Keywords
arthroplasty, replacement, knee - knee joint - magnetic resonance imaging - rotation
Introduction
Total knee arthroplasty (TKA) is the gold-standard surgical procedure for advanced
tricompartmental knee osteoarthritis, with satisfactory pain and function improvement.[1] Multiple factors determine the survival and successful functional outcomes of TKA,
including correct component alignment.[1]
[2]
[3]
[4]
[5]
In the axial plane, rotational alignment of the femoral component greatly influences
postoperative knee kinematics, particularly patellar tracking. To define femoral rotational
alignment, several anatomical parameters are discussed, including the transepicondylar
axis (TEE), the trochlear groove line (Whiteside line), the posterior condylar axis
(PCA) or, even, according to the balanced extension gap.[2] In an arthroplasty following the principle of mechanical alignment based on measured
resection, the literature widely accepts the femoral component positioning parallel
to the surgical TEA.[3]
[4]
[5] Therefore, to achieve the final parallelism of the TEA component, the standard instrumentation
of commercially-available prostheses is typically designed for a section at 3° of
external rotation regarding the PCA, since the femoral condylar posterior surface
supports the guide.
However, the bone morphology of the distal third of the femur may differ in populations
or ethnic groups, as previously reported in Eastern populations.[5]
[6]
[7]
[8] With this context in mind, it is possible that the universal application of 3° of
external rotation regarding the PCA is not adequate for all populational groups.
The main objective of the present study was to evaluate the angular relationship between
the femoral TEA and femoral PCA in a Brazilian population to identify whether the
empirical application of 3° of external rotation of the femoral component is appropriate
for this population.
Materials and Methods
Sample and eligibility criteria
We retrospectively analyzed magnetic resonance imaging (MRI) scans from the Radiology
Department of a quaternary hospital after approval by the Institutional Ethics Committee.
The study included previously-collected images of patients older than 60 years of
age (a clinically-relevant age range for TKA indication) complaining of knee pain
and without osteoarthritis greater than grade 2 on the Kellgren-Lawrence classification
or intra- or extra-articular deformities. We excluded patients with a history of lower
limb surgery, fractures, or pre-existing systemic inflammatory disease.
Sample size determination considered a 5% significance level. The correlation between
the analyzed variables was deemed moderate, and the statistical test power was of
80%. Thus, the minimum number of patients was of 55 for each sex, considering the
significant anthropometric differences between male and female subjects.
Radiological analysis
We analyzed axial T2-weighted MRI scans in a single section containing the topographies
of the medial and lateral epicondyles and posterior femoral condyles. The TEA definition
relied on the apices of the medial and lateral epicondylar prominences, and the PCA
definition relied on a line tangent to the most posterior extension of the femoral
condyles. The primary study outcome was the angular relationships between the TEA
and the PCA. To identify the potential influence of the posterior condylar cartilage
on PCA measurement, we used four methods with different anatomical landmarks:
-
Measurement A – defided by the most posterior chondral prominences of the posterior
medial and lateral femoral condyles ([Fig. 1A]).
-
Measurement B – defided by the most posterior bony prominences of the posterior medial
and lateral femoral condyles ([Fig. 1B]).
-
Measurement C – defided by the medial femoral condylar chondral prominence and lateral
femoral condylar bony prominence ([Fig. 1C]).
-
Measurement D – defided by the medial femoral condylar bony prominence and lateral
femoral condylar chondral prominence ([Fig. 1D]).
Fig. 1 Measurement of the posterior condylar axis (PCA) according to several anatomical
references. (A) PCA defined by the most posterior chondral prominences of the medial and lateral
femoral condyles; (B) PCA defined by the most posterior bony prominences of the medial and lateral femoral
condyles; (C) PCA defined by the medial femoral condylar chondral prominence and lateral femoral
condylar bony prominence; and (D) PCA defined by the medial femoral condylar bony prominence and lateral femoral condylar
chondral prominence.
The same orthopedist performed all angular measurements using a specific software
(Synapse Radiology PACS, Fujifilm, Minato, Tokyo, Japan).
Statistical analysis
To assess whether the angular measurements of external rotation between the TEA and
the PCA presented statistically significant differences from the universal value of
3° of external rotation used in traditional TKA instrumentation, we used the Student's
t-test for a single sample, which is suitable for comparing means between a sample
and an established target value. The significance level was of 5% (p < 0.05).
