced ligaments and a good quality remaining bone. If you take another patient where the same 16° of global varus deformity is due exclusively to the bone wear on the medial tibial plateau, a 90° cut would allow for a tibial baseplate to be fitted, the gap to be filled, and compensate for the bone wear, ultimately restoring the native anatomy. However, if a varus cut is performed in that patient, the medial tibial plateau will sit on weak bone and has a high risk of failure.
3. Off-the-shelf TKA implants are not designed to be positioned differently in each patient: When using KA technique, the same standard prosthesis will be implanted slightly differently in all patients, which can create an overhang of posterior condyles or trochlea, and a patellofemoral maltracking due to the deviation of the trochlea [6] and to the internal femoral rotation relative to the transepicondylar axis [7].

How do surgeons overcome these issues? Bonnin and his colleagues strongly believe that there is a need to bridge the two concepts so that if a surgeon adopts a CA, then a custom TKA should also be used [8, 9]. To address the above three problems, surgeons need to adapt and incorporate planning from CT scans and 3D analysis, generate anatomical implants, and design implants that adapt to the patient's native alignment. Thus, the aim of a custom implant is to reproduce the native shape of the knee, including the native radii of curvature both in the coronal and sagittal planes, the size of the trochlear, the rotation of the femur, the asymmetry of the femoral condyles and tibia, and the native orientation of the joint line.

Considerations and techniques for custom total knee arthroplasty

There are many aspects a surgeon needs to consider when using personalized alignment and custom TKA. Here, Bonnin goes through his preoperative planning and tips for custom TKA.

Types of patients: The advantage of CA and custom implants is that a majority of patients requiring a TKA can be considered candidates for this approach, including patients presenting with lateral, medial, global, or patellofemoral osteoarthritis (OA), and those with rheumatoid arthritis. In general, indication for surgery is end-stage OA, and usually body mass index or residual hardware from previous procedures are not a contraindication for surgery. There are some exceptions, though, as patients with severe coronal deformities > 15°, stiff knees with extension deficit > 15°, severe medial laxity > 10°, or severe lateral laxity > 15° may not be considered for custom TKA.

A cohort of patients with prior osteotomies and/or extraarticular fracture sequelae with mild residual deformities have been treated with this technology, with satisfactory outcomes [10]. However, the place of custom implants in such complex situations needs to be clarified and is limited to minor/mild deformities. More studies like this will give better insight into the feasibility of these techniques in difficult patients.

Preoperative planning: The beauty of custom implants comes from the advances in 3D imaging and modeling that enable engineers and surgeons to estimate the bony wear of the articular surfaces due to OA and by comparing this wear with the nonworn areas of the bone, to deduce the native prearthritic shape of the femur and tibia (Figure 1). For a precise 3D model to be generated, all patients must have a CT scan according to the implant manufacturer's protocol, with a 6- to 8-week waiting period for the design and manufacturing process. The CT images are used to assess the native femoral mechanical angle (FMA), tibial mechanical angle (TMA), and hip-knee-ankle (HKA) angle. Several of the manufacturer's systems design the prosthesis based on 3D analysis of bony anatomy, arthritic deformities, and native limb alignment using these CT scans. The objective of the implant design and production is that [8]:

• The femoral component reproduces the contours, sagittal radii of curvature, and joint line obliquity of the native femur.
• The tibial baseplate reproduces the contours of the native tibial plateau.
• The polyethylene insert has up to 2 mm difference in thickness between the medial and lateral compartments.

The manufacturers will also produce single-use patient-specific cutting guides.