MATERIAL & METHOD

Data Acquisition

The single-center retrospective study was conducted to collect data on the arthrometry of distal femoral and proximal tibial resection surfaces. The dataset consisted of 40 knees from 40 female patients undergoing total knee replacement. The patients underwent CT scans of the entire leg, from the iliac crest to the heel, while in full extension. The original data were stored in Digital Imaging and Communications in Medicine (DICOM) format and subsequently imported into Mimics 25.0 software (Materialise, Leuven, Belgium) for segmentation and creation of 3D models of the knee. Bone cuts were made using the mechanical alignment method, and morphological data of the femoral and tibial components were taken. Measurements were independently performed by four senior surgeons (including…), and the average values were used in the study. All dimensions were recorded in millimeters and rounded to two decimal places. This study was approved by the Ethical Review Board of Hanoi Medical University.

Femoral Measurements

The distal femoral cut plane was positioned perpendicular to the femoral mechanical axis on the coronal plane and perpendicular to the anatomical axis of the lower femoral shaft on the sagittal plane. The thickness of the bone slices was set at 7mm from the most distal point of the femoral condyles. All femoral size measurements were performed on this resected plane.

To determine the femoral anteroposterior (fAP) dimension, two lines parallel to the surgical transepicondylar axis (sTEA) were drawn, tangent to the most anterior and most posterior points of the lateral condyle. The fAP was defined as the length between these two lines. The femoral mediolateral (fML) dimension was determined as the maximum length perpendicular to the sTEA on the resected plane. The femoral aspect ratio (fAR), defined as the ratio of the mediolateral (fML) to anteroposterior (fAP) dimensions of the distal femur, was first introduced by Hitt et al. [1]. This ratio was calculated for each patient and included in the morphological data analysis.

Tibial Measurements

The simulation cut plane for the proximal tibia was positioned perpendicular to the tibial mechanical axis on the coronal plane, with the slope angle on the sagittal plane set at 3°. The thickness of the bone cut was set at 8mm from the lateral plateau. On the simulated tibial cut plane, a rectangle was drawn such that its two horizontal sides were parallel to the sTEA and tangent to the most anterior and most posterior points. The other two sides were tangent to the most lateral and medial points of the tibia. The AP and ML dimensions were determined as the dimensions of the bounding rectangle. Osteophytes surrounding the resected surface were excluded from the measurements. The ratio between tML and tAP was called the tibial aspect ratio (tAR).

Correlation Between Bone Resection Size and Implant Size

We chosed 6 prosthesis systems: Attune (DePuy Synthes, Johnson & Johnson MedTech), Scorpio (Stryker Corporation), Legion (Smith & Nephew), Link (Waldemar Link GmbH & Co. KG), Microport (Microport Orthopedics), UOC – U2 (United Orthopedic Corporation). Then the anteroposterior, mediolateral and aspect ratio of femoral and tibial resection surfaces were compared with the corresponding dimensions of femoral and tibial components (Figure 1).

Figure 1: Illustation how to measure the parameters on (a) simulated 3D resection surfaces of femur and tibia (b) the femoral and tibial components of prosthesis systems

Calculation the percentage of patient coverage by implant sizespopulation coverage by the prosthesis systems

On the femoral side, choosing the implant size with the best fit in the AP dimension of the resection surface was part of the anterior/posterior referencing in the TKR procedure. Inadequate AP fit increased the risk of altering the range of motion and laxity of the knee. After choosing the appropriate AP dimension of the component, the matching ML dimension was determined to achieve similarity to the femur. Mahoney et al. showed that an ML overhang exceeding 3mm could result in significant knee pain [2]. We decided to tolerate an AP to ML mismatch ratio of 1:2, as detailed in the study by Grothues et al. [3]. Based on the principle that optimal AP length fit and avoidance of overhang in ML dimension, Grothues utilized this tolerance ratio for size mismatch in their methodological approach of 1049 femoral CT scan datasets [3]. For instance, if the maximum acceptable AP mismatch is 1.5 mm, the maximum ML difference would be 3 mm. We analyzed allowed mismatch for AP dimension (from ± 0.5mm to ± 3mm) and ML dimension (from ± 1mm to ± 6mm) in 0.1mm increments. To evaluate the percentage of suitable knee implants for each allowed mismatch level, we implemented an algorithm to match implant sizes with each case that simultaneously satisfied both the AP and ML mismatch criteria, and then calculated the percentage of cases that met the criteria.

For the tibial component, the general consensus is to externally rotate and respect the posterior tibial rim to reduce the risk of patellar pain and instability. However, according to Bonnin et al., implant placement planning varies case by case, requiring a compromise between undersizing the medial plateau or overhanging the posterior lateral plateau. Therefore, we assessed the percentage of suitable tibial components with increased levels of AP and ML dimensional mismatch in a 1:1 ratio, increasing from ± 0.5 mm to ± 6 mm.

Statistical analysis

This study compared femoral anteroposterior (AP), femoral mediolateral (ML), tibial anteroposterior (AP), and tibial mediolateral (ML) measurements between patients with implants and those without using descriptive statistics (median, interquartile range [IQR]), and assessed normality with Shapiro-Wilk tests, followed by Wilcoxon rank-sum tests for non-normal data to determine significant differences.