We performed analysis of variance (ANOVA) to detect whether there was any difference
between the different forms of rotation of the TEA regarding the PCA. When the ANOVA
revealed statistically significant differences, the Tukey test was used for post-hoc
comparisons between the groups to identify actual differences comparing each pair
of measurements (A-B, A-C, A-D, B-C etc.).
Results
The present study included 167 knees, 106 from female and 61 from male patients, with
95 left and 72 right knees. The mean age of the final sample was of 67.57 ± 6.22 (range|:
60–85) years. [Figs. 2]
[3]
[4] show the distribution of the measurements for the total sample, and for the male
and female patients respectively.
Fig. 2 Graph of the distribution of the rotation angles of the transepicondylar axis regarding
each posterior condylar axis measurement (A, B, C, and D).
Fig. 3 Graph of the distribution of the rotation angles of the transepicondylar axis regarding
each posterior condylar axis measurement (A, B, C, and D) in male patients.
Fig. 4 Graph of the distribution of the rotation angles of the transepicondylar axis regarding
each posterior condylar axis measurement (A, B, C, and D) in female patients.
The mean angular values of external rotation of the TEA regarding the PCA, according
to measurements A, B, C, and D, were of 5.44°, 4.94°, 8.57°, and 2.33° respectively.
These values, in all forms of rotation measurement (A, B, C, D), presented statistically
significant differences (p < 0.0001) from the universal value of 3° of external rotation used in the standard
TKA instrumentation ([Table 1]). There was no statistically significant difference between male and female patients
in any measurement (p > 0.05).
Table 1
|
TEA-PCA external rotation: mean ± standard deviation
|
p versus universal value of 3° of external rotation
|
|
Measurement A
|
|
|
|
Male sex
|
5.590° ± 2.404°
|
|
|
Female sex
|
5.358° ± 2.391°
|
|
|
Total
|
5.443° ± 2.384°
|
p < 0.0001
|
|
Measurement B
|
|
|
|
Male sex
|
5.115° ± 2.119°
|
|
|
Female sex
|
4.846° ± 2.051°
|
|
|
Total
|
4.946° ± 2.097°
|
p < 0.0001
|
|
Measurement C
|
|
|
|
Male sex
|
8.334° ± 2.190°
|
|
|
Female sex
|
8.698° ± 2.132°
|
|
|
Total
|
8.347° ± 2.273°
|
p < 0.0001
|
|
Measurement D
|
|
|
|
Male sex
|
2.639° ± 1.684°
|
|
|
Female sex
|
2.263° ± 2.213°
|
|
|
Total
|
2.335° ± 2.002°
|
p < 0.0001
|
Among the different measurement methods, based on combinations of bone or chondral
limits of the condyles, ANOVA revealed statistically significant differences between
the groups (p < 0.0001), with a large effect size (f = 1.02). Next, we compared each group using
the Tukey's (Honest Significant Difference, HSD) test. This test detected statistically
significant differences (p < 0.0001) for all measurement pairs (A-C, A-D, B-C, B-D, C-D), except for A and B
(p = 0.161; [Table 2]).
Table 2
|
Comparison
|
Difference
|
Standard error
|
Confidence interval
|
p-value
|
|
A-B
|
0.497
|
0.1691
|
−0.119–1.13
|
0.1614
|
|
A-C
|
3.1257
|
0.1691
|
2.5097–3.7418
|
< 0.0001
|
|
A-D
|
3.1078
|
0.1691
|
2.4918–3.7238
|
< 0.0001
|
|
B-C
|
3.6228
|
0.1691
|
3.0067–4.2388
|
< 0.0001
|
|
B-D
|
2.6108
|
0.1691
|
1.9947–3.2268
|
< 0.0001
|
Discussion
The most significant finding of the present study is the external rotation profile
of the TEA regarding PCA specifically for a Brazilian population, which presents a
significant difference to the universal value of 3°. Our results show that using the
parameters previously established in the literature, such as 3° of external rotation
for the axial positioning of the femoral component, may not be an accurate method
for the Brazilian population.
