Search tips
Search criteria 


Logo of intorthopspringer.comThis journalToc AlertsOpen ChoiceSubmit Online
Int Orthop. 2009 June; 33(3): 679–685.
Published online 2008 July 11. doi:  10.1007/s00264-008-0562-7
PMCID: PMC2903080

Language: English | French

Comparison of tibial rotation in fixed and mobile bearing total knee arthroplasty using computer navigation


Tibial rotation is an important aspect of knee function and can be altered after total knee arthroplasty (TKA). These alterations include decreased internal rotation with knee flexion as compared to the normal state and paradoxical external rotation with flexion. Mobile bearing total knee prostheses may allow greater unconstrained tibial rotation. I compared tibial rotation after fixed bearing or mobile bearing total knee arthroplasty in 82 patients who underwent TKA with the tibia cut first technique to ascertain any differences. Using intraoperative imageless computer navigation, measurements included the determination of tibial rotation from extension to 90° flexion before and after prosthetic implantation with non-weight-bearing range of motion. I found that tibial rotation was significantly reduced after fixed bearing total knee replacement as compared to mobile bearing. In addition, the tibial position compared to the distal femur in extension was more external in fixed bearings compared to mobile bearings. Placing the fixed tibial tray with increased internal rotation could explain this difference.


La rotation du tibia est un élément important sur le plan fonctionnel. Elle peut être altérée après une prothèse totale du genou. Ces altérations de la rotation portent sur une diminution de la rotation interne et paradoxalement une augmentation de la rotation externe en flexion. Une prothèse totale du genou à plateau mobile permet une libération plus importante de la rotation tibiale. Nous avons comparé cette rotation à partir de prothèses à la plateau fixe et à plateau mobile sur une série de 82 patients ayant bénéficie de cette arthroplastie avec une coupe première du tibia. Nous avons utilisé pour cela un système de navigation per opératoire. Nous avons trouvé à l’issue de cette étude que les rotations tibiales étaient significativement réduites après prothèse totale à plateau fixe versus prothèses à plateau mobile. Par ailleurs, la position du tibia en extension est en rotation externe par rapport au fémur distal. Cette rotation est plus importante dans les prothèses à plateau fixe que dans les prothèses à plateau mobile. Le positionnement du plateau tibial en rotation interne lors de l’intervention peut expliquer cette différence.


Axial femorotibial rotation during flexion of the healthy knee has been seen in numerous previous in vitro and in vivo kinematic analyses [5, 8, 10, 16]. With knee flexion, the tibia typically internally rotates relative to the femur, and conversely, externally rotates with knee extension (i.e. normal screw-home mechanism) [8, 10, 15, 16]. Previous total knee arthroplasty (TKA) studies have been limited and have analysed small numbers of patients, often using non-weight-bearing conditions or only throughout a limited percentage of the entire flexion range [1, 2, 7, 11, 12, 18, 19]. It is assumed that different axial rotation magnitudes and patterns (i.e. direction of rotation, internal or external tibial rotation versus the femur) may occur after TKA because of removal or alteration of the cruciate ligaments and failure to exactly duplicate geometry of the medial and lateral femoral and tibial condyles. Fluoroscopic video kinematic studies have demonstrated that while some cases demonstrate the expected internal rotation with knee flexion, others will actually exhibit paradoxical external rotation with knee flexion [6, 24]. Knowledge of rotational movement is an important consideration for understanding polyethylene wear patterns where exaggerated sliding motion coupled with rotation may produce detrimental delamination wear [4, 17, 25].

