Preoperative planning of total knee arthroplasty: reliability of axial alignment using a three-dimensional planning approach

Abstract

Background

Preoperative templating of total knee arthroplasty (TKA) can nowadays be performed three-dimensionally with software solutions using computed tomography (CT) datasets. Currently there is no consensus concerning the axial orientation of TKA components in three-dimensional (3D) planning.

Purpose

To assess intra-/inter-observer reliability of detection of different bony landmarks in planning axial component alignment using axial CT images and 3D reconstructions.

Material and Methods

Intra- and inter-observer reliability of determination of four predefined axial femoral and tibial axes was calculated using data from CT scans. Axes determination was performed on the axial slices and on the 3D reconstruction using preoperative planning software. In summary, 61 datasets were analyzed by one medical student (intra-observer reliability) and 15 datasets were analyzed by four different observers independently (inter-observer reliability).

Results

For the femur, clinical epicondylar axis and posterior condylar axis showed the best reliability with an inter-observer variability of 0.7° and 0.5°, respectively. For the tibia, posterior condylar axis provided best reliability (inter-observer variability: 1.7°). Overall variability was greater for tibial than for femoral axes. Reliability of axis determination was more accurate using axial CT slices rather than 3D reconstructions.

Conclusion

The femoral clinical epicondylar axis is highly reliable. Landmarks for the tibia are not as easily identifiable as for the femur. The tibial posterior condylar axis presents the axis with highest reliability. Based on these results, clinical epicondylar axis for orientation of the femoral TKA component and posterior condylar axis for the tibial implant, both defined on axial slices can be recommended.

Keywords

Total knee arthroplasty reliability preoperative planning axial orientation

Introduction

Approximately 10% of men and 18% of women aged over 60 years suffer from osteoarthritis (OA) (1). The final therapeutic treatment option of OA is surgical resurfacing by total knee arthroplasty (TKA) to restore joint functionality and to reduce pain. Templating is an essential element in the preoperative planning process for TKA for the following reasons: (i) it supports the surgeon in preparing the surgical procedure; and (ii) it should be done in order to achieve the greatest benefit from it (2). Classically, preoperative planning of TKA is performed using standard X-rays in anteroposterior, mediolateral view (3) and long-leg radiographs (4) in combination with manufacturers’ templates of implant components on printed X-ray films. Software-based templating can be done two-dimensionally (2D) using the abovementioned standard X-rays in combination with a calibration device. However, a three-dimensional (3D) planning approach could prevent known problems, e.g. rotational or projection errors within 2D radiographs (5,6).

Several factors can cause inaccuracy in the prediction of the correct sizing of the implant. Using 2D X-rays, limb rotation and contractures at the time of radiographic image acquisition will alter the measurement of alignment and affect the outcome of a well-aligned TKA (6). 3D imaging could prevent the aforementioned problems as it accurately represents the actual intraoperative situation (5). Especially in cases of revision or complex anatomical situations, computed tomography (CT) allows a more detailed evaluation of bony landmarks and supports the surgeon’s orientation, even if common landmarks are absent.

In contrast to conventional 2D radiographs, the axial orientation of the prostheses can additionally be aligned on CT scans besides anteroposterior and mediolateral orientations. The literature confirms that axial malrotation leads to pain, altered kinematics, and early failure of the prosthesis (7). However, to the authors’ knowledge, no consensus currently exists concerning the axial orientation of the components for TKA in the 3D planning, especially for the tibial component (8).

The authors hypothesize that: (i) there is a significant difference in intra-/inter-observer reliability for the detection of different landmarks and axes of the distal femur and the proximal tibia; (ii) planning axial component alignment using sliced CT data delivers better intra-/inter-observer reliability compared to a 3D reconstruction approach; and (iii) slice thickness of the CT data will influence reliability – slice thicknesses ≤2 mm provide a better reliability than a slice thickness >2 mm.

Knowledge of the clinical reproducibility of femoral and tibial component axial orientation could contribute to a joint consensus within the framework of preoperative planning.

