Title: Why is the Tibial Tuberosity an Unreliable Landmark for Rotational Alignment of Tibial Component in Total Knee Arthroplasty?
Author(s): Murat Bozkurt MD 1*, M. Gulbiz Kartal MD 2, Safa Gursoy MD 1, Mustafa Akkaya MD 1, Cetin Isik MD 1, Nurdan Cay MD 3
1 Yildirim Beyazit University, Dept. of Orthopaedics and Traumatology, Ankara, Turkey 2 Istanbul University, Dept. of Radiology, Ankara, Turkey 3 Yildirim Beyazit University, Dept. of Radiology, Ankara, Turkey
* Corresponding Author
Vol 2, Num 2, April 2015
Introduction:The rotational alignment of the tibia is an unresolved issue in knee replacement. Several anatomic landmarks have been used as reference for rotational positioning of the components. Among these, mostly preferred one by the surgeons is the tibial tuberosity. Previous studies have reported that tibial tuberosity is not a reliable reference. The aim of this study is to evaluate variations in tibial tuberosity position and proximal tibial torsion.
Materials and Methods:Thirty-nine dry tibias were scanned in MDCT. Axial slices were taken parallel to the short axis of the bone. To evaluate proximal tibia we have measured (1) the position of the tubercle with respect to tibial spines, (2) degree of rotation between the tibial tuberosity and geometric center of the bone, (3) degree of rotation between the tibial tuberosity and anatomic axis of the bone, (4) degree of torsion of the tibia and proximal tibia.
Results:In 6 out of 39 bones (15.38%), the tibial tubercle was positioned medial to the lateral eminence. The distance between the tubercle and lateral eminence changed between 0.7 – 17 mm (mean = 6.2 ± 5.3 mm). The torsion showed significant variance in each bone.
Conclusion:Tibial tubercle shows great variations in sagittal axis, due to the variations in structure of the bones. Individual planning for each bone may be assessed particularly with CT before the operation.
Keyword:Rotational alignment, Total knee arthroplasty, Tibial component, Tibial tuberosity, Orthopaedics, A.7, A.10
Long term follow-up after total knee arthroplasty (TKA) procedures have shown that patellar instability is a major cause of postoperative pain and functional limitation for which revision surgery may be necessary. Proper rotational alignment of the tibia and femur is one of the major components contributing the outcome.[3,4,5,9,10,15] Previous studies have demonstrated direct correlation of the femoral and tibial component rotation to the severity of the patellofemoral complication suggesting that the femoral and tibial component rotation combined may be the predominant cause of patellofemoral complications in patients with normal axial alignment. In cases where malalignment is present, revision is usually required and less favorable results are obtained.[9,10]
The transepicondylar axis is the widely accepted reference for rotational alignment of femoral component. As for the tibia there is no consensus regarding the anatomic reference to be used. This makes tibial component rotation a challenging issue in TKA procedures. In a study by Uehara et al. 109 knees were reviewed and in 50% of these tibial malrotation was found to be more than 5º. In a review, rotation of tibial component varied by 25º, whereas it only varied by 9º in the femoral component.
In order to determine position of the tibial component, most of the surgeons prefer to use the tibial tuberosity. Among the three points recommended on the tubercle, which are the medial border of the tubercle, medial third of the tubercle and the most prominent point of the tubercle, one third medial of the tubercle has been issued as the most significant. On the other hand, in another study the rotational position of the medial 1/3 of the tibial tuberosity relative to femoral epicondylar line and relative to the center of the ankle was assessed and the range was over 40º and it was conclude that in the treatment of patients medial torsion of the tibial component the angle between tibial tuberosity and the center of the ankle should have been taken into account.
As for the sagittal axis, there have been a few studies. A sagittal axis passing through the second metatarsal and a line perpendicular to the posterior surface through the medial third of the tubercle have been assessed.[16,13] In a study by Cobb et al. the line perpendicular at the mid-point of the line joining the medial and lateral condylar axis was defined as the most reliable one. However, none of these measurements were accepted universally.