Computer-assisted surgical navigation (CAOS) offers a unique opportunity to evaluate tibial rotation intraoperatively along with numerous other parameters such as mechanical alignment, joint flexion, ligament balance and tibial axis alignment in flexion. This report retrospectively reviewed tibial rotation documented before and after mobile and fixed total knee arthroplasty in a group of patients where the “tibia cut first” surgical technique was used with both implants. While the mobile bearing prosthesis tested offers the potential for unconstrained rotational motion, it was believed that the semi-constrained fixed bearing prosthesis has enough geometric freedom to allow similar rotational movement. The hypothesis was that there should be no significant difference between the fixed and mobile bearing tibial insert with regard to rotational movement in total knee arthroplasty. More specifically, a fixed bearing total knee prosthesis should easily accommodate the typical amount of tibial rotation that has been shown to occur by analysing prior kinematic studies assessing arthritic knee rotation.


A group of 82 patients underwent primary total knee arthroplasty using the “tibia cut first” technique performed by a single surgeon experienced with this technique (JBS). Implants used were either the LCS mobile bearing prosthesis or the NexGen LPS Flex prosthesis. The low contact stress mobile prosthesis is a total condylar design that offers very high conformity of the femoral and tibial insert from 0 to 40° of flexion followed by less congruity with deeper flexion resulting from diminished radii of curvature of the posterior femoral condyles. The tibial insert has a central cone that articulates with a matching reverse cone on the tibial tray. Tibial rotation is unconstrained with this device. The NexGen LPS Flex prosthesis is a total condylar design that articulates with semi-constrained fixed tibial insert. The condylar geometry is reasonably conforming but will allow up to 7° of coronal plane “lift-off” and up to 12° of tibial internal or external rotation.

The groups were selected from a consecutive series of navigated cases performed from 2003 to 2005 and were chosen such that the amount of deformity was matched for the two groups. For the mobile bearing cohort, the average amount of preoperative deformity was 7° of varus. There were 25 males and 16 females. The average age was 56 and average body mass index (BMI) was 31. Postoperative, the average deformity was corrected to an average of 0.7° varus. For the fixed bearing group, the average deformity was 6.2° of varus. There were 15 males and 26 females. The average age was 69 and BMI was 31. Postoperative, the deformity was corrected to an average of 0.5° of varus (Table 1). The Medtronic Stealth Treon system with the Universal Imageless Total Knee software (Medtronic, Inc., Louisville, CO, USA) was used in all cases with dynamic reference base markers attached to either the medial proximal tibia or the distal medial femur over the medial epicondyle. The tibial rotation was defined mathematically by the relationship of the transepicondylar axis and a vector measured from the tibial centre to the midpoint prominence of the tibial tubercle. More recently, the Medtronic imageless protocol added the femoral anterior/posterior axis of Whiteside as an additional mark to determine femoral rotation eliminating the need for the transepicondylar axis reference.

Table 1
Demographics of both cohorts

Navigation referencing protocol

The specifics of the navigation referencing are an important element of the technique and require detailed description. Hip centre determination is done using the kinematic method originally described by Saragaglia et al. [20]. Femoral referencing is done with the two most important points being the femoral centre and the cortical reference of the anterior femoral cortex (Fig. 1). For the tibia reference, the tibial centre is defined as the bisection of the transverse tibial axis [3]. The transverse tibial axis is a line that connects the anterior/posterior (AP) midpoints of the medial and lateral condylar surfaces. The tibial centre approximates the lateral insertion of the anterior cruciate ligament (ACL). The anterior/posterior tibial axis is a perpendicular extension of the tibial centre of the transverse tibial axis (Fig. 2). This point typically matches the extension of the femoral AP axis that may be extended onto the anterior surface of the tibia. The computer algorithm then picks a point on the transmalleolar axis which is 40% from the most medial point which has been shown by anatomical studies to approximate the centre of the dome of the talus.

Fig. 1
Femoral centre reference point match which is located under the roof of the intercondylar notch and lies both on the anterior/posterior axis of Whiteside and the transepicondylar axis
Fig. 2
Proximal tibial centre point match which is located at the intersection of the transcondylar axis (line bisecting the anterior/posterior midpoints of the medial and lateral tibial condyles) and the anterior/posterior axis which bisects and is perpendicular ...