Material and Methods

Axial CT scans of knee joints recorded within the period from November 2010 to May 2017 were retrospectively identified by the hospital’s radiological information system (RIS) and picture archiving and communication system (PACS). Inclusion criteria were defined as axial CT data of the distal femur and proximal tibia without artifacts or dislocated bone fractures. The medical indication for CT imaging was not only OA but fracture, tumor, infection, rheumatoid arthritis, and cysts as well.

The present study was approved by the Ethics Committee of the Friedrich-Alexander-University Erlangen-Nürnberg (approval no. 6_17B) during their meeting on 14 February 2017.

A total of 61 CT datasets from 51 different patients were selected with slice thicknesses in the range of 0.4–4.0 mm (cases ≤2 mm, n = 36; >2 mm, n = 25) (Table 1).

Table 1.

Number of available CT datasets classified according to medical indication and slice thickness.

Reliability	Indication for CT	No. of datasets	Slice thickness (mm)
Reliability	Indication for CT	No. of datasets	0.4	0.5	0.6	0.7	0.8	2.0	2.5	3.0	4.0
Intra-observer	Osteoarthritis	6			1	1		2		1	1
	Fractures/Trauma	20		1		2	1	10	1	5
	Tumor/Cysts	19	2	3	1	2	1	1		9
	Infection	1								1
	Pain	2						1		1
	Gout arthritis	3				2		1
	Osteonecrosis	2								1	1
	Synovitis	1								1
	Subchondral edema	1								1
	Rotational analysis	2				1				1
	Fibrous dysplasia	2	2
	No information available	2						1		1
∑		61	4	4	2	8	2	16	1	22	2
Inter-observer	Osteoarthritis	2			1	1
	Fractures/Trauma	6		1			1	2		2
	Tumor/Cysts	4	1			1		1		1
	Pain	1								1
	No information available	2						1		1
∑		15	1	1	1	2	1	4		5

CT, computed tomography.

Intra- and inter-observer reliability of four common axial axes for TKA component alignment were assessed (Table 2) for the femur and tibia, respectively (Fig. 1) (8 –13).

Table 2.

Definition of femoral and tibial axes for axial TKA component alignment and related landmarks according to Fig. 1.

Bone	Axis	Landmark	Source
Femur	CEA	Medial epicondylar prominence Lateral epicondylar prominence	(9)
	SEA	Sulcus of the medial epicondyle Lateral epicondylar prominence	(9)
	PCAF	Most posterior point of the medial condyleMost posterior point of the lateral condyle	(10)
	AP	Center of the intercondylar notchDeepest part of the patellar groove	(11)
Tibia	PCAT	Most posterior point of the medial condyleMost posterior point of the lateral condyle	(12)
	ML	Most medial point on the tibial plateauMost lateral point on the tibial plateau	(8)
	STA	Midpoint between the medial and lateral tibial spineThe most anterior point of the tibial tubercle	(13)
	TTA	The center of the best-fit circle around the edge of the cortex on the medial tibial plateauThe center of the best-fit circle around the edge of the cortex on the lateral tibial plateau	(13)

AP, Whiteside Line; CEA, clinical epicondylar axis; ML, mediolateral axis; PCAF, posterior condylar axis of the femur; PCAT, posterior condylar axis of the tibia; SEA, surgical epicondylar axis; STA, sagittal tubercle axis; TKA, total knee arthroplasty; TTA, transverse tibial axis.

Fig. 1.

Predefined landmarks and axes. (a) Femur: CEA (point A to B), SEA (point C to D), PCAF (point E to F), and Whiteside Line (AP; point G to H); and (b) Tibia: PCAT (point I to J), ML (point K to L), STA (point M to N), and TTA (point O to P). CEA, clinical epicondylar axis; ML, mediolateral axis; PCAF, posterior condylar axis of the femur; PCAT, posterior condylar axis of the tibia; SEA, surgical epicondylar axis; STA, sagittal tubercle axis; TTA, transverse tibial axis.

Analysis protocol

Analyses were performed using 3D planning software (mediCAD®Knee 3D version 1.5.15.5647; mediCAD Hectec GmbH, Altdorf, Germany) on axial CT slices and on corresponding 3D reconstructions.