Although there are several studies determining the accuracy of the anatomic references, none of the studies have investigated the role of tibial torsion, in variative positioning of the tubercle and sagittal axis. We hypothesized that the reasons that tibial tuberosity is not a reliable landmark in positioning the tibial component during TKA are the anatomic variations in position of the tibial tuberosity and the lateral offset of the tibial tuberosity and torsion in the proximal tibia. The failure in the outcome may be reduced by taking anatomic variations into consideration during arthroplasty.
Materials and Methods
Thirty nine tibias which carry landmarks of required for the measurements of this study were scanned in Toshiba Aquillon 64-row multidetector computerized tomography (MDCT). Axial slices were taken parallel to the short axis of the bone. Scanning parameters were as following: Slice thickness: 1 mm, mA: 40, kV: 120, rotation time: 0.5 sec. While evaluating proximal tibia we have measured (1) the position of the tubercle with respect to the tibial spines, (2) degree of rotation between the tibial tuberosity and geometric center of the bone, (3) degree of rotation between the tibial tuberosity and anatomic axis of the bone, (4) degree of torsion of the proximal tibia. The measurements were taken by the two radiologists (MGK, NC) and were repeated thrice in order to assess intra- and inter-observer variability.
1. In order to determine position of the tubercle with respect to the lateral tibial spine we have taken the most proximal slice where the tibial spines are completely visualized and overlaid it on the slice that passes through the most distal part of the tubercle. A line was drawn joining the two most lateral points on the tubercle and a line perpendicular to this line passing through one third medial of the first line was drawn. The point where the second line intersects the tubercle was determined (Figure 1a) A line was drawn joining the two most posterior points on the condyles (Figure 1b). Two parallel lines were drawn perpendicular to this line. One of them passed through tangential to the lateral side of the lateral tibial spine and the other passed through the one third medial of the tubercle. The distance between the two lines were measured (Figure 1c). For those tubercles which are located medial to the lateral eminence, the distance was shown as (-).
2. As for the angle of rotation of the tubercle with respect to geometric center (ROTGC), the method used by Cobb et al. was modified. The slice where the transverse diameter of the tibial plateau is the widest was taken. On this slice we overlaid the most proximal slice where the tibial spines are completely visualized in order to determine the borders of the lateral and medial condyles. Circles best fitted in the lateral and medial condyles were drawn. Geometric centers of the both condyles were detected (GCc). The mid-point of the line joining the two GCc was determined the geometric center of the bone (GC) (Fig. 2a). We have drawn the line between the two most posterior points of the plateau on the same slice and the line perpendicular to this line was named as the tibial central axis (TCA) (Fig. 2b).
The distal slice was chosen as the most distal one passing through the tubercle. A point on 1/3 medial of the tubercle was determined as explained in figures 1a. Two slices were overlaid and the angle between the line passing through the one third medial of the tubercle to the geometric center and the tibial axis was determined as rotgc (Figure 2c).
3. Thirdly, angle of rotation of tubercle with respect to anatomical axis (AA) was determined according to the method by Han et al.  We have determined the AA of the tibia by taking the slices 7 cm below the tibial plateau and 7 cm above the plafond. On axial slices, the centers were determined as the center of the circles that fitted the best on the slices (Figure 3a, 3b). A sagittal line was drawn connecting the two centers on the selected slices. We have again determined TCA as defined previously. The distal slice was chosen as the most distal one passing through the tubercle. Two slices were overlaid and the angle between line passing through the one third medial of the tubercle and geometric center and tibial axis. The angle determined was determined as the rotaa (Figure 3c).
4. Tibial torsion is assessed by taking the difference between the distal tibial axis and proximal tibial axis. The former is determined by Jend method  and the latter one is determined by a tangential line drawn posterior to the tibial plateau. Proximal tibial torsion is determined by the angle between the lines drawn through the two most posterior points on the slice where the transverse diameter of the plateau is the widest and the slice through the most distal part of the tubercle. Beside this, graphics of tibial torsion were obtained. Nine slices were taken (the most proximal slice through the plateau, the slice where the transverse diameter of the plateau is the widest, the slice distal to the plateau, the slice through the most proximal part of the tubercle, the slice where apex of the tubercle is the most prominent, the slice through the most distal part of the tubercle, the slice passing through 70 cm below the plateau, slice passing through 70 cm above the plafond, the slice passing through the plafond where fibular incisura is seen. All the lines drawn were overlayed on the last slice (Figure 4a). Lines were drawn between the two most posterior points on the slices. The angles between these lines and x axis were measured. A graphic was obtained for each individual bone where x axis shows the position and y axis the angles (Figure 4b).
