For a long time, the alignment goal for total knee arthroplasty (TKA) was a neutrally aligned limb with orthogonal joint lines to the mechanical axes (mechanical alignment [MA]). This had mainly to do with the belief of a better long-term durability because of more equal load distribution in mechanically aligned TKA. Recognizing the variability in individual knee alignment and compromised functional outcomes in TKA, there is an increasing interest among knee surgeons for more personalized, more anatomical alignment (AA) methods. Among those newer alignment concepts, the kinematic alignment (KA) method is the most promising one. , The goal of the KA concept is to restore the prearthritic alignment of a patient. To achieve this goal, surgeons need to have a profound knowledge of the individual native anatomy of the knee and its variability.
This chapter provides a basis for all alignment methods and a detailed overview of the current knowledge regarding the variability of lower limb alignment. First, a review of the literature will be presented. Following the discussion of the literature, the functional knee phenotype concept is introduced. Based on these phenotypes, the difference between three alignment goals of the most common alignment concepts (mechanical, anatomical, and restricted kinematic) and the native alignment will be presented and thoroughly discussed.
Variability of the native knee alignment
When performing TKA, a knee surgeon should consider the native knee alignment of the patient. Only with an extended knowledge of native knee alignment in all planes (coronal, sagittal, and axial) is the surgeon able to achieve the optimum outcome for the individual patient. The basic principle in knee surgery is to be as anatomical as possible. This principle has been neglected in TKA in terms of alignment for decades and all knees were forced into an alignment not because of better function and anatomy but because of other considerations such as implant durability and implant design, as well as easier and more reliable instrumentation. However, the only constant in anatomy is its variability, as one of the godfathers in knee surgery, Werner Müller, once stated.
What do we know so far?
Numerous studies have investigated the native coronal knee alignment. The reported overall knee alignment, usually represented as the hip-knee-ankle angle (HKA), averaged around 180 degrees. However, several studies have found that the individual coronal alignment is highly variable, and a significant number of patients appears to have either a varus or valgus alignment. Bellemans et al. investigated the incidence of constitutional varus (HKA <177 degrees) in 250 patients (male:female 125:125), aged between 20 and 27 years, on long leg radiographs (LLRs) in a Belgian population. They found that 32.0% of the males and 17.2% of the females had a constitutional varus. Most of the patients (66% of the males and 80% of the females) had a neutral aligned lower limb (177 degrees – HKA – 183 degrees). Only 2% of males and 2.8% of the females had a so-called “constitutional valgus.” Although the ethnicity of the participants has not been described, most of them can be regarded as Caucasians. Shetty et al. investigated the distribution of HKA based on the same criteria as Bellemans et al. in an Asian population and found more knees in varus (34% Indian, 35% Korean). The authors suggested that femoral bowing of the Asian population might be one reason for this shift. Interestingly, Song et al. found contradictory results investigating only Korean females. More patients had a valgus than varus (25% vs. 20%) alignment. Clearly, to date, the influence of ethnicity and its potential interaction with sex on HKA remains unclear.
Even though there is a great interest in the coronal alignment, most previous studies only used conventional radiographs for their measurements and few used two-dimensional magnetic resonance images (MRI) or EOS images. More importantly, a recent systematic review by Moser et al. identified a number of limitations of the existing literature on native coronal alignment. For example, the systematic review found only four studies investigating the orientation of the femoral joint line (usually measured as femoral mechanical angle [FMA]) and all studies only used LLRs. The reported mean values varied considerably (92.1 ± 1.9 degrees to 97.2 ± 2.7 degrees), and no study reported any ranges. The orientation of the tibial joint line (usually measured as tibial mechanical angle [TMA]) has been investigated by 10 studies, but sample sizes, reported mean values, and ranges varied widely (mean values ranged from 84.6 ± 2.5 degrees to 89.6 degrees). Additionally, many of these studies did not distinguish between male and female patients, although sex-related differences are well known. Therefore it seems that in the past, the research was focused on the overall alignment, whereas details about the orientation of the joint lines have been only superficially covered. As in many fields of knee surgery, for the sake of simplicity and reproducibility, more complexity was avoided. However, in the context of TKA, the alignment of the joint lines is very important because they are not only directly accessible during surgery but also define the overall alignment. As shown by Cooke et al., the overall alignment defined by the HKA equals the sum of the FMA, TMA, and the joint line convergence angle (angle between joint surfaces [JLCA]). Fig. 2.1 shows a detailed description of these angles and the equation.
As a consequence of this lack of knowledge, the authors of the present chapter investigated the coronal alignment parameters of a young, nonosteoarthritic population using three-dimensional reconstructed computed tomography images. Their reported mean values support the findings of previous studies, but the reported ranges also demonstrate that the variability of the native anatomy had been underestimated. Mean values and ranges for HKA, FMA, and TMA are shown in Table 2.1 . Based on these findings, it seems clear that a systematic MA or AA method will not fit all knees. In some knees the use of these systematic approaches will result in a major change of the alignment, which cannot be balanced by bone cuts only—extensive ligament balancing techniques need to be executed.