Four observers independently performed the analyses after sufficient software training by the manufacturer. Three observers were medical students, one observer was an engineer. One observer (skills: medical student) performed the analyses three times on 61 CT datasets to assess intra-observer reliability with an interval of at least one week between the analyses. All four observers determined the axes on 15 different CT datasets on axial slices and on the 3D reconstruction respectively to assess inter-observer reliability. The 15 cases were selected out of the 61 available datasets in order to achieve a mixture of slice thicknesses (resulting medical indication: fracture, OA, tumor, cysts) (Table 1).

The coordinates of each landmark were obtained, and the variability of the points were calculated as described by Victor et al. (13).

The landmark coordinate calculation for intra-observer reliability was as follows:

\bar{X} = \frac{x_{a} + {x^{'}}_{a} = {x^{″}}_{a}}{no . repeated analysis}

\bar{Y} = \frac{y_{a} + {y^{'}}_{a} = {y^{″}}_{a}}{no . repeated analysis}

\bar{Z} = \frac{z_{a} + {z^{'}}_{a} = {z^{″}}_{a}}{no . repeated analysis}

The landmark coordinate calculation for inter-observer reliability was as follows:

\bar{X} = \frac{x_{a} + x_{b} + x_{c} + x_{d}}{no . of observers}

\bar{Y} = \frac{y_{a} + y_{b} + y_{c} + y_{d}}{no . of observers}

\bar{Z} = \frac{z_{a} + z_{b} + z_{c} + z_{d}}{no . of observers}

A virtual midpoint between all determined points for every single landmark was created and the distances to the individual points were calculated.

Each axis was generated as a vector linking the two landmarks. The vector was projected onto the axial/transverse plane (Fig. 2). The angular deviation was measured by comparing the vector to the x-axis given by the CT dataset:

angular deviation = \arctan 2 (x_{a} - x_{b}; y_{a} - y_{b})

Fig. 2.

Schematical representation of PCAF determination (landmark E and F; resulting Line₁) on axial slices on different heights and consequent computation of the angles against the x-axis given (Line₂) by the CT dataset. CT, computed tomography; PCAF, posterior condylar axis.

Standard deviations for each CT dataset were calculated.

Since CT datasets with different slice thicknesses were analyzed, results from CT datasets with high resolution (axial slice thickness ≤2 mm) and low resolution (axial slice thickness >2 mm) could be compared (Fig. 3). The threshold was set at 2 mm, as the software manufacturer proposed CT recordings with a slice thickness of 2 mm or less.

Fig. 3.

CT data quality. (a, b) 3D reconstruction and corresponding (c, d) axial slice image of (a, c) 0.4 mm and (b, d) 4.0 mm slice thickness, respectively. CT, computed tomography.

Statistics

Mean and standard deviations were calculated to determine the variability of distances of the landmarks. Median, quartiles, and ranges were calculated to determine the angular deviation of each axis and dataset.

The Wilcoxon signed rank test was used to compare the bony landmark detection (in axial slices and 3D reconstruction), the angular deviation of the axes, and the variability between the different axes. The Mann–Whitney U test was used to compare the two different groups of slice thicknesses. The intraclass correlation coefficient (ICC; two-way mixed effects, consistency, single measurement for intra-observer and average measurement for inter-observer reliability) was used in accordance with Koo et al. (14) to calculate reliability of the determined angular variability. For intra-/inter-observer reliability, ICC values were interpreted as follows: > 0.75 = excellent reliability; 0.40–0.75 = fair to good reliability; and <0.40 = poor reliability (15). Statistical analysis was performed using SPSS version 24 (SPSS Inc., Chicago, IL, USA). The level of significance was set to P = 0.05.

Results

In general, variability for tibial parameters (maximum 6.5 mm and 9.3° axial) was higher than for the femur (maximum 3.1 mm and 4.6° axial) (Figs. 4 and 5).

Fig. 4.

Registration variability of anatomical landmarks for the tibia and the femur – distance deviations (in mm) are shown as mean value ± SD for femur and tibia: (a) intra-observer variability femur; (b) intra-observer variability tibia; (c) inter-observer variability femur; and (d) inter-observer variability tibia for axial slices (dark gray) and 3D (gray) planning.