Statistical analyses were performed by using computer software (SPSS for Windows V 15.0; SPSS, Chicago). Intra-observer variability was assessed with ICC test. Two groups (right/left) were compared with Student's and Man-Whitney tests. Accuracy of the angle and distance measurements was assessed with Levene's test for equality of variances. Torsion analysis within two different slices was performed with Wilcoxon test using Bonforoni correction. Pearson and Spearman tests were used in order to evaluate any association between the rotational angle measurements and distance between the tubercle to the geometric center and between torsional differences and the two described measurements.
Thirty nine dry tibia bones were evaluated. Twenty four (60.5%) of the 39 tibias were rightsided and 15 (39.5%) were left-sided. In 6 of the 39 bones (15.38%), tubercle was positioned medial to the lateral eminence. The distance between the tubercle and lateral eminence changed between 0.7 – 17 mm (mean = 6.2 ± 5.3 mm). Rotgc varied between 18.12º - 41.72º (mean = 29.41º ± 6.86º). ROTAA varied between 8.14º - 47.51º (mean = 30.96º ± 9.07º). The intra-observer reliability for the rotgc, rotaa and distance measurements were high with an intra-observer correlation coefficient of 0.95 where the correlation interval was 0.91 - 0.97 (p<0.001), which indicated absolute agreement.
Equality of variance test showed that the rotgc values did not vary significantly, whereas rota did (F = 1.326, p = 0.257 and F = 2.082, p = 0.158, respectively). Distance did not correlate neither with rotgc nor with rotaa (r2= 0.071, p = 0.100).
Torsion of the proximal tibia was defined as the difference between the angles in slice 2 and slice 6. The torsion of the bones and proximal part of the bones showed significant variance (F = 9.539, p = 0.04) and the former ranged between 13.17º and 41.23º, while the latter ranged between -17.83º and 23.59º (mean = 0.67º ± 11.12º). There was no correlation between torsion and proximal torsion of the bones (r2= 0.040, p = 0.220). However, in all the bones torsion in proximal part of the bones was less than overall torsion showing less internal rotation in the proximal part of the bone. The torsion was not correlated with rotgc and rotaa, and distance measurements (r2= 0.021, p = 0.384; r2= 0.012, p = 0.485; r2 = 0.009, p = 0.561, respectively). The difference between the first slice through the tubercle (slice 4) and the last slice through the tubercle (slice 6) was moderately associated with distance and rotgc (r = -0.466, p = 0.003 and 0.400, p = 0.013, respectively).
The differences in between two consecutive slices were evaluated. There was significant difference between the slice 1 and 2 (p = 0.08), 2 and 3 (p<0.01), 3 and 4 (p =0.04). There were no differences in terms of torsion between the slices 4 and 5 and the slices 5 and 6. Differences in slices with respect to the first slice were assessed and it was concludedthat torsional difference between slice 3 and 1 increased between slice 2 and 1. Between slice 4 and 1 the difference was less than the difference between slice 3 and 1. As for the slices 4, 5 and 6, the difference relative to the slice 1 was constant. The direction of the torsion did not change all through the selected slices.