|Mean ± SD (in degrees)||Range (in degrees, varus − valgus)||Mean ± SD (in degrees)||Range (in degrees, varus – valgus)|
|Hip-knee-ankle angle (HKA)||179.2 ± 2.8||172.6–184.9||180.5 ± 2.8||172.9–187.1|
|Femoral mechanical angle (FMA)||93.1 ± 2.1||87.9–100.0||93.8 ± 1.8||90.1–98.1|
|Tibial mechanical angle (TMA)||86.7 ± 2.3||81.3–94.6||88.0 ± 2.4||82.3–94.0|
Looking at this huge mass of data on coronal knee alignment, one has to recognize that the conventional mindset of knee surgeons, which simply differentiates knees in varus, valgus, or neutral alignment, is too short-sighted. However, this myopic approach focusing on coronal limb alignment based on HKA is still the mainstay in TKA. Starting a more differentiated perspective by making subgroups of varus and valgus knees and including the joint lines draws a different picture, as our data show. There is a myriad of different combinations of the FMA and TMA, and the HKA is only a subordinate parameter of the FMA, TMA, and JLCA. It is therefore not sufficient to only consider HKA, and it is a prerequisite to include joint line orientation of the distal femur and proximal tibia, as well as JLCA, accordingly. An overall neutral, varus, or valgus HKA could have the theoretical combinations of the FMA and TMA shown in Table 2.2 .
|Category||Deviation||Name||Mean value (degrees)||Range (degrees)|
|Lim phenotypes |
(Hip-knee-ankle angle; HKA)
|VAR HKA||VAR HKA 9 degrees||171||169.5<HKA<172.5|
|VAR HKA 6 degrees||174||172.5<HKA<175.5|
|VAR HKA 3 degrees||177||175.5<HKA<178.5|
|NEU HKA||NEU HKA 0 degrees||180||178.5<HKA<181.5|
|VAL HKA||VAL HKA 3 degrees||183||181.5<HKA<184.5|
|VAL HKA 6 degrees||186||184.5<HKA<187.5|
|VAL HKA 9 degrees||189||187.5<HKA<190.5|
|Femur phenotypes (Femoral mechanical angle; FMA)||VAR FMA||VAR FMA 6 degrees||87||85.5<FMA<88.5|
|VAR FMA 3 degrees||90||88.5<FMA<91.5|
|NEU FMA||NEU FMA 0 degrees||93||91.5<FMA<94.5|
|VAL FMA||VAL FMA 3 degrees||96||94.5<FMA<97.5|
|VAL FMA 6 degrees||99||97.5<FMA<100.5|
|Tibia phenotypes |
(Tibial mechanical angle; TMA)
|VAR TMA||VAR TMA 6 degrees||81||79.5<TMA<82.5|
|VAR TMA 3 degrees||84||82.5<TMA<85.5|
|NEU TMA||NEU TMA 0 degrees||87||85.5<TMA<88.5|
|VAL TMA||VAL TMA 3 degrees||90||88.5<TMA<91.5|
|VAL TMA 6 degrees||93||91.5<TMA<94.5|
In fact, this might be one reason for conflicting findings in scientific articles. Some papers did not find any difference for mechanically aligned knees versus nonmechanically aligned knees, whereas others did. Clearly, the problem is that when looking at the overall coronal alignments, one fails to find any difference in outcomes with regards to alignment because patients with a distinctly different alignment are grouped together. The mass of data is in or around the mean and conceals the truth in a big foggy data cloud. We are just not able to see it yet.
Based on the aforementioned, it is clear that there is need for a more detailed analysis of coronal alignment. Consequently, we have developed a new classification system for knee alignment based on a native healthy population and have named it the “knee phenotype system.” This system aims first to allow a better understanding of variability in knee alignment and second, in a later stage, to give guidance for an optimal alignment in TKA for each individual knee.
The knee phenotype system
For all alignment parameters (HKA, FMA, TMA), groups with a range of 3 degrees, so-called phenotypes, were defined. The HKA groups were called “limb phenotypes,” the FMA groups “femoral phenotypes,” and the TMA “tibial phenotypes.”
The family of FMA, TMA, and HKA phenotypes are defined by the deviation from the mean value of a random sampling of a population of subjects with nonosteoarthritic knees and covers a range of ±1.5 degrees from this mean (e.g., 180 ± 1.5 degrees). To state it another way, the phenotypes represent 3-degree increments of the angle starting from the rounded overall mean angle (HKA: 180 degrees; FMA: 93 degrees; TMA: 87 degrees). Table 2.1 shows the definition of all these phenotypes.
The nomenclature of the phenotypes is organized as follows: The first part (NEU, VAR, VAL) defines the direction of alignment. The second subscripted part (HKA, FMA, and TMA) states the phenotype group. The last part (0, 3, and 6 degrees) shows the angular deviation from the mean value.
Until now, the different aspects of the coronal alignment have been investigated separately from each other. However, the coronal alignment is defined by the myriad of combinations of all these angles and, to see the bigger picture, it therefore seems important to assess the combination of the different parameters. Thus, in the next step, the authors defined combinations of the three phenotypes, so-called “knee phenotypes” (combination of femoral and tibial phenotype) and functional knee phenotypes (combination of all three).
The alignment of the young, nonosteoarthritic population was phenotyped according to this new system. A total of 18 knee phenotypes were found out of the theoretically possible 25 knee phenotypes (by combining 5 femoral and 5 tibial phenotypes). Some 17 different knee phenotypes were found in the male population, whereas 12 different ones were found in the female population, including 11 mutual phenotypes. Table 2.3 shows the found knee phenotypes and the percentage of patients they represent.