Fig. 5.

Variability of angular deviations (in degrees) for femur and tibia: (a) intra-observer variability femur; (b) intra-observer variability tibia; (c) inter-observer variability femur; and (d) inter-observer variability tibia for axial (dark gray) and 3D (gray) planning.

Intra-observer reliability was significantly higher using axial slices than using 3D reconstruction in 10 out of 16 anatomical landmarks (P < 0.005) (Fig. 4) and in seven out of eight predefined axes (P < 0.009), respectively (Fig. 5). Inter-observer reliability was significantly higher using axial slices in 13 out of 16 anatomical landmarks (P < 0.014) (Fig. 4) and in six out of eight predefined axes (P < 0.032) (Fig. 5).

For reasons of clarity and readability, only angular and distance deviations measured in axial slices are described in the text. Nevertheless, the results derived from 3D reconstructions are also illustrated (Table 3 and Figs. 4 and 5).

Table 3.

Intra- and inter-observer mean variability of landmark detection and angular deviation separated in high (≤2 mm) and low resolution (>2 mm).

	Intra-observer				Inter-observer
	ASL		3D		ASL		3D
Slice thickness (mm)	≤ 2	> 2	≤ 2	> 2	≤ 2	> 2	≤ 2	> 2
CEA (deg)	0.6	0.6	1.0	0.9	0.6	0.9	2.3	1.9
A (mm)	0.9	1.0	1.7	1.3	1.7	1.9	3.0	2.7
B (mm)	1.7	1.7	1.6*	1.0	2.0	2.2	3.7	2.8
SEA (deg)	0.7	0.8	1.5	1.5	1.1	1.4	3.3	3.6
C (mm)	1.2	1.4	1.7	1.8	1.9	2.1	3.8	3.4
D (mm)	1.6	1.6	1.6*	1.2	2.1*	3.5	3.7	3.9
PCAF (deg)	0.3	0.3	0.4	0.6	0.6*	0.3	0.6	0.6
E (mm)	1.5	1.4	2.6*	2.2	2.3	1.8	3.0	3.5
F (mm)	1.8*	1.2	2.3	2.3	2.6	2.0	3.1	3.2
AP (deg)	1.6	1.8	3.6	3.8	1.6	1.9	7.4	6.0
G (mm)	2.5	2.6	2.5*	1.7	2.9	3.3	2.9	2.4
H (mm)	2.1	2.3	2.7	2.9	3.0	3.3	9.3	7.3
PCAT (deg)	1.1	1.0	2.3	3.3	1.9	1.3	4.2	3.8
I (mm)	2.2*	1.6	4.1	4.8	3.8	2.7	4.2*	5.3
J (mm)	2.6	2.1	3.1*	4.0	4.2	3.9	4.7	4.9
ML (deg)	2.5	2.9	4.3	3.8	2.4	2.7	8.4	5.8
K (mm)	2.1	2.4	2.9	2.4	2.8	2.4	5.8	4.6
L (mm)	2.0	2.3	4.4	3.6	2.6	3.0	7.0*	4.2
STA (deg)	1.5	1.9	3.6	3.0	4.3	2.9	6.4	4.3
M (mm)	1.0	1.3	3.0	2.8	2.0*	2.6	4.4	3.7
N (mm)	2.6	2.0	3.2*	1.7	7.6	4.3	7.3	5.2
TTA (deg)	1.5	1.0	1.8	1,8	2.8	3.1	3.2	2.4
O (mm)	1.6*	1.3	1.6	2.0	2.6	2.1	3.4	3.0
P (mm)	1.5	1.3	1.6	1.9	2.3	2.7	3.1	3.3

*Mean value significant difference (P < 0.05).