Rotational malalignment is an important factor in failure of success in TKA. In a study by Bedard et al. the tibia was found to be affected in 33 of 34 cases with an average internal rotation of 13.7º. Berger et al. have concluded that the effect of the tibial and femoral components on patellar maltracking were exerted individually and additively. Although transepicondylar axis is the accepted reference for the femoral component, there is no consensus on the reference to be used for tibial component. This makes the rotational positioning of the tibial component as one of the major drawbacks of the rotational alignment. In recent studies, rotation of tibial component has been shown to have variation in a wider range than femoral component does.[2,3,5,16]
Most of the studies evaluating internal rotation of the tibial component have mostly focused on the tibial tuberosity. Bindeglass has studied tibial component rotation by taking one third medial of the tubercle as the reference. Later Lützner et al. have demonstrated that rotational alignment of the tibial component in TKA was better at the medial third of the tibial tuberosity than at the medial border. Akagi et al. have determined the AP axis as the line connecting the middle of the PCL and the medial border of the patellar tendon.
However, there are only a few studies determining sagittal axis of the tibia.[7,8] Cobb et al. described a sagittal axis perpendicular to the mid-point of the line joining centers of the medial and lateral condyles. This line was found to be more reliable than posterior condylar line. In another study by Han et al. anatomic references for tibial sagittal alignment were evaluated. However, this study was focused on relationship between sagittal mechanical axes of tibia and medial and lateral slopes of the tibia.
In the current study, we aimed to determine the factors related to the tibia leading to variations in results when the tibial tuberosity is taken as the reference. We have evaluated dry bones to be able to exclude other factors that may have impact on variation of the results, such as other osseous structures and soft tissues.
In this study, geometric center was used in order to determine structural differences and the anatomic axis was used in order to demonstrate the clinical presentation since the anatomic axis is usually used during surgery. There was a great variation within the results obtained using anatomic axis of the tibia, whereas the measurements taken with respect to geometric axis did not vary significantly. This may be because in the method with the geometric center asymmetry of the condyles were taken into consideration. Besides in the measurements with anatomic axis, the reference points used to obtain the sagittal axis were selected on the slices passing through diaphysis. However, as shown in this study in proximal part of the tibia starting from the plateau to the distal part of the tubercle, the bone demonstrates torsion, which varies significantly on each bone.
When the differences in torsion in each slice with respect to the first slice which passes through the most proximal part of the tibial plateau was assessed, it was concluded that torsional difference between slices pass through and distal to the plateau and first slice varied. However, for the slices difference relative to the slice 1 was constant. Interestingly, although degree of the torsion for the whole proximal tibia did not correlate with rotgc and rotaa, and distance measurements, the difference between the first slice through the tubercle (slice 4) and the last slice through the tubercle (slice 6) was moderately associated with distance and ROTGC.
Theoretically, tibial component should be positioned on the line where sagittal anatomical axis intersects the articular surface of the tibia. Various methods have been defined in order to determine this line. However, discrepancies exist in between these methods. The mostly used method by the surgeons is using an external guide to join 1/3 medial of the tubercle with the mid-point of the 2nd metatarsal and ankle. On the other hand, it has been shown in several studies that tibial tuberosity is not a reliable landmark. In this study, we have evaluated tubercle position with respect to the sagittal anatomical axis and demonstrated the variations in this position. Besides this, we have determined the variations in torsion in the proximal tibia. We believe that variations in position of the tubercle and torsion in the proximal part of the tibia may explain why the tubercle is not a reliable marker.
One of the limitations in this study was the limited sample size. The other was that the dry bones we have evaluated do not necessarily correspond to an osteoarthritic tibia. We acknowledge the fact that in most of the patients undergoing TKA, landmarks on the bones cannot always be accurately determined due to severe osteoarthritic changes. On the other hand, since our aim was to investigate rotational variations, using non-deformed dry bones helped to avoid any variations that might have been attributed to degenerative changes.
Variations in osteoarthritic knee may be a subject to other studies. We believe the data obtained in this study which extends and expands previous studies on variations of tibial tuberosity may be useful in planning TKA procedures.
The tibial tuberosity shows great variations in sagittal axis, due to the variations in structure of the bones. Individual planning for each bone may be assessed particularly with CT before the operation.