Morphological parameters of Vietnamese female knees
Parameters N = 401
Mediolateral Femur (fML) 63.07 ± 3.05 (mm)
Anteroposterior Femur (fAP) 57.46 ± 3.06 (mm)
Anteroposterior Tibia (tAP) 48.95 ± 3.01 (mm)
Mediolateral Tibia (tML) 67.14 ± 3.49 (mm)
Femoral aspect ratio (fML/fAP) 1.10 ± 0.05 (mm)
Tibial aspect ratio (tML/tAP) 1.37 ± 0.06 (mm)
1 Mean ± SD (mm)

1. Correlation between the femoral mediolateral width (fML) and the femoral anteroposterior length (fAP) among six implants and femoral size

2. Correlation between the femoral aspect ratio (fML/fAP) and the femoral anteroposterior length (fAP) among six implants and tibial size

3. Correlation between the tibial mediolateral width (tML) and the tibial anteroposterior length (tAP) among six implants and tibial size

4. Correlation between the tibial aspect ratio (tML/tAP) and the tibial anteroposterior length (tAP) among six implants and tibial size

Femur combined AP/ML

Distribution of coverage rate across six implants for femoral AP and ML mismatch
(fAP: ±0.5 to ±3mm; fML: ±1 to ±6mm)
Implant ±0.5/1 mm ±1/2 mm ±1.5/3 mm ±2/4 mm ±2.5/5 mm ±3/6 mm
Link 7.5 30.0 65.0 85 92.5 97.5
Microport 17.5 37.5 67.5 85 100.0 100.0
UOC - U2 12.5 40.0 62.5 90 97.5 97.5
Legion 5.0 30.0 52.5 70 90.0 97.5
Attune 10.0 32.5 57.5 85 100.0 100.0
Scorpio 10.0 45.0 67.5 90 97.5 100.0

Percentage of femoral component coverage over the allowed mismatch thresholds across six different implants

Mediolateral (ML) mismatch is twice the anteroposterior (AP) mismatch

Tibia combined AP/2ML

Distribution of coverage rate across six implants for allowed tibial AP/ML mismatch
(tAP: ±0.5 to ±3mm; tML: ±1 to ±6mm)
Implant ±0.5/1 mm ±1/2 mm ±1.5/3 mm ±2/4 mm ±2.5/5 mm ±3/6 mm ±3.5/7 mm ±4/8 mm
Link 2.5 2.5 12.5 20 37.5 62.5 82.5 82.5
Microport 10.0 40.0 77.5 80 95.0 100.0 100.0 100.0
UOC - U2 12.5 17.5 32.5 60 80.0 82.5 90.0 92.5
Legion 17.5 35.0 67.5 80 90.0 95.0 97.5 100.0
Attune 0.0 7.5 17.5 35 52.5 75.0 82.5 90.0
Scorpio 0.0 10.0 17.5 35 55.0 77.5 82.5 90.0

Tibia AP and ML

Distribution of coverage rate across six implants for allowed tibial AP mismatch
(AP: ±0.5 to ±3mm)
Implant ±1 mm ±2 mm ±3 mm ±4 mm ±5 mm ±6 mm
Link 40.0 82.5 100 100.0 100.0 100
Microport 87.5 100.0 100 100.0 100.0 100
UOC - U2 65.0 87.5 100 100.0 100.0 100
Legion 85.0 100.0 100 100.0 100.0 100
Attune 95.0 95.0 95 97.5 97.5 100
Scorpio 65.0 100.0 100 100.0 100.0 100

Distribution of coverage rate across six implants for allowed tibial ML mismatch
(ML: ±1 to ±6mm)
Implant ±0.5 mm ±1 mm ±1.5 mm ±2 mm ±2.5 mm ±3 mm ±3.5 mm ±4 mm
Link 15.0 25.0 35.0 57.5 82.5 95 97.5 100
Microport 35.0 50.0 62.5 82.5 92.5 100 100.0 100
UOC - U2 42.5 75.0 97.5 100.0 100.0 100 100.0 100
Legion 37.5 70.0 82.5 100.0 100.0 100 100.0 100
Attune 35.0 72.5 100.0 100.0 100.0 100 100.0 100
Scorpio 37.5 75.0 97.5 97.5 97.5 100 100.0 100

Tibia combined AP/ML

Distribution of coverage rate across six implants for allowed tibial AP/ML mismatch
tAP: ±0.5 to ±3mm; tML: ±1 to ±6mm
Implant ±1 mm ±2 mm ±3 mm ±4 mm ±5 mm ±6 mm
Link 2.5 7.5 37.5 57.5 85.0 92.5
Microport 27.5 50.0 92.5 97.5 100.0 100.0
UOC - U2 15.0 47.5 70.0 85.0 95.0 97.5
Legion 27.5 65.0 82.5 90.0 97.5 100.0
Attune 0.0 12.5 40.0 77.5 85.0 95.0
Scorpio 7.5 15.0 40.0 65.0 85.0 92.5

Characteristic Group p-value2
Implant1 Patient1
Femoral anteroposterior 62 (55, 68) 58 (55, 60) 0.007
Femoral mediolateral 68 (62, 73) 63 (61, 65) 0.004
Tibial anteroposterior 50.2 (44.8, 56.0) 48.7 (46.8, 51.1) 0.3
Tibial mediolateral 74 (66, 82) 67 (65, 70) 0.001
1 Median (IQR)
2 Wilcoxon rank sum test

The table presents the median and interquartile range (IQR) for four measurements comparing the Implant and Patient groups. The femoral anteroposterior (62 [55, 68] vs. 58 [55, 60], p=0.007), femoral mediolateral (68 [62, 73] vs. 63 [61, 65], p=0.004), and tibial mediolateral (74 [66, 82] vs. 67 [65, 70], p=0.001) measurements were significantly higher in the Implant group, while the tibial anteroposterior measurement (50.2 [44.8, 56.0] vs. 48.7 [46.8, 51.1], p=0.3) showed no significant difference between the two groups.