A, medial epicondylar prominence; AP, Whiteside Line; ASL, axial slices; B, lateral epicondylar prominence; C, sulcus of the medial epicondyle; CEA, clinical epicondylar axis; D, lateral epicondylar prominence; E, most posterior point of the medial condyle; F, most posterior point of the lateral condyle; G, center of the intercondylar notch; H, deepest part of the patellar groove; I, most posterior point of the medial condyle; J, most posterior point of the lateral condyle; ML, mediolateral axis; K, most medial point on the tibial plateau; L, most lateral point on the tibial plateau; M, midpoint between medial und lateral tibial spine; N, most anterior point of the tibial tubercle; O, the center of the best-fit circle around the edge of the cortex on the medial tibial plateau; P, the center of the best-fit circle around the edge of the cortex on the lateral tibial plateau; PCAF, posterior condylar axis of the femur; PCAT, posterior condylar axis of the tibia; SEA, surgical epicondylar axis; STA, sagittal tubercle axis; TTA, transverse tibial axis.

Femur

While the determination of the landmarks showed only marginal variability (intra-observer mean range = 0.9–2.5 mm, inter-observer mean range = 1.8–3.1 mm) (Fig. 4), the determination of the angular deviations indicates that the clinical epicondylar axis and the posterior condylar axis are the most reliable axes for intra-observer as well as for inter-observer reliability (Fig. 5). The clinical epicondylar axis delivered a reliability with a mean angular variability of 0.6° for intra-observer and 0.7° for inter-observer. In comparison, the posterior condylar axis provided higher reliability with 0.3° intra-observer and 0.5° inter-observer variability. The ICC for the clinical epicondylar axis and the posterior condylar axis was 1.000 each. The two axes were significantly more reliable compared to the other axes for the distal femur on axial slices (P < 0.008). Only the intra-observer reliability for the posterior condylar axis showed a significantly better result than the clinical epicondylar axis (P = 0.002). Results for inter-observer reliability were not significantly different (P = 0.390). The mean intra-observer variability of the Whiteside Line and the surgical epicondylar axis were 1.7° and 0.8°, respectively; for inter-observer reliability the Whiteside Line was 1.7° and the surgical epicondylar axis was 1.2°.

Tibia

The posterior condylar axis of the tibia indicated best data variability of 1.0° (intra-observer) and 1.7° (inter-observer). In comparison, the transverse tibial axis variability was marginally higher (intra-observer = 1.6°, inter-observer = 2.9°). Data spreading was higher for the sagittal tubercle axis (maximum = 9.3°) and the mediolateral axis (maximum = 9.3°). The ICC for the posterior condylar axis and the transverse tibial axis was 1.000 for intra- as well as for inter-observer reliability; in contrast, the sagittal tubercle axis (intra-observer = 0.994, inter-observer = 0.987) and the mediolateral axis (intra-observer = 0.999, inter-observer = 1.000) showed a marginally lower ICC.

The posterior condylar axis detection showed the highest reliability, although the determination of the landmarks was not consistently stable (intra-observer mean range of bony landmarks I = 2.0 mm and J = 2.4 mm, inter-observer mean range of bony landmarks I = 3.4 mm and J = 4.1 mm) (Fig. 4). The determined angular deviation for the posterior condylar axis indicated a significant difference compared to other axes on axial slices (P < 0.009) except for inter-observer reliability of the mediolateral axis (P = 0.055).

Slice thickness

There was no significant difference for intra-observer reliability in 13 out of 16 anatomical landmarks (P > 0.065) and for all eight predefined axes (P > 0.203) (Table 3). In addition, the inter-observer reliability did not show statistically significant differences in 14 out of 16 anatomical landmarks (P > 0.062) and in seven out of eight predefined axes (P > 0.154) (Table 3). Furthermore, in four out of six cases the CT dataset with low resolution was significantly more reliable compared to high-resolution CT.

Discussion

The clinical epicondylar axis for the femur and the posterior condylar axis for the tibia are the most reliable axes. The determination based on axial slices is more reliable than on 3D reconstruction. The slice thickness has no influence on the reliability.

The axis used for component alignment must be a true and reliable representation of joint kinematics. The transepicondylar axis has been hypothesized as the rotation center of the knee. However, there are misleading datasets, whether the clinical epicondylar axis or the surgical epicondylar axis correspond more accurately to the optimal knee flexion axis (16). Apparently, the surgical epicondylar axis correlates better with the femorotibial articulation and the clinical epicondylar axis with the patellofemoral articulation (17). Eckhoff et al. (18) proposed a cylindrical axis through the condyles that may fit more accurately to the optimal knee flexion. However, this axial orientation was not investigated in this study while this approach was not realizable with applied planning tool.