“Tibia cut first” surgical technique

The “tibia cut first” method with total knee arthroplasty follows the original technique of Insall where ligament balancing is done initially in extension before any bone cuts are made [9]. The tibia cut is made perpendicular to the mechanical axis with a 7° posterior slope to the proximal tibia. The anterior distal femoral cut is made precisely at the distal anterior surface of the femur, and the flexion gap is cut with a block that removes the posterior condyles after ligament tensioning has been done. Ligament tension is determined either with a gap spacer or a custom tensioner that adjusts and measures the amount of tension to cut a specific gap. Distal femoral chamfer and notch cuts complete femoral preparation. Following final preparation for femoral implantation, trials are inserted to assess the tension of the gaps that are created. These gaps typically will not have laxity over 3 mm, with a maximum allowed laxity in any plane of 5–6 mm.

Statistical and error analysis

Statistical analysis was performed using the Data Analysis tools in Excel 2004 for MAC. This included the Student’s t-test for two samples assuming equal variances and the F-test for two samples for determining variances. Statistical significance was determined at the P < 0.05 level for all measures.

An error analysis was determined for the computer navigation system where multiple tests were measured in the operating room setting using a single normal cadaver. After initial referencing, measurements of tibial rotation were done at full extension and 90° of flexion. Repeatability of one surgeon performing eight consecutive referencing trials revealed a mean of 13.1° internal rotation [SD: 1.5; range: 11–16°; 95% confidence interval (CI): 1.2]. Reproducibility of three surgeons performing the same tests on the same cadaver revealed a mean 12.7° internal rotation (SD: 2.1°; range: 9–16°; 95% CI: 0.9).


Compared to the baseline measurement of the arthritic knee, tibial rotation was significantly reduced with postoperative fixed bearing total knee replacement. The variation of rotation was significantly reduced from baseline to the post-implant measurement in the fixed bearing with an F ratio of 2.79 (P < 0.0007). Comparing the fixed and mobile prostheses groups, there was significantly lower rotation and variance in rotation in the fixed bearing group with an F ratio of 0.36 (P < 0.0008) (Table 2). These results reject the hypothesis that there should be no difference between the two implant types with regard to post-implantation tibial rotation.

Table 2
Outcome data from this study

A post hoc power analysis was performed to determine if the conclusion of difference in tibial rotation between the two devices was reliable. With the fixed bearing group, the mean rotation regardless of direction was 4.9° (SD = 2.8°). For the mobile group, mean rotation regardless of direction was 6.9° (SD = 5.3°). Using Student’s t-test, these two group were different at P < .02. Using PC-Size and a pooled standard deviation, this investigation had a post hoc power of 79% to demonstrate a difference, if the difference were present.

From the baseline position of the tibia as measured in relation to the distal femur at full extension to the position after total knee arthroplasty, the tibia position changed and moved more internally in 68% of knees. Comparing the fixed and mobile bearing prosthesis, we found significantly greater change for the mobile prosthesis, i.e. the tibia moved internally to a greater degree. Again, this finding rejects the hypothesis that there would no difference between the two prosthetic types. Two explanations for this difference could be that the surgeon did not optimise the tibial tray position with the fixed bearing prosthesis placing it more internally on the proximal tibia. Alternatively, the mobile bearing device, being unconstrained, may allow for alteration of the external tibial placement found at baseline in arthritic knees that have loss of the anterior cruciate ligament.

At baseline, 23% of all arthritic knees demonstrated tibial external rotation (reverse screw-home) with flexion while 23% had tibial external rotation after TKA. In 38% of cases there was a change in the direction in tibial rotation from the baseline measurement to the final measurement after total knee arthroplasty.