In a pivotal study, Berger et al.[7] demonstrated, in the axial plane, the requirement of approximately 3° of external
rotation of the femoral component regarding the plane of the posterior condyles as
a landmark. However, the results from this American population do not necessarily
reflect the reality of populations with different ethnic compositions. Using similar
measurements, Murgier et al.[8] described angles of 6.4° of external rotation in the Asian population. In another
study, Pun et al.[9] found 4.6° of rotation for the Indian population. Therefore, given the discrepancy
in values presented in the literature according to the ethnic group studied, the surgeon
should question whether using anatomical references based on other populations may
interfere with the planning for their patients.
In the present study, when the reference for the measurements was the intact chondral
and bicondylar bone structures (groups A and B), there were no significant differences
between the mean angles of 5.44° and 4.94°. Given the conditions of equal posterior
condylar wear, the chance of error in rotational positioning is lower if the surgeon
applies an external rotation greater than that traditionally stipulated in the cutting
guides, using an approximate value of 5° of external rotation empirically. The results
of a previous study by Loures et al.[10] corroborate the need for higher external rotation of the femoral component in the
Brazilian population, for they identified a mean value of 6.89°. However, the authors[10] analyzed the PCA with a single measurement.
Specifically regarding the rotational positioning of the femoral component using the
PCA and considering the anatomical parameters of the lateral and medial chondral bony
prominence of the posterior femoral condyles (group C), we found relevant results
for the difference in mean measurements of 8.56°. Thus, one must pay attention to
the chondral profile of the posterior femoral condyles for rotational planning of
the femoral component when using the condylar support guide. It is worth noting that
the findings of the present study may be valuable in lateral posterior chondral wear,
which was more prevalent in knees with valgus overload and resembled measurement C
in the current study. In this scenario, the approximate mean divergence would be an
additional 5° of external rotation regarding the universal empirical value of 3°.
This fact would result in the positioning of the femoral component in internal rotation,
which would contribute to an increase in lateral patellofemoral conflict and increase
retinacular tension and patellar tilt, conditions that favor post-arthroplasty patellofemoral
syndrome.[11]
[12]
[13]
[14] The resulting non-physiological knee kinematics entails the potential for worse
clinical outcomes and postoperative anterior pain.[3]
[4]
[5]
Nam et al.[15] showed that the average thickness of the lateral posterior femoral condyle cartilage
is greater than the cartilage thickness of the medial posterior femoral condyle, with
an average difference of 0.4 mm. This discrepancy may justify the significant difference
in the angular values between the different forms of PCA measurements. In addition,
the inequality in the chondral thickness of the femoral condyles in knee osteoarthritis
may vary according to limb axis deviations and contribute as a causal factor for the
error in determining the PCA.[16] Tashiro et al.[17] and Yang et al.[18], when comparing the chondral thickness in the posterior aspect of the lateral and
medial condyles, demonstrated that a thinning of just 2 mm of the cartilage in the
lateral condyle can generate divergences of approximately 2° in the rotational alignment
of the femoral implant.
The results of the present study provide additional support for the recommendation
to use more than one method to measure the rotational positioning of the femoral component
to reduce potential alignment errors due to anatomical variations, in addition to
potential interobserver differences when defining the anatomical parameters used as
landmarks.[19]
[20]
The current study has some limitations. Limbs with valgus may present hypoplasia of
the lateral femoral condyle, and we did not analyze the mechanical axes of the lower
limbs, introducing a potential bias. Another limitation is that the group of the Brazilian
population studied, in terms of number and age, may not reflect anatomically the entire
Brazilian population due to the great ethnic diversity found in Brazil.
Conclusion
Considering exclusively chondral or exclusively bony posterior references, the TEA
in the Brazilian population studied presents approximately 5° of external rotation
regarding the PCA. Therefore, the empirical application of 3° of external rotation
in the femoral component may be insufficient for adequate rotational alignment in
the Brazilian population, which may be even more pronounced in genu valgum. Thus,
it is critical to consider the need for individualized techniques for optimal rotational
positioning of the femoral component to obtain better accuracy.
Bibliographical Record
Enzo Mameri, Isabelle Kaptzky Ballarini, Maria Beatriz Pinheiro Leonel, Marcio de
Castro Ferreira, Marcus Vinicius Malheiros Luzo, Marcelo Seiji Kubota. Morfologia
do fêmur distal em uma população brasileira: Desafiando o uso universal de 3° de rotação
externa. Rev Bras Ortop (Sao Paulo) 2025; 60: s00441800923.
DOI: 10.1055/s-0044-1800923