The higher detection variability of the surgical epicondylar axis compared to the clinical epicondylar axis within this study was caused by identification difficulties of the medial sulcus. Several studies (19) have described low detection rates of the medial sulcus previously, especially in OA knees. In contrast, the clinical epicondylar axis was always identifiable (20).

Within this study, the posterior condylar axis detection as well as the clinical epicondylar axis detection delivered excellent reliability. However, careful attention must be paid when aligning along the posterior condylar axis. As mentioned before, the transepicondylar axis is hypothesized as the optimal femoral rotational component axis. The relationship between the posterior condylar axis and the transepicondylar axis is extremely variable. Especially in valgus knees, the angle between these two axes is significantly larger, likely because of a smaller lateral condyle. Variability within 5° were described by several authors (21).

The detection of the Whiteside Line showed the lowest reliability with more outliers. One reason might be that this line is a curved and short line and often affected by OA and/or dysplasia (22). Furthermore, it was necessary to identify the intercondylar notch and patella groove, which are often located in different slices. If the Whiteside Line is used for alignment in OA knees, the femoral component is likely to be orientated in significant external rotation (10).

There is no consensus about the orientation of the tibial component for TKA (2). Bonnin et al. (23) postulated two main criteria for the tibial orientation: optimal knee kinematics and patellofemoral tracking as well as optimal tibial coverage of the prosthesis. However, the rotational alignment must be taken into account primarily (24). Compared to the distal femur there is an absence of existing landmarks on the proximal tibia (25). In addition, the tibial component alignment is more difficult because the tibial slope must also be taken into account.

The posterior condylar axis of the tibia showed the highest reliability within this study despite the fact that this axis may be affected by osteophytes (2). The posterior condylar axis also remains a stable parameter for deeper resection depths, e.g. in revision cases. The literature proposes that the tibial component should be orientated with 10° of external rotation to the posterior condylar axis as this orientation corresponds with the transepicondylar axis of the femur (26).

The transverse tibial axis was previously described as the most reliable axis for axial orientation of the tibial TKA component (12,13). However, this could not be confirmed in this study, since variability was higher compared to the posterior condylar axis. A possible reason for the higher variability of this axis compared to literature data might be that the tibial plateau level was chosen for determining the transverse tibial axis in this study. Cobb et al. (12) stated that the tibia condyles are more circular at a resection depth of 8 mm than at the height of the plateau surface. If the tibial TKA component is orientated along the transverse tibial axis, 4° of additional external rotation to this axis is recommended (26).

Low reliability was measured for the sagittal tubercle axis. The tibial tubercle in particular showed high variability, although the anterior point of the tibial tubercle was chosen, which, in the authors’ opinion, is easier to define than the one-third medial or medial border of the tibial tubercle (27).

Determination of landmarks was in general more reliable in axial CT slices than in 3D reconstruction. Hung et al. (28) demonstrated that the determination of the surgical epicondylar axis in axial CT slices is more accurate than in the 3D reconstructed CT images compared to the true surgical epicondylar axis measured during cadaveric dissection, which is in line with the results of the present study.

Lower reliability of landmark detection of this study compared to Victor et al. (13) could be explained by the different software packages applied for computing the 3D bony surface model. In addition, the data from Victor et al. were derived from cadaveric experiments using standardized CT protocols. Within this study, available CT data (retrospectively collected) present patients from the Department of Orthopaedics. The CT data come from several CT scanners from different brands, meaning the scanning protocol is not controlled or standardized. However, the aim of the present study was to assess the reliability of the axes’ determination under different clinical conditions with regard to image acquisition settings (independent from scanning protocol and scanner brands), available slice thicknesses, and analyzing environments.

The medical skills of the observers (medical students, engineer) present an additional constraint. Reliability protocol was chosen with medical students and not clinical professionals, because preoperative templating is often a “learning” tool when starting arthroplasties in residency (29) and as Ettinger et al. (30) demonstrated, the level of experience does not influence the accuracy of templating.