The results of this study would indicate that the amount of tibial rotation after total knee arthroplasty was significantly less than the baseline for the fixed bearing prosthesis but not for the mobile bearing prosthesis. Contrary to the original hypothesis, when comparing fixed and mobile bearing knees done using a similar “tibia cut first” technique, there was significantly lower tibial rotation in the fixed bearing knees. In addition, there was significantly lower variation of rotation in the fixed bearing group. The results parallel those of an earlier in vivo fluoroscopic kinematic study comparing a fixed bearing posterior stabilised design with an unconstrained posterior cruciate retaining prosthesis where the PS design had significantly lower rotation. Those authors also believed that the PS design was more sensitive to alignment of the prosthesis in the patient’s knee (Table 3).

Table 3
Comparison of kinematic and clinical data regarding tibial rotation after total knee arthroplasty

A limitation of the current study is that all measurements were made non-weight-bearing. However, most cadaveric studies and kinematic studies with roentgenographic spectrophotogrammetry (RSA) have also been done non-weight-bearing. Secondly, the inaccuracy of referencing the transepicondylar axis and the tibial tubercle are such that only relative numbers are possible. Siston et al. have shown significant variability with attempting to identify the transepicondylar axis during surgery and when attempting to use computer navigation [23]. The same authors evaluated methods to determine tibial tray rotational alignment finding that tibial tubercle referencing in computer navigation produced greater variability than even conventional total knee instrumentation [22]. Another source of error could be positioning of the knee and how the leg is held as the surgeon passes the knee through a passive range of motion. The error analysis would suggest that the potential variation from this factor is limited with a standard deviation of 1.5–2.0° for measured rotation. Finally, the author confirms that results were not changed by the selection criteria of matching the two groups for preoperative deformity. This selection was done as the surgeon who collected this series did not typically use the LCS prosthesis in cases with severe valgus deformity. However, there was a significant difference in age and sex between the two groups with the LCS prosthesis being chosen in greater numbers for younger males. It is unknown if differences in sex and age could affect the outcome.

In this study, baseline intraoperative tibial rotation and measurement after total knee arthroplasty were less than typically shown in the normal knee. Cadaveric studies have shown 14–19° of internal tibial rotation (positive screw-home) occurs throughout the arc of knee flexion in the normal knee [10, 15, 16] (Table 3). Several studies have assessed the effect of anterior cruciate ligament disruption on non-weight-bearing tibial rotation finding that rotation is typically diminished compared to normal [13, 14]. Throughout the flexion range tested, ACL-deficient knee tibias were positioned more externally relative to the distal femur compared to that seen in normal knees. Additionally, although an average positive screw-home rotation pattern was noted in ACL-deficient knees, other reports note that abherrent external rotation (reverse screw-home) may occur in some knees (Table 3). Results of this paper would confirm the presence of reverse screw-home in abnormal arthritic knees as this was identified in 23% of arthritic knees at baseline. While the current methodology cannot show that the arthritic tibias were more externally placed than normal at baseline measurement, a trend towards reduction of external tibial position was seen after placement of a prosthesis.

Kinematic studies have shown diminished tibial rotation with flexion in patients after total knee replacement (Table 3). Studies using RSA have found that rotation may vary from 1 to 10° depending on the prosthesis and surgical technique [12, 18, 19]. In vivo weight-bearing video fluoroscopy has been used to evaluate tibial rotation in a large number of patients following total knee arthroplasty compared to normal and ACL-deficient knees [6]. In one study, screw-home rotation in normal knees averaged 16.5° while that found in ACL-deficient knees averaged 8.1° and following total knee replacement 3.7°. However, the maximum range of motion on deep knee bend was notable with internal rotation of 21.3° internal rotation and external rotation of 22.3° in total knees.