Murat Bozkurt Yildirim Beyazit University, Faculty of Medicine, Dept. of Orthopaedics and Traumatology, Ankara, Turkey Corresponding Author firstname.lastname@example.org
M. Gulbiz Kartal Istanbul University, Faculty of Medicine, Dept. of Radiology, Ankara, Turkey
Safa Gursoy Yildirim Beyazit University, Faculty of Medicine, Dept. of Orthopaedics and Traumatology, Ankara, Turkey
Mustafa Akkaya Yildirim Beyazit University, Faculty of Medicine, Dept. of Orthopaedics and Traumatology, Ankara, Turkey
Cetin Isik Yildirim Beyazit University, Faculty of Medicine, Dept. of Orthopaedics and Traumatology, Ankara, Turkey
Nurdan Cay Yildirim Beyazit University, Faculty of Medicine, Dept. of Radiology, Ankara, Turkey
Acknowledgements: None declared.
Financial disclosure: None declared.
1. Akagi M, Oh M, Nonaka T, Tsujimoto H, Asano T, Hamanishi C. An anteroposterior axis of the tibia for total knee arthroplasty. Clin Orthop. 2004;420:213-19.
2. Asano T, Akagi M, Koike K, Nakamura T. In vivo three-dimensional patellar tracking on the femur. Clin Orthop Relat Res. 2003;413:222-32.
3. Barrack RL, Schrader T, Bertot AJ, Wolfe MW, Myers L. Component rotation and anterior knee pain after total knee arthroplasty. Clin Orthop Relat Res. 2001;392:46-55.
4. Bédard M, Vince KG, Redfern J, Collen SR. Internal rotation of tibial component is frequent in stiff total knee arthroplasty. Clin Orthop Relat Res. 2011;469(8): 2346-55.
5. Berger RA, Crossett LS, Jacobs JJ, Rubash HE. Malrotation causing patellofemoral complications after total knee arthroplasty. Clin Orthop Relat Res.1998;356:144-53.
6. Bindeglass DF. Rotational alignment of the tibial component in total knee arthroplasty. Orthopedics. 2001;24(11):1049-51.
7. Cobb JP, Dixon H, Dandachli W, Iranpour F. The anatomical tibial axis- reliable rotational orientation knee replacement. J Bone Joint Surg Br. 2001;90(8):1032-8.
8. Han HS, Chang CB, Seong SC, Lee S, Lee MC. Evaluation of anatomic references for tibial sagittal alignment in total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc. 2008;16(4):373-7.
9. Hofmann S, Romero J, Roth-Schiffl E, Albrecht T. Rotational malalignment of the components may cause chronic pain or early failure in TKA. Orthopade. 2003;32(6):469-76.
10. Incavo SJ, Wild JJ, Coughlin KM, Beynnon BD. Early revision for component malrotation in total knee arthroplasty. Clin Orthop Rel Res. 2007;458:131-6.
11. Jend HH, Heller M, Dallek M, Schoettle H. Measurement of tibial torsion by computer tomography. Acta Radiol Diagn (Stockh). 1989;22(3A):271-6.
12. Lützner J, Krummenauer F, Günther KP, Kirschner S. Rotational alignment of the tibial component in total knee arthroplasty is better at the medial third of the tibial tuberosity than at the medial border. BMC Musculoskelet Disord. 2010;25;11:57.
13. Mizu-uchi H, Matsuda S, Miura H, Higaki H, Okazaki K, Iwamoto Y. The effect of ankle rotation on cutting of the tibia in total knee arthroplasty. J Bone Joint Surg Am. 2006;88(12):2632-6.
14. Nagamine R, Miyanishi K, Miura H, Urabe K, Matsuda S, Iwamoto Y. Medial torsion of the tibia in Japanese patients with osteoarthritis of the knee. Clin Orthop Relat Res. 2003;408:218-24.
15. Sahin N, Atıcı T, Öztürk A, Özkaya G, Özkan Y, Avcu B. Accuracy of anatomical references used for rotational malalignment. Knee Surg Sports Traumatol Arhtrosc. 2012;20(3):565-570.
16. Uehara K, Kadoya Y, Kobayashi A, Ohashi H, Yamano Y. Bone anatomy and rotational alignment in total knee arthroplasty. Clin Orthop Relat Res. 2002;402:196-201.