By assessing preoperative CT scans, the observers were able to check the bony situation without interference from the soft tissues such as pathologically altered ligaments or cartilage (3).

The accuracy achieved in preoperative templating must be transferred to the intraoperative situation. However, intraoperative reproducibility imposes a great challenge for surgeons, especially as the used software is designed for conventional TKA templating without later computer-assisted surgery. Jenny and Boeri (31) measured a mean intra-observer variability of 5° and inter-observer variability of 9° by determining the transepicondylar axis intraoperatively. Siston et al. (32) described a small mean error for the tibia but high standard deviations of 12°. Hence, low differences in reliability between the clinical epicondylar axis and the posterior condylar axis for the femur or even the transverse tibial axis and posterior condylar axis for the tibia may not influence the outcome of TKA, which is gained by preoperative templating.

In conclusion, according to postulated hypotheses, the authors can state that: (i) using the clinical epicondylar axis for orientation of the femur and the posterior condylar axis for the orientation of the tibia could be recommended because of advantageous intra- and inter-observer reliabilities; (ii) it could be confirmed that templating with axial CT slices results in a significantly better intra- and inter-observer reliability compared to an approach using 3D reconstruction of CTs; and (iii) it could be rejected that the slice thickness of the CT datasets did not influence reliability of determination of axial knee axes. However, from the data in the present study, it is not possible to conclude that it is unnecessary to perform CT recordings with high resolution (slice thickness <2 mm), as the axial orientation is just one part of preoperative planning and templating.

Footnotes

Acknowledgments

The authors thank Monya Reinmuth and Nico Strödick for processing the axial and 3D reconstruction planning within this study for inter-observer reliability assessment. The present work was performed in fulfilment of the requirements for obtaining the degree “Dr. med.” (ST) at the Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU), Germany.

Declaration of conflicting interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: MediCAD Hectec GmbH (Altdorf, Germany) supported this study by providing the 3D planning software. The company took part neither in the analysis and interpretation of data nor in the preparation of the manuscript. None of the authors had a further conflict of interest that may have biased the study.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Frank Seehaus

References

Glyn-Jones

Palmer

AJR

Agricola

, et al. Osteoarthritis. Lancet 2015; 386:376–387.

Pietsch

Djahani

Hochegger

, et al. Patient-specific total knee arthroplasty: the importance of planning by the surgeon. Knee Surg Sports Traumatol Arthrosc 2013; 21:2220–2226.

Kniesel

Konstantinidis

Hirschmuller

, et al. Digital templating in total knee and hip replacement: an analysis of planning accuracy. Int Orthop 2014; 38:733–739.

Chou

Siu

, et al. Preoperative templating and computer-assisted total knee arthroplasty for arthritic valgus knee. J Arthroplasty 2013; 28:1781–1787.

Okamoto

Mizu-uchi

Okazaki

, et al. Two-dimensional planning can result in internal rotation of the femoral component in total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc 2016; 24:229–235.

Kawakami

Sugano

Yonenobu

, et al. Effects of rotation on measurement of lower limb alignment for knee osteotomy. J Orthop Res 2004; 22:1248–1253.

Kim

Park

Kim

, et al. The relationship between the survival of total knee arthroplasty and postoperative coronal, sagittal and rotational alignment of knee prosthesis. Int Orthop 2014; 38:379–385.

Hutter

Granger

Beal

, et al. Is there a gold standard for TKA tibial component rotational alignment? Clin Orthop Relat Res 2013; 471:1646–1653.

Berger

Rubash

Seel

, et al . Determining the rotational alignment of the femoral component in total knee arthroplasty using the epicondylar axis. Clin Orthop Rel Res 1993; 286:40–47.

10.

Nagamine

Miura

Inoue

, et al. Reliability of the anteroposterior axis and the posterior condylar axis for determining totational alignment of the femoral component in total knee arthroplasty. J Orthop Sci 1998; 3:194–198.

11.

Whiteside

Arima

The anteroposterior axis for femoral rotation alignment in valgus total knee arthroplasty. Clin Orthop Rel Res 1995; 321:168–172.

12.

Cobb

Dixon

Dandachli

, et al. The anatomical tibial axis: reliable rotational orientation in knee replacement. J Bone Joint Surg Br 2008; 90:1032–1038.

13.

Victor

Van Doninck

Labey

, et al. How precise can bony landmarks be determined on a CT scan of the knee? Knee 2009; 16:358–365.

14.

Koo

MY.

A guideline of selecting and reporting intraclass correlation coefficients for reliability research. J Chiropr Med 2016; 15:155–163.

15.

Shrout

Fleiss

JL.

Intraclass correlations: uses in assessing rater reliability. Psychol Bull 1979; 86:420–428.

16.

Kobayashi

Akamatsu

Kumagai

, et al. The surgical epicondylar axis is a consistent reference of the distal femur in the coronal and axial planes. Knee Surg Sports Traumatol Arthrosc 2014; 22:2947–2953.

17.

Paternostre

Schwab

Thienpont

The combined Whiteside’s and posterior condylar line as a reliable reference to describe axial distal femoral anatomy in patient-specific instrument planning. Knee Surg Sports Traumatol Arthrosc 2014; 22:3054–3059.

18.

Eckhoff

Bach

Spitzer

, et al. Three-dimensional mechanics, kinematics, and morphology of the knee viewed in virtual reality. J Bone Joint Surg Am 2005; 87 Suppl 2:71–80.

19.

Uehara

Kadoya

Kobayashi

, et al. Bone anatomy and rotational alignment in total knee arthroplasty. Clin Orthop Rel Res 2002; 402:196–201.

20.

Yoshino

Takai

Ohtsuki

, et al. Computed tomography measurement of the surgical and clinical transepicondylar axis of the distal femur in osteoarthritic knees. J Arthroplasty 2001; 16:493–497.

21.

Iijima

Ohi

Isho

, et al. Association of bilateral flat feet with knee pain and disability in patients with knee osteoarthritis: A cross-sectional study. J Orthop Res 2017; 35:2490–2498.

22.

Talbot

Dimitriou

Radic

, et al. The sulcus line of the trochlear groove is more accurate than Whiteside’s Line in determining femoral component rotation. Knee Surg Sports Traumatol Arthrosc 2015; 23:3306–3316.

23.

Bonnin

Saffarini

Mercier

, et al. Is the anterior tibial tuberosity a reliable rotational landmark for the tibial component in total knee arthroplasty? J Arthroplasty 2011; 26:260–267.

24.

Martin

Saurez

Ismaily

, et al. Maximizing tibial coverage is detrimental to proper rotational alignment. Clin Orthop Relat Res 2014; 472:121–125.

25.

Drexler

Backstein

Studler

, et al. The medial border of the tibial tuberosity as an auxiliary tool for tibial component rotational alignment during total knee arthroplasty (TKA). Knee Surg Sports Traumatol Arthrosc 2017; 25:1736–1742.

26.

Forster-Horvath

Kremo

Muller-Gerbl

, et al. Using the anatomical tibial axis for total knee arthroplasty alignment may lead to an internal rotation error. Int Orthop 2015; 39:2347–2353.

27.

Akagi

Nonaka

, et al . An anteroposterior axis of the tibia for total knee arthroplasty. Clin Orthop 2004; 420:213–219.

28.

Hung

CLW

Pan

Yuen

, et al. Interobserver and intraobserver error in distal femur transepicondylar axis measurement with computed tomography. J Arthroplasty 2009; 24:96–100.

29.

Curtin

Fehring

Lauber

Errors in knee alignment using fixed femoral resection angles. Orthopedics 2014; 37:e644–648.

30.

Ettinger

Claassen

Paes

, et al. 2D versus 3D templating in total knee arthroplasty. Knee 2016; 23:149–151.

31.

Jenny

Boeri

Low reproducibility of the intra-operative measurement of the transepicondylar axis during total knee replacement. Acta Orthop Scand 2004; 75:74–77.

32.

Siston

Goodman

Patel

, et al. The high variability of tibial rotational alignment in total knee arthroplasty. Clin Orthop Relat Res 2006; 452:65–69.