The results noted in this study are consistent with most prior studies and offer new insights when comparing the baseline and postoperative total knee findings. It has been stated in the recent literature that total knee arthroplasty will result in abnormal kinematics. Siston et al. have shown in cadavers and intraoperative patients that osteoarthritic knees have reduced normal screw-home rotation which also persists after total knee arthroplasty [21]. One could argue that arthritic joints often have loss of normal anterior cruciate ligament function, altered articular load bearing position, and altered ligament tension, and are not likely to perform as a normal joint would. These changes could explain the tendency of the tibia to be externally rotated in the baseline arthritic state.

In conclusion, the most significant finding of this study is that normal screw-home rotation is reduced in postoperative total knee arthroplasty. Fixed bearing total knees demonstrated significantly less rotation than mobile bearing total knees and tended to have a greater external position of the tibia in relationship to the distal femur as compared to mobile bearing total knees. Factors such as prosthetic constraint and surgical technique for tibial tray placement may explain some of these differences.


1. Argenson JN, Scuderi GR, Komistek RD, Scott WN, Kelly MA, Aubaniac JM. In vivo kinematic evaluations and design considerations related to high flexion in total knee arthroplasty. J Biomech. 2005;38:277–284. doi: 10.1016/j.jbiomech.2004.02.027. [PubMed] [Cross Ref]
2. Banks SA, Markovich GD, Hodge WA. In vivo kinematics of cruciate-retaining and –substituting knee arthroplasties. J Arthroplasty. 1997;12:297–304. doi: 10.1016/S0883-5403(97)90026-7. [PubMed] [Cross Ref]
3. Berger RA, Crossett LS, Jacobs JJ, Rubash HE. Rotation causing patellofemoral complications after total knee arthroplasty. Clin Orthop. 1988;356:144–153. [PubMed]
4. Blunn GW, Joshi AB, Minns RJ, Lidgren L, Liley R, Ryd L, Engelbrecht E, Walker PS. Wear in retrieved condylar knee arthroplasties. A comparison of wear in different designs of 280 retrieved condylar knee prostheses. J Arthroplasty. 1997;12:281–290. doi: 10.1016/S0883-5403(97)90024-3. [PubMed] [Cross Ref]
5. Coughlin KM, Incavo SJ, Churchill DL, Beynnon BD. Tibial axis and patellar position relative to the femoral epicondylar axis during squatting. J Arthroplasty. 2003;18:1048–1055. doi: 10.1016/S0883-5403(03)00449-2. [PubMed] [Cross Ref]
6. Dennis DA, Komistek RD, Mahfouz MR, Walker SA, Tucker A. A multicenter analysis of axial femorotibial rotation after total knee arthroplasty. Clin Orthop. 2004;428:180–189. doi: 10.1097/01.blo.0000148777.98244.84. [PubMed] [Cross Ref]
7. Nahass B, Madson MM, Walker PS. Motion of the knee after condylar resurfacing—an in vivo study. J Biomech. 1991;24:1107–1117. doi: 10.1016/0021-9290(91)90003-6. [PubMed] [Cross Ref]
8. Hoff WA, Komistek RD, Dennis DA, Gabriel SM. Three-dimensional determination of femoral-tibial contact positions under in vivo conditions using fluoroscopy. Clin Biomech (Bristol, Avon) 1998;13:455–472. doi: 10.1016/S0268-0033(98)00009-6. [PubMed] [Cross Ref]
9. Insall JA, Binazzi R, Soudry M, Mestriner LA. Total knee arthroplasty. Clin Orthop. 1983;192:13–17. [PubMed]
10. Ishii Y, Terajima K, Terashima S, Koga Y. Three-dimensional kinematics of the human knee with intracortical pin fixation. Clin Orthop. 1997;343:144–150. [PubMed]
11. Jonsson H, Karrholm J, Elmqvist LG. Kinematics of active knee extension after tear of the anterior cruciate ligament. Am J Sports Med. 1989;17:796–802. doi: 10.1177/036354658901700613. [PubMed] [Cross Ref]
12. Karrholm J, Jonsson H, Nilsson KG, Soderqvist I. Kinematics of successful knee prosthesis during weight-bearing: three dimensional movements and positions of screw axes in the Tricon-M and Miller-Galante designs. Knee Surg Sports Traumatol Arthrosc. 1994;2:50–59. doi: 10.1007/BF01552655. [PubMed] [Cross Ref]
13. Karrholm J, Selvik G, Elmqvist LG, Hansson LI. Active knee motion after cruciate ligament rupture. Stereoradiography. Acta Orthop Scand. 1988;59:158–164. [PubMed]
14. Karrholm J, Selvik G, Elmqvist LG, et al. Three-dimensional instability of the anterior cruciate deficient knee. J Bone Joint Surg Br. 1988;70:777–783. [PubMed]
15. Komistek RD, Dennis DA, Mahfouz MR. In vivo fluoroscopic analysis of the normal human knee. Clin Orthop. 2003;410:69–81. doi: 10.1097/01.blo.0000062384.79828.3b. [PubMed] [Cross Ref]
16. Lafortune MA, Cavanagh PR, Sommer HJ, III, Kalenak A. Three-dimensional kinematics of the human knee during walking. J Biomech. 1992;25:347–357. doi: 10.1016/0021-9290(92)90254-X. [PubMed] [Cross Ref]
17. Lewis P, Rorabeck CH, Bourne RB, Devane P. Posteromedial tibial polyethylene failure in total knee replacements. Clin Orthop. 1994;299:11–17. [PubMed]
18. Nilsson KG, Karrholm J, Ekelund L. Knee motion in total knee arthroplasty. A roentgen stereophotogrammetric analysis of the kinematics of the Tricon-M knee prosthesis. Clin Orthop. 1990;256:147–161. [PubMed]
19. Nilsson KG, Karrholm J, Gadegaard P. Abnormal kinematics of the artificial knee. Roentgen stereophotogrammetric analysis of 10 Miller-Galante and five New Jersey LCS knees. Acta Orthop Scand. 1991;62:440–446. doi: 10.3109/17453679108996640. [PubMed] [Cross Ref]
20. Saragaglia D, Picard F. Computer-assisted implantation of total knee endoprosthesis with no preoperative imaging: the kinematic model. In: Stiehl JB, Konermann WH, Haaker RG, editors. Navigation and robotics in total joint and spinal surgery. Berlin Heidelberg New York: Springer; 2003.
21. Siston RA, Giori DJ, Goodman SB, Delp SL. Intraoperative passive kinematics of osteoarthritic knees before and after total knee arthroplasty. J Orthop Res. 2006;24:1607–1614. doi: 10.1002/jor.20163. [PubMed] [Cross Ref]
22. Siston RA, Goodman SB, Patel JJ, Delp SL, Giori NJ. High variability of tibial rotational alignment in total knee arthroplasty. Clin Orthop. 2006;452:65–69. doi: 10.1097/01.blo.0000229335.36900.a0. [PubMed] [Cross Ref]
23. Siston RA, Patel JJ, Goodman SB, Delp SL, Giori NJ. True variability of rotational alignment in total knee arthroplasty. J Bone Joint Surg Am. 2005;87:2276–2280. doi: 10.2106/JBJS.D.02945. [PubMed] [Cross Ref]
24. Stiehl JB, Dennis DA, Komistek RD, Crane HS. In vivo determination of condylar lift off and screw-home in a mobile-bearing total knee arthroplasty. J Arthroplasty. 1999;14:293–299. doi: 10.1016/S0883-5403(99)90054-2. [PubMed] [Cross Ref]
25. Stiehl JB, Komistek RD, Dennis DA. Detrimental kinematics of a flat on flat total condylar knee arthroplasty. Clin Orthop. 1999;365:139–148. doi: 10.1097/00003086-199908000-00019. [PubMed] [Cross Ref]

Articles from International Orthopaedics are provided here courtesy of Springer-